[kernel] r18312 - in people/ukleinek/3.0-rt/linux-2.6/debian: . patches/features/all/rt patches/series
Uwe Kleine-König
ukleinek-guest at alioth.debian.org
Tue Nov 22 20:00:13 UTC 2011
Author: ukleinek-guest
Date: Tue Nov 22 20:00:09 2011
New Revision: 18312
Log:
[amd64] Update rt featureset to 3.0.9-rt26
Added:
people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt26.patch
- copied, changed from r18296, people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt25.patch
people/ukleinek/3.0-rt/linux-2.6/debian/patches/series/6ptx2-extra
Deleted:
people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt25.patch
people/ukleinek/3.0-rt/linux-2.6/debian/patches/series/6ptx1-extra
Modified:
people/ukleinek/3.0-rt/linux-2.6/debian/changelog
Modified: people/ukleinek/3.0-rt/linux-2.6/debian/changelog
==============================================================================
--- people/ukleinek/3.0-rt/linux-2.6/debian/changelog Tue Nov 22 15:26:19 2011 (r18311)
+++ people/ukleinek/3.0-rt/linux-2.6/debian/changelog Tue Nov 22 20:00:09 2011 (r18312)
@@ -1,3 +1,9 @@
+linux-2.6 (3.0.0-6ptx2) unstable; urgency=low
+
+ * [amd64] Update rt featureset to 3.0.9-rt26
+
+ -- Uwe Kleine-König <u.kleine-koenig at pengutronix.de> Tue, 22 Nov 2011 20:37:47 +0100
+
linux-2.6 (3.0.0-6ptx1) unstable; urgency=low
[ Aurelien Jarno ]
Copied and modified: people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt26.patch (from r18296, people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt25.patch)
==============================================================================
--- people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt25.patch Fri Nov 18 13:19:10 2011 (r18296, copy source)
+++ people/ukleinek/3.0-rt/linux-2.6/debian/patches/features/all/rt/patch-3.0.9-rt26.patch Tue Nov 22 20:00:09 2011 (r18312)
@@ -1,3103 +1,2369 @@
-[bwh: Dropped fix to kernel/taskstats.c applied separately in
- bugfix/all/Make-TASKSTATS-require-root-access.patch]
-
-Index: linux-3.0/mm/memory.c
-===================================================================
---- linux-3.0.orig/mm/memory.c
-+++ linux-3.0/mm/memory.c
-@@ -1290,13 +1290,6 @@ static unsigned long unmap_page_range(st
- return addr;
- }
-
--#ifdef CONFIG_PREEMPT
--# define ZAP_BLOCK_SIZE (8 * PAGE_SIZE)
--#else
--/* No preempt: go for improved straight-line efficiency */
--# define ZAP_BLOCK_SIZE (1024 * PAGE_SIZE)
--#endif
--
- /**
- * unmap_vmas - unmap a range of memory covered by a list of vma's
- * @tlb: address of the caller's struct mmu_gather
-@@ -3435,6 +3428,32 @@ unlock:
- return 0;
- }
-
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+void pagefault_disable(void)
-+{
-+ migrate_disable();
-+ current->pagefault_disabled++;
-+ /*
-+ * make sure to have issued the store before a pagefault
-+ * can hit.
-+ */
-+ barrier();
-+}
-+EXPORT_SYMBOL_GPL(pagefault_disable);
+diff --git a/Documentation/hwlat_detector.txt b/Documentation/hwlat_detector.txt
+new file mode 100644
+index 0000000..cb61516
+--- /dev/null
++++ b/Documentation/hwlat_detector.txt
+@@ -0,0 +1,64 @@
++Introduction:
++-------------
+
-+void pagefault_enable(void)
-+{
-+ /*
-+ * make sure to issue those last loads/stores before enabling
-+ * the pagefault handler again.
-+ */
-+ barrier();
-+ current->pagefault_disabled--;
-+ migrate_enable();
-+}
-+EXPORT_SYMBOL_GPL(pagefault_enable);
-+#endif
++The module hwlat_detector is a special purpose kernel module that is used to
++detect large system latencies induced by the behavior of certain underlying
++hardware or firmware, independent of Linux itself. The code was developed
++originally to detect SMIs (System Management Interrupts) on x86 systems,
++however there is nothing x86 specific about this patchset. It was
++originally written for use by the "RT" patch since the Real Time
++kernel is highly latency sensitive.
+
- /*
- * By the time we get here, we already hold the mm semaphore
- */
-@@ -3983,3 +4002,35 @@ void copy_user_huge_page(struct page *ds
- }
- }
- #endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */
++SMIs are usually not serviced by the Linux kernel, which typically does not
++even know that they are occuring. SMIs are instead are set up by BIOS code
++and are serviced by BIOS code, usually for "critical" events such as
++management of thermal sensors and fans. Sometimes though, SMIs are used for
++other tasks and those tasks can spend an inordinate amount of time in the
++handler (sometimes measured in milliseconds). Obviously this is a problem if
++you are trying to keep event service latencies down in the microsecond range.
+
-+#if defined(CONFIG_PREEMPT_RT_FULL) && (USE_SPLIT_PTLOCKS > 0)
-+/*
-+ * Heinous hack, relies on the caller doing something like:
-+ *
-+ * pte = alloc_pages(PGALLOC_GFP, 0);
-+ * if (pte)
-+ * pgtable_page_ctor(pte);
-+ * return pte;
-+ *
-+ * This ensures we release the page and return NULL when the
-+ * lock allocation fails.
-+ */
-+struct page *pte_lock_init(struct page *page)
-+{
-+ page->ptl = kmalloc(sizeof(spinlock_t), GFP_KERNEL);
-+ if (page->ptl) {
-+ spin_lock_init(__pte_lockptr(page));
-+ } else {
-+ __free_page(page);
-+ page = NULL;
-+ }
-+ return page;
-+}
++The hardware latency detector works by hogging all of the cpus for configurable
++amounts of time (by calling stop_machine()), polling the CPU Time Stamp Counter
++for some period, then looking for gaps in the TSC data. Any gap indicates a
++time when the polling was interrupted and since the machine is stopped and
++interrupts turned off the only thing that could do that would be an SMI.
+
-+void pte_lock_deinit(struct page *page)
-+{
-+ kfree(page->ptl);
-+ page->mapping = NULL;
-+}
++Note that the SMI detector should *NEVER* be used in a production environment.
++It is intended to be run manually to determine if the hardware platform has a
++problem with long system firmware service routines.
+
-+#endif
-Index: linux-3.0/kernel/sched_cpupri.c
-===================================================================
---- linux-3.0.orig/kernel/sched_cpupri.c
-+++ linux-3.0/kernel/sched_cpupri.c
-@@ -47,9 +47,6 @@ static int convert_prio(int prio)
- return cpupri;
- }
-
--#define for_each_cpupri_active(array, idx) \
-- for_each_set_bit(idx, array, CPUPRI_NR_PRIORITIES)
--
- /**
- * cpupri_find - find the best (lowest-pri) CPU in the system
- * @cp: The cpupri context
-@@ -71,11 +68,38 @@ int cpupri_find(struct cpupri *cp, struc
- int idx = 0;
- int task_pri = convert_prio(p->prio);
-
-- for_each_cpupri_active(cp->pri_active, idx) {
-+ if (task_pri >= MAX_RT_PRIO)
-+ return 0;
++Usage:
++------
+
-+ for (idx = 0; idx < task_pri; idx++) {
- struct cpupri_vec *vec = &cp->pri_to_cpu[idx];
-+ int skip = 0;
-
-- if (idx >= task_pri)
-- break;
-+ if (!atomic_read(&(vec)->count))
-+ skip = 1;
-+ /*
-+ * When looking at the vector, we need to read the counter,
-+ * do a memory barrier, then read the mask.
-+ *
-+ * Note: This is still all racey, but we can deal with it.
-+ * Ideally, we only want to look at masks that are set.
-+ *
-+ * If a mask is not set, then the only thing wrong is that we
-+ * did a little more work than necessary.
-+ *
-+ * If we read a zero count but the mask is set, because of the
-+ * memory barriers, that can only happen when the highest prio
-+ * task for a run queue has left the run queue, in which case,
-+ * it will be followed by a pull. If the task we are processing
-+ * fails to find a proper place to go, that pull request will
-+ * pull this task if the run queue is running at a lower
-+ * priority.
-+ */
-+ smp_rmb();
++Loading the module hwlat_detector passing the parameter "enabled=1" (or by
++setting the "enable" entry in "hwlat_detector" debugfs toggled on) is the only
++step required to start the hwlat_detector. It is possible to redefine the
++threshold in microseconds (us) above which latency spikes will be taken
++into account (parameter "threshold=").
+
-+ /* Need to do the rmb for every iteration */
-+ if (skip)
-+ continue;
-
- if (cpumask_any_and(&p->cpus_allowed, vec->mask) >= nr_cpu_ids)
- continue;
-@@ -115,7 +139,7 @@ void cpupri_set(struct cpupri *cp, int c
- {
- int *currpri = &cp->cpu_to_pri[cpu];
- int oldpri = *currpri;
-- unsigned long flags;
-+ int do_mb = 0;
-
- newpri = convert_prio(newpri);
-
-@@ -134,26 +158,41 @@ void cpupri_set(struct cpupri *cp, int c
- if (likely(newpri != CPUPRI_INVALID)) {
- struct cpupri_vec *vec = &cp->pri_to_cpu[newpri];
++Example:
++
++ # modprobe hwlat_detector enabled=1 threshold=100
++
++After the module is loaded, it creates a directory named "hwlat_detector" under
++the debugfs mountpoint, "/debug/hwlat_detector" for this text. It is necessary
++to have debugfs mounted, which might be on /sys/debug on your system.
++
++The /debug/hwlat_detector interface contains the following files:
++
++count - number of latency spikes observed since last reset
++enable - a global enable/disable toggle (0/1), resets count
++max - maximum hardware latency actually observed (usecs)
++sample - a pipe from which to read current raw sample data
++ in the format <timestamp> <latency observed usecs>
++ (can be opened O_NONBLOCK for a single sample)
++threshold - minimum latency value to be considered (usecs)
++width - time period to sample with CPUs held (usecs)
++ must be less than the total window size (enforced)
++window - total period of sampling, width being inside (usecs)
++
++By default we will set width to 500,000 and window to 1,000,000, meaning that
++we will sample every 1,000,000 usecs (1s) for 500,000 usecs (0.5s). If we
++observe any latencies that exceed the threshold (initially 100 usecs),
++then we write to a global sample ring buffer of 8K samples, which is
++consumed by reading from the "sample" (pipe) debugfs file interface.
+diff --git a/Documentation/sysrq.txt b/Documentation/sysrq.txt
+index 312e375..9981f30 100644
+--- a/Documentation/sysrq.txt
++++ b/Documentation/sysrq.txt
+@@ -57,10 +57,17 @@ On PowerPC - Press 'ALT - Print Screen (or F13) - <command key>,
+ On other - If you know of the key combos for other architectures, please
+ let me know so I can add them to this section.
-- raw_spin_lock_irqsave(&vec->lock, flags);
--
- cpumask_set_cpu(cpu, vec->mask);
-- vec->count++;
-- if (vec->count == 1)
-- set_bit(newpri, cp->pri_active);
+-On all - write a character to /proc/sysrq-trigger. e.g.:
-
-- raw_spin_unlock_irqrestore(&vec->lock, flags);
-+ /*
-+ * When adding a new vector, we update the mask first,
-+ * do a write memory barrier, and then update the count, to
-+ * make sure the vector is visible when count is set.
-+ */
-+ smp_mb__before_atomic_inc();
-+ atomic_inc(&(vec)->count);
-+ do_mb = 1;
- }
- if (likely(oldpri != CPUPRI_INVALID)) {
- struct cpupri_vec *vec = &cp->pri_to_cpu[oldpri];
++On all - write a character to /proc/sysrq-trigger, e.g.:
+ echo t > /proc/sysrq-trigger
-- raw_spin_lock_irqsave(&vec->lock, flags);
--
-- vec->count--;
-- if (!vec->count)
-- clear_bit(oldpri, cp->pri_active);
-+ /*
-+ * Because the order of modification of the vec->count
-+ * is important, we must make sure that the update
-+ * of the new prio is seen before we decrement the
-+ * old prio. This makes sure that the loop sees
-+ * one or the other when we raise the priority of
-+ * the run queue. We don't care about when we lower the
-+ * priority, as that will trigger an rt pull anyway.
-+ *
-+ * We only need to do a memory barrier if we updated
-+ * the new priority vec.
-+ */
-+ if (do_mb)
-+ smp_mb__after_atomic_inc();
++On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g.
++ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq
++ Send an ICMP echo request with this pattern plus the particular
++ SysRq command key. Example:
++ # ping -c1 -s57 -p0102030468
++ will trigger the SysRq-H (help) command.
+
-+ /*
-+ * When removing from the vector, we decrement the counter first
-+ * do a memory barrier and then clear the mask.
-+ */
-+ atomic_dec(&(vec)->count);
-+ smp_mb__after_atomic_inc();
- cpumask_clear_cpu(cpu, vec->mask);
--
-- raw_spin_unlock_irqrestore(&vec->lock, flags);
- }
-
- *currpri = newpri;
-@@ -175,8 +214,7 @@ int cpupri_init(struct cpupri *cp)
- for (i = 0; i < CPUPRI_NR_PRIORITIES; i++) {
- struct cpupri_vec *vec = &cp->pri_to_cpu[i];
-
-- raw_spin_lock_init(&vec->lock);
-- vec->count = 0;
-+ atomic_set(&vec->count, 0);
- if (!zalloc_cpumask_var(&vec->mask, GFP_KERNEL))
- goto cleanup;
- }
-Index: linux-3.0/kernel/sched_cpupri.h
-===================================================================
---- linux-3.0.orig/kernel/sched_cpupri.h
-+++ linux-3.0/kernel/sched_cpupri.h
-@@ -12,9 +12,8 @@
- /* values 2-101 are RT priorities 0-99 */
-
- struct cpupri_vec {
-- raw_spinlock_t lock;
-- int count;
-- cpumask_var_t mask;
-+ atomic_t count;
-+ cpumask_var_t mask;
- };
-
- struct cpupri {
-Index: linux-3.0/mm/slab.c
-===================================================================
---- linux-3.0.orig/mm/slab.c
-+++ linux-3.0/mm/slab.c
-@@ -116,6 +116,7 @@
- #include <linux/kmemcheck.h>
- #include <linux/memory.h>
- #include <linux/prefetch.h>
-+#include <linux/locallock.h>
-
- #include <asm/cacheflush.h>
- #include <asm/tlbflush.h>
-@@ -620,6 +621,51 @@ int slab_is_available(void)
- static struct lock_class_key on_slab_l3_key;
- static struct lock_class_key on_slab_alc_key;
-
-+static struct lock_class_key debugobj_l3_key;
-+static struct lock_class_key debugobj_alc_key;
+
-+static void slab_set_lock_classes(struct kmem_cache *cachep,
-+ struct lock_class_key *l3_key, struct lock_class_key *alc_key,
-+ int q)
-+{
-+ struct array_cache **alc;
-+ struct kmem_list3 *l3;
-+ int r;
+ * What are the 'command' keys?
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ 'b' - Will immediately reboot the system without syncing or unmounting
+diff --git a/Documentation/trace/histograms.txt b/Documentation/trace/histograms.txt
+new file mode 100644
+index 0000000..6f2aeab
+--- /dev/null
++++ b/Documentation/trace/histograms.txt
+@@ -0,0 +1,186 @@
++ Using the Linux Kernel Latency Histograms
+
-+ l3 = cachep->nodelists[q];
-+ if (!l3)
-+ return;
+
-+ lockdep_set_class(&l3->list_lock, l3_key);
-+ alc = l3->alien;
-+ /*
-+ * FIXME: This check for BAD_ALIEN_MAGIC
-+ * should go away when common slab code is taught to
-+ * work even without alien caches.
-+ * Currently, non NUMA code returns BAD_ALIEN_MAGIC
-+ * for alloc_alien_cache,
-+ */
-+ if (!alc || (unsigned long)alc == BAD_ALIEN_MAGIC)
-+ return;
-+ for_each_node(r) {
-+ if (alc[r])
-+ lockdep_set_class(&alc[r]->lock, alc_key);
-+ }
-+}
++This document gives a short explanation how to enable, configure and use
++latency histograms. Latency histograms are primarily relevant in the
++context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT)
++and are used in the quality management of the Linux real-time
++capabilities.
+
-+static void slab_set_debugobj_lock_classes_node(struct kmem_cache *cachep, int node)
-+{
-+ slab_set_lock_classes(cachep, &debugobj_l3_key, &debugobj_alc_key, node);
-+}
+
-+static void slab_set_debugobj_lock_classes(struct kmem_cache *cachep)
-+{
-+ int node;
++* Purpose of latency histograms
+
-+ for_each_online_node(node)
-+ slab_set_debugobj_lock_classes_node(cachep, node);
-+}
++A latency histogram continuously accumulates the frequencies of latency
++data. There are two types of histograms
++- potential sources of latencies
++- effective latencies
+
- static void init_node_lock_keys(int q)
- {
- struct cache_sizes *s = malloc_sizes;
-@@ -628,29 +674,14 @@ static void init_node_lock_keys(int q)
- return;
-
- for (s = malloc_sizes; s->cs_size != ULONG_MAX; s++) {
-- struct array_cache **alc;
- struct kmem_list3 *l3;
-- int r;
-
- l3 = s->cs_cachep->nodelists[q];
- if (!l3 || OFF_SLAB(s->cs_cachep))
- continue;
-- lockdep_set_class(&l3->list_lock, &on_slab_l3_key);
-- alc = l3->alien;
-- /*
-- * FIXME: This check for BAD_ALIEN_MAGIC
-- * should go away when common slab code is taught to
-- * work even without alien caches.
-- * Currently, non NUMA code returns BAD_ALIEN_MAGIC
-- * for alloc_alien_cache,
-- */
-- if (!alc || (unsigned long)alc == BAD_ALIEN_MAGIC)
-- continue;
-- for_each_node(r) {
-- if (alc[r])
-- lockdep_set_class(&alc[r]->lock,
-- &on_slab_alc_key);
-- }
+
-+ slab_set_lock_classes(s->cs_cachep, &on_slab_l3_key,
-+ &on_slab_alc_key, q);
- }
- }
-
-@@ -669,6 +700,14 @@ static void init_node_lock_keys(int q)
- static inline void init_lock_keys(void)
- {
- }
++* Potential sources of latencies
+
-+static void slab_set_debugobj_lock_classes_node(struct kmem_cache *cachep, int node)
-+{
-+}
++Potential sources of latencies are code segments where interrupts,
++preemption or both are disabled (aka critical sections). To create
++histograms of potential sources of latency, the kernel stores the time
++stamp at the start of a critical section, determines the time elapsed
++when the end of the section is reached, and increments the frequency
++counter of that latency value - irrespective of whether any concurrently
++running process is affected by latency or not.
++- Configuration items (in the Kernel hacking/Tracers submenu)
++ CONFIG_INTERRUPT_OFF_LATENCY
++ CONFIG_PREEMPT_OFF_LATENCY
+
-+static void slab_set_debugobj_lock_classes(struct kmem_cache *cachep)
-+{
-+}
- #endif
-
- /*
-@@ -678,12 +717,66 @@ static DEFINE_MUTEX(cache_chain_mutex);
- static struct list_head cache_chain;
-
- static DEFINE_PER_CPU(struct delayed_work, slab_reap_work);
-+static DEFINE_PER_CPU(struct list_head, slab_free_list);
-+static DEFINE_LOCAL_IRQ_LOCK(slab_lock);
+
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+# define slab_on_each_cpu(func, cp) on_each_cpu(func, cp, 1)
-+#else
-+/*
-+ * execute func() for all CPUs. On PREEMPT_RT we dont actually have
-+ * to run on the remote CPUs - we only have to take their CPU-locks.
-+ * (This is a rare operation, so cacheline bouncing is not an issue.)
-+ */
-+static void
-+slab_on_each_cpu(void (*func)(void *arg, int this_cpu), void *arg)
-+{
-+ unsigned int i;
++* Effective latencies
+
-+ for_each_online_cpu(i)
-+ func(arg, i);
-+}
-+#endif
++Effective latencies are actually occuring during wakeup of a process. To
++determine effective latencies, the kernel stores the time stamp when a
++process is scheduled to be woken up, and determines the duration of the
++wakeup time shortly before control is passed over to this process. Note
++that the apparent latency in user space may be somewhat longer, since the
++process may be interrupted after control is passed over to it but before
++the execution in user space takes place. Simply measuring the interval
++between enqueuing and wakeup may also not appropriate in cases when a
++process is scheduled as a result of a timer expiration. The timer may have
++missed its deadline, e.g. due to disabled interrupts, but this latency
++would not be registered. Therefore, the offsets of missed timers are
++recorded in a separate histogram. If both wakeup latency and missed timer
++offsets are configured and enabled, a third histogram may be enabled that
++records the overall latency as a sum of the timer latency, if any, and the
++wakeup latency. This histogram is called "timerandwakeup".
++- Configuration items (in the Kernel hacking/Tracers submenu)
++ CONFIG_WAKEUP_LATENCY
++ CONFIG_MISSED_TIMER_OFSETS
+
-+static void free_delayed(struct list_head *h)
-+{
-+ while(!list_empty(h)) {
-+ struct page *page = list_first_entry(h, struct page, lru);
+
-+ list_del(&page->lru);
-+ __free_pages(page, page->index);
-+ }
-+}
++* Usage
+
-+static void unlock_l3_and_free_delayed(spinlock_t *list_lock)
-+{
-+ LIST_HEAD(tmp);
++The interface to the administration of the latency histograms is located
++in the debugfs file system. To mount it, either enter
+
-+ list_splice_init(&__get_cpu_var(slab_free_list), &tmp);
-+ local_spin_unlock_irq(slab_lock, list_lock);
-+ free_delayed(&tmp);
-+}
++mount -t sysfs nodev /sys
++mount -t debugfs nodev /sys/kernel/debug
+
-+static void unlock_slab_and_free_delayed(unsigned long flags)
-+{
-+ LIST_HEAD(tmp);
++from shell command line level, or add
+
-+ list_splice_init(&__get_cpu_var(slab_free_list), &tmp);
-+ local_unlock_irqrestore(slab_lock, flags);
-+ free_delayed(&tmp);
-+}
-
- static inline struct array_cache *cpu_cache_get(struct kmem_cache *cachep)
- {
- return cachep->array[smp_processor_id()];
- }
-
-+static inline struct array_cache *cpu_cache_get_on_cpu(struct kmem_cache *cachep,
-+ int cpu)
-+{
-+ return cachep->array[cpu];
-+}
++nodev /sys sysfs defaults 0 0
++nodev /sys/kernel/debug debugfs defaults 0 0
+
- static inline struct kmem_cache *__find_general_cachep(size_t size,
- gfp_t gfpflags)
- {
-@@ -1021,9 +1114,10 @@ static void reap_alien(struct kmem_cache
- if (l3->alien) {
- struct array_cache *ac = l3->alien[node];
-
-- if (ac && ac->avail && spin_trylock_irq(&ac->lock)) {
-+ if (ac && ac->avail &&
-+ local_spin_trylock_irq(slab_lock, &ac->lock)) {
- __drain_alien_cache(cachep, ac, node);
-- spin_unlock_irq(&ac->lock);
-+ local_spin_unlock_irq(slab_lock, &ac->lock);
- }
- }
- }
-@@ -1038,9 +1132,9 @@ static void drain_alien_cache(struct kme
- for_each_online_node(i) {
- ac = alien[i];
- if (ac) {
-- spin_lock_irqsave(&ac->lock, flags);
-+ local_spin_lock_irqsave(slab_lock, &ac->lock, flags);
- __drain_alien_cache(cachep, ac, i);
-- spin_unlock_irqrestore(&ac->lock, flags);
-+ local_spin_unlock_irqrestore(slab_lock, &ac->lock, flags);
- }
- }
- }
-@@ -1119,11 +1213,11 @@ static int init_cache_nodelists_node(int
- cachep->nodelists[node] = l3;
- }
-
-- spin_lock_irq(&cachep->nodelists[node]->list_lock);
-+ local_spin_lock_irq(slab_lock, &cachep->nodelists[node]->list_lock);
- cachep->nodelists[node]->free_limit =
- (1 + nr_cpus_node(node)) *
- cachep->batchcount + cachep->num;
-- spin_unlock_irq(&cachep->nodelists[node]->list_lock);
-+ local_spin_unlock_irq(slab_lock, &cachep->nodelists[node]->list_lock);
- }
- return 0;
- }
-@@ -1148,7 +1242,7 @@ static void __cpuinit cpuup_canceled(lon
- if (!l3)
- goto free_array_cache;
-
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
-
- /* Free limit for this kmem_list3 */
- l3->free_limit -= cachep->batchcount;
-@@ -1156,7 +1250,7 @@ static void __cpuinit cpuup_canceled(lon
- free_block(cachep, nc->entry, nc->avail, node);
-
- if (!cpumask_empty(mask)) {
-- spin_unlock_irq(&l3->list_lock);
-+ unlock_l3_and_free_delayed(&l3->list_lock);
- goto free_array_cache;
- }
-
-@@ -1170,7 +1264,7 @@ static void __cpuinit cpuup_canceled(lon
- alien = l3->alien;
- l3->alien = NULL;
-
-- spin_unlock_irq(&l3->list_lock);
-+ unlock_l3_and_free_delayed(&l3->list_lock);
-
- kfree(shared);
- if (alien) {
-@@ -1244,7 +1338,7 @@ static int __cpuinit cpuup_prepare(long
- l3 = cachep->nodelists[node];
- BUG_ON(!l3);
-
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
- if (!l3->shared) {
- /*
- * We are serialised from CPU_DEAD or
-@@ -1259,9 +1353,11 @@ static int __cpuinit cpuup_prepare(long
- alien = NULL;
- }
- #endif
-- spin_unlock_irq(&l3->list_lock);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
- kfree(shared);
- free_alien_cache(alien);
-+ if (cachep->flags & SLAB_DEBUG_OBJECTS)
-+ slab_set_debugobj_lock_classes_node(cachep, node);
- }
- init_node_lock_keys(node);
-
-@@ -1448,6 +1544,10 @@ void __init kmem_cache_init(void)
- if (num_possible_nodes() == 1)
- use_alien_caches = 0;
-
-+ local_irq_lock_init(slab_lock);
-+ for_each_possible_cpu(i)
-+ INIT_LIST_HEAD(&per_cpu(slab_free_list, i));
++to the file /etc/fstab. All latency histogram related files are then
++available in the directory /sys/kernel/debug/tracing/latency_hist. A
++particular histogram type is enabled by writing non-zero to the related
++variable in the /sys/kernel/debug/tracing/latency_hist/enable directory.
++Select "preemptirqsoff" for the histograms of potential sources of
++latencies and "wakeup" for histograms of effective latencies etc. The
++histogram data - one per CPU - are available in the files
+
- for (i = 0; i < NUM_INIT_LISTS; i++) {
- kmem_list3_init(&initkmem_list3[i]);
- if (i < MAX_NUMNODES)
-@@ -1625,6 +1725,9 @@ void __init kmem_cache_init_late(void)
- {
- struct kmem_cache *cachep;
++/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx
++/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx
++/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx
++/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx
++/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx
++/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx
++/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx
++
++The histograms are reset by writing non-zero to the file "reset" in a
++particular latency directory. To reset all latency data, use
++
++#!/bin/sh
++
++TRACINGDIR=/sys/kernel/debug/tracing
++HISTDIR=$TRACINGDIR/latency_hist
++
++if test -d $HISTDIR
++then
++ cd $HISTDIR
++ for i in `find . | grep /reset$`
++ do
++ echo 1 >$i
++ done
++fi
++
++
++* Data format
++
++Latency data are stored with a resolution of one microsecond. The
++maximum latency is 10,240 microseconds. The data are only valid, if the
++overflow register is empty. Every output line contains the latency in
++microseconds in the first row and the number of samples in the second
++row. To display only lines with a positive latency count, use, for
++example,
++
++grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0
++
++#Minimum latency: 0 microseconds.
++#Average latency: 0 microseconds.
++#Maximum latency: 25 microseconds.
++#Total samples: 3104770694
++#There are 0 samples greater or equal than 10240 microseconds
++#usecs samples
++ 0 2984486876
++ 1 49843506
++ 2 58219047
++ 3 5348126
++ 4 2187960
++ 5 3388262
++ 6 959289
++ 7 208294
++ 8 40420
++ 9 4485
++ 10 14918
++ 11 18340
++ 12 25052
++ 13 19455
++ 14 5602
++ 15 969
++ 16 47
++ 17 18
++ 18 14
++ 19 1
++ 20 3
++ 21 2
++ 22 5
++ 23 2
++ 25 1
++
++
++* Wakeup latency of a selected process
++
++To only collect wakeup latency data of a particular process, write the
++PID of the requested process to
++
++/sys/kernel/debug/tracing/latency_hist/wakeup/pid
++
++PIDs are not considered, if this variable is set to 0.
++
++
++* Details of the process with the highest wakeup latency so far
++
++Selected data of the process that suffered from the highest wakeup
++latency that occurred in a particular CPU are available in the file
++
++/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx.
++
++In addition, other relevant system data at the time when the
++latency occurred are given.
++
++The format of the data is (all in one line):
++<PID> <Priority> <Latency> (<Timeroffset>) <Command> \
++<- <PID> <Priority> <Command> <Timestamp>
++
++The value of <Timeroffset> is only relevant in the combined timer
++and wakeup latency recording. In the wakeup recording, it is
++always 0, in the missed_timer_offsets recording, it is the same
++as <Latency>.
++
++When retrospectively searching for the origin of a latency and
++tracing was not enabled, it may be helpful to know the name and
++some basic data of the task that (finally) was switching to the
++late real-tlme task. In addition to the victim's data, also the
++data of the possible culprit are therefore displayed after the
++"<-" symbol.
++
++Finally, the timestamp of the time when the latency occurred
++in <seconds>.<microseconds> after the most recent system boot
++is provided.
++
++These data are also reset when the wakeup histogram is reset.
+diff --git a/MAINTAINERS b/MAINTAINERS
+index 34e2418..fed9c84 100644
+--- a/MAINTAINERS
++++ b/MAINTAINERS
+@@ -2911,6 +2911,15 @@ L: linuxppc-dev at lists.ozlabs.org
+ S: Odd Fixes
+ F: drivers/tty/hvc/
-+ /* Annotate slab for lockdep -- annotate the malloc caches */
-+ init_lock_keys();
++HARDWARE LATENCY DETECTOR
++P: Jon Masters
++M: jcm at jonmasters.org
++W: http://www.kernel.org/pub/linux/kernel/people/jcm/hwlat_detector/
++S: Supported
++L: linux-kernel at vger.kernel.org
++F: Documentation/hwlat_detector.txt
++F: drivers/misc/hwlat_detector.c
+
- /* 6) resize the head arrays to their final sizes */
- mutex_lock(&cache_chain_mutex);
- list_for_each_entry(cachep, &cache_chain, next)
-@@ -1635,9 +1738,6 @@ void __init kmem_cache_init_late(void)
- /* Done! */
- g_cpucache_up = FULL;
+ HARDWARE MONITORING
+ M: Jean Delvare <khali at linux-fr.org>
+ M: Guenter Roeck <guenter.roeck at ericsson.com>
+diff --git a/arch/Kconfig b/arch/Kconfig
+index 26b0e23..40112de 100644
+--- a/arch/Kconfig
++++ b/arch/Kconfig
+@@ -6,6 +6,7 @@ config OPROFILE
+ tristate "OProfile system profiling"
+ depends on PROFILING
+ depends on HAVE_OPROFILE
++ depends on !PREEMPT_RT_FULL
+ select RING_BUFFER
+ select RING_BUFFER_ALLOW_SWAP
+ help
+diff --git a/arch/alpha/include/asm/rwsem.h b/arch/alpha/include/asm/rwsem.h
+index a83bbea..0f880ca 100644
+--- a/arch/alpha/include/asm/rwsem.h
++++ b/arch/alpha/include/asm/rwsem.h
+@@ -21,7 +21,7 @@
+ #define RWSEM_ACTIVE_READ_BIAS RWSEM_ACTIVE_BIAS
+ #define RWSEM_ACTIVE_WRITE_BIAS (RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
-- /* Annotate slab for lockdep -- annotate the malloc caches */
-- init_lock_keys();
--
- /*
- * Register a cpu startup notifier callback that initializes
- * cpu_cache_get for all new cpus
-@@ -1725,12 +1825,14 @@ static void *kmem_getpages(struct kmem_c
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
+ {
+ long oldcount;
+ #ifndef CONFIG_SMP
+@@ -48,7 +48,7 @@ static inline void __down_read(struct rw_semaphore *sem)
/*
- * Interface to system's page release.
+ * trylock for reading -- returns 1 if successful, 0 if contention
*/
--static void kmem_freepages(struct kmem_cache *cachep, void *addr)
-+static void kmem_freepages(struct kmem_cache *cachep, void *addr, bool delayed)
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
{
- unsigned long i = (1 << cachep->gfporder);
-- struct page *page = virt_to_page(addr);
-+ struct page *page, *basepage = virt_to_page(addr);
- const unsigned long nr_freed = i;
-
-+ page = basepage;
-+
- kmemcheck_free_shadow(page, cachep->gfporder);
+ long old, new, res;
- if (cachep->flags & SLAB_RECLAIM_ACCOUNT)
-@@ -1746,7 +1848,13 @@ static void kmem_freepages(struct kmem_c
- }
- if (current->reclaim_state)
- current->reclaim_state->reclaimed_slab += nr_freed;
-- free_pages((unsigned long)addr, cachep->gfporder);
-+
-+ if (!delayed) {
-+ free_pages((unsigned long)addr, cachep->gfporder);
-+ } else {
-+ basepage->index = cachep->gfporder;
-+ list_add(&basepage->lru, &__get_cpu_var(slab_free_list));
-+ }
+@@ -63,7 +63,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ return res >= 0 ? 1 : 0;
}
- static void kmem_rcu_free(struct rcu_head *head)
-@@ -1754,7 +1862,7 @@ static void kmem_rcu_free(struct rcu_hea
- struct slab_rcu *slab_rcu = (struct slab_rcu *)head;
- struct kmem_cache *cachep = slab_rcu->cachep;
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
+ {
+ long oldcount;
+ #ifndef CONFIG_SMP
+@@ -90,7 +90,7 @@ static inline void __down_write(struct rw_semaphore *sem)
+ /*
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ long ret = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
+ RWSEM_ACTIVE_WRITE_BIAS);
+@@ -99,7 +99,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ return 0;
+ }
-- kmem_freepages(cachep, slab_rcu->addr);
-+ kmem_freepages(cachep, slab_rcu->addr, false);
- if (OFF_SLAB(cachep))
- kmem_cache_free(cachep->slabp_cache, slab_rcu);
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ long oldcount;
+ #ifndef CONFIG_SMP
+@@ -124,7 +124,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ rwsem_wake(sem);
}
-@@ -1973,7 +2081,8 @@ static void slab_destroy_debugcheck(stru
- * Before calling the slab must have been unlinked from the cache. The
- * cache-lock is not held/needed.
+
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
+ {
+ long count;
+ #ifndef CONFIG_SMP
+@@ -153,7 +153,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
*/
--static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp)
-+static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp,
-+ bool delayed)
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
{
- void *addr = slabp->s_mem - slabp->colouroff;
-
-@@ -1986,7 +2095,7 @@ static void slab_destroy(struct kmem_cac
- slab_rcu->addr = addr;
- call_rcu(&slab_rcu->head, kmem_rcu_free);
- } else {
-- kmem_freepages(cachep, addr);
-+ kmem_freepages(cachep, addr, delayed);
- if (OFF_SLAB(cachep))
- kmem_cache_free(cachep->slabp_cache, slabp);
- }
-@@ -2424,6 +2533,16 @@ kmem_cache_create (const char *name, siz
- goto oops;
- }
+ long oldcount;
+ #ifndef CONFIG_SMP
+@@ -177,7 +177,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
+ rwsem_downgrade_wake(sem);
+ }
-+ if (flags & SLAB_DEBUG_OBJECTS) {
-+ /*
-+ * Would deadlock through slab_destroy()->call_rcu()->
-+ * debug_object_activate()->kmem_cache_alloc().
-+ */
-+ WARN_ON_ONCE(flags & SLAB_DESTROY_BY_RCU);
-+
-+ slab_set_debugobj_lock_classes(cachep);
-+ }
-+
- /* cache setup completed, link it into the list */
- list_add(&cachep->next, &cache_chain);
- oops:
-@@ -2441,7 +2560,7 @@ EXPORT_SYMBOL(kmem_cache_create);
- #if DEBUG
- static void check_irq_off(void)
+-static inline void rwsem_atomic_add(long val, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(long val, struct rw_anon_semaphore *sem)
{
-- BUG_ON(!irqs_disabled());
-+ BUG_ON_NONRT(!irqs_disabled());
+ #ifndef CONFIG_SMP
+ sem->count += val;
+@@ -196,7 +196,7 @@ static inline void rwsem_atomic_add(long val, struct rw_semaphore *sem)
+ #endif
}
- static void check_irq_on(void)
-@@ -2476,26 +2595,43 @@ static void drain_array(struct kmem_cach
- struct array_cache *ac,
- int force, int node);
+-static inline long rwsem_atomic_update(long val, struct rw_semaphore *sem)
++static inline long rwsem_atomic_update(long val, struct rw_anon_semaphore *sem)
+ {
+ #ifndef CONFIG_SMP
+ sem->count += val;
+diff --git a/arch/alpha/mm/fault.c b/arch/alpha/mm/fault.c
+index fadd5f8..4a0a0af 100644
+--- a/arch/alpha/mm/fault.c
++++ b/arch/alpha/mm/fault.c
+@@ -107,7 +107,7 @@ do_page_fault(unsigned long address, unsigned long mmcsr,
--static void do_drain(void *arg)
-+static void __do_drain(void *arg, unsigned int cpu)
+ /* If we're in an interrupt context, or have no user context,
+ we must not take the fault. */
+- if (!mm || in_atomic())
++ if (!mm || pagefault_disabled())
+ goto no_context;
+
+ #ifdef CONFIG_ALPHA_LARGE_VMALLOC
+diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
+index 91c84cb..5c224f5 100644
+--- a/arch/arm/Kconfig
++++ b/arch/arm/Kconfig
+@@ -29,6 +29,7 @@ config ARM
+ select HAVE_GENERIC_HARDIRQS
+ select HAVE_SPARSE_IRQ
+ select GENERIC_IRQ_SHOW
++ select IRQ_FORCED_THREADING
+ help
+ The ARM series is a line of low-power-consumption RISC chip designs
+ licensed by ARM Ltd and targeted at embedded applications and
+@@ -1524,7 +1525,7 @@ config HAVE_ARCH_PFN_VALID
+
+ config HIGHMEM
+ bool "High Memory Support"
+- depends on MMU
++ depends on MMU && !PREEMPT_RT_FULL
+ help
+ The address space of ARM processors is only 4 Gigabytes large
+ and it has to accommodate user address space, kernel address
+diff --git a/arch/arm/common/gic.c b/arch/arm/common/gic.c
+index 4ddd0a6..5c1dd07 100644
+--- a/arch/arm/common/gic.c
++++ b/arch/arm/common/gic.c
+@@ -33,7 +33,7 @@
+ #include <asm/mach/irq.h>
+ #include <asm/hardware/gic.h>
+
+-static DEFINE_SPINLOCK(irq_controller_lock);
++static DEFINE_RAW_SPINLOCK(irq_controller_lock);
+
+ /* Address of GIC 0 CPU interface */
+ void __iomem *gic_cpu_base_addr __read_mostly;
+@@ -88,30 +88,30 @@ static void gic_mask_irq(struct irq_data *d)
{
- struct kmem_cache *cachep = arg;
- struct array_cache *ac;
-- int node = numa_mem_id();
-+ int node = cpu_to_mem(cpu);
+ u32 mask = 1 << (d->irq % 32);
-- check_irq_off();
-- ac = cpu_cache_get(cachep);
-+ ac = cpu_cache_get_on_cpu(cachep, cpu);
- spin_lock(&cachep->nodelists[node]->list_lock);
- free_block(cachep, ac->entry, ac->avail, node);
- spin_unlock(&cachep->nodelists[node]->list_lock);
- ac->avail = 0;
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
+ writel_relaxed(mask, gic_dist_base(d) + GIC_DIST_ENABLE_CLEAR + (gic_irq(d) / 32) * 4);
+ if (gic_arch_extn.irq_mask)
+ gic_arch_extn.irq_mask(d);
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
}
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+static void do_drain(void *arg)
-+{
-+ __do_drain(arg, smp_processor_id());
-+}
-+#else
-+static void do_drain(void *arg, int cpu)
-+{
-+ LIST_HEAD(tmp);
-+
-+ spin_lock_irq(&per_cpu(slab_lock, cpu).lock);
-+ __do_drain(arg, cpu);
-+ list_splice_init(&per_cpu(slab_free_list, cpu), &tmp);
-+ spin_unlock_irq(&per_cpu(slab_lock, cpu).lock);
-+ free_delayed(&tmp);
-+}
-+#endif
-+
- static void drain_cpu_caches(struct kmem_cache *cachep)
+ static void gic_unmask_irq(struct irq_data *d)
{
- struct kmem_list3 *l3;
- int node;
-
-- on_each_cpu(do_drain, cachep, 1);
-+ slab_on_each_cpu(do_drain, cachep);
- check_irq_on();
- for_each_online_node(node) {
- l3 = cachep->nodelists[node];
-@@ -2526,10 +2662,10 @@ static int drain_freelist(struct kmem_ca
- nr_freed = 0;
- while (nr_freed < tofree && !list_empty(&l3->slabs_free)) {
+ u32 mask = 1 << (d->irq % 32);
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
- p = l3->slabs_free.prev;
- if (p == &l3->slabs_free) {
-- spin_unlock_irq(&l3->list_lock);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
- goto out;
- }
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
+ if (gic_arch_extn.irq_unmask)
+ gic_arch_extn.irq_unmask(d);
+ writel_relaxed(mask, gic_dist_base(d) + GIC_DIST_ENABLE_SET + (gic_irq(d) / 32) * 4);
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
+ }
-@@ -2543,8 +2679,8 @@ static int drain_freelist(struct kmem_ca
- * to the cache.
- */
- l3->free_objects -= cache->num;
-- spin_unlock_irq(&l3->list_lock);
-- slab_destroy(cache, slabp);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
-+ slab_destroy(cache, slabp, false);
- nr_freed++;
+ static void gic_eoi_irq(struct irq_data *d)
+ {
+ if (gic_arch_extn.irq_eoi) {
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
+ gic_arch_extn.irq_eoi(d);
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
}
- out:
-@@ -2838,7 +2974,7 @@ static int cache_grow(struct kmem_cache
- offset *= cachep->colour_off;
- if (local_flags & __GFP_WAIT)
-- local_irq_enable();
-+ local_unlock_irq(slab_lock);
+ writel_relaxed(gic_irq(d), gic_cpu_base(d) + GIC_CPU_EOI);
+@@ -135,7 +135,7 @@ static int gic_set_type(struct irq_data *d, unsigned int type)
+ if (type != IRQ_TYPE_LEVEL_HIGH && type != IRQ_TYPE_EDGE_RISING)
+ return -EINVAL;
- /*
- * The test for missing atomic flag is performed here, rather than
-@@ -2868,7 +3004,7 @@ static int cache_grow(struct kmem_cache
- cache_init_objs(cachep, slabp);
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
- if (local_flags & __GFP_WAIT)
-- local_irq_disable();
-+ local_lock_irq(slab_lock);
- check_irq_off();
- spin_lock(&l3->list_lock);
+ if (gic_arch_extn.irq_set_type)
+ gic_arch_extn.irq_set_type(d, type);
+@@ -160,7 +160,7 @@ static int gic_set_type(struct irq_data *d, unsigned int type)
+ if (enabled)
+ writel_relaxed(enablemask, base + GIC_DIST_ENABLE_SET + enableoff);
+
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
-@@ -2879,10 +3015,10 @@ static int cache_grow(struct kmem_cache
- spin_unlock(&l3->list_lock);
- return 1;
- opps1:
-- kmem_freepages(cachep, objp);
-+ kmem_freepages(cachep, objp, false);
- failed:
- if (local_flags & __GFP_WAIT)
-- local_irq_disable();
-+ local_lock_irq(slab_lock);
return 0;
}
+@@ -188,11 +188,11 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
+ mask = 0xff << shift;
+ bit = 1 << (cpu + shift);
-@@ -3280,11 +3416,11 @@ retry:
- * set and go into memory reserves if necessary.
- */
- if (local_flags & __GFP_WAIT)
-- local_irq_enable();
-+ local_unlock_irq(slab_lock);
- kmem_flagcheck(cache, flags);
- obj = kmem_getpages(cache, local_flags, numa_mem_id());
- if (local_flags & __GFP_WAIT)
-- local_irq_disable();
-+ local_lock_irq(slab_lock);
- if (obj) {
- /*
- * Insert into the appropriate per node queues
-@@ -3400,7 +3536,7 @@ __cache_alloc_node(struct kmem_cache *ca
- return NULL;
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
+ d->node = cpu;
+ val = readl_relaxed(reg) & ~mask;
+ writel_relaxed(val | bit, reg);
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
- cache_alloc_debugcheck_before(cachep, flags);
-- local_irq_save(save_flags);
-+ local_lock_irqsave(slab_lock, save_flags);
+ return 0;
+ }
+@@ -222,9 +222,9 @@ static void gic_handle_cascade_irq(unsigned int irq, struct irq_desc *desc)
- if (nodeid == -1)
- nodeid = slab_node;
-@@ -3425,7 +3561,7 @@ __cache_alloc_node(struct kmem_cache *ca
- /* ___cache_alloc_node can fall back to other nodes */
- ptr = ____cache_alloc_node(cachep, flags, nodeid);
- out:
-- local_irq_restore(save_flags);
-+ local_unlock_irqrestore(slab_lock, save_flags);
- ptr = cache_alloc_debugcheck_after(cachep, flags, ptr, caller);
- kmemleak_alloc_recursive(ptr, obj_size(cachep), 1, cachep->flags,
- flags);
-@@ -3485,9 +3621,9 @@ __cache_alloc(struct kmem_cache *cachep,
- return NULL;
+ chained_irq_enter(chip, desc);
- cache_alloc_debugcheck_before(cachep, flags);
-- local_irq_save(save_flags);
-+ local_lock_irqsave(slab_lock, save_flags);
- objp = __do_cache_alloc(cachep, flags);
-- local_irq_restore(save_flags);
-+ local_unlock_irqrestore(slab_lock, save_flags);
- objp = cache_alloc_debugcheck_after(cachep, flags, objp, caller);
- kmemleak_alloc_recursive(objp, obj_size(cachep), 1, cachep->flags,
- flags);
-@@ -3535,7 +3671,7 @@ static void free_block(struct kmem_cache
- * a different cache, refer to comments before
- * alloc_slabmgmt.
- */
-- slab_destroy(cachep, slabp);
-+ slab_destroy(cachep, slabp, true);
- } else {
- list_add(&slabp->list, &l3->slabs_free);
- }
-@@ -3798,12 +3934,12 @@ void kmem_cache_free(struct kmem_cache *
- {
- unsigned long flags;
+- spin_lock(&irq_controller_lock);
++ raw_spin_lock(&irq_controller_lock);
+ status = readl_relaxed(chip_data->cpu_base + GIC_CPU_INTACK);
+- spin_unlock(&irq_controller_lock);
++ raw_spin_unlock(&irq_controller_lock);
-- local_irq_save(flags);
- debug_check_no_locks_freed(objp, obj_size(cachep));
- if (!(cachep->flags & SLAB_DEBUG_OBJECTS))
- debug_check_no_obj_freed(objp, obj_size(cachep));
-+ local_lock_irqsave(slab_lock, flags);
- __cache_free(cachep, objp, __builtin_return_address(0));
-- local_irq_restore(flags);
-+ unlock_slab_and_free_delayed(flags);
+ gic_irq = (status & 0x3ff);
+ if (gic_irq == 1023)
+diff --git a/arch/arm/include/asm/dma.h b/arch/arm/include/asm/dma.h
+index 4200554..8a54b7d 100644
+--- a/arch/arm/include/asm/dma.h
++++ b/arch/arm/include/asm/dma.h
+@@ -33,18 +33,18 @@
+ #define DMA_MODE_CASCADE 0xc0
+ #define DMA_AUTOINIT 0x10
- trace_kmem_cache_free(_RET_IP_, objp);
- }
-@@ -3827,13 +3963,13 @@ void kfree(const void *objp)
+-extern spinlock_t dma_spin_lock;
++extern raw_spinlock_t dma_spin_lock;
- if (unlikely(ZERO_OR_NULL_PTR(objp)))
- return;
-- local_irq_save(flags);
- kfree_debugcheck(objp);
- c = virt_to_cache(objp);
- debug_check_no_locks_freed(objp, obj_size(c));
- debug_check_no_obj_freed(objp, obj_size(c));
-+ local_lock_irqsave(slab_lock, flags);
- __cache_free(c, (void *)objp, __builtin_return_address(0));
-- local_irq_restore(flags);
-+ unlock_slab_and_free_delayed(flags);
+ static inline unsigned long claim_dma_lock(void)
+ {
+ unsigned long flags;
+- spin_lock_irqsave(&dma_spin_lock, flags);
++ raw_spin_lock_irqsave(&dma_spin_lock, flags);
+ return flags;
}
- EXPORT_SYMBOL(kfree);
-@@ -3876,7 +4012,7 @@ static int alloc_kmemlist(struct kmem_ca
- if (l3) {
- struct array_cache *shared = l3->shared;
+ static inline void release_dma_lock(unsigned long flags)
+ {
+- spin_unlock_irqrestore(&dma_spin_lock, flags);
++ raw_spin_unlock_irqrestore(&dma_spin_lock, flags);
+ }
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
+ /* Clear the 'DMA Pointer Flip Flop'.
+diff --git a/arch/arm/include/asm/mmu.h b/arch/arm/include/asm/mmu.h
+index b4ffe9d..1496565 100644
+--- a/arch/arm/include/asm/mmu.h
++++ b/arch/arm/include/asm/mmu.h
+@@ -6,7 +6,7 @@
+ typedef struct {
+ #ifdef CONFIG_CPU_HAS_ASID
+ unsigned int id;
+- spinlock_t id_lock;
++ raw_spinlock_t id_lock;
+ #endif
+ unsigned int kvm_seq;
+ } mm_context_t;
+@@ -16,7 +16,7 @@ typedef struct {
- if (shared)
- free_block(cachep, shared->entry,
-@@ -3889,7 +4025,8 @@ static int alloc_kmemlist(struct kmem_ca
- }
- l3->free_limit = (1 + nr_cpus_node(node)) *
- cachep->batchcount + cachep->num;
-- spin_unlock_irq(&l3->list_lock);
-+ unlock_l3_and_free_delayed(&l3->list_lock);
-+
- kfree(shared);
- free_alien_cache(new_alien);
- continue;
-@@ -3936,17 +4073,30 @@ struct ccupdate_struct {
- struct array_cache *new[NR_CPUS];
- };
+ /* init_mm.context.id_lock should be initialized. */
+ #define INIT_MM_CONTEXT(name) \
+- .context.id_lock = __SPIN_LOCK_UNLOCKED(name.context.id_lock),
++ .context.id_lock = __RAW_SPIN_LOCK_UNLOCKED(name.context.id_lock),
+ #else
+ #define ASID(mm) (0)
+ #endif
+diff --git a/arch/arm/kernel/dma.c b/arch/arm/kernel/dma.c
+index 2c4a185..7b829d9 100644
+--- a/arch/arm/kernel/dma.c
++++ b/arch/arm/kernel/dma.c
+@@ -23,7 +23,7 @@
--static void do_ccupdate_local(void *info)
-+static void __do_ccupdate_local(void *info, int cpu)
- {
- struct ccupdate_struct *new = info;
- struct array_cache *old;
+ #include <asm/mach/dma.h>
-- check_irq_off();
-- old = cpu_cache_get(new->cachep);
-+ old = cpu_cache_get_on_cpu(new->cachep, cpu);
+-DEFINE_SPINLOCK(dma_spin_lock);
++DEFINE_RAW_SPINLOCK(dma_spin_lock);
+ EXPORT_SYMBOL(dma_spin_lock);
-- new->cachep->array[smp_processor_id()] = new->new[smp_processor_id()];
-- new->new[smp_processor_id()] = old;
-+ new->cachep->array[cpu] = new->new[cpu];
-+ new->new[cpu] = old;
-+}
-+
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+static void do_ccupdate_local(void *info)
-+{
-+ __do_ccupdate_local(info, smp_processor_id());
-+}
-+#else
-+static void do_ccupdate_local(void *info, int cpu)
-+{
-+ spin_lock_irq(&per_cpu(slab_lock, cpu).lock);
-+ __do_ccupdate_local(info, cpu);
-+ spin_unlock_irq(&per_cpu(slab_lock, cpu).lock);
+ static dma_t *dma_chan[MAX_DMA_CHANNELS];
+diff --git a/arch/arm/kernel/early_printk.c b/arch/arm/kernel/early_printk.c
+index 85aa2b2..4307653 100644
+--- a/arch/arm/kernel/early_printk.c
++++ b/arch/arm/kernel/early_printk.c
+@@ -29,28 +29,17 @@ static void early_console_write(struct console *con, const char *s, unsigned n)
+ early_write(s, n);
}
-+#endif
- /* Always called with the cache_chain_mutex held */
- static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
-@@ -3971,7 +4121,7 @@ static int do_tune_cpucache(struct kmem_
- }
- new->cachep = cachep;
-
-- on_each_cpu(do_ccupdate_local, (void *)new, 1);
-+ slab_on_each_cpu(do_ccupdate_local, (void *)new);
+-static struct console early_console = {
++static struct console early_console_dev = {
+ .name = "earlycon",
+ .write = early_console_write,
+ .flags = CON_PRINTBUFFER | CON_BOOT,
+ .index = -1,
+ };
- check_irq_on();
- cachep->batchcount = batchcount;
-@@ -3982,9 +4132,11 @@ static int do_tune_cpucache(struct kmem_
- struct array_cache *ccold = new->new[i];
- if (!ccold)
- continue;
-- spin_lock_irq(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
-+ local_spin_lock_irq(slab_lock,
-+ &cachep->nodelists[cpu_to_mem(i)]->list_lock);
- free_block(cachep, ccold->entry, ccold->avail, cpu_to_mem(i));
-- spin_unlock_irq(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
-+
-+ unlock_l3_and_free_delayed(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
- kfree(ccold);
- }
- kfree(new);
-@@ -4060,7 +4212,7 @@ static void drain_array(struct kmem_cach
- if (ac->touched && !force) {
- ac->touched = 0;
- } else {
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
- if (ac->avail) {
- tofree = force ? ac->avail : (ac->limit + 4) / 5;
- if (tofree > ac->avail)
-@@ -4070,7 +4222,7 @@ static void drain_array(struct kmem_cach
- memmove(ac->entry, &(ac->entry[tofree]),
- sizeof(void *) * ac->avail);
- }
-- spin_unlock_irq(&l3->list_lock);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
- }
+-asmlinkage void early_printk(const char *fmt, ...)
+-{
+- char buf[512];
+- int n;
+- va_list ap;
+-
+- va_start(ap, fmt);
+- n = vscnprintf(buf, sizeof(buf), fmt, ap);
+- early_write(buf, n);
+- va_end(ap);
+-}
+-
+ static int __init setup_early_printk(char *buf)
+ {
+- register_console(&early_console);
++ early_console = &early_console_dev;
++ register_console(&early_console_dev);
+ return 0;
}
-@@ -4209,7 +4361,7 @@ static int s_show(struct seq_file *m, vo
+diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c
+index 2b5b142..7460d53 100644
+--- a/arch/arm/kernel/perf_event.c
++++ b/arch/arm/kernel/perf_event.c
+@@ -420,7 +420,7 @@ armpmu_reserve_hardware(void)
continue;
- check_irq_on();
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
-
- list_for_each_entry(slabp, &l3->slabs_full, list) {
- if (slabp->inuse != cachep->num && !error)
-@@ -4234,7 +4386,7 @@ static int s_show(struct seq_file *m, vo
- if (l3->shared)
- shared_avail += l3->shared->avail;
-
-- spin_unlock_irq(&l3->list_lock);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
+ err = request_irq(irq, handle_irq,
+- IRQF_DISABLED | IRQF_NOBALANCING,
++ IRQF_NOBALANCING | IRQF_NO_THREAD,
+ "armpmu", NULL);
+ if (err) {
+ pr_warning("unable to request IRQ%d for ARM perf "
+diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c
+index 5e1e541..fd9477d 100644
+--- a/arch/arm/kernel/process.c
++++ b/arch/arm/kernel/process.c
+@@ -209,9 +209,7 @@ void cpu_idle(void)
+ }
+ leds_event(led_idle_end);
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
- num_slabs += active_slabs;
- num_objs = num_slabs * cachep->num;
-@@ -4463,13 +4615,13 @@ static int leaks_show(struct seq_file *m
- continue;
+ }
- check_irq_on();
-- spin_lock_irq(&l3->list_lock);
-+ local_spin_lock_irq(slab_lock, &l3->list_lock);
+@@ -486,6 +484,31 @@ unsigned long arch_randomize_brk(struct mm_struct *mm)
+ }
- list_for_each_entry(slabp, &l3->slabs_full, list)
- handle_slab(n, cachep, slabp);
- list_for_each_entry(slabp, &l3->slabs_partial, list)
- handle_slab(n, cachep, slabp);
-- spin_unlock_irq(&l3->list_lock);
-+ local_spin_unlock_irq(slab_lock, &l3->list_lock);
- }
- name = cachep->name;
- if (n[0] == n[1]) {
-Index: linux-3.0/kernel/lockdep.c
-===================================================================
---- linux-3.0.orig/kernel/lockdep.c
-+++ linux-3.0/kernel/lockdep.c
-@@ -2859,10 +2859,7 @@ static int mark_lock(struct task_struct
- void lockdep_init_map(struct lockdep_map *lock, const char *name,
- struct lock_class_key *key, int subclass)
- {
-- int i;
--
-- for (i = 0; i < NR_LOCKDEP_CACHING_CLASSES; i++)
-- lock->class_cache[i] = NULL;
-+ memset(lock, 0, sizeof(*lock));
-
- #ifdef CONFIG_LOCK_STAT
- lock->cpu = raw_smp_processor_id();
-@@ -3341,6 +3338,7 @@ static void check_flags(unsigned long fl
- }
- }
+ #ifdef CONFIG_MMU
++
++/*
++ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not
++ * initialized by pgtable_page_ctor() then a coredump of the vector page will
++ * fail.
++ */
++static int __init vectors_user_mapping_init_page(void)
++{
++ struct page *page;
++ unsigned long addr = 0xffff0000;
++ pgd_t *pgd;
++ pud_t *pud;
++ pmd_t *pmd;
++
++ pgd = pgd_offset_k(addr);
++ pud = pud_offset(pgd, addr);
++ pmd = pmd_offset(pud, addr);
++ page = pmd_page(*(pmd));
++
++ pgtable_page_ctor(page);
++
++ return 0;
++}
++late_initcall(vectors_user_mapping_init_page);
++
+ /*
+ * The vectors page is always readable from user space for the
+ * atomic helpers and the signal restart code. Let's declare a mapping
+diff --git a/arch/arm/kernel/signal.c b/arch/arm/kernel/signal.c
+index 0340224..5c9c403 100644
+--- a/arch/arm/kernel/signal.c
++++ b/arch/arm/kernel/signal.c
+@@ -673,6 +673,9 @@ static void do_signal(struct pt_regs *regs, int syscall)
+ if (!user_mode(regs))
+ return;
-+#ifndef CONFIG_PREEMPT_RT_FULL
++ local_irq_enable();
++ preempt_check_resched();
++
/*
- * We dont accurately track softirq state in e.g.
- * hardirq contexts (such as on 4KSTACKS), so only
-@@ -3352,6 +3350,7 @@ static void check_flags(unsigned long fl
- else
- DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled);
- }
-+#endif
-
- if (!debug_locks)
- print_irqtrace_events(current);
-Index: linux-3.0/arch/x86/kernel/apic/apic.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/apic/apic.c
-+++ linux-3.0/arch/x86/kernel/apic/apic.c
-@@ -856,8 +856,8 @@ void __irq_entry smp_apic_timer_interrup
- * Besides, if we don't timer interrupts ignore the global
- * interrupt lock, which is the WrongThing (tm) to do.
+ * If we were from a system call, check for system call restarting...
*/
-- exit_idle();
- irq_enter();
-+ exit_idle();
- local_apic_timer_interrupt();
- irq_exit();
+diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c
+index e7f92a4..a9710d3 100644
+--- a/arch/arm/kernel/smp.c
++++ b/arch/arm/kernel/smp.c
+@@ -305,6 +305,18 @@ asmlinkage void __cpuinit secondary_start_kernel(void)
+ * Enable local interrupts.
+ */
+ notify_cpu_starting(cpu);
++
++ /*
++ * OK, now it's safe to let the boot CPU continue. Wait for
++ * the CPU migration code to notice that the CPU is online
++ * before we continue. We need to do that before we enable
++ * interrupts otherwise a wakeup of a kernel thread affine to
++ * this CPU might break the affinity and let hell break lose.
++ */
++ set_cpu_online(cpu, true);
++ while (!cpu_active(cpu))
++ cpu_relax();
++
+ local_irq_enable();
+ local_fiq_enable();
-@@ -1790,8 +1790,8 @@ void smp_spurious_interrupt(struct pt_re
- {
- u32 v;
+@@ -318,15 +330,6 @@ asmlinkage void __cpuinit secondary_start_kernel(void)
+ smp_store_cpu_info(cpu);
-- exit_idle();
- irq_enter();
-+ exit_idle();
/*
- * Check if this really is a spurious interrupt and ACK it
- * if it is a vectored one. Just in case...
-@@ -1827,8 +1827,8 @@ void smp_error_interrupt(struct pt_regs
- "Illegal register address", /* APIC Error Bit 7 */
- };
-
-- exit_idle();
- irq_enter();
-+ exit_idle();
- /* First tickle the hardware, only then report what went on. -- REW */
- v0 = apic_read(APIC_ESR);
- apic_write(APIC_ESR, 0);
-Index: linux-3.0/arch/x86/kernel/apic/io_apic.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/apic/io_apic.c
-+++ linux-3.0/arch/x86/kernel/apic/io_apic.c
-@@ -2275,8 +2275,8 @@ asmlinkage void smp_irq_move_cleanup_int
- unsigned vector, me;
+- * OK, now it's safe to let the boot CPU continue. Wait for
+- * the CPU migration code to notice that the CPU is online
+- * before we continue.
+- */
+- set_cpu_online(cpu, true);
+- while (!cpu_active(cpu))
+- cpu_relax();
+-
+- /*
+ * OK, it's off to the idle thread for us
+ */
+ cpu_idle();
+@@ -531,7 +534,7 @@ static void percpu_timer_stop(void)
+ }
+ #endif
- ack_APIC_irq();
-- exit_idle();
- irq_enter();
-+ exit_idle();
+-static DEFINE_SPINLOCK(stop_lock);
++static DEFINE_RAW_SPINLOCK(stop_lock);
- me = smp_processor_id();
- for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
-@@ -2417,7 +2417,8 @@ static void ack_apic_level(struct irq_da
- irq_complete_move(cfg);
- #ifdef CONFIG_GENERIC_PENDING_IRQ
- /* If we are moving the irq we need to mask it */
-- if (unlikely(irqd_is_setaffinity_pending(data))) {
-+ if (unlikely(irqd_is_setaffinity_pending(data) &&
-+ !irqd_irq_inprogress(data))) {
- do_unmask_irq = 1;
- mask_ioapic(cfg);
+ /*
+ * ipi_cpu_stop - handle IPI from smp_send_stop()
+@@ -540,10 +543,10 @@ static void ipi_cpu_stop(unsigned int cpu)
+ {
+ if (system_state == SYSTEM_BOOTING ||
+ system_state == SYSTEM_RUNNING) {
+- spin_lock(&stop_lock);
++ raw_spin_lock(&stop_lock);
+ printk(KERN_CRIT "CPU%u: stopping\n", cpu);
+ dump_stack();
+- spin_unlock(&stop_lock);
++ raw_spin_unlock(&stop_lock);
}
-Index: linux-3.0/arch/x86/kernel/cpu/mcheck/mce.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/mcheck/mce.c
-+++ linux-3.0/arch/x86/kernel/cpu/mcheck/mce.c
-@@ -38,6 +38,7 @@
- #include <linux/mm.h>
- #include <linux/debugfs.h>
- #include <linux/edac_mce.h>
-+#include <linux/jiffies.h>
- #include <asm/processor.h>
- #include <asm/hw_irq.h>
-@@ -470,8 +471,8 @@ static inline void mce_get_rip(struct mc
- asmlinkage void smp_mce_self_interrupt(struct pt_regs *regs)
- {
- ack_APIC_irq();
-- exit_idle();
- irq_enter();
-+ exit_idle();
- mce_notify_irq();
- mce_schedule_work();
- irq_exit();
-@@ -1139,17 +1140,14 @@ void mce_log_therm_throt_event(__u64 sta
- * poller finds an MCE, poll 2x faster. When the poller finds no more
- * errors, poll 2x slower (up to check_interval seconds).
- */
--static int check_interval = 5 * 60; /* 5 minutes */
-+static unsigned long check_interval = 5 * 60; /* 5 minutes */
+ set_cpu_online(cpu, false);
+diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
+index 6807cb1..f91a5a4 100644
+--- a/arch/arm/kernel/traps.c
++++ b/arch/arm/kernel/traps.c
+@@ -255,7 +255,7 @@ static int __die(const char *str, int err, struct thread_info *thread, struct pt
+ return ret;
+ }
--static DEFINE_PER_CPU(int, mce_next_interval); /* in jiffies */
--static DEFINE_PER_CPU(struct timer_list, mce_timer);
-+static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */
-+static DEFINE_PER_CPU(struct hrtimer, mce_timer);
+-static DEFINE_SPINLOCK(die_lock);
++static DEFINE_RAW_SPINLOCK(die_lock);
--static void mce_start_timer(unsigned long data)
-+static enum hrtimer_restart mce_start_timer(struct hrtimer *timer)
- {
-- struct timer_list *t = &per_cpu(mce_timer, data);
-- int *n;
--
-- WARN_ON(smp_processor_id() != data);
-+ unsigned long *n;
+ /*
+ * This function is protected against re-entrancy.
+@@ -267,7 +267,7 @@ void die(const char *str, struct pt_regs *regs, int err)
- if (mce_available(__this_cpu_ptr(&cpu_info))) {
- machine_check_poll(MCP_TIMESTAMP,
-@@ -1162,12 +1160,13 @@ static void mce_start_timer(unsigned lon
- */
- n = &__get_cpu_var(mce_next_interval);
- if (mce_notify_irq())
-- *n = max(*n/2, HZ/100);
-+ *n = max(*n/2, HZ/100UL);
- else
-- *n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ));
-+ *n = min(*n*2, round_jiffies_relative(check_interval*HZ));
+ oops_enter();
-- t->expires = jiffies + *n;
-- add_timer_on(t, smp_processor_id());
-+ hrtimer_forward(timer, timer->base->get_time(),
-+ ns_to_ktime(jiffies_to_usecs(*n) * 1000));
-+ return HRTIMER_RESTART;
+- spin_lock_irq(&die_lock);
++ raw_spin_lock_irq(&die_lock);
+ console_verbose();
+ bust_spinlocks(1);
+ ret = __die(str, err, thread, regs);
+@@ -277,7 +277,7 @@ void die(const char *str, struct pt_regs *regs, int err)
+
+ bust_spinlocks(0);
+ add_taint(TAINT_DIE);
+- spin_unlock_irq(&die_lock);
++ raw_spin_unlock_irq(&die_lock);
+ oops_exit();
+
+ if (in_interrupt())
+@@ -302,24 +302,24 @@ void arm_notify_die(const char *str, struct pt_regs *regs,
}
- static void mce_do_trigger(struct work_struct *work)
-@@ -1393,10 +1392,11 @@ static void __mcheck_cpu_init_vendor(str
+ static LIST_HEAD(undef_hook);
+-static DEFINE_SPINLOCK(undef_lock);
++static DEFINE_RAW_SPINLOCK(undef_lock);
- static void __mcheck_cpu_init_timer(void)
+ void register_undef_hook(struct undef_hook *hook)
{
-- struct timer_list *t = &__get_cpu_var(mce_timer);
-- int *n = &__get_cpu_var(mce_next_interval);
-+ struct hrtimer *t = &__get_cpu_var(mce_timer);
-+ unsigned long *n = &__get_cpu_var(mce_next_interval);
-
-- setup_timer(t, mce_start_timer, smp_processor_id());
-+ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
-+ t->function = mce_start_timer;
+ unsigned long flags;
- if (mce_ignore_ce)
- return;
-@@ -1404,8 +1404,9 @@ static void __mcheck_cpu_init_timer(void
- *n = check_interval * HZ;
- if (!*n)
- return;
-- t->expires = round_jiffies(jiffies + *n);
-- add_timer_on(t, smp_processor_id());
-+
-+ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(*n) * 1000),
-+ 0 , HRTIMER_MODE_REL_PINNED);
+- spin_lock_irqsave(&undef_lock, flags);
++ raw_spin_lock_irqsave(&undef_lock, flags);
+ list_add(&hook->node, &undef_hook);
+- spin_unlock_irqrestore(&undef_lock, flags);
++ raw_spin_unlock_irqrestore(&undef_lock, flags);
}
- /* Handle unconfigured int18 (should never happen) */
-@@ -1768,7 +1769,7 @@ static struct syscore_ops mce_syscore_op
-
- static void mce_cpu_restart(void *data)
+ void unregister_undef_hook(struct undef_hook *hook)
{
-- del_timer_sync(&__get_cpu_var(mce_timer));
-+ hrtimer_cancel(&__get_cpu_var(mce_timer));
- if (!mce_available(__this_cpu_ptr(&cpu_info)))
- return;
- __mcheck_cpu_init_generic();
-@@ -1787,7 +1788,7 @@ static void mce_disable_ce(void *all)
- if (!mce_available(__this_cpu_ptr(&cpu_info)))
- return;
- if (all)
-- del_timer_sync(&__get_cpu_var(mce_timer));
-+ hrtimer_cancel(&__get_cpu_var(mce_timer));
- cmci_clear();
+ unsigned long flags;
+
+- spin_lock_irqsave(&undef_lock, flags);
++ raw_spin_lock_irqsave(&undef_lock, flags);
+ list_del(&hook->node);
+- spin_unlock_irqrestore(&undef_lock, flags);
++ raw_spin_unlock_irqrestore(&undef_lock, flags);
}
-@@ -2016,6 +2017,8 @@ static void __cpuinit mce_disable_cpu(vo
- if (!mce_available(__this_cpu_ptr(&cpu_info)))
- return;
+ static int call_undef_hook(struct pt_regs *regs, unsigned int instr)
+@@ -328,12 +328,12 @@ static int call_undef_hook(struct pt_regs *regs, unsigned int instr)
+ unsigned long flags;
+ int (*fn)(struct pt_regs *regs, unsigned int instr) = NULL;
-+ hrtimer_cancel(&__get_cpu_var(mce_timer));
-+
- if (!(action & CPU_TASKS_FROZEN))
- cmci_clear();
- for (i = 0; i < banks; i++) {
-@@ -2042,6 +2045,7 @@ static void __cpuinit mce_reenable_cpu(v
- if (b->init)
- wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
- }
-+ __mcheck_cpu_init_timer();
+- spin_lock_irqsave(&undef_lock, flags);
++ raw_spin_lock_irqsave(&undef_lock, flags);
+ list_for_each_entry(hook, &undef_hook, node)
+ if ((instr & hook->instr_mask) == hook->instr_val &&
+ (regs->ARM_cpsr & hook->cpsr_mask) == hook->cpsr_val)
+ fn = hook->fn;
+- spin_unlock_irqrestore(&undef_lock, flags);
++ raw_spin_unlock_irqrestore(&undef_lock, flags);
+
+ return fn ? fn(regs, instr) : 1;
}
+diff --git a/arch/arm/mach-at91/at91rm9200_time.c b/arch/arm/mach-at91/at91rm9200_time.c
+index 1dd69c8..0666570 100644
+--- a/arch/arm/mach-at91/at91rm9200_time.c
++++ b/arch/arm/mach-at91/at91rm9200_time.c
+@@ -114,6 +114,7 @@ clkevt32k_mode(enum clock_event_mode mode, struct clock_event_device *dev)
+ last_crtr = read_CRTR();
+ switch (mode) {
+ case CLOCK_EVT_MODE_PERIODIC:
++ setup_irq(AT91_ID_SYS, &at91rm9200_timer_irq);
+ /* PIT for periodic irqs; fixed rate of 1/HZ */
+ irqmask = AT91_ST_PITS;
+ at91_sys_write(AT91_ST_PIMR, LATCH);
+@@ -127,6 +128,7 @@ clkevt32k_mode(enum clock_event_mode mode, struct clock_event_device *dev)
+ break;
+ case CLOCK_EVT_MODE_SHUTDOWN:
+ case CLOCK_EVT_MODE_UNUSED:
++ remove_irq(AT91_ID_SYS, &at91rm9200_timer_irq);
+ case CLOCK_EVT_MODE_RESUME:
+ irqmask = 0;
+ break;
+diff --git a/arch/arm/mach-at91/at91sam926x_time.c b/arch/arm/mach-at91/at91sam926x_time.c
+index 4ba8549..97d1e14 100644
+--- a/arch/arm/mach-at91/at91sam926x_time.c
++++ b/arch/arm/mach-at91/at91sam926x_time.c
+@@ -54,7 +54,7 @@ static struct clocksource pit_clk = {
+ .flags = CLOCK_SOURCE_IS_CONTINUOUS,
+ };
- /* Get notified when a cpu comes on/off. Be hotplug friendly. */
-@@ -2049,7 +2053,6 @@ static int __cpuinit
- mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
+-
++static struct irqaction at91sam926x_pit_irq;
+ /*
+ * Clockevent device: interrupts every 1/HZ (== pit_cycles * MCK/16)
+ */
+@@ -63,6 +63,9 @@ pit_clkevt_mode(enum clock_event_mode mode, struct clock_event_device *dev)
{
- unsigned int cpu = (unsigned long)hcpu;
-- struct timer_list *t = &per_cpu(mce_timer, cpu);
-
- switch (action) {
- case CPU_ONLINE:
-@@ -2066,16 +2069,10 @@ mce_cpu_callback(struct notifier_block *
- break;
- case CPU_DOWN_PREPARE:
- case CPU_DOWN_PREPARE_FROZEN:
-- del_timer_sync(t);
- smp_call_function_single(cpu, mce_disable_cpu, &action, 1);
+ switch (mode) {
+ case CLOCK_EVT_MODE_PERIODIC:
++ /* Set up irq handler */
++ setup_irq(AT91_ID_SYS, &at91sam926x_pit_irq);
++
+ /* update clocksource counter */
+ pit_cnt += pit_cycle * PIT_PICNT(at91_sys_read(AT91_PIT_PIVR));
+ at91_sys_write(AT91_PIT_MR, (pit_cycle - 1) | AT91_PIT_PITEN
+@@ -75,6 +78,7 @@ pit_clkevt_mode(enum clock_event_mode mode, struct clock_event_device *dev)
+ case CLOCK_EVT_MODE_UNUSED:
+ /* disable irq, leaving the clocksource active */
+ at91_sys_write(AT91_PIT_MR, (pit_cycle - 1) | AT91_PIT_PITEN);
++ remove_irq(AT91_ID_SYS, &at91sam926x_pit_irq);
break;
- case CPU_DOWN_FAILED:
- case CPU_DOWN_FAILED_FROZEN:
-- if (!mce_ignore_ce && check_interval) {
-- t->expires = round_jiffies(jiffies +
-- __get_cpu_var(mce_next_interval));
-- add_timer_on(t, cpu);
-- }
- smp_call_function_single(cpu, mce_reenable_cpu, &action, 1);
+ case CLOCK_EVT_MODE_RESUME:
break;
- case CPU_POST_DEAD:
-Index: linux-3.0/arch/x86/kernel/cpu/mcheck/therm_throt.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/mcheck/therm_throt.c
-+++ linux-3.0/arch/x86/kernel/cpu/mcheck/therm_throt.c
-@@ -396,8 +396,8 @@ static void (*smp_thermal_vector)(void)
+diff --git a/arch/arm/mach-exynos4/platsmp.c b/arch/arm/mach-exynos4/platsmp.c
+index c5e65a0..6579fdd 100644
+--- a/arch/arm/mach-exynos4/platsmp.c
++++ b/arch/arm/mach-exynos4/platsmp.c
+@@ -56,7 +56,7 @@ static void __iomem *scu_base_addr(void)
+ return (void __iomem *)(S5P_VA_SCU);
+ }
- asmlinkage void smp_thermal_interrupt(struct pt_regs *regs)
- {
-- exit_idle();
- irq_enter();
-+ exit_idle();
- inc_irq_stat(irq_thermal_count);
- smp_thermal_vector();
- irq_exit();
-Index: linux-3.0/arch/x86/kernel/cpu/mcheck/threshold.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/mcheck/threshold.c
-+++ linux-3.0/arch/x86/kernel/cpu/mcheck/threshold.c
-@@ -19,8 +19,8 @@ void (*mce_threshold_vector)(void) = def
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
- asmlinkage void smp_threshold_interrupt(void)
+ void __cpuinit platform_secondary_init(unsigned int cpu)
{
-- exit_idle();
- irq_enter();
-+ exit_idle();
- inc_irq_stat(irq_threshold_count);
- mce_threshold_vector();
- irq_exit();
-Index: linux-3.0/arch/x86/kernel/irq.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/irq.c
-+++ linux-3.0/arch/x86/kernel/irq.c
-@@ -180,8 +180,8 @@ unsigned int __irq_entry do_IRQ(struct p
- unsigned vector = ~regs->orig_ax;
- unsigned irq;
-
-- exit_idle();
- irq_enter();
-+ exit_idle();
-
- irq = __this_cpu_read(vector_irq[vector]);
-
-@@ -208,10 +208,10 @@ void smp_x86_platform_ipi(struct pt_regs
-
- ack_APIC_irq();
-
-- exit_idle();
--
- irq_enter();
-
-+ exit_idle();
-+
- inc_irq_stat(x86_platform_ipis);
-
- if (x86_platform_ipi_callback)
- static struct genl_ops cgroupstats_ops = {
-Index: linux-3.0/kernel/trace/ftrace.c
-===================================================================
---- linux-3.0.orig/kernel/trace/ftrace.c
-+++ linux-3.0/kernel/trace/ftrace.c
-@@ -1182,8 +1182,14 @@ alloc_and_copy_ftrace_hash(int size_bits
- return NULL;
+@@ -76,8 +76,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
+ /*
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
}
-+static void
-+ftrace_hash_rec_disable(struct ftrace_ops *ops, int filter_hash);
-+static void
-+ftrace_hash_rec_enable(struct ftrace_ops *ops, int filter_hash);
-+
- static int
--ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src)
-+ftrace_hash_move(struct ftrace_ops *ops, int enable,
-+ struct ftrace_hash **dst, struct ftrace_hash *src)
- {
- struct ftrace_func_entry *entry;
- struct hlist_node *tp, *tn;
-@@ -1193,9 +1199,16 @@ ftrace_hash_move(struct ftrace_hash **ds
- unsigned long key;
- int size = src->count;
- int bits = 0;
-+ int ret;
- int i;
+ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+@@ -88,7 +88,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * Set synchronisation state between this boot processor
+ * and the secondary one
+ */
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
/*
-+ * Remove the current set, update the hash and add
-+ * them back.
-+ */
-+ ftrace_hash_rec_disable(ops, enable);
-+
-+ /*
- * If the new source is empty, just free dst and assign it
- * the empty_hash.
+ * The secondary processor is waiting to be released from
+@@ -120,7 +120,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
*/
-@@ -1215,9 +1228,10 @@ ftrace_hash_move(struct ftrace_hash **ds
- if (bits > FTRACE_HASH_MAX_BITS)
- bits = FTRACE_HASH_MAX_BITS;
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
-+ ret = -ENOMEM;
- new_hash = alloc_ftrace_hash(bits);
- if (!new_hash)
-- return -ENOMEM;
-+ goto out;
+ return pen_release != -1 ? -ENOSYS : 0;
+ }
+diff --git a/arch/arm/mach-footbridge/include/mach/hardware.h b/arch/arm/mach-footbridge/include/mach/hardware.h
+index b6fdf23..14db5a0 100644
+--- a/arch/arm/mach-footbridge/include/mach/hardware.h
++++ b/arch/arm/mach-footbridge/include/mach/hardware.h
+@@ -93,7 +93,7 @@
+ #define CPLD_FLASH_WR_ENABLE 1
- size = 1 << src->size_bits;
- for (i = 0; i < size; i++) {
-@@ -1236,7 +1250,16 @@ ftrace_hash_move(struct ftrace_hash **ds
- rcu_assign_pointer(*dst, new_hash);
- free_ftrace_hash_rcu(old_hash);
+ #ifndef __ASSEMBLY__
+-extern spinlock_t nw_gpio_lock;
++extern raw_spinlock_t nw_gpio_lock;
+ extern void nw_gpio_modify_op(unsigned int mask, unsigned int set);
+ extern void nw_gpio_modify_io(unsigned int mask, unsigned int in);
+ extern unsigned int nw_gpio_read(void);
+diff --git a/arch/arm/mach-footbridge/netwinder-hw.c b/arch/arm/mach-footbridge/netwinder-hw.c
+index 06e514f..5b73190 100644
+--- a/arch/arm/mach-footbridge/netwinder-hw.c
++++ b/arch/arm/mach-footbridge/netwinder-hw.c
+@@ -68,7 +68,7 @@ static inline void wb977_ww(int reg, int val)
+ /*
+ * This is a lock for accessing ports GP1_IO_BASE and GP2_IO_BASE
+ */
+-DEFINE_SPINLOCK(nw_gpio_lock);
++DEFINE_RAW_SPINLOCK(nw_gpio_lock);
+ EXPORT_SYMBOL(nw_gpio_lock);
-- return 0;
-+ ret = 0;
-+ out:
-+ /*
-+ * Enable regardless of ret:
-+ * On success, we enable the new hash.
-+ * On failure, we re-enable the original hash.
-+ */
-+ ftrace_hash_rec_enable(ops, enable);
-+
-+ return ret;
+ static unsigned int current_gpio_op;
+@@ -327,9 +327,9 @@ static inline void wb977_init_gpio(void)
+ /*
+ * Set Group1/Group2 outputs
+ */
+- spin_lock_irqsave(&nw_gpio_lock, flags);
++ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
+ nw_gpio_modify_op(-1, GPIO_RED_LED | GPIO_FAN);
+- spin_unlock_irqrestore(&nw_gpio_lock, flags);
++ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
}
/*
-@@ -2877,7 +2900,7 @@ ftrace_set_regex(struct ftrace_ops *ops,
- ftrace_match_records(hash, buf, len);
+@@ -390,9 +390,9 @@ static void __init cpld_init(void)
+ {
+ unsigned long flags;
- mutex_lock(&ftrace_lock);
-- ret = ftrace_hash_move(orig_hash, hash);
-+ ret = ftrace_hash_move(ops, enable, orig_hash, hash);
- mutex_unlock(&ftrace_lock);
+- spin_lock_irqsave(&nw_gpio_lock, flags);
++ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
+ nw_cpld_modify(-1, CPLD_UNMUTE | CPLD_7111_DISABLE);
+- spin_unlock_irqrestore(&nw_gpio_lock, flags);
++ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
+ }
- mutex_unlock(&ftrace_regex_lock);
-@@ -3060,18 +3083,12 @@ ftrace_regex_release(struct inode *inode
- orig_hash = &iter->ops->notrace_hash;
+ static unsigned char rwa_unlock[] __initdata =
+@@ -616,9 +616,9 @@ static int __init nw_hw_init(void)
+ cpld_init();
+ rwa010_init();
- mutex_lock(&ftrace_lock);
-- /*
-- * Remove the current set, update the hash and add
-- * them back.
-- */
-- ftrace_hash_rec_disable(iter->ops, filter_hash);
-- ret = ftrace_hash_move(orig_hash, iter->hash);
-- if (!ret) {
-- ftrace_hash_rec_enable(iter->ops, filter_hash);
-- if (iter->ops->flags & FTRACE_OPS_FL_ENABLED
-- && ftrace_enabled)
-- ftrace_run_update_code(FTRACE_ENABLE_CALLS);
-- }
-+ ret = ftrace_hash_move(iter->ops, filter_hash,
-+ orig_hash, iter->hash);
-+ if (!ret && (iter->ops->flags & FTRACE_OPS_FL_ENABLED)
-+ && ftrace_enabled)
-+ ftrace_run_update_code(FTRACE_ENABLE_CALLS);
-+
- mutex_unlock(&ftrace_lock);
+- spin_lock_irqsave(&nw_gpio_lock, flags);
++ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
+ nw_gpio_modify_op(GPIO_RED_LED|GPIO_GREEN_LED, DEFAULT_LEDS);
+- spin_unlock_irqrestore(&nw_gpio_lock, flags);
++ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
}
- free_ftrace_hash(iter->hash);
-Index: linux-3.0/drivers/gpu/drm/drm_irq.c
-===================================================================
---- linux-3.0.orig/drivers/gpu/drm/drm_irq.c
-+++ linux-3.0/drivers/gpu/drm/drm_irq.c
-@@ -109,10 +109,7 @@ static void vblank_disable_and_save(stru
- /* Prevent vblank irq processing while disabling vblank irqs,
- * so no updates of timestamps or count can happen after we've
- * disabled. Needed to prevent races in case of delayed irq's.
-- * Disable preemption, so vblank_time_lock is held as short as
-- * possible, even under a kernel with PREEMPT_RT patches.
- */
-- preempt_disable();
- spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
-
- dev->driver->disable_vblank(dev, crtc);
-@@ -163,7 +160,6 @@ static void vblank_disable_and_save(stru
- clear_vblank_timestamps(dev, crtc);
-
- spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
-- preempt_enable();
+ return 0;
}
+diff --git a/arch/arm/mach-footbridge/netwinder-leds.c b/arch/arm/mach-footbridge/netwinder-leds.c
+index 00269fe..e57102e 100644
+--- a/arch/arm/mach-footbridge/netwinder-leds.c
++++ b/arch/arm/mach-footbridge/netwinder-leds.c
+@@ -31,13 +31,13 @@
+ static char led_state;
+ static char hw_led_state;
- static void vblank_disable_fn(unsigned long arg)
-@@ -875,10 +871,6 @@ int drm_vblank_get(struct drm_device *de
- spin_lock_irqsave(&dev->vbl_lock, irqflags);
- /* Going from 0->1 means we have to enable interrupts again */
- if (atomic_add_return(1, &dev->vblank_refcount[crtc]) == 1) {
-- /* Disable preemption while holding vblank_time_lock. Do
-- * it explicitely to guard against PREEMPT_RT kernel.
-- */
-- preempt_disable();
- spin_lock_irqsave(&dev->vblank_time_lock, irqflags2);
- if (!dev->vblank_enabled[crtc]) {
- /* Enable vblank irqs under vblank_time_lock protection.
-@@ -898,7 +890,6 @@ int drm_vblank_get(struct drm_device *de
- }
- }
- spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags2);
-- preempt_enable();
- } else {
- if (!dev->vblank_enabled[crtc]) {
- atomic_dec(&dev->vblank_refcount[crtc]);
-Index: linux-3.0/arch/x86/kernel/kprobes.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/kprobes.c
-+++ linux-3.0/arch/x86/kernel/kprobes.c
-@@ -477,7 +477,6 @@ static void __kprobes setup_singlestep(s
- * stepping.
- */
- regs->ip = (unsigned long)p->ainsn.insn;
-- preempt_enable_no_resched();
- return;
- }
- #endif
-Index: linux-3.0/drivers/ide/ide_platform.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide_platform.c
-+++ linux-3.0/drivers/ide/ide_platform.c
-@@ -95,7 +95,7 @@ static int __devinit plat_ide_probe(stru
- plat_ide_setup_ports(&hw, base, alt_base, pdata, res_irq->start);
- hw.dev = &pdev->dev;
-
-- d.irq_flags = res_irq->flags;
-+ d.irq_flags = 0;
- if (mmio)
- d.host_flags |= IDE_HFLAG_MMIO;
-
-Index: linux-3.0/arch/x86/kernel/hpet.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/hpet.c
-+++ linux-3.0/arch/x86/kernel/hpet.c
-@@ -7,6 +7,7 @@
- #include <linux/slab.h>
- #include <linux/hpet.h>
- #include <linux/init.h>
-+#include <linux/dmi.h>
- #include <linux/cpu.h>
- #include <linux/pm.h>
- #include <linux/io.h>
-@@ -566,6 +567,30 @@ static void init_one_hpet_msi_clockevent
- #define RESERVE_TIMERS 0
- #endif
+-static DEFINE_SPINLOCK(leds_lock);
++static DEFINE_RAW_SPINLOCK(leds_lock);
-+static int __init dmi_disable_hpet_msi(const struct dmi_system_id *d)
-+{
-+ hpet_msi_disable = 1;
-+ return 0;
-+}
-+
-+static struct dmi_system_id __initdata dmi_hpet_table[] = {
-+ /*
-+ * MSI based per cpu timers lose interrupts when intel_idle()
-+ * is enabled - independent of the c-state. With idle=poll the
-+ * problem cannot be observed. We have no idea yet, whether
-+ * this is a W510 specific issue or a general chipset oddity.
-+ */
-+ {
-+ .callback = dmi_disable_hpet_msi,
-+ .ident = "Lenovo W510",
-+ .matches = {
-+ DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
-+ DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad W510"),
-+ },
-+ },
-+ {}
-+};
-+
- static void hpet_msi_capability_lookup(unsigned int start_timer)
+ static void netwinder_leds_event(led_event_t evt)
{
- unsigned int id;
-@@ -573,6 +598,8 @@ static void hpet_msi_capability_lookup(u
- unsigned int num_timers_used = 0;
- int i;
-
-+ dmi_check_system(dmi_hpet_table);
-+
- if (hpet_msi_disable)
- return;
+ unsigned long flags;
-Index: linux-3.0/block/blk-core.c
-===================================================================
---- linux-3.0.orig/block/blk-core.c
-+++ linux-3.0/block/blk-core.c
-@@ -236,7 +236,7 @@ EXPORT_SYMBOL(blk_delay_queue);
- **/
- void blk_start_queue(struct request_queue *q)
- {
-- WARN_ON(!irqs_disabled());
-+ WARN_ON_NONRT(!irqs_disabled());
+- spin_lock_irqsave(&leds_lock, flags);
++ raw_spin_lock_irqsave(&leds_lock, flags);
- queue_flag_clear(QUEUE_FLAG_STOPPED, q);
- __blk_run_queue(q);
-@@ -301,7 +301,11 @@ void __blk_run_queue(struct request_queu
- {
- if (unlikely(blk_queue_stopped(q)))
- return;
--
-+ /*
-+ * q->request_fn() can drop q->queue_lock and reenable
-+ * interrupts, but must return with q->queue_lock held and
-+ * interrupts disabled.
-+ */
- q->request_fn(q);
- }
- EXPORT_SYMBOL(__blk_run_queue);
-@@ -2669,11 +2673,11 @@ static void queue_unplugged(struct reque
- * this lock).
- */
- if (from_schedule) {
-- spin_unlock(q->queue_lock);
-+ spin_unlock_irq(q->queue_lock);
- blk_run_queue_async(q);
- } else {
- __blk_run_queue(q);
-- spin_unlock(q->queue_lock);
-+ spin_unlock_irq(q->queue_lock);
+ switch (evt) {
+ case led_start:
+@@ -117,12 +117,12 @@ static void netwinder_leds_event(led_event_t evt)
+ break;
}
- }
-@@ -2699,7 +2703,6 @@ static void flush_plug_callbacks(struct
- void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
- {
- struct request_queue *q;
-- unsigned long flags;
- struct request *rq;
- LIST_HEAD(list);
- unsigned int depth;
-@@ -2720,11 +2723,6 @@ void blk_flush_plug_list(struct blk_plug
- q = NULL;
- depth = 0;
+- spin_unlock_irqrestore(&leds_lock, flags);
++ raw_spin_unlock_irqrestore(&leds_lock, flags);
-- /*
-- * Save and disable interrupts here, to avoid doing it for every
-- * queue lock we have to take.
-- */
-- local_irq_save(flags);
- while (!list_empty(&list)) {
- rq = list_entry_rq(list.next);
- list_del_init(&rq->queuelist);
-@@ -2737,7 +2735,7 @@ void blk_flush_plug_list(struct blk_plug
- queue_unplugged(q, depth, from_schedule);
- q = rq->q;
- depth = 0;
-- spin_lock(q->queue_lock);
-+ spin_lock_irq(q->queue_lock);
- }
- /*
- * rq is already accounted, so use raw insert
-@@ -2755,8 +2753,6 @@ void blk_flush_plug_list(struct blk_plug
- */
- if (q)
- queue_unplugged(q, depth, from_schedule);
--
-- local_irq_restore(flags);
+ if (led_state & LED_STATE_ENABLED) {
+- spin_lock_irqsave(&nw_gpio_lock, flags);
++ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
+ nw_gpio_modify_op(GPIO_RED_LED | GPIO_GREEN_LED, hw_led_state);
+- spin_unlock_irqrestore(&nw_gpio_lock, flags);
++ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
+ }
}
- void blk_finish_plug(struct blk_plug *plug)
-Index: linux-3.0/kernel/sched.c
-===================================================================
---- linux-3.0.orig/kernel/sched.c
-+++ linux-3.0/kernel/sched.c
-@@ -71,6 +71,7 @@
- #include <linux/ctype.h>
- #include <linux/ftrace.h>
- #include <linux/slab.h>
-+#include <linux/init_task.h>
+diff --git a/arch/arm/mach-integrator/core.c b/arch/arm/mach-integrator/core.c
+index 77315b9..0c20cf6 100644
+--- a/arch/arm/mach-integrator/core.c
++++ b/arch/arm/mach-integrator/core.c
+@@ -205,7 +205,7 @@ static struct amba_pl010_data integrator_uart_data = {
- #include <asm/tlb.h>
- #include <asm/irq_regs.h>
-@@ -185,6 +186,7 @@ void init_rt_bandwidth(struct rt_bandwid
+ #define CM_CTRL IO_ADDRESS(INTEGRATOR_HDR_CTRL)
- hrtimer_init(&rt_b->rt_period_timer,
- CLOCK_MONOTONIC, HRTIMER_MODE_REL);
-+ rt_b->rt_period_timer.irqsafe = 1;
- rt_b->rt_period_timer.function = sched_rt_period_timer;
+-static DEFINE_SPINLOCK(cm_lock);
++static DEFINE_RAW_SPINLOCK(cm_lock);
+
+ /**
+ * cm_control - update the CM_CTRL register.
+@@ -217,10 +217,10 @@ void cm_control(u32 mask, u32 set)
+ unsigned long flags;
+ u32 val;
+
+- spin_lock_irqsave(&cm_lock, flags);
++ raw_spin_lock_irqsave(&cm_lock, flags);
+ val = readl(CM_CTRL) & ~mask;
+ writel(val | set, CM_CTRL);
+- spin_unlock_irqrestore(&cm_lock, flags);
++ raw_spin_unlock_irqrestore(&cm_lock, flags);
}
-@@ -800,7 +802,11 @@ late_initcall(sched_init_debug);
- * Number of tasks to iterate in a single balance run.
- * Limited because this is done with IRQs disabled.
+ EXPORT_SYMBOL(cm_control);
+diff --git a/arch/arm/mach-integrator/pci_v3.c b/arch/arm/mach-integrator/pci_v3.c
+index 6467d99..a9b90a0 100644
+--- a/arch/arm/mach-integrator/pci_v3.c
++++ b/arch/arm/mach-integrator/pci_v3.c
+@@ -163,7 +163,7 @@
+ * 7:2 register number
+ *
*/
-+#ifndef CONFIG_PREEMPT_RT_FULL
- const_debug unsigned int sysctl_sched_nr_migrate = 32;
-+#else
-+const_debug unsigned int sysctl_sched_nr_migrate = 8;
-+#endif
+-static DEFINE_SPINLOCK(v3_lock);
++static DEFINE_RAW_SPINLOCK(v3_lock);
- /*
- * period over which we average the RT time consumption, measured
-@@ -1136,6 +1142,7 @@ static void init_rq_hrtick(struct rq *rq
+ #define PCI_BUS_NONMEM_START 0x00000000
+ #define PCI_BUS_NONMEM_SIZE SZ_256M
+@@ -284,7 +284,7 @@ static int v3_read_config(struct pci_bus *bus, unsigned int devfn, int where,
+ unsigned long flags;
+ u32 v;
- hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
- rq->hrtick_timer.function = hrtick;
-+ rq->hrtick_timer.irqsafe = 1;
- }
- #else /* CONFIG_SCHED_HRTICK */
- static inline void hrtick_clear(struct rq *rq)
-@@ -2378,11 +2385,11 @@ static int select_fallback_rq(int cpu, s
+- spin_lock_irqsave(&v3_lock, flags);
++ raw_spin_lock_irqsave(&v3_lock, flags);
+ addr = v3_open_config_window(bus, devfn, where);
- /* Look for allowed, online CPU in same node. */
- for_each_cpu_and(dest_cpu, nodemask, cpu_active_mask)
-- if (cpumask_test_cpu(dest_cpu, &p->cpus_allowed))
-+ if (cpumask_test_cpu(dest_cpu, tsk_cpus_allowed(p)))
- return dest_cpu;
+ switch (size) {
+@@ -302,7 +302,7 @@ static int v3_read_config(struct pci_bus *bus, unsigned int devfn, int where,
+ }
- /* Any allowed, online CPU? */
-- dest_cpu = cpumask_any_and(&p->cpus_allowed, cpu_active_mask);
-+ dest_cpu = cpumask_any_and(tsk_cpus_allowed(p), cpu_active_mask);
- if (dest_cpu < nr_cpu_ids)
- return dest_cpu;
+ v3_close_config_window();
+- spin_unlock_irqrestore(&v3_lock, flags);
++ raw_spin_unlock_irqrestore(&v3_lock, flags);
-@@ -2397,7 +2404,12 @@ static int select_fallback_rq(int cpu, s
- printk(KERN_INFO "process %d (%s) no longer affine to cpu%d\n",
- task_pid_nr(p), p->comm, cpu);
+ *val = v;
+ return PCIBIOS_SUCCESSFUL;
+@@ -314,7 +314,7 @@ static int v3_write_config(struct pci_bus *bus, unsigned int devfn, int where,
+ unsigned long addr;
+ unsigned long flags;
+
+- spin_lock_irqsave(&v3_lock, flags);
++ raw_spin_lock_irqsave(&v3_lock, flags);
+ addr = v3_open_config_window(bus, devfn, where);
+
+ switch (size) {
+@@ -335,7 +335,7 @@ static int v3_write_config(struct pci_bus *bus, unsigned int devfn, int where,
}
--
-+ /*
-+ * Clear PF_THREAD_BOUND, otherwise we wreckage
-+ * migrate_disable/enable. See optimization for
-+ * PF_THREAD_BOUND tasks there.
-+ */
-+ p->flags &= ~PF_THREAD_BOUND;
- return dest_cpu;
+
+ v3_close_config_window();
+- spin_unlock_irqrestore(&v3_lock, flags);
++ raw_spin_unlock_irqrestore(&v3_lock, flags);
+
+ return PCIBIOS_SUCCESSFUL;
}
+@@ -510,7 +510,7 @@ void __init pci_v3_preinit(void)
+ hook_fault_code(8, v3_pci_fault, SIGBUS, 0, "external abort on non-linefetch");
+ hook_fault_code(10, v3_pci_fault, SIGBUS, 0, "external abort on non-linefetch");
-@@ -2419,7 +2431,7 @@ int select_task_rq(struct task_struct *p
- * [ this allows ->select_task() to simply return task_cpu(p) and
- * not worry about this generic constraint ]
- */
-- if (unlikely(!cpumask_test_cpu(cpu, &p->cpus_allowed) ||
-+ if (unlikely(!cpumask_test_cpu(cpu, tsk_cpus_allowed(p)) ||
- !cpu_online(cpu)))
- cpu = select_fallback_rq(task_cpu(p), p);
+- spin_lock_irqsave(&v3_lock, flags);
++ raw_spin_lock_irqsave(&v3_lock, flags);
+
+ /*
+ * Unlock V3 registers, but only if they were previously locked.
+@@ -583,7 +583,7 @@ void __init pci_v3_preinit(void)
+ printk(KERN_ERR "PCI: unable to grab PCI error "
+ "interrupt: %d\n", ret);
+
+- spin_unlock_irqrestore(&v3_lock, flags);
++ raw_spin_unlock_irqrestore(&v3_lock, flags);
+ }
+
+ void __init pci_v3_postinit(void)
+diff --git a/arch/arm/mach-ixp4xx/common-pci.c b/arch/arm/mach-ixp4xx/common-pci.c
+index e9a5893..ab19445 100644
+--- a/arch/arm/mach-ixp4xx/common-pci.c
++++ b/arch/arm/mach-ixp4xx/common-pci.c
+@@ -54,7 +54,7 @@ unsigned long ixp4xx_pci_reg_base = 0;
+ * these transactions are atomic or we will end up
+ * with corrupt data on the bus or in a driver.
+ */
+-static DEFINE_SPINLOCK(ixp4xx_pci_lock);
++static DEFINE_RAW_SPINLOCK(ixp4xx_pci_lock);
-@@ -2477,10 +2489,6 @@ static void ttwu_activate(struct rq *rq,
+ /*
+ * Read from PCI config space
+@@ -62,10 +62,10 @@ static DEFINE_SPINLOCK(ixp4xx_pci_lock);
+ static void crp_read(u32 ad_cbe, u32 *data)
{
- activate_task(rq, p, en_flags);
- p->on_rq = 1;
--
-- /* if a worker is waking up, notify workqueue */
-- if (p->flags & PF_WQ_WORKER)
-- wq_worker_waking_up(p, cpu_of(rq));
+ unsigned long flags;
+- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
++ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
+ *PCI_CRP_AD_CBE = ad_cbe;
+ *data = *PCI_CRP_RDATA;
+- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
++ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
}
/*
-@@ -2678,8 +2686,25 @@ try_to_wake_up(struct task_struct *p, un
+@@ -74,10 +74,10 @@ static void crp_read(u32 ad_cbe, u32 *data)
+ static void crp_write(u32 ad_cbe, u32 data)
+ {
+ unsigned long flags;
+- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
++ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
+ *PCI_CRP_AD_CBE = CRP_AD_CBE_WRITE | ad_cbe;
+ *PCI_CRP_WDATA = data;
+- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
++ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
+ }
- smp_wmb();
- raw_spin_lock_irqsave(&p->pi_lock, flags);
-- if (!(p->state & state))
-+ if (!(p->state & state)) {
-+ /*
-+ * The task might be running due to a spinlock sleeper
-+ * wakeup. Check the saved state and set it to running
-+ * if the wakeup condition is true.
-+ */
-+ if (!(wake_flags & WF_LOCK_SLEEPER)) {
-+ if (p->saved_state & state)
-+ p->saved_state = TASK_RUNNING;
-+ }
- goto out;
-+ }
-+
-+ /*
-+ * If this is a regular wakeup, then we can unconditionally
-+ * clear the saved state of a "lock sleeper".
-+ */
-+ if (!(wake_flags & WF_LOCK_SLEEPER))
-+ p->saved_state = TASK_RUNNING;
+ static inline int check_master_abort(void)
+@@ -101,7 +101,7 @@ int ixp4xx_pci_read_errata(u32 addr, u32 cmd, u32* data)
+ int retval = 0;
+ int i;
- success = 1; /* we're going to change ->state */
- cpu = task_cpu(p);
-@@ -2735,40 +2760,6 @@ out:
+- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
++ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
+
+ *PCI_NP_AD = addr;
+
+@@ -118,7 +118,7 @@ int ixp4xx_pci_read_errata(u32 addr, u32 cmd, u32* data)
+ if(check_master_abort())
+ retval = 1;
+
+- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
++ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
+ return retval;
}
- /**
-- * try_to_wake_up_local - try to wake up a local task with rq lock held
-- * @p: the thread to be awakened
-- *
-- * Put @p on the run-queue if it's not already there. The caller must
-- * ensure that this_rq() is locked, @p is bound to this_rq() and not
-- * the current task.
-- */
--static void try_to_wake_up_local(struct task_struct *p)
--{
-- struct rq *rq = task_rq(p);
--
-- BUG_ON(rq != this_rq());
-- BUG_ON(p == current);
-- lockdep_assert_held(&rq->lock);
--
-- if (!raw_spin_trylock(&p->pi_lock)) {
-- raw_spin_unlock(&rq->lock);
-- raw_spin_lock(&p->pi_lock);
-- raw_spin_lock(&rq->lock);
-- }
--
-- if (!(p->state & TASK_NORMAL))
-- goto out;
--
-- if (!p->on_rq)
-- ttwu_activate(rq, p, ENQUEUE_WAKEUP);
--
-- ttwu_do_wakeup(rq, p, 0);
-- ttwu_stat(p, smp_processor_id(), 0);
--out:
-- raw_spin_unlock(&p->pi_lock);
--}
--
--/**
- * wake_up_process - Wake up a specific process
- * @p: The process to be woken up.
- *
-@@ -2785,6 +2776,18 @@ int wake_up_process(struct task_struct *
+@@ -127,7 +127,7 @@ int ixp4xx_pci_read_no_errata(u32 addr, u32 cmd, u32* data)
+ unsigned long flags;
+ int retval = 0;
+
+- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
++ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
+
+ *PCI_NP_AD = addr;
+
+@@ -140,7 +140,7 @@ int ixp4xx_pci_read_no_errata(u32 addr, u32 cmd, u32* data)
+ if(check_master_abort())
+ retval = 1;
+
+- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
++ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
+ return retval;
}
- EXPORT_SYMBOL(wake_up_process);
-+/**
-+ * wake_up_lock_sleeper - Wake up a specific process blocked on a "sleeping lock"
-+ * @p: The process to be woken up.
-+ *
-+ * Same as wake_up_process() above, but wake_flags=WF_LOCK_SLEEPER to indicate
-+ * the nature of the wakeup.
-+ */
-+int wake_up_lock_sleeper(struct task_struct *p)
-+{
-+ return try_to_wake_up(p, TASK_ALL, WF_LOCK_SLEEPER);
-+}
-+
- int wake_up_state(struct task_struct *p, unsigned int state)
- {
- return try_to_wake_up(p, state, 0);
-@@ -3060,8 +3063,12 @@ static void finish_task_switch(struct rq
- finish_lock_switch(rq, prev);
+@@ -149,7 +149,7 @@ int ixp4xx_pci_write(u32 addr, u32 cmd, u32 data)
+ unsigned long flags;
+ int retval = 0;
- fire_sched_in_preempt_notifiers(current);
-+ /*
-+ * We use mmdrop_delayed() here so we don't have to do the
-+ * full __mmdrop() when we are the last user.
-+ */
- if (mm)
-- mmdrop(mm);
-+ mmdrop_delayed(mm);
- if (unlikely(prev_state == TASK_DEAD)) {
- /*
- * Remove function-return probe instances associated with this
-@@ -4182,6 +4189,126 @@ static inline void schedule_debug(struct
- schedstat_inc(this_rq(), sched_count);
+- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
++ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
+
+ *PCI_NP_AD = addr;
+
+@@ -162,7 +162,7 @@ int ixp4xx_pci_write(u32 addr, u32 cmd, u32 data)
+ if(check_master_abort())
+ retval = 1;
+
+- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
++ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
+ return retval;
}
-+#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
-+#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
-+#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
-+#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
-+
-+static inline void update_migrate_disable(struct task_struct *p)
-+{
-+ const struct cpumask *mask;
-+
-+ if (likely(!p->migrate_disable))
-+ return;
-+
-+ /* Did we already update affinity? */
-+ if (unlikely(migrate_disabled_updated(p)))
-+ return;
-+
-+ /*
-+ * Since this is always current we can get away with only locking
-+ * rq->lock, the ->cpus_allowed value can normally only be changed
-+ * while holding both p->pi_lock and rq->lock, but seeing that this
-+ * is current, we cannot actually be waking up, so all code that
-+ * relies on serialization against p->pi_lock is out of scope.
-+ *
-+ * Having rq->lock serializes us against things like
-+ * set_cpus_allowed_ptr() that can still happen concurrently.
-+ */
-+ mask = tsk_cpus_allowed(p);
-+
-+ if (p->sched_class->set_cpus_allowed)
-+ p->sched_class->set_cpus_allowed(p, mask);
-+ p->rt.nr_cpus_allowed = cpumask_weight(mask);
-+
-+ /* Let migrate_enable know to fix things back up */
-+ p->migrate_disable |= MIGRATE_DISABLE_SET_AFFIN;
-+}
-+
-+void migrate_disable(void)
-+{
-+ struct task_struct *p = current;
-+
-+ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
-+#ifdef CONFIG_SCHED_DEBUG
-+ p->migrate_disable_atomic++;
-+#endif
-+ return;
-+ }
-+
-+#ifdef CONFIG_SCHED_DEBUG
-+ WARN_ON_ONCE(p->migrate_disable_atomic);
-+#endif
-+
-+ preempt_disable();
-+ if (p->migrate_disable) {
-+ p->migrate_disable++;
-+ preempt_enable();
-+ return;
-+ }
-+
-+ pin_current_cpu();
-+ p->migrate_disable = 1;
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL_GPL(migrate_disable);
-+
-+void migrate_enable(void)
-+{
-+ struct task_struct *p = current;
-+ const struct cpumask *mask;
-+ unsigned long flags;
-+ struct rq *rq;
-+
-+ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
-+#ifdef CONFIG_SCHED_DEBUG
-+ p->migrate_disable_atomic--;
-+#endif
-+ return;
-+ }
-+
-+#ifdef CONFIG_SCHED_DEBUG
-+ WARN_ON_ONCE(p->migrate_disable_atomic);
-+#endif
-+ WARN_ON_ONCE(p->migrate_disable <= 0);
-+
-+ preempt_disable();
-+ if (migrate_disable_count(p) > 1) {
-+ p->migrate_disable--;
-+ preempt_enable();
-+ return;
-+ }
-+
-+ if (unlikely(migrate_disabled_updated(p))) {
-+ /*
-+ * Undo whatever update_migrate_disable() did, also see there
-+ * about locking.
-+ */
-+ rq = this_rq();
-+ raw_spin_lock_irqsave(&rq->lock, flags);
-+
-+ /*
-+ * Clearing migrate_disable causes tsk_cpus_allowed to
-+ * show the tasks original cpu affinity.
-+ */
-+ p->migrate_disable = 0;
-+ mask = tsk_cpus_allowed(p);
-+ if (p->sched_class->set_cpus_allowed)
-+ p->sched_class->set_cpus_allowed(p, mask);
-+ p->rt.nr_cpus_allowed = cpumask_weight(mask);
-+ raw_spin_unlock_irqrestore(&rq->lock, flags);
-+ } else
-+ p->migrate_disable = 0;
-+
-+ unpin_current_cpu();
-+ preempt_enable();
-+}
-+EXPORT_SYMBOL_GPL(migrate_enable);
-+#else
-+static inline void update_migrate_disable(struct task_struct *p) { }
-+#define migrate_disabled_updated(p) 0
-+#endif
-+
- static void put_prev_task(struct rq *rq, struct task_struct *prev)
- {
- if (prev->on_rq || rq->skip_clock_update < 0)
-@@ -4241,6 +4368,8 @@ need_resched:
+diff --git a/arch/arm/mach-msm/platsmp.c b/arch/arm/mach-msm/platsmp.c
+index 2034098..3804971 100644
+--- a/arch/arm/mach-msm/platsmp.c
++++ b/arch/arm/mach-msm/platsmp.c
+@@ -38,7 +38,7 @@ extern void msm_secondary_startup(void);
+ */
+ volatile int pen_release = -1;
- raw_spin_lock_irq(&rq->lock);
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
-+ update_migrate_disable(prev);
-+
- switch_count = &prev->nivcsw;
- if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) {
- if (unlikely(signal_pending_state(prev->state, prev))) {
-@@ -4248,19 +4377,6 @@ need_resched:
- } else {
- deactivate_task(rq, prev, DEQUEUE_SLEEP);
- prev->on_rq = 0;
--
-- /*
-- * If a worker went to sleep, notify and ask workqueue
-- * whether it wants to wake up a task to maintain
-- * concurrency.
-- */
-- if (prev->flags & PF_WQ_WORKER) {
-- struct task_struct *to_wakeup;
--
-- to_wakeup = wq_worker_sleeping(prev, cpu);
-- if (to_wakeup)
-- try_to_wake_up_local(to_wakeup);
-- }
- }
- switch_count = &prev->nvcsw;
- }
-@@ -4294,15 +4410,23 @@ need_resched:
+ void __cpuinit platform_secondary_init(unsigned int cpu)
+ {
+@@ -62,8 +62,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
+ /*
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+ }
- post_schedule(rq);
+ static __cpuinit void prepare_cold_cpu(unsigned int cpu)
+@@ -100,7 +100,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * set synchronisation state between this boot processor
+ * and the secondary one
+ */
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
-- preempt_enable_no_resched();
-+ __preempt_enable_no_resched();
- if (need_resched())
- goto need_resched;
+ /*
+ * The secondary processor is waiting to be released from
+@@ -134,7 +134,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
+ */
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+
+ return pen_release != -1 ? -ENOSYS : 0;
}
+diff --git a/arch/arm/mach-omap2/omap-smp.c b/arch/arm/mach-omap2/omap-smp.c
+index ecfe93c..6aa135c 100644
+--- a/arch/arm/mach-omap2/omap-smp.c
++++ b/arch/arm/mach-omap2/omap-smp.c
+@@ -29,7 +29,7 @@
+ /* SCU base address */
+ static void __iomem *scu_base;
- static inline void sched_submit_work(struct task_struct *tsk)
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
+
+ void __cpuinit platform_secondary_init(unsigned int cpu)
{
-- if (!tsk->state)
-+ if (!tsk->state || tsk_is_pi_blocked(tsk))
- return;
-+
-+ /*
-+ * If a worker went to sleep, notify and ask workqueue whether
-+ * it wants to wake up a task to maintain concurrency.
-+ */
-+ if (tsk->flags & PF_WQ_WORKER)
-+ wq_worker_sleeping(tsk);
-+
+@@ -43,8 +43,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
/*
- * If we are going to sleep and we have plugged IO queued,
- * make sure to submit it to avoid deadlocks.
-@@ -4311,15 +4435,37 @@ static inline void sched_submit_work(str
- blk_schedule_flush_plug(tsk);
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
}
-+static inline void sched_update_worker(struct task_struct *tsk)
-+{
-+ if (tsk_is_pi_blocked(tsk))
-+ return;
-+
-+ if (tsk->flags & PF_WQ_WORKER)
-+ wq_worker_running(tsk);
-+}
-+
- asmlinkage void __sched schedule(void)
- {
- struct task_struct *tsk = current;
+ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+@@ -53,7 +53,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * Set synchronisation state between this boot processor
+ * and the secondary one
+ */
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
- sched_submit_work(tsk);
- __schedule();
-+ sched_update_worker(tsk);
+ /*
+ * Update the AuxCoreBoot0 with boot state for secondary core.
+@@ -70,7 +70,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * Now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
+ */
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+
+ return 0;
}
- EXPORT_SYMBOL(schedule);
+diff --git a/arch/arm/mach-shark/leds.c b/arch/arm/mach-shark/leds.c
+index c9e32de..ccd4918 100644
+--- a/arch/arm/mach-shark/leds.c
++++ b/arch/arm/mach-shark/leds.c
+@@ -36,7 +36,7 @@ static char led_state;
+ static short hw_led_state;
+ static short saved_state;
-+/**
-+ * schedule_preempt_disabled - called with preemption disabled
-+ *
-+ * Returns with preemption disabled. Note: preempt_count must be 1
-+ */
-+void __sched schedule_preempt_disabled(void)
-+{
-+ __preempt_enable_no_resched();
-+ schedule();
-+ preempt_disable();
-+}
-+
- #ifdef CONFIG_MUTEX_SPIN_ON_OWNER
+-static DEFINE_SPINLOCK(leds_lock);
++static DEFINE_RAW_SPINLOCK(leds_lock);
- static inline bool owner_running(struct mutex *lock, struct task_struct *owner)
-@@ -4391,7 +4537,16 @@ asmlinkage void __sched notrace preempt_
+ short sequoia_read(int addr) {
+ outw(addr,0x24);
+@@ -52,7 +52,7 @@ static void sequoia_leds_event(led_event_t evt)
+ {
+ unsigned long flags;
- do {
- add_preempt_count_notrace(PREEMPT_ACTIVE);
-+ /*
-+ * The add/subtract must not be traced by the function
-+ * tracer. But we still want to account for the
-+ * preempt off latency tracer. Since the _notrace versions
-+ * of add/subtract skip the accounting for latency tracer
-+ * we must force it manually.
-+ */
-+ start_critical_timings();
- __schedule();
-+ stop_critical_timings();
- sub_preempt_count_notrace(PREEMPT_ACTIVE);
+- spin_lock_irqsave(&leds_lock, flags);
++ raw_spin_lock_irqsave(&leds_lock, flags);
- /*
-@@ -4814,9 +4969,8 @@ long __sched sleep_on_timeout(wait_queue
- EXPORT_SYMBOL(sleep_on_timeout);
+ hw_led_state = sequoia_read(0x09);
- #ifdef CONFIG_RT_MUTEXES
--
- /*
-- * rt_mutex_setprio - set the current priority of a task
-+ * task_setprio - set the current priority of a task
- * @p: task
- * @prio: prio value (kernel-internal form)
- *
-@@ -4825,7 +4979,7 @@ EXPORT_SYMBOL(sleep_on_timeout);
- *
- * Used by the rt_mutex code to implement priority inheritance logic.
- */
--void rt_mutex_setprio(struct task_struct *p, int prio)
-+void task_setprio(struct task_struct *p, int prio)
- {
- int oldprio, on_rq, running;
- struct rq *rq;
-@@ -4835,6 +4989,24 @@ void rt_mutex_setprio(struct task_struct
+@@ -144,7 +144,7 @@ static void sequoia_leds_event(led_event_t evt)
+ if (led_state & LED_STATE_ENABLED)
+ sequoia_write(hw_led_state,0x09);
- rq = __task_rq_lock(p);
+- spin_unlock_irqrestore(&leds_lock, flags);
++ raw_spin_unlock_irqrestore(&leds_lock, flags);
+ }
-+ /*
-+ * Idle task boosting is a nono in general. There is one
-+ * exception, when PREEMPT_RT and NOHZ is active:
-+ *
-+ * The idle task calls get_next_timer_interrupt() and holds
-+ * the timer wheel base->lock on the CPU and another CPU wants
-+ * to access the timer (probably to cancel it). We can safely
-+ * ignore the boosting request, as the idle CPU runs this code
-+ * with interrupts disabled and will complete the lock
-+ * protected section without being interrupted. So there is no
-+ * real need to boost.
-+ */
-+ if (unlikely(p == rq->idle)) {
-+ WARN_ON(p != rq->curr);
-+ WARN_ON(p->pi_blocked_on);
-+ goto out_unlock;
-+ }
-+
- trace_sched_pi_setprio(p, prio);
- oldprio = p->prio;
- prev_class = p->sched_class;
-@@ -4858,9 +5030,9 @@ void rt_mutex_setprio(struct task_struct
- enqueue_task(rq, p, oldprio < prio ? ENQUEUE_HEAD : 0);
+ static int __init leds_init(void)
+diff --git a/arch/arm/mach-tegra/platsmp.c b/arch/arm/mach-tegra/platsmp.c
+index b8ae3c9..6662075 100644
+--- a/arch/arm/mach-tegra/platsmp.c
++++ b/arch/arm/mach-tegra/platsmp.c
+@@ -29,7 +29,7 @@
- check_class_changed(rq, p, prev_class, oldprio);
-+out_unlock:
- __task_rq_unlock(rq);
- }
--
- #endif
+ extern void tegra_secondary_startup(void);
- void set_user_nice(struct task_struct *p, long nice)
-@@ -4995,7 +5167,13 @@ EXPORT_SYMBOL(task_nice);
- */
- int idle_cpu(int cpu)
- {
-- return cpu_curr(cpu) == cpu_rq(cpu)->idle;
-+ struct rq *rq = cpu_rq(cpu);
-+
-+#ifdef CONFIG_SMP
-+ return rq->curr == rq->idle && !rq->nr_running && !rq->wake_list;
-+#else
-+ return rq->curr == rq->idle && !rq->nr_running;
-+#endif
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
+ static void __iomem *scu_base = IO_ADDRESS(TEGRA_ARM_PERIF_BASE);
+
+ #define EVP_CPU_RESET_VECTOR \
+@@ -51,8 +51,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
+ /*
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
}
- /**
-@@ -5529,7 +5707,7 @@ SYSCALL_DEFINE0(sched_yield)
- __release(rq->lock);
- spin_release(&rq->lock.dep_map, 1, _THIS_IP_);
- do_raw_spin_unlock(&rq->lock);
-- preempt_enable_no_resched();
-+ __preempt_enable_no_resched();
+ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+@@ -66,7 +66,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * set synchronisation state between this boot processor
+ * and the secondary one
+ */
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
- schedule();
-@@ -5543,9 +5721,17 @@ static inline int should_resched(void)
+ /* set the reset vector to point to the secondary_startup routine */
+@@ -102,7 +102,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
+ */
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
- static void __cond_resched(void)
- {
-- add_preempt_count(PREEMPT_ACTIVE);
-- __schedule();
-- sub_preempt_count(PREEMPT_ACTIVE);
-+ do {
-+ add_preempt_count(PREEMPT_ACTIVE);
-+ __schedule();
-+ sub_preempt_count(PREEMPT_ACTIVE);
-+ /*
-+ * Check again in case we missed a preemption
-+ * opportunity between schedule and now.
-+ */
-+ barrier();
-+
-+ } while (need_resched());
+ return 0;
}
-
- int __sched _cond_resched(void)
-@@ -5586,6 +5772,7 @@ int __cond_resched_lock(spinlock_t *lock
+diff --git a/arch/arm/mach-ux500/platsmp.c b/arch/arm/mach-ux500/platsmp.c
+index 0c527fe..64a9edb 100644
+--- a/arch/arm/mach-ux500/platsmp.c
++++ b/arch/arm/mach-ux500/platsmp.c
+@@ -57,7 +57,7 @@ static void __iomem *scu_base_addr(void)
+ return NULL;
}
- EXPORT_SYMBOL(__cond_resched_lock);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- int __sched __cond_resched_softirq(void)
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
+
+ void __cpuinit platform_secondary_init(unsigned int cpu)
{
- BUG_ON(!in_softirq());
-@@ -5599,6 +5786,7 @@ int __sched __cond_resched_softirq(void)
- return 0;
+@@ -77,8 +77,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
+ /*
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
}
- EXPORT_SYMBOL(__cond_resched_softirq);
-+#endif
- /**
- * yield - yield the current processor to other threads.
-@@ -5845,7 +6033,7 @@ void show_state_filter(unsigned long sta
- printk(KERN_INFO
- " task PC stack pid father\n");
- #endif
-- read_lock(&tasklist_lock);
-+ rcu_read_lock();
- do_each_thread(g, p) {
- /*
- * reset the NMI-timeout, listing all files on a slow
-@@ -5861,7 +6049,7 @@ void show_state_filter(unsigned long sta
- #ifdef CONFIG_SCHED_DEBUG
- sysrq_sched_debug_show();
- #endif
-- read_unlock(&tasklist_lock);
-+ rcu_read_unlock();
- /*
- * Only show locks if all tasks are dumped:
+ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+@@ -89,7 +89,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * set synchronisation state between this boot processor
+ * and the secondary one
*/
-@@ -5922,6 +6110,9 @@ void __cpuinit init_idle(struct task_str
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
+
+ /*
+ * The secondary processor is waiting to be released from
+@@ -110,7 +110,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
*/
- idle->sched_class = &idle_sched_class;
- ftrace_graph_init_idle_task(idle, cpu);
-+#if defined(CONFIG_SMP)
-+ sprintf(idle->comm, "%s/%d", INIT_TASK_COMM, cpu);
-+#endif
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+
+ return pen_release != -1 ? -ENOSYS : 0;
}
+diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c
+index 44c0867..f080cd5 100644
+--- a/arch/arm/mm/cache-l2x0.c
++++ b/arch/arm/mm/cache-l2x0.c
+@@ -26,7 +26,7 @@
+ #define CACHE_LINE_SIZE 32
- /*
-@@ -5983,12 +6174,12 @@ static inline void sched_init_granularit
- #ifdef CONFIG_SMP
- void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
+ static void __iomem *l2x0_base;
+-static DEFINE_SPINLOCK(l2x0_lock);
++static DEFINE_RAW_SPINLOCK(l2x0_lock);
+ static uint32_t l2x0_way_mask; /* Bitmask of active ways */
+ static uint32_t l2x0_size;
+
+@@ -115,9 +115,9 @@ static void l2x0_cache_sync(void)
{
-- if (p->sched_class && p->sched_class->set_cpus_allowed)
-- p->sched_class->set_cpus_allowed(p, new_mask);
-- else {
-- cpumask_copy(&p->cpus_allowed, new_mask);
-+ if (!migrate_disabled_updated(p)) {
-+ if (p->sched_class && p->sched_class->set_cpus_allowed)
-+ p->sched_class->set_cpus_allowed(p, new_mask);
- p->rt.nr_cpus_allowed = cpumask_weight(new_mask);
- }
-+ cpumask_copy(&p->cpus_allowed, new_mask);
- }
+ unsigned long flags;
- /*
-@@ -6039,7 +6230,7 @@ int set_cpus_allowed_ptr(struct task_str
- do_set_cpus_allowed(p, new_mask);
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
+ }
- /* Can the task run on the task's current CPU? If so, we're done */
-- if (cpumask_test_cpu(task_cpu(p), new_mask))
-+ if (cpumask_test_cpu(task_cpu(p), new_mask) || __migrate_disabled(p))
- goto out;
+ static void __l2x0_flush_all(void)
+@@ -134,9 +134,9 @@ static void l2x0_flush_all(void)
+ unsigned long flags;
- dest_cpu = cpumask_any_and(cpu_active_mask, new_mask);
-@@ -6086,7 +6277,7 @@ static int __migrate_task(struct task_st
- if (task_cpu(p) != src_cpu)
- goto done;
- /* Affinity changed (again). */
-- if (!cpumask_test_cpu(dest_cpu, &p->cpus_allowed))
-+ if (!cpumask_test_cpu(dest_cpu, tsk_cpus_allowed(p)))
- goto fail;
+ /* clean all ways */
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ __l2x0_flush_all();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
+ }
- /*
-@@ -6128,6 +6319,8 @@ static int migration_cpu_stop(void *data
+ static void l2x0_clean_all(void)
+@@ -144,11 +144,11 @@ static void l2x0_clean_all(void)
+ unsigned long flags;
- #ifdef CONFIG_HOTPLUG_CPU
+ /* clean all ways */
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_CLEAN_WAY);
+ cache_wait_way(l2x0_base + L2X0_CLEAN_WAY, l2x0_way_mask);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
+ }
-+static DEFINE_PER_CPU(struct mm_struct *, idle_last_mm);
-+
- /*
- * Ensures that the idle task is using init_mm right before its cpu goes
- * offline.
-@@ -6140,7 +6333,12 @@ void idle_task_exit(void)
+ static void l2x0_inv_all(void)
+@@ -156,13 +156,13 @@ static void l2x0_inv_all(void)
+ unsigned long flags;
- if (mm != &init_mm)
- switch_mm(mm, &init_mm, current);
-- mmdrop(mm);
-+
-+ /*
-+ * Defer the cleanup to an alive cpu. On RT we can neither
-+ * call mmdrop() nor mmdrop_delayed() from here.
-+ */
-+ per_cpu(idle_last_mm, smp_processor_id()) = mm;
+ /* invalidate all ways */
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ /* Invalidating when L2 is enabled is a nono */
+ BUG_ON(readl(l2x0_base + L2X0_CTRL) & 1);
+ writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
+ cache_wait_way(l2x0_base + L2X0_INV_WAY, l2x0_way_mask);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
}
- /*
-@@ -6458,6 +6656,12 @@ migration_call(struct notifier_block *nf
- migrate_nr_uninterruptible(rq);
- calc_global_load_remove(rq);
- break;
-+ case CPU_DEAD:
-+ if (per_cpu(idle_last_mm, cpu)) {
-+ mmdrop(per_cpu(idle_last_mm, cpu));
-+ per_cpu(idle_last_mm, cpu) = NULL;
-+ }
-+ break;
- #endif
- }
+ static void l2x0_inv_range(unsigned long start, unsigned long end)
+@@ -170,7 +170,7 @@ static void l2x0_inv_range(unsigned long start, unsigned long end)
+ void __iomem *base = l2x0_base;
+ unsigned long flags;
-@@ -8175,7 +8379,8 @@ void __init sched_init(void)
- #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP
- static inline int preempt_count_equals(int preempt_offset)
- {
-- int nested = (preempt_count() & ~PREEMPT_ACTIVE) + rcu_preempt_depth();
-+ int nested = (preempt_count() & ~PREEMPT_ACTIVE) +
-+ sched_rcu_preempt_depth();
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ if (start & (CACHE_LINE_SIZE - 1)) {
+ start &= ~(CACHE_LINE_SIZE - 1);
+ debug_writel(0x03);
+@@ -195,13 +195,13 @@ static void l2x0_inv_range(unsigned long start, unsigned long end)
+ }
- return (nested == preempt_offset);
+ if (blk_end < end) {
+- spin_unlock_irqrestore(&l2x0_lock, flags);
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ }
+ }
+ cache_wait(base + L2X0_INV_LINE_PA, 1);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
}
-Index: linux-3.0/kernel/workqueue.c
-===================================================================
---- linux-3.0.orig/kernel/workqueue.c
-+++ linux-3.0/kernel/workqueue.c
-@@ -41,6 +41,7 @@
- #include <linux/debug_locks.h>
- #include <linux/lockdep.h>
- #include <linux/idr.h>
-+#include <linux/delay.h>
-
- #include "workqueue_sched.h"
-
-@@ -57,20 +58,10 @@ enum {
- WORKER_DIE = 1 << 1, /* die die die */
- WORKER_IDLE = 1 << 2, /* is idle */
- WORKER_PREP = 1 << 3, /* preparing to run works */
-- WORKER_ROGUE = 1 << 4, /* not bound to any cpu */
-- WORKER_REBIND = 1 << 5, /* mom is home, come back */
-- WORKER_CPU_INTENSIVE = 1 << 6, /* cpu intensive */
-- WORKER_UNBOUND = 1 << 7, /* worker is unbound */
--
-- WORKER_NOT_RUNNING = WORKER_PREP | WORKER_ROGUE | WORKER_REBIND |
-- WORKER_CPU_INTENSIVE | WORKER_UNBOUND,
--
-- /* gcwq->trustee_state */
-- TRUSTEE_START = 0, /* start */
-- TRUSTEE_IN_CHARGE = 1, /* trustee in charge of gcwq */
-- TRUSTEE_BUTCHER = 2, /* butcher workers */
-- TRUSTEE_RELEASE = 3, /* release workers */
-- TRUSTEE_DONE = 4, /* trustee is done */
-+ WORKER_CPU_INTENSIVE = 1 << 4, /* cpu intensive */
-+ WORKER_UNBOUND = 1 << 5, /* worker is unbound */
-+
-+ WORKER_NOT_RUNNING = WORKER_PREP | WORKER_CPU_INTENSIVE | WORKER_UNBOUND,
- BUSY_WORKER_HASH_ORDER = 6, /* 64 pointers */
- BUSY_WORKER_HASH_SIZE = 1 << BUSY_WORKER_HASH_ORDER,
-@@ -84,7 +75,6 @@ enum {
- (min two ticks) */
- MAYDAY_INTERVAL = HZ / 10, /* and then every 100ms */
- CREATE_COOLDOWN = HZ, /* time to breath after fail */
-- TRUSTEE_COOLDOWN = HZ / 10, /* for trustee draining */
+ static void l2x0_clean_range(unsigned long start, unsigned long end)
+@@ -214,7 +214,7 @@ static void l2x0_clean_range(unsigned long start, unsigned long end)
+ return;
+ }
- /*
- * Rescue workers are used only on emergencies and shared by
-@@ -136,7 +126,7 @@ struct worker {
- unsigned long last_active; /* L: last active timestamp */
- unsigned int flags; /* X: flags */
- int id; /* I: worker id */
-- struct work_struct rebind_work; /* L: rebind worker to cpu */
-+ int sleeping; /* None */
- };
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ start &= ~(CACHE_LINE_SIZE - 1);
+ while (start < end) {
+ unsigned long blk_end = start + min(end - start, 4096UL);
+@@ -225,13 +225,13 @@ static void l2x0_clean_range(unsigned long start, unsigned long end)
+ }
- /*
-@@ -163,10 +153,8 @@ struct global_cwq {
+ if (blk_end < end) {
+- spin_unlock_irqrestore(&l2x0_lock, flags);
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ }
+ }
+ cache_wait(base + L2X0_CLEAN_LINE_PA, 1);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
+ }
- struct ida worker_ida; /* L: for worker IDs */
+ static void l2x0_flush_range(unsigned long start, unsigned long end)
+@@ -244,7 +244,7 @@ static void l2x0_flush_range(unsigned long start, unsigned long end)
+ return;
+ }
-- struct task_struct *trustee; /* L: for gcwq shutdown */
-- unsigned int trustee_state; /* L: trustee state */
-- wait_queue_head_t trustee_wait; /* trustee wait */
- struct worker *first_idle; /* L: first idle worker */
-+ wait_queue_head_t idle_wait;
- } ____cacheline_aligned_in_smp;
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ start &= ~(CACHE_LINE_SIZE - 1);
+ while (start < end) {
+ unsigned long blk_end = start + min(end - start, 4096UL);
+@@ -257,24 +257,24 @@ static void l2x0_flush_range(unsigned long start, unsigned long end)
+ debug_writel(0x00);
- /*
-@@ -657,66 +645,58 @@ static void wake_up_worker(struct global
+ if (blk_end < end) {
+- spin_unlock_irqrestore(&l2x0_lock, flags);
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ }
+ }
+ cache_wait(base + L2X0_CLEAN_INV_LINE_PA, 1);
+ cache_sync();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
}
- /**
-- * wq_worker_waking_up - a worker is waking up
-- * @task: task waking up
-- * @cpu: CPU @task is waking up to
-- *
-- * This function is called during try_to_wake_up() when a worker is
-- * being awoken.
-+ * wq_worker_running - a worker is running again
-+ * @task: task returning from sleep
- *
-- * CONTEXT:
-- * spin_lock_irq(rq->lock)
-+ * This function is called when a worker returns from schedule()
- */
--void wq_worker_waking_up(struct task_struct *task, unsigned int cpu)
-+void wq_worker_running(struct task_struct *task)
+ static void l2x0_disable(void)
{
- struct worker *worker = kthread_data(task);
+ unsigned long flags;
-+ if (!worker->sleeping)
-+ return;
- if (!(worker->flags & WORKER_NOT_RUNNING))
-- atomic_inc(get_gcwq_nr_running(cpu));
-+ atomic_inc(get_gcwq_nr_running(smp_processor_id()));
-+ worker->sleeping = 0;
+- spin_lock_irqsave(&l2x0_lock, flags);
++ raw_spin_lock_irqsave(&l2x0_lock, flags);
+ __l2x0_flush_all();
+ writel_relaxed(0, l2x0_base + L2X0_CTRL);
+ dsb();
+- spin_unlock_irqrestore(&l2x0_lock, flags);
++ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
}
- /**
- * wq_worker_sleeping - a worker is going to sleep
- * @task: task going to sleep
-- * @cpu: CPU in question, must be the current CPU number
-- *
-- * This function is called during schedule() when a busy worker is
-- * going to sleep. Worker on the same cpu can be woken up by
-- * returning pointer to its task.
- *
-- * CONTEXT:
-- * spin_lock_irq(rq->lock)
-- *
-- * RETURNS:
-- * Worker task on @cpu to wake up, %NULL if none.
-+ * This function is called from schedule() when a busy worker is
-+ * going to sleep.
- */
--struct task_struct *wq_worker_sleeping(struct task_struct *task,
-- unsigned int cpu)
-+void wq_worker_sleeping(struct task_struct *task)
- {
-- struct worker *worker = kthread_data(task), *to_wakeup = NULL;
-- struct global_cwq *gcwq = get_gcwq(cpu);
-- atomic_t *nr_running = get_gcwq_nr_running(cpu);
-+ struct worker *worker = kthread_data(task);
-+ struct global_cwq *gcwq;
-+ int cpu;
+ void __init l2x0_init(void __iomem *base, __u32 aux_val, __u32 aux_mask)
+diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c
+index b0ee9ba..93aac06 100644
+--- a/arch/arm/mm/context.c
++++ b/arch/arm/mm/context.c
+@@ -16,7 +16,7 @@
+ #include <asm/mmu_context.h>
+ #include <asm/tlbflush.h>
- if (worker->flags & WORKER_NOT_RUNNING)
-- return NULL;
-+ return;
+-static DEFINE_SPINLOCK(cpu_asid_lock);
++static DEFINE_RAW_SPINLOCK(cpu_asid_lock);
+ unsigned int cpu_last_asid = ASID_FIRST_VERSION;
+ #ifdef CONFIG_SMP
+ DEFINE_PER_CPU(struct mm_struct *, current_mm);
+@@ -31,7 +31,7 @@ DEFINE_PER_CPU(struct mm_struct *, current_mm);
+ void __init_new_context(struct task_struct *tsk, struct mm_struct *mm)
+ {
+ mm->context.id = 0;
+- spin_lock_init(&mm->context.id_lock);
++ raw_spin_lock_init(&mm->context.id_lock);
+ }
-- /* this can only happen on the local cpu */
-- BUG_ON(cpu != raw_smp_processor_id());
-+ if (WARN_ON_ONCE(worker->sleeping))
-+ return;
-+
-+ worker->sleeping = 1;
+ static void flush_context(void)
+@@ -58,7 +58,7 @@ static void set_mm_context(struct mm_struct *mm, unsigned int asid)
+ * the broadcast. This function is also called via IPI so the
+ * mm->context.id_lock has to be IRQ-safe.
+ */
+- spin_lock_irqsave(&mm->context.id_lock, flags);
++ raw_spin_lock_irqsave(&mm->context.id_lock, flags);
+ if (likely((mm->context.id ^ cpu_last_asid) >> ASID_BITS)) {
+ /*
+ * Old version of ASID found. Set the new one and
+@@ -67,7 +67,7 @@ static void set_mm_context(struct mm_struct *mm, unsigned int asid)
+ mm->context.id = asid;
+ cpumask_clear(mm_cpumask(mm));
+ }
+- spin_unlock_irqrestore(&mm->context.id_lock, flags);
++ raw_spin_unlock_irqrestore(&mm->context.id_lock, flags);
-+ cpu = smp_processor_id();
-+ gcwq = get_gcwq(cpu);
-+ spin_lock_irq(&gcwq->lock);
/*
- * The counterpart of the following dec_and_test, implied mb,
- * worklist not empty test sequence is in insert_work().
- * Please read comment there.
-- *
-- * NOT_RUNNING is clear. This means that trustee is not in
-- * charge and we're running on the local cpu w/ rq lock held
-- * and preemption disabled, which in turn means that none else
-- * could be manipulating idle_list, so dereferencing idle_list
-- * without gcwq lock is safe.
-- */
-- if (atomic_dec_and_test(nr_running) && !list_empty(&gcwq->worklist))
-- to_wakeup = first_worker(gcwq);
-- return to_wakeup ? to_wakeup->task : NULL;
-+ */
-+ if (atomic_dec_and_test(get_gcwq_nr_running(cpu)) &&
-+ !list_empty(&gcwq->worklist)) {
-+ worker = first_worker(gcwq);
-+ if (worker)
-+ wake_up_process(worker->task);
-+ }
-+ spin_unlock_irq(&gcwq->lock);
- }
+ * Set the mm_cpumask(mm) bit for the current CPU.
+@@ -117,7 +117,7 @@ void __new_context(struct mm_struct *mm)
+ {
+ unsigned int asid;
- /**
-@@ -978,13 +958,38 @@ static bool is_chained_work(struct workq
- return false;
+- spin_lock(&cpu_asid_lock);
++ raw_spin_lock(&cpu_asid_lock);
+ #ifdef CONFIG_SMP
+ /*
+ * Check the ASID again, in case the change was broadcast from
+@@ -125,7 +125,7 @@ void __new_context(struct mm_struct *mm)
+ */
+ if (unlikely(((mm->context.id ^ cpu_last_asid) >> ASID_BITS) == 0)) {
+ cpumask_set_cpu(smp_processor_id(), mm_cpumask(mm));
+- spin_unlock(&cpu_asid_lock);
++ raw_spin_unlock(&cpu_asid_lock);
+ return;
+ }
+ #endif
+@@ -153,5 +153,5 @@ void __new_context(struct mm_struct *mm)
+ }
+
+ set_mm_context(mm, asid);
+- spin_unlock(&cpu_asid_lock);
++ raw_spin_unlock(&cpu_asid_lock);
}
+diff --git a/arch/arm/mm/copypage-v4mc.c b/arch/arm/mm/copypage-v4mc.c
+index b806151..7d0a8c2 100644
+--- a/arch/arm/mm/copypage-v4mc.c
++++ b/arch/arm/mm/copypage-v4mc.c
+@@ -30,7 +30,7 @@
+ #define minicache_pgprot __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | \
+ L_PTE_MT_MINICACHE)
--static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
-- struct work_struct *work)
-+static void ___queue_work(struct workqueue_struct *wq, struct global_cwq *gcwq,
-+ struct work_struct *work)
- {
-- struct global_cwq *gcwq;
- struct cpu_workqueue_struct *cwq;
- struct list_head *worklist;
- unsigned int work_flags;
-+
-+ /* gcwq determined, get cwq and queue */
-+ cwq = get_cwq(gcwq->cpu, wq);
-+ trace_workqueue_queue_work(gcwq->cpu, cwq, work);
-+
-+ BUG_ON(!list_empty(&work->entry));
-+
-+ cwq->nr_in_flight[cwq->work_color]++;
-+ work_flags = work_color_to_flags(cwq->work_color);
-+
-+ if (likely(cwq->nr_active < cwq->max_active)) {
-+ trace_workqueue_activate_work(work);
-+ cwq->nr_active++;
-+ worklist = gcwq_determine_ins_pos(gcwq, cwq);
-+ } else {
-+ work_flags |= WORK_STRUCT_DELAYED;
-+ worklist = &cwq->delayed_works;
-+ }
-+
-+ insert_work(cwq, work, worklist, work_flags);
-+}
-+
-+static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
-+ struct work_struct *work)
-+{
-+ struct global_cwq *gcwq;
- unsigned long flags;
+-static DEFINE_SPINLOCK(minicache_lock);
++static DEFINE_RAW_SPINLOCK(minicache_lock);
- debug_work_activate(work);
-@@ -1030,27 +1035,32 @@ static void __queue_work(unsigned int cp
- spin_lock_irqsave(&gcwq->lock, flags);
- }
+ /*
+ * ARMv4 mini-dcache optimised copy_user_highpage
+@@ -76,14 +76,14 @@ void v4_mc_copy_user_highpage(struct page *to, struct page *from,
+ if (!test_and_set_bit(PG_dcache_clean, &from->flags))
+ __flush_dcache_page(page_mapping(from), from);
-- /* gcwq determined, get cwq and queue */
-- cwq = get_cwq(gcwq->cpu, wq);
-- trace_workqueue_queue_work(cpu, cwq, work);
-+ ___queue_work(wq, gcwq, work);
+- spin_lock(&minicache_lock);
++ raw_spin_lock(&minicache_lock);
-- BUG_ON(!list_empty(&work->entry));
-+ spin_unlock_irqrestore(&gcwq->lock, flags);
-+}
+ set_pte_ext(TOP_PTE(0xffff8000), pfn_pte(page_to_pfn(from), minicache_pgprot), 0);
+ flush_tlb_kernel_page(0xffff8000);
-- cwq->nr_in_flight[cwq->work_color]++;
-- work_flags = work_color_to_flags(cwq->work_color);
-+/**
-+ * queue_work_on - queue work on specific cpu
-+ * @cpu: CPU number to execute work on
-+ * @wq: workqueue to use
-+ * @work: work to queue
-+ *
-+ * Returns 0 if @work was already on a queue, non-zero otherwise.
-+ *
-+ * We queue the work to a specific CPU, the caller must ensure it
-+ * can't go away.
-+ */
-+static int
-+__queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
-+{
-+ int ret = 0;
+ mc_copy_user_page((void *)0xffff8000, kto);
-- if (likely(cwq->nr_active < cwq->max_active)) {
-- trace_workqueue_activate_work(work);
-- cwq->nr_active++;
-- worklist = gcwq_determine_ins_pos(gcwq, cwq);
-- } else {
-- work_flags |= WORK_STRUCT_DELAYED;
-- worklist = &cwq->delayed_works;
-+ if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
-+ __queue_work(cpu, wq, work);
-+ ret = 1;
- }
--
-- insert_work(cwq, work, worklist, work_flags);
--
-- spin_unlock_irqrestore(&gcwq->lock, flags);
-+ return ret;
+- spin_unlock(&minicache_lock);
++ raw_spin_unlock(&minicache_lock);
+
+ kunmap_atomic(kto, KM_USER1);
}
+diff --git a/arch/arm/mm/copypage-v6.c b/arch/arm/mm/copypage-v6.c
+index bdba6c6..b2a8f9c 100644
+--- a/arch/arm/mm/copypage-v6.c
++++ b/arch/arm/mm/copypage-v6.c
+@@ -27,7 +27,7 @@
+ #define from_address (0xffff8000)
+ #define to_address (0xffffc000)
- /**
-@@ -1067,34 +1077,19 @@ int queue_work(struct workqueue_struct *
- {
- int ret;
+-static DEFINE_SPINLOCK(v6_lock);
++static DEFINE_RAW_SPINLOCK(v6_lock);
-- ret = queue_work_on(get_cpu(), wq, work);
-- put_cpu();
-+ ret = __queue_work_on(get_cpu_light(), wq, work);
-+ put_cpu_light();
+ /*
+ * Copy the user page. No aliasing to deal with so we can just
+@@ -89,7 +89,7 @@ static void v6_copy_user_highpage_aliasing(struct page *to,
+ * Now copy the page using the same cache colour as the
+ * pages ultimate destination.
+ */
+- spin_lock(&v6_lock);
++ raw_spin_lock(&v6_lock);
- return ret;
- }
- EXPORT_SYMBOL_GPL(queue_work);
+ set_pte_ext(TOP_PTE(from_address) + offset, pfn_pte(page_to_pfn(from), PAGE_KERNEL), 0);
+ set_pte_ext(TOP_PTE(to_address) + offset, pfn_pte(page_to_pfn(to), PAGE_KERNEL), 0);
+@@ -102,7 +102,7 @@ static void v6_copy_user_highpage_aliasing(struct page *to,
--/**
-- * queue_work_on - queue work on specific cpu
-- * @cpu: CPU number to execute work on
-- * @wq: workqueue to use
-- * @work: work to queue
-- *
-- * Returns 0 if @work was already on a queue, non-zero otherwise.
-- *
-- * We queue the work to a specific CPU, the caller must ensure it
-- * can't go away.
-- */
- int
- queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
- {
-- int ret = 0;
-+ WARN_ON(wq->flags & WQ_NON_AFFINE);
+ copy_page((void *)kto, (void *)kfrom);
-- if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
-- __queue_work(cpu, wq, work);
-- ret = 1;
-- }
-- return ret;
-+ return __queue_work_on(cpu, wq, work);
+- spin_unlock(&v6_lock);
++ raw_spin_unlock(&v6_lock);
}
- EXPORT_SYMBOL_GPL(queue_work_on);
-@@ -1140,6 +1135,8 @@ int queue_delayed_work_on(int cpu, struc
- struct timer_list *timer = &dwork->timer;
- struct work_struct *work = &dwork->work;
+ /*
+@@ -122,13 +122,13 @@ static void v6_clear_user_highpage_aliasing(struct page *page, unsigned long vad
+ * Now clear the page using the same cache colour as
+ * the pages ultimate destination.
+ */
+- spin_lock(&v6_lock);
++ raw_spin_lock(&v6_lock);
-+ WARN_ON((wq->flags & WQ_NON_AFFINE) && cpu != -1);
-+
- if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
- unsigned int lcpu;
+ set_pte_ext(TOP_PTE(to_address) + offset, pfn_pte(page_to_pfn(page), PAGE_KERNEL), 0);
+ flush_tlb_kernel_page(to);
+ clear_page((void *)to);
-@@ -1205,12 +1202,13 @@ static void worker_enter_idle(struct wor
- /* idle_list is LIFO */
- list_add(&worker->entry, &gcwq->idle_list);
+- spin_unlock(&v6_lock);
++ raw_spin_unlock(&v6_lock);
+ }
-- if (likely(!(worker->flags & WORKER_ROGUE))) {
-- if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer))
-- mod_timer(&gcwq->idle_timer,
-- jiffies + IDLE_WORKER_TIMEOUT);
-- } else
-- wake_up_all(&gcwq->trustee_wait);
-+ if (gcwq->nr_idle == gcwq->nr_workers)
-+ wake_up_all(&gcwq->idle_wait);
-+
-+ if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer)) {
-+ mod_timer(&gcwq->idle_timer,
-+ jiffies + IDLE_WORKER_TIMEOUT);
-+ }
+ struct cpu_user_fns v6_user_fns __initdata = {
+diff --git a/arch/arm/mm/copypage-xscale.c b/arch/arm/mm/copypage-xscale.c
+index 649bbcd..610c24c 100644
+--- a/arch/arm/mm/copypage-xscale.c
++++ b/arch/arm/mm/copypage-xscale.c
+@@ -32,7 +32,7 @@
+ #define minicache_pgprot __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | \
+ L_PTE_MT_MINICACHE)
- /* sanity check nr_running */
- WARN_ON_ONCE(gcwq->nr_workers == gcwq->nr_idle &&
-@@ -1287,8 +1285,14 @@ __acquires(&gcwq->lock)
- return false;
- if (task_cpu(task) == gcwq->cpu &&
- cpumask_equal(¤t->cpus_allowed,
-- get_cpu_mask(gcwq->cpu)))
-+ get_cpu_mask(gcwq->cpu))) {
-+ /*
-+ * Since we're binding to a particular cpu and need to
-+ * stay there for correctness, mark us PF_THREAD_BOUND.
-+ */
-+ task->flags |= PF_THREAD_BOUND;
- return true;
-+ }
- spin_unlock_irq(&gcwq->lock);
+-static DEFINE_SPINLOCK(minicache_lock);
++static DEFINE_RAW_SPINLOCK(minicache_lock);
- /*
-@@ -1302,20 +1306,15 @@ __acquires(&gcwq->lock)
- }
- }
+ /*
+ * XScale mini-dcache optimised copy_user_highpage
+@@ -98,14 +98,14 @@ void xscale_mc_copy_user_highpage(struct page *to, struct page *from,
+ if (!test_and_set_bit(PG_dcache_clean, &from->flags))
+ __flush_dcache_page(page_mapping(from), from);
--/*
-- * Function for worker->rebind_work used to rebind rogue busy workers
-- * to the associated cpu which is coming back online. This is
-- * scheduled by cpu up but can race with other cpu hotplug operations
-- * and may be executed twice without intervening cpu down.
-- */
--static void worker_rebind_fn(struct work_struct *work)
-+static void worker_unbind_and_unlock(struct worker *worker)
- {
-- struct worker *worker = container_of(work, struct worker, rebind_work);
- struct global_cwq *gcwq = worker->gcwq;
-+ struct task_struct *task = worker->task;
+- spin_lock(&minicache_lock);
++ raw_spin_lock(&minicache_lock);
-- if (worker_maybe_bind_and_lock(worker))
-- worker_clr_flags(worker, WORKER_REBIND);
--
-+ /*
-+ * Its no longer required we're PF_THREAD_BOUND, the work is done.
-+ */
-+ task->flags &= ~PF_THREAD_BOUND;
- spin_unlock_irq(&gcwq->lock);
+ set_pte_ext(TOP_PTE(COPYPAGE_MINICACHE), pfn_pte(page_to_pfn(from), minicache_pgprot), 0);
+ flush_tlb_kernel_page(COPYPAGE_MINICACHE);
+
+ mc_copy_user_page((void *)COPYPAGE_MINICACHE, kto);
+
+- spin_unlock(&minicache_lock);
++ raw_spin_unlock(&minicache_lock);
+
+ kunmap_atomic(kto, KM_USER1);
+ }
+diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c
+index bc0e1d8..9afc70a 100644
+--- a/arch/arm/mm/fault.c
++++ b/arch/arm/mm/fault.c
+@@ -289,7 +289,7 @@ do_page_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs)
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
+
+ /*
+diff --git a/arch/arm/plat-versatile/platsmp.c b/arch/arm/plat-versatile/platsmp.c
+index 51ecfea..6096db8 100644
+--- a/arch/arm/plat-versatile/platsmp.c
++++ b/arch/arm/plat-versatile/platsmp.c
+@@ -37,7 +37,7 @@ static void __cpuinit write_pen_release(int val)
+ outer_clean_range(__pa(&pen_release), __pa(&pen_release + 1));
}
-@@ -1327,7 +1326,6 @@ static struct worker *alloc_worker(void)
- if (worker) {
- INIT_LIST_HEAD(&worker->entry);
- INIT_LIST_HEAD(&worker->scheduled);
-- INIT_WORK(&worker->rebind_work, worker_rebind_fn);
- /* on creation a worker is in !idle && prep state */
- worker->flags = WORKER_PREP;
- }
-@@ -1382,15 +1380,9 @@ static struct worker *create_worker(stru
- if (IS_ERR(worker->task))
- goto fail;
+-static DEFINE_SPINLOCK(boot_lock);
++static DEFINE_RAW_SPINLOCK(boot_lock);
-- /*
-- * A rogue worker will become a regular one if CPU comes
-- * online later on. Make sure every worker has
-- * PF_THREAD_BOUND set.
-- */
- if (bind && !on_unbound_cpu)
- kthread_bind(worker->task, gcwq->cpu);
- else {
-- worker->task->flags |= PF_THREAD_BOUND;
- if (on_unbound_cpu)
- worker->flags |= WORKER_UNBOUND;
+ void __cpuinit platform_secondary_init(unsigned int cpu)
+ {
+@@ -57,8 +57,8 @@ void __cpuinit platform_secondary_init(unsigned int cpu)
+ /*
+ * Synchronise with the boot thread.
+ */
+- spin_lock(&boot_lock);
+- spin_unlock(&boot_lock);
++ raw_spin_lock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+ }
+
+ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+@@ -69,7 +69,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * Set synchronisation state between this boot processor
+ * and the secondary one
+ */
+- spin_lock(&boot_lock);
++ raw_spin_lock(&boot_lock);
+
+ /*
+ * This is really belt and braces; we hold unintended secondary
+@@ -99,7 +99,7 @@ int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
+ * now the secondary core is starting up let it run its
+ * calibrations, then wait for it to finish
+ */
+- spin_unlock(&boot_lock);
++ raw_spin_unlock(&boot_lock);
+
+ return pen_release != -1 ? -ENOSYS : 0;
+ }
+diff --git a/arch/avr32/kernel/process.c b/arch/avr32/kernel/process.c
+index ef5a2a0..c8724c9 100644
+--- a/arch/avr32/kernel/process.c
++++ b/arch/avr32/kernel/process.c
+@@ -38,9 +38,7 @@ void cpu_idle(void)
+ while (!need_resched())
+ cpu_idle_sleep();
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
-@@ -1667,13 +1659,6 @@ static bool manage_workers(struct worker
+ }
- gcwq->flags &= ~GCWQ_MANAGING_WORKERS;
+diff --git a/arch/avr32/mm/fault.c b/arch/avr32/mm/fault.c
+index f7040a1..155ad8d 100644
+--- a/arch/avr32/mm/fault.c
++++ b/arch/avr32/mm/fault.c
+@@ -81,7 +81,7 @@ asmlinkage void do_page_fault(unsigned long ecr, struct pt_regs *regs)
+ * If we're in an interrupt or have no user context, we must
+ * not take the fault...
+ */
+- if (in_atomic() || !mm || regs->sr & SYSREG_BIT(GM))
++ if (!mm || regs->sr & SYSREG_BIT(GM) || pagefault_disabled())
+ goto no_context;
-- /*
-- * The trustee might be waiting to take over the manager
-- * position, tell it we're done.
-- */
-- if (unlikely(gcwq->trustee))
-- wake_up_all(&gcwq->trustee_wait);
+ local_irq_enable();
+diff --git a/arch/blackfin/kernel/early_printk.c b/arch/blackfin/kernel/early_printk.c
+index 84ed837..61fbd2d 100644
+--- a/arch/blackfin/kernel/early_printk.c
++++ b/arch/blackfin/kernel/early_printk.c
+@@ -25,8 +25,6 @@ extern struct console *bfin_earlyserial_init(unsigned int port,
+ extern struct console *bfin_jc_early_init(void);
+ #endif
+
+-static struct console *early_console;
-
- return ret;
+ /* Default console */
+ #define DEFAULT_PORT 0
+ #define DEFAULT_CFLAG CS8|B57600
+diff --git a/arch/blackfin/kernel/process.c b/arch/blackfin/kernel/process.c
+index 6a660fa..ecc3b92 100644
+--- a/arch/blackfin/kernel/process.c
++++ b/arch/blackfin/kernel/process.c
+@@ -92,9 +92,7 @@ void cpu_idle(void)
+ while (!need_resched())
+ idle();
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
}
-@@ -2074,7 +2059,7 @@ repeat:
- if (keep_working(gcwq))
- wake_up_worker(gcwq);
+diff --git a/arch/cris/kernel/process.c b/arch/cris/kernel/process.c
+index c99aeab..f88f435 100644
+--- a/arch/cris/kernel/process.c
++++ b/arch/cris/kernel/process.c
+@@ -115,9 +115,7 @@ void cpu_idle (void)
+ idle = default_idle;
+ idle();
+ }
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
-- spin_unlock_irq(&gcwq->lock);
-+ worker_unbind_and_unlock(rescuer);
+diff --git a/arch/cris/mm/fault.c b/arch/cris/mm/fault.c
+index 9dcac8e..ba9cfbe 100644
+--- a/arch/cris/mm/fault.c
++++ b/arch/cris/mm/fault.c
+@@ -111,7 +111,7 @@ do_page_fault(unsigned long address, struct pt_regs *regs,
+ * user context, we must not take the fault.
+ */
+
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
+
+ down_read(&mm->mmap_sem);
+diff --git a/arch/frv/kernel/process.c b/arch/frv/kernel/process.c
+index 9d35975..f6b9b60 100644
+--- a/arch/frv/kernel/process.c
++++ b/arch/frv/kernel/process.c
+@@ -92,9 +92,7 @@ void cpu_idle(void)
+ idle();
+ }
+
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
+ }
- schedule();
-@@ -2970,7 +2955,6 @@ struct workqueue_struct *__alloc_workque
- if (IS_ERR(rescuer->task))
- goto err;
+diff --git a/arch/frv/mm/fault.c b/arch/frv/mm/fault.c
+index a325d57..a9ce0f0 100644
+--- a/arch/frv/mm/fault.c
++++ b/arch/frv/mm/fault.c
+@@ -79,7 +79,7 @@ asmlinkage void do_page_fault(int datammu, unsigned long esr0, unsigned long ear
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
-- rescuer->task->flags |= PF_THREAD_BOUND;
- wake_up_process(rescuer->task);
+ down_read(&mm->mmap_sem);
+diff --git a/arch/h8300/kernel/process.c b/arch/h8300/kernel/process.c
+index 933bd38..1a173b3 100644
+--- a/arch/h8300/kernel/process.c
++++ b/arch/h8300/kernel/process.c
+@@ -81,9 +81,7 @@ void cpu_idle(void)
+ while (1) {
+ while (!need_resched())
+ idle();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
+ }
-@@ -3189,171 +3173,76 @@ EXPORT_SYMBOL_GPL(work_busy);
- * gcwqs serve mix of short, long and very long running works making
- * blocked draining impractical.
- *
-- * This is solved by allowing a gcwq to be detached from CPU, running
-- * it with unbound (rogue) workers and allowing it to be reattached
-- * later if the cpu comes back online. A separate thread is created
-- * to govern a gcwq in such state and is called the trustee of the
-- * gcwq.
-- *
-- * Trustee states and their descriptions.
-- *
-- * START Command state used on startup. On CPU_DOWN_PREPARE, a
-- * new trustee is started with this state.
-- *
-- * IN_CHARGE Once started, trustee will enter this state after
-- * assuming the manager role and making all existing
-- * workers rogue. DOWN_PREPARE waits for trustee to
-- * enter this state. After reaching IN_CHARGE, trustee
-- * tries to execute the pending worklist until it's empty
-- * and the state is set to BUTCHER, or the state is set
-- * to RELEASE.
-- *
-- * BUTCHER Command state which is set by the cpu callback after
-- * the cpu has went down. Once this state is set trustee
-- * knows that there will be no new works on the worklist
-- * and once the worklist is empty it can proceed to
-- * killing idle workers.
-- *
-- * RELEASE Command state which is set by the cpu callback if the
-- * cpu down has been canceled or it has come online
-- * again. After recognizing this state, trustee stops
-- * trying to drain or butcher and clears ROGUE, rebinds
-- * all remaining workers back to the cpu and releases
-- * manager role.
-- *
-- * DONE Trustee will enter this state after BUTCHER or RELEASE
-- * is complete.
-- *
-- * trustee CPU draining
-- * took over down complete
-- * START -----------> IN_CHARGE -----------> BUTCHER -----------> DONE
-- * | | ^
-- * | CPU is back online v return workers |
-- * ----------------> RELEASE --------------
+diff --git a/arch/ia64/include/asm/rwsem.h b/arch/ia64/include/asm/rwsem.h
+index 3027e75..885fddb 100644
+--- a/arch/ia64/include/asm/rwsem.h
++++ b/arch/ia64/include/asm/rwsem.h
+@@ -38,7 +38,7 @@
+ * lock for reading
*/
+ static inline void
+-__down_read (struct rw_semaphore *sem)
++__down_read (struct rw_anon_semaphore *sem)
+ {
+ long result = ia64_fetchadd8_acq((unsigned long *)&sem->count, 1);
--/**
-- * trustee_wait_event_timeout - timed event wait for trustee
-- * @cond: condition to wait for
-- * @timeout: timeout in jiffies
-- *
-- * wait_event_timeout() for trustee to use. Handles locking and
-- * checks for RELEASE request.
-- *
-- * CONTEXT:
-- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
-- * multiple times. To be used by trustee.
-- *
-- * RETURNS:
-- * Positive indicating left time if @cond is satisfied, 0 if timed
-- * out, -1 if canceled.
-- */
--#define trustee_wait_event_timeout(cond, timeout) ({ \
-- long __ret = (timeout); \
-- while (!((cond) || (gcwq->trustee_state == TRUSTEE_RELEASE)) && \
-- __ret) { \
-- spin_unlock_irq(&gcwq->lock); \
-- __wait_event_timeout(gcwq->trustee_wait, (cond) || \
-- (gcwq->trustee_state == TRUSTEE_RELEASE), \
-- __ret); \
-- spin_lock_irq(&gcwq->lock); \
-- } \
-- gcwq->trustee_state == TRUSTEE_RELEASE ? -1 : (__ret); \
--})
-+static int __devinit workqueue_cpu_up_callback(struct notifier_block *nfb,
-+ unsigned long action,
-+ void *hcpu)
-+{
-+ unsigned int cpu = (unsigned long)hcpu;
-+ struct global_cwq *gcwq = get_gcwq(cpu);
-+ struct worker *uninitialized_var(new_worker);
-+ unsigned long flags;
-
--/**
-- * trustee_wait_event - event wait for trustee
-- * @cond: condition to wait for
-- *
-- * wait_event() for trustee to use. Automatically handles locking and
-- * checks for CANCEL request.
-- *
-- * CONTEXT:
-- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
-- * multiple times. To be used by trustee.
-- *
-- * RETURNS:
-- * 0 if @cond is satisfied, -1 if canceled.
-- */
--#define trustee_wait_event(cond) ({ \
-- long __ret1; \
-- __ret1 = trustee_wait_event_timeout(cond, MAX_SCHEDULE_TIMEOUT);\
-- __ret1 < 0 ? -1 : 0; \
--})
-+ action &= ~CPU_TASKS_FROZEN;
+@@ -50,7 +50,7 @@ __down_read (struct rw_semaphore *sem)
+ * lock for writing
+ */
+ static inline void
+-__down_write (struct rw_semaphore *sem)
++__down_write (struct rw_anon_semaphore *sem)
+ {
+ long old, new;
--static int __cpuinit trustee_thread(void *__gcwq)
--{
-- struct global_cwq *gcwq = __gcwq;
-- struct worker *worker;
-- struct work_struct *work;
-- struct hlist_node *pos;
-- long rc;
-- int i;
-+ switch (action) {
-+ case CPU_UP_PREPARE:
-+ BUG_ON(gcwq->first_idle);
-+ new_worker = create_worker(gcwq, false);
-+ if (!new_worker)
-+ return NOTIFY_BAD;
-+ case CPU_UP_CANCELED:
-+ case CPU_ONLINE:
-+ break;
-+ default:
-+ return notifier_from_errno(0);
-+ }
+@@ -67,7 +67,7 @@ __down_write (struct rw_semaphore *sem)
+ * unlock after reading
+ */
+ static inline void
+-__up_read (struct rw_semaphore *sem)
++__up_read (struct rw_anon_semaphore *sem)
+ {
+ long result = ia64_fetchadd8_rel((unsigned long *)&sem->count, -1);
-- BUG_ON(gcwq->cpu != smp_processor_id());
-+ /* some are called w/ irq disabled, don't disturb irq status */
-+ spin_lock_irqsave(&gcwq->lock, flags);
+@@ -79,7 +79,7 @@ __up_read (struct rw_semaphore *sem)
+ * unlock after writing
+ */
+ static inline void
+-__up_write (struct rw_semaphore *sem)
++__up_write (struct rw_anon_semaphore *sem)
+ {
+ long old, new;
-- spin_lock_irq(&gcwq->lock);
-- /*
-- * Claim the manager position and make all workers rogue.
-- * Trustee must be bound to the target cpu and can't be
-- * cancelled.
-- */
-- BUG_ON(gcwq->cpu != smp_processor_id());
-- rc = trustee_wait_event(!(gcwq->flags & GCWQ_MANAGING_WORKERS));
-- BUG_ON(rc < 0);
-+ switch (action) {
-+ case CPU_UP_PREPARE:
-+ BUG_ON(gcwq->first_idle);
-+ gcwq->first_idle = new_worker;
-+ break;
+@@ -96,7 +96,7 @@ __up_write (struct rw_semaphore *sem)
+ * trylock for reading -- returns 1 if successful, 0 if contention
+ */
+ static inline int
+-__down_read_trylock (struct rw_semaphore *sem)
++__down_read_trylock (struct rw_anon_semaphore *sem)
+ {
+ long tmp;
+ while ((tmp = sem->count) >= 0) {
+@@ -111,7 +111,7 @@ __down_read_trylock (struct rw_semaphore *sem)
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+ static inline int
+-__down_write_trylock (struct rw_semaphore *sem)
++__down_write_trylock (struct rw_anon_semaphore *sem)
+ {
+ long tmp = cmpxchg_acq(&sem->count, RWSEM_UNLOCKED_VALUE,
+ RWSEM_ACTIVE_WRITE_BIAS);
+@@ -122,7 +122,7 @@ __down_write_trylock (struct rw_semaphore *sem)
+ * downgrade write lock to read lock
+ */
+ static inline void
+-__downgrade_write (struct rw_semaphore *sem)
++__downgrade_write (struct rw_anon_semaphore *sem)
+ {
+ long old, new;
-- gcwq->flags |= GCWQ_MANAGING_WORKERS;
-+ case CPU_UP_CANCELED:
-+ destroy_worker(gcwq->first_idle);
-+ gcwq->first_idle = NULL;
-+ break;
+diff --git a/arch/ia64/kernel/process.c b/arch/ia64/kernel/process.c
+index 6d33c5c..9dc52b6 100644
+--- a/arch/ia64/kernel/process.c
++++ b/arch/ia64/kernel/process.c
+@@ -330,9 +330,7 @@ cpu_idle (void)
+ normal_xtp();
+ #endif
+ }
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ check_pgt_cache();
+ if (cpu_is_offline(cpu))
+ play_dead();
+diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
+index 85118df..604a636 100644
+--- a/arch/ia64/kernel/time.c
++++ b/arch/ia64/kernel/time.c
+@@ -36,7 +36,7 @@
+ static cycle_t itc_get_cycles(struct clocksource *cs);
-- list_for_each_entry(worker, &gcwq->idle_list, entry)
-- worker->flags |= WORKER_ROGUE;
-+ case CPU_ONLINE:
-+ spin_unlock_irq(&gcwq->lock);
-+ kthread_bind(gcwq->first_idle->task, cpu);
-+ spin_lock_irq(&gcwq->lock);
-+ gcwq->flags |= GCWQ_MANAGE_WORKERS;
-+ start_worker(gcwq->first_idle);
-+ gcwq->first_idle = NULL;
-+ break;
-+ }
+ struct fsyscall_gtod_data_t fsyscall_gtod_data = {
+- .lock = __SEQLOCK_UNLOCKED(fsyscall_gtod_data.lock),
++ .lock = __RAW_SEQLOCK_UNLOCKED(fsyscall_gtod_data.lock),
+ };
-- for_each_busy_worker(worker, i, pos, gcwq)
-- worker->flags |= WORKER_ROGUE;
-+ spin_unlock_irqrestore(&gcwq->lock, flags);
+ struct itc_jitter_data_t itc_jitter_data;
+@@ -462,7 +462,7 @@ void update_vsyscall(struct timespec *wall, struct timespec *wtm,
+ {
+ unsigned long flags;
-- /*
-- * Call schedule() so that we cross rq->lock and thus can
-- * guarantee sched callbacks see the rogue flag. This is
-- * necessary as scheduler callbacks may be invoked from other
-- * cpus.
-- */
-- spin_unlock_irq(&gcwq->lock);
-- schedule();
-- spin_lock_irq(&gcwq->lock);
-+ return notifier_from_errno(0);
-+}
+- write_seqlock_irqsave(&fsyscall_gtod_data.lock, flags);
++ raw_write_seqlock_irqsave(&fsyscall_gtod_data.lock, flags);
-- /*
-- * Sched callbacks are disabled now. Zap nr_running. After
-- * this, nr_running stays zero and need_more_worker() and
-- * keep_working() are always true as long as the worklist is
-- * not empty.
-- */
-- atomic_set(get_gcwq_nr_running(gcwq->cpu), 0);
-+static void flush_gcwq(struct global_cwq *gcwq)
-+{
-+ struct work_struct *work, *nw;
-+ struct worker *worker, *n;
-+ LIST_HEAD(non_affine_works);
+ /* copy fsyscall clock data */
+ fsyscall_gtod_data.clk_mask = c->mask;
+@@ -485,6 +485,6 @@ void update_vsyscall(struct timespec *wall, struct timespec *wtm,
+ fsyscall_gtod_data.monotonic_time.tv_sec++;
+ }
-- spin_unlock_irq(&gcwq->lock);
-- del_timer_sync(&gcwq->idle_timer);
- spin_lock_irq(&gcwq->lock);
-+ list_for_each_entry_safe(work, nw, &gcwq->worklist, entry) {
-+ struct workqueue_struct *wq = get_work_cwq(work)->wq;
+- write_sequnlock_irqrestore(&fsyscall_gtod_data.lock, flags);
++ raw_write_sequnlock_irqrestore(&fsyscall_gtod_data.lock, flags);
+ }
-- /*
-- * We're now in charge. Notify and proceed to drain. We need
-- * to keep the gcwq running during the whole CPU down
-- * procedure as other cpu hotunplug callbacks may need to
-- * flush currently running tasks.
-- */
-- gcwq->trustee_state = TRUSTEE_IN_CHARGE;
-- wake_up_all(&gcwq->trustee_wait);
--
-- /*
-- * The original cpu is in the process of dying and may go away
-- * anytime now. When that happens, we and all workers would
-- * be migrated to other cpus. Try draining any left work. We
-- * want to get it over with ASAP - spam rescuers, wake up as
-- * many idlers as necessary and create new ones till the
-- * worklist is empty. Note that if the gcwq is frozen, there
-- * may be frozen works in freezable cwqs. Don't declare
-- * completion while frozen.
-- */
-- while (gcwq->nr_workers != gcwq->nr_idle ||
-- gcwq->flags & GCWQ_FREEZING ||
-- gcwq->trustee_state == TRUSTEE_IN_CHARGE) {
-+ if (wq->flags & WQ_NON_AFFINE)
-+ list_move(&work->entry, &non_affine_works);
-+ }
-+
-+ while (!list_empty(&gcwq->worklist)) {
- int nr_works = 0;
+diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c
+index 20b3593..05946c2 100644
+--- a/arch/ia64/mm/fault.c
++++ b/arch/ia64/mm/fault.c
+@@ -89,7 +89,7 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re
+ /*
+ * If we're in an interrupt or have no user context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
- list_for_each_entry(work, &gcwq->worklist, entry) {
-@@ -3367,189 +3256,54 @@ static int __cpuinit trustee_thread(void
- wake_up_process(worker->task);
- }
+ #ifdef CONFIG_VIRTUAL_MEM_MAP
+diff --git a/arch/m32r/kernel/process.c b/arch/m32r/kernel/process.c
+index 422bea9..3a4a32b 100644
+--- a/arch/m32r/kernel/process.c
++++ b/arch/m32r/kernel/process.c
+@@ -90,9 +90,7 @@ void cpu_idle (void)
-+ spin_unlock_irq(&gcwq->lock);
-+
- if (need_to_create_worker(gcwq)) {
-- spin_unlock_irq(&gcwq->lock);
-- worker = create_worker(gcwq, false);
-- spin_lock_irq(&gcwq->lock);
-- if (worker) {
-- worker->flags |= WORKER_ROGUE;
-+ worker = create_worker(gcwq, true);
-+ if (worker)
- start_worker(worker);
-- }
+ idle();
}
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
-- /* give a breather */
-- if (trustee_wait_event_timeout(false, TRUSTEE_COOLDOWN) < 0)
-- break;
-- }
--
-- /*
-- * Either all works have been scheduled and cpu is down, or
-- * cpu down has already been canceled. Wait for and butcher
-- * all workers till we're canceled.
-- */
-- do {
-- rc = trustee_wait_event(!list_empty(&gcwq->idle_list));
-- while (!list_empty(&gcwq->idle_list))
-- destroy_worker(list_first_entry(&gcwq->idle_list,
-- struct worker, entry));
-- } while (gcwq->nr_workers && rc >= 0);
--
-- /*
-- * At this point, either draining has completed and no worker
-- * is left, or cpu down has been canceled or the cpu is being
-- * brought back up. There shouldn't be any idle one left.
-- * Tell the remaining busy ones to rebind once it finishes the
-- * currently scheduled works by scheduling the rebind_work.
-- */
-- WARN_ON(!list_empty(&gcwq->idle_list));
--
-- for_each_busy_worker(worker, i, pos, gcwq) {
-- struct work_struct *rebind_work = &worker->rebind_work;
-+ wait_event_timeout(gcwq->idle_wait,
-+ gcwq->nr_idle == gcwq->nr_workers, HZ/10);
-
-- /*
-- * Rebind_work may race with future cpu hotplug
-- * operations. Use a separate flag to mark that
-- * rebinding is scheduled.
-- */
-- worker->flags |= WORKER_REBIND;
-- worker->flags &= ~WORKER_ROGUE;
-+ spin_lock_irq(&gcwq->lock);
-+ }
-
-- /* queue rebind_work, wq doesn't matter, use the default one */
-- if (test_and_set_bit(WORK_STRUCT_PENDING_BIT,
-- work_data_bits(rebind_work)))
-- continue;
-+ WARN_ON(gcwq->nr_workers != gcwq->nr_idle);
-
-- debug_work_activate(rebind_work);
-- insert_work(get_cwq(gcwq->cpu, system_wq), rebind_work,
-- worker->scheduled.next,
-- work_color_to_flags(WORK_NO_COLOR));
-- }
-+ list_for_each_entry_safe(worker, n, &gcwq->idle_list, entry)
-+ destroy_worker(worker);
-
-- /* relinquish manager role */
-- gcwq->flags &= ~GCWQ_MANAGING_WORKERS;
-+ WARN_ON(gcwq->nr_workers || gcwq->nr_idle);
+diff --git a/arch/m32r/mm/fault.c b/arch/m32r/mm/fault.c
+index 2c9aeb4..6d763f6 100644
+--- a/arch/m32r/mm/fault.c
++++ b/arch/m32r/mm/fault.c
+@@ -115,7 +115,7 @@ asmlinkage void do_page_fault(struct pt_regs *regs, unsigned long error_code,
+ * If we're in an interrupt or have no user context or are running in an
+ * atomic region then we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto bad_area_nosemaphore;
-- /* notify completion */
-- gcwq->trustee = NULL;
-- gcwq->trustee_state = TRUSTEE_DONE;
-- wake_up_all(&gcwq->trustee_wait);
- spin_unlock_irq(&gcwq->lock);
-- return 0;
--}
+ /* When running in the kernel we expect faults to occur only to
+diff --git a/arch/m68k/kernel/process_mm.c b/arch/m68k/kernel/process_mm.c
+index c2a1fc2..65d7bcd 100644
+--- a/arch/m68k/kernel/process_mm.c
++++ b/arch/m68k/kernel/process_mm.c
+@@ -94,9 +94,7 @@ void cpu_idle(void)
+ while (1) {
+ while (!need_resched())
+ idle();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
--/**
-- * wait_trustee_state - wait for trustee to enter the specified state
-- * @gcwq: gcwq the trustee of interest belongs to
-- * @state: target state to wait for
-- *
-- * Wait for the trustee to reach @state. DONE is already matched.
-- *
-- * CONTEXT:
-- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
-- * multiple times. To be used by cpu_callback.
-- */
--static void __cpuinit wait_trustee_state(struct global_cwq *gcwq, int state)
--__releases(&gcwq->lock)
--__acquires(&gcwq->lock)
--{
-- if (!(gcwq->trustee_state == state ||
-- gcwq->trustee_state == TRUSTEE_DONE)) {
-- spin_unlock_irq(&gcwq->lock);
-- __wait_event(gcwq->trustee_wait,
-- gcwq->trustee_state == state ||
-- gcwq->trustee_state == TRUSTEE_DONE);
-- spin_lock_irq(&gcwq->lock);
-+ gcwq = get_gcwq(get_cpu_light());
-+ spin_lock_irq(&gcwq->lock);
-+ list_for_each_entry_safe(work, nw, &non_affine_works, entry) {
-+ list_del_init(&work->entry);
-+ ___queue_work(get_work_cwq(work)->wq, gcwq, work);
+diff --git a/arch/m68k/kernel/process_no.c b/arch/m68k/kernel/process_no.c
+index 9b86ad1..d7877f2 100644
+--- a/arch/m68k/kernel/process_no.c
++++ b/arch/m68k/kernel/process_no.c
+@@ -73,9 +73,7 @@ void cpu_idle(void)
+ /* endless idle loop with no priority at all */
+ while (1) {
+ idle();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
-+ spin_unlock_irq(&gcwq->lock);
-+ put_cpu_light();
}
--static int __devinit workqueue_cpu_callback(struct notifier_block *nfb,
-+static int __devinit workqueue_cpu_down_callback(struct notifier_block *nfb,
- unsigned long action,
- void *hcpu)
- {
- unsigned int cpu = (unsigned long)hcpu;
- struct global_cwq *gcwq = get_gcwq(cpu);
-- struct task_struct *new_trustee = NULL;
-- struct worker *uninitialized_var(new_worker);
-- unsigned long flags;
+diff --git a/arch/m68k/mm/fault.c b/arch/m68k/mm/fault.c
+index 2db6099..74fe559 100644
+--- a/arch/m68k/mm/fault.c
++++ b/arch/m68k/mm/fault.c
+@@ -85,7 +85,7 @@ int do_page_fault(struct pt_regs *regs, unsigned long address,
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
- action &= ~CPU_TASKS_FROZEN;
+ down_read(&mm->mmap_sem);
+diff --git a/arch/microblaze/kernel/early_printk.c b/arch/microblaze/kernel/early_printk.c
+index c3616a0..565f02a 100644
+--- a/arch/microblaze/kernel/early_printk.c
++++ b/arch/microblaze/kernel/early_printk.c
+@@ -21,7 +21,6 @@
+ #include <asm/setup.h>
+ #include <asm/prom.h>
-- switch (action) {
-- case CPU_DOWN_PREPARE:
-- new_trustee = kthread_create(trustee_thread, gcwq,
-- "workqueue_trustee/%d\n", cpu);
-- if (IS_ERR(new_trustee))
-- return notifier_from_errno(PTR_ERR(new_trustee));
-- kthread_bind(new_trustee, cpu);
-- /* fall through */
-- case CPU_UP_PREPARE:
-- BUG_ON(gcwq->first_idle);
-- new_worker = create_worker(gcwq, false);
-- if (!new_worker) {
-- if (new_trustee)
-- kthread_stop(new_trustee);
-- return NOTIFY_BAD;
-- }
-- }
-+ switch (action) {
-+ case CPU_DOWN_PREPARE:
-+ flush_gcwq(gcwq);
-+ break;
-+ }
+-static u32 early_console_initialized;
+ static u32 base_addr;
-- /* some are called w/ irq disabled, don't disturb irq status */
-- spin_lock_irqsave(&gcwq->lock, flags);
--
-- switch (action) {
-- case CPU_DOWN_PREPARE:
-- /* initialize trustee and tell it to acquire the gcwq */
-- BUG_ON(gcwq->trustee || gcwq->trustee_state != TRUSTEE_DONE);
-- gcwq->trustee = new_trustee;
-- gcwq->trustee_state = TRUSTEE_START;
-- wake_up_process(gcwq->trustee);
-- wait_trustee_state(gcwq, TRUSTEE_IN_CHARGE);
-- /* fall through */
-- case CPU_UP_PREPARE:
-- BUG_ON(gcwq->first_idle);
-- gcwq->first_idle = new_worker;
-- break;
--
-- case CPU_DYING:
-- /*
-- * Before this, the trustee and all workers except for
-- * the ones which are still executing works from
-- * before the last CPU down must be on the cpu. After
-- * this, they'll all be diasporas.
-- */
-- gcwq->flags |= GCWQ_DISASSOCIATED;
-- break;
--
-- case CPU_POST_DEAD:
-- gcwq->trustee_state = TRUSTEE_BUTCHER;
-- /* fall through */
-- case CPU_UP_CANCELED:
-- destroy_worker(gcwq->first_idle);
-- gcwq->first_idle = NULL;
-- break;
+ #ifdef CONFIG_SERIAL_UARTLITE_CONSOLE
+@@ -109,31 +108,14 @@ static struct console early_serial_uart16550_console = {
+ };
+ #endif /* CONFIG_SERIAL_8250_CONSOLE */
+
+-static struct console *early_console;
-
-- case CPU_DOWN_FAILED:
-- case CPU_ONLINE:
-- gcwq->flags &= ~GCWQ_DISASSOCIATED;
-- if (gcwq->trustee_state != TRUSTEE_DONE) {
-- gcwq->trustee_state = TRUSTEE_RELEASE;
-- wake_up_process(gcwq->trustee);
-- wait_trustee_state(gcwq, TRUSTEE_DONE);
-- }
+-void early_printk(const char *fmt, ...)
+-{
+- char buf[512];
+- int n;
+- va_list ap;
-
-- /*
-- * Trustee is done and there might be no worker left.
-- * Put the first_idle in and request a real manager to
-- * take a look.
-- */
-- spin_unlock_irq(&gcwq->lock);
-- kthread_bind(gcwq->first_idle->task, cpu);
-- spin_lock_irq(&gcwq->lock);
-- gcwq->flags |= GCWQ_MANAGE_WORKERS;
-- start_worker(gcwq->first_idle);
-- gcwq->first_idle = NULL;
-- break;
+- if (early_console_initialized) {
+- va_start(ap, fmt);
+- n = vscnprintf(buf, 512, fmt, ap);
+- early_console->write(early_console, buf, n);
+- va_end(ap);
- }
+-}
-
-- spin_unlock_irqrestore(&gcwq->lock, flags);
-
- return notifier_from_errno(0);
- }
-@@ -3747,7 +3501,8 @@ static int __init init_workqueues(void)
- unsigned int cpu;
- int i;
-
-- cpu_notifier(workqueue_cpu_callback, CPU_PRI_WORKQUEUE);
-+ cpu_notifier(workqueue_cpu_up_callback, CPU_PRI_WORKQUEUE_ACTIVE);
-+ hotcpu_notifier(workqueue_cpu_down_callback, CPU_PRI_WORKQUEUE_INACTIVE);
-
- /* initialize gcwqs */
- for_each_gcwq_cpu(cpu) {
-@@ -3770,9 +3525,7 @@ static int __init init_workqueues(void)
- (unsigned long)gcwq);
+ int __init setup_early_printk(char *opt)
+ {
+- if (early_console_initialized)
++ if (early_console)
+ return 1;
- ida_init(&gcwq->worker_ida);
--
-- gcwq->trustee_state = TRUSTEE_DONE;
-- init_waitqueue_head(&gcwq->trustee_wait);
-+ init_waitqueue_head(&gcwq->idle_wait);
+ #ifdef CONFIG_SERIAL_UARTLITE_CONSOLE
+ base_addr = early_uartlite_console();
+ if (base_addr) {
+- early_console_initialized = 1;
+ #ifdef CONFIG_MMU
+ early_console_reg_tlb_alloc(base_addr);
+ #endif
+@@ -151,7 +133,6 @@ int __init setup_early_printk(char *opt)
+ base_addr = early_uart16550_console();
+ base_addr &= ~3; /* clear register offset */
+ if (base_addr) {
+- early_console_initialized = 1;
+ #ifdef CONFIG_MMU
+ early_console_reg_tlb_alloc(base_addr);
+ #endif
+@@ -171,9 +152,9 @@ int __init setup_early_printk(char *opt)
+
+ void __init disable_early_printk(void)
+ {
+- if (!early_console_initialized || !early_console)
++ if (!early_console)
+ return;
+ printk(KERN_WARNING "disabling early console\n");
+ unregister_console(early_console);
+- early_console_initialized = 0;
++ early_console = NULL;
+ }
+diff --git a/arch/microblaze/kernel/process.c b/arch/microblaze/kernel/process.c
+index 968648a..b17087e 100644
+--- a/arch/microblaze/kernel/process.c
++++ b/arch/microblaze/kernel/process.c
+@@ -108,9 +108,7 @@ void cpu_idle(void)
+ idle();
+ tick_nohz_restart_sched_tick();
+
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ check_pgt_cache();
}
+ }
+diff --git a/arch/microblaze/mm/fault.c b/arch/microblaze/mm/fault.c
+index ae97d2c..4cdd84d 100644
+--- a/arch/microblaze/mm/fault.c
++++ b/arch/microblaze/mm/fault.c
+@@ -107,7 +107,7 @@ void do_page_fault(struct pt_regs *regs, unsigned long address,
+ if ((error_code & 0x13) == 0x13 || (error_code & 0x11) == 0x11)
+ is_write = 0;
- /* create the initial worker */
-Index: linux-3.0/kernel/workqueue_sched.h
-===================================================================
---- linux-3.0.orig/kernel/workqueue_sched.h
-+++ linux-3.0/kernel/workqueue_sched.h
-@@ -4,6 +4,5 @@
- * Scheduler hooks for concurrency managed workqueue. Only to be
- * included from sched.c and workqueue.c.
- */
--void wq_worker_waking_up(struct task_struct *task, unsigned int cpu);
--struct task_struct *wq_worker_sleeping(struct task_struct *task,
-- unsigned int cpu);
-+void wq_worker_running(struct task_struct *task);
-+void wq_worker_sleeping(struct task_struct *task);
-Index: linux-3.0/arch/mips/sibyte/sb1250/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/sibyte/sb1250/irq.c
-+++ linux-3.0/arch/mips/sibyte/sb1250/irq.c
-@@ -178,7 +178,7 @@ static void ack_sb1250_irq(struct irq_da
+- if (unlikely(in_atomic() || !mm)) {
++ if (unlikely(!mm || pagefault_disabled())) {
+ if (kernel_mode(regs))
+ goto bad_area_nosemaphore;
- static struct irq_chip sb1250_irq_type = {
- .name = "SB1250-IMR",
-- .irq_mask_ack = ack_sb1250_irq,
-+ .irq_mask = ack_sb1250_irq,
- .irq_unmask = enable_sb1250_irq,
- #ifdef CONFIG_SMP
- .irq_set_affinity = sb1250_set_affinity
-Index: linux-3.0/arch/mips/kernel/ftrace.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/ftrace.c
-+++ linux-3.0/arch/mips/kernel/ftrace.c
+diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
+index 653da62..b762223 100644
+--- a/arch/mips/Kconfig
++++ b/arch/mips/Kconfig
+@@ -24,6 +24,7 @@ config MIPS
+ select GENERIC_IRQ_PROBE
+ select GENERIC_IRQ_SHOW
+ select HAVE_ARCH_JUMP_LABEL
++ select IRQ_FORCED_THREADING
+
+ menu "Machine selection"
+
+@@ -2038,7 +2039,7 @@ config CPU_R4400_WORKAROUNDS
+ #
+ config HIGHMEM
+ bool "High Memory Support"
+- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM
++ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !PREEMPT_RT_FULL
+
+ config CPU_SUPPORTS_HIGHMEM
+ bool
+diff --git a/arch/mips/ar7/irq.c b/arch/mips/ar7/irq.c
+index 03db3da..88c4bab 100644
+--- a/arch/mips/ar7/irq.c
++++ b/arch/mips/ar7/irq.c
+@@ -98,7 +98,8 @@ static struct irq_chip ar7_sec_irq_type = {
+
+ static struct irqaction ar7_cascade_action = {
+ .handler = no_action,
+- .name = "AR7 cascade interrupt"
++ .name = "AR7 cascade interrupt",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ static void __init ar7_irq_init(int base)
+diff --git a/arch/mips/bcm63xx/irq.c b/arch/mips/bcm63xx/irq.c
+index cea6021c..162e11b 100644
+--- a/arch/mips/bcm63xx/irq.c
++++ b/arch/mips/bcm63xx/irq.c
+@@ -222,6 +222,7 @@ static struct irq_chip bcm63xx_external_irq_chip = {
+ static struct irqaction cpu_ip2_cascade_action = {
+ .handler = no_action,
+ .name = "cascade_ip2",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ void __init arch_init_irq(void)
+diff --git a/arch/mips/cavium-octeon/smp.c b/arch/mips/cavium-octeon/smp.c
+index 8b60642..efcfff4 100644
+--- a/arch/mips/cavium-octeon/smp.c
++++ b/arch/mips/cavium-octeon/smp.c
+@@ -207,8 +207,9 @@ void octeon_prepare_cpus(unsigned int max_cpus)
+ * the other bits alone.
+ */
+ cvmx_write_csr(CVMX_CIU_MBOX_CLRX(cvmx_get_core_num()), 0xffff);
+- if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt, IRQF_DISABLED,
+- "SMP-IPI", mailbox_interrupt)) {
++ if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt,
++ IRQF_PERCPU | IRQF_NO_THREAD, "SMP-IPI",
++ mailbox_interrupt)) {
+ panic("Cannot request_irq(OCTEON_IRQ_MBOX0)\n");
+ }
+ }
+diff --git a/arch/mips/cobalt/irq.c b/arch/mips/cobalt/irq.c
+index cb9bf82..965c777 100644
+--- a/arch/mips/cobalt/irq.c
++++ b/arch/mips/cobalt/irq.c
+@@ -48,6 +48,7 @@ asmlinkage void plat_irq_dispatch(void)
+ static struct irqaction cascade = {
+ .handler = no_action,
+ .name = "cascade",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ void __init arch_init_irq(void)
+diff --git a/arch/mips/dec/setup.c b/arch/mips/dec/setup.c
+index fa45e92..f7b7ba6 100644
+--- a/arch/mips/dec/setup.c
++++ b/arch/mips/dec/setup.c
+@@ -101,20 +101,24 @@ int cpu_fpu_mask = DEC_CPU_IRQ_MASK(DEC_CPU_INR_FPU);
+ static struct irqaction ioirq = {
+ .handler = no_action,
+ .name = "cascade",
++ .flags = IRQF_NO_THREAD,
+ };
+ static struct irqaction fpuirq = {
+ .handler = no_action,
+ .name = "fpu",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ static struct irqaction busirq = {
+ .flags = IRQF_DISABLED,
+ .name = "bus error",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ static struct irqaction haltirq = {
+ .handler = dec_intr_halt,
+ .name = "halt",
++ .flags = IRQF_NO_THREAD,
+ };
+
+
+diff --git a/arch/mips/emma/markeins/irq.c b/arch/mips/emma/markeins/irq.c
+index 3dbd7a5..7798887 100644
+--- a/arch/mips/emma/markeins/irq.c
++++ b/arch/mips/emma/markeins/irq.c
+@@ -169,7 +169,7 @@ void emma2rh_gpio_irq_init(void)
+
+ static struct irqaction irq_cascade = {
+ .handler = no_action,
+- .flags = 0,
++ .flags = IRQF_NO_THREAD,
+ .name = "cascade",
+ .dev_id = NULL,
+ .next = NULL,
+diff --git a/arch/mips/kernel/early_printk.c b/arch/mips/kernel/early_printk.c
+index 9ae813e..973c995 100644
+--- a/arch/mips/kernel/early_printk.c
++++ b/arch/mips/kernel/early_printk.c
+@@ -25,20 +25,18 @@ early_console_write(struct console *con, const char *s, unsigned n)
+ }
+ }
+
+-static struct console early_console __initdata = {
++static struct console early_console_prom = {
+ .name = "early",
+ .write = early_console_write,
+ .flags = CON_PRINTBUFFER | CON_BOOT,
+ .index = -1
+ };
+
+-static int early_console_initialized __initdata;
+-
+ void __init setup_early_printk(void)
+ {
+- if (early_console_initialized)
++ if (early_console)
+ return;
+- early_console_initialized = 1;
++ early_console = &early_console_prom;
+
+- register_console(&early_console);
++ register_console(&early_console_prom);
+ }
+diff --git a/arch/mips/kernel/ftrace.c b/arch/mips/kernel/ftrace.c
+index feb8021..6a2d758 100644
+--- a/arch/mips/kernel/ftrace.c
++++ b/arch/mips/kernel/ftrace.c
@@ -19,6 +19,26 @@
#include <asm-generic/sections.h>
@@ -3125,7 +2391,7 @@
#ifdef CONFIG_DYNAMIC_FTRACE
#define JAL 0x0c000000 /* jump & link: ip --> ra, jump to target */
-@@ -54,20 +74,6 @@ static inline void ftrace_dyn_arch_init_
+@@ -54,20 +74,6 @@ static inline void ftrace_dyn_arch_init_insns(void)
#endif
}
@@ -3146,7 +2412,7 @@
static int ftrace_modify_code(unsigned long ip, unsigned int new_code)
{
int faulted;
-@@ -112,11 +118,6 @@ static int ftrace_modify_code(unsigned l
+@@ -112,11 +118,6 @@ static int ftrace_modify_code(unsigned long ip, unsigned int new_code)
* 1: offset = 4 instructions
*/
@@ -3158,11 +2424,95 @@
#define INSN_B_1F (0x10000000 | MCOUNT_OFFSET_INSNS)
int ftrace_make_nop(struct module *mod,
-Index: linux-3.0/arch/mips/loongson/fuloong-2e/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/loongson/fuloong-2e/irq.c
-+++ linux-3.0/arch/mips/loongson/fuloong-2e/irq.c
-@@ -42,6 +42,7 @@ asmlinkage void mach_irq_dispatch(unsign
+diff --git a/arch/mips/kernel/i8259.c b/arch/mips/kernel/i8259.c
+index 5c74eb7..fb338db 100644
+--- a/arch/mips/kernel/i8259.c
++++ b/arch/mips/kernel/i8259.c
+@@ -295,6 +295,7 @@ static void init_8259A(int auto_eoi)
+ static struct irqaction irq2 = {
+ .handler = no_action,
+ .name = "cascade",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ static struct resource pic1_io_resource = {
+diff --git a/arch/mips/kernel/process.c b/arch/mips/kernel/process.c
+index d2112d3..5d7c19c 100644
+--- a/arch/mips/kernel/process.c
++++ b/arch/mips/kernel/process.c
+@@ -78,9 +78,7 @@ void __noreturn cpu_idle(void)
+ play_dead();
+ #endif
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
+
+diff --git a/arch/mips/kernel/signal.c b/arch/mips/kernel/signal.c
+index dbbe0ce..395623c 100644
+--- a/arch/mips/kernel/signal.c
++++ b/arch/mips/kernel/signal.c
+@@ -603,6 +603,9 @@ static void do_signal(struct pt_regs *regs)
+ if (!user_mode(regs))
+ return;
+
++ local_irq_enable();
++ preempt_check_resched();
++
+ if (test_thread_flag(TIF_RESTORE_SIGMASK))
+ oldset = ¤t->saved_sigmask;
+ else
+diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c
+index e9b3af2..ba033e3 100644
+--- a/arch/mips/kernel/traps.c
++++ b/arch/mips/kernel/traps.c
+@@ -364,7 +364,7 @@ static int regs_to_trapnr(struct pt_regs *regs)
+ return (regs->cp0_cause >> 2) & 0x1f;
+ }
+
+-static DEFINE_SPINLOCK(die_lock);
++static DEFINE_RAW_SPINLOCK(die_lock);
+
+ void __noreturn die(const char *str, struct pt_regs *regs)
+ {
+@@ -378,7 +378,7 @@ void __noreturn die(const char *str, struct pt_regs *regs)
+ sig = 0;
+
+ console_verbose();
+- spin_lock_irq(&die_lock);
++ raw_spin_lock_irq(&die_lock);
+ bust_spinlocks(1);
+ #ifdef CONFIG_MIPS_MT_SMTC
+ mips_mt_regdump(dvpret);
+@@ -387,7 +387,7 @@ void __noreturn die(const char *str, struct pt_regs *regs)
+ printk("%s[#%d]:\n", str, ++die_counter);
+ show_registers(regs);
+ add_taint(TAINT_DIE);
+- spin_unlock_irq(&die_lock);
++ raw_spin_unlock_irq(&die_lock);
+
+ if (in_interrupt())
+ panic("Fatal exception in interrupt");
+diff --git a/arch/mips/lasat/interrupt.c b/arch/mips/lasat/interrupt.c
+index de4c165..d608b6e 100644
+--- a/arch/mips/lasat/interrupt.c
++++ b/arch/mips/lasat/interrupt.c
+@@ -105,6 +105,7 @@ asmlinkage void plat_irq_dispatch(void)
+ static struct irqaction cascade = {
+ .handler = no_action,
+ .name = "cascade",
++ .flags = IRQF_NO_THREAD,
+ };
+
+ void __init arch_init_irq(void)
+diff --git a/arch/mips/loongson/fuloong-2e/irq.c b/arch/mips/loongson/fuloong-2e/irq.c
+index d61a042..3cf1fef 100644
+--- a/arch/mips/loongson/fuloong-2e/irq.c
++++ b/arch/mips/loongson/fuloong-2e/irq.c
+@@ -42,6 +42,7 @@ asmlinkage void mach_irq_dispatch(unsigned int pending)
static struct irqaction cascade_irqaction = {
.handler = no_action,
.name = "cascade",
@@ -3170,11 +2520,11 @@
};
void __init mach_init_irq(void)
-Index: linux-3.0/arch/mips/loongson/lemote-2f/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/loongson/lemote-2f/irq.c
-+++ linux-3.0/arch/mips/loongson/lemote-2f/irq.c
-@@ -96,12 +96,13 @@ static irqreturn_t ip6_action(int cpl, v
+diff --git a/arch/mips/loongson/lemote-2f/irq.c b/arch/mips/loongson/lemote-2f/irq.c
+index 081db10..14b0818 100644
+--- a/arch/mips/loongson/lemote-2f/irq.c
++++ b/arch/mips/loongson/lemote-2f/irq.c
+@@ -96,12 +96,13 @@ static irqreturn_t ip6_action(int cpl, void *dev_id)
struct irqaction ip6_irqaction = {
.handler = ip6_action,
.name = "cascade",
@@ -3189,124 +2539,45 @@
};
void __init mach_init_irq(void)
-Index: linux-3.0/arch/mips/ar7/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/ar7/irq.c
-+++ linux-3.0/arch/mips/ar7/irq.c
-@@ -98,7 +98,8 @@ static struct irq_chip ar7_sec_irq_type
+diff --git a/arch/mips/mm/fault.c b/arch/mips/mm/fault.c
+index 137ee76..5dc20c4 100644
+--- a/arch/mips/mm/fault.c
++++ b/arch/mips/mm/fault.c
+@@ -88,7 +88,7 @@ asmlinkage void __kprobes do_page_fault(struct pt_regs *regs, unsigned long writ
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto bad_area_nosemaphore;
- static struct irqaction ar7_cascade_action = {
+ down_read(&mm->mmap_sem);
+diff --git a/arch/mips/mti-malta/malta-int.c b/arch/mips/mti-malta/malta-int.c
+index 1d36c511..d53ff91 100644
+--- a/arch/mips/mti-malta/malta-int.c
++++ b/arch/mips/mti-malta/malta-int.c
+@@ -350,12 +350,14 @@ unsigned int plat_ipi_resched_int_xlate(unsigned int cpu)
+
+ static struct irqaction i8259irq = {
.handler = no_action,
-- .name = "AR7 cascade interrupt"
-+ .name = "AR7 cascade interrupt",
+- .name = "XT-PIC cascade"
++ .name = "XT-PIC cascade",
+ .flags = IRQF_NO_THREAD,
};
- static void __init ar7_irq_init(int base)
-Index: linux-3.0/arch/mips/bcm63xx/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/bcm63xx/irq.c
-+++ linux-3.0/arch/mips/bcm63xx/irq.c
-@@ -222,6 +222,7 @@ static struct irq_chip bcm63xx_external_
- static struct irqaction cpu_ip2_cascade_action = {
- .handler = no_action,
- .name = "cascade_ip2",
-+ .flags = IRQF_NO_THREAD,
- };
-
- void __init arch_init_irq(void)
-Index: linux-3.0/arch/mips/cobalt/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/cobalt/irq.c
-+++ linux-3.0/arch/mips/cobalt/irq.c
-@@ -48,6 +48,7 @@ asmlinkage void plat_irq_dispatch(void)
- static struct irqaction cascade = {
- .handler = no_action,
- .name = "cascade",
-+ .flags = IRQF_NO_THREAD,
- };
-
- void __init arch_init_irq(void)
-Index: linux-3.0/arch/mips/dec/setup.c
-===================================================================
---- linux-3.0.orig/arch/mips/dec/setup.c
-+++ linux-3.0/arch/mips/dec/setup.c
-@@ -101,20 +101,24 @@ int cpu_fpu_mask = DEC_CPU_IRQ_MASK(DEC_
- static struct irqaction ioirq = {
- .handler = no_action,
- .name = "cascade",
-+ .flags = IRQF_NO_THREAD,
- };
- static struct irqaction fpuirq = {
- .handler = no_action,
- .name = "fpu",
-+ .flags = IRQF_NO_THREAD,
- };
-
- static struct irqaction busirq = {
- .flags = IRQF_DISABLED,
- .name = "bus error",
-+ .flags = IRQF_NO_THREAD,
- };
-
- static struct irqaction haltirq = {
- .handler = dec_intr_halt,
- .name = "halt",
-+ .flags = IRQF_NO_THREAD,
- };
-
-
-Index: linux-3.0/arch/mips/emma/markeins/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/emma/markeins/irq.c
-+++ linux-3.0/arch/mips/emma/markeins/irq.c
-@@ -169,7 +169,7 @@ void emma2rh_gpio_irq_init(void)
-
- static struct irqaction irq_cascade = {
- .handler = no_action,
-- .flags = 0,
-+ .flags = IRQF_NO_THREAD,
- .name = "cascade",
- .dev_id = NULL,
- .next = NULL,
-Index: linux-3.0/arch/mips/lasat/interrupt.c
-===================================================================
---- linux-3.0.orig/arch/mips/lasat/interrupt.c
-+++ linux-3.0/arch/mips/lasat/interrupt.c
-@@ -105,6 +105,7 @@ asmlinkage void plat_irq_dispatch(void)
- static struct irqaction cascade = {
- .handler = no_action,
- .name = "cascade",
-+ .flags = IRQF_NO_THREAD,
- };
-
- void __init arch_init_irq(void)
-Index: linux-3.0/arch/mips/mti-malta/malta-int.c
-===================================================================
---- linux-3.0.orig/arch/mips/mti-malta/malta-int.c
-+++ linux-3.0/arch/mips/mti-malta/malta-int.c
-@@ -350,12 +350,14 @@ unsigned int plat_ipi_resched_int_xlate(
-
- static struct irqaction i8259irq = {
- .handler = no_action,
-- .name = "XT-PIC cascade"
-+ .name = "XT-PIC cascade",
-+ .flags = IRQF_NO_THREAD,
- };
-
- static struct irqaction corehi_irqaction = {
- .handler = no_action,
-- .name = "CoreHi"
-+ .name = "CoreHi",
-+ .flags = IRQF_NO_THREAD,
+ static struct irqaction corehi_irqaction = {
+ .handler = no_action,
+- .name = "CoreHi"
++ .name = "CoreHi",
++ .flags = IRQF_NO_THREAD,
};
static msc_irqmap_t __initdata msc_irqmap[] = {
-Index: linux-3.0/arch/mips/pmc-sierra/msp71xx/msp_irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/pmc-sierra/msp71xx/msp_irq.c
-+++ linux-3.0/arch/mips/pmc-sierra/msp71xx/msp_irq.c
-@@ -109,11 +109,13 @@ asmlinkage void plat_irq_dispatch(struct
+diff --git a/arch/mips/pmc-sierra/msp71xx/msp_irq.c b/arch/mips/pmc-sierra/msp71xx/msp_irq.c
+index 4531c4a..42a4c04 100644
+--- a/arch/mips/pmc-sierra/msp71xx/msp_irq.c
++++ b/arch/mips/pmc-sierra/msp71xx/msp_irq.c
+@@ -109,11 +109,13 @@ asmlinkage void plat_irq_dispatch(struct pt_regs *regs)
static struct irqaction cic_cascade_msp = {
.handler = no_action,
.name = "MSP CIC cascade"
@@ -3320,11 +2591,11 @@
};
void __init arch_init_irq(void)
-Index: linux-3.0/arch/mips/pnx8550/common/int.c
-===================================================================
---- linux-3.0.orig/arch/mips/pnx8550/common/int.c
-+++ linux-3.0/arch/mips/pnx8550/common/int.c
-@@ -167,7 +167,7 @@ static struct irq_chip level_irq_type =
+diff --git a/arch/mips/pnx8550/common/int.c b/arch/mips/pnx8550/common/int.c
+index 6b93c81..1ebe22b 100644
+--- a/arch/mips/pnx8550/common/int.c
++++ b/arch/mips/pnx8550/common/int.c
+@@ -167,7 +167,7 @@ static struct irq_chip level_irq_type = {
static struct irqaction gic_action = {
.handler = no_action,
@@ -3333,11 +2604,11 @@
.name = "GIC",
};
-Index: linux-3.0/arch/mips/sgi-ip22/ip22-int.c
-===================================================================
---- linux-3.0.orig/arch/mips/sgi-ip22/ip22-int.c
-+++ linux-3.0/arch/mips/sgi-ip22/ip22-int.c
-@@ -155,32 +155,32 @@ static void __irq_entry indy_buserror_ir
+diff --git a/arch/mips/sgi-ip22/ip22-int.c b/arch/mips/sgi-ip22/ip22-int.c
+index b4d08e4..f72c336 100644
+--- a/arch/mips/sgi-ip22/ip22-int.c
++++ b/arch/mips/sgi-ip22/ip22-int.c
+@@ -155,32 +155,32 @@ static void __irq_entry indy_buserror_irq(void)
static struct irqaction local0_cascade = {
.handler = no_action,
@@ -3375,10 +2646,23 @@
.name = "mapable1 cascade",
};
#define SGI_INTERRUPTS SGINT_END
-Index: linux-3.0/arch/mips/sni/rm200.c
-===================================================================
---- linux-3.0.orig/arch/mips/sni/rm200.c
-+++ linux-3.0/arch/mips/sni/rm200.c
+diff --git a/arch/mips/sibyte/sb1250/irq.c b/arch/mips/sibyte/sb1250/irq.c
+index be4460a..5cef2a0 100644
+--- a/arch/mips/sibyte/sb1250/irq.c
++++ b/arch/mips/sibyte/sb1250/irq.c
+@@ -178,7 +178,7 @@ static void ack_sb1250_irq(struct irq_data *d)
+
+ static struct irq_chip sb1250_irq_type = {
+ .name = "SB1250-IMR",
+- .irq_mask_ack = ack_sb1250_irq,
++ .irq_mask = ack_sb1250_irq,
+ .irq_unmask = enable_sb1250_irq,
+ #ifdef CONFIG_SMP
+ .irq_set_affinity = sb1250_set_affinity
+diff --git a/arch/mips/sni/rm200.c b/arch/mips/sni/rm200.c
+index a7e5a6d..3ab5b5d 100644
+--- a/arch/mips/sni/rm200.c
++++ b/arch/mips/sni/rm200.c
@@ -359,6 +359,7 @@ void sni_rm200_init_8259A(void)
static struct irqaction sni_rm200_irq2 = {
.handler = no_action,
@@ -3387,11 +2671,11 @@
};
static struct resource sni_rm200_pic1_resource = {
-Index: linux-3.0/arch/mips/vr41xx/common/irq.c
-===================================================================
---- linux-3.0.orig/arch/mips/vr41xx/common/irq.c
-+++ linux-3.0/arch/mips/vr41xx/common/irq.c
-@@ -34,6 +34,7 @@ static irq_cascade_t irq_cascade[NR_IRQS
+diff --git a/arch/mips/vr41xx/common/irq.c b/arch/mips/vr41xx/common/irq.c
+index 70a3b85..fad2bef 100644
+--- a/arch/mips/vr41xx/common/irq.c
++++ b/arch/mips/vr41xx/common/irq.c
+@@ -34,6 +34,7 @@ static irq_cascade_t irq_cascade[NR_IRQS] __cacheline_aligned;
static struct irqaction cascade_irqaction = {
.handler = no_action,
.name = "cascade",
@@ -3399,565 +2683,327 @@
};
int cascade_irq(unsigned int irq, int (*get_irq)(unsigned int))
-Index: linux-3.0/arch/mips/Kconfig
-===================================================================
---- linux-3.0.orig/arch/mips/Kconfig
-+++ linux-3.0/arch/mips/Kconfig
-@@ -24,6 +24,7 @@ config MIPS
- select GENERIC_IRQ_PROBE
- select GENERIC_IRQ_SHOW
- select HAVE_ARCH_JUMP_LABEL
-+ select IRQ_FORCED_THREADING
+diff --git a/arch/mn10300/kernel/process.c b/arch/mn10300/kernel/process.c
+index 28eec31..cac401d 100644
+--- a/arch/mn10300/kernel/process.c
++++ b/arch/mn10300/kernel/process.c
+@@ -123,9 +123,7 @@ void cpu_idle(void)
+ idle();
+ }
- menu "Machine selection"
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
-@@ -2038,7 +2039,7 @@ config CPU_R4400_WORKAROUNDS
- #
- config HIGHMEM
- bool "High Memory Support"
-- depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM
-+ depends on 32BIT && CPU_SUPPORTS_HIGHMEM && SYS_SUPPORTS_HIGHMEM && !PREEMPT_RT_FULL
+diff --git a/arch/mn10300/mm/fault.c b/arch/mn10300/mm/fault.c
+index 0945409..2fea01c 100644
+--- a/arch/mn10300/mm/fault.c
++++ b/arch/mn10300/mm/fault.c
+@@ -168,7 +168,7 @@ asmlinkage void do_page_fault(struct pt_regs *regs, unsigned long fault_code,
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
- config CPU_SUPPORTS_HIGHMEM
- bool
-Index: linux-3.0/arch/mips/kernel/traps.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/traps.c
-+++ linux-3.0/arch/mips/kernel/traps.c
-@@ -364,7 +364,7 @@ static int regs_to_trapnr(struct pt_regs
- return (regs->cp0_cause >> 2) & 0x1f;
+ down_read(&mm->mmap_sem);
+diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c
+index 4b4b918..f6eb367 100644
+--- a/arch/parisc/kernel/process.c
++++ b/arch/parisc/kernel/process.c
+@@ -71,9 +71,7 @@ void cpu_idle(void)
+ while (1) {
+ while (!need_resched())
+ barrier();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ check_pgt_cache();
+ }
}
+diff --git a/arch/parisc/mm/fault.c b/arch/parisc/mm/fault.c
+index 18162ce..df22f39 100644
+--- a/arch/parisc/mm/fault.c
++++ b/arch/parisc/mm/fault.c
+@@ -176,7 +176,7 @@ void do_page_fault(struct pt_regs *regs, unsigned long code,
+ unsigned long acc_type;
+ int fault;
--static DEFINE_SPINLOCK(die_lock);
-+static DEFINE_RAW_SPINLOCK(die_lock);
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
- void __noreturn die(const char *str, struct pt_regs *regs)
- {
-@@ -378,7 +378,7 @@ void __noreturn die(const char *str, str
- sig = 0;
+ down_read(&mm->mmap_sem);
+diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
+index 2729c66..80c73b6 100644
+--- a/arch/powerpc/Kconfig
++++ b/arch/powerpc/Kconfig
+@@ -69,10 +69,11 @@ config LOCKDEP_SUPPORT
- console_verbose();
-- spin_lock_irq(&die_lock);
-+ raw_spin_lock_irq(&die_lock);
- bust_spinlocks(1);
- #ifdef CONFIG_MIPS_MT_SMTC
- mips_mt_regdump(dvpret);
-@@ -387,7 +387,7 @@ void __noreturn die(const char *str, str
- printk("%s[#%d]:\n", str, ++die_counter);
- show_registers(regs);
- add_taint(TAINT_DIE);
-- spin_unlock_irq(&die_lock);
-+ raw_spin_unlock_irq(&die_lock);
+ config RWSEM_GENERIC_SPINLOCK
+ bool
++ default y if PREEMPT_RT_FULL
- if (in_interrupt())
- panic("Fatal exception in interrupt");
-Index: linux-3.0/arch/mips/kernel/signal.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/signal.c
-+++ linux-3.0/arch/mips/kernel/signal.c
-@@ -603,6 +603,9 @@ static void do_signal(struct pt_regs *re
- if (!user_mode(regs))
- return;
+ config RWSEM_XCHGADD_ALGORITHM
+ bool
+- default y
++ default y if !PREEMPT_RT_FULL
-+ local_irq_enable();
-+ preempt_check_resched();
-+
- if (test_thread_flag(TIF_RESTORE_SIGMASK))
- oldset = ¤t->saved_sigmask;
- else
-Index: linux-3.0/arch/mips/kernel/i8259.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/i8259.c
-+++ linux-3.0/arch/mips/kernel/i8259.c
-@@ -295,6 +295,7 @@ static void init_8259A(int auto_eoi)
- static struct irqaction irq2 = {
- .handler = no_action,
- .name = "cascade",
-+ .flags = IRQF_NO_THREAD,
- };
+ config GENERIC_LOCKBREAK
+ bool
+@@ -134,6 +135,7 @@ config PPC
+ select GENERIC_IRQ_SHOW_LEVEL
+ select HAVE_RCU_TABLE_FREE if SMP
+ select HAVE_SYSCALL_TRACEPOINTS
++ select IRQ_FORCED_THREADING
- static struct resource pic1_io_resource = {
-Index: linux-3.0/drivers/watchdog/octeon-wdt-main.c
-===================================================================
---- linux-3.0.orig/drivers/watchdog/octeon-wdt-main.c
-+++ linux-3.0/drivers/watchdog/octeon-wdt-main.c
-@@ -402,7 +402,7 @@ static void octeon_wdt_setup_interrupt(i
- irq = OCTEON_IRQ_WDOG0 + core;
+ config EARLY_PRINTK
+ bool
+@@ -271,7 +273,7 @@ menu "Kernel options"
- if (request_irq(irq, octeon_wdt_poke_irq,
-- IRQF_DISABLED, "octeon_wdt", octeon_wdt_poke_irq))
-+ IRQF_NO_THREAD, "octeon_wdt", octeon_wdt_poke_irq))
- panic("octeon_wdt: Couldn't obtain irq %d", irq);
+ config HIGHMEM
+ bool "High memory support"
+- depends on PPC32
++ depends on PPC32 && !PREEMPT_RT_FULL
- cpumask_set_cpu(cpu, &irq_enabled_cpus);
-Index: linux-3.0/arch/mips/cavium-octeon/smp.c
-===================================================================
---- linux-3.0.orig/arch/mips/cavium-octeon/smp.c
-+++ linux-3.0/arch/mips/cavium-octeon/smp.c
-@@ -207,8 +207,9 @@ void octeon_prepare_cpus(unsigned int ma
- * the other bits alone.
- */
- cvmx_write_csr(CVMX_CIU_MBOX_CLRX(cvmx_get_core_num()), 0xffff);
-- if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt, IRQF_DISABLED,
-- "SMP-IPI", mailbox_interrupt)) {
-+ if (request_irq(OCTEON_IRQ_MBOX0, mailbox_interrupt,
-+ IRQF_PERCPU | IRQF_NO_THREAD, "SMP-IPI",
-+ mailbox_interrupt)) {
- panic("Cannot request_irq(OCTEON_IRQ_MBOX0)\n");
- }
+ source kernel/time/Kconfig
+ source kernel/Kconfig.hz
+diff --git a/arch/powerpc/include/asm/rwsem.h b/arch/powerpc/include/asm/rwsem.h
+index bb1e2cd..06b28a7 100644
+--- a/arch/powerpc/include/asm/rwsem.h
++++ b/arch/powerpc/include/asm/rwsem.h
+@@ -31,13 +31,13 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
+ {
+ if (unlikely(atomic_long_inc_return((atomic_long_t *)&sem->count) <= 0))
+ rwsem_down_read_failed(sem);
}
-Index: linux-3.0/arch/arm/kernel/signal.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/signal.c
-+++ linux-3.0/arch/arm/kernel/signal.c
-@@ -673,6 +673,9 @@ static void do_signal(struct pt_regs *re
- if (!user_mode(regs))
- return;
-+ local_irq_enable();
-+ preempt_check_resched();
-+
- /*
- * If we were from a system call, check for system call restarting...
- */
-Index: linux-3.0/arch/arm/kernel/smp.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/smp.c
-+++ linux-3.0/arch/arm/kernel/smp.c
-@@ -305,6 +305,18 @@ asmlinkage void __cpuinit secondary_star
- * Enable local interrupts.
- */
- notify_cpu_starting(cpu);
-+
-+ /*
-+ * OK, now it's safe to let the boot CPU continue. Wait for
-+ * the CPU migration code to notice that the CPU is online
-+ * before we continue. We need to do that before we enable
-+ * interrupts otherwise a wakeup of a kernel thread affine to
-+ * this CPU might break the affinity and let hell break lose.
-+ */
-+ set_cpu_online(cpu, true);
-+ while (!cpu_active(cpu))
-+ cpu_relax();
-+
- local_irq_enable();
- local_fiq_enable();
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
-@@ -318,15 +330,6 @@ asmlinkage void __cpuinit secondary_star
- smp_store_cpu_info(cpu);
+@@ -53,7 +53,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ /*
+ * lock for writing
+ */
+-static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
++static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ long tmp;
- /*
-- * OK, now it's safe to let the boot CPU continue. Wait for
-- * the CPU migration code to notice that the CPU is online
-- * before we continue.
-- */
-- set_cpu_online(cpu, true);
-- while (!cpu_active(cpu))
-- cpu_relax();
--
-- /*
- * OK, it's off to the idle thread for us
- */
- cpu_idle();
-@@ -531,7 +534,7 @@ static void percpu_timer_stop(void)
+@@ -63,12 +63,12 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
+ rwsem_down_write_failed(sem);
}
- #endif
--static DEFINE_SPINLOCK(stop_lock);
-+static DEFINE_RAW_SPINLOCK(stop_lock);
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
+ {
+ __down_write_nested(sem, 0);
+ }
+
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
+@@ -80,7 +80,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
/*
- * ipi_cpu_stop - handle IPI from smp_send_stop()
-@@ -540,10 +543,10 @@ static void ipi_cpu_stop(unsigned int cp
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
{
- if (system_state == SYSTEM_BOOTING ||
- system_state == SYSTEM_RUNNING) {
-- spin_lock(&stop_lock);
-+ raw_spin_lock(&stop_lock);
- printk(KERN_CRIT "CPU%u: stopping\n", cpu);
- dump_stack();
-- spin_unlock(&stop_lock);
-+ raw_spin_unlock(&stop_lock);
- }
+ long tmp;
- set_cpu_online(cpu, false);
-Index: linux-3.0/include/linux/jbd.h
-===================================================================
---- linux-3.0.orig/include/linux/jbd.h
-+++ linux-3.0/include/linux/jbd.h
-@@ -244,6 +244,7 @@ typedef struct journal_superblock_s
+@@ -92,7 +92,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
+ {
+ if (unlikely(atomic_long_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+ (atomic_long_t *)&sem->count) < 0))
+@@ -102,7 +102,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
+ {
+ atomic_long_add(delta, (atomic_long_t *)&sem->count);
+ }
+@@ -110,7 +110,7 @@ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
- #include <linux/fs.h>
- #include <linux/sched.h>
-+#include <linux/jbd_common.h>
+@@ -123,7 +123,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
+ /*
+ * implement exchange and add functionality
+ */
+-static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
++static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
+ {
+ return atomic_long_add_return(delta, (atomic_long_t *)&sem->count);
+ }
+diff --git a/arch/powerpc/kernel/idle.c b/arch/powerpc/kernel/idle.c
+index 39a2baa..5d70d10 100644
+--- a/arch/powerpc/kernel/idle.c
++++ b/arch/powerpc/kernel/idle.c
+@@ -94,11 +94,11 @@ void cpu_idle(void)
+ HMT_medium();
+ ppc64_runlatch_on();
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- if (cpu_should_die())
++ if (cpu_should_die()) {
++ __preempt_enable_no_resched();
+ cpu_die();
+- schedule();
+- preempt_disable();
++ }
++ schedule_preempt_disabled();
+ }
+ }
- #define J_ASSERT(assert) BUG_ON(!(assert))
+diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c
+index 5b428e3..70e88b2 100644
+--- a/arch/powerpc/kernel/irq.c
++++ b/arch/powerpc/kernel/irq.c
+@@ -449,6 +449,7 @@ void irq_ctx_init(void)
+ }
+ }
-@@ -270,69 +271,6 @@ typedef struct journal_superblock_s
- #define J_EXPECT_JH(jh, expr, why...) __journal_expect(expr, ## why)
- #endif
++#ifndef CONFIG_PREEMPT_RT_FULL
+ static inline void do_softirq_onstack(void)
+ {
+ struct thread_info *curtp, *irqtp;
+@@ -478,7 +479,7 @@ void do_softirq(void)
--enum jbd_state_bits {
-- BH_JBD /* Has an attached ext3 journal_head */
-- = BH_PrivateStart,
-- BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
-- BH_Freed, /* Has been freed (truncated) */
-- BH_Revoked, /* Has been revoked from the log */
-- BH_RevokeValid, /* Revoked flag is valid */
-- BH_JBDDirty, /* Is dirty but journaled */
-- BH_State, /* Pins most journal_head state */
-- BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
-- BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
--};
--
--BUFFER_FNS(JBD, jbd)
--BUFFER_FNS(JWrite, jwrite)
--BUFFER_FNS(JBDDirty, jbddirty)
--TAS_BUFFER_FNS(JBDDirty, jbddirty)
--BUFFER_FNS(Revoked, revoked)
--TAS_BUFFER_FNS(Revoked, revoked)
--BUFFER_FNS(RevokeValid, revokevalid)
--TAS_BUFFER_FNS(RevokeValid, revokevalid)
--BUFFER_FNS(Freed, freed)
--
--static inline struct buffer_head *jh2bh(struct journal_head *jh)
--{
-- return jh->b_bh;
--}
--
--static inline struct journal_head *bh2jh(struct buffer_head *bh)
--{
-- return bh->b_private;
--}
--
--static inline void jbd_lock_bh_state(struct buffer_head *bh)
--{
-- bit_spin_lock(BH_State, &bh->b_state);
--}
--
--static inline int jbd_trylock_bh_state(struct buffer_head *bh)
--{
-- return bit_spin_trylock(BH_State, &bh->b_state);
--}
--
--static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
--{
-- return bit_spin_is_locked(BH_State, &bh->b_state);
--}
--
--static inline void jbd_unlock_bh_state(struct buffer_head *bh)
--{
-- bit_spin_unlock(BH_State, &bh->b_state);
--}
--
--static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
--{
-- bit_spin_lock(BH_JournalHead, &bh->b_state);
--}
--
--static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
--{
-- bit_spin_unlock(BH_JournalHead, &bh->b_state);
--}
+ local_irq_restore(flags);
+ }
-
- struct jbd_revoke_table_s;
++#endif
- /**
-Index: linux-3.0/include/linux/jbd2.h
-===================================================================
---- linux-3.0.orig/include/linux/jbd2.h
-+++ linux-3.0/include/linux/jbd2.h
-@@ -275,6 +275,7 @@ typedef struct journal_superblock_s
+ /*
+ * IRQ controller and virtual interrupts
+diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S
+index 998a100..c997f75 100644
+--- a/arch/powerpc/kernel/misc_32.S
++++ b/arch/powerpc/kernel/misc_32.S
+@@ -34,6 +34,7 @@
- #include <linux/fs.h>
- #include <linux/sched.h>
-+#include <linux/jbd_common.h>
+ .text
- #define J_ASSERT(assert) BUG_ON(!(assert))
++#ifndef CONFIG_PREEMPT_RT_FULL
+ _GLOBAL(call_do_softirq)
+ mflr r0
+ stw r0,4(r1)
+@@ -44,6 +45,7 @@ _GLOBAL(call_do_softirq)
+ lwz r0,4(r1)
+ mtlr r0
+ blr
++#endif
-@@ -302,70 +303,6 @@ typedef struct journal_superblock_s
- #define J_EXPECT_JH(jh, expr, why...) __journal_expect(expr, ## why)
+ _GLOBAL(call_handle_irq)
+ mflr r0
+diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S
+index e89df59..b8b9e20 100644
+--- a/arch/powerpc/kernel/misc_64.S
++++ b/arch/powerpc/kernel/misc_64.S
+@@ -29,6 +29,7 @@
+
+ .text
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ _GLOBAL(call_do_softirq)
+ mflr r0
+ std r0,16(r1)
+@@ -39,6 +40,7 @@ _GLOBAL(call_do_softirq)
+ ld r0,16(r1)
+ mtlr r0
+ blr
++#endif
+
+ _GLOBAL(call_handle_irq)
+ ld r8,0(r6)
+diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c
+index 8ebc670..8fb0cf2 100644
+--- a/arch/powerpc/kernel/smp.c
++++ b/arch/powerpc/kernel/smp.c
+@@ -170,7 +170,7 @@ int smp_request_message_ipi(int virq, int msg)
+ return 1;
+ }
#endif
+- err = request_irq(virq, smp_ipi_action[msg], IRQF_DISABLED|IRQF_PERCPU,
++ err = request_irq(virq, smp_ipi_action[msg], IRQF_NO_THREAD|IRQF_PERCPU,
+ smp_ipi_name[msg], 0);
+ WARN(err < 0, "unable to request_irq %d for %s (rc %d)\n",
+ virq, smp_ipi_name[msg], err);
+diff --git a/arch/powerpc/kernel/udbg.c b/arch/powerpc/kernel/udbg.c
+index 23d65ab..d6875fe 100644
+--- a/arch/powerpc/kernel/udbg.c
++++ b/arch/powerpc/kernel/udbg.c
+@@ -169,15 +169,13 @@ static struct console udbg_console = {
+ .index = 0,
+ };
--enum jbd_state_bits {
-- BH_JBD /* Has an attached ext3 journal_head */
-- = BH_PrivateStart,
-- BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
-- BH_Freed, /* Has been freed (truncated) */
-- BH_Revoked, /* Has been revoked from the log */
-- BH_RevokeValid, /* Revoked flag is valid */
-- BH_JBDDirty, /* Is dirty but journaled */
-- BH_State, /* Pins most journal_head state */
-- BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
-- BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
-- BH_JBDPrivateStart, /* First bit available for private use by FS */
--};
--
--BUFFER_FNS(JBD, jbd)
--BUFFER_FNS(JWrite, jwrite)
--BUFFER_FNS(JBDDirty, jbddirty)
--TAS_BUFFER_FNS(JBDDirty, jbddirty)
--BUFFER_FNS(Revoked, revoked)
--TAS_BUFFER_FNS(Revoked, revoked)
--BUFFER_FNS(RevokeValid, revokevalid)
--TAS_BUFFER_FNS(RevokeValid, revokevalid)
--BUFFER_FNS(Freed, freed)
--
--static inline struct buffer_head *jh2bh(struct journal_head *jh)
--{
-- return jh->b_bh;
--}
--
--static inline struct journal_head *bh2jh(struct buffer_head *bh)
--{
-- return bh->b_private;
--}
--
--static inline void jbd_lock_bh_state(struct buffer_head *bh)
--{
-- bit_spin_lock(BH_State, &bh->b_state);
--}
--
--static inline int jbd_trylock_bh_state(struct buffer_head *bh)
--{
-- return bit_spin_trylock(BH_State, &bh->b_state);
--}
--
--static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
--{
-- return bit_spin_is_locked(BH_State, &bh->b_state);
--}
+-static int early_console_initialized;
-
--static inline void jbd_unlock_bh_state(struct buffer_head *bh)
--{
-- bit_spin_unlock(BH_State, &bh->b_state);
--}
--
--static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
--{
-- bit_spin_lock(BH_JournalHead, &bh->b_state);
--}
--
--static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
--{
-- bit_spin_unlock(BH_JournalHead, &bh->b_state);
--}
--
- /* Flags in jbd_inode->i_flags */
- #define __JI_COMMIT_RUNNING 0
- /* Commit of the inode data in progress. We use this flag to protect us from
-Index: linux-3.0/include/linux/jbd_common.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/jbd_common.h
-@@ -0,0 +1,92 @@
-+#ifndef _LINUX_JBD_STATE_H
-+#define _LINUX_JBD_STATE_H
-+
-+enum jbd_state_bits {
-+ BH_JBD /* Has an attached ext3 journal_head */
-+ = BH_PrivateStart,
-+ BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
-+ BH_Freed, /* Has been freed (truncated) */
-+ BH_Revoked, /* Has been revoked from the log */
-+ BH_RevokeValid, /* Revoked flag is valid */
-+ BH_JBDDirty, /* Is dirty but journaled */
-+ BH_State, /* Pins most journal_head state */
-+ BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
-+ BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
-+ BH_JBDPrivateStart, /* First bit available for private use by FS */
-+};
-+
-+BUFFER_FNS(JBD, jbd)
-+BUFFER_FNS(JWrite, jwrite)
-+BUFFER_FNS(JBDDirty, jbddirty)
-+TAS_BUFFER_FNS(JBDDirty, jbddirty)
-+BUFFER_FNS(Revoked, revoked)
-+TAS_BUFFER_FNS(Revoked, revoked)
-+BUFFER_FNS(RevokeValid, revokevalid)
-+TAS_BUFFER_FNS(RevokeValid, revokevalid)
-+BUFFER_FNS(Freed, freed)
-+
-+static inline struct buffer_head *jh2bh(struct journal_head *jh)
-+{
-+ return jh->b_bh;
-+}
-+
-+static inline struct journal_head *bh2jh(struct buffer_head *bh)
-+{
-+ return bh->b_private;
-+}
-+
-+static inline void jbd_lock_bh_state(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ bit_spin_lock(BH_State, &bh->b_state);
-+#else
-+ spin_lock(&bh->b_state_lock);
-+#endif
-+}
-+
-+static inline int jbd_trylock_bh_state(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ return bit_spin_trylock(BH_State, &bh->b_state);
-+#else
-+ return spin_trylock(&bh->b_state_lock);
-+#endif
-+}
-+
-+static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ return bit_spin_is_locked(BH_State, &bh->b_state);
-+#else
-+ return spin_is_locked(&bh->b_state_lock);
-+#endif
-+}
-+
-+static inline void jbd_unlock_bh_state(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ bit_spin_unlock(BH_State, &bh->b_state);
-+#else
-+ spin_unlock(&bh->b_state_lock);
-+#endif
-+}
-+
-+static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ bit_spin_lock(BH_JournalHead, &bh->b_state);
-+#else
-+ spin_lock(&bh->b_journal_head_lock);
-+#endif
-+}
-+
-+static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ bit_spin_unlock(BH_JournalHead, &bh->b_state);
-+#else
-+ spin_unlock(&bh->b_journal_head_lock);
-+#endif
-+}
-+
-+#endif
-Index: linux-3.0/kernel/sched_rt.c
-===================================================================
---- linux-3.0.orig/kernel/sched_rt.c
-+++ linux-3.0/kernel/sched_rt.c
-@@ -536,6 +536,9 @@ static int balance_runtime(struct rt_rq
- {
- int more = 0;
-
-+ if (!sched_feat(RT_RUNTIME_SHARE))
-+ return more;
-+
- if (rt_rq->rt_time > rt_rq->rt_runtime) {
- raw_spin_unlock(&rt_rq->rt_runtime_lock);
- more = do_balance_runtime(rt_rq);
-@@ -553,12 +556,9 @@ static inline int balance_runtime(struct
-
- static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
+ /*
+ * Called by setup_system after ppc_md->probe and ppc_md->early_init.
+ * Call it again after setting udbg_putc in ppc_md->setup_arch.
+ */
+ void __init register_early_udbg_console(void)
{
-- int i, idle = 1;
-+ int i, idle = 1, throttled = 0;
- const struct cpumask *span;
-
-- if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF)
-- return 1;
--
- span = sched_rt_period_mask();
- for_each_cpu(i, span) {
- int enqueue = 0;
-@@ -593,12 +593,17 @@ static int do_sched_rt_period_timer(stru
- if (!rt_rq_throttled(rt_rq))
- enqueue = 1;
- }
-+ if (rt_rq->rt_throttled)
-+ throttled = 1;
+- if (early_console_initialized)
++ if (early_console)
+ return;
- if (enqueue)
- sched_rt_rq_enqueue(rt_rq);
- raw_spin_unlock(&rq->lock);
+ if (!udbg_putc)
+@@ -187,7 +185,7 @@ void __init register_early_udbg_console(void)
+ printk(KERN_INFO "early console immortal !\n");
+ udbg_console.flags &= ~CON_BOOT;
}
-
-+ if (!throttled && (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF))
-+ return 1;
-+
- return idle;
+- early_console_initialized = 1;
++ early_console = &udbg_console;
+ register_console(&udbg_console);
}
-@@ -630,7 +635,24 @@ static int sched_rt_runtime_exceeded(str
- return 0;
-
- if (rt_rq->rt_time > runtime) {
-- rt_rq->rt_throttled = 1;
-+ struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
-+
-+ /*
-+ * Don't actually throttle groups that have no runtime assigned
-+ * but accrue some time due to boosting.
-+ */
-+ if (likely(rt_b->rt_runtime)) {
-+ rt_rq->rt_throttled = 1;
-+ printk_once(KERN_WARNING "sched: RT throttling activated\n");
-+ } else {
-+ /*
-+ * In case we did anyway, make it go away,
-+ * replenishment is a joke, since it will replenish us
-+ * with exactly 0 ns.
-+ */
-+ rt_rq->rt_time = 0;
-+ }
-+
- if (rt_rq_throttled(rt_rq)) {
- sched_rt_rq_dequeue(rt_rq);
- return 1;
-@@ -658,7 +680,8 @@ static void update_curr_rt(struct rq *rq
- if (unlikely((s64)delta_exec < 0))
- delta_exec = 0;
-
-- schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec));
-+ schedstat_set(curr->se.statistics.exec_max,
-+ max(curr->se.statistics.exec_max, delta_exec));
-
- curr->se.sum_exec_runtime += delta_exec;
- account_group_exec_runtime(curr, delta_exec);
-@@ -1186,7 +1209,7 @@ static void deactivate_task(struct rq *r
- static int pick_rt_task(struct rq *rq, struct task_struct *p, int cpu)
- {
- if (!task_running(rq, p) &&
-- (cpu < 0 || cpumask_test_cpu(cpu, &p->cpus_allowed)) &&
-+ (cpu < 0 || cpumask_test_cpu(cpu, tsk_cpus_allowed(p))) &&
- (p->rt.nr_cpus_allowed > 1))
- return 1;
- return 0;
-@@ -1331,7 +1354,7 @@ static struct rq *find_lock_lowest_rq(st
- */
- if (unlikely(task_rq(task) != rq ||
- !cpumask_test_cpu(lowest_rq->cpu,
-- &task->cpus_allowed) ||
-+ tsk_cpus_allowed(task)) ||
- task_running(rq, task) ||
- !task->on_rq)) {
-
-@@ -1614,9 +1637,6 @@ static void set_cpus_allowed_rt(struct t
-
- update_rt_migration(&rq->rt);
+diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
+index ad35f66..c5dc467 100644
+--- a/arch/powerpc/mm/fault.c
++++ b/arch/powerpc/mm/fault.c
+@@ -162,7 +162,7 @@ int __kprobes do_page_fault(struct pt_regs *regs, unsigned long address,
}
--
-- cpumask_copy(&p->cpus_allowed, new_mask);
-- p->rt.nr_cpus_allowed = weight;
- }
+ #endif
- /* Assumes rq->lock is held */
-Index: linux-3.0/arch/powerpc/platforms/85xx/mpc85xx_cds.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/platforms/85xx/mpc85xx_cds.c
-+++ linux-3.0/arch/powerpc/platforms/85xx/mpc85xx_cds.c
-@@ -178,7 +178,7 @@ static irqreturn_t mpc85xx_8259_cascade_
+- if (in_atomic() || mm == NULL) {
++ if (!mm || pagefault_disabled()) {
+ if (!user_mode(regs))
+ return SIGSEGV;
+ /* in_atomic() in user mode is really bad,
+diff --git a/arch/powerpc/platforms/85xx/mpc85xx_cds.c b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
+index 6299a2a..95b9777 100644
+--- a/arch/powerpc/platforms/85xx/mpc85xx_cds.c
++++ b/arch/powerpc/platforms/85xx/mpc85xx_cds.c
+@@ -178,7 +178,7 @@ static irqreturn_t mpc85xx_8259_cascade_action(int irq, void *dev_id)
static struct irqaction mpc85xxcds_8259_irqaction = {
.handler = mpc85xx_8259_cascade_action,
@@ -3966,38 +3012,37 @@
.name = "8259 cascade",
};
#endif /* PPC_I8259 */
-Index: linux-3.0/arch/powerpc/platforms/wsp/opb_pic.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/platforms/wsp/opb_pic.c
-+++ linux-3.0/arch/powerpc/platforms/wsp/opb_pic.c
-@@ -320,7 +320,8 @@ void __init opb_pic_init(void)
- }
+diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c
+index c25a081..a499888 100644
+--- a/arch/powerpc/platforms/iseries/setup.c
++++ b/arch/powerpc/platforms/iseries/setup.c
+@@ -581,9 +581,7 @@ static void iseries_shared_idle(void)
+ if (hvlpevent_is_pending())
+ process_iSeries_events();
- /* Attach opb interrupt handler to new virtual IRQ */
-- rc = request_irq(virq, opb_irq_handler, 0, "OPB LS Cascade", opb);
-+ rc = request_irq(virq, opb_irq_handler, IRQF_NO_THREAD,
-+ "OPB LS Cascade", opb);
- if (rc) {
- printk("opb: request_irq failed: %d\n", rc);
- continue;
-Index: linux-3.0/arch/powerpc/kernel/smp.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/smp.c
-+++ linux-3.0/arch/powerpc/kernel/smp.c
-@@ -170,7 +170,7 @@ int smp_request_message_ipi(int virq, in
- return 1;
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
- #endif
-- err = request_irq(virq, smp_ipi_action[msg], IRQF_DISABLED|IRQF_PERCPU,
-+ err = request_irq(virq, smp_ipi_action[msg], IRQF_NO_THREAD|IRQF_PERCPU,
- smp_ipi_name[msg], 0);
- WARN(err < 0, "unable to request_irq %d for %s (rc %d)\n",
- virq, smp_ipi_name[msg], err);
-Index: linux-3.0/arch/powerpc/platforms/powermac/smp.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/platforms/powermac/smp.c
-+++ linux-3.0/arch/powerpc/platforms/powermac/smp.c
-@@ -200,7 +200,7 @@ static int psurge_secondary_ipi_init(voi
+ }
+
+@@ -610,9 +608,7 @@ static void iseries_dedicated_idle(void)
+
+ ppc64_runlatch_on();
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
+
+diff --git a/arch/powerpc/platforms/powermac/smp.c b/arch/powerpc/platforms/powermac/smp.c
+index db092d7..0c49480 100644
+--- a/arch/powerpc/platforms/powermac/smp.c
++++ b/arch/powerpc/platforms/powermac/smp.c
+@@ -200,7 +200,7 @@ static int psurge_secondary_ipi_init(void)
if (psurge_secondary_virq)
rc = request_irq(psurge_secondary_virq, psurge_ipi_intr,
@@ -4006,7 +3051,7 @@
if (rc)
pr_err("Failed to setup secondary cpu IPI\n");
-@@ -408,7 +408,7 @@ static int __init smp_psurge_kick_cpu(in
+@@ -408,7 +408,7 @@ static int __init smp_psurge_kick_cpu(int nr)
static struct irqaction psurge_irqaction = {
.handler = psurge_ipi_intr,
@@ -4015,10 +3060,125 @@
.name = "primary IPI",
};
-Index: linux-3.0/arch/powerpc/sysdev/xics/xics-common.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/sysdev/xics/xics-common.c
-+++ linux-3.0/arch/powerpc/sysdev/xics/xics-common.c
+diff --git a/arch/powerpc/platforms/wsp/opb_pic.c b/arch/powerpc/platforms/wsp/opb_pic.c
+index be05631..19f353d 100644
+--- a/arch/powerpc/platforms/wsp/opb_pic.c
++++ b/arch/powerpc/platforms/wsp/opb_pic.c
+@@ -320,7 +320,8 @@ void __init opb_pic_init(void)
+ }
+
+ /* Attach opb interrupt handler to new virtual IRQ */
+- rc = request_irq(virq, opb_irq_handler, 0, "OPB LS Cascade", opb);
++ rc = request_irq(virq, opb_irq_handler, IRQF_NO_THREAD,
++ "OPB LS Cascade", opb);
+ if (rc) {
+ printk("opb: request_irq failed: %d\n", rc);
+ continue;
+diff --git a/arch/powerpc/sysdev/uic.c b/arch/powerpc/sysdev/uic.c
+index 984cd20..3330fec 100644
+--- a/arch/powerpc/sysdev/uic.c
++++ b/arch/powerpc/sysdev/uic.c
+@@ -47,7 +47,7 @@ struct uic {
+ int index;
+ int dcrbase;
+
+- spinlock_t lock;
++ raw_spinlock_t lock;
+
+ /* The remapper for this UIC */
+ struct irq_host *irqhost;
+@@ -61,14 +61,14 @@ static void uic_unmask_irq(struct irq_data *d)
+ u32 er, sr;
+
+ sr = 1 << (31-src);
+- spin_lock_irqsave(&uic->lock, flags);
++ raw_spin_lock_irqsave(&uic->lock, flags);
+ /* ack level-triggered interrupts here */
+ if (irqd_is_level_type(d))
+ mtdcr(uic->dcrbase + UIC_SR, sr);
+ er = mfdcr(uic->dcrbase + UIC_ER);
+ er |= sr;
+ mtdcr(uic->dcrbase + UIC_ER, er);
+- spin_unlock_irqrestore(&uic->lock, flags);
++ raw_spin_unlock_irqrestore(&uic->lock, flags);
+ }
+
+ static void uic_mask_irq(struct irq_data *d)
+@@ -78,11 +78,11 @@ static void uic_mask_irq(struct irq_data *d)
+ unsigned long flags;
+ u32 er;
+
+- spin_lock_irqsave(&uic->lock, flags);
++ raw_spin_lock_irqsave(&uic->lock, flags);
+ er = mfdcr(uic->dcrbase + UIC_ER);
+ er &= ~(1 << (31 - src));
+ mtdcr(uic->dcrbase + UIC_ER, er);
+- spin_unlock_irqrestore(&uic->lock, flags);
++ raw_spin_unlock_irqrestore(&uic->lock, flags);
+ }
+
+ static void uic_ack_irq(struct irq_data *d)
+@@ -91,9 +91,9 @@ static void uic_ack_irq(struct irq_data *d)
+ unsigned int src = irqd_to_hwirq(d);
+ unsigned long flags;
+
+- spin_lock_irqsave(&uic->lock, flags);
++ raw_spin_lock_irqsave(&uic->lock, flags);
+ mtdcr(uic->dcrbase + UIC_SR, 1 << (31-src));
+- spin_unlock_irqrestore(&uic->lock, flags);
++ raw_spin_unlock_irqrestore(&uic->lock, flags);
+ }
+
+ static void uic_mask_ack_irq(struct irq_data *d)
+@@ -104,7 +104,7 @@ static void uic_mask_ack_irq(struct irq_data *d)
+ u32 er, sr;
+
+ sr = 1 << (31-src);
+- spin_lock_irqsave(&uic->lock, flags);
++ raw_spin_lock_irqsave(&uic->lock, flags);
+ er = mfdcr(uic->dcrbase + UIC_ER);
+ er &= ~sr;
+ mtdcr(uic->dcrbase + UIC_ER, er);
+@@ -118,7 +118,7 @@ static void uic_mask_ack_irq(struct irq_data *d)
+ */
+ if (!irqd_is_level_type(d))
+ mtdcr(uic->dcrbase + UIC_SR, sr);
+- spin_unlock_irqrestore(&uic->lock, flags);
++ raw_spin_unlock_irqrestore(&uic->lock, flags);
+ }
+
+ static int uic_set_irq_type(struct irq_data *d, unsigned int flow_type)
+@@ -152,7 +152,7 @@ static int uic_set_irq_type(struct irq_data *d, unsigned int flow_type)
+
+ mask = ~(1 << (31 - src));
+
+- spin_lock_irqsave(&uic->lock, flags);
++ raw_spin_lock_irqsave(&uic->lock, flags);
+ tr = mfdcr(uic->dcrbase + UIC_TR);
+ pr = mfdcr(uic->dcrbase + UIC_PR);
+ tr = (tr & mask) | (trigger << (31-src));
+@@ -161,7 +161,7 @@ static int uic_set_irq_type(struct irq_data *d, unsigned int flow_type)
+ mtdcr(uic->dcrbase + UIC_PR, pr);
+ mtdcr(uic->dcrbase + UIC_TR, tr);
+
+- spin_unlock_irqrestore(&uic->lock, flags);
++ raw_spin_unlock_irqrestore(&uic->lock, flags);
+
+ return 0;
+ }
+@@ -254,7 +254,7 @@ static struct uic * __init uic_init_one(struct device_node *node)
+ if (! uic)
+ return NULL; /* FIXME: panic? */
+
+- spin_lock_init(&uic->lock);
++ raw_spin_lock_init(&uic->lock);
+ indexp = of_get_property(node, "cell-index", &len);
+ if (!indexp || (len != sizeof(u32))) {
+ printk(KERN_ERR "uic: Device node %s has missing or invalid "
+diff --git a/arch/powerpc/sysdev/xics/xics-common.c b/arch/powerpc/sysdev/xics/xics-common.c
+index 445c5a0..0f4d153 100644
+--- a/arch/powerpc/sysdev/xics/xics-common.c
++++ b/arch/powerpc/sysdev/xics/xics-common.c
@@ -134,11 +134,11 @@ static void xics_request_ipi(void)
BUG_ON(ipi == NO_IRQ);
@@ -4034,10331 +3194,7441 @@
}
int __init xics_smp_probe(void)
-Index: linux-3.0/arch/powerpc/Kconfig
-===================================================================
---- linux-3.0.orig/arch/powerpc/Kconfig
-+++ linux-3.0/arch/powerpc/Kconfig
-@@ -69,10 +69,11 @@ config LOCKDEP_SUPPORT
-
- config RWSEM_GENERIC_SPINLOCK
- bool
-+ default y if PREEMPT_RT_FULL
-
- config RWSEM_XCHGADD_ALGORITHM
- bool
-- default y
-+ default y if !PREEMPT_RT_FULL
+diff --git a/arch/s390/include/asm/rwsem.h b/arch/s390/include/asm/rwsem.h
+index d0eb465..059b0d2 100644
+--- a/arch/s390/include/asm/rwsem.h
++++ b/arch/s390/include/asm/rwsem.h
+@@ -60,7 +60,7 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
+ {
+ signed long old, new;
- config GENERIC_LOCKBREAK
- bool
-@@ -134,6 +135,7 @@ config PPC
- select GENERIC_IRQ_SHOW_LEVEL
- select HAVE_RCU_TABLE_FREE if SMP
- select HAVE_SYSCALL_TRACEPOINTS
-+ select IRQ_FORCED_THREADING
+@@ -88,7 +88,7 @@ static inline void __down_read(struct rw_semaphore *sem)
+ /*
+ * trylock for reading -- returns 1 if successful, 0 if contention
+ */
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ signed long old, new;
- config EARLY_PRINTK
- bool
-@@ -271,7 +273,7 @@ menu "Kernel options"
+@@ -119,7 +119,8 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ /*
+ * lock for writing
+ */
+-static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
++static inline void
++__down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ signed long old, new, tmp;
- config HIGHMEM
- bool "High memory support"
-- depends on PPC32
-+ depends on PPC32 && !PREEMPT_RT_FULL
+@@ -145,7 +146,7 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
+ rwsem_down_write_failed(sem);
+ }
- source kernel/time/Kconfig
- source kernel/Kconfig.hz
-Index: linux-3.0/net/netfilter/ipvs/ip_vs_ctl.c
-===================================================================
---- linux-3.0.orig/net/netfilter/ipvs/ip_vs_ctl.c
-+++ linux-3.0/net/netfilter/ipvs/ip_vs_ctl.c
-@@ -3679,7 +3679,7 @@ int __net_init __ip_vs_control_init(stru
- int idx;
- struct netns_ipvs *ipvs = net_ipvs(net);
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
+ {
+ __down_write_nested(sem, 0);
+ }
+@@ -153,7 +154,7 @@ static inline void __down_write(struct rw_semaphore *sem)
+ /*
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ signed long old;
-- ipvs->rs_lock = __RW_LOCK_UNLOCKED(ipvs->rs_lock);
-+ rwlock_init(&ipvs->rs_lock);
+@@ -181,7 +182,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ signed long old, new;
- /* Initialize rs_table */
- for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++)
-Index: linux-3.0/kernel/watchdog.c
-===================================================================
---- linux-3.0.orig/kernel/watchdog.c
-+++ linux-3.0/kernel/watchdog.c
-@@ -208,6 +208,8 @@ static struct perf_event_attr wd_hw_attr
- .disabled = 1,
- };
+@@ -210,7 +211,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
+ {
+ signed long old, new, tmp;
-+static DEFINE_RAW_SPINLOCK(watchdog_output_lock);
-+
- /* Callback function for perf event subsystem */
- static void watchdog_overflow_callback(struct perf_event *event, int nmi,
- struct perf_sample_data *data,
-@@ -234,10 +236,19 @@ static void watchdog_overflow_callback(s
- if (__this_cpu_read(hard_watchdog_warn) == true)
- return;
+@@ -240,7 +241,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ signed long old, new, tmp;
-- if (hardlockup_panic)
-+ /*
-+ * If early-printk is enabled then make sure we do not
-+ * lock up in printk() and kill console logging:
-+ */
-+ printk_kill();
-+
-+ if (hardlockup_panic) {
- panic("Watchdog detected hard LOCKUP on cpu %d", this_cpu);
-- else
-+ } else {
-+ raw_spin_lock(&watchdog_output_lock);
- WARN(1, "Watchdog detected hard LOCKUP on cpu %d", this_cpu);
-+ raw_spin_unlock(&watchdog_output_lock);
-+ }
+@@ -269,7 +270,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
+ {
+ signed long old, new;
- __this_cpu_write(hard_watchdog_warn, true);
- return;
-@@ -320,7 +331,7 @@ static enum hrtimer_restart watchdog_tim
+@@ -295,7 +296,8 @@ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
+ /*
+ * implement exchange and add functionality
*/
- static int watchdog(void *unused)
+-static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
++static inline long
++rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
{
-- static struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
-+ struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
- struct hrtimer *hrtimer = &__raw_get_cpu_var(watchdog_hrtimer);
+ signed long old, new;
- sched_setscheduler(current, SCHED_FIFO, ¶m);
-@@ -349,7 +360,8 @@ static int watchdog(void *unused)
- set_current_state(TASK_INTERRUPTIBLE);
+diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c
+index 541a750..a77214f 100644
+--- a/arch/s390/kernel/process.c
++++ b/arch/s390/kernel/process.c
+@@ -94,9 +94,7 @@ void cpu_idle(void)
+ while (!need_resched())
+ default_idle();
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
- __set_current_state(TASK_RUNNING);
--
-+ param.sched_priority = 0;
-+ sched_setscheduler(current, SCHED_NORMAL, ¶m);
- return 0;
}
-@@ -422,6 +434,7 @@ static void watchdog_prepare_cpu(int cpu
- WARN_ON(per_cpu(softlockup_watchdog, cpu));
- hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
- hrtimer->function = watchdog_timer_fn;
-+ hrtimer->irqsafe = 1;
- }
-
- static int watchdog_enable(int cpu)
-Index: linux-3.0/kernel/time/clocksource.c
-===================================================================
---- linux-3.0.orig/kernel/time/clocksource.c
-+++ linux-3.0/kernel/time/clocksource.c
-@@ -186,6 +186,7 @@ static struct timer_list watchdog_timer;
- static DECLARE_WORK(watchdog_work, clocksource_watchdog_work);
- static DEFINE_SPINLOCK(watchdog_lock);
- static int watchdog_running;
-+static atomic_t watchdog_reset_pending;
-
- static int clocksource_watchdog_kthread(void *data);
- static void __clocksource_change_rating(struct clocksource *cs, int rating);
-@@ -247,12 +248,14 @@ static void clocksource_watchdog(unsigne
- struct clocksource *cs;
- cycle_t csnow, wdnow;
- int64_t wd_nsec, cs_nsec;
-- int next_cpu;
-+ int next_cpu, reset_pending;
-
- spin_lock(&watchdog_lock);
- if (!watchdog_running)
+diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
+index fe103e8..608caa1 100644
+--- a/arch/s390/mm/fault.c
++++ b/arch/s390/mm/fault.c
+@@ -295,7 +295,8 @@ static inline int do_exception(struct pt_regs *regs, int access,
+ * user context.
+ */
+ fault = VM_FAULT_BADCONTEXT;
+- if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
++ if (unlikely(!user_space_fault(trans_exc_code) ||
++ !mm || pagefault_disabled()))
goto out;
-+ reset_pending = atomic_read(&watchdog_reset_pending);
-+
- list_for_each_entry(cs, &watchdog_list, wd_list) {
+ address = trans_exc_code & __FAIL_ADDR_MASK;
+@@ -410,7 +411,8 @@ void __kprobes do_asce_exception(struct pt_regs *regs, long pgm_int_code,
+ struct mm_struct *mm = current->mm;
+ struct vm_area_struct *vma;
- /* Clocksource already marked unstable? */
-@@ -268,7 +271,8 @@ static void clocksource_watchdog(unsigne
- local_irq_enable();
+- if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
++ if (unlikely(!user_space_fault(trans_exc_code) ||
++ !mm || pagefault_disabled()))
+ goto no_context;
- /* Clocksource initialized ? */
-- if (!(cs->flags & CLOCK_SOURCE_WATCHDOG)) {
-+ if (!(cs->flags & CLOCK_SOURCE_WATCHDOG) ||
-+ atomic_read(&watchdog_reset_pending)) {
- cs->flags |= CLOCK_SOURCE_WATCHDOG;
- cs->wd_last = wdnow;
- cs->cs_last = csnow;
-@@ -283,8 +287,11 @@ static void clocksource_watchdog(unsigne
- cs->cs_last = csnow;
- cs->wd_last = wdnow;
+ down_read(&mm->mmap_sem);
+diff --git a/arch/score/kernel/process.c b/arch/score/kernel/process.c
+index 25d0803..2707023 100644
+--- a/arch/score/kernel/process.c
++++ b/arch/score/kernel/process.c
+@@ -53,9 +53,7 @@ void __noreturn cpu_idle(void)
+ while (!need_resched())
+ barrier();
-+ if (atomic_read(&watchdog_reset_pending))
-+ continue;
-+
- /* Check the deviation from the watchdog clocksource. */
-- if (abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD) {
-+ if ((abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD)) {
- clocksource_unstable(cs, cs_nsec - wd_nsec);
- continue;
- }
-@@ -303,6 +310,13 @@ static void clocksource_watchdog(unsigne
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
-
- /*
-+ * We only clear the watchdog_reset_pending, when we did a
-+ * full cycle through all clocksources.
-+ */
-+ if (reset_pending)
-+ atomic_dec(&watchdog_reset_pending);
-+
-+ /*
- * Cycle through CPUs to check if the CPUs stay synchronized
- * to each other.
- */
-@@ -344,23 +358,7 @@ static inline void clocksource_reset_wat
-
- static void clocksource_resume_watchdog(void)
- {
-- unsigned long flags;
--
-- /*
-- * We use trylock here to avoid a potential dead lock when
-- * kgdb calls this code after the kernel has been stopped with
-- * watchdog_lock held. When watchdog_lock is held we just
-- * return and accept, that the watchdog might trigger and mark
-- * the monitored clock source (usually TSC) unstable.
-- *
-- * This does not affect the other caller clocksource_resume()
-- * because at this point the kernel is UP, interrupts are
-- * disabled and nothing can hold watchdog_lock.
-- */
-- if (!spin_trylock_irqsave(&watchdog_lock, flags))
-- return;
-- clocksource_reset_watchdog();
-- spin_unlock_irqrestore(&watchdog_lock, flags);
-+ atomic_inc(&watchdog_reset_pending);
}
- static void clocksource_enqueue_watchdog(struct clocksource *cs)
-Index: linux-3.0/kernel/rtmutex-debug.c
-===================================================================
---- linux-3.0.orig/kernel/rtmutex-debug.c
-+++ linux-3.0/kernel/rtmutex-debug.c
-@@ -29,61 +29,6 @@
-
- #include "rtmutex_common.h"
+diff --git a/arch/score/mm/fault.c b/arch/score/mm/fault.c
+index 47b600e..59fccbe 100644
+--- a/arch/score/mm/fault.c
++++ b/arch/score/mm/fault.c
+@@ -72,7 +72,7 @@ asmlinkage void do_page_fault(struct pt_regs *regs, unsigned long write,
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto bad_area_nosemaphore;
--# define TRACE_WARN_ON(x) WARN_ON(x)
--# define TRACE_BUG_ON(x) BUG_ON(x)
--
--# define TRACE_OFF() \
--do { \
-- if (rt_trace_on) { \
-- rt_trace_on = 0; \
-- console_verbose(); \
-- if (raw_spin_is_locked(¤t->pi_lock)) \
-- raw_spin_unlock(¤t->pi_lock); \
-- } \
--} while (0)
--
--# define TRACE_OFF_NOLOCK() \
--do { \
-- if (rt_trace_on) { \
-- rt_trace_on = 0; \
-- console_verbose(); \
-- } \
--} while (0)
--
--# define TRACE_BUG_LOCKED() \
--do { \
-- TRACE_OFF(); \
-- BUG(); \
--} while (0)
--
--# define TRACE_WARN_ON_LOCKED(c) \
--do { \
-- if (unlikely(c)) { \
-- TRACE_OFF(); \
-- WARN_ON(1); \
-- } \
--} while (0)
--
--# define TRACE_BUG_ON_LOCKED(c) \
--do { \
-- if (unlikely(c)) \
-- TRACE_BUG_LOCKED(); \
--} while (0)
--
--#ifdef CONFIG_SMP
--# define SMP_TRACE_BUG_ON_LOCKED(c) TRACE_BUG_ON_LOCKED(c)
--#else
--# define SMP_TRACE_BUG_ON_LOCKED(c) do { } while (0)
--#endif
--
--/*
-- * deadlock detection flag. We turn it off when we detect
-- * the first problem because we dont want to recurse back
-- * into the tracing code when doing error printk or
-- * executing a BUG():
-- */
--static int rt_trace_on = 1;
--
- static void printk_task(struct task_struct *p)
+ down_read(&mm->mmap_sem);
+diff --git a/arch/sh/include/asm/rwsem.h b/arch/sh/include/asm/rwsem.h
+index edab572..3085cdb 100644
+--- a/arch/sh/include/asm/rwsem.h
++++ b/arch/sh/include/asm/rwsem.h
+@@ -22,7 +22,7 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
{
- if (p)
-@@ -111,8 +56,8 @@ static void printk_lock(struct rt_mutex
+ if (atomic_inc_return((atomic_t *)(&sem->count)) > 0)
+ smp_wmb();
+@@ -30,7 +30,7 @@ static inline void __down_read(struct rw_semaphore *sem)
+ rwsem_down_read_failed(sem);
+ }
- void rt_mutex_debug_task_free(struct task_struct *task)
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
{
-- WARN_ON(!plist_head_empty(&task->pi_waiters));
-- WARN_ON(task->pi_blocked_on);
-+ DEBUG_LOCKS_WARN_ON(!plist_head_empty(&task->pi_waiters));
-+ DEBUG_LOCKS_WARN_ON(task->pi_blocked_on);
- }
+ int tmp;
+@@ -47,7 +47,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
/*
-@@ -125,7 +70,7 @@ void debug_rt_mutex_deadlock(int detect,
+ * lock for writing
+ */
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
{
- struct task_struct *task;
+ int tmp;
-- if (!rt_trace_on || detect || !act_waiter)
-+ if (!debug_locks || detect || !act_waiter)
- return;
+@@ -59,7 +59,7 @@ static inline void __down_write(struct rw_semaphore *sem)
+ rwsem_down_write_failed(sem);
+ }
- task = rt_mutex_owner(act_waiter->lock);
-@@ -139,7 +84,7 @@ void debug_rt_mutex_print_deadlock(struc
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
{
- struct task_struct *task;
-
-- if (!waiter->deadlock_lock || !rt_trace_on)
-+ if (!waiter->deadlock_lock || !debug_locks)
- return;
-
- rcu_read_lock();
-@@ -149,7 +94,10 @@ void debug_rt_mutex_print_deadlock(struc
- return;
- }
-
-- TRACE_OFF_NOLOCK();
-+ if (!debug_locks_off()) {
-+ rcu_read_unlock();
-+ return;
-+ }
-
- printk("\n============================================\n");
- printk( "[ BUG: circular locking deadlock detected! ]\n");
-@@ -180,7 +128,6 @@ void debug_rt_mutex_print_deadlock(struc
-
- printk("[ turning off deadlock detection."
- "Please report this trace. ]\n\n");
-- local_irq_disable();
- }
+ int tmp;
- void debug_rt_mutex_lock(struct rt_mutex *lock)
-@@ -189,7 +136,7 @@ void debug_rt_mutex_lock(struct rt_mutex
+@@ -72,7 +72,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ int tmp;
- void debug_rt_mutex_unlock(struct rt_mutex *lock)
+@@ -85,7 +85,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
{
-- TRACE_WARN_ON_LOCKED(rt_mutex_owner(lock) != current);
-+ DEBUG_LOCKS_WARN_ON(rt_mutex_owner(lock) != current);
+ smp_wmb();
+ if (atomic_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+@@ -96,7 +96,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(int delta, struct rw_anon_semaphore *sem)
+ {
+ atomic_add(delta, (atomic_t *)(&sem->count));
}
-
- void
-@@ -199,7 +146,7 @@ debug_rt_mutex_proxy_lock(struct rt_mute
-
- void debug_rt_mutex_proxy_unlock(struct rt_mutex *lock)
+@@ -104,7 +104,7 @@ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
{
-- TRACE_WARN_ON_LOCKED(!rt_mutex_owner(lock));
-+ DEBUG_LOCKS_WARN_ON(!rt_mutex_owner(lock));
+ int tmp;
+
+@@ -114,7 +114,8 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
+ rwsem_downgrade_wake(sem);
}
- void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter)
-@@ -213,8 +160,8 @@ void debug_rt_mutex_init_waiter(struct r
- void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter)
+-static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
++static inline void
++ __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
{
- put_pid(waiter->deadlock_task_pid);
-- TRACE_WARN_ON(!plist_node_empty(&waiter->list_entry));
-- TRACE_WARN_ON(!plist_node_empty(&waiter->pi_list_entry));
-+ DEBUG_LOCKS_WARN_ON(!plist_node_empty(&waiter->list_entry));
-+ DEBUG_LOCKS_WARN_ON(!plist_node_empty(&waiter->pi_list_entry));
- memset(waiter, 0x22, sizeof(*waiter));
+ __down_write(sem);
}
+@@ -122,7 +123,7 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
+ /*
+ * implement exchange and add functionality
+ */
+-static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
++static inline int rwsem_atomic_update(int delta, struct rw_anon_semaphore *sem)
+ {
+ smp_mb();
+ return atomic_add_return(delta, (atomic_t *)(&sem->count));
+diff --git a/arch/sh/kernel/idle.c b/arch/sh/kernel/idle.c
+index 425d604..179e92e 100644
+--- a/arch/sh/kernel/idle.c
++++ b/arch/sh/kernel/idle.c
+@@ -110,9 +110,7 @@ void cpu_idle(void)
+ }
-Index: linux-3.0/include/linux/kprobes.h
-===================================================================
---- linux-3.0.orig/include/linux/kprobes.h
-+++ linux-3.0/include/linux/kprobes.h
-@@ -181,7 +181,7 @@ struct kretprobe {
- int nmissed;
- size_t data_size;
- struct hlist_head free_instances;
-- spinlock_t lock;
-+ raw_spinlock_t lock;
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
+
+diff --git a/arch/sh/kernel/irq.c b/arch/sh/kernel/irq.c
+index a3ee919..9127bc0 100644
+--- a/arch/sh/kernel/irq.c
++++ b/arch/sh/kernel/irq.c
+@@ -149,6 +149,7 @@ void irq_ctx_exit(int cpu)
+ hardirq_ctx[cpu] = NULL;
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ asmlinkage void do_softirq(void)
+ {
+ unsigned long flags;
+@@ -191,6 +192,7 @@ asmlinkage void do_softirq(void)
+
+ local_irq_restore(flags);
+ }
++#endif
+ #else
+ static inline void handle_one_irq(unsigned int irq)
+ {
+diff --git a/arch/sh/kernel/sh_bios.c b/arch/sh/kernel/sh_bios.c
+index 47475cc..a5b51b9 100644
+--- a/arch/sh/kernel/sh_bios.c
++++ b/arch/sh/kernel/sh_bios.c
+@@ -144,8 +144,6 @@ static struct console bios_console = {
+ .index = -1,
};
- struct kretprobe_instance {
-Index: linux-3.0/kernel/kprobes.c
-===================================================================
---- linux-3.0.orig/kernel/kprobes.c
-+++ linux-3.0/kernel/kprobes.c
-@@ -78,10 +78,10 @@ static bool kprobes_all_disarmed;
- static DEFINE_MUTEX(kprobe_mutex);
- static DEFINE_PER_CPU(struct kprobe *, kprobe_instance) = NULL;
- static struct {
-- spinlock_t lock ____cacheline_aligned_in_smp;
-+ raw_spinlock_t lock ____cacheline_aligned_in_smp;
- } kretprobe_table_locks[KPROBE_TABLE_SIZE];
+-static struct console *early_console;
+-
+ static int __init setup_early_printk(char *buf)
+ {
+ int keep_early = 0;
+diff --git a/arch/sh/mm/fault_32.c b/arch/sh/mm/fault_32.c
+index d4c34d7..8bd2407 100644
+--- a/arch/sh/mm/fault_32.c
++++ b/arch/sh/mm/fault_32.c
+@@ -166,7 +166,7 @@ asmlinkage void __kprobes do_page_fault(struct pt_regs *regs,
+ * If we're in an interrupt, have no user context or are running
+ * in an atomic region then we must not take the fault:
+ */
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto no_context;
--static spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
-+static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
+ down_read(&mm->mmap_sem);
+diff --git a/arch/sparc/include/asm/rwsem.h b/arch/sparc/include/asm/rwsem.h
+index 069bf4d..235fa89 100644
+--- a/arch/sparc/include/asm/rwsem.h
++++ b/arch/sparc/include/asm/rwsem.h
+@@ -23,13 +23,13 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
{
- return &(kretprobe_table_locks[hash].lock);
+ if (unlikely(atomic64_inc_return((atomic64_t *)(&sem->count)) <= 0L))
+ rwsem_down_read_failed(sem);
}
-@@ -1013,9 +1013,9 @@ void __kprobes recycle_rp_inst(struct kr
- hlist_del(&ri->hlist);
- INIT_HLIST_NODE(&ri->hlist);
- if (likely(rp)) {
-- spin_lock(&rp->lock);
-+ raw_spin_lock(&rp->lock);
- hlist_add_head(&ri->hlist, &rp->free_instances);
-- spin_unlock(&rp->lock);
-+ raw_spin_unlock(&rp->lock);
- } else
- /* Unregistering */
- hlist_add_head(&ri->hlist, head);
-@@ -1026,19 +1026,19 @@ void __kprobes kretprobe_hash_lock(struc
- __acquires(hlist_lock)
+
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
{
- unsigned long hash = hash_ptr(tsk, KPROBE_HASH_BITS);
-- spinlock_t *hlist_lock;
-+ raw_spinlock_t *hlist_lock;
+ long tmp;
- *head = &kretprobe_inst_table[hash];
- hlist_lock = kretprobe_table_lock_ptr(hash);
-- spin_lock_irqsave(hlist_lock, *flags);
-+ raw_spin_lock_irqsave(hlist_lock, *flags);
+@@ -45,7 +45,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ /*
+ * lock for writing
+ */
+-static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
++static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ long tmp;
+
+@@ -55,12 +55,12 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
+ rwsem_down_write_failed(sem);
}
- static void __kprobes kretprobe_table_lock(unsigned long hash,
- unsigned long *flags)
- __acquires(hlist_lock)
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
{
-- spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
-- spin_lock_irqsave(hlist_lock, *flags);
-+ raw_spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
-+ raw_spin_lock_irqsave(hlist_lock, *flags);
+ __down_write_nested(sem, 0);
}
- void __kprobes kretprobe_hash_unlock(struct task_struct *tsk,
-@@ -1046,18 +1046,18 @@ void __kprobes kretprobe_hash_unlock(str
- __releases(hlist_lock)
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
{
- unsigned long hash = hash_ptr(tsk, KPROBE_HASH_BITS);
-- spinlock_t *hlist_lock;
-+ raw_spinlock_t *hlist_lock;
+ long tmp;
- hlist_lock = kretprobe_table_lock_ptr(hash);
-- spin_unlock_irqrestore(hlist_lock, *flags);
-+ raw_spin_unlock_irqrestore(hlist_lock, *flags);
- }
+@@ -72,7 +72,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
- static void __kprobes kretprobe_table_unlock(unsigned long hash,
- unsigned long *flags)
- __releases(hlist_lock)
+@@ -84,7 +84,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
{
-- spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
-- spin_unlock_irqrestore(hlist_lock, *flags);
-+ raw_spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
-+ raw_spin_unlock_irqrestore(hlist_lock, *flags);
+ if (unlikely(atomic64_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+ (atomic64_t *)(&sem->count)) < 0L))
+@@ -94,7 +94,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
+ {
+ atomic64_add(delta, (atomic64_t *)(&sem->count));
}
+@@ -102,7 +102,7 @@ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
+@@ -114,7 +114,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
/*
-@@ -1650,12 +1650,12 @@ static int __kprobes pre_handler_kretpro
+ * implement exchange and add functionality
+ */
+-static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
++static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
+ {
+ return atomic64_add_return(delta, (atomic64_t *)(&sem->count));
+ }
+diff --git a/arch/sparc/kernel/irq_64.c b/arch/sparc/kernel/irq_64.c
+index 4e78862..202a831 100644
+--- a/arch/sparc/kernel/irq_64.c
++++ b/arch/sparc/kernel/irq_64.c
+@@ -700,6 +700,7 @@ void __irq_entry handler_irq(int pil, struct pt_regs *regs)
+ set_irq_regs(old_regs);
+ }
- /*TODO: consider to only swap the RA after the last pre_handler fired */
- hash = hash_ptr(current, KPROBE_HASH_BITS);
-- spin_lock_irqsave(&rp->lock, flags);
-+ raw_spin_lock_irqsave(&rp->lock, flags);
- if (!hlist_empty(&rp->free_instances)) {
- ri = hlist_entry(rp->free_instances.first,
- struct kretprobe_instance, hlist);
- hlist_del(&ri->hlist);
-- spin_unlock_irqrestore(&rp->lock, flags);
-+ raw_spin_unlock_irqrestore(&rp->lock, flags);
++#ifndef CONFIG_PREEMPT_RT_FULL
+ void do_softirq(void)
+ {
+ unsigned long flags;
+@@ -725,6 +726,7 @@ void do_softirq(void)
- ri->rp = rp;
- ri->task = current;
-@@ -1672,7 +1672,7 @@ static int __kprobes pre_handler_kretpro
- kretprobe_table_unlock(hash, &flags);
- } else {
- rp->nmissed++;
-- spin_unlock_irqrestore(&rp->lock, flags);
-+ raw_spin_unlock_irqrestore(&rp->lock, flags);
- }
- return 0;
+ local_irq_restore(flags);
}
-@@ -1708,7 +1708,7 @@ int __kprobes register_kretprobe(struct
- rp->maxactive = num_possible_cpus();
- #endif
++#endif
+
+ #ifdef CONFIG_HOTPLUG_CPU
+ void fixup_irqs(void)
+diff --git a/arch/sparc/kernel/process_32.c b/arch/sparc/kernel/process_32.c
+index c8cc461..e1b53c1 100644
+--- a/arch/sparc/kernel/process_32.c
++++ b/arch/sparc/kernel/process_32.c
+@@ -113,9 +113,7 @@ void cpu_idle(void)
+ while (!need_resched())
+ cpu_relax();
+ }
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ check_pgt_cache();
}
-- spin_lock_init(&rp->lock);
-+ raw_spin_lock_init(&rp->lock);
- INIT_HLIST_HEAD(&rp->free_instances);
- for (i = 0; i < rp->maxactive; i++) {
- inst = kmalloc(sizeof(struct kretprobe_instance) +
-@@ -1946,7 +1946,7 @@ static int __init init_kprobes(void)
- for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
- INIT_HLIST_HEAD(&kprobe_table[i]);
- INIT_HLIST_HEAD(&kretprobe_inst_table[i]);
-- spin_lock_init(&(kretprobe_table_locks[i].lock));
-+ raw_spin_lock_init(&(kretprobe_table_locks[i].lock));
+ }
+@@ -138,9 +136,7 @@ void cpu_idle(void)
+ while (!need_resched())
+ cpu_relax();
+ }
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ check_pgt_cache();
}
+ }
+diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c
+index c158a95..2b98b84 100644
+--- a/arch/sparc/kernel/process_64.c
++++ b/arch/sparc/kernel/process_64.c
+@@ -102,15 +102,13 @@ void cpu_idle(void)
- /*
-Index: linux-3.0/include/linux/percpu_counter.h
-===================================================================
---- linux-3.0.orig/include/linux/percpu_counter.h
-+++ linux-3.0/include/linux/percpu_counter.h
-@@ -16,7 +16,7 @@
- #ifdef CONFIG_SMP
+ tick_nohz_restart_sched_tick();
- struct percpu_counter {
-- spinlock_t lock;
-+ raw_spinlock_t lock;
- s64 count;
+- preempt_enable_no_resched();
+-
#ifdef CONFIG_HOTPLUG_CPU
- struct list_head list; /* All percpu_counters are on a list */
-Index: linux-3.0/lib/percpu_counter.c
-===================================================================
---- linux-3.0.orig/lib/percpu_counter.c
-+++ linux-3.0/lib/percpu_counter.c
-@@ -59,13 +59,13 @@ void percpu_counter_set(struct percpu_co
- {
- int cpu;
-
-- spin_lock(&fbc->lock);
-+ raw_spin_lock(&fbc->lock);
- for_each_possible_cpu(cpu) {
- s32 *pcount = per_cpu_ptr(fbc->counters, cpu);
- *pcount = 0;
+- if (cpu_is_offline(cpu))
++ if (cpu_is_offline(cpu)) {
++ __preempt_enable_no_resched();
+ cpu_play_dead();
++ }
+ #endif
+-
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
}
- fbc->count = amount;
-- spin_unlock(&fbc->lock);
-+ raw_spin_unlock(&fbc->lock);
}
- EXPORT_SYMBOL(percpu_counter_set);
-@@ -76,10 +76,10 @@ void __percpu_counter_add(struct percpu_
- preempt_disable();
- count = __this_cpu_read(*fbc->counters) + amount;
- if (count >= batch || count <= -batch) {
-- spin_lock(&fbc->lock);
-+ raw_spin_lock(&fbc->lock);
- fbc->count += count;
- __this_cpu_write(*fbc->counters, 0);
-- spin_unlock(&fbc->lock);
-+ raw_spin_unlock(&fbc->lock);
- } else {
- __this_cpu_write(*fbc->counters, count);
- }
-@@ -96,13 +96,13 @@ s64 __percpu_counter_sum(struct percpu_c
- s64 ret;
- int cpu;
+diff --git a/arch/sparc/kernel/prom_common.c b/arch/sparc/kernel/prom_common.c
+index ed25834..ec3747f 100644
+--- a/arch/sparc/kernel/prom_common.c
++++ b/arch/sparc/kernel/prom_common.c
+@@ -67,7 +67,7 @@ int of_set_property(struct device_node *dp, const char *name, void *val, int len
+ err = -ENODEV;
-- spin_lock(&fbc->lock);
-+ raw_spin_lock(&fbc->lock);
- ret = fbc->count;
- for_each_online_cpu(cpu) {
- s32 *pcount = per_cpu_ptr(fbc->counters, cpu);
- ret += *pcount;
+ mutex_lock(&of_set_property_mutex);
+- write_lock(&devtree_lock);
++ raw_spin_lock(&devtree_lock);
+ prevp = &dp->properties;
+ while (*prevp) {
+ struct property *prop = *prevp;
+@@ -94,7 +94,7 @@ int of_set_property(struct device_node *dp, const char *name, void *val, int len
+ }
+ prevp = &(*prevp)->next;
}
-- spin_unlock(&fbc->lock);
-+ raw_spin_unlock(&fbc->lock);
- return ret;
- }
- EXPORT_SYMBOL(__percpu_counter_sum);
-@@ -110,7 +110,7 @@ EXPORT_SYMBOL(__percpu_counter_sum);
- int __percpu_counter_init(struct percpu_counter *fbc, s64 amount,
- struct lock_class_key *key)
- {
-- spin_lock_init(&fbc->lock);
-+ raw_spin_lock_init(&fbc->lock);
- lockdep_set_class(&fbc->lock, key);
- fbc->count = amount;
- fbc->counters = alloc_percpu(s32);
-@@ -173,11 +173,11 @@ static int __cpuinit percpu_counter_hotc
- s32 *pcount;
- unsigned long flags;
+- write_unlock(&devtree_lock);
++ raw_spin_unlock(&devtree_lock);
+ mutex_unlock(&of_set_property_mutex);
-- spin_lock_irqsave(&fbc->lock, flags);
-+ raw_spin_lock_irqsave(&fbc->lock, flags);
- pcount = per_cpu_ptr(fbc->counters, cpu);
- fbc->count += *pcount;
- *pcount = 0;
-- spin_unlock_irqrestore(&fbc->lock, flags);
-+ raw_spin_unlock_irqrestore(&fbc->lock, flags);
- }
- mutex_unlock(&percpu_counters_lock);
- #endif
-Index: linux-3.0/kernel/cgroup.c
-===================================================================
---- linux-3.0.orig/kernel/cgroup.c
-+++ linux-3.0/kernel/cgroup.c
-@@ -263,7 +263,7 @@ list_for_each_entry(_root, &roots, root_
- /* the list of cgroups eligible for automatic release. Protected by
- * release_list_lock */
- static LIST_HEAD(release_list);
--static DEFINE_SPINLOCK(release_list_lock);
-+static DEFINE_RAW_SPINLOCK(release_list_lock);
- static void cgroup_release_agent(struct work_struct *work);
- static DECLARE_WORK(release_agent_work, cgroup_release_agent);
- static void check_for_release(struct cgroup *cgrp);
-@@ -4010,11 +4010,11 @@ again:
- finish_wait(&cgroup_rmdir_waitq, &wait);
- clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags);
+ /* XXX Upate procfs if necessary... */
+diff --git a/arch/sparc/kernel/setup_32.c b/arch/sparc/kernel/setup_32.c
+index d26e1f6..441a441 100644
+--- a/arch/sparc/kernel/setup_32.c
++++ b/arch/sparc/kernel/setup_32.c
+@@ -220,6 +220,7 @@ void __init setup_arch(char **cmdline_p)
-- spin_lock(&release_list_lock);
-+ raw_spin_lock(&release_list_lock);
- set_bit(CGRP_REMOVED, &cgrp->flags);
- if (!list_empty(&cgrp->release_list))
- list_del_init(&cgrp->release_list);
-- spin_unlock(&release_list_lock);
-+ raw_spin_unlock(&release_list_lock);
+ boot_flags_init(*cmdline_p);
- cgroup_lock_hierarchy(cgrp->root);
- /* delete this cgroup from parent->children */
-@@ -4667,13 +4667,13 @@ static void check_for_release(struct cgr
- * already queued for a userspace notification, queue
- * it now */
- int need_schedule_work = 0;
-- spin_lock(&release_list_lock);
-+ raw_spin_lock(&release_list_lock);
- if (!cgroup_is_removed(cgrp) &&
- list_empty(&cgrp->release_list)) {
- list_add(&cgrp->release_list, &release_list);
- need_schedule_work = 1;
- }
-- spin_unlock(&release_list_lock);
-+ raw_spin_unlock(&release_list_lock);
- if (need_schedule_work)
- schedule_work(&release_agent_work);
- }
-@@ -4725,7 +4725,7 @@ static void cgroup_release_agent(struct
- {
- BUG_ON(work != &release_agent_work);
- mutex_lock(&cgroup_mutex);
-- spin_lock(&release_list_lock);
-+ raw_spin_lock(&release_list_lock);
- while (!list_empty(&release_list)) {
- char *argv[3], *envp[3];
- int i;
-@@ -4734,7 +4734,7 @@ static void cgroup_release_agent(struct
- struct cgroup,
- release_list);
- list_del_init(&cgrp->release_list);
-- spin_unlock(&release_list_lock);
-+ raw_spin_unlock(&release_list_lock);
- pathbuf = kmalloc(PAGE_SIZE, GFP_KERNEL);
- if (!pathbuf)
- goto continue_free;
-@@ -4764,9 +4764,9 @@ static void cgroup_release_agent(struct
- continue_free:
- kfree(pathbuf);
- kfree(agentbuf);
-- spin_lock(&release_list_lock);
-+ raw_spin_lock(&release_list_lock);
- }
-- spin_unlock(&release_list_lock);
-+ raw_spin_unlock(&release_list_lock);
- mutex_unlock(&cgroup_mutex);
++ early_console = &prom_early_console;
+ register_console(&prom_early_console);
+
+ /* Set sparc_cpu_model */
+diff --git a/arch/sparc/kernel/setup_64.c b/arch/sparc/kernel/setup_64.c
+index 3c5bb78..409cb28 100644
+--- a/arch/sparc/kernel/setup_64.c
++++ b/arch/sparc/kernel/setup_64.c
+@@ -469,6 +469,12 @@ static void __init init_sparc64_elf_hwcap(void)
+ popc_patch();
}
-Index: linux-3.0/include/linux/proportions.h
-===================================================================
---- linux-3.0.orig/include/linux/proportions.h
-+++ linux-3.0/include/linux/proportions.h
-@@ -58,7 +58,7 @@ struct prop_local_percpu {
++static inline void register_prom_console(void)
++{
++ early_console = &prom_early_console;
++ register_console(&prom_early_console);
++}
++
+ void __init setup_arch(char **cmdline_p)
+ {
+ /* Initialize PROM console and command line. */
+@@ -480,7 +486,7 @@ void __init setup_arch(char **cmdline_p)
+ #ifdef CONFIG_EARLYFB
+ if (btext_find_display())
+ #endif
+- register_console(&prom_early_console);
++ register_prom_console();
+
+ if (tlb_type == hypervisor)
+ printk("ARCH: SUN4V\n");
+diff --git a/arch/sparc/mm/fault_32.c b/arch/sparc/mm/fault_32.c
+index 7543ddb..d03ad11 100644
+--- a/arch/sparc/mm/fault_32.c
++++ b/arch/sparc/mm/fault_32.c
+@@ -248,8 +248,8 @@ asmlinkage void do_sparc_fault(struct pt_regs *regs, int text_fault, int write,
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
*/
- int shift;
- unsigned long period;
-- spinlock_t lock; /* protect the snapshot state */
-+ raw_spinlock_t lock; /* protect the snapshot state */
- };
+- if (in_atomic() || !mm)
+- goto no_context;
++ if (!mm || pagefault_disabled())
++ goto no_context;
- int prop_local_init_percpu(struct prop_local_percpu *pl);
-@@ -106,11 +106,11 @@ struct prop_local_single {
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
+
+diff --git a/arch/sparc/mm/fault_64.c b/arch/sparc/mm/fault_64.c
+index f92ce56..2a47747 100644
+--- a/arch/sparc/mm/fault_64.c
++++ b/arch/sparc/mm/fault_64.c
+@@ -322,7 +322,7 @@ asmlinkage void __kprobes do_sparc64_fault(struct pt_regs *regs)
+ * If we're in an interrupt or have no user
+ * context, we must not take the fault..
*/
- unsigned long period;
- int shift;
-- spinlock_t lock; /* protect the snapshot state */
-+ raw_spinlock_t lock; /* protect the snapshot state */
- };
+- if (in_atomic() || !mm)
++ if (!mm || pagefault_disabled())
+ goto intr_or_no_mm;
- #define INIT_PROP_LOCAL_SINGLE(name) \
--{ .lock = __SPIN_LOCK_UNLOCKED(name.lock), \
-+{ .lock = __RAW_SPIN_LOCK_UNLOCKED(name.lock), \
- }
+ perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
+diff --git a/arch/tile/kernel/early_printk.c b/arch/tile/kernel/early_printk.c
+index 493a0e6..ba2ac00 100644
+--- a/arch/tile/kernel/early_printk.c
++++ b/arch/tile/kernel/early_printk.c
+@@ -32,25 +32,8 @@ static struct console early_hv_console = {
+ };
- int prop_local_init_single(struct prop_local_single *pl);
-Index: linux-3.0/lib/proportions.c
-===================================================================
---- linux-3.0.orig/lib/proportions.c
-+++ linux-3.0/lib/proportions.c
-@@ -190,7 +190,7 @@ prop_adjust_shift(int *pl_shift, unsigne
+ /* Direct interface for emergencies */
+-static struct console *early_console = &early_hv_console;
+-static int early_console_initialized;
+ static int early_console_complete;
- int prop_local_init_percpu(struct prop_local_percpu *pl)
+-static void early_vprintk(const char *fmt, va_list ap)
+-{
+- char buf[512];
+- int n = vscnprintf(buf, sizeof(buf), fmt, ap);
+- early_console->write(early_console, buf, n);
+-}
+-
+-void early_printk(const char *fmt, ...)
+-{
+- va_list ap;
+- va_start(ap, fmt);
+- early_vprintk(fmt, ap);
+- va_end(ap);
+-}
+-
+ void early_panic(const char *fmt, ...)
{
-- spin_lock_init(&pl->lock);
-+ raw_spin_lock_init(&pl->lock);
- pl->shift = 0;
- pl->period = 0;
- return percpu_counter_init(&pl->events, 0);
-@@ -226,7 +226,7 @@ void prop_norm_percpu(struct prop_global
- if (pl->period == global_period)
- return;
-
-- spin_lock_irqsave(&pl->lock, flags);
-+ raw_spin_lock_irqsave(&pl->lock, flags);
- prop_adjust_shift(&pl->shift, &pl->period, pg->shift);
+ va_list ap;
+@@ -68,14 +51,13 @@ static int __initdata keep_early;
- /*
-@@ -247,7 +247,7 @@ void prop_norm_percpu(struct prop_global
- percpu_counter_set(&pl->events, 0);
+ static int __init setup_early_printk(char *str)
+ {
+- if (early_console_initialized)
++ if (early_console)
+ return 1;
- pl->period = global_period;
-- spin_unlock_irqrestore(&pl->lock, flags);
-+ raw_spin_unlock_irqrestore(&pl->lock, flags);
- }
+ if (str != NULL && strncmp(str, "keep", 4) == 0)
+ keep_early = 1;
- /*
-@@ -324,7 +324,7 @@ void prop_fraction_percpu(struct prop_de
+ early_console = &early_hv_console;
+- early_console_initialized = 1;
+ register_console(early_console);
- int prop_local_init_single(struct prop_local_single *pl)
+ return 0;
+@@ -84,12 +66,12 @@ static int __init setup_early_printk(char *str)
+ void __init disable_early_printk(void)
{
-- spin_lock_init(&pl->lock);
-+ raw_spin_lock_init(&pl->lock);
- pl->shift = 0;
- pl->period = 0;
- pl->events = 0;
-@@ -356,7 +356,7 @@ void prop_norm_single(struct prop_global
- if (pl->period == global_period)
+ early_console_complete = 1;
+- if (!early_console_initialized || !early_console)
++ if (!early_console)
return;
+ if (!keep_early) {
+ early_printk("disabling early console\n");
+ unregister_console(early_console);
+- early_console_initialized = 0;
++ early_console = NULL;
+ } else {
+ early_printk("keeping early console\n");
+ }
+@@ -97,7 +79,7 @@ void __init disable_early_printk(void)
-- spin_lock_irqsave(&pl->lock, flags);
-+ raw_spin_lock_irqsave(&pl->lock, flags);
- prop_adjust_shift(&pl->shift, &pl->period, pg->shift);
- /*
- * For each missed period, we half the local counter.
-@@ -367,7 +367,7 @@ void prop_norm_single(struct prop_global
- else
- pl->events = 0;
- pl->period = global_period;
-- spin_unlock_irqrestore(&pl->lock, flags);
-+ raw_spin_unlock_irqrestore(&pl->lock, flags);
+ void warn_early_printk(void)
+ {
+- if (early_console_complete || early_console_initialized)
++ if (early_console_complete || early_console)
+ return;
+ early_printk("\
+ Machine shutting down before console output is fully initialized.\n\
+diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c
+index 9c45d8b..5a8b631 100644
+--- a/arch/tile/kernel/process.c
++++ b/arch/tile/kernel/process.c
+@@ -106,9 +106,7 @@ void cpu_idle(void)
+ current_thread_info()->status |= TS_POLLING;
+ }
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
}
- /*
-Index: linux-3.0/kernel/trace/ring_buffer.c
-===================================================================
---- linux-3.0.orig/kernel/trace/ring_buffer.c
-+++ linux-3.0/kernel/trace/ring_buffer.c
-@@ -1040,6 +1040,44 @@ static int rb_allocate_pages(struct ring
- return -ENOMEM;
+diff --git a/arch/tile/mm/fault.c b/arch/tile/mm/fault.c
+index 25b7b90..1b6fa51 100644
+--- a/arch/tile/mm/fault.c
++++ b/arch/tile/mm/fault.c
+@@ -346,7 +346,7 @@ static int handle_page_fault(struct pt_regs *regs,
+ * If we're in an interrupt, have no user context or are running in an
+ * atomic region then we must not take the fault.
+ */
+- if (in_atomic() || !mm) {
++ if (!mm || pagefault_disabled()) {
+ vma = NULL; /* happy compiler */
+ goto bad_area_nosemaphore;
+ }
+diff --git a/arch/um/kernel/early_printk.c b/arch/um/kernel/early_printk.c
+index ec649bf..183060f 100644
+--- a/arch/um/kernel/early_printk.c
++++ b/arch/um/kernel/early_printk.c
+@@ -16,7 +16,7 @@ static void early_console_write(struct console *con, const char *s, unsigned int
+ um_early_printk(s, n);
}
-+static inline int ok_to_lock(void)
-+{
-+ if (in_nmi())
-+ return 0;
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ if (in_atomic())
-+ return 0;
-+#endif
-+ return 1;
-+}
-+
-+static int
-+read_buffer_lock(struct ring_buffer_per_cpu *cpu_buffer,
-+ unsigned long *flags)
-+{
-+ /*
-+ * If an NMI die dumps out the content of the ring buffer
-+ * do not grab locks. We also permanently disable the ring
-+ * buffer too. A one time deal is all you get from reading
-+ * the ring buffer from an NMI.
-+ */
-+ if (!ok_to_lock()) {
-+ if (spin_trylock_irqsave(&cpu_buffer->reader_lock, *flags))
-+ return 1;
-+ tracing_off_permanent();
-+ return 0;
-+ }
-+ spin_lock_irqsave(&cpu_buffer->reader_lock, *flags);
-+ return 1;
-+}
-+
-+static void
-+read_buffer_unlock(struct ring_buffer_per_cpu *cpu_buffer,
-+ unsigned long flags, int locked)
-+{
-+ if (locked)
-+ spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+}
- static struct ring_buffer_per_cpu *
- rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu)
- {
-@@ -1250,9 +1288,11 @@ rb_remove_pages(struct ring_buffer_per_c
- {
- struct buffer_page *bpage;
- struct list_head *p;
-+ unsigned long flags;
- unsigned i;
-+ int locked;
+-static struct console early_console = {
++static struct console early_console_dev = {
+ .name = "earlycon",
+ .write = early_console_write,
+ .flags = CON_BOOT,
+@@ -25,8 +25,10 @@ static struct console early_console = {
-- spin_lock_irq(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- rb_head_page_deactivate(cpu_buffer);
+ static int __init setup_early_printk(char *buf)
+ {
+- register_console(&early_console);
+-
++ if (!early_console) {
++ early_console = &early_console_dev;
++ register_console(&early_console_dev);
++ }
+ return 0;
+ }
- for (i = 0; i < nr_pages; i++) {
-@@ -1270,7 +1310,7 @@ rb_remove_pages(struct ring_buffer_per_c
- rb_check_pages(cpu_buffer);
+diff --git a/arch/um/kernel/trap.c b/arch/um/kernel/trap.c
+index 8c7b882..a4eaf40 100644
+--- a/arch/um/kernel/trap.c
++++ b/arch/um/kernel/trap.c
+@@ -37,7 +37,7 @@ int handle_page_fault(unsigned long address, unsigned long ip,
+ * If the fault was during atomic operation, don't take the fault, just
+ * fail.
+ */
+- if (in_atomic())
++ if (!mm || pagefault_disabled())
+ goto out_nosemaphore;
- out:
-- spin_unlock_irq(&cpu_buffer->reader_lock);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
- }
+ down_read(&mm->mmap_sem);
+diff --git a/arch/unicore32/kernel/early_printk.c b/arch/unicore32/kernel/early_printk.c
+index 3922255..9be0d5d 100644
+--- a/arch/unicore32/kernel/early_printk.c
++++ b/arch/unicore32/kernel/early_printk.c
+@@ -33,21 +33,17 @@ static struct console early_ocd_console = {
+ .index = -1,
+ };
- static void
-@@ -1279,9 +1319,11 @@ rb_insert_pages(struct ring_buffer_per_c
+-/* Direct interface for emergencies */
+-static struct console *early_console = &early_ocd_console;
+-
+-static int __initdata keep_early;
+-
+ static int __init setup_early_printk(char *buf)
{
- struct buffer_page *bpage;
- struct list_head *p;
-+ unsigned long flags;
- unsigned i;
-+ int locked;
+- if (!buf)
++ int keep_early;
++
++ if (!buf || early_console)
+ return 0;
-- spin_lock_irq(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- rb_head_page_deactivate(cpu_buffer);
+ if (strstr(buf, "keep"))
+ keep_early = 1;
- for (i = 0; i < nr_pages; i++) {
-@@ -1296,7 +1338,7 @@ rb_insert_pages(struct ring_buffer_per_c
- rb_check_pages(cpu_buffer);
+- if (!strncmp(buf, "ocd", 3))
+- early_console = &early_ocd_console;
++ early_console = &early_ocd_console;
- out:
-- spin_unlock_irq(&cpu_buffer->reader_lock);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
- }
+ if (keep_early)
+ early_console->flags &= ~CON_BOOT;
+diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
+index 37357a5..fe4ccf6 100644
+--- a/arch/x86/Kconfig
++++ b/arch/x86/Kconfig
+@@ -152,10 +152,10 @@ config ARCH_MAY_HAVE_PC_FDC
+ def_bool ISA_DMA_API
- /**
-@@ -2784,15 +2826,16 @@ void ring_buffer_iter_reset(struct ring_
- {
- struct ring_buffer_per_cpu *cpu_buffer;
- unsigned long flags;
-+ int locked;
+ config RWSEM_GENERIC_SPINLOCK
+- def_bool !X86_XADD
++ def_bool !X86_XADD || PREEMPT_RT_FULL
- if (!iter)
- return;
+ config RWSEM_XCHGADD_ALGORITHM
+- def_bool X86_XADD
++ def_bool X86_XADD && !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL
- cpu_buffer = iter->cpu_buffer;
+ config ARCH_HAS_CPU_IDLE_WAIT
+ def_bool y
+diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c
+index feee8ff..81ec77c 100644
+--- a/arch/x86/crypto/aesni-intel_glue.c
++++ b/arch/x86/crypto/aesni-intel_glue.c
+@@ -288,14 +288,14 @@ static int ecb_encrypt(struct blkcipher_desc *desc,
+ err = blkcipher_walk_virt(desc, &walk);
+ desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+
+- kernel_fpu_begin();
+ while ((nbytes = walk.nbytes)) {
++ kernel_fpu_begin();
+ aesni_ecb_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr,
+- nbytes & AES_BLOCK_MASK);
++ nbytes & AES_BLOCK_MASK);
++ kernel_fpu_end();
+ nbytes &= AES_BLOCK_SIZE - 1;
+ err = blkcipher_walk_done(desc, &walk, nbytes);
+ }
+- kernel_fpu_end();
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- rb_iter_reset(iter);
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ return err;
}
- EXPORT_SYMBOL_GPL(ring_buffer_iter_reset);
+@@ -312,14 +312,14 @@ static int ecb_decrypt(struct blkcipher_desc *desc,
+ err = blkcipher_walk_virt(desc, &walk);
+ desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+
+- kernel_fpu_begin();
+ while ((nbytes = walk.nbytes)) {
++ kernel_fpu_begin();
+ aesni_ecb_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr,
+ nbytes & AES_BLOCK_MASK);
++ kernel_fpu_end();
+ nbytes &= AES_BLOCK_SIZE - 1;
+ err = blkcipher_walk_done(desc, &walk, nbytes);
+ }
+- kernel_fpu_end();
-@@ -3210,21 +3253,6 @@ rb_iter_peek(struct ring_buffer_iter *it
+ return err;
}
- EXPORT_SYMBOL_GPL(ring_buffer_iter_peek);
-
--static inline int rb_ok_to_lock(void)
--{
-- /*
-- * If an NMI die dumps out the content of the ring buffer
-- * do not grab locks. We also permanently disable the ring
-- * buffer too. A one time deal is all you get from reading
-- * the ring buffer from an NMI.
-- */
-- if (likely(!in_nmi()))
-- return 1;
--
-- tracing_off_permanent();
-- return 0;
--}
--
- /**
- * ring_buffer_peek - peek at the next event to be read
- * @buffer: The ring buffer to read
-@@ -3242,22 +3270,17 @@ ring_buffer_peek(struct ring_buffer *buf
- struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
- struct ring_buffer_event *event;
- unsigned long flags;
-- int dolock;
-+ int locked;
+@@ -358,14 +358,14 @@ static int cbc_encrypt(struct blkcipher_desc *desc,
+ err = blkcipher_walk_virt(desc, &walk);
+ desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+
+- kernel_fpu_begin();
+ while ((nbytes = walk.nbytes)) {
++ kernel_fpu_begin();
+ aesni_cbc_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr,
+ nbytes & AES_BLOCK_MASK, walk.iv);
++ kernel_fpu_end();
+ nbytes &= AES_BLOCK_SIZE - 1;
+ err = blkcipher_walk_done(desc, &walk, nbytes);
+ }
+- kernel_fpu_end();
- if (!cpumask_test_cpu(cpu, buffer->cpumask))
- return NULL;
+ return err;
+ }
+@@ -382,14 +382,14 @@ static int cbc_decrypt(struct blkcipher_desc *desc,
+ err = blkcipher_walk_virt(desc, &walk);
+ desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+
+- kernel_fpu_begin();
+ while ((nbytes = walk.nbytes)) {
++ kernel_fpu_begin();
+ aesni_cbc_dec(ctx, walk.dst.virt.addr, walk.src.virt.addr,
+ nbytes & AES_BLOCK_MASK, walk.iv);
++ kernel_fpu_end();
+ nbytes &= AES_BLOCK_SIZE - 1;
+ err = blkcipher_walk_done(desc, &walk, nbytes);
+ }
+- kernel_fpu_end();
-- dolock = rb_ok_to_lock();
- again:
-- local_irq_save(flags);
-- if (dolock)
-- spin_lock(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- event = rb_buffer_peek(cpu_buffer, ts, lost_events);
- if (event && event->type_len == RINGBUF_TYPE_PADDING)
- rb_advance_reader(cpu_buffer);
-- if (dolock)
-- spin_unlock(&cpu_buffer->reader_lock);
-- local_irq_restore(flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ return err;
+ }
+@@ -444,18 +444,20 @@ static int ctr_crypt(struct blkcipher_desc *desc,
+ err = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE);
+ desc->flags &= ~CRYPTO_TFM_REQ_MAY_SLEEP;
+
+- kernel_fpu_begin();
+ while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) {
++ kernel_fpu_begin();
+ aesni_ctr_enc(ctx, walk.dst.virt.addr, walk.src.virt.addr,
+ nbytes & AES_BLOCK_MASK, walk.iv);
++ kernel_fpu_end();
+ nbytes &= AES_BLOCK_SIZE - 1;
+ err = blkcipher_walk_done(desc, &walk, nbytes);
+ }
+ if (walk.nbytes) {
++ kernel_fpu_begin();
+ ctr_crypt_final(ctx, &walk);
++ kernel_fpu_end();
+ err = blkcipher_walk_done(desc, &walk, 0);
+ }
+- kernel_fpu_end();
- if (event && event->type_len == RINGBUF_TYPE_PADDING)
- goto again;
-@@ -3279,11 +3302,12 @@ ring_buffer_iter_peek(struct ring_buffer
- struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
- struct ring_buffer_event *event;
- unsigned long flags;
-+ int locked;
+ return err;
+ }
+diff --git a/arch/x86/include/asm/acpi.h b/arch/x86/include/asm/acpi.h
+index 610001d..c1c23d2 100644
+--- a/arch/x86/include/asm/acpi.h
++++ b/arch/x86/include/asm/acpi.h
+@@ -51,8 +51,8 @@
- again:
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- event = rb_iter_peek(iter, ts);
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
-
- if (event && event->type_len == RINGBUF_TYPE_PADDING)
- goto again;
-@@ -3309,9 +3333,7 @@ ring_buffer_consume(struct ring_buffer *
- struct ring_buffer_per_cpu *cpu_buffer;
- struct ring_buffer_event *event = NULL;
- unsigned long flags;
-- int dolock;
--
-- dolock = rb_ok_to_lock();
-+ int locked;
-
- again:
- /* might be called in atomic */
-@@ -3321,9 +3343,7 @@ ring_buffer_consume(struct ring_buffer *
- goto out;
-
- cpu_buffer = buffer->buffers[cpu];
-- local_irq_save(flags);
-- if (dolock)
-- spin_lock(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
+ #define ACPI_ASM_MACROS
+ #define BREAKPOINT3
+-#define ACPI_DISABLE_IRQS() local_irq_disable()
+-#define ACPI_ENABLE_IRQS() local_irq_enable()
++#define ACPI_DISABLE_IRQS() local_irq_disable_nort()
++#define ACPI_ENABLE_IRQS() local_irq_enable_nort()
+ #define ACPI_FLUSH_CPU_CACHE() wbinvd()
- event = rb_buffer_peek(cpu_buffer, ts, lost_events);
- if (event) {
-@@ -3331,9 +3351,8 @@ ring_buffer_consume(struct ring_buffer *
- rb_advance_reader(cpu_buffer);
- }
+ int __acpi_acquire_global_lock(unsigned int *lock);
+diff --git a/arch/x86/include/asm/amd_nb.h b/arch/x86/include/asm/amd_nb.h
+index 67f87f2..8e41071 100644
+--- a/arch/x86/include/asm/amd_nb.h
++++ b/arch/x86/include/asm/amd_nb.h
+@@ -19,9 +19,15 @@ extern int amd_numa_init(void);
+ extern int amd_get_subcaches(int);
+ extern int amd_set_subcaches(int, int);
-- if (dolock)
-- spin_unlock(&cpu_buffer->reader_lock);
-- local_irq_restore(flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
++struct amd_l3_cache {
++ unsigned indices;
++ u8 subcaches[4];
++};
+
+ struct amd_northbridge {
+ struct pci_dev *misc;
+ struct pci_dev *link;
++ struct amd_l3_cache l3_cache;
+ };
- out:
- preempt_enable();
-@@ -3418,17 +3437,18 @@ ring_buffer_read_start(struct ring_buffe
- {
- struct ring_buffer_per_cpu *cpu_buffer;
- unsigned long flags;
-+ int locked;
+ struct amd_northbridge_info {
+diff --git a/arch/x86/include/asm/irqflags.h b/arch/x86/include/asm/irqflags.h
+index 5745ce8..bba3cf8 100644
+--- a/arch/x86/include/asm/irqflags.h
++++ b/arch/x86/include/asm/irqflags.h
+@@ -60,23 +60,24 @@ static inline void native_halt(void)
+ #include <asm/paravirt.h>
+ #else
+ #ifndef __ASSEMBLY__
++#include <linux/types.h>
- if (!iter)
- return;
+-static inline unsigned long arch_local_save_flags(void)
++static inline notrace unsigned long arch_local_save_flags(void)
+ {
+ return native_save_fl();
+ }
- cpu_buffer = iter->cpu_buffer;
+-static inline void arch_local_irq_restore(unsigned long flags)
++static inline notrace void arch_local_irq_restore(unsigned long flags)
+ {
+ native_restore_fl(flags);
+ }
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- arch_spin_lock(&cpu_buffer->lock);
- rb_iter_reset(iter);
- arch_spin_unlock(&cpu_buffer->lock);
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+-static inline void arch_local_irq_disable(void)
++static inline notrace void arch_local_irq_disable(void)
+ {
+ native_irq_disable();
}
- EXPORT_SYMBOL_GPL(ring_buffer_read_start);
-@@ -3462,8 +3482,9 @@ ring_buffer_read(struct ring_buffer_iter
- struct ring_buffer_event *event;
- struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
- unsigned long flags;
-+ int locked;
+-static inline void arch_local_irq_enable(void)
++static inline notrace void arch_local_irq_enable(void)
+ {
+ native_irq_enable();
+ }
+@@ -102,7 +103,7 @@ static inline void halt(void)
+ /*
+ * For spinlocks, etc:
+ */
+-static inline unsigned long arch_local_irq_save(void)
++static inline notrace unsigned long arch_local_irq_save(void)
+ {
+ unsigned long flags = arch_local_save_flags();
+ arch_local_irq_disable();
+diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h
+index 7639dbf..0883ecd 100644
+--- a/arch/x86/include/asm/page_64_types.h
++++ b/arch/x86/include/asm/page_64_types.h
+@@ -14,12 +14,21 @@
+ #define IRQ_STACK_ORDER 2
+ #define IRQ_STACK_SIZE (PAGE_SIZE << IRQ_STACK_ORDER)
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- again:
- event = rb_iter_peek(iter, ts);
- if (!event)
-@@ -3474,7 +3495,7 @@ ring_buffer_read(struct ring_buffer_iter
+-#define STACKFAULT_STACK 1
+-#define DOUBLEFAULT_STACK 2
+-#define NMI_STACK 3
+-#define DEBUG_STACK 4
+-#define MCE_STACK 5
+-#define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
++#ifdef CONFIG_PREEMPT_RT_FULL
++# define STACKFAULT_STACK 0
++# define DOUBLEFAULT_STACK 1
++# define NMI_STACK 2
++# define DEBUG_STACK 0
++# define MCE_STACK 3
++# define N_EXCEPTION_STACKS 3 /* hw limit: 7 */
++#else
++# define STACKFAULT_STACK 1
++# define DOUBLEFAULT_STACK 2
++# define NMI_STACK 3
++# define DEBUG_STACK 4
++# define MCE_STACK 5
++# define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
++#endif
- rb_advance_iter(iter);
- out:
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ #define PUD_PAGE_SIZE (_AC(1, UL) << PUD_SHIFT)
+ #define PUD_PAGE_MASK (~(PUD_PAGE_SIZE-1))
+diff --git a/arch/x86/include/asm/rwsem.h b/arch/x86/include/asm/rwsem.h
+index df4cd32..6fc303a 100644
+--- a/arch/x86/include/asm/rwsem.h
++++ b/arch/x86/include/asm/rwsem.h
+@@ -60,7 +60,7 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
+ {
+ asm volatile("# beginning down_read\n\t"
+ LOCK_PREFIX _ASM_INC "(%1)\n\t"
+@@ -77,7 +77,7 @@ static inline void __down_read(struct rw_semaphore *sem)
+ /*
+ * trylock for reading -- returns 1 if successful, 0 if contention
+ */
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ long result, tmp;
+ asm volatile("# beginning __down_read_trylock\n\t"
+@@ -99,7 +99,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ /*
+ * lock for writing
+ */
+-static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
++static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ long tmp;
+ asm volatile("# beginning down_write\n\t"
+@@ -116,7 +116,7 @@ static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
+ : "memory", "cc");
+ }
- return event;
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
+ {
+ __down_write_nested(sem, 0);
}
-@@ -3537,13 +3558,14 @@ void ring_buffer_reset_cpu(struct ring_b
+@@ -124,7 +124,7 @@ static inline void __down_write(struct rw_semaphore *sem)
+ /*
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
{
- struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
- unsigned long flags;
-+ int locked;
+ long ret = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
+ RWSEM_ACTIVE_WRITE_BIAS);
+@@ -136,7 +136,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
+ asm volatile("# beginning __up_read\n\t"
+@@ -154,7 +154,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
+ {
+ long tmp;
+ asm volatile("# beginning __up_write\n\t"
+@@ -172,7 +172,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ asm volatile("# beginning __downgrade_write\n\t"
+ LOCK_PREFIX _ASM_ADD "%2,(%1)\n\t"
+@@ -192,7 +192,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
+ {
+ asm volatile(LOCK_PREFIX _ASM_ADD "%1,%0"
+ : "+m" (sem->count)
+@@ -202,7 +202,7 @@ static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
+ /*
+ * implement exchange and add functionality
+ */
+-static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
++static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
+ {
+ long tmp = delta;
- if (!cpumask_test_cpu(cpu, buffer->cpumask))
- return;
+diff --git a/arch/x86/include/asm/stackprotector.h b/arch/x86/include/asm/stackprotector.h
+index 1575177..ac0703b 100644
+--- a/arch/x86/include/asm/stackprotector.h
++++ b/arch/x86/include/asm/stackprotector.h
+@@ -58,7 +58,7 @@
+ */
+ static __always_inline void boot_init_stack_canary(void)
+ {
+- u64 canary;
++ u64 uninitialized_var(canary);
+ u64 tsc;
- atomic_inc(&cpu_buffer->record_disabled);
+ #ifdef CONFIG_X86_64
+@@ -69,8 +69,16 @@ static __always_inline void boot_init_stack_canary(void)
+ * of randomness. The TSC only matters for very early init,
+ * there it already has some randomness on most systems. Later
+ * on during the bootup the random pool has true entropy too.
++ *
++ * For preempt-rt we need to weaken the randomness a bit, as
++ * we can't call into the random generator from atomic context
++ * due to locking constraints. We just leave canary
++ * uninitialized and use the TSC based randomness on top of
++ * it.
+ */
++#ifndef CONFIG_PREEMPT_RT_FULL
+ get_random_bytes(&canary, sizeof(canary));
++#endif
+ tsc = __native_read_tsc();
+ canary += tsc + (tsc << 32UL);
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
+diff --git a/arch/x86/include/asm/vgtod.h b/arch/x86/include/asm/vgtod.h
+index 646b4c1..4f72846 100644
+--- a/arch/x86/include/asm/vgtod.h
++++ b/arch/x86/include/asm/vgtod.h
+@@ -5,7 +5,7 @@
+ #include <linux/clocksource.h>
- if (RB_WARN_ON(cpu_buffer, local_read(&cpu_buffer->committing)))
- goto out;
-@@ -3555,7 +3577,7 @@ void ring_buffer_reset_cpu(struct ring_b
- arch_spin_unlock(&cpu_buffer->lock);
+ struct vsyscall_gtod_data {
+- seqlock_t lock;
++ raw_seqlock_t lock;
- out:
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
-
- atomic_dec(&cpu_buffer->record_disabled);
- }
-@@ -3582,22 +3604,16 @@ int ring_buffer_empty(struct ring_buffer
- {
- struct ring_buffer_per_cpu *cpu_buffer;
- unsigned long flags;
-- int dolock;
-+ int locked;
- int cpu;
- int ret;
-
-- dolock = rb_ok_to_lock();
--
- /* yes this is racy, but if you don't like the race, lock the buffer */
- for_each_buffer_cpu(buffer, cpu) {
- cpu_buffer = buffer->buffers[cpu];
-- local_irq_save(flags);
-- if (dolock)
-- spin_lock(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- ret = rb_per_cpu_empty(cpu_buffer);
-- if (dolock)
-- spin_unlock(&cpu_buffer->reader_lock);
-- local_irq_restore(flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ /* open coded 'struct timespec' */
+ time_t wall_time_sec;
+diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c
+index b9338b8..82af921 100644
+--- a/arch/x86/kernel/apic/apic.c
++++ b/arch/x86/kernel/apic/apic.c
+@@ -856,8 +856,8 @@ void __irq_entry smp_apic_timer_interrupt(struct pt_regs *regs)
+ * Besides, if we don't timer interrupts ignore the global
+ * interrupt lock, which is the WrongThing (tm) to do.
+ */
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ local_apic_timer_interrupt();
+ irq_exit();
- if (!ret)
- return 0;
-@@ -3616,22 +3632,16 @@ int ring_buffer_empty_cpu(struct ring_bu
+@@ -1790,8 +1790,8 @@ void smp_spurious_interrupt(struct pt_regs *regs)
{
- struct ring_buffer_per_cpu *cpu_buffer;
- unsigned long flags;
-- int dolock;
-+ int locked;
- int ret;
-
- if (!cpumask_test_cpu(cpu, buffer->cpumask))
- return 1;
-
-- dolock = rb_ok_to_lock();
--
- cpu_buffer = buffer->buffers[cpu];
-- local_irq_save(flags);
-- if (dolock)
-- spin_lock(&cpu_buffer->reader_lock);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
- ret = rb_per_cpu_empty(cpu_buffer);
-- if (dolock)
-- spin_unlock(&cpu_buffer->reader_lock);
-- local_irq_restore(flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ u32 v;
- return ret;
- }
-@@ -3805,6 +3815,7 @@ int ring_buffer_read_page(struct ring_bu
- unsigned int commit;
- unsigned int read;
- u64 save_timestamp;
-+ int locked;
- int ret = -1;
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ /*
+ * Check if this really is a spurious interrupt and ACK it
+ * if it is a vectored one. Just in case...
+@@ -1827,8 +1827,8 @@ void smp_error_interrupt(struct pt_regs *regs)
+ "Illegal register address", /* APIC Error Bit 7 */
+ };
- if (!cpumask_test_cpu(cpu, buffer->cpumask))
-@@ -3826,7 +3837,7 @@ int ring_buffer_read_page(struct ring_bu
- if (!bpage)
- goto out;
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ /* First tickle the hardware, only then report what went on. -- REW */
+ v0 = apic_read(APIC_ESR);
+ apic_write(APIC_ESR, 0);
+diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
+index e529339..8f3f458 100644
+--- a/arch/x86/kernel/apic/io_apic.c
++++ b/arch/x86/kernel/apic/io_apic.c
+@@ -2275,8 +2275,8 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void)
+ unsigned vector, me;
-- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
-+ locked = read_buffer_lock(cpu_buffer, &flags);
+ ack_APIC_irq();
+- exit_idle();
+ irq_enter();
++ exit_idle();
- reader = rb_get_reader_page(cpu_buffer);
- if (!reader)
-@@ -3949,7 +3960,7 @@ int ring_buffer_read_page(struct ring_bu
- memset(&bpage->data[commit], 0, BUF_PAGE_SIZE - commit);
+ me = smp_processor_id();
+ for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) {
+@@ -2417,7 +2417,8 @@ static void ack_apic_level(struct irq_data *data)
+ irq_complete_move(cfg);
+ #ifdef CONFIG_GENERIC_PENDING_IRQ
+ /* If we are moving the irq we need to mask it */
+- if (unlikely(irqd_is_setaffinity_pending(data))) {
++ if (unlikely(irqd_is_setaffinity_pending(data) &&
++ !irqd_irq_inprogress(data))) {
+ do_unmask_irq = 1;
+ mask_ioapic(cfg);
+ }
+diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
+index 22a073d..707a7be 100644
+--- a/arch/x86/kernel/cpu/common.c
++++ b/arch/x86/kernel/cpu/common.c
+@@ -1050,7 +1050,9 @@ DEFINE_PER_CPU(unsigned int, irq_count) = -1;
+ */
+ static const unsigned int exception_stack_sizes[N_EXCEPTION_STACKS] = {
+ [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ,
++#if DEBUG_STACK > 0
+ [DEBUG_STACK - 1] = DEBUG_STKSZ
++#endif
+ };
- out_unlock:
-- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
-+ read_buffer_unlock(cpu_buffer, flags, locked);
+ static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks
+diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c
+index c105c53..951820f 100644
+--- a/arch/x86/kernel/cpu/intel_cacheinfo.c
++++ b/arch/x86/kernel/cpu/intel_cacheinfo.c
+@@ -151,28 +151,17 @@ union _cpuid4_leaf_ecx {
+ u32 full;
+ };
- out:
- return ret;
-Index: linux-3.0/kernel/trace/trace.c
-===================================================================
---- linux-3.0.orig/kernel/trace/trace.c
-+++ linux-3.0/kernel/trace/trace.c
-@@ -341,7 +341,7 @@ unsigned long trace_flags = TRACE_ITER_P
- TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE;
+-struct amd_l3_cache {
+- struct amd_northbridge *nb;
+- unsigned indices;
+- u8 subcaches[4];
+-};
+-
+-struct _cpuid4_info {
++struct _cpuid4_info_regs {
+ union _cpuid4_leaf_eax eax;
+ union _cpuid4_leaf_ebx ebx;
+ union _cpuid4_leaf_ecx ecx;
+ unsigned long size;
+- struct amd_l3_cache *l3;
+- DECLARE_BITMAP(shared_cpu_map, NR_CPUS);
++ struct amd_northbridge *nb;
+ };
- static int trace_stop_count;
--static DEFINE_SPINLOCK(tracing_start_lock);
-+static DEFINE_RAW_SPINLOCK(tracing_start_lock);
+-/* subset of above _cpuid4_info w/o shared_cpu_map */
+-struct _cpuid4_info_regs {
+- union _cpuid4_leaf_eax eax;
+- union _cpuid4_leaf_ebx ebx;
+- union _cpuid4_leaf_ecx ecx;
+- unsigned long size;
+- struct amd_l3_cache *l3;
++struct _cpuid4_info {
++ struct _cpuid4_info_regs base;
++ DECLARE_BITMAP(shared_cpu_map, NR_CPUS);
+ };
- /**
- * trace_wake_up - wake up tasks waiting for trace input
-@@ -351,6 +351,7 @@ static DEFINE_SPINLOCK(tracing_start_loc
+ unsigned short num_cache_leaves;
+@@ -314,12 +303,13 @@ struct _cache_attr {
+ /*
+ * L3 cache descriptors
*/
- void trace_wake_up(void)
+-static void __cpuinit amd_calc_l3_indices(struct amd_l3_cache *l3)
++static void __cpuinit amd_calc_l3_indices(struct amd_northbridge *nb)
{
-+#ifndef CONFIG_PREEMPT_RT_FULL
- int cpu;
++ struct amd_l3_cache *l3 = &nb->l3_cache;
+ unsigned int sc0, sc1, sc2, sc3;
+ u32 val = 0;
- if (trace_flags & TRACE_ITER_BLOCK)
-@@ -363,6 +364,7 @@ void trace_wake_up(void)
- if (!runqueue_is_locked(cpu))
- wake_up(&trace_wait);
- put_cpu();
-+#endif
- }
+- pci_read_config_dword(l3->nb->misc, 0x1C4, &val);
++ pci_read_config_dword(nb->misc, 0x1C4, &val);
- static int __init set_buf_size(char *str)
-@@ -716,6 +718,12 @@ update_max_tr_single(struct trace_array
- }
- #endif /* CONFIG_TRACER_MAX_TRACE */
+ /* calculate subcache sizes */
+ l3->subcaches[0] = sc0 = !(val & BIT(0));
+@@ -333,33 +323,16 @@ static void __cpuinit amd_calc_l3_indices(struct amd_l3_cache *l3)
+ static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf,
+ int index)
+ {
+- static struct amd_l3_cache *__cpuinitdata l3_caches;
+ int node;
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+static void default_wait_pipe(struct trace_iterator *iter);
-+#else
-+#define default_wait_pipe poll_wait_pipe
-+#endif
-+
- /**
- * register_tracer - register a tracer with the ftrace system.
- * @type - the plugin for the tracer
-@@ -958,7 +966,7 @@ void tracing_start(void)
- if (tracing_disabled)
+ /* only for L3, and not in virtualized environments */
+- if (index < 3 || amd_nb_num() == 0)
++ if (index < 3)
return;
-- spin_lock_irqsave(&tracing_start_lock, flags);
-+ raw_spin_lock_irqsave(&tracing_start_lock, flags);
- if (--trace_stop_count) {
- if (trace_stop_count < 0) {
- /* Someone screwed up their debugging */
-@@ -983,7 +991,7 @@ void tracing_start(void)
-
- ftrace_start();
- out:
-- spin_unlock_irqrestore(&tracing_start_lock, flags);
-+ raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
+- /*
+- * Strictly speaking, the amount in @size below is leaked since it is
+- * never freed but this is done only on shutdown so it doesn't matter.
+- */
+- if (!l3_caches) {
+- int size = amd_nb_num() * sizeof(struct amd_l3_cache);
+-
+- l3_caches = kzalloc(size, GFP_ATOMIC);
+- if (!l3_caches)
+- return;
+- }
+-
+ node = amd_get_nb_id(smp_processor_id());
+-
+- if (!l3_caches[node].nb) {
+- l3_caches[node].nb = node_to_amd_nb(node);
+- amd_calc_l3_indices(&l3_caches[node]);
+- }
+-
+- this_leaf->l3 = &l3_caches[node];
++ this_leaf->nb = node_to_amd_nb(node);
++ if (this_leaf->nb && !this_leaf->nb->l3_cache.indices)
++ amd_calc_l3_indices(this_leaf->nb);
}
- /**
-@@ -998,7 +1006,7 @@ void tracing_stop(void)
- unsigned long flags;
-
- ftrace_stop();
-- spin_lock_irqsave(&tracing_start_lock, flags);
-+ raw_spin_lock_irqsave(&tracing_start_lock, flags);
- if (trace_stop_count++)
- goto out;
+ /*
+@@ -369,11 +342,11 @@ static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf,
+ *
+ * @returns: the disabled index if used or negative value if slot free.
+ */
+-int amd_get_l3_disable_slot(struct amd_l3_cache *l3, unsigned slot)
++int amd_get_l3_disable_slot(struct amd_northbridge *nb, unsigned slot)
+ {
+ unsigned int reg = 0;
-@@ -1016,7 +1024,7 @@ void tracing_stop(void)
- arch_spin_unlock(&ftrace_max_lock);
+- pci_read_config_dword(l3->nb->misc, 0x1BC + slot * 4, ®);
++ pci_read_config_dword(nb->misc, 0x1BC + slot * 4, ®);
- out:
-- spin_unlock_irqrestore(&tracing_start_lock, flags);
-+ raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
- }
+ /* check whether this slot is activated already */
+ if (reg & (3UL << 30))
+@@ -387,11 +360,10 @@ static ssize_t show_cache_disable(struct _cpuid4_info *this_leaf, char *buf,
+ {
+ int index;
- void trace_stop_cmdline_recording(void);
-@@ -1120,6 +1128,8 @@ tracing_generic_entry_update(struct trac
- ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
- ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
- (need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
-+
-+ entry->migrate_disable = (tsk) ? __migrate_disabled(tsk) & 0xFF : 0;
- }
- EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
+- if (!this_leaf->l3 ||
+- !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
++ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
+ return -EINVAL;
-@@ -1757,9 +1767,10 @@ static void print_lat_help_header(struct
- seq_puts(m, "# | / _----=> need-resched \n");
- seq_puts(m, "# || / _---=> hardirq/softirq \n");
- seq_puts(m, "# ||| / _--=> preempt-depth \n");
-- seq_puts(m, "# |||| / delay \n");
-- seq_puts(m, "# cmd pid ||||| time | caller \n");
-- seq_puts(m, "# \\ / ||||| \\ | / \n");
-+ seq_puts(m, "# |||| / _--=> migrate-disable\n");
-+ seq_puts(m, "# ||||| / delay \n");
-+ seq_puts(m, "# cmd pid |||||| time | caller \n");
-+ seq_puts(m, "# \\ / ||||| \\ | / \n");
- }
+- index = amd_get_l3_disable_slot(this_leaf->l3, slot);
++ index = amd_get_l3_disable_slot(this_leaf->base.nb, slot);
+ if (index >= 0)
+ return sprintf(buf, "%d\n", index);
- static void print_func_help_header(struct seq_file *m)
-@@ -3067,6 +3078,7 @@ static int tracing_release_pipe(struct i
- return 0;
- }
+@@ -408,7 +380,7 @@ show_cache_disable_##slot(struct _cpuid4_info *this_leaf, char *buf, \
+ SHOW_CACHE_DISABLE(0)
+ SHOW_CACHE_DISABLE(1)
-+#ifndef CONFIG_PREEMPT_RT_FULL
- static unsigned int
- tracing_poll_pipe(struct file *filp, poll_table *poll_table)
+-static void amd_l3_disable_index(struct amd_l3_cache *l3, int cpu,
++static void amd_l3_disable_index(struct amd_northbridge *nb, int cpu,
+ unsigned slot, unsigned long idx)
{
-@@ -3088,8 +3100,7 @@ tracing_poll_pipe(struct file *filp, pol
+ int i;
+@@ -421,10 +393,10 @@ static void amd_l3_disable_index(struct amd_l3_cache *l3, int cpu,
+ for (i = 0; i < 4; i++) {
+ u32 reg = idx | (i << 20);
+
+- if (!l3->subcaches[i])
++ if (!nb->l3_cache.subcaches[i])
+ continue;
+
+- pci_write_config_dword(l3->nb->misc, 0x1BC + slot * 4, reg);
++ pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg);
+
+ /*
+ * We need to WBINVD on a core on the node containing the L3
+@@ -434,7 +406,7 @@ static void amd_l3_disable_index(struct amd_l3_cache *l3, int cpu,
+ wbinvd_on_cpu(cpu);
+
+ reg |= BIT(31);
+- pci_write_config_dword(l3->nb->misc, 0x1BC + slot * 4, reg);
++ pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg);
}
}
--
--void default_wait_pipe(struct trace_iterator *iter)
-+static void default_wait_pipe(struct trace_iterator *iter)
+@@ -448,24 +420,24 @@ static void amd_l3_disable_index(struct amd_l3_cache *l3, int cpu,
+ *
+ * @return: 0 on success, error status on failure
+ */
+-int amd_set_l3_disable_slot(struct amd_l3_cache *l3, int cpu, unsigned slot,
++int amd_set_l3_disable_slot(struct amd_northbridge *nb, int cpu, unsigned slot,
+ unsigned long index)
{
- DEFINE_WAIT(wait);
+ int ret = 0;
-@@ -3100,6 +3111,20 @@ void default_wait_pipe(struct trace_iter
+ /* check if @slot is already used or the index is already disabled */
+- ret = amd_get_l3_disable_slot(l3, slot);
++ ret = amd_get_l3_disable_slot(nb, slot);
+ if (ret >= 0)
+ return -EINVAL;
- finish_wait(&trace_wait, &wait);
- }
-+#else
-+static unsigned int
-+tracing_poll_pipe(struct file *filp, poll_table *poll_table)
-+{
-+ struct trace_iterator *iter = filp->private_data;
-+
-+ if ((trace_flags & TRACE_ITER_BLOCK) || !trace_empty(iter))
-+ return POLLIN | POLLRDNORM;
-+ poll_wait_pipe(iter);
-+ if (!trace_empty(iter))
-+ return POLLIN | POLLRDNORM;
-+ return 0;
-+}
-+#endif
+- if (index > l3->indices)
++ if (index > nb->l3_cache.indices)
+ return -EINVAL;
- /*
- * This is a make-shift waitqueue.
-Index: linux-3.0/kernel/trace/trace_irqsoff.c
-===================================================================
---- linux-3.0.orig/kernel/trace/trace_irqsoff.c
-+++ linux-3.0/kernel/trace/trace_irqsoff.c
-@@ -17,13 +17,14 @@
- #include <linux/fs.h>
+ /* check whether the other slot has disabled the same index already */
+- if (index == amd_get_l3_disable_slot(l3, !slot))
++ if (index == amd_get_l3_disable_slot(nb, !slot))
+ return -EINVAL;
- #include "trace.h"
-+#include <trace/events/hist.h>
+- amd_l3_disable_index(l3, cpu, slot, index);
++ amd_l3_disable_index(nb, cpu, slot, index);
- static struct trace_array *irqsoff_trace __read_mostly;
- static int tracer_enabled __read_mostly;
+ return 0;
+ }
+@@ -480,8 +452,7 @@ static ssize_t store_cache_disable(struct _cpuid4_info *this_leaf,
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
- static DEFINE_PER_CPU(int, tracing_cpu);
+- if (!this_leaf->l3 ||
+- !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
++ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
+ return -EINVAL;
--static DEFINE_SPINLOCK(max_trace_lock);
-+static DEFINE_RAW_SPINLOCK(max_trace_lock);
+ cpu = cpumask_first(to_cpumask(this_leaf->shared_cpu_map));
+@@ -489,7 +460,7 @@ static ssize_t store_cache_disable(struct _cpuid4_info *this_leaf,
+ if (strict_strtoul(buf, 10, &val) < 0)
+ return -EINVAL;
- enum {
- TRACER_IRQS_OFF = (1 << 1),
-@@ -319,7 +320,7 @@ check_critical_timing(struct trace_array
- if (!report_latency(delta))
- goto out;
+- err = amd_set_l3_disable_slot(this_leaf->l3, cpu, slot, val);
++ err = amd_set_l3_disable_slot(this_leaf->base.nb, cpu, slot, val);
+ if (err) {
+ if (err == -EEXIST)
+ printk(KERN_WARNING "L3 disable slot %d in use!\n",
+@@ -518,7 +489,7 @@ static struct _cache_attr cache_disable_1 = __ATTR(cache_disable_1, 0644,
+ static ssize_t
+ show_subcaches(struct _cpuid4_info *this_leaf, char *buf, unsigned int cpu)
+ {
+- if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
++ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
+ return -EINVAL;
-- spin_lock_irqsave(&max_trace_lock, flags);
-+ raw_spin_lock_irqsave(&max_trace_lock, flags);
+ return sprintf(buf, "%x\n", amd_get_subcaches(cpu));
+@@ -533,7 +504,7 @@ store_subcaches(struct _cpuid4_info *this_leaf, const char *buf, size_t count,
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
- /* check if we are still the max latency */
- if (!report_latency(delta))
-@@ -342,7 +343,7 @@ check_critical_timing(struct trace_array
- max_sequence++;
+- if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
++ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
+ return -EINVAL;
- out_unlock:
-- spin_unlock_irqrestore(&max_trace_lock, flags);
-+ raw_spin_unlock_irqrestore(&max_trace_lock, flags);
+ if (strict_strtoul(buf, 16, &val) < 0)
+@@ -769,7 +740,7 @@ static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index)
+ return;
+ }
+ this_leaf = CPUID4_INFO_IDX(cpu, index);
+- num_threads_sharing = 1 + this_leaf->eax.split.num_threads_sharing;
++ num_threads_sharing = 1 + this_leaf->base.eax.split.num_threads_sharing;
- out:
- data->critical_sequence = max_sequence;
-@@ -424,11 +425,13 @@ void start_critical_timings(void)
- {
- if (preempt_trace() || irq_trace())
- start_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
-+ trace_preemptirqsoff_hist(TRACE_START, 1);
- }
- EXPORT_SYMBOL_GPL(start_critical_timings);
+ if (num_threads_sharing == 1)
+ cpumask_set_cpu(cpu, to_cpumask(this_leaf->shared_cpu_map));
+@@ -820,29 +791,19 @@ static void __cpuinit free_cache_attributes(unsigned int cpu)
+ for (i = 0; i < num_cache_leaves; i++)
+ cache_remove_shared_cpu_map(cpu, i);
- void stop_critical_timings(void)
- {
-+ trace_preemptirqsoff_hist(TRACE_STOP, 0);
- if (preempt_trace() || irq_trace())
- stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
- }
-@@ -438,6 +441,7 @@ EXPORT_SYMBOL_GPL(stop_critical_timings)
- #ifdef CONFIG_PROVE_LOCKING
- void time_hardirqs_on(unsigned long a0, unsigned long a1)
- {
-+ trace_preemptirqsoff_hist(IRQS_ON, 0);
- if (!preempt_trace() && irq_trace())
- stop_critical_timing(a0, a1);
- }
-@@ -446,6 +450,7 @@ void time_hardirqs_off(unsigned long a0,
- {
- if (!preempt_trace() && irq_trace())
- start_critical_timing(a0, a1);
-+ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+- kfree(per_cpu(ici_cpuid4_info, cpu)->l3);
+ kfree(per_cpu(ici_cpuid4_info, cpu));
+ per_cpu(ici_cpuid4_info, cpu) = NULL;
}
- #else /* !CONFIG_PROVE_LOCKING */
-@@ -471,6 +476,7 @@ inline void print_irqtrace_events(struct
- */
- void trace_hardirqs_on(void)
- {
-+ trace_preemptirqsoff_hist(IRQS_ON, 0);
- if (!preempt_trace() && irq_trace())
- stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
- }
-@@ -480,11 +486,13 @@ void trace_hardirqs_off(void)
+-static int
+-__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
+-{
+- struct _cpuid4_info_regs *leaf_regs =
+- (struct _cpuid4_info_regs *)this_leaf;
+-
+- return cpuid4_cache_lookup_regs(index, leaf_regs);
+-}
+-
+ static void __cpuinit get_cpu_leaves(void *_retval)
{
- if (!preempt_trace() && irq_trace())
- start_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
-+ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+ int j, *retval = _retval, cpu = smp_processor_id();
+
+ /* Do cpuid and store the results */
+ for (j = 0; j < num_cache_leaves; j++) {
+- struct _cpuid4_info *this_leaf;
+- this_leaf = CPUID4_INFO_IDX(cpu, j);
+- *retval = cpuid4_cache_lookup(j, this_leaf);
++ struct _cpuid4_info *this_leaf = CPUID4_INFO_IDX(cpu, j);
++
++ *retval = cpuid4_cache_lookup_regs(j, &this_leaf->base);
+ if (unlikely(*retval < 0)) {
+ int i;
+
+@@ -900,16 +861,16 @@ static ssize_t show_##file_name(struct _cpuid4_info *this_leaf, char *buf, \
+ return sprintf(buf, "%lu\n", (unsigned long)this_leaf->object + val); \
}
- EXPORT_SYMBOL(trace_hardirqs_off);
- void trace_hardirqs_on_caller(unsigned long caller_addr)
+-show_one_plus(level, eax.split.level, 0);
+-show_one_plus(coherency_line_size, ebx.split.coherency_line_size, 1);
+-show_one_plus(physical_line_partition, ebx.split.physical_line_partition, 1);
+-show_one_plus(ways_of_associativity, ebx.split.ways_of_associativity, 1);
+-show_one_plus(number_of_sets, ecx.split.number_of_sets, 1);
++show_one_plus(level, base.eax.split.level, 0);
++show_one_plus(coherency_line_size, base.ebx.split.coherency_line_size, 1);
++show_one_plus(physical_line_partition, base.ebx.split.physical_line_partition, 1);
++show_one_plus(ways_of_associativity, base.ebx.split.ways_of_associativity, 1);
++show_one_plus(number_of_sets, base.ecx.split.number_of_sets, 1);
+
+ static ssize_t show_size(struct _cpuid4_info *this_leaf, char *buf,
+ unsigned int cpu)
{
-+ trace_preemptirqsoff_hist(IRQS_ON, 0);
- if (!preempt_trace() && irq_trace())
- stop_critical_timing(CALLER_ADDR0, caller_addr);
+- return sprintf(buf, "%luK\n", this_leaf->size / 1024);
++ return sprintf(buf, "%luK\n", this_leaf->base.size / 1024);
}
-@@ -494,6 +502,7 @@ void trace_hardirqs_off_caller(unsigned
+
+ static ssize_t show_shared_cpu_map_func(struct _cpuid4_info *this_leaf,
+@@ -946,7 +907,7 @@ static inline ssize_t show_shared_cpu_list(struct _cpuid4_info *leaf, char *buf,
+ static ssize_t show_type(struct _cpuid4_info *this_leaf, char *buf,
+ unsigned int cpu)
{
- if (!preempt_trace() && irq_trace())
- start_critical_timing(CALLER_ADDR0, caller_addr);
-+ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+- switch (this_leaf->eax.split.type) {
++ switch (this_leaf->base.eax.split.type) {
+ case CACHE_TYPE_DATA:
+ return sprintf(buf, "Data\n");
+ case CACHE_TYPE_INST:
+@@ -1135,7 +1096,7 @@ static int __cpuinit cache_add_dev(struct sys_device * sys_dev)
+
+ ktype_cache.default_attrs = default_attrs;
+ #ifdef CONFIG_AMD_NB
+- if (this_leaf->l3)
++ if (this_leaf->base.nb)
+ ktype_cache.default_attrs = amd_l3_attrs();
+ #endif
+ retval = kobject_init_and_add(&(this_object->kobj),
+diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
+index ff1ae9b..1b3d7eb 100644
+--- a/arch/x86/kernel/cpu/mcheck/mce.c
++++ b/arch/x86/kernel/cpu/mcheck/mce.c
+@@ -38,6 +38,7 @@
+ #include <linux/mm.h>
+ #include <linux/debugfs.h>
+ #include <linux/edac_mce.h>
++#include <linux/jiffies.h>
+
+ #include <asm/processor.h>
+ #include <asm/hw_irq.h>
+@@ -470,8 +471,8 @@ static inline void mce_get_rip(struct mce *m, struct pt_regs *regs)
+ asmlinkage void smp_mce_self_interrupt(struct pt_regs *regs)
+ {
+ ack_APIC_irq();
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ mce_notify_irq();
+ mce_schedule_work();
+ irq_exit();
+@@ -1139,17 +1140,14 @@ void mce_log_therm_throt_event(__u64 status)
+ * poller finds an MCE, poll 2x faster. When the poller finds no more
+ * errors, poll 2x slower (up to check_interval seconds).
+ */
+-static int check_interval = 5 * 60; /* 5 minutes */
++static unsigned long check_interval = 5 * 60; /* 5 minutes */
+
+-static DEFINE_PER_CPU(int, mce_next_interval); /* in jiffies */
+-static DEFINE_PER_CPU(struct timer_list, mce_timer);
++static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */
++static DEFINE_PER_CPU(struct hrtimer, mce_timer);
+
+-static void mce_start_timer(unsigned long data)
++static enum hrtimer_restart mce_start_timer(struct hrtimer *timer)
+ {
+- struct timer_list *t = &per_cpu(mce_timer, data);
+- int *n;
+-
+- WARN_ON(smp_processor_id() != data);
++ unsigned long *n;
+
+ if (mce_available(__this_cpu_ptr(&cpu_info))) {
+ machine_check_poll(MCP_TIMESTAMP,
+@@ -1162,12 +1160,13 @@ static void mce_start_timer(unsigned long data)
+ */
+ n = &__get_cpu_var(mce_next_interval);
+ if (mce_notify_irq())
+- *n = max(*n/2, HZ/100);
++ *n = max(*n/2, HZ/100UL);
+ else
+- *n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ));
++ *n = min(*n*2, round_jiffies_relative(check_interval*HZ));
+
+- t->expires = jiffies + *n;
+- add_timer_on(t, smp_processor_id());
++ hrtimer_forward(timer, timer->base->get_time(),
++ ns_to_ktime(jiffies_to_usecs(*n) * 1000));
++ return HRTIMER_RESTART;
}
- EXPORT_SYMBOL(trace_hardirqs_off_caller);
-@@ -503,13 +512,15 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller)
- #ifdef CONFIG_PREEMPT_TRACER
- void trace_preempt_on(unsigned long a0, unsigned long a1)
+ static void mce_do_trigger(struct work_struct *work)
+@@ -1393,10 +1392,11 @@ static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c)
+
+ static void __mcheck_cpu_init_timer(void)
{
-- if (preempt_trace())
-+ trace_preemptirqsoff_hist(PREEMPT_ON, 0);
-+ if (preempt_trace() && !irq_trace())
- stop_critical_timing(a0, a1);
+- struct timer_list *t = &__get_cpu_var(mce_timer);
+- int *n = &__get_cpu_var(mce_next_interval);
++ struct hrtimer *t = &__get_cpu_var(mce_timer);
++ unsigned long *n = &__get_cpu_var(mce_next_interval);
+
+- setup_timer(t, mce_start_timer, smp_processor_id());
++ hrtimer_init(t, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
++ t->function = mce_start_timer;
+
+ if (mce_ignore_ce)
+ return;
+@@ -1404,8 +1404,9 @@ static void __mcheck_cpu_init_timer(void)
+ *n = check_interval * HZ;
+ if (!*n)
+ return;
+- t->expires = round_jiffies(jiffies + *n);
+- add_timer_on(t, smp_processor_id());
++
++ hrtimer_start_range_ns(t, ns_to_ktime(jiffies_to_usecs(*n) * 1000),
++ 0 , HRTIMER_MODE_REL_PINNED);
}
- void trace_preempt_off(unsigned long a0, unsigned long a1)
+ /* Handle unconfigured int18 (should never happen) */
+@@ -1768,7 +1769,7 @@ static struct syscore_ops mce_syscore_ops = {
+
+ static void mce_cpu_restart(void *data)
{
-- if (preempt_trace())
-+ trace_preemptirqsoff_hist(PREEMPT_ON, 1);
-+ if (preempt_trace() && !irq_trace())
- start_critical_timing(a0, a1);
+- del_timer_sync(&__get_cpu_var(mce_timer));
++ hrtimer_cancel(&__get_cpu_var(mce_timer));
+ if (!mce_available(__this_cpu_ptr(&cpu_info)))
+ return;
+ __mcheck_cpu_init_generic();
+@@ -1787,7 +1788,7 @@ static void mce_disable_ce(void *all)
+ if (!mce_available(__this_cpu_ptr(&cpu_info)))
+ return;
+ if (all)
+- del_timer_sync(&__get_cpu_var(mce_timer));
++ hrtimer_cancel(&__get_cpu_var(mce_timer));
+ cmci_clear();
}
- #endif /* CONFIG_PREEMPT_TRACER */
-Index: linux-3.0/include/linux/ratelimit.h
-===================================================================
---- linux-3.0.orig/include/linux/ratelimit.h
-+++ linux-3.0/include/linux/ratelimit.h
-@@ -8,7 +8,7 @@
- #define DEFAULT_RATELIMIT_BURST 10
- struct ratelimit_state {
-- spinlock_t lock; /* protect the state */
-+ raw_spinlock_t lock; /* protect the state */
+@@ -2016,6 +2017,8 @@ static void __cpuinit mce_disable_cpu(void *h)
+ if (!mce_available(__this_cpu_ptr(&cpu_info)))
+ return;
- int interval;
- int burst;
-@@ -20,7 +20,7 @@ struct ratelimit_state {
- #define DEFINE_RATELIMIT_STATE(name, interval_init, burst_init) \
- \
- struct ratelimit_state name = { \
-- .lock = __SPIN_LOCK_UNLOCKED(name.lock), \
-+ .lock = __RAW_SPIN_LOCK_UNLOCKED(name.lock), \
- .interval = interval_init, \
- .burst = burst_init, \
++ hrtimer_cancel(&__get_cpu_var(mce_timer));
++
+ if (!(action & CPU_TASKS_FROZEN))
+ cmci_clear();
+ for (i = 0; i < banks; i++) {
+@@ -2042,6 +2045,7 @@ static void __cpuinit mce_reenable_cpu(void *h)
+ if (b->init)
+ wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
}
-@@ -28,7 +28,7 @@ struct ratelimit_state {
- static inline void ratelimit_state_init(struct ratelimit_state *rs,
- int interval, int burst)
++ __mcheck_cpu_init_timer();
+ }
+
+ /* Get notified when a cpu comes on/off. Be hotplug friendly. */
+@@ -2049,7 +2053,6 @@ static int __cpuinit
+ mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
{
-- spin_lock_init(&rs->lock);
-+ raw_spin_lock_init(&rs->lock);
- rs->interval = interval;
- rs->burst = burst;
- rs->printed = 0;
-Index: linux-3.0/kernel/printk.c
-===================================================================
---- linux-3.0.orig/kernel/printk.c
-+++ linux-3.0/kernel/printk.c
-@@ -21,6 +21,7 @@
- #include <linux/tty.h>
- #include <linux/tty_driver.h>
- #include <linux/console.h>
-+#include <linux/sysrq.h>
- #include <linux/init.h>
- #include <linux/jiffies.h>
- #include <linux/nmi.h>
-@@ -44,13 +45,6 @@
+ unsigned int cpu = (unsigned long)hcpu;
+- struct timer_list *t = &per_cpu(mce_timer, cpu);
- #include <asm/uaccess.h>
+ switch (action) {
+ case CPU_ONLINE:
+@@ -2066,16 +2069,10 @@ mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu)
+ break;
+ case CPU_DOWN_PREPARE:
+ case CPU_DOWN_PREPARE_FROZEN:
+- del_timer_sync(t);
+ smp_call_function_single(cpu, mce_disable_cpu, &action, 1);
+ break;
+ case CPU_DOWN_FAILED:
+ case CPU_DOWN_FAILED_FROZEN:
+- if (!mce_ignore_ce && check_interval) {
+- t->expires = round_jiffies(jiffies +
+- __get_cpu_var(mce_next_interval));
+- add_timer_on(t, cpu);
+- }
+ smp_call_function_single(cpu, mce_reenable_cpu, &action, 1);
+ break;
+ case CPU_POST_DEAD:
+diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c
+index 8694ef56..38e49bc 100644
+--- a/arch/x86/kernel/cpu/mcheck/mce_intel.c
++++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c
+@@ -28,7 +28,7 @@ static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned);
+ * cmci_discover_lock protects against parallel discovery attempts
+ * which could race against each other.
+ */
+-static DEFINE_SPINLOCK(cmci_discover_lock);
++static DEFINE_RAW_SPINLOCK(cmci_discover_lock);
--/*
-- * Architectures can override it:
-- */
--void asmlinkage __attribute__((weak)) early_printk(const char *fmt, ...)
--{
--}
--
- #define __LOG_BUF_LEN (1 << CONFIG_LOG_BUF_SHIFT)
+ #define CMCI_THRESHOLD 1
- /* printk's without a loglevel use this.. */
-@@ -100,7 +94,7 @@ static int console_locked, console_suspe
- * It is also used in interesting ways to provide interlocking in
- * console_unlock();.
- */
--static DEFINE_SPINLOCK(logbuf_lock);
-+static DEFINE_RAW_SPINLOCK(logbuf_lock);
+@@ -85,7 +85,7 @@ static void cmci_discover(int banks, int boot)
+ int hdr = 0;
+ int i;
- #define LOG_BUF_MASK (log_buf_len-1)
- #define LOG_BUF(idx) (log_buf[(idx) & LOG_BUF_MASK])
-@@ -212,7 +206,7 @@ void __init setup_log_buf(int early)
+- spin_lock_irqsave(&cmci_discover_lock, flags);
++ raw_spin_lock_irqsave(&cmci_discover_lock, flags);
+ for (i = 0; i < banks; i++) {
+ u64 val;
+
+@@ -116,7 +116,7 @@ static void cmci_discover(int banks, int boot)
+ WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks)));
+ }
+ }
+- spin_unlock_irqrestore(&cmci_discover_lock, flags);
++ raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
+ if (hdr)
+ printk(KERN_CONT "\n");
+ }
+@@ -150,7 +150,7 @@ void cmci_clear(void)
+
+ if (!cmci_supported(&banks))
return;
+- spin_lock_irqsave(&cmci_discover_lock, flags);
++ raw_spin_lock_irqsave(&cmci_discover_lock, flags);
+ for (i = 0; i < banks; i++) {
+ if (!test_bit(i, __get_cpu_var(mce_banks_owned)))
+ continue;
+@@ -160,7 +160,7 @@ void cmci_clear(void)
+ wrmsrl(MSR_IA32_MCx_CTL2(i), val);
+ __clear_bit(i, __get_cpu_var(mce_banks_owned));
}
+- spin_unlock_irqrestore(&cmci_discover_lock, flags);
++ raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
+ }
-- spin_lock_irqsave(&logbuf_lock, flags);
-+ raw_spin_lock_irqsave(&logbuf_lock, flags);
- log_buf_len = new_log_buf_len;
- log_buf = new_log_buf;
- new_log_buf_len = 0;
-@@ -230,7 +224,7 @@ void __init setup_log_buf(int early)
- log_start -= offset;
- con_start -= offset;
- log_end -= offset;
-- spin_unlock_irqrestore(&logbuf_lock, flags);
-+ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+ /*
+diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
+index 27c6251..f6bbc64 100644
+--- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
++++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
+@@ -396,8 +396,8 @@ static void (*smp_thermal_vector)(void) = unexpected_thermal_interrupt;
- pr_info("log_buf_len: %d\n", log_buf_len);
- pr_info("early log buf free: %d(%d%%)\n",
-@@ -363,18 +357,18 @@ int do_syslog(int type, char __user *buf
- if (error)
- goto out;
- i = 0;
-- spin_lock_irq(&logbuf_lock);
-+ raw_spin_lock_irq(&logbuf_lock);
- while (!error && (log_start != log_end) && i < len) {
- c = LOG_BUF(log_start);
- log_start++;
-- spin_unlock_irq(&logbuf_lock);
-+ raw_spin_unlock_irq(&logbuf_lock);
- error = __put_user(c,buf);
- buf++;
- i++;
- cond_resched();
-- spin_lock_irq(&logbuf_lock);
-+ raw_spin_lock_irq(&logbuf_lock);
- }
-- spin_unlock_irq(&logbuf_lock);
-+ raw_spin_unlock_irq(&logbuf_lock);
- if (!error)
- error = i;
- break;
-@@ -397,7 +391,7 @@ int do_syslog(int type, char __user *buf
- count = len;
- if (count > log_buf_len)
- count = log_buf_len;
-- spin_lock_irq(&logbuf_lock);
-+ raw_spin_lock_irq(&logbuf_lock);
- if (count > logged_chars)
- count = logged_chars;
- if (do_clear)
-@@ -414,12 +408,12 @@ int do_syslog(int type, char __user *buf
- if (j + log_buf_len < log_end)
- break;
- c = LOG_BUF(j);
-- spin_unlock_irq(&logbuf_lock);
-+ raw_spin_unlock_irq(&logbuf_lock);
- error = __put_user(c,&buf[count-1-i]);
- cond_resched();
-- spin_lock_irq(&logbuf_lock);
-+ raw_spin_lock_irq(&logbuf_lock);
- }
-- spin_unlock_irq(&logbuf_lock);
-+ raw_spin_unlock_irq(&logbuf_lock);
- if (error)
- break;
- error = i;
-@@ -509,6 +503,7 @@ static void __call_console_drivers(unsig
+ asmlinkage void smp_thermal_interrupt(struct pt_regs *regs)
{
- struct console *con;
-
-+ migrate_disable();
- for_each_console(con) {
- if (exclusive_console && con != exclusive_console)
- continue;
-@@ -517,8 +512,62 @@ static void __call_console_drivers(unsig
- (con->flags & CON_ANYTIME)))
- con->write(con, &LOG_BUF(start), end - start);
- }
-+ migrate_enable();
-+}
-+
-+#ifdef CONFIG_EARLY_PRINTK
-+struct console *early_console;
-+
-+static void early_vprintk(const char *fmt, va_list ap)
-+{
-+ char buf[512];
-+ int n = vscnprintf(buf, sizeof(buf), fmt, ap);
-+ if (early_console)
-+ early_console->write(early_console, buf, n);
-+}
-+
-+asmlinkage void early_printk(const char *fmt, ...)
-+{
-+ va_list ap;
-+ va_start(ap, fmt);
-+ early_vprintk(fmt, ap);
-+ va_end(ap);
- }
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ inc_irq_stat(irq_thermal_count);
+ smp_thermal_vector();
+ irq_exit();
+diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c
+index d746df2..aa578ca 100644
+--- a/arch/x86/kernel/cpu/mcheck/threshold.c
++++ b/arch/x86/kernel/cpu/mcheck/threshold.c
+@@ -19,8 +19,8 @@ void (*mce_threshold_vector)(void) = default_threshold_interrupt;
-+/*
-+ * This is independent of any log levels - a global
-+ * kill switch that turns off all of printk.
-+ *
-+ * Used by the NMI watchdog if early-printk is enabled.
-+ */
-+static int __read_mostly printk_killswitch;
-+
-+static int __init force_early_printk_setup(char *str)
-+{
-+ printk_killswitch = 1;
-+ return 0;
-+}
-+early_param("force_early_printk", force_early_printk_setup);
-+
-+void printk_kill(void)
-+{
-+ printk_killswitch = 1;
-+}
-+
-+static int forced_early_printk(const char *fmt, va_list ap)
-+{
-+ if (!printk_killswitch)
-+ return 0;
-+ early_vprintk(fmt, ap);
-+ return 1;
-+}
-+#else
-+static inline int forced_early_printk(const char *fmt, va_list ap)
-+{
-+ return 0;
-+}
-+#endif
+ asmlinkage void smp_threshold_interrupt(void)
+ {
+- exit_idle();
+ irq_enter();
++ exit_idle();
+ inc_irq_stat(irq_threshold_count);
+ mce_threshold_vector();
+ irq_exit();
+diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c
+index bf6d692..6b63c01 100644
+--- a/arch/x86/kernel/cpu/perf_event.c
++++ b/arch/x86/kernel/cpu/perf_event.c
+@@ -142,6 +142,8 @@ struct cpu_hw_events {
+ * AMD specific bits
+ */
+ struct amd_nb *amd_nb;
+
- static int __read_mostly ignore_loglevel;
-
- static int __init ignore_loglevel_setup(char *str)
-@@ -687,7 +736,7 @@ static void zap_locks(void)
- oops_timestamp = jiffies;
++ void *kfree_on_online;
+ };
- /* If a crash is occurring, make sure we can't deadlock */
-- spin_lock_init(&logbuf_lock);
-+ raw_spin_lock_init(&logbuf_lock);
- /* And make sure that we print immediately */
- sema_init(&console_sem, 1);
- }
-@@ -779,12 +828,18 @@ static inline int can_use_console(unsign
- * interrupts disabled. It should return with 'lockbuf_lock'
- * released but interrupts still disabled.
- */
--static int console_trylock_for_printk(unsigned int cpu)
-+static int console_trylock_for_printk(unsigned int cpu, unsigned long flags)
- __releases(&logbuf_lock)
+ #define __EVENT_CONSTRAINT(c, n, m, w) {\
+@@ -1452,10 +1454,12 @@ static int __cpuinit
+ x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
{
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ int lock = (!early_boot_irqs_disabled && !irqs_disabled_flags(flags) &&
-+ !preempt_count()) || sysrq_in_progress;
-+#else
-+ int lock = 1;
-+#endif
- int retval = 0;
-
-- if (console_trylock()) {
-+ if (lock && console_trylock()) {
- retval = 1;
+ unsigned int cpu = (long)hcpu;
++ struct cpu_hw_events *cpuc = &per_cpu(cpu_hw_events, cpu);
+ int ret = NOTIFY_OK;
- /*
-@@ -800,7 +855,7 @@ static int console_trylock_for_printk(un
- }
- }
- printk_cpu = UINT_MAX;
-- spin_unlock(&logbuf_lock);
-+ raw_spin_unlock(&logbuf_lock);
- return retval;
- }
- static const char recursion_bug_msg [] =
-@@ -833,6 +888,13 @@ asmlinkage int vprintk(const char *fmt,
- size_t plen;
- char special;
+ switch (action & ~CPU_TASKS_FROZEN) {
+ case CPU_UP_PREPARE:
++ cpuc->kfree_on_online = NULL;
+ if (x86_pmu.cpu_prepare)
+ ret = x86_pmu.cpu_prepare(cpu);
+ break;
+@@ -1465,6 +1469,10 @@ x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
+ x86_pmu.cpu_starting(cpu);
+ break;
-+ /*
-+ * Fall back to early_printk if a debugging subsystem has
-+ * killed printk output
-+ */
-+ if (unlikely(forced_early_printk(fmt, args)))
-+ return 1;
++ case CPU_ONLINE:
++ kfree(cpuc->kfree_on_online);
++ break;
+
- boot_delay_msec();
- printk_delay();
+ case CPU_DYING:
+ if (x86_pmu.cpu_dying)
+ x86_pmu.cpu_dying(cpu);
+diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c
+index fe29c1d..eaa5f72 100644
+--- a/arch/x86/kernel/cpu/perf_event_amd.c
++++ b/arch/x86/kernel/cpu/perf_event_amd.c
+@@ -336,7 +336,7 @@ static void amd_pmu_cpu_starting(int cpu)
+ continue;
-@@ -860,7 +922,7 @@ asmlinkage int vprintk(const char *fmt,
- }
+ if (nb->nb_id == nb_id) {
+- kfree(cpuc->amd_nb);
++ cpuc->kfree_on_online = cpuc->amd_nb;
+ cpuc->amd_nb = nb;
+ break;
+ }
+diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
+index dd208a8..fa3ca78 100644
+--- a/arch/x86/kernel/cpu/perf_event_intel.c
++++ b/arch/x86/kernel/cpu/perf_event_intel.c
+@@ -1267,7 +1267,7 @@ static void intel_pmu_cpu_starting(int cpu)
+ struct intel_percore *pc = per_cpu(cpu_hw_events, i).per_core;
- lockdep_off();
-- spin_lock(&logbuf_lock);
-+ raw_spin_lock(&logbuf_lock);
- printk_cpu = this_cpu;
+ if (pc && pc->core_id == core_id) {
+- kfree(cpuc->per_core);
++ cpuc->kfree_on_online = cpuc->per_core;
+ cpuc->per_core = pc;
+ break;
+ }
+diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c
+index e71c98d..d53c7ae 100644
+--- a/arch/x86/kernel/dumpstack_64.c
++++ b/arch/x86/kernel/dumpstack_64.c
+@@ -21,10 +21,14 @@
+ (N_EXCEPTION_STACKS + DEBUG_STKSZ/EXCEPTION_STKSZ - 2)
- if (recursion_bug) {
-@@ -953,8 +1015,15 @@ asmlinkage int vprintk(const char *fmt,
- * will release 'logbuf_lock' regardless of whether it
- * actually gets the semaphore or not.
- */
-- if (console_trylock_for_printk(this_cpu))
-+ if (console_trylock_for_printk(this_cpu, flags)) {
-+#ifndef CONFIG_PREEMPT_RT_FULL
- console_unlock();
-+#else
-+ raw_local_irq_restore(flags);
-+ console_unlock();
-+ raw_local_irq_save(flags);
+ static char x86_stack_ids[][8] = {
++#if DEBUG_STACK > 0
+ [ DEBUG_STACK-1 ] = "#DB",
+#endif
-+ }
-
- lockdep_on();
- out_restore_irqs:
-@@ -1213,8 +1282,8 @@ void printk_tick(void)
-
- int printk_needs_cpu(int cpu)
- {
-- if (cpu_is_offline(cpu))
-- printk_tick();
-+ if (unlikely(cpu_is_offline(cpu)))
-+ __this_cpu_write(printk_pending, 0);
- return __this_cpu_read(printk_pending);
- }
-
-@@ -1252,18 +1321,23 @@ void console_unlock(void)
- console_may_schedule = 0;
+ [ NMI_STACK-1 ] = "NMI",
+ [ DOUBLEFAULT_STACK-1 ] = "#DF",
++#if STACKFAULT_STACK > 0
+ [ STACKFAULT_STACK-1 ] = "#SS",
++#endif
+ [ MCE_STACK-1 ] = "#MC",
+ #if DEBUG_STKSZ > EXCEPTION_STKSZ
+ [ N_EXCEPTION_STACKS ...
+diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c
+index cd28a35..5f3d9c5 100644
+--- a/arch/x86/kernel/early_printk.c
++++ b/arch/x86/kernel/early_printk.c
+@@ -169,25 +169,9 @@ static struct console early_serial_console = {
+ .index = -1,
+ };
- for ( ; ; ) {
-- spin_lock_irqsave(&logbuf_lock, flags);
-+ raw_spin_lock_irqsave(&logbuf_lock, flags);
- wake_klogd |= log_start - log_end;
- if (con_start == log_end)
- break; /* Nothing to print */
- _con_start = con_start;
- _log_end = log_end;
- con_start = log_end; /* Flush */
-- spin_unlock(&logbuf_lock);
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+ raw_spin_unlock(&logbuf_lock);
- stop_critical_timings(); /* don't trace print latency */
- call_console_drivers(_con_start, _log_end);
- start_critical_timings();
- local_irq_restore(flags);
-+#else
-+ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
-+ call_console_drivers(_con_start, _log_end);
-+#endif
- }
- console_locked = 0;
-
-@@ -1272,7 +1346,7 @@ void console_unlock(void)
- exclusive_console = NULL;
-
- up(&console_sem);
-- spin_unlock_irqrestore(&logbuf_lock, flags);
-+ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
- if (wake_klogd)
- wake_up_klogd();
- }
-@@ -1502,9 +1576,9 @@ void register_console(struct console *ne
- * console_unlock(); will print out the buffered messages
- * for us.
- */
-- spin_lock_irqsave(&logbuf_lock, flags);
-+ raw_spin_lock_irqsave(&logbuf_lock, flags);
- con_start = log_start;
-- spin_unlock_irqrestore(&logbuf_lock, flags);
-+ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
- /*
- * We're about to replay the log buffer. Only do this to the
- * just-registered console to avoid excessive message spam to
-@@ -1711,10 +1785,10 @@ void kmsg_dump(enum kmsg_dump_reason rea
- /* Theoretically, the log could move on after we do this, but
- there's not a lot we can do about that. The new messages
- will overwrite the start of what we dump. */
-- spin_lock_irqsave(&logbuf_lock, flags);
-+ raw_spin_lock_irqsave(&logbuf_lock, flags);
- end = log_end & LOG_BUF_MASK;
- chars = logged_chars;
-- spin_unlock_irqrestore(&logbuf_lock, flags);
-+ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+-/* Direct interface for emergencies */
+-static struct console *early_console = &early_vga_console;
+-static int __initdata early_console_initialized;
+-
+-asmlinkage void early_printk(const char *fmt, ...)
+-{
+- char buf[512];
+- int n;
+- va_list ap;
+-
+- va_start(ap, fmt);
+- n = vscnprintf(buf, sizeof(buf), fmt, ap);
+- early_console->write(early_console, buf, n);
+- va_end(ap);
+-}
+-
+ static inline void early_console_register(struct console *con, int keep_early)
+ {
+- if (early_console->index != -1) {
++ if (con->index != -1) {
+ printk(KERN_CRIT "ERROR: earlyprintk= %s already used\n",
+ con->name);
+ return;
+@@ -207,9 +191,8 @@ static int __init setup_early_printk(char *buf)
+ if (!buf)
+ return 0;
- if (chars > end) {
- s1 = log_buf + log_buf_len - chars + end;
-Index: linux-3.0/lib/ratelimit.c
-===================================================================
---- linux-3.0.orig/lib/ratelimit.c
-+++ linux-3.0/lib/ratelimit.c
-@@ -39,7 +39,7 @@ int ___ratelimit(struct ratelimit_state
- * in addition to the one that will be printed by
- * the entity that is holding the lock already:
- */
-- if (!spin_trylock_irqsave(&rs->lock, flags))
-+ if (!raw_spin_trylock_irqsave(&rs->lock, flags))
+- if (early_console_initialized)
++ if (early_console)
return 0;
+- early_console_initialized = 1;
- if (!rs->begin)
-@@ -60,7 +60,7 @@ int ___ratelimit(struct ratelimit_state
- rs->missed++;
- ret = 0;
- }
-- spin_unlock_irqrestore(&rs->lock, flags);
-+ raw_spin_unlock_irqrestore(&rs->lock, flags);
+ keep = (strstr(buf, "keep") != NULL);
- return ret;
- }
-Index: linux-3.0/include/linux/init_task.h
-===================================================================
---- linux-3.0.orig/include/linux/init_task.h
-+++ linux-3.0/include/linux/init_task.h
-@@ -42,7 +42,7 @@ extern struct fs_struct init_fs;
- .cputimer = { \
- .cputime = INIT_CPUTIME, \
- .running = 0, \
-- .lock = __SPIN_LOCK_UNLOCKED(sig.cputimer.lock), \
-+ .lock = __RAW_SPIN_LOCK_UNLOCKED(sig.cputimer.lock), \
- }, \
- .cred_guard_mutex = \
- __MUTEX_INITIALIZER(sig.cred_guard_mutex), \
-@@ -126,6 +126,14 @@ extern struct cred init_cred;
- # define INIT_PERF_EVENTS(tsk)
+diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S
+index 5c1a9197..5974741 100644
+--- a/arch/x86/kernel/entry_32.S
++++ b/arch/x86/kernel/entry_32.S
+@@ -625,7 +625,11 @@ work_notifysig: # deal with pending signals and
+ jne work_notifysig_v86 # returning to kernel-space or
+ # vm86-space
+ xorl %edx, %edx
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
+ call do_notify_resume
++ DISABLE_INTERRUPTS(CLBR_ANY)
++ TRACE_IRQS_OFF
+ jmp resume_userspace_sig
+
+ ALIGN
+@@ -638,7 +642,11 @@ work_notifysig_v86:
+ movl %esp, %eax
#endif
+ xorl %edx, %edx
++ TRACE_IRQS_ON
++ ENABLE_INTERRUPTS(CLBR_NONE)
+ call do_notify_resume
++ DISABLE_INTERRUPTS(CLBR_ANY)
++ TRACE_IRQS_OFF
+ jmp resume_userspace_sig
+ END(work_pending)
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+# define INIT_TIMER_LIST .posix_timer_list = NULL,
-+#else
-+# define INIT_TIMER_LIST
+diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
+index 8a445a0..1689be7 100644
+--- a/arch/x86/kernel/entry_64.S
++++ b/arch/x86/kernel/entry_64.S
+@@ -1197,6 +1197,7 @@ ENTRY(kernel_execve)
+ CFI_ENDPROC
+ END(kernel_execve)
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /* Call softirq on interrupt stack. Interrupts are off. */
+ ENTRY(call_softirq)
+ CFI_STARTPROC
+@@ -1216,6 +1217,7 @@ ENTRY(call_softirq)
+ ret
+ CFI_ENDPROC
+ END(call_softirq)
+#endif
+
+ #ifdef CONFIG_XEN
+ zeroentry xen_hypervisor_callback xen_do_hypervisor_callback
+diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c
+index 6781765..91e83b8 100644
+--- a/arch/x86/kernel/hpet.c
++++ b/arch/x86/kernel/hpet.c
+@@ -7,6 +7,7 @@
+ #include <linux/slab.h>
+ #include <linux/hpet.h>
+ #include <linux/init.h>
++#include <linux/dmi.h>
+ #include <linux/cpu.h>
+ #include <linux/pm.h>
+ #include <linux/io.h>
+@@ -566,6 +567,30 @@ static void init_one_hpet_msi_clockevent(struct hpet_dev *hdev, int cpu)
+ #define RESERVE_TIMERS 0
+ #endif
+
++static int __init dmi_disable_hpet_msi(const struct dmi_system_id *d)
++{
++ hpet_msi_disable = 1;
++ return 0;
++}
+
-+#define INIT_TASK_COMM "swapper"
++static struct dmi_system_id __initdata dmi_hpet_table[] = {
++ /*
++ * MSI based per cpu timers lose interrupts when intel_idle()
++ * is enabled - independent of the c-state. With idle=poll the
++ * problem cannot be observed. We have no idea yet, whether
++ * this is a W510 specific issue or a general chipset oddity.
++ */
++ {
++ .callback = dmi_disable_hpet_msi,
++ .ident = "Lenovo W510",
++ .matches = {
++ DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
++ DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad W510"),
++ },
++ },
++ {}
++};
+
- /*
- * INIT_TASK is used to set up the first task table, touch at
- * your own risk!. Base=0, limit=0x1fffff (=2MB)
-@@ -162,7 +170,7 @@ extern struct cred init_cred;
- .group_leader = &tsk, \
- RCU_INIT_POINTER(.real_cred, &init_cred), \
- RCU_INIT_POINTER(.cred, &init_cred), \
-- .comm = "swapper", \
-+ .comm = INIT_TASK_COMM, \
- .thread = INIT_THREAD, \
- .fs = &init_fs, \
- .files = &init_files, \
-@@ -179,6 +187,7 @@ extern struct cred init_cred;
- .fs_excl = ATOMIC_INIT(0), \
- .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \
- .timer_slack_ns = 50000, /* 50 usec default slack */ \
-+ INIT_TIMER_LIST \
- .pids = { \
- [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \
- [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \
-Index: linux-3.0/include/linux/sched.h
-===================================================================
---- linux-3.0.orig/include/linux/sched.h
-+++ linux-3.0/include/linux/sched.h
-@@ -63,6 +63,7 @@ struct sched_param {
- #include <linux/nodemask.h>
- #include <linux/mm_types.h>
+ static void hpet_msi_capability_lookup(unsigned int start_timer)
+ {
+ unsigned int id;
+@@ -573,6 +598,8 @@ static void hpet_msi_capability_lookup(unsigned int start_timer)
+ unsigned int num_timers_used = 0;
+ int i;
-+#include <asm/kmap_types.h>
- #include <asm/system.h>
- #include <asm/page.h>
- #include <asm/ptrace.h>
-@@ -90,6 +91,7 @@ struct sched_param {
- #include <linux/task_io_accounting.h>
- #include <linux/latencytop.h>
- #include <linux/cred.h>
-+#include <linux/hardirq.h>
++ dmi_check_system(dmi_hpet_table);
++
+ if (hpet_msi_disable)
+ return;
- #include <asm/processor.h>
+diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c
+index 6c0802e..73cf928 100644
+--- a/arch/x86/kernel/irq.c
++++ b/arch/x86/kernel/irq.c
+@@ -180,8 +180,8 @@ unsigned int __irq_entry do_IRQ(struct pt_regs *regs)
+ unsigned vector = ~regs->orig_ax;
+ unsigned irq;
-@@ -359,6 +361,7 @@ extern signed long schedule_timeout_inte
- extern signed long schedule_timeout_killable(signed long timeout);
- extern signed long schedule_timeout_uninterruptible(signed long timeout);
- asmlinkage void schedule(void);
-+extern void schedule_preempt_disabled(void);
- extern int mutex_spin_on_owner(struct mutex *lock, struct task_struct *owner);
+- exit_idle();
+ irq_enter();
++ exit_idle();
- struct nsproxy;
-@@ -510,7 +513,7 @@ struct task_cputime {
- struct thread_group_cputimer {
- struct task_cputime cputime;
- int running;
-- spinlock_t lock;
-+ raw_spinlock_t lock;
- };
+ irq = __this_cpu_read(vector_irq[vector]);
- #include <linux/rwsem.h>
-@@ -1070,6 +1073,7 @@ struct sched_domain;
- #define WF_SYNC 0x01 /* waker goes to sleep after wakup */
- #define WF_FORK 0x02 /* child wakeup after fork */
- #define WF_MIGRATED 0x04 /* internal use, task got migrated */
-+#define WF_LOCK_SLEEPER 0x08 /* wakeup spinlock "sleeper" */
+@@ -208,10 +208,10 @@ void smp_x86_platform_ipi(struct pt_regs *regs)
- #define ENQUEUE_WAKEUP 1
- #define ENQUEUE_HEAD 2
-@@ -1219,6 +1223,7 @@ enum perf_event_task_context {
+ ack_APIC_irq();
- struct task_struct {
- volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */
-+ volatile long saved_state; /* saved state for "spinlock sleepers" */
- void *stack;
- atomic_t usage;
- unsigned int flags; /* per process flags, defined below */
-@@ -1255,14 +1260,17 @@ struct task_struct {
- #endif
-
- unsigned int policy;
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ int migrate_disable;
-+#ifdef CONFIG_SCHED_DEBUG
-+ int migrate_disable_atomic;
-+#endif
-+#endif
- cpumask_t cpus_allowed;
-
- #ifdef CONFIG_PREEMPT_RCU
- int rcu_read_lock_nesting;
- char rcu_read_unlock_special;
--#if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU)
-- int rcu_boosted;
--#endif /* #if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU) */
- struct list_head rcu_node_entry;
- #endif /* #ifdef CONFIG_PREEMPT_RCU */
- #ifdef CONFIG_TREE_PREEMPT_RCU
-@@ -1356,6 +1364,9 @@ struct task_struct {
-
- struct task_cputime cputime_expires;
- struct list_head cpu_timers[3];
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ struct task_struct *posix_timer_list;
-+#endif
-
- /* process credentials */
- const struct cred __rcu *real_cred; /* objective and real subjective task
-@@ -1389,6 +1400,7 @@ struct task_struct {
- /* signal handlers */
- struct signal_struct *signal;
- struct sighand_struct *sighand;
-+ struct sigqueue *sigqueue_cache;
-
- sigset_t blocked, real_blocked;
- sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */
-@@ -1432,6 +1444,9 @@ struct task_struct {
- /* mutex deadlock detection */
- struct mutex_waiter *blocked_on;
- #endif
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ int pagefault_disabled;
-+#endif
- #ifdef CONFIG_TRACE_IRQFLAGS
- unsigned int irq_events;
- unsigned long hardirq_enable_ip;
-@@ -1558,6 +1573,12 @@ struct task_struct {
- unsigned long trace;
- /* bitmask and counter of trace recursion */
- unsigned long trace_recursion;
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ u64 preempt_timestamp_hist;
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ unsigned long timer_offset;
-+#endif
-+#endif
- #endif /* CONFIG_TRACING */
- #ifdef CONFIG_CGROUP_MEM_RES_CTLR /* memcg uses this to do batch job */
- struct memcg_batch_info {
-@@ -1570,10 +1591,26 @@ struct task_struct {
- #ifdef CONFIG_HAVE_HW_BREAKPOINT
- atomic_t ptrace_bp_refcnt;
- #endif
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ struct rcu_head put_rcu;
-+ int softirq_nestcnt;
-+#endif
-+#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM
-+ int kmap_idx;
-+ pte_t kmap_pte[KM_TYPE_NR];
-+#endif
- };
+- exit_idle();
+-
+ irq_enter();
--/* Future-safe accessor for struct task_struct's cpus_allowed. */
--#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed)
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+static inline bool cur_pf_disabled(void) { return current->pagefault_disabled; }
-+#else
-+static inline bool cur_pf_disabled(void) { return false; }
-+#endif
++ exit_idle();
+
-+static inline bool pagefault_disabled(void)
-+{
-+ return in_atomic() || cur_pf_disabled();
-+}
+ inc_irq_stat(x86_platform_ipis);
- /*
- * Priority of a process goes from 0..MAX_PRIO-1, valid RT
-@@ -1743,6 +1780,15 @@ extern struct pid *cad_pid;
- extern void free_task(struct task_struct *tsk);
- #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0)
+ if (x86_platform_ipi_callback)
+diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c
+index 7209070..84417a2 100644
+--- a/arch/x86/kernel/irq_32.c
++++ b/arch/x86/kernel/irq_32.c
+@@ -149,6 +149,7 @@ void __cpuinit irq_ctx_init(int cpu)
+ cpu, per_cpu(hardirq_ctx, cpu), per_cpu(softirq_ctx, cpu));
+ }
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+extern void __put_task_struct_cb(struct rcu_head *rhp);
-+
-+static inline void put_task_struct(struct task_struct *t)
-+{
-+ if (atomic_dec_and_test(&t->usage))
-+ call_rcu(&t->put_rcu, __put_task_struct_cb);
-+}
-+#else
- extern void __put_task_struct(struct task_struct *t);
++#ifndef CONFIG_PREEMPT_RT_FULL
+ asmlinkage void do_softirq(void)
+ {
+ unsigned long flags;
+@@ -179,6 +180,7 @@ asmlinkage void do_softirq(void)
- static inline void put_task_struct(struct task_struct *t)
-@@ -1750,6 +1796,7 @@ static inline void put_task_struct(struc
- if (atomic_dec_and_test(&t->usage))
- __put_task_struct(t);
+ local_irq_restore(flags);
}
+#endif
- extern void task_times(struct task_struct *p, cputime_t *ut, cputime_t *st);
- extern void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *st);
-@@ -1774,6 +1821,7 @@ extern void thread_group_times(struct ta
- #define PF_FROZEN 0x00010000 /* frozen for system suspend */
- #define PF_FSTRANS 0x00020000 /* inside a filesystem transaction */
- #define PF_KSWAPD 0x00040000 /* I am kswapd */
-+#define PF_STOMPER 0x00080000 /* I am a stomp machine thread */
- #define PF_LESS_THROTTLE 0x00100000 /* Throttle me less: I clean memory */
- #define PF_KTHREAD 0x00200000 /* I am a kernel thread */
- #define PF_RANDOMIZE 0x00400000 /* randomize virtual address space */
-@@ -2021,15 +2069,27 @@ static inline void sched_autogroup_exit(
- #endif
-
- #ifdef CONFIG_RT_MUTEXES
-+extern void task_setprio(struct task_struct *p, int prio);
- extern int rt_mutex_getprio(struct task_struct *p);
--extern void rt_mutex_setprio(struct task_struct *p, int prio);
-+static inline void rt_mutex_setprio(struct task_struct *p, int prio)
-+{
-+ task_setprio(p, prio);
-+}
- extern void rt_mutex_adjust_pi(struct task_struct *p);
-+static inline bool tsk_is_pi_blocked(struct task_struct *tsk)
-+{
-+ return tsk->pi_blocked_on != NULL;
-+}
- #else
- static inline int rt_mutex_getprio(struct task_struct *p)
+ bool handle_irq(unsigned irq, struct pt_regs *regs)
{
- return p->normal_prio;
+diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c
+index acf8fbf..aa8ae57 100644
+--- a/arch/x86/kernel/irq_64.c
++++ b/arch/x86/kernel/irq_64.c
+@@ -62,7 +62,7 @@ bool handle_irq(unsigned irq, struct pt_regs *regs)
+ return true;
}
- # define rt_mutex_adjust_pi(p) do { } while (0)
-+static inline bool tsk_is_pi_blocked(struct task_struct *tsk)
-+{
-+ return false;
-+}
- #endif
-
- extern bool yield_to(struct task_struct *p, bool preempt);
-@@ -2109,6 +2169,7 @@ extern void xtime_update(unsigned long t
- extern int wake_up_state(struct task_struct *tsk, unsigned int state);
- extern int wake_up_process(struct task_struct *tsk);
-+extern int wake_up_lock_sleeper(struct task_struct * tsk);
- extern void wake_up_new_task(struct task_struct *tsk);
- #ifdef CONFIG_SMP
- extern void kick_process(struct task_struct *tsk);
-@@ -2198,12 +2259,24 @@ extern struct mm_struct * mm_alloc(void)
+-
++#ifndef CONFIG_PREEMPT_RT_FULL
+ extern void call_softirq(void);
- /* mmdrop drops the mm and the page tables */
- extern void __mmdrop(struct mm_struct *);
-+
- static inline void mmdrop(struct mm_struct * mm)
- {
- if (unlikely(atomic_dec_and_test(&mm->mm_count)))
- __mmdrop(mm);
+ asmlinkage void do_softirq(void)
+@@ -82,3 +82,4 @@ asmlinkage void do_softirq(void)
+ }
+ local_irq_restore(flags);
}
-
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+extern void __mmdrop_delayed(struct rcu_head *rhp);
-+static inline void mmdrop_delayed(struct mm_struct *mm)
-+{
-+ if (atomic_dec_and_test(&mm->mm_count))
-+ call_rcu(&mm->delayed_drop, __mmdrop_delayed);
-+}
-+#else
-+# define mmdrop_delayed(mm) mmdrop(mm)
+#endif
-+
- /* mmput gets rid of the mappings and all user-space */
- extern void mmput(struct mm_struct *);
- /* Grab a reference to a task's mm, if it is not already going away */
-@@ -2509,7 +2582,7 @@ extern int _cond_resched(void);
-
- extern int __cond_resched_lock(spinlock_t *lock);
-
--#ifdef CONFIG_PREEMPT
-+#if defined(CONFIG_PREEMPT) && !defined(CONFIG_PREEMPT_RT_FULL)
- #define PREEMPT_LOCK_OFFSET PREEMPT_OFFSET
- #else
- #define PREEMPT_LOCK_OFFSET 0
-@@ -2520,12 +2593,16 @@ extern int __cond_resched_lock(spinlock_
- __cond_resched_lock(lock); \
- })
+diff --git a/arch/x86/kernel/irq_work.c b/arch/x86/kernel/irq_work.c
+index ca8f703..129b8bb 100644
+--- a/arch/x86/kernel/irq_work.c
++++ b/arch/x86/kernel/irq_work.c
+@@ -18,6 +18,7 @@ void smp_irq_work_interrupt(struct pt_regs *regs)
+ irq_exit();
+ }
+#ifndef CONFIG_PREEMPT_RT_FULL
- extern int __cond_resched_softirq(void);
-
- #define cond_resched_softirq() ({ \
- __might_sleep(__FILE__, __LINE__, SOFTIRQ_DISABLE_OFFSET); \
- __cond_resched_softirq(); \
- })
-+#else
-+# define cond_resched_softirq() cond_resched()
-+#endif
-
- /*
- * Does a critical section need to be broken due to another
-@@ -2549,7 +2626,7 @@ void thread_group_cputimer(struct task_s
-
- static inline void thread_group_cputime_init(struct signal_struct *sig)
+ void arch_irq_work_raise(void)
{
-- spin_lock_init(&sig->cputimer.lock);
-+ raw_spin_lock_init(&sig->cputimer.lock);
+ #ifdef CONFIG_X86_LOCAL_APIC
+@@ -28,3 +29,4 @@ void arch_irq_work_raise(void)
+ apic_wait_icr_idle();
+ #endif
}
++#endif
+diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c
+index 794bc95..d334d77 100644
+--- a/arch/x86/kernel/kprobes.c
++++ b/arch/x86/kernel/kprobes.c
+@@ -477,7 +477,6 @@ static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs,
+ * stepping.
+ */
+ regs->ip = (unsigned long)p->ainsn.insn;
+- preempt_enable_no_resched();
+ return;
+ }
+ #endif
+diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c
+index a3d0dc5..8380813 100644
+--- a/arch/x86/kernel/process_32.c
++++ b/arch/x86/kernel/process_32.c
+@@ -38,6 +38,7 @@
+ #include <linux/uaccess.h>
+ #include <linux/io.h>
+ #include <linux/kdebug.h>
++#include <linux/highmem.h>
- /*
-@@ -2588,6 +2665,26 @@ static inline void set_task_cpu(struct t
+ #include <asm/pgtable.h>
+ #include <asm/system.h>
+@@ -113,9 +114,7 @@ void cpu_idle(void)
+ start_critical_timings();
+ }
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
+ }
- #endif /* CONFIG_SMP */
+@@ -348,6 +347,41 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p)
+ task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT))
+ __switch_to_xtra(prev_p, next_p, tss);
-+static inline int __migrate_disabled(struct task_struct *p)
-+{
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ return p->migrate_disable;
-+#else
-+ return 0;
-+#endif
-+}
++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM
++ /*
++ * Save @prev's kmap_atomic stack
++ */
++ prev_p->kmap_idx = __this_cpu_read(__kmap_atomic_idx);
++ if (unlikely(prev_p->kmap_idx)) {
++ int i;
+
-+/* Future-safe accessor for struct task_struct's cpus_allowed. */
-+static inline const struct cpumask *tsk_cpus_allowed(struct task_struct *p)
-+{
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ if (p->migrate_disable)
-+ return cpumask_of(task_cpu(p));
-+#endif
++ for (i = 0; i < prev_p->kmap_idx; i++) {
++ int idx = i + KM_TYPE_NR * smp_processor_id();
+
-+ return &p->cpus_allowed;
-+}
++ pte_t *ptep = kmap_pte - idx;
++ prev_p->kmap_pte[i] = *ptep;
++ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx));
++ }
+
- extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask);
- extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
++ __this_cpu_write(__kmap_atomic_idx, 0);
++ }
++
++ /*
++ * Restore @next_p's kmap_atomic stack
++ */
++ if (unlikely(next_p->kmap_idx)) {
++ int i;
++
++ __this_cpu_write(__kmap_atomic_idx, next_p->kmap_idx);
++
++ for (i = 0; i < next_p->kmap_idx; i++) {
++ int idx = i + KM_TYPE_NR * smp_processor_id();
++
++ set_pte(kmap_pte - idx, next_p->kmap_pte[i]);
++ }
++ }
++#endif
++
+ /* If we're going to preload the fpu context, make sure clts
+ is run while we're batching the cpu state updates. */
+ if (preload_fpu)
+diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c
+index ca6f7ab..7d26b37 100644
+--- a/arch/x86/kernel/process_64.c
++++ b/arch/x86/kernel/process_64.c
+@@ -146,9 +146,7 @@ void cpu_idle(void)
+ }
-Index: linux-3.0/kernel/posix-cpu-timers.c
-===================================================================
---- linux-3.0.orig/kernel/posix-cpu-timers.c
-+++ linux-3.0/kernel/posix-cpu-timers.c
-@@ -282,13 +282,13 @@ void thread_group_cputimer(struct task_s
- * it.
- */
- thread_group_cputime(tsk, &sum);
-- spin_lock_irqsave(&cputimer->lock, flags);
-+ raw_spin_lock_irqsave(&cputimer->lock, flags);
- cputimer->running = 1;
- update_gt_cputime(&cputimer->cputime, &sum);
- } else
-- spin_lock_irqsave(&cputimer->lock, flags);
-+ raw_spin_lock_irqsave(&cputimer->lock, flags);
- *times = cputimer->cputime;
-- spin_unlock_irqrestore(&cputimer->lock, flags);
-+ raw_spin_unlock_irqrestore(&cputimer->lock, flags);
+ tick_nohz_restart_sched_tick();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
}
- /*
-@@ -701,7 +701,7 @@ static int posix_cpu_timer_set(struct k_
- /*
- * Disarm any old timer after extracting its expiry time.
- */
-- BUG_ON(!irqs_disabled());
-+ BUG_ON_NONRT(!irqs_disabled());
+diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c
+index 3e68218..bbcbaaa 100644
+--- a/arch/x86/kernel/vsyscall_64.c
++++ b/arch/x86/kernel/vsyscall_64.c
+@@ -52,7 +52,7 @@
+ DEFINE_VVAR(int, vgetcpu_mode);
+ DEFINE_VVAR(struct vsyscall_gtod_data, vsyscall_gtod_data) =
+ {
+- .lock = __SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock),
++ .lock = __RAW_SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock),
+ .sysctl_enabled = 1,
+ };
- ret = 0;
- old_incr = timer->it.cpu.incr;
-@@ -999,9 +999,9 @@ static void stop_process_timers(struct s
- struct thread_group_cputimer *cputimer = &sig->cputimer;
+@@ -60,10 +60,10 @@ void update_vsyscall_tz(void)
+ {
unsigned long flags;
-- spin_lock_irqsave(&cputimer->lock, flags);
-+ raw_spin_lock_irqsave(&cputimer->lock, flags);
- cputimer->running = 0;
-- spin_unlock_irqrestore(&cputimer->lock, flags);
-+ raw_spin_unlock_irqrestore(&cputimer->lock, flags);
+- write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
++ raw_write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
+ /* sys_tz has changed */
+ vsyscall_gtod_data.sys_tz = sys_tz;
+- write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
++ raw_write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
}
- static u32 onecputick;
-@@ -1223,7 +1223,7 @@ void posix_cpu_timer_schedule(struct k_i
- /*
- * Now re-arm for the new expiry time.
+ void update_vsyscall(struct timespec *wall_time, struct timespec *wtm,
+@@ -71,7 +71,7 @@ void update_vsyscall(struct timespec *wall_time, struct timespec *wtm,
+ {
+ unsigned long flags;
+
+- write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
++ raw_write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
+ /* copy vsyscall data */
+ vsyscall_gtod_data.clock.vread = clock->vread;
+ vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
+@@ -82,7 +82,7 @@ void update_vsyscall(struct timespec *wall_time, struct timespec *wtm,
+ vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
+ vsyscall_gtod_data.wall_to_monotonic = *wtm;
+ vsyscall_gtod_data.wall_time_coarse = __current_kernel_time();
+- write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
++ raw_write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
+ }
+
+ /* RED-PEN may want to readd seq locking, but then the variable should be
+diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
+index 77c9d86..545c61b 100644
+--- a/arch/x86/kvm/x86.c
++++ b/arch/x86/kvm/x86.c
+@@ -4900,6 +4900,13 @@ int kvm_arch_init(void *opaque)
+ goto out;
+ }
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) {
++ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n");
++ return -EOPNOTSUPP;
++ }
++#endif
++
+ r = kvm_mmu_module_init();
+ if (r)
+ goto out;
+diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
+index 2dbf6bf..b9bc727 100644
+--- a/arch/x86/mm/fault.c
++++ b/arch/x86/mm/fault.c
+@@ -1065,7 +1065,7 @@ do_page_fault(struct pt_regs *regs, unsigned long error_code)
+ * If we're in an interrupt, have no user context or are running
+ * in an atomic region then we must not take the fault:
*/
-- BUG_ON(!irqs_disabled());
-+ BUG_ON_NONRT(!irqs_disabled());
- arm_timer(timer);
- spin_unlock(&p->sighand->siglock);
+- if (unlikely(in_atomic() || !mm)) {
++ if (unlikely(!mm || pagefault_disabled())) {
+ bad_area_nosemaphore(regs, error_code, address);
+ return;
+ }
+diff --git a/arch/x86/mm/highmem_32.c b/arch/x86/mm/highmem_32.c
+index b499626..29dd703 100644
+--- a/arch/x86/mm/highmem_32.c
++++ b/arch/x86/mm/highmem_32.c
+@@ -43,7 +43,7 @@ void *kmap_atomic_prot(struct page *page, pgprot_t prot)
+ type = kmap_atomic_idx_push();
+ idx = type + KM_TYPE_NR*smp_processor_id();
+ vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
+- BUG_ON(!pte_none(*(kmap_pte-idx)));
++ WARN_ON(!pte_none(*(kmap_pte-idx)));
+ set_pte(kmap_pte-idx, mk_pte(page, prot));
-@@ -1290,10 +1290,11 @@ static inline int fastpath_timer_check(s
- sig = tsk->signal;
- if (sig->cputimer.running) {
- struct task_cputime group_sample;
-+ unsigned long flags;
+ return (void *)vaddr;
+diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c
+index 68894fd..96646b3 100644
+--- a/arch/x86/oprofile/nmi_int.c
++++ b/arch/x86/oprofile/nmi_int.c
+@@ -355,10 +355,10 @@ static void nmi_cpu_setup(void *dummy)
+ int cpu = smp_processor_id();
+ struct op_msrs *msrs = &per_cpu(cpu_msrs, cpu);
+ nmi_cpu_save_registers(msrs);
+- spin_lock(&oprofilefs_lock);
++ raw_spin_lock(&oprofilefs_lock);
+ model->setup_ctrs(model, msrs);
+ nmi_cpu_setup_mux(cpu, msrs);
+- spin_unlock(&oprofilefs_lock);
++ raw_spin_unlock(&oprofilefs_lock);
+ per_cpu(saved_lvtpc, cpu) = apic_read(APIC_LVTPC);
+ apic_write(APIC_LVTPC, APIC_DM_NMI);
+ }
+diff --git a/arch/xtensa/include/asm/rwsem.h b/arch/xtensa/include/asm/rwsem.h
+index 249619e..ea7ee21 100644
+--- a/arch/xtensa/include/asm/rwsem.h
++++ b/arch/xtensa/include/asm/rwsem.h
+@@ -27,7 +27,7 @@
+ /*
+ * lock for reading
+ */
+-static inline void __down_read(struct rw_semaphore *sem)
++static inline void __down_read(struct rw_anon_semaphore *sem)
+ {
+ if (atomic_add_return(1,(atomic_t *)(&sem->count)) > 0)
+ smp_wmb();
+@@ -35,7 +35,7 @@ static inline void __down_read(struct rw_semaphore *sem)
+ rwsem_down_read_failed(sem);
+ }
-- spin_lock(&sig->cputimer.lock);
-+ raw_spin_lock_irqsave(&sig->cputimer.lock, flags);
- group_sample = sig->cputimer.cputime;
-- spin_unlock(&sig->cputimer.lock);
-+ raw_spin_unlock_irqrestore(&sig->cputimer.lock, flags);
+-static inline int __down_read_trylock(struct rw_semaphore *sem)
++static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ int tmp;
- if (task_cputime_expired(&group_sample, &sig->cputime_expires))
- return 1;
-@@ -1307,13 +1308,13 @@ static inline int fastpath_timer_check(s
- * already updated our counts. We need to check if any timers fire now.
- * Interrupts are disabled.
+@@ -52,7 +52,7 @@ static inline int __down_read_trylock(struct rw_semaphore *sem)
+ /*
+ * lock for writing
*/
--void run_posix_cpu_timers(struct task_struct *tsk)
-+static void __run_posix_cpu_timers(struct task_struct *tsk)
+-static inline void __down_write(struct rw_semaphore *sem)
++static inline void __down_write(struct rw_anon_semaphore *sem)
{
- LIST_HEAD(firing);
- struct k_itimer *timer, *next;
- unsigned long flags;
+ int tmp;
-- BUG_ON(!irqs_disabled());
-+ BUG_ON_NONRT(!irqs_disabled());
+@@ -64,7 +64,7 @@ static inline void __down_write(struct rw_semaphore *sem)
+ rwsem_down_write_failed(sem);
+ }
- /*
- * The fast path checks that there are no expired thread or thread
-@@ -1371,6 +1372,190 @@ void run_posix_cpu_timers(struct task_st
- }
+-static inline int __down_write_trylock(struct rw_semaphore *sem)
++static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ int tmp;
+
+@@ -77,7 +77,7 @@ static inline int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * unlock after reading
+ */
+-static inline void __up_read(struct rw_semaphore *sem)
++static inline void __up_read(struct rw_anon_semaphore *sem)
+ {
+ int tmp;
+
+@@ -90,7 +90,7 @@ static inline void __up_read(struct rw_semaphore *sem)
+ /*
+ * unlock after writing
+ */
+-static inline void __up_write(struct rw_semaphore *sem)
++static inline void __up_write(struct rw_anon_semaphore *sem)
+ {
+ smp_wmb();
+ if (atomic_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
+@@ -101,7 +101,7 @@ static inline void __up_write(struct rw_semaphore *sem)
+ /*
+ * implement atomic add functionality
+ */
+-static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
++static inline void rwsem_atomic_add(int delta, struct rw_anon_semaphore *sem)
+ {
+ atomic_add(delta, (atomic_t *)(&sem->count));
}
+@@ -109,7 +109,7 @@ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
+ /*
+ * downgrade write lock to read lock
+ */
+-static inline void __downgrade_write(struct rw_semaphore *sem)
++static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ int tmp;
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+#include <linux/kthread.h>
-+#include <linux/cpu.h>
-+DEFINE_PER_CPU(struct task_struct *, posix_timer_task);
-+DEFINE_PER_CPU(struct task_struct *, posix_timer_tasklist);
-+
-+static int posix_cpu_timers_thread(void *data)
-+{
-+ int cpu = (long)data;
-+
-+ BUG_ON(per_cpu(posix_timer_task,cpu) != current);
-+
-+ while (!kthread_should_stop()) {
-+ struct task_struct *tsk = NULL;
-+ struct task_struct *next = NULL;
-+
-+ if (cpu_is_offline(cpu))
-+ goto wait_to_die;
-+
-+ /* grab task list */
-+ raw_local_irq_disable();
-+ tsk = per_cpu(posix_timer_tasklist, cpu);
-+ per_cpu(posix_timer_tasklist, cpu) = NULL;
-+ raw_local_irq_enable();
-+
-+ /* its possible the list is empty, just return */
-+ if (!tsk) {
-+ set_current_state(TASK_INTERRUPTIBLE);
-+ schedule();
-+ __set_current_state(TASK_RUNNING);
-+ continue;
-+ }
-+
-+ /* Process task list */
-+ while (1) {
-+ /* save next */
-+ next = tsk->posix_timer_list;
-+
-+ /* run the task timers, clear its ptr and
-+ * unreference it
-+ */
-+ __run_posix_cpu_timers(tsk);
-+ tsk->posix_timer_list = NULL;
-+ put_task_struct(tsk);
-+
-+ /* check if this is the last on the list */
-+ if (next == tsk)
-+ break;
-+ tsk = next;
-+ }
-+ }
-+ return 0;
-+
-+wait_to_die:
-+ /* Wait for kthread_stop */
-+ set_current_state(TASK_INTERRUPTIBLE);
-+ while (!kthread_should_stop()) {
-+ schedule();
-+ set_current_state(TASK_INTERRUPTIBLE);
-+ }
-+ __set_current_state(TASK_RUNNING);
-+ return 0;
-+}
-+
-+static inline int __fastpath_timer_check(struct task_struct *tsk)
-+{
-+ /* tsk == current, ensure it is safe to use ->signal/sighand */
-+ if (unlikely(tsk->exit_state))
-+ return 0;
-+
-+ if (!task_cputime_zero(&tsk->cputime_expires))
-+ return 1;
-+
-+ if (!task_cputime_zero(&tsk->signal->cputime_expires))
-+ return 1;
-+
-+ return 0;
-+}
-+
-+void run_posix_cpu_timers(struct task_struct *tsk)
-+{
-+ unsigned long cpu = smp_processor_id();
-+ struct task_struct *tasklist;
-+
-+ BUG_ON(!irqs_disabled());
-+ if(!per_cpu(posix_timer_task, cpu))
-+ return;
-+ /* get per-cpu references */
-+ tasklist = per_cpu(posix_timer_tasklist, cpu);
-+
-+ /* check to see if we're already queued */
-+ if (!tsk->posix_timer_list && __fastpath_timer_check(tsk)) {
-+ get_task_struct(tsk);
-+ if (tasklist) {
-+ tsk->posix_timer_list = tasklist;
-+ } else {
-+ /*
-+ * The list is terminated by a self-pointing
-+ * task_struct
-+ */
-+ tsk->posix_timer_list = tsk;
-+ }
-+ per_cpu(posix_timer_tasklist, cpu) = tsk;
-+
-+ wake_up_process(per_cpu(posix_timer_task, cpu));
-+ }
-+}
-+
-+/*
-+ * posix_cpu_thread_call - callback that gets triggered when a CPU is added.
-+ * Here we can start up the necessary migration thread for the new CPU.
-+ */
-+static int posix_cpu_thread_call(struct notifier_block *nfb,
-+ unsigned long action, void *hcpu)
-+{
-+ int cpu = (long)hcpu;
-+ struct task_struct *p;
-+ struct sched_param param;
-+
-+ switch (action) {
-+ case CPU_UP_PREPARE:
-+ p = kthread_create(posix_cpu_timers_thread, hcpu,
-+ "posixcputmr/%d",cpu);
-+ if (IS_ERR(p))
-+ return NOTIFY_BAD;
-+ p->flags |= PF_NOFREEZE;
-+ kthread_bind(p, cpu);
-+ /* Must be high prio to avoid getting starved */
-+ param.sched_priority = MAX_RT_PRIO-1;
-+ sched_setscheduler(p, SCHED_FIFO, ¶m);
-+ per_cpu(posix_timer_task,cpu) = p;
-+ break;
-+ case CPU_ONLINE:
-+ /* Strictly unneccessary, as first user will wake it. */
-+ wake_up_process(per_cpu(posix_timer_task,cpu));
-+ break;
-+#ifdef CONFIG_HOTPLUG_CPU
-+ case CPU_UP_CANCELED:
-+ /* Unbind it from offline cpu so it can run. Fall thru. */
-+ kthread_bind(per_cpu(posix_timer_task,cpu),
-+ any_online_cpu(cpu_online_map));
-+ kthread_stop(per_cpu(posix_timer_task,cpu));
-+ per_cpu(posix_timer_task,cpu) = NULL;
-+ break;
-+ case CPU_DEAD:
-+ kthread_stop(per_cpu(posix_timer_task,cpu));
-+ per_cpu(posix_timer_task,cpu) = NULL;
-+ break;
-+#endif
-+ }
-+ return NOTIFY_OK;
-+}
-+
-+/* Register at highest priority so that task migration (migrate_all_tasks)
-+ * happens before everything else.
-+ */
-+static struct notifier_block __devinitdata posix_cpu_thread_notifier = {
-+ .notifier_call = posix_cpu_thread_call,
-+ .priority = 10
-+};
-+
-+static int __init posix_cpu_thread_init(void)
-+{
-+ void *hcpu = (void *)(long)smp_processor_id();
-+ /* Start one for boot CPU. */
-+ unsigned long cpu;
-+
-+ /* init the per-cpu posix_timer_tasklets */
-+ for_each_cpu_mask(cpu, cpu_possible_map)
-+ per_cpu(posix_timer_tasklist, cpu) = NULL;
-+
-+ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_UP_PREPARE, hcpu);
-+ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_ONLINE, hcpu);
-+ register_cpu_notifier(&posix_cpu_thread_notifier);
-+ return 0;
-+}
-+early_initcall(posix_cpu_thread_init);
-+#else /* CONFIG_PREEMPT_RT_BASE */
-+void run_posix_cpu_timers(struct task_struct *tsk)
-+{
-+ __run_posix_cpu_timers(tsk);
-+}
-+#endif /* CONFIG_PREEMPT_RT_BASE */
-+
+@@ -122,7 +122,7 @@ static inline void __downgrade_write(struct rw_semaphore *sem)
/*
- * Set one of the process-wide special case CPU timers or RLIMIT_CPU.
- * The tsk->sighand->siglock must be held by the caller.
-Index: linux-3.0/kernel/sched_stats.h
-===================================================================
---- linux-3.0.orig/kernel/sched_stats.h
-+++ linux-3.0/kernel/sched_stats.h
-@@ -282,10 +282,10 @@ static inline void account_group_user_ti
- if (!cputimer->running)
- return;
-
-- spin_lock(&cputimer->lock);
-+ raw_spin_lock(&cputimer->lock);
- cputimer->cputime.utime =
- cputime_add(cputimer->cputime.utime, cputime);
-- spin_unlock(&cputimer->lock);
-+ raw_spin_unlock(&cputimer->lock);
+ * implement exchange and add functionality
+ */
+-static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
++static inline int rwsem_atomic_update(int delta, struct rw_anon_semaphore *sem)
+ {
+ smp_mb();
+ return atomic_add_return(delta, (atomic_t *)(&sem->count));
+diff --git a/arch/xtensa/kernel/process.c b/arch/xtensa/kernel/process.c
+index e3558b9..005dd36 100644
+--- a/arch/xtensa/kernel/process.c
++++ b/arch/xtensa/kernel/process.c
+@@ -113,9 +113,7 @@ void cpu_idle(void)
+ while (1) {
+ while (!need_resched())
+ platform_idle();
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ }
}
- /**
-@@ -306,10 +306,10 @@ static inline void account_group_system_
- if (!cputimer->running)
+diff --git a/arch/xtensa/mm/fault.c b/arch/xtensa/mm/fault.c
+index e367e30..8f3f52a 100644
+--- a/arch/xtensa/mm/fault.c
++++ b/arch/xtensa/mm/fault.c
+@@ -57,7 +57,7 @@ void do_page_fault(struct pt_regs *regs)
+ /* If we're in an interrupt or have no user
+ * context, we must not take the fault..
+ */
+- if (in_atomic() || !mm) {
++ if (!mm || pagefault_disabled()) {
+ bad_page_fault(regs, address, SIGSEGV);
return;
+ }
+diff --git a/block/blk-core.c b/block/blk-core.c
+index 847d04e..69ef3f7 100644
+--- a/block/blk-core.c
++++ b/block/blk-core.c
+@@ -236,7 +236,7 @@ EXPORT_SYMBOL(blk_delay_queue);
+ **/
+ void blk_start_queue(struct request_queue *q)
+ {
+- WARN_ON(!irqs_disabled());
++ WARN_ON_NONRT(!irqs_disabled());
-- spin_lock(&cputimer->lock);
-+ raw_spin_lock(&cputimer->lock);
- cputimer->cputime.stime =
- cputime_add(cputimer->cputime.stime, cputime);
-- spin_unlock(&cputimer->lock);
-+ raw_spin_unlock(&cputimer->lock);
- }
-
- /**
-@@ -330,7 +330,7 @@ static inline void account_group_exec_ru
- if (!cputimer->running)
+ queue_flag_clear(QUEUE_FLAG_STOPPED, q);
+ __blk_run_queue(q);
+@@ -301,7 +301,11 @@ void __blk_run_queue(struct request_queue *q)
+ {
+ if (unlikely(blk_queue_stopped(q)))
return;
-
-- spin_lock(&cputimer->lock);
-+ raw_spin_lock(&cputimer->lock);
- cputimer->cputime.sum_exec_runtime += ns;
-- spin_unlock(&cputimer->lock);
-+ raw_spin_unlock(&cputimer->lock);
+-
++ /*
++ * q->request_fn() can drop q->queue_lock and reenable
++ * interrupts, but must return with q->queue_lock held and
++ * interrupts disabled.
++ */
+ q->request_fn(q);
}
-Index: linux-3.0/include/linux/semaphore.h
-===================================================================
---- linux-3.0.orig/include/linux/semaphore.h
-+++ linux-3.0/include/linux/semaphore.h
-@@ -14,14 +14,14 @@
-
- /* Please don't access any members of this structure directly */
- struct semaphore {
-- spinlock_t lock;
-+ raw_spinlock_t lock;
- unsigned int count;
- struct list_head wait_list;
- };
-
- #define __SEMAPHORE_INITIALIZER(name, n) \
- { \
-- .lock = __SPIN_LOCK_UNLOCKED((name).lock), \
-+ .lock = __RAW_SPIN_LOCK_UNLOCKED((name).lock), \
- .count = n, \
- .wait_list = LIST_HEAD_INIT((name).wait_list), \
+ EXPORT_SYMBOL(__blk_run_queue);
+@@ -2669,11 +2673,11 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth,
+ * this lock).
+ */
+ if (from_schedule) {
+- spin_unlock(q->queue_lock);
++ spin_unlock_irq(q->queue_lock);
+ blk_run_queue_async(q);
+ } else {
+ __blk_run_queue(q);
+- spin_unlock(q->queue_lock);
++ spin_unlock_irq(q->queue_lock);
+ }
+
}
-Index: linux-3.0/kernel/semaphore.c
-===================================================================
---- linux-3.0.orig/kernel/semaphore.c
-+++ linux-3.0/kernel/semaphore.c
-@@ -54,12 +54,12 @@ void down(struct semaphore *sem)
+@@ -2699,7 +2703,6 @@ static void flush_plug_callbacks(struct blk_plug *plug)
+ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
{
- unsigned long flags;
+ struct request_queue *q;
+- unsigned long flags;
+ struct request *rq;
+ LIST_HEAD(list);
+ unsigned int depth;
+@@ -2720,11 +2723,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
+ q = NULL;
+ depth = 0;
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- if (likely(sem->count > 0))
- sem->count--;
- else
- __down(sem);
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
+- /*
+- * Save and disable interrupts here, to avoid doing it for every
+- * queue lock we have to take.
+- */
+- local_irq_save(flags);
+ while (!list_empty(&list)) {
+ rq = list_entry_rq(list.next);
+ list_del_init(&rq->queuelist);
+@@ -2737,7 +2735,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
+ queue_unplugged(q, depth, from_schedule);
+ q = rq->q;
+ depth = 0;
+- spin_lock(q->queue_lock);
++ spin_lock_irq(q->queue_lock);
+ }
+ /*
+ * rq is already accounted, so use raw insert
+@@ -2755,8 +2753,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
+ */
+ if (q)
+ queue_unplugged(q, depth, from_schedule);
+-
+- local_irq_restore(flags);
}
- EXPORT_SYMBOL(down);
-@@ -77,12 +77,12 @@ int down_interruptible(struct semaphore
- unsigned long flags;
- int result = 0;
+ void blk_finish_plug(struct blk_plug *plug)
+diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
+index 431ab11..3e05314 100644
+--- a/drivers/acpi/processor_idle.c
++++ b/drivers/acpi/processor_idle.c
+@@ -852,7 +852,7 @@ static int acpi_idle_enter_simple(struct cpuidle_device *dev,
+ }
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- if (likely(sem->count > 0))
- sem->count--;
- else
- result = __down_interruptible(sem);
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
+ static int c3_cpu_count;
+-static DEFINE_SPINLOCK(c3_lock);
++static DEFINE_RAW_SPINLOCK(c3_lock);
- return result;
- }
-@@ -103,12 +103,12 @@ int down_killable(struct semaphore *sem)
+ /**
+ * acpi_idle_enter_bm - enters C3 with proper BM handling
+@@ -930,12 +930,12 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev,
+ * without doing anything.
+ */
+ if (pr->flags.bm_check && pr->flags.bm_control) {
+- spin_lock(&c3_lock);
++ raw_spin_lock(&c3_lock);
+ c3_cpu_count++;
+ /* Disable bus master arbitration when all CPUs are in C3 */
+ if (c3_cpu_count == num_online_cpus())
+ acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 1);
+- spin_unlock(&c3_lock);
++ raw_spin_unlock(&c3_lock);
+ } else if (!pr->flags.bm_check) {
+ ACPI_FLUSH_CPU_CACHE();
+ }
+@@ -944,10 +944,10 @@ static int acpi_idle_enter_bm(struct cpuidle_device *dev,
+
+ /* Re-enable bus master arbitration */
+ if (pr->flags.bm_check && pr->flags.bm_control) {
+- spin_lock(&c3_lock);
++ raw_spin_lock(&c3_lock);
+ acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 0);
+ c3_cpu_count--;
+- spin_unlock(&c3_lock);
++ raw_spin_unlock(&c3_lock);
+ }
+ kt2 = ktime_get_real();
+ idle_time_ns = ktime_to_ns(ktime_sub(kt2, kt1));
+diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
+index b1b926c..5ea190f 100644
+--- a/drivers/ata/libata-sff.c
++++ b/drivers/ata/libata-sff.c
+@@ -677,9 +677,9 @@ unsigned int ata_sff_data_xfer_noirq(struct ata_device *dev, unsigned char *buf,
unsigned long flags;
- int result = 0;
+ unsigned int consumed;
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- if (likely(sem->count > 0))
- sem->count--;
- else
- result = __down_killable(sem);
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ consumed = ata_sff_data_xfer(dev, buf, buflen, rw);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
- return result;
+ return consumed;
}
-@@ -132,11 +132,11 @@ int down_trylock(struct semaphore *sem)
- unsigned long flags;
- int count;
+@@ -718,7 +718,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc)
+ unsigned long flags;
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- count = sem->count - 1;
- if (likely(count >= 0))
- sem->count = count;
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
+ /* FIXME: use a bounce buffer */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ buf = kmap_atomic(page, KM_IRQ0);
- return (count < 0);
- }
-@@ -157,12 +157,12 @@ int down_timeout(struct semaphore *sem,
- unsigned long flags;
- int result = 0;
+ /* do the actual data transfer */
+@@ -726,7 +726,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc)
+ do_write);
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- if (likely(sem->count > 0))
- sem->count--;
- else
- result = __down_timeout(sem, jiffies);
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
+ kunmap_atomic(buf, KM_IRQ0);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ } else {
+ buf = page_address(page);
+ ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size,
+@@ -863,7 +863,7 @@ next_sg:
+ unsigned long flags;
- return result;
- }
-@@ -179,12 +179,12 @@ void up(struct semaphore *sem)
- {
- unsigned long flags;
+ /* FIXME: use bounce buffer */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ buf = kmap_atomic(page, KM_IRQ0);
-- spin_lock_irqsave(&sem->lock, flags);
-+ raw_spin_lock_irqsave(&sem->lock, flags);
- if (likely(list_empty(&sem->wait_list)))
- sem->count++;
- else
- __up(sem);
-- spin_unlock_irqrestore(&sem->lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->lock, flags);
- }
- EXPORT_SYMBOL(up);
+ /* do the actual data transfer */
+@@ -871,7 +871,7 @@ next_sg:
+ count, rw);
-@@ -217,9 +217,9 @@ static inline int __sched __down_common(
- if (timeout <= 0)
- goto timed_out;
- __set_task_state(task, state);
-- spin_unlock_irq(&sem->lock);
-+ raw_spin_unlock_irq(&sem->lock);
- timeout = schedule_timeout(timeout);
-- spin_lock_irq(&sem->lock);
-+ raw_spin_lock_irq(&sem->lock);
- if (waiter.up)
- return 0;
- }
-Index: linux-3.0/include/linux/rwsem-spinlock.h
-===================================================================
---- linux-3.0.orig/include/linux/rwsem-spinlock.h
-+++ linux-3.0/include/linux/rwsem-spinlock.h
-@@ -20,26 +20,42 @@
- * - if activity is -1 then there is one active writer
- * - if wait_list is not empty, then there are processes waiting for the semaphore
- */
-+struct rw_anon_semaphore {
-+ __s32 activity;
-+ raw_spinlock_t wait_lock;
-+ struct list_head wait_list;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+};
-+
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+/*
-+ * Non preempt-rt implementation of rw_semaphore. Same as above, but
-+ * restricted vs. ownership. i.e. ownerless locked state and non owner
-+ * release not allowed.
-+ */
- struct rw_semaphore {
- __s32 activity;
-- spinlock_t wait_lock;
-+ raw_spinlock_t wait_lock;
- struct list_head wait_list;
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
- struct lockdep_map dep_map;
- #endif
+ kunmap_atomic(buf, KM_IRQ0);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ } else {
+ buf = page_address(page);
+ consumed = ap->ops->sff_data_xfer(dev, buf + offset,
+diff --git a/drivers/char/random.c b/drivers/char/random.c
+index c35a785..52221d2 100644
+--- a/drivers/char/random.c
++++ b/drivers/char/random.c
+@@ -433,7 +433,7 @@ static struct entropy_store input_pool = {
+ .poolinfo = &poolinfo_table[0],
+ .name = "input",
+ .limit = 1,
+- .lock = __SPIN_LOCK_UNLOCKED(&input_pool.lock),
++ .lock = __SPIN_LOCK_UNLOCKED(input_pool.lock),
+ .pool = input_pool_data
};
-+#endif /* PREEMPT_RT_FULL */
- #define RWSEM_UNLOCKED_VALUE 0x00000000
+@@ -442,7 +442,7 @@ static struct entropy_store blocking_pool = {
+ .name = "blocking",
+ .limit = 1,
+ .pull = &input_pool,
+- .lock = __SPIN_LOCK_UNLOCKED(&blocking_pool.lock),
++ .lock = __SPIN_LOCK_UNLOCKED(blocking_pool.lock),
+ .pool = blocking_pool_data
+ };
--extern void __down_read(struct rw_semaphore *sem);
--extern int __down_read_trylock(struct rw_semaphore *sem);
--extern void __down_write(struct rw_semaphore *sem);
--extern void __down_write_nested(struct rw_semaphore *sem, int subclass);
--extern int __down_write_trylock(struct rw_semaphore *sem);
--extern void __up_read(struct rw_semaphore *sem);
--extern void __up_write(struct rw_semaphore *sem);
--extern void __downgrade_write(struct rw_semaphore *sem);
--extern int rwsem_is_locked(struct rw_semaphore *sem);
-+extern void __down_read(struct rw_anon_semaphore *sem);
-+extern int __down_read_trylock(struct rw_anon_semaphore *sem);
-+extern void __down_write(struct rw_anon_semaphore *sem);
-+extern void __down_write_nested(struct rw_anon_semaphore *sem, int subclass);
-+extern int __down_write_trylock(struct rw_anon_semaphore *sem);
-+extern void __up_read(struct rw_anon_semaphore *sem);
-+extern void __up_write(struct rw_anon_semaphore *sem);
-+extern void __downgrade_write(struct rw_anon_semaphore *sem);
-+extern int anon_rwsem_is_locked(struct rw_anon_semaphore *sem);
+@@ -450,7 +450,7 @@ static struct entropy_store nonblocking_pool = {
+ .poolinfo = &poolinfo_table[1],
+ .name = "nonblocking",
+ .pull = &input_pool,
+- .lock = __SPIN_LOCK_UNLOCKED(&nonblocking_pool.lock),
++ .lock = __SPIN_LOCK_UNLOCKED(nonblocking_pool.lock),
+ .pool = nonblocking_pool_data
+ };
- #endif /* __KERNEL__ */
- #endif /* _LINUX_RWSEM_SPINLOCK_H */
-Index: linux-3.0/include/linux/rwsem.h
-===================================================================
---- linux-3.0.orig/include/linux/rwsem.h
-+++ linux-3.0/include/linux/rwsem.h
-@@ -17,37 +17,50 @@
- #include <asm/system.h>
- #include <asm/atomic.h>
+@@ -633,8 +633,11 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num)
+ preempt_disable();
+ /* if over the trickle threshold, use only 1 in 4096 samples */
+ if (input_pool.entropy_count > trickle_thresh &&
+- ((__this_cpu_inc_return(trickle_count) - 1) & 0xfff))
+- goto out;
++ ((__this_cpu_inc_return(trickle_count) - 1) & 0xfff)) {
++ preempt_enable();
++ return;
++ }
++ preempt_enable();
-+struct rw_anon_semaphore;
- struct rw_semaphore;
+ sample.jiffies = jiffies;
+ sample.cycles = get_cycles();
+@@ -676,8 +679,6 @@ static void add_timer_randomness(struct timer_rand_state *state, unsigned num)
+ credit_entropy_bits(&input_pool,
+ min_t(int, fls(delta>>1), 11));
+ }
+-out:
+- preempt_enable();
+ }
- #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK
- #include <linux/rwsem-spinlock.h> /* use a generic implementation */
--#else
-+#else /* RWSEM_GENERIC_SPINLOCK */
-+
- /* All arch specific implementations share the same struct */
--struct rw_semaphore {
-+struct rw_anon_semaphore {
- long count;
-- spinlock_t wait_lock;
-+ raw_spinlock_t wait_lock;
- struct list_head wait_list;
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
- struct lockdep_map dep_map;
- #endif
+ void add_input_randomness(unsigned int type, unsigned int code,
+diff --git a/drivers/clocksource/tcb_clksrc.c b/drivers/clocksource/tcb_clksrc.c
+index 79c47e8..8976b3d 100644
+--- a/drivers/clocksource/tcb_clksrc.c
++++ b/drivers/clocksource/tcb_clksrc.c
+@@ -21,8 +21,7 @@
+ * resolution better than 200 nsec).
+ *
+ * - The third channel may be used to provide a 16-bit clockevent
+- * source, used in either periodic or oneshot mode. This runs
+- * at 32 KiHZ, and can handle delays of up to two seconds.
++ * source, used in either periodic or oneshot mode.
+ *
+ * A boot clocksource and clockevent source are also currently needed,
+ * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so
+@@ -68,6 +67,7 @@ static struct clocksource clksrc = {
+ struct tc_clkevt_device {
+ struct clock_event_device clkevt;
+ struct clk *clk;
++ u32 freq;
+ void __iomem *regs;
};
--extern struct rw_semaphore *rwsem_down_read_failed(struct rw_semaphore *sem);
--extern struct rw_semaphore *rwsem_down_write_failed(struct rw_semaphore *sem);
--extern struct rw_semaphore *rwsem_wake(struct rw_semaphore *);
--extern struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem);
-+extern struct rw_anon_semaphore *rwsem_down_read_failed(struct rw_anon_semaphore *sem);
-+extern struct rw_anon_semaphore *rwsem_down_write_failed(struct rw_anon_semaphore *sem);
-+extern struct rw_anon_semaphore *rwsem_wake(struct rw_anon_semaphore *);
-+extern struct rw_anon_semaphore *rwsem_downgrade_wake(struct rw_anon_semaphore *sem);
+@@ -76,13 +76,6 @@ static struct tc_clkevt_device *to_tc_clkevt(struct clock_event_device *clkevt)
+ return container_of(clkevt, struct tc_clkevt_device, clkevt);
+ }
- /* Include the arch specific part */
- #include <asm/rwsem.h>
+-/* For now, we always use the 32K clock ... this optimizes for NO_HZ,
+- * because using one of the divided clocks would usually mean the
+- * tick rate can never be less than several dozen Hz (vs 0.5 Hz).
+- *
+- * A divided clock could be good for high resolution timers, since
+- * 30.5 usec resolution can seem "low".
+- */
+ static u32 timer_clock;
- /* In all implementations count != 0 means locked */
--static inline int rwsem_is_locked(struct rw_semaphore *sem)
-+static inline int anon_rwsem_is_locked(struct rw_anon_semaphore *sem)
- {
- return sem->count != 0;
- }
+ static void tc_mode(enum clock_event_mode m, struct clock_event_device *d)
+@@ -105,11 +98,12 @@ static void tc_mode(enum clock_event_mode m, struct clock_event_device *d)
+ case CLOCK_EVT_MODE_PERIODIC:
+ clk_enable(tcd->clk);
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+struct rw_semaphore {
-+ long count;
-+ raw_spinlock_t wait_lock;
-+ struct list_head wait_list;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
+- /* slow clock, count up to RC, then irq and restart */
++ /* count up to RC, then irq and restart */
+ __raw_writel(timer_clock
+ | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO,
+ regs + ATMEL_TC_REG(2, CMR));
+- __raw_writel((32768 + HZ/2) / HZ, tcaddr + ATMEL_TC_REG(2, RC));
++ __raw_writel((tcd->freq + HZ/2)/HZ,
++ tcaddr + ATMEL_TC_REG(2, RC));
+
+ /* Enable clock and interrupts on RC compare */
+ __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER));
+@@ -122,7 +116,7 @@ static void tc_mode(enum clock_event_mode m, struct clock_event_device *d)
+ case CLOCK_EVT_MODE_ONESHOT:
+ clk_enable(tcd->clk);
+
+- /* slow clock, count up to RC, then irq and stop */
++ /* count up to RC, then irq and stop */
+ __raw_writel(timer_clock | ATMEL_TC_CPCSTOP
+ | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO,
+ regs + ATMEL_TC_REG(2, CMR));
+@@ -152,8 +146,12 @@ static struct tc_clkevt_device clkevt = {
+ .features = CLOCK_EVT_FEAT_PERIODIC
+ | CLOCK_EVT_FEAT_ONESHOT,
+ .shift = 32,
++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
+ /* Should be lower than at91rm9200's system timer */
+ .rating = 125,
++#else
++ .rating = 200,
+#endif
-+};
- #endif
+ .set_next_event = tc_next_event,
+ .set_mode = tc_mode,
+ },
+@@ -179,8 +177,9 @@ static struct irqaction tc_irqaction = {
+ .handler = ch2_irq,
+ };
-+#endif /* !RWSEM_GENERIC_SPINLOCK */
-+
- /* Common initializer macros and functions */
+-static void __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx)
++static void __init setup_clkevents(struct atmel_tc *tc, int divisor_idx)
+ {
++ unsigned divisor = atmel_tc_divisors[divisor_idx];
+ struct clk *t2_clk = tc->clk[2];
+ int irq = tc->irq[2];
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
-@@ -56,57 +69,59 @@ static inline int rwsem_is_locked(struct
- # define __RWSEM_DEP_MAP_INIT(lockname)
- #endif
+@@ -188,11 +187,17 @@ static void __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx)
+ clkevt.clk = t2_clk;
+ tc_irqaction.dev_id = &clkevt;
--#define __RWSEM_INITIALIZER(name) \
-- { RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED(name.wait_lock), \
-- LIST_HEAD_INIT((name).wait_list) __RWSEM_DEP_MAP_INIT(name) }
-+#define __RWSEM_ANON_INITIALIZER(name) \
-+ { RWSEM_UNLOCKED_VALUE, \
-+ __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
-+ LIST_HEAD_INIT((name).wait_list) \
-+ __RWSEM_DEP_MAP_INIT(name) }
+- timer_clock = clk32k_divisor_idx;
++ timer_clock = divisor_idx;
--#define DECLARE_RWSEM(name) \
-- struct rw_semaphore name = __RWSEM_INITIALIZER(name)
-+#define DECLARE_ANON_RWSEM(name) \
-+ struct rw_anon_semaphore name = __RWSEM_INITIALIZER(name)
+- clkevt.clkevt.mult = div_sc(32768, NSEC_PER_SEC, clkevt.clkevt.shift);
+- clkevt.clkevt.max_delta_ns
+- = clockevent_delta2ns(0xffff, &clkevt.clkevt);
++ if (!divisor)
++ clkevt.freq = 32768;
++ else
++ clkevt.freq = clk_get_rate(t2_clk)/divisor;
++
++ clkevt.clkevt.mult = div_sc(clkevt.freq, NSEC_PER_SEC,
++ clkevt.clkevt.shift);
++ clkevt.clkevt.max_delta_ns =
++ clockevent_delta2ns(0xffff, &clkevt.clkevt);
+ clkevt.clkevt.min_delta_ns = clockevent_delta2ns(1, &clkevt.clkevt) + 1;
+ clkevt.clkevt.cpumask = cpumask_of(0);
--extern void __init_rwsem(struct rw_semaphore *sem, const char *name,
-- struct lock_class_key *key);
-+extern void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
-+ struct lock_class_key *key);
+@@ -295,8 +300,11 @@ static int __init tcb_clksrc_init(void)
+ clocksource_register(&clksrc);
--#define init_rwsem(sem) \
-+#define init_anon_rwsem(sem) \
- do { \
- static struct lock_class_key __key; \
- \
-- __init_rwsem((sem), #sem, &__key); \
-+ __init_anon_rwsem((sem), #sem, &__key); \
- } while (0)
+ /* channel 2: periodic and oneshot timer support */
++#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
+ setup_clkevents(tc, clk32k_divisor_idx);
+-
++#else
++ setup_clkevents(tc, best_divisor_idx);
++#endif
+ return 0;
+ }
+ arch_initcall(tcb_clksrc_init);
+diff --git a/drivers/dca/dca-core.c b/drivers/dca/dca-core.c
+index 4abd089..25ec0bb 100644
+--- a/drivers/dca/dca-core.c
++++ b/drivers/dca/dca-core.c
+@@ -35,7 +35,7 @@ MODULE_VERSION(DCA_VERSION);
+ MODULE_LICENSE("GPL");
+ MODULE_AUTHOR("Intel Corporation");
- /*
- * lock for reading
- */
--extern void down_read(struct rw_semaphore *sem);
-+extern void anon_down_read(struct rw_anon_semaphore *sem);
+-static DEFINE_SPINLOCK(dca_lock);
++static DEFINE_RAW_SPINLOCK(dca_lock);
- /*
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
--extern int down_read_trylock(struct rw_semaphore *sem);
-+extern int anon_down_read_trylock(struct rw_anon_semaphore *sem);
+ static LIST_HEAD(dca_domains);
- /*
- * lock for writing
- */
--extern void down_write(struct rw_semaphore *sem);
-+extern void anon_down_write(struct rw_anon_semaphore *sem);
+@@ -101,10 +101,10 @@ static void unregister_dca_providers(void)
- /*
- * trylock for writing -- returns 1 if successful, 0 if contention
- */
--extern int down_write_trylock(struct rw_semaphore *sem);
-+extern int anon_down_write_trylock(struct rw_anon_semaphore *sem);
+ INIT_LIST_HEAD(&unregistered_providers);
- /*
- * release a read lock
- */
--extern void up_read(struct rw_semaphore *sem);
-+extern void anon_up_read(struct rw_anon_semaphore *sem);
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
- /*
- * release a write lock
- */
--extern void up_write(struct rw_semaphore *sem);
-+extern void anon_up_write(struct rw_anon_semaphore *sem);
+ if (list_empty(&dca_domains)) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return;
+ }
- /*
- * downgrade write lock to read lock
- */
--extern void downgrade_write(struct rw_semaphore *sem);
-+extern void anon_downgrade_write(struct rw_anon_semaphore *sem);
+@@ -116,7 +116,7 @@ static void unregister_dca_providers(void)
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
- /*
-@@ -122,21 +137,101 @@ extern void downgrade_write(struct rw_se
- * lockdep_set_class() at lock initialization time.
- * See Documentation/lockdep-design.txt for more details.)
- */
--extern void down_read_nested(struct rw_semaphore *sem, int subclass);
--extern void down_write_nested(struct rw_semaphore *sem, int subclass);
-+extern void anon_down_read_nested(struct rw_anon_semaphore *sem, int subclass);
-+extern void anon_down_write_nested(struct rw_anon_semaphore *sem, int subclass);
- /*
- * Take/release a lock when not the owner will release it.
- *
- * [ This API should be avoided as much as possible - the
- * proper abstraction for this case is completions. ]
- */
--extern void down_read_non_owner(struct rw_semaphore *sem);
--extern void up_read_non_owner(struct rw_semaphore *sem);
-+extern void anon_down_read_non_owner(struct rw_anon_semaphore *sem);
-+extern void anon_up_read_non_owner(struct rw_anon_semaphore *sem);
- #else
--# define down_read_nested(sem, subclass) down_read(sem)
--# define down_write_nested(sem, subclass) down_write(sem)
--# define down_read_non_owner(sem) down_read(sem)
--# define up_read_non_owner(sem) up_read(sem)
-+# define anon_down_read_nested(sem, subclass) anon_down_read(sem)
-+# define anon_down_write_nested(sem, subclass) anon_down_write(sem)
-+# define anon_down_read_non_owner(sem) anon_down_read(sem)
-+# define anon_up_read_non_owner(sem) anon_up_read(sem)
- #endif
+ dca_free_domain(domain);
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+#include <linux/rwsem_rt.h>
-+#else /* PREEMPT_RT_FULL */
-+/*
-+ * Non preempt-rt implementations
-+ */
-+#define __RWSEM_INITIALIZER(name) \
-+ { RWSEM_UNLOCKED_VALUE, \
-+ __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
-+ LIST_HEAD_INIT((name).wait_list) \
-+ __RWSEM_DEP_MAP_INIT(name) }
-+
-+#define DECLARE_RWSEM(name) \
-+ struct rw_semaphore name = __RWSEM_INITIALIZER(name)
-+
-+static inline void __init_rwsem(struct rw_semaphore *sem, const char *name,
-+ struct lock_class_key *key)
-+{
-+ __init_anon_rwsem((struct rw_anon_semaphore *)sem, name, key);
-+}
-+
-+#define init_rwsem(sem) \
-+do { \
-+ static struct lock_class_key __key; \
-+ \
-+ __init_rwsem((sem), #sem, &__key); \
-+} while (0)
-+
-+static inline void down_read(struct rw_semaphore *sem)
-+{
-+ anon_down_read((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline int down_read_trylock(struct rw_semaphore *sem)
-+{
-+ return anon_down_read_trylock((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline void down_write(struct rw_semaphore *sem)
-+{
-+ anon_down_write((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline int down_write_trylock(struct rw_semaphore *sem)
-+{
-+ return anon_down_write_trylock((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline void up_read(struct rw_semaphore *sem)
-+{
-+ anon_up_read((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline void up_write(struct rw_semaphore *sem)
-+{
-+ anon_up_write((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline void downgrade_write(struct rw_semaphore *sem)
-+{
-+ anon_downgrade_write((struct rw_anon_semaphore *)sem);
-+}
-+
-+static inline void down_read_nested(struct rw_semaphore *sem, int subclass)
-+{
-+ return anon_down_read_nested((struct rw_anon_semaphore *)sem, subclass);
-+}
-+
-+static inline void down_write_nested(struct rw_semaphore *sem, int subclass)
-+{
-+ anon_down_write_nested((struct rw_anon_semaphore *)sem, subclass);
-+}
-+
-+static inline int rwsem_is_locked(struct rw_semaphore *sem)
-+{
-+ return anon_rwsem_is_locked((struct rw_anon_semaphore *)sem);
-+}
-+#endif /* !PREEMPT_RT_FULL */
-+
- #endif /* _LINUX_RWSEM_H */
-+
-Index: linux-3.0/lib/rwsem-spinlock.c
-===================================================================
---- linux-3.0.orig/lib/rwsem-spinlock.c
-+++ linux-3.0/lib/rwsem-spinlock.c
-@@ -17,24 +17,24 @@ struct rwsem_waiter {
- #define RWSEM_WAITING_FOR_WRITE 0x00000002
- };
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
--int rwsem_is_locked(struct rw_semaphore *sem)
-+int anon_rwsem_is_locked(struct rw_anon_semaphore *sem)
- {
- int ret = 1;
- unsigned long flags;
+ list_for_each_entry_safe(dca, _dca, &unregistered_providers, node) {
+ dca_sysfs_remove_provider(dca);
+@@ -144,13 +144,8 @@ static struct dca_domain *dca_get_domain(struct device *dev)
+ domain = dca_find_domain(rc);
-- if (spin_trylock_irqsave(&sem->wait_lock, flags)) {
-+ if (raw_spin_trylock_irqsave(&sem->wait_lock, flags)) {
- ret = (sem->activity != 0);
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ if (!domain) {
+- if (dca_provider_ioat_ver_3_0(dev) && !list_empty(&dca_domains)) {
++ if (dca_provider_ioat_ver_3_0(dev) && !list_empty(&dca_domains))
+ dca_providers_blocked = 1;
+- } else {
+- domain = dca_allocate_domain(rc);
+- if (domain)
+- list_add(&domain->node, &dca_domains);
+- }
}
- return ret;
- }
--EXPORT_SYMBOL(rwsem_is_locked);
-+EXPORT_SYMBOL(anon_rwsem_is_locked);
-
- /*
- * initialise the semaphore
- */
--void __init_rwsem(struct rw_semaphore *sem, const char *name,
-- struct lock_class_key *key)
-+void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
-+ struct lock_class_key *key)
- {
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
- /*
-@@ -44,10 +44,10 @@ void __init_rwsem(struct rw_semaphore *s
- lockdep_init_map(&sem->dep_map, name, key, 0);
- #endif
- sem->activity = 0;
-- spin_lock_init(&sem->wait_lock);
-+ raw_spin_lock_init(&sem->wait_lock);
- INIT_LIST_HEAD(&sem->wait_list);
- }
--EXPORT_SYMBOL(__init_rwsem);
-+EXPORT_SYMBOL(__init_anon_rwsem);
- /*
- * handle the lock release when processes blocked on it that can now run
-@@ -58,8 +58,8 @@ EXPORT_SYMBOL(__init_rwsem);
- * - woken process blocks are discarded from the list after having task zeroed
- * - writers are only woken if wakewrite is non-zero
- */
--static inline struct rw_semaphore *
--__rwsem_do_wake(struct rw_semaphore *sem, int wakewrite)
-+static inline struct rw_anon_semaphore *
-+__rwsem_do_wake(struct rw_anon_semaphore *sem, int wakewrite)
- {
- struct rwsem_waiter *waiter;
- struct task_struct *tsk;
-@@ -117,8 +117,8 @@ __rwsem_do_wake(struct rw_semaphore *sem
- /*
- * wake a single writer
- */
--static inline struct rw_semaphore *
--__rwsem_wake_one_writer(struct rw_semaphore *sem)
-+static inline struct rw_anon_semaphore *
-+__rwsem_wake_one_writer(struct rw_anon_semaphore *sem)
- {
- struct rwsem_waiter *waiter;
- struct task_struct *tsk;
-@@ -139,18 +139,18 @@ __rwsem_wake_one_writer(struct rw_semaph
- /*
- * get a read lock on the semaphore
- */
--void __sched __down_read(struct rw_semaphore *sem)
-+void __sched __down_read(struct rw_anon_semaphore *sem)
- {
- struct rwsem_waiter waiter;
- struct task_struct *tsk;
- unsigned long flags;
+ return domain;
+@@ -198,19 +193,19 @@ int dca_add_requester(struct device *dev)
+ if (!dev)
+ return -EFAULT;
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
- if (sem->activity >= 0 && list_empty(&sem->wait_list)) {
- /* granted */
- sem->activity++;
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
- goto out;
+ /* check if the requester has not been added already */
+ dca = dca_find_provider_by_dev(dev);
+ if (dca) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return -EEXIST;
}
-@@ -165,7 +165,7 @@ void __sched __down_read(struct rw_semap
- list_add_tail(&waiter.list, &sem->wait_list);
-
- /* we don't need to touch the semaphore struct anymore */
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ pci_rc = dca_pci_rc_from_dev(dev);
+ domain = dca_find_domain(pci_rc);
+ if (!domain) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return -ENODEV;
+ }
- /* wait to be given the lock */
- for (;;) {
-@@ -183,13 +183,13 @@ void __sched __down_read(struct rw_semap
- /*
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
--int __down_read_trylock(struct rw_semaphore *sem)
-+int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- unsigned long flags;
- int ret = 0;
+@@ -220,17 +215,17 @@ int dca_add_requester(struct device *dev)
+ break;
+ }
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+ if (slot < 0)
+ return slot;
- if (sem->activity >= 0 && list_empty(&sem->wait_list)) {
- /* granted */
-@@ -197,7 +197,7 @@ int __down_read_trylock(struct rw_semaph
- ret = 1;
+ err = dca_sysfs_add_req(dca, dev, slot);
+ if (err) {
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
+ if (dca == dca_find_provider_by_dev(dev))
+ dca->ops->remove_requester(dca, dev);
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return err;
}
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+@@ -251,14 +246,14 @@ int dca_remove_requester(struct device *dev)
+ if (!dev)
+ return -EFAULT;
- return ret;
- }
-@@ -206,18 +206,18 @@ int __down_read_trylock(struct rw_semaph
- * get a write lock on the semaphore
- * - we increment the waiting count anyway to indicate an exclusive lock
- */
--void __sched __down_write_nested(struct rw_semaphore *sem, int subclass)
-+void __sched __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- struct rwsem_waiter waiter;
- struct task_struct *tsk;
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
+ dca = dca_find_provider_by_dev(dev);
+ if (!dca) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return -ENODEV;
+ }
+ slot = dca->ops->remove_requester(dca, dev);
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+
+ if (slot < 0)
+ return slot;
+@@ -280,16 +275,16 @@ u8 dca_common_get_tag(struct device *dev, int cpu)
+ u8 tag;
unsigned long flags;
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
- if (sem->activity == 0 && list_empty(&sem->wait_list)) {
- /* granted */
- sem->activity = -1;
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
- goto out;
+ dca = dca_find_provider_by_dev(dev);
+ if (!dca) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return -ENODEV;
}
+ tag = dca->ops->get_tag(dca, dev, cpu);
-@@ -232,7 +232,7 @@ void __sched __down_write_nested(struct
- list_add_tail(&waiter.list, &sem->wait_list);
-
- /* we don't need to touch the semaphore struct anymore */
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
-
- /* wait to be given the lock */
- for (;;) {
-@@ -247,7 +247,7 @@ void __sched __down_write_nested(struct
- ;
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return tag;
}
--void __sched __down_write(struct rw_semaphore *sem)
-+void __sched __down_write(struct rw_anon_semaphore *sem)
- {
- __down_write_nested(sem, 0);
- }
-@@ -255,12 +255,12 @@ void __sched __down_write(struct rw_sema
- /*
- * trylock for writing -- returns 1 if successful, 0 if contention
- */
--int __down_write_trylock(struct rw_semaphore *sem)
-+int __down_write_trylock(struct rw_anon_semaphore *sem)
+@@ -360,36 +355,51 @@ int register_dca_provider(struct dca_provider *dca, struct device *dev)
{
+ int err;
unsigned long flags;
- int ret = 0;
-
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- struct dca_domain *domain;
++ struct dca_domain *domain, *newdomain = NULL;
- if (sem->activity == 0 && list_empty(&sem->wait_list)) {
- /* granted */
-@@ -268,7 +268,7 @@ int __down_write_trylock(struct rw_semap
- ret = 1;
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
+ if (dca_providers_blocked) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ return -ENODEV;
}
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ err = dca_sysfs_add_provider(dca, dev);
+ if (err)
+ return err;
- return ret;
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
+ domain = dca_get_domain(dev);
+ if (!domain) {
++ struct pci_bus *rc;
++
+ if (dca_providers_blocked) {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+ dca_sysfs_remove_provider(dca);
+ unregister_dca_providers();
+- } else {
+- spin_unlock_irqrestore(&dca_lock, flags);
++ return -ENODEV;
++ }
++
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
++ rc = dca_pci_rc_from_dev(dev);
++ newdomain = dca_allocate_domain(rc);
++ if (!newdomain)
++ return -ENODEV;
++ raw_spin_lock_irqsave(&dca_lock, flags);
++ /* Recheck, we might have raced after dropping the lock */
++ domain = dca_get_domain(dev);
++ if (!domain) {
++ domain = newdomain;
++ newdomain = NULL;
++ list_add(&domain->node, &dca_domains);
+ }
+- return -ENODEV;
+ }
+ list_add(&dca->node, &domain->dca_providers);
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+
+ blocking_notifier_call_chain(&dca_provider_chain,
+ DCA_PROVIDER_ADD, NULL);
++ kfree(newdomain);
+ return 0;
}
-@@ -276,48 +276,48 @@ int __down_write_trylock(struct rw_semap
- /*
- * release a read lock on the semaphore
- */
--void __up_read(struct rw_semaphore *sem)
-+void __up_read(struct rw_anon_semaphore *sem)
- {
- unsigned long flags;
+ EXPORT_SYMBOL_GPL(register_dca_provider);
+@@ -407,7 +417,7 @@ void unregister_dca_provider(struct dca_provider *dca, struct device *dev)
+ blocking_notifier_call_chain(&dca_provider_chain,
+ DCA_PROVIDER_REMOVE, NULL);
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&dca_lock, flags);
++ raw_spin_lock_irqsave(&dca_lock, flags);
- if (--sem->activity == 0 && !list_empty(&sem->wait_list))
- sem = __rwsem_wake_one_writer(sem);
+ list_del(&dca->node);
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+@@ -416,7 +426,7 @@ void unregister_dca_provider(struct dca_provider *dca, struct device *dev)
+ if (list_empty(&domain->dca_providers))
+ dca_free_domain(domain);
+
+- spin_unlock_irqrestore(&dca_lock, flags);
++ raw_spin_unlock_irqrestore(&dca_lock, flags);
+
+ dca_sysfs_remove_provider(dca);
}
+diff --git a/drivers/dma/ipu/ipu_irq.c b/drivers/dma/ipu/ipu_irq.c
+index ab8a4ef..a71f55e 100644
+--- a/drivers/dma/ipu/ipu_irq.c
++++ b/drivers/dma/ipu/ipu_irq.c
+@@ -81,7 +81,7 @@ static struct ipu_irq_map irq_map[CONFIG_MX3_IPU_IRQS];
+ /* Protects allocations from the above array of maps */
+ static DEFINE_MUTEX(map_lock);
+ /* Protects register accesses and individual mappings */
+-static DEFINE_SPINLOCK(bank_lock);
++static DEFINE_RAW_SPINLOCK(bank_lock);
- /*
- * release a write lock on the semaphore
- */
--void __up_write(struct rw_semaphore *sem)
-+void __up_write(struct rw_anon_semaphore *sem)
+ static struct ipu_irq_map *src2map(unsigned int src)
{
- unsigned long flags;
+@@ -101,11 +101,11 @@ static void ipu_irq_unmask(struct irq_data *d)
+ uint32_t reg;
+ unsigned long lock_flags;
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- sem->activity = 0;
- if (!list_empty(&sem->wait_list))
- sem = __rwsem_do_wake(sem, 1);
+ bank = map->bank;
+ if (!bank) {
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
+ pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
+ return;
+ }
+@@ -114,7 +114,7 @@ static void ipu_irq_unmask(struct irq_data *d)
+ reg |= (1UL << (map->source & 31));
+ ipu_write_reg(bank->ipu, reg, bank->control);
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
}
- /*
- * downgrade a write lock into a read lock
- * - just wake up any readers at the front of the queue
- */
--void __downgrade_write(struct rw_semaphore *sem)
-+void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- unsigned long flags;
+ static void ipu_irq_mask(struct irq_data *d)
+@@ -124,11 +124,11 @@ static void ipu_irq_mask(struct irq_data *d)
+ uint32_t reg;
+ unsigned long lock_flags;
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- sem->activity = 1;
- if (!list_empty(&sem->wait_list))
- sem = __rwsem_do_wake(sem, 0);
+ bank = map->bank;
+ if (!bank) {
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
+ pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
+ return;
+ }
+@@ -137,7 +137,7 @@ static void ipu_irq_mask(struct irq_data *d)
+ reg &= ~(1UL << (map->source & 31));
+ ipu_write_reg(bank->ipu, reg, bank->control);
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
}
-Index: linux-3.0/lib/rwsem.c
-===================================================================
---- linux-3.0.orig/lib/rwsem.c
-+++ linux-3.0/lib/rwsem.c
-@@ -11,8 +11,8 @@
- /*
- * Initialize an rwsem:
- */
--void __init_rwsem(struct rw_semaphore *sem, const char *name,
-- struct lock_class_key *key)
-+void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
-+ struct lock_class_key *key)
- {
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
- /*
-@@ -22,11 +22,11 @@ void __init_rwsem(struct rw_semaphore *s
- lockdep_init_map(&sem->dep_map, name, key, 0);
- #endif
- sem->count = RWSEM_UNLOCKED_VALUE;
-- spin_lock_init(&sem->wait_lock);
-+ raw_spin_lock_init(&sem->wait_lock);
- INIT_LIST_HEAD(&sem->wait_list);
- }
+ static void ipu_irq_ack(struct irq_data *d)
+@@ -146,17 +146,17 @@ static void ipu_irq_ack(struct irq_data *d)
+ struct ipu_irq_bank *bank;
+ unsigned long lock_flags;
--EXPORT_SYMBOL(__init_rwsem);
-+EXPORT_SYMBOL(__init_anon_rwsem);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- struct rwsem_waiter {
- struct list_head list;
-@@ -54,8 +54,8 @@ struct rwsem_waiter {
- * - woken process blocks are discarded from the list after having task zeroed
- * - writers are only woken if downgrading is false
- */
--static struct rw_semaphore *
--__rwsem_do_wake(struct rw_semaphore *sem, int wake_type)
-+static struct rw_anon_semaphore *
-+__rwsem_do_wake(struct rw_anon_semaphore *sem, int wake_type)
- {
- struct rwsem_waiter *waiter;
- struct task_struct *tsk;
-@@ -169,8 +169,8 @@ __rwsem_do_wake(struct rw_semaphore *sem
- /*
- * wait for a lock to be granted
- */
--static struct rw_semaphore __sched *
--rwsem_down_failed_common(struct rw_semaphore *sem,
-+static struct rw_anon_semaphore __sched *
-+rwsem_down_failed_common(struct rw_anon_semaphore *sem,
- unsigned int flags, signed long adjustment)
- {
- struct rwsem_waiter waiter;
-@@ -180,7 +180,7 @@ rwsem_down_failed_common(struct rw_semap
- set_task_state(tsk, TASK_UNINTERRUPTIBLE);
-
- /* set up my own style of waitqueue */
-- spin_lock_irq(&sem->wait_lock);
-+ raw_spin_lock_irq(&sem->wait_lock);
- waiter.task = tsk;
- waiter.flags = flags;
- get_task_struct(tsk);
-@@ -204,7 +204,7 @@ rwsem_down_failed_common(struct rw_semap
- adjustment == -RWSEM_ACTIVE_WRITE_BIAS)
- sem = __rwsem_do_wake(sem, RWSEM_WAKE_READ_OWNED);
+ bank = map->bank;
+ if (!bank) {
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
+ pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
+ return;
+ }
-- spin_unlock_irq(&sem->wait_lock);
-+ raw_spin_unlock_irq(&sem->wait_lock);
+ ipu_write_reg(bank->ipu, 1UL << (map->source & 31), bank->status);
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
+ }
- /* wait to be given the lock */
- for (;;) {
-@@ -222,7 +222,8 @@ rwsem_down_failed_common(struct rw_semap
- /*
- * wait for the read lock to be granted
- */
--struct rw_semaphore __sched *rwsem_down_read_failed(struct rw_semaphore *sem)
-+struct rw_anon_semaphore __sched *
-+rwsem_down_read_failed(struct rw_anon_semaphore *sem)
- {
- return rwsem_down_failed_common(sem, RWSEM_WAITING_FOR_READ,
- -RWSEM_ACTIVE_READ_BIAS);
-@@ -231,7 +232,8 @@ struct rw_semaphore __sched *rwsem_down_
- /*
- * wait for the write lock to be granted
- */
--struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem)
-+struct rw_anon_semaphore __sched *
-+rwsem_down_write_failed(struct rw_anon_semaphore *sem)
- {
- return rwsem_down_failed_common(sem, RWSEM_WAITING_FOR_WRITE,
- -RWSEM_ACTIVE_WRITE_BIAS);
-@@ -241,17 +243,17 @@ struct rw_semaphore __sched *rwsem_down_
- * handle waking up a waiter on the semaphore
- * - up_read/up_write has decremented the active part of count if we come here
- */
--struct rw_semaphore *rwsem_wake(struct rw_semaphore *sem)
-+struct rw_anon_semaphore *rwsem_wake(struct rw_anon_semaphore *sem)
- {
- unsigned long flags;
+ /**
+@@ -172,11 +172,11 @@ bool ipu_irq_status(unsigned int irq)
+ unsigned long lock_flags;
+ bool ret;
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
+ bank = map->bank;
+ ret = bank && ipu_read_reg(bank->ipu, bank->status) &
+ (1UL << (map->source & 31));
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- /* do nothing if list empty */
- if (!list_empty(&sem->wait_list))
- sem = __rwsem_do_wake(sem, RWSEM_WAKE_ANY);
+ return ret;
+ }
+@@ -213,10 +213,10 @@ int ipu_irq_map(unsigned int source)
+ if (irq_map[i].source < 0) {
+ unsigned long lock_flags;
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
+ irq_map[i].source = source;
+ irq_map[i].bank = irq_bank + source / 32;
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- return sem;
- }
-@@ -261,17 +263,17 @@ struct rw_semaphore *rwsem_wake(struct r
- * - caller incremented waiting part of count and discovered it still negative
- * - just wake up any readers at the front of the queue
- */
--struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem)
-+struct rw_anon_semaphore *rwsem_downgrade_wake(struct rw_anon_semaphore *sem)
- {
- unsigned long flags;
+ ret = irq_map[i].irq;
+ pr_debug("IPU: mapped source %u to IRQ %u\n",
+@@ -252,10 +252,10 @@ int ipu_irq_unmap(unsigned int source)
+ pr_debug("IPU: unmapped source %u from IRQ %u\n",
+ source, irq_map[i].irq);
-- spin_lock_irqsave(&sem->wait_lock, flags);
-+ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+- spin_lock_irqsave(&bank_lock, lock_flags);
++ raw_spin_lock_irqsave(&bank_lock, lock_flags);
+ irq_map[i].source = -EINVAL;
+ irq_map[i].bank = NULL;
+- spin_unlock_irqrestore(&bank_lock, lock_flags);
++ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- /* do nothing if list empty */
- if (!list_empty(&sem->wait_list))
- sem = __rwsem_do_wake(sem, RWSEM_WAKE_READ_OWNED);
+ ret = 0;
+ break;
+@@ -276,7 +276,7 @@ static void ipu_irq_err(unsigned int irq, struct irq_desc *desc)
+ for (i = IPU_IRQ_NR_FN_BANKS; i < IPU_IRQ_NR_BANKS; i++) {
+ struct ipu_irq_bank *bank = irq_bank + i;
-- spin_unlock_irqrestore(&sem->wait_lock, flags);
-+ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+- spin_lock(&bank_lock);
++ raw_spin_lock(&bank_lock);
+ status = ipu_read_reg(ipu, bank->status);
+ /*
+ * Don't think we have to clear all interrupts here, they will
+@@ -284,18 +284,18 @@ static void ipu_irq_err(unsigned int irq, struct irq_desc *desc)
+ * might want to clear unhandled interrupts after the loop...
+ */
+ status &= ipu_read_reg(ipu, bank->control);
+- spin_unlock(&bank_lock);
++ raw_spin_unlock(&bank_lock);
+ while ((line = ffs(status))) {
+ struct ipu_irq_map *map;
- return sem;
- }
-Index: linux-3.0/kernel/time/timer_stats.c
-===================================================================
---- linux-3.0.orig/kernel/time/timer_stats.c
-+++ linux-3.0/kernel/time/timer_stats.c
-@@ -81,7 +81,7 @@ struct entry {
- /*
- * Spinlock protecting the tables - not taken during lookup:
- */
--static DEFINE_SPINLOCK(table_lock);
-+static DEFINE_RAW_SPINLOCK(table_lock);
+ line--;
+ status &= ~(1UL << line);
- /*
- * Per-CPU lookup locks for fast hash lookup:
-@@ -188,7 +188,7 @@ static struct entry *tstat_lookup(struct
- prev = NULL;
- curr = *head;
+- spin_lock(&bank_lock);
++ raw_spin_lock(&bank_lock);
+ map = src2map(32 * i + line);
+ if (map)
+ irq = map->irq;
+- spin_unlock(&bank_lock);
++ raw_spin_unlock(&bank_lock);
-- spin_lock(&table_lock);
-+ raw_spin_lock(&table_lock);
- /*
- * Make sure we have not raced with another CPU:
- */
-@@ -215,7 +215,7 @@ static struct entry *tstat_lookup(struct
- *head = curr;
- }
- out_unlock:
-- spin_unlock(&table_lock);
-+ raw_spin_unlock(&table_lock);
+ if (!map) {
+ pr_err("IPU: Interrupt on unmapped source %u bank %d\n",
+@@ -317,22 +317,22 @@ static void ipu_irq_fn(unsigned int irq, struct irq_desc *desc)
+ for (i = 0; i < IPU_IRQ_NR_FN_BANKS; i++) {
+ struct ipu_irq_bank *bank = irq_bank + i;
- return curr;
- }
-Index: linux-3.0/kernel/latencytop.c
-===================================================================
---- linux-3.0.orig/kernel/latencytop.c
-+++ linux-3.0/kernel/latencytop.c
-@@ -58,7 +58,7 @@
- #include <linux/list.h>
- #include <linux/stacktrace.h>
+- spin_lock(&bank_lock);
++ raw_spin_lock(&bank_lock);
+ status = ipu_read_reg(ipu, bank->status);
+ /* Not clearing all interrupts, see above */
+ status &= ipu_read_reg(ipu, bank->control);
+- spin_unlock(&bank_lock);
++ raw_spin_unlock(&bank_lock);
+ while ((line = ffs(status))) {
+ struct ipu_irq_map *map;
--static DEFINE_SPINLOCK(latency_lock);
-+static DEFINE_RAW_SPINLOCK(latency_lock);
+ line--;
+ status &= ~(1UL << line);
- #define MAXLR 128
- static struct latency_record latency_record[MAXLR];
-@@ -72,19 +72,19 @@ void clear_all_latency_tracing(struct ta
- if (!latencytop_enabled)
- return;
+- spin_lock(&bank_lock);
++ raw_spin_lock(&bank_lock);
+ map = src2map(32 * i + line);
+ if (map)
+ irq = map->irq;
+- spin_unlock(&bank_lock);
++ raw_spin_unlock(&bank_lock);
-- spin_lock_irqsave(&latency_lock, flags);
-+ raw_spin_lock_irqsave(&latency_lock, flags);
- memset(&p->latency_record, 0, sizeof(p->latency_record));
- p->latency_record_count = 0;
-- spin_unlock_irqrestore(&latency_lock, flags);
-+ raw_spin_unlock_irqrestore(&latency_lock, flags);
- }
+ if (!map) {
+ pr_err("IPU: Interrupt on unmapped source %u bank %d\n",
+diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
+index 2022a5c..ea65ca6 100644
+--- a/drivers/gpu/drm/drm_irq.c
++++ b/drivers/gpu/drm/drm_irq.c
+@@ -109,10 +109,7 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
+ /* Prevent vblank irq processing while disabling vblank irqs,
+ * so no updates of timestamps or count can happen after we've
+ * disabled. Needed to prevent races in case of delayed irq's.
+- * Disable preemption, so vblank_time_lock is held as short as
+- * possible, even under a kernel with PREEMPT_RT patches.
+ */
+- preempt_disable();
+ spin_lock_irqsave(&dev->vblank_time_lock, irqflags);
- static void clear_global_latency_tracing(void)
- {
- unsigned long flags;
+ dev->driver->disable_vblank(dev, crtc);
+@@ -163,7 +160,6 @@ static void vblank_disable_and_save(struct drm_device *dev, int crtc)
+ clear_vblank_timestamps(dev, crtc);
-- spin_lock_irqsave(&latency_lock, flags);
-+ raw_spin_lock_irqsave(&latency_lock, flags);
- memset(&latency_record, 0, sizeof(latency_record));
-- spin_unlock_irqrestore(&latency_lock, flags);
-+ raw_spin_unlock_irqrestore(&latency_lock, flags);
+ spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
+- preempt_enable();
}
- static void __sched
-@@ -190,7 +190,7 @@ __account_scheduler_latency(struct task_
- lat.max = usecs;
- store_stacktrace(tsk, &lat);
-
-- spin_lock_irqsave(&latency_lock, flags);
-+ raw_spin_lock_irqsave(&latency_lock, flags);
+ static void vblank_disable_fn(unsigned long arg)
+@@ -875,10 +871,6 @@ int drm_vblank_get(struct drm_device *dev, int crtc)
+ spin_lock_irqsave(&dev->vbl_lock, irqflags);
+ /* Going from 0->1 means we have to enable interrupts again */
+ if (atomic_add_return(1, &dev->vblank_refcount[crtc]) == 1) {
+- /* Disable preemption while holding vblank_time_lock. Do
+- * it explicitely to guard against PREEMPT_RT kernel.
+- */
+- preempt_disable();
+ spin_lock_irqsave(&dev->vblank_time_lock, irqflags2);
+ if (!dev->vblank_enabled[crtc]) {
+ /* Enable vblank irqs under vblank_time_lock protection.
+@@ -898,7 +890,6 @@ int drm_vblank_get(struct drm_device *dev, int crtc)
+ }
+ }
+ spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags2);
+- preempt_enable();
+ } else {
+ if (!dev->vblank_enabled[crtc]) {
+ atomic_dec(&dev->vblank_refcount[crtc]);
+diff --git a/drivers/ide/alim15x3.c b/drivers/ide/alim15x3.c
+index 2c8016a..6fd6037 100644
+--- a/drivers/ide/alim15x3.c
++++ b/drivers/ide/alim15x3.c
+@@ -234,7 +234,7 @@ static int init_chipset_ali15x3(struct pci_dev *dev)
- account_global_scheduler_latency(tsk, &lat);
+ isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL);
-@@ -231,7 +231,7 @@ __account_scheduler_latency(struct task_
- memcpy(&tsk->latency_record[i], &lat, sizeof(struct latency_record));
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
- out_unlock:
-- spin_unlock_irqrestore(&latency_lock, flags);
-+ raw_spin_unlock_irqrestore(&latency_lock, flags);
- }
-
- static int lstats_show(struct seq_file *m, void *v)
-Index: linux-3.0/drivers/video/console/vgacon.c
-===================================================================
---- linux-3.0.orig/drivers/video/console/vgacon.c
-+++ linux-3.0/drivers/video/console/vgacon.c
-@@ -50,7 +50,7 @@
- #include <video/vga.h>
- #include <asm/io.h>
-
--static DEFINE_SPINLOCK(vga_lock);
-+static DEFINE_RAW_SPINLOCK(vga_lock);
- static int cursor_size_lastfrom;
- static int cursor_size_lastto;
- static u32 vgacon_xres;
-@@ -157,7 +157,7 @@ static inline void write_vga(unsigned ch
- * ddprintk might set the console position from interrupt
- * handlers, thus the write has to be IRQ-atomic.
- */
-- spin_lock_irqsave(&vga_lock, flags);
-+ raw_spin_lock_irqsave(&vga_lock, flags);
-
- #ifndef SLOW_VGA
- v1 = reg + (val & 0xff00);
-@@ -170,7 +170,7 @@ static inline void write_vga(unsigned ch
- outb_p(reg + 1, vga_video_port_reg);
- outb_p(val & 0xff, vga_video_port_val);
- #endif
-- spin_unlock_irqrestore(&vga_lock, flags);
-+ raw_spin_unlock_irqrestore(&vga_lock, flags);
+ if (m5229_revision < 0xC2) {
+ /*
+@@ -325,7 +325,7 @@ out:
+ }
+ pci_dev_put(north);
+ pci_dev_put(isa_dev);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ return 0;
}
- static inline void vga_set_mem_top(struct vc_data *c)
-@@ -664,7 +664,7 @@ static void vgacon_set_cursor_size(int x
- cursor_size_lastfrom = from;
- cursor_size_lastto = to;
+diff --git a/drivers/ide/hpt366.c b/drivers/ide/hpt366.c
+index 58c51cd..d2a4059 100644
+--- a/drivers/ide/hpt366.c
++++ b/drivers/ide/hpt366.c
+@@ -1241,7 +1241,7 @@ static int __devinit init_dma_hpt366(ide_hwif_t *hwif,
-- spin_lock_irqsave(&vga_lock, flags);
-+ raw_spin_lock_irqsave(&vga_lock, flags);
- if (vga_video_type >= VIDEO_TYPE_VGAC) {
- outb_p(VGA_CRTC_CURSOR_START, vga_video_port_reg);
- curs = inb_p(vga_video_port_val);
-@@ -682,7 +682,7 @@ static void vgacon_set_cursor_size(int x
- outb_p(curs, vga_video_port_val);
- outb_p(VGA_CRTC_CURSOR_END, vga_video_port_reg);
- outb_p(cure, vga_video_port_val);
-- spin_unlock_irqrestore(&vga_lock, flags);
-+ raw_spin_unlock_irqrestore(&vga_lock, flags);
- }
+ dma_old = inb(base + 2);
- static void vgacon_cursor(struct vc_data *c, int mode)
-@@ -757,7 +757,7 @@ static int vgacon_doresize(struct vc_dat
- unsigned int scanlines = height * c->vc_font.height;
- u8 scanlines_lo = 0, r7 = 0, vsync_end = 0, mode, max_scan;
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
-- spin_lock_irqsave(&vga_lock, flags);
-+ raw_spin_lock_irqsave(&vga_lock, flags);
+ dma_new = dma_old;
+ pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma);
+@@ -1252,7 +1252,7 @@ static int __devinit init_dma_hpt366(ide_hwif_t *hwif,
+ if (dma_new != dma_old)
+ outb(dma_new, base + 2);
- vgacon_xres = width * VGA_FONTWIDTH;
- vgacon_yres = height * c->vc_font.height;
-@@ -808,7 +808,7 @@ static int vgacon_doresize(struct vc_dat
- outb_p(vsync_end, vga_video_port_val);
- }
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
-- spin_unlock_irqrestore(&vga_lock, flags);
-+ raw_spin_unlock_irqrestore(&vga_lock, flags);
- return 0;
- }
+ printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n",
+ hwif->name, base, base + 7);
+diff --git a/drivers/ide/ide-io-std.c b/drivers/ide/ide-io-std.c
+index 46721c4..b6f114a 100644
+--- a/drivers/ide/ide-io-std.c
++++ b/drivers/ide/ide-io-std.c
+@@ -174,7 +174,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf,
+ unsigned long uninitialized_var(flags);
-@@ -891,11 +891,11 @@ static void vga_vesa_blank(struct vgasta
- {
- /* save original values of VGA controller registers */
- if (!vga_vesa_blanked) {
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- vga_state.SeqCtrlIndex = vga_r(state->vgabase, VGA_SEQ_I);
- vga_state.CrtCtrlIndex = inb_p(vga_video_port_reg);
- vga_state.CrtMiscIO = vga_r(state->vgabase, VGA_MIS_R);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
+ if ((io_32bit & 2) && !mmio) {
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ ata_vlb_sync(io_ports->nsect_addr);
+ }
- outb_p(0x00, vga_video_port_reg); /* HorizontalTotal */
- vga_state.HorizontalTotal = inb_p(vga_video_port_val);
-@@ -918,7 +918,7 @@ static void vga_vesa_blank(struct vgasta
+@@ -185,7 +185,7 @@ void ide_input_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf,
+ insl(data_addr, buf, words);
- /* assure that video is enabled */
- /* "0x20" is VIDEO_ENABLE_bit in register 01 of sequencer */
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- vga_wseq(state->vgabase, VGA_SEQ_CLOCK_MODE, vga_state.ClockingMode | 0x20);
+ if ((io_32bit & 2) && !mmio)
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
- /* test for vertical retrace in process.... */
-@@ -954,13 +954,13 @@ static void vga_vesa_blank(struct vgasta
- /* restore both index registers */
- vga_w(state->vgabase, VGA_SEQ_I, vga_state.SeqCtrlIndex);
- outb_p(vga_state.CrtCtrlIndex, vga_video_port_reg);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
- }
+ if (((len + 1) & 3) < 2)
+ return;
+@@ -218,7 +218,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf,
+ unsigned long uninitialized_var(flags);
- static void vga_vesa_unblank(struct vgastate *state)
- {
- /* restore original values of VGA controller registers */
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- vga_w(state->vgabase, VGA_MIS_W, vga_state.CrtMiscIO);
+ if ((io_32bit & 2) && !mmio) {
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ ata_vlb_sync(io_ports->nsect_addr);
+ }
- outb_p(0x00, vga_video_port_reg); /* HorizontalTotal */
-@@ -985,7 +985,7 @@ static void vga_vesa_unblank(struct vgas
- /* restore index/control registers */
- vga_w(state->vgabase, VGA_SEQ_I, vga_state.SeqCtrlIndex);
- outb_p(vga_state.CrtCtrlIndex, vga_video_port_reg);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
- }
+@@ -229,7 +229,7 @@ void ide_output_data(ide_drive_t *drive, struct ide_cmd *cmd, void *buf,
+ outsl(data_addr, buf, words);
- static void vga_pal_blank(struct vgastate *state)
-@@ -1104,7 +1104,7 @@ static int vgacon_do_font_op(struct vgas
- charmap += 4 * cmapsz;
- #endif
+ if ((io_32bit & 2) && !mmio)
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- /* First, the Sequencer */
- vga_wseq(state->vgabase, VGA_SEQ_RESET, 0x1);
- /* CPU writes only to map 2 */
-@@ -1120,7 +1120,7 @@ static int vgacon_do_font_op(struct vgas
- vga_wgfx(state->vgabase, VGA_GFX_MODE, 0x00);
- /* map start at A000:0000 */
- vga_wgfx(state->vgabase, VGA_GFX_MISC, 0x00);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
+ if (((len + 1) & 3) < 2)
+ return;
+diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c
+index 177db6d..079ae6b 100644
+--- a/drivers/ide/ide-io.c
++++ b/drivers/ide/ide-io.c
+@@ -659,7 +659,7 @@ void ide_timer_expiry (unsigned long data)
+ /* disable_irq_nosync ?? */
+ disable_irq(hwif->irq);
+ /* local CPU only, as if we were handling an interrupt */
+- local_irq_disable();
++ local_irq_disable_nort();
+ if (hwif->polling) {
+ startstop = handler(drive);
+ } else if (drive_is_ready(drive)) {
+diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c
+index 376f2dc..f014dd1 100644
+--- a/drivers/ide/ide-iops.c
++++ b/drivers/ide/ide-iops.c
+@@ -129,12 +129,12 @@ int __ide_wait_stat(ide_drive_t *drive, u8 good, u8 bad,
+ if ((stat & ATA_BUSY) == 0)
+ break;
- if (arg) {
- if (set)
-@@ -1147,7 +1147,7 @@ static int vgacon_do_font_op(struct vgas
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ *rstat = stat;
+ return -EBUSY;
+ }
}
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
}
+ /*
+ * Allow status to settle, then read it again.
+diff --git a/drivers/ide/ide-probe.c b/drivers/ide/ide-probe.c
+index 068cef0..38e69e1 100644
+--- a/drivers/ide/ide-probe.c
++++ b/drivers/ide/ide-probe.c
+@@ -196,10 +196,10 @@ static void do_identify(ide_drive_t *drive, u8 cmd, u16 *id)
+ int bswap = 1;
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- /* First, the sequencer, Synchronous reset */
- vga_wseq(state->vgabase, VGA_SEQ_RESET, 0x01);
- /* CPU writes to maps 0 and 1 */
-@@ -1186,7 +1186,7 @@ static int vgacon_do_font_op(struct vgas
- inb_p(video_port_status);
- vga_wattr(state->vgabase, VGA_AR_ENABLE_DISPLAY, 0);
- }
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
- return 0;
- }
+ /* local CPU only; some systems need this */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ /* read 512 bytes of id info */
+ hwif->tp_ops->input_data(drive, NULL, id, SECTOR_SIZE);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
-@@ -1211,26 +1211,26 @@ static int vgacon_adjust_height(struct v
- registers; they are write-only on EGA, but it appears that they
- are all don't care bits on EGA, so I guess it doesn't matter. */
+ drive->dev_flags |= IDE_DFLAG_ID_READ;
+ #ifdef DEBUG
+diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c
+index 600c89a..af308a9 100644
+--- a/drivers/ide/ide-taskfile.c
++++ b/drivers/ide/ide-taskfile.c
+@@ -250,7 +250,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd,
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- outb_p(0x07, vga_video_port_reg); /* CRTC overflow register */
- ovr = inb_p(vga_video_port_val);
- outb_p(0x09, vga_video_port_reg); /* Font size register */
- fsr = inb_p(vga_video_port_val);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
+ page_is_high = PageHighMem(page);
+ if (page_is_high)
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
- vde = maxscan & 0xff; /* Vertical display end reg */
- ovr = (ovr & 0xbd) + /* Overflow register */
- ((maxscan & 0x100) >> 7) + ((maxscan & 0x200) >> 3);
- fsr = (fsr & 0xe0) + (fontheight - 1); /* Font size register */
+ buf = kmap_atomic(page, KM_BIO_SRC_IRQ) + offset;
-- spin_lock_irq(&vga_lock);
-+ raw_spin_lock_irq(&vga_lock);
- outb_p(0x07, vga_video_port_reg); /* CRTC overflow register */
- outb_p(ovr, vga_video_port_val);
- outb_p(0x09, vga_video_port_reg); /* Font size */
- outb_p(fsr, vga_video_port_val);
- outb_p(0x12, vga_video_port_reg); /* Vertical display limit */
- outb_p(vde, vga_video_port_val);
-- spin_unlock_irq(&vga_lock);
-+ raw_spin_unlock_irq(&vga_lock);
- vga_video_font_height = fontheight;
+@@ -271,7 +271,7 @@ void ide_pio_bytes(ide_drive_t *drive, struct ide_cmd *cmd,
+ kunmap_atomic(buf, KM_BIO_SRC_IRQ);
- for (i = 0; i < MAX_NR_CONSOLES; i++) {
-Index: linux-3.0/arch/x86/oprofile/nmi_int.c
-===================================================================
---- linux-3.0.orig/arch/x86/oprofile/nmi_int.c
-+++ linux-3.0/arch/x86/oprofile/nmi_int.c
-@@ -355,10 +355,10 @@ static void nmi_cpu_setup(void *dummy)
- int cpu = smp_processor_id();
- struct op_msrs *msrs = &per_cpu(cpu_msrs, cpu);
- nmi_cpu_save_registers(msrs);
-- spin_lock(&oprofilefs_lock);
-+ raw_spin_lock(&oprofilefs_lock);
- model->setup_ctrs(model, msrs);
- nmi_cpu_setup_mux(cpu, msrs);
-- spin_unlock(&oprofilefs_lock);
-+ raw_spin_unlock(&oprofilefs_lock);
- per_cpu(saved_lvtpc, cpu) = apic_read(APIC_LVTPC);
- apic_write(APIC_LVTPC, APIC_DM_NMI);
- }
-Index: linux-3.0/drivers/oprofile/event_buffer.c
-===================================================================
---- linux-3.0.orig/drivers/oprofile/event_buffer.c
-+++ linux-3.0/drivers/oprofile/event_buffer.c
-@@ -82,10 +82,10 @@ int alloc_event_buffer(void)
- {
- unsigned long flags;
+ if (page_is_high)
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
-- spin_lock_irqsave(&oprofilefs_lock, flags);
-+ raw_spin_lock_irqsave(&oprofilefs_lock, flags);
- buffer_size = oprofile_buffer_size;
- buffer_watershed = oprofile_buffer_watershed;
-- spin_unlock_irqrestore(&oprofilefs_lock, flags);
-+ raw_spin_unlock_irqrestore(&oprofilefs_lock, flags);
+ len -= nr_bytes;
+ }
+@@ -414,7 +414,7 @@ static ide_startstop_t pre_task_out_intr(ide_drive_t *drive,
+ }
- if (buffer_watershed >= buffer_size)
- return -EINVAL;
-Index: linux-3.0/drivers/oprofile/oprofile_perf.c
-===================================================================
---- linux-3.0.orig/drivers/oprofile/oprofile_perf.c
-+++ linux-3.0/drivers/oprofile/oprofile_perf.c
-@@ -160,9 +160,9 @@ static int oprofile_perf_create_files(st
+ if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0)
+- local_irq_disable();
++ local_irq_disable_nort();
- static int oprofile_perf_setup(void)
- {
-- spin_lock(&oprofilefs_lock);
-+ raw_spin_lock(&oprofilefs_lock);
- op_perf_setup();
-- spin_unlock(&oprofilefs_lock);
-+ raw_spin_unlock(&oprofilefs_lock);
- return 0;
- }
-
-Index: linux-3.0/drivers/oprofile/oprofilefs.c
-===================================================================
---- linux-3.0.orig/drivers/oprofile/oprofilefs.c
-+++ linux-3.0/drivers/oprofile/oprofilefs.c
-@@ -21,7 +21,7 @@
+ ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE);
- #define OPROFILEFS_MAGIC 0x6f70726f
+diff --git a/drivers/ide/ide_platform.c b/drivers/ide/ide_platform.c
+index 542603b..83a190f 100644
+--- a/drivers/ide/ide_platform.c
++++ b/drivers/ide/ide_platform.c
+@@ -95,7 +95,7 @@ static int __devinit plat_ide_probe(struct platform_device *pdev)
+ plat_ide_setup_ports(&hw, base, alt_base, pdata, res_irq->start);
+ hw.dev = &pdev->dev;
--DEFINE_SPINLOCK(oprofilefs_lock);
-+DEFINE_RAW_SPINLOCK(oprofilefs_lock);
+- d.irq_flags = res_irq->flags;
++ d.irq_flags = 0;
+ if (mmio)
+ d.host_flags |= IDE_HFLAG_MMIO;
- static struct inode *oprofilefs_get_inode(struct super_block *sb, int mode)
- {
-@@ -76,9 +76,9 @@ int oprofilefs_ulong_from_user(unsigned
- if (copy_from_user(tmpbuf, buf, count))
- return -EFAULT;
+diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+index 3871ac6..79960b1 100644
+--- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
++++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
+@@ -794,7 +794,7 @@ void ipoib_mcast_restart_task(struct work_struct *work)
-- spin_lock_irqsave(&oprofilefs_lock, flags);
-+ raw_spin_lock_irqsave(&oprofilefs_lock, flags);
- *val = simple_strtoul(tmpbuf, NULL, 0);
-- spin_unlock_irqrestore(&oprofilefs_lock, flags);
-+ raw_spin_unlock_irqrestore(&oprofilefs_lock, flags);
- return 0;
- }
+ ipoib_mcast_stop_thread(dev, 0);
-Index: linux-3.0/include/linux/oprofile.h
-===================================================================
---- linux-3.0.orig/include/linux/oprofile.h
-+++ linux-3.0/include/linux/oprofile.h
-@@ -166,7 +166,7 @@ ssize_t oprofilefs_ulong_to_user(unsigne
- int oprofilefs_ulong_from_user(unsigned long * val, char const __user * buf, size_t count);
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ netif_addr_lock(dev);
+ spin_lock(&priv->lock);
- /** lock for read/write safety */
--extern spinlock_t oprofilefs_lock;
-+extern raw_spinlock_t oprofilefs_lock;
+@@ -876,7 +876,7 @@ void ipoib_mcast_restart_task(struct work_struct *work)
- /**
- * Add the contents of a circular buffer to the event buffer.
-Index: linux-3.0/drivers/acpi/processor_idle.c
-===================================================================
---- linux-3.0.orig/drivers/acpi/processor_idle.c
-+++ linux-3.0/drivers/acpi/processor_idle.c
-@@ -852,7 +852,7 @@ static int acpi_idle_enter_simple(struct
- }
+ spin_unlock(&priv->lock);
+ netif_addr_unlock(dev);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
- static int c3_cpu_count;
--static DEFINE_SPINLOCK(c3_lock);
-+static DEFINE_RAW_SPINLOCK(c3_lock);
+ /* We have to cancel outside of the spinlock */
+ list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
+diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c
+index 5b8f59d..80e44ce 100644
+--- a/drivers/input/gameport/gameport.c
++++ b/drivers/input/gameport/gameport.c
+@@ -87,12 +87,12 @@ static int gameport_measure_speed(struct gameport *gameport)
+ tx = 1 << 30;
- /**
- * acpi_idle_enter_bm - enters C3 with proper BM handling
-@@ -930,12 +930,12 @@ static int acpi_idle_enter_bm(struct cpu
- * without doing anything.
- */
- if (pr->flags.bm_check && pr->flags.bm_control) {
-- spin_lock(&c3_lock);
-+ raw_spin_lock(&c3_lock);
- c3_cpu_count++;
- /* Disable bus master arbitration when all CPUs are in C3 */
- if (c3_cpu_count == num_online_cpus())
- acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 1);
-- spin_unlock(&c3_lock);
-+ raw_spin_unlock(&c3_lock);
- } else if (!pr->flags.bm_check) {
- ACPI_FLUSH_CPU_CACHE();
+ for(i = 0; i < 50; i++) {
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ GET_TIME(t1);
+ for (t = 0; t < 50; t++) gameport_read(gameport);
+ GET_TIME(t2);
+ GET_TIME(t3);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ udelay(i * 10);
+ if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t;
}
-@@ -944,10 +944,10 @@ static int acpi_idle_enter_bm(struct cpu
+@@ -111,11 +111,11 @@ static int gameport_measure_speed(struct gameport *gameport)
+ tx = 1 << 30;
- /* Re-enable bus master arbitration */
- if (pr->flags.bm_check && pr->flags.bm_control) {
-- spin_lock(&c3_lock);
-+ raw_spin_lock(&c3_lock);
- acpi_write_bit_register(ACPI_BITREG_ARB_DISABLE, 0);
- c3_cpu_count--;
-- spin_unlock(&c3_lock);
-+ raw_spin_unlock(&c3_lock);
+ for(i = 0; i < 50; i++) {
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ rdtscl(t1);
+ for (t = 0; t < 50; t++) gameport_read(gameport);
+ rdtscl(t2);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ udelay(i * 10);
+ if (t2 - t1 < tx) tx = t2 - t1;
+ }
+diff --git a/drivers/md/dm.c b/drivers/md/dm.c
+index 41abc6d..67207d6 100644
+--- a/drivers/md/dm.c
++++ b/drivers/md/dm.c
+@@ -1644,14 +1644,14 @@ static void dm_request_fn(struct request_queue *q)
+ if (map_request(ti, clone, md))
+ goto requeued;
+
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
+ spin_lock(q->queue_lock);
}
- kt2 = ktime_get_real();
- idle_time_ns = ktime_to_ns(ktime_sub(kt2, kt1));
-Index: linux-3.0/arch/x86/kernel/cpu/mcheck/mce_intel.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/mcheck/mce_intel.c
-+++ linux-3.0/arch/x86/kernel/cpu/mcheck/mce_intel.c
-@@ -28,7 +28,7 @@ static DEFINE_PER_CPU(mce_banks_t, mce_b
- * cmci_discover_lock protects against parallel discovery attempts
- * which could race against each other.
- */
--static DEFINE_SPINLOCK(cmci_discover_lock);
-+static DEFINE_RAW_SPINLOCK(cmci_discover_lock);
- #define CMCI_THRESHOLD 1
+ goto out;
-@@ -85,7 +85,7 @@ static void cmci_discover(int banks, int
- int hdr = 0;
- int i;
+ requeued:
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
+ spin_lock(q->queue_lock);
-- spin_lock_irqsave(&cmci_discover_lock, flags);
-+ raw_spin_lock_irqsave(&cmci_discover_lock, flags);
- for (i = 0; i < banks; i++) {
- u64 val;
+ delay_and_out:
+diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
+index e509147..777a79d 100644
+--- a/drivers/md/raid5.c
++++ b/drivers/md/raid5.c
+@@ -1226,8 +1226,9 @@ static void __raid_run_ops(struct stripe_head *sh, unsigned long ops_request)
+ struct raid5_percpu *percpu;
+ unsigned long cpu;
-@@ -116,7 +116,7 @@ static void cmci_discover(int banks, int
- WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks)));
+- cpu = get_cpu();
++ cpu = get_cpu_light();
+ percpu = per_cpu_ptr(conf->percpu, cpu);
++ spin_lock(&percpu->lock);
+ if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) {
+ ops_run_biofill(sh);
+ overlap_clear++;
+@@ -1279,7 +1280,8 @@ static void __raid_run_ops(struct stripe_head *sh, unsigned long ops_request)
+ if (test_and_clear_bit(R5_Overlap, &dev->flags))
+ wake_up(&sh->raid_conf->wait_for_overlap);
}
- }
-- spin_unlock_irqrestore(&cmci_discover_lock, flags);
-+ raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
- if (hdr)
- printk(KERN_CONT "\n");
+- put_cpu();
++ spin_unlock(&percpu->lock);
++ put_cpu_light();
}
-@@ -150,7 +150,7 @@ void cmci_clear(void)
- if (!cmci_supported(&banks))
- return;
-- spin_lock_irqsave(&cmci_discover_lock, flags);
-+ raw_spin_lock_irqsave(&cmci_discover_lock, flags);
- for (i = 0; i < banks; i++) {
- if (!test_bit(i, __get_cpu_var(mce_banks_owned)))
- continue;
-@@ -160,7 +160,7 @@ void cmci_clear(void)
- wrmsrl(MSR_IA32_MCx_CTL2(i), val);
- __clear_bit(i, __get_cpu_var(mce_banks_owned));
+ #ifdef CONFIG_MULTICORE_RAID456
+@@ -4756,6 +4758,7 @@ static int raid5_alloc_percpu(raid5_conf_t *conf)
+ break;
+ }
+ per_cpu_ptr(conf->percpu, cpu)->scribble = scribble;
++ spin_lock_init(&per_cpu_ptr(conf->percpu, cpu)->lock);
}
-- spin_unlock_irqrestore(&cmci_discover_lock, flags);
-+ raw_spin_unlock_irqrestore(&cmci_discover_lock, flags);
- }
-
- /*
-Index: linux-3.0/arch/powerpc/sysdev/uic.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/sysdev/uic.c
-+++ linux-3.0/arch/powerpc/sysdev/uic.c
-@@ -47,7 +47,7 @@ struct uic {
- int index;
- int dcrbase;
-
-- spinlock_t lock;
-+ raw_spinlock_t lock;
-
- /* The remapper for this UIC */
- struct irq_host *irqhost;
-@@ -61,14 +61,14 @@ static void uic_unmask_irq(struct irq_da
- u32 er, sr;
-
- sr = 1 << (31-src);
-- spin_lock_irqsave(&uic->lock, flags);
-+ raw_spin_lock_irqsave(&uic->lock, flags);
- /* ack level-triggered interrupts here */
- if (irqd_is_level_type(d))
- mtdcr(uic->dcrbase + UIC_SR, sr);
- er = mfdcr(uic->dcrbase + UIC_ER);
- er |= sr;
- mtdcr(uic->dcrbase + UIC_ER, er);
-- spin_unlock_irqrestore(&uic->lock, flags);
-+ raw_spin_unlock_irqrestore(&uic->lock, flags);
- }
-
- static void uic_mask_irq(struct irq_data *d)
-@@ -78,11 +78,11 @@ static void uic_mask_irq(struct irq_data
- unsigned long flags;
- u32 er;
-
-- spin_lock_irqsave(&uic->lock, flags);
-+ raw_spin_lock_irqsave(&uic->lock, flags);
- er = mfdcr(uic->dcrbase + UIC_ER);
- er &= ~(1 << (31 - src));
- mtdcr(uic->dcrbase + UIC_ER, er);
-- spin_unlock_irqrestore(&uic->lock, flags);
-+ raw_spin_unlock_irqrestore(&uic->lock, flags);
- }
+ #ifdef CONFIG_HOTPLUG_CPU
+ conf->cpu_notify.notifier_call = raid456_cpu_notify;
+diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h
+index 3ca77a2..b1156c2 100644
+--- a/drivers/md/raid5.h
++++ b/drivers/md/raid5.h
+@@ -402,6 +402,7 @@ struct raid5_private_data {
- static void uic_ack_irq(struct irq_data *d)
-@@ -91,9 +91,9 @@ static void uic_ack_irq(struct irq_data
- unsigned int src = irqd_to_hwirq(d);
- unsigned long flags;
+ /* per cpu variables */
+ struct raid5_percpu {
++ spinlock_t lock; /* Protection for -RT */
+ struct page *spare_page; /* Used when checking P/Q in raid6 */
+ void *scribble; /* space for constructing buffer
+ * lists and performing address
+diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig
+index 3546474..7254a4f 100644
+--- a/drivers/misc/Kconfig
++++ b/drivers/misc/Kconfig
+@@ -82,6 +82,7 @@ config AB8500_PWM
+ config ATMEL_TCLIB
+ bool "Atmel AT32/AT91 Timer/Counter Library"
+ depends on (AVR32 || ARCH_AT91)
++ default y if PREEMPT_RT_FULL
+ help
+ Select this if you want a library to allocate the Timer/Counter
+ blocks found on many Atmel processors. This facilitates using
+@@ -97,8 +98,7 @@ config ATMEL_TCB_CLKSRC
+ are combined to make a single 32-bit timer.
-- spin_lock_irqsave(&uic->lock, flags);
-+ raw_spin_lock_irqsave(&uic->lock, flags);
- mtdcr(uic->dcrbase + UIC_SR, 1 << (31-src));
-- spin_unlock_irqrestore(&uic->lock, flags);
-+ raw_spin_unlock_irqrestore(&uic->lock, flags);
- }
+ When GENERIC_CLOCKEVENTS is defined, the third timer channel
+- may be used as a clock event device supporting oneshot mode
+- (delays of up to two seconds) based on the 32 KiHz clock.
++ may be used as a clock event device supporting oneshot mode.
- static void uic_mask_ack_irq(struct irq_data *d)
-@@ -104,7 +104,7 @@ static void uic_mask_ack_irq(struct irq_
- u32 er, sr;
+ config ATMEL_TCB_CLKSRC_BLOCK
+ int
+@@ -112,6 +112,14 @@ config ATMEL_TCB_CLKSRC_BLOCK
+ TC can be used for other purposes, such as PWM generation and
+ interval timing.
- sr = 1 << (31-src);
-- spin_lock_irqsave(&uic->lock, flags);
-+ raw_spin_lock_irqsave(&uic->lock, flags);
- er = mfdcr(uic->dcrbase + UIC_ER);
- er &= ~sr;
- mtdcr(uic->dcrbase + UIC_ER, er);
-@@ -118,7 +118,7 @@ static void uic_mask_ack_irq(struct irq_
- */
- if (!irqd_is_level_type(d))
- mtdcr(uic->dcrbase + UIC_SR, sr);
-- spin_unlock_irqrestore(&uic->lock, flags);
-+ raw_spin_unlock_irqrestore(&uic->lock, flags);
- }
++config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
++ bool "TC Block use 32 KiHz clock"
++ depends on ATMEL_TCB_CLKSRC
++ default y if !PREEMPT_RT_FULL
++ help
++ Select this to use 32 KiHz base clock rate as TC block clock
++ source for clock events.
++
+ config IBM_ASM
+ tristate "Device driver for IBM RSA service processor"
+ depends on X86 && PCI && INPUT && EXPERIMENTAL
+@@ -133,6 +141,35 @@ config IBM_ASM
+ for information on the specific driver level and support statement
+ for your IBM server.
- static int uic_set_irq_type(struct irq_data *d, unsigned int flow_type)
-@@ -152,7 +152,7 @@ static int uic_set_irq_type(struct irq_d
-
- mask = ~(1 << (31 - src));
-
-- spin_lock_irqsave(&uic->lock, flags);
-+ raw_spin_lock_irqsave(&uic->lock, flags);
- tr = mfdcr(uic->dcrbase + UIC_TR);
- pr = mfdcr(uic->dcrbase + UIC_PR);
- tr = (tr & mask) | (trigger << (31-src));
-@@ -161,7 +161,7 @@ static int uic_set_irq_type(struct irq_d
- mtdcr(uic->dcrbase + UIC_PR, pr);
- mtdcr(uic->dcrbase + UIC_TR, tr);
-
-- spin_unlock_irqrestore(&uic->lock, flags);
-+ raw_spin_unlock_irqrestore(&uic->lock, flags);
-
- return 0;
- }
-@@ -254,7 +254,7 @@ static struct uic * __init uic_init_one(
- if (! uic)
- return NULL; /* FIXME: panic? */
-
-- spin_lock_init(&uic->lock);
-+ raw_spin_lock_init(&uic->lock);
- indexp = of_get_property(node, "cell-index", &len);
- if (!indexp || (len != sizeof(u32))) {
- printk(KERN_ERR "uic: Device node %s has missing or invalid "
-Index: linux-3.0/drivers/dca/dca-core.c
-===================================================================
---- linux-3.0.orig/drivers/dca/dca-core.c
-+++ linux-3.0/drivers/dca/dca-core.c
-@@ -35,7 +35,7 @@ MODULE_VERSION(DCA_VERSION);
- MODULE_LICENSE("GPL");
- MODULE_AUTHOR("Intel Corporation");
-
--static DEFINE_SPINLOCK(dca_lock);
-+static DEFINE_RAW_SPINLOCK(dca_lock);
-
- static LIST_HEAD(dca_domains);
-
-@@ -101,10 +101,10 @@ static void unregister_dca_providers(voi
-
- INIT_LIST_HEAD(&unregistered_providers);
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
-
- if (list_empty(&dca_domains)) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return;
- }
-
-@@ -116,7 +116,7 @@ static void unregister_dca_providers(voi
-
- dca_free_domain(domain);
-
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- list_for_each_entry_safe(dca, _dca, &unregistered_providers, node) {
- dca_sysfs_remove_provider(dca);
-@@ -144,13 +144,8 @@ static struct dca_domain *dca_get_domain
- domain = dca_find_domain(rc);
-
- if (!domain) {
-- if (dca_provider_ioat_ver_3_0(dev) && !list_empty(&dca_domains)) {
-+ if (dca_provider_ioat_ver_3_0(dev) && !list_empty(&dca_domains))
- dca_providers_blocked = 1;
-- } else {
-- domain = dca_allocate_domain(rc);
-- if (domain)
-- list_add(&domain->node, &dca_domains);
-- }
- }
-
- return domain;
-@@ -198,19 +193,19 @@ int dca_add_requester(struct device *dev
- if (!dev)
- return -EFAULT;
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
-
- /* check if the requester has not been added already */
- dca = dca_find_provider_by_dev(dev);
- if (dca) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return -EEXIST;
- }
-
- pci_rc = dca_pci_rc_from_dev(dev);
- domain = dca_find_domain(pci_rc);
- if (!domain) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return -ENODEV;
- }
-
-@@ -220,17 +215,17 @@ int dca_add_requester(struct device *dev
- break;
- }
-
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- if (slot < 0)
- return slot;
-
- err = dca_sysfs_add_req(dca, dev, slot);
- if (err) {
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
- if (dca == dca_find_provider_by_dev(dev))
- dca->ops->remove_requester(dca, dev);
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return err;
- }
-
-@@ -251,14 +246,14 @@ int dca_remove_requester(struct device *
- if (!dev)
- return -EFAULT;
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
- dca = dca_find_provider_by_dev(dev);
- if (!dca) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return -ENODEV;
- }
- slot = dca->ops->remove_requester(dca, dev);
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- if (slot < 0)
- return slot;
-@@ -280,16 +275,16 @@ u8 dca_common_get_tag(struct device *dev
- u8 tag;
- unsigned long flags;
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
-
- dca = dca_find_provider_by_dev(dev);
- if (!dca) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return -ENODEV;
- }
- tag = dca->ops->get_tag(dca, dev, cpu);
-
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return tag;
- }
-
-@@ -360,36 +355,51 @@ int register_dca_provider(struct dca_pro
- {
- int err;
- unsigned long flags;
-- struct dca_domain *domain;
-+ struct dca_domain *domain, *newdomain = NULL;
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
- if (dca_providers_blocked) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- return -ENODEV;
- }
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- err = dca_sysfs_add_provider(dca, dev);
- if (err)
- return err;
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
- domain = dca_get_domain(dev);
- if (!domain) {
-+ struct pci_bus *rc;
++config HWLAT_DETECTOR
++ tristate "Testing module to detect hardware-induced latencies"
++ depends on DEBUG_FS
++ depends on RING_BUFFER
++ default m
++ ---help---
++ A simple hardware latency detector. Use this module to detect
++ large latencies introduced by the behavior of the underlying
++ system firmware external to Linux. We do this using periodic
++ use of stop_machine to grab all available CPUs and measure
++ for unexplainable gaps in the CPU timestamp counter(s). By
++ default, the module is not enabled until the "enable" file
++ within the "hwlat_detector" debugfs directory is toggled.
+
- if (dca_providers_blocked) {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
- dca_sysfs_remove_provider(dca);
- unregister_dca_providers();
-- } else {
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ return -ENODEV;
-+ }
++ This module is often used to detect SMI (System Management
++ Interrupts) on x86 systems, though is not x86 specific. To
++ this end, we default to using a sample window of 1 second,
++ during which we will sample for 0.5 seconds. If an SMI or
++ similar event occurs during that time, it is recorded
++ into an 8K samples global ring buffer until retreived.
+
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-+ rc = dca_pci_rc_from_dev(dev);
-+ newdomain = dca_allocate_domain(rc);
-+ if (!newdomain)
-+ return -ENODEV;
-+ raw_spin_lock_irqsave(&dca_lock, flags);
-+ /* Recheck, we might have raced after dropping the lock */
-+ domain = dca_get_domain(dev);
-+ if (!domain) {
-+ domain = newdomain;
-+ newdomain = NULL;
-+ list_add(&domain->node, &dca_domains);
- }
-- return -ENODEV;
- }
- list_add(&dca->node, &domain->dca_providers);
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- blocking_notifier_call_chain(&dca_provider_chain,
- DCA_PROVIDER_ADD, NULL);
-+ kfree(newdomain);
- return 0;
- }
- EXPORT_SYMBOL_GPL(register_dca_provider);
-@@ -407,7 +417,7 @@ void unregister_dca_provider(struct dca_
- blocking_notifier_call_chain(&dca_provider_chain,
- DCA_PROVIDER_REMOVE, NULL);
-
-- spin_lock_irqsave(&dca_lock, flags);
-+ raw_spin_lock_irqsave(&dca_lock, flags);
-
- list_del(&dca->node);
-
-@@ -416,7 +426,7 @@ void unregister_dca_provider(struct dca_
- if (list_empty(&domain->dca_providers))
- dca_free_domain(domain);
-
-- spin_unlock_irqrestore(&dca_lock, flags);
-+ raw_spin_unlock_irqrestore(&dca_lock, flags);
-
- dca_sysfs_remove_provider(dca);
- }
-Index: linux-3.0/arch/arm/common/gic.c
-===================================================================
---- linux-3.0.orig/arch/arm/common/gic.c
-+++ linux-3.0/arch/arm/common/gic.c
-@@ -33,7 +33,7 @@
- #include <asm/mach/irq.h>
- #include <asm/hardware/gic.h>
-
--static DEFINE_SPINLOCK(irq_controller_lock);
-+static DEFINE_RAW_SPINLOCK(irq_controller_lock);
-
- /* Address of GIC 0 CPU interface */
- void __iomem *gic_cpu_base_addr __read_mostly;
-@@ -88,30 +88,30 @@ static void gic_mask_irq(struct irq_data
- {
- u32 mask = 1 << (d->irq % 32);
-
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
- writel_relaxed(mask, gic_dist_base(d) + GIC_DIST_ENABLE_CLEAR + (gic_irq(d) / 32) * 4);
- if (gic_arch_extn.irq_mask)
- gic_arch_extn.irq_mask(d);
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
- }
-
- static void gic_unmask_irq(struct irq_data *d)
- {
- u32 mask = 1 << (d->irq % 32);
-
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
- if (gic_arch_extn.irq_unmask)
- gic_arch_extn.irq_unmask(d);
- writel_relaxed(mask, gic_dist_base(d) + GIC_DIST_ENABLE_SET + (gic_irq(d) / 32) * 4);
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
- }
-
- static void gic_eoi_irq(struct irq_data *d)
- {
- if (gic_arch_extn.irq_eoi) {
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
- gic_arch_extn.irq_eoi(d);
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
- }
-
- writel_relaxed(gic_irq(d), gic_cpu_base(d) + GIC_CPU_EOI);
-@@ -135,7 +135,7 @@ static int gic_set_type(struct irq_data
- if (type != IRQ_TYPE_LEVEL_HIGH && type != IRQ_TYPE_EDGE_RISING)
- return -EINVAL;
-
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
-
- if (gic_arch_extn.irq_set_type)
- gic_arch_extn.irq_set_type(d, type);
-@@ -160,7 +160,7 @@ static int gic_set_type(struct irq_data
- if (enabled)
- writel_relaxed(enablemask, base + GIC_DIST_ENABLE_SET + enableoff);
-
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
-
- return 0;
- }
-@@ -188,11 +188,11 @@ static int gic_set_affinity(struct irq_d
- mask = 0xff << shift;
- bit = 1 << (cpu + shift);
-
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
- d->node = cpu;
- val = readl_relaxed(reg) & ~mask;
- writel_relaxed(val | bit, reg);
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
-
- return 0;
- }
-@@ -222,9 +222,9 @@ static void gic_handle_cascade_irq(unsig
-
- chained_irq_enter(chip, desc);
-
-- spin_lock(&irq_controller_lock);
-+ raw_spin_lock(&irq_controller_lock);
- status = readl_relaxed(chip_data->cpu_base + GIC_CPU_INTACK);
-- spin_unlock(&irq_controller_lock);
-+ raw_spin_unlock(&irq_controller_lock);
-
- gic_irq = (status & 0x3ff);
- if (gic_irq == 1023)
-Index: linux-3.0/arch/arm/include/asm/dma.h
-===================================================================
---- linux-3.0.orig/arch/arm/include/asm/dma.h
-+++ linux-3.0/arch/arm/include/asm/dma.h
-@@ -33,18 +33,18 @@
- #define DMA_MODE_CASCADE 0xc0
- #define DMA_AUTOINIT 0x10
-
--extern spinlock_t dma_spin_lock;
-+extern raw_spinlock_t dma_spin_lock;
-
- static inline unsigned long claim_dma_lock(void)
- {
- unsigned long flags;
-- spin_lock_irqsave(&dma_spin_lock, flags);
-+ raw_spin_lock_irqsave(&dma_spin_lock, flags);
- return flags;
- }
-
- static inline void release_dma_lock(unsigned long flags)
- {
-- spin_unlock_irqrestore(&dma_spin_lock, flags);
-+ raw_spin_unlock_irqrestore(&dma_spin_lock, flags);
- }
-
- /* Clear the 'DMA Pointer Flip Flop'.
-Index: linux-3.0/arch/arm/include/asm/mmu.h
-===================================================================
---- linux-3.0.orig/arch/arm/include/asm/mmu.h
-+++ linux-3.0/arch/arm/include/asm/mmu.h
-@@ -6,7 +6,7 @@
- typedef struct {
- #ifdef CONFIG_CPU_HAS_ASID
- unsigned int id;
-- spinlock_t id_lock;
-+ raw_spinlock_t id_lock;
- #endif
- unsigned int kvm_seq;
- } mm_context_t;
-@@ -16,7 +16,7 @@ typedef struct {
-
- /* init_mm.context.id_lock should be initialized. */
- #define INIT_MM_CONTEXT(name) \
-- .context.id_lock = __SPIN_LOCK_UNLOCKED(name.context.id_lock),
-+ .context.id_lock = __RAW_SPIN_LOCK_UNLOCKED(name.context.id_lock),
- #else
- #define ASID(mm) (0)
- #endif
-Index: linux-3.0/arch/arm/kernel/dma.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/dma.c
-+++ linux-3.0/arch/arm/kernel/dma.c
-@@ -23,7 +23,7 @@
-
- #include <asm/mach/dma.h>
-
--DEFINE_SPINLOCK(dma_spin_lock);
-+DEFINE_RAW_SPINLOCK(dma_spin_lock);
- EXPORT_SYMBOL(dma_spin_lock);
-
- static dma_t *dma_chan[MAX_DMA_CHANNELS];
-Index: linux-3.0/arch/arm/kernel/traps.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/traps.c
-+++ linux-3.0/arch/arm/kernel/traps.c
-@@ -255,7 +255,7 @@ static int __die(const char *str, int er
- return ret;
- }
-
--static DEFINE_SPINLOCK(die_lock);
-+static DEFINE_RAW_SPINLOCK(die_lock);
-
- /*
- * This function is protected against re-entrancy.
-@@ -267,7 +267,7 @@ void die(const char *str, struct pt_regs
-
- oops_enter();
-
-- spin_lock_irq(&die_lock);
-+ raw_spin_lock_irq(&die_lock);
- console_verbose();
- bust_spinlocks(1);
- ret = __die(str, err, thread, regs);
-@@ -277,7 +277,7 @@ void die(const char *str, struct pt_regs
-
- bust_spinlocks(0);
- add_taint(TAINT_DIE);
-- spin_unlock_irq(&die_lock);
-+ raw_spin_unlock_irq(&die_lock);
- oops_exit();
-
- if (in_interrupt())
-@@ -302,24 +302,24 @@ void arm_notify_die(const char *str, str
- }
-
- static LIST_HEAD(undef_hook);
--static DEFINE_SPINLOCK(undef_lock);
-+static DEFINE_RAW_SPINLOCK(undef_lock);
-
- void register_undef_hook(struct undef_hook *hook)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&undef_lock, flags);
-+ raw_spin_lock_irqsave(&undef_lock, flags);
- list_add(&hook->node, &undef_hook);
-- spin_unlock_irqrestore(&undef_lock, flags);
-+ raw_spin_unlock_irqrestore(&undef_lock, flags);
- }
-
- void unregister_undef_hook(struct undef_hook *hook)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&undef_lock, flags);
-+ raw_spin_lock_irqsave(&undef_lock, flags);
- list_del(&hook->node);
-- spin_unlock_irqrestore(&undef_lock, flags);
-+ raw_spin_unlock_irqrestore(&undef_lock, flags);
- }
-
- static int call_undef_hook(struct pt_regs *regs, unsigned int instr)
-@@ -328,12 +328,12 @@ static int call_undef_hook(struct pt_reg
- unsigned long flags;
- int (*fn)(struct pt_regs *regs, unsigned int instr) = NULL;
-
-- spin_lock_irqsave(&undef_lock, flags);
-+ raw_spin_lock_irqsave(&undef_lock, flags);
- list_for_each_entry(hook, &undef_hook, node)
- if ((instr & hook->instr_mask) == hook->instr_val &&
- (regs->ARM_cpsr & hook->cpsr_mask) == hook->cpsr_val)
- fn = hook->fn;
-- spin_unlock_irqrestore(&undef_lock, flags);
-+ raw_spin_unlock_irqrestore(&undef_lock, flags);
-
- return fn ? fn(regs, instr) : 1;
- }
-Index: linux-3.0/arch/arm/mach-footbridge/include/mach/hardware.h
-===================================================================
---- linux-3.0.orig/arch/arm/mach-footbridge/include/mach/hardware.h
-+++ linux-3.0/arch/arm/mach-footbridge/include/mach/hardware.h
-@@ -93,7 +93,7 @@
- #define CPLD_FLASH_WR_ENABLE 1
-
- #ifndef __ASSEMBLY__
--extern spinlock_t nw_gpio_lock;
-+extern raw_spinlock_t nw_gpio_lock;
- extern void nw_gpio_modify_op(unsigned int mask, unsigned int set);
- extern void nw_gpio_modify_io(unsigned int mask, unsigned int in);
- extern unsigned int nw_gpio_read(void);
-Index: linux-3.0/arch/arm/mach-footbridge/netwinder-hw.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-footbridge/netwinder-hw.c
-+++ linux-3.0/arch/arm/mach-footbridge/netwinder-hw.c
-@@ -68,7 +68,7 @@ static inline void wb977_ww(int reg, int
- /*
- * This is a lock for accessing ports GP1_IO_BASE and GP2_IO_BASE
- */
--DEFINE_SPINLOCK(nw_gpio_lock);
-+DEFINE_RAW_SPINLOCK(nw_gpio_lock);
- EXPORT_SYMBOL(nw_gpio_lock);
-
- static unsigned int current_gpio_op;
-@@ -327,9 +327,9 @@ static inline void wb977_init_gpio(void)
- /*
- * Set Group1/Group2 outputs
- */
-- spin_lock_irqsave(&nw_gpio_lock, flags);
-+ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
- nw_gpio_modify_op(-1, GPIO_RED_LED | GPIO_FAN);
-- spin_unlock_irqrestore(&nw_gpio_lock, flags);
-+ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
- }
-
- /*
-@@ -390,9 +390,9 @@ static void __init cpld_init(void)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&nw_gpio_lock, flags);
-+ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
- nw_cpld_modify(-1, CPLD_UNMUTE | CPLD_7111_DISABLE);
-- spin_unlock_irqrestore(&nw_gpio_lock, flags);
-+ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
- }
-
- static unsigned char rwa_unlock[] __initdata =
-@@ -616,9 +616,9 @@ static int __init nw_hw_init(void)
- cpld_init();
- rwa010_init();
-
-- spin_lock_irqsave(&nw_gpio_lock, flags);
-+ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
- nw_gpio_modify_op(GPIO_RED_LED|GPIO_GREEN_LED, DEFAULT_LEDS);
-- spin_unlock_irqrestore(&nw_gpio_lock, flags);
-+ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
- }
- return 0;
- }
-Index: linux-3.0/arch/arm/mach-footbridge/netwinder-leds.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-footbridge/netwinder-leds.c
-+++ linux-3.0/arch/arm/mach-footbridge/netwinder-leds.c
-@@ -31,13 +31,13 @@
- static char led_state;
- static char hw_led_state;
-
--static DEFINE_SPINLOCK(leds_lock);
-+static DEFINE_RAW_SPINLOCK(leds_lock);
-
- static void netwinder_leds_event(led_event_t evt)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&leds_lock, flags);
-+ raw_spin_lock_irqsave(&leds_lock, flags);
-
- switch (evt) {
- case led_start:
-@@ -117,12 +117,12 @@ static void netwinder_leds_event(led_eve
- break;
- }
-
-- spin_unlock_irqrestore(&leds_lock, flags);
-+ raw_spin_unlock_irqrestore(&leds_lock, flags);
-
- if (led_state & LED_STATE_ENABLED) {
-- spin_lock_irqsave(&nw_gpio_lock, flags);
-+ raw_spin_lock_irqsave(&nw_gpio_lock, flags);
- nw_gpio_modify_op(GPIO_RED_LED | GPIO_GREEN_LED, hw_led_state);
-- spin_unlock_irqrestore(&nw_gpio_lock, flags);
-+ raw_spin_unlock_irqrestore(&nw_gpio_lock, flags);
- }
- }
-
-Index: linux-3.0/arch/arm/mach-integrator/core.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-integrator/core.c
-+++ linux-3.0/arch/arm/mach-integrator/core.c
-@@ -205,7 +205,7 @@ static struct amba_pl010_data integrator
-
- #define CM_CTRL IO_ADDRESS(INTEGRATOR_HDR_CTRL)
-
--static DEFINE_SPINLOCK(cm_lock);
-+static DEFINE_RAW_SPINLOCK(cm_lock);
-
- /**
- * cm_control - update the CM_CTRL register.
-@@ -217,10 +217,10 @@ void cm_control(u32 mask, u32 set)
- unsigned long flags;
- u32 val;
-
-- spin_lock_irqsave(&cm_lock, flags);
-+ raw_spin_lock_irqsave(&cm_lock, flags);
- val = readl(CM_CTRL) & ~mask;
- writel(val | set, CM_CTRL);
-- spin_unlock_irqrestore(&cm_lock, flags);
-+ raw_spin_unlock_irqrestore(&cm_lock, flags);
- }
-
- EXPORT_SYMBOL(cm_control);
-Index: linux-3.0/arch/arm/mach-integrator/pci_v3.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-integrator/pci_v3.c
-+++ linux-3.0/arch/arm/mach-integrator/pci_v3.c
-@@ -163,7 +163,7 @@
- * 7:2 register number
- *
- */
--static DEFINE_SPINLOCK(v3_lock);
-+static DEFINE_RAW_SPINLOCK(v3_lock);
-
- #define PCI_BUS_NONMEM_START 0x00000000
- #define PCI_BUS_NONMEM_SIZE SZ_256M
-@@ -284,7 +284,7 @@ static int v3_read_config(struct pci_bus
- unsigned long flags;
- u32 v;
-
-- spin_lock_irqsave(&v3_lock, flags);
-+ raw_spin_lock_irqsave(&v3_lock, flags);
- addr = v3_open_config_window(bus, devfn, where);
-
- switch (size) {
-@@ -302,7 +302,7 @@ static int v3_read_config(struct pci_bus
- }
-
- v3_close_config_window();
-- spin_unlock_irqrestore(&v3_lock, flags);
-+ raw_spin_unlock_irqrestore(&v3_lock, flags);
-
- *val = v;
- return PCIBIOS_SUCCESSFUL;
-@@ -314,7 +314,7 @@ static int v3_write_config(struct pci_bu
- unsigned long addr;
- unsigned long flags;
-
-- spin_lock_irqsave(&v3_lock, flags);
-+ raw_spin_lock_irqsave(&v3_lock, flags);
- addr = v3_open_config_window(bus, devfn, where);
-
- switch (size) {
-@@ -335,7 +335,7 @@ static int v3_write_config(struct pci_bu
- }
-
- v3_close_config_window();
-- spin_unlock_irqrestore(&v3_lock, flags);
-+ raw_spin_unlock_irqrestore(&v3_lock, flags);
-
- return PCIBIOS_SUCCESSFUL;
- }
-@@ -510,7 +510,7 @@ void __init pci_v3_preinit(void)
- hook_fault_code(8, v3_pci_fault, SIGBUS, 0, "external abort on non-linefetch");
- hook_fault_code(10, v3_pci_fault, SIGBUS, 0, "external abort on non-linefetch");
-
-- spin_lock_irqsave(&v3_lock, flags);
-+ raw_spin_lock_irqsave(&v3_lock, flags);
-
- /*
- * Unlock V3 registers, but only if they were previously locked.
-@@ -583,7 +583,7 @@ void __init pci_v3_preinit(void)
- printk(KERN_ERR "PCI: unable to grab PCI error "
- "interrupt: %d\n", ret);
-
-- spin_unlock_irqrestore(&v3_lock, flags);
-+ raw_spin_unlock_irqrestore(&v3_lock, flags);
- }
-
- void __init pci_v3_postinit(void)
-Index: linux-3.0/arch/arm/mach-ixp4xx/common-pci.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-ixp4xx/common-pci.c
-+++ linux-3.0/arch/arm/mach-ixp4xx/common-pci.c
-@@ -54,7 +54,7 @@ unsigned long ixp4xx_pci_reg_base = 0;
- * these transactions are atomic or we will end up
- * with corrupt data on the bus or in a driver.
- */
--static DEFINE_SPINLOCK(ixp4xx_pci_lock);
-+static DEFINE_RAW_SPINLOCK(ixp4xx_pci_lock);
-
- /*
- * Read from PCI config space
-@@ -62,10 +62,10 @@ static DEFINE_SPINLOCK(ixp4xx_pci_lock);
- static void crp_read(u32 ad_cbe, u32 *data)
- {
- unsigned long flags;
-- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-+ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
- *PCI_CRP_AD_CBE = ad_cbe;
- *data = *PCI_CRP_RDATA;
-- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
-+ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
- }
-
- /*
-@@ -74,10 +74,10 @@ static void crp_read(u32 ad_cbe, u32 *da
- static void crp_write(u32 ad_cbe, u32 data)
- {
- unsigned long flags;
-- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-+ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
- *PCI_CRP_AD_CBE = CRP_AD_CBE_WRITE | ad_cbe;
- *PCI_CRP_WDATA = data;
-- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
-+ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
- }
-
- static inline int check_master_abort(void)
-@@ -101,7 +101,7 @@ int ixp4xx_pci_read_errata(u32 addr, u32
- int retval = 0;
- int i;
-
-- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-+ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-
- *PCI_NP_AD = addr;
-
-@@ -118,7 +118,7 @@ int ixp4xx_pci_read_errata(u32 addr, u32
- if(check_master_abort())
- retval = 1;
-
-- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
-+ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
- return retval;
- }
-
-@@ -127,7 +127,7 @@ int ixp4xx_pci_read_no_errata(u32 addr,
- unsigned long flags;
- int retval = 0;
-
-- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-+ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-
- *PCI_NP_AD = addr;
-
-@@ -140,7 +140,7 @@ int ixp4xx_pci_read_no_errata(u32 addr,
- if(check_master_abort())
- retval = 1;
-
-- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
-+ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
- return retval;
- }
-
-@@ -149,7 +149,7 @@ int ixp4xx_pci_write(u32 addr, u32 cmd,
- unsigned long flags;
- int retval = 0;
-
-- spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-+ raw_spin_lock_irqsave(&ixp4xx_pci_lock, flags);
-
- *PCI_NP_AD = addr;
-
-@@ -162,7 +162,7 @@ int ixp4xx_pci_write(u32 addr, u32 cmd,
- if(check_master_abort())
- retval = 1;
-
-- spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
-+ raw_spin_unlock_irqrestore(&ixp4xx_pci_lock, flags);
- return retval;
- }
-
-Index: linux-3.0/arch/arm/mach-shark/leds.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-shark/leds.c
-+++ linux-3.0/arch/arm/mach-shark/leds.c
-@@ -36,7 +36,7 @@ static char led_state;
- static short hw_led_state;
- static short saved_state;
-
--static DEFINE_SPINLOCK(leds_lock);
-+static DEFINE_RAW_SPINLOCK(leds_lock);
-
- short sequoia_read(int addr) {
- outw(addr,0x24);
-@@ -52,7 +52,7 @@ static void sequoia_leds_event(led_event
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&leds_lock, flags);
-+ raw_spin_lock_irqsave(&leds_lock, flags);
-
- hw_led_state = sequoia_read(0x09);
-
-@@ -144,7 +144,7 @@ static void sequoia_leds_event(led_event
- if (led_state & LED_STATE_ENABLED)
- sequoia_write(hw_led_state,0x09);
-
-- spin_unlock_irqrestore(&leds_lock, flags);
-+ raw_spin_unlock_irqrestore(&leds_lock, flags);
- }
-
- static int __init leds_init(void)
-Index: linux-3.0/arch/arm/mm/cache-l2x0.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/cache-l2x0.c
-+++ linux-3.0/arch/arm/mm/cache-l2x0.c
-@@ -26,7 +26,7 @@
- #define CACHE_LINE_SIZE 32
-
- static void __iomem *l2x0_base;
--static DEFINE_SPINLOCK(l2x0_lock);
-+static DEFINE_RAW_SPINLOCK(l2x0_lock);
- static uint32_t l2x0_way_mask; /* Bitmask of active ways */
- static uint32_t l2x0_size;
-
-@@ -115,9 +115,9 @@ static void l2x0_cache_sync(void)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void __l2x0_flush_all(void)
-@@ -134,9 +134,9 @@ static void l2x0_flush_all(void)
- unsigned long flags;
-
- /* clean all ways */
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- __l2x0_flush_all();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_clean_all(void)
-@@ -144,11 +144,11 @@ static void l2x0_clean_all(void)
- unsigned long flags;
-
- /* clean all ways */
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_CLEAN_WAY);
- cache_wait_way(l2x0_base + L2X0_CLEAN_WAY, l2x0_way_mask);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_inv_all(void)
-@@ -156,13 +156,13 @@ static void l2x0_inv_all(void)
- unsigned long flags;
-
- /* invalidate all ways */
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- /* Invalidating when L2 is enabled is a nono */
- BUG_ON(readl(l2x0_base + L2X0_CTRL) & 1);
- writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
- cache_wait_way(l2x0_base + L2X0_INV_WAY, l2x0_way_mask);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_inv_range(unsigned long start, unsigned long end)
-@@ -170,7 +170,7 @@ static void l2x0_inv_range(unsigned long
- void __iomem *base = l2x0_base;
- unsigned long flags;
-
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- if (start & (CACHE_LINE_SIZE - 1)) {
- start &= ~(CACHE_LINE_SIZE - 1);
- debug_writel(0x03);
-@@ -195,13 +195,13 @@ static void l2x0_inv_range(unsigned long
- }
-
- if (blk_end < end) {
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- }
- }
- cache_wait(base + L2X0_INV_LINE_PA, 1);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_clean_range(unsigned long start, unsigned long end)
-@@ -214,7 +214,7 @@ static void l2x0_clean_range(unsigned lo
- return;
- }
-
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- start &= ~(CACHE_LINE_SIZE - 1);
- while (start < end) {
- unsigned long blk_end = start + min(end - start, 4096UL);
-@@ -225,13 +225,13 @@ static void l2x0_clean_range(unsigned lo
- }
-
- if (blk_end < end) {
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- }
- }
- cache_wait(base + L2X0_CLEAN_LINE_PA, 1);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_flush_range(unsigned long start, unsigned long end)
-@@ -244,7 +244,7 @@ static void l2x0_flush_range(unsigned lo
- return;
- }
-
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- start &= ~(CACHE_LINE_SIZE - 1);
- while (start < end) {
- unsigned long blk_end = start + min(end - start, 4096UL);
-@@ -257,24 +257,24 @@ static void l2x0_flush_range(unsigned lo
- debug_writel(0x00);
-
- if (blk_end < end) {
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- }
- }
- cache_wait(base + L2X0_CLEAN_INV_LINE_PA, 1);
- cache_sync();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- static void l2x0_disable(void)
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&l2x0_lock, flags);
-+ raw_spin_lock_irqsave(&l2x0_lock, flags);
- __l2x0_flush_all();
- writel_relaxed(0, l2x0_base + L2X0_CTRL);
- dsb();
-- spin_unlock_irqrestore(&l2x0_lock, flags);
-+ raw_spin_unlock_irqrestore(&l2x0_lock, flags);
- }
-
- void __init l2x0_init(void __iomem *base, __u32 aux_val, __u32 aux_mask)
-Index: linux-3.0/arch/arm/mm/context.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/context.c
-+++ linux-3.0/arch/arm/mm/context.c
-@@ -16,7 +16,7 @@
- #include <asm/mmu_context.h>
- #include <asm/tlbflush.h>
-
--static DEFINE_SPINLOCK(cpu_asid_lock);
-+static DEFINE_RAW_SPINLOCK(cpu_asid_lock);
- unsigned int cpu_last_asid = ASID_FIRST_VERSION;
- #ifdef CONFIG_SMP
- DEFINE_PER_CPU(struct mm_struct *, current_mm);
-@@ -31,7 +31,7 @@ DEFINE_PER_CPU(struct mm_struct *, curre
- void __init_new_context(struct task_struct *tsk, struct mm_struct *mm)
- {
- mm->context.id = 0;
-- spin_lock_init(&mm->context.id_lock);
-+ raw_spin_lock_init(&mm->context.id_lock);
- }
-
- static void flush_context(void)
-@@ -58,7 +58,7 @@ static void set_mm_context(struct mm_str
- * the broadcast. This function is also called via IPI so the
- * mm->context.id_lock has to be IRQ-safe.
- */
-- spin_lock_irqsave(&mm->context.id_lock, flags);
-+ raw_spin_lock_irqsave(&mm->context.id_lock, flags);
- if (likely((mm->context.id ^ cpu_last_asid) >> ASID_BITS)) {
- /*
- * Old version of ASID found. Set the new one and
-@@ -67,7 +67,7 @@ static void set_mm_context(struct mm_str
- mm->context.id = asid;
- cpumask_clear(mm_cpumask(mm));
- }
-- spin_unlock_irqrestore(&mm->context.id_lock, flags);
-+ raw_spin_unlock_irqrestore(&mm->context.id_lock, flags);
-
- /*
- * Set the mm_cpumask(mm) bit for the current CPU.
-@@ -117,7 +117,7 @@ void __new_context(struct mm_struct *mm)
- {
- unsigned int asid;
-
-- spin_lock(&cpu_asid_lock);
-+ raw_spin_lock(&cpu_asid_lock);
- #ifdef CONFIG_SMP
- /*
- * Check the ASID again, in case the change was broadcast from
-@@ -125,7 +125,7 @@ void __new_context(struct mm_struct *mm)
- */
- if (unlikely(((mm->context.id ^ cpu_last_asid) >> ASID_BITS) == 0)) {
- cpumask_set_cpu(smp_processor_id(), mm_cpumask(mm));
-- spin_unlock(&cpu_asid_lock);
-+ raw_spin_unlock(&cpu_asid_lock);
- return;
- }
- #endif
-@@ -153,5 +153,5 @@ void __new_context(struct mm_struct *mm)
- }
-
- set_mm_context(mm, asid);
-- spin_unlock(&cpu_asid_lock);
-+ raw_spin_unlock(&cpu_asid_lock);
- }
-Index: linux-3.0/arch/arm/mm/copypage-v4mc.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/copypage-v4mc.c
-+++ linux-3.0/arch/arm/mm/copypage-v4mc.c
-@@ -30,7 +30,7 @@
- #define minicache_pgprot __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | \
- L_PTE_MT_MINICACHE)
-
--static DEFINE_SPINLOCK(minicache_lock);
-+static DEFINE_RAW_SPINLOCK(minicache_lock);
-
- /*
- * ARMv4 mini-dcache optimised copy_user_highpage
-@@ -76,14 +76,14 @@ void v4_mc_copy_user_highpage(struct pag
- if (!test_and_set_bit(PG_dcache_clean, &from->flags))
- __flush_dcache_page(page_mapping(from), from);
-
-- spin_lock(&minicache_lock);
-+ raw_spin_lock(&minicache_lock);
-
- set_pte_ext(TOP_PTE(0xffff8000), pfn_pte(page_to_pfn(from), minicache_pgprot), 0);
- flush_tlb_kernel_page(0xffff8000);
-
- mc_copy_user_page((void *)0xffff8000, kto);
-
-- spin_unlock(&minicache_lock);
-+ raw_spin_unlock(&minicache_lock);
-
- kunmap_atomic(kto, KM_USER1);
- }
-Index: linux-3.0/arch/arm/mm/copypage-v6.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/copypage-v6.c
-+++ linux-3.0/arch/arm/mm/copypage-v6.c
-@@ -27,7 +27,7 @@
- #define from_address (0xffff8000)
- #define to_address (0xffffc000)
-
--static DEFINE_SPINLOCK(v6_lock);
-+static DEFINE_RAW_SPINLOCK(v6_lock);
-
- /*
- * Copy the user page. No aliasing to deal with so we can just
-@@ -89,7 +89,7 @@ static void v6_copy_user_highpage_aliasi
- * Now copy the page using the same cache colour as the
- * pages ultimate destination.
- */
-- spin_lock(&v6_lock);
-+ raw_spin_lock(&v6_lock);
-
- set_pte_ext(TOP_PTE(from_address) + offset, pfn_pte(page_to_pfn(from), PAGE_KERNEL), 0);
- set_pte_ext(TOP_PTE(to_address) + offset, pfn_pte(page_to_pfn(to), PAGE_KERNEL), 0);
-@@ -102,7 +102,7 @@ static void v6_copy_user_highpage_aliasi
-
- copy_page((void *)kto, (void *)kfrom);
-
-- spin_unlock(&v6_lock);
-+ raw_spin_unlock(&v6_lock);
- }
-
- /*
-@@ -122,13 +122,13 @@ static void v6_clear_user_highpage_alias
- * Now clear the page using the same cache colour as
- * the pages ultimate destination.
- */
-- spin_lock(&v6_lock);
-+ raw_spin_lock(&v6_lock);
-
- set_pte_ext(TOP_PTE(to_address) + offset, pfn_pte(page_to_pfn(page), PAGE_KERNEL), 0);
- flush_tlb_kernel_page(to);
- clear_page((void *)to);
-
-- spin_unlock(&v6_lock);
-+ raw_spin_unlock(&v6_lock);
- }
-
- struct cpu_user_fns v6_user_fns __initdata = {
-Index: linux-3.0/arch/arm/mm/copypage-xscale.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/copypage-xscale.c
-+++ linux-3.0/arch/arm/mm/copypage-xscale.c
-@@ -32,7 +32,7 @@
- #define minicache_pgprot __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | \
- L_PTE_MT_MINICACHE)
-
--static DEFINE_SPINLOCK(minicache_lock);
-+static DEFINE_RAW_SPINLOCK(minicache_lock);
-
- /*
- * XScale mini-dcache optimised copy_user_highpage
-@@ -98,14 +98,14 @@ void xscale_mc_copy_user_highpage(struct
- if (!test_and_set_bit(PG_dcache_clean, &from->flags))
- __flush_dcache_page(page_mapping(from), from);
-
-- spin_lock(&minicache_lock);
-+ raw_spin_lock(&minicache_lock);
-
- set_pte_ext(TOP_PTE(COPYPAGE_MINICACHE), pfn_pte(page_to_pfn(from), minicache_pgprot), 0);
- flush_tlb_kernel_page(COPYPAGE_MINICACHE);
-
- mc_copy_user_page((void *)COPYPAGE_MINICACHE, kto);
-
-- spin_unlock(&minicache_lock);
-+ raw_spin_unlock(&minicache_lock);
-
- kunmap_atomic(kto, KM_USER1);
- }
-Index: linux-3.0/drivers/dma/ipu/ipu_irq.c
-===================================================================
---- linux-3.0.orig/drivers/dma/ipu/ipu_irq.c
-+++ linux-3.0/drivers/dma/ipu/ipu_irq.c
-@@ -81,7 +81,7 @@ static struct ipu_irq_map irq_map[CONFIG
- /* Protects allocations from the above array of maps */
- static DEFINE_MUTEX(map_lock);
- /* Protects register accesses and individual mappings */
--static DEFINE_SPINLOCK(bank_lock);
-+static DEFINE_RAW_SPINLOCK(bank_lock);
-
- static struct ipu_irq_map *src2map(unsigned int src)
- {
-@@ -101,11 +101,11 @@ static void ipu_irq_unmask(struct irq_da
- uint32_t reg;
- unsigned long lock_flags;
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
-
- bank = map->bank;
- if (!bank) {
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
- return;
- }
-@@ -114,7 +114,7 @@ static void ipu_irq_unmask(struct irq_da
- reg |= (1UL << (map->source & 31));
- ipu_write_reg(bank->ipu, reg, bank->control);
-
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- }
-
- static void ipu_irq_mask(struct irq_data *d)
-@@ -124,11 +124,11 @@ static void ipu_irq_mask(struct irq_data
- uint32_t reg;
- unsigned long lock_flags;
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
-
- bank = map->bank;
- if (!bank) {
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
- return;
- }
-@@ -137,7 +137,7 @@ static void ipu_irq_mask(struct irq_data
- reg &= ~(1UL << (map->source & 31));
- ipu_write_reg(bank->ipu, reg, bank->control);
-
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- }
-
- static void ipu_irq_ack(struct irq_data *d)
-@@ -146,17 +146,17 @@ static void ipu_irq_ack(struct irq_data
- struct ipu_irq_bank *bank;
- unsigned long lock_flags;
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
-
- bank = map->bank;
- if (!bank) {
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- pr_err("IPU: %s(%u) - unmapped!\n", __func__, d->irq);
- return;
- }
-
- ipu_write_reg(bank->ipu, 1UL << (map->source & 31), bank->status);
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
- }
-
- /**
-@@ -172,11 +172,11 @@ bool ipu_irq_status(unsigned int irq)
- unsigned long lock_flags;
- bool ret;
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- bank = map->bank;
- ret = bank && ipu_read_reg(bank->ipu, bank->status) &
- (1UL << (map->source & 31));
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
-
- return ret;
- }
-@@ -213,10 +213,10 @@ int ipu_irq_map(unsigned int source)
- if (irq_map[i].source < 0) {
- unsigned long lock_flags;
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- irq_map[i].source = source;
- irq_map[i].bank = irq_bank + source / 32;
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
-
- ret = irq_map[i].irq;
- pr_debug("IPU: mapped source %u to IRQ %u\n",
-@@ -252,10 +252,10 @@ int ipu_irq_unmap(unsigned int source)
- pr_debug("IPU: unmapped source %u from IRQ %u\n",
- source, irq_map[i].irq);
-
-- spin_lock_irqsave(&bank_lock, lock_flags);
-+ raw_spin_lock_irqsave(&bank_lock, lock_flags);
- irq_map[i].source = -EINVAL;
- irq_map[i].bank = NULL;
-- spin_unlock_irqrestore(&bank_lock, lock_flags);
-+ raw_spin_unlock_irqrestore(&bank_lock, lock_flags);
-
- ret = 0;
- break;
-@@ -276,7 +276,7 @@ static void ipu_irq_err(unsigned int irq
- for (i = IPU_IRQ_NR_FN_BANKS; i < IPU_IRQ_NR_BANKS; i++) {
- struct ipu_irq_bank *bank = irq_bank + i;
-
-- spin_lock(&bank_lock);
-+ raw_spin_lock(&bank_lock);
- status = ipu_read_reg(ipu, bank->status);
- /*
- * Don't think we have to clear all interrupts here, they will
-@@ -284,18 +284,18 @@ static void ipu_irq_err(unsigned int irq
- * might want to clear unhandled interrupts after the loop...
- */
- status &= ipu_read_reg(ipu, bank->control);
-- spin_unlock(&bank_lock);
-+ raw_spin_unlock(&bank_lock);
- while ((line = ffs(status))) {
- struct ipu_irq_map *map;
-
- line--;
- status &= ~(1UL << line);
-
-- spin_lock(&bank_lock);
-+ raw_spin_lock(&bank_lock);
- map = src2map(32 * i + line);
- if (map)
- irq = map->irq;
-- spin_unlock(&bank_lock);
-+ raw_spin_unlock(&bank_lock);
-
- if (!map) {
- pr_err("IPU: Interrupt on unmapped source %u bank %d\n",
-@@ -317,22 +317,22 @@ static void ipu_irq_fn(unsigned int irq,
- for (i = 0; i < IPU_IRQ_NR_FN_BANKS; i++) {
- struct ipu_irq_bank *bank = irq_bank + i;
-
-- spin_lock(&bank_lock);
-+ raw_spin_lock(&bank_lock);
- status = ipu_read_reg(ipu, bank->status);
- /* Not clearing all interrupts, see above */
- status &= ipu_read_reg(ipu, bank->control);
-- spin_unlock(&bank_lock);
-+ raw_spin_unlock(&bank_lock);
- while ((line = ffs(status))) {
- struct ipu_irq_map *map;
-
- line--;
- status &= ~(1UL << line);
-
-- spin_lock(&bank_lock);
-+ raw_spin_lock(&bank_lock);
- map = src2map(32 * i + line);
- if (map)
- irq = map->irq;
-- spin_unlock(&bank_lock);
-+ raw_spin_unlock(&bank_lock);
-
- if (!map) {
- pr_err("IPU: Interrupt on unmapped source %u bank %d\n",
-Index: linux-3.0/drivers/pci/dmar.c
-===================================================================
---- linux-3.0.orig/drivers/pci/dmar.c
-+++ linux-3.0/drivers/pci/dmar.c
-@@ -800,7 +800,7 @@ int alloc_iommu(struct dmar_drhd_unit *d
- (unsigned long long)iommu->cap,
- (unsigned long long)iommu->ecap);
-
-- spin_lock_init(&iommu->register_lock);
-+ raw_spin_lock_init(&iommu->register_lock);
-
- drhd->iommu = iommu;
- return 0;
-@@ -921,11 +921,11 @@ int qi_submit_sync(struct qi_desc *desc,
- restart:
- rc = 0;
-
-- spin_lock_irqsave(&qi->q_lock, flags);
-+ raw_spin_lock_irqsave(&qi->q_lock, flags);
- while (qi->free_cnt < 3) {
-- spin_unlock_irqrestore(&qi->q_lock, flags);
-+ raw_spin_unlock_irqrestore(&qi->q_lock, flags);
- cpu_relax();
-- spin_lock_irqsave(&qi->q_lock, flags);
-+ raw_spin_lock_irqsave(&qi->q_lock, flags);
- }
-
- index = qi->free_head;
-@@ -965,15 +965,15 @@ restart:
- if (rc)
- break;
-
-- spin_unlock(&qi->q_lock);
-+ raw_spin_unlock(&qi->q_lock);
- cpu_relax();
-- spin_lock(&qi->q_lock);
-+ raw_spin_lock(&qi->q_lock);
- }
-
- qi->desc_status[index] = QI_DONE;
-
- reclaim_free_desc(qi);
-- spin_unlock_irqrestore(&qi->q_lock, flags);
-+ raw_spin_unlock_irqrestore(&qi->q_lock, flags);
-
- if (rc == -EAGAIN)
- goto restart;
-@@ -1062,7 +1062,7 @@ void dmar_disable_qi(struct intel_iommu
- if (!ecap_qis(iommu->ecap))
- return;
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-
- sts = dmar_readq(iommu->reg + DMAR_GSTS_REG);
- if (!(sts & DMA_GSTS_QIES))
-@@ -1082,7 +1082,7 @@ void dmar_disable_qi(struct intel_iommu
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl,
- !(sts & DMA_GSTS_QIES), sts);
- end:
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- }
-
- /*
-@@ -1097,7 +1097,7 @@ static void __dmar_enable_qi(struct inte
- qi->free_head = qi->free_tail = 0;
- qi->free_cnt = QI_LENGTH;
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-
- /* write zero to the tail reg */
- writel(0, iommu->reg + DMAR_IQT_REG);
-@@ -1110,7 +1110,7 @@ static void __dmar_enable_qi(struct inte
- /* Make sure hardware complete it */
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_QIES), sts);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- }
-
- /*
-@@ -1159,7 +1159,7 @@ int dmar_enable_qi(struct intel_iommu *i
- qi->free_head = qi->free_tail = 0;
- qi->free_cnt = QI_LENGTH;
-
-- spin_lock_init(&qi->q_lock);
-+ raw_spin_lock_init(&qi->q_lock);
-
- __dmar_enable_qi(iommu);
-
-@@ -1225,11 +1225,11 @@ void dmar_msi_unmask(struct irq_data *da
- unsigned long flag;
-
- /* unmask it */
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- writel(0, iommu->reg + DMAR_FECTL_REG);
- /* Read a reg to force flush the post write */
- readl(iommu->reg + DMAR_FECTL_REG);
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- void dmar_msi_mask(struct irq_data *data)
-@@ -1238,11 +1238,11 @@ void dmar_msi_mask(struct irq_data *data
- struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
-
- /* mask it */
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- writel(DMA_FECTL_IM, iommu->reg + DMAR_FECTL_REG);
- /* Read a reg to force flush the post write */
- readl(iommu->reg + DMAR_FECTL_REG);
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- void dmar_msi_write(int irq, struct msi_msg *msg)
-@@ -1250,11 +1250,11 @@ void dmar_msi_write(int irq, struct msi_
- struct intel_iommu *iommu = irq_get_handler_data(irq);
- unsigned long flag;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- writel(msg->data, iommu->reg + DMAR_FEDATA_REG);
- writel(msg->address_lo, iommu->reg + DMAR_FEADDR_REG);
- writel(msg->address_hi, iommu->reg + DMAR_FEUADDR_REG);
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- void dmar_msi_read(int irq, struct msi_msg *msg)
-@@ -1262,11 +1262,11 @@ void dmar_msi_read(int irq, struct msi_m
- struct intel_iommu *iommu = irq_get_handler_data(irq);
- unsigned long flag;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- msg->data = readl(iommu->reg + DMAR_FEDATA_REG);
- msg->address_lo = readl(iommu->reg + DMAR_FEADDR_REG);
- msg->address_hi = readl(iommu->reg + DMAR_FEUADDR_REG);
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- static int dmar_fault_do_one(struct intel_iommu *iommu, int type,
-@@ -1303,7 +1303,7 @@ irqreturn_t dmar_fault(int irq, void *de
- u32 fault_status;
- unsigned long flag;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- fault_status = readl(iommu->reg + DMAR_FSTS_REG);
- if (fault_status)
- printk(KERN_ERR "DRHD: handling fault status reg %x\n",
-@@ -1342,7 +1342,7 @@ irqreturn_t dmar_fault(int irq, void *de
- writel(DMA_FRCD_F, iommu->reg + reg +
- fault_index * PRIMARY_FAULT_REG_LEN + 12);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
-
- dmar_fault_do_one(iommu, type, fault_reason,
- source_id, guest_addr);
-@@ -1350,14 +1350,14 @@ irqreturn_t dmar_fault(int irq, void *de
- fault_index++;
- if (fault_index >= cap_num_fault_regs(iommu->cap))
- fault_index = 0;
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- }
- clear_rest:
- /* clear all the other faults */
- fault_status = readl(iommu->reg + DMAR_FSTS_REG);
- writel(fault_status, iommu->reg + DMAR_FSTS_REG);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- return IRQ_HANDLED;
- }
-
-Index: linux-3.0/drivers/pci/intel-iommu.c
-===================================================================
---- linux-3.0.orig/drivers/pci/intel-iommu.c
-+++ linux-3.0/drivers/pci/intel-iommu.c
-@@ -933,7 +933,7 @@ static void iommu_set_root_entry(struct
-
- addr = iommu->root_entry;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- dmar_writeq(iommu->reg + DMAR_RTADDR_REG, virt_to_phys(addr));
-
- writel(iommu->gcmd | DMA_GCMD_SRTP, iommu->reg + DMAR_GCMD_REG);
-@@ -942,7 +942,7 @@ static void iommu_set_root_entry(struct
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (sts & DMA_GSTS_RTPS), sts);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- static void iommu_flush_write_buffer(struct intel_iommu *iommu)
-@@ -953,14 +953,14 @@ static void iommu_flush_write_buffer(str
- if (!rwbf_quirk && !cap_rwbf(iommu->cap))
- return;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- writel(iommu->gcmd | DMA_GCMD_WBF, iommu->reg + DMAR_GCMD_REG);
-
- /* Make sure hardware complete it */
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (!(val & DMA_GSTS_WBFS)), val);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- /* return value determine if we need a write buffer flush */
-@@ -987,14 +987,14 @@ static void __iommu_flush_context(struct
- }
- val |= DMA_CCMD_ICC;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- dmar_writeq(iommu->reg + DMAR_CCMD_REG, val);
-
- /* Make sure hardware complete it */
- IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG,
- dmar_readq, (!(val & DMA_CCMD_ICC)), val);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- /* return value determine if we need a write buffer flush */
-@@ -1033,7 +1033,7 @@ static void __iommu_flush_iotlb(struct i
- if (cap_write_drain(iommu->cap))
- val |= DMA_TLB_WRITE_DRAIN;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- /* Note: Only uses first TLB reg currently */
- if (val_iva)
- dmar_writeq(iommu->reg + tlb_offset, val_iva);
-@@ -1043,7 +1043,7 @@ static void __iommu_flush_iotlb(struct i
- IOMMU_WAIT_OP(iommu, tlb_offset + 8,
- dmar_readq, (!(val & DMA_TLB_IVT)), val);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
-
- /* check IOTLB invalidation granularity */
- if (DMA_TLB_IAIG(val) == 0)
-@@ -1159,7 +1159,7 @@ static void iommu_disable_protect_mem_re
- u32 pmen;
- unsigned long flags;
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
- pmen = readl(iommu->reg + DMAR_PMEN_REG);
- pmen &= ~DMA_PMEN_EPM;
- writel(pmen, iommu->reg + DMAR_PMEN_REG);
-@@ -1168,7 +1168,7 @@ static void iommu_disable_protect_mem_re
- IOMMU_WAIT_OP(iommu, DMAR_PMEN_REG,
- readl, !(pmen & DMA_PMEN_PRS), pmen);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- }
-
- static int iommu_enable_translation(struct intel_iommu *iommu)
-@@ -1176,7 +1176,7 @@ static int iommu_enable_translation(stru
- u32 sts;
- unsigned long flags;
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
- iommu->gcmd |= DMA_GCMD_TE;
- writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
-
-@@ -1184,7 +1184,7 @@ static int iommu_enable_translation(stru
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (sts & DMA_GSTS_TES), sts);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- return 0;
- }
-
-@@ -1193,7 +1193,7 @@ static int iommu_disable_translation(str
- u32 sts;
- unsigned long flag;
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- iommu->gcmd &= ~DMA_GCMD_TE;
- writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
-
-@@ -1201,7 +1201,7 @@ static int iommu_disable_translation(str
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (!(sts & DMA_GSTS_TES)), sts);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- return 0;
- }
-
-@@ -3321,7 +3321,7 @@ static int iommu_suspend(void)
- for_each_active_iommu(iommu, drhd) {
- iommu_disable_translation(iommu);
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
-
- iommu->iommu_state[SR_DMAR_FECTL_REG] =
- readl(iommu->reg + DMAR_FECTL_REG);
-@@ -3332,7 +3332,7 @@ static int iommu_suspend(void)
- iommu->iommu_state[SR_DMAR_FEUADDR_REG] =
- readl(iommu->reg + DMAR_FEUADDR_REG);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
- return 0;
-
-@@ -3359,7 +3359,7 @@ static void iommu_resume(void)
-
- for_each_active_iommu(iommu, drhd) {
-
-- spin_lock_irqsave(&iommu->register_lock, flag);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flag);
-
- writel(iommu->iommu_state[SR_DMAR_FECTL_REG],
- iommu->reg + DMAR_FECTL_REG);
-@@ -3370,7 +3370,7 @@ static void iommu_resume(void)
- writel(iommu->iommu_state[SR_DMAR_FEUADDR_REG],
- iommu->reg + DMAR_FEUADDR_REG);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flag);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
- }
-
- for_each_active_iommu(iommu, drhd)
-Index: linux-3.0/drivers/pci/intr_remapping.c
-===================================================================
---- linux-3.0.orig/drivers/pci/intr_remapping.c
-+++ linux-3.0/drivers/pci/intr_remapping.c
-@@ -46,7 +46,7 @@ static __init int setup_intremap(char *s
- }
- early_param("intremap", setup_intremap);
-
--static DEFINE_SPINLOCK(irq_2_ir_lock);
-+static DEFINE_RAW_SPINLOCK(irq_2_ir_lock);
-
- static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
- {
-@@ -63,12 +63,12 @@ int get_irte(int irq, struct irte *entry
- if (!entry || !irq_iommu)
- return -1;
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
-
- index = irq_iommu->irte_index + irq_iommu->sub_handle;
- *entry = *(irq_iommu->iommu->ir_table->base + index);
-
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
- return 0;
- }
-
-@@ -102,7 +102,7 @@ int alloc_irte(struct intel_iommu *iommu
- return -1;
- }
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
- do {
- for (i = index; i < index + count; i++)
- if (table->base[i].present)
-@@ -114,7 +114,7 @@ int alloc_irte(struct intel_iommu *iommu
- index = (index + count) % INTR_REMAP_TABLE_ENTRIES;
-
- if (index == start_index) {
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
- printk(KERN_ERR "can't allocate an IRTE\n");
- return -1;
- }
-@@ -128,7 +128,7 @@ int alloc_irte(struct intel_iommu *iommu
- irq_iommu->sub_handle = 0;
- irq_iommu->irte_mask = mask;
-
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
- return index;
- }
-@@ -153,10 +153,10 @@ int map_irq_to_irte_handle(int irq, u16
- if (!irq_iommu)
- return -1;
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
- *sub_handle = irq_iommu->sub_handle;
- index = irq_iommu->irte_index;
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
- return index;
- }
-
-@@ -168,14 +168,14 @@ int set_irte_irq(int irq, struct intel_i
- if (!irq_iommu)
- return -1;
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
-
- irq_iommu->iommu = iommu;
- irq_iommu->irte_index = index;
- irq_iommu->sub_handle = subhandle;
- irq_iommu->irte_mask = 0;
-
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
- return 0;
- }
-@@ -191,7 +191,7 @@ int modify_irte(int irq, struct irte *ir
- if (!irq_iommu)
- return -1;
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
-
- iommu = irq_iommu->iommu;
-
-@@ -203,7 +203,7 @@ int modify_irte(int irq, struct irte *ir
- __iommu_flush_cache(iommu, irte, sizeof(*irte));
-
- rc = qi_flush_iec(iommu, index, 0);
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
- return rc;
- }
-@@ -271,7 +271,7 @@ int free_irte(int irq)
- if (!irq_iommu)
- return -1;
-
-- spin_lock_irqsave(&irq_2_ir_lock, flags);
-+ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
-
- rc = clear_entries(irq_iommu);
-
-@@ -280,7 +280,7 @@ int free_irte(int irq)
- irq_iommu->sub_handle = 0;
- irq_iommu->irte_mask = 0;
-
-- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-+ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-
- return rc;
- }
-@@ -410,7 +410,7 @@ static void iommu_set_intr_remapping(str
-
- addr = virt_to_phys((void *)iommu->ir_table->base);
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-
- dmar_writeq(iommu->reg + DMAR_IRTA_REG,
- (addr) | IR_X2APIC_MODE(mode) | INTR_REMAP_TABLE_REG_SIZE);
-@@ -421,7 +421,7 @@ static void iommu_set_intr_remapping(str
-
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (sts & DMA_GSTS_IRTPS), sts);
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
-
- /*
- * global invalidation of interrupt entry cache before enabling
-@@ -429,7 +429,7 @@ static void iommu_set_intr_remapping(str
- */
- qi_global_iec(iommu);
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-
- /* Enable interrupt-remapping */
- iommu->gcmd |= DMA_GCMD_IRE;
-@@ -438,7 +438,7 @@ static void iommu_set_intr_remapping(str
- IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
- readl, (sts & DMA_GSTS_IRES), sts);
-
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- }
-
-
-@@ -486,7 +486,7 @@ static void iommu_disable_intr_remapping
- */
- qi_global_iec(iommu);
-
-- spin_lock_irqsave(&iommu->register_lock, flags);
-+ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-
- sts = dmar_readq(iommu->reg + DMAR_GSTS_REG);
- if (!(sts & DMA_GSTS_IRES))
-@@ -499,7 +499,7 @@ static void iommu_disable_intr_remapping
- readl, !(sts & DMA_GSTS_IRES), sts);
-
- end:
-- spin_unlock_irqrestore(&iommu->register_lock, flags);
-+ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- }
-
- int __init intr_remapping_supported(void)
-Index: linux-3.0/include/linux/intel-iommu.h
-===================================================================
---- linux-3.0.orig/include/linux/intel-iommu.h
-+++ linux-3.0/include/linux/intel-iommu.h
-@@ -271,7 +271,7 @@ struct qi_desc {
- };
-
- struct q_inval {
-- spinlock_t q_lock;
-+ raw_spinlock_t q_lock;
- struct qi_desc *desc; /* invalidation queue */
- int *desc_status; /* desc status */
- int free_head; /* first free entry */
-@@ -311,7 +311,7 @@ struct intel_iommu {
- u64 cap;
- u64 ecap;
- u32 gcmd; /* Holds TE, EAFL. Don't need SRTP, SFL, WBF */
-- spinlock_t register_lock; /* protect register handling */
-+ raw_spinlock_t register_lock; /* protect register handling */
- int seq_id; /* sequence id of the iommu */
- int agaw; /* agaw of this iommu */
- int msagaw; /* max sagaw of this iommu */
-Index: linux-3.0/lib/atomic64.c
-===================================================================
---- linux-3.0.orig/lib/atomic64.c
-+++ linux-3.0/lib/atomic64.c
-@@ -29,11 +29,11 @@
- * Ensure each lock is in a separate cacheline.
- */
- static union {
-- spinlock_t lock;
-+ raw_spinlock_t lock;
- char pad[L1_CACHE_BYTES];
- } atomic64_lock[NR_LOCKS] __cacheline_aligned_in_smp;
-
--static inline spinlock_t *lock_addr(const atomic64_t *v)
-+static inline raw_spinlock_t *lock_addr(const atomic64_t *v)
- {
- unsigned long addr = (unsigned long) v;
-
-@@ -45,12 +45,12 @@ static inline spinlock_t *lock_addr(cons
- long long atomic64_read(const atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_read);
-@@ -58,34 +58,34 @@ EXPORT_SYMBOL(atomic64_read);
- void atomic64_set(atomic64_t *v, long long i)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- v->counter = i;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- }
- EXPORT_SYMBOL(atomic64_set);
-
- void atomic64_add(long long a, atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- v->counter += a;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- }
- EXPORT_SYMBOL(atomic64_add);
-
- long long atomic64_add_return(long long a, atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter += a;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_add_return);
-@@ -93,23 +93,23 @@ EXPORT_SYMBOL(atomic64_add_return);
- void atomic64_sub(long long a, atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- v->counter -= a;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- }
- EXPORT_SYMBOL(atomic64_sub);
-
- long long atomic64_sub_return(long long a, atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter -= a;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_sub_return);
-@@ -117,14 +117,14 @@ EXPORT_SYMBOL(atomic64_sub_return);
- long long atomic64_dec_if_positive(atomic64_t *v)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter - 1;
- if (val >= 0)
- v->counter = val;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_dec_if_positive);
-@@ -132,14 +132,14 @@ EXPORT_SYMBOL(atomic64_dec_if_positive);
- long long atomic64_cmpxchg(atomic64_t *v, long long o, long long n)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter;
- if (val == o)
- v->counter = n;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_cmpxchg);
-@@ -147,13 +147,13 @@ EXPORT_SYMBOL(atomic64_cmpxchg);
- long long atomic64_xchg(atomic64_t *v, long long new)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- long long val;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- val = v->counter;
- v->counter = new;
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return val;
- }
- EXPORT_SYMBOL(atomic64_xchg);
-@@ -161,15 +161,15 @@ EXPORT_SYMBOL(atomic64_xchg);
- int atomic64_add_unless(atomic64_t *v, long long a, long long u)
- {
- unsigned long flags;
-- spinlock_t *lock = lock_addr(v);
-+ raw_spinlock_t *lock = lock_addr(v);
- int ret = 0;
-
-- spin_lock_irqsave(lock, flags);
-+ raw_spin_lock_irqsave(lock, flags);
- if (v->counter != u) {
- v->counter += a;
- ret = 1;
- }
-- spin_unlock_irqrestore(lock, flags);
-+ raw_spin_unlock_irqrestore(lock, flags);
- return ret;
- }
- EXPORT_SYMBOL(atomic64_add_unless);
-@@ -179,7 +179,7 @@ static int init_atomic64_lock(void)
- int i;
-
- for (i = 0; i < NR_LOCKS; ++i)
-- spin_lock_init(&atomic64_lock[i].lock);
-+ raw_spin_lock_init(&atomic64_lock[i].lock);
- return 0;
- }
-
-Index: linux-3.0/arch/x86/kernel/cpu/intel_cacheinfo.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/intel_cacheinfo.c
-+++ linux-3.0/arch/x86/kernel/cpu/intel_cacheinfo.c
-@@ -151,28 +151,17 @@ union _cpuid4_leaf_ecx {
- u32 full;
- };
-
--struct amd_l3_cache {
-- struct amd_northbridge *nb;
-- unsigned indices;
-- u8 subcaches[4];
--};
--
--struct _cpuid4_info {
-+struct _cpuid4_info_regs {
- union _cpuid4_leaf_eax eax;
- union _cpuid4_leaf_ebx ebx;
- union _cpuid4_leaf_ecx ecx;
- unsigned long size;
-- struct amd_l3_cache *l3;
-- DECLARE_BITMAP(shared_cpu_map, NR_CPUS);
-+ struct amd_northbridge *nb;
- };
-
--/* subset of above _cpuid4_info w/o shared_cpu_map */
--struct _cpuid4_info_regs {
-- union _cpuid4_leaf_eax eax;
-- union _cpuid4_leaf_ebx ebx;
-- union _cpuid4_leaf_ecx ecx;
-- unsigned long size;
-- struct amd_l3_cache *l3;
-+struct _cpuid4_info {
-+ struct _cpuid4_info_regs base;
-+ DECLARE_BITMAP(shared_cpu_map, NR_CPUS);
- };
-
- unsigned short num_cache_leaves;
-@@ -314,12 +303,13 @@ struct _cache_attr {
- /*
- * L3 cache descriptors
- */
--static void __cpuinit amd_calc_l3_indices(struct amd_l3_cache *l3)
-+static void __cpuinit amd_calc_l3_indices(struct amd_northbridge *nb)
- {
-+ struct amd_l3_cache *l3 = &nb->l3_cache;
- unsigned int sc0, sc1, sc2, sc3;
- u32 val = 0;
-
-- pci_read_config_dword(l3->nb->misc, 0x1C4, &val);
-+ pci_read_config_dword(nb->misc, 0x1C4, &val);
-
- /* calculate subcache sizes */
- l3->subcaches[0] = sc0 = !(val & BIT(0));
-@@ -333,33 +323,16 @@ static void __cpuinit amd_calc_l3_indice
- static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf,
- int index)
- {
-- static struct amd_l3_cache *__cpuinitdata l3_caches;
- int node;
-
- /* only for L3, and not in virtualized environments */
-- if (index < 3 || amd_nb_num() == 0)
-+ if (index < 3)
- return;
-
-- /*
-- * Strictly speaking, the amount in @size below is leaked since it is
-- * never freed but this is done only on shutdown so it doesn't matter.
-- */
-- if (!l3_caches) {
-- int size = amd_nb_num() * sizeof(struct amd_l3_cache);
--
-- l3_caches = kzalloc(size, GFP_ATOMIC);
-- if (!l3_caches)
-- return;
-- }
--
- node = amd_get_nb_id(smp_processor_id());
--
-- if (!l3_caches[node].nb) {
-- l3_caches[node].nb = node_to_amd_nb(node);
-- amd_calc_l3_indices(&l3_caches[node]);
-- }
--
-- this_leaf->l3 = &l3_caches[node];
-+ this_leaf->nb = node_to_amd_nb(node);
-+ if (this_leaf->nb && !this_leaf->nb->l3_cache.indices)
-+ amd_calc_l3_indices(this_leaf->nb);
- }
-
- /*
-@@ -369,11 +342,11 @@ static void __cpuinit amd_init_l3_cache(
- *
- * @returns: the disabled index if used or negative value if slot free.
- */
--int amd_get_l3_disable_slot(struct amd_l3_cache *l3, unsigned slot)
-+int amd_get_l3_disable_slot(struct amd_northbridge *nb, unsigned slot)
- {
- unsigned int reg = 0;
-
-- pci_read_config_dword(l3->nb->misc, 0x1BC + slot * 4, ®);
-+ pci_read_config_dword(nb->misc, 0x1BC + slot * 4, ®);
-
- /* check whether this slot is activated already */
- if (reg & (3UL << 30))
-@@ -387,11 +360,10 @@ static ssize_t show_cache_disable(struct
- {
- int index;
-
-- if (!this_leaf->l3 ||
-- !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
-+ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
- return -EINVAL;
-
-- index = amd_get_l3_disable_slot(this_leaf->l3, slot);
-+ index = amd_get_l3_disable_slot(this_leaf->base.nb, slot);
- if (index >= 0)
- return sprintf(buf, "%d\n", index);
-
-@@ -408,7 +380,7 @@ show_cache_disable_##slot(struct _cpuid4
- SHOW_CACHE_DISABLE(0)
- SHOW_CACHE_DISABLE(1)
-
--static void amd_l3_disable_index(struct amd_l3_cache *l3, int cpu,
-+static void amd_l3_disable_index(struct amd_northbridge *nb, int cpu,
- unsigned slot, unsigned long idx)
- {
- int i;
-@@ -421,10 +393,10 @@ static void amd_l3_disable_index(struct
- for (i = 0; i < 4; i++) {
- u32 reg = idx | (i << 20);
-
-- if (!l3->subcaches[i])
-+ if (!nb->l3_cache.subcaches[i])
- continue;
-
-- pci_write_config_dword(l3->nb->misc, 0x1BC + slot * 4, reg);
-+ pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg);
-
- /*
- * We need to WBINVD on a core on the node containing the L3
-@@ -434,7 +406,7 @@ static void amd_l3_disable_index(struct
- wbinvd_on_cpu(cpu);
-
- reg |= BIT(31);
-- pci_write_config_dword(l3->nb->misc, 0x1BC + slot * 4, reg);
-+ pci_write_config_dword(nb->misc, 0x1BC + slot * 4, reg);
- }
- }
-
-@@ -448,24 +420,24 @@ static void amd_l3_disable_index(struct
- *
- * @return: 0 on success, error status on failure
- */
--int amd_set_l3_disable_slot(struct amd_l3_cache *l3, int cpu, unsigned slot,
-+int amd_set_l3_disable_slot(struct amd_northbridge *nb, int cpu, unsigned slot,
- unsigned long index)
- {
- int ret = 0;
-
- /* check if @slot is already used or the index is already disabled */
-- ret = amd_get_l3_disable_slot(l3, slot);
-+ ret = amd_get_l3_disable_slot(nb, slot);
- if (ret >= 0)
- return -EINVAL;
-
-- if (index > l3->indices)
-+ if (index > nb->l3_cache.indices)
- return -EINVAL;
-
- /* check whether the other slot has disabled the same index already */
-- if (index == amd_get_l3_disable_slot(l3, !slot))
-+ if (index == amd_get_l3_disable_slot(nb, !slot))
- return -EINVAL;
-
-- amd_l3_disable_index(l3, cpu, slot, index);
-+ amd_l3_disable_index(nb, cpu, slot, index);
-
- return 0;
- }
-@@ -480,8 +452,7 @@ static ssize_t store_cache_disable(struc
- if (!capable(CAP_SYS_ADMIN))
- return -EPERM;
-
-- if (!this_leaf->l3 ||
-- !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
-+ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE))
- return -EINVAL;
-
- cpu = cpumask_first(to_cpumask(this_leaf->shared_cpu_map));
-@@ -489,7 +460,7 @@ static ssize_t store_cache_disable(struc
- if (strict_strtoul(buf, 10, &val) < 0)
- return -EINVAL;
-
-- err = amd_set_l3_disable_slot(this_leaf->l3, cpu, slot, val);
-+ err = amd_set_l3_disable_slot(this_leaf->base.nb, cpu, slot, val);
- if (err) {
- if (err == -EEXIST)
- printk(KERN_WARNING "L3 disable slot %d in use!\n",
-@@ -518,7 +489,7 @@ static struct _cache_attr cache_disable_
- static ssize_t
- show_subcaches(struct _cpuid4_info *this_leaf, char *buf, unsigned int cpu)
- {
-- if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
-+ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
- return -EINVAL;
-
- return sprintf(buf, "%x\n", amd_get_subcaches(cpu));
-@@ -533,7 +504,7 @@ store_subcaches(struct _cpuid4_info *thi
- if (!capable(CAP_SYS_ADMIN))
- return -EPERM;
-
-- if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
-+ if (!this_leaf->base.nb || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING))
- return -EINVAL;
-
- if (strict_strtoul(buf, 16, &val) < 0)
-@@ -769,7 +740,7 @@ static void __cpuinit cache_shared_cpu_m
- return;
- }
- this_leaf = CPUID4_INFO_IDX(cpu, index);
-- num_threads_sharing = 1 + this_leaf->eax.split.num_threads_sharing;
-+ num_threads_sharing = 1 + this_leaf->base.eax.split.num_threads_sharing;
-
- if (num_threads_sharing == 1)
- cpumask_set_cpu(cpu, to_cpumask(this_leaf->shared_cpu_map));
-@@ -820,29 +791,19 @@ static void __cpuinit free_cache_attribu
- for (i = 0; i < num_cache_leaves; i++)
- cache_remove_shared_cpu_map(cpu, i);
-
-- kfree(per_cpu(ici_cpuid4_info, cpu)->l3);
- kfree(per_cpu(ici_cpuid4_info, cpu));
- per_cpu(ici_cpuid4_info, cpu) = NULL;
- }
-
--static int
--__cpuinit cpuid4_cache_lookup(int index, struct _cpuid4_info *this_leaf)
--{
-- struct _cpuid4_info_regs *leaf_regs =
-- (struct _cpuid4_info_regs *)this_leaf;
--
-- return cpuid4_cache_lookup_regs(index, leaf_regs);
--}
--
- static void __cpuinit get_cpu_leaves(void *_retval)
- {
- int j, *retval = _retval, cpu = smp_processor_id();
-
- /* Do cpuid and store the results */
- for (j = 0; j < num_cache_leaves; j++) {
-- struct _cpuid4_info *this_leaf;
-- this_leaf = CPUID4_INFO_IDX(cpu, j);
-- *retval = cpuid4_cache_lookup(j, this_leaf);
-+ struct _cpuid4_info *this_leaf = CPUID4_INFO_IDX(cpu, j);
-+
-+ *retval = cpuid4_cache_lookup_regs(j, &this_leaf->base);
- if (unlikely(*retval < 0)) {
- int i;
-
-@@ -900,16 +861,16 @@ static ssize_t show_##file_name(struct _
- return sprintf(buf, "%lu\n", (unsigned long)this_leaf->object + val); \
- }
-
--show_one_plus(level, eax.split.level, 0);
--show_one_plus(coherency_line_size, ebx.split.coherency_line_size, 1);
--show_one_plus(physical_line_partition, ebx.split.physical_line_partition, 1);
--show_one_plus(ways_of_associativity, ebx.split.ways_of_associativity, 1);
--show_one_plus(number_of_sets, ecx.split.number_of_sets, 1);
-+show_one_plus(level, base.eax.split.level, 0);
-+show_one_plus(coherency_line_size, base.ebx.split.coherency_line_size, 1);
-+show_one_plus(physical_line_partition, base.ebx.split.physical_line_partition, 1);
-+show_one_plus(ways_of_associativity, base.ebx.split.ways_of_associativity, 1);
-+show_one_plus(number_of_sets, base.ecx.split.number_of_sets, 1);
-
- static ssize_t show_size(struct _cpuid4_info *this_leaf, char *buf,
- unsigned int cpu)
- {
-- return sprintf(buf, "%luK\n", this_leaf->size / 1024);
-+ return sprintf(buf, "%luK\n", this_leaf->base.size / 1024);
- }
-
- static ssize_t show_shared_cpu_map_func(struct _cpuid4_info *this_leaf,
-@@ -946,7 +907,7 @@ static inline ssize_t show_shared_cpu_li
- static ssize_t show_type(struct _cpuid4_info *this_leaf, char *buf,
- unsigned int cpu)
- {
-- switch (this_leaf->eax.split.type) {
-+ switch (this_leaf->base.eax.split.type) {
- case CACHE_TYPE_DATA:
- return sprintf(buf, "Data\n");
- case CACHE_TYPE_INST:
-@@ -1135,7 +1096,7 @@ static int __cpuinit cache_add_dev(struc
-
- ktype_cache.default_attrs = default_attrs;
- #ifdef CONFIG_AMD_NB
-- if (this_leaf->l3)
-+ if (this_leaf->base.nb)
- ktype_cache.default_attrs = amd_l3_attrs();
- #endif
- retval = kobject_init_and_add(&(this_object->kobj),
-Index: linux-3.0/arch/x86/include/asm/amd_nb.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/amd_nb.h
-+++ linux-3.0/arch/x86/include/asm/amd_nb.h
-@@ -19,9 +19,15 @@ extern int amd_numa_init(void);
- extern int amd_get_subcaches(int);
- extern int amd_set_subcaches(int, int);
-
-+struct amd_l3_cache {
-+ unsigned indices;
-+ u8 subcaches[4];
-+};
-+
- struct amd_northbridge {
- struct pci_dev *misc;
- struct pci_dev *link;
-+ struct amd_l3_cache l3_cache;
- };
-
- struct amd_northbridge_info {
-Index: linux-3.0/kernel/sched_fair.c
-===================================================================
---- linux-3.0.orig/kernel/sched_fair.c
-+++ linux-3.0/kernel/sched_fair.c
-@@ -1565,7 +1565,7 @@ find_idlest_group(struct sched_domain *s
-
- /* Skip over this group if it has no CPUs allowed */
- if (!cpumask_intersects(sched_group_cpus(group),
-- &p->cpus_allowed))
-+ tsk_cpus_allowed(p)))
- continue;
-
- local_group = cpumask_test_cpu(this_cpu,
-@@ -1611,7 +1611,7 @@ find_idlest_cpu(struct sched_group *grou
- int i;
-
- /* Traverse only the allowed CPUs */
-- for_each_cpu_and(i, sched_group_cpus(group), &p->cpus_allowed) {
-+ for_each_cpu_and(i, sched_group_cpus(group), tsk_cpus_allowed(p)) {
- load = weighted_cpuload(i);
-
- if (load < min_load || (load == min_load && i == this_cpu)) {
-@@ -1655,7 +1655,7 @@ static int select_idle_sibling(struct ta
- if (!(sd->flags & SD_SHARE_PKG_RESOURCES))
- break;
-
-- for_each_cpu_and(i, sched_domain_span(sd), &p->cpus_allowed) {
-+ for_each_cpu_and(i, sched_domain_span(sd), tsk_cpus_allowed(p)) {
- if (idle_cpu(i)) {
- target = i;
- break;
-@@ -1698,7 +1698,7 @@ select_task_rq_fair(struct task_struct *
- int sync = wake_flags & WF_SYNC;
-
- if (sd_flag & SD_BALANCE_WAKE) {
-- if (cpumask_test_cpu(cpu, &p->cpus_allowed))
-+ if (cpumask_test_cpu(cpu, tsk_cpus_allowed(p)))
- want_affine = 1;
- new_cpu = prev_cpu;
- }
-@@ -2067,7 +2067,7 @@ int can_migrate_task(struct task_struct
- * 2) cannot be migrated to this CPU due to cpus_allowed, or
- * 3) are cache-hot on their current CPU.
- */
-- if (!cpumask_test_cpu(this_cpu, &p->cpus_allowed)) {
-+ if (!cpumask_test_cpu(this_cpu, tsk_cpus_allowed(p))) {
- schedstat_inc(p, se.statistics.nr_failed_migrations_affine);
- return 0;
- }
-@@ -2173,6 +2173,10 @@ balance_tasks(struct rq *this_rq, int th
- */
- if (idle == CPU_NEWLY_IDLE)
- break;
-+
-+ if (raw_spin_is_contended(&this_rq->lock) ||
-+ raw_spin_is_contended(&busiest->lock))
-+ break;
- #endif
-
- /*
-@@ -2277,6 +2281,20 @@ load_balance_fair(struct rq *this_rq, in
- rem_load_move -= moved_load;
- if (rem_load_move < 0)
- break;
-+
-+#ifdef CONFIG_PREEMPT
-+ /*
-+ * NEWIDLE balancing is a source of latency, so preemptible
-+ * kernels will stop after the first task is pulled to minimize
-+ * the critical section.
-+ */
-+ if (idle == CPU_NEWLY_IDLE && this_rq->nr_running)
-+ break;
-+
-+ if (raw_spin_is_contended(&this_rq->lock) ||
-+ raw_spin_is_contended(&busiest->lock))
-+ break;
-+#endif
- }
- rcu_read_unlock();
-
-@@ -3418,7 +3436,7 @@ redo:
- * moved to this_cpu
- */
- if (!cpumask_test_cpu(this_cpu,
-- &busiest->curr->cpus_allowed)) {
-+ tsk_cpus_allowed(busiest->curr))) {
- raw_spin_unlock_irqrestore(&busiest->lock,
- flags);
- all_pinned = 1;
-Index: linux-3.0/lib/smp_processor_id.c
-===================================================================
---- linux-3.0.orig/lib/smp_processor_id.c
-+++ linux-3.0/lib/smp_processor_id.c
-@@ -22,7 +22,7 @@ notrace unsigned int debug_smp_processor
- * Kernel threads bound to a single CPU can safely use
- * smp_processor_id():
- */
-- if (cpumask_equal(¤t->cpus_allowed, cpumask_of(this_cpu)))
-+ if (cpumask_equal(tsk_cpus_allowed(current), cpumask_of(this_cpu)))
- goto out;
-
- /*
-@@ -39,9 +39,9 @@ notrace unsigned int debug_smp_processor
- if (!printk_ratelimit())
- goto out_enable;
-
-- printk(KERN_ERR "BUG: using smp_processor_id() in preemptible [%08x] "
-- "code: %s/%d\n",
-- preempt_count() - 1, current->comm, current->pid);
-+ printk(KERN_ERR "BUG: using smp_processor_id() in preemptible [%08x %08x] "
-+ "code: %s/%d\n", preempt_count() - 1,
-+ __migrate_disabled(current), current->comm, current->pid);
- print_symbol("caller is %s\n", (long)__builtin_return_address(0));
- dump_stack();
-
-Index: linux-3.0/arch/x86/include/asm/irqflags.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/irqflags.h
-+++ linux-3.0/arch/x86/include/asm/irqflags.h
-@@ -60,23 +60,24 @@ static inline void native_halt(void)
- #include <asm/paravirt.h>
- #else
- #ifndef __ASSEMBLY__
-+#include <linux/types.h>
-
--static inline unsigned long arch_local_save_flags(void)
-+static inline notrace unsigned long arch_local_save_flags(void)
- {
- return native_save_fl();
- }
-
--static inline void arch_local_irq_restore(unsigned long flags)
-+static inline notrace void arch_local_irq_restore(unsigned long flags)
- {
- native_restore_fl(flags);
- }
-
--static inline void arch_local_irq_disable(void)
-+static inline notrace void arch_local_irq_disable(void)
- {
- native_irq_disable();
- }
-
--static inline void arch_local_irq_enable(void)
-+static inline notrace void arch_local_irq_enable(void)
- {
- native_irq_enable();
- }
-@@ -102,7 +103,7 @@ static inline void halt(void)
- /*
- * For spinlocks, etc:
- */
--static inline unsigned long arch_local_irq_save(void)
-+static inline notrace unsigned long arch_local_irq_save(void)
- {
- unsigned long flags = arch_local_save_flags();
- arch_local_irq_disable();
-Index: linux-3.0/kernel/signal.c
-===================================================================
---- linux-3.0.orig/kernel/signal.c
-+++ linux-3.0/kernel/signal.c
-@@ -300,13 +300,45 @@ static bool task_participate_group_stop(
- return false;
- }
-
-+#ifdef __HAVE_ARCH_CMPXCHG
-+static inline struct sigqueue *get_task_cache(struct task_struct *t)
-+{
-+ struct sigqueue *q = t->sigqueue_cache;
-+
-+ if (cmpxchg(&t->sigqueue_cache, q, NULL) != q)
-+ return NULL;
-+ return q;
-+}
-+
-+static inline int put_task_cache(struct task_struct *t, struct sigqueue *q)
-+{
-+ if (cmpxchg(&t->sigqueue_cache, NULL, q) == NULL)
-+ return 0;
-+ return 1;
-+}
-+
-+#else
-+
-+static inline struct sigqueue *get_task_cache(struct task_struct *t)
-+{
-+ return NULL;
-+}
-+
-+static inline int put_task_cache(struct task_struct *t, struct sigqueue *q)
-+{
-+ return 1;
-+}
-+
-+#endif
-+
- /*
- * allocate a new signal queue record
- * - this may be called without locks if and only if t == current, otherwise an
- * appropriate lock must be held to stop the target task from exiting
- */
- static struct sigqueue *
--__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
-+__sigqueue_do_alloc(int sig, struct task_struct *t, gfp_t flags,
-+ int override_rlimit, int fromslab)
- {
- struct sigqueue *q = NULL;
- struct user_struct *user;
-@@ -323,7 +355,10 @@ __sigqueue_alloc(int sig, struct task_st
- if (override_rlimit ||
- atomic_read(&user->sigpending) <=
- task_rlimit(t, RLIMIT_SIGPENDING)) {
-- q = kmem_cache_alloc(sigqueue_cachep, flags);
-+ if (!fromslab)
-+ q = get_task_cache(t);
-+ if (!q)
-+ q = kmem_cache_alloc(sigqueue_cachep, flags);
- } else {
- print_dropped_signal(sig);
- }
-@@ -340,6 +375,13 @@ __sigqueue_alloc(int sig, struct task_st
- return q;
- }
-
-+static struct sigqueue *
-+__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags,
-+ int override_rlimit)
-+{
-+ return __sigqueue_do_alloc(sig, t, flags, override_rlimit, 0);
-+}
-+
- static void __sigqueue_free(struct sigqueue *q)
- {
- if (q->flags & SIGQUEUE_PREALLOC)
-@@ -349,6 +391,21 @@ static void __sigqueue_free(struct sigqu
- kmem_cache_free(sigqueue_cachep, q);
- }
-
-+static void sigqueue_free_current(struct sigqueue *q)
-+{
-+ struct user_struct *up;
-+
-+ if (q->flags & SIGQUEUE_PREALLOC)
-+ return;
-+
-+ up = q->user;
-+ if (rt_prio(current->normal_prio) && !put_task_cache(current, q)) {
-+ atomic_dec(&up->sigpending);
-+ free_uid(up);
-+ } else
-+ __sigqueue_free(q);
-+}
-+
- void flush_sigqueue(struct sigpending *queue)
- {
- struct sigqueue *q;
-@@ -362,6 +419,21 @@ void flush_sigqueue(struct sigpending *q
- }
-
- /*
-+ * Called from __exit_signal. Flush tsk->pending and
-+ * tsk->sigqueue_cache
-+ */
-+void flush_task_sigqueue(struct task_struct *tsk)
-+{
-+ struct sigqueue *q;
-+
-+ flush_sigqueue(&tsk->pending);
-+
-+ q = get_task_cache(tsk);
-+ if (q)
-+ kmem_cache_free(sigqueue_cachep, q);
-+}
-+
-+/*
- * Flush all pending signals for a task.
- */
- void __flush_signals(struct task_struct *t)
-@@ -509,7 +581,7 @@ static void collect_signal(int sig, stru
- still_pending:
- list_del_init(&first->list);
- copy_siginfo(info, &first->info);
-- __sigqueue_free(first);
-+ sigqueue_free_current(first);
- } else {
- /*
- * Ok, it wasn't in the queue. This must be
-@@ -555,6 +627,8 @@ int dequeue_signal(struct task_struct *t
- {
- int signr;
-
-+ WARN_ON_ONCE(tsk != current);
-+
- /* We only dequeue private signals from ourselves, we don't let
- * signalfd steal them
- */
-@@ -637,6 +711,9 @@ void signal_wake_up(struct task_struct *
-
- set_tsk_thread_flag(t, TIF_SIGPENDING);
-
-+ if (unlikely(t == current))
-+ return;
-+
- /*
- * For SIGKILL, we want to wake it up in the stopped/traced/killable
- * case. We don't check t->state here because there is a race with it
-@@ -1179,12 +1256,12 @@ struct sighand_struct *__lock_task_sigha
- struct sighand_struct *sighand;
-
- for (;;) {
-- local_irq_save(*flags);
-+ local_irq_save_nort(*flags);
- rcu_read_lock();
- sighand = rcu_dereference(tsk->sighand);
- if (unlikely(sighand == NULL)) {
- rcu_read_unlock();
-- local_irq_restore(*flags);
-+ local_irq_restore_nort(*flags);
- break;
- }
-
-@@ -1195,7 +1272,7 @@ struct sighand_struct *__lock_task_sigha
- }
- spin_unlock(&sighand->siglock);
- rcu_read_unlock();
-- local_irq_restore(*flags);
-+ local_irq_restore_nort(*flags);
- }
-
- return sighand;
-@@ -1434,7 +1511,8 @@ EXPORT_SYMBOL(kill_pid);
- */
- struct sigqueue *sigqueue_alloc(void)
- {
-- struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
-+ /* Preallocated sigqueue objects always from the slabcache ! */
-+ struct sigqueue *q = __sigqueue_do_alloc(-1, current, GFP_KERNEL, 0, 1);
-
- if (q)
- q->flags |= SIGQUEUE_PREALLOC;
-@@ -1782,15 +1860,7 @@ static void ptrace_stop(int exit_code, i
- if (gstop_done && !real_parent_is_ptracer(current))
- do_notify_parent_cldstop(current, false, why);
-
-- /*
-- * Don't want to allow preemption here, because
-- * sys_ptrace() needs this task to be inactive.
-- *
-- * XXX: implement read_unlock_no_resched().
-- */
-- preempt_disable();
- read_unlock(&tasklist_lock);
-- preempt_enable_no_resched();
- schedule();
- } else {
- /*
-Index: linux-3.0/arch/arm/kernel/perf_event.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/perf_event.c
-+++ linux-3.0/arch/arm/kernel/perf_event.c
-@@ -420,7 +420,7 @@ armpmu_reserve_hardware(void)
- continue;
-
- err = request_irq(irq, handle_irq,
-- IRQF_DISABLED | IRQF_NOBALANCING,
-+ IRQF_NOBALANCING | IRQF_NO_THREAD,
- "armpmu", NULL);
- if (err) {
- pr_warning("unable to request IRQ%d for ARM perf "
-Index: linux-3.0/arch/arm/Kconfig
-===================================================================
---- linux-3.0.orig/arch/arm/Kconfig
-+++ linux-3.0/arch/arm/Kconfig
-@@ -29,6 +29,7 @@ config ARM
- select HAVE_GENERIC_HARDIRQS
- select HAVE_SPARSE_IRQ
- select GENERIC_IRQ_SHOW
-+ select IRQ_FORCED_THREADING
- help
- The ARM series is a line of low-power-consumption RISC chip designs
- licensed by ARM Ltd and targeted at embedded applications and
-@@ -1524,7 +1525,7 @@ config HAVE_ARCH_PFN_VALID
-
- config HIGHMEM
- bool "High Memory Support"
-- depends on MMU
-+ depends on MMU && !PREEMPT_RT_FULL
- help
- The address space of ARM processors is only 4 Gigabytes large
- and it has to accommodate user address space, kernel address
-Index: linux-3.0/arch/arm/mach-exynos4/platsmp.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-exynos4/platsmp.c
-+++ linux-3.0/arch/arm/mach-exynos4/platsmp.c
-@@ -56,7 +56,7 @@ static void __iomem *scu_base_addr(void)
- return (void __iomem *)(S5P_VA_SCU);
- }
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
-
- void __cpuinit platform_secondary_init(unsigned int cpu)
- {
-@@ -76,8 +76,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
-@@ -88,7 +88,7 @@ int __cpuinit boot_secondary(unsigned in
- * Set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
- /*
- * The secondary processor is waiting to be released from
-@@ -120,7 +120,7 @@ int __cpuinit boot_secondary(unsigned in
- * now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return pen_release != -1 ? -ENOSYS : 0;
- }
-Index: linux-3.0/arch/arm/mach-msm/platsmp.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-msm/platsmp.c
-+++ linux-3.0/arch/arm/mach-msm/platsmp.c
-@@ -38,7 +38,7 @@ extern void msm_secondary_startup(void);
- */
- volatile int pen_release = -1;
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
-
- void __cpuinit platform_secondary_init(unsigned int cpu)
- {
-@@ -62,8 +62,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- static __cpuinit void prepare_cold_cpu(unsigned int cpu)
-@@ -100,7 +100,7 @@ int __cpuinit boot_secondary(unsigned in
- * set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
- /*
- * The secondary processor is waiting to be released from
-@@ -134,7 +134,7 @@ int __cpuinit boot_secondary(unsigned in
- * now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return pen_release != -1 ? -ENOSYS : 0;
- }
-Index: linux-3.0/arch/arm/mach-omap2/omap-smp.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-omap2/omap-smp.c
-+++ linux-3.0/arch/arm/mach-omap2/omap-smp.c
-@@ -29,7 +29,7 @@
- /* SCU base address */
- static void __iomem *scu_base;
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
-
- void __cpuinit platform_secondary_init(unsigned int cpu)
- {
-@@ -43,8 +43,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
-@@ -53,7 +53,7 @@ int __cpuinit boot_secondary(unsigned in
- * Set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
- /*
- * Update the AuxCoreBoot0 with boot state for secondary core.
-@@ -70,7 +70,7 @@ int __cpuinit boot_secondary(unsigned in
- * Now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return 0;
- }
-Index: linux-3.0/arch/arm/mach-tegra/platsmp.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-tegra/platsmp.c
-+++ linux-3.0/arch/arm/mach-tegra/platsmp.c
-@@ -29,7 +29,7 @@
-
- extern void tegra_secondary_startup(void);
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
- static void __iomem *scu_base = IO_ADDRESS(TEGRA_ARM_PERIF_BASE);
-
- #define EVP_CPU_RESET_VECTOR \
-@@ -51,8 +51,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
-@@ -66,7 +66,7 @@ int __cpuinit boot_secondary(unsigned in
- * set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
-
- /* set the reset vector to point to the secondary_startup routine */
-@@ -102,7 +102,7 @@ int __cpuinit boot_secondary(unsigned in
- * now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return 0;
- }
-Index: linux-3.0/arch/arm/mach-ux500/platsmp.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-ux500/platsmp.c
-+++ linux-3.0/arch/arm/mach-ux500/platsmp.c
-@@ -57,7 +57,7 @@ static void __iomem *scu_base_addr(void)
- return NULL;
- }
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
-
- void __cpuinit platform_secondary_init(unsigned int cpu)
- {
-@@ -77,8 +77,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
-@@ -89,7 +89,7 @@ int __cpuinit boot_secondary(unsigned in
- * set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
- /*
- * The secondary processor is waiting to be released from
-@@ -110,7 +110,7 @@ int __cpuinit boot_secondary(unsigned in
- * now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return pen_release != -1 ? -ENOSYS : 0;
- }
-Index: linux-3.0/arch/arm/plat-versatile/platsmp.c
-===================================================================
---- linux-3.0.orig/arch/arm/plat-versatile/platsmp.c
-+++ linux-3.0/arch/arm/plat-versatile/platsmp.c
-@@ -37,7 +37,7 @@ static void __cpuinit write_pen_release(
- outer_clean_range(__pa(&pen_release), __pa(&pen_release + 1));
- }
-
--static DEFINE_SPINLOCK(boot_lock);
-+static DEFINE_RAW_SPINLOCK(boot_lock);
-
- void __cpuinit platform_secondary_init(unsigned int cpu)
- {
-@@ -57,8 +57,8 @@ void __cpuinit platform_secondary_init(u
- /*
- * Synchronise with the boot thread.
- */
-- spin_lock(&boot_lock);
-- spin_unlock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
- }
-
- int __cpuinit boot_secondary(unsigned int cpu, struct task_struct *idle)
-@@ -69,7 +69,7 @@ int __cpuinit boot_secondary(unsigned in
- * Set synchronisation state between this boot processor
- * and the secondary one
- */
-- spin_lock(&boot_lock);
-+ raw_spin_lock(&boot_lock);
-
- /*
- * This is really belt and braces; we hold unintended secondary
-@@ -99,7 +99,7 @@ int __cpuinit boot_secondary(unsigned in
- * now the secondary core is starting up let it run its
- * calibrations, then wait for it to finish
- */
-- spin_unlock(&boot_lock);
-+ raw_spin_unlock(&boot_lock);
-
- return pen_release != -1 ? -ENOSYS : 0;
- }
-Index: linux-3.0/arch/arm/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/process.c
-+++ linux-3.0/arch/arm/kernel/process.c
-@@ -209,9 +209,7 @@ void cpu_idle(void)
- }
- leds_event(led_idle_end);
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-@@ -486,6 +484,31 @@ unsigned long arch_randomize_brk(struct
- }
-
- #ifdef CONFIG_MMU
++ WARNING: This software should never be enabled (it can be built
++ but should not be turned on after it is loaded) in a production
++ environment where high latencies are a concern since the
++ sampling mechanism actually introduces latencies for
++ regular tasks while the CPU(s) are being held.
++
++ If unsure, say N
++
+ config PHANTOM
+ tristate "Sensable PHANToM (PCI)"
+ depends on PCI
+diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile
+index 5f03172..d456181 100644
+--- a/drivers/misc/Makefile
++++ b/drivers/misc/Makefile
+@@ -46,3 +46,4 @@ obj-y += ti-st/
+ obj-$(CONFIG_AB8500_PWM) += ab8500-pwm.o
+ obj-y += lis3lv02d/
+ obj-y += carma/
++obj-$(CONFIG_HWLAT_DETECTOR) += hwlat_detector.o
+diff --git a/drivers/misc/hwlat_detector.c b/drivers/misc/hwlat_detector.c
+new file mode 100644
+index 0000000..b7b7c90
+--- /dev/null
++++ b/drivers/misc/hwlat_detector.c
+@@ -0,0 +1,1212 @@
++/*
++ * hwlat_detector.c - A simple Hardware Latency detector.
++ *
++ * Use this module to detect large system latencies induced by the behavior of
++ * certain underlying system hardware or firmware, independent of Linux itself.
++ * The code was developed originally to detect the presence of SMIs on Intel
++ * and AMD systems, although there is no dependency upon x86 herein.
++ *
++ * The classical example usage of this module is in detecting the presence of
++ * SMIs or System Management Interrupts on Intel and AMD systems. An SMI is a
++ * somewhat special form of hardware interrupt spawned from earlier CPU debug
++ * modes in which the (BIOS/EFI/etc.) firmware arranges for the South Bridge
++ * LPC (or other device) to generate a special interrupt under certain
++ * circumstances, for example, upon expiration of a special SMI timer device,
++ * due to certain external thermal readings, on certain I/O address accesses,
++ * and other situations. An SMI hits a special CPU pin, triggers a special
++ * SMI mode (complete with special memory map), and the OS is unaware.
++ *
++ * Although certain hardware-inducing latencies are necessary (for example,
++ * a modern system often requires an SMI handler for correct thermal control
++ * and remote management) they can wreak havoc upon any OS-level performance
++ * guarantees toward low-latency, especially when the OS is not even made
++ * aware of the presence of these interrupts. For this reason, we need a
++ * somewhat brute force mechanism to detect these interrupts. In this case,
++ * we do it by hogging all of the CPU(s) for configurable timer intervals,
++ * sampling the built-in CPU timer, looking for discontiguous readings.
++ *
++ * WARNING: This implementation necessarily introduces latencies. Therefore,
++ * you should NEVER use this module in a production environment
++ * requiring any kind of low-latency performance guarantee(s).
++ *
++ * Copyright (C) 2008-2009 Jon Masters, Red Hat, Inc. <jcm at redhat.com>
++ *
++ * Includes useful feedback from Clark Williams <clark at redhat.com>
++ *
++ * This file is licensed under the terms of the GNU General Public
++ * License version 2. This program is licensed "as is" without any
++ * warranty of any kind, whether express or implied.
++ */
++
++#include <linux/module.h>
++#include <linux/init.h>
++#include <linux/ring_buffer.h>
++#include <linux/stop_machine.h>
++#include <linux/time.h>
++#include <linux/hrtimer.h>
++#include <linux/kthread.h>
++#include <linux/debugfs.h>
++#include <linux/seq_file.h>
++#include <linux/uaccess.h>
++#include <linux/version.h>
++#include <linux/delay.h>
++#include <linux/slab.h>
++
++#define BUF_SIZE_DEFAULT 262144UL /* 8K*(sizeof(entry)) */
++#define BUF_FLAGS (RB_FL_OVERWRITE) /* no block on full */
++#define U64STR_SIZE 22 /* 20 digits max */
++
++#define VERSION "1.0.0"
++#define BANNER "hwlat_detector: "
++#define DRVNAME "hwlat_detector"
++#define DEFAULT_SAMPLE_WINDOW 1000000 /* 1s */
++#define DEFAULT_SAMPLE_WIDTH 500000 /* 0.5s */
++#define DEFAULT_LAT_THRESHOLD 10 /* 10us */
++
++/* Module metadata */
++
++MODULE_LICENSE("GPL");
++MODULE_AUTHOR("Jon Masters <jcm at redhat.com>");
++MODULE_DESCRIPTION("A simple hardware latency detector");
++MODULE_VERSION(VERSION);
++
++/* Module parameters */
++
++static int debug;
++static int enabled;
++static int threshold;
++
++module_param(debug, int, 0); /* enable debug */
++module_param(enabled, int, 0); /* enable detector */
++module_param(threshold, int, 0); /* latency threshold */
++
++/* Buffering and sampling */
++
++static struct ring_buffer *ring_buffer; /* sample buffer */
++static DEFINE_MUTEX(ring_buffer_mutex); /* lock changes */
++static unsigned long buf_size = BUF_SIZE_DEFAULT;
++static struct task_struct *kthread; /* sampling thread */
++
++/* DebugFS filesystem entries */
++
++static struct dentry *debug_dir; /* debugfs directory */
++static struct dentry *debug_max; /* maximum TSC delta */
++static struct dentry *debug_count; /* total detect count */
++static struct dentry *debug_sample_width; /* sample width us */
++static struct dentry *debug_sample_window; /* sample window us */
++static struct dentry *debug_sample; /* raw samples us */
++static struct dentry *debug_threshold; /* threshold us */
++static struct dentry *debug_enable; /* enable/disable */
++
++/* Individual samples and global state */
++
++struct sample; /* latency sample */
++struct data; /* Global state */
++
++/* Sampling functions */
++static int __buffer_add_sample(struct sample *sample);
++static struct sample *buffer_get_sample(struct sample *sample);
++static int get_sample(void *unused);
++
++/* Threading and state */
++static int kthread_fn(void *unused);
++static int start_kthread(void);
++static int stop_kthread(void);
++static void __reset_stats(void);
++static int init_stats(void);
++
++/* Debugfs interface */
++static ssize_t simple_data_read(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos, const u64 *entry);
++static ssize_t simple_data_write(struct file *filp, const char __user *ubuf,
++ size_t cnt, loff_t *ppos, u64 *entry);
++static int debug_sample_fopen(struct inode *inode, struct file *filp);
++static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos);
++static int debug_sample_release(struct inode *inode, struct file *filp);
++static int debug_enable_fopen(struct inode *inode, struct file *filp);
++static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos);
++static ssize_t debug_enable_fwrite(struct file *file,
++ const char __user *user_buffer,
++ size_t user_size, loff_t *offset);
++
++/* Initialization functions */
++static int init_debugfs(void);
++static void free_debugfs(void);
++static int detector_init(void);
++static void detector_exit(void);
++
++/* Individual latency samples are stored here when detected and packed into
++ * the ring_buffer circular buffer, where they are overwritten when
++ * more than buf_size/sizeof(sample) samples are received. */
++struct sample {
++ u64 seqnum; /* unique sequence */
++ u64 duration; /* ktime delta */
++ struct timespec timestamp; /* wall time */
++ unsigned long lost;
++};
++
++/* keep the global state somewhere. Mostly used under stop_machine. */
++static struct data {
++
++ struct mutex lock; /* protect changes */
++
++ u64 count; /* total since reset */
++ u64 max_sample; /* max hardware latency */
++ u64 threshold; /* sample threshold level */
++
++ u64 sample_window; /* total sampling window (on+off) */
++ u64 sample_width; /* active sampling portion of window */
++
++ atomic_t sample_open; /* whether the sample file is open */
++
++ wait_queue_head_t wq; /* waitqeue for new sample values */
++
++} data;
++
++/**
++ * __buffer_add_sample - add a new latency sample recording to the ring buffer
++ * @sample: The new latency sample value
++ *
++ * This receives a new latency sample and records it in a global ring buffer.
++ * No additional locking is used in this case - suited for stop_machine use.
++ */
++static int __buffer_add_sample(struct sample *sample)
++{
++ return ring_buffer_write(ring_buffer,
++ sizeof(struct sample), sample);
++}
++
++/**
++ * buffer_get_sample - remove a hardware latency sample from the ring buffer
++ * @sample: Pre-allocated storage for the sample
++ *
++ * This retrieves a hardware latency sample from the global circular buffer
++ */
++static struct sample *buffer_get_sample(struct sample *sample)
++{
++ struct ring_buffer_event *e = NULL;
++ struct sample *s = NULL;
++ unsigned int cpu = 0;
++
++ if (!sample)
++ return NULL;
++
++ mutex_lock(&ring_buffer_mutex);
++ for_each_online_cpu(cpu) {
++ e = ring_buffer_consume(ring_buffer, cpu, NULL, &sample->lost);
++ if (e)
++ break;
++ }
++
++ if (e) {
++ s = ring_buffer_event_data(e);
++ memcpy(sample, s, sizeof(struct sample));
++ } else
++ sample = NULL;
++ mutex_unlock(&ring_buffer_mutex);
++
++ return sample;
++}
++
++/**
++ * get_sample - sample the CPU TSC and look for likely hardware latencies
++ * @unused: This is not used but is a part of the stop_machine API
++ *
++ * Used to repeatedly capture the CPU TSC (or similar), looking for potential
++ * hardware-induced latency. Called under stop_machine, with data.lock held.
++ */
++static int get_sample(void *unused)
++{
++ ktime_t start, t1, t2;
++ s64 diff, total = 0;
++ u64 sample = 0;
++ int ret = 1;
++
++ start = ktime_get(); /* start timestamp */
++
++ do {
++
++ t1 = ktime_get(); /* we'll look for a discontinuity */
++ t2 = ktime_get();
++
++ total = ktime_to_us(ktime_sub(t2, start)); /* sample width */
++ diff = ktime_to_us(ktime_sub(t2, t1)); /* current diff */
++
++ /* This shouldn't happen */
++ if (diff < 0) {
++ printk(KERN_ERR BANNER "time running backwards\n");
++ goto out;
++ }
++
++ if (diff > sample)
++ sample = diff; /* only want highest value */
++
++ } while (total <= data.sample_width);
++
++ /* If we exceed the threshold value, we have found a hardware latency */
++ if (sample > data.threshold) {
++ struct sample s;
++
++ data.count++;
++ s.seqnum = data.count;
++ s.duration = sample;
++ s.timestamp = CURRENT_TIME;
++ __buffer_add_sample(&s);
++
++ /* Keep a running maximum ever recorded hardware latency */
++ if (sample > data.max_sample)
++ data.max_sample = sample;
++ }
++
++ ret = 0;
++out:
++ return ret;
++}
+
+/*
-+ * CONFIG_SPLIT_PTLOCK_CPUS results in a page->ptl lock. If the lock is not
-+ * initialized by pgtable_page_ctor() then a coredump of the vector page will
-+ * fail.
++ * kthread_fn - The CPU time sampling/hardware latency detection kernel thread
++ * @unused: A required part of the kthread API.
++ *
++ * Used to periodically sample the CPU TSC via a call to get_sample. We
++ * use stop_machine, whith does (intentionally) introduce latency since we
++ * need to ensure nothing else might be running (and thus pre-empting).
++ * Obviously this should never be used in production environments.
++ *
++ * stop_machine will schedule us typically only on CPU0 which is fine for
++ * almost every real-world hardware latency situation - but we might later
++ * generalize this if we find there are any actualy systems with alternate
++ * SMI delivery or other non CPU0 hardware latencies.
+ */
-+static int __init vectors_user_mapping_init_page(void)
++static int kthread_fn(void *unused)
+{
-+ struct page *page;
-+ unsigned long addr = 0xffff0000;
-+ pgd_t *pgd;
-+ pud_t *pud;
-+ pmd_t *pmd;
++ int err = 0;
++ u64 interval = 0;
+
-+ pgd = pgd_offset_k(addr);
-+ pud = pud_offset(pgd, addr);
-+ pmd = pmd_offset(pud, addr);
-+ page = pmd_page(*(pmd));
++ while (!kthread_should_stop()) {
+
-+ pgtable_page_ctor(page);
++ mutex_lock(&data.lock);
++
++ err = stop_machine(get_sample, unused, 0);
++ if (err) {
++ /* Houston, we have a problem */
++ mutex_unlock(&data.lock);
++ goto err_out;
++ }
++
++ wake_up(&data.wq); /* wake up reader(s) */
++
++ interval = data.sample_window - data.sample_width;
++ do_div(interval, USEC_PER_MSEC); /* modifies interval value */
++
++ mutex_unlock(&data.lock);
++
++ if (msleep_interruptible(interval))
++ goto out;
++ }
++ goto out;
++err_out:
++ printk(KERN_ERR BANNER "could not call stop_machine, disabling\n");
++ enabled = 0;
++out:
++ return err;
++
++}
++
++/**
++ * start_kthread - Kick off the hardware latency sampling/detector kthread
++ *
++ * This starts a kernel thread that will sit and sample the CPU timestamp
++ * counter (TSC or similar) and look for potential hardware latencies.
++ */
++static int start_kthread(void)
++{
++ kthread = kthread_run(kthread_fn, NULL,
++ DRVNAME);
++ if (IS_ERR(kthread)) {
++ printk(KERN_ERR BANNER "could not start sampling thread\n");
++ enabled = 0;
++ return -ENOMEM;
++ }
++
++ return 0;
++}
++
++/**
++ * stop_kthread - Inform the hardware latency samping/detector kthread to stop
++ *
++ * This kicks the running hardware latency sampling/detector kernel thread and
++ * tells it to stop sampling now. Use this on unload and at system shutdown.
++ */
++static int stop_kthread(void)
++{
++ int ret;
++
++ ret = kthread_stop(kthread);
++
++ return ret;
++}
++
++/**
++ * __reset_stats - Reset statistics for the hardware latency detector
++ *
++ * We use data to store various statistics and global state. We call this
++ * function in order to reset those when "enable" is toggled on or off, and
++ * also at initialization. Should be called with data.lock held.
++ */
++static void __reset_stats(void)
++{
++ data.count = 0;
++ data.max_sample = 0;
++ ring_buffer_reset(ring_buffer); /* flush out old sample entries */
++}
++
++/**
++ * init_stats - Setup global state statistics for the hardware latency detector
++ *
++ * We use data to store various statistics and global state. We also use
++ * a global ring buffer (ring_buffer) to keep raw samples of detected hardware
++ * induced system latencies. This function initializes these structures and
++ * allocates the global ring buffer also.
++ */
++static int init_stats(void)
++{
++ int ret = -ENOMEM;
++
++ mutex_init(&data.lock);
++ init_waitqueue_head(&data.wq);
++ atomic_set(&data.sample_open, 0);
++
++ ring_buffer = ring_buffer_alloc(buf_size, BUF_FLAGS);
++
++ if (WARN(!ring_buffer, KERN_ERR BANNER
++ "failed to allocate ring buffer!\n"))
++ goto out;
++
++ __reset_stats();
++ data.threshold = DEFAULT_LAT_THRESHOLD; /* threshold us */
++ data.sample_window = DEFAULT_SAMPLE_WINDOW; /* window us */
++ data.sample_width = DEFAULT_SAMPLE_WIDTH; /* width us */
++
++ ret = 0;
++
++out:
++ return ret;
++
++}
++
++/*
++ * simple_data_read - Wrapper read function for global state debugfs entries
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ * @entry: The entry to read from
++ *
++ * This function provides a generic read implementation for the global state
++ * "data" structure debugfs filesystem entries. It would be nice to use
++ * simple_attr_read directly, but we need to make sure that the data.lock
++ * spinlock is held during the actual read (even though we likely won't ever
++ * actually race here as the updater runs under a stop_machine context).
++ */
++static ssize_t simple_data_read(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos, const u64 *entry)
++{
++ char buf[U64STR_SIZE];
++ u64 val = 0;
++ int len = 0;
++
++ memset(buf, 0, sizeof(buf));
++
++ if (!entry)
++ return -EFAULT;
++
++ mutex_lock(&data.lock);
++ val = *entry;
++ mutex_unlock(&data.lock);
++
++ len = snprintf(buf, sizeof(buf), "%llu\n", (unsigned long long)val);
++
++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, len);
++
++}
++
++/*
++ * simple_data_write - Wrapper write function for global state debugfs entries
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to write value from
++ * @cnt: The maximum number of bytes to write
++ * @ppos: The current "file" position
++ * @entry: The entry to write to
++ *
++ * This function provides a generic write implementation for the global state
++ * "data" structure debugfs filesystem entries. It would be nice to use
++ * simple_attr_write directly, but we need to make sure that the data.lock
++ * spinlock is held during the actual write (even though we likely won't ever
++ * actually race here as the updater runs under a stop_machine context).
++ */
++static ssize_t simple_data_write(struct file *filp, const char __user *ubuf,
++ size_t cnt, loff_t *ppos, u64 *entry)
++{
++ char buf[U64STR_SIZE];
++ int csize = min(cnt, sizeof(buf));
++ u64 val = 0;
++ int err = 0;
++
++ memset(buf, '\0', sizeof(buf));
++ if (copy_from_user(buf, ubuf, csize))
++ return -EFAULT;
++
++ buf[U64STR_SIZE-1] = '\0'; /* just in case */
++ err = strict_strtoull(buf, 10, &val);
++ if (err)
++ return -EINVAL;
++
++ mutex_lock(&data.lock);
++ *entry = val;
++ mutex_unlock(&data.lock);
++
++ return csize;
++}
+
++/**
++ * debug_count_fopen - Open function for "count" debugfs entry
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "count" debugfs
++ * interface to the hardware latency detector.
++ */
++static int debug_count_fopen(struct inode *inode, struct file *filp)
++{
+ return 0;
+}
-+late_initcall(vectors_user_mapping_init_page);
+
- /*
- * The vectors page is always readable from user space for the
- * atomic helpers and the signal restart code. Let's declare a mapping
-Index: linux-3.0/arch/avr32/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/avr32/kernel/process.c
-+++ linux-3.0/arch/avr32/kernel/process.c
-@@ -38,9 +38,7 @@ void cpu_idle(void)
- while (!need_resched())
- cpu_idle_sleep();
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/blackfin/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/blackfin/kernel/process.c
-+++ linux-3.0/arch/blackfin/kernel/process.c
-@@ -92,9 +92,7 @@ void cpu_idle(void)
- while (!need_resched())
- idle();
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/cris/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/cris/kernel/process.c
-+++ linux-3.0/arch/cris/kernel/process.c
-@@ -115,9 +115,7 @@ void cpu_idle (void)
- idle = default_idle;
- idle();
- }
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/frv/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/frv/kernel/process.c
-+++ linux-3.0/arch/frv/kernel/process.c
-@@ -92,9 +92,7 @@ void cpu_idle(void)
- idle();
- }
-
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/h8300/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/h8300/kernel/process.c
-+++ linux-3.0/arch/h8300/kernel/process.c
-@@ -81,9 +81,7 @@ void cpu_idle(void)
- while (1) {
- while (!need_resched())
- idle();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/ia64/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/ia64/kernel/process.c
-+++ linux-3.0/arch/ia64/kernel/process.c
-@@ -330,9 +330,7 @@ cpu_idle (void)
- normal_xtp();
- #endif
- }
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- check_pgt_cache();
- if (cpu_is_offline(cpu))
- play_dead();
-Index: linux-3.0/arch/m32r/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/m32r/kernel/process.c
-+++ linux-3.0/arch/m32r/kernel/process.c
-@@ -90,9 +90,7 @@ void cpu_idle (void)
-
- idle();
- }
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/m68k/kernel/process_mm.c
-===================================================================
---- linux-3.0.orig/arch/m68k/kernel/process_mm.c
-+++ linux-3.0/arch/m68k/kernel/process_mm.c
-@@ -94,9 +94,7 @@ void cpu_idle(void)
- while (1) {
- while (!need_resched())
- idle();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/m68k/kernel/process_no.c
-===================================================================
---- linux-3.0.orig/arch/m68k/kernel/process_no.c
-+++ linux-3.0/arch/m68k/kernel/process_no.c
-@@ -73,9 +73,7 @@ void cpu_idle(void)
- /* endless idle loop with no priority at all */
- while (1) {
- idle();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/microblaze/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/microblaze/kernel/process.c
-+++ linux-3.0/arch/microblaze/kernel/process.c
-@@ -108,9 +108,7 @@ void cpu_idle(void)
- idle();
- tick_nohz_restart_sched_tick();
-
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- check_pgt_cache();
- }
- }
-Index: linux-3.0/arch/mips/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/process.c
-+++ linux-3.0/arch/mips/kernel/process.c
-@@ -78,9 +78,7 @@ void __noreturn cpu_idle(void)
- play_dead();
- #endif
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/mn10300/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/mn10300/kernel/process.c
-+++ linux-3.0/arch/mn10300/kernel/process.c
-@@ -123,9 +123,7 @@ void cpu_idle(void)
- idle();
- }
-
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/parisc/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/parisc/kernel/process.c
-+++ linux-3.0/arch/parisc/kernel/process.c
-@@ -71,9 +71,7 @@ void cpu_idle(void)
- while (1) {
- while (!need_resched())
- barrier();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- check_pgt_cache();
- }
- }
-Index: linux-3.0/arch/powerpc/kernel/idle.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/idle.c
-+++ linux-3.0/arch/powerpc/kernel/idle.c
-@@ -94,11 +94,11 @@ void cpu_idle(void)
- HMT_medium();
- ppc64_runlatch_on();
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- if (cpu_should_die())
-+ if (cpu_should_die()) {
-+ __preempt_enable_no_resched();
- cpu_die();
-- schedule();
-- preempt_disable();
-+ }
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/powerpc/platforms/iseries/setup.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/platforms/iseries/setup.c
-+++ linux-3.0/arch/powerpc/platforms/iseries/setup.c
-@@ -581,9 +581,7 @@ static void iseries_shared_idle(void)
- if (hvlpevent_is_pending())
- process_iSeries_events();
-
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-@@ -610,9 +608,7 @@ static void iseries_dedicated_idle(void)
-
- ppc64_runlatch_on();
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/s390/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/s390/kernel/process.c
-+++ linux-3.0/arch/s390/kernel/process.c
-@@ -94,9 +94,7 @@ void cpu_idle(void)
- while (!need_resched())
- default_idle();
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/score/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/score/kernel/process.c
-+++ linux-3.0/arch/score/kernel/process.c
-@@ -53,9 +53,7 @@ void __noreturn cpu_idle(void)
- while (!need_resched())
- barrier();
-
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/sh/kernel/idle.c
-===================================================================
---- linux-3.0.orig/arch/sh/kernel/idle.c
-+++ linux-3.0/arch/sh/kernel/idle.c
-@@ -110,9 +110,7 @@ void cpu_idle(void)
- }
-
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/sparc/kernel/process_32.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/process_32.c
-+++ linux-3.0/arch/sparc/kernel/process_32.c
-@@ -113,9 +113,7 @@ void cpu_idle(void)
- while (!need_resched())
- cpu_relax();
- }
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- check_pgt_cache();
- }
- }
-@@ -138,9 +136,7 @@ void cpu_idle(void)
- while (!need_resched())
- cpu_relax();
- }
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- check_pgt_cache();
- }
- }
-Index: linux-3.0/arch/sparc/kernel/process_64.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/process_64.c
-+++ linux-3.0/arch/sparc/kernel/process_64.c
-@@ -102,15 +102,13 @@ void cpu_idle(void)
-
- tick_nohz_restart_sched_tick();
-
-- preempt_enable_no_resched();
--
- #ifdef CONFIG_HOTPLUG_CPU
-- if (cpu_is_offline(cpu))
-+ if (cpu_is_offline(cpu)) {
-+ __preempt_enable_no_resched();
- cpu_play_dead();
++/**
++ * debug_count_fread - Read function for "count" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "count" debugfs
++ * interface to the hardware latency detector. Can be used to read the
++ * number of latency readings exceeding the configured threshold since
++ * the detector was last reset (e.g. by writing a zero into "count").
++ */
++static ssize_t debug_count_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ return simple_data_read(filp, ubuf, cnt, ppos, &data.count);
++}
++
++/**
++ * debug_count_fwrite - Write function for "count" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "count" debugfs
++ * interface to the hardware latency detector. Can be used to write a
++ * desired value, especially to zero the total count.
++ */
++static ssize_t debug_count_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ return simple_data_write(filp, ubuf, cnt, ppos, &data.count);
++}
++
++/**
++ * debug_enable_fopen - Dummy open function for "enable" debugfs interface
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "enable" debugfs
++ * interface to the hardware latency detector.
++ */
++static int debug_enable_fopen(struct inode *inode, struct file *filp)
++{
++ return 0;
++}
++
++/**
++ * debug_enable_fread - Read function for "enable" debugfs interface
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "enable" debugfs
++ * interface to the hardware latency detector. Can be used to determine
++ * whether the detector is currently enabled ("0\n" or "1\n" returned).
++ */
++static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ char buf[4];
++
++ if ((cnt < sizeof(buf)) || (*ppos))
++ return 0;
++
++ buf[0] = enabled ? '1' : '0';
++ buf[1] = '\n';
++ buf[2] = '\0';
++ if (copy_to_user(ubuf, buf, strlen(buf)))
++ return -EFAULT;
++ return *ppos = strlen(buf);
++}
++
++/**
++ * debug_enable_fwrite - Write function for "enable" debugfs interface
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "enable" debugfs
++ * interface to the hardware latency detector. Can be used to enable or
++ * disable the detector, which will have the side-effect of possibly
++ * also resetting the global stats and kicking off the measuring
++ * kthread (on an enable) or the converse (upon a disable).
++ */
++static ssize_t debug_enable_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ char buf[4];
++ int csize = min(cnt, sizeof(buf));
++ long val = 0;
++ int err = 0;
++
++ memset(buf, '\0', sizeof(buf));
++ if (copy_from_user(buf, ubuf, csize))
++ return -EFAULT;
++
++ buf[sizeof(buf)-1] = '\0'; /* just in case */
++ err = strict_strtoul(buf, 10, &val);
++ if (0 != err)
++ return -EINVAL;
++
++ if (val) {
++ if (enabled)
++ goto unlock;
++ enabled = 1;
++ __reset_stats();
++ if (start_kthread())
++ return -EFAULT;
++ } else {
++ if (!enabled)
++ goto unlock;
++ enabled = 0;
++ err = stop_kthread();
++ if (err) {
++ printk(KERN_ERR BANNER "cannot stop kthread\n");
++ return -EFAULT;
+ }
- #endif
--
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/tile/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/tile/kernel/process.c
-+++ linux-3.0/arch/tile/kernel/process.c
-@@ -106,9 +106,7 @@ void cpu_idle(void)
- current_thread_info()->status |= TS_POLLING;
- }
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/x86/kernel/process_32.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/process_32.c
-+++ linux-3.0/arch/x86/kernel/process_32.c
-@@ -38,6 +38,7 @@
- #include <linux/uaccess.h>
- #include <linux/io.h>
- #include <linux/kdebug.h>
-+#include <linux/highmem.h>
-
- #include <asm/pgtable.h>
- #include <asm/system.h>
-@@ -113,9 +114,7 @@ void cpu_idle(void)
- start_critical_timings();
- }
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-@@ -348,6 +347,41 @@ __switch_to(struct task_struct *prev_p,
- task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT))
- __switch_to_xtra(prev_p, next_p, tss);
-
-+#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM
-+ /*
-+ * Save @prev's kmap_atomic stack
-+ */
-+ prev_p->kmap_idx = __this_cpu_read(__kmap_atomic_idx);
-+ if (unlikely(prev_p->kmap_idx)) {
-+ int i;
++ wake_up(&data.wq); /* reader(s) should return */
++ }
++unlock:
++ return csize;
++}
++
++/**
++ * debug_max_fopen - Open function for "max" debugfs entry
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "max" debugfs
++ * interface to the hardware latency detector.
++ */
++static int debug_max_fopen(struct inode *inode, struct file *filp)
++{
++ return 0;
++}
++
++/**
++ * debug_max_fread - Read function for "max" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "max" debugfs
++ * interface to the hardware latency detector. Can be used to determine
++ * the maximum latency value observed since it was last reset.
++ */
++static ssize_t debug_max_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ return simple_data_read(filp, ubuf, cnt, ppos, &data.max_sample);
++}
++
++/**
++ * debug_max_fwrite - Write function for "max" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "max" debugfs
++ * interface to the hardware latency detector. Can be used to reset the
++ * maximum or set it to some other desired value - if, then, subsequent
++ * measurements exceed this value, the maximum will be updated.
++ */
++static ssize_t debug_max_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ return simple_data_write(filp, ubuf, cnt, ppos, &data.max_sample);
++}
++
++
++/**
++ * debug_sample_fopen - An open function for "sample" debugfs interface
++ * @inode: The in-kernel inode representation of this debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function handles opening the "sample" file within the hardware
++ * latency detector debugfs directory interface. This file is used to read
++ * raw samples from the global ring_buffer and allows the user to see a
++ * running latency history. Can be opened blocking or non-blocking,
++ * affecting whether it behaves as a buffer read pipe, or does not.
++ * Implements simple locking to prevent multiple simultaneous use.
++ */
++static int debug_sample_fopen(struct inode *inode, struct file *filp)
++{
++ if (!atomic_add_unless(&data.sample_open, 1, 1))
++ return -EBUSY;
++ else
++ return 0;
++}
++
++/**
++ * debug_sample_fread - A read function for "sample" debugfs interface
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that will contain the samples read
++ * @cnt: The maximum bytes to read from the debugfs "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function handles reading from the "sample" file within the hardware
++ * latency detector debugfs directory interface. This file is used to read
++ * raw samples from the global ring_buffer and allows the user to see a
++ * running latency history. By default this will block pending a new
++ * value written into the sample buffer, unless there are already a
++ * number of value(s) waiting in the buffer, or the sample file was
++ * previously opened in a non-blocking mode of operation.
++ */
++static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ int len = 0;
++ char buf[64];
++ struct sample *sample = NULL;
++
++ if (!enabled)
++ return 0;
++
++ sample = kzalloc(sizeof(struct sample), GFP_KERNEL);
++ if (!sample)
++ return -ENOMEM;
++
++ while (!buffer_get_sample(sample)) {
++
++ DEFINE_WAIT(wait);
++
++ if (filp->f_flags & O_NONBLOCK) {
++ len = -EAGAIN;
++ goto out;
++ }
++
++ prepare_to_wait(&data.wq, &wait, TASK_INTERRUPTIBLE);
++ schedule();
++ finish_wait(&data.wq, &wait);
++
++ if (signal_pending(current)) {
++ len = -EINTR;
++ goto out;
++ }
++
++ if (!enabled) { /* enable was toggled */
++ len = 0;
++ goto out;
++ }
++ }
++
++ len = snprintf(buf, sizeof(buf), "%010lu.%010lu\t%llu\n",
++ sample->timestamp.tv_sec,
++ sample->timestamp.tv_nsec,
++ sample->duration);
++
++
++ /* handling partial reads is more trouble than it's worth */
++ if (len > cnt)
++ goto out;
++
++ if (copy_to_user(ubuf, buf, len))
++ len = -EFAULT;
++
++out:
++ kfree(sample);
++ return len;
++}
++
++/**
++ * debug_sample_release - Release function for "sample" debugfs interface
++ * @inode: The in-kernel inode represenation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function completes the close of the debugfs interface "sample" file.
++ * Frees the sample_open "lock" so that other users may open the interface.
++ */
++static int debug_sample_release(struct inode *inode, struct file *filp)
++{
++ atomic_dec(&data.sample_open);
++
++ return 0;
++}
++
++/**
++ * debug_threshold_fopen - Open function for "threshold" debugfs entry
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "threshold" debugfs
++ * interface to the hardware latency detector.
++ */
++static int debug_threshold_fopen(struct inode *inode, struct file *filp)
++{
++ return 0;
++}
++
++/**
++ * debug_threshold_fread - Read function for "threshold" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "threshold" debugfs
++ * interface to the hardware latency detector. It can be used to determine
++ * the current threshold level at which a latency will be recorded in the
++ * global ring buffer, typically on the order of 10us.
++ */
++static ssize_t debug_threshold_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ return simple_data_read(filp, ubuf, cnt, ppos, &data.threshold);
++}
++
++/**
++ * debug_threshold_fwrite - Write function for "threshold" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "threshold" debugfs
++ * interface to the hardware latency detector. It can be used to configure
++ * the threshold level at which any subsequently detected latencies will
++ * be recorded into the global ring buffer.
++ */
++static ssize_t debug_threshold_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ int ret;
++
++ ret = simple_data_write(filp, ubuf, cnt, ppos, &data.threshold);
++
++ if (enabled)
++ wake_up_process(kthread);
++
++ return ret;
++}
++
++/**
++ * debug_width_fopen - Open function for "width" debugfs entry
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "width" debugfs
++ * interface to the hardware latency detector.
++ */
++static int debug_width_fopen(struct inode *inode, struct file *filp)
++{
++ return 0;
++}
++
++/**
++ * debug_width_fread - Read function for "width" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "width" debugfs
++ * interface to the hardware latency detector. It can be used to determine
++ * for how many us of the total window us we will actively sample for any
++ * hardware-induced latecy periods. Obviously, it is not possible to
++ * sample constantly and have the system respond to a sample reader, or,
++ * worse, without having the system appear to have gone out to lunch.
++ */
++static ssize_t debug_width_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_width);
++}
++
++/**
++ * debug_width_fwrite - Write function for "width" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "width" debugfs
++ * interface to the hardware latency detector. It can be used to configure
++ * for how many us of the total window us we will actively sample for any
++ * hardware-induced latency periods. Obviously, it is not possible to
++ * sample constantly and have the system respond to a sample reader, or,
++ * worse, without having the system appear to have gone out to lunch. It
++ * is enforced that width is less that the total window size.
++ */
++static ssize_t debug_width_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ char buf[U64STR_SIZE];
++ int csize = min(cnt, sizeof(buf));
++ u64 val = 0;
++ int err = 0;
++
++ memset(buf, '\0', sizeof(buf));
++ if (copy_from_user(buf, ubuf, csize))
++ return -EFAULT;
++
++ buf[U64STR_SIZE-1] = '\0'; /* just in case */
++ err = strict_strtoull(buf, 10, &val);
++ if (0 != err)
++ return -EINVAL;
++
++ mutex_lock(&data.lock);
++ if (val < data.sample_window)
++ data.sample_width = val;
++ else {
++ mutex_unlock(&data.lock);
++ return -EINVAL;
++ }
++ mutex_unlock(&data.lock);
++
++ if (enabled)
++ wake_up_process(kthread);
++
++ return csize;
++}
++
++/**
++ * debug_window_fopen - Open function for "window" debugfs entry
++ * @inode: The in-kernel inode representation of the debugfs "file"
++ * @filp: The active open file structure for the debugfs "file"
++ *
++ * This function provides an open implementation for the "window" debugfs
++ * interface to the hardware latency detector. The window is the total time
++ * in us that will be considered one sample period. Conceptually, windows
++ * occur back-to-back and contain a sample width period during which
++ * actual sampling occurs.
++ */
++static int debug_window_fopen(struct inode *inode, struct file *filp)
++{
++ return 0;
++}
++
++/**
++ * debug_window_fread - Read function for "window" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The userspace provided buffer to read value into
++ * @cnt: The maximum number of bytes to read
++ * @ppos: The current "file" position
++ *
++ * This function provides a read implementation for the "window" debugfs
++ * interface to the hardware latency detector. The window is the total time
++ * in us that will be considered one sample period. Conceptually, windows
++ * occur back-to-back and contain a sample width period during which
++ * actual sampling occurs. Can be used to read the total window size.
++ */
++static ssize_t debug_window_fread(struct file *filp, char __user *ubuf,
++ size_t cnt, loff_t *ppos)
++{
++ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_window);
++}
++
++/**
++ * debug_window_fwrite - Write function for "window" debugfs entry
++ * @filp: The active open file structure for the debugfs "file"
++ * @ubuf: The user buffer that contains the value to write
++ * @cnt: The maximum number of bytes to write to "file"
++ * @ppos: The current position in the debugfs "file"
++ *
++ * This function provides a write implementation for the "window" debufds
++ * interface to the hardware latency detetector. The window is the total time
++ * in us that will be considered one sample period. Conceptually, windows
++ * occur back-to-back and contain a sample width period during which
++ * actual sampling occurs. Can be used to write a new total window size. It
++ * is enfoced that any value written must be greater than the sample width
++ * size, or an error results.
++ */
++static ssize_t debug_window_fwrite(struct file *filp,
++ const char __user *ubuf,
++ size_t cnt,
++ loff_t *ppos)
++{
++ char buf[U64STR_SIZE];
++ int csize = min(cnt, sizeof(buf));
++ u64 val = 0;
++ int err = 0;
+
-+ for (i = 0; i < prev_p->kmap_idx; i++) {
-+ int idx = i + KM_TYPE_NR * smp_processor_id();
++ memset(buf, '\0', sizeof(buf));
++ if (copy_from_user(buf, ubuf, csize))
++ return -EFAULT;
+
-+ pte_t *ptep = kmap_pte - idx;
-+ prev_p->kmap_pte[i] = *ptep;
-+ kpte_clear_flush(ptep, __fix_to_virt(FIX_KMAP_BEGIN + idx));
-+ }
++ buf[U64STR_SIZE-1] = '\0'; /* just in case */
++ err = strict_strtoull(buf, 10, &val);
++ if (0 != err)
++ return -EINVAL;
+
-+ __this_cpu_write(__kmap_atomic_idx, 0);
++ mutex_lock(&data.lock);
++ if (data.sample_width < val)
++ data.sample_window = val;
++ else {
++ mutex_unlock(&data.lock);
++ return -EINVAL;
+ }
++ mutex_unlock(&data.lock);
+
-+ /*
-+ * Restore @next_p's kmap_atomic stack
-+ */
-+ if (unlikely(next_p->kmap_idx)) {
-+ int i;
++ return csize;
++}
+
-+ __this_cpu_write(__kmap_atomic_idx, next_p->kmap_idx);
++/*
++ * Function pointers for the "count" debugfs file operations
++ */
++static const struct file_operations count_fops = {
++ .open = debug_count_fopen,
++ .read = debug_count_fread,
++ .write = debug_count_fwrite,
++ .owner = THIS_MODULE,
++};
+
-+ for (i = 0; i < next_p->kmap_idx; i++) {
-+ int idx = i + KM_TYPE_NR * smp_processor_id();
++/*
++ * Function pointers for the "enable" debugfs file operations
++ */
++static const struct file_operations enable_fops = {
++ .open = debug_enable_fopen,
++ .read = debug_enable_fread,
++ .write = debug_enable_fwrite,
++ .owner = THIS_MODULE,
++};
+
-+ set_pte(kmap_pte - idx, next_p->kmap_pte[i]);
-+ }
-+ }
-+#endif
++/*
++ * Function pointers for the "max" debugfs file operations
++ */
++static const struct file_operations max_fops = {
++ .open = debug_max_fopen,
++ .read = debug_max_fread,
++ .write = debug_max_fwrite,
++ .owner = THIS_MODULE,
++};
+
- /* If we're going to preload the fpu context, make sure clts
- is run while we're batching the cpu state updates. */
- if (preload_fpu)
-Index: linux-3.0/arch/x86/kernel/process_64.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/process_64.c
-+++ linux-3.0/arch/x86/kernel/process_64.c
-@@ -146,9 +146,7 @@ void cpu_idle(void)
- }
-
- tick_nohz_restart_sched_tick();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/arch/xtensa/kernel/process.c
-===================================================================
---- linux-3.0.orig/arch/xtensa/kernel/process.c
-+++ linux-3.0/arch/xtensa/kernel/process.c
-@@ -113,9 +113,7 @@ void cpu_idle(void)
- while (1) {
- while (!need_resched())
- platform_idle();
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- }
- }
-
-Index: linux-3.0/init/main.c
-===================================================================
---- linux-3.0.orig/init/main.c
-+++ linux-3.0/init/main.c
-@@ -68,6 +68,7 @@
- #include <linux/shmem_fs.h>
- #include <linux/slab.h>
- #include <linux/perf_event.h>
-+#include <linux/posix-timers.h>
-
- #include <asm/io.h>
- #include <asm/bugs.h>
-@@ -367,9 +368,7 @@ static noinline void __init_refok rest_i
- * at least once to get things moving:
- */
- init_idle_bootup_task(current);
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
-
- /* Call into cpu_idle with preempt disabled */
- cpu_idle();
-@@ -501,6 +500,7 @@ asmlinkage void __init start_kernel(void
- parse_args("Booting kernel", static_command_line, __start___param,
- __stop___param - __start___param,
- &unknown_bootoption);
-+ softirq_early_init();
- /*
- * These use large bootmem allocations and must precede
- * kmem_cache_init()
-Index: linux-3.0/kernel/mutex.c
-===================================================================
---- linux-3.0.orig/kernel/mutex.c
-+++ linux-3.0/kernel/mutex.c
-@@ -240,9 +240,7 @@ __mutex_lock_common(struct mutex *lock,
-
- /* didn't get the lock, go to sleep: */
- spin_unlock_mutex(&lock->wait_lock, flags);
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-+ schedule_preempt_disabled();
- spin_lock_mutex(&lock->wait_lock, flags);
- }
-
-Index: linux-3.0/kernel/softirq.c
-===================================================================
---- linux-3.0.orig/kernel/softirq.c
-+++ linux-3.0/kernel/softirq.c
-@@ -24,6 +24,7 @@
- #include <linux/ftrace.h>
- #include <linux/smp.h>
- #include <linux/tick.h>
-+#include <linux/locallock.h>
-
- #define CREATE_TRACE_POINTS
- #include <trace/events/irq.h>
-@@ -61,6 +62,67 @@ char *softirq_to_name[NR_SOFTIRQS] = {
- "TASKLET", "SCHED", "HRTIMER", "RCU"
- };
-
-+#ifdef CONFIG_NO_HZ
-+# ifdef CONFIG_PREEMPT_RT_FULL
+/*
-+ * On preempt-rt a softirq might be blocked on a lock. There might be
-+ * no other runnable task on this CPU because the lock owner runs on
-+ * some other CPU. So we have to go into idle with the pending bit
-+ * set. Therefor we need to check this otherwise we warn about false
-+ * positives which confuses users and defeats the whole purpose of
-+ * this test.
++ * Function pointers for the "sample" debugfs file operations
++ */
++static const struct file_operations sample_fops = {
++ .open = debug_sample_fopen,
++ .read = debug_sample_fread,
++ .release = debug_sample_release,
++ .owner = THIS_MODULE,
++};
++
++/*
++ * Function pointers for the "threshold" debugfs file operations
++ */
++static const struct file_operations threshold_fops = {
++ .open = debug_threshold_fopen,
++ .read = debug_threshold_fread,
++ .write = debug_threshold_fwrite,
++ .owner = THIS_MODULE,
++};
++
++/*
++ * Function pointers for the "width" debugfs file operations
++ */
++static const struct file_operations width_fops = {
++ .open = debug_width_fopen,
++ .read = debug_width_fread,
++ .write = debug_width_fwrite,
++ .owner = THIS_MODULE,
++};
++
++/*
++ * Function pointers for the "window" debugfs file operations
++ */
++static const struct file_operations window_fops = {
++ .open = debug_window_fopen,
++ .read = debug_window_fread,
++ .write = debug_window_fwrite,
++ .owner = THIS_MODULE,
++};
++
++/**
++ * init_debugfs - A function to initialize the debugfs interface files
+ *
-+ * This code is called with interrupts disabled.
++ * This function creates entries in debugfs for "hwlat_detector", including
++ * files to read values from the detector, current samples, and the
++ * maximum sample that has been captured since the hardware latency
++ * dectector was started.
++ */
++static int init_debugfs(void)
++{
++ int ret = -ENOMEM;
++
++ debug_dir = debugfs_create_dir(DRVNAME, NULL);
++ if (!debug_dir)
++ goto err_debug_dir;
++
++ debug_sample = debugfs_create_file("sample", 0444,
++ debug_dir, NULL,
++ &sample_fops);
++ if (!debug_sample)
++ goto err_sample;
++
++ debug_count = debugfs_create_file("count", 0444,
++ debug_dir, NULL,
++ &count_fops);
++ if (!debug_count)
++ goto err_count;
++
++ debug_max = debugfs_create_file("max", 0444,
++ debug_dir, NULL,
++ &max_fops);
++ if (!debug_max)
++ goto err_max;
++
++ debug_sample_window = debugfs_create_file("window", 0644,
++ debug_dir, NULL,
++ &window_fops);
++ if (!debug_sample_window)
++ goto err_window;
++
++ debug_sample_width = debugfs_create_file("width", 0644,
++ debug_dir, NULL,
++ &width_fops);
++ if (!debug_sample_width)
++ goto err_width;
++
++ debug_threshold = debugfs_create_file("threshold", 0644,
++ debug_dir, NULL,
++ &threshold_fops);
++ if (!debug_threshold)
++ goto err_threshold;
++
++ debug_enable = debugfs_create_file("enable", 0644,
++ debug_dir, &enabled,
++ &enable_fops);
++ if (!debug_enable)
++ goto err_enable;
++
++ else {
++ ret = 0;
++ goto out;
++ }
++
++err_enable:
++ debugfs_remove(debug_threshold);
++err_threshold:
++ debugfs_remove(debug_sample_width);
++err_width:
++ debugfs_remove(debug_sample_window);
++err_window:
++ debugfs_remove(debug_max);
++err_max:
++ debugfs_remove(debug_count);
++err_count:
++ debugfs_remove(debug_sample);
++err_sample:
++ debugfs_remove(debug_dir);
++err_debug_dir:
++out:
++ return ret;
++}
++
++/**
++ * free_debugfs - A function to cleanup the debugfs file interface
+ */
-+void softirq_check_pending_idle(void)
++static void free_debugfs(void)
+{
-+ static int rate_limit;
-+ u32 warnpending = 0, pending = local_softirq_pending();
++ /* could also use a debugfs_remove_recursive */
++ debugfs_remove(debug_enable);
++ debugfs_remove(debug_threshold);
++ debugfs_remove(debug_sample_width);
++ debugfs_remove(debug_sample_window);
++ debugfs_remove(debug_max);
++ debugfs_remove(debug_count);
++ debugfs_remove(debug_sample);
++ debugfs_remove(debug_dir);
++}
+
-+ if (rate_limit >= 10)
-+ return;
++/**
++ * detector_init - Standard module initialization code
++ */
++static int detector_init(void)
++{
++ int ret = -ENOMEM;
+
-+ if (pending) {
-+ struct task_struct *tsk;
++ printk(KERN_INFO BANNER "version %s\n", VERSION);
+
-+ tsk = __get_cpu_var(ksoftirqd);
-+ /*
-+ * The wakeup code in rtmutex.c wakes up the task
-+ * _before_ it sets pi_blocked_on to NULL under
-+ * tsk->pi_lock. So we need to check for both: state
-+ * and pi_blocked_on.
-+ */
-+ raw_spin_lock(&tsk->pi_lock);
++ ret = init_stats();
++ if (0 != ret)
++ goto out;
+
-+ if (!tsk->pi_blocked_on && !(tsk->state == TASK_RUNNING))
-+ warnpending = 1;
++ ret = init_debugfs();
++ if (0 != ret)
++ goto err_stats;
+
-+ raw_spin_unlock(&tsk->pi_lock);
-+ }
++ if (enabled)
++ ret = start_kthread();
++
++ goto out;
++
++err_stats:
++ ring_buffer_free(ring_buffer);
++out:
++ return ret;
+
-+ if (warnpending) {
-+ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
-+ pending);
-+ rate_limit++;
-+ }
+}
-+# else
-+/*
-+ * On !PREEMPT_RT we just printk rate limited:
++
++/**
++ * detector_exit - Standard module cleanup code
+ */
-+void softirq_check_pending_idle(void)
++static void detector_exit(void)
+{
-+ static int rate_limit;
++ int err;
+
-+ if (rate_limit < 10) {
-+ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
-+ local_softirq_pending());
-+ rate_limit++;
++ if (enabled) {
++ enabled = 0;
++ err = stop_kthread();
++ if (err)
++ printk(KERN_ERR BANNER "cannot stop kthread\n");
+ }
++
++ free_debugfs();
++ ring_buffer_free(ring_buffer); /* free up the ring buffer */
++
+}
-+# endif
-+#endif
+
- /*
- * we cannot loop indefinitely here to avoid userspace starvation,
- * but we also don't want to introduce a worst case 1/HZ latency
-@@ -76,6 +138,36 @@ static void wakeup_softirqd(void)
- wake_up_process(tsk);
++module_init(detector_init);
++module_exit(detector_exit);
+diff --git a/drivers/net/3c59x.c b/drivers/net/3c59x.c
+index 8cc2256..1c9ef0f 100644
+--- a/drivers/net/3c59x.c
++++ b/drivers/net/3c59x.c
+@@ -843,9 +843,9 @@ static void poll_vortex(struct net_device *dev)
+ {
+ struct vortex_private *vp = netdev_priv(dev);
+ unsigned long flags;
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ (vp->full_bus_master_rx ? boomerang_interrupt:vortex_interrupt)(dev->irq,dev);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
}
+ #endif
-+static void handle_pending_softirqs(u32 pending, int cpu, int need_rcu_bh_qs)
-+{
-+ struct softirq_action *h = softirq_vec;
-+ unsigned int prev_count = preempt_count();
-+
-+ local_irq_enable();
-+ for ( ; pending; h++, pending >>= 1) {
-+ unsigned int vec_nr = h - softirq_vec;
-+
-+ if (!(pending & 1))
-+ continue;
-+
-+ kstat_incr_softirqs_this_cpu(vec_nr);
-+ trace_softirq_entry(vec_nr);
-+ h->action(h);
-+ trace_softirq_exit(vec_nr);
-+ if (unlikely(prev_count != preempt_count())) {
-+ printk(KERN_ERR
-+ "huh, entered softirq %u %s %p with preempt_count %08x exited with %08x?\n",
-+ vec_nr, softirq_to_name[vec_nr], h->action,
-+ prev_count, (unsigned int) preempt_count());
-+ preempt_count() = prev_count;
-+ }
-+ if (need_rcu_bh_qs)
-+ rcu_bh_qs(cpu);
-+ }
-+ local_irq_disable();
-+}
+@@ -1921,12 +1921,12 @@ static void vortex_tx_timeout(struct net_device *dev)
+ * Block interrupts because vortex_interrupt does a bare spin_lock()
+ */
+ unsigned long flags;
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ if (vp->full_bus_master_tx)
+ boomerang_interrupt(dev->irq, dev);
+ else
+ vortex_interrupt(dev->irq, dev);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ }
+ }
+
+diff --git a/drivers/net/8139too.c b/drivers/net/8139too.c
+index e3bad82..a1e5834 100644
+--- a/drivers/net/8139too.c
++++ b/drivers/net/8139too.c
+@@ -2173,7 +2173,11 @@ static irqreturn_t rtl8139_interrupt (int irq, void *dev_instance)
+ */
+ static void rtl8139_poll_controller(struct net_device *dev)
+ {
+- disable_irq(dev->irq);
++ /*
++ * use _nosync() variant - might be used by netconsole
++ * from atomic contexts:
++ */
++ disable_irq_nosync(dev->irq);
+ rtl8139_interrupt(dev->irq, dev);
+ enable_irq(dev->irq);
+ }
+diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
+index 93359fa..046b9f5 100644
+--- a/drivers/net/Kconfig
++++ b/drivers/net/Kconfig
+@@ -3410,6 +3410,7 @@ config NET_FC
+
+ config NETCONSOLE
+ tristate "Network console logging support"
++ depends on !PREEMPT_RT_FULL
+ ---help---
+ If you want to log kernel messages over the network, enable this.
+ See <file:Documentation/networking/netconsole.txt> for details.
+diff --git a/drivers/net/arm/at91_ether.c b/drivers/net/arm/at91_ether.c
+index e07b314..ef581d9 100644
+--- a/drivers/net/arm/at91_ether.c
++++ b/drivers/net/arm/at91_ether.c
+@@ -199,7 +199,9 @@ static irqreturn_t at91ether_phy_interrupt(int irq, void *dev_id)
+ struct net_device *dev = (struct net_device *) dev_id;
+ struct at91_private *lp = netdev_priv(dev);
+ unsigned int phy;
++ unsigned long flags;
+
++ spin_lock_irqsave(&lp->lock, flags);
+ /*
+ * This hander is triggered on both edges, but the PHY chips expect
+ * level-triggering. We therefore have to check if the PHY actually has
+@@ -241,6 +243,7 @@ static irqreturn_t at91ether_phy_interrupt(int irq, void *dev_id)
+
+ done:
+ disable_mdi();
++ spin_unlock_irqrestore(&lp->lock, flags);
+
+ return IRQ_HANDLED;
+ }
+@@ -397,9 +400,11 @@ static void at91ether_check_link(unsigned long dev_id)
+ struct net_device *dev = (struct net_device *) dev_id;
+ struct at91_private *lp = netdev_priv(dev);
+
++ spin_lock_irq(&lp->lock);
+ enable_mdi();
+ update_linkspeed(dev, 1);
+ disable_mdi();
++ spin_unlock_irq(&lp->lock);
+
+ mod_timer(&lp->check_timer, jiffies + LINK_POLL_INTERVAL);
+ }
+diff --git a/drivers/net/atl1c/atl1c_main.c b/drivers/net/atl1c/atl1c_main.c
+index 1269ba5..f63d31d 100644
+--- a/drivers/net/atl1c/atl1c_main.c
++++ b/drivers/net/atl1c/atl1c_main.c
+@@ -2218,11 +2218,8 @@ static netdev_tx_t atl1c_xmit_frame(struct sk_buff *skb,
+ }
+
+ tpd_req = atl1c_cal_tpd_req(skb);
+- if (!spin_trylock_irqsave(&adapter->tx_lock, flags)) {
+- if (netif_msg_pktdata(adapter))
+- dev_info(&adapter->pdev->dev, "tx locked\n");
+- return NETDEV_TX_LOCKED;
+- }
++ spin_lock_irqsave(&adapter->tx_lock, flags);
+
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /*
- * preempt_count and SOFTIRQ_OFFSET usage:
- * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving
-@@ -206,7 +298,6 @@ EXPORT_SYMBOL(local_bh_enable_ip);
+ if (skb->mark == 0x01)
+ type = atl1c_trans_high;
+ else
+diff --git a/drivers/net/atl1e/atl1e_main.c b/drivers/net/atl1e/atl1e_main.c
+index 86a9122..df092670 100644
+--- a/drivers/net/atl1e/atl1e_main.c
++++ b/drivers/net/atl1e/atl1e_main.c
+@@ -1800,8 +1800,7 @@ static netdev_tx_t atl1e_xmit_frame(struct sk_buff *skb,
+ return NETDEV_TX_OK;
+ }
+ tpd_req = atl1e_cal_tdp_req(skb);
+- if (!spin_trylock_irqsave(&adapter->tx_lock, flags))
+- return NETDEV_TX_LOCKED;
++ spin_lock_irqsave(&adapter->tx_lock, flags);
+
+ if (atl1e_tpd_avail(adapter) < tpd_req) {
+ /* no enough descriptor, just stop queue */
+diff --git a/drivers/net/chelsio/sge.c b/drivers/net/chelsio/sge.c
+index 58380d2..6c2f480 100644
+--- a/drivers/net/chelsio/sge.c
++++ b/drivers/net/chelsio/sge.c
+@@ -1679,8 +1679,7 @@ static int t1_sge_tx(struct sk_buff *skb, struct adapter *adapter,
+ struct cmdQ *q = &sge->cmdQ[qid];
+ unsigned int credits, pidx, genbit, count, use_sched_skb = 0;
+
+- if (!spin_trylock(&q->lock))
+- return NETDEV_TX_LOCKED;
++ spin_lock(&q->lock);
+
+ reclaim_completed_tx(sge, q);
+
+diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c
+index 3fd5a24..6df4fbb 100644
+--- a/drivers/net/ehea/ehea_main.c
++++ b/drivers/net/ehea/ehea_main.c
+@@ -1369,7 +1369,7 @@ static int ehea_reg_interrupts(struct net_device *dev)
+ "%s-queue%d", dev->name, i);
+ ret = ibmebus_request_irq(pr->eq->attr.ist1,
+ ehea_recv_irq_handler,
+- IRQF_DISABLED, pr->int_send_name,
++ IRQF_NO_THREAD, pr->int_send_name,
+ pr);
+ if (ret) {
+ netdev_err(dev, "failed registering irq for ehea_queue port_res_nr:%d, ist=%X\n",
+diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
+index dfa55f9..7697c61 100644
+--- a/drivers/net/gianfar.c
++++ b/drivers/net/gianfar.c
+@@ -1668,7 +1668,7 @@ void stop_gfar(struct net_device *dev)
- asmlinkage void __do_softirq(void)
- {
-- struct softirq_action *h;
- __u32 pending;
- int max_restart = MAX_SOFTIRQ_RESTART;
- int cpu;
-@@ -215,7 +306,7 @@ asmlinkage void __do_softirq(void)
- account_system_vtime(current);
- __local_bh_disable((unsigned long)__builtin_return_address(0),
-- SOFTIRQ_OFFSET);
-+ SOFTIRQ_OFFSET);
- lockdep_softirq_enter();
+ /* Lock it down */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ lock_tx_qs(priv);
+ lock_rx_qs(priv);
- cpu = smp_processor_id();
-@@ -223,36 +314,7 @@ restart:
- /* Reset the pending bitmask before enabling irqs */
- set_softirq_pending(0);
+@@ -1676,7 +1676,7 @@ void stop_gfar(struct net_device *dev)
-- local_irq_enable();
--
-- h = softirq_vec;
--
-- do {
-- if (pending & 1) {
-- unsigned int vec_nr = h - softirq_vec;
-- int prev_count = preempt_count();
--
-- kstat_incr_softirqs_this_cpu(vec_nr);
--
-- trace_softirq_entry(vec_nr);
-- h->action(h);
-- trace_softirq_exit(vec_nr);
-- if (unlikely(prev_count != preempt_count())) {
-- printk(KERN_ERR "huh, entered softirq %u %s %p"
-- "with preempt_count %08x,"
-- " exited with %08x?\n", vec_nr,
-- softirq_to_name[vec_nr], h->action,
-- prev_count, preempt_count());
-- preempt_count() = prev_count;
-- }
--
-- rcu_bh_qs(cpu);
-- }
-- h++;
-- pending >>= 1;
-- } while (pending);
--
-- local_irq_disable();
-+ handle_pending_softirqs(pending, cpu, 1);
+ unlock_rx_qs(priv);
+ unlock_tx_qs(priv);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
- pending = local_softirq_pending();
- if (pending && --max_restart)
-@@ -267,6 +329,26 @@ restart:
- __local_bh_enable(SOFTIRQ_OFFSET);
+ /* Free the IRQs */
+ if (priv->device_flags & FSL_GIANFAR_DEV_HAS_MULTI_INTR) {
+@@ -2951,7 +2951,7 @@ static void adjust_link(struct net_device *dev)
+ struct phy_device *phydev = priv->phydev;
+ int new_state = 0;
+
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ lock_tx_qs(priv);
+
+ if (phydev->link) {
+@@ -3019,7 +3019,7 @@ static void adjust_link(struct net_device *dev)
+ if (new_state && netif_msg_link(priv))
+ phy_print_status(phydev);
+ unlock_tx_qs(priv);
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
}
-+/*
-+ * Called with preemption disabled from run_ksoftirqd()
-+ */
-+static int ksoftirqd_do_softirq(int cpu)
-+{
-+ /*
-+ * Preempt disable stops cpu going offline.
-+ * If already offline, we'll be on wrong CPU:
-+ * don't process.
-+ */
-+ if (cpu_is_offline(cpu))
-+ return -1;
-+
-+ local_irq_disable();
-+ if (local_softirq_pending())
-+ __do_softirq();
-+ local_irq_enable();
-+ return 0;
-+}
-+
- #ifndef __ARCH_HAS_DO_SOFTIRQ
+ /* Update the hash table based on the current list of multicast
+diff --git a/drivers/net/rionet.c b/drivers/net/rionet.c
+index 1f421d7..3e123cd 100644
+--- a/drivers/net/rionet.c
++++ b/drivers/net/rionet.c
+@@ -176,11 +176,7 @@ static int rionet_start_xmit(struct sk_buff *skb, struct net_device *ndev)
+ u16 destid;
+ unsigned long flags;
- asmlinkage void do_softirq(void)
-@@ -289,6 +371,183 @@ asmlinkage void do_softirq(void)
+- local_irq_save(flags);
+- if (!spin_trylock(&rnet->tx_lock)) {
+- local_irq_restore(flags);
+- return NETDEV_TX_LOCKED;
+- }
++ spin_lock_irqsave(&rnet->tx_lock, flags);
- #endif
+ if ((rnet->tx_cnt + 1) > RIONET_TX_RING_SIZE) {
+ netif_stop_queue(ndev);
+diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c
+index df0d2c8..f51d322 100644
+--- a/drivers/net/s2io.c
++++ b/drivers/net/s2io.c
+@@ -4142,12 +4142,7 @@ static netdev_tx_t s2io_xmit(struct sk_buff *skb, struct net_device *dev)
+ [skb->priority & (MAX_TX_FIFOS - 1)];
+ fifo = &mac_control->fifos[queue];
-+static inline void local_bh_disable_nort(void) { local_bh_disable(); }
-+static inline void _local_bh_enable_nort(void) { _local_bh_enable(); }
-+static inline void ksoftirqd_set_sched_params(void) { }
-+static inline void ksoftirqd_clr_sched_params(void) { }
-+
-+#else /* !PREEMPT_RT_FULL */
-+
-+/*
-+ * On RT we serialize softirq execution with a cpu local lock
-+ */
-+static DEFINE_LOCAL_IRQ_LOCK(local_softirq_lock);
-+static DEFINE_PER_CPU(struct task_struct *, local_softirq_runner);
-+
-+static void __do_softirq_common(int need_rcu_bh_qs);
-+
-+void __do_softirq(void)
-+{
-+ __do_softirq_common(0);
-+}
-+
-+void __init softirq_early_init(void)
-+{
-+ local_irq_lock_init(local_softirq_lock);
-+}
-+
-+void local_bh_disable(void)
-+{
-+ migrate_disable();
-+ current->softirq_nestcnt++;
-+}
-+EXPORT_SYMBOL(local_bh_disable);
-+
-+void local_bh_enable(void)
-+{
-+ if (WARN_ON(current->softirq_nestcnt == 0))
-+ return;
+- if (do_spin_lock)
+- spin_lock_irqsave(&fifo->tx_lock, flags);
+- else {
+- if (unlikely(!spin_trylock_irqsave(&fifo->tx_lock, flags)))
+- return NETDEV_TX_LOCKED;
+- }
++ spin_lock_irqsave(&fifo->tx_lock, flags);
+
+ if (sp->config.multiq) {
+ if (__netif_subqueue_stopped(dev, fifo->fifo_no)) {
+diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c
+index ab59300..55394e6 100644
+--- a/drivers/net/sungem.c
++++ b/drivers/net/sungem.c
+@@ -1013,10 +1013,8 @@ static netdev_tx_t gem_start_xmit(struct sk_buff *skb,
+ (csum_stuff_off << 21));
+ }
+
+- if (!spin_trylock_irqsave(&gp->tx_lock, flags)) {
+- /* Tell upper layer to requeue */
+- return NETDEV_TX_LOCKED;
+- }
++ spin_lock_irqsave(&gp->tx_lock, flags);
+
-+ if ((current->softirq_nestcnt == 1) &&
-+ local_softirq_pending() &&
-+ local_trylock(local_softirq_lock)) {
+ /* We raced with gem_do_stop() */
+ if (!gp->running) {
+ spin_unlock_irqrestore(&gp->tx_lock, flags);
+diff --git a/drivers/net/tehuti.c b/drivers/net/tehuti.c
+index 80fbee0..9b92642 100644
+--- a/drivers/net/tehuti.c
++++ b/drivers/net/tehuti.c
+@@ -1627,13 +1627,8 @@ static netdev_tx_t bdx_tx_transmit(struct sk_buff *skb,
+ unsigned long flags;
+
+ ENTER;
+- local_irq_save(flags);
+- if (!spin_trylock(&priv->tx_lock)) {
+- local_irq_restore(flags);
+- DBG("%s[%s]: TX locked, returning NETDEV_TX_LOCKED\n",
+- BDX_DRV_NAME, ndev->name);
+- return NETDEV_TX_LOCKED;
+- }
+
-+ local_irq_disable();
-+ if (local_softirq_pending())
-+ __do_softirq();
-+ local_irq_enable();
-+ local_unlock(local_softirq_lock);
-+ WARN_ON(current->softirq_nestcnt != 1);
-+ }
-+ current->softirq_nestcnt--;
-+ migrate_enable();
-+}
-+EXPORT_SYMBOL(local_bh_enable);
++ spin_lock_irqsave(&priv->tx_lock, flags);
+
+ /* build tx descriptor */
+ BDX_ASSERT(f->m.wptr >= f->m.memsz); /* started with valid wptr */
+diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c
+index 82f8764..fd31008 100644
+--- a/drivers/net/tulip/tulip_core.c
++++ b/drivers/net/tulip/tulip_core.c
+@@ -1951,6 +1951,7 @@ static void __devexit tulip_remove_one (struct pci_dev *pdev)
+ pci_iounmap(pdev, tp->base_addr);
+ free_netdev (dev);
+ pci_release_regions (pdev);
++ pci_disable_device (pdev);
+ pci_set_drvdata (pdev, NULL);
+
+ /* pci_power_off (pdev, -1); */
+diff --git a/drivers/of/base.c b/drivers/of/base.c
+index 632ebae..64a3aca 100644
+--- a/drivers/of/base.c
++++ b/drivers/of/base.c
+@@ -29,7 +29,7 @@ struct device_node *of_chosen;
+ /* use when traversing tree through the allnext, child, sibling,
+ * or parent members of struct device_node.
+ */
+-DEFINE_RWLOCK(devtree_lock);
++DEFINE_RAW_SPINLOCK(devtree_lock);
+
+ int of_n_addr_cells(struct device_node *np)
+ {
+@@ -138,16 +138,14 @@ void of_node_put(struct device_node *node)
+ EXPORT_SYMBOL(of_node_put);
+ #endif /* !CONFIG_SPARC */
+
+-struct property *of_find_property(const struct device_node *np,
+- const char *name,
+- int *lenp)
++static struct property *__of_find_property(const struct device_node *np,
++ const char *name, int *lenp)
+ {
+ struct property *pp;
+
+ if (!np)
+ return NULL;
+
+- read_lock(&devtree_lock);
+ for (pp = np->properties; pp != 0; pp = pp->next) {
+ if (of_prop_cmp(pp->name, name) == 0) {
+ if (lenp != 0)
+@@ -155,7 +153,20 @@ struct property *of_find_property(const struct device_node *np,
+ break;
+ }
+ }
+- read_unlock(&devtree_lock);
+
-+void local_bh_enable_ip(unsigned long ip)
-+{
-+ local_bh_enable();
++ return pp;
+}
-+EXPORT_SYMBOL(local_bh_enable_ip);
+
-+/* For tracing */
-+int notrace __in_softirq(void)
++struct property *of_find_property(const struct device_node *np,
++ const char *name,
++ int *lenp)
+{
-+ if (__get_cpu_var(local_softirq_lock).owner == current)
-+ return __get_cpu_var(local_softirq_lock).nestcnt;
-+ return 0;
-+}
++ struct property *pp;
++ unsigned long flags;
+
-+int in_serving_softirq(void)
++ raw_spin_lock_irqsave(&devtree_lock, flags);
++ pp = __of_find_property(np, name, lenp);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+
+ return pp;
+ }
+@@ -173,13 +184,13 @@ struct device_node *of_find_all_nodes(struct device_node *prev)
+ {
+ struct device_node *np;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock(&devtree_lock);
+ np = prev ? prev->allnext : allnodes;
+ for (; np != NULL; np = np->allnext)
+ if (of_node_get(np))
+ break;
+ of_node_put(prev);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock(&devtree_lock);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_all_nodes);
+@@ -188,8 +199,20 @@ EXPORT_SYMBOL(of_find_all_nodes);
+ * Find a property with a given name for a given node
+ * and return the value.
+ */
++static const void *__of_get_property(const struct device_node *np,
++ const char *name, int *lenp)
+{
-+ int res;
++ struct property *pp = __of_find_property(np, name, lenp);
+
-+ preempt_disable();
-+ res = __get_cpu_var(local_softirq_runner) == current;
-+ preempt_enable();
-+ return res;
++ return pp ? pp->value : NULL;
+}
+
+/*
-+ * Called with bh and local interrupts disabled. For full RT cpu must
-+ * be pinned.
++ * Find a property with a given name for a given node
++ * and return the value.
+ */
-+static void __do_softirq_common(int need_rcu_bh_qs)
-+{
-+ u32 pending = local_softirq_pending();
-+ int cpu = smp_processor_id();
-+
-+ current->softirq_nestcnt++;
-+
-+ /* Reset the pending bitmask before enabling irqs */
-+ set_softirq_pending(0);
-+
-+ __get_cpu_var(local_softirq_runner) = current;
-+
-+ lockdep_softirq_enter();
-+
-+ handle_pending_softirqs(pending, cpu, need_rcu_bh_qs);
-+
-+ pending = local_softirq_pending();
-+ if (pending)
-+ wakeup_softirqd();
-+
-+ lockdep_softirq_exit();
-+ __get_cpu_var(local_softirq_runner) = NULL;
-+
-+ current->softirq_nestcnt--;
-+}
+ const void *of_get_property(const struct device_node *np, const char *name,
+- int *lenp)
++ int *lenp)
+ {
+ struct property *pp = of_find_property(np, name, lenp);
+
+@@ -200,13 +223,13 @@ EXPORT_SYMBOL(of_get_property);
+ /** Checks if the given "compat" string matches one of the strings in
+ * the device's "compatible" property
+ */
+-int of_device_is_compatible(const struct device_node *device,
+- const char *compat)
++static int __of_device_is_compatible(const struct device_node *device,
++ const char *compat)
+ {
+ const char* cp;
+- int cplen, l;
++ int uninitialized_var(cplen), l;
+
+- cp = of_get_property(device, "compatible", &cplen);
++ cp = __of_get_property(device, "compatible", &cplen);
+ if (cp == NULL)
+ return 0;
+ while (cplen > 0) {
+@@ -219,6 +242,21 @@ int of_device_is_compatible(const struct device_node *device,
+
+ return 0;
+ }
+
-+static int __thread_do_softirq(int cpu)
++/** Checks if the given "compat" string matches one of the strings in
++ * the device's "compatible" property
++ */
++int of_device_is_compatible(const struct device_node *device,
++ const char *compat)
+{
-+ /*
-+ * Prevent the current cpu from going offline.
-+ * pin_current_cpu() can reenable preemption and block on the
-+ * hotplug mutex. When it returns, the current cpu is
-+ * pinned. It might be the wrong one, but the offline check
-+ * below catches that.
-+ */
-+ pin_current_cpu();
-+ /*
-+ * If called from ksoftirqd (cpu >= 0) we need to check
-+ * whether we are on the wrong cpu due to cpu offlining. If
-+ * called via thread_do_softirq() no action required.
-+ */
-+ if (cpu >= 0 && cpu_is_offline(cpu)) {
-+ unpin_current_cpu();
-+ return -1;
-+ }
-+ preempt_enable();
-+ local_lock(local_softirq_lock);
-+ local_irq_disable();
-+ /*
-+ * We cannot switch stacks on RT as we want to be able to
-+ * schedule!
-+ */
-+ if (local_softirq_pending())
-+ __do_softirq_common(cpu >= 0);
-+ local_unlock(local_softirq_lock);
-+ unpin_current_cpu();
-+ preempt_disable();
-+ local_irq_enable();
-+ return 0;
++ unsigned long flags;
++ int res;
++
++ raw_spin_lock_irqsave(&devtree_lock, flags);
++ res = __of_device_is_compatible(device, compat);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
++ return res;
+}
+ EXPORT_SYMBOL(of_device_is_compatible);
+
+ /**
+@@ -278,13 +316,14 @@ EXPORT_SYMBOL(of_device_is_available);
+ struct device_node *of_get_parent(const struct device_node *node)
+ {
+ struct device_node *np;
++ unsigned long flags;
+
+ if (!node)
+ return NULL;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = of_node_get(node->parent);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_get_parent);
+@@ -303,14 +342,15 @@ EXPORT_SYMBOL(of_get_parent);
+ struct device_node *of_get_next_parent(struct device_node *node)
+ {
+ struct device_node *parent;
++ unsigned long flags;
+
+ if (!node)
+ return NULL;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ parent = of_node_get(node->parent);
+ of_node_put(node);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return parent;
+ }
+
+@@ -326,14 +366,15 @@ struct device_node *of_get_next_child(const struct device_node *node,
+ struct device_node *prev)
+ {
+ struct device_node *next;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ next = prev ? prev->sibling : node->child;
+ for (; next; next = next->sibling)
+ if (of_node_get(next))
+ break;
+ of_node_put(prev);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return next;
+ }
+ EXPORT_SYMBOL(of_get_next_child);
+@@ -348,14 +389,15 @@ EXPORT_SYMBOL(of_get_next_child);
+ struct device_node *of_find_node_by_path(const char *path)
+ {
+ struct device_node *np = allnodes;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ for (; np; np = np->allnext) {
+ if (np->full_name && (of_node_cmp(np->full_name, path) == 0)
+ && of_node_get(np))
+ break;
+ }
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_node_by_path);
+@@ -375,15 +417,16 @@ struct device_node *of_find_node_by_name(struct device_node *from,
+ const char *name)
+ {
+ struct device_node *np;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = from ? from->allnext : allnodes;
+ for (; np; np = np->allnext)
+ if (np->name && (of_node_cmp(np->name, name) == 0)
+ && of_node_get(np))
+ break;
+ of_node_put(from);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_node_by_name);
+@@ -404,15 +447,16 @@ struct device_node *of_find_node_by_type(struct device_node *from,
+ const char *type)
+ {
+ struct device_node *np;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = from ? from->allnext : allnodes;
+ for (; np; np = np->allnext)
+ if (np->type && (of_node_cmp(np->type, type) == 0)
+ && of_node_get(np))
+ break;
+ of_node_put(from);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_node_by_type);
+@@ -435,18 +479,20 @@ struct device_node *of_find_compatible_node(struct device_node *from,
+ const char *type, const char *compatible)
+ {
+ struct device_node *np;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = from ? from->allnext : allnodes;
+ for (; np; np = np->allnext) {
+ if (type
+ && !(np->type && (of_node_cmp(np->type, type) == 0)))
+ continue;
+- if (of_device_is_compatible(np, compatible) && of_node_get(np))
++ if (__of_device_is_compatible(np, compatible) &&
++ of_node_get(np))
+ break;
+ }
+ of_node_put(from);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_compatible_node);
+@@ -468,8 +514,9 @@ struct device_node *of_find_node_with_property(struct device_node *from,
+ {
+ struct device_node *np;
+ struct property *pp;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = from ? from->allnext : allnodes;
+ for (; np; np = np->allnext) {
+ for (pp = np->properties; pp != 0; pp = pp->next) {
+@@ -481,20 +528,14 @@ struct device_node *of_find_node_with_property(struct device_node *from,
+ }
+ out:
+ of_node_put(from);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
+ }
+ EXPORT_SYMBOL(of_find_node_with_property);
+
+-/**
+- * of_match_node - Tell if an device_node has a matching of_match structure
+- * @matches: array of of device match structures to search in
+- * @node: the of device structure to match against
+- *
+- * Low level utility function used by device matching.
+- */
+-const struct of_device_id *of_match_node(const struct of_device_id *matches,
+- const struct device_node *node)
++static
++const struct of_device_id *__of_match_node(const struct of_device_id *matches,
++ const struct device_node *node)
+ {
+ if (!matches)
+ return NULL;
+@@ -508,14 +549,33 @@ const struct of_device_id *of_match_node(const struct of_device_id *matches,
+ match &= node->type
+ && !strcmp(matches->type, node->type);
+ if (matches->compatible[0])
+- match &= of_device_is_compatible(node,
+- matches->compatible);
++ match &= __of_device_is_compatible(node,
++ matches->compatible);
+ if (match)
+ return matches;
+ matches++;
+ }
+ return NULL;
+ }
+
-+/*
-+ * Called from netif_rx_ni(). Preemption enabled.
++/**
++ * of_match_node - Tell if an device_node has a matching of_match structure
++ * @matches: array of of device match structures to search in
++ * @node: the of device structure to match against
++ *
++ * Low level utility function used by device matching.
+ */
-+void thread_do_softirq(void)
-+{
-+ if (!in_serving_softirq()) {
-+ preempt_disable();
-+ __thread_do_softirq(-1);
-+ preempt_enable();
-+ }
-+}
-+
-+static int ksoftirqd_do_softirq(int cpu)
-+{
-+ return __thread_do_softirq(cpu);
-+}
-+
-+static inline void local_bh_disable_nort(void) { }
-+static inline void _local_bh_enable_nort(void) { }
-+
-+static inline void ksoftirqd_set_sched_params(void)
-+{
-+ struct sched_param param = { .sched_priority = 1 };
-+
-+ sched_setscheduler(current, SCHED_FIFO, ¶m);
-+}
-+
-+static inline void ksoftirqd_clr_sched_params(void)
++const struct of_device_id *of_match_node(const struct of_device_id *matches,
++ const struct device_node *node)
+{
-+ struct sched_param param = { .sched_priority = 0 };
++ const struct of_device_id *match;
++ unsigned long flags;
+
-+ sched_setscheduler(current, SCHED_NORMAL, ¶m);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
++ match = __of_match_node(matches, node);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
++ return match;
+}
-+
-+#endif /* PREEMPT_RT_FULL */
- /*
- * Enter an interrupt context.
- */
-@@ -302,9 +561,9 @@ void irq_enter(void)
- * Prevent raise_softirq from needlessly waking up ksoftirqd
- * here, as softirq will be serviced on return from interrupt.
- */
-- local_bh_disable();
-+ local_bh_disable_nort();
- tick_check_idle(cpu);
-- _local_bh_enable();
-+ _local_bh_enable_nort();
- }
+ EXPORT_SYMBOL(of_match_node);
- __irq_enter();
-@@ -313,6 +572,7 @@ void irq_enter(void)
- #ifdef __ARCH_IRQ_EXIT_IRQS_DISABLED
- static inline void invoke_softirq(void)
+ /**
+@@ -534,15 +594,16 @@ struct device_node *of_find_matching_node(struct device_node *from,
+ const struct of_device_id *matches)
{
-+#ifndef CONFIG_PREEMPT_RT_FULL
- if (!force_irqthreads)
- __do_softirq();
- else {
-@@ -321,10 +581,14 @@ static inline void invoke_softirq(void)
- wakeup_softirqd();
- __local_bh_enable(SOFTIRQ_OFFSET);
+ struct device_node *np;
++ unsigned long flags;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np = from ? from->allnext : allnodes;
+ for (; np; np = np->allnext) {
+- if (of_match_node(matches, np) && of_node_get(np))
++ if (__of_match_node(matches, np) && of_node_get(np))
+ break;
}
-+#else
-+ wakeup_softirqd();
-+#endif
+ of_node_put(from);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return np;
}
- #else
- static inline void invoke_softirq(void)
+ EXPORT_SYMBOL(of_find_matching_node);
+@@ -585,12 +646,12 @@ struct device_node *of_find_node_by_phandle(phandle handle)
{
-+#ifndef CONFIG_PREEMPT_RT_FULL
- if (!force_irqthreads)
- do_softirq();
- else {
-@@ -333,6 +597,9 @@ static inline void invoke_softirq(void)
- wakeup_softirqd();
- __local_bh_enable(SOFTIRQ_OFFSET);
- }
-+#else
-+ wakeup_softirqd();
-+#endif
+ struct device_node *np;
+
+- read_lock(&devtree_lock);
++ raw_spin_lock(&devtree_lock);
+ for (np = allnodes; np; np = np->allnext)
+ if (np->phandle == handle)
+ break;
+ of_node_get(np);
+- read_unlock(&devtree_lock);
++ raw_spin_unlock(&devtree_lock);
+ return np;
}
- #endif
+ EXPORT_SYMBOL(of_find_node_by_phandle);
+@@ -745,18 +806,18 @@ int prom_add_property(struct device_node *np, struct property *prop)
+ unsigned long flags;
-@@ -353,7 +620,7 @@ void irq_exit(void)
- if (idle_cpu(smp_processor_id()) && !in_interrupt() && !need_resched())
- tick_nohz_stop_sched_tick(0);
- #endif
-- preempt_enable_no_resched();
-+ __preempt_enable_no_resched();
+ prop->next = NULL;
+- write_lock_irqsave(&devtree_lock, flags);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ next = &np->properties;
+ while (*next) {
+ if (strcmp(prop->name, (*next)->name) == 0) {
+ /* duplicate ! don't insert it */
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ return -1;
+ }
+ next = &(*next)->next;
+ }
+ *next = prop;
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+
+ #ifdef CONFIG_PROC_DEVICETREE
+ /* try to add to proc as well if it was initialized */
+@@ -781,7 +842,7 @@ int prom_remove_property(struct device_node *np, struct property *prop)
+ unsigned long flags;
+ int found = 0;
+
+- write_lock_irqsave(&devtree_lock, flags);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ next = &np->properties;
+ while (*next) {
+ if (*next == prop) {
+@@ -794,7 +855,7 @@ int prom_remove_property(struct device_node *np, struct property *prop)
+ }
+ next = &(*next)->next;
+ }
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+
+ if (!found)
+ return -ENODEV;
+@@ -824,7 +885,7 @@ int prom_update_property(struct device_node *np,
+ unsigned long flags;
+ int found = 0;
+
+- write_lock_irqsave(&devtree_lock, flags);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ next = &np->properties;
+ while (*next) {
+ if (*next == oldprop) {
+@@ -838,7 +899,7 @@ int prom_update_property(struct device_node *np,
+ }
+ next = &(*next)->next;
+ }
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+
+ if (!found)
+ return -ENODEV;
+@@ -868,12 +929,12 @@ void of_attach_node(struct device_node *np)
+ {
+ unsigned long flags;
+
+- write_lock_irqsave(&devtree_lock, flags);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+ np->sibling = np->parent->child;
+ np->allnext = allnodes;
+ np->parent->child = np;
+ allnodes = np;
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
}
- /*
-@@ -739,29 +1006,21 @@ void __init softirq_init(void)
+ /**
+@@ -887,7 +948,7 @@ void of_detach_node(struct device_node *np)
+ struct device_node *parent;
+ unsigned long flags;
- static int run_ksoftirqd(void * __bind_cpu)
+- write_lock_irqsave(&devtree_lock, flags);
++ raw_spin_lock_irqsave(&devtree_lock, flags);
+
+ parent = np->parent;
+ if (!parent)
+@@ -918,7 +979,7 @@ void of_detach_node(struct device_node *np)
+ of_node_set_flag(np, OF_DETACHED);
+
+ out_unlock:
+- write_unlock_irqrestore(&devtree_lock, flags);
++ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ }
+ #endif /* defined(CONFIG_OF_DYNAMIC) */
+
+diff --git a/drivers/oprofile/event_buffer.c b/drivers/oprofile/event_buffer.c
+index dd87e86..c0cc4e7 100644
+--- a/drivers/oprofile/event_buffer.c
++++ b/drivers/oprofile/event_buffer.c
+@@ -82,10 +82,10 @@ int alloc_event_buffer(void)
{
-+ ksoftirqd_set_sched_params();
-+
- set_current_state(TASK_INTERRUPTIBLE);
+ unsigned long flags;
- while (!kthread_should_stop()) {
- preempt_disable();
-- if (!local_softirq_pending()) {
-- preempt_enable_no_resched();
-- schedule();
-- preempt_disable();
-- }
-+ if (!local_softirq_pending())
-+ schedule_preempt_disabled();
+- spin_lock_irqsave(&oprofilefs_lock, flags);
++ raw_spin_lock_irqsave(&oprofilefs_lock, flags);
+ buffer_size = oprofile_buffer_size;
+ buffer_watershed = oprofile_buffer_watershed;
+- spin_unlock_irqrestore(&oprofilefs_lock, flags);
++ raw_spin_unlock_irqrestore(&oprofilefs_lock, flags);
- __set_current_state(TASK_RUNNING);
+ if (buffer_watershed >= buffer_size)
+ return -EINVAL;
+diff --git a/drivers/oprofile/oprofile_perf.c b/drivers/oprofile/oprofile_perf.c
+index 9046f7b..f088cdc 100644
+--- a/drivers/oprofile/oprofile_perf.c
++++ b/drivers/oprofile/oprofile_perf.c
+@@ -160,9 +160,9 @@ static int oprofile_perf_create_files(struct super_block *sb, struct dentry *roo
- while (local_softirq_pending()) {
-- /* Preempt disable stops cpu going offline.
-- If already offline, we'll be on wrong CPU:
-- don't process */
-- if (cpu_is_offline((long)__bind_cpu))
-+ if (ksoftirqd_do_softirq((long) __bind_cpu))
- goto wait_to_die;
-- local_irq_disable();
-- if (local_softirq_pending())
-- __do_softirq();
-- local_irq_enable();
-- preempt_enable_no_resched();
-+ __preempt_enable_no_resched();
- cond_resched();
- preempt_disable();
- rcu_note_context_switch((long)__bind_cpu);
-@@ -774,6 +1033,7 @@ static int run_ksoftirqd(void * __bind_c
+ static int oprofile_perf_setup(void)
+ {
+- spin_lock(&oprofilefs_lock);
++ raw_spin_lock(&oprofilefs_lock);
+ op_perf_setup();
+- spin_unlock(&oprofilefs_lock);
++ raw_spin_unlock(&oprofilefs_lock);
+ return 0;
+ }
- wait_to_die:
- preempt_enable();
-+ ksoftirqd_clr_sched_params();
- /* Wait for kthread_stop */
- set_current_state(TASK_INTERRUPTIBLE);
- while (!kthread_should_stop()) {
-@@ -850,9 +1110,8 @@ static int __cpuinit cpu_callback(struct
- int hotcpu = (unsigned long)hcpu;
- struct task_struct *p;
+diff --git a/drivers/oprofile/oprofilefs.c b/drivers/oprofile/oprofilefs.c
+index e9ff6f7..d0de6cc 100644
+--- a/drivers/oprofile/oprofilefs.c
++++ b/drivers/oprofile/oprofilefs.c
+@@ -21,7 +21,7 @@
-- switch (action) {
-+ switch (action & ~CPU_TASKS_FROZEN) {
- case CPU_UP_PREPARE:
-- case CPU_UP_PREPARE_FROZEN:
- p = kthread_create_on_node(run_ksoftirqd,
- hcpu,
- cpu_to_node(hotcpu),
-@@ -865,19 +1124,16 @@ static int __cpuinit cpu_callback(struct
- per_cpu(ksoftirqd, hotcpu) = p;
- break;
- case CPU_ONLINE:
-- case CPU_ONLINE_FROZEN:
- wake_up_process(per_cpu(ksoftirqd, hotcpu));
- break;
- #ifdef CONFIG_HOTPLUG_CPU
- case CPU_UP_CANCELED:
-- case CPU_UP_CANCELED_FROZEN:
- if (!per_cpu(ksoftirqd, hotcpu))
- break;
- /* Unbind so it can run. Fall thru. */
- kthread_bind(per_cpu(ksoftirqd, hotcpu),
- cpumask_any(cpu_online_mask));
-- case CPU_DEAD:
-- case CPU_DEAD_FROZEN: {
-+ case CPU_POST_DEAD: {
- static const struct sched_param param = {
- .sched_priority = MAX_RT_PRIO-1
- };
-Index: linux-3.0/kernel/posix-timers.c
-===================================================================
---- linux-3.0.orig/kernel/posix-timers.c
-+++ linux-3.0/kernel/posix-timers.c
-@@ -439,6 +439,7 @@ static enum hrtimer_restart posix_timer_
- static struct pid *good_sigevent(sigevent_t * event)
- {
- struct task_struct *rtn = current->group_leader;
-+ int sig = event->sigev_signo;
+ #define OPROFILEFS_MAGIC 0x6f70726f
- if ((event->sigev_notify & SIGEV_THREAD_ID ) &&
- (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) ||
-@@ -447,7 +448,8 @@ static struct pid *good_sigevent(sigeven
- return NULL;
+-DEFINE_SPINLOCK(oprofilefs_lock);
++DEFINE_RAW_SPINLOCK(oprofilefs_lock);
- if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) &&
-- ((event->sigev_signo <= 0) || (event->sigev_signo > SIGRTMAX)))
-+ (sig <= 0 || sig > SIGRTMAX || sig_kernel_only(sig) ||
-+ sig_kernel_coredump(sig)))
- return NULL;
+ static struct inode *oprofilefs_get_inode(struct super_block *sb, int mode)
+ {
+@@ -76,9 +76,9 @@ int oprofilefs_ulong_from_user(unsigned long *val, char const __user *buf, size_
+ if (copy_from_user(tmpbuf, buf, count))
+ return -EFAULT;
- return task_pid(rtn);
-@@ -764,6 +766,20 @@ SYSCALL_DEFINE1(timer_getoverrun, timer_
- return overrun;
+- spin_lock_irqsave(&oprofilefs_lock, flags);
++ raw_spin_lock_irqsave(&oprofilefs_lock, flags);
+ *val = simple_strtoul(tmpbuf, NULL, 0);
+- spin_unlock_irqrestore(&oprofilefs_lock, flags);
++ raw_spin_unlock_irqrestore(&oprofilefs_lock, flags);
+ return 0;
}
-+/*
-+ * Protected by RCU!
-+ */
-+static void timer_wait_for_callback(struct k_clock *kc, struct k_itimer *timr)
-+{
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ if (kc->timer_set == common_timer_set)
-+ hrtimer_wait_for_timer(&timr->it.real.timer);
-+ else
-+ /* FIXME: Whacky hack for posix-cpu-timers */
-+ schedule_timeout(1);
-+#endif
-+}
-+
- /* Set a POSIX.1b interval timer. */
- /* timr->it_lock is taken. */
- static int
-@@ -841,6 +857,7 @@ retry:
- if (!timr)
- return -EINVAL;
+diff --git a/drivers/pci/dmar.c b/drivers/pci/dmar.c
+index 6dcc7e2..82dd604 100644
+--- a/drivers/pci/dmar.c
++++ b/drivers/pci/dmar.c
+@@ -800,7 +800,7 @@ int alloc_iommu(struct dmar_drhd_unit *drhd)
+ (unsigned long long)iommu->cap,
+ (unsigned long long)iommu->ecap);
-+ rcu_read_lock();
- kc = clockid_to_kclock(timr->it_clock);
- if (WARN_ON_ONCE(!kc || !kc->timer_set))
- error = -EINVAL;
-@@ -849,9 +866,12 @@ retry:
+- spin_lock_init(&iommu->register_lock);
++ raw_spin_lock_init(&iommu->register_lock);
- unlock_timer(timr, flag);
- if (error == TIMER_RETRY) {
-+ timer_wait_for_callback(kc, timr);
- rtn = NULL; // We already got the old time...
-+ rcu_read_unlock();
- goto retry;
+ drhd->iommu = iommu;
+ return 0;
+@@ -921,11 +921,11 @@ int qi_submit_sync(struct qi_desc *desc, struct intel_iommu *iommu)
+ restart:
+ rc = 0;
+
+- spin_lock_irqsave(&qi->q_lock, flags);
++ raw_spin_lock_irqsave(&qi->q_lock, flags);
+ while (qi->free_cnt < 3) {
+- spin_unlock_irqrestore(&qi->q_lock, flags);
++ raw_spin_unlock_irqrestore(&qi->q_lock, flags);
+ cpu_relax();
+- spin_lock_irqsave(&qi->q_lock, flags);
++ raw_spin_lock_irqsave(&qi->q_lock, flags);
}
-+ rcu_read_unlock();
- if (old_setting && !error &&
- copy_to_user(old_setting, &old_spec, sizeof (old_spec)))
-@@ -889,10 +909,15 @@ retry_delete:
- if (!timer)
- return -EINVAL;
+ index = qi->free_head;
+@@ -965,15 +965,15 @@ restart:
+ if (rc)
+ break;
-+ rcu_read_lock();
- if (timer_delete_hook(timer) == TIMER_RETRY) {
- unlock_timer(timer, flags);
-+ timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
-+ timer);
-+ rcu_read_unlock();
- goto retry_delete;
+- spin_unlock(&qi->q_lock);
++ raw_spin_unlock(&qi->q_lock);
+ cpu_relax();
+- spin_lock(&qi->q_lock);
++ raw_spin_lock(&qi->q_lock);
}
-+ rcu_read_unlock();
- spin_lock(¤t->sighand->siglock);
- list_del(&timer->list);
-@@ -918,8 +943,18 @@ static void itimer_delete(struct k_itime
- retry_delete:
- spin_lock_irqsave(&timer->it_lock, flags);
+ qi->desc_status[index] = QI_DONE;
-+ /* On RT we can race with a deletion */
-+ if (!timer->it_signal) {
-+ unlock_timer(timer, flags);
-+ return;
-+ }
-+
- if (timer_delete_hook(timer) == TIMER_RETRY) {
-+ rcu_read_lock();
- unlock_timer(timer, flags);
-+ timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
-+ timer);
-+ rcu_read_unlock();
- goto retry_delete;
- }
- list_del(&timer->list);
-Index: linux-3.0/include/linux/signal.h
-===================================================================
---- linux-3.0.orig/include/linux/signal.h
-+++ linux-3.0/include/linux/signal.h
-@@ -229,6 +229,7 @@ static inline void init_sigpending(struc
- }
+ reclaim_free_desc(qi);
+- spin_unlock_irqrestore(&qi->q_lock, flags);
++ raw_spin_unlock_irqrestore(&qi->q_lock, flags);
- extern void flush_sigqueue(struct sigpending *queue);
-+extern void flush_task_sigqueue(struct task_struct *tsk);
+ if (rc == -EAGAIN)
+ goto restart;
+@@ -1062,7 +1062,7 @@ void dmar_disable_qi(struct intel_iommu *iommu)
+ if (!ecap_qis(iommu->ecap))
+ return;
- /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */
- static inline int valid_signal(unsigned long sig)
-Index: linux-3.0/kernel/exit.c
-===================================================================
---- linux-3.0.orig/kernel/exit.c
-+++ linux-3.0/kernel/exit.c
-@@ -142,7 +142,7 @@ static void __exit_signal(struct task_st
- * Do this under ->siglock, we can race with another thread
- * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
- */
-- flush_sigqueue(&tsk->pending);
-+ flush_task_sigqueue(tsk);
- tsk->sighand = NULL;
- spin_unlock(&sighand->siglock);
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-Index: linux-3.0/kernel/fork.c
-===================================================================
---- linux-3.0.orig/kernel/fork.c
-+++ linux-3.0/kernel/fork.c
-@@ -87,7 +87,7 @@ int max_threads; /* tunable limit on nr
+ sts = dmar_readq(iommu->reg + DMAR_GSTS_REG);
+ if (!(sts & DMA_GSTS_QIES))
+@@ -1082,7 +1082,7 @@ void dmar_disable_qi(struct intel_iommu *iommu)
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl,
+ !(sts & DMA_GSTS_QIES), sts);
+ end:
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
+ }
- DEFINE_PER_CPU(unsigned long, process_counts) = 0;
+ /*
+@@ -1097,7 +1097,7 @@ static void __dmar_enable_qi(struct intel_iommu *iommu)
+ qi->free_head = qi->free_tail = 0;
+ qi->free_cnt = QI_LENGTH;
--__cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */
-+DEFINE_RWLOCK(tasklist_lock); /* outer */
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
- #ifdef CONFIG_PROVE_RCU
- int lockdep_tasklist_lock_is_held(void)
-@@ -198,7 +198,18 @@ void __put_task_struct(struct task_struc
- if (!profile_handoff_task(tsk))
- free_task(tsk);
- }
-+#ifndef CONFIG_PREEMPT_RT_BASE
- EXPORT_SYMBOL_GPL(__put_task_struct);
-+#else
-+void __put_task_struct_cb(struct rcu_head *rhp)
-+{
-+ struct task_struct *tsk = container_of(rhp, struct task_struct, put_rcu);
-+
-+ __put_task_struct(tsk);
-+
-+}
-+EXPORT_SYMBOL_GPL(__put_task_struct_cb);
-+#endif
+ /* write zero to the tail reg */
+ writel(0, iommu->reg + DMAR_IQT_REG);
+@@ -1110,7 +1110,7 @@ static void __dmar_enable_qi(struct intel_iommu *iommu)
+ /* Make sure hardware complete it */
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, readl, (sts & DMA_GSTS_QIES), sts);
- /*
- * macro override instead of weak attribute alias, to workaround
-@@ -546,6 +557,19 @@ void __mmdrop(struct mm_struct *mm)
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
}
- EXPORT_SYMBOL_GPL(__mmdrop);
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+/*
-+ * RCU callback for delayed mm drop. Not strictly rcu, but we don't
-+ * want another facility to make this work.
-+ */
-+void __mmdrop_delayed(struct rcu_head *rhp)
-+{
-+ struct mm_struct *mm = container_of(rhp, struct mm_struct, delayed_drop);
-+
-+ __mmdrop(mm);
-+}
-+#endif
-+
/*
- * Decrement the use count and release all resources for an mm.
- */
-@@ -1030,6 +1054,9 @@ void mm_init_owner(struct mm_struct *mm,
- */
- static void posix_cpu_timers_init(struct task_struct *tsk)
- {
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ tsk->posix_timer_list = NULL;
-+#endif
- tsk->cputime_expires.prof_exp = cputime_zero;
- tsk->cputime_expires.virt_exp = cputime_zero;
- tsk->cputime_expires.sched_exp = 0;
-@@ -1137,6 +1164,7 @@ static struct task_struct *copy_process(
- spin_lock_init(&p->alloc_lock);
+@@ -1159,7 +1159,7 @@ int dmar_enable_qi(struct intel_iommu *iommu)
+ qi->free_head = qi->free_tail = 0;
+ qi->free_cnt = QI_LENGTH;
- init_sigpending(&p->pending);
-+ p->sigqueue_cache = NULL;
+- spin_lock_init(&qi->q_lock);
++ raw_spin_lock_init(&qi->q_lock);
- p->utime = cputime_zero;
- p->stime = cputime_zero;
-@@ -1194,6 +1222,9 @@ static struct task_struct *copy_process(
- p->hardirq_context = 0;
- p->softirq_context = 0;
- #endif
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ p->pagefault_disabled = 0;
-+#endif
- #ifdef CONFIG_LOCKDEP
- p->lockdep_depth = 0; /* no locks held yet */
- p->curr_chain_key = 0;
-Index: linux-3.0/kernel/sched_features.h
-===================================================================
---- linux-3.0.orig/kernel/sched_features.h
-+++ linux-3.0/kernel/sched_features.h
-@@ -65,10 +65,15 @@ SCHED_FEAT(OWNER_SPIN, 1)
- */
- SCHED_FEAT(NONIRQ_POWER, 1)
+ __dmar_enable_qi(iommu);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /*
- * Queue remote wakeups on the target CPU and process them
- * using the scheduler IPI. Reduces rq->lock contention/bounces.
- */
- SCHED_FEAT(TTWU_QUEUE, 1)
-+#else
-+SCHED_FEAT(TTWU_QUEUE, 0)
-+#endif
+@@ -1225,11 +1225,11 @@ void dmar_msi_unmask(struct irq_data *data)
+ unsigned long flag;
- SCHED_FEAT(FORCE_SD_OVERLAP, 0)
-+SCHED_FEAT(RT_RUNTIME_SHARE, 1)
-Index: linux-3.0/include/asm-generic/cmpxchg-local.h
-===================================================================
---- linux-3.0.orig/include/asm-generic/cmpxchg-local.h
-+++ linux-3.0/include/asm-generic/cmpxchg-local.h
-@@ -21,7 +21,7 @@ static inline unsigned long __cmpxchg_lo
- if (size == 8 && sizeof(unsigned long) != 8)
- wrong_size_cmpxchg(ptr);
+ /* unmask it */
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ writel(0, iommu->reg + DMAR_FECTL_REG);
+ /* Read a reg to force flush the post write */
+ readl(iommu->reg + DMAR_FECTL_REG);
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
-- local_irq_save(flags);
-+ raw_local_irq_save(flags);
- switch (size) {
- case 1: prev = *(u8 *)ptr;
- if (prev == old)
-@@ -42,7 +42,7 @@ static inline unsigned long __cmpxchg_lo
- default:
- wrong_size_cmpxchg(ptr);
- }
-- local_irq_restore(flags);
-+ raw_local_irq_restore(flags);
- return prev;
+ void dmar_msi_mask(struct irq_data *data)
+@@ -1238,11 +1238,11 @@ void dmar_msi_mask(struct irq_data *data)
+ struct intel_iommu *iommu = irq_data_get_irq_handler_data(data);
+
+ /* mask it */
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ writel(DMA_FECTL_IM, iommu->reg + DMAR_FECTL_REG);
+ /* Read a reg to force flush the post write */
+ readl(iommu->reg + DMAR_FECTL_REG);
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
}
-@@ -55,11 +55,11 @@ static inline u64 __cmpxchg64_local_gene
- u64 prev;
- unsigned long flags;
+ void dmar_msi_write(int irq, struct msi_msg *msg)
+@@ -1250,11 +1250,11 @@ void dmar_msi_write(int irq, struct msi_msg *msg)
+ struct intel_iommu *iommu = irq_get_handler_data(irq);
+ unsigned long flag;
-- local_irq_save(flags);
-+ raw_local_irq_save(flags);
- prev = *(u64 *)ptr;
- if (prev == old)
- *(u64 *)ptr = new;
-- local_irq_restore(flags);
-+ raw_local_irq_restore(flags);
- return prev;
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ writel(msg->data, iommu->reg + DMAR_FEDATA_REG);
+ writel(msg->address_lo, iommu->reg + DMAR_FEADDR_REG);
+ writel(msg->address_hi, iommu->reg + DMAR_FEUADDR_REG);
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
}
-Index: linux-3.0/drivers/char/random.c
-===================================================================
---- linux-3.0.orig/drivers/char/random.c
-+++ linux-3.0/drivers/char/random.c
-@@ -433,7 +433,7 @@ static struct entropy_store input_pool =
- .poolinfo = &poolinfo_table[0],
- .name = "input",
- .limit = 1,
-- .lock = __SPIN_LOCK_UNLOCKED(&input_pool.lock),
-+ .lock = __SPIN_LOCK_UNLOCKED(input_pool.lock),
- .pool = input_pool_data
- };
+ void dmar_msi_read(int irq, struct msi_msg *msg)
+@@ -1262,11 +1262,11 @@ void dmar_msi_read(int irq, struct msi_msg *msg)
+ struct intel_iommu *iommu = irq_get_handler_data(irq);
+ unsigned long flag;
-@@ -442,7 +442,7 @@ static struct entropy_store blocking_poo
- .name = "blocking",
- .limit = 1,
- .pull = &input_pool,
-- .lock = __SPIN_LOCK_UNLOCKED(&blocking_pool.lock),
-+ .lock = __SPIN_LOCK_UNLOCKED(blocking_pool.lock),
- .pool = blocking_pool_data
- };
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ msg->data = readl(iommu->reg + DMAR_FEDATA_REG);
+ msg->address_lo = readl(iommu->reg + DMAR_FEADDR_REG);
+ msg->address_hi = readl(iommu->reg + DMAR_FEUADDR_REG);
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
-@@ -450,7 +450,7 @@ static struct entropy_store nonblocking_
- .poolinfo = &poolinfo_table[1],
- .name = "nonblocking",
- .pull = &input_pool,
-- .lock = __SPIN_LOCK_UNLOCKED(&nonblocking_pool.lock),
-+ .lock = __SPIN_LOCK_UNLOCKED(nonblocking_pool.lock),
- .pool = nonblocking_pool_data
- };
+ static int dmar_fault_do_one(struct intel_iommu *iommu, int type,
+@@ -1303,7 +1303,7 @@ irqreturn_t dmar_fault(int irq, void *dev_id)
+ u32 fault_status;
+ unsigned long flag;
-@@ -633,8 +633,11 @@ static void add_timer_randomness(struct
- preempt_disable();
- /* if over the trickle threshold, use only 1 in 4096 samples */
- if (input_pool.entropy_count > trickle_thresh &&
-- ((__this_cpu_inc_return(trickle_count) - 1) & 0xfff))
-- goto out;
-+ ((__this_cpu_inc_return(trickle_count) - 1) & 0xfff)) {
-+ preempt_enable();
-+ return;
-+ }
-+ preempt_enable();
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ fault_status = readl(iommu->reg + DMAR_FSTS_REG);
+ if (fault_status)
+ printk(KERN_ERR "DRHD: handling fault status reg %x\n",
+@@ -1342,7 +1342,7 @@ irqreturn_t dmar_fault(int irq, void *dev_id)
+ writel(DMA_FRCD_F, iommu->reg + reg +
+ fault_index * PRIMARY_FAULT_REG_LEN + 12);
- sample.jiffies = jiffies;
- sample.cycles = get_cycles();
-@@ -676,8 +679,6 @@ static void add_timer_randomness(struct
- credit_entropy_bits(&input_pool,
- min_t(int, fls(delta>>1), 11));
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+
+ dmar_fault_do_one(iommu, type, fault_reason,
+ source_id, guest_addr);
+@@ -1350,14 +1350,14 @@ irqreturn_t dmar_fault(int irq, void *dev_id)
+ fault_index++;
+ if (fault_index >= cap_num_fault_regs(iommu->cap))
+ fault_index = 0;
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
}
--out:
-- preempt_enable();
- }
+ clear_rest:
+ /* clear all the other faults */
+ fault_status = readl(iommu->reg + DMAR_FSTS_REG);
+ writel(fault_status, iommu->reg + DMAR_FSTS_REG);
- void add_input_randomness(unsigned int type, unsigned int code,
-Index: linux-3.0/arch/arm/mach-at91/at91rm9200_time.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-at91/at91rm9200_time.c
-+++ linux-3.0/arch/arm/mach-at91/at91rm9200_time.c
-@@ -114,6 +114,7 @@ clkevt32k_mode(enum clock_event_mode mod
- last_crtr = read_CRTR();
- switch (mode) {
- case CLOCK_EVT_MODE_PERIODIC:
-+ setup_irq(AT91_ID_SYS, &at91rm9200_timer_irq);
- /* PIT for periodic irqs; fixed rate of 1/HZ */
- irqmask = AT91_ST_PITS;
- at91_sys_write(AT91_ST_PIMR, LATCH);
-@@ -127,6 +128,7 @@ clkevt32k_mode(enum clock_event_mode mod
- break;
- case CLOCK_EVT_MODE_SHUTDOWN:
- case CLOCK_EVT_MODE_UNUSED:
-+ remove_irq(AT91_ID_SYS, &at91rm9200_timer_irq);
- case CLOCK_EVT_MODE_RESUME:
- irqmask = 0;
- break;
-Index: linux-3.0/arch/arm/mach-at91/at91sam926x_time.c
-===================================================================
---- linux-3.0.orig/arch/arm/mach-at91/at91sam926x_time.c
-+++ linux-3.0/arch/arm/mach-at91/at91sam926x_time.c
-@@ -54,7 +54,7 @@ static struct clocksource pit_clk = {
- .flags = CLOCK_SOURCE_IS_CONTINUOUS,
- };
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ return IRQ_HANDLED;
+ }
--
-+static struct irqaction at91sam926x_pit_irq;
- /*
- * Clockevent device: interrupts every 1/HZ (== pit_cycles * MCK/16)
- */
-@@ -63,6 +63,9 @@ pit_clkevt_mode(enum clock_event_mode mo
- {
- switch (mode) {
- case CLOCK_EVT_MODE_PERIODIC:
-+ /* Set up irq handler */
-+ setup_irq(AT91_ID_SYS, &at91sam926x_pit_irq);
-+
- /* update clocksource counter */
- pit_cnt += pit_cycle * PIT_PICNT(at91_sys_read(AT91_PIT_PIVR));
- at91_sys_write(AT91_PIT_MR, (pit_cycle - 1) | AT91_PIT_PITEN
-@@ -75,6 +78,7 @@ pit_clkevt_mode(enum clock_event_mode mo
- case CLOCK_EVT_MODE_UNUSED:
- /* disable irq, leaving the clocksource active */
- at91_sys_write(AT91_PIT_MR, (pit_cycle - 1) | AT91_PIT_PITEN);
-+ remove_irq(AT91_ID_SYS, &at91sam926x_pit_irq);
- break;
- case CLOCK_EVT_MODE_RESUME:
- break;
-Index: linux-3.0/drivers/clocksource/tcb_clksrc.c
-===================================================================
---- linux-3.0.orig/drivers/clocksource/tcb_clksrc.c
-+++ linux-3.0/drivers/clocksource/tcb_clksrc.c
-@@ -21,8 +21,7 @@
- * resolution better than 200 nsec).
- *
- * - The third channel may be used to provide a 16-bit clockevent
-- * source, used in either periodic or oneshot mode. This runs
-- * at 32 KiHZ, and can handle delays of up to two seconds.
-+ * source, used in either periodic or oneshot mode.
- *
- * A boot clocksource and clockevent source are also currently needed,
- * unless the relevant platforms (ARM/AT91, AVR32/AT32) are changed so
-@@ -68,6 +67,7 @@ static struct clocksource clksrc = {
- struct tc_clkevt_device {
- struct clock_event_device clkevt;
- struct clk *clk;
-+ u32 freq;
- void __iomem *regs;
- };
+diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c
+index f02c34d..8c2564d 100644
+--- a/drivers/pci/intel-iommu.c
++++ b/drivers/pci/intel-iommu.c
+@@ -933,7 +933,7 @@ static void iommu_set_root_entry(struct intel_iommu *iommu)
-@@ -76,13 +76,6 @@ static struct tc_clkevt_device *to_tc_cl
- return container_of(clkevt, struct tc_clkevt_device, clkevt);
- }
+ addr = iommu->root_entry;
--/* For now, we always use the 32K clock ... this optimizes for NO_HZ,
-- * because using one of the divided clocks would usually mean the
-- * tick rate can never be less than several dozen Hz (vs 0.5 Hz).
-- *
-- * A divided clock could be good for high resolution timers, since
-- * 30.5 usec resolution can seem "low".
-- */
- static u32 timer_clock;
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ dmar_writeq(iommu->reg + DMAR_RTADDR_REG, virt_to_phys(addr));
- static void tc_mode(enum clock_event_mode m, struct clock_event_device *d)
-@@ -105,11 +98,12 @@ static void tc_mode(enum clock_event_mod
- case CLOCK_EVT_MODE_PERIODIC:
- clk_enable(tcd->clk);
+ writel(iommu->gcmd | DMA_GCMD_SRTP, iommu->reg + DMAR_GCMD_REG);
+@@ -942,7 +942,7 @@ static void iommu_set_root_entry(struct intel_iommu *iommu)
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (sts & DMA_GSTS_RTPS), sts);
-- /* slow clock, count up to RC, then irq and restart */
-+ /* count up to RC, then irq and restart */
- __raw_writel(timer_clock
- | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO,
- regs + ATMEL_TC_REG(2, CMR));
-- __raw_writel((32768 + HZ/2) / HZ, tcaddr + ATMEL_TC_REG(2, RC));
-+ __raw_writel((tcd->freq + HZ/2)/HZ,
-+ tcaddr + ATMEL_TC_REG(2, RC));
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
- /* Enable clock and interrupts on RC compare */
- __raw_writel(ATMEL_TC_CPCS, regs + ATMEL_TC_REG(2, IER));
-@@ -122,7 +116,7 @@ static void tc_mode(enum clock_event_mod
- case CLOCK_EVT_MODE_ONESHOT:
- clk_enable(tcd->clk);
+ static void iommu_flush_write_buffer(struct intel_iommu *iommu)
+@@ -953,14 +953,14 @@ static void iommu_flush_write_buffer(struct intel_iommu *iommu)
+ if (!rwbf_quirk && !cap_rwbf(iommu->cap))
+ return;
-- /* slow clock, count up to RC, then irq and stop */
-+ /* count up to RC, then irq and stop */
- __raw_writel(timer_clock | ATMEL_TC_CPCSTOP
- | ATMEL_TC_WAVE | ATMEL_TC_WAVESEL_UP_AUTO,
- regs + ATMEL_TC_REG(2, CMR));
-@@ -152,8 +146,12 @@ static struct tc_clkevt_device clkevt =
- .features = CLOCK_EVT_FEAT_PERIODIC
- | CLOCK_EVT_FEAT_ONESHOT,
- .shift = 32,
-+#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
- /* Should be lower than at91rm9200's system timer */
- .rating = 125,
-+#else
-+ .rating = 200,
-+#endif
- .set_next_event = tc_next_event,
- .set_mode = tc_mode,
- },
-@@ -179,8 +177,9 @@ static struct irqaction tc_irqaction = {
- .handler = ch2_irq,
- };
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ writel(iommu->gcmd | DMA_GCMD_WBF, iommu->reg + DMAR_GCMD_REG);
--static void __init setup_clkevents(struct atmel_tc *tc, int clk32k_divisor_idx)
-+static void __init setup_clkevents(struct atmel_tc *tc, int divisor_idx)
- {
-+ unsigned divisor = atmel_tc_divisors[divisor_idx];
- struct clk *t2_clk = tc->clk[2];
- int irq = tc->irq[2];
+ /* Make sure hardware complete it */
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (!(val & DMA_GSTS_WBFS)), val);
-@@ -188,11 +187,17 @@ static void __init setup_clkevents(struc
- clkevt.clk = t2_clk;
- tc_irqaction.dev_id = &clkevt;
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
-- timer_clock = clk32k_divisor_idx;
-+ timer_clock = divisor_idx;
+ /* return value determine if we need a write buffer flush */
+@@ -987,14 +987,14 @@ static void __iommu_flush_context(struct intel_iommu *iommu,
+ }
+ val |= DMA_CCMD_ICC;
-- clkevt.clkevt.mult = div_sc(32768, NSEC_PER_SEC, clkevt.clkevt.shift);
-- clkevt.clkevt.max_delta_ns
-- = clockevent_delta2ns(0xffff, &clkevt.clkevt);
-+ if (!divisor)
-+ clkevt.freq = 32768;
-+ else
-+ clkevt.freq = clk_get_rate(t2_clk)/divisor;
-+
-+ clkevt.clkevt.mult = div_sc(clkevt.freq, NSEC_PER_SEC,
-+ clkevt.clkevt.shift);
-+ clkevt.clkevt.max_delta_ns =
-+ clockevent_delta2ns(0xffff, &clkevt.clkevt);
- clkevt.clkevt.min_delta_ns = clockevent_delta2ns(1, &clkevt.clkevt) + 1;
- clkevt.clkevt.cpumask = cpumask_of(0);
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ dmar_writeq(iommu->reg + DMAR_CCMD_REG, val);
-@@ -295,8 +300,11 @@ static int __init tcb_clksrc_init(void)
- clocksource_register(&clksrc);
+ /* Make sure hardware complete it */
+ IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG,
+ dmar_readq, (!(val & DMA_CCMD_ICC)), val);
- /* channel 2: periodic and oneshot timer support */
-+#ifdef CONFIG_ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
- setup_clkevents(tc, clk32k_divisor_idx);
--
-+#else
-+ setup_clkevents(tc, best_divisor_idx);
-+#endif
- return 0;
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
}
- arch_initcall(tcb_clksrc_init);
-Index: linux-3.0/drivers/misc/Kconfig
-===================================================================
---- linux-3.0.orig/drivers/misc/Kconfig
-+++ linux-3.0/drivers/misc/Kconfig
-@@ -82,6 +82,7 @@ config AB8500_PWM
- config ATMEL_TCLIB
- bool "Atmel AT32/AT91 Timer/Counter Library"
- depends on (AVR32 || ARCH_AT91)
-+ default y if PREEMPT_RT_FULL
- help
- Select this if you want a library to allocate the Timer/Counter
- blocks found on many Atmel processors. This facilitates using
-@@ -97,8 +98,7 @@ config ATMEL_TCB_CLKSRC
- are combined to make a single 32-bit timer.
- When GENERIC_CLOCKEVENTS is defined, the third timer channel
-- may be used as a clock event device supporting oneshot mode
-- (delays of up to two seconds) based on the 32 KiHz clock.
-+ may be used as a clock event device supporting oneshot mode.
+ /* return value determine if we need a write buffer flush */
+@@ -1033,7 +1033,7 @@ static void __iommu_flush_iotlb(struct intel_iommu *iommu, u16 did,
+ if (cap_write_drain(iommu->cap))
+ val |= DMA_TLB_WRITE_DRAIN;
- config ATMEL_TCB_CLKSRC_BLOCK
- int
-@@ -112,6 +112,14 @@ config ATMEL_TCB_CLKSRC_BLOCK
- TC can be used for other purposes, such as PWM generation and
- interval timing.
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ /* Note: Only uses first TLB reg currently */
+ if (val_iva)
+ dmar_writeq(iommu->reg + tlb_offset, val_iva);
+@@ -1043,7 +1043,7 @@ static void __iommu_flush_iotlb(struct intel_iommu *iommu, u16 did,
+ IOMMU_WAIT_OP(iommu, tlb_offset + 8,
+ dmar_readq, (!(val & DMA_TLB_IVT)), val);
-+config ATMEL_TCB_CLKSRC_USE_SLOW_CLOCK
-+ bool "TC Block use 32 KiHz clock"
-+ depends on ATMEL_TCB_CLKSRC
-+ default y if !PREEMPT_RT_FULL
-+ help
-+ Select this to use 32 KiHz base clock rate as TC block clock
-+ source for clock events.
-+
- config IBM_ASM
- tristate "Device driver for IBM RSA service processor"
- depends on X86 && PCI && INPUT && EXPERIMENTAL
-@@ -133,6 +141,35 @@ config IBM_ASM
- for information on the specific driver level and support statement
- for your IBM server.
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
-+config HWLAT_DETECTOR
-+ tristate "Testing module to detect hardware-induced latencies"
-+ depends on DEBUG_FS
-+ depends on RING_BUFFER
-+ default m
-+ ---help---
-+ A simple hardware latency detector. Use this module to detect
-+ large latencies introduced by the behavior of the underlying
-+ system firmware external to Linux. We do this using periodic
-+ use of stop_machine to grab all available CPUs and measure
-+ for unexplainable gaps in the CPU timestamp counter(s). By
-+ default, the module is not enabled until the "enable" file
-+ within the "hwlat_detector" debugfs directory is toggled.
-+
-+ This module is often used to detect SMI (System Management
-+ Interrupts) on x86 systems, though is not x86 specific. To
-+ this end, we default to using a sample window of 1 second,
-+ during which we will sample for 0.5 seconds. If an SMI or
-+ similar event occurs during that time, it is recorded
-+ into an 8K samples global ring buffer until retreived.
-+
-+ WARNING: This software should never be enabled (it can be built
-+ but should not be turned on after it is loaded) in a production
-+ environment where high latencies are a concern since the
-+ sampling mechanism actually introduces latencies for
-+ regular tasks while the CPU(s) are being held.
-+
-+ If unsure, say N
-+
- config PHANTOM
- tristate "Sensable PHANToM (PCI)"
- depends on PCI
-Index: linux-3.0/drivers/net/tulip/tulip_core.c
-===================================================================
---- linux-3.0.orig/drivers/net/tulip/tulip_core.c
-+++ linux-3.0/drivers/net/tulip/tulip_core.c
-@@ -1951,6 +1951,7 @@ static void __devexit tulip_remove_one (
- pci_iounmap(pdev, tp->base_addr);
- free_netdev (dev);
- pci_release_regions (pdev);
-+ pci_disable_device (pdev);
- pci_set_drvdata (pdev, NULL);
+ /* check IOTLB invalidation granularity */
+ if (DMA_TLB_IAIG(val) == 0)
+@@ -1159,7 +1159,7 @@ static void iommu_disable_protect_mem_regions(struct intel_iommu *iommu)
+ u32 pmen;
+ unsigned long flags;
- /* pci_power_off (pdev, -1); */
-Index: linux-3.0/drivers/net/8139too.c
-===================================================================
---- linux-3.0.orig/drivers/net/8139too.c
-+++ linux-3.0/drivers/net/8139too.c
-@@ -2173,7 +2173,11 @@ static irqreturn_t rtl8139_interrupt (in
- */
- static void rtl8139_poll_controller(struct net_device *dev)
- {
-- disable_irq(dev->irq);
-+ /*
-+ * use _nosync() variant - might be used by netconsole
-+ * from atomic contexts:
-+ */
-+ disable_irq_nosync(dev->irq);
- rtl8139_interrupt(dev->irq, dev);
- enable_irq(dev->irq);
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
+ pmen = readl(iommu->reg + DMAR_PMEN_REG);
+ pmen &= ~DMA_PMEN_EPM;
+ writel(pmen, iommu->reg + DMAR_PMEN_REG);
+@@ -1168,7 +1168,7 @@ static void iommu_disable_protect_mem_regions(struct intel_iommu *iommu)
+ IOMMU_WAIT_OP(iommu, DMAR_PMEN_REG,
+ readl, !(pmen & DMA_PMEN_PRS), pmen);
+
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
}
-Index: linux-3.0/drivers/net/ehea/ehea_main.c
-===================================================================
---- linux-3.0.orig/drivers/net/ehea/ehea_main.c
-+++ linux-3.0/drivers/net/ehea/ehea_main.c
-@@ -1369,7 +1369,7 @@ static int ehea_reg_interrupts(struct ne
- "%s-queue%d", dev->name, i);
- ret = ibmebus_request_irq(pr->eq->attr.ist1,
- ehea_recv_irq_handler,
-- IRQF_DISABLED, pr->int_send_name,
-+ IRQF_NO_THREAD, pr->int_send_name,
- pr);
- if (ret) {
- netdev_err(dev, "failed registering irq for ehea_queue port_res_nr:%d, ist=%X\n",
-Index: linux-3.0/drivers/net/arm/at91_ether.c
-===================================================================
---- linux-3.0.orig/drivers/net/arm/at91_ether.c
-+++ linux-3.0/drivers/net/arm/at91_ether.c
-@@ -199,7 +199,9 @@ static irqreturn_t at91ether_phy_interru
- struct net_device *dev = (struct net_device *) dev_id;
- struct at91_private *lp = netdev_priv(dev);
- unsigned int phy;
-+ unsigned long flags;
-+ spin_lock_irqsave(&lp->lock, flags);
- /*
- * This hander is triggered on both edges, but the PHY chips expect
- * level-triggering. We therefore have to check if the PHY actually has
-@@ -241,6 +243,7 @@ static irqreturn_t at91ether_phy_interru
+ static int iommu_enable_translation(struct intel_iommu *iommu)
+@@ -1176,7 +1176,7 @@ static int iommu_enable_translation(struct intel_iommu *iommu)
+ u32 sts;
+ unsigned long flags;
- done:
- disable_mdi();
-+ spin_unlock_irqrestore(&lp->lock, flags);
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
+ iommu->gcmd |= DMA_GCMD_TE;
+ writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
- return IRQ_HANDLED;
+@@ -1184,7 +1184,7 @@ static int iommu_enable_translation(struct intel_iommu *iommu)
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (sts & DMA_GSTS_TES), sts);
+
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
+ return 0;
}
-@@ -397,9 +400,11 @@ static void at91ether_check_link(unsigne
- struct net_device *dev = (struct net_device *) dev_id;
- struct at91_private *lp = netdev_priv(dev);
-+ spin_lock_irq(&lp->lock);
- enable_mdi();
- update_linkspeed(dev, 1);
- disable_mdi();
-+ spin_unlock_irq(&lp->lock);
+@@ -1193,7 +1193,7 @@ static int iommu_disable_translation(struct intel_iommu *iommu)
+ u32 sts;
+ unsigned long flag;
+
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
+ iommu->gcmd &= ~DMA_GCMD_TE;
+ writel(iommu->gcmd, iommu->reg + DMAR_GCMD_REG);
- mod_timer(&lp->check_timer, jiffies + LINK_POLL_INTERVAL);
+@@ -1201,7 +1201,7 @@ static int iommu_disable_translation(struct intel_iommu *iommu)
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (!(sts & DMA_GSTS_TES)), sts);
+
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ return 0;
}
-Index: linux-3.0/include/linux/preempt.h
-===================================================================
---- linux-3.0.orig/include/linux/preempt.h
-+++ linux-3.0/include/linux/preempt.h
-@@ -33,12 +33,18 @@ do { \
- barrier(); \
- } while (0)
--#define preempt_enable_no_resched() \
-+#define __preempt_enable_no_resched() \
- do { \
- barrier(); \
- dec_preempt_count(); \
- } while (0)
+@@ -3321,7 +3321,7 @@ static int iommu_suspend(void)
+ for_each_active_iommu(iommu, drhd) {
+ iommu_disable_translation(iommu);
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+# define preempt_enable_no_resched() __preempt_enable_no_resched()
-+#else
-+# define preempt_enable_no_resched() preempt_enable()
-+#endif
-+
- #define preempt_check_resched() \
- do { \
- if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) \
-@@ -47,7 +53,7 @@ do { \
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- #define preempt_enable() \
- do { \
-- preempt_enable_no_resched(); \
-+ __preempt_enable_no_resched(); \
- barrier(); \
- preempt_check_resched(); \
- } while (0)
-@@ -83,6 +89,7 @@ do { \
- #else
+ iommu->iommu_state[SR_DMAR_FECTL_REG] =
+ readl(iommu->reg + DMAR_FECTL_REG);
+@@ -3332,7 +3332,7 @@ static int iommu_suspend(void)
+ iommu->iommu_state[SR_DMAR_FEUADDR_REG] =
+ readl(iommu->reg + DMAR_FEUADDR_REG);
- #define preempt_disable() do { } while (0)
-+#define __preempt_enable_no_resched() do { } while (0)
- #define preempt_enable_no_resched() do { } while (0)
- #define preempt_enable() do { } while (0)
- #define preempt_check_resched() do { } while (0)
-@@ -93,6 +100,27 @@ do { \
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
+ return 0;
- #endif
+@@ -3359,7 +3359,7 @@ static void iommu_resume(void)
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# define preempt_disable_rt() preempt_disable()
-+# define preempt_enable_rt() preempt_enable()
-+# define preempt_disable_nort() do { } while (0)
-+# define preempt_enable_nort() do { } while (0)
-+# ifdef CONFIG_SMP
-+ extern void migrate_disable(void);
-+ extern void migrate_enable(void);
-+# else /* CONFIG_SMP */
-+# define migrate_disable() do { } while (0)
-+# define migrate_enable() do { } while (0)
-+# endif /* CONFIG_SMP */
-+#else
-+# define preempt_disable_rt() do { } while (0)
-+# define preempt_enable_rt() do { } while (0)
-+# define preempt_disable_nort() preempt_disable()
-+# define preempt_enable_nort() preempt_enable()
-+# define migrate_disable() preempt_disable()
-+# define migrate_enable() preempt_enable()
-+#endif
-+
- #ifdef CONFIG_PREEMPT_NOTIFIERS
+ for_each_active_iommu(iommu, drhd) {
- struct preempt_notifier;
-Index: linux-3.0/include/linux/uaccess.h
-===================================================================
---- linux-3.0.orig/include/linux/uaccess.h
-+++ linux-3.0/include/linux/uaccess.h
-@@ -6,38 +6,37 @@
+- spin_lock_irqsave(&iommu->register_lock, flag);
++ raw_spin_lock_irqsave(&iommu->register_lock, flag);
- /*
- * These routines enable/disable the pagefault handler in that
-- * it will not take any locks and go straight to the fixup table.
-- *
-- * They have great resemblance to the preempt_disable/enable calls
-- * and in fact they are identical; this is because currently there is
-- * no other way to make the pagefault handlers do this. So we do
-- * disable preemption but we don't necessarily care about that.
-+ * it will not take any MM locks and go straight to the fixup table.
- */
--static inline void pagefault_disable(void)
-+static inline void raw_pagefault_disable(void)
- {
- inc_preempt_count();
-- /*
-- * make sure to have issued the store before a pagefault
-- * can hit.
-- */
- barrier();
- }
+ writel(iommu->iommu_state[SR_DMAR_FECTL_REG],
+ iommu->reg + DMAR_FECTL_REG);
+@@ -3370,7 +3370,7 @@ static void iommu_resume(void)
+ writel(iommu->iommu_state[SR_DMAR_FEUADDR_REG],
+ iommu->reg + DMAR_FEUADDR_REG);
--static inline void pagefault_enable(void)
-+static inline void raw_pagefault_enable(void)
- {
-- /*
-- * make sure to issue those last loads/stores before enabling
-- * the pagefault handler again.
-- */
- barrier();
- dec_preempt_count();
-- /*
-- * make sure we do..
-- */
- barrier();
- preempt_check_resched();
+- spin_unlock_irqrestore(&iommu->register_lock, flag);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flag);
+ }
+
+ for_each_active_iommu(iommu, drhd)
+diff --git a/drivers/pci/intr_remapping.c b/drivers/pci/intr_remapping.c
+index 3607faf..c0c1510 100644
+--- a/drivers/pci/intr_remapping.c
++++ b/drivers/pci/intr_remapping.c
+@@ -46,7 +46,7 @@ static __init int setup_intremap(char *str)
}
+ early_param("intremap", setup_intremap);
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+static inline void pagefault_disable(void)
-+{
-+ raw_pagefault_disable();
-+}
-+
-+static inline void pagefault_enable(void)
-+{
-+ raw_pagefault_enable();
-+}
-+#else
-+extern void pagefault_disable(void);
-+extern void pagefault_enable(void);
-+#endif
-+
- #ifndef ARCH_HAS_NOCACHE_UACCESS
+-static DEFINE_SPINLOCK(irq_2_ir_lock);
++static DEFINE_RAW_SPINLOCK(irq_2_ir_lock);
- static inline unsigned long __copy_from_user_inatomic_nocache(void *to,
-@@ -77,9 +76,9 @@ static inline unsigned long __copy_from_
- mm_segment_t old_fs = get_fs(); \
- \
- set_fs(KERNEL_DS); \
-- pagefault_disable(); \
-+ raw_pagefault_disable(); \
- ret = __copy_from_user_inatomic(&(retval), (__force typeof(retval) __user *)(addr), sizeof(retval)); \
-- pagefault_enable(); \
-+ raw_pagefault_enable(); \
- set_fs(old_fs); \
- ret; \
- })
-Index: linux-3.0/arch/alpha/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/alpha/mm/fault.c
-+++ linux-3.0/arch/alpha/mm/fault.c
-@@ -107,7 +107,7 @@ do_page_fault(unsigned long address, uns
+ static struct irq_2_iommu *irq_2_iommu(unsigned int irq)
+ {
+@@ -63,12 +63,12 @@ int get_irte(int irq, struct irte *entry)
+ if (!entry || !irq_iommu)
+ return -1;
- /* If we're in an interrupt context, or have no user context,
- we must not take the fault. */
-- if (!mm || in_atomic())
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
- #ifdef CONFIG_ALPHA_LARGE_VMALLOC
-Index: linux-3.0/arch/arm/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/arm/mm/fault.c
-+++ linux-3.0/arch/arm/mm/fault.c
-@@ -289,7 +289,7 @@ do_page_fault(unsigned long addr, unsign
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+ index = irq_iommu->irte_index + irq_iommu->sub_handle;
+ *entry = *(irq_iommu->iommu->ir_table->base + index);
- /*
-Index: linux-3.0/arch/avr32/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/avr32/mm/fault.c
-+++ linux-3.0/arch/avr32/mm/fault.c
-@@ -81,7 +81,7 @@ asmlinkage void do_page_fault(unsigned l
- * If we're in an interrupt or have no user context, we must
- * not take the fault...
- */
-- if (in_atomic() || !mm || regs->sr & SYSREG_BIT(GM))
-+ if (!mm || regs->sr & SYSREG_BIT(GM) || pagefault_disabled())
- goto no_context;
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+ return 0;
+ }
- local_irq_enable();
-Index: linux-3.0/arch/cris/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/cris/mm/fault.c
-+++ linux-3.0/arch/cris/mm/fault.c
-@@ -111,7 +111,7 @@ do_page_fault(unsigned long address, str
- * user context, we must not take the fault.
- */
+@@ -102,7 +102,7 @@ int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
+ return -1;
+ }
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
+ do {
+ for (i = index; i < index + count; i++)
+ if (table->base[i].present)
+@@ -114,7 +114,7 @@ int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
+ index = (index + count) % INTR_REMAP_TABLE_ENTRIES;
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/frv/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/frv/mm/fault.c
-+++ linux-3.0/arch/frv/mm/fault.c
-@@ -79,7 +79,7 @@ asmlinkage void do_page_fault(int datamm
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+ if (index == start_index) {
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+ printk(KERN_ERR "can't allocate an IRTE\n");
+ return -1;
+ }
+@@ -128,7 +128,7 @@ int alloc_irte(struct intel_iommu *iommu, int irq, u16 count)
+ irq_iommu->sub_handle = 0;
+ irq_iommu->irte_mask = mask;
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/ia64/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/ia64/mm/fault.c
-+++ linux-3.0/arch/ia64/mm/fault.c
-@@ -89,7 +89,7 @@ ia64_do_page_fault (unsigned long addres
- /*
- * If we're in an interrupt or have no user context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
- #ifdef CONFIG_VIRTUAL_MEM_MAP
-Index: linux-3.0/arch/m32r/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/m32r/mm/fault.c
-+++ linux-3.0/arch/m32r/mm/fault.c
-@@ -115,7 +115,7 @@ asmlinkage void do_page_fault(struct pt_
- * If we're in an interrupt or have no user context or are running in an
- * atomic region then we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto bad_area_nosemaphore;
+ return index;
+ }
+@@ -153,10 +153,10 @@ int map_irq_to_irte_handle(int irq, u16 *sub_handle)
+ if (!irq_iommu)
+ return -1;
- /* When running in the kernel we expect faults to occur only to
-Index: linux-3.0/arch/m68k/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/m68k/mm/fault.c
-+++ linux-3.0/arch/m68k/mm/fault.c
-@@ -85,7 +85,7 @@ int do_page_fault(struct pt_regs *regs,
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
+ *sub_handle = irq_iommu->sub_handle;
+ index = irq_iommu->irte_index;
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
+ return index;
+ }
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/microblaze/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/microblaze/mm/fault.c
-+++ linux-3.0/arch/microblaze/mm/fault.c
-@@ -107,7 +107,7 @@ void do_page_fault(struct pt_regs *regs,
- if ((error_code & 0x13) == 0x13 || (error_code & 0x11) == 0x11)
- is_write = 0;
+@@ -168,14 +168,14 @@ int set_irte_irq(int irq, struct intel_iommu *iommu, u16 index, u16 subhandle)
+ if (!irq_iommu)
+ return -1;
-- if (unlikely(in_atomic() || !mm)) {
-+ if (unlikely(!mm || pagefault_disabled())) {
- if (kernel_mode(regs))
- goto bad_area_nosemaphore;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
-Index: linux-3.0/arch/mips/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/mips/mm/fault.c
-+++ linux-3.0/arch/mips/mm/fault.c
-@@ -88,7 +88,7 @@ asmlinkage void __kprobes do_page_fault(
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto bad_area_nosemaphore;
+ irq_iommu->iommu = iommu;
+ irq_iommu->irte_index = index;
+ irq_iommu->sub_handle = subhandle;
+ irq_iommu->irte_mask = 0;
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/mn10300/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/mn10300/mm/fault.c
-+++ linux-3.0/arch/mn10300/mm/fault.c
-@@ -168,7 +168,7 @@ asmlinkage void do_page_fault(struct pt_
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/parisc/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/parisc/mm/fault.c
-+++ linux-3.0/arch/parisc/mm/fault.c
-@@ -176,7 +176,7 @@ void do_page_fault(struct pt_regs *regs,
- unsigned long acc_type;
- int fault;
+ return 0;
+ }
+@@ -191,7 +191,7 @@ int modify_irte(int irq, struct irte *irte_modified)
+ if (!irq_iommu)
+ return -1;
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/powerpc/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/mm/fault.c
-+++ linux-3.0/arch/powerpc/mm/fault.c
-@@ -162,7 +162,7 @@ int __kprobes do_page_fault(struct pt_re
- }
- #endif
+ iommu = irq_iommu->iommu;
-- if (in_atomic() || mm == NULL) {
-+ if (!mm || pagefault_disabled()) {
- if (!user_mode(regs))
- return SIGSEGV;
- /* in_atomic() in user mode is really bad,
-Index: linux-3.0/arch/s390/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/s390/mm/fault.c
-+++ linux-3.0/arch/s390/mm/fault.c
-@@ -295,7 +295,8 @@ static inline int do_exception(struct pt
- * user context.
- */
- fault = VM_FAULT_BADCONTEXT;
-- if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
-+ if (unlikely(!user_space_fault(trans_exc_code) ||
-+ !mm || pagefault_disabled()))
- goto out;
+@@ -203,7 +203,7 @@ int modify_irte(int irq, struct irte *irte_modified)
+ __iommu_flush_cache(iommu, irte, sizeof(*irte));
- address = trans_exc_code & __FAIL_ADDR_MASK;
-@@ -410,7 +411,8 @@ void __kprobes do_asce_exception(struct
- struct mm_struct *mm = current->mm;
- struct vm_area_struct *vma;
+ rc = qi_flush_iec(iommu, index, 0);
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-- if (unlikely(!user_space_fault(trans_exc_code) || in_atomic() || !mm))
-+ if (unlikely(!user_space_fault(trans_exc_code) ||
-+ !mm || pagefault_disabled()))
- goto no_context;
+ return rc;
+ }
+@@ -271,7 +271,7 @@ int free_irte(int irq)
+ if (!irq_iommu)
+ return -1;
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/score/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/score/mm/fault.c
-+++ linux-3.0/arch/score/mm/fault.c
-@@ -72,7 +72,7 @@ asmlinkage void do_page_fault(struct pt_
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto bad_area_nosemaphore;
+- spin_lock_irqsave(&irq_2_ir_lock, flags);
++ raw_spin_lock_irqsave(&irq_2_ir_lock, flags);
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/sh/mm/fault_32.c
-===================================================================
---- linux-3.0.orig/arch/sh/mm/fault_32.c
-+++ linux-3.0/arch/sh/mm/fault_32.c
-@@ -166,7 +166,7 @@ asmlinkage void __kprobes do_page_fault(
- * If we're in an interrupt, have no user context or are running
- * in an atomic region then we must not take the fault:
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto no_context;
+ rc = clear_entries(irq_iommu);
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/sparc/mm/fault_32.c
-===================================================================
---- linux-3.0.orig/arch/sparc/mm/fault_32.c
-+++ linux-3.0/arch/sparc/mm/fault_32.c
-@@ -248,8 +248,8 @@ asmlinkage void do_sparc_fault(struct pt
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-- goto no_context;
-+ if (!mm || pagefault_disabled())
-+ goto no_context;
+@@ -280,7 +280,7 @@ int free_irte(int irq)
+ irq_iommu->sub_handle = 0;
+ irq_iommu->irte_mask = 0;
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
+- spin_unlock_irqrestore(&irq_2_ir_lock, flags);
++ raw_spin_unlock_irqrestore(&irq_2_ir_lock, flags);
-Index: linux-3.0/arch/sparc/mm/fault_64.c
-===================================================================
---- linux-3.0.orig/arch/sparc/mm/fault_64.c
-+++ linux-3.0/arch/sparc/mm/fault_64.c
-@@ -322,7 +322,7 @@ asmlinkage void __kprobes do_sparc64_fau
- * If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm)
-+ if (!mm || pagefault_disabled())
- goto intr_or_no_mm;
+ return rc;
+ }
+@@ -410,7 +410,7 @@ static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode)
- perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, 0, regs, address);
-Index: linux-3.0/arch/tile/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/tile/mm/fault.c
-+++ linux-3.0/arch/tile/mm/fault.c
-@@ -346,7 +346,7 @@ static int handle_page_fault(struct pt_r
- * If we're in an interrupt, have no user context or are running in an
- * atomic region then we must not take the fault.
- */
-- if (in_atomic() || !mm) {
-+ if (!mm || pagefault_disabled()) {
- vma = NULL; /* happy compiler */
- goto bad_area_nosemaphore;
- }
-Index: linux-3.0/arch/um/kernel/trap.c
-===================================================================
---- linux-3.0.orig/arch/um/kernel/trap.c
-+++ linux-3.0/arch/um/kernel/trap.c
-@@ -37,7 +37,7 @@ int handle_page_fault(unsigned long addr
- * If the fault was during atomic operation, don't take the fault, just
- * fail.
- */
-- if (in_atomic())
-+ if (!mm || pagefault_disabled())
- goto out_nosemaphore;
+ addr = virt_to_phys((void *)iommu->ir_table->base);
- down_read(&mm->mmap_sem);
-Index: linux-3.0/arch/x86/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/x86/mm/fault.c
-+++ linux-3.0/arch/x86/mm/fault.c
-@@ -1065,7 +1065,7 @@ do_page_fault(struct pt_regs *regs, unsi
- * If we're in an interrupt, have no user context or are running
- * in an atomic region then we must not take the fault:
- */
-- if (unlikely(in_atomic() || !mm)) {
-+ if (unlikely(!mm || pagefault_disabled())) {
- bad_area_nosemaphore(regs, error_code, address);
- return;
- }
-Index: linux-3.0/arch/xtensa/mm/fault.c
-===================================================================
---- linux-3.0.orig/arch/xtensa/mm/fault.c
-+++ linux-3.0/arch/xtensa/mm/fault.c
-@@ -57,7 +57,7 @@ void do_page_fault(struct pt_regs *regs)
- /* If we're in an interrupt or have no user
- * context, we must not take the fault..
- */
-- if (in_atomic() || !mm) {
-+ if (!mm || pagefault_disabled()) {
- bad_page_fault(regs, address, SIGSEGV);
- return;
- }
-Index: linux-3.0/mm/filemap.c
-===================================================================
---- linux-3.0.orig/mm/filemap.c
-+++ linux-3.0/mm/filemap.c
-@@ -2040,7 +2040,7 @@ size_t iov_iter_copy_from_user_atomic(st
- char *kaddr;
- size_t copied;
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-- BUG_ON(!in_atomic());
-+ BUG_ON(!pagefault_disabled());
- kaddr = kmap_atomic(page, KM_USER0);
- if (likely(i->nr_segs == 1)) {
- int left;
-Index: linux-3.0/arch/x86/mm/highmem_32.c
-===================================================================
---- linux-3.0.orig/arch/x86/mm/highmem_32.c
-+++ linux-3.0/arch/x86/mm/highmem_32.c
-@@ -43,7 +43,7 @@ void *kmap_atomic_prot(struct page *page
- type = kmap_atomic_idx_push();
- idx = type + KM_TYPE_NR*smp_processor_id();
- vaddr = __fix_to_virt(FIX_KMAP_BEGIN + idx);
-- BUG_ON(!pte_none(*(kmap_pte-idx)));
-+ WARN_ON(!pte_none(*(kmap_pte-idx)));
- set_pte(kmap_pte-idx, mk_pte(page, prot));
+ dmar_writeq(iommu->reg + DMAR_IRTA_REG,
+ (addr) | IR_X2APIC_MODE(mode) | INTR_REMAP_TABLE_REG_SIZE);
+@@ -421,7 +421,7 @@ static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode)
- return (void *)vaddr;
-Index: linux-3.0/include/linux/kernel.h
-===================================================================
---- linux-3.0.orig/include/linux/kernel.h
-+++ linux-3.0/include/linux/kernel.h
-@@ -346,7 +346,7 @@ extern enum system_states {
- SYSTEM_HALT,
- SYSTEM_POWER_OFF,
- SYSTEM_RESTART,
-- SYSTEM_SUSPEND_DISK,
-+ SYSTEM_SUSPEND,
- } system_state;
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (sts & DMA_GSTS_IRTPS), sts);
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
- #define TAINT_PROPRIETARY_MODULE 0
-Index: linux-3.0/kernel/power/hibernate.c
-===================================================================
---- linux-3.0.orig/kernel/power/hibernate.c
-+++ linux-3.0/kernel/power/hibernate.c
-@@ -278,6 +278,8 @@ static int create_image(int platform_mod
+ /*
+ * global invalidation of interrupt entry cache before enabling
+@@ -429,7 +429,7 @@ static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode)
+ */
+ qi_global_iec(iommu);
- local_irq_disable();
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
-+ system_state = SYSTEM_SUSPEND;
-+
- error = syscore_suspend();
- if (error) {
- printk(KERN_ERR "PM: Some system devices failed to power down, "
-@@ -305,6 +307,7 @@ static int create_image(int platform_mod
- syscore_resume();
+ /* Enable interrupt-remapping */
+ iommu->gcmd |= DMA_GCMD_IRE;
+@@ -438,7 +438,7 @@ static void iommu_set_intr_remapping(struct intel_iommu *iommu, int mode)
+ IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG,
+ readl, (sts & DMA_GSTS_IRES), sts);
- Enable_irqs:
-+ system_state = SYSTEM_RUNNING;
- local_irq_enable();
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
+ }
- Enable_cpus:
-@@ -412,6 +415,7 @@ static int resume_target_kernel(bool pla
- goto Enable_cpus;
- local_irq_disable();
-+ system_state = SYSTEM_SUSPEND;
+@@ -486,7 +486,7 @@ static void iommu_disable_intr_remapping(struct intel_iommu *iommu)
+ */
+ qi_global_iec(iommu);
- error = syscore_suspend();
- if (error)
-@@ -445,6 +449,7 @@ static int resume_target_kernel(bool pla
- syscore_resume();
+- spin_lock_irqsave(&iommu->register_lock, flags);
++ raw_spin_lock_irqsave(&iommu->register_lock, flags);
- Enable_irqs:
-+ system_state = SYSTEM_RUNNING;
- local_irq_enable();
+ sts = dmar_readq(iommu->reg + DMAR_GSTS_REG);
+ if (!(sts & DMA_GSTS_IRES))
+@@ -499,7 +499,7 @@ static void iommu_disable_intr_remapping(struct intel_iommu *iommu)
+ readl, !(sts & DMA_GSTS_IRES), sts);
- Enable_cpus:
-@@ -524,6 +529,7 @@ int hibernation_platform_enter(void)
- goto Platform_finish;
+ end:
+- spin_unlock_irqrestore(&iommu->register_lock, flags);
++ raw_spin_unlock_irqrestore(&iommu->register_lock, flags);
+ }
- local_irq_disable();
-+ system_state = SYSTEM_SUSPEND;
- syscore_suspend();
- if (pm_wakeup_pending()) {
- error = -EAGAIN;
-@@ -536,6 +542,7 @@ int hibernation_platform_enter(void)
+ int __init intr_remapping_supported(void)
+diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
+index 8885b3e..3e5b58c 100644
+--- a/drivers/scsi/fcoe/fcoe.c
++++ b/drivers/scsi/fcoe/fcoe.c
+@@ -1113,7 +1113,7 @@ static void fcoe_percpu_thread_destroy(unsigned int cpu)
+ struct sk_buff *skb;
+ #ifdef CONFIG_SMP
+ struct fcoe_percpu_s *p0;
+- unsigned targ_cpu = get_cpu();
++ unsigned targ_cpu = get_cpu_light();
+ #endif /* CONFIG_SMP */
- Power_up:
- syscore_resume();
-+ system_state = SYSTEM_RUNNING;
- local_irq_enable();
- enable_nonboot_cpus();
+ FCOE_DBG("Destroying receive thread for CPU %d\n", cpu);
+@@ -1169,7 +1169,7 @@ static void fcoe_percpu_thread_destroy(unsigned int cpu)
+ kfree_skb(skb);
+ spin_unlock_bh(&p->fcoe_rx_list.lock);
+ }
+- put_cpu();
++ put_cpu_light();
+ #else
+ /*
+ * This a non-SMP scenario where the singular Rx thread is
+@@ -1365,11 +1365,11 @@ err2:
+ static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen)
+ {
+ struct fcoe_percpu_s *fps;
+- int rc;
++ int rc, cpu = get_cpu_light();
-Index: linux-3.0/kernel/power/suspend.c
-===================================================================
---- linux-3.0.orig/kernel/power/suspend.c
-+++ linux-3.0/kernel/power/suspend.c
-@@ -163,6 +163,8 @@ static int suspend_enter(suspend_state_t
- arch_suspend_disable_irqs();
- BUG_ON(!irqs_disabled());
+- fps = &get_cpu_var(fcoe_percpu);
++ fps = &per_cpu(fcoe_percpu, cpu);
+ rc = fcoe_get_paged_crc_eof(skb, tlen, fps);
+- put_cpu_var(fcoe_percpu);
++ put_cpu_light();
-+ system_state = SYSTEM_SUSPEND;
-+
- error = syscore_suspend();
- if (!error) {
- if (!(suspend_test(TEST_CORE) || pm_wakeup_pending())) {
-@@ -172,6 +174,8 @@ static int suspend_enter(suspend_state_t
- syscore_resume();
- }
+ return rc;
+ }
+@@ -1561,6 +1561,7 @@ static inline int fcoe_filter_frames(struct fc_lport *lport,
+ stats->InvalidCRCCount++;
+ if (stats->InvalidCRCCount < 5)
+ printk(KERN_WARNING "fcoe: dropping frame with CRC error\n");
++ put_cpu();
+ return -EINVAL;
+ }
-+ system_state = SYSTEM_RUNNING;
-+
- arch_suspend_enable_irqs();
- BUG_ON(irqs_disabled());
+@@ -1605,7 +1606,7 @@ static void fcoe_recv_frame(struct sk_buff *skb)
+ */
+ hp = (struct fcoe_hdr *) skb_network_header(skb);
-Index: linux-3.0/drivers/of/base.c
-===================================================================
---- linux-3.0.orig/drivers/of/base.c
-+++ linux-3.0/drivers/of/base.c
-@@ -29,7 +29,7 @@ struct device_node *of_chosen;
- /* use when traversing tree through the allnext, child, sibling,
- * or parent members of struct device_node.
- */
--DEFINE_RWLOCK(devtree_lock);
-+DEFINE_RAW_SPINLOCK(devtree_lock);
+- stats = per_cpu_ptr(lport->dev_stats, get_cpu());
++ stats = per_cpu_ptr(lport->dev_stats, get_cpu_light());
+ if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) {
+ if (stats->ErrorFrames < 5)
+ printk(KERN_WARNING "fcoe: FCoE version "
+@@ -1637,13 +1638,13 @@ static void fcoe_recv_frame(struct sk_buff *skb)
+ goto drop;
- int of_n_addr_cells(struct device_node *np)
- {
-@@ -138,16 +138,14 @@ void of_node_put(struct device_node *nod
- EXPORT_SYMBOL(of_node_put);
- #endif /* !CONFIG_SPARC */
+ if (!fcoe_filter_frames(lport, fp)) {
+- put_cpu();
++ put_cpu_light();
+ fc_exch_recv(lport, fp);
+ return;
+ }
+ drop:
+ stats->ErrorFrames++;
+- put_cpu();
++ put_cpu_light();
+ kfree_skb(skb);
+ }
--struct property *of_find_property(const struct device_node *np,
-- const char *name,
-- int *lenp)
-+static struct property *__of_find_property(const struct device_node *np,
-+ const char *name, int *lenp)
- {
- struct property *pp;
+diff --git a/drivers/scsi/fcoe/fcoe_ctlr.c b/drivers/scsi/fcoe/fcoe_ctlr.c
+index c74c4b8..dc739e7 100644
+--- a/drivers/scsi/fcoe/fcoe_ctlr.c
++++ b/drivers/scsi/fcoe/fcoe_ctlr.c
+@@ -716,7 +716,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip)
+ unsigned long sel_time = 0;
+ struct fcoe_dev_stats *stats;
- if (!np)
- return NULL;
+- stats = per_cpu_ptr(fip->lp->dev_stats, get_cpu());
++ stats = per_cpu_ptr(fip->lp->dev_stats, get_cpu_light());
-- read_lock(&devtree_lock);
- for (pp = np->properties; pp != 0; pp = pp->next) {
- if (of_prop_cmp(pp->name, name) == 0) {
- if (lenp != 0)
-@@ -155,7 +153,20 @@ struct property *of_find_property(const
- break;
+ list_for_each_entry_safe(fcf, next, &fip->fcfs, list) {
+ deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2;
+@@ -749,7 +749,7 @@ static unsigned long fcoe_ctlr_age_fcfs(struct fcoe_ctlr *fip)
+ sel_time = fcf->time;
}
}
-- read_unlock(&devtree_lock);
-+
-+ return pp;
-+}
-+
-+struct property *of_find_property(const struct device_node *np,
-+ const char *name,
-+ int *lenp)
-+{
-+ struct property *pp;
-+ unsigned long flags;
-+
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
-+ pp = __of_find_property(np, name, lenp);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+- put_cpu();
++ put_cpu_light();
+ if (sel_time && !fip->sel_fcf && !fip->sel_time) {
+ sel_time += msecs_to_jiffies(FCOE_CTLR_START_DELAY);
+ fip->sel_time = sel_time;
+diff --git a/drivers/scsi/libfc/fc_exch.c b/drivers/scsi/libfc/fc_exch.c
+index 3b8a645..0b4c4a5 100644
+--- a/drivers/scsi/libfc/fc_exch.c
++++ b/drivers/scsi/libfc/fc_exch.c
+@@ -705,10 +705,10 @@ static struct fc_exch *fc_exch_em_alloc(struct fc_lport *lport,
+ }
+ memset(ep, 0, sizeof(*ep));
- return pp;
- }
-@@ -173,13 +184,13 @@ struct device_node *of_find_all_nodes(st
- {
- struct device_node *np;
+- cpu = get_cpu();
++ cpu = get_cpu_light();
+ pool = per_cpu_ptr(mp->pool, cpu);
+ spin_lock_bh(&pool->lock);
+- put_cpu();
++ put_cpu_light();
-- read_lock(&devtree_lock);
-+ raw_spin_lock(&devtree_lock);
- np = prev ? prev->allnext : allnodes;
- for (; np != NULL; np = np->allnext)
- if (of_node_get(np))
- break;
- of_node_put(prev);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock(&devtree_lock);
- return np;
- }
- EXPORT_SYMBOL(of_find_all_nodes);
-@@ -188,8 +199,20 @@ EXPORT_SYMBOL(of_find_all_nodes);
- * Find a property with a given name for a given node
- * and return the value.
- */
-+static const void *__of_get_property(const struct device_node *np,
-+ const char *name, int *lenp)
-+{
-+ struct property *pp = __of_find_property(np, name, lenp);
-+
-+ return pp ? pp->value : NULL;
-+}
-+
-+/*
-+ * Find a property with a given name for a given node
-+ * and return the value.
-+ */
- const void *of_get_property(const struct device_node *np, const char *name,
-- int *lenp)
-+ int *lenp)
- {
- struct property *pp = of_find_property(np, name, lenp);
+ /* peek cache of free slot */
+ if (pool->left != FC_XID_UNKNOWN) {
+diff --git a/drivers/tty/serial/8250.c b/drivers/tty/serial/8250.c
+index 762ce72..b89144e 100644
+--- a/drivers/tty/serial/8250.c
++++ b/drivers/tty/serial/8250.c
+@@ -38,6 +38,7 @@
+ #include <linux/nmi.h>
+ #include <linux/mutex.h>
+ #include <linux/slab.h>
++#include <linux/kdb.h>
-@@ -200,13 +223,13 @@ EXPORT_SYMBOL(of_get_property);
- /** Checks if the given "compat" string matches one of the strings in
- * the device's "compatible" property
- */
--int of_device_is_compatible(const struct device_node *device,
-- const char *compat)
-+static int __of_device_is_compatible(const struct device_node *device,
-+ const char *compat)
- {
- const char* cp;
-- int cplen, l;
-+ int uninitialized_var(cplen), l;
+ #include <asm/io.h>
+ #include <asm/irq.h>
+@@ -1678,12 +1679,14 @@ static irqreturn_t serial8250_interrupt(int irq, void *dev_id)
-- cp = of_get_property(device, "compatible", &cplen);
-+ cp = __of_get_property(device, "compatible", &cplen);
- if (cp == NULL)
- return 0;
- while (cplen > 0) {
-@@ -219,6 +242,21 @@ int of_device_is_compatible(const struct
+ l = l->next;
- return 0;
- }
-+
-+/** Checks if the given "compat" string matches one of the strings in
-+ * the device's "compatible" property
-+ */
-+int of_device_is_compatible(const struct device_node *device,
-+ const char *compat)
-+{
-+ unsigned long flags;
-+ int res;
-+
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
-+ res = __of_device_is_compatible(device, compat);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
-+ return res;
-+}
- EXPORT_SYMBOL(of_device_is_compatible);
++#ifndef CONFIG_PREEMPT_RT_FULL
+ if (l == i->head && pass_counter++ > PASS_LIMIT) {
+ /* If we hit this, we're dead. */
+ printk_ratelimited(KERN_ERR
+ "serial8250: too much work for irq%d\n", irq);
+ break;
+ }
++#endif
+ } while (l != end);
- /**
-@@ -278,13 +316,14 @@ EXPORT_SYMBOL(of_device_is_available);
- struct device_node *of_get_parent(const struct device_node *node)
- {
- struct device_node *np;
-+ unsigned long flags;
+ spin_unlock(&i->lock);
+@@ -2894,14 +2897,14 @@ serial8250_console_write(struct console *co, const char *s, unsigned int count)
- if (!node)
- return NULL;
+ touch_nmi_watchdog();
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = of_node_get(node->parent);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
- }
- EXPORT_SYMBOL(of_get_parent);
-@@ -303,14 +342,15 @@ EXPORT_SYMBOL(of_get_parent);
- struct device_node *of_get_next_parent(struct device_node *node)
- {
- struct device_node *parent;
-+ unsigned long flags;
+- local_irq_save(flags);
+- if (up->port.sysrq) {
+- /* serial8250_handle_port() already took the lock */
++ if (unlikely(in_kdb_printk())) {
+ locked = 0;
+- } else if (oops_in_progress) {
+- locked = spin_trylock(&up->port.lock);
+- } else
+- spin_lock(&up->port.lock);
++ } else {
++ if (up->port.sysrq || oops_in_progress)
++ locked = spin_trylock_irqsave(&up->port.lock, flags);
++ else
++ spin_lock_irqsave(&up->port.lock, flags);
++ }
- if (!node)
- return NULL;
+ /*
+ * First save the IER then disable the interrupts
+@@ -2933,8 +2936,7 @@ serial8250_console_write(struct console *co, const char *s, unsigned int count)
+ check_modem_status(up);
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- parent = of_node_get(node->parent);
- of_node_put(node);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return parent;
+ if (locked)
+- spin_unlock(&up->port.lock);
+- local_irq_restore(flags);
++ spin_unlock_irqrestore(&up->port.lock, flags);
}
-@@ -326,14 +366,15 @@ struct device_node *of_get_next_child(co
- struct device_node *prev)
+ static int __init serial8250_console_setup(struct console *co, char *options)
+diff --git a/drivers/tty/serial/cpm_uart/cpm_uart_core.c b/drivers/tty/serial/cpm_uart/cpm_uart_core.c
+index 9488da7..2a75c84 100644
+--- a/drivers/tty/serial/cpm_uart/cpm_uart_core.c
++++ b/drivers/tty/serial/cpm_uart/cpm_uart_core.c
+@@ -1225,7 +1225,7 @@ static void cpm_uart_console_write(struct console *co, const char *s,
{
- struct device_node *next;
-+ unsigned long flags;
+ struct uart_cpm_port *pinfo = &cpm_uart_ports[co->index];
+ unsigned long flags;
+- int nolock = oops_in_progress;
++ int nolock = oops_in_progress || sysrq_in_progress;
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- next = prev ? prev->sibling : node->child;
- for (; next; next = next->sibling)
- if (of_node_get(next))
- break;
- of_node_put(prev);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return next;
- }
- EXPORT_SYMBOL(of_get_next_child);
-@@ -348,14 +389,15 @@ EXPORT_SYMBOL(of_get_next_child);
- struct device_node *of_find_node_by_path(const char *path)
- {
- struct device_node *np = allnodes;
-+ unsigned long flags;
+ if (unlikely(nolock)) {
+ local_irq_save(flags);
+diff --git a/drivers/tty/serial/omap-serial.c b/drivers/tty/serial/omap-serial.c
+index 6d3ec14..a8effec 100644
+--- a/drivers/tty/serial/omap-serial.c
++++ b/drivers/tty/serial/omap-serial.c
+@@ -946,13 +946,12 @@ serial_omap_console_write(struct console *co, const char *s,
+ unsigned int ier;
+ int locked = 1;
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- for (; np; np = np->allnext) {
- if (np->full_name && (of_node_cmp(np->full_name, path) == 0)
- && of_node_get(np))
- break;
- }
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
- }
- EXPORT_SYMBOL(of_find_node_by_path);
-@@ -375,15 +417,16 @@ struct device_node *of_find_node_by_name
- const char *name)
- {
- struct device_node *np;
-+ unsigned long flags;
+- local_irq_save(flags);
+ if (up->port.sysrq)
+ locked = 0;
+ else if (oops_in_progress)
+- locked = spin_trylock(&up->port.lock);
++ locked = spin_trylock_irqsave(&up->port.lock, flags);
+ else
+- spin_lock(&up->port.lock);
++ spin_lock_irqsave(&up->port.lock, flags);
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = from ? from->allnext : allnodes;
- for (; np; np = np->allnext)
- if (np->name && (of_node_cmp(np->name, name) == 0)
- && of_node_get(np))
- break;
- of_node_put(from);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
+ /*
+ * First save the IER then disable the interrupts
+@@ -979,8 +978,7 @@ serial_omap_console_write(struct console *co, const char *s,
+ check_modem_status(up);
+
+ if (locked)
+- spin_unlock(&up->port.lock);
+- local_irq_restore(flags);
++ spin_unlock_irqrestore(&up->port.lock, flags);
}
- EXPORT_SYMBOL(of_find_node_by_name);
-@@ -404,15 +447,16 @@ struct device_node *of_find_node_by_type
- const char *type)
- {
- struct device_node *np;
-+ unsigned long flags;
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = from ? from->allnext : allnodes;
- for (; np; np = np->allnext)
- if (np->type && (of_node_cmp(np->type, type) == 0)
- && of_node_get(np))
- break;
- of_node_put(from);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
+ static int __init
+diff --git a/drivers/tty/sysrq.c b/drivers/tty/sysrq.c
+index 43db715..5219738 100644
+--- a/drivers/tty/sysrq.c
++++ b/drivers/tty/sysrq.c
+@@ -492,6 +492,23 @@ static void __sysrq_put_key_op(int key, struct sysrq_key_op *op_p)
+ sysrq_key_table[i] = op_p;
}
- EXPORT_SYMBOL(of_find_node_by_type);
-@@ -435,18 +479,20 @@ struct device_node *of_find_compatible_n
- const char *type, const char *compatible)
+
++#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
++
++int sysrq_in_progress;
++
++static void set_sysrq_in_progress(int value)
++{
++ sysrq_in_progress = value;
++}
++
++#else
++
++static void set_sysrq_in_progress(int value)
++{
++}
++
++#endif
++
+ void __handle_sysrq(int key, bool check_mask)
{
- struct device_node *np;
-+ unsigned long flags;
+ struct sysrq_key_op *op_p;
+@@ -500,6 +517,9 @@ void __handle_sysrq(int key, bool check_mask)
+ unsigned long flags;
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = from ? from->allnext : allnodes;
- for (; np; np = np->allnext) {
- if (type
- && !(np->type && (of_node_cmp(np->type, type) == 0)))
- continue;
-- if (of_device_is_compatible(np, compatible) && of_node_get(np))
-+ if (__of_device_is_compatible(np, compatible) &&
-+ of_node_get(np))
- break;
+ spin_lock_irqsave(&sysrq_key_table_lock, flags);
++
++ set_sysrq_in_progress(1);
++
+ /*
+ * Raise the apparent loglevel to maximum so that the sysrq header
+ * is shown to provide the user with positive feedback. We do not
+@@ -541,6 +561,9 @@ void __handle_sysrq(int key, bool check_mask)
+ printk("\n");
+ console_loglevel = orig_log_level;
}
- of_node_put(from);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
++
++ set_sysrq_in_progress(0);
++
+ spin_unlock_irqrestore(&sysrq_key_table_lock, flags);
}
- EXPORT_SYMBOL(of_find_compatible_node);
-@@ -468,8 +514,9 @@ struct device_node *of_find_node_with_pr
- {
- struct device_node *np;
- struct property *pp;
-+ unsigned long flags;
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = from ? from->allnext : allnodes;
- for (; np; np = np->allnext) {
- for (pp = np->properties; pp != 0; pp = pp->next) {
-@@ -481,20 +528,14 @@ struct device_node *of_find_node_with_pr
+diff --git a/drivers/tty/tty_buffer.c b/drivers/tty/tty_buffer.c
+index 6c9b7cd..a56c223 100644
+--- a/drivers/tty/tty_buffer.c
++++ b/drivers/tty/tty_buffer.c
+@@ -493,10 +493,14 @@ void tty_flip_buffer_push(struct tty_struct *tty)
+ tty->buf.tail->commit = tty->buf.tail->used;
+ spin_unlock_irqrestore(&tty->buf.lock, flags);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ if (tty->low_latency)
+ flush_to_ldisc(&tty->buf.work);
+ else
+ schedule_work(&tty->buf.work);
++#else
++ flush_to_ldisc(&tty->buf.work);
++#endif
+ }
+ EXPORT_SYMBOL(tty_flip_buffer_push);
+
+diff --git a/drivers/tty/tty_ldisc.c b/drivers/tty/tty_ldisc.c
+index ef925d5..e5e5902 100644
+--- a/drivers/tty/tty_ldisc.c
++++ b/drivers/tty/tty_ldisc.c
+@@ -70,7 +70,7 @@ static void put_ldisc(struct tty_ldisc *ld)
+ * We really want an "atomic_dec_and_lock_irqsave()",
+ * but we don't have it, so this does it by hand.
+ */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ if (atomic_dec_and_lock(&ld->users, &tty_ldisc_lock)) {
+ struct tty_ldisc_ops *ldo = ld->ops;
+
+@@ -81,7 +81,7 @@ static void put_ldisc(struct tty_ldisc *ld)
+ kfree(ld);
+ return;
}
- out:
- of_node_put(from);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ wake_up(&tty_ldisc_idle);
}
- EXPORT_SYMBOL(of_find_node_with_property);
--/**
-- * of_match_node - Tell if an device_node has a matching of_match structure
-- * @matches: array of of device match structures to search in
-- * @node: the of device structure to match against
-- *
-- * Low level utility function used by device matching.
-- */
--const struct of_device_id *of_match_node(const struct of_device_id *matches,
-- const struct device_node *node)
-+static
-+const struct of_device_id *__of_match_node(const struct of_device_id *matches,
-+ const struct device_node *node)
- {
- if (!matches)
- return NULL;
-@@ -508,14 +549,33 @@ const struct of_device_id *of_match_node
- match &= node->type
- && !strcmp(matches->type, node->type);
- if (matches->compatible[0])
-- match &= of_device_is_compatible(node,
-- matches->compatible);
-+ match &= __of_device_is_compatible(node,
-+ matches->compatible);
- if (match)
- return matches;
- matches++;
+diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
+index 39ea00b..cec6dfd 100644
+--- a/drivers/usb/core/hcd.c
++++ b/drivers/usb/core/hcd.c
+@@ -2120,7 +2120,7 @@ irqreturn_t usb_hcd_irq (int irq, void *__hcd)
+ * when the first handler doesn't use it. So let's just
+ * assume it's never used.
+ */
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+
+ if (unlikely(HCD_DEAD(hcd) || !HCD_HW_ACCESSIBLE(hcd))) {
+ rc = IRQ_NONE;
+@@ -2133,7 +2133,7 @@ irqreturn_t usb_hcd_irq (int irq, void *__hcd)
+ rc = IRQ_HANDLED;
}
- return NULL;
+
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ return rc;
}
-+
-+/**
-+ * of_match_node - Tell if an device_node has a matching of_match structure
-+ * @matches: array of of device match structures to search in
-+ * @node: the of device structure to match against
-+ *
-+ * Low level utility function used by device matching.
-+ */
-+const struct of_device_id *of_match_node(const struct of_device_id *matches,
-+ const struct device_node *node)
-+{
-+ const struct of_device_id *match;
-+ unsigned long flags;
-+
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
-+ match = __of_match_node(matches, node);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
-+ return match;
-+}
- EXPORT_SYMBOL(of_match_node);
+ EXPORT_SYMBOL_GPL(usb_hcd_irq);
+diff --git a/drivers/usb/gadget/ci13xxx_udc.c b/drivers/usb/gadget/ci13xxx_udc.c
+index baaf87e..ded9da1 100644
+--- a/drivers/usb/gadget/ci13xxx_udc.c
++++ b/drivers/usb/gadget/ci13xxx_udc.c
+@@ -816,7 +816,7 @@ static struct {
+ } dbg_data = {
+ .idx = 0,
+ .tty = 0,
+- .lck = __RW_LOCK_UNLOCKED(lck)
++ .lck = __RW_LOCK_UNLOCKED(dbg_data.lck)
+ };
/**
-@@ -534,15 +594,16 @@ struct device_node *of_find_matching_nod
- const struct of_device_id *matches)
- {
- struct device_node *np;
-+ unsigned long flags;
+diff --git a/drivers/usb/host/ohci-hcd.c b/drivers/usb/host/ohci-hcd.c
+index f9cf3f0..5fc952d 100644
+--- a/drivers/usb/host/ohci-hcd.c
++++ b/drivers/usb/host/ohci-hcd.c
+@@ -833,9 +833,13 @@ static irqreturn_t ohci_irq (struct usb_hcd *hcd)
+ }
-- read_lock(&devtree_lock);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np = from ? from->allnext : allnodes;
- for (; np; np = np->allnext) {
-- if (of_match_node(matches, np) && of_node_get(np))
-+ if (__of_match_node(matches, np) && of_node_get(np))
- break;
+ if (ints & OHCI_INTR_WDH) {
+- spin_lock (&ohci->lock);
+- dl_done_list (ohci);
+- spin_unlock (&ohci->lock);
++ if (ohci->hcca->done_head == 0) {
++ ints &= ~OHCI_INTR_WDH;
++ } else {
++ spin_lock (&ohci->lock);
++ dl_done_list (ohci);
++ spin_unlock (&ohci->lock);
++ }
}
- of_node_put(from);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return np;
- }
- EXPORT_SYMBOL(of_find_matching_node);
-@@ -585,12 +646,12 @@ struct device_node *of_find_node_by_phan
- {
- struct device_node *np;
-- read_lock(&devtree_lock);
-+ raw_spin_lock(&devtree_lock);
- for (np = allnodes; np; np = np->allnext)
- if (np->phandle == handle)
- break;
- of_node_get(np);
-- read_unlock(&devtree_lock);
-+ raw_spin_unlock(&devtree_lock);
- return np;
+ if (quirk_zfmicro(ohci) && (ints & OHCI_INTR_SF)) {
+diff --git a/drivers/video/console/vgacon.c b/drivers/video/console/vgacon.c
+index 915fd74..d449a74 100644
+--- a/drivers/video/console/vgacon.c
++++ b/drivers/video/console/vgacon.c
+@@ -50,7 +50,7 @@
+ #include <video/vga.h>
+ #include <asm/io.h>
+
+-static DEFINE_SPINLOCK(vga_lock);
++static DEFINE_RAW_SPINLOCK(vga_lock);
+ static int cursor_size_lastfrom;
+ static int cursor_size_lastto;
+ static u32 vgacon_xres;
+@@ -157,7 +157,7 @@ static inline void write_vga(unsigned char reg, unsigned int val)
+ * ddprintk might set the console position from interrupt
+ * handlers, thus the write has to be IRQ-atomic.
+ */
+- spin_lock_irqsave(&vga_lock, flags);
++ raw_spin_lock_irqsave(&vga_lock, flags);
+
+ #ifndef SLOW_VGA
+ v1 = reg + (val & 0xff00);
+@@ -170,7 +170,7 @@ static inline void write_vga(unsigned char reg, unsigned int val)
+ outb_p(reg + 1, vga_video_port_reg);
+ outb_p(val & 0xff, vga_video_port_val);
+ #endif
+- spin_unlock_irqrestore(&vga_lock, flags);
++ raw_spin_unlock_irqrestore(&vga_lock, flags);
}
- EXPORT_SYMBOL(of_find_node_by_phandle);
-@@ -745,18 +806,18 @@ int prom_add_property(struct device_node
- unsigned long flags;
- prop->next = NULL;
-- write_lock_irqsave(&devtree_lock, flags);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- next = &np->properties;
- while (*next) {
- if (strcmp(prop->name, (*next)->name) == 0) {
- /* duplicate ! don't insert it */
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- return -1;
- }
- next = &(*next)->next;
- }
- *next = prop;
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ static inline void vga_set_mem_top(struct vc_data *c)
+@@ -664,7 +664,7 @@ static void vgacon_set_cursor_size(int xpos, int from, int to)
+ cursor_size_lastfrom = from;
+ cursor_size_lastto = to;
- #ifdef CONFIG_PROC_DEVICETREE
- /* try to add to proc as well if it was initialized */
-@@ -781,7 +842,7 @@ int prom_remove_property(struct device_n
- unsigned long flags;
- int found = 0;
+- spin_lock_irqsave(&vga_lock, flags);
++ raw_spin_lock_irqsave(&vga_lock, flags);
+ if (vga_video_type >= VIDEO_TYPE_VGAC) {
+ outb_p(VGA_CRTC_CURSOR_START, vga_video_port_reg);
+ curs = inb_p(vga_video_port_val);
+@@ -682,7 +682,7 @@ static void vgacon_set_cursor_size(int xpos, int from, int to)
+ outb_p(curs, vga_video_port_val);
+ outb_p(VGA_CRTC_CURSOR_END, vga_video_port_reg);
+ outb_p(cure, vga_video_port_val);
+- spin_unlock_irqrestore(&vga_lock, flags);
++ raw_spin_unlock_irqrestore(&vga_lock, flags);
+ }
-- write_lock_irqsave(&devtree_lock, flags);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- next = &np->properties;
- while (*next) {
- if (*next == prop) {
-@@ -794,7 +855,7 @@ int prom_remove_property(struct device_n
- }
- next = &(*next)->next;
- }
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ static void vgacon_cursor(struct vc_data *c, int mode)
+@@ -757,7 +757,7 @@ static int vgacon_doresize(struct vc_data *c,
+ unsigned int scanlines = height * c->vc_font.height;
+ u8 scanlines_lo = 0, r7 = 0, vsync_end = 0, mode, max_scan;
- if (!found)
- return -ENODEV;
-@@ -824,7 +885,7 @@ int prom_update_property(struct device_n
- unsigned long flags;
- int found = 0;
+- spin_lock_irqsave(&vga_lock, flags);
++ raw_spin_lock_irqsave(&vga_lock, flags);
-- write_lock_irqsave(&devtree_lock, flags);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- next = &np->properties;
- while (*next) {
- if (*next == oldprop) {
-@@ -838,7 +899,7 @@ int prom_update_property(struct device_n
- }
- next = &(*next)->next;
+ vgacon_xres = width * VGA_FONTWIDTH;
+ vgacon_yres = height * c->vc_font.height;
+@@ -808,7 +808,7 @@ static int vgacon_doresize(struct vc_data *c,
+ outb_p(vsync_end, vga_video_port_val);
}
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- if (!found)
- return -ENODEV;
-@@ -868,12 +929,12 @@ void of_attach_node(struct device_node *
+- spin_unlock_irqrestore(&vga_lock, flags);
++ raw_spin_unlock_irqrestore(&vga_lock, flags);
+ return 0;
+ }
+
+@@ -891,11 +891,11 @@ static void vga_vesa_blank(struct vgastate *state, int mode)
{
- unsigned long flags;
+ /* save original values of VGA controller registers */
+ if (!vga_vesa_blanked) {
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ vga_state.SeqCtrlIndex = vga_r(state->vgabase, VGA_SEQ_I);
+ vga_state.CrtCtrlIndex = inb_p(vga_video_port_reg);
+ vga_state.CrtMiscIO = vga_r(state->vgabase, VGA_MIS_R);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
-- write_lock_irqsave(&devtree_lock, flags);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
- np->sibling = np->parent->child;
- np->allnext = allnodes;
- np->parent->child = np;
- allnodes = np;
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
- }
+ outb_p(0x00, vga_video_port_reg); /* HorizontalTotal */
+ vga_state.HorizontalTotal = inb_p(vga_video_port_val);
+@@ -918,7 +918,7 @@ static void vga_vesa_blank(struct vgastate *state, int mode)
- /**
-@@ -887,7 +948,7 @@ void of_detach_node(struct device_node *
- struct device_node *parent;
- unsigned long flags;
+ /* assure that video is enabled */
+ /* "0x20" is VIDEO_ENABLE_bit in register 01 of sequencer */
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ vga_wseq(state->vgabase, VGA_SEQ_CLOCK_MODE, vga_state.ClockingMode | 0x20);
-- write_lock_irqsave(&devtree_lock, flags);
-+ raw_spin_lock_irqsave(&devtree_lock, flags);
+ /* test for vertical retrace in process.... */
+@@ -954,13 +954,13 @@ static void vga_vesa_blank(struct vgastate *state, int mode)
+ /* restore both index registers */
+ vga_w(state->vgabase, VGA_SEQ_I, vga_state.SeqCtrlIndex);
+ outb_p(vga_state.CrtCtrlIndex, vga_video_port_reg);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
+ }
- parent = np->parent;
- if (!parent)
-@@ -918,7 +979,7 @@ void of_detach_node(struct device_node *
- of_node_set_flag(np, OF_DETACHED);
+ static void vga_vesa_unblank(struct vgastate *state)
+ {
+ /* restore original values of VGA controller registers */
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ vga_w(state->vgabase, VGA_MIS_W, vga_state.CrtMiscIO);
- out_unlock:
-- write_unlock_irqrestore(&devtree_lock, flags);
-+ raw_spin_unlock_irqrestore(&devtree_lock, flags);
+ outb_p(0x00, vga_video_port_reg); /* HorizontalTotal */
+@@ -985,7 +985,7 @@ static void vga_vesa_unblank(struct vgastate *state)
+ /* restore index/control registers */
+ vga_w(state->vgabase, VGA_SEQ_I, vga_state.SeqCtrlIndex);
+ outb_p(vga_state.CrtCtrlIndex, vga_video_port_reg);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
}
- #endif /* defined(CONFIG_OF_DYNAMIC) */
-Index: linux-3.0/arch/sparc/kernel/prom_common.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/prom_common.c
-+++ linux-3.0/arch/sparc/kernel/prom_common.c
-@@ -67,7 +67,7 @@ int of_set_property(struct device_node *
- err = -ENODEV;
+ static void vga_pal_blank(struct vgastate *state)
+@@ -1104,7 +1104,7 @@ static int vgacon_do_font_op(struct vgastate *state,char *arg,int set,int ch512)
+ charmap += 4 * cmapsz;
+ #endif
- mutex_lock(&of_set_property_mutex);
-- write_lock(&devtree_lock);
-+ raw_spin_lock(&devtree_lock);
- prevp = &dp->properties;
- while (*prevp) {
- struct property *prop = *prevp;
-@@ -94,7 +94,7 @@ int of_set_property(struct device_node *
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ /* First, the Sequencer */
+ vga_wseq(state->vgabase, VGA_SEQ_RESET, 0x1);
+ /* CPU writes only to map 2 */
+@@ -1120,7 +1120,7 @@ static int vgacon_do_font_op(struct vgastate *state,char *arg,int set,int ch512)
+ vga_wgfx(state->vgabase, VGA_GFX_MODE, 0x00);
+ /* map start at A000:0000 */
+ vga_wgfx(state->vgabase, VGA_GFX_MISC, 0x00);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
+
+ if (arg) {
+ if (set)
+@@ -1147,7 +1147,7 @@ static int vgacon_do_font_op(struct vgastate *state,char *arg,int set,int ch512)
}
- prevp = &(*prevp)->next;
}
-- write_unlock(&devtree_lock);
-+ raw_spin_unlock(&devtree_lock);
- mutex_unlock(&of_set_property_mutex);
- /* XXX Upate procfs if necessary... */
-Index: linux-3.0/include/linux/of.h
-===================================================================
---- linux-3.0.orig/include/linux/of.h
-+++ linux-3.0/include/linux/of.h
-@@ -68,7 +68,7 @@ struct device_node {
- /* Pointer for first entry in chain of all nodes. */
- extern struct device_node *allnodes;
- extern struct device_node *of_chosen;
--extern rwlock_t devtree_lock;
-+extern raw_spinlock_t devtree_lock;
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ /* First, the sequencer, Synchronous reset */
+ vga_wseq(state->vgabase, VGA_SEQ_RESET, 0x01);
+ /* CPU writes to maps 0 and 1 */
+@@ -1186,7 +1186,7 @@ static int vgacon_do_font_op(struct vgastate *state,char *arg,int set,int ch512)
+ inb_p(video_port_status);
+ vga_wattr(state->vgabase, VGA_AR_ENABLE_DISPLAY, 0);
+ }
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
+ return 0;
+ }
+
+@@ -1211,26 +1211,26 @@ static int vgacon_adjust_height(struct vc_data *vc, unsigned fontheight)
+ registers; they are write-only on EGA, but it appears that they
+ are all don't care bits on EGA, so I guess it doesn't matter. */
+
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ outb_p(0x07, vga_video_port_reg); /* CRTC overflow register */
+ ovr = inb_p(vga_video_port_val);
+ outb_p(0x09, vga_video_port_reg); /* Font size register */
+ fsr = inb_p(vga_video_port_val);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
+
+ vde = maxscan & 0xff; /* Vertical display end reg */
+ ovr = (ovr & 0xbd) + /* Overflow register */
+ ((maxscan & 0x100) >> 7) + ((maxscan & 0x200) >> 3);
+ fsr = (fsr & 0xe0) + (fontheight - 1); /* Font size register */
+
+- spin_lock_irq(&vga_lock);
++ raw_spin_lock_irq(&vga_lock);
+ outb_p(0x07, vga_video_port_reg); /* CRTC overflow register */
+ outb_p(ovr, vga_video_port_val);
+ outb_p(0x09, vga_video_port_reg); /* Font size */
+ outb_p(fsr, vga_video_port_val);
+ outb_p(0x12, vga_video_port_reg); /* Vertical display limit */
+ outb_p(vde, vga_video_port_val);
+- spin_unlock_irq(&vga_lock);
++ raw_spin_unlock_irq(&vga_lock);
+ vga_video_font_height = fontheight;
+
+ for (i = 0; i < MAX_NR_CONSOLES; i++) {
+diff --git a/drivers/watchdog/octeon-wdt-main.c b/drivers/watchdog/octeon-wdt-main.c
+index 945ee83..7c0d863 100644
+--- a/drivers/watchdog/octeon-wdt-main.c
++++ b/drivers/watchdog/octeon-wdt-main.c
+@@ -402,7 +402,7 @@ static void octeon_wdt_setup_interrupt(int cpu)
+ irq = OCTEON_IRQ_WDOG0 + core;
- static inline bool of_have_populated_dt(void)
- {
-Index: linux-3.0/include/linux/list.h
-===================================================================
---- linux-3.0.orig/include/linux/list.h
-+++ linux-3.0/include/linux/list.h
-@@ -362,6 +362,17 @@ static inline void list_splice_tail_init
- list_entry((ptr)->next, type, member)
+ if (request_irq(irq, octeon_wdt_poke_irq,
+- IRQF_DISABLED, "octeon_wdt", octeon_wdt_poke_irq))
++ IRQF_NO_THREAD, "octeon_wdt", octeon_wdt_poke_irq))
+ panic("octeon_wdt: Couldn't obtain irq %d", irq);
- /**
-+ * list_last_entry - get the last element from a list
-+ * @ptr: the list head to take the element from.
-+ * @type: the type of the struct this is embedded in.
-+ * @member: the name of the list_struct within the struct.
-+ *
-+ * Note, that list is expected to be not empty.
-+ */
-+#define list_last_entry(ptr, type, member) \
-+ list_entry((ptr)->prev, type, member)
-+
-+/**
- * list_for_each - iterate over a list
- * @pos: the &struct list_head to use as a loop cursor.
- * @head: the head for your list.
-Index: linux-3.0/mm/page_alloc.c
-===================================================================
---- linux-3.0.orig/mm/page_alloc.c
-+++ linux-3.0/mm/page_alloc.c
-@@ -57,6 +57,7 @@
- #include <linux/ftrace_event.h>
- #include <linux/memcontrol.h>
- #include <linux/prefetch.h>
-+#include <linux/locallock.h>
+ cpumask_set_cpu(cpu, &irq_enabled_cpus);
+diff --git a/fs/attr.c b/fs/attr.c
+index caf2aa5..8d56f3c 100644
+--- a/fs/attr.c
++++ b/fs/attr.c
+@@ -233,7 +233,7 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
+ return error;
- #include <asm/tlbflush.h>
- #include <asm/div64.h>
-@@ -222,6 +223,18 @@ EXPORT_SYMBOL(nr_node_ids);
- EXPORT_SYMBOL(nr_online_nodes);
- #endif
+ if (ia_valid & ATTR_SIZE)
+- down_write(&dentry->d_inode->i_alloc_sem);
++ anon_down_write(&dentry->d_inode->i_alloc_sem);
-+static DEFINE_LOCAL_IRQ_LOCK(pa_lock);
-+
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+# define cpu_lock_irqsave(cpu, flags) \
-+ spin_lock_irqsave(&per_cpu(pa_lock, cpu).lock, flags)
-+# define cpu_unlock_irqrestore(cpu, flags) \
-+ spin_unlock_irqrestore(&per_cpu(pa_lock, cpu).lock, flags)
-+#else
-+# define cpu_lock_irqsave(cpu, flags) local_irq_save(flags)
-+# define cpu_unlock_irqrestore(cpu, flags) local_irq_restore(flags)
-+#endif
-+
- int page_group_by_mobility_disabled __read_mostly;
+ if (inode->i_op->setattr)
+ error = inode->i_op->setattr(dentry, attr);
+@@ -241,7 +241,7 @@ int notify_change(struct dentry * dentry, struct iattr * attr)
+ error = simple_setattr(dentry, attr);
- static void set_pageblock_migratetype(struct page *page, int migratetype)
-@@ -580,7 +593,7 @@ static inline int free_pages_check(struc
- }
+ if (ia_valid & ATTR_SIZE)
+- up_write(&dentry->d_inode->i_alloc_sem);
++ anon_up_write(&dentry->d_inode->i_alloc_sem);
- /*
-- * Frees a number of pages from the PCP lists
-+ * Frees a number of pages which have been collected from the pcp lists.
- * Assumes all pages on list are in same zone, and of same order.
- * count is the number of pages to free.
- *
-@@ -591,16 +604,42 @@ static inline int free_pages_check(struc
- * pinned" detection logic.
+ if (!error)
+ fsnotify_change(dentry, ia_valid);
+diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c
+index 66fa43d..3b02fb1 100644
+--- a/fs/btrfs/locking.c
++++ b/fs/btrfs/locking.c
+@@ -93,6 +93,7 @@ static int btrfs_spin_on_block(struct extent_buffer *eb)
*/
- static void free_pcppages_bulk(struct zone *zone, int count,
-- struct per_cpu_pages *pcp)
-+ struct list_head *list)
+ int btrfs_try_spin_lock(struct extent_buffer *eb)
{
-- int migratetype = 0;
-- int batch_free = 0;
- int to_free = count;
-+ unsigned long flags;
-
-- spin_lock(&zone->lock);
-+ spin_lock_irqsave(&zone->lock, flags);
- zone->all_unreclaimable = 0;
- zone->pages_scanned = 0;
-
-+ while (!list_empty(list)) {
-+ struct page *page = list_first_entry(list, struct page, lru);
-+
-+ /* must delete as __free_one_page list manipulates */
-+ list_del(&page->lru);
-+ /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
-+ __free_one_page(page, zone, 0, page_private(page));
-+ trace_mm_page_pcpu_drain(page, 0, page_private(page));
-+ to_free--;
-+ }
-+ WARN_ON(to_free != 0);
-+ __mod_zone_page_state(zone, NR_FREE_PAGES, count);
-+ spin_unlock_irqrestore(&zone->lock, flags);
-+}
-+
-+/*
-+ * Moves a number of pages from the PCP lists to free list which
-+ * is freed outside of the locked region.
-+ *
-+ * Assumes all pages on list are in same zone, and of same order.
-+ * count is the number of pages to free.
-+ */
-+static void isolate_pcp_pages(int to_free, struct per_cpu_pages *src,
-+ struct list_head *dst)
-+{
-+ int migratetype = 0, batch_free = 0;
-+
- while (to_free) {
- struct page *page;
- struct list_head *list;
-@@ -616,7 +655,7 @@ static void free_pcppages_bulk(struct zo
- batch_free++;
- if (++migratetype == MIGRATE_PCPTYPES)
- migratetype = 0;
-- list = &pcp->lists[migratetype];
-+ list = &src->lists[migratetype];
- } while (list_empty(list));
-
- /* This is the only non-empty list. Free them all. */
-@@ -624,28 +663,25 @@ static void free_pcppages_bulk(struct zo
- batch_free = to_free;
++#ifndef CONFIG_PREEMPT_RT_FULL
+ int i;
- do {
-- page = list_entry(list->prev, struct page, lru);
-- /* must delete as __free_one_page list manipulates */
-+ page = list_last_entry(list, struct page, lru);
- list_del(&page->lru);
-- /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
-- __free_one_page(page, zone, 0, page_private(page));
-- trace_mm_page_pcpu_drain(page, 0, page_private(page));
-+ list_add(&page->lru, dst);
- } while (--to_free && --batch_free && !list_empty(list));
+ if (btrfs_spin_on_block(eb)) {
+@@ -112,6 +113,7 @@ int btrfs_try_spin_lock(struct extent_buffer *eb)
+ return 1;
+ spin_unlock(&eb->lock);
}
-- __mod_zone_page_state(zone, NR_FREE_PAGES, count);
-- spin_unlock(&zone->lock);
++#endif
+ return 0;
}
- static void free_one_page(struct zone *zone, struct page *page, int order,
- int migratetype)
- {
-- spin_lock(&zone->lock);
-+ unsigned long flags;
-+
-+ spin_lock_irqsave(&zone->lock, flags);
- zone->all_unreclaimable = 0;
- zone->pages_scanned = 0;
-
- __free_one_page(page, zone, order, migratetype);
- __mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order);
-- spin_unlock(&zone->lock);
-+ spin_unlock_irqrestore(&zone->lock, flags);
- }
+diff --git a/fs/buffer.c b/fs/buffer.c
+index 1a80b04..85f7895 100644
+--- a/fs/buffer.c
++++ b/fs/buffer.c
+@@ -328,8 +328,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ * decide that the page is now completely done.
+ */
+ first = page_buffers(page);
+- local_irq_save(flags);
+- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
++ flags = bh_uptodate_lock_irqsave(first);
+ clear_buffer_async_read(bh);
+ unlock_buffer(bh);
+ tmp = bh;
+@@ -342,8 +341,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ }
+ tmp = tmp->b_this_page;
+ } while (tmp != bh);
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+- local_irq_restore(flags);
++ bh_uptodate_unlock_irqrestore(first, flags);
- static bool free_pages_prepare(struct page *page, unsigned int order)
-@@ -682,13 +718,13 @@ static void __free_pages_ok(struct page
- if (!free_pages_prepare(page, order))
- return;
+ /*
+ * If none of the buffers had errors and they are all
+@@ -355,9 +353,7 @@ static void end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ return;
-- local_irq_save(flags);
-+ local_lock_irqsave(pa_lock, flags);
- if (unlikely(wasMlocked))
- free_page_mlock(page);
- __count_vm_events(PGFREE, 1 << order);
- free_one_page(page_zone(page), page, order,
- get_pageblock_migratetype(page));
+ still_busy:
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
- local_irq_restore(flags);
-+ local_unlock_irqrestore(pa_lock, flags);
+- return;
++ bh_uptodate_unlock_irqrestore(first, flags);
}
/*
-@@ -1064,16 +1100,18 @@ static int rmqueue_bulk(struct zone *zon
- void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp)
- {
- unsigned long flags;
-+ LIST_HEAD(dst);
- int to_drain;
+@@ -391,8 +387,7 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate)
+ }
+ first = page_buffers(page);
- local_irq_save(flags);
-+ local_lock_irqsave(pa_lock, flags);
- if (pcp->count >= pcp->batch)
- to_drain = pcp->batch;
- else
- to_drain = pcp->count;
-- free_pcppages_bulk(zone, to_drain, pcp);
-+ isolate_pcp_pages(to_drain, pcp, &dst);
- pcp->count -= to_drain;
+- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
++ flags = bh_uptodate_lock_irqsave(first);
+
+ clear_buffer_async_write(bh);
+ unlock_buffer(bh);
+@@ -404,15 +399,12 @@ void end_buffer_async_write(struct buffer_head *bh, int uptodate)
+ }
+ tmp = tmp->b_this_page;
+ }
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
- local_irq_restore(flags);
-+ local_unlock_irqrestore(pa_lock, flags);
-+ free_pcppages_bulk(zone, to_drain, &dst);
++ bh_uptodate_unlock_irqrestore(first, flags);
+ end_page_writeback(page);
+ return;
+
+ still_busy:
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+- local_irq_restore(flags);
+- return;
++ bh_uptodate_unlock_irqrestore(first, flags);
}
- #endif
+ EXPORT_SYMBOL(end_buffer_async_write);
-@@ -1092,16 +1130,21 @@ static void drain_pages(unsigned int cpu
- for_each_populated_zone(zone) {
- struct per_cpu_pageset *pset;
- struct per_cpu_pages *pcp;
-+ LIST_HEAD(dst);
-+ int count;
+@@ -3220,6 +3212,7 @@ struct buffer_head *alloc_buffer_head(gfp_t gfp_flags)
+ struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags);
+ if (ret) {
+ INIT_LIST_HEAD(&ret->b_assoc_buffers);
++ buffer_head_init_locks(ret);
+ preempt_disable();
+ __this_cpu_inc(bh_accounting.nr);
+ recalc_bh_state();
+diff --git a/fs/direct-io.c b/fs/direct-io.c
+index ac5f164..91ea211 100644
+--- a/fs/direct-io.c
++++ b/fs/direct-io.c
+@@ -255,7 +255,7 @@ static ssize_t dio_complete(struct dio *dio, loff_t offset, ssize_t ret, bool is
-- local_irq_save(flags);
-+ cpu_lock_irqsave(cpu, flags);
- pset = per_cpu_ptr(zone->pageset, cpu);
+ if (dio->flags & DIO_LOCKING)
+ /* lockdep: non-owner release */
+- up_read_non_owner(&dio->inode->i_alloc_sem);
++ anon_up_read_non_owner(&dio->inode->i_alloc_sem);
- pcp = &pset->pcp;
-- if (pcp->count) {
-- free_pcppages_bulk(zone, pcp->count, pcp);
-+ count = pcp->count;
-+ if (count) {
-+ isolate_pcp_pages(count, pcp, &dst);
- pcp->count = 0;
- }
-- local_irq_restore(flags);
-+ cpu_unlock_irqrestore(cpu, flags);
-+ if (count)
-+ free_pcppages_bulk(zone, count, &dst);
- }
+ return ret;
}
+@@ -1234,7 +1234,7 @@ __blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode,
+ * Will be released at I/O completion, possibly in a
+ * different thread.
+ */
+- down_read_non_owner(&inode->i_alloc_sem);
++ anon_down_read_non_owner(&inode->i_alloc_sem);
+ }
-@@ -1118,7 +1161,14 @@ void drain_local_pages(void *arg)
+ /*
+diff --git a/fs/eventpoll.c b/fs/eventpoll.c
+index 2acaf60..0652d4f 100644
+--- a/fs/eventpoll.c
++++ b/fs/eventpoll.c
+@@ -438,12 +438,12 @@ static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests)
*/
- void drain_all_pages(void)
+ static void ep_poll_safewake(wait_queue_head_t *wq)
{
-+#ifndef CONFIG_PREEMPT_RT_BASE
- on_each_cpu(drain_local_pages, NULL, 1);
-+#else
-+ int i;
-+
-+ for_each_online_cpu(i)
-+ drain_pages(i);
-+#endif
- }
-
- #ifdef CONFIG_HIBERNATION
-@@ -1174,7 +1224,7 @@ void free_hot_cold_page(struct page *pag
+- int this_cpu = get_cpu();
++ int this_cpu = get_cpu_light();
- migratetype = get_pageblock_migratetype(page);
- set_page_private(page, migratetype);
-- local_irq_save(flags);
-+ local_lock_irqsave(pa_lock, flags);
- if (unlikely(wasMlocked))
- free_page_mlock(page);
- __count_vm_event(PGFREE);
-@@ -1201,12 +1251,19 @@ void free_hot_cold_page(struct page *pag
- list_add(&page->lru, &pcp->lists[migratetype]);
- pcp->count++;
- if (pcp->count >= pcp->high) {
-- free_pcppages_bulk(zone, pcp->batch, pcp);
-+ LIST_HEAD(dst);
-+ int count;
-+
-+ isolate_pcp_pages(pcp->batch, pcp, &dst);
- pcp->count -= pcp->batch;
-+ count = pcp->batch;
-+ local_unlock_irqrestore(pa_lock, flags);
-+ free_pcppages_bulk(zone, count, &dst);
-+ return;
- }
+ ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS,
+ ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu);
- out:
-- local_irq_restore(flags);
-+ local_unlock_irqrestore(pa_lock, flags);
+- put_cpu();
++ put_cpu_light();
}
/*
-@@ -1301,7 +1358,7 @@ again:
- struct per_cpu_pages *pcp;
- struct list_head *list;
-
-- local_irq_save(flags);
-+ local_lock_irqsave(pa_lock, flags);
- pcp = &this_cpu_ptr(zone->pageset)->pcp;
- list = &pcp->lists[migratetype];
- if (list_empty(list)) {
-@@ -1333,17 +1390,19 @@ again:
- */
- WARN_ON_ONCE(order > 1);
+diff --git a/fs/exec.c b/fs/exec.c
+index 044c13f..709557e 100644
+--- a/fs/exec.c
++++ b/fs/exec.c
+@@ -844,6 +844,7 @@ static int exec_mmap(struct mm_struct *mm)
}
-- spin_lock_irqsave(&zone->lock, flags);
-+ local_spin_lock_irqsave(pa_lock, &zone->lock, flags);
- page = __rmqueue(zone, order, migratetype);
-- spin_unlock(&zone->lock);
-- if (!page)
-+ if (!page) {
-+ spin_unlock(&zone->lock);
- goto failed;
-+ }
- __mod_zone_page_state(zone, NR_FREE_PAGES, -(1 << order));
-+ spin_unlock(&zone->lock);
+ }
+ task_lock(tsk);
++ local_irq_disable_rt();
+ active_mm = tsk->active_mm;
+ tsk->mm = mm;
+ tsk->active_mm = mm;
+@@ -852,6 +853,7 @@ static int exec_mmap(struct mm_struct *mm)
+ atomic_dec(&old_mm->oom_disable_count);
+ atomic_inc(&tsk->mm->oom_disable_count);
+ }
++ local_irq_enable_rt();
+ task_unlock(tsk);
+ arch_pick_mmap_layout(mm);
+ if (old_mm) {
+diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
+index 1265904..ddf500d 100644
+--- a/fs/ext4/inode.c
++++ b/fs/ext4/inode.c
+@@ -5872,7 +5872,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
+ * Get i_alloc_sem to stop truncates messing with the inode. We cannot
+ * get i_mutex because we are already holding mmap_sem.
+ */
+- down_read(&inode->i_alloc_sem);
++ anon_down_read(&inode->i_alloc_sem);
+ size = i_size_read(inode);
+ if (page->mapping != mapping || size <= page_offset(page)
+ || !PageUptodate(page)) {
+@@ -5884,7 +5884,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
+ lock_page(page);
+ wait_on_page_writeback(page);
+ if (PageMappedToDisk(page)) {
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ return VM_FAULT_LOCKED;
}
- __count_zone_vm_events(PGALLOC, zone, 1 << order);
- zone_statistics(preferred_zone, zone, gfp_flags);
-- local_irq_restore(flags);
-+ local_unlock_irqrestore(pa_lock, flags);
-
- VM_BUG_ON(bad_range(zone, page));
- if (prep_new_page(page, order, gfp_flags))
-@@ -1351,7 +1410,7 @@ again:
- return page;
-
- failed:
-- local_irq_restore(flags);
-+ local_unlock_irqrestore(pa_lock, flags);
- return NULL;
+@@ -5902,7 +5902,7 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
+ if (page_has_buffers(page)) {
+ if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,
+ ext4_bh_unmapped)) {
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ return VM_FAULT_LOCKED;
+ }
+ }
+@@ -5931,11 +5931,11 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
+ */
+ lock_page(page);
+ wait_on_page_writeback(page);
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ return VM_FAULT_LOCKED;
+ out_unlock:
+ if (ret)
+ ret = VM_FAULT_SIGBUS;
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ return ret;
}
+diff --git a/fs/fat/inode.c b/fs/fat/inode.c
+index cb8d839..44ab357 100644
+--- a/fs/fat/inode.c
++++ b/fs/fat/inode.c
+@@ -224,9 +224,9 @@ static sector_t _fat_bmap(struct address_space *mapping, sector_t block)
+ sector_t blocknr;
-@@ -1912,8 +1971,8 @@ __alloc_pages_direct_compact(gfp_t gfp_m
- if (*did_some_progress != COMPACT_SKIPPED) {
+ /* fat_get_cluster() assumes the requested blocknr isn't truncated. */
+- down_read(&mapping->host->i_alloc_sem);
++ anon_down_read(&mapping->host->i_alloc_sem);
+ blocknr = generic_block_bmap(mapping, block, fat_get_block);
+- up_read(&mapping->host->i_alloc_sem);
++ anon_up_read(&mapping->host->i_alloc_sem);
- /* Page migration frees to the PCP lists but we want merging */
-- drain_pages(get_cpu());
-- put_cpu();
-+ drain_pages(get_cpu_light());
+ return blocknr;
+ }
+diff --git a/fs/file.c b/fs/file.c
+index 4c6992d..fd03258 100644
+--- a/fs/file.c
++++ b/fs/file.c
+@@ -105,14 +105,14 @@ void free_fdtable_rcu(struct rcu_head *rcu)
+ kfree(fdt->open_fds);
+ kfree(fdt);
+ } else {
+- fddef = &get_cpu_var(fdtable_defer_list);
++ fddef = &per_cpu(fdtable_defer_list, get_cpu_light());
+ spin_lock(&fddef->lock);
+ fdt->next = fddef->next;
+ fddef->next = fdt;
+ /* vmallocs are handled from the workqueue context */
+ schedule_work(&fddef->wq);
+ spin_unlock(&fddef->lock);
+- put_cpu_var(fdtable_defer_list);
+ put_cpu_light();
+ }
+ }
- page = get_page_from_freelist(gfp_mask, nodemask,
- order, zonelist, high_zoneidx,
-@@ -3685,14 +3744,16 @@ static int __zone_pcp_update(void *data)
- for_each_possible_cpu(cpu) {
- struct per_cpu_pageset *pset;
- struct per_cpu_pages *pcp;
-+ LIST_HEAD(dst);
+@@ -422,7 +422,7 @@ struct files_struct init_files = {
+ .close_on_exec = (fd_set *)&init_files.close_on_exec_init,
+ .open_fds = (fd_set *)&init_files.open_fds_init,
+ },
+- .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock),
++ .file_lock = __SPIN_LOCK_UNLOCKED(init_files.file_lock),
+ };
- pset = per_cpu_ptr(zone->pageset, cpu);
- pcp = &pset->pcp;
+ /*
+diff --git a/fs/inode.c b/fs/inode.c
+index 43566d1..731debd 100644
+--- a/fs/inode.c
++++ b/fs/inode.c
+@@ -176,7 +176,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
+ mutex_init(&inode->i_mutex);
+ lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key);
+
+- init_rwsem(&inode->i_alloc_sem);
++ init_anon_rwsem(&inode->i_alloc_sem);
+ lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key);
+ mapping->a_ops = &empty_aops;
+diff --git a/fs/namespace.c b/fs/namespace.c
+index 537dd96..480df48 100644
+--- a/fs/namespace.c
++++ b/fs/namespace.c
+@@ -341,8 +341,14 @@ int mnt_want_write(struct vfsmount *mnt)
+ * incremented count after it has set MNT_WRITE_HOLD.
+ */
+ smp_mb();
+- while (mnt->mnt_flags & MNT_WRITE_HOLD)
++ /*
++ * No need to keep preemption disabled accross the spin loop.
++ */
++ while (mnt->mnt_flags & MNT_WRITE_HOLD) {
++ preempt_enable();
+ cpu_relax();
++ preempt_disable();
++ }
+ /*
+ * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will
+ * be set to match its requirements. So we must not load that until
+@@ -352,9 +358,7 @@ int mnt_want_write(struct vfsmount *mnt)
+ if (__mnt_is_readonly(mnt)) {
+ mnt_dec_writers(mnt);
+ ret = -EROFS;
+- goto out;
+ }
+-out:
+ preempt_enable();
+ return ret;
+ }
+diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c
+index 0b1e885b..4c8095c 100644
+--- a/fs/ntfs/aops.c
++++ b/fs/ntfs/aops.c
+@@ -108,8 +108,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ "0x%llx.", (unsigned long long)bh->b_blocknr);
+ }
+ first = page_buffers(page);
+- local_irq_save(flags);
+- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
++ flags = bh_uptodate_lock_irqsave(first);
+ clear_buffer_async_read(bh);
+ unlock_buffer(bh);
+ tmp = bh;
+@@ -124,8 +123,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ }
+ tmp = tmp->b_this_page;
+ } while (tmp != bh);
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+- local_irq_restore(flags);
++ bh_uptodate_unlock_irqrestore(first, flags);
+ /*
+ * If none of the buffers had errors then we can set the page uptodate,
+ * but we first have to perform the post read mst fixups, if the
+@@ -146,13 +144,13 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ recs = PAGE_CACHE_SIZE / rec_size;
+ /* Should have been verified before we got here... */
+ BUG_ON(!recs);
- local_irq_save(flags);
-- free_pcppages_bulk(zone, pcp->count, pcp);
-+ cpu_lock_irqsave(cpu, flags);
-+ isolate_pcp_pages(pcp->count, pcp, &dst);
-+ free_pcppages_bulk(zone, pcp->count, &dst);
- setup_pageset(pset, batch);
++ local_irq_save_nort(flags);
+ kaddr = kmap_atomic(page, KM_BIO_SRC_IRQ);
+ for (i = 0; i < recs; i++)
+ post_read_mst_fixup((NTFS_RECORD*)(kaddr +
+ i * rec_size), rec_size);
+ kunmap_atomic(kaddr, KM_BIO_SRC_IRQ);
- local_irq_restore(flags);
-+ cpu_unlock_irqrestore(cpu, flags);
- }
- return 0;
- }
-@@ -5004,6 +5065,7 @@ static int page_alloc_cpu_notify(struct
- void __init page_alloc_init(void)
- {
- hotcpu_notifier(page_alloc_cpu_notify, 0);
-+ local_irq_lock_init(pa_lock);
++ local_irq_restore_nort(flags);
+ flush_dcache_page(page);
+ if (likely(page_uptodate && !PageError(page)))
+ SetPageUptodate(page);
+@@ -160,9 +158,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate)
+ unlock_page(page);
+ return;
+ still_busy:
+- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
+- local_irq_restore(flags);
+- return;
++ bh_uptodate_unlock_irqrestore(first, flags);
}
- /*
-Index: linux-3.0/include/linux/pid.h
-===================================================================
---- linux-3.0.orig/include/linux/pid.h
-+++ linux-3.0/include/linux/pid.h
-@@ -2,6 +2,7 @@
- #define _LINUX_PID_H
-
- #include <linux/rcupdate.h>
-+#include <linux/atomic.h>
-
- enum pid_type
- {
-Index: linux-3.0/net/core/sock.c
-===================================================================
---- linux-3.0.orig/net/core/sock.c
-+++ linux-3.0/net/core/sock.c
-@@ -2035,12 +2035,11 @@ void lock_sock_nested(struct sock *sk, i
- if (sk->sk_lock.owned)
- __lock_sock(sk);
- sk->sk_lock.owned = 1;
-- spin_unlock(&sk->sk_lock.slock);
-+ spin_unlock_bh(&sk->sk_lock.slock);
- /*
- * The sk_lock has mutex_lock() semantics here:
+ /**
+diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
+index f4b1057..fbd9093 100644
+--- a/fs/ntfs/file.c
++++ b/fs/ntfs/file.c
+@@ -1832,9 +1832,9 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb,
+ * fails again.
*/
- mutex_acquire(&sk->sk_lock.dep_map, subclass, 0, _RET_IP_);
-- local_bh_enable();
- }
- EXPORT_SYMBOL(lock_sock_nested);
-
-Index: linux-3.0/include/linux/interrupt.h
-===================================================================
---- linux-3.0.orig/include/linux/interrupt.h
-+++ linux-3.0/include/linux/interrupt.h
-@@ -205,7 +205,7 @@ extern void devm_free_irq(struct device
- #ifdef CONFIG_LOCKDEP
- # define local_irq_enable_in_hardirq() do { } while (0)
- #else
--# define local_irq_enable_in_hardirq() local_irq_enable()
-+# define local_irq_enable_in_hardirq() local_irq_enable_nort()
- #endif
-
- extern void disable_irq_nosync(unsigned int irq);
-@@ -380,9 +380,13 @@ static inline int disable_irq_wake(unsig
-
-
- #ifdef CONFIG_IRQ_FORCED_THREADING
--extern bool force_irqthreads;
-+# ifndef CONFIG_PREEMPT_RT_BASE
-+ extern bool force_irqthreads;
-+# else
-+# define force_irqthreads (true)
-+# endif
- #else
--#define force_irqthreads (0)
-+#define force_irqthreads (false)
- #endif
-
- #ifndef __ARCH_SET_SOFTIRQ_PENDING
-@@ -436,8 +440,14 @@ struct softirq_action
- void (*action)(struct softirq_action *);
- };
+ if (unlikely(NInoTruncateFailed(ni))) {
+- down_write(&vi->i_alloc_sem);
++ anon_down_write(&vi->i_alloc_sem);
+ err = ntfs_truncate(vi);
+- up_write(&vi->i_alloc_sem);
++ anon_up_write(&vi->i_alloc_sem);
+ if (err || NInoTruncateFailed(ni)) {
+ if (!err)
+ err = -EIO;
+diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
+index ac97bca..1c679a9 100644
+--- a/fs/ocfs2/aops.c
++++ b/fs/ocfs2/aops.c
+@@ -569,7 +569,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
+ BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
-+#ifndef CONFIG_PREEMPT_RT_FULL
- asmlinkage void do_softirq(void);
- asmlinkage void __do_softirq(void);
-+static inline void thread_do_softirq(void) { do_softirq(); }
-+#else
-+extern void thread_do_softirq(void);
-+#endif
-+
- extern void open_softirq(int nr, void (*action)(struct softirq_action *));
- extern void softirq_init(void);
- static inline void __raise_softirq_irqoff(unsigned int nr)
-@@ -449,6 +459,8 @@ static inline void __raise_softirq_irqof
- extern void raise_softirq_irqoff(unsigned int nr);
- extern void raise_softirq(unsigned int nr);
+ if (ocfs2_iocb_is_sem_locked(iocb)) {
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ ocfs2_iocb_clear_sem_locked(iocb);
+ }
-+extern void softirq_check_pending_idle(void);
-+
- /* This is the worklist that queues up per-cpu softirq work.
- *
- * send_remote_sendirq() adds work to these lists, and
-@@ -626,6 +638,12 @@ void tasklet_hrtimer_cancel(struct taskl
- tasklet_kill(&ttimer->tasklet);
- }
+diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
+index b1e35a3..c26eafe 100644
+--- a/fs/ocfs2/file.c
++++ b/fs/ocfs2/file.c
+@@ -2238,7 +2238,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
+ relock:
+ /* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
+ if (direct_io) {
+- down_read(&inode->i_alloc_sem);
++ anon_down_read(&inode->i_alloc_sem);
+ have_alloc_sem = 1;
+ /* communicate with ocfs2_dio_end_io */
+ ocfs2_iocb_set_sem_locked(iocb);
+@@ -2290,7 +2290,7 @@ relock:
+ */
+ if (direct_io && !can_do_direct) {
+ ocfs2_rw_unlock(inode, rw_level);
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+extern void softirq_early_init(void);
-+#else
-+static inline void softirq_early_init(void) { }
-+#endif
-+
- /*
- * Autoprobing for irqs:
- *
-Index: linux-3.0/net/core/dev.c
-===================================================================
---- linux-3.0.orig/net/core/dev.c
-+++ linux-3.0/net/core/dev.c
-@@ -213,14 +213,14 @@ static inline struct hlist_head *dev_ind
- static inline void rps_lock(struct softnet_data *sd)
- {
- #ifdef CONFIG_RPS
-- spin_lock(&sd->input_pkt_queue.lock);
-+ raw_spin_lock(&sd->input_pkt_queue.raw_lock);
- #endif
- }
+ have_alloc_sem = 0;
+ rw_level = -1;
+@@ -2379,7 +2379,7 @@ out:
- static inline void rps_unlock(struct softnet_data *sd)
- {
- #ifdef CONFIG_RPS
-- spin_unlock(&sd->input_pkt_queue.lock);
-+ raw_spin_unlock(&sd->input_pkt_queue.raw_lock);
- #endif
- }
+ out_sems:
+ if (have_alloc_sem) {
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ ocfs2_iocb_clear_sem_locked(iocb);
+ }
-@@ -2902,16 +2902,46 @@ int netif_rx_ni(struct sk_buff *skb)
- {
- int err;
+@@ -2531,7 +2531,7 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
+ * need locks to protect pending reads from racing with truncate.
+ */
+ if (filp->f_flags & O_DIRECT) {
+- down_read(&inode->i_alloc_sem);
++ anon_down_read(&inode->i_alloc_sem);
+ have_alloc_sem = 1;
+ ocfs2_iocb_set_sem_locked(iocb);
-- preempt_disable();
-+ migrate_disable();
- err = netif_rx(skb);
- if (local_softirq_pending())
-- do_softirq();
-- preempt_enable();
-+ thread_do_softirq();
-+ migrate_enable();
+@@ -2575,7 +2575,7 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
- return err;
- }
- EXPORT_SYMBOL(netif_rx_ni);
+ bail:
+ if (have_alloc_sem) {
+- up_read(&inode->i_alloc_sem);
++ anon_up_read(&inode->i_alloc_sem);
+ ocfs2_iocb_clear_sem_locked(iocb);
+ }
+ if (rw_level != -1)
+diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c
+index d780896..1446033 100644
+--- a/fs/reiserfs/xattr.c
++++ b/fs/reiserfs/xattr.c
+@@ -555,11 +555,11 @@ reiserfs_xattr_set_handle(struct reiserfs_transaction_handle *th,
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+/*
-+ * RT runs ksoftirqd as a real time thread and the root_lock is a
-+ * "sleeping spinlock". If the trylock fails then we can go into an
-+ * infinite loop when ksoftirqd preempted the task which actually
-+ * holds the lock, because we requeue q and raise NET_TX softirq
-+ * causing ksoftirqd to loop forever.
-+ *
-+ * It's safe to use spin_lock on RT here as softirqs run in thread
-+ * context and cannot deadlock against the thread which is holding
-+ * root_lock.
-+ *
-+ * On !RT the trylock might fail, but there we bail out from the
-+ * softirq loop after 10 attempts which we can't do on RT. And the
-+ * task holding root_lock cannot be preempted, so the only downside of
-+ * that trylock is that we need 10 loops to decide that we should have
-+ * given up in the first one :)
-+ */
-+static inline int take_root_lock(spinlock_t *lock)
-+{
-+ spin_lock(lock);
-+ return 1;
-+}
-+#else
-+static inline int take_root_lock(spinlock_t *lock)
-+{
-+ return spin_trylock(lock);
-+}
-+#endif
-+
- static void net_tx_action(struct softirq_action *h)
- {
- struct softnet_data *sd = &__get_cpu_var(softnet_data);
-@@ -2950,7 +2980,7 @@ static void net_tx_action(struct softirq
- head = head->next_sched;
+ reiserfs_write_unlock(inode->i_sb);
+ mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_XATTR);
+- down_write(&dentry->d_inode->i_alloc_sem);
++ anon_down_write(&dentry->d_inode->i_alloc_sem);
+ reiserfs_write_lock(inode->i_sb);
- root_lock = qdisc_lock(q);
-- if (spin_trylock(root_lock)) {
-+ if (take_root_lock(root_lock)) {
- smp_mb__before_clear_bit();
- clear_bit(__QDISC_STATE_SCHED,
- &q->state);
-@@ -3276,7 +3306,7 @@ static void flush_backlog(void *arg)
- skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) {
- if (skb->dev == dev) {
- __skb_unlink(skb, &sd->input_pkt_queue);
-- kfree_skb(skb);
-+ __skb_queue_tail(&sd->tofree_queue, skb);
- input_queue_head_incr(sd);
- }
- }
-@@ -3285,10 +3315,13 @@ static void flush_backlog(void *arg)
- skb_queue_walk_safe(&sd->process_queue, skb, tmp) {
- if (skb->dev == dev) {
- __skb_unlink(skb, &sd->process_queue);
-- kfree_skb(skb);
-+ __skb_queue_tail(&sd->tofree_queue, skb);
- input_queue_head_incr(sd);
- }
- }
-+
-+ if (!skb_queue_empty(&sd->tofree_queue))
-+ raise_softirq_irqoff(NET_RX_SOFTIRQ);
- }
+ err = reiserfs_setattr(dentry, &newattrs);
+- up_write(&dentry->d_inode->i_alloc_sem);
++ anon_up_write(&dentry->d_inode->i_alloc_sem);
+ mutex_unlock(&dentry->d_inode->i_mutex);
+ } else
+ update_ctime(inode);
+diff --git a/fs/xfs/linux-2.6/mrlock.h b/fs/xfs/linux-2.6/mrlock.h
+index ff6a198..c2f7017 100644
+--- a/fs/xfs/linux-2.6/mrlock.h
++++ b/fs/xfs/linux-2.6/mrlock.h
+@@ -21,18 +21,18 @@
+ #include <linux/rwsem.h>
- static int napi_gro_complete(struct sk_buff *skb)
-@@ -3766,10 +3799,17 @@ static void net_rx_action(struct softirq
- struct softnet_data *sd = &__get_cpu_var(softnet_data);
- unsigned long time_limit = jiffies + 2;
- int budget = netdev_budget;
-+ struct sk_buff *skb;
- void *have;
+ typedef struct {
+- struct rw_semaphore mr_lock;
++ struct rw_anon_semaphore mr_lock;
+ #ifdef DEBUG
+- int mr_writer;
++ int mr_writer;
+ #endif
+ } mrlock_t;
- local_irq_disable();
+ #ifdef DEBUG
+ #define mrinit(mrp, name) \
+- do { (mrp)->mr_writer = 0; init_rwsem(&(mrp)->mr_lock); } while (0)
++ do { (mrp)->mr_writer = 0; init_anon_rwsem(&(mrp)->mr_lock); } while (0)
+ #else
+ #define mrinit(mrp, name) \
+- do { init_rwsem(&(mrp)->mr_lock); } while (0)
++ do { init_anon_rwsem(&(mrp)->mr_lock); } while (0)
+ #endif
-+ while ((skb = __skb_dequeue(&sd->tofree_queue))) {
-+ local_irq_enable();
-+ kfree_skb(skb);
-+ local_irq_disable();
-+ }
-+
- while (!list_empty(&sd->poll_list)) {
- struct napi_struct *n;
- int work, weight;
-@@ -6197,6 +6237,9 @@ static int dev_cpu_callback(struct notif
- netif_rx(skb);
- input_queue_head_incr(oldsd);
- }
-+ while ((skb = __skb_dequeue(&oldsd->tofree_queue))) {
-+ kfree_skb(skb);
-+ }
+ #define mrlock_init(mrp, t,n,s) mrinit(mrp, n)
+@@ -40,12 +40,12 @@ typedef struct {
- return NOTIFY_OK;
+ static inline void mraccess_nested(mrlock_t *mrp, int subclass)
+ {
+- down_read_nested(&mrp->mr_lock, subclass);
++ anon_down_read_nested(&mrp->mr_lock, subclass);
}
-@@ -6462,8 +6505,9 @@ static int __init net_dev_init(void)
- struct softnet_data *sd = &per_cpu(softnet_data, i);
-
- memset(sd, 0, sizeof(*sd));
-- skb_queue_head_init(&sd->input_pkt_queue);
-- skb_queue_head_init(&sd->process_queue);
-+ skb_queue_head_init_raw(&sd->input_pkt_queue);
-+ skb_queue_head_init_raw(&sd->process_queue);
-+ skb_queue_head_init_raw(&sd->tofree_queue);
- sd->completion_queue = NULL;
- INIT_LIST_HEAD(&sd->poll_list);
- sd->output_queue = NULL;
-Index: linux-3.0/arch/x86/kernel/entry_32.S
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/entry_32.S
-+++ linux-3.0/arch/x86/kernel/entry_32.S
-@@ -625,7 +625,11 @@ work_notifysig: # deal with pending s
- jne work_notifysig_v86 # returning to kernel-space or
- # vm86-space
- xorl %edx, %edx
-+ TRACE_IRQS_ON
-+ ENABLE_INTERRUPTS(CLBR_NONE)
- call do_notify_resume
-+ DISABLE_INTERRUPTS(CLBR_ANY)
-+ TRACE_IRQS_OFF
- jmp resume_userspace_sig
- ALIGN
-@@ -638,7 +642,11 @@ work_notifysig_v86:
- movl %esp, %eax
+ static inline void mrupdate_nested(mrlock_t *mrp, int subclass)
+ {
+- down_write_nested(&mrp->mr_lock, subclass);
++ anon_down_write_nested(&mrp->mr_lock, subclass);
+ #ifdef DEBUG
+ mrp->mr_writer = 1;
#endif
- xorl %edx, %edx
-+ TRACE_IRQS_ON
-+ ENABLE_INTERRUPTS(CLBR_NONE)
- call do_notify_resume
-+ DISABLE_INTERRUPTS(CLBR_ANY)
-+ TRACE_IRQS_OFF
- jmp resume_userspace_sig
- END(work_pending)
+@@ -53,12 +53,12 @@ static inline void mrupdate_nested(mrlock_t *mrp, int subclass)
-Index: linux-3.0/kernel/rcutree.c
-===================================================================
---- linux-3.0.orig/kernel/rcutree.c
-+++ linux-3.0/kernel/rcutree.c
-@@ -166,6 +166,12 @@ void rcu_sched_qs(int cpu)
- rdp->passed_quiesc = 1;
+ static inline int mrtryaccess(mrlock_t *mrp)
+ {
+- return down_read_trylock(&mrp->mr_lock);
++ return anon_down_read_trylock(&mrp->mr_lock);
}
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+void rcu_bh_qs(int cpu)
-+{
-+ rcu_preempt_qs(cpu);
-+}
-+#else
- void rcu_bh_qs(int cpu)
+ static inline int mrtryupdate(mrlock_t *mrp)
{
- struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu);
-@@ -174,6 +180,7 @@ void rcu_bh_qs(int cpu)
- barrier();
- rdp->passed_quiesc = 1;
+- if (!down_write_trylock(&mrp->mr_lock))
++ if (!anon_down_write_trylock(&mrp->mr_lock))
+ return 0;
+ #ifdef DEBUG
+ mrp->mr_writer = 1;
+@@ -71,12 +71,12 @@ static inline void mrunlock_excl(mrlock_t *mrp)
+ #ifdef DEBUG
+ mrp->mr_writer = 0;
+ #endif
+- up_write(&mrp->mr_lock);
++ anon_up_write(&mrp->mr_lock);
}
-+#endif
- /*
- * Note a context switch. This is a quiescent state for RCU-sched,
-@@ -216,6 +223,7 @@ long rcu_batches_completed_sched(void)
+ static inline void mrunlock_shared(mrlock_t *mrp)
+ {
+- up_read(&mrp->mr_lock);
++ anon_up_read(&mrp->mr_lock);
}
- EXPORT_SYMBOL_GPL(rcu_batches_completed_sched);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /*
- * Return the number of RCU BH batches processed thus far for debug & stats.
- */
-@@ -233,6 +241,7 @@ void rcu_bh_force_quiescent_state(void)
- force_quiescent_state(&rcu_bh_state, 0);
+ static inline void mrdemote(mrlock_t *mrp)
+@@ -84,7 +84,7 @@ static inline void mrdemote(mrlock_t *mrp)
+ #ifdef DEBUG
+ mrp->mr_writer = 0;
+ #endif
+- downgrade_write(&mrp->mr_lock);
++ anon_downgrade_write(&mrp->mr_lock);
}
- EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state);
-+#endif
- /*
- * Record the number of times rcutorture tests have been initiated and
-@@ -1153,7 +1162,7 @@ static void __rcu_offline_cpu(int cpu, s
- else
- raw_spin_unlock_irqrestore(&rnp->lock, flags);
- if (need_report & RCU_OFL_TASKS_EXP_GP)
-- rcu_report_exp_rnp(rsp, rnp);
-+ rcu_report_exp_rnp(rsp, rnp, true);
- rcu_node_kthread_setaffinity(rnp, -1);
- }
+ #endif /* __XFS_SUPPORT_MRLOCK_H__ */
+diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
+index 347cae9..dd45aae 100644
+--- a/fs/xfs/linux-2.6/xfs_super.c
++++ b/fs/xfs/linux-2.6/xfs_super.c
+@@ -1002,7 +1002,7 @@ xfs_fs_evict_inode(
+ * (and basically indicate what we are doing), we explicitly
+ * re-init the iolock here.
+ */
+- ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
++ ASSERT(!anon_rwsem_is_locked(&ip->i_iolock.mr_lock));
+ mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
+ lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
+ &xfs_iolock_reclaimable, "xfs_iolock_reclaimable");
+diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
+index 3631783..55d3702 100644
+--- a/fs/xfs/xfs_iget.c
++++ b/fs/xfs/xfs_iget.c
+@@ -272,7 +272,7 @@ xfs_iget_cache_hit(
+ __xfs_inode_clear_reclaim_tag(mp, pag, ip);
+ inode->i_state = I_NEW;
-@@ -1579,6 +1588,7 @@ void call_rcu_sched(struct rcu_head *hea
- }
- EXPORT_SYMBOL_GPL(call_rcu_sched);
+- ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
++ ASSERT(!anon_rwsem_is_locked(&ip->i_iolock.mr_lock));
+ mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
+ lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
+ &xfs_iolock_active, "xfs_iolock_active");
+@@ -706,13 +706,13 @@ xfs_isilocked(
+ if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) {
+ if (!(lock_flags & XFS_ILOCK_SHARED))
+ return !!ip->i_lock.mr_writer;
+- return rwsem_is_locked(&ip->i_lock.mr_lock);
++ return anon_rwsem_is_locked(&ip->i_lock.mr_lock);
+ }
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /*
- * Queue an RCU for invocation after a quicker grace period.
- */
-@@ -1587,6 +1597,7 @@ void call_rcu_bh(struct rcu_head *head,
- __call_rcu(head, func, &rcu_bh_state);
- }
- EXPORT_SYMBOL_GPL(call_rcu_bh);
+ if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) {
+ if (!(lock_flags & XFS_IOLOCK_SHARED))
+ return !!ip->i_iolock.mr_writer;
+- return rwsem_is_locked(&ip->i_iolock.mr_lock);
++ return anon_rwsem_is_locked(&ip->i_iolock.mr_lock);
+ }
+
+ ASSERT(0);
+diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h
+index dfb0ec6..f65690a 100644
+--- a/include/asm-generic/bug.h
++++ b/include/asm-generic/bug.h
+@@ -3,6 +3,10 @@
+
+ #include <linux/compiler.h>
+
++#ifndef __ASSEMBLY__
++extern void __WARN_ON(const char *func, const char *file, const int line);
++#endif /* __ASSEMBLY__ */
++
+ #ifdef CONFIG_BUG
+
+ #ifdef CONFIG_GENERIC_BUG
+@@ -201,4 +205,18 @@ extern void warn_slowpath_null(const char *file, const int line);
+ # define WARN_ON_SMP(x) ({0;})
+ #endif
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++# define BUG_ON_RT(c) BUG_ON(c)
++# define BUG_ON_NONRT(c) do { } while (0)
++# define WARN_ON_RT(condition) WARN_ON(condition)
++# define WARN_ON_NONRT(condition) do { } while (0)
++# define WARN_ON_ONCE_NONRT(condition) do { } while (0)
++#else
++# define BUG_ON_RT(c) do { } while (0)
++# define BUG_ON_NONRT(c) BUG_ON(c)
++# define WARN_ON_RT(condition) do { } while (0)
++# define WARN_ON_NONRT(condition) WARN_ON(condition)
++# define WARN_ON_ONCE_NONRT(condition) WARN_ON_ONCE(condition)
+#endif
++
+ #endif
+diff --git a/include/asm-generic/cmpxchg-local.h b/include/asm-generic/cmpxchg-local.h
+index 2533fdd..d8d4c89 100644
+--- a/include/asm-generic/cmpxchg-local.h
++++ b/include/asm-generic/cmpxchg-local.h
+@@ -21,7 +21,7 @@ static inline unsigned long __cmpxchg_local_generic(volatile void *ptr,
+ if (size == 8 && sizeof(unsigned long) != 8)
+ wrong_size_cmpxchg(ptr);
- /**
- * synchronize_sched - wait until an rcu-sched grace period has elapsed.
-@@ -1628,6 +1639,7 @@ void synchronize_sched(void)
+- local_irq_save(flags);
++ raw_local_irq_save(flags);
+ switch (size) {
+ case 1: prev = *(u8 *)ptr;
+ if (prev == old)
+@@ -42,7 +42,7 @@ static inline unsigned long __cmpxchg_local_generic(volatile void *ptr,
+ default:
+ wrong_size_cmpxchg(ptr);
+ }
+- local_irq_restore(flags);
++ raw_local_irq_restore(flags);
+ return prev;
}
- EXPORT_SYMBOL_GPL(synchronize_sched);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /**
- * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed.
- *
-@@ -1653,6 +1665,7 @@ void synchronize_rcu_bh(void)
- destroy_rcu_head_on_stack(&rcu.head);
+@@ -55,11 +55,11 @@ static inline u64 __cmpxchg64_local_generic(volatile void *ptr,
+ u64 prev;
+ unsigned long flags;
+
+- local_irq_save(flags);
++ raw_local_irq_save(flags);
+ prev = *(u64 *)ptr;
+ if (prev == old)
+ *(u64 *)ptr = new;
+- local_irq_restore(flags);
++ raw_local_irq_restore(flags);
+ return prev;
}
- EXPORT_SYMBOL_GPL(synchronize_rcu_bh);
+
+diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
+index 503c8a6..dbaeaec 100644
+--- a/include/linux/buffer_head.h
++++ b/include/linux/buffer_head.h
+@@ -72,8 +72,52 @@ struct buffer_head {
+ struct address_space *b_assoc_map; /* mapping this buffer is
+ associated with */
+ atomic_t b_count; /* users using this buffer_head */
++#ifdef CONFIG_PREEMPT_RT_BASE
++ spinlock_t b_uptodate_lock;
++#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \
++ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE)
++ spinlock_t b_state_lock;
++ spinlock_t b_journal_head_lock;
+#endif
++#endif
+ };
++static inline unsigned long bh_uptodate_lock_irqsave(struct buffer_head *bh)
++{
++ unsigned long flags;
++
++#ifndef CONFIG_PREEMPT_RT_BASE
++ local_irq_save(flags);
++ bit_spin_lock(BH_Uptodate_Lock, &bh->b_state);
++#else
++ spin_lock_irqsave(&bh->b_uptodate_lock, flags);
++#endif
++ return flags;
++}
++
++static inline void
++bh_uptodate_unlock_irqrestore(struct buffer_head *bh, unsigned long flags)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ bit_spin_unlock(BH_Uptodate_Lock, &bh->b_state);
++ local_irq_restore(flags);
++#else
++ spin_unlock_irqrestore(&bh->b_uptodate_lock, flags);
++#endif
++}
++
++static inline void buffer_head_init_locks(struct buffer_head *bh)
++{
++#ifdef CONFIG_PREEMPT_RT_BASE
++ spin_lock_init(&bh->b_uptodate_lock);
++#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \
++ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE)
++ spin_lock_init(&bh->b_state_lock);
++ spin_lock_init(&bh->b_journal_head_lock);
++#endif
++#endif
++}
++
/*
- * Check to see if there is any immediate RCU-related work to be done
-@@ -1806,6 +1819,7 @@ static void _rcu_barrier(struct rcu_stat
- mutex_unlock(&rcu_barrier_mutex);
- }
+ * macro tricks to expand the set_buffer_foo(), clear_buffer_foo()
+ * and buffer_foo() functions.
+diff --git a/include/linux/console.h b/include/linux/console.h
+index 7453cfd..e5b5dc0 100644
+--- a/include/linux/console.h
++++ b/include/linux/console.h
+@@ -133,6 +133,7 @@ struct console {
+ for (con = console_drivers; con != NULL; con = con->next)
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /**
- * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete.
- */
-@@ -1814,6 +1828,7 @@ void rcu_barrier_bh(void)
- _rcu_barrier(&rcu_bh_state, call_rcu_bh);
- }
- EXPORT_SYMBOL_GPL(rcu_barrier_bh);
-+#endif
+ extern int console_set_on_cmdline;
++extern struct console *early_console;
- /**
- * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks.
-Index: linux-3.0/kernel/rcutree.h
-===================================================================
---- linux-3.0.orig/kernel/rcutree.h
-+++ linux-3.0/kernel/rcutree.h
-@@ -422,6 +422,7 @@ DECLARE_PER_CPU(struct rcu_data, rcu_pre
- /* Forward declarations for rcutree_plugin.h */
- static void rcu_bootup_announce(void);
- long rcu_batches_completed(void);
-+static void rcu_preempt_qs(int cpu);
- static void rcu_preempt_note_context_switch(int cpu);
- static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp);
- #ifdef CONFIG_HOTPLUG_CPU
-@@ -443,7 +444,8 @@ static void rcu_preempt_check_callbacks(
- static void rcu_preempt_process_callbacks(void);
- void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
- #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU)
--static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp);
-+static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
-+ bool wake);
- #endif /* #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU) */
- static int rcu_preempt_pending(int cpu);
- static int rcu_preempt_needs_cpu(int cpu);
-Index: linux-3.0/kernel/rcutree_plugin.h
-===================================================================
---- linux-3.0.orig/kernel/rcutree_plugin.h
-+++ linux-3.0/kernel/rcutree_plugin.h
-@@ -299,6 +299,9 @@ static noinline void rcu_read_unlock_spe
- int empty_exp;
- unsigned long flags;
- struct list_head *np;
-+#ifdef CONFIG_RCU_BOOST
-+ struct rt_mutex *rbmp = NULL;
-+#endif /* #ifdef CONFIG_RCU_BOOST */
- struct rcu_node *rnp;
- int special;
+ extern int add_preferred_console(char *name, int idx, char *options);
+ extern int update_console_cmdline(char *name, int idx, char *name_new, int idx_new, char *options);
+diff --git a/include/linux/cpu.h b/include/linux/cpu.h
+index 5f09323..d7d7a12 100644
+--- a/include/linux/cpu.h
++++ b/include/linux/cpu.h
+@@ -60,14 +60,16 @@ enum {
+ */
+ CPU_PRI_SCHED_ACTIVE = INT_MAX,
+ CPU_PRI_CPUSET_ACTIVE = INT_MAX - 1,
+- CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
+- CPU_PRI_CPUSET_INACTIVE = INT_MIN,
-@@ -318,7 +321,7 @@ static noinline void rcu_read_unlock_spe
- }
+ /* migration should happen before other stuff but after perf */
+- CPU_PRI_PERF = 20,
+- CPU_PRI_MIGRATION = 10,
+- /* prepare workqueues for other notifiers */
+- CPU_PRI_WORKQUEUE = 5,
++ CPU_PRI_PERF = 20,
++ CPU_PRI_MIGRATION = 10,
++ CPU_PRI_WORKQUEUE_ACTIVE = 5, /* prepare workqueues for others */
++ CPU_PRI_NORMAL = 0,
++ CPU_PRI_WORKQUEUE_INACTIVE = -5, /* flush workqueues after others */
++
++ CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
++ CPU_PRI_CPUSET_INACTIVE = INT_MIN,
+ };
- /* Hardware IRQ handlers cannot block. */
-- if (in_irq() || in_serving_softirq()) {
-+ if (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET)) {
- local_irq_restore(flags);
- return;
- }
-@@ -344,6 +347,7 @@ static noinline void rcu_read_unlock_spe
- smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
- np = rcu_next_node_entry(t, rnp);
- list_del_init(&t->rcu_node_entry);
-+ t->rcu_blocked_node = NULL;
- if (&t->rcu_node_entry == rnp->gp_tasks)
- rnp->gp_tasks = np;
- if (&t->rcu_node_entry == rnp->exp_tasks)
-@@ -351,13 +355,12 @@ static noinline void rcu_read_unlock_spe
- #ifdef CONFIG_RCU_BOOST
- if (&t->rcu_node_entry == rnp->boost_tasks)
- rnp->boost_tasks = np;
-- /* Snapshot and clear ->rcu_boosted with rcu_node lock held. */
-- if (t->rcu_boosted) {
-- special |= RCU_READ_UNLOCK_BOOSTED;
-- t->rcu_boosted = 0;
-+ /* Snapshot/clear ->rcu_boost_mutex with rcu_node lock held. */
-+ if (t->rcu_boost_mutex) {
-+ rbmp = t->rcu_boost_mutex;
-+ t->rcu_boost_mutex = NULL;
- }
- #endif /* #ifdef CONFIG_RCU_BOOST */
-- t->rcu_blocked_node = NULL;
+ #ifdef CONFIG_SMP
+@@ -134,6 +136,8 @@ extern struct sysdev_class cpu_sysdev_class;
- /*
- * If this was the last task on the current list, and if
-@@ -371,10 +374,8 @@ static noinline void rcu_read_unlock_spe
+ extern void get_online_cpus(void);
+ extern void put_online_cpus(void);
++extern void pin_current_cpu(void);
++extern void unpin_current_cpu(void);
+ #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri)
+ #define register_hotcpu_notifier(nb) register_cpu_notifier(nb)
+ #define unregister_hotcpu_notifier(nb) unregister_cpu_notifier(nb)
+@@ -156,6 +160,8 @@ static inline void cpu_hotplug_driver_unlock(void)
- #ifdef CONFIG_RCU_BOOST
- /* Unboost if we were boosted. */
-- if (special & RCU_READ_UNLOCK_BOOSTED) {
-- rt_mutex_unlock(t->rcu_boost_mutex);
-- t->rcu_boost_mutex = NULL;
-- }
-+ if (rbmp)
-+ rt_mutex_unlock(rbmp);
- #endif /* #ifdef CONFIG_RCU_BOOST */
+ #define get_online_cpus() do { } while (0)
+ #define put_online_cpus() do { } while (0)
++static inline void pin_current_cpu(void) { }
++static inline void unpin_current_cpu(void) { }
+ #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0)
+ /* These aren't inline functions due to a GCC bug. */
+ #define register_hotcpu_notifier(nb) ({ (void)(nb); 0; })
+diff --git a/include/linux/fs.h b/include/linux/fs.h
+index b5b9792..3a023ec 100644
+--- a/include/linux/fs.h
++++ b/include/linux/fs.h
+@@ -777,7 +777,7 @@ struct inode {
+ struct timespec i_ctime;
+ blkcnt_t i_blocks;
+ unsigned short i_bytes;
+- struct rw_semaphore i_alloc_sem;
++ struct rw_anon_semaphore i_alloc_sem;
+ const struct file_operations *i_fop; /* former ->i_op->default_file_ops */
+ struct file_lock *i_flock;
+ struct address_space *i_mapping;
+diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h
+index 59d3ef1..004ebc8 100644
+--- a/include/linux/ftrace_event.h
++++ b/include/linux/ftrace_event.h
+@@ -49,7 +49,8 @@ struct trace_entry {
+ unsigned char flags;
+ unsigned char preempt_count;
+ int pid;
+- int padding;
++ unsigned short migrate_disable;
++ unsigned short padding;
+ };
- /*
-@@ -382,7 +383,7 @@ static noinline void rcu_read_unlock_spe
- * then we need to report up the rcu_node hierarchy.
- */
- if (!empty_exp && !rcu_preempted_readers_exp(rnp))
-- rcu_report_exp_rnp(&rcu_preempt_state, rnp);
-+ rcu_report_exp_rnp(&rcu_preempt_state, rnp, true);
- } else {
- local_irq_restore(flags);
- }
-@@ -711,7 +712,8 @@ static int sync_rcu_preempt_exp_done(str
- *
- * Caller must hold sync_rcu_preempt_exp_mutex.
- */
--static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
-+static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
-+ bool wake)
- {
- unsigned long flags;
- unsigned long mask;
-@@ -724,7 +726,8 @@ static void rcu_report_exp_rnp(struct rc
- }
- if (rnp->parent == NULL) {
- raw_spin_unlock_irqrestore(&rnp->lock, flags);
-- wake_up(&sync_rcu_preempt_exp_wq);
-+ if (wake)
-+ wake_up(&sync_rcu_preempt_exp_wq);
- break;
- }
- mask = rnp->grpmask;
-@@ -757,7 +760,7 @@ sync_rcu_preempt_exp_init(struct rcu_sta
- must_wait = 1;
- }
- if (!must_wait)
-- rcu_report_exp_rnp(rsp, rnp);
-+ rcu_report_exp_rnp(rsp, rnp, false);
- }
+ #define FTRACE_MAX_EVENT \
+diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
+index ba36217..2711055 100644
+--- a/include/linux/hardirq.h
++++ b/include/linux/hardirq.h
+@@ -60,7 +60,11 @@
+ #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT)
+ #define NMI_OFFSET (1UL << NMI_SHIFT)
- /*
-@@ -1048,9 +1051,9 @@ EXPORT_SYMBOL_GPL(synchronize_rcu_expedi
- * report on tasks preempted in RCU read-side critical sections during
- * expedited RCU grace periods.
- */
--static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
-+static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
-+ bool wake)
- {
-- return;
- }
+-#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET)
++#ifndef CONFIG_PREEMPT_RT_FULL
++# define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET)
++#else
++# define SOFTIRQ_DISABLE_OFFSET (0)
++#endif
- #endif /* #ifdef CONFIG_HOTPLUG_CPU */
-@@ -1199,7 +1202,6 @@ static int rcu_boost(struct rcu_node *rn
- t = container_of(tb, struct task_struct, rcu_node_entry);
- rt_mutex_init_proxy_locked(&mtx, t);
- t->rcu_boost_mutex = &mtx;
-- t->rcu_boosted = 1;
- raw_spin_unlock_irqrestore(&rnp->lock, flags);
- rt_mutex_lock(&mtx); /* Side effect: boosts task t's priority. */
- rt_mutex_unlock(&mtx); /* Keep lockdep happy. */
-@@ -1890,7 +1892,7 @@ EXPORT_SYMBOL_GPL(synchronize_sched_expe
+ #ifndef PREEMPT_ACTIVE
+ #define PREEMPT_ACTIVE_BITS 1
+@@ -73,10 +77,17 @@
+ #endif
- #endif /* #else #ifndef CONFIG_SMP */
+ #define hardirq_count() (preempt_count() & HARDIRQ_MASK)
+-#define softirq_count() (preempt_count() & SOFTIRQ_MASK)
+ #define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \
+ | NMI_MASK))
--#if !defined(CONFIG_RCU_FAST_NO_HZ)
-+#if 1 /* !defined(CONFIG_RCU_FAST_NO_HZ) */
++#ifndef CONFIG_PREEMPT_RT_FULL
++# define softirq_count() (preempt_count() & SOFTIRQ_MASK)
++# define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET)
++#else
++# define softirq_count() (0UL)
++extern int in_serving_softirq(void);
++#endif
++
+ /*
+ * Are we doing bottom half or hardware interrupt processing?
+ * Are we in a softirq context? Interrupt context?
+@@ -86,7 +97,6 @@
+ #define in_irq() (hardirq_count())
+ #define in_softirq() (softirq_count())
+ #define in_interrupt() (irq_count())
+-#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET)
/*
- * Check to see if any future RCU-related work will need to be done
-Index: linux-3.0/drivers/usb/gadget/ci13xxx_udc.c
-===================================================================
---- linux-3.0.orig/drivers/usb/gadget/ci13xxx_udc.c
-+++ linux-3.0/drivers/usb/gadget/ci13xxx_udc.c
-@@ -816,7 +816,7 @@ static struct {
- } dbg_data = {
- .idx = 0,
- .tty = 0,
-- .lck = __RW_LOCK_UNLOCKED(lck)
-+ .lck = __RW_LOCK_UNLOCKED(dbg_data.lck)
+ * Are we in NMI context?
+diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h
+index fd0dc30..0e37086 100644
+--- a/include/linux/hrtimer.h
++++ b/include/linux/hrtimer.h
+@@ -111,6 +111,8 @@ struct hrtimer {
+ enum hrtimer_restart (*function)(struct hrtimer *);
+ struct hrtimer_clock_base *base;
+ unsigned long state;
++ struct list_head cb_entry;
++ int irqsafe;
+ #ifdef CONFIG_TIMER_STATS
+ int start_pid;
+ void *start_site;
+@@ -147,6 +149,7 @@ struct hrtimer_clock_base {
+ int index;
+ clockid_t clockid;
+ struct timerqueue_head active;
++ struct list_head expired;
+ ktime_t resolution;
+ ktime_t (*get_time)(void);
+ ktime_t softirq_time;
+@@ -187,6 +190,9 @@ struct hrtimer_cpu_base {
+ unsigned long nr_hangs;
+ ktime_t max_hang_time;
+ #endif
++#ifdef CONFIG_PREEMPT_RT_BASE
++ wait_queue_head_t wait;
++#endif
+ struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES];
};
- /**
-Index: linux-3.0/fs/file.c
-===================================================================
---- linux-3.0.orig/fs/file.c
-+++ linux-3.0/fs/file.c
-@@ -105,14 +105,14 @@ void free_fdtable_rcu(struct rcu_head *r
- kfree(fdt->open_fds);
- kfree(fdt);
- } else {
-- fddef = &get_cpu_var(fdtable_defer_list);
-+ fddef = &per_cpu(fdtable_defer_list, get_cpu_light());
- spin_lock(&fddef->lock);
- fdt->next = fddef->next;
- fddef->next = fdt;
- /* vmallocs are handled from the workqueue context */
- schedule_work(&fddef->wq);
- spin_unlock(&fddef->lock);
-- put_cpu_var(fdtable_defer_list);
-+ put_cpu_light();
- }
+@@ -374,6 +380,13 @@ static inline int hrtimer_restart(struct hrtimer *timer)
+ return hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
}
-@@ -422,7 +422,7 @@ struct files_struct init_files = {
- .close_on_exec = (fd_set *)&init_files.close_on_exec_init,
- .open_fds = (fd_set *)&init_files.open_fds_init,
- },
-- .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock),
-+ .file_lock = __SPIN_LOCK_UNLOCKED(init_files.file_lock),
- };
-
- /*
-Index: linux-3.0/include/linux/idr.h
-===================================================================
---- linux-3.0.orig/include/linux/idr.h
-+++ linux-3.0/include/linux/idr.h
++/* Softirq preemption could deadlock timer removal */
++#ifdef CONFIG_PREEMPT_RT_BASE
++ extern void hrtimer_wait_for_timer(const struct hrtimer *timer);
++#else
++# define hrtimer_wait_for_timer(timer) do { cpu_relax(); } while (0)
++#endif
++
+ /* Query timers: */
+ extern ktime_t hrtimer_get_remaining(const struct hrtimer *timer);
+ extern int hrtimer_get_res(const clockid_t which_clock, struct timespec *tp);
+diff --git a/include/linux/idr.h b/include/linux/idr.h
+index 13a801f..261ef9e 100644
+--- a/include/linux/idr.h
++++ b/include/linux/idr.h
@@ -136,7 +136,7 @@ struct ida {
struct ida_bitmap *free_bitmap;
};
@@ -14368,11820 +10638,15950 @@
#define DEFINE_IDA(name) struct ida name = IDA_INIT(name)
int ida_pre_get(struct ida *ida, gfp_t gfp_mask);
-Index: linux-3.0/kernel/cred.c
-===================================================================
---- linux-3.0.orig/kernel/cred.c
-+++ linux-3.0/kernel/cred.c
-@@ -35,7 +35,7 @@ static struct kmem_cache *cred_jar;
- static struct thread_group_cred init_tgcred = {
- .usage = ATOMIC_INIT(2),
- .tgid = 0,
-- .lock = __SPIN_LOCK_UNLOCKED(init_cred.tgcred.lock),
-+ .lock = __SPIN_LOCK_UNLOCKED(init_tgcred.lock),
+diff --git a/include/linux/init_task.h b/include/linux/init_task.h
+index 580f70c..7b75096 100644
+--- a/include/linux/init_task.h
++++ b/include/linux/init_task.h
+@@ -42,7 +42,7 @@ extern struct fs_struct init_fs;
+ .cputimer = { \
+ .cputime = INIT_CPUTIME, \
+ .running = 0, \
+- .lock = __SPIN_LOCK_UNLOCKED(sig.cputimer.lock), \
++ .lock = __RAW_SPIN_LOCK_UNLOCKED(sig.cputimer.lock), \
+ }, \
+ .cred_guard_mutex = \
+ __MUTEX_INITIALIZER(sig.cred_guard_mutex), \
+@@ -126,6 +126,14 @@ extern struct cred init_cred;
+ # define INIT_PERF_EVENTS(tsk)
+ #endif
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++# define INIT_TIMER_LIST .posix_timer_list = NULL,
++#else
++# define INIT_TIMER_LIST
++#endif
++
++#define INIT_TASK_COMM "swapper"
++
+ /*
+ * INIT_TASK is used to set up the first task table, touch at
+ * your own risk!. Base=0, limit=0x1fffff (=2MB)
+@@ -162,7 +170,7 @@ extern struct cred init_cred;
+ .group_leader = &tsk, \
+ RCU_INIT_POINTER(.real_cred, &init_cred), \
+ RCU_INIT_POINTER(.cred, &init_cred), \
+- .comm = "swapper", \
++ .comm = INIT_TASK_COMM, \
+ .thread = INIT_THREAD, \
+ .fs = &init_fs, \
+ .files = &init_files, \
+@@ -179,6 +187,7 @@ extern struct cred init_cred;
+ .fs_excl = ATOMIC_INIT(0), \
+ .pi_lock = __RAW_SPIN_LOCK_UNLOCKED(tsk.pi_lock), \
+ .timer_slack_ns = 50000, /* 50 usec default slack */ \
++ INIT_TIMER_LIST \
+ .pids = { \
+ [PIDTYPE_PID] = INIT_PID_LINK(PIDTYPE_PID), \
+ [PIDTYPE_PGID] = INIT_PID_LINK(PIDTYPE_PGID), \
+diff --git a/include/linux/intel-iommu.h b/include/linux/intel-iommu.h
+index 9310c69..8b9b5d3 100644
+--- a/include/linux/intel-iommu.h
++++ b/include/linux/intel-iommu.h
+@@ -271,7 +271,7 @@ struct qi_desc {
+ };
+
+ struct q_inval {
+- spinlock_t q_lock;
++ raw_spinlock_t q_lock;
+ struct qi_desc *desc; /* invalidation queue */
+ int *desc_status; /* desc status */
+ int free_head; /* first free entry */
+@@ -311,7 +311,7 @@ struct intel_iommu {
+ u64 cap;
+ u64 ecap;
+ u32 gcmd; /* Holds TE, EAFL. Don't need SRTP, SFL, WBF */
+- spinlock_t register_lock; /* protect register handling */
++ raw_spinlock_t register_lock; /* protect register handling */
+ int seq_id; /* sequence id of the iommu */
+ int agaw; /* agaw of this iommu */
+ int msagaw; /* max sagaw of this iommu */
+diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h
+index b9490bf..a62158f 100644
+--- a/include/linux/interrupt.h
++++ b/include/linux/interrupt.h
+@@ -205,7 +205,7 @@ extern void devm_free_irq(struct device *dev, unsigned int irq, void *dev_id);
+ #ifdef CONFIG_LOCKDEP
+ # define local_irq_enable_in_hardirq() do { } while (0)
+ #else
+-# define local_irq_enable_in_hardirq() local_irq_enable()
++# define local_irq_enable_in_hardirq() local_irq_enable_nort()
+ #endif
+
+ extern void disable_irq_nosync(unsigned int irq);
+@@ -380,9 +380,13 @@ static inline int disable_irq_wake(unsigned int irq)
+
+
+ #ifdef CONFIG_IRQ_FORCED_THREADING
+-extern bool force_irqthreads;
++# ifndef CONFIG_PREEMPT_RT_BASE
++ extern bool force_irqthreads;
++# else
++# define force_irqthreads (true)
++# endif
+ #else
+-#define force_irqthreads (0)
++#define force_irqthreads (false)
+ #endif
+
+ #ifndef __ARCH_SET_SOFTIRQ_PENDING
+@@ -436,8 +440,14 @@ struct softirq_action
+ void (*action)(struct softirq_action *);
};
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ asmlinkage void do_softirq(void);
+ asmlinkage void __do_softirq(void);
++static inline void thread_do_softirq(void) { do_softirq(); }
++#else
++extern void thread_do_softirq(void);
++#endif
++
+ extern void open_softirq(int nr, void (*action)(struct softirq_action *));
+ extern void softirq_init(void);
+ static inline void __raise_softirq_irqoff(unsigned int nr)
+@@ -449,6 +459,8 @@ static inline void __raise_softirq_irqoff(unsigned int nr)
+ extern void raise_softirq_irqoff(unsigned int nr);
+ extern void raise_softirq(unsigned int nr);
+
++extern void softirq_check_pending_idle(void);
++
+ /* This is the worklist that queues up per-cpu softirq work.
+ *
+ * send_remote_sendirq() adds work to these lists, and
+@@ -626,6 +638,12 @@ void tasklet_hrtimer_cancel(struct tasklet_hrtimer *ttimer)
+ tasklet_kill(&ttimer->tasklet);
+ }
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++extern void softirq_early_init(void);
++#else
++static inline void softirq_early_init(void) { }
++#endif
++
+ /*
+ * Autoprobing for irqs:
+ *
+diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h
+index d176d65..a52b35d 100644
+--- a/include/linux/irqflags.h
++++ b/include/linux/irqflags.h
+@@ -25,8 +25,6 @@
+ # define trace_softirqs_enabled(p) ((p)->softirqs_enabled)
+ # define trace_hardirq_enter() do { current->hardirq_context++; } while (0)
+ # define trace_hardirq_exit() do { current->hardirq_context--; } while (0)
+-# define lockdep_softirq_enter() do { current->softirq_context++; } while (0)
+-# define lockdep_softirq_exit() do { current->softirq_context--; } while (0)
+ # define INIT_TRACE_IRQFLAGS .softirqs_enabled = 1,
+ #else
+ # define trace_hardirqs_on() do { } while (0)
+@@ -39,9 +37,15 @@
+ # define trace_softirqs_enabled(p) 0
+ # define trace_hardirq_enter() do { } while (0)
+ # define trace_hardirq_exit() do { } while (0)
++# define INIT_TRACE_IRQFLAGS
++#endif
++
++#if defined(CONFIG_TRACE_IRQFLAGS) && !defined(CONFIG_PREEMPT_RT_FULL)
++# define lockdep_softirq_enter() do { current->softirq_context++; } while (0)
++# define lockdep_softirq_exit() do { current->softirq_context--; } while (0)
++#else
+ # define lockdep_softirq_enter() do { } while (0)
+ # define lockdep_softirq_exit() do { } while (0)
+-# define INIT_TRACE_IRQFLAGS
+ #endif
+
+ #if defined(CONFIG_IRQSOFF_TRACER) || \
+@@ -147,4 +151,23 @@
+
+ #endif /* CONFIG_TRACE_IRQFLAGS_SUPPORT */
+
++/*
++ * local_irq* variants depending on RT/!RT
++ */
++#ifdef CONFIG_PREEMPT_RT_FULL
++# define local_irq_disable_nort() do { } while (0)
++# define local_irq_enable_nort() do { } while (0)
++# define local_irq_save_nort(flags) do { local_save_flags(flags); } while (0)
++# define local_irq_restore_nort(flags) do { (void)(flags); } while (0)
++# define local_irq_disable_rt() local_irq_disable()
++# define local_irq_enable_rt() local_irq_enable()
++#else
++# define local_irq_disable_nort() local_irq_disable()
++# define local_irq_enable_nort() local_irq_enable()
++# define local_irq_save_nort(flags) local_irq_save(flags)
++# define local_irq_restore_nort(flags) local_irq_restore(flags)
++# define local_irq_disable_rt() do { } while (0)
++# define local_irq_enable_rt() do { } while (0)
++#endif
++
#endif
+diff --git a/include/linux/jbd.h b/include/linux/jbd.h
+index e069650..d9def47 100644
+--- a/include/linux/jbd.h
++++ b/include/linux/jbd.h
+@@ -244,6 +244,7 @@ typedef struct journal_superblock_s
+
+ #include <linux/fs.h>
+ #include <linux/sched.h>
++#include <linux/jbd_common.h>
-Index: linux-3.0/include/linux/seqlock.h
-===================================================================
---- linux-3.0.orig/include/linux/seqlock.h
-+++ linux-3.0/include/linux/seqlock.h
-@@ -30,92 +30,12 @@
- #include <linux/preempt.h>
- #include <asm/processor.h>
+ #define J_ASSERT(assert) BUG_ON(!(assert))
--typedef struct {
-- unsigned sequence;
-- spinlock_t lock;
--} seqlock_t;
+@@ -270,69 +271,6 @@ typedef struct journal_superblock_s
+ #define J_EXPECT_JH(jh, expr, why...) __journal_expect(expr, ## why)
+ #endif
+
+-enum jbd_state_bits {
+- BH_JBD /* Has an attached ext3 journal_head */
+- = BH_PrivateStart,
+- BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
+- BH_Freed, /* Has been freed (truncated) */
+- BH_Revoked, /* Has been revoked from the log */
+- BH_RevokeValid, /* Revoked flag is valid */
+- BH_JBDDirty, /* Is dirty but journaled */
+- BH_State, /* Pins most journal_head state */
+- BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
+- BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
+-};
-
--/*
-- * These macros triggered gcc-3.x compile-time problems. We think these are
-- * OK now. Be cautious.
-- */
--#define __SEQLOCK_UNLOCKED(lockname) \
-- { 0, __SPIN_LOCK_UNLOCKED(lockname) }
+-BUFFER_FNS(JBD, jbd)
+-BUFFER_FNS(JWrite, jwrite)
+-BUFFER_FNS(JBDDirty, jbddirty)
+-TAS_BUFFER_FNS(JBDDirty, jbddirty)
+-BUFFER_FNS(Revoked, revoked)
+-TAS_BUFFER_FNS(Revoked, revoked)
+-BUFFER_FNS(RevokeValid, revokevalid)
+-TAS_BUFFER_FNS(RevokeValid, revokevalid)
+-BUFFER_FNS(Freed, freed)
-
--#define seqlock_init(x) \
-- do { \
-- (x)->sequence = 0; \
-- spin_lock_init(&(x)->lock); \
-- } while (0)
+-static inline struct buffer_head *jh2bh(struct journal_head *jh)
+-{
+- return jh->b_bh;
+-}
-
--#define DEFINE_SEQLOCK(x) \
-- seqlock_t x = __SEQLOCK_UNLOCKED(x)
+-static inline struct journal_head *bh2jh(struct buffer_head *bh)
+-{
+- return bh->b_private;
+-}
-
--/* Lock out other writers and update the count.
-- * Acts like a normal spin_lock/unlock.
-- * Don't need preempt_disable() because that is in the spin_lock already.
-- */
--static inline void write_seqlock(seqlock_t *sl)
+-static inline void jbd_lock_bh_state(struct buffer_head *bh)
-{
-- spin_lock(&sl->lock);
-- ++sl->sequence;
-- smp_wmb();
+- bit_spin_lock(BH_State, &bh->b_state);
-}
-
--static inline void write_sequnlock(seqlock_t *sl)
+-static inline int jbd_trylock_bh_state(struct buffer_head *bh)
-{
-- smp_wmb();
-- sl->sequence++;
-- spin_unlock(&sl->lock);
+- return bit_spin_trylock(BH_State, &bh->b_state);
-}
-
--static inline int write_tryseqlock(seqlock_t *sl)
+-static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
-{
-- int ret = spin_trylock(&sl->lock);
+- return bit_spin_is_locked(BH_State, &bh->b_state);
+-}
-
-- if (ret) {
-- ++sl->sequence;
-- smp_wmb();
-- }
-- return ret;
+-static inline void jbd_unlock_bh_state(struct buffer_head *bh)
+-{
+- bit_spin_unlock(BH_State, &bh->b_state);
-}
-
--/* Start of read calculation -- fetch last complete writer token */
--static __always_inline unsigned read_seqbegin(const seqlock_t *sl)
+-static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
-{
-- unsigned ret;
+- bit_spin_lock(BH_JournalHead, &bh->b_state);
+-}
-
--repeat:
-- ret = ACCESS_ONCE(sl->sequence);
-- if (unlikely(ret & 1)) {
-- cpu_relax();
-- goto repeat;
-- }
-- smp_rmb();
+-static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
+-{
+- bit_spin_unlock(BH_JournalHead, &bh->b_state);
+-}
-
-- return ret;
+ struct jbd_revoke_table_s;
+
+ /**
+diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
+index d087c2e..599d9ed 100644
+--- a/include/linux/jbd2.h
++++ b/include/linux/jbd2.h
+@@ -275,6 +275,7 @@ typedef struct journal_superblock_s
+
+ #include <linux/fs.h>
+ #include <linux/sched.h>
++#include <linux/jbd_common.h>
+
+ #define J_ASSERT(assert) BUG_ON(!(assert))
+
+@@ -302,70 +303,6 @@ typedef struct journal_superblock_s
+ #define J_EXPECT_JH(jh, expr, why...) __journal_expect(expr, ## why)
+ #endif
+
+-enum jbd_state_bits {
+- BH_JBD /* Has an attached ext3 journal_head */
+- = BH_PrivateStart,
+- BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
+- BH_Freed, /* Has been freed (truncated) */
+- BH_Revoked, /* Has been revoked from the log */
+- BH_RevokeValid, /* Revoked flag is valid */
+- BH_JBDDirty, /* Is dirty but journaled */
+- BH_State, /* Pins most journal_head state */
+- BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
+- BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
+- BH_JBDPrivateStart, /* First bit available for private use by FS */
+-};
+-
+-BUFFER_FNS(JBD, jbd)
+-BUFFER_FNS(JWrite, jwrite)
+-BUFFER_FNS(JBDDirty, jbddirty)
+-TAS_BUFFER_FNS(JBDDirty, jbddirty)
+-BUFFER_FNS(Revoked, revoked)
+-TAS_BUFFER_FNS(Revoked, revoked)
+-BUFFER_FNS(RevokeValid, revokevalid)
+-TAS_BUFFER_FNS(RevokeValid, revokevalid)
+-BUFFER_FNS(Freed, freed)
+-
+-static inline struct buffer_head *jh2bh(struct journal_head *jh)
+-{
+- return jh->b_bh;
-}
-
--/*
-- * Test if reader processed invalid data.
-- *
-- * If sequence value changed then writer changed data while in section.
-- */
--static __always_inline int read_seqretry(const seqlock_t *sl, unsigned start)
+-static inline struct journal_head *bh2jh(struct buffer_head *bh)
-{
-- smp_rmb();
+- return bh->b_private;
+-}
-
-- return unlikely(sl->sequence != start);
+-static inline void jbd_lock_bh_state(struct buffer_head *bh)
+-{
+- bit_spin_lock(BH_State, &bh->b_state);
+-}
+-
+-static inline int jbd_trylock_bh_state(struct buffer_head *bh)
+-{
+- return bit_spin_trylock(BH_State, &bh->b_state);
-}
-
+-static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
+-{
+- return bit_spin_is_locked(BH_State, &bh->b_state);
+-}
-
- /*
- * Version using sequence counter only.
- * This can be used when code has its own mutex protecting the
- * updating starting before the write_seqcountbeqin() and ending
- * after the write_seqcount_end().
- */
+-static inline void jbd_unlock_bh_state(struct buffer_head *bh)
+-{
+- bit_spin_unlock(BH_State, &bh->b_state);
+-}
-
- typedef struct seqcount {
- unsigned sequence;
- } seqcount_t;
-@@ -197,7 +117,6 @@ static inline int __read_seqcount_retry(
- static inline int read_seqcount_retry(const seqcount_t *s, unsigned start)
- {
- smp_rmb();
+-static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
+-{
+- bit_spin_lock(BH_JournalHead, &bh->b_state);
+-}
-
- return __read_seqcount_retry(s, start);
- }
+-static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
+-{
+- bit_spin_unlock(BH_JournalHead, &bh->b_state);
+-}
+-
+ /* Flags in jbd_inode->i_flags */
+ #define __JI_COMMIT_RUNNING 0
+ /* Commit of the inode data in progress. We use this flag to protect us from
+diff --git a/include/linux/jbd_common.h b/include/linux/jbd_common.h
+new file mode 100644
+index 0000000..11c313e
+--- /dev/null
++++ b/include/linux/jbd_common.h
+@@ -0,0 +1,92 @@
++#ifndef _LINUX_JBD_STATE_H
++#define _LINUX_JBD_STATE_H
++
++enum jbd_state_bits {
++ BH_JBD /* Has an attached ext3 journal_head */
++ = BH_PrivateStart,
++ BH_JWrite, /* Being written to log (@@@ DEBUGGING) */
++ BH_Freed, /* Has been freed (truncated) */
++ BH_Revoked, /* Has been revoked from the log */
++ BH_RevokeValid, /* Revoked flag is valid */
++ BH_JBDDirty, /* Is dirty but journaled */
++ BH_State, /* Pins most journal_head state */
++ BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
++ BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
++ BH_JBDPrivateStart, /* First bit available for private use by FS */
++};
++
++BUFFER_FNS(JBD, jbd)
++BUFFER_FNS(JWrite, jwrite)
++BUFFER_FNS(JBDDirty, jbddirty)
++TAS_BUFFER_FNS(JBDDirty, jbddirty)
++BUFFER_FNS(Revoked, revoked)
++TAS_BUFFER_FNS(Revoked, revoked)
++BUFFER_FNS(RevokeValid, revokevalid)
++TAS_BUFFER_FNS(RevokeValid, revokevalid)
++BUFFER_FNS(Freed, freed)
++
++static inline struct buffer_head *jh2bh(struct journal_head *jh)
++{
++ return jh->b_bh;
++}
++
++static inline struct journal_head *bh2jh(struct buffer_head *bh)
++{
++ return bh->b_private;
++}
++
++static inline void jbd_lock_bh_state(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ bit_spin_lock(BH_State, &bh->b_state);
++#else
++ spin_lock(&bh->b_state_lock);
++#endif
++}
++
++static inline int jbd_trylock_bh_state(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ return bit_spin_trylock(BH_State, &bh->b_state);
++#else
++ return spin_trylock(&bh->b_state_lock);
++#endif
++}
++
++static inline int jbd_is_locked_bh_state(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ return bit_spin_is_locked(BH_State, &bh->b_state);
++#else
++ return spin_is_locked(&bh->b_state_lock);
++#endif
++}
++
++static inline void jbd_unlock_bh_state(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ bit_spin_unlock(BH_State, &bh->b_state);
++#else
++ spin_unlock(&bh->b_state_lock);
++#endif
++}
++
++static inline void jbd_lock_bh_journal_head(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ bit_spin_lock(BH_JournalHead, &bh->b_state);
++#else
++ spin_lock(&bh->b_journal_head_lock);
++#endif
++}
++
++static inline void jbd_unlock_bh_journal_head(struct buffer_head *bh)
++{
++#ifndef CONFIG_PREEMPT_RT_BASE
++ bit_spin_unlock(BH_JournalHead, &bh->b_state);
++#else
++ spin_unlock(&bh->b_journal_head_lock);
++#endif
++}
++
++#endif
+diff --git a/include/linux/jump_label.h b/include/linux/jump_label.h
+index 83e745f..4ec8a0b 100644
+--- a/include/linux/jump_label.h
++++ b/include/linux/jump_label.h
+@@ -4,7 +4,7 @@
+ #include <linux/types.h>
+ #include <linux/compiler.h>
-@@ -231,31 +150,154 @@ static inline void write_seqcount_barrie
- s->sequence+=2;
- }
+-#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
++#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL) && !defined(CONFIG_PREEMPT_BASE)
-+typedef struct {
-+ struct seqcount seqcount;
+ struct jump_label_key {
+ atomic_t enabled;
+diff --git a/include/linux/kdb.h b/include/linux/kdb.h
+index aadff7c..18d20dd 100644
+--- a/include/linux/kdb.h
++++ b/include/linux/kdb.h
+@@ -153,12 +153,14 @@ extern int kdb_register(char *, kdb_func_t, char *, char *, short);
+ extern int kdb_register_repeat(char *, kdb_func_t, char *, char *,
+ short, kdb_repeat_t);
+ extern int kdb_unregister(char *);
++#define in_kdb_printk() (kdb_trap_printk)
+ #else /* ! CONFIG_KGDB_KDB */
+ #define kdb_printf(...)
+ #define kdb_init(x)
+ #define kdb_register(...)
+ #define kdb_register_repeat(...)
+ #define kdb_uregister(x)
++#define in_kdb_printk() (0)
+ #endif /* CONFIG_KGDB_KDB */
+ enum {
+ KDB_NOT_INITIALIZED,
+diff --git a/include/linux/kernel.h b/include/linux/kernel.h
+index 953352a..06dcdb4 100644
+--- a/include/linux/kernel.h
++++ b/include/linux/kernel.h
+@@ -346,7 +346,7 @@ extern enum system_states {
+ SYSTEM_HALT,
+ SYSTEM_POWER_OFF,
+ SYSTEM_RESTART,
+- SYSTEM_SUSPEND_DISK,
++ SYSTEM_SUSPEND,
+ } system_state;
+
+ #define TAINT_PROPRIETARY_MODULE 0
+diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h
+index dd7c12e..dce6e4d 100644
+--- a/include/linux/kprobes.h
++++ b/include/linux/kprobes.h
+@@ -181,7 +181,7 @@ struct kretprobe {
+ int nmissed;
+ size_t data_size;
+ struct hlist_head free_instances;
+- spinlock_t lock;
+ raw_spinlock_t lock;
-+} raw_seqlock_t;
+ };
+
+ struct kretprobe_instance {
+diff --git a/include/linux/lglock.h b/include/linux/lglock.h
+index f549056..433f12d 100644
+--- a/include/linux/lglock.h
++++ b/include/linux/lglock.h
+@@ -70,6 +70,9 @@
+ extern void name##_global_lock_online(void); \
+ extern void name##_global_unlock_online(void); \
+
+
-+typedef struct {
-+ struct seqcount seqcount;
-+ spinlock_t lock;
-+} seqlock_t;
++#ifndef CONFIG_PREEMPT_RT_FULL
+
-+/*
-+ * These macros triggered gcc-3.x compile-time problems. We think these are
-+ * OK now. Be cautious.
+ #define DEFINE_LGLOCK(name) \
+ \
+ DEFINE_PER_CPU(arch_spinlock_t, name##_lock); \
+@@ -169,4 +172,101 @@
+ preempt_enable(); \
+ } \
+ EXPORT_SYMBOL(name##_global_unlock);
++
++#else /* !PREEMPT_RT_FULL */
++#define DEFINE_LGLOCK(name) \
++ \
++ DEFINE_PER_CPU(struct rt_mutex, name##_lock); \
++ DEFINE_LGLOCK_LOCKDEP(name); \
++ \
++ void name##_lock_init(void) { \
++ int i; \
++ LOCKDEP_INIT_MAP(&name##_lock_dep_map, #name, &name##_lock_key, 0); \
++ for_each_possible_cpu(i) { \
++ struct rt_mutex *lock; \
++ lock = &per_cpu(name##_lock, i); \
++ rt_mutex_init(lock); \
++ } \
++ } \
++ EXPORT_SYMBOL(name##_lock_init); \
++ \
++ void name##_local_lock(void) { \
++ struct rt_mutex *lock; \
++ migrate_disable(); \
++ rwlock_acquire_read(&name##_lock_dep_map, 0, 0, _THIS_IP_); \
++ lock = &__get_cpu_var(name##_lock); \
++ __rt_spin_lock(lock); \
++ } \
++ EXPORT_SYMBOL(name##_local_lock); \
++ \
++ void name##_local_unlock(void) { \
++ struct rt_mutex *lock; \
++ rwlock_release(&name##_lock_dep_map, 1, _THIS_IP_); \
++ lock = &__get_cpu_var(name##_lock); \
++ __rt_spin_unlock(lock); \
++ migrate_enable(); \
++ } \
++ EXPORT_SYMBOL(name##_local_unlock); \
++ \
++ void name##_local_lock_cpu(int cpu) { \
++ struct rt_mutex *lock; \
++ rwlock_acquire_read(&name##_lock_dep_map, 0, 0, _THIS_IP_); \
++ lock = &per_cpu(name##_lock, cpu); \
++ __rt_spin_lock(lock); \
++ } \
++ EXPORT_SYMBOL(name##_local_lock_cpu); \
++ \
++ void name##_local_unlock_cpu(int cpu) { \
++ struct rt_mutex *lock; \
++ rwlock_release(&name##_lock_dep_map, 1, _THIS_IP_); \
++ lock = &per_cpu(name##_lock, cpu); \
++ __rt_spin_unlock(lock); \
++ } \
++ EXPORT_SYMBOL(name##_local_unlock_cpu); \
++ \
++ void name##_global_lock_online(void) { \
++ int i; \
++ rwlock_acquire(&name##_lock_dep_map, 0, 0, _RET_IP_); \
++ for_each_online_cpu(i) { \
++ struct rt_mutex *lock; \
++ lock = &per_cpu(name##_lock, i); \
++ __rt_spin_lock(lock); \
++ } \
++ } \
++ EXPORT_SYMBOL(name##_global_lock_online); \
++ \
++ void name##_global_unlock_online(void) { \
++ int i; \
++ rwlock_release(&name##_lock_dep_map, 1, _RET_IP_); \
++ for_each_online_cpu(i) { \
++ struct rt_mutex *lock; \
++ lock = &per_cpu(name##_lock, i); \
++ __rt_spin_unlock(lock); \
++ } \
++ } \
++ EXPORT_SYMBOL(name##_global_unlock_online); \
++ \
++ void name##_global_lock(void) { \
++ int i; \
++ rwlock_acquire(&name##_lock_dep_map, 0, 0, _RET_IP_); \
++ for_each_possible_cpu(i) { \
++ struct rt_mutex *lock; \
++ lock = &per_cpu(name##_lock, i); \
++ __rt_spin_lock(lock); \
++ } \
++ } \
++ EXPORT_SYMBOL(name##_global_lock); \
++ \
++ void name##_global_unlock(void) { \
++ int i; \
++ rwlock_release(&name##_lock_dep_map, 1, _RET_IP_); \
++ for_each_possible_cpu(i) { \
++ struct rt_mutex *lock; \
++ lock = &per_cpu(name##_lock, i); \
++ __rt_spin_unlock(lock); \
++ } \
++ } \
++ EXPORT_SYMBOL(name##_global_unlock);
++#endif /* PRREMPT_RT_FULL */
++
+ #endif
+diff --git a/include/linux/list.h b/include/linux/list.h
+index cc6d2aa..7a9851b 100644
+--- a/include/linux/list.h
++++ b/include/linux/list.h
+@@ -362,6 +362,17 @@ static inline void list_splice_tail_init(struct list_head *list,
+ list_entry((ptr)->next, type, member)
+
+ /**
++ * list_last_entry - get the last element from a list
++ * @ptr: the list head to take the element from.
++ * @type: the type of the struct this is embedded in.
++ * @member: the name of the list_struct within the struct.
++ *
++ * Note, that list is expected to be not empty.
+ */
-+#define __RAW_SEQLOCK_UNLOCKED(lockname) \
-+ { \
-+ .seqcount = SEQCNT_ZERO, \
-+ .lock = __RAW_SPIN_LOCK_UNLOCKED(lockname) \
-+ }
++#define list_last_entry(ptr, type, member) \
++ list_entry((ptr)->prev, type, member)
+
-+#define raw_seqlock_init(x) \
-+ do { \
-+ seqcount_init(&(x)->seqcount); \
-+ raw_spin_lock_init(&(x)->lock); \
-+ } while (0)
++/**
+ * list_for_each - iterate over a list
+ * @pos: the &struct list_head to use as a loop cursor.
+ * @head: the head for your list.
+diff --git a/include/linux/locallock.h b/include/linux/locallock.h
+new file mode 100644
+index 0000000..8fbc393
+--- /dev/null
++++ b/include/linux/locallock.h
+@@ -0,0 +1,230 @@
++#ifndef _LINUX_LOCALLOCK_H
++#define _LINUX_LOCALLOCK_H
+
-+#define DEFINE_RAW_SEQLOCK(x) \
-+ raw_seqlock_t x = __RAW_SEQLOCK_UNLOCKED(x)
++#include <linux/spinlock.h>
+
-+#define __SEQLOCK_UNLOCKED(lockname) \
-+ { \
-+ .seqcount = SEQCNT_ZERO, \
-+ .lock = __SPIN_LOCK_UNLOCKED(lockname) \
-+ }
++#ifdef CONFIG_PREEMPT_RT_BASE
+
-+#define seqlock_init(x) \
-+ do { \
-+ seqcount_init(&(x)->seqcount); \
-+ spin_lock_init(&(x)->lock); \
-+ } while (0)
++#ifdef CONFIG_DEBUG_SPINLOCK
++# define LL_WARN(cond) WARN_ON(cond)
++#else
++# define LL_WARN(cond) do { } while (0)
++#endif
+
-+#define DEFINE_SEQLOCK(x) \
-+ seqlock_t x = __SEQLOCK_UNLOCKED(x)
++/*
++ * per cpu lock based substitute for local_irq_*()
++ */
++struct local_irq_lock {
++ spinlock_t lock;
++ struct task_struct *owner;
++ int nestcnt;
++ unsigned long flags;
++};
++
++#define DEFINE_LOCAL_IRQ_LOCK(lvar) \
++ DEFINE_PER_CPU(struct local_irq_lock, lvar) = { \
++ .lock = __SPIN_LOCK_UNLOCKED((lvar).lock) }
+
-+#define read_seqbegin(sl) read_seqcount_begin(&(sl)->seqcount)
-+#define read_seqretry(sl, start) read_seqcount_retry(&(sl)->seqcount, start)
++#define local_irq_lock_init(lvar) \
++ do { \
++ int __cpu; \
++ for_each_possible_cpu(__cpu) \
++ spin_lock_init(&per_cpu(lvar, __cpu).lock); \
++ } while (0)
+
- /*
-- * Possible sw/hw IRQ protected versions of the interfaces.
-+ * Lock out other writers and update the count.
-+ * Acts like a normal spin_lock/unlock.
-+ * Don't need preempt_disable() because that is in the spin_lock already.
- */
-+static inline void raw_write_seqlock(raw_seqlock_t *sl)
++static inline void __local_lock(struct local_irq_lock *lv)
+{
-+ raw_spin_lock(&sl->lock);
-+ write_seqcount_begin(&sl->seqcount);
++ if (lv->owner != current) {
++ spin_lock(&lv->lock);
++ LL_WARN(lv->owner);
++ LL_WARN(lv->nestcnt);
++ lv->owner = current;
++ }
++ lv->nestcnt++;
+}
+
-+static inline void raw_write_sequnlock(raw_seqlock_t *sl)
-+{
-+ write_seqcount_end(&sl->seqcount);
-+ raw_spin_unlock(&sl->lock);
-+}
++#define local_lock(lvar) \
++ do { __local_lock(&get_local_var(lvar)); } while (0)
+
-+static inline void raw_write_seqlock_irq(raw_seqlock_t *sl)
++static inline int __local_trylock(struct local_irq_lock *lv)
+{
-+ raw_spin_lock_irq(&sl->lock);
-+ write_seqcount_begin(&sl->seqcount);
++ if (lv->owner != current && spin_trylock(&lv->lock)) {
++ LL_WARN(lv->owner);
++ LL_WARN(lv->nestcnt);
++ lv->owner = current;
++ lv->nestcnt = 1;
++ return 1;
++ }
++ return 0;
+}
+
-+static inline void raw_write_sequnlock_irq(raw_seqlock_t *sl)
-+{
-+ write_seqcount_end(&sl->seqcount);
-+ raw_spin_unlock_irq(&sl->lock);
-+}
++#define local_trylock(lvar) \
++ ({ \
++ int __locked; \
++ __locked = __local_trylock(&get_local_var(lvar)); \
++ if (!__locked) \
++ put_local_var(lvar); \
++ __locked; \
++ })
+
-+static inline unsigned long __raw_write_seqlock_irqsave(raw_seqlock_t *sl)
++static inline void __local_unlock(struct local_irq_lock *lv)
+{
-+ unsigned long flags;
++ LL_WARN(lv->nestcnt == 0);
++ LL_WARN(lv->owner != current);
++ if (--lv->nestcnt)
++ return;
+
-+ raw_spin_lock_irqsave(&sl->lock, flags);
-+ write_seqcount_begin(&sl->seqcount);
-+ return flags;
++ lv->owner = NULL;
++ spin_unlock(&lv->lock);
+}
+
-+#define raw_write_seqlock_irqsave(lock, flags) \
-+ do { flags = __raw_write_seqlock_irqsave(lock); } while (0)
++#define local_unlock(lvar) \
++ do { \
++ __local_unlock(&__get_cpu_var(lvar)); \
++ put_local_var(lvar); \
++ } while (0)
+
-+static inline void
-+raw_write_sequnlock_irqrestore(raw_seqlock_t *sl, unsigned long flags)
++static inline void __local_lock_irq(struct local_irq_lock *lv)
+{
-+ write_seqcount_end(&sl->seqcount);
-+ raw_spin_unlock_irqrestore(&sl->lock, flags);
++ spin_lock_irqsave(&lv->lock, lv->flags);
++ LL_WARN(lv->owner);
++ LL_WARN(lv->nestcnt);
++ lv->owner = current;
++ lv->nestcnt = 1;
+}
+
-+static inline void write_seqlock(seqlock_t *sl)
-+{
-+ spin_lock(&sl->lock);
-+ write_seqcount_begin(&sl->seqcount);
-+}
++#define local_lock_irq(lvar) \
++ do { __local_lock_irq(&get_local_var(lvar)); } while (0)
+
-+static inline void write_sequnlock(seqlock_t *sl)
++static inline void __local_unlock_irq(struct local_irq_lock *lv)
+{
-+ write_seqcount_end(&sl->seqcount);
-+ spin_unlock(&sl->lock);
++ LL_WARN(!lv->nestcnt);
++ LL_WARN(lv->owner != current);
++ lv->owner = NULL;
++ lv->nestcnt = 0;
++ spin_unlock_irq(&lv->lock);
+}
+
-+static inline void write_seqlock_bh(seqlock_t *sl)
-+{
-+ spin_lock_bh(&sl->lock);
-+ write_seqcount_begin(&sl->seqcount);
-+}
++#define local_unlock_irq(lvar) \
++ do { \
++ __local_unlock_irq(&__get_cpu_var(lvar)); \
++ put_local_var(lvar); \
++ } while (0)
+
-+static inline void write_sequnlock_bh(seqlock_t *sl)
++static inline int __local_lock_irqsave(struct local_irq_lock *lv)
+{
-+ write_seqcount_end(&sl->seqcount);
-+ spin_unlock_bh(&sl->lock);
++ if (lv->owner != current) {
++ __local_lock_irq(lv);
++ return 0;
++ } else {
++ lv->nestcnt++;
++ return 1;
++ }
+}
+
-+static inline void write_seqlock_irq(seqlock_t *sl)
-+{
-+ spin_lock_irq(&sl->lock);
-+ write_seqcount_begin(&sl->seqcount);
-+}
++#define local_lock_irqsave(lvar, _flags) \
++ do { \
++ if (__local_lock_irqsave(&get_local_var(lvar))) \
++ put_local_var(lvar); \
++ _flags = __get_cpu_var(lvar).flags; \
++ } while (0)
+
-+static inline void write_sequnlock_irq(seqlock_t *sl)
++static inline int __local_unlock_irqrestore(struct local_irq_lock *lv,
++ unsigned long flags)
+{
-+ write_seqcount_end(&sl->seqcount);
-+ spin_unlock_irq(&sl->lock);
++ LL_WARN(!lv->nestcnt);
++ LL_WARN(lv->owner != current);
++ if (--lv->nestcnt)
++ return 0;
++
++ lv->owner = NULL;
++ spin_unlock_irqrestore(&lv->lock, lv->flags);
++ return 1;
+}
+
-+static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl)
-+{
-+ unsigned long flags;
++#define local_unlock_irqrestore(lvar, flags) \
++ do { \
++ if (__local_unlock_irqrestore(&__get_cpu_var(lvar), flags)) \
++ put_local_var(lvar); \
++ } while (0)
+
-+ spin_lock_irqsave(&sl->lock, flags);
-+ write_seqcount_begin(&sl->seqcount);
-+ return flags;
-+}
++#define local_spin_trylock_irq(lvar, lock) \
++ ({ \
++ int __locked; \
++ local_lock_irq(lvar); \
++ __locked = spin_trylock(lock); \
++ if (!__locked) \
++ local_unlock_irq(lvar); \
++ __locked; \
++ })
+
- #define write_seqlock_irqsave(lock, flags) \
-- do { local_irq_save(flags); write_seqlock(lock); } while (0)
--#define write_seqlock_irq(lock) \
-- do { local_irq_disable(); write_seqlock(lock); } while (0)
--#define write_seqlock_bh(lock) \
-- do { local_bh_disable(); write_seqlock(lock); } while (0)
--
--#define write_sequnlock_irqrestore(lock, flags) \
-- do { write_sequnlock(lock); local_irq_restore(flags); } while(0)
--#define write_sequnlock_irq(lock) \
-- do { write_sequnlock(lock); local_irq_enable(); } while(0)
--#define write_sequnlock_bh(lock) \
-- do { write_sequnlock(lock); local_bh_enable(); } while(0)
--
--#define read_seqbegin_irqsave(lock, flags) \
-- ({ local_irq_save(flags); read_seqbegin(lock); })
--
--#define read_seqretry_irqrestore(lock, iv, flags) \
-- ({ \
-- int ret = read_seqretry(lock, iv); \
-- local_irq_restore(flags); \
-- ret; \
-- })
-+ do { flags = __write_seqlock_irqsave(lock); } while (0)
++#define local_spin_lock_irq(lvar, lock) \
++ do { \
++ local_lock_irq(lvar); \
++ spin_lock(lock); \
++ } while (0)
++
++#define local_spin_unlock_irq(lvar, lock) \
++ do { \
++ spin_unlock(lock); \
++ local_unlock_irq(lvar); \
++ } while (0)
++
++#define local_spin_lock_irqsave(lvar, lock, flags) \
++ do { \
++ local_lock_irqsave(lvar, flags); \
++ spin_lock(lock); \
++ } while (0)
++
++#define local_spin_unlock_irqrestore(lvar, lock, flags) \
++ do { \
++ spin_unlock(lock); \
++ local_unlock_irqrestore(lvar, flags); \
++ } while (0)
++
++#define get_locked_var(lvar, var) \
++ (*({ \
++ local_lock(lvar); \
++ &__get_cpu_var(var); \
++ }))
++
++#define put_locked_var(lvar, var) local_unlock(lvar)
++
++#define local_lock_cpu(lvar) \
++ ({ \
++ local_lock(lvar); \
++ smp_processor_id(); \
++ })
++
++#define local_unlock_cpu(lvar) local_unlock(lvar)
++
++#else /* PREEMPT_RT_BASE */
++
++#define DEFINE_LOCAL_IRQ_LOCK(lvar) __typeof__(const int) lvar
++
++static inline void local_irq_lock_init(int lvar) { }
++
++#define local_lock(lvar) preempt_disable()
++#define local_unlock(lvar) preempt_enable()
++#define local_lock_irq(lvar) local_irq_disable()
++#define local_unlock_irq(lvar) local_irq_enable()
++#define local_lock_irqsave(lvar, flags) local_irq_save(flags)
++#define local_unlock_irqrestore(lvar, flags) local_irq_restore(flags)
++
++#define local_spin_trylock_irq(lvar, lock) spin_trylock_irq(lock)
++#define local_spin_lock_irq(lvar, lock) spin_lock_irq(lock)
++#define local_spin_unlock_irq(lvar, lock) spin_unlock_irq(lock)
++#define local_spin_lock_irqsave(lvar, lock, flags) \
++ spin_lock_irqsave(lock, flags)
++#define local_spin_unlock_irqrestore(lvar, lock, flags) \
++ spin_unlock_irqrestore(lock, flags)
++
++#define get_locked_var(lvar, var) get_cpu_var(var)
++#define put_locked_var(lvar, var) put_cpu_var(var)
+
-+static inline void
-+write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags)
++#define local_lock_cpu(lvar) get_cpu()
++#define local_unlock_cpu(lvar) put_cpu()
++
++#endif
++
++#endif
+diff --git a/include/linux/mm.h b/include/linux/mm.h
+index 18eea05..9ce5766 100644
+--- a/include/linux/mm.h
++++ b/include/linux/mm.h
+@@ -1229,27 +1229,59 @@ static inline pmd_t *pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long a
+ * overflow into the next struct page (as it might with DEBUG_SPINLOCK).
+ * When freeing, reset page->mapping so free_pages_check won't complain.
+ */
++#ifndef CONFIG_PREEMPT_RT_FULL
++
+ #define __pte_lockptr(page) &((page)->ptl)
+-#define pte_lock_init(_page) do { \
+- spin_lock_init(__pte_lockptr(_page)); \
+-} while (0)
++
++static inline struct page *pte_lock_init(struct page *page)
+{
-+ write_seqcount_end(&sl->seqcount);
-+ spin_unlock_irqrestore(&sl->lock, flags);
++ spin_lock_init(__pte_lockptr(page));
++ return page;
+}
-
- #endif /* __LINUX_SEQLOCK_H */
-Index: linux-3.0/arch/ia64/kernel/time.c
-===================================================================
---- linux-3.0.orig/arch/ia64/kernel/time.c
-+++ linux-3.0/arch/ia64/kernel/time.c
-@@ -36,7 +36,7 @@
- static cycle_t itc_get_cycles(struct clocksource *cs);
-
- struct fsyscall_gtod_data_t fsyscall_gtod_data = {
-- .lock = __SEQLOCK_UNLOCKED(fsyscall_gtod_data.lock),
-+ .lock = __RAW_SEQLOCK_UNLOCKED(fsyscall_gtod_data.lock),
- };
-
- struct itc_jitter_data_t itc_jitter_data;
-@@ -462,7 +462,7 @@ void update_vsyscall(struct timespec *wa
- {
- unsigned long flags;
-
-- write_seqlock_irqsave(&fsyscall_gtod_data.lock, flags);
-+ raw_write_seqlock_irqsave(&fsyscall_gtod_data.lock, flags);
-
- /* copy fsyscall clock data */
- fsyscall_gtod_data.clk_mask = c->mask;
-@@ -485,6 +485,6 @@ void update_vsyscall(struct timespec *wa
- fsyscall_gtod_data.monotonic_time.tv_sec++;
- }
-
-- write_sequnlock_irqrestore(&fsyscall_gtod_data.lock, flags);
-+ raw_write_sequnlock_irqrestore(&fsyscall_gtod_data.lock, flags);
- }
-
-Index: linux-3.0/arch/x86/include/asm/vgtod.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/vgtod.h
-+++ linux-3.0/arch/x86/include/asm/vgtod.h
-@@ -5,7 +5,7 @@
- #include <linux/clocksource.h>
-
- struct vsyscall_gtod_data {
-- seqlock_t lock;
-+ raw_seqlock_t lock;
-
- /* open coded 'struct timespec' */
- time_t wall_time_sec;
-Index: linux-3.0/arch/x86/kernel/vsyscall_64.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/vsyscall_64.c
-+++ linux-3.0/arch/x86/kernel/vsyscall_64.c
-@@ -52,7 +52,7 @@
- DEFINE_VVAR(int, vgetcpu_mode);
- DEFINE_VVAR(struct vsyscall_gtod_data, vsyscall_gtod_data) =
- {
-- .lock = __SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock),
-+ .lock = __RAW_SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock),
- .sysctl_enabled = 1,
- };
-
-@@ -60,10 +60,10 @@ void update_vsyscall_tz(void)
- {
- unsigned long flags;
-
-- write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
-+ raw_write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
- /* sys_tz has changed */
- vsyscall_gtod_data.sys_tz = sys_tz;
-- write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
-+ raw_write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
- }
-
- void update_vsyscall(struct timespec *wall_time, struct timespec *wtm,
-@@ -71,7 +71,7 @@ void update_vsyscall(struct timespec *wa
- {
- unsigned long flags;
-
-- write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
-+ raw_write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags);
- /* copy vsyscall data */
- vsyscall_gtod_data.clock.vread = clock->vread;
- vsyscall_gtod_data.clock.cycle_last = clock->cycle_last;
-@@ -82,7 +82,7 @@ void update_vsyscall(struct timespec *wa
- vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec;
- vsyscall_gtod_data.wall_to_monotonic = *wtm;
- vsyscall_gtod_data.wall_time_coarse = __current_kernel_time();
-- write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
-+ raw_write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags);
- }
-
- /* RED-PEN may want to readd seq locking, but then the variable should be
-Index: linux-3.0/kernel/time/ntp.c
-===================================================================
---- linux-3.0.orig/kernel/time/ntp.c
-+++ linux-3.0/kernel/time/ntp.c
-@@ -358,7 +358,7 @@ static enum hrtimer_restart ntp_leap_sec
- {
- enum hrtimer_restart res = HRTIMER_NORESTART;
-
-- write_seqlock(&xtime_lock);
-+ raw_write_seqlock(&xtime_lock);
-
- switch (time_state) {
- case TIME_OK:
-@@ -388,7 +388,7 @@ static enum hrtimer_restart ntp_leap_sec
- break;
- }
-
-- write_sequnlock(&xtime_lock);
-+ raw_write_sequnlock(&xtime_lock);
-
- return res;
- }
-@@ -663,7 +663,7 @@ int do_adjtimex(struct timex *txc)
-
- getnstimeofday(&ts);
-
-- write_seqlock_irq(&xtime_lock);
-+ raw_write_seqlock_irq(&xtime_lock);
-
- if (txc->modes & ADJ_ADJTIME) {
- long save_adjust = time_adjust;
-@@ -705,7 +705,7 @@ int do_adjtimex(struct timex *txc)
- /* fill PPS status fields */
- pps_fill_timex(txc);
-
-- write_sequnlock_irq(&xtime_lock);
-+ raw_write_sequnlock_irq(&xtime_lock);
-
- txc->time.tv_sec = ts.tv_sec;
- txc->time.tv_usec = ts.tv_nsec;
-@@ -903,7 +903,7 @@ void hardpps(const struct timespec *phas
-
- pts_norm = pps_normalize_ts(*phase_ts);
-
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
-
- /* clear the error bits, they will be set again if needed */
- time_status &= ~(STA_PPSJITTER | STA_PPSWANDER | STA_PPSERROR);
-@@ -916,7 +916,7 @@ void hardpps(const struct timespec *phas
- * just start the frequency interval */
- if (unlikely(pps_fbase.tv_sec == 0)) {
- pps_fbase = *raw_ts;
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
- return;
- }
-
-@@ -931,7 +931,7 @@ void hardpps(const struct timespec *phas
- time_status |= STA_PPSJITTER;
- /* restart the frequency calibration interval */
- pps_fbase = *raw_ts;
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
- pr_err("hardpps: PPSJITTER: bad pulse\n");
- return;
- }
-@@ -948,7 +948,7 @@ void hardpps(const struct timespec *phas
-
- hardpps_update_phase(pts_norm.nsec);
-
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
- }
- EXPORT_SYMBOL(hardpps);
-
-Index: linux-3.0/kernel/time/tick-common.c
-===================================================================
---- linux-3.0.orig/kernel/time/tick-common.c
-+++ linux-3.0/kernel/time/tick-common.c
-@@ -63,13 +63,13 @@ int tick_is_oneshot_available(void)
- static void tick_periodic(int cpu)
- {
- if (tick_do_timer_cpu == cpu) {
-- write_seqlock(&xtime_lock);
-+ raw_write_seqlock(&xtime_lock);
-
- /* Keep track of the next tick event */
- tick_next_period = ktime_add(tick_next_period, tick_period);
-
- do_timer(1);
-- write_sequnlock(&xtime_lock);
-+ raw_write_sequnlock(&xtime_lock);
- }
-
- update_process_times(user_mode(get_irq_regs()));
-Index: linux-3.0/kernel/time/tick-internal.h
-===================================================================
---- linux-3.0.orig/kernel/time/tick-internal.h
-+++ linux-3.0/kernel/time/tick-internal.h
-@@ -143,4 +143,4 @@ static inline int tick_device_is_functio
- #endif
-
- extern void do_timer(unsigned long ticks);
--extern seqlock_t xtime_lock;
-+extern raw_seqlock_t xtime_lock;
-Index: linux-3.0/kernel/time/tick-sched.c
-===================================================================
---- linux-3.0.orig/kernel/time/tick-sched.c
-+++ linux-3.0/kernel/time/tick-sched.c
-@@ -56,7 +56,7 @@ static void tick_do_update_jiffies64(kti
- return;
-
- /* Reevalute with xtime_lock held */
-- write_seqlock(&xtime_lock);
-+ raw_write_seqlock(&xtime_lock);
-
- delta = ktime_sub(now, last_jiffies_update);
- if (delta.tv64 >= tick_period.tv64) {
-@@ -79,7 +79,7 @@ static void tick_do_update_jiffies64(kti
- /* Keep the tick_next_period variable up to date */
- tick_next_period = ktime_add(last_jiffies_update, tick_period);
- }
-- write_sequnlock(&xtime_lock);
-+ raw_write_sequnlock(&xtime_lock);
- }
-
- /*
-@@ -89,12 +89,12 @@ static ktime_t tick_init_jiffy_update(vo
- {
- ktime_t period;
-
-- write_seqlock(&xtime_lock);
-+ raw_write_seqlock(&xtime_lock);
- /* Did we start the jiffies update yet ? */
- if (last_jiffies_update.tv64 == 0)
- last_jiffies_update = tick_next_period;
- period = last_jiffies_update;
-- write_sequnlock(&xtime_lock);
-+ raw_write_sequnlock(&xtime_lock);
- return period;
- }
-
-@@ -304,13 +304,7 @@ void tick_nohz_stop_sched_tick(int inidl
- goto end;
-
- if (unlikely(local_softirq_pending() && cpu_online(cpu))) {
-- static int ratelimit;
--
-- if (ratelimit < 10) {
-- printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
-- (unsigned int) local_softirq_pending());
-- ratelimit++;
-- }
-+ softirq_check_pending_idle();
- goto end;
- }
-
-@@ -777,6 +771,7 @@ void tick_setup_sched_timer(void)
- * Emulate tick processing via per-CPU hrtimers:
- */
- hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
-+ ts->sched_timer.irqsafe = 1;
- ts->sched_timer.function = tick_sched_timer;
-
- /* Get the next period (per cpu) */
-Index: linux-3.0/kernel/time/timekeeping.c
-===================================================================
---- linux-3.0.orig/kernel/time/timekeeping.c
-+++ linux-3.0/kernel/time/timekeeping.c
-@@ -139,8 +139,7 @@ static inline s64 timekeeping_get_ns_raw
- * This read-write spinlock protects us from races in SMP while
- * playing with xtime.
- */
--__cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock);
--
-+__cacheline_aligned_in_smp DEFINE_RAW_SEQLOCK(xtime_lock);
-
++
+ #define pte_lock_deinit(page) ((page)->mapping = NULL)
++
++#else /* !PREEMPT_RT_FULL */
++
++/*
++ * On PREEMPT_RT_FULL the spinlock_t's are too large to embed in the
++ * page frame, hence it only has a pointer and we need to dynamically
++ * allocate the lock when we allocate PTE-pages.
++ *
++ * This is an overall win, since only a small fraction of the pages
++ * will be PTE pages under normal circumstances.
++ */
++
++#define __pte_lockptr(page) ((page)->ptl)
++
++extern struct page *pte_lock_init(struct page *page);
++extern void pte_lock_deinit(struct page *page);
++
++#endif /* PREEMPT_RT_FULL */
++
+ #define pte_lockptr(mm, pmd) ({(void)(mm); __pte_lockptr(pmd_page(*(pmd)));})
+ #else /* !USE_SPLIT_PTLOCKS */
/*
- * The current time
-@@ -361,7 +360,7 @@ int do_settimeofday(const struct timespe
- if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
- return -EINVAL;
-
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
-
- timekeeping_forward_now();
-
-@@ -377,7 +376,7 @@ int do_settimeofday(const struct timespe
- update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
- timekeeper.mult);
-
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
-
- /* signal hrtimers about time change */
- clock_was_set();
-@@ -401,7 +400,7 @@ int timekeeping_inject_offset(struct tim
- if ((unsigned long)ts->tv_nsec >= NSEC_PER_SEC)
- return -EINVAL;
-
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
-
- timekeeping_forward_now();
-
-@@ -414,7 +413,7 @@ int timekeeping_inject_offset(struct tim
- update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
- timekeeper.mult);
-
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
-
- /* signal hrtimers about time change */
- clock_was_set();
-@@ -568,7 +567,7 @@ void __init timekeeping_init(void)
- read_persistent_clock(&now);
- read_boot_clock(&boot);
-
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
-
- ntp_init();
+ * We use mm->page_table_lock to guard all pagetable pages of the mm.
+ */
+-#define pte_lock_init(page) do {} while (0)
++static inline struct page *pte_lock_init(struct page *page) { return page; }
+ #define pte_lock_deinit(page) do {} while (0)
+ #define pte_lockptr(mm, pmd) ({(void)(pmd); &(mm)->page_table_lock;})
+ #endif /* USE_SPLIT_PTLOCKS */
-@@ -589,7 +588,7 @@ void __init timekeeping_init(void)
- -boot.tv_sec, -boot.tv_nsec);
- total_sleep_time.tv_sec = 0;
- total_sleep_time.tv_nsec = 0;
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+-static inline void pgtable_page_ctor(struct page *page)
++static inline struct page *__pgtable_page_ctor(struct page *page)
+ {
+- pte_lock_init(page);
+- inc_zone_page_state(page, NR_PAGETABLE);
++ page = pte_lock_init(page);
++ if (page)
++ inc_zone_page_state(page, NR_PAGETABLE);
++ return page;
}
- /* time in seconds when suspend began */
-@@ -630,7 +629,7 @@ void timekeeping_inject_sleeptime(struct
- if (!(ts.tv_sec == 0 && ts.tv_nsec == 0))
- return;
-
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
- timekeeping_forward_now();
-
- __timekeeping_inject_sleeptime(delta);
-@@ -640,7 +639,7 @@ void timekeeping_inject_sleeptime(struct
- update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
- timekeeper.mult);
++#define pgtable_page_ctor(page) \
++do { \
++ page = __pgtable_page_ctor(page); \
++} while (0)
++
+ static inline void pgtable_page_dtor(struct page *page)
+ {
+ pte_lock_deinit(page);
+diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
+index 059839c..8708014 100644
+--- a/include/linux/mm_types.h
++++ b/include/linux/mm_types.h
+@@ -12,6 +12,7 @@
+ #include <linux/completion.h>
+ #include <linux/cpumask.h>
+ #include <linux/page-debug-flags.h>
++#include <linux/rcupdate.h>
+ #include <asm/page.h>
+ #include <asm/mmu.h>
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+@@ -77,7 +78,11 @@ struct page {
+ */
+ };
+ #if USE_SPLIT_PTLOCKS
++#ifndef CONFIG_PREEMPT_RT_FULL
+ spinlock_t ptl;
++#else
++ spinlock_t *ptl;
++#endif
+ #endif
+ struct kmem_cache *slab; /* SLUB: Pointer to slab */
+ struct page *first_page; /* Compound tail pages */
+@@ -330,6 +335,9 @@ struct mm_struct {
+ #ifdef CONFIG_CPUMASK_OFFSTACK
+ struct cpumask cpumask_allocation;
+ #endif
++#ifdef CONFIG_PREEMPT_RT_BASE
++ struct rcu_head delayed_drop;
++#endif
+ };
- /* signal hrtimers about time change */
- clock_was_set();
-@@ -663,7 +662,7 @@ static void timekeeping_resume(void)
+ static inline void mm_init_cpumask(struct mm_struct *mm)
+diff --git a/include/linux/mutex.h b/include/linux/mutex.h
+index a940fe4..30bd8f2 100644
+--- a/include/linux/mutex.h
++++ b/include/linux/mutex.h
+@@ -17,6 +17,17 @@
- clocksource_resume();
+ #include <asm/atomic.h>
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \
++ , .dep_map = { .name = #lockname }
++#else
++# define __DEP_MAP_MUTEX_INITIALIZER(lockname)
++#endif
++
++#ifdef CONFIG_PREEMPT_RT_FULL
++#include <linux/mutex_rt.h>
++#else
++
+ /*
+ * Simple, straightforward mutexes with strict semantics:
+ *
+@@ -95,13 +106,6 @@ do { \
+ # define mutex_destroy(mutex) do { } while (0)
+ #endif
- if (timespec_compare(&ts, &timekeeping_suspend_time) > 0) {
- ts = timespec_sub(ts, timekeeping_suspend_time);
-@@ -673,7 +672,7 @@ static void timekeeping_resume(void)
- timekeeper.clock->cycle_last = timekeeper.clock->read(timekeeper.clock);
- timekeeper.ntp_error = 0;
- timekeeping_suspended = 0;
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+-#ifdef CONFIG_DEBUG_LOCK_ALLOC
+-# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \
+- , .dep_map = { .name = #lockname }
+-#else
+-# define __DEP_MAP_MUTEX_INITIALIZER(lockname)
+-#endif
+-
+ #define __MUTEX_INITIALIZER(lockname) \
+ { .count = ATOMIC_INIT(1) \
+ , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \
+@@ -167,6 +171,9 @@ extern int __must_check mutex_lock_killable(struct mutex *lock);
+ */
+ extern int mutex_trylock(struct mutex *lock);
+ extern void mutex_unlock(struct mutex *lock);
++
++#endif /* !PREEMPT_RT_FULL */
++
+ extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock);
- touch_softlockup_watchdog();
+ #ifndef CONFIG_HAVE_ARCH_MUTEX_CPU_RELAX
+diff --git a/include/linux/mutex_rt.h b/include/linux/mutex_rt.h
+new file mode 100644
+index 0000000..c38a44b
+--- /dev/null
++++ b/include/linux/mutex_rt.h
+@@ -0,0 +1,84 @@
++#ifndef __LINUX_MUTEX_RT_H
++#define __LINUX_MUTEX_RT_H
++
++#ifndef __LINUX_MUTEX_H
++#error "Please include mutex.h"
++#endif
++
++#include <linux/rtmutex.h>
++
++/* FIXME: Just for __lockfunc */
++#include <linux/spinlock.h>
++
++struct mutex {
++ struct rt_mutex lock;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++};
++
++#define __MUTEX_INITIALIZER(mutexname) \
++ { \
++ .lock = __RT_MUTEX_INITIALIZER(mutexname.lock) \
++ __DEP_MAP_MUTEX_INITIALIZER(mutexname) \
++ }
++
++#define DEFINE_MUTEX(mutexname) \
++ struct mutex mutexname = __MUTEX_INITIALIZER(mutexname)
++
++extern void __mutex_do_init(struct mutex *lock, const char *name, struct lock_class_key *key);
++extern void __lockfunc _mutex_lock(struct mutex *lock);
++extern int __lockfunc _mutex_lock_interruptible(struct mutex *lock);
++extern int __lockfunc _mutex_lock_killable(struct mutex *lock);
++extern void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass);
++extern void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest_lock);
++extern int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass);
++extern int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass);
++extern int __lockfunc _mutex_trylock(struct mutex *lock);
++extern void __lockfunc _mutex_unlock(struct mutex *lock);
++
++#define mutex_is_locked(l) rt_mutex_is_locked(&(l)->lock)
++#define mutex_lock(l) _mutex_lock(l)
++#define mutex_lock_interruptible(l) _mutex_lock_interruptible(l)
++#define mutex_lock_killable(l) _mutex_lock_killable(l)
++#define mutex_trylock(l) _mutex_trylock(l)
++#define mutex_unlock(l) _mutex_unlock(l)
++#define mutex_destroy(l) rt_mutex_destroy(&(l)->lock)
++
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define mutex_lock_nested(l, s) _mutex_lock_nested(l, s)
++# define mutex_lock_interruptible_nested(l, s) \
++ _mutex_lock_interruptible_nested(l, s)
++# define mutex_lock_killable_nested(l, s) \
++ _mutex_lock_killable_nested(l, s)
++
++# define mutex_lock_nest_lock(lock, nest_lock) \
++do { \
++ typecheck(struct lockdep_map *, &(nest_lock)->dep_map); \
++ _mutex_lock_nest_lock(lock, &(nest_lock)->dep_map); \
++} while (0)
++
++#else
++# define mutex_lock_nested(l, s) _mutex_lock(l)
++# define mutex_lock_interruptible_nested(l, s) \
++ _mutex_lock_interruptible(l)
++# define mutex_lock_killable_nested(l, s) \
++ _mutex_lock_killable(l)
++# define mutex_lock_nest_lock(lock, nest_lock) mutex_lock(lock)
++#endif
++
++# define mutex_init(mutex) \
++do { \
++ static struct lock_class_key __key; \
++ \
++ rt_mutex_init(&(mutex)->lock); \
++ __mutex_do_init((mutex), #mutex, &__key); \
++} while (0)
++
++# define __mutex_init(mutex, name, key) \
++do { \
++ rt_mutex_init(&(mutex)->lock); \
++ __mutex_do_init((mutex), name, key); \
++} while (0)
++
++#endif
+diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
+index 33b5968..0decca2 100644
+--- a/include/linux/netdevice.h
++++ b/include/linux/netdevice.h
+@@ -1764,6 +1764,7 @@ struct softnet_data {
+ unsigned dropped;
+ struct sk_buff_head input_pkt_queue;
+ struct napi_struct backlog;
++ struct sk_buff_head tofree_queue;
+ };
-@@ -689,10 +688,10 @@ static int timekeeping_suspend(void)
+ static inline void input_queue_head_incr(struct softnet_data *sd)
+diff --git a/include/linux/of.h b/include/linux/of.h
+index bfc0ed1..00ddacd 100644
+--- a/include/linux/of.h
++++ b/include/linux/of.h
+@@ -68,7 +68,7 @@ struct device_node {
+ /* Pointer for first entry in chain of all nodes. */
+ extern struct device_node *allnodes;
+ extern struct device_node *of_chosen;
+-extern rwlock_t devtree_lock;
++extern raw_spinlock_t devtree_lock;
- read_persistent_clock(&timekeeping_suspend_time);
+ static inline bool of_have_populated_dt(void)
+ {
+diff --git a/include/linux/oprofile.h b/include/linux/oprofile.h
+index 7f5cfd3..fdb984f 100644
+--- a/include/linux/oprofile.h
++++ b/include/linux/oprofile.h
+@@ -166,7 +166,7 @@ ssize_t oprofilefs_ulong_to_user(unsigned long val, char __user * buf, size_t co
+ int oprofilefs_ulong_from_user(unsigned long * val, char const __user * buf, size_t count);
-- write_seqlock_irqsave(&xtime_lock, flags);
-+ raw_write_seqlock_irqsave(&xtime_lock, flags);
- timekeeping_forward_now();
- timekeeping_suspended = 1;
-- write_sequnlock_irqrestore(&xtime_lock, flags);
-+ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+ /** lock for read/write safety */
+-extern spinlock_t oprofilefs_lock;
++extern raw_spinlock_t oprofilefs_lock;
- clockevents_notify(CLOCK_EVT_NOTIFY_SUSPEND, NULL);
- clocksource_suspend();
-@@ -1121,7 +1120,7 @@ ktime_t ktime_get_monotonic_offset(void)
+ /**
+ * Add the contents of a circular buffer to the event buffer.
+diff --git a/include/linux/page_cgroup.h b/include/linux/page_cgroup.h
+index 961ecc7..2927c08 100644
+--- a/include/linux/page_cgroup.h
++++ b/include/linux/page_cgroup.h
+@@ -30,6 +30,10 @@ enum {
*/
- void xtime_update(unsigned long ticks)
- {
-- write_seqlock(&xtime_lock);
-+ raw_write_seqlock(&xtime_lock);
- do_timer(ticks);
-- write_sequnlock(&xtime_lock);
-+ raw_write_sequnlock(&xtime_lock);
+ struct page_cgroup {
+ unsigned long flags;
++#ifdef CONFIG_PREEMPT_RT_BASE
++ spinlock_t pcg_lock;
++ spinlock_t pcm_lock;
++#endif
+ struct mem_cgroup *mem_cgroup;
+ struct list_head lru; /* per cgroup LRU list */
+ };
+@@ -96,30 +100,54 @@ static inline void lock_page_cgroup(struct page_cgroup *pc)
+ * Don't take this lock in IRQ context.
+ * This lock is for pc->mem_cgroup, USED, CACHE, MIGRATION
+ */
++#ifndef CONFIG_PREEMPT_RT_BASE
+ bit_spin_lock(PCG_LOCK, &pc->flags);
++#else
++ spin_lock(&pc->pcg_lock);
++#endif
}
-Index: linux-3.0/arch/alpha/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/alpha/include/asm/rwsem.h
-+++ linux-3.0/arch/alpha/include/asm/rwsem.h
-@@ -21,7 +21,7 @@
- #define RWSEM_ACTIVE_READ_BIAS RWSEM_ACTIVE_BIAS
- #define RWSEM_ACTIVE_WRITE_BIAS (RWSEM_WAITING_BIAS + RWSEM_ACTIVE_BIAS)
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- long oldcount;
- #ifndef CONFIG_SMP
-@@ -48,7 +48,7 @@ static inline void __down_read(struct rw
- /*
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ static inline void unlock_page_cgroup(struct page_cgroup *pc)
{
- long old, new, res;
-
-@@ -63,7 +63,7 @@ static inline int __down_read_trylock(st
- return res >= 0 ? 1 : 0;
++#ifndef CONFIG_PREEMPT_RT_BASE
+ bit_spin_unlock(PCG_LOCK, &pc->flags);
++#else
++ spin_unlock(&pc->pcg_lock);
++#endif
}
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- long oldcount;
- #ifndef CONFIG_SMP
-@@ -90,7 +90,7 @@ static inline void __down_write(struct r
- /*
- * trylock for writing -- returns 1 if successful, 0 if contention
- */
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
+ static inline void move_lock_page_cgroup(struct page_cgroup *pc,
+ unsigned long *flags)
{
- long ret = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
- RWSEM_ACTIVE_WRITE_BIAS);
-@@ -99,7 +99,7 @@ static inline int __down_write_trylock(s
- return 0;
++#ifndef CONFIG_PREEMPT_RT_BASE
+ /*
+ * We know updates to pc->flags of page cache's stats are from both of
+ * usual context or IRQ context. Disable IRQ to avoid deadlock.
+ */
+ local_irq_save(*flags);
+ bit_spin_lock(PCG_MOVE_LOCK, &pc->flags);
++#else
++ spin_lock_irqsave(&pc->pcm_lock, *flags);
++#endif
}
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
+ static inline void move_unlock_page_cgroup(struct page_cgroup *pc,
+ unsigned long *flags)
{
- long oldcount;
- #ifndef CONFIG_SMP
-@@ -124,7 +124,7 @@ static inline void __up_read(struct rw_s
- rwsem_wake(sem);
++#ifndef CONFIG_PREEMPT_RT_BASE
+ bit_spin_unlock(PCG_MOVE_LOCK, &pc->flags);
+ local_irq_restore(*flags);
++#else
++ spin_unlock_irqrestore(&pc->pcm_lock, *flags);
++#endif
++}
++
++static inline void page_cgroup_lock_init(struct page_cgroup *pc)
++{
++#ifdef CONFIG_PREEMPT_RT_BASE
++ spin_lock_init(&pc->pcg_lock);
++ spin_lock_init(&pc->pcm_lock);
++#endif
}
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- long count;
- #ifndef CONFIG_SMP
-@@ -153,7 +153,7 @@ static inline void __up_write(struct rw_
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- long oldcount;
- #ifndef CONFIG_SMP
-@@ -177,7 +177,7 @@ static inline void __downgrade_write(str
- rwsem_downgrade_wake(sem);
- }
+ #ifdef CONFIG_SPARSEMEM
+diff --git a/include/linux/percpu.h b/include/linux/percpu.h
+index 9ca008f..58163cd 100644
+--- a/include/linux/percpu.h
++++ b/include/linux/percpu.h
+@@ -48,6 +48,31 @@
+ preempt_enable(); \
+ } while (0)
--static inline void rwsem_atomic_add(long val, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(long val, struct rw_anon_semaphore *sem)
- {
- #ifndef CONFIG_SMP
- sem->count += val;
-@@ -196,7 +196,7 @@ static inline void rwsem_atomic_add(long
- #endif
- }
++#ifndef CONFIG_PREEMPT_RT_FULL
++# define get_local_var(var) get_cpu_var(var)
++# define put_local_var(var) put_cpu_var(var)
++# define get_local_ptr(var) get_cpu_ptr(var)
++# define put_local_ptr(var) put_cpu_ptr(var)
++#else
++# define get_local_var(var) (*({ \
++ migrate_disable(); \
++ &__get_cpu_var(var); }))
++
++# define put_local_var(var) do { \
++ (void)&(var); \
++ migrate_enable(); \
++} while (0)
++
++# define get_local_ptr(var) ({ \
++ migrate_disable(); \
++ this_cpu_ptr(var); })
++
++# define put_local_ptr(var) do { \
++ (void)(var); \
++ migrate_enable(); \
++} while (0)
++#endif
++
+ /* minimum unit size, also is the maximum supported allocation size */
+ #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10)
--static inline long rwsem_atomic_update(long val, struct rw_semaphore *sem)
-+static inline long rwsem_atomic_update(long val, struct rw_anon_semaphore *sem)
- {
- #ifndef CONFIG_SMP
- sem->count += val;
-Index: linux-3.0/arch/ia64/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/ia64/include/asm/rwsem.h
-+++ linux-3.0/arch/ia64/include/asm/rwsem.h
-@@ -38,7 +38,7 @@
- * lock for reading
- */
- static inline void
--__down_read (struct rw_semaphore *sem)
-+__down_read (struct rw_anon_semaphore *sem)
- {
- long result = ia64_fetchadd8_acq((unsigned long *)&sem->count, 1);
+diff --git a/include/linux/percpu_counter.h b/include/linux/percpu_counter.h
+index 5edc901..b9df9ed 100644
+--- a/include/linux/percpu_counter.h
++++ b/include/linux/percpu_counter.h
+@@ -16,7 +16,7 @@
+ #ifdef CONFIG_SMP
-@@ -50,7 +50,7 @@ __down_read (struct rw_semaphore *sem)
- * lock for writing
- */
- static inline void
--__down_write (struct rw_semaphore *sem)
-+__down_write (struct rw_anon_semaphore *sem)
- {
- long old, new;
+ struct percpu_counter {
+- spinlock_t lock;
++ raw_spinlock_t lock;
+ s64 count;
+ #ifdef CONFIG_HOTPLUG_CPU
+ struct list_head list; /* All percpu_counters are on a list */
+diff --git a/include/linux/pid.h b/include/linux/pid.h
+index b152d44..7f33683 100644
+--- a/include/linux/pid.h
++++ b/include/linux/pid.h
+@@ -2,6 +2,7 @@
+ #define _LINUX_PID_H
-@@ -67,7 +67,7 @@ __down_write (struct rw_semaphore *sem)
- * unlock after reading
- */
- static inline void
--__up_read (struct rw_semaphore *sem)
-+__up_read (struct rw_anon_semaphore *sem)
- {
- long result = ia64_fetchadd8_rel((unsigned long *)&sem->count, -1);
+ #include <linux/rcupdate.h>
++#include <linux/atomic.h>
-@@ -79,7 +79,7 @@ __up_read (struct rw_semaphore *sem)
- * unlock after writing
- */
- static inline void
--__up_write (struct rw_semaphore *sem)
-+__up_write (struct rw_anon_semaphore *sem)
+ enum pid_type
{
- long old, new;
+diff --git a/include/linux/plist.h b/include/linux/plist.h
+index c9b9f32..66ba13c 100644
+--- a/include/linux/plist.h
++++ b/include/linux/plist.h
+@@ -77,13 +77,15 @@
-@@ -96,7 +96,7 @@ __up_write (struct rw_semaphore *sem)
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
- static inline int
--__down_read_trylock (struct rw_semaphore *sem)
-+__down_read_trylock (struct rw_anon_semaphore *sem)
- {
- long tmp;
- while ((tmp = sem->count) >= 0) {
-@@ -111,7 +111,7 @@ __down_read_trylock (struct rw_semaphore
- * trylock for writing -- returns 1 if successful, 0 if contention
+ #include <linux/kernel.h>
+ #include <linux/list.h>
+-#include <linux/spinlock_types.h>
++
++struct raw_spinlock;
++struct spinlock;
+
+ struct plist_head {
+ struct list_head node_list;
+ #ifdef CONFIG_DEBUG_PI_LIST
+- raw_spinlock_t *rawlock;
+- spinlock_t *spinlock;
++ struct raw_spinlock *rawlock;
++ struct spinlock *spinlock;
+ #endif
+ };
+
+@@ -144,7 +146,7 @@ struct plist_node {
+ * @lock: spinlock protecting the list (debugging)
*/
- static inline int
--__down_write_trylock (struct rw_semaphore *sem)
-+__down_write_trylock (struct rw_anon_semaphore *sem)
+ static inline void
+-plist_head_init(struct plist_head *head, spinlock_t *lock)
++plist_head_init(struct plist_head *head, struct spinlock *lock)
{
- long tmp = cmpxchg_acq(&sem->count, RWSEM_UNLOCKED_VALUE,
- RWSEM_ACTIVE_WRITE_BIAS);
-@@ -122,7 +122,7 @@ __down_write_trylock (struct rw_semaphor
- * downgrade write lock to read lock
+ INIT_LIST_HEAD(&head->node_list);
+ #ifdef CONFIG_DEBUG_PI_LIST
+@@ -159,7 +161,7 @@ plist_head_init(struct plist_head *head, spinlock_t *lock)
+ * @lock: raw_spinlock protecting the list (debugging)
*/
static inline void
--__downgrade_write (struct rw_semaphore *sem)
-+__downgrade_write (struct rw_anon_semaphore *sem)
+-plist_head_init_raw(struct plist_head *head, raw_spinlock_t *lock)
++plist_head_init_raw(struct plist_head *head, struct raw_spinlock *lock)
{
- long old, new;
+ INIT_LIST_HEAD(&head->node_list);
+ #ifdef CONFIG_DEBUG_PI_LIST
+diff --git a/include/linux/preempt.h b/include/linux/preempt.h
+index 2e681d9..7a2bbbb 100644
+--- a/include/linux/preempt.h
++++ b/include/linux/preempt.h
+@@ -33,12 +33,18 @@ do { \
+ barrier(); \
+ } while (0)
-Index: linux-3.0/arch/powerpc/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/powerpc/include/asm/rwsem.h
-+++ linux-3.0/arch/powerpc/include/asm/rwsem.h
-@@ -31,13 +31,13 @@
- /*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- if (unlikely(atomic_long_inc_return((atomic_long_t *)&sem->count) <= 0))
- rwsem_down_read_failed(sem);
- }
+-#define preempt_enable_no_resched() \
++#define __preempt_enable_no_resched() \
+ do { \
+ barrier(); \
+ dec_preempt_count(); \
+ } while (0)
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- long tmp;
++#ifndef CONFIG_PREEMPT_RT_BASE
++# define preempt_enable_no_resched() __preempt_enable_no_resched()
++#else
++# define preempt_enable_no_resched() preempt_enable()
++#endif
++
+ #define preempt_check_resched() \
+ do { \
+ if (unlikely(test_thread_flag(TIF_NEED_RESCHED))) \
+@@ -47,7 +53,7 @@ do { \
-@@ -53,7 +53,7 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
- */
--static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
-+static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- long tmp;
+ #define preempt_enable() \
+ do { \
+- preempt_enable_no_resched(); \
++ __preempt_enable_no_resched(); \
+ barrier(); \
+ preempt_check_resched(); \
+ } while (0)
+@@ -83,6 +89,7 @@ do { \
+ #else
-@@ -63,12 +63,12 @@ static inline void __down_write_nested(s
- rwsem_down_write_failed(sem);
- }
+ #define preempt_disable() do { } while (0)
++#define __preempt_enable_no_resched() do { } while (0)
+ #define preempt_enable_no_resched() do { } while (0)
+ #define preempt_enable() do { } while (0)
+ #define preempt_check_resched() do { } while (0)
+@@ -93,6 +100,27 @@ do { \
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- __down_write_nested(sem, 0);
+ #endif
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++# define preempt_disable_rt() preempt_disable()
++# define preempt_enable_rt() preempt_enable()
++# define preempt_disable_nort() do { } while (0)
++# define preempt_enable_nort() do { } while (0)
++# ifdef CONFIG_SMP
++ extern void migrate_disable(void);
++ extern void migrate_enable(void);
++# else /* CONFIG_SMP */
++# define migrate_disable() do { } while (0)
++# define migrate_enable() do { } while (0)
++# endif /* CONFIG_SMP */
++#else
++# define preempt_disable_rt() do { } while (0)
++# define preempt_enable_rt() do { } while (0)
++# define preempt_disable_nort() preempt_disable()
++# define preempt_enable_nort() preempt_enable()
++# define migrate_disable() preempt_disable()
++# define migrate_enable() preempt_enable()
++#endif
++
+ #ifdef CONFIG_PREEMPT_NOTIFIERS
+
+ struct preempt_notifier;
+diff --git a/include/linux/printk.h b/include/linux/printk.h
+index 0101d55..bf508b6 100644
+--- a/include/linux/printk.h
++++ b/include/linux/printk.h
+@@ -88,8 +88,15 @@ int no_printk(const char *fmt, ...)
+ return 0;
}
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- long tmp;
++#ifdef CONFIG_EARLY_PRINTK
+ extern asmlinkage __attribute__ ((format (printf, 1, 2)))
+ void early_printk(const char *fmt, ...);
++extern void printk_kill(void);
++#else
++static inline __attribute__ ((format (printf, 1, 2))) __cold
++void early_printk(const char *s, ...) { }
++static inline void printk_kill(void) { }
++#endif
-@@ -80,7 +80,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
- {
- long tmp;
+ extern int printk_needs_cpu(int cpu);
+ extern void printk_tick(void);
+@@ -109,7 +116,6 @@ extern int __printk_ratelimit(const char *func);
+ #define printk_ratelimit() __printk_ratelimit(__func__)
+ extern bool printk_timed_ratelimit(unsigned long *caller_jiffies,
+ unsigned int interval_msec);
+-
+ extern int printk_delay_msec;
+ extern int dmesg_restrict;
+ extern int kptr_restrict;
+diff --git a/include/linux/proportions.h b/include/linux/proportions.h
+index cf793bb..ef35bb7 100644
+--- a/include/linux/proportions.h
++++ b/include/linux/proportions.h
+@@ -58,7 +58,7 @@ struct prop_local_percpu {
+ */
+ int shift;
+ unsigned long period;
+- spinlock_t lock; /* protect the snapshot state */
++ raw_spinlock_t lock; /* protect the snapshot state */
+ };
-@@ -92,7 +92,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
- */
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- if (unlikely(atomic_long_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
- (atomic_long_t *)&sem->count) < 0))
-@@ -102,7 +102,7 @@ static inline void __up_write(struct rw_
- /*
- * implement atomic add functionality
- */
--static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
- {
- atomic_long_add(delta, (atomic_long_t *)&sem->count);
+ int prop_local_init_percpu(struct prop_local_percpu *pl);
+@@ -106,11 +106,11 @@ struct prop_local_single {
+ */
+ unsigned long period;
+ int shift;
+- spinlock_t lock; /* protect the snapshot state */
++ raw_spinlock_t lock; /* protect the snapshot state */
+ };
+
+ #define INIT_PROP_LOCAL_SINGLE(name) \
+-{ .lock = __SPIN_LOCK_UNLOCKED(name.lock), \
++{ .lock = __RAW_SPIN_LOCK_UNLOCKED(name.lock), \
}
-@@ -110,7 +110,7 @@ static inline void rwsem_atomic_add(long
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- long tmp;
-@@ -123,7 +123,7 @@ static inline void __downgrade_write(str
- /*
- * implement exchange and add functionality
- */
--static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
-+static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
+ int prop_local_init_single(struct prop_local_single *pl);
+diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h
+index 23241c2..7d36716 100644
+--- a/include/linux/radix-tree.h
++++ b/include/linux/radix-tree.h
+@@ -201,7 +201,13 @@ unsigned long radix_tree_next_hole(struct radix_tree_root *root,
+ unsigned long index, unsigned long max_scan);
+ unsigned long radix_tree_prev_hole(struct radix_tree_root *root,
+ unsigned long index, unsigned long max_scan);
++
++#ifndef CONFIG_PREEMPT_RT_FULL
+ int radix_tree_preload(gfp_t gfp_mask);
++#else
++static inline int radix_tree_preload(gfp_t gm) { return 0; }
++#endif
++
+ void radix_tree_init(void);
+ void *radix_tree_tag_set(struct radix_tree_root *root,
+ unsigned long index, unsigned int tag);
+@@ -225,7 +231,7 @@ int radix_tree_tagged(struct radix_tree_root *root, unsigned int tag);
+
+ static inline void radix_tree_preload_end(void)
{
- return atomic_long_add_return(delta, (atomic_long_t *)&sem->count);
+- preempt_enable();
++ preempt_enable_nort();
}
-Index: linux-3.0/arch/s390/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/s390/include/asm/rwsem.h
-+++ linux-3.0/arch/s390/include/asm/rwsem.h
-@@ -60,7 +60,7 @@
- /*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- signed long old, new;
-@@ -88,7 +88,7 @@ static inline void __down_read(struct rw
- /*
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
+ #endif /* _LINUX_RADIX_TREE_H */
+diff --git a/include/linux/ratelimit.h b/include/linux/ratelimit.h
+index 2f00715..e11ccb4 100644
+--- a/include/linux/ratelimit.h
++++ b/include/linux/ratelimit.h
+@@ -8,7 +8,7 @@
+ #define DEFAULT_RATELIMIT_BURST 10
+
+ struct ratelimit_state {
+- spinlock_t lock; /* protect the state */
++ raw_spinlock_t lock; /* protect the state */
+
+ int interval;
+ int burst;
+@@ -20,7 +20,7 @@ struct ratelimit_state {
+ #define DEFINE_RATELIMIT_STATE(name, interval_init, burst_init) \
+ \
+ struct ratelimit_state name = { \
+- .lock = __SPIN_LOCK_UNLOCKED(name.lock), \
++ .lock = __RAW_SPIN_LOCK_UNLOCKED(name.lock), \
+ .interval = interval_init, \
+ .burst = burst_init, \
+ }
+@@ -28,7 +28,7 @@ struct ratelimit_state {
+ static inline void ratelimit_state_init(struct ratelimit_state *rs,
+ int interval, int burst)
{
- signed long old, new;
+- spin_lock_init(&rs->lock);
++ raw_spin_lock_init(&rs->lock);
+ rs->interval = interval;
+ rs->burst = burst;
+ rs->printed = 0;
+diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
+index 99f9aa7..e8dc1d8 100644
+--- a/include/linux/rcupdate.h
++++ b/include/linux/rcupdate.h
+@@ -78,7 +78,13 @@ struct rcu_head {
+ extern void call_rcu_sched(struct rcu_head *head,
+ void (*func)(struct rcu_head *rcu));
+ extern void synchronize_sched(void);
++
++#ifdef CONFIG_PREEMPT_RT_FULL
++# define rcu_barrier_bh rcu_barrier
++#else
+ extern void rcu_barrier_bh(void);
++#endif
++
+ extern void rcu_barrier_sched(void);
-@@ -119,7 +119,8 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
+ static inline void __rcu_read_lock_bh(void)
+@@ -104,6 +110,11 @@ void synchronize_rcu(void);
+ * types of kernel builds, the rcu_read_lock() nesting depth is unknowable.
*/
--static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
-+static inline void
-+__down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- signed long old, new, tmp;
+ #define rcu_preempt_depth() (current->rcu_read_lock_nesting)
++#ifndef CONFIG_PREEMPT_RT_FULL
++#define sched_rcu_preempt_depth() rcu_preempt_depth()
++#else
++static inline int sched_rcu_preempt_depth(void) { return 0; }
++#endif
-@@ -145,7 +146,7 @@ static inline void __down_write_nested(s
- rwsem_down_write_failed(sem);
- }
+ #else /* #ifdef CONFIG_PREEMPT_RCU */
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- __down_write_nested(sem, 0);
+@@ -127,6 +138,8 @@ static inline int rcu_preempt_depth(void)
+ return 0;
}
-@@ -153,7 +154,7 @@ static inline void __down_write(struct r
- /*
- * trylock for writing -- returns 1 if successful, 0 if contention
- */
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- signed long old;
-@@ -181,7 +182,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
- {
- signed long old, new;
++#define sched_rcu_preempt_depth() rcu_preempt_depth()
++
+ #endif /* #else #ifdef CONFIG_PREEMPT_RCU */
-@@ -210,7 +211,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
+ /* Internal to kernel */
+@@ -222,7 +235,14 @@ static inline int rcu_read_lock_held(void)
+ * rcu_read_lock_bh_held() is defined out of line to avoid #include-file
+ * hell.
*/
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- signed long old, new, tmp;
++#ifdef CONFIG_PREEMPT_RT_FULL
++static inline int rcu_read_lock_bh_held(void)
++{
++ return rcu_read_lock_held();
++}
++#else
+ extern int rcu_read_lock_bh_held(void);
++#endif
-@@ -240,7 +241,7 @@ static inline void __up_write(struct rw_
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
+ /**
+ * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section?
+@@ -631,8 +651,13 @@ static inline void rcu_read_unlock(void)
+ static inline void rcu_read_lock_bh(void)
{
- signed long old, new, tmp;
+ __rcu_read_lock_bh();
++
++#ifdef CONFIG_PREEMPT_RT_FULL
++ rcu_read_lock();
++#else
+ __acquire(RCU_BH);
+ rcu_read_acquire_bh();
++#endif
+ }
-@@ -269,7 +270,7 @@ static inline void __downgrade_write(str
/*
- * implement atomic add functionality
+@@ -642,8 +667,12 @@ static inline void rcu_read_lock_bh(void)
*/
--static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
+ static inline void rcu_read_unlock_bh(void)
{
- signed long old, new;
++#ifdef CONFIG_PREEMPT_RT_FULL
++ rcu_read_unlock();
++#else
+ rcu_read_release_bh();
+ __release(RCU_BH);
++#endif
+ __rcu_read_unlock_bh();
+ }
-@@ -295,7 +296,8 @@ static inline void rwsem_atomic_add(long
- /*
- * implement exchange and add functionality
+@@ -750,6 +779,9 @@ extern void call_rcu(struct rcu_head *head,
+
+ #endif /* #else #ifdef CONFIG_PREEMPT_RCU */
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++#define call_rcu_bh call_rcu
++#else
+ /**
+ * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period.
+ * @head: structure to be used for queueing the RCU updates.
+@@ -770,6 +802,7 @@ extern void call_rcu(struct rcu_head *head,
*/
--static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
-+static inline long
-+rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
- {
- signed long old, new;
+ extern void call_rcu_bh(struct rcu_head *head,
+ void (*func)(struct rcu_head *head));
++#endif
-Index: linux-3.0/arch/sh/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/sh/include/asm/rwsem.h
-+++ linux-3.0/arch/sh/include/asm/rwsem.h
-@@ -22,7 +22,7 @@
/*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- if (atomic_inc_return((atomic_t *)(&sem->count)) > 0)
- smp_wmb();
-@@ -30,7 +30,7 @@ static inline void __down_read(struct rw
- rwsem_down_read_failed(sem);
- }
+ * debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
+diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h
+index e65d066..3bf47dc 100644
+--- a/include/linux/rcutree.h
++++ b/include/linux/rcutree.h
+@@ -57,7 +57,11 @@ static inline void exit_rcu(void)
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- int tmp;
+ #endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
-@@ -47,7 +47,7 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
- */
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- int tmp;
++#ifndef CONFIG_PREEMPT_RT_FULL
+ extern void synchronize_rcu_bh(void);
++#else
++# define synchronize_rcu_bh synchronize_rcu
++#endif
+ extern void synchronize_sched_expedited(void);
+ extern void synchronize_rcu_expedited(void);
-@@ -59,7 +59,7 @@ static inline void __down_write(struct r
- rwsem_down_write_failed(sem);
- }
+@@ -71,13 +75,19 @@ extern void rcu_barrier(void);
+ extern unsigned long rcutorture_testseq;
+ extern unsigned long rcutorture_vernum;
+ extern long rcu_batches_completed(void);
+-extern long rcu_batches_completed_bh(void);
+ extern long rcu_batches_completed_sched(void);
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- int tmp;
+ extern void rcu_force_quiescent_state(void);
+-extern void rcu_bh_force_quiescent_state(void);
+ extern void rcu_sched_force_quiescent_state(void);
-@@ -72,7 +72,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
++#ifndef CONFIG_PREEMPT_RT_FULL
++extern void rcu_bh_force_quiescent_state(void);
++extern long rcu_batches_completed_bh(void);
++#else
++# define rcu_bh_force_quiescent_state rcu_force_quiescent_state
++# define rcu_batches_completed_bh rcu_batches_completed
++#endif
++
+ /* A context switch is a grace period for RCU-sched and RCU-bh. */
+ static inline int rcu_blocking_is_gp(void)
{
- int tmp;
+diff --git a/include/linux/rtmutex.h b/include/linux/rtmutex.h
+index 8d522ff..8951029 100644
+--- a/include/linux/rtmutex.h
++++ b/include/linux/rtmutex.h
+@@ -14,7 +14,7 @@
-@@ -85,7 +85,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
- */
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- smp_wmb();
- if (atomic_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
-@@ -96,7 +96,7 @@ static inline void __up_write(struct rw_
- /*
- * implement atomic add functionality
- */
--static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(int delta, struct rw_anon_semaphore *sem)
- {
- atomic_add(delta, (atomic_t *)(&sem->count));
- }
-@@ -104,7 +104,7 @@ static inline void rwsem_atomic_add(int
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- int tmp;
+ #include <linux/linkage.h>
+ #include <linux/plist.h>
+-#include <linux/spinlock_types.h>
++#include <linux/spinlock_types_raw.h>
+
+ extern int max_lock_depth; /* for sysctl */
+
+@@ -29,9 +29,10 @@ struct rt_mutex {
+ raw_spinlock_t wait_lock;
+ struct plist_head wait_list;
+ struct task_struct *owner;
+-#ifdef CONFIG_DEBUG_RT_MUTEXES
+ int save_state;
+- const char *name, *file;
++#ifdef CONFIG_DEBUG_RT_MUTEXES
++ const char *file;
++ const char *name;
+ int line;
+ void *magic;
+ #endif
+@@ -56,19 +57,39 @@ struct hrtimer_sleeper;
+ #ifdef CONFIG_DEBUG_RT_MUTEXES
+ # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) \
+ , .name = #mutexname, .file = __FILE__, .line = __LINE__
+-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, __func__)
++
++# define rt_mutex_init(mutex) \
++ do { \
++ raw_spin_lock_init(&(mutex)->wait_lock); \
++ __rt_mutex_init(mutex, #mutex); \
++ } while (0)
++
+ extern void rt_mutex_debug_task_free(struct task_struct *tsk);
+ #else
+ # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname)
+-# define rt_mutex_init(mutex) __rt_mutex_init(mutex, NULL)
++
++# define rt_mutex_init(mutex) \
++ do { \
++ raw_spin_lock_init(&(mutex)->wait_lock); \
++ __rt_mutex_init(mutex, #mutex); \
++ } while (0)
++
+ # define rt_mutex_debug_task_free(t) do { } while (0)
+ #endif
+
+-#define __RT_MUTEX_INITIALIZER(mutexname) \
+- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
++#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
+ , .wait_list = PLIST_HEAD_INIT_RAW(mutexname.wait_list, mutexname.wait_lock) \
+ , .owner = NULL \
+- __DEBUG_RT_MUTEX_INITIALIZER(mutexname)}
++ __DEBUG_RT_MUTEX_INITIALIZER(mutexname)
++
++
++#define __RT_MUTEX_INITIALIZER(mutexname) \
++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) }
++
++#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \
++ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
++ , .save_state = 1 }
-@@ -114,7 +114,8 @@ static inline void __downgrade_write(str
- rwsem_downgrade_wake(sem);
- }
+ #define DEFINE_RT_MUTEX(mutexname) \
+ struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname)
+@@ -90,6 +111,7 @@ extern void rt_mutex_destroy(struct rt_mutex *lock);
+ extern void rt_mutex_lock(struct rt_mutex *lock);
+ extern int rt_mutex_lock_interruptible(struct rt_mutex *lock,
+ int detect_deadlock);
++extern int rt_mutex_lock_killable(struct rt_mutex *lock, int detect_deadlock);
+ extern int rt_mutex_timed_lock(struct rt_mutex *lock,
+ struct hrtimer_sleeper *timeout,
+ int detect_deadlock);
+diff --git a/include/linux/rwlock_rt.h b/include/linux/rwlock_rt.h
+new file mode 100644
+index 0000000..853ee36
+--- /dev/null
++++ b/include/linux/rwlock_rt.h
+@@ -0,0 +1,123 @@
++#ifndef __LINUX_RWLOCK_RT_H
++#define __LINUX_RWLOCK_RT_H
++
++#ifndef __LINUX_SPINLOCK_H
++#error Do not include directly. Use spinlock.h
++#endif
++
++#define rwlock_init(rwl) \
++do { \
++ static struct lock_class_key __key; \
++ \
++ rt_mutex_init(&(rwl)->lock); \
++ __rt_rwlock_init(rwl, #rwl, &__key); \
++} while (0)
++
++extern void __lockfunc rt_write_lock(rwlock_t *rwlock);
++extern void __lockfunc rt_read_lock(rwlock_t *rwlock);
++extern int __lockfunc rt_write_trylock(rwlock_t *rwlock);
++extern int __lockfunc rt_write_trylock_irqsave(rwlock_t *trylock, unsigned long *flags);
++extern int __lockfunc rt_read_trylock(rwlock_t *rwlock);
++extern void __lockfunc rt_write_unlock(rwlock_t *rwlock);
++extern void __lockfunc rt_read_unlock(rwlock_t *rwlock);
++extern unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock);
++extern unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock);
++extern void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key);
++
++#define read_trylock(lock) __cond_lock(lock, rt_read_trylock(lock))
++#define write_trylock(lock) __cond_lock(lock, rt_write_trylock(lock))
++
++#define write_trylock_irqsave(lock, flags) \
++ __cond_lock(lock, rt_write_trylock_irqsave(lock, &flags))
++
++#define read_lock_irqsave(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ migrate_disable(); \
++ flags = rt_read_lock_irqsave(lock); \
++ } while (0)
++
++#define write_lock_irqsave(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ migrate_disable(); \
++ flags = rt_write_lock_irqsave(lock); \
++ } while (0)
++
++#define read_lock(lock) \
++ do { \
++ migrate_disable(); \
++ rt_read_lock(lock); \
++ } while (0)
++
++#define read_lock_bh(lock) \
++ do { \
++ local_bh_disable(); \
++ migrate_disable(); \
++ rt_read_lock(lock); \
++ } while (0)
++
++#define read_lock_irq(lock) read_lock(lock)
++
++#define write_lock(lock) \
++ do { \
++ migrate_disable(); \
++ rt_write_lock(lock); \
++ } while (0)
++
++#define write_lock_bh(lock) \
++ do { \
++ local_bh_disable(); \
++ migrate_disable(); \
++ rt_write_lock(lock); \
++ } while (0)
++
++#define write_lock_irq(lock) write_lock(lock)
++
++#define read_unlock(lock) \
++ do { \
++ rt_read_unlock(lock); \
++ migrate_enable(); \
++ } while (0)
++
++#define read_unlock_bh(lock) \
++ do { \
++ rt_read_unlock(lock); \
++ migrate_enable(); \
++ local_bh_enable(); \
++ } while (0)
++
++#define read_unlock_irq(lock) read_unlock(lock)
++
++#define write_unlock(lock) \
++ do { \
++ rt_write_unlock(lock); \
++ migrate_enable(); \
++ } while (0)
++
++#define write_unlock_bh(lock) \
++ do { \
++ rt_write_unlock(lock); \
++ migrate_enable(); \
++ local_bh_enable(); \
++ } while (0)
++
++#define write_unlock_irq(lock) write_unlock(lock)
++
++#define read_unlock_irqrestore(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ (void) flags; \
++ rt_read_unlock(lock); \
++ migrate_enable(); \
++ } while (0)
++
++#define write_unlock_irqrestore(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ (void) flags; \
++ rt_write_unlock(lock); \
++ migrate_enable(); \
++ } while (0)
++
++#endif
+diff --git a/include/linux/rwlock_types.h b/include/linux/rwlock_types.h
+index cc0072e..d0da966 100644
+--- a/include/linux/rwlock_types.h
++++ b/include/linux/rwlock_types.h
+@@ -1,6 +1,10 @@
+ #ifndef __LINUX_RWLOCK_TYPES_H
+ #define __LINUX_RWLOCK_TYPES_H
--static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
-+static inline void
-+ __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- __down_write(sem);
- }
-@@ -122,7 +123,7 @@ static inline void __down_write_nested(s
- /*
- * implement exchange and add functionality
- */
--static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
-+static inline int rwsem_atomic_update(int delta, struct rw_anon_semaphore *sem)
- {
- smp_mb();
- return atomic_add_return(delta, (atomic_t *)(&sem->count));
-Index: linux-3.0/arch/sparc/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/sparc/include/asm/rwsem.h
-+++ linux-3.0/arch/sparc/include/asm/rwsem.h
-@@ -23,13 +23,13 @@
++#if !defined(__LINUX_SPINLOCK_TYPES_H)
++# error "Do not include directly, include spinlock_types.h"
++#endif
++
/*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- if (unlikely(atomic64_inc_return((atomic64_t *)(&sem->count)) <= 0L))
- rwsem_down_read_failed(sem);
- }
+ * include/linux/rwlock_types.h - generic rwlock type definitions
+ * and initializers
+@@ -43,6 +47,7 @@ typedef struct {
+ RW_DEP_MAP_INIT(lockname) }
+ #endif
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- long tmp;
+-#define DEFINE_RWLOCK(x) rwlock_t x = __RW_LOCK_UNLOCKED(x)
++#define DEFINE_RWLOCK(name) \
++ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name)
-@@ -45,7 +45,7 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
+ #endif /* __LINUX_RWLOCK_TYPES_H */
+diff --git a/include/linux/rwlock_types_rt.h b/include/linux/rwlock_types_rt.h
+new file mode 100644
+index 0000000..b138321
+--- /dev/null
++++ b/include/linux/rwlock_types_rt.h
+@@ -0,0 +1,33 @@
++#ifndef __LINUX_RWLOCK_TYPES_RT_H
++#define __LINUX_RWLOCK_TYPES_RT_H
++
++#ifndef __LINUX_SPINLOCK_TYPES_H
++#error "Do not include directly. Include spinlock_types.h instead"
++#endif
++
++/*
++ * rwlocks - rtmutex which allows single reader recursion
++ */
++typedef struct {
++ struct rt_mutex lock;
++ int read_depth;
++ unsigned int break_lock;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++} rwlock_t;
++
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define RW_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
++#else
++# define RW_DEP_MAP_INIT(lockname)
++#endif
++
++#define __RW_LOCK_UNLOCKED(name) \
++ { .lock = __RT_MUTEX_INITIALIZER_SAVE_STATE(name.lock), \
++ RW_DEP_MAP_INIT(name) }
++
++#define DEFINE_RWLOCK(name) \
++ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name)
++
++#endif
+diff --git a/include/linux/rwsem-spinlock.h b/include/linux/rwsem-spinlock.h
+index 3470124..2d4529b 100644
+--- a/include/linux/rwsem-spinlock.h
++++ b/include/linux/rwsem-spinlock.h
+@@ -20,26 +20,42 @@
+ * - if activity is -1 then there is one active writer
+ * - if wait_list is not empty, then there are processes waiting for the semaphore
*/
--static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
-+static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- long tmp;
-
-@@ -55,12 +55,12 @@ static inline void __down_write_nested(s
- rwsem_down_write_failed(sem);
- }
-
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- __down_write_nested(sem, 0);
- }
++struct rw_anon_semaphore {
++ __s32 activity;
++ raw_spinlock_t wait_lock;
++ struct list_head wait_list;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++};
++
++#ifndef CONFIG_PREEMPT_RT_FULL
++/*
++ * Non preempt-rt implementation of rw_semaphore. Same as above, but
++ * restricted vs. ownership. i.e. ownerless locked state and non owner
++ * release not allowed.
++ */
+ struct rw_semaphore {
+ __s32 activity;
+- spinlock_t wait_lock;
++ raw_spinlock_t wait_lock;
+ struct list_head wait_list;
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ struct lockdep_map dep_map;
+ #endif
+ };
++#endif /* PREEMPT_RT_FULL */
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- long tmp;
+ #define RWSEM_UNLOCKED_VALUE 0x00000000
-@@ -72,7 +72,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
- {
- long tmp;
+-extern void __down_read(struct rw_semaphore *sem);
+-extern int __down_read_trylock(struct rw_semaphore *sem);
+-extern void __down_write(struct rw_semaphore *sem);
+-extern void __down_write_nested(struct rw_semaphore *sem, int subclass);
+-extern int __down_write_trylock(struct rw_semaphore *sem);
+-extern void __up_read(struct rw_semaphore *sem);
+-extern void __up_write(struct rw_semaphore *sem);
+-extern void __downgrade_write(struct rw_semaphore *sem);
+-extern int rwsem_is_locked(struct rw_semaphore *sem);
++extern void __down_read(struct rw_anon_semaphore *sem);
++extern int __down_read_trylock(struct rw_anon_semaphore *sem);
++extern void __down_write(struct rw_anon_semaphore *sem);
++extern void __down_write_nested(struct rw_anon_semaphore *sem, int subclass);
++extern int __down_write_trylock(struct rw_anon_semaphore *sem);
++extern void __up_read(struct rw_anon_semaphore *sem);
++extern void __up_write(struct rw_anon_semaphore *sem);
++extern void __downgrade_write(struct rw_anon_semaphore *sem);
++extern int anon_rwsem_is_locked(struct rw_anon_semaphore *sem);
-@@ -84,7 +84,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
- */
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- if (unlikely(atomic64_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
- (atomic64_t *)(&sem->count)) < 0L))
-@@ -94,7 +94,7 @@ static inline void __up_write(struct rw_
- /*
- * implement atomic add functionality
- */
--static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
- {
- atomic64_add(delta, (atomic64_t *)(&sem->count));
- }
-@@ -102,7 +102,7 @@ static inline void rwsem_atomic_add(long
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- long tmp;
+ #endif /* __KERNEL__ */
+ #endif /* _LINUX_RWSEM_SPINLOCK_H */
+diff --git a/include/linux/rwsem.h b/include/linux/rwsem.h
+index a8afe9c..ae50875 100644
+--- a/include/linux/rwsem.h
++++ b/include/linux/rwsem.h
+@@ -17,37 +17,50 @@
+ #include <asm/system.h>
+ #include <asm/atomic.h>
-@@ -114,7 +114,7 @@ static inline void __downgrade_write(str
- /*
- * implement exchange and add functionality
- */
--static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
-+static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
- {
- return atomic64_add_return(delta, (atomic64_t *)(&sem->count));
- }
-Index: linux-3.0/arch/x86/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/rwsem.h
-+++ linux-3.0/arch/x86/include/asm/rwsem.h
-@@ -60,7 +60,7 @@
- /*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- asm volatile("# beginning down_read\n\t"
- LOCK_PREFIX _ASM_INC "(%1)\n\t"
-@@ -77,7 +77,7 @@ static inline void __down_read(struct rw
- /*
- * trylock for reading -- returns 1 if successful, 0 if contention
- */
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- long result, tmp;
- asm volatile("# beginning __down_read_trylock\n\t"
-@@ -99,7 +99,7 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
- */
--static inline void __down_write_nested(struct rw_semaphore *sem, int subclass)
-+static inline void __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- long tmp;
- asm volatile("# beginning down_write\n\t"
-@@ -116,7 +116,7 @@ static inline void __down_write_nested(s
- : "memory", "cc");
- }
++struct rw_anon_semaphore;
+ struct rw_semaphore;
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
- {
- __down_write_nested(sem, 0);
- }
-@@ -124,7 +124,7 @@ static inline void __down_write(struct r
- /*
- * trylock for writing -- returns 1 if successful, 0 if contention
- */
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- long ret = cmpxchg(&sem->count, RWSEM_UNLOCKED_VALUE,
- RWSEM_ACTIVE_WRITE_BIAS);
-@@ -136,7 +136,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
- {
- long tmp;
- asm volatile("# beginning __up_read\n\t"
-@@ -154,7 +154,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
- */
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- long tmp;
- asm volatile("# beginning __up_write\n\t"
-@@ -172,7 +172,7 @@ static inline void __up_write(struct rw_
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- asm volatile("# beginning __downgrade_write\n\t"
- LOCK_PREFIX _ASM_ADD "%2,(%1)\n\t"
-@@ -192,7 +192,7 @@ static inline void __downgrade_write(str
- /*
- * implement atomic add functionality
- */
--static inline void rwsem_atomic_add(long delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(long delta, struct rw_anon_semaphore *sem)
- {
- asm volatile(LOCK_PREFIX _ASM_ADD "%1,%0"
- : "+m" (sem->count)
-@@ -202,7 +202,7 @@ static inline void rwsem_atomic_add(long
- /*
- * implement exchange and add functionality
- */
--static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem)
-+static inline long rwsem_atomic_update(long delta, struct rw_anon_semaphore *sem)
- {
- long tmp = delta;
+ #ifdef CONFIG_RWSEM_GENERIC_SPINLOCK
+ #include <linux/rwsem-spinlock.h> /* use a generic implementation */
+-#else
++#else /* RWSEM_GENERIC_SPINLOCK */
++
+ /* All arch specific implementations share the same struct */
+-struct rw_semaphore {
++struct rw_anon_semaphore {
+ long count;
+- spinlock_t wait_lock;
++ raw_spinlock_t wait_lock;
+ struct list_head wait_list;
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ struct lockdep_map dep_map;
+ #endif
+ };
-Index: linux-3.0/arch/xtensa/include/asm/rwsem.h
-===================================================================
---- linux-3.0.orig/arch/xtensa/include/asm/rwsem.h
-+++ linux-3.0/arch/xtensa/include/asm/rwsem.h
-@@ -27,7 +27,7 @@
- /*
- * lock for reading
- */
--static inline void __down_read(struct rw_semaphore *sem)
-+static inline void __down_read(struct rw_anon_semaphore *sem)
- {
- if (atomic_add_return(1,(atomic_t *)(&sem->count)) > 0)
- smp_wmb();
-@@ -35,7 +35,7 @@ static inline void __down_read(struct rw
- rwsem_down_read_failed(sem);
- }
+-extern struct rw_semaphore *rwsem_down_read_failed(struct rw_semaphore *sem);
+-extern struct rw_semaphore *rwsem_down_write_failed(struct rw_semaphore *sem);
+-extern struct rw_semaphore *rwsem_wake(struct rw_semaphore *);
+-extern struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem);
++extern struct rw_anon_semaphore *rwsem_down_read_failed(struct rw_anon_semaphore *sem);
++extern struct rw_anon_semaphore *rwsem_down_write_failed(struct rw_anon_semaphore *sem);
++extern struct rw_anon_semaphore *rwsem_wake(struct rw_anon_semaphore *);
++extern struct rw_anon_semaphore *rwsem_downgrade_wake(struct rw_anon_semaphore *sem);
--static inline int __down_read_trylock(struct rw_semaphore *sem)
-+static inline int __down_read_trylock(struct rw_anon_semaphore *sem)
- {
- int tmp;
+ /* Include the arch specific part */
+ #include <asm/rwsem.h>
-@@ -52,7 +52,7 @@ static inline int __down_read_trylock(st
- /*
- * lock for writing
- */
--static inline void __down_write(struct rw_semaphore *sem)
-+static inline void __down_write(struct rw_anon_semaphore *sem)
+ /* In all implementations count != 0 means locked */
+-static inline int rwsem_is_locked(struct rw_semaphore *sem)
++static inline int anon_rwsem_is_locked(struct rw_anon_semaphore *sem)
{
- int tmp;
-
-@@ -64,7 +64,7 @@ static inline void __down_write(struct r
- rwsem_down_write_failed(sem);
+ return sem->count != 0;
}
--static inline int __down_write_trylock(struct rw_semaphore *sem)
-+static inline int __down_write_trylock(struct rw_anon_semaphore *sem)
- {
- int tmp;
++#ifndef CONFIG_PREEMPT_RT_FULL
++struct rw_semaphore {
++ long count;
++ raw_spinlock_t wait_lock;
++ struct list_head wait_list;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++};
+ #endif
-@@ -77,7 +77,7 @@ static inline int __down_write_trylock(s
- /*
- * unlock after reading
- */
--static inline void __up_read(struct rw_semaphore *sem)
-+static inline void __up_read(struct rw_anon_semaphore *sem)
- {
- int tmp;
++#endif /* !RWSEM_GENERIC_SPINLOCK */
++
+ /* Common initializer macros and functions */
-@@ -90,7 +90,7 @@ static inline void __up_read(struct rw_s
- /*
- * unlock after writing
- */
--static inline void __up_write(struct rw_semaphore *sem)
-+static inline void __up_write(struct rw_anon_semaphore *sem)
- {
- smp_wmb();
- if (atomic_sub_return(RWSEM_ACTIVE_WRITE_BIAS,
-@@ -101,7 +101,7 @@ static inline void __up_write(struct rw_
- /*
- * implement atomic add functionality
- */
--static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem)
-+static inline void rwsem_atomic_add(int delta, struct rw_anon_semaphore *sem)
- {
- atomic_add(delta, (atomic_t *)(&sem->count));
- }
-@@ -109,7 +109,7 @@ static inline void rwsem_atomic_add(int
- /*
- * downgrade write lock to read lock
- */
--static inline void __downgrade_write(struct rw_semaphore *sem)
-+static inline void __downgrade_write(struct rw_anon_semaphore *sem)
- {
- int tmp;
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+@@ -56,57 +69,59 @@ static inline int rwsem_is_locked(struct rw_semaphore *sem)
+ # define __RWSEM_DEP_MAP_INIT(lockname)
+ #endif
+
+-#define __RWSEM_INITIALIZER(name) \
+- { RWSEM_UNLOCKED_VALUE, __SPIN_LOCK_UNLOCKED(name.wait_lock), \
+- LIST_HEAD_INIT((name).wait_list) __RWSEM_DEP_MAP_INIT(name) }
++#define __RWSEM_ANON_INITIALIZER(name) \
++ { RWSEM_UNLOCKED_VALUE, \
++ __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
++ LIST_HEAD_INIT((name).wait_list) \
++ __RWSEM_DEP_MAP_INIT(name) }
+
+-#define DECLARE_RWSEM(name) \
+- struct rw_semaphore name = __RWSEM_INITIALIZER(name)
++#define DECLARE_ANON_RWSEM(name) \
++ struct rw_anon_semaphore name = __RWSEM_INITIALIZER(name)
+
+-extern void __init_rwsem(struct rw_semaphore *sem, const char *name,
+- struct lock_class_key *key);
++extern void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
++ struct lock_class_key *key);
+
+-#define init_rwsem(sem) \
++#define init_anon_rwsem(sem) \
+ do { \
+ static struct lock_class_key __key; \
+ \
+- __init_rwsem((sem), #sem, &__key); \
++ __init_anon_rwsem((sem), #sem, &__key); \
+ } while (0)
-@@ -122,7 +122,7 @@ static inline void __downgrade_write(str
- /*
- * implement exchange and add functionality
- */
--static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem)
-+static inline int rwsem_atomic_update(int delta, struct rw_anon_semaphore *sem)
- {
- smp_mb();
- return atomic_add_return(delta, (atomic_t *)(&sem->count));
-Index: linux-3.0/kernel/rwsem.c
-===================================================================
---- linux-3.0.orig/kernel/rwsem.c
-+++ linux-3.0/kernel/rwsem.c
-@@ -16,20 +16,19 @@
/*
* lock for reading
*/
--void __sched down_read(struct rw_semaphore *sem)
-+void __sched anon_down_read(struct rw_anon_semaphore *sem)
- {
- might_sleep();
- rwsem_acquire_read(&sem->dep_map, 0, 0, _RET_IP_);
-
- LOCK_CONTENDED(sem, __down_read_trylock, __down_read);
- }
--
--EXPORT_SYMBOL(down_read);
-+EXPORT_SYMBOL(anon_down_read);
+-extern void down_read(struct rw_semaphore *sem);
++extern void anon_down_read(struct rw_anon_semaphore *sem);
/*
* trylock for reading -- returns 1 if successful, 0 if contention
*/
--int down_read_trylock(struct rw_semaphore *sem)
-+int anon_down_read_trylock(struct rw_anon_semaphore *sem)
- {
- int ret = __down_read_trylock(sem);
-
-@@ -37,26 +36,24 @@ int down_read_trylock(struct rw_semaphor
- rwsem_acquire_read(&sem->dep_map, 0, 1, _RET_IP_);
- return ret;
- }
--
--EXPORT_SYMBOL(down_read_trylock);
-+EXPORT_SYMBOL(anon_down_read_trylock);
+-extern int down_read_trylock(struct rw_semaphore *sem);
++extern int anon_down_read_trylock(struct rw_anon_semaphore *sem);
/*
* lock for writing
*/
--void __sched down_write(struct rw_semaphore *sem)
-+void __sched anon_down_write(struct rw_anon_semaphore *sem)
- {
- might_sleep();
- rwsem_acquire(&sem->dep_map, 0, 0, _RET_IP_);
-
- LOCK_CONTENDED(sem, __down_write_trylock, __down_write);
- }
--
--EXPORT_SYMBOL(down_write);
-+EXPORT_SYMBOL(anon_down_write);
+-extern void down_write(struct rw_semaphore *sem);
++extern void anon_down_write(struct rw_anon_semaphore *sem);
/*
* trylock for writing -- returns 1 if successful, 0 if contention
*/
--int down_write_trylock(struct rw_semaphore *sem)
-+int anon_down_write_trylock(struct rw_anon_semaphore *sem)
- {
- int ret = __down_write_trylock(sem);
-
-@@ -64,37 +61,34 @@ int down_write_trylock(struct rw_semapho
- rwsem_acquire(&sem->dep_map, 0, 1, _RET_IP_);
- return ret;
- }
--
--EXPORT_SYMBOL(down_write_trylock);
-+EXPORT_SYMBOL(anon_down_write_trylock);
+-extern int down_write_trylock(struct rw_semaphore *sem);
++extern int anon_down_write_trylock(struct rw_anon_semaphore *sem);
/*
* release a read lock
*/
--void up_read(struct rw_semaphore *sem)
-+void anon_up_read(struct rw_anon_semaphore *sem)
- {
- rwsem_release(&sem->dep_map, 1, _RET_IP_);
-
- __up_read(sem);
- }
--
--EXPORT_SYMBOL(up_read);
-+EXPORT_SYMBOL(anon_up_read);
+-extern void up_read(struct rw_semaphore *sem);
++extern void anon_up_read(struct rw_anon_semaphore *sem);
/*
* release a write lock
*/
--void up_write(struct rw_semaphore *sem)
-+void anon_up_write(struct rw_anon_semaphore *sem)
- {
- rwsem_release(&sem->dep_map, 1, _RET_IP_);
-
- __up_write(sem);
- }
--
--EXPORT_SYMBOL(up_write);
-+EXPORT_SYMBOL(anon_up_write);
-
- /*
- * downgrade write lock to read lock
- */
--void downgrade_write(struct rw_semaphore *sem)
-+void anon_downgrade_write(struct rw_anon_semaphore *sem)
- {
- /*
- * lockdep: a downgraded write will live on as a write
-@@ -102,46 +96,41 @@ void downgrade_write(struct rw_semaphore
- */
- __downgrade_write(sem);
- }
--
--EXPORT_SYMBOL(downgrade_write);
-+EXPORT_SYMBOL(anon_downgrade_write);
-
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
-
--void down_read_nested(struct rw_semaphore *sem, int subclass)
-+void anon_down_read_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- might_sleep();
- rwsem_acquire_read(&sem->dep_map, subclass, 0, _RET_IP_);
-
- LOCK_CONTENDED(sem, __down_read_trylock, __down_read);
- }
-+EXPORT_SYMBOL(anon_down_read_nested);
-
--EXPORT_SYMBOL(down_read_nested);
--
--void down_read_non_owner(struct rw_semaphore *sem)
-+void anon_down_read_non_owner(struct rw_anon_semaphore *sem)
- {
- might_sleep();
-
- __down_read(sem);
- }
-+EXPORT_SYMBOL(anon_down_read_non_owner);
-
--EXPORT_SYMBOL(down_read_non_owner);
--
--void down_write_nested(struct rw_semaphore *sem, int subclass)
-+void anon_down_write_nested(struct rw_anon_semaphore *sem, int subclass)
- {
- might_sleep();
- rwsem_acquire(&sem->dep_map, subclass, 0, _RET_IP_);
-
- LOCK_CONTENDED(sem, __down_write_trylock, __down_write);
- }
-+EXPORT_SYMBOL(anon_down_write_nested);
-
--EXPORT_SYMBOL(down_write_nested);
--
--void up_read_non_owner(struct rw_semaphore *sem)
-+void anon_up_read_non_owner(struct rw_anon_semaphore *sem)
- {
- __up_read(sem);
- }
--
--EXPORT_SYMBOL(up_read_non_owner);
-+EXPORT_SYMBOL(anon_up_read_non_owner);
-
- #endif
-
-Index: linux-3.0/fs/attr.c
-===================================================================
---- linux-3.0.orig/fs/attr.c
-+++ linux-3.0/fs/attr.c
-@@ -233,7 +233,7 @@ int notify_change(struct dentry * dentry
- return error;
-
- if (ia_valid & ATTR_SIZE)
-- down_write(&dentry->d_inode->i_alloc_sem);
-+ anon_down_write(&dentry->d_inode->i_alloc_sem);
-
- if (inode->i_op->setattr)
- error = inode->i_op->setattr(dentry, attr);
-@@ -241,7 +241,7 @@ int notify_change(struct dentry * dentry
- error = simple_setattr(dentry, attr);
-
- if (ia_valid & ATTR_SIZE)
-- up_write(&dentry->d_inode->i_alloc_sem);
-+ anon_up_write(&dentry->d_inode->i_alloc_sem);
-
- if (!error)
- fsnotify_change(dentry, ia_valid);
-Index: linux-3.0/fs/direct-io.c
-===================================================================
---- linux-3.0.orig/fs/direct-io.c
-+++ linux-3.0/fs/direct-io.c
-@@ -255,7 +255,7 @@ static ssize_t dio_complete(struct dio *
-
- if (dio->flags & DIO_LOCKING)
- /* lockdep: non-owner release */
-- up_read_non_owner(&dio->inode->i_alloc_sem);
-+ anon_up_read_non_owner(&dio->inode->i_alloc_sem);
-
- return ret;
- }
-@@ -1234,7 +1234,7 @@ __blockdev_direct_IO(int rw, struct kioc
- * Will be released at I/O completion, possibly in a
- * different thread.
- */
-- down_read_non_owner(&inode->i_alloc_sem);
-+ anon_down_read_non_owner(&inode->i_alloc_sem);
- }
-
- /*
-Index: linux-3.0/fs/ext4/inode.c
-===================================================================
---- linux-3.0.orig/fs/ext4/inode.c
-+++ linux-3.0/fs/ext4/inode.c
-@@ -5872,7 +5872,7 @@ int ext4_page_mkwrite(struct vm_area_str
- * Get i_alloc_sem to stop truncates messing with the inode. We cannot
- * get i_mutex because we are already holding mmap_sem.
- */
-- down_read(&inode->i_alloc_sem);
-+ anon_down_read(&inode->i_alloc_sem);
- size = i_size_read(inode);
- if (page->mapping != mapping || size <= page_offset(page)
- || !PageUptodate(page)) {
-@@ -5884,7 +5884,7 @@ int ext4_page_mkwrite(struct vm_area_str
- lock_page(page);
- wait_on_page_writeback(page);
- if (PageMappedToDisk(page)) {
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- return VM_FAULT_LOCKED;
- }
-
-@@ -5902,7 +5902,7 @@ int ext4_page_mkwrite(struct vm_area_str
- if (page_has_buffers(page)) {
- if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL,
- ext4_bh_unmapped)) {
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- return VM_FAULT_LOCKED;
- }
- }
-@@ -5931,11 +5931,11 @@ int ext4_page_mkwrite(struct vm_area_str
- */
- lock_page(page);
- wait_on_page_writeback(page);
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- return VM_FAULT_LOCKED;
- out_unlock:
- if (ret)
- ret = VM_FAULT_SIGBUS;
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- return ret;
- }
-Index: linux-3.0/fs/fat/inode.c
-===================================================================
---- linux-3.0.orig/fs/fat/inode.c
-+++ linux-3.0/fs/fat/inode.c
-@@ -224,9 +224,9 @@ static sector_t _fat_bmap(struct address
- sector_t blocknr;
-
- /* fat_get_cluster() assumes the requested blocknr isn't truncated. */
-- down_read(&mapping->host->i_alloc_sem);
-+ anon_down_read(&mapping->host->i_alloc_sem);
- blocknr = generic_block_bmap(mapping, block, fat_get_block);
-- up_read(&mapping->host->i_alloc_sem);
-+ anon_up_read(&mapping->host->i_alloc_sem);
-
- return blocknr;
- }
-Index: linux-3.0/fs/inode.c
-===================================================================
---- linux-3.0.orig/fs/inode.c
-+++ linux-3.0/fs/inode.c
-@@ -176,7 +176,7 @@ int inode_init_always(struct super_block
- mutex_init(&inode->i_mutex);
- lockdep_set_class(&inode->i_mutex, &sb->s_type->i_mutex_key);
+-extern void up_write(struct rw_semaphore *sem);
++extern void anon_up_write(struct rw_anon_semaphore *sem);
-- init_rwsem(&inode->i_alloc_sem);
-+ init_anon_rwsem(&inode->i_alloc_sem);
- lockdep_set_class(&inode->i_alloc_sem, &sb->s_type->i_alloc_sem_key);
+ /*
+ * downgrade write lock to read lock
+ */
+-extern void downgrade_write(struct rw_semaphore *sem);
++extern void anon_downgrade_write(struct rw_anon_semaphore *sem);
- mapping->a_ops = &empty_aops;
-Index: linux-3.0/fs/ntfs/file.c
-===================================================================
---- linux-3.0.orig/fs/ntfs/file.c
-+++ linux-3.0/fs/ntfs/file.c
-@@ -1832,9 +1832,9 @@ static ssize_t ntfs_file_buffered_write(
- * fails again.
- */
- if (unlikely(NInoTruncateFailed(ni))) {
-- down_write(&vi->i_alloc_sem);
-+ anon_down_write(&vi->i_alloc_sem);
- err = ntfs_truncate(vi);
-- up_write(&vi->i_alloc_sem);
-+ anon_up_write(&vi->i_alloc_sem);
- if (err || NInoTruncateFailed(ni)) {
- if (!err)
- err = -EIO;
-Index: linux-3.0/fs/ocfs2/aops.c
-===================================================================
---- linux-3.0.orig/fs/ocfs2/aops.c
-+++ linux-3.0/fs/ocfs2/aops.c
-@@ -569,7 +569,7 @@ static void ocfs2_dio_end_io(struct kioc
- BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ /*
+@@ -122,21 +137,101 @@ extern void downgrade_write(struct rw_semaphore *sem);
+ * lockdep_set_class() at lock initialization time.
+ * See Documentation/lockdep-design.txt for more details.)
+ */
+-extern void down_read_nested(struct rw_semaphore *sem, int subclass);
+-extern void down_write_nested(struct rw_semaphore *sem, int subclass);
++extern void anon_down_read_nested(struct rw_anon_semaphore *sem, int subclass);
++extern void anon_down_write_nested(struct rw_anon_semaphore *sem, int subclass);
+ /*
+ * Take/release a lock when not the owner will release it.
+ *
+ * [ This API should be avoided as much as possible - the
+ * proper abstraction for this case is completions. ]
+ */
+-extern void down_read_non_owner(struct rw_semaphore *sem);
+-extern void up_read_non_owner(struct rw_semaphore *sem);
++extern void anon_down_read_non_owner(struct rw_anon_semaphore *sem);
++extern void anon_up_read_non_owner(struct rw_anon_semaphore *sem);
+ #else
+-# define down_read_nested(sem, subclass) down_read(sem)
+-# define down_write_nested(sem, subclass) down_write(sem)
+-# define down_read_non_owner(sem) down_read(sem)
+-# define up_read_non_owner(sem) up_read(sem)
++# define anon_down_read_nested(sem, subclass) anon_down_read(sem)
++# define anon_down_write_nested(sem, subclass) anon_down_write(sem)
++# define anon_down_read_non_owner(sem) anon_down_read(sem)
++# define anon_up_read_non_owner(sem) anon_up_read(sem)
+ #endif
- if (ocfs2_iocb_is_sem_locked(iocb)) {
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- ocfs2_iocb_clear_sem_locked(iocb);
- }
++#ifdef CONFIG_PREEMPT_RT_FULL
++#include <linux/rwsem_rt.h>
++#else /* PREEMPT_RT_FULL */
++/*
++ * Non preempt-rt implementations
++ */
++#define __RWSEM_INITIALIZER(name) \
++ { RWSEM_UNLOCKED_VALUE, \
++ __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
++ LIST_HEAD_INIT((name).wait_list) \
++ __RWSEM_DEP_MAP_INIT(name) }
++
++#define DECLARE_RWSEM(name) \
++ struct rw_semaphore name = __RWSEM_INITIALIZER(name)
++
++static inline void __init_rwsem(struct rw_semaphore *sem, const char *name,
++ struct lock_class_key *key)
++{
++ __init_anon_rwsem((struct rw_anon_semaphore *)sem, name, key);
++}
++
++#define init_rwsem(sem) \
++do { \
++ static struct lock_class_key __key; \
++ \
++ __init_rwsem((sem), #sem, &__key); \
++} while (0)
++
++static inline void down_read(struct rw_semaphore *sem)
++{
++ anon_down_read((struct rw_anon_semaphore *)sem);
++}
++
++static inline int down_read_trylock(struct rw_semaphore *sem)
++{
++ return anon_down_read_trylock((struct rw_anon_semaphore *)sem);
++}
++
++static inline void down_write(struct rw_semaphore *sem)
++{
++ anon_down_write((struct rw_anon_semaphore *)sem);
++}
++
++static inline int down_write_trylock(struct rw_semaphore *sem)
++{
++ return anon_down_write_trylock((struct rw_anon_semaphore *)sem);
++}
++
++static inline void up_read(struct rw_semaphore *sem)
++{
++ anon_up_read((struct rw_anon_semaphore *)sem);
++}
++
++static inline void up_write(struct rw_semaphore *sem)
++{
++ anon_up_write((struct rw_anon_semaphore *)sem);
++}
++
++static inline void downgrade_write(struct rw_semaphore *sem)
++{
++ anon_downgrade_write((struct rw_anon_semaphore *)sem);
++}
++
++static inline void down_read_nested(struct rw_semaphore *sem, int subclass)
++{
++ return anon_down_read_nested((struct rw_anon_semaphore *)sem, subclass);
++}
++
++static inline void down_write_nested(struct rw_semaphore *sem, int subclass)
++{
++ anon_down_write_nested((struct rw_anon_semaphore *)sem, subclass);
++}
++
++static inline int rwsem_is_locked(struct rw_semaphore *sem)
++{
++ return anon_rwsem_is_locked((struct rw_anon_semaphore *)sem);
++}
++#endif /* !PREEMPT_RT_FULL */
++
+ #endif /* _LINUX_RWSEM_H */
++
+diff --git a/include/linux/rwsem_rt.h b/include/linux/rwsem_rt.h
+new file mode 100644
+index 0000000..802c690
+--- /dev/null
++++ b/include/linux/rwsem_rt.h
+@@ -0,0 +1,105 @@
++#ifndef _LINUX_RWSEM_RT_H
++#define _LINUX_RWSEM_RT_H
++
++#ifndef _LINUX_RWSEM_H
++#error "Include rwsem.h"
++#endif
++
++/*
++ * RW-semaphores are a spinlock plus a reader-depth count.
++ *
++ * Note that the semantics are different from the usual
++ * Linux rw-sems, in PREEMPT_RT mode we do not allow
++ * multiple readers to hold the lock at once, we only allow
++ * a read-lock owner to read-lock recursively. This is
++ * better for latency, makes the implementation inherently
++ * fair and makes it simpler as well.
++ */
++
++#include <linux/rtmutex.h>
++
++struct rw_semaphore {
++ struct rt_mutex lock;
++ int read_depth;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++};
++
++#define __RWSEM_INITIALIZER(name) \
++ { .lock = __RT_MUTEX_INITIALIZER(name.lock), \
++ RW_DEP_MAP_INIT(name) }
++
++#define DECLARE_RWSEM(lockname) \
++ struct rw_semaphore lockname = __RWSEM_INITIALIZER(lockname)
++
++extern void __rt_rwsem_init(struct rw_semaphore *rwsem, char *name,
++ struct lock_class_key *key);
++
++# define rt_init_rwsem(sem) \
++do { \
++ static struct lock_class_key __key; \
++ \
++ rt_mutex_init(&(sem)->lock); \
++ __rt_rwsem_init((sem), #sem, &__key); \
++} while (0)
++
++extern void rt_down_write(struct rw_semaphore *rwsem);
++extern void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass);
++extern void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass);
++extern void rt_down_read(struct rw_semaphore *rwsem);
++extern int rt_down_write_trylock(struct rw_semaphore *rwsem);
++extern int rt_down_read_trylock(struct rw_semaphore *rwsem);
++extern void rt_up_read(struct rw_semaphore *rwsem);
++extern void rt_up_write(struct rw_semaphore *rwsem);
++extern void rt_downgrade_write(struct rw_semaphore *rwsem);
++
++#define init_rwsem(sem) rt_init_rwsem(sem)
++#define rwsem_is_locked(s) rt_mutex_is_locked(&(s)->lock)
++
++static inline void down_read(struct rw_semaphore *sem)
++{
++ rt_down_read(sem);
++}
++
++static inline int down_read_trylock(struct rw_semaphore *sem)
++{
++ return rt_down_read_trylock(sem);
++}
++
++static inline void down_write(struct rw_semaphore *sem)
++{
++ rt_down_write(sem);
++}
++
++static inline int down_write_trylock(struct rw_semaphore *sem)
++{
++ return rt_down_write_trylock(sem);
++}
++
++static inline void up_read(struct rw_semaphore *sem)
++{
++ rt_up_read(sem);
++}
++
++static inline void up_write(struct rw_semaphore *sem)
++{
++ rt_up_write(sem);
++}
++
++static inline void downgrade_write(struct rw_semaphore *sem)
++{
++ rt_downgrade_write(sem);
++}
++
++static inline void down_read_nested(struct rw_semaphore *sem, int subclass)
++{
++ return rt_down_read_nested(sem, subclass);
++}
++
++static inline void down_write_nested(struct rw_semaphore *sem, int subclass)
++{
++ rt_down_write_nested(sem, subclass);
++}
++
++#endif
+diff --git a/include/linux/sched.h b/include/linux/sched.h
+index 4ef452b..4cd3278 100644
+--- a/include/linux/sched.h
++++ b/include/linux/sched.h
+@@ -63,6 +63,7 @@ struct sched_param {
+ #include <linux/nodemask.h>
+ #include <linux/mm_types.h>
+
++#include <asm/kmap_types.h>
+ #include <asm/system.h>
+ #include <asm/page.h>
+ #include <asm/ptrace.h>
+@@ -90,6 +91,7 @@ struct sched_param {
+ #include <linux/task_io_accounting.h>
+ #include <linux/latencytop.h>
+ #include <linux/cred.h>
++#include <linux/hardirq.h>
-Index: linux-3.0/fs/ocfs2/file.c
-===================================================================
---- linux-3.0.orig/fs/ocfs2/file.c
-+++ linux-3.0/fs/ocfs2/file.c
-@@ -2238,7 +2238,7 @@ static ssize_t ocfs2_file_aio_write(stru
- relock:
- /* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
- if (direct_io) {
-- down_read(&inode->i_alloc_sem);
-+ anon_down_read(&inode->i_alloc_sem);
- have_alloc_sem = 1;
- /* communicate with ocfs2_dio_end_io */
- ocfs2_iocb_set_sem_locked(iocb);
-@@ -2290,7 +2290,7 @@ relock:
- */
- if (direct_io && !can_do_direct) {
- ocfs2_rw_unlock(inode, rw_level);
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
+ #include <asm/processor.h>
- have_alloc_sem = 0;
- rw_level = -1;
-@@ -2379,7 +2379,7 @@ out:
+@@ -359,6 +361,7 @@ extern signed long schedule_timeout_interruptible(signed long timeout);
+ extern signed long schedule_timeout_killable(signed long timeout);
+ extern signed long schedule_timeout_uninterruptible(signed long timeout);
+ asmlinkage void schedule(void);
++extern void schedule_preempt_disabled(void);
+ extern int mutex_spin_on_owner(struct mutex *lock, struct task_struct *owner);
- out_sems:
- if (have_alloc_sem) {
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- ocfs2_iocb_clear_sem_locked(iocb);
- }
+ struct nsproxy;
+@@ -510,7 +513,7 @@ struct task_cputime {
+ struct thread_group_cputimer {
+ struct task_cputime cputime;
+ int running;
+- spinlock_t lock;
++ raw_spinlock_t lock;
+ };
-@@ -2531,7 +2531,7 @@ static ssize_t ocfs2_file_aio_read(struc
- * need locks to protect pending reads from racing with truncate.
- */
- if (filp->f_flags & O_DIRECT) {
-- down_read(&inode->i_alloc_sem);
-+ anon_down_read(&inode->i_alloc_sem);
- have_alloc_sem = 1;
- ocfs2_iocb_set_sem_locked(iocb);
+ #include <linux/rwsem.h>
+@@ -1070,6 +1073,7 @@ struct sched_domain;
+ #define WF_SYNC 0x01 /* waker goes to sleep after wakup */
+ #define WF_FORK 0x02 /* child wakeup after fork */
+ #define WF_MIGRATED 0x04 /* internal use, task got migrated */
++#define WF_LOCK_SLEEPER 0x08 /* wakeup spinlock "sleeper" */
-@@ -2575,7 +2575,7 @@ static ssize_t ocfs2_file_aio_read(struc
+ #define ENQUEUE_WAKEUP 1
+ #define ENQUEUE_HEAD 2
+@@ -1219,6 +1223,7 @@ enum perf_event_task_context {
- bail:
- if (have_alloc_sem) {
-- up_read(&inode->i_alloc_sem);
-+ anon_up_read(&inode->i_alloc_sem);
- ocfs2_iocb_clear_sem_locked(iocb);
- }
- if (rw_level != -1)
-Index: linux-3.0/fs/reiserfs/xattr.c
-===================================================================
---- linux-3.0.orig/fs/reiserfs/xattr.c
-+++ linux-3.0/fs/reiserfs/xattr.c
-@@ -555,11 +555,11 @@ reiserfs_xattr_set_handle(struct reiserf
+ struct task_struct {
+ volatile long state; /* -1 unrunnable, 0 runnable, >0 stopped */
++ volatile long saved_state; /* saved state for "spinlock sleepers" */
+ void *stack;
+ atomic_t usage;
+ unsigned int flags; /* per process flags, defined below */
+@@ -1255,14 +1260,17 @@ struct task_struct {
+ #endif
- reiserfs_write_unlock(inode->i_sb);
- mutex_lock_nested(&dentry->d_inode->i_mutex, I_MUTEX_XATTR);
-- down_write(&dentry->d_inode->i_alloc_sem);
-+ anon_down_write(&dentry->d_inode->i_alloc_sem);
- reiserfs_write_lock(inode->i_sb);
+ unsigned int policy;
++#ifdef CONFIG_PREEMPT_RT_FULL
++ int migrate_disable;
++#ifdef CONFIG_SCHED_DEBUG
++ int migrate_disable_atomic;
++#endif
++#endif
+ cpumask_t cpus_allowed;
- err = reiserfs_setattr(dentry, &newattrs);
-- up_write(&dentry->d_inode->i_alloc_sem);
-+ anon_up_write(&dentry->d_inode->i_alloc_sem);
- mutex_unlock(&dentry->d_inode->i_mutex);
- } else
- update_ctime(inode);
-Index: linux-3.0/include/linux/fs.h
-===================================================================
---- linux-3.0.orig/include/linux/fs.h
-+++ linux-3.0/include/linux/fs.h
-@@ -777,7 +777,7 @@ struct inode {
- struct timespec i_ctime;
- blkcnt_t i_blocks;
- unsigned short i_bytes;
-- struct rw_semaphore i_alloc_sem;
-+ struct rw_anon_semaphore i_alloc_sem;
- const struct file_operations *i_fop; /* former ->i_op->default_file_ops */
- struct file_lock *i_flock;
- struct address_space *i_mapping;
-Index: linux-3.0/mm/truncate.c
-===================================================================
---- linux-3.0.orig/mm/truncate.c
-+++ linux-3.0/mm/truncate.c
-@@ -622,12 +622,12 @@ int vmtruncate_range(struct inode *inode
- return -ENOSYS;
+ #ifdef CONFIG_PREEMPT_RCU
+ int rcu_read_lock_nesting;
+ char rcu_read_unlock_special;
+-#if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU)
+- int rcu_boosted;
+-#endif /* #if defined(CONFIG_RCU_BOOST) && defined(CONFIG_TREE_PREEMPT_RCU) */
+ struct list_head rcu_node_entry;
+ #endif /* #ifdef CONFIG_PREEMPT_RCU */
+ #ifdef CONFIG_TREE_PREEMPT_RCU
+@@ -1356,6 +1364,9 @@ struct task_struct {
- mutex_lock(&inode->i_mutex);
-- down_write(&inode->i_alloc_sem);
-+ anon_down_write(&inode->i_alloc_sem);
- unmap_mapping_range(mapping, offset, (end - offset), 1);
- inode->i_op->truncate_range(inode, offset, end);
- /* unmap again to remove racily COWed private pages */
- unmap_mapping_range(mapping, offset, (end - offset), 1);
-- up_write(&inode->i_alloc_sem);
-+ anon_up_write(&inode->i_alloc_sem);
- mutex_unlock(&inode->i_mutex);
+ struct task_cputime cputime_expires;
+ struct list_head cpu_timers[3];
++#ifdef CONFIG_PREEMPT_RT_BASE
++ struct task_struct *posix_timer_list;
++#endif
- return 0;
-Index: linux-3.0/fs/xfs/linux-2.6/mrlock.h
-===================================================================
---- linux-3.0.orig/fs/xfs/linux-2.6/mrlock.h
-+++ linux-3.0/fs/xfs/linux-2.6/mrlock.h
-@@ -21,18 +21,18 @@
- #include <linux/rwsem.h>
+ /* process credentials */
+ const struct cred __rcu *real_cred; /* objective and real subjective task
+@@ -1389,6 +1400,7 @@ struct task_struct {
+ /* signal handlers */
+ struct signal_struct *signal;
+ struct sighand_struct *sighand;
++ struct sigqueue *sigqueue_cache;
- typedef struct {
-- struct rw_semaphore mr_lock;
-+ struct rw_anon_semaphore mr_lock;
- #ifdef DEBUG
-- int mr_writer;
-+ int mr_writer;
+ sigset_t blocked, real_blocked;
+ sigset_t saved_sigmask; /* restored if set_restore_sigmask() was used */
+@@ -1432,6 +1444,9 @@ struct task_struct {
+ /* mutex deadlock detection */
+ struct mutex_waiter *blocked_on;
#endif
- } mrlock_t;
-
- #ifdef DEBUG
- #define mrinit(mrp, name) \
-- do { (mrp)->mr_writer = 0; init_rwsem(&(mrp)->mr_lock); } while (0)
-+ do { (mrp)->mr_writer = 0; init_anon_rwsem(&(mrp)->mr_lock); } while (0)
- #else
- #define mrinit(mrp, name) \
-- do { init_rwsem(&(mrp)->mr_lock); } while (0)
-+ do { init_anon_rwsem(&(mrp)->mr_lock); } while (0)
++#ifdef CONFIG_PREEMPT_RT_FULL
++ int pagefault_disabled;
++#endif
+ #ifdef CONFIG_TRACE_IRQFLAGS
+ unsigned int irq_events;
+ unsigned long hardirq_enable_ip;
+@@ -1558,6 +1573,12 @@ struct task_struct {
+ unsigned long trace;
+ /* bitmask and counter of trace recursion */
+ unsigned long trace_recursion;
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ u64 preempt_timestamp_hist;
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ unsigned long timer_offset;
++#endif
++#endif
+ #endif /* CONFIG_TRACING */
+ #ifdef CONFIG_CGROUP_MEM_RES_CTLR /* memcg uses this to do batch job */
+ struct memcg_batch_info {
+@@ -1570,10 +1591,26 @@ struct task_struct {
+ #ifdef CONFIG_HAVE_HW_BREAKPOINT
+ atomic_t ptrace_bp_refcnt;
#endif
++#ifdef CONFIG_PREEMPT_RT_BASE
++ struct rcu_head put_rcu;
++ int softirq_nestcnt;
++#endif
++#if defined CONFIG_PREEMPT_RT_FULL && defined CONFIG_HIGHMEM
++ int kmap_idx;
++ pte_t kmap_pte[KM_TYPE_NR];
++#endif
+ };
- #define mrlock_init(mrp, t,n,s) mrinit(mrp, n)
-@@ -40,12 +40,12 @@ typedef struct {
+-/* Future-safe accessor for struct task_struct's cpus_allowed. */
+-#define tsk_cpus_allowed(tsk) (&(tsk)->cpus_allowed)
++#ifdef CONFIG_PREEMPT_RT_FULL
++static inline bool cur_pf_disabled(void) { return current->pagefault_disabled; }
++#else
++static inline bool cur_pf_disabled(void) { return false; }
++#endif
++
++static inline bool pagefault_disabled(void)
++{
++ return in_atomic() || cur_pf_disabled();
++}
- static inline void mraccess_nested(mrlock_t *mrp, int subclass)
- {
-- down_read_nested(&mrp->mr_lock, subclass);
-+ anon_down_read_nested(&mrp->mr_lock, subclass);
- }
+ /*
+ * Priority of a process goes from 0..MAX_PRIO-1, valid RT
+@@ -1743,6 +1780,15 @@ extern struct pid *cad_pid;
+ extern void free_task(struct task_struct *tsk);
+ #define get_task_struct(tsk) do { atomic_inc(&(tsk)->usage); } while(0)
- static inline void mrupdate_nested(mrlock_t *mrp, int subclass)
- {
-- down_write_nested(&mrp->mr_lock, subclass);
-+ anon_down_write_nested(&mrp->mr_lock, subclass);
- #ifdef DEBUG
- mrp->mr_writer = 1;
- #endif
-@@ -53,12 +53,12 @@ static inline void mrupdate_nested(mrloc
++#ifdef CONFIG_PREEMPT_RT_BASE
++extern void __put_task_struct_cb(struct rcu_head *rhp);
++
++static inline void put_task_struct(struct task_struct *t)
++{
++ if (atomic_dec_and_test(&t->usage))
++ call_rcu(&t->put_rcu, __put_task_struct_cb);
++}
++#else
+ extern void __put_task_struct(struct task_struct *t);
- static inline int mrtryaccess(mrlock_t *mrp)
- {
-- return down_read_trylock(&mrp->mr_lock);
-+ return anon_down_read_trylock(&mrp->mr_lock);
+ static inline void put_task_struct(struct task_struct *t)
+@@ -1750,6 +1796,7 @@ static inline void put_task_struct(struct task_struct *t)
+ if (atomic_dec_and_test(&t->usage))
+ __put_task_struct(t);
}
++#endif
- static inline int mrtryupdate(mrlock_t *mrp)
- {
-- if (!down_write_trylock(&mrp->mr_lock))
-+ if (!anon_down_write_trylock(&mrp->mr_lock))
- return 0;
- #ifdef DEBUG
- mrp->mr_writer = 1;
-@@ -71,12 +71,12 @@ static inline void mrunlock_excl(mrlock_
- #ifdef DEBUG
- mrp->mr_writer = 0;
- #endif
-- up_write(&mrp->mr_lock);
-+ anon_up_write(&mrp->mr_lock);
- }
+ extern void task_times(struct task_struct *p, cputime_t *ut, cputime_t *st);
+ extern void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *st);
+@@ -1774,6 +1821,7 @@ extern void thread_group_times(struct task_struct *p, cputime_t *ut, cputime_t *
+ #define PF_FROZEN 0x00010000 /* frozen for system suspend */
+ #define PF_FSTRANS 0x00020000 /* inside a filesystem transaction */
+ #define PF_KSWAPD 0x00040000 /* I am kswapd */
++#define PF_STOMPER 0x00080000 /* I am a stomp machine thread */
+ #define PF_LESS_THROTTLE 0x00100000 /* Throttle me less: I clean memory */
+ #define PF_KTHREAD 0x00200000 /* I am a kernel thread */
+ #define PF_RANDOMIZE 0x00400000 /* randomize virtual address space */
+@@ -2021,15 +2069,27 @@ static inline void sched_autogroup_exit(struct signal_struct *sig) { }
+ #endif
- static inline void mrunlock_shared(mrlock_t *mrp)
+ #ifdef CONFIG_RT_MUTEXES
++extern void task_setprio(struct task_struct *p, int prio);
+ extern int rt_mutex_getprio(struct task_struct *p);
+-extern void rt_mutex_setprio(struct task_struct *p, int prio);
++static inline void rt_mutex_setprio(struct task_struct *p, int prio)
++{
++ task_setprio(p, prio);
++}
+ extern void rt_mutex_adjust_pi(struct task_struct *p);
++static inline bool tsk_is_pi_blocked(struct task_struct *tsk)
++{
++ return tsk->pi_blocked_on != NULL;
++}
+ #else
+ static inline int rt_mutex_getprio(struct task_struct *p)
{
-- up_read(&mrp->mr_lock);
-+ anon_up_read(&mrp->mr_lock);
+ return p->normal_prio;
}
-
- static inline void mrdemote(mrlock_t *mrp)
-@@ -84,7 +84,7 @@ static inline void mrdemote(mrlock_t *mr
- #ifdef DEBUG
- mrp->mr_writer = 0;
+ # define rt_mutex_adjust_pi(p) do { } while (0)
++static inline bool tsk_is_pi_blocked(struct task_struct *tsk)
++{
++ return false;
++}
#endif
-- downgrade_write(&mrp->mr_lock);
-+ anon_downgrade_write(&mrp->mr_lock);
- }
-
- #endif /* __XFS_SUPPORT_MRLOCK_H__ */
-Index: linux-3.0/fs/xfs/linux-2.6/xfs_super.c
-===================================================================
---- linux-3.0.orig/fs/xfs/linux-2.6/xfs_super.c
-+++ linux-3.0/fs/xfs/linux-2.6/xfs_super.c
-@@ -1002,7 +1002,7 @@ xfs_fs_evict_inode(
- * (and basically indicate what we are doing), we explicitly
- * re-init the iolock here.
- */
-- ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
-+ ASSERT(!anon_rwsem_is_locked(&ip->i_iolock.mr_lock));
- mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
- lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
- &xfs_iolock_reclaimable, "xfs_iolock_reclaimable");
-Index: linux-3.0/fs/xfs/xfs_iget.c
-===================================================================
---- linux-3.0.orig/fs/xfs/xfs_iget.c
-+++ linux-3.0/fs/xfs/xfs_iget.c
-@@ -272,7 +272,7 @@ xfs_iget_cache_hit(
- __xfs_inode_clear_reclaim_tag(mp, pag, ip);
- inode->i_state = I_NEW;
-- ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock));
-+ ASSERT(!anon_rwsem_is_locked(&ip->i_iolock.mr_lock));
- mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino);
- lockdep_set_class_and_name(&ip->i_iolock.mr_lock,
- &xfs_iolock_active, "xfs_iolock_active");
-@@ -706,13 +706,13 @@ xfs_isilocked(
- if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) {
- if (!(lock_flags & XFS_ILOCK_SHARED))
- return !!ip->i_lock.mr_writer;
-- return rwsem_is_locked(&ip->i_lock.mr_lock);
-+ return anon_rwsem_is_locked(&ip->i_lock.mr_lock);
- }
+ extern bool yield_to(struct task_struct *p, bool preempt);
+@@ -2109,6 +2169,7 @@ extern void xtime_update(unsigned long ticks);
- if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) {
- if (!(lock_flags & XFS_IOLOCK_SHARED))
- return !!ip->i_iolock.mr_writer;
-- return rwsem_is_locked(&ip->i_iolock.mr_lock);
-+ return anon_rwsem_is_locked(&ip->i_iolock.mr_lock);
- }
+ extern int wake_up_state(struct task_struct *tsk, unsigned int state);
+ extern int wake_up_process(struct task_struct *tsk);
++extern int wake_up_lock_sleeper(struct task_struct * tsk);
+ extern void wake_up_new_task(struct task_struct *tsk);
+ #ifdef CONFIG_SMP
+ extern void kick_process(struct task_struct *tsk);
+@@ -2198,12 +2259,24 @@ extern struct mm_struct * mm_alloc(void);
- ASSERT(0);
-Index: linux-3.0/Documentation/trace/histograms.txt
-===================================================================
---- /dev/null
-+++ linux-3.0/Documentation/trace/histograms.txt
-@@ -0,0 +1,186 @@
-+ Using the Linux Kernel Latency Histograms
-+
-+
-+This document gives a short explanation how to enable, configure and use
-+latency histograms. Latency histograms are primarily relevant in the
-+context of real-time enabled kernels (CONFIG_PREEMPT/CONFIG_PREEMPT_RT)
-+and are used in the quality management of the Linux real-time
-+capabilities.
-+
-+
-+* Purpose of latency histograms
-+
-+A latency histogram continuously accumulates the frequencies of latency
-+data. There are two types of histograms
-+- potential sources of latencies
-+- effective latencies
-+
-+
-+* Potential sources of latencies
-+
-+Potential sources of latencies are code segments where interrupts,
-+preemption or both are disabled (aka critical sections). To create
-+histograms of potential sources of latency, the kernel stores the time
-+stamp at the start of a critical section, determines the time elapsed
-+when the end of the section is reached, and increments the frequency
-+counter of that latency value - irrespective of whether any concurrently
-+running process is affected by latency or not.
-+- Configuration items (in the Kernel hacking/Tracers submenu)
-+ CONFIG_INTERRUPT_OFF_LATENCY
-+ CONFIG_PREEMPT_OFF_LATENCY
-+
-+
-+* Effective latencies
-+
-+Effective latencies are actually occuring during wakeup of a process. To
-+determine effective latencies, the kernel stores the time stamp when a
-+process is scheduled to be woken up, and determines the duration of the
-+wakeup time shortly before control is passed over to this process. Note
-+that the apparent latency in user space may be somewhat longer, since the
-+process may be interrupted after control is passed over to it but before
-+the execution in user space takes place. Simply measuring the interval
-+between enqueuing and wakeup may also not appropriate in cases when a
-+process is scheduled as a result of a timer expiration. The timer may have
-+missed its deadline, e.g. due to disabled interrupts, but this latency
-+would not be registered. Therefore, the offsets of missed timers are
-+recorded in a separate histogram. If both wakeup latency and missed timer
-+offsets are configured and enabled, a third histogram may be enabled that
-+records the overall latency as a sum of the timer latency, if any, and the
-+wakeup latency. This histogram is called "timerandwakeup".
-+- Configuration items (in the Kernel hacking/Tracers submenu)
-+ CONFIG_WAKEUP_LATENCY
-+ CONFIG_MISSED_TIMER_OFSETS
-+
-+
-+* Usage
-+
-+The interface to the administration of the latency histograms is located
-+in the debugfs file system. To mount it, either enter
-+
-+mount -t sysfs nodev /sys
-+mount -t debugfs nodev /sys/kernel/debug
-+
-+from shell command line level, or add
-+
-+nodev /sys sysfs defaults 0 0
-+nodev /sys/kernel/debug debugfs defaults 0 0
-+
-+to the file /etc/fstab. All latency histogram related files are then
-+available in the directory /sys/kernel/debug/tracing/latency_hist. A
-+particular histogram type is enabled by writing non-zero to the related
-+variable in the /sys/kernel/debug/tracing/latency_hist/enable directory.
-+Select "preemptirqsoff" for the histograms of potential sources of
-+latencies and "wakeup" for histograms of effective latencies etc. The
-+histogram data - one per CPU - are available in the files
-+
-+/sys/kernel/debug/tracing/latency_hist/preemptoff/CPUx
-+/sys/kernel/debug/tracing/latency_hist/irqsoff/CPUx
-+/sys/kernel/debug/tracing/latency_hist/preemptirqsoff/CPUx
-+/sys/kernel/debug/tracing/latency_hist/wakeup/CPUx
-+/sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio/CPUx
-+/sys/kernel/debug/tracing/latency_hist/missed_timer_offsets/CPUx
-+/sys/kernel/debug/tracing/latency_hist/timerandwakeup/CPUx
-+
-+The histograms are reset by writing non-zero to the file "reset" in a
-+particular latency directory. To reset all latency data, use
-+
-+#!/bin/sh
-+
-+TRACINGDIR=/sys/kernel/debug/tracing
-+HISTDIR=$TRACINGDIR/latency_hist
-+
-+if test -d $HISTDIR
-+then
-+ cd $HISTDIR
-+ for i in `find . | grep /reset$`
-+ do
-+ echo 1 >$i
-+ done
-+fi
-+
-+
-+* Data format
-+
-+Latency data are stored with a resolution of one microsecond. The
-+maximum latency is 10,240 microseconds. The data are only valid, if the
-+overflow register is empty. Every output line contains the latency in
-+microseconds in the first row and the number of samples in the second
-+row. To display only lines with a positive latency count, use, for
-+example,
-+
-+grep -v " 0$" /sys/kernel/debug/tracing/latency_hist/preemptoff/CPU0
-+
-+#Minimum latency: 0 microseconds.
-+#Average latency: 0 microseconds.
-+#Maximum latency: 25 microseconds.
-+#Total samples: 3104770694
-+#There are 0 samples greater or equal than 10240 microseconds
-+#usecs samples
-+ 0 2984486876
-+ 1 49843506
-+ 2 58219047
-+ 3 5348126
-+ 4 2187960
-+ 5 3388262
-+ 6 959289
-+ 7 208294
-+ 8 40420
-+ 9 4485
-+ 10 14918
-+ 11 18340
-+ 12 25052
-+ 13 19455
-+ 14 5602
-+ 15 969
-+ 16 47
-+ 17 18
-+ 18 14
-+ 19 1
-+ 20 3
-+ 21 2
-+ 22 5
-+ 23 2
-+ 25 1
-+
-+
-+* Wakeup latency of a selected process
-+
-+To only collect wakeup latency data of a particular process, write the
-+PID of the requested process to
-+
-+/sys/kernel/debug/tracing/latency_hist/wakeup/pid
-+
-+PIDs are not considered, if this variable is set to 0.
-+
+ /* mmdrop drops the mm and the page tables */
+ extern void __mmdrop(struct mm_struct *);
+
-+* Details of the process with the highest wakeup latency so far
+ static inline void mmdrop(struct mm_struct * mm)
+ {
+ if (unlikely(atomic_dec_and_test(&mm->mm_count)))
+ __mmdrop(mm);
+ }
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++extern void __mmdrop_delayed(struct rcu_head *rhp);
++static inline void mmdrop_delayed(struct mm_struct *mm)
++{
++ if (atomic_dec_and_test(&mm->mm_count))
++ call_rcu(&mm->delayed_drop, __mmdrop_delayed);
++}
++#else
++# define mmdrop_delayed(mm) mmdrop(mm)
++#endif
+
-+Selected data of the process that suffered from the highest wakeup
-+latency that occurred in a particular CPU are available in the file
+ /* mmput gets rid of the mappings and all user-space */
+ extern void mmput(struct mm_struct *);
+ /* Grab a reference to a task's mm, if it is not already going away */
+@@ -2509,7 +2582,7 @@ extern int _cond_resched(void);
+
+ extern int __cond_resched_lock(spinlock_t *lock);
+
+-#ifdef CONFIG_PREEMPT
++#if defined(CONFIG_PREEMPT) && !defined(CONFIG_PREEMPT_RT_FULL)
+ #define PREEMPT_LOCK_OFFSET PREEMPT_OFFSET
+ #else
+ #define PREEMPT_LOCK_OFFSET 0
+@@ -2520,12 +2593,16 @@ extern int __cond_resched_lock(spinlock_t *lock);
+ __cond_resched_lock(lock); \
+ })
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ extern int __cond_resched_softirq(void);
+
+ #define cond_resched_softirq() ({ \
+ __might_sleep(__FILE__, __LINE__, SOFTIRQ_DISABLE_OFFSET); \
+ __cond_resched_softirq(); \
+ })
++#else
++# define cond_resched_softirq() cond_resched()
++#endif
+
+ /*
+ * Does a critical section need to be broken due to another
+@@ -2549,7 +2626,7 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
+
+ static inline void thread_group_cputime_init(struct signal_struct *sig)
+ {
+- spin_lock_init(&sig->cputimer.lock);
++ raw_spin_lock_init(&sig->cputimer.lock);
+ }
+
+ /*
+@@ -2588,6 +2665,26 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu)
+
+ #endif /* CONFIG_SMP */
+
++static inline int __migrate_disabled(struct task_struct *p)
++{
++#ifdef CONFIG_PREEMPT_RT_FULL
++ return p->migrate_disable;
++#else
++ return 0;
++#endif
++}
+
-+/sys/kernel/debug/tracing/latency_hist/wakeup/max_latency-CPUx.
++/* Future-safe accessor for struct task_struct's cpus_allowed. */
++static inline const struct cpumask *tsk_cpus_allowed(struct task_struct *p)
++{
++#ifdef CONFIG_PREEMPT_RT_FULL
++ if (p->migrate_disable)
++ return cpumask_of(task_cpu(p));
++#endif
+
-+In addition, other relevant system data at the time when the
-+latency occurred are given.
++ return &p->cpus_allowed;
++}
+
-+The format of the data is (all in one line):
-+<PID> <Priority> <Latency> (<Timeroffset>) <Command> \
-+<- <PID> <Priority> <Command> <Timestamp>
+ extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask);
+ extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
+
+diff --git a/include/linux/semaphore.h b/include/linux/semaphore.h
+index 39fa049..dc368b8 100644
+--- a/include/linux/semaphore.h
++++ b/include/linux/semaphore.h
+@@ -14,14 +14,14 @@
+
+ /* Please don't access any members of this structure directly */
+ struct semaphore {
+- spinlock_t lock;
++ raw_spinlock_t lock;
+ unsigned int count;
+ struct list_head wait_list;
+ };
+
+ #define __SEMAPHORE_INITIALIZER(name, n) \
+ { \
+- .lock = __SPIN_LOCK_UNLOCKED((name).lock), \
++ .lock = __RAW_SPIN_LOCK_UNLOCKED((name).lock), \
+ .count = n, \
+ .wait_list = LIST_HEAD_INIT((name).wait_list), \
+ }
+diff --git a/include/linux/seqlock.h b/include/linux/seqlock.h
+index c6db9fb..e262353 100644
+--- a/include/linux/seqlock.h
++++ b/include/linux/seqlock.h
+@@ -30,92 +30,12 @@
+ #include <linux/preempt.h>
+ #include <asm/processor.h>
+
+-typedef struct {
+- unsigned sequence;
+- spinlock_t lock;
+-} seqlock_t;
+-
+-/*
+- * These macros triggered gcc-3.x compile-time problems. We think these are
+- * OK now. Be cautious.
+- */
+-#define __SEQLOCK_UNLOCKED(lockname) \
+- { 0, __SPIN_LOCK_UNLOCKED(lockname) }
+-
+-#define seqlock_init(x) \
+- do { \
+- (x)->sequence = 0; \
+- spin_lock_init(&(x)->lock); \
+- } while (0)
+-
+-#define DEFINE_SEQLOCK(x) \
+- seqlock_t x = __SEQLOCK_UNLOCKED(x)
+-
+-/* Lock out other writers and update the count.
+- * Acts like a normal spin_lock/unlock.
+- * Don't need preempt_disable() because that is in the spin_lock already.
+- */
+-static inline void write_seqlock(seqlock_t *sl)
+-{
+- spin_lock(&sl->lock);
+- ++sl->sequence;
+- smp_wmb();
+-}
+-
+-static inline void write_sequnlock(seqlock_t *sl)
+-{
+- smp_wmb();
+- sl->sequence++;
+- spin_unlock(&sl->lock);
+-}
+-
+-static inline int write_tryseqlock(seqlock_t *sl)
+-{
+- int ret = spin_trylock(&sl->lock);
+-
+- if (ret) {
+- ++sl->sequence;
+- smp_wmb();
+- }
+- return ret;
+-}
+-
+-/* Start of read calculation -- fetch last complete writer token */
+-static __always_inline unsigned read_seqbegin(const seqlock_t *sl)
+-{
+- unsigned ret;
+-
+-repeat:
+- ret = ACCESS_ONCE(sl->sequence);
+- if (unlikely(ret & 1)) {
+- cpu_relax();
+- goto repeat;
+- }
+- smp_rmb();
+-
+- return ret;
+-}
+-
+-/*
+- * Test if reader processed invalid data.
+- *
+- * If sequence value changed then writer changed data while in section.
+- */
+-static __always_inline int read_seqretry(const seqlock_t *sl, unsigned start)
+-{
+- smp_rmb();
+-
+- return unlikely(sl->sequence != start);
+-}
+-
+-
+ /*
+ * Version using sequence counter only.
+ * This can be used when code has its own mutex protecting the
+ * updating starting before the write_seqcountbeqin() and ending
+ * after the write_seqcount_end().
+ */
+-
+ typedef struct seqcount {
+ unsigned sequence;
+ } seqcount_t;
+@@ -197,7 +117,6 @@ static inline int __read_seqcount_retry(const seqcount_t *s, unsigned start)
+ static inline int read_seqcount_retry(const seqcount_t *s, unsigned start)
+ {
+ smp_rmb();
+-
+ return __read_seqcount_retry(s, start);
+ }
+
+@@ -231,31 +150,154 @@ static inline void write_seqcount_barrier(seqcount_t *s)
+ s->sequence+=2;
+ }
+
++typedef struct {
++ struct seqcount seqcount;
++ raw_spinlock_t lock;
++} raw_seqlock_t;
+
-+The value of <Timeroffset> is only relevant in the combined timer
-+and wakeup latency recording. In the wakeup recording, it is
-+always 0, in the missed_timer_offsets recording, it is the same
-+as <Latency>.
++typedef struct {
++ struct seqcount seqcount;
++ spinlock_t lock;
++} seqlock_t;
+
-+When retrospectively searching for the origin of a latency and
-+tracing was not enabled, it may be helpful to know the name and
-+some basic data of the task that (finally) was switching to the
-+late real-tlme task. In addition to the victim's data, also the
-+data of the possible culprit are therefore displayed after the
-+"<-" symbol.
++/*
++ * These macros triggered gcc-3.x compile-time problems. We think these are
++ * OK now. Be cautious.
++ */
++#define __RAW_SEQLOCK_UNLOCKED(lockname) \
++ { \
++ .seqcount = SEQCNT_ZERO, \
++ .lock = __RAW_SPIN_LOCK_UNLOCKED(lockname) \
++ }
+
-+Finally, the timestamp of the time when the latency occurred
-+in <seconds>.<microseconds> after the most recent system boot
-+is provided.
++#define raw_seqlock_init(x) \
++ do { \
++ seqcount_init(&(x)->seqcount); \
++ raw_spin_lock_init(&(x)->lock); \
++ } while (0)
+
-+These data are also reset when the wakeup histogram is reset.
-Index: linux-3.0/include/trace/events/hist.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/trace/events/hist.h
-@@ -0,0 +1,69 @@
-+#undef TRACE_SYSTEM
-+#define TRACE_SYSTEM hist
++#define DEFINE_RAW_SEQLOCK(x) \
++ raw_seqlock_t x = __RAW_SEQLOCK_UNLOCKED(x)
+
-+#if !defined(_TRACE_HIST_H) || defined(TRACE_HEADER_MULTI_READ)
-+#define _TRACE_HIST_H
++#define __SEQLOCK_UNLOCKED(lockname) \
++ { \
++ .seqcount = SEQCNT_ZERO, \
++ .lock = __SPIN_LOCK_UNLOCKED(lockname) \
++ }
+
-+#include "latency_hist.h"
-+#include <linux/tracepoint.h>
++#define seqlock_init(x) \
++ do { \
++ seqcount_init(&(x)->seqcount); \
++ spin_lock_init(&(x)->lock); \
++ } while (0)
+
-+#if !defined(CONFIG_PREEMPT_OFF_HIST) && !defined(CONFIG_INTERRUPT_OFF_HIST)
-+#define trace_preemptirqsoff_hist(a,b)
-+#else
-+TRACE_EVENT(preemptirqsoff_hist,
++#define DEFINE_SEQLOCK(x) \
++ seqlock_t x = __SEQLOCK_UNLOCKED(x)
+
-+ TP_PROTO(int reason, int starthist),
++#define read_seqbegin(sl) read_seqcount_begin(&(sl)->seqcount)
++#define read_seqretry(sl, start) read_seqcount_retry(&(sl)->seqcount, start)
+
-+ TP_ARGS(reason, starthist),
+ /*
+- * Possible sw/hw IRQ protected versions of the interfaces.
++ * Lock out other writers and update the count.
++ * Acts like a normal spin_lock/unlock.
++ * Don't need preempt_disable() because that is in the spin_lock already.
+ */
++static inline void raw_write_seqlock(raw_seqlock_t *sl)
++{
++ raw_spin_lock(&sl->lock);
++ write_seqcount_begin(&sl->seqcount);
++}
+
-+ TP_STRUCT__entry(
-+ __field(int, reason )
-+ __field(int, starthist )
-+ ),
++static inline void raw_write_sequnlock(raw_seqlock_t *sl)
++{
++ write_seqcount_end(&sl->seqcount);
++ raw_spin_unlock(&sl->lock);
++}
+
-+ TP_fast_assign(
-+ __entry->reason = reason;
-+ __entry->starthist = starthist;
-+ ),
++static inline void raw_write_seqlock_irq(raw_seqlock_t *sl)
++{
++ raw_spin_lock_irq(&sl->lock);
++ write_seqcount_begin(&sl->seqcount);
++}
+
-+ TP_printk("reason=%s starthist=%s", getaction(__entry->reason),
-+ __entry->starthist ? "start" : "stop")
-+);
-+#endif
++static inline void raw_write_sequnlock_irq(raw_seqlock_t *sl)
++{
++ write_seqcount_end(&sl->seqcount);
++ raw_spin_unlock_irq(&sl->lock);
++}
+
-+#ifndef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+#define trace_hrtimer_interrupt(a,b,c,d)
-+#else
-+TRACE_EVENT(hrtimer_interrupt,
++static inline unsigned long __raw_write_seqlock_irqsave(raw_seqlock_t *sl)
++{
++ unsigned long flags;
+
-+ TP_PROTO(int cpu, long long offset, struct task_struct *curr, struct task_struct *task),
++ raw_spin_lock_irqsave(&sl->lock, flags);
++ write_seqcount_begin(&sl->seqcount);
++ return flags;
++}
+
-+ TP_ARGS(cpu, offset, curr, task),
++#define raw_write_seqlock_irqsave(lock, flags) \
++ do { flags = __raw_write_seqlock_irqsave(lock); } while (0)
+
-+ TP_STRUCT__entry(
-+ __field(int, cpu )
-+ __field(long long, offset )
-+ __array(char, ccomm, TASK_COMM_LEN)
-+ __field(int, cprio )
-+ __array(char, tcomm, TASK_COMM_LEN)
-+ __field(int, tprio )
-+ ),
++static inline void
++raw_write_sequnlock_irqrestore(raw_seqlock_t *sl, unsigned long flags)
++{
++ write_seqcount_end(&sl->seqcount);
++ raw_spin_unlock_irqrestore(&sl->lock, flags);
++}
+
-+ TP_fast_assign(
-+ __entry->cpu = cpu;
-+ __entry->offset = offset;
-+ memcpy(__entry->ccomm, curr->comm, TASK_COMM_LEN);
-+ __entry->cprio = curr->prio;
-+ memcpy(__entry->tcomm, task != NULL ? task->comm : "<none>", task != NULL ? TASK_COMM_LEN : 7);
-+ __entry->tprio = task != NULL ? task->prio : -1;
-+ ),
++static inline void write_seqlock(seqlock_t *sl)
++{
++ spin_lock(&sl->lock);
++ write_seqcount_begin(&sl->seqcount);
++}
+
-+ TP_printk("cpu=%d offset=%lld curr=%s[%d] thread=%s[%d]",
-+ __entry->cpu, __entry->offset, __entry->ccomm, __entry->cprio, __entry->tcomm, __entry->tprio)
-+);
-+#endif
++static inline void write_sequnlock(seqlock_t *sl)
++{
++ write_seqcount_end(&sl->seqcount);
++ spin_unlock(&sl->lock);
++}
+
-+#endif /* _TRACE_HIST_H */
++static inline void write_seqlock_bh(seqlock_t *sl)
++{
++ spin_lock_bh(&sl->lock);
++ write_seqcount_begin(&sl->seqcount);
++}
+
-+/* This part must be outside protection */
-+#include <trace/define_trace.h>
-Index: linux-3.0/include/trace/events/latency_hist.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/trace/events/latency_hist.h
-@@ -0,0 +1,30 @@
-+#ifndef _LATENCY_HIST_H
-+#define _LATENCY_HIST_H
++static inline void write_sequnlock_bh(seqlock_t *sl)
++{
++ write_seqcount_end(&sl->seqcount);
++ spin_unlock_bh(&sl->lock);
++}
+
-+enum hist_action {
-+ IRQS_ON,
-+ PREEMPT_ON,
-+ TRACE_STOP,
-+ IRQS_OFF,
-+ PREEMPT_OFF,
-+ TRACE_START,
-+};
++static inline void write_seqlock_irq(seqlock_t *sl)
++{
++ spin_lock_irq(&sl->lock);
++ write_seqcount_begin(&sl->seqcount);
++}
+
-+static char *actions[] = {
-+ "IRQS_ON",
-+ "PREEMPT_ON",
-+ "TRACE_STOP",
-+ "IRQS_OFF",
-+ "PREEMPT_OFF",
-+ "TRACE_START",
-+};
++static inline void write_sequnlock_irq(seqlock_t *sl)
++{
++ write_seqcount_end(&sl->seqcount);
++ spin_unlock_irq(&sl->lock);
++}
+
-+static inline char *getaction(int action)
++static inline unsigned long __write_seqlock_irqsave(seqlock_t *sl)
+{
-+ if (action >= 0 && action <= sizeof(actions)/sizeof(actions[0]))
-+ return(actions[action]);
-+ return("unknown");
++ unsigned long flags;
++
++ spin_lock_irqsave(&sl->lock, flags);
++ write_seqcount_begin(&sl->seqcount);
++ return flags;
+}
+
-+#endif /* _LATENCY_HIST_H */
+ #define write_seqlock_irqsave(lock, flags) \
+- do { local_irq_save(flags); write_seqlock(lock); } while (0)
+-#define write_seqlock_irq(lock) \
+- do { local_irq_disable(); write_seqlock(lock); } while (0)
+-#define write_seqlock_bh(lock) \
+- do { local_bh_disable(); write_seqlock(lock); } while (0)
+-
+-#define write_sequnlock_irqrestore(lock, flags) \
+- do { write_sequnlock(lock); local_irq_restore(flags); } while(0)
+-#define write_sequnlock_irq(lock) \
+- do { write_sequnlock(lock); local_irq_enable(); } while(0)
+-#define write_sequnlock_bh(lock) \
+- do { write_sequnlock(lock); local_bh_enable(); } while(0)
+-
+-#define read_seqbegin_irqsave(lock, flags) \
+- ({ local_irq_save(flags); read_seqbegin(lock); })
+-
+-#define read_seqretry_irqrestore(lock, iv, flags) \
+- ({ \
+- int ret = read_seqretry(lock, iv); \
+- local_irq_restore(flags); \
+- ret; \
+- })
++ do { flags = __write_seqlock_irqsave(lock); } while (0)
+
-Index: linux-3.0/kernel/hrtimer.c
-===================================================================
---- linux-3.0.orig/kernel/hrtimer.c
-+++ linux-3.0/kernel/hrtimer.c
-@@ -49,6 +49,7 @@
- #include <asm/uaccess.h>
-
- #include <trace/events/timer.h>
-+#include <trace/events/hist.h>
++static inline void
++write_sequnlock_irqrestore(seqlock_t *sl, unsigned long flags)
++{
++ write_seqcount_end(&sl->seqcount);
++ spin_unlock_irqrestore(&sl->lock, flags);
++}
- /*
- * The timer bases:
-@@ -588,8 +589,7 @@ static int hrtimer_reprogram(struct hrti
- * When the callback is running, we do not reprogram the clock event
- * device. The timer callback is either running on a different CPU or
- * the callback is executed in the hrtimer_interrupt context. The
-- * reprogramming is handled either by the softirq, which called the
-- * callback or at the end of the hrtimer_interrupt.
-+ * reprogramming is handled at the end of the hrtimer_interrupt.
- */
- if (hrtimer_callback_running(timer))
- return 0;
-@@ -624,6 +624,9 @@ static int hrtimer_reprogram(struct hrti
- return res;
+ #endif /* __LINUX_SEQLOCK_H */
+diff --git a/include/linux/signal.h b/include/linux/signal.h
+index a822300..a448900 100644
+--- a/include/linux/signal.h
++++ b/include/linux/signal.h
+@@ -229,6 +229,7 @@ static inline void init_sigpending(struct sigpending *sig)
}
-+static void __run_hrtimer(struct hrtimer *timer, ktime_t *now);
-+static int hrtimer_rt_defer(struct hrtimer *timer);
-+
- /*
- * Initialize the high resolution related parts of cpu_base
- */
-@@ -644,14 +647,23 @@ static inline int hrtimer_enqueue_reprog
- int wakeup)
- {
- if (base->cpu_base->hres_active && hrtimer_reprogram(timer, base)) {
-- if (wakeup) {
-- raw_spin_unlock(&base->cpu_base->lock);
-- raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-- raw_spin_lock(&base->cpu_base->lock);
-- } else
-- __raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-+ if (!wakeup)
-+ return -ETIME;
-
-- return 1;
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ /*
-+ * Move softirq based timers away from the rbtree in
-+ * case it expired already. Otherwise we would have a
-+ * stale base->first entry until the softirq runs.
-+ */
-+ if (!hrtimer_rt_defer(timer))
-+ return -ETIME;
-+#endif
-+ raw_spin_unlock(&base->cpu_base->lock);
-+ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
-+ raw_spin_lock(&base->cpu_base->lock);
-+
-+ return 0;
- }
+ extern void flush_sigqueue(struct sigpending *queue);
++extern void flush_task_sigqueue(struct task_struct *tsk);
- return 0;
-@@ -732,6 +744,11 @@ static inline int hrtimer_enqueue_reprog
- }
- static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { }
- static inline void retrigger_next_event(void *arg) { }
-+static inline int hrtimer_reprogram(struct hrtimer *timer,
-+ struct hrtimer_clock_base *base)
-+{
-+ return 0;
-+}
+ /* Test if 'sig' is valid signal. Use this instead of testing _NSIG directly */
+ static inline int valid_signal(unsigned long sig)
+diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
+index b920a72..14113248 100644
+--- a/include/linux/skbuff.h
++++ b/include/linux/skbuff.h
+@@ -118,6 +118,7 @@ struct sk_buff_head {
- #endif /* CONFIG_HIGH_RES_TIMERS */
+ __u32 qlen;
+ spinlock_t lock;
++ raw_spinlock_t raw_lock;
+ };
-@@ -846,6 +863,32 @@ u64 hrtimer_forward(struct hrtimer *time
+ struct sk_buff;
+@@ -875,6 +876,12 @@ static inline void skb_queue_head_init(struct sk_buff_head *list)
+ __skb_queue_head_init(list);
}
- EXPORT_SYMBOL_GPL(hrtimer_forward);
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+# define wake_up_timer_waiters(b) wake_up(&(b)->wait)
-+
-+/**
-+ * hrtimer_wait_for_timer - Wait for a running timer
-+ *
-+ * @timer: timer to wait for
-+ *
-+ * The function waits in case the timers callback function is
-+ * currently executed on the waitqueue of the timer base. The
-+ * waitqueue is woken up after the timer callback function has
-+ * finished execution.
-+ */
-+void hrtimer_wait_for_timer(const struct hrtimer *timer)
-+{
-+ struct hrtimer_clock_base *base = timer->base;
-+
-+ if (base && base->cpu_base && !timer->irqsafe)
-+ wait_event(base->cpu_base->wait,
-+ !(timer->state & HRTIMER_STATE_CALLBACK));
++static inline void skb_queue_head_init_raw(struct sk_buff_head *list)
++{
++ raw_spin_lock_init(&list->raw_lock);
++ __skb_queue_head_init(list);
+}
+
-+#else
-+# define wake_up_timer_waiters(b) do { } while (0)
-+#endif
-+
+ static inline void skb_queue_head_init_class(struct sk_buff_head *list,
+ struct lock_class_key *class)
+ {
+diff --git a/include/linux/smp.h b/include/linux/smp.h
+index 8cc38d3..78fd0a2 100644
+--- a/include/linux/smp.h
++++ b/include/linux/smp.h
+@@ -80,7 +80,6 @@ void __smp_call_function_single(int cpuid, struct call_single_data *data,
+
+ int smp_call_function_any(const struct cpumask *mask,
+ smp_call_func_t func, void *info, int wait);
+-
/*
- * enqueue_hrtimer - internal function to (re)start a timer
- *
-@@ -888,6 +931,11 @@ static void __remove_hrtimer(struct hrti
- if (!(timer->state & HRTIMER_STATE_ENQUEUED))
- goto out;
+ * Generic and arch helpers
+ */
+@@ -173,6 +172,9 @@ smp_call_function_any(const struct cpumask *mask, smp_call_func_t func,
+ #define get_cpu() ({ preempt_disable(); smp_processor_id(); })
+ #define put_cpu() preempt_enable()
-+ if (unlikely(!list_empty(&timer->cb_entry))) {
-+ list_del_init(&timer->cb_entry);
-+ goto out;
-+ }
++#define get_cpu_light() ({ migrate_disable(); smp_processor_id(); })
++#define put_cpu_light() migrate_enable()
+
- if (&timer->node == timerqueue_getnext(&base->active)) {
- #ifdef CONFIG_HIGH_RES_TIMERS
- /* Reprogram the clock event device. if enabled */
-@@ -983,8 +1031,20 @@ int __hrtimer_start_range_ns(struct hrti
- *
- * XXX send_remote_softirq() ?
- */
-- if (leftmost && new_base->cpu_base == &__get_cpu_var(hrtimer_bases))
-- hrtimer_enqueue_reprogram(timer, new_base, wakeup);
-+ if (leftmost && new_base->cpu_base == &__get_cpu_var(hrtimer_bases)) {
-+ ret = hrtimer_enqueue_reprogram(timer, new_base, wakeup);
-+ if (ret) {
-+ /*
-+ * In case we failed to reprogram the timer (mostly
-+ * because out current timer is already elapsed),
-+ * remove it again and report a failure. This avoids
-+ * stale base->first entries.
-+ */
-+ debug_deactivate(timer);
-+ __remove_hrtimer(timer, new_base,
-+ timer->state & HRTIMER_STATE_CALLBACK, 0);
-+ }
-+ }
-
- unlock_hrtimer_base(timer, &flags);
+ /*
+ * Callback to arch code if there's nosmp or maxcpus=0 on the
+ * boot command line:
+diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h
+index 0b22d51..87b185c 100644
+--- a/include/linux/spinlock.h
++++ b/include/linux/spinlock.h
+@@ -254,7 +254,11 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock)
+ #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock))
-@@ -1070,7 +1130,7 @@ int hrtimer_cancel(struct hrtimer *timer
+ /* Include rwlock functions */
+-#include <linux/rwlock.h>
++#ifdef CONFIG_PREEMPT_RT_FULL
++# include <linux/rwlock_rt.h>
++#else
++# include <linux/rwlock.h>
++#endif
- if (ret >= 0)
- return ret;
-- cpu_relax();
-+ hrtimer_wait_for_timer(timer);
- }
- }
- EXPORT_SYMBOL_GPL(hrtimer_cancel);
-@@ -1149,6 +1209,7 @@ static void __hrtimer_init(struct hrtime
+ /*
+ * Pull the _spin_*()/_read_*()/_write_*() functions/declarations:
+@@ -265,6 +269,10 @@ static inline void do_raw_spin_unlock(raw_spinlock_t *lock) __releases(lock)
+ # include <linux/spinlock_api_up.h>
+ #endif
- base = hrtimer_clockid_to_base(clock_id);
- timer->base = &cpu_base->clock_base[base];
-+ INIT_LIST_HEAD(&timer->cb_entry);
- timerqueue_init(&timer->node);
++#ifdef CONFIG_PREEMPT_RT_FULL
++# include <linux/spinlock_rt.h>
++#else /* PREEMPT_RT_FULL */
++
+ /*
+ * Map the spin_lock functions to the raw variants for PREEMPT_RT=n
+ */
+@@ -397,4 +405,6 @@ extern int _atomic_dec_and_lock(atomic_t *atomic, spinlock_t *lock);
+ #define atomic_dec_and_lock(atomic, lock) \
+ __cond_lock(lock, _atomic_dec_and_lock(atomic, lock))
- #ifdef CONFIG_TIMER_STATS
-@@ -1232,6 +1293,122 @@ static void __run_hrtimer(struct hrtimer
- timer->state &= ~HRTIMER_STATE_CALLBACK;
++#endif /* !PREEMPT_RT_FULL */
++
+ #endif /* __LINUX_SPINLOCK_H */
+diff --git a/include/linux/spinlock_api_smp.h b/include/linux/spinlock_api_smp.h
+index e253ccd..2a5ba05 100644
+--- a/include/linux/spinlock_api_smp.h
++++ b/include/linux/spinlock_api_smp.h
+@@ -191,6 +191,8 @@ static inline int __raw_spin_trylock_bh(raw_spinlock_t *lock)
+ return 0;
}
-+static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer);
-+
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+static void hrtimer_rt_reprogram(int restart, struct hrtimer *timer,
-+ struct hrtimer_clock_base *base)
-+{
-+ /*
-+ * Note, we clear the callback flag before we requeue the
-+ * timer otherwise we trigger the callback_running() check
-+ * in hrtimer_reprogram().
-+ */
-+ timer->state &= ~HRTIMER_STATE_CALLBACK;
+-#include <linux/rwlock_api_smp.h>
++#ifndef CONFIG_PREEMPT_RT_FULL
++# include <linux/rwlock_api_smp.h>
++#endif
+
+ #endif /* __LINUX_SPINLOCK_API_SMP_H */
+diff --git a/include/linux/spinlock_rt.h b/include/linux/spinlock_rt.h
+new file mode 100644
+index 0000000..3b555b4
+--- /dev/null
++++ b/include/linux/spinlock_rt.h
+@@ -0,0 +1,166 @@
++#ifndef __LINUX_SPINLOCK_RT_H
++#define __LINUX_SPINLOCK_RT_H
+
-+ if (restart != HRTIMER_NORESTART) {
-+ BUG_ON(hrtimer_active(timer));
-+ /*
-+ * Enqueue the timer, if it's the leftmost timer then
-+ * we need to reprogram it.
-+ */
-+ if (!enqueue_hrtimer(timer, base))
-+ return;
++#ifndef __LINUX_SPINLOCK_H
++#error Do not include directly. Use spinlock.h
++#endif
+
-+#ifndef CONFIG_HIGH_RES_TIMERS
-+ }
-+#else
-+ if (base->cpu_base->hres_active &&
-+ hrtimer_reprogram(timer, base))
-+ goto requeue;
++extern void
++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key);
+
-+ } else if (hrtimer_active(timer)) {
-+ /*
-+ * If the timer was rearmed on another CPU, reprogram
-+ * the event device.
-+ */
-+ if (&timer->node == base->active.next &&
-+ base->cpu_base->hres_active &&
-+ hrtimer_reprogram(timer, base))
-+ goto requeue;
-+ }
-+ return;
++#define spin_lock_init(slock) \
++do { \
++ static struct lock_class_key __key; \
++ \
++ rt_mutex_init(&(slock)->lock); \
++ __rt_spin_lock_init(slock, #slock, &__key); \
++} while (0)
+
-+requeue:
-+ /*
-+ * Timer is expired. Thus move it from tree to pending list
-+ * again.
-+ */
-+ __remove_hrtimer(timer, base, timer->state, 0);
-+ list_add_tail(&timer->cb_entry, &base->expired);
-+#endif
-+}
++extern void __lockfunc rt_spin_lock(spinlock_t *lock);
++extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock);
++extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass);
++extern void __lockfunc rt_spin_unlock(spinlock_t *lock);
++extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock);
++extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags);
++extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock);
++extern int __lockfunc rt_spin_trylock(spinlock_t *lock);
++extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock);
+
+/*
-+ * The changes in mainline which removed the callback modes from
-+ * hrtimer are not yet working with -rt. The non wakeup_process()
-+ * based callbacks which involve sleeping locks need to be treated
-+ * seperately.
++ * lockdep-less calls, for derived types like rwlock:
++ * (for trylock they can use rt_mutex_trylock() directly.
+ */
-+static void hrtimer_rt_run_pending(void)
-+{
-+ enum hrtimer_restart (*fn)(struct hrtimer *);
-+ struct hrtimer_cpu_base *cpu_base;
-+ struct hrtimer_clock_base *base;
-+ struct hrtimer *timer;
-+ int index, restart;
++extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock);
++extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock);
+
-+ local_irq_disable();
-+ cpu_base = &per_cpu(hrtimer_bases, smp_processor_id());
++#define spin_lock_local(lock) rt_spin_lock(lock)
++#define spin_unlock_local(lock) rt_spin_unlock(lock)
+
-+ raw_spin_lock(&cpu_base->lock);
++#define spin_lock(lock) \
++ do { \
++ migrate_disable(); \
++ rt_spin_lock(lock); \
++ } while (0)
+
-+ for (index = 0; index < HRTIMER_MAX_CLOCK_BASES; index++) {
-+ base = &cpu_base->clock_base[index];
++#define spin_lock_bh(lock) \
++ do { \
++ local_bh_disable(); \
++ migrate_disable(); \
++ rt_spin_lock(lock); \
++ } while (0)
++
++#define spin_lock_irq(lock) spin_lock(lock)
++
++#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock))
++
++#define spin_trylock(lock) \
++({ \
++ int __locked; \
++ migrate_disable(); \
++ __locked = spin_do_trylock(lock); \
++ if (!__locked) \
++ migrate_enable(); \
++ __locked; \
++})
++
++#ifdef CONFIG_LOCKDEP
++# define spin_lock_nested(lock, subclass) \
++ do { \
++ migrate_disable(); \
++ rt_spin_lock_nested(lock, subclass); \
++ } while (0)
++
++# define spin_lock_irqsave_nested(lock, flags, subclass) \
++ do { \
++ typecheck(unsigned long, flags); \
++ flags = 0; \
++ migrate_disable(); \
++ rt_spin_lock_nested(lock, subclass); \
++ } while (0)
++#else
++# define spin_lock_nested(lock, subclass) spin_lock(lock)
++
++# define spin_lock_irqsave_nested(lock, flags, subclass) \
++ do { \
++ typecheck(unsigned long, flags); \
++ flags = 0; \
++ spin_lock(lock); \
++ } while (0)
++#endif
++
++#define spin_lock_irqsave(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ flags = 0; \
++ spin_lock(lock); \
++ } while (0)
++
++static inline unsigned long spin_lock_trace_flags(spinlock_t *lock)
++{
++ unsigned long flags = 0;
++#ifdef CONFIG_TRACE_IRQFLAGS
++ flags = rt_spin_lock_trace_flags(lock);
++#else
++ spin_lock(lock); /* lock_local */
++#endif
++ return flags;
++}
+
-+ while (!list_empty(&base->expired)) {
-+ timer = list_first_entry(&base->expired,
-+ struct hrtimer, cb_entry);
++/* FIXME: we need rt_spin_lock_nest_lock */
++#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0)
+
-+ /*
-+ * Same as the above __run_hrtimer function
-+ * just we run with interrupts enabled.
-+ */
-+ debug_hrtimer_deactivate(timer);
-+ __remove_hrtimer(timer, base, HRTIMER_STATE_CALLBACK, 0);
-+ timer_stats_account_hrtimer(timer);
-+ fn = timer->function;
++#define spin_unlock(lock) \
++ do { \
++ rt_spin_unlock(lock); \
++ migrate_enable(); \
++ } while (0)
+
-+ raw_spin_unlock_irq(&cpu_base->lock);
-+ restart = fn(timer);
-+ raw_spin_lock_irq(&cpu_base->lock);
++#define spin_unlock_bh(lock) \
++ do { \
++ rt_spin_unlock(lock); \
++ migrate_enable(); \
++ local_bh_enable(); \
++ } while (0)
+
-+ hrtimer_rt_reprogram(restart, timer, base);
-+ }
-+ }
++#define spin_unlock_irq(lock) spin_unlock(lock)
+
-+ raw_spin_unlock_irq(&cpu_base->lock);
++#define spin_unlock_irqrestore(lock, flags) \
++ do { \
++ typecheck(unsigned long, flags); \
++ (void) flags; \
++ spin_unlock(lock); \
++ } while (0)
+
-+ wake_up_timer_waiters(cpu_base);
-+}
++#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock))
++#define spin_trylock_irq(lock) spin_trylock(lock)
+
-+static int hrtimer_rt_defer(struct hrtimer *timer)
-+{
-+ if (timer->irqsafe)
-+ return 0;
++#define spin_trylock_irqsave(lock, flags) \
++ rt_spin_trylock_irqsave(lock, &(flags))
+
-+ __remove_hrtimer(timer, timer->base, timer->state, 0);
-+ list_add_tail(&timer->cb_entry, &timer->base->expired);
-+ return 1;
-+}
++#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock)
+
++#ifdef CONFIG_GENERIC_LOCKBREAK
++# define spin_is_contended(lock) ((lock)->break_lock)
+#else
-+
-+static inline void hrtimer_rt_run_pending(void) { }
-+static inline int hrtimer_rt_defer(struct hrtimer *timer) { return 0; }
-+
++# define spin_is_contended(lock) (((void)(lock), 0))
+#endif
+
- #ifdef CONFIG_HIGH_RES_TIMERS
-
- /*
-@@ -1242,7 +1419,7 @@ void hrtimer_interrupt(struct clock_even
- {
- struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
- ktime_t expires_next, now, entry_time, delta;
-- int i, retries = 0;
-+ int i, retries = 0, raise = 0;
-
- BUG_ON(!cpu_base->hres_active);
- cpu_base->nr_events++;
-@@ -1278,6 +1455,14 @@ retry:
-
- timer = container_of(node, struct hrtimer, node);
-
-+ trace_hrtimer_interrupt(raw_smp_processor_id(),
-+ ktime_to_ns(ktime_sub(
-+ hrtimer_get_expires(timer), basenow)),
-+ current,
-+ timer->function == hrtimer_wakeup ?
-+ container_of(timer, struct hrtimer_sleeper,
-+ timer)->task : NULL);
-+
- /*
- * The immediate goal for using the softexpires is
- * minimizing wakeups, not running timers at the
-@@ -1301,7 +1486,10 @@ retry:
- break;
- }
-
-- __run_hrtimer(timer, &basenow);
-+ if (!hrtimer_rt_defer(timer))
-+ __run_hrtimer(timer, &basenow);
-+ else
-+ raise = 1;
- }
- }
-
-@@ -1316,6 +1504,10 @@ retry:
- if (expires_next.tv64 == KTIME_MAX ||
- !tick_program_event(expires_next, 0)) {
- cpu_base->hang_detected = 0;
++static inline int spin_can_lock(spinlock_t *lock)
++{
++ return !rt_mutex_is_locked(&lock->lock);
++}
+
-+ if (raise)
-+ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
++static inline int spin_is_locked(spinlock_t *lock)
++{
++ return rt_mutex_is_locked(&lock->lock);
++}
+
- return;
- }
-
-@@ -1391,17 +1583,17 @@ void hrtimer_peek_ahead_timers(void)
- local_irq_restore(flags);
- }
-
--static void run_hrtimer_softirq(struct softirq_action *h)
--{
-- hrtimer_peek_ahead_timers();
--}
--
- #else /* CONFIG_HIGH_RES_TIMERS */
-
- static inline void __hrtimer_peek_ahead_timers(void) { }
-
- #endif /* !CONFIG_HIGH_RES_TIMERS */
-
-+static void run_hrtimer_softirq(struct softirq_action *h)
++static inline void assert_spin_locked(spinlock_t *lock)
+{
-+ hrtimer_rt_run_pending();
++ BUG_ON(!spin_is_locked(lock));
+}
+
- /*
- * Called from timer softirq every jiffy, expire hrtimers:
- *
-@@ -1434,7 +1626,7 @@ void hrtimer_run_queues(void)
- struct timerqueue_node *node;
- struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
- struct hrtimer_clock_base *base;
-- int index, gettime = 1;
-+ int index, gettime = 1, raise = 0;
-
- if (hrtimer_hres_active())
- return;
-@@ -1459,10 +1651,16 @@ void hrtimer_run_queues(void)
- hrtimer_get_expires_tv64(timer))
- break;
-
-- __run_hrtimer(timer, &base->softirq_time);
-+ if (!hrtimer_rt_defer(timer))
-+ __run_hrtimer(timer, &base->softirq_time);
-+ else
-+ raise = 1;
- }
- raw_spin_unlock(&cpu_base->lock);
- }
++#define atomic_dec_and_lock(atomic, lock) \
++ atomic_dec_and_spin_lock(atomic, lock)
+
-+ if (raise)
-+ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
- }
++#endif
+diff --git a/include/linux/spinlock_types.h b/include/linux/spinlock_types.h
+index 73548eb..10bac71 100644
+--- a/include/linux/spinlock_types.h
++++ b/include/linux/spinlock_types.h
+@@ -9,80 +9,15 @@
+ * Released under the General Public License (GPL).
+ */
- /*
-@@ -1484,6 +1682,7 @@ static enum hrtimer_restart hrtimer_wake
- void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task)
- {
- sl->timer.function = hrtimer_wakeup;
-+ sl->timer.irqsafe = 1;
- sl->task = task;
- }
- EXPORT_SYMBOL_GPL(hrtimer_init_sleeper);
-@@ -1622,9 +1821,13 @@ static void __cpuinit init_hrtimers_cpu(
- for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
- cpu_base->clock_base[i].cpu_base = cpu_base;
- timerqueue_init_head(&cpu_base->clock_base[i].active);
-+ INIT_LIST_HEAD(&cpu_base->clock_base[i].expired);
- }
+-#if defined(CONFIG_SMP)
+-# include <asm/spinlock_types.h>
+-#else
+-# include <linux/spinlock_types_up.h>
+-#endif
+-
+-#include <linux/lockdep.h>
+-
+-typedef struct raw_spinlock {
+- arch_spinlock_t raw_lock;
+-#ifdef CONFIG_GENERIC_LOCKBREAK
+- unsigned int break_lock;
+-#endif
+-#ifdef CONFIG_DEBUG_SPINLOCK
+- unsigned int magic, owner_cpu;
+- void *owner;
+-#endif
+-#ifdef CONFIG_DEBUG_LOCK_ALLOC
+- struct lockdep_map dep_map;
+-#endif
+-} raw_spinlock_t;
+-
+-#define SPINLOCK_MAGIC 0xdead4ead
+-
+-#define SPINLOCK_OWNER_INIT ((void *)-1L)
+-
+-#ifdef CONFIG_DEBUG_LOCK_ALLOC
+-# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
+-#else
+-# define SPIN_DEP_MAP_INIT(lockname)
+-#endif
++#include <linux/spinlock_types_raw.h>
- hrtimer_init_hres(cpu_base);
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ init_waitqueue_head(&cpu_base->wait);
-+#endif
- }
+-#ifdef CONFIG_DEBUG_SPINLOCK
+-# define SPIN_DEBUG_INIT(lockname) \
+- .magic = SPINLOCK_MAGIC, \
+- .owner_cpu = -1, \
+- .owner = SPINLOCK_OWNER_INIT,
++#ifndef CONFIG_PREEMPT_RT_FULL
++# include <linux/spinlock_types_nort.h>
++# include <linux/rwlock_types.h>
+ #else
+-# define SPIN_DEBUG_INIT(lockname)
++# include <linux/rtmutex.h>
++# include <linux/spinlock_types_rt.h>
++# include <linux/rwlock_types_rt.h>
+ #endif
- #ifdef CONFIG_HOTPLUG_CPU
-@@ -1737,9 +1940,7 @@ void __init hrtimers_init(void)
- hrtimer_cpu_notify(&hrtimers_nb, (unsigned long)CPU_UP_PREPARE,
- (void *)(long)smp_processor_id());
- register_cpu_notifier(&hrtimers_nb);
--#ifdef CONFIG_HIGH_RES_TIMERS
- open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq);
+-#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \
+- { \
+- .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \
+- SPIN_DEBUG_INIT(lockname) \
+- SPIN_DEP_MAP_INIT(lockname) }
+-
+-#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \
+- (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname)
+-
+-#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x)
+-
+-typedef struct spinlock {
+- union {
+- struct raw_spinlock rlock;
+-
+-#ifdef CONFIG_DEBUG_LOCK_ALLOC
+-# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map))
+- struct {
+- u8 __padding[LOCK_PADSIZE];
+- struct lockdep_map dep_map;
+- };
-#endif
- }
-
- /**
-Index: linux-3.0/kernel/trace/Kconfig
-===================================================================
---- linux-3.0.orig/kernel/trace/Kconfig
-+++ linux-3.0/kernel/trace/Kconfig
-@@ -192,6 +192,24 @@ config IRQSOFF_TRACER
- enabled. This option and the preempt-off timing option can be
- used together or separately.)
-
-+config INTERRUPT_OFF_HIST
-+ bool "Interrupts-off Latency Histogram"
-+ depends on IRQSOFF_TRACER
-+ help
-+ This option generates continuously updated histograms (one per cpu)
-+ of the duration of time periods with interrupts disabled. The
-+ histograms are disabled by default. To enable them, write a non-zero
-+ number to
+- };
+-} spinlock_t;
+-
+-#define __SPIN_LOCK_INITIALIZER(lockname) \
+- { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } }
+-
+-#define __SPIN_LOCK_UNLOCKED(lockname) \
+- (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname)
+-
+-#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x)
+-
+-#include <linux/rwlock_types.h>
+-
+ #endif /* __LINUX_SPINLOCK_TYPES_H */
+diff --git a/include/linux/spinlock_types_nort.h b/include/linux/spinlock_types_nort.h
+new file mode 100644
+index 0000000..f1dac1f
+--- /dev/null
++++ b/include/linux/spinlock_types_nort.h
+@@ -0,0 +1,33 @@
++#ifndef __LINUX_SPINLOCK_TYPES_NORT_H
++#define __LINUX_SPINLOCK_TYPES_NORT_H
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff
++#ifndef __LINUX_SPINLOCK_TYPES_H
++#error "Do not include directly. Include spinlock_types.h instead"
++#endif
+
-+ If PREEMPT_OFF_HIST is also selected, additional histograms (one
-+ per cpu) are generated that accumulate the duration of time periods
-+ when both interrupts and preemption are disabled. The histogram data
-+ will be located in the debug file system at
++/*
++ * The non RT version maps spinlocks to raw_spinlocks
++ */
++typedef struct spinlock {
++ union {
++ struct raw_spinlock rlock;
+
-+ /sys/kernel/debug/tracing/latency_hist/irqsoff
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map))
++ struct {
++ u8 __padding[LOCK_PADSIZE];
++ struct lockdep_map dep_map;
++ };
++#endif
++ };
++} spinlock_t;
+
- config PREEMPT_TRACER
- bool "Preemption-off Latency Tracer"
- default n
-@@ -214,6 +232,24 @@ config PREEMPT_TRACER
- enabled. This option and the irqs-off timing option can be
- used together or separately.)
-
-+config PREEMPT_OFF_HIST
-+ bool "Preemption-off Latency Histogram"
-+ depends on PREEMPT_TRACER
-+ help
-+ This option generates continuously updated histograms (one per cpu)
-+ of the duration of time periods with preemption disabled. The
-+ histograms are disabled by default. To enable them, write a non-zero
-+ number to
++#define __SPIN_LOCK_INITIALIZER(lockname) \
++ { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } }
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff
++#define __SPIN_LOCK_UNLOCKED(lockname) \
++ (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname)
+
-+ If INTERRUPT_OFF_HIST is also selected, additional histograms (one
-+ per cpu) are generated that accumulate the duration of time periods
-+ when both interrupts and preemption are disabled. The histogram data
-+ will be located in the debug file system at
++#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x)
+
-+ /sys/kernel/debug/tracing/latency_hist/preemptoff
++#endif
+diff --git a/include/linux/spinlock_types_raw.h b/include/linux/spinlock_types_raw.h
+new file mode 100644
+index 0000000..edffc4d
+--- /dev/null
++++ b/include/linux/spinlock_types_raw.h
+@@ -0,0 +1,56 @@
++#ifndef __LINUX_SPINLOCK_TYPES_RAW_H
++#define __LINUX_SPINLOCK_TYPES_RAW_H
+
- config SCHED_TRACER
- bool "Scheduling Latency Tracer"
- select GENERIC_TRACER
-@@ -223,6 +259,74 @@ config SCHED_TRACER
- This tracer tracks the latency of the highest priority task
- to be scheduled in, starting from the point it has woken up.
-
-+config WAKEUP_LATENCY_HIST
-+ bool "Scheduling Latency Histogram"
-+ depends on SCHED_TRACER
-+ help
-+ This option generates continuously updated histograms (one per cpu)
-+ of the scheduling latency of the highest priority task.
-+ The histograms are disabled by default. To enable them, write a
-+ non-zero number to
++#if defined(CONFIG_SMP)
++# include <asm/spinlock_types.h>
++#else
++# include <linux/spinlock_types_up.h>
++#endif
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/wakeup
++#include <linux/lockdep.h>
+
-+ Two different algorithms are used, one to determine the latency of
-+ processes that exclusively use the highest priority of the system and
-+ another one to determine the latency of processes that share the
-+ highest system priority with other processes. The former is used to
-+ improve hardware and system software, the latter to optimize the
-+ priority design of a given system. The histogram data will be
-+ located in the debug file system at
++typedef struct raw_spinlock {
++ arch_spinlock_t raw_lock;
++#ifdef CONFIG_GENERIC_LOCKBREAK
++ unsigned int break_lock;
++#endif
++#ifdef CONFIG_DEBUG_SPINLOCK
++ unsigned int magic, owner_cpu;
++ void *owner;
++#endif
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++} raw_spinlock_t;
+
-+ /sys/kernel/debug/tracing/latency_hist/wakeup
++#define SPINLOCK_MAGIC 0xdead4ead
+
-+ and
++#define SPINLOCK_OWNER_INIT ((void *)-1L)
+
-+ /sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
++#else
++# define SPIN_DEP_MAP_INIT(lockname)
++#endif
+
-+ If both Scheduling Latency Histogram and Missed Timer Offsets
-+ Histogram are selected, additional histogram data will be collected
-+ that contain, in addition to the wakeup latency, the timer latency, in
-+ case the wakeup was triggered by an expired timer. These histograms
-+ are available in the
++#ifdef CONFIG_DEBUG_SPINLOCK
++# define SPIN_DEBUG_INIT(lockname) \
++ .magic = SPINLOCK_MAGIC, \
++ .owner_cpu = -1, \
++ .owner = SPINLOCK_OWNER_INIT,
++#else
++# define SPIN_DEBUG_INIT(lockname)
++#endif
+
-+ /sys/kernel/debug/tracing/latency_hist/timerandwakeup
++#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \
++ { \
++ .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \
++ SPIN_DEBUG_INIT(lockname) \
++ SPIN_DEP_MAP_INIT(lockname) }
+
-+ directory. They reflect the apparent interrupt and scheduling latency
-+ and are best suitable to determine the worst-case latency of a given
-+ system. To enable these histograms, write a non-zero number to
++#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \
++ (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname)
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup
++#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x)
+
-+config MISSED_TIMER_OFFSETS_HIST
-+ depends on HIGH_RES_TIMERS
-+ select GENERIC_TRACER
-+ bool "Missed Timer Offsets Histogram"
-+ help
-+ Generate a histogram of missed timer offsets in microseconds. The
-+ histograms are disabled by default. To enable them, write a non-zero
-+ number to
++#endif
+diff --git a/include/linux/spinlock_types_rt.h b/include/linux/spinlock_types_rt.h
+new file mode 100644
+index 0000000..1fe8fc0
+--- /dev/null
++++ b/include/linux/spinlock_types_rt.h
+@@ -0,0 +1,49 @@
++#ifndef __LINUX_SPINLOCK_TYPES_RT_H
++#define __LINUX_SPINLOCK_TYPES_RT_H
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/missed_timer_offsets
++#ifndef __LINUX_SPINLOCK_TYPES_H
++#error "Do not include directly. Include spinlock_types.h instead"
++#endif
+
-+ The histogram data will be located in the debug file system at
++/*
++ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field:
++ */
++typedef struct spinlock {
++ struct rt_mutex lock;
++ unsigned int break_lock;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ struct lockdep_map dep_map;
++#endif
++} spinlock_t;
+
-+ /sys/kernel/debug/tracing/latency_hist/missed_timer_offsets
++#ifdef CONFIG_DEBUG_RT_MUTEXES
++# define __RT_SPIN_INITIALIZER(name) \
++ { \
++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
++ .save_state = 1, \
++ .file = __FILE__, \
++ .line = __LINE__ , \
++ }
++#else
++# define __RT_SPIN_INITIALIZER(name) \
++ { \
++ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
++ .save_state = 1, \
++ }
++#endif
+
-+ If both Scheduling Latency Histogram and Missed Timer Offsets
-+ Histogram are selected, additional histogram data will be collected
-+ that contain, in addition to the wakeup latency, the timer latency, in
-+ case the wakeup was triggered by an expired timer. These histograms
-+ are available in the
++/*
++.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock)
++*/
+
-+ /sys/kernel/debug/tracing/latency_hist/timerandwakeup
++#define __SPIN_LOCK_UNLOCKED(name) \
++ { .lock = __RT_SPIN_INITIALIZER(name.lock), \
++ SPIN_DEP_MAP_INIT(name) }
+
-+ directory. They reflect the apparent interrupt and scheduling latency
-+ and are best suitable to determine the worst-case latency of a given
-+ system. To enable these histograms, write a non-zero number to
++#define __DEFINE_SPINLOCK(name) \
++ spinlock_t name = __SPIN_LOCK_UNLOCKED(name)
++
++#define DEFINE_SPINLOCK(name) \
++ spinlock_t name __cacheline_aligned_in_smp = __SPIN_LOCK_UNLOCKED(name)
++
++#endif
+diff --git a/include/linux/sysrq.h b/include/linux/sysrq.h
+index 7faf933..d224c0b 100644
+--- a/include/linux/sysrq.h
++++ b/include/linux/sysrq.h
+@@ -38,6 +38,11 @@ struct sysrq_key_op {
+ int enable_mask;
+ };
+
++#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
++extern int sysrq_in_progress;
++#else
++#define sysrq_in_progress 0
++#endif
+ #ifdef CONFIG_MAGIC_SYSRQ
+
+ /* Generic SysRq interface -- you may call it from any device driver, supplying
+diff --git a/include/linux/timer.h b/include/linux/timer.h
+index 6abd913..b703477 100644
+--- a/include/linux/timer.h
++++ b/include/linux/timer.h
+@@ -276,7 +276,7 @@ extern void add_timer(struct timer_list *timer);
+
+ extern int try_to_del_timer_sync(struct timer_list *timer);
+
+-#ifdef CONFIG_SMP
++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL)
+ extern int del_timer_sync(struct timer_list *timer);
+ #else
+ # define del_timer_sync(t) del_timer(t)
+diff --git a/include/linux/uaccess.h b/include/linux/uaccess.h
+index 5ca0951..44b3751 100644
+--- a/include/linux/uaccess.h
++++ b/include/linux/uaccess.h
+@@ -6,38 +6,37 @@
+
+ /*
+ * These routines enable/disable the pagefault handler in that
+- * it will not take any locks and go straight to the fixup table.
+- *
+- * They have great resemblance to the preempt_disable/enable calls
+- * and in fact they are identical; this is because currently there is
+- * no other way to make the pagefault handlers do this. So we do
+- * disable preemption but we don't necessarily care about that.
++ * it will not take any MM locks and go straight to the fixup table.
+ */
+-static inline void pagefault_disable(void)
++static inline void raw_pagefault_disable(void)
+ {
+ inc_preempt_count();
+- /*
+- * make sure to have issued the store before a pagefault
+- * can hit.
+- */
+ barrier();
+ }
+
+-static inline void pagefault_enable(void)
++static inline void raw_pagefault_enable(void)
+ {
+- /*
+- * make sure to issue those last loads/stores before enabling
+- * the pagefault handler again.
+- */
+ barrier();
+ dec_preempt_count();
+- /*
+- * make sure we do..
+- */
+ barrier();
+ preempt_check_resched();
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
++static inline void pagefault_disable(void)
++{
++ raw_pagefault_disable();
++}
+
-+ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup
++static inline void pagefault_enable(void)
++{
++ raw_pagefault_enable();
++}
++#else
++extern void pagefault_disable(void);
++extern void pagefault_enable(void);
++#endif
+
- config ENABLE_DEFAULT_TRACERS
- bool "Trace process context switches and events"
- depends on !GENERIC_TRACER
-Index: linux-3.0/kernel/trace/Makefile
-===================================================================
---- linux-3.0.orig/kernel/trace/Makefile
-+++ linux-3.0/kernel/trace/Makefile
-@@ -34,6 +34,10 @@ obj-$(CONFIG_FUNCTION_TRACER) += trace_f
- obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o
- obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o
- obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o
-+obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o
-+obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o
-+obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o
-+obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o
- obj-$(CONFIG_NOP_TRACER) += trace_nop.o
- obj-$(CONFIG_STACK_TRACER) += trace_stack.o
- obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o
-Index: linux-3.0/kernel/trace/latency_hist.c
-===================================================================
+ #ifndef ARCH_HAS_NOCACHE_UACCESS
+
+ static inline unsigned long __copy_from_user_inatomic_nocache(void *to,
+@@ -77,9 +76,9 @@ static inline unsigned long __copy_from_user_nocache(void *to,
+ mm_segment_t old_fs = get_fs(); \
+ \
+ set_fs(KERNEL_DS); \
+- pagefault_disable(); \
++ raw_pagefault_disable(); \
+ ret = __copy_from_user_inatomic(&(retval), (__force typeof(retval) __user *)(addr), sizeof(retval)); \
+- pagefault_enable(); \
++ raw_pagefault_enable(); \
+ set_fs(old_fs); \
+ ret; \
+ })
+diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h
+index bcd942f..12c7d8b 100644
+--- a/include/linux/vmstat.h
++++ b/include/linux/vmstat.h
+@@ -29,7 +29,9 @@ DECLARE_PER_CPU(struct vm_event_state, vm_event_states);
+
+ static inline void __count_vm_event(enum vm_event_item item)
+ {
++ preempt_disable_rt();
+ __this_cpu_inc(vm_event_states.event[item]);
++ preempt_enable_rt();
+ }
+
+ static inline void count_vm_event(enum vm_event_item item)
+@@ -39,7 +41,9 @@ static inline void count_vm_event(enum vm_event_item item)
+
+ static inline void __count_vm_events(enum vm_event_item item, long delta)
+ {
++ preempt_disable_rt();
+ __this_cpu_add(vm_event_states.event[item], delta);
++ preempt_enable_rt();
+ }
+
+ static inline void count_vm_events(enum vm_event_item item, long delta)
+diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h
+index f584aba..eda8db8 100644
+--- a/include/linux/workqueue.h
++++ b/include/linux/workqueue.h
+@@ -254,9 +254,10 @@ enum {
+ WQ_MEM_RECLAIM = 1 << 3, /* may be used for memory reclaim */
+ WQ_HIGHPRI = 1 << 4, /* high priority */
+ WQ_CPU_INTENSIVE = 1 << 5, /* cpu instensive workqueue */
++ WQ_NON_AFFINE = 1 << 6, /* free to move works around cpus */
+
+- WQ_DYING = 1 << 6, /* internal: workqueue is dying */
+- WQ_RESCUER = 1 << 7, /* internal: workqueue has rescuer */
++ WQ_DYING = 1 << 7, /* internal: workqueue is dying */
++ WQ_RESCUER = 1 << 8, /* internal: workqueue has rescuer */
+
+ WQ_MAX_ACTIVE = 512, /* I like 512, better ideas? */
+ WQ_MAX_UNBOUND_PER_CPU = 4, /* 4 * #cpus for unbound wq */
+diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
+index d786b4f..8cef1d1 100644
+--- a/include/net/netns/ipv4.h
++++ b/include/net/netns/ipv4.h
+@@ -47,6 +47,7 @@ struct netns_ipv4 {
+
+ int sysctl_icmp_echo_ignore_all;
+ int sysctl_icmp_echo_ignore_broadcasts;
++ int sysctl_icmp_echo_sysrq;
+ int sysctl_icmp_ignore_bogus_error_responses;
+ int sysctl_icmp_ratelimit;
+ int sysctl_icmp_ratemask;
+diff --git a/include/trace/events/hist.h b/include/trace/events/hist.h
+new file mode 100644
+index 0000000..28646db
--- /dev/null
-+++ linux-3.0/kernel/trace/latency_hist.c
-@@ -0,0 +1,1170 @@
-+/*
-+ * kernel/trace/latency_hist.c
-+ *
-+ * Add support for histograms of preemption-off latency and
-+ * interrupt-off latency and wakeup latency, it depends on
-+ * Real-Time Preemption Support.
-+ *
-+ * Copyright (C) 2005 MontaVista Software, Inc.
-+ * Yi Yang <yyang at ch.mvista.com>
-+ *
-+ * Converted to work with the new latency tracer.
-+ * Copyright (C) 2008 Red Hat, Inc.
-+ * Steven Rostedt <srostedt at redhat.com>
-+ *
-+ */
-+#include <linux/module.h>
-+#include <linux/debugfs.h>
-+#include <linux/seq_file.h>
-+#include <linux/percpu.h>
-+#include <linux/kallsyms.h>
-+#include <linux/uaccess.h>
-+#include <linux/sched.h>
-+#include <linux/slab.h>
-+#include <asm/atomic.h>
-+#include <asm/div64.h>
++++ b/include/trace/events/hist.h
+@@ -0,0 +1,69 @@
++#undef TRACE_SYSTEM
++#define TRACE_SYSTEM hist
+
-+#include "trace.h"
-+#include <trace/events/sched.h>
++#if !defined(_TRACE_HIST_H) || defined(TRACE_HEADER_MULTI_READ)
++#define _TRACE_HIST_H
+
-+#define CREATE_TRACE_POINTS
-+#include <trace/events/hist.h>
++#include "latency_hist.h"
++#include <linux/tracepoint.h>
+
-+enum {
-+ IRQSOFF_LATENCY = 0,
-+ PREEMPTOFF_LATENCY,
-+ PREEMPTIRQSOFF_LATENCY,
-+ WAKEUP_LATENCY,
-+ WAKEUP_LATENCY_SHAREDPRIO,
-+ MISSED_TIMER_OFFSETS,
-+ TIMERANDWAKEUP_LATENCY,
-+ MAX_LATENCY_TYPE,
-+};
++#if !defined(CONFIG_PREEMPT_OFF_HIST) && !defined(CONFIG_INTERRUPT_OFF_HIST)
++#define trace_preemptirqsoff_hist(a,b)
++#else
++TRACE_EVENT(preemptirqsoff_hist,
+
-+#define MAX_ENTRY_NUM 10240
++ TP_PROTO(int reason, int starthist),
+
-+struct hist_data {
-+ atomic_t hist_mode; /* 0 log, 1 don't log */
-+ long offset; /* set it to MAX_ENTRY_NUM/2 for a bipolar scale */
-+ unsigned long min_lat;
-+ unsigned long max_lat;
-+ unsigned long long below_hist_bound_samples;
-+ unsigned long long above_hist_bound_samples;
-+ unsigned long long accumulate_lat;
-+ unsigned long long total_samples;
-+ unsigned long long hist_array[MAX_ENTRY_NUM];
-+};
++ TP_ARGS(reason, starthist),
+
-+struct enable_data {
-+ int latency_type;
-+ int enabled;
-+};
++ TP_STRUCT__entry(
++ __field(int, reason )
++ __field(int, starthist )
++ ),
+
-+static char *latency_hist_dir_root = "latency_hist";
++ TP_fast_assign(
++ __entry->reason = reason;
++ __entry->starthist = starthist;
++ ),
+
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+static DEFINE_PER_CPU(struct hist_data, irqsoff_hist);
-+static char *irqsoff_hist_dir = "irqsoff";
-+static DEFINE_PER_CPU(cycles_t, hist_irqsoff_start);
-+static DEFINE_PER_CPU(int, hist_irqsoff_counting);
++ TP_printk("reason=%s starthist=%s", getaction(__entry->reason),
++ __entry->starthist ? "start" : "stop")
++);
+#endif
+
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+static DEFINE_PER_CPU(struct hist_data, preemptoff_hist);
-+static char *preemptoff_hist_dir = "preemptoff";
-+static DEFINE_PER_CPU(cycles_t, hist_preemptoff_start);
-+static DEFINE_PER_CPU(int, hist_preemptoff_counting);
-+#endif
++#ifndef CONFIG_MISSED_TIMER_OFFSETS_HIST
++#define trace_hrtimer_interrupt(a,b,c,d)
++#else
++TRACE_EVENT(hrtimer_interrupt,
+
-+#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST)
-+static DEFINE_PER_CPU(struct hist_data, preemptirqsoff_hist);
-+static char *preemptirqsoff_hist_dir = "preemptirqsoff";
-+static DEFINE_PER_CPU(cycles_t, hist_preemptirqsoff_start);
-+static DEFINE_PER_CPU(int, hist_preemptirqsoff_counting);
-+#endif
++ TP_PROTO(int cpu, long long offset, struct task_struct *curr, struct task_struct *task),
+
-+#if defined(CONFIG_PREEMPT_OFF_HIST) || defined(CONFIG_INTERRUPT_OFF_HIST)
-+static notrace void probe_preemptirqsoff_hist(void *v, int reason, int start);
-+static struct enable_data preemptirqsoff_enabled_data = {
-+ .latency_type = PREEMPTIRQSOFF_LATENCY,
-+ .enabled = 0,
-+};
-+#endif
++ TP_ARGS(cpu, offset, curr, task),
+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+struct maxlatproc_data {
-+ char comm[FIELD_SIZEOF(struct task_struct, comm)];
-+ char current_comm[FIELD_SIZEOF(struct task_struct, comm)];
-+ int pid;
-+ int current_pid;
-+ int prio;
-+ int current_prio;
-+ long latency;
-+ long timeroffset;
-+ cycle_t timestamp;
-+};
-+#endif
++ TP_STRUCT__entry(
++ __field(int, cpu )
++ __field(long long, offset )
++ __array(char, ccomm, TASK_COMM_LEN)
++ __field(int, cprio )
++ __array(char, tcomm, TASK_COMM_LEN)
++ __field(int, tprio )
++ ),
+
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist);
-+static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist_sharedprio);
-+static char *wakeup_latency_hist_dir = "wakeup";
-+static char *wakeup_latency_hist_dir_sharedprio = "sharedprio";
-+static notrace void probe_wakeup_latency_hist_start(void *v,
-+ struct task_struct *p, int success);
-+static notrace void probe_wakeup_latency_hist_stop(void *v,
-+ struct task_struct *prev, struct task_struct *next);
-+static notrace void probe_sched_migrate_task(void *,
-+ struct task_struct *task, int cpu);
-+static struct enable_data wakeup_latency_enabled_data = {
-+ .latency_type = WAKEUP_LATENCY,
-+ .enabled = 0,
-+};
-+static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc);
-+static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc_sharedprio);
-+static DEFINE_PER_CPU(struct task_struct *, wakeup_task);
-+static DEFINE_PER_CPU(int, wakeup_sharedprio);
-+static unsigned long wakeup_pid;
++ TP_fast_assign(
++ __entry->cpu = cpu;
++ __entry->offset = offset;
++ memcpy(__entry->ccomm, curr->comm, TASK_COMM_LEN);
++ __entry->cprio = curr->prio;
++ memcpy(__entry->tcomm, task != NULL ? task->comm : "<none>", task != NULL ? TASK_COMM_LEN : 7);
++ __entry->tprio = task != NULL ? task->prio : -1;
++ ),
++
++ TP_printk("cpu=%d offset=%lld curr=%s[%d] thread=%s[%d]",
++ __entry->cpu, __entry->offset, __entry->ccomm, __entry->cprio, __entry->tcomm, __entry->tprio)
++);
+#endif
+
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+static DEFINE_PER_CPU(struct hist_data, missed_timer_offsets);
-+static char *missed_timer_offsets_dir = "missed_timer_offsets";
-+static notrace void probe_hrtimer_interrupt(void *v, int cpu,
-+ long long offset, struct task_struct *curr, struct task_struct *task);
-+static struct enable_data missed_timer_offsets_enabled_data = {
-+ .latency_type = MISSED_TIMER_OFFSETS,
-+ .enabled = 0,
++#endif /* _TRACE_HIST_H */
++
++/* This part must be outside protection */
++#include <trace/define_trace.h>
+diff --git a/include/trace/events/latency_hist.h b/include/trace/events/latency_hist.h
+new file mode 100644
+index 0000000..d6b5d77
+--- /dev/null
++++ b/include/trace/events/latency_hist.h
+@@ -0,0 +1,30 @@
++#ifndef _LATENCY_HIST_H
++#define _LATENCY_HIST_H
++
++enum hist_action {
++ IRQS_ON,
++ PREEMPT_ON,
++ TRACE_STOP,
++ IRQS_OFF,
++ PREEMPT_OFF,
++ TRACE_START,
+};
-+static DEFINE_PER_CPU(struct maxlatproc_data, missed_timer_offsets_maxlatproc);
-+static unsigned long missed_timer_offsets_pid;
-+#endif
+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+static DEFINE_PER_CPU(struct hist_data, timerandwakeup_latency_hist);
-+static char *timerandwakeup_latency_hist_dir = "timerandwakeup";
-+static struct enable_data timerandwakeup_enabled_data = {
-+ .latency_type = TIMERANDWAKEUP_LATENCY,
-+ .enabled = 0,
++static char *actions[] = {
++ "IRQS_ON",
++ "PREEMPT_ON",
++ "TRACE_STOP",
++ "IRQS_OFF",
++ "PREEMPT_OFF",
++ "TRACE_START",
+};
-+static DEFINE_PER_CPU(struct maxlatproc_data, timerandwakeup_maxlatproc);
-+#endif
+
-+void notrace latency_hist(int latency_type, int cpu, unsigned long latency,
-+ unsigned long timeroffset, cycle_t stop,
-+ struct task_struct *p)
++static inline char *getaction(int action)
+{
-+ struct hist_data *my_hist;
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ struct maxlatproc_data *mp = NULL;
++ if (action >= 0 && action <= sizeof(actions)/sizeof(actions[0]))
++ return(actions[action]);
++ return("unknown");
++}
++
++#endif /* _LATENCY_HIST_H */
++
+diff --git a/init/Kconfig b/init/Kconfig
+index 412c21b..89e40a4 100644
+--- a/init/Kconfig
++++ b/init/Kconfig
+@@ -401,7 +401,7 @@ config TREE_PREEMPT_RCU
+
+ config TINY_RCU
+ bool "UP-only small-memory-footprint RCU"
+- depends on !SMP
++ depends on !SMP && !PREEMPT_RT_FULL
+ help
+ This option selects the RCU implementation that is
+ designed for UP systems from which real-time response
+@@ -410,7 +410,7 @@ config TINY_RCU
+
+ config TINY_PREEMPT_RCU
+ bool "Preemptible UP-only small-memory-footprint RCU"
+- depends on !SMP && PREEMPT
++ depends on !SMP && PREEMPT && !PREEMPT_RT_FULL
+ help
+ This option selects the RCU implementation that is designed
+ for real-time UP systems. This option greatly reduces the
+@@ -1220,6 +1220,7 @@ config SLAB
+
+ config SLUB
+ bool "SLUB (Unqueued Allocator)"
++ depends on !PREEMPT_RT_FULL
+ help
+ SLUB is a slab allocator that minimizes cache line usage
+ instead of managing queues of cached objects (SLAB approach).
+@@ -1231,6 +1232,7 @@ config SLUB
+ config SLOB
+ depends on EXPERT
+ bool "SLOB (Simple Allocator)"
++ depends on !PREEMPT_RT_FULL
+ help
+ SLOB replaces the stock allocator with a drastically simpler
+ allocator. SLOB is generally more space efficient but
+diff --git a/init/Makefile b/init/Makefile
+index 0bf677a..6b473cd 100644
+--- a/init/Makefile
++++ b/init/Makefile
+@@ -29,4 +29,4 @@ silent_chk_compile.h = :
+ include/generated/compile.h: FORCE
+ @$($(quiet)chk_compile.h)
+ $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \
+- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)"
++ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)"
+diff --git a/init/main.c b/init/main.c
+index d7211fa..34ff139 100644
+--- a/init/main.c
++++ b/init/main.c
+@@ -68,6 +68,7 @@
+ #include <linux/shmem_fs.h>
+ #include <linux/slab.h>
+ #include <linux/perf_event.h>
++#include <linux/posix-timers.h>
+
+ #include <asm/io.h>
+ #include <asm/bugs.h>
+@@ -367,9 +368,7 @@ static noinline void __init_refok rest_init(void)
+ * at least once to get things moving:
+ */
+ init_idle_bootup_task(current);
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+
+ /* Call into cpu_idle with preempt disabled */
+ cpu_idle();
+@@ -501,6 +500,7 @@ asmlinkage void __init start_kernel(void)
+ parse_args("Booting kernel", static_command_line, __start___param,
+ __stop___param - __start___param,
+ &unknown_bootoption);
++ softirq_early_init();
+ /*
+ * These use large bootmem allocations and must precede
+ * kmem_cache_init()
+diff --git a/ipc/mqueue.c b/ipc/mqueue.c
+index ed049ea..0360149 100644
+--- a/ipc/mqueue.c
++++ b/ipc/mqueue.c
+@@ -820,12 +820,17 @@ static inline void pipelined_send(struct mqueue_inode_info *info,
+ struct msg_msg *message,
+ struct ext_wait_queue *receiver)
+ {
++ /*
++ * Keep them in one critical section for PREEMPT_RT:
++ */
++ preempt_disable_rt();
+ receiver->msg = message;
+ list_del(&receiver->list);
+ receiver->state = STATE_PENDING;
+ wake_up_process(receiver->task);
+ smp_wmb();
+ receiver->state = STATE_READY;
++ preempt_enable_rt();
+ }
+
+ /* pipelined_receive() - if there is task waiting in sys_mq_timedsend()
+@@ -839,15 +844,19 @@ static inline void pipelined_receive(struct mqueue_inode_info *info)
+ wake_up_interruptible(&info->wait_q);
+ return;
+ }
++ /*
++ * Keep them in one critical section for PREEMPT_RT:
++ */
++ preempt_disable_rt();
+ msg_insert(sender->msg, info);
+ list_del(&sender->list);
+ sender->state = STATE_PENDING;
+ wake_up_process(sender->task);
+ smp_wmb();
+ sender->state = STATE_READY;
++ preempt_enable_rt();
+ }
+-
+-SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr,
++ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr,
+ size_t, msg_len, unsigned int, msg_prio,
+ const struct timespec __user *, u_abs_timeout)
+ {
+diff --git a/ipc/msg.c b/ipc/msg.c
+index 7385de2..06642ac 100644
+--- a/ipc/msg.c
++++ b/ipc/msg.c
+@@ -259,12 +259,20 @@ static void expunge_all(struct msg_queue *msq, int res)
+ while (tmp != &msq->q_receivers) {
+ struct msg_receiver *msr;
+
++ /*
++ * Make sure that the wakeup doesnt preempt
++ * this CPU prematurely. (on PREEMPT_RT)
++ */
++ preempt_disable_rt();
++
+ msr = list_entry(tmp, struct msg_receiver, r_list);
+ tmp = tmp->next;
+ msr->r_msg = NULL;
+ wake_up_process(msr->r_tsk);
+ smp_mb();
+ msr->r_msg = ERR_PTR(res);
++
++ preempt_enable_rt();
+ }
+ }
+
+@@ -611,6 +619,12 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
+ !security_msg_queue_msgrcv(msq, msg, msr->r_tsk,
+ msr->r_msgtype, msr->r_mode)) {
+
++ /*
++ * Make sure that the wakeup doesnt preempt
++ * this CPU prematurely. (on PREEMPT_RT)
++ */
++ preempt_disable_rt();
++
+ list_del(&msr->r_list);
+ if (msr->r_maxsize < msg->m_ts) {
+ msr->r_msg = NULL;
+@@ -624,9 +638,11 @@ static inline int pipelined_send(struct msg_queue *msq, struct msg_msg *msg)
+ wake_up_process(msr->r_tsk);
+ smp_mb();
+ msr->r_msg = msg;
++ preempt_enable_rt();
+
+ return 1;
+ }
++ preempt_enable_rt();
+ }
+ }
+ return 0;
+diff --git a/ipc/sem.c b/ipc/sem.c
+index e68a8f5..3da3049 100644
+--- a/ipc/sem.c
++++ b/ipc/sem.c
+@@ -415,6 +415,13 @@ undo:
+ static void wake_up_sem_queue_prepare(struct list_head *pt,
+ struct sem_queue *q, int error)
+ {
++#ifdef CONFIG_PREEMPT_RT_BASE
++ struct task_struct *p = q->sleeper;
++ get_task_struct(p);
++ q->status = error;
++ wake_up_process(p);
++ put_task_struct(p);
++#else
+ if (list_empty(pt)) {
+ /*
+ * Hold preempt off so that we don't get preempted and have the
+@@ -426,6 +433,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt,
+ q->pid = error;
+
+ list_add_tail(&q->simple_list, pt);
++#endif
+ }
+
+ /**
+@@ -439,6 +447,7 @@ static void wake_up_sem_queue_prepare(struct list_head *pt,
+ */
+ static void wake_up_sem_queue_do(struct list_head *pt)
+ {
++#ifndef CONFIG_PREEMPT_RT_BASE
+ struct sem_queue *q, *t;
+ int did_something;
+
+@@ -451,6 +460,7 @@ static void wake_up_sem_queue_do(struct list_head *pt)
+ }
+ if (did_something)
+ preempt_enable();
+#endif
+ }
+
+ static void unlink_queue(struct sem_array *sma, struct sem_queue *q)
+diff --git a/kernel/Kconfig.locks b/kernel/Kconfig.locks
+index 5068e2a..7bd0598 100644
+--- a/kernel/Kconfig.locks
++++ b/kernel/Kconfig.locks
+@@ -199,4 +199,4 @@ config INLINE_WRITE_UNLOCK_IRQRESTORE
+ def_bool !DEBUG_SPINLOCK && ARCH_INLINE_WRITE_UNLOCK_IRQRESTORE
+
+ config MUTEX_SPIN_ON_OWNER
+- def_bool SMP && !DEBUG_MUTEXES
++ def_bool SMP && !DEBUG_MUTEXES && !PREEMPT_RT_FULL
+diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt
+index bf987b9..7a1698a 100644
+--- a/kernel/Kconfig.preempt
++++ b/kernel/Kconfig.preempt
+@@ -1,3 +1,9 @@
++config PREEMPT
++ bool
+
-+ if (cpu < 0 || cpu >= NR_CPUS || latency_type < 0 ||
-+ latency_type >= MAX_LATENCY_TYPE)
-+ return;
++config PREEMPT_RT_BASE
++ bool
++ select PREEMPT
+
+ choice
+ prompt "Preemption Model"
+@@ -33,8 +39,9 @@ config PREEMPT_VOLUNTARY
+
+ Select this if you are building a kernel for a desktop system.
+
+-config PREEMPT
++config PREEMPT_LL
+ bool "Preemptible Kernel (Low-Latency Desktop)"
++ select PREEMPT
+ help
+ This option reduces the latency of the kernel by making
+ all kernel code (that is not executing in a critical section)
+@@ -50,5 +57,20 @@ config PREEMPT
+ embedded system with latency requirements in the milliseconds
+ range.
+
++config PREEMPT_RTB
++ bool "Preemptible Kernel (Basic RT)"
++ select PREEMPT_RT_BASE
++ help
++ This option is basically the same as (Low-Latency Desktop) but
++ enables changes which are preliminary for the full preemptiple
++ RT kernel.
+
-+ switch (latency_type) {
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ case IRQSOFF_LATENCY:
-+ my_hist = &per_cpu(irqsoff_hist, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ case PREEMPTOFF_LATENCY:
-+ my_hist = &per_cpu(preemptoff_hist, cpu);
-+ break;
-+#endif
-+#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST)
-+ case PREEMPTIRQSOFF_LATENCY:
-+ my_hist = &per_cpu(preemptirqsoff_hist, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ case WAKEUP_LATENCY:
-+ my_hist = &per_cpu(wakeup_latency_hist, cpu);
-+ mp = &per_cpu(wakeup_maxlatproc, cpu);
-+ break;
-+ case WAKEUP_LATENCY_SHAREDPRIO:
-+ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu);
-+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ case MISSED_TIMER_OFFSETS:
-+ my_hist = &per_cpu(missed_timer_offsets, cpu);
-+ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu);
-+ break;
-+#endif
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ case TIMERANDWAKEUP_LATENCY:
-+ my_hist = &per_cpu(timerandwakeup_latency_hist, cpu);
-+ mp = &per_cpu(timerandwakeup_maxlatproc, cpu);
-+ break;
-+#endif
++config PREEMPT_RT_FULL
++ bool "Fully Preemptible Kernel (RT)"
++ depends on IRQ_FORCED_THREADING
++ select PREEMPT_RT_BASE
++ help
++ All and everything
+
-+ default:
-+ return;
-+ }
+ endchoice
+
+diff --git a/kernel/Makefile b/kernel/Makefile
+index 2d64cfc..11949f1 100644
+--- a/kernel/Makefile
++++ b/kernel/Makefile
+@@ -7,7 +7,7 @@ obj-y = sched.o fork.o exec_domain.o panic.o printk.o \
+ sysctl.o sysctl_binary.o capability.o ptrace.o timer.o user.o \
+ signal.o sys.o kmod.o workqueue.o pid.o \
+ rcupdate.o extable.o params.o posix-timers.o \
+- kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \
++ kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o \
+ hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \
+ notifier.o ksysfs.o pm_qos_params.o sched_clock.o cred.o \
+ async.o range.o jump_label.o
+@@ -29,7 +29,10 @@ obj-$(CONFIG_PROFILING) += profile.o
+ obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o
+ obj-$(CONFIG_STACKTRACE) += stacktrace.o
+ obj-y += time/
++ifneq ($(CONFIG_PREEMPT_RT_FULL),y)
++obj-y += mutex.o
+ obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o
++endif
+ obj-$(CONFIG_LOCKDEP) += lockdep.o
+ ifeq ($(CONFIG_PROC_FS),y)
+ obj-$(CONFIG_LOCKDEP) += lockdep_proc.o
+@@ -41,6 +44,7 @@ endif
+ obj-$(CONFIG_RT_MUTEXES) += rtmutex.o
+ obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o
+ obj-$(CONFIG_RT_MUTEX_TESTER) += rtmutex-tester.o
++obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o
+ obj-$(CONFIG_GENERIC_ISA_DMA) += dma.o
+ obj-$(CONFIG_SMP) += smp.o
+ ifneq ($(CONFIG_SMP),y)
+diff --git a/kernel/cgroup.c b/kernel/cgroup.c
+index 2731d11..2d7503c 100644
+--- a/kernel/cgroup.c
++++ b/kernel/cgroup.c
+@@ -263,7 +263,7 @@ list_for_each_entry(_root, &roots, root_list)
+ /* the list of cgroups eligible for automatic release. Protected by
+ * release_list_lock */
+ static LIST_HEAD(release_list);
+-static DEFINE_SPINLOCK(release_list_lock);
++static DEFINE_RAW_SPINLOCK(release_list_lock);
+ static void cgroup_release_agent(struct work_struct *work);
+ static DECLARE_WORK(release_agent_work, cgroup_release_agent);
+ static void check_for_release(struct cgroup *cgrp);
+@@ -4010,11 +4010,11 @@ again:
+ finish_wait(&cgroup_rmdir_waitq, &wait);
+ clear_bit(CGRP_WAIT_ON_RMDIR, &cgrp->flags);
+
+- spin_lock(&release_list_lock);
++ raw_spin_lock(&release_list_lock);
+ set_bit(CGRP_REMOVED, &cgrp->flags);
+ if (!list_empty(&cgrp->release_list))
+ list_del_init(&cgrp->release_list);
+- spin_unlock(&release_list_lock);
++ raw_spin_unlock(&release_list_lock);
+
+ cgroup_lock_hierarchy(cgrp->root);
+ /* delete this cgroup from parent->children */
+@@ -4667,13 +4667,13 @@ static void check_for_release(struct cgroup *cgrp)
+ * already queued for a userspace notification, queue
+ * it now */
+ int need_schedule_work = 0;
+- spin_lock(&release_list_lock);
++ raw_spin_lock(&release_list_lock);
+ if (!cgroup_is_removed(cgrp) &&
+ list_empty(&cgrp->release_list)) {
+ list_add(&cgrp->release_list, &release_list);
+ need_schedule_work = 1;
+ }
+- spin_unlock(&release_list_lock);
++ raw_spin_unlock(&release_list_lock);
+ if (need_schedule_work)
+ schedule_work(&release_agent_work);
+ }
+@@ -4725,7 +4725,7 @@ static void cgroup_release_agent(struct work_struct *work)
+ {
+ BUG_ON(work != &release_agent_work);
+ mutex_lock(&cgroup_mutex);
+- spin_lock(&release_list_lock);
++ raw_spin_lock(&release_list_lock);
+ while (!list_empty(&release_list)) {
+ char *argv[3], *envp[3];
+ int i;
+@@ -4734,7 +4734,7 @@ static void cgroup_release_agent(struct work_struct *work)
+ struct cgroup,
+ release_list);
+ list_del_init(&cgrp->release_list);
+- spin_unlock(&release_list_lock);
++ raw_spin_unlock(&release_list_lock);
+ pathbuf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+ if (!pathbuf)
+ goto continue_free;
+@@ -4764,9 +4764,9 @@ static void cgroup_release_agent(struct work_struct *work)
+ continue_free:
+ kfree(pathbuf);
+ kfree(agentbuf);
+- spin_lock(&release_list_lock);
++ raw_spin_lock(&release_list_lock);
+ }
+- spin_unlock(&release_list_lock);
++ raw_spin_unlock(&release_list_lock);
+ mutex_unlock(&cgroup_mutex);
+ }
+
+diff --git a/kernel/cpu.c b/kernel/cpu.c
+index 12b7458..13066a3 100644
+--- a/kernel/cpu.c
++++ b/kernel/cpu.c
+@@ -57,6 +57,104 @@ static struct {
+ .refcount = 0,
+ };
+
++struct hotplug_pcp {
++ struct task_struct *unplug;
++ int refcount;
++ struct completion synced;
++};
+
-+ latency += my_hist->offset;
++static DEFINE_PER_CPU(struct hotplug_pcp, hotplug_pcp);
+
-+ if (atomic_read(&my_hist->hist_mode) == 0)
-+ return;
++/**
++ * pin_current_cpu - Prevent the current cpu from being unplugged
++ *
++ * Lightweight version of get_online_cpus() to prevent cpu from being
++ * unplugged when code runs in a migration disabled region.
++ *
++ * Must be called with preemption disabled (preempt_count = 1)!
++ */
++void pin_current_cpu(void)
++{
++ struct hotplug_pcp *hp;
+
-+ if (latency < 0 || latency >= MAX_ENTRY_NUM) {
-+ if (latency < 0)
-+ my_hist->below_hist_bound_samples++;
-+ else
-+ my_hist->above_hist_bound_samples++;
-+ } else
-+ my_hist->hist_array[latency]++;
++retry:
++ hp = &__get_cpu_var(hotplug_pcp);
+
-+ if (unlikely(latency > my_hist->max_lat ||
-+ my_hist->min_lat == ULONG_MAX)) {
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ if (latency_type == WAKEUP_LATENCY ||
-+ latency_type == WAKEUP_LATENCY_SHAREDPRIO ||
-+ latency_type == MISSED_TIMER_OFFSETS ||
-+ latency_type == TIMERANDWAKEUP_LATENCY) {
-+ strncpy(mp->comm, p->comm, sizeof(mp->comm));
-+ strncpy(mp->current_comm, current->comm,
-+ sizeof(mp->current_comm));
-+ mp->pid = task_pid_nr(p);
-+ mp->current_pid = task_pid_nr(current);
-+ mp->prio = p->prio;
-+ mp->current_prio = current->prio;
-+ mp->latency = latency;
-+ mp->timeroffset = timeroffset;
-+ mp->timestamp = stop;
-+ }
-+#endif
-+ my_hist->max_lat = latency;
-+ }
-+ if (unlikely(latency < my_hist->min_lat))
-+ my_hist->min_lat = latency;
-+ my_hist->total_samples++;
-+ my_hist->accumulate_lat += latency;
++ if (!hp->unplug || hp->refcount || preempt_count() > 1 ||
++ hp->unplug == current || (current->flags & PF_STOMPER)) {
++ hp->refcount++;
++ return;
++ }
++ preempt_enable();
++ mutex_lock(&cpu_hotplug.lock);
++ mutex_unlock(&cpu_hotplug.lock);
++ preempt_disable();
++ goto retry;
+}
+
-+static void *l_start(struct seq_file *m, loff_t *pos)
++/**
++ * unpin_current_cpu - Allow unplug of current cpu
++ *
++ * Must be called with preemption or interrupts disabled!
++ */
++void unpin_current_cpu(void)
+{
-+ loff_t *index_ptr = NULL;
-+ loff_t index = *pos;
-+ struct hist_data *my_hist = m->private;
++ struct hotplug_pcp *hp = &__get_cpu_var(hotplug_pcp);
+
-+ if (index == 0) {
-+ char minstr[32], avgstr[32], maxstr[32];
++ WARN_ON(hp->refcount <= 0);
+
-+ atomic_dec(&my_hist->hist_mode);
++ /* This is safe. sync_unplug_thread is pinned to this cpu */
++ if (!--hp->refcount && hp->unplug && hp->unplug != current &&
++ !(current->flags & PF_STOMPER))
++ wake_up_process(hp->unplug);
++}
+
-+ if (likely(my_hist->total_samples)) {
-+ unsigned long avg = (unsigned long)
-+ div64_u64(my_hist->accumulate_lat,
-+ my_hist->total_samples);
-+ snprintf(minstr, sizeof(minstr), "%ld",
-+ (long) my_hist->min_lat - my_hist->offset);
-+ snprintf(avgstr, sizeof(avgstr), "%ld",
-+ (long) avg - my_hist->offset);
-+ snprintf(maxstr, sizeof(maxstr), "%ld",
-+ (long) my_hist->max_lat - my_hist->offset);
-+ } else {
-+ strcpy(minstr, "<undef>");
-+ strcpy(avgstr, minstr);
-+ strcpy(maxstr, minstr);
-+ }
++/*
++ * FIXME: Is this really correct under all circumstances ?
++ */
++static int sync_unplug_thread(void *data)
++{
++ struct hotplug_pcp *hp = data;
+
-+ seq_printf(m, "#Minimum latency: %s microseconds\n"
-+ "#Average latency: %s microseconds\n"
-+ "#Maximum latency: %s microseconds\n"
-+ "#Total samples: %llu\n"
-+ "#There are %llu samples lower than %ld"
-+ " microseconds.\n"
-+ "#There are %llu samples greater or equal"
-+ " than %ld microseconds.\n"
-+ "#usecs\t%16s\n",
-+ minstr, avgstr, maxstr,
-+ my_hist->total_samples,
-+ my_hist->below_hist_bound_samples,
-+ -my_hist->offset,
-+ my_hist->above_hist_bound_samples,
-+ MAX_ENTRY_NUM - my_hist->offset,
-+ "samples");
-+ }
-+ if (index < MAX_ENTRY_NUM) {
-+ index_ptr = kmalloc(sizeof(loff_t), GFP_KERNEL);
-+ if (index_ptr)
-+ *index_ptr = index;
++ preempt_disable();
++ hp->unplug = current;
++ set_current_state(TASK_UNINTERRUPTIBLE);
++ while (hp->refcount) {
++ schedule_preempt_disabled();
++ set_current_state(TASK_UNINTERRUPTIBLE);
+ }
-+
-+ return index_ptr;
++ set_current_state(TASK_RUNNING);
++ preempt_enable();
++ complete(&hp->synced);
++ return 0;
+}
+
-+static void *l_next(struct seq_file *m, void *p, loff_t *pos)
++/*
++ * Start the sync_unplug_thread on the target cpu and wait for it to
++ * complete.
++ */
++static int cpu_unplug_begin(unsigned int cpu)
+{
-+ loff_t *index_ptr = p;
-+ struct hist_data *my_hist = m->private;
++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu);
++ struct task_struct *tsk;
+
-+ if (++*pos >= MAX_ENTRY_NUM) {
-+ atomic_inc(&my_hist->hist_mode);
-+ return NULL;
-+ }
-+ *index_ptr = *pos;
-+ return index_ptr;
++ init_completion(&hp->synced);
++ tsk = kthread_create(sync_unplug_thread, hp, "sync_unplug/%d", cpu);
++ if (IS_ERR(tsk))
++ return (PTR_ERR(tsk));
++ kthread_bind(tsk, cpu);
++ wake_up_process(tsk);
++ wait_for_completion(&hp->synced);
++ return 0;
+}
+
-+static void l_stop(struct seq_file *m, void *p)
++static void cpu_unplug_done(unsigned int cpu)
+{
-+ kfree(p);
++ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu);
++
++ hp->unplug = NULL;
+}
+
-+static int l_show(struct seq_file *m, void *p)
+ void get_online_cpus(void)
+ {
+ might_sleep();
+@@ -210,13 +308,14 @@ static int __ref take_cpu_down(void *_param)
+ /* Requires cpu_add_remove_lock to be held */
+ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
+ {
+- int err, nr_calls = 0;
++ int mycpu, err, nr_calls = 0;
+ void *hcpu = (void *)(long)cpu;
+ unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0;
+ struct take_cpu_down_param tcd_param = {
+ .mod = mod,
+ .hcpu = hcpu,
+ };
++ cpumask_var_t cpumask;
+
+ if (num_online_cpus() == 1)
+ return -EBUSY;
+@@ -224,7 +323,26 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
+ if (!cpu_online(cpu))
+ return -EINVAL;
+
++ /* Move the downtaker off the unplug cpu */
++ if (!alloc_cpumask_var(&cpumask, GFP_KERNEL))
++ return -ENOMEM;
++ cpumask_andnot(cpumask, cpu_online_mask, cpumask_of(cpu));
++ set_cpus_allowed_ptr(current, cpumask);
++ free_cpumask_var(cpumask);
++ migrate_disable();
++ mycpu = smp_processor_id();
++ if (mycpu == cpu) {
++ printk(KERN_ERR "Yuck! Still on unplug CPU\n!");
++ migrate_enable();
++ return -EBUSY;
++ }
++
+ cpu_hotplug_begin();
++ err = cpu_unplug_begin(cpu);
++ if (err) {
++ printk("cpu_unplug_begin(%d) failed\n", cpu);
++ goto out_cancel;
++ }
+
+ err = __cpu_notify(CPU_DOWN_PREPARE | mod, hcpu, -1, &nr_calls);
+ if (err) {
+@@ -263,6 +381,9 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
+ check_for_tasks(cpu);
+
+ out_release:
++ cpu_unplug_done(cpu);
++out_cancel:
++ migrate_enable();
+ cpu_hotplug_done();
+ if (!err)
+ cpu_notify_nofail(CPU_POST_DEAD | mod, hcpu);
+diff --git a/kernel/cred.c b/kernel/cred.c
+index 174fa84..da48112 100644
+--- a/kernel/cred.c
++++ b/kernel/cred.c
+@@ -35,7 +35,7 @@ static struct kmem_cache *cred_jar;
+ static struct thread_group_cred init_tgcred = {
+ .usage = ATOMIC_INIT(2),
+ .tgid = 0,
+- .lock = __SPIN_LOCK_UNLOCKED(init_cred.tgcred.lock),
++ .lock = __SPIN_LOCK_UNLOCKED(init_tgcred.lock),
+ };
+ #endif
+
+diff --git a/kernel/debug/kdb/kdb_io.c b/kernel/debug/kdb/kdb_io.c
+index 96fdaac..ffd5eb2 100644
+--- a/kernel/debug/kdb/kdb_io.c
++++ b/kernel/debug/kdb/kdb_io.c
+@@ -539,7 +539,6 @@ int vkdb_printf(const char *fmt, va_list ap)
+ int diag;
+ int linecount;
+ int logging, saved_loglevel = 0;
+- int saved_trap_printk;
+ int got_printf_lock = 0;
+ int retlen = 0;
+ int fnd, len;
+@@ -550,8 +549,6 @@ int vkdb_printf(const char *fmt, va_list ap)
+ unsigned long uninitialized_var(flags);
+
+ preempt_disable();
+- saved_trap_printk = kdb_trap_printk;
+- kdb_trap_printk = 0;
+
+ /* Serialize kdb_printf if multiple cpus try to write at once.
+ * But if any cpu goes recursive in kdb, just print the output,
+@@ -807,7 +804,6 @@ kdb_print_out:
+ } else {
+ __release(kdb_printf_lock);
+ }
+- kdb_trap_printk = saved_trap_printk;
+ preempt_enable();
+ return retlen;
+ }
+@@ -817,9 +813,11 @@ int kdb_printf(const char *fmt, ...)
+ va_list ap;
+ int r;
+
++ kdb_trap_printk++;
+ va_start(ap, fmt);
+ r = vkdb_printf(fmt, ap);
+ va_end(ap);
++ kdb_trap_printk--;
+
+ return r;
+ }
+diff --git a/kernel/events/core.c b/kernel/events/core.c
+index 32a6151..08315ad 100644
+--- a/kernel/events/core.c
++++ b/kernel/events/core.c
+@@ -3916,9 +3916,15 @@ static bool perf_output_space(struct perf_buffer *buffer, unsigned long tail,
+
+ static void perf_output_wakeup(struct perf_output_handle *handle)
+ {
++ int queue = handle->nmi;
++
++#ifdef CONFIG_PREEMPT_RT_FULL
++ queue |= irqs_disabled() || preempt_count();
++#endif
++
+ atomic_set(&handle->buffer->poll, POLL_IN);
+
+- if (handle->nmi) {
++ if (queue) {
+ handle->event->pending_wakeup = 1;
+ irq_work_queue(&handle->event->pending);
+ } else
+diff --git a/kernel/exit.c b/kernel/exit.c
+index f2b321b..7e8cbe8 100644
+--- a/kernel/exit.c
++++ b/kernel/exit.c
+@@ -142,7 +142,7 @@ static void __exit_signal(struct task_struct *tsk)
+ * Do this under ->siglock, we can race with another thread
+ * doing sigqueue_free() if we have SIGQUEUE_PREALLOC signals.
+ */
+- flush_sigqueue(&tsk->pending);
++ flush_task_sigqueue(tsk);
+ tsk->sighand = NULL;
+ spin_unlock(&sighand->siglock);
+
+diff --git a/kernel/fork.c b/kernel/fork.c
+index 0276c30..8487ae7 100644
+--- a/kernel/fork.c
++++ b/kernel/fork.c
+@@ -87,7 +87,7 @@ int max_threads; /* tunable limit on nr_threads */
+
+ DEFINE_PER_CPU(unsigned long, process_counts) = 0;
+
+-__cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */
++DEFINE_RWLOCK(tasklist_lock); /* outer */
+
+ #ifdef CONFIG_PROVE_RCU
+ int lockdep_tasklist_lock_is_held(void)
+@@ -198,7 +198,18 @@ void __put_task_struct(struct task_struct *tsk)
+ if (!profile_handoff_task(tsk))
+ free_task(tsk);
+ }
++#ifndef CONFIG_PREEMPT_RT_BASE
+ EXPORT_SYMBOL_GPL(__put_task_struct);
++#else
++void __put_task_struct_cb(struct rcu_head *rhp)
++{
++ struct task_struct *tsk = container_of(rhp, struct task_struct, put_rcu);
++
++ __put_task_struct(tsk);
++
++}
++EXPORT_SYMBOL_GPL(__put_task_struct_cb);
++#endif
+
+ /*
+ * macro override instead of weak attribute alias, to workaround
+@@ -546,6 +557,19 @@ void __mmdrop(struct mm_struct *mm)
+ }
+ EXPORT_SYMBOL_GPL(__mmdrop);
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++/*
++ * RCU callback for delayed mm drop. Not strictly rcu, but we don't
++ * want another facility to make this work.
++ */
++void __mmdrop_delayed(struct rcu_head *rhp)
+{
-+ int index = *(loff_t *) p;
-+ struct hist_data *my_hist = m->private;
++ struct mm_struct *mm = container_of(rhp, struct mm_struct, delayed_drop);
+
-+ seq_printf(m, "%6ld\t%16llu\n", index - my_hist->offset,
-+ my_hist->hist_array[index]);
-+ return 0;
++ __mmdrop(mm);
+}
++#endif
+
-+static struct seq_operations latency_hist_seq_op = {
-+ .start = l_start,
-+ .next = l_next,
-+ .stop = l_stop,
-+ .show = l_show
-+};
+ /*
+ * Decrement the use count and release all resources for an mm.
+ */
+@@ -1030,6 +1054,9 @@ void mm_init_owner(struct mm_struct *mm, struct task_struct *p)
+ */
+ static void posix_cpu_timers_init(struct task_struct *tsk)
+ {
++#ifdef CONFIG_PREEMPT_RT_BASE
++ tsk->posix_timer_list = NULL;
++#endif
+ tsk->cputime_expires.prof_exp = cputime_zero;
+ tsk->cputime_expires.virt_exp = cputime_zero;
+ tsk->cputime_expires.sched_exp = 0;
+@@ -1137,6 +1164,7 @@ static struct task_struct *copy_process(unsigned long clone_flags,
+ spin_lock_init(&p->alloc_lock);
+
+ init_sigpending(&p->pending);
++ p->sigqueue_cache = NULL;
+
+ p->utime = cputime_zero;
+ p->stime = cputime_zero;
+@@ -1194,6 +1222,9 @@ static struct task_struct *copy_process(unsigned long clone_flags,
+ p->hardirq_context = 0;
+ p->softirq_context = 0;
+ #endif
++#ifdef CONFIG_PREEMPT_RT_FULL
++ p->pagefault_disabled = 0;
++#endif
+ #ifdef CONFIG_LOCKDEP
+ p->lockdep_depth = 0; /* no locks held yet */
+ p->curr_chain_key = 0;
+diff --git a/kernel/futex.c b/kernel/futex.c
+index 8b6da25..30b238c 100644
+--- a/kernel/futex.c
++++ b/kernel/futex.c
+@@ -1410,6 +1410,16 @@ retry_private:
+ requeue_pi_wake_futex(this, &key2, hb2);
+ drop_count++;
+ continue;
++ } else if (ret == -EAGAIN) {
++ /*
++ * Waiter was woken by timeout or
++ * signal and has set pi_blocked_on to
++ * PI_WAKEUP_INPROGRESS before we
++ * tried to enqueue it on the rtmutex.
++ */
++ this->pi_state = NULL;
++ free_pi_state(pi_state);
++ continue;
+ } else if (ret) {
+ /* -EDEADLK */
+ this->pi_state = NULL;
+@@ -2254,7 +2264,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
+ struct hrtimer_sleeper timeout, *to = NULL;
+ struct rt_mutex_waiter rt_waiter;
+ struct rt_mutex *pi_mutex = NULL;
+- struct futex_hash_bucket *hb;
++ struct futex_hash_bucket *hb, *hb2;
+ union futex_key key2 = FUTEX_KEY_INIT;
+ struct futex_q q = futex_q_init;
+ int res, ret;
+@@ -2276,8 +2286,7 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
+ * The waiter is allocated on our stack, manipulated by the requeue
+ * code while we sleep on uaddr.
+ */
+- debug_rt_mutex_init_waiter(&rt_waiter);
+- rt_waiter.task = NULL;
++ rt_mutex_init_waiter(&rt_waiter, false);
+
+ ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE);
+ if (unlikely(ret != 0))
+@@ -2298,20 +2307,55 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
+ /* Queue the futex_q, drop the hb lock, wait for wakeup. */
+ futex_wait_queue_me(hb, &q, to);
+
+- spin_lock(&hb->lock);
+- ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to);
+- spin_unlock(&hb->lock);
+- if (ret)
+- goto out_put_keys;
++ /*
++ * On RT we must avoid races with requeue and trying to block
++ * on two mutexes (hb->lock and uaddr2's rtmutex) by
++ * serializing access to pi_blocked_on with pi_lock.
++ */
++ raw_spin_lock_irq(¤t->pi_lock);
++ if (current->pi_blocked_on) {
++ /*
++ * We have been requeued or are in the process of
++ * being requeued.
++ */
++ raw_spin_unlock_irq(¤t->pi_lock);
++ } else {
++ /*
++ * Setting pi_blocked_on to PI_WAKEUP_INPROGRESS
++ * prevents a concurrent requeue from moving us to the
++ * uaddr2 rtmutex. After that we can safely acquire
++ * (and possibly block on) hb->lock.
++ */
++ current->pi_blocked_on = PI_WAKEUP_INPROGRESS;
++ raw_spin_unlock_irq(¤t->pi_lock);
+
-+static int latency_hist_open(struct inode *inode, struct file *file)
-+{
-+ int ret;
++ spin_lock(&hb->lock);
+
-+ ret = seq_open(file, &latency_hist_seq_op);
-+ if (!ret) {
-+ struct seq_file *seq = file->private_data;
-+ seq->private = inode->i_private;
++ /*
++ * Clean up pi_blocked_on. We might leak it otherwise
++ * when we succeeded with the hb->lock in the fast
++ * path.
++ */
++ raw_spin_lock_irq(¤t->pi_lock);
++ current->pi_blocked_on = NULL;
++ raw_spin_unlock_irq(¤t->pi_lock);
++
++ ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to);
++ spin_unlock(&hb->lock);
++ if (ret)
++ goto out_put_keys;
+ }
-+ return ret;
-+}
+
+ /*
+- * In order for us to be here, we know our q.key == key2, and since
+- * we took the hb->lock above, we also know that futex_requeue() has
+- * completed and we no longer have to concern ourselves with a wakeup
+- * race with the atomic proxy lock acquisition by the requeue code. The
+- * futex_requeue dropped our key1 reference and incremented our key2
+- * reference count.
++ * In order to be here, we have either been requeued, are in
++ * the process of being requeued, or requeue successfully
++ * acquired uaddr2 on our behalf. If pi_blocked_on was
++ * non-null above, we may be racing with a requeue. Do not
++ * rely on q->lock_ptr to be hb2->lock until after blocking on
++ * hb->lock or hb2->lock. The futex_requeue dropped our key1
++ * reference and incremented our key2 reference count.
+ */
++ hb2 = hash_futex(&key2);
+
+ /* Check if the requeue code acquired the second futex for us. */
+ if (!q.rt_waiter) {
+@@ -2320,9 +2364,10 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
+ * did a lock-steal - fix up the PI-state in that case.
+ */
+ if (q.pi_state && (q.pi_state->owner != current)) {
+- spin_lock(q.lock_ptr);
++ spin_lock(&hb2->lock);
++ BUG_ON(&hb2->lock != q.lock_ptr);
+ ret = fixup_pi_state_owner(uaddr2, &q, current);
+- spin_unlock(q.lock_ptr);
++ spin_unlock(&hb2->lock);
+ }
+ } else {
+ /*
+@@ -2335,7 +2380,8 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags,
+ ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter, 1);
+ debug_rt_mutex_free_waiter(&rt_waiter);
+
+- spin_lock(q.lock_ptr);
++ spin_lock(&hb2->lock);
++ BUG_ON(&hb2->lock != q.lock_ptr);
+ /*
+ * Fixup the pi_state owner and possibly acquire the lock if we
+ * haven't already.
+diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c
+index a9205e3..af89591 100644
+--- a/kernel/hrtimer.c
++++ b/kernel/hrtimer.c
+@@ -49,6 +49,7 @@
+ #include <asm/uaccess.h>
+
+ #include <trace/events/timer.h>
++#include <trace/events/hist.h>
+
+ /*
+ * The timer bases:
+@@ -588,8 +589,7 @@ static int hrtimer_reprogram(struct hrtimer *timer,
+ * When the callback is running, we do not reprogram the clock event
+ * device. The timer callback is either running on a different CPU or
+ * the callback is executed in the hrtimer_interrupt context. The
+- * reprogramming is handled either by the softirq, which called the
+- * callback or at the end of the hrtimer_interrupt.
++ * reprogramming is handled at the end of the hrtimer_interrupt.
+ */
+ if (hrtimer_callback_running(timer))
+ return 0;
+@@ -624,6 +624,9 @@ static int hrtimer_reprogram(struct hrtimer *timer,
+ return res;
+ }
+
++static void __run_hrtimer(struct hrtimer *timer, ktime_t *now);
++static int hrtimer_rt_defer(struct hrtimer *timer);
+
-+static struct file_operations latency_hist_fops = {
-+ .open = latency_hist_open,
-+ .read = seq_read,
-+ .llseek = seq_lseek,
-+ .release = seq_release,
-+};
+ /*
+ * Initialize the high resolution related parts of cpu_base
+ */
+@@ -644,14 +647,23 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
+ int wakeup)
+ {
+ if (base->cpu_base->hres_active && hrtimer_reprogram(timer, base)) {
+- if (wakeup) {
+- raw_spin_unlock(&base->cpu_base->lock);
+- raise_softirq_irqoff(HRTIMER_SOFTIRQ);
+- raw_spin_lock(&base->cpu_base->lock);
+- } else
+- __raise_softirq_irqoff(HRTIMER_SOFTIRQ);
++ if (!wakeup)
++ return -ETIME;
+
+- return 1;
++#ifdef CONFIG_PREEMPT_RT_BASE
++ /*
++ * Move softirq based timers away from the rbtree in
++ * case it expired already. Otherwise we would have a
++ * stale base->first entry until the softirq runs.
++ */
++ if (!hrtimer_rt_defer(timer))
++ return -ETIME;
++#endif
++ raw_spin_unlock(&base->cpu_base->lock);
++ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
++ raw_spin_lock(&base->cpu_base->lock);
+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+static void clear_maxlatprocdata(struct maxlatproc_data *mp)
++ return 0;
+ }
+
+ return 0;
+@@ -732,6 +744,11 @@ static inline int hrtimer_enqueue_reprogram(struct hrtimer *timer,
+ }
+ static inline void hrtimer_init_hres(struct hrtimer_cpu_base *base) { }
+ static inline void retrigger_next_event(void *arg) { }
++static inline int hrtimer_reprogram(struct hrtimer *timer,
++ struct hrtimer_clock_base *base)
+{
-+ mp->comm[0] = mp->current_comm[0] = '\0';
-+ mp->prio = mp->current_prio = mp->pid = mp->current_pid =
-+ mp->latency = mp->timeroffset = -1;
-+ mp->timestamp = 0;
++ return 0;
+}
-+#endif
+
+ #endif /* CONFIG_HIGH_RES_TIMERS */
+
+@@ -846,6 +863,32 @@ u64 hrtimer_forward(struct hrtimer *timer, ktime_t now, ktime_t interval)
+ }
+ EXPORT_SYMBOL_GPL(hrtimer_forward);
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++# define wake_up_timer_waiters(b) wake_up(&(b)->wait)
+
-+static void hist_reset(struct hist_data *hist)
++/**
++ * hrtimer_wait_for_timer - Wait for a running timer
++ *
++ * @timer: timer to wait for
++ *
++ * The function waits in case the timers callback function is
++ * currently executed on the waitqueue of the timer base. The
++ * waitqueue is woken up after the timer callback function has
++ * finished execution.
++ */
++void hrtimer_wait_for_timer(const struct hrtimer *timer)
+{
-+ atomic_dec(&hist->hist_mode);
-+
-+ memset(hist->hist_array, 0, sizeof(hist->hist_array));
-+ hist->below_hist_bound_samples = 0ULL;
-+ hist->above_hist_bound_samples = 0ULL;
-+ hist->min_lat = ULONG_MAX;
-+ hist->max_lat = 0UL;
-+ hist->total_samples = 0ULL;
-+ hist->accumulate_lat = 0ULL;
++ struct hrtimer_clock_base *base = timer->base;
+
-+ atomic_inc(&hist->hist_mode);
++ if (base && base->cpu_base && !timer->irqsafe)
++ wait_event(base->cpu_base->wait,
++ !(timer->state & HRTIMER_STATE_CALLBACK));
+}
+
-+static ssize_t
-+latency_hist_reset(struct file *file, const char __user *a,
-+ size_t size, loff_t *off)
-+{
-+ int cpu;
-+ struct hist_data *hist = NULL;
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ struct maxlatproc_data *mp = NULL;
++#else
++# define wake_up_timer_waiters(b) do { } while (0)
+#endif
-+ off_t latency_type = (off_t) file->private_data;
+
-+ for_each_online_cpu(cpu) {
+ /*
+ * enqueue_hrtimer - internal function to (re)start a timer
+ *
+@@ -888,6 +931,11 @@ static void __remove_hrtimer(struct hrtimer *timer,
+ if (!(timer->state & HRTIMER_STATE_ENQUEUED))
+ goto out;
+
++ if (unlikely(!list_empty(&timer->cb_entry))) {
++ list_del_init(&timer->cb_entry);
++ goto out;
++ }
+
-+ switch (latency_type) {
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ case PREEMPTOFF_LATENCY:
-+ hist = &per_cpu(preemptoff_hist, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ case IRQSOFF_LATENCY:
-+ hist = &per_cpu(irqsoff_hist, cpu);
-+ break;
-+#endif
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
-+ case PREEMPTIRQSOFF_LATENCY:
-+ hist = &per_cpu(preemptirqsoff_hist, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ case WAKEUP_LATENCY:
-+ hist = &per_cpu(wakeup_latency_hist, cpu);
-+ mp = &per_cpu(wakeup_maxlatproc, cpu);
-+ break;
-+ case WAKEUP_LATENCY_SHAREDPRIO:
-+ hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu);
-+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu);
-+ break;
-+#endif
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ case MISSED_TIMER_OFFSETS:
-+ hist = &per_cpu(missed_timer_offsets, cpu);
-+ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu);
-+ break;
-+#endif
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ case TIMERANDWAKEUP_LATENCY:
-+ hist = &per_cpu(timerandwakeup_latency_hist, cpu);
-+ mp = &per_cpu(timerandwakeup_maxlatproc, cpu);
-+ break;
-+#endif
+ if (&timer->node == timerqueue_getnext(&base->active)) {
+ #ifdef CONFIG_HIGH_RES_TIMERS
+ /* Reprogram the clock event device. if enabled */
+@@ -983,8 +1031,20 @@ int __hrtimer_start_range_ns(struct hrtimer *timer, ktime_t tim,
+ *
+ * XXX send_remote_softirq() ?
+ */
+- if (leftmost && new_base->cpu_base == &__get_cpu_var(hrtimer_bases))
+- hrtimer_enqueue_reprogram(timer, new_base, wakeup);
++ if (leftmost && new_base->cpu_base == &__get_cpu_var(hrtimer_bases)) {
++ ret = hrtimer_enqueue_reprogram(timer, new_base, wakeup);
++ if (ret) {
++ /*
++ * In case we failed to reprogram the timer (mostly
++ * because out current timer is already elapsed),
++ * remove it again and report a failure. This avoids
++ * stale base->first entries.
++ */
++ debug_deactivate(timer);
++ __remove_hrtimer(timer, new_base,
++ timer->state & HRTIMER_STATE_CALLBACK, 0);
+ }
-+
-+ hist_reset(hist);
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ if (latency_type == WAKEUP_LATENCY ||
-+ latency_type == WAKEUP_LATENCY_SHAREDPRIO ||
-+ latency_type == MISSED_TIMER_OFFSETS ||
-+ latency_type == TIMERANDWAKEUP_LATENCY)
-+ clear_maxlatprocdata(mp);
-+#endif
+ }
+
+ unlock_hrtimer_base(timer, &flags);
+
+@@ -1070,7 +1130,7 @@ int hrtimer_cancel(struct hrtimer *timer)
+
+ if (ret >= 0)
+ return ret;
+- cpu_relax();
++ hrtimer_wait_for_timer(timer);
+ }
+ }
+ EXPORT_SYMBOL_GPL(hrtimer_cancel);
+@@ -1149,6 +1209,7 @@ static void __hrtimer_init(struct hrtimer *timer, clockid_t clock_id,
+
+ base = hrtimer_clockid_to_base(clock_id);
+ timer->base = &cpu_base->clock_base[base];
++ INIT_LIST_HEAD(&timer->cb_entry);
+ timerqueue_init(&timer->node);
+
+ #ifdef CONFIG_TIMER_STATS
+@@ -1232,6 +1293,122 @@ static void __run_hrtimer(struct hrtimer *timer, ktime_t *now)
+ timer->state &= ~HRTIMER_STATE_CALLBACK;
+ }
+
++static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer);
+
-+ return size;
-+}
-+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+static ssize_t
-+show_pid(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
++#ifdef CONFIG_PREEMPT_RT_BASE
++static void hrtimer_rt_reprogram(int restart, struct hrtimer *timer,
++ struct hrtimer_clock_base *base)
+{
-+ char buf[64];
-+ int r;
-+ unsigned long *this_pid = file->private_data;
++ /*
++ * Note, we clear the callback flag before we requeue the
++ * timer otherwise we trigger the callback_running() check
++ * in hrtimer_reprogram().
++ */
++ timer->state &= ~HRTIMER_STATE_CALLBACK;
+
-+ r = snprintf(buf, sizeof(buf), "%lu\n", *this_pid);
-+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
-+}
++ if (restart != HRTIMER_NORESTART) {
++ BUG_ON(hrtimer_active(timer));
++ /*
++ * Enqueue the timer, if it's the leftmost timer then
++ * we need to reprogram it.
++ */
++ if (!enqueue_hrtimer(timer, base))
++ return;
+
-+static ssize_t do_pid(struct file *file, const char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
-+{
-+ char buf[64];
-+ unsigned long pid;
-+ unsigned long *this_pid = file->private_data;
++#ifndef CONFIG_HIGH_RES_TIMERS
++ }
++#else
++ if (base->cpu_base->hres_active &&
++ hrtimer_reprogram(timer, base))
++ goto requeue;
+
-+ if (cnt >= sizeof(buf))
-+ return -EINVAL;
++ } else if (hrtimer_active(timer)) {
++ /*
++ * If the timer was rearmed on another CPU, reprogram
++ * the event device.
++ */
++ if (&timer->node == base->active.next &&
++ base->cpu_base->hres_active &&
++ hrtimer_reprogram(timer, base))
++ goto requeue;
++ }
++ return;
+
-+ if (copy_from_user(&buf, ubuf, cnt))
-+ return -EFAULT;
++requeue:
++ /*
++ * Timer is expired. Thus move it from tree to pending list
++ * again.
++ */
++ __remove_hrtimer(timer, base, timer->state, 0);
++ list_add_tail(&timer->cb_entry, &base->expired);
++#endif
++}
+
-+ buf[cnt] = '\0';
++/*
++ * The changes in mainline which removed the callback modes from
++ * hrtimer are not yet working with -rt. The non wakeup_process()
++ * based callbacks which involve sleeping locks need to be treated
++ * seperately.
++ */
++static void hrtimer_rt_run_pending(void)
++{
++ enum hrtimer_restart (*fn)(struct hrtimer *);
++ struct hrtimer_cpu_base *cpu_base;
++ struct hrtimer_clock_base *base;
++ struct hrtimer *timer;
++ int index, restart;
+
-+ if (strict_strtoul(buf, 10, &pid))
-+ return(-EINVAL);
++ local_irq_disable();
++ cpu_base = &per_cpu(hrtimer_bases, smp_processor_id());
+
-+ *this_pid = pid;
++ raw_spin_lock(&cpu_base->lock);
+
-+ return cnt;
-+}
-+#endif
++ for (index = 0; index < HRTIMER_MAX_CLOCK_BASES; index++) {
++ base = &cpu_base->clock_base[index];
+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+static ssize_t
-+show_maxlatproc(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
-+{
-+ int r;
-+ struct maxlatproc_data *mp = file->private_data;
-+ int strmaxlen = (TASK_COMM_LEN * 2) + (8 * 8);
-+ unsigned long long t;
-+ unsigned long usecs, secs;
-+ char *buf;
++ while (!list_empty(&base->expired)) {
++ timer = list_first_entry(&base->expired,
++ struct hrtimer, cb_entry);
++
++ /*
++ * Same as the above __run_hrtimer function
++ * just we run with interrupts enabled.
++ */
++ debug_hrtimer_deactivate(timer);
++ __remove_hrtimer(timer, base, HRTIMER_STATE_CALLBACK, 0);
++ timer_stats_account_hrtimer(timer);
++ fn = timer->function;
+
-+ if (mp->pid == -1 || mp->current_pid == -1) {
-+ buf = "(none)\n";
-+ return simple_read_from_buffer(ubuf, cnt, ppos, buf,
-+ strlen(buf));
++ raw_spin_unlock_irq(&cpu_base->lock);
++ restart = fn(timer);
++ raw_spin_lock_irq(&cpu_base->lock);
++
++ hrtimer_rt_reprogram(restart, timer, base);
++ }
+ }
+
-+ buf = kmalloc(strmaxlen, GFP_KERNEL);
-+ if (buf == NULL)
-+ return -ENOMEM;
++ raw_spin_unlock_irq(&cpu_base->lock);
+
-+ t = ns2usecs(mp->timestamp);
-+ usecs = do_div(t, USEC_PER_SEC);
-+ secs = (unsigned long) t;
-+ r = snprintf(buf, strmaxlen,
-+ "%d %d %ld (%ld) %s <- %d %d %s %lu.%06lu\n", mp->pid,
-+ MAX_RT_PRIO-1 - mp->prio, mp->latency, mp->timeroffset, mp->comm,
-+ mp->current_pid, MAX_RT_PRIO-1 - mp->current_prio, mp->current_comm,
-+ secs, usecs);
-+ r = simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
-+ kfree(buf);
-+ return r;
++ wake_up_timer_waiters(cpu_base);
+}
-+#endif
+
-+static ssize_t
-+show_enable(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
++static int hrtimer_rt_defer(struct hrtimer *timer)
+{
-+ char buf[64];
-+ struct enable_data *ed = file->private_data;
-+ int r;
++ if (timer->irqsafe)
++ return 0;
+
-+ r = snprintf(buf, sizeof(buf), "%d\n", ed->enabled);
-+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
++ __remove_hrtimer(timer, timer->base, timer->state, 0);
++ list_add_tail(&timer->cb_entry, &timer->base->expired);
++ return 1;
+}
+
-+static ssize_t
-+do_enable(struct file *file, const char __user *ubuf, size_t cnt, loff_t *ppos)
-+{
-+ char buf[64];
-+ long enable;
-+ struct enable_data *ed = file->private_data;
-+
-+ if (cnt >= sizeof(buf))
-+ return -EINVAL;
-+
-+ if (copy_from_user(&buf, ubuf, cnt))
-+ return -EFAULT;
-+
-+ buf[cnt] = 0;
++#else
+
-+ if (strict_strtol(buf, 10, &enable))
-+ return(-EINVAL);
++static inline void hrtimer_rt_run_pending(void) { }
++static inline int hrtimer_rt_defer(struct hrtimer *timer) { return 0; }
+
-+ if ((enable && ed->enabled) || (!enable && !ed->enabled))
-+ return cnt;
++#endif
+
-+ if (enable) {
-+ int ret;
+ #ifdef CONFIG_HIGH_RES_TIMERS
+
+ /*
+@@ -1242,7 +1419,7 @@ void hrtimer_interrupt(struct clock_event_device *dev)
+ {
+ struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
+ ktime_t expires_next, now, entry_time, delta;
+- int i, retries = 0;
++ int i, retries = 0, raise = 0;
+
+ BUG_ON(!cpu_base->hres_active);
+ cpu_base->nr_events++;
+@@ -1278,6 +1455,14 @@ retry:
+
+ timer = container_of(node, struct hrtimer, node);
+
++ trace_hrtimer_interrupt(raw_smp_processor_id(),
++ ktime_to_ns(ktime_sub(
++ hrtimer_get_expires(timer), basenow)),
++ current,
++ timer->function == hrtimer_wakeup ?
++ container_of(timer, struct hrtimer_sleeper,
++ timer)->task : NULL);
+
-+ switch (ed->latency_type) {
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
-+ case PREEMPTIRQSOFF_LATENCY:
-+ ret = register_trace_preemptirqsoff_hist(
-+ probe_preemptirqsoff_hist, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_preemptirqsoff_hist "
-+ "to trace_preemptirqsoff_hist\n");
-+ return ret;
-+ }
-+ break;
-+#endif
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ case WAKEUP_LATENCY:
-+ ret = register_trace_sched_wakeup(
-+ probe_wakeup_latency_hist_start, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_wakeup_latency_hist_start "
-+ "to trace_sched_wakeup\n");
-+ return ret;
-+ }
-+ ret = register_trace_sched_wakeup_new(
-+ probe_wakeup_latency_hist_start, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_wakeup_latency_hist_start "
-+ "to trace_sched_wakeup_new\n");
-+ unregister_trace_sched_wakeup(
-+ probe_wakeup_latency_hist_start, NULL);
-+ return ret;
-+ }
-+ ret = register_trace_sched_switch(
-+ probe_wakeup_latency_hist_stop, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_wakeup_latency_hist_stop "
-+ "to trace_sched_switch\n");
-+ unregister_trace_sched_wakeup(
-+ probe_wakeup_latency_hist_start, NULL);
-+ unregister_trace_sched_wakeup_new(
-+ probe_wakeup_latency_hist_start, NULL);
-+ return ret;
-+ }
-+ ret = register_trace_sched_migrate_task(
-+ probe_sched_migrate_task, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_sched_migrate_task "
-+ "to trace_sched_migrate_task\n");
-+ unregister_trace_sched_wakeup(
-+ probe_wakeup_latency_hist_start, NULL);
-+ unregister_trace_sched_wakeup_new(
-+ probe_wakeup_latency_hist_start, NULL);
-+ unregister_trace_sched_switch(
-+ probe_wakeup_latency_hist_stop, NULL);
-+ return ret;
-+ }
-+ break;
-+#endif
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ case MISSED_TIMER_OFFSETS:
-+ ret = register_trace_hrtimer_interrupt(
-+ probe_hrtimer_interrupt, NULL);
-+ if (ret) {
-+ pr_info("wakeup trace: Couldn't assign "
-+ "probe_hrtimer_interrupt "
-+ "to trace_hrtimer_interrupt\n");
-+ return ret;
-+ }
-+ break;
-+#endif
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ case TIMERANDWAKEUP_LATENCY:
-+ if (!wakeup_latency_enabled_data.enabled ||
-+ !missed_timer_offsets_enabled_data.enabled)
-+ return -EINVAL;
-+ break;
-+#endif
-+ default:
-+ break;
-+ }
-+ } else {
-+ switch (ed->latency_type) {
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
-+ case PREEMPTIRQSOFF_LATENCY:
-+ {
-+ int cpu;
+ /*
+ * The immediate goal for using the softexpires is
+ * minimizing wakeups, not running timers at the
+@@ -1301,7 +1486,10 @@ retry:
+ break;
+ }
+
+- __run_hrtimer(timer, &basenow);
++ if (!hrtimer_rt_defer(timer))
++ __run_hrtimer(timer, &basenow);
++ else
++ raise = 1;
+ }
+ }
+
+@@ -1316,6 +1504,10 @@ retry:
+ if (expires_next.tv64 == KTIME_MAX ||
+ !tick_program_event(expires_next, 0)) {
+ cpu_base->hang_detected = 0;
+
-+ unregister_trace_preemptirqsoff_hist(
-+ probe_preemptirqsoff_hist, NULL);
-+ for_each_online_cpu(cpu) {
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ per_cpu(hist_irqsoff_counting,
-+ cpu) = 0;
-+#endif
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ per_cpu(hist_preemptoff_counting,
-+ cpu) = 0;
-+#endif
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
-+ per_cpu(hist_preemptirqsoff_counting,
-+ cpu) = 0;
-+#endif
-+ }
-+ }
-+ break;
-+#endif
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ case WAKEUP_LATENCY:
-+ {
-+ int cpu;
++ if (raise)
++ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
+
-+ unregister_trace_sched_wakeup(
-+ probe_wakeup_latency_hist_start, NULL);
-+ unregister_trace_sched_wakeup_new(
-+ probe_wakeup_latency_hist_start, NULL);
-+ unregister_trace_sched_switch(
-+ probe_wakeup_latency_hist_stop, NULL);
-+ unregister_trace_sched_migrate_task(
-+ probe_sched_migrate_task, NULL);
+ return;
+ }
+
+@@ -1391,17 +1583,17 @@ void hrtimer_peek_ahead_timers(void)
+ local_irq_restore(flags);
+ }
+
+-static void run_hrtimer_softirq(struct softirq_action *h)
+-{
+- hrtimer_peek_ahead_timers();
+-}
+-
+ #else /* CONFIG_HIGH_RES_TIMERS */
+
+ static inline void __hrtimer_peek_ahead_timers(void) { }
+
+ #endif /* !CONFIG_HIGH_RES_TIMERS */
+
++static void run_hrtimer_softirq(struct softirq_action *h)
++{
++ hrtimer_rt_run_pending();
++}
+
-+ for_each_online_cpu(cpu) {
-+ per_cpu(wakeup_task, cpu) = NULL;
-+ per_cpu(wakeup_sharedprio, cpu) = 0;
-+ }
-+ }
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ timerandwakeup_enabled_data.enabled = 0;
+ /*
+ * Called from timer softirq every jiffy, expire hrtimers:
+ *
+@@ -1434,7 +1626,7 @@ void hrtimer_run_queues(void)
+ struct timerqueue_node *node;
+ struct hrtimer_cpu_base *cpu_base = &__get_cpu_var(hrtimer_bases);
+ struct hrtimer_clock_base *base;
+- int index, gettime = 1;
++ int index, gettime = 1, raise = 0;
+
+ if (hrtimer_hres_active())
+ return;
+@@ -1459,10 +1651,16 @@ void hrtimer_run_queues(void)
+ hrtimer_get_expires_tv64(timer))
+ break;
+
+- __run_hrtimer(timer, &base->softirq_time);
++ if (!hrtimer_rt_defer(timer))
++ __run_hrtimer(timer, &base->softirq_time);
++ else
++ raise = 1;
+ }
+ raw_spin_unlock(&cpu_base->lock);
+ }
++
++ if (raise)
++ raise_softirq_irqoff(HRTIMER_SOFTIRQ);
+ }
+
+ /*
+@@ -1484,6 +1682,7 @@ static enum hrtimer_restart hrtimer_wakeup(struct hrtimer *timer)
+ void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, struct task_struct *task)
+ {
+ sl->timer.function = hrtimer_wakeup;
++ sl->timer.irqsafe = 1;
+ sl->task = task;
+ }
+ EXPORT_SYMBOL_GPL(hrtimer_init_sleeper);
+@@ -1622,9 +1821,13 @@ static void __cpuinit init_hrtimers_cpu(int cpu)
+ for (i = 0; i < HRTIMER_MAX_CLOCK_BASES; i++) {
+ cpu_base->clock_base[i].cpu_base = cpu_base;
+ timerqueue_init_head(&cpu_base->clock_base[i].active);
++ INIT_LIST_HEAD(&cpu_base->clock_base[i].expired);
+ }
+
+ hrtimer_init_hres(cpu_base);
++#ifdef CONFIG_PREEMPT_RT_BASE
++ init_waitqueue_head(&cpu_base->wait);
+#endif
-+ break;
+ }
+
+ #ifdef CONFIG_HOTPLUG_CPU
+@@ -1737,9 +1940,7 @@ void __init hrtimers_init(void)
+ hrtimer_cpu_notify(&hrtimers_nb, (unsigned long)CPU_UP_PREPARE,
+ (void *)(long)smp_processor_id());
+ register_cpu_notifier(&hrtimers_nb);
+-#ifdef CONFIG_HIGH_RES_TIMERS
+ open_softirq(HRTIMER_SOFTIRQ, run_hrtimer_softirq);
+-#endif
+ }
+
+ /**
+diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c
+index 470d08c..634620c 100644
+--- a/kernel/irq/handle.c
++++ b/kernel/irq/handle.c
+@@ -156,8 +156,11 @@ handle_irq_event_percpu(struct irq_desc *desc, struct irqaction *action)
+ action = action->next;
+ } while (action);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
++ /* FIXME: Can we unbreak that ? */
+ if (random & IRQF_SAMPLE_RANDOM)
+ add_interrupt_randomness(irq);
+#endif
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ case MISSED_TIMER_OFFSETS:
-+ unregister_trace_hrtimer_interrupt(
-+ probe_hrtimer_interrupt, NULL);
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ timerandwakeup_enabled_data.enabled = 0;
+
+ if (!noirqdebug)
+ note_interrupt(irq, desc, retval);
+diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c
+index 0a7840ae..d624790 100644
+--- a/kernel/irq/manage.c
++++ b/kernel/irq/manage.c
+@@ -18,6 +18,7 @@
+ #include "internals.h"
+
+ #ifdef CONFIG_IRQ_FORCED_THREADING
++# ifndef CONFIG_PREEMPT_RT_BASE
+ __read_mostly bool force_irqthreads;
+
+ static int __init setup_forced_irqthreads(char *arg)
+@@ -26,6 +27,7 @@ static int __init setup_forced_irqthreads(char *arg)
+ return 0;
+ }
+ early_param("threadirqs", setup_forced_irqthreads);
++# endif
+ #endif
+
+ /**
+diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c
+index aa57d5d..e57f1b3 100644
+--- a/kernel/irq/spurious.c
++++ b/kernel/irq/spurious.c
+@@ -339,6 +339,11 @@ MODULE_PARM_DESC(noirqdebug, "Disable irq lockup detection when true");
+
+ static int __init irqfixup_setup(char *str)
+ {
++#ifdef CONFIG_PREEMPT_RT_BASE
++ printk(KERN_WARNING "irqfixup boot option not supported "
++ "w/ CONFIG_PREEMPT_RT\n");
++ return 1;
+#endif
-+ break;
+ irqfixup = 1;
+ printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n");
+ printk(KERN_WARNING "This may impact system performance.\n");
+@@ -351,6 +356,11 @@ module_param(irqfixup, int, 0644);
+
+ static int __init irqpoll_setup(char *str)
+ {
++#ifdef CONFIG_PREEMPT_RT_BASE
++ printk(KERN_WARNING "irqpoll boot option not supported "
++ "w/ CONFIG_PREEMPT_RT\n");
++ return 1;
+#endif
-+ default:
-+ break;
-+ }
-+ }
-+ ed->enabled = enable;
-+ return cnt;
+ irqfixup = 2;
+ printk(KERN_WARNING "Misrouted IRQ fixup and polling support "
+ "enabled\n");
+diff --git a/kernel/irq_work.c b/kernel/irq_work.c
+index c58fa7d..75ea006 100644
+--- a/kernel/irq_work.c
++++ b/kernel/irq_work.c
+@@ -125,8 +125,10 @@ void irq_work_run(void)
+ if (this_cpu_read(irq_work_list) == NULL)
+ return;
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ BUG_ON(!in_irq());
+ BUG_ON(!irqs_disabled());
++#endif
+
+ list = this_cpu_xchg(irq_work_list, NULL);
+
+diff --git a/kernel/itimer.c b/kernel/itimer.c
+index d802883..2c582fc 100644
+--- a/kernel/itimer.c
++++ b/kernel/itimer.c
+@@ -214,6 +214,7 @@ again:
+ /* We are sharing ->siglock with it_real_fn() */
+ if (hrtimer_try_to_cancel(timer) < 0) {
+ spin_unlock_irq(&tsk->sighand->siglock);
++ hrtimer_wait_for_timer(&tsk->signal->real_timer);
+ goto again;
+ }
+ expires = timeval_to_ktime(value->it_value);
+diff --git a/kernel/kprobes.c b/kernel/kprobes.c
+index 7798181..3553e90 100644
+--- a/kernel/kprobes.c
++++ b/kernel/kprobes.c
+@@ -78,10 +78,10 @@ static bool kprobes_all_disarmed;
+ static DEFINE_MUTEX(kprobe_mutex);
+ static DEFINE_PER_CPU(struct kprobe *, kprobe_instance) = NULL;
+ static struct {
+- spinlock_t lock ____cacheline_aligned_in_smp;
++ raw_spinlock_t lock ____cacheline_aligned_in_smp;
+ } kretprobe_table_locks[KPROBE_TABLE_SIZE];
+
+-static spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
++static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
+ {
+ return &(kretprobe_table_locks[hash].lock);
+ }
+@@ -1013,9 +1013,9 @@ void __kprobes recycle_rp_inst(struct kretprobe_instance *ri,
+ hlist_del(&ri->hlist);
+ INIT_HLIST_NODE(&ri->hlist);
+ if (likely(rp)) {
+- spin_lock(&rp->lock);
++ raw_spin_lock(&rp->lock);
+ hlist_add_head(&ri->hlist, &rp->free_instances);
+- spin_unlock(&rp->lock);
++ raw_spin_unlock(&rp->lock);
+ } else
+ /* Unregistering */
+ hlist_add_head(&ri->hlist, head);
+@@ -1026,19 +1026,19 @@ void __kprobes kretprobe_hash_lock(struct task_struct *tsk,
+ __acquires(hlist_lock)
+ {
+ unsigned long hash = hash_ptr(tsk, KPROBE_HASH_BITS);
+- spinlock_t *hlist_lock;
++ raw_spinlock_t *hlist_lock;
+
+ *head = &kretprobe_inst_table[hash];
+ hlist_lock = kretprobe_table_lock_ptr(hash);
+- spin_lock_irqsave(hlist_lock, *flags);
++ raw_spin_lock_irqsave(hlist_lock, *flags);
+ }
+
+ static void __kprobes kretprobe_table_lock(unsigned long hash,
+ unsigned long *flags)
+ __acquires(hlist_lock)
+ {
+- spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
+- spin_lock_irqsave(hlist_lock, *flags);
++ raw_spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
++ raw_spin_lock_irqsave(hlist_lock, *flags);
+ }
+
+ void __kprobes kretprobe_hash_unlock(struct task_struct *tsk,
+@@ -1046,18 +1046,18 @@ void __kprobes kretprobe_hash_unlock(struct task_struct *tsk,
+ __releases(hlist_lock)
+ {
+ unsigned long hash = hash_ptr(tsk, KPROBE_HASH_BITS);
+- spinlock_t *hlist_lock;
++ raw_spinlock_t *hlist_lock;
+
+ hlist_lock = kretprobe_table_lock_ptr(hash);
+- spin_unlock_irqrestore(hlist_lock, *flags);
++ raw_spin_unlock_irqrestore(hlist_lock, *flags);
+ }
+
+ static void __kprobes kretprobe_table_unlock(unsigned long hash,
+ unsigned long *flags)
+ __releases(hlist_lock)
+ {
+- spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
+- spin_unlock_irqrestore(hlist_lock, *flags);
++ raw_spinlock_t *hlist_lock = kretprobe_table_lock_ptr(hash);
++ raw_spin_unlock_irqrestore(hlist_lock, *flags);
+ }
+
+ /*
+@@ -1650,12 +1650,12 @@ static int __kprobes pre_handler_kretprobe(struct kprobe *p,
+
+ /*TODO: consider to only swap the RA after the last pre_handler fired */
+ hash = hash_ptr(current, KPROBE_HASH_BITS);
+- spin_lock_irqsave(&rp->lock, flags);
++ raw_spin_lock_irqsave(&rp->lock, flags);
+ if (!hlist_empty(&rp->free_instances)) {
+ ri = hlist_entry(rp->free_instances.first,
+ struct kretprobe_instance, hlist);
+ hlist_del(&ri->hlist);
+- spin_unlock_irqrestore(&rp->lock, flags);
++ raw_spin_unlock_irqrestore(&rp->lock, flags);
+
+ ri->rp = rp;
+ ri->task = current;
+@@ -1672,7 +1672,7 @@ static int __kprobes pre_handler_kretprobe(struct kprobe *p,
+ kretprobe_table_unlock(hash, &flags);
+ } else {
+ rp->nmissed++;
+- spin_unlock_irqrestore(&rp->lock, flags);
++ raw_spin_unlock_irqrestore(&rp->lock, flags);
+ }
+ return 0;
+ }
+@@ -1708,7 +1708,7 @@ int __kprobes register_kretprobe(struct kretprobe *rp)
+ rp->maxactive = num_possible_cpus();
+ #endif
+ }
+- spin_lock_init(&rp->lock);
++ raw_spin_lock_init(&rp->lock);
+ INIT_HLIST_HEAD(&rp->free_instances);
+ for (i = 0; i < rp->maxactive; i++) {
+ inst = kmalloc(sizeof(struct kretprobe_instance) +
+@@ -1946,7 +1946,7 @@ static int __init init_kprobes(void)
+ for (i = 0; i < KPROBE_TABLE_SIZE; i++) {
+ INIT_HLIST_HEAD(&kprobe_table[i]);
+ INIT_HLIST_HEAD(&kretprobe_inst_table[i]);
+- spin_lock_init(&(kretprobe_table_locks[i].lock));
++ raw_spin_lock_init(&(kretprobe_table_locks[i].lock));
+ }
+
+ /*
+diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c
+index 3b053c0..915b522 100644
+--- a/kernel/ksysfs.c
++++ b/kernel/ksysfs.c
+@@ -132,6 +132,15 @@ KERNEL_ATTR_RO(vmcoreinfo);
+
+ #endif /* CONFIG_KEXEC */
+
++#if defined(CONFIG_PREEMPT_RT_FULL)
++static ssize_t realtime_show(struct kobject *kobj,
++ struct kobj_attribute *attr, char *buf)
++{
++ return sprintf(buf, "%d\n", 1);
+}
-+
-+static const struct file_operations latency_hist_reset_fops = {
-+ .open = tracing_open_generic,
-+ .write = latency_hist_reset,
-+};
-+
-+static const struct file_operations enable_fops = {
-+ .open = tracing_open_generic,
-+ .read = show_enable,
-+ .write = do_enable,
-+};
-+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+static const struct file_operations pid_fops = {
-+ .open = tracing_open_generic,
-+ .read = show_pid,
-+ .write = do_pid,
-+};
-+
-+static const struct file_operations maxlatproc_fops = {
-+ .open = tracing_open_generic,
-+ .read = show_maxlatproc,
-+};
++KERNEL_ATTR_RO(realtime);
+#endif
+
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
-+static notrace void probe_preemptirqsoff_hist(void *v, int reason,
-+ int starthist)
-+{
-+ int cpu = raw_smp_processor_id();
-+ int time_set = 0;
-+
-+ if (starthist) {
-+ cycle_t uninitialized_var(start);
-+
-+ if (!preempt_count() && !irqs_disabled())
-+ return;
-+
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ if ((reason == IRQS_OFF || reason == TRACE_START) &&
-+ !per_cpu(hist_irqsoff_counting, cpu)) {
-+ per_cpu(hist_irqsoff_counting, cpu) = 1;
-+ start = ftrace_now(cpu);
-+ time_set++;
-+ per_cpu(hist_irqsoff_start, cpu) = start;
-+ }
+ /* whether file capabilities are enabled */
+ static ssize_t fscaps_show(struct kobject *kobj,
+ struct kobj_attribute *attr, char *buf)
+@@ -181,6 +190,9 @@ static struct attribute * kernel_attrs[] = {
+ &kexec_crash_size_attr.attr,
+ &vmcoreinfo_attr.attr,
+ #endif
++#ifdef CONFIG_PREEMPT_RT_FULL
++ &realtime_attr.attr,
+#endif
-+
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ if ((reason == PREEMPT_OFF || reason == TRACE_START) &&
-+ !per_cpu(hist_preemptoff_counting, cpu)) {
-+ per_cpu(hist_preemptoff_counting, cpu) = 1;
-+ if (!(time_set++))
-+ start = ftrace_now(cpu);
-+ per_cpu(hist_preemptoff_start, cpu) = start;
-+ }
+ NULL
+ };
+
+diff --git a/kernel/latencytop.c b/kernel/latencytop.c
+index 376066e..4ac8ebf 100644
+--- a/kernel/latencytop.c
++++ b/kernel/latencytop.c
+@@ -58,7 +58,7 @@
+ #include <linux/list.h>
+ #include <linux/stacktrace.h>
+
+-static DEFINE_SPINLOCK(latency_lock);
++static DEFINE_RAW_SPINLOCK(latency_lock);
+
+ #define MAXLR 128
+ static struct latency_record latency_record[MAXLR];
+@@ -72,19 +72,19 @@ void clear_all_latency_tracing(struct task_struct *p)
+ if (!latencytop_enabled)
+ return;
+
+- spin_lock_irqsave(&latency_lock, flags);
++ raw_spin_lock_irqsave(&latency_lock, flags);
+ memset(&p->latency_record, 0, sizeof(p->latency_record));
+ p->latency_record_count = 0;
+- spin_unlock_irqrestore(&latency_lock, flags);
++ raw_spin_unlock_irqrestore(&latency_lock, flags);
+ }
+
+ static void clear_global_latency_tracing(void)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&latency_lock, flags);
++ raw_spin_lock_irqsave(&latency_lock, flags);
+ memset(&latency_record, 0, sizeof(latency_record));
+- spin_unlock_irqrestore(&latency_lock, flags);
++ raw_spin_unlock_irqrestore(&latency_lock, flags);
+ }
+
+ static void __sched
+@@ -190,7 +190,7 @@ __account_scheduler_latency(struct task_struct *tsk, int usecs, int inter)
+ lat.max = usecs;
+ store_stacktrace(tsk, &lat);
+
+- spin_lock_irqsave(&latency_lock, flags);
++ raw_spin_lock_irqsave(&latency_lock, flags);
+
+ account_global_scheduler_latency(tsk, &lat);
+
+@@ -231,7 +231,7 @@ __account_scheduler_latency(struct task_struct *tsk, int usecs, int inter)
+ memcpy(&tsk->latency_record[i], &lat, sizeof(struct latency_record));
+
+ out_unlock:
+- spin_unlock_irqrestore(&latency_lock, flags);
++ raw_spin_unlock_irqrestore(&latency_lock, flags);
+ }
+
+ static int lstats_show(struct seq_file *m, void *v)
+diff --git a/kernel/lockdep.c b/kernel/lockdep.c
+index 298c927..a8d5a46 100644
+--- a/kernel/lockdep.c
++++ b/kernel/lockdep.c
+@@ -2859,10 +2859,7 @@ static int mark_lock(struct task_struct *curr, struct held_lock *this,
+ void lockdep_init_map(struct lockdep_map *lock, const char *name,
+ struct lock_class_key *key, int subclass)
+ {
+- int i;
+-
+- for (i = 0; i < NR_LOCKDEP_CACHING_CLASSES; i++)
+- lock->class_cache[i] = NULL;
++ memset(lock, 0, sizeof(*lock));
+
+ #ifdef CONFIG_LOCK_STAT
+ lock->cpu = raw_smp_processor_id();
+@@ -3341,6 +3338,7 @@ static void check_flags(unsigned long flags)
+ }
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * We dont accurately track softirq state in e.g.
+ * hardirq contexts (such as on 4KSTACKS), so only
+@@ -3352,6 +3350,7 @@ static void check_flags(unsigned long flags)
+ else
+ DEBUG_LOCKS_WARN_ON(!current->softirqs_enabled);
+ }
+#endif
-+
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
-+ if (per_cpu(hist_irqsoff_counting, cpu) &&
-+ per_cpu(hist_preemptoff_counting, cpu) &&
-+ !per_cpu(hist_preemptirqsoff_counting, cpu)) {
-+ per_cpu(hist_preemptirqsoff_counting, cpu) = 1;
-+ if (!time_set)
-+ start = ftrace_now(cpu);
-+ per_cpu(hist_preemptirqsoff_start, cpu) = start;
-+ }
+
+ if (!debug_locks)
+ print_irqtrace_events(current);
+diff --git a/kernel/mutex.c b/kernel/mutex.c
+index d607ed5..abb9291 100644
+--- a/kernel/mutex.c
++++ b/kernel/mutex.c
+@@ -240,9 +240,7 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
+
+ /* didn't get the lock, go to sleep: */
+ spin_unlock_mutex(&lock->wait_lock, flags);
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
++ schedule_preempt_disabled();
+ spin_lock_mutex(&lock->wait_lock, flags);
+ }
+
+diff --git a/kernel/panic.c b/kernel/panic.c
+index 6923167..df22c71 100644
+--- a/kernel/panic.c
++++ b/kernel/panic.c
+@@ -330,9 +330,11 @@ static u64 oops_id;
+
+ static int init_oops_id(void)
+ {
++#ifndef CONFIG_PREEMPT_RT_FULL
+ if (!oops_id)
+ get_random_bytes(&oops_id, sizeof(oops_id));
+ else
+#endif
-+ } else {
-+ cycle_t uninitialized_var(stop);
+ oops_id++;
+
+ return 0;
+diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
+index 640ded8..1d4c609 100644
+--- a/kernel/posix-cpu-timers.c
++++ b/kernel/posix-cpu-timers.c
+@@ -282,13 +282,13 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
+ * it.
+ */
+ thread_group_cputime(tsk, &sum);
+- spin_lock_irqsave(&cputimer->lock, flags);
++ raw_spin_lock_irqsave(&cputimer->lock, flags);
+ cputimer->running = 1;
+ update_gt_cputime(&cputimer->cputime, &sum);
+ } else
+- spin_lock_irqsave(&cputimer->lock, flags);
++ raw_spin_lock_irqsave(&cputimer->lock, flags);
+ *times = cputimer->cputime;
+- spin_unlock_irqrestore(&cputimer->lock, flags);
++ raw_spin_unlock_irqrestore(&cputimer->lock, flags);
+ }
+
+ /*
+@@ -701,7 +701,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int flags,
+ /*
+ * Disarm any old timer after extracting its expiry time.
+ */
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
+
+ ret = 0;
+ old_incr = timer->it.cpu.incr;
+@@ -999,9 +999,9 @@ static void stop_process_timers(struct signal_struct *sig)
+ struct thread_group_cputimer *cputimer = &sig->cputimer;
+ unsigned long flags;
+
+- spin_lock_irqsave(&cputimer->lock, flags);
++ raw_spin_lock_irqsave(&cputimer->lock, flags);
+ cputimer->running = 0;
+- spin_unlock_irqrestore(&cputimer->lock, flags);
++ raw_spin_unlock_irqrestore(&cputimer->lock, flags);
+ }
+
+ static u32 onecputick;
+@@ -1223,7 +1223,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
+ /*
+ * Now re-arm for the new expiry time.
+ */
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
+ arm_timer(timer);
+ spin_unlock(&p->sighand->siglock);
+
+@@ -1290,10 +1290,11 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
+ sig = tsk->signal;
+ if (sig->cputimer.running) {
+ struct task_cputime group_sample;
++ unsigned long flags;
+
+- spin_lock(&sig->cputimer.lock);
++ raw_spin_lock_irqsave(&sig->cputimer.lock, flags);
+ group_sample = sig->cputimer.cputime;
+- spin_unlock(&sig->cputimer.lock);
++ raw_spin_unlock_irqrestore(&sig->cputimer.lock, flags);
+
+ if (task_cputime_expired(&group_sample, &sig->cputime_expires))
+ return 1;
+@@ -1307,13 +1308,13 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
+ * already updated our counts. We need to check if any timers fire now.
+ * Interrupts are disabled.
+ */
+-void run_posix_cpu_timers(struct task_struct *tsk)
++static void __run_posix_cpu_timers(struct task_struct *tsk)
+ {
+ LIST_HEAD(firing);
+ struct k_itimer *timer, *next;
+ unsigned long flags;
+
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
+
+ /*
+ * The fast path checks that there are no expired thread or thread
+@@ -1371,6 +1372,190 @@ void run_posix_cpu_timers(struct task_struct *tsk)
+ }
+ }
+
++#ifdef CONFIG_PREEMPT_RT_BASE
++#include <linux/kthread.h>
++#include <linux/cpu.h>
++DEFINE_PER_CPU(struct task_struct *, posix_timer_task);
++DEFINE_PER_CPU(struct task_struct *, posix_timer_tasklist);
+
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ if ((reason == IRQS_ON || reason == TRACE_STOP) &&
-+ per_cpu(hist_irqsoff_counting, cpu)) {
-+ cycle_t start = per_cpu(hist_irqsoff_start, cpu);
++static int posix_cpu_timers_thread(void *data)
++{
++ int cpu = (long)data;
+
-+ stop = ftrace_now(cpu);
-+ time_set++;
-+ if (start && stop >= start) {
-+ unsigned long latency =
-+ nsecs_to_usecs(stop - start);
++ BUG_ON(per_cpu(posix_timer_task,cpu) != current);
+
-+ latency_hist(IRQSOFF_LATENCY, cpu, latency, 0,
-+ stop, NULL);
-+ }
-+ per_cpu(hist_irqsoff_counting, cpu) = 0;
-+ }
-+#endif
++ while (!kthread_should_stop()) {
++ struct task_struct *tsk = NULL;
++ struct task_struct *next = NULL;
+
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ if ((reason == PREEMPT_ON || reason == TRACE_STOP) &&
-+ per_cpu(hist_preemptoff_counting, cpu)) {
-+ cycle_t start = per_cpu(hist_preemptoff_start, cpu);
++ if (cpu_is_offline(cpu))
++ goto wait_to_die;
+
-+ if (!(time_set++))
-+ stop = ftrace_now(cpu);
-+ if (start && stop >= start) {
-+ unsigned long latency =
-+ nsecs_to_usecs(stop - start);
++ /* grab task list */
++ raw_local_irq_disable();
++ tsk = per_cpu(posix_timer_tasklist, cpu);
++ per_cpu(posix_timer_tasklist, cpu) = NULL;
++ raw_local_irq_enable();
+
-+ latency_hist(PREEMPTOFF_LATENCY, cpu, latency,
-+ 0, stop, NULL);
-+ }
-+ per_cpu(hist_preemptoff_counting, cpu) = 0;
++ /* its possible the list is empty, just return */
++ if (!tsk) {
++ set_current_state(TASK_INTERRUPTIBLE);
++ schedule();
++ __set_current_state(TASK_RUNNING);
++ continue;
+ }
-+#endif
+
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
-+ if ((!per_cpu(hist_irqsoff_counting, cpu) ||
-+ !per_cpu(hist_preemptoff_counting, cpu)) &&
-+ per_cpu(hist_preemptirqsoff_counting, cpu)) {
-+ cycle_t start = per_cpu(hist_preemptirqsoff_start, cpu);
++ /* Process task list */
++ while (1) {
++ /* save next */
++ next = tsk->posix_timer_list;
+
-+ if (!time_set)
-+ stop = ftrace_now(cpu);
-+ if (start && stop >= start) {
-+ unsigned long latency =
-+ nsecs_to_usecs(stop - start);
-+ latency_hist(PREEMPTIRQSOFF_LATENCY, cpu,
-+ latency, 0, stop, NULL);
-+ }
-+ per_cpu(hist_preemptirqsoff_counting, cpu) = 0;
++ /* run the task timers, clear its ptr and
++ * unreference it
++ */
++ __run_posix_cpu_timers(tsk);
++ tsk->posix_timer_list = NULL;
++ put_task_struct(tsk);
++
++ /* check if this is the last on the list */
++ if (next == tsk)
++ break;
++ tsk = next;
+ }
-+#endif
+ }
++ return 0;
++
++wait_to_die:
++ /* Wait for kthread_stop */
++ set_current_state(TASK_INTERRUPTIBLE);
++ while (!kthread_should_stop()) {
++ schedule();
++ set_current_state(TASK_INTERRUPTIBLE);
++ }
++ __set_current_state(TASK_RUNNING);
++ return 0;
+}
-+#endif
+
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+static DEFINE_RAW_SPINLOCK(wakeup_lock);
-+static notrace void probe_sched_migrate_task(void *v, struct task_struct *task,
-+ int cpu)
++static inline int __fastpath_timer_check(struct task_struct *tsk)
+{
-+ int old_cpu = task_cpu(task);
++ /* tsk == current, ensure it is safe to use ->signal/sighand */
++ if (unlikely(tsk->exit_state))
++ return 0;
+
-+ if (cpu != old_cpu) {
-+ unsigned long flags;
-+ struct task_struct *cpu_wakeup_task;
++ if (!task_cputime_zero(&tsk->cputime_expires))
++ return 1;
+
-+ raw_spin_lock_irqsave(&wakeup_lock, flags);
++ if (!task_cputime_zero(&tsk->signal->cputime_expires))
++ return 1;
+
-+ cpu_wakeup_task = per_cpu(wakeup_task, old_cpu);
-+ if (task == cpu_wakeup_task) {
-+ put_task_struct(cpu_wakeup_task);
-+ per_cpu(wakeup_task, old_cpu) = NULL;
-+ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = task;
-+ get_task_struct(cpu_wakeup_task);
++ return 0;
++}
++
++void run_posix_cpu_timers(struct task_struct *tsk)
++{
++ unsigned long cpu = smp_processor_id();
++ struct task_struct *tasklist;
++
++ BUG_ON(!irqs_disabled());
++ if(!per_cpu(posix_timer_task, cpu))
++ return;
++ /* get per-cpu references */
++ tasklist = per_cpu(posix_timer_tasklist, cpu);
++
++ /* check to see if we're already queued */
++ if (!tsk->posix_timer_list && __fastpath_timer_check(tsk)) {
++ get_task_struct(tsk);
++ if (tasklist) {
++ tsk->posix_timer_list = tasklist;
++ } else {
++ /*
++ * The list is terminated by a self-pointing
++ * task_struct
++ */
++ tsk->posix_timer_list = tsk;
+ }
++ per_cpu(posix_timer_tasklist, cpu) = tsk;
+
-+ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
++ wake_up_process(per_cpu(posix_timer_task, cpu));
+ }
+}
+
-+static notrace void probe_wakeup_latency_hist_start(void *v,
-+ struct task_struct *p, int success)
++/*
++ * posix_cpu_thread_call - callback that gets triggered when a CPU is added.
++ * Here we can start up the necessary migration thread for the new CPU.
++ */
++static int posix_cpu_thread_call(struct notifier_block *nfb,
++ unsigned long action, void *hcpu)
+{
-+ unsigned long flags;
-+ struct task_struct *curr = current;
-+ int cpu = task_cpu(p);
-+ struct task_struct *cpu_wakeup_task;
++ int cpu = (long)hcpu;
++ struct task_struct *p;
++ struct sched_param param;
+
-+ raw_spin_lock_irqsave(&wakeup_lock, flags);
++ switch (action) {
++ case CPU_UP_PREPARE:
++ p = kthread_create(posix_cpu_timers_thread, hcpu,
++ "posixcputmr/%d",cpu);
++ if (IS_ERR(p))
++ return NOTIFY_BAD;
++ p->flags |= PF_NOFREEZE;
++ kthread_bind(p, cpu);
++ /* Must be high prio to avoid getting starved */
++ param.sched_priority = MAX_RT_PRIO-1;
++ sched_setscheduler(p, SCHED_FIFO, ¶m);
++ per_cpu(posix_timer_task,cpu) = p;
++ break;
++ case CPU_ONLINE:
++ /* Strictly unneccessary, as first user will wake it. */
++ wake_up_process(per_cpu(posix_timer_task,cpu));
++ break;
++#ifdef CONFIG_HOTPLUG_CPU
++ case CPU_UP_CANCELED:
++ /* Unbind it from offline cpu so it can run. Fall thru. */
++ kthread_bind(per_cpu(posix_timer_task,cpu),
++ any_online_cpu(cpu_online_map));
++ kthread_stop(per_cpu(posix_timer_task,cpu));
++ per_cpu(posix_timer_task,cpu) = NULL;
++ break;
++ case CPU_DEAD:
++ kthread_stop(per_cpu(posix_timer_task,cpu));
++ per_cpu(posix_timer_task,cpu) = NULL;
++ break;
++#endif
++ }
++ return NOTIFY_OK;
++}
+
-+ cpu_wakeup_task = per_cpu(wakeup_task, cpu);
++/* Register at highest priority so that task migration (migrate_all_tasks)
++ * happens before everything else.
++ */
++static struct notifier_block __devinitdata posix_cpu_thread_notifier = {
++ .notifier_call = posix_cpu_thread_call,
++ .priority = 10
++};
+
-+ if (wakeup_pid) {
-+ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) ||
-+ p->prio == curr->prio)
-+ per_cpu(wakeup_sharedprio, cpu) = 1;
-+ if (likely(wakeup_pid != task_pid_nr(p)))
-+ goto out;
-+ } else {
-+ if (likely(!rt_task(p)) ||
-+ (cpu_wakeup_task && p->prio > cpu_wakeup_task->prio) ||
-+ p->prio > curr->prio)
-+ goto out;
-+ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) ||
-+ p->prio == curr->prio)
-+ per_cpu(wakeup_sharedprio, cpu) = 1;
-+ }
++static int __init posix_cpu_thread_init(void)
++{
++ void *hcpu = (void *)(long)smp_processor_id();
++ /* Start one for boot CPU. */
++ unsigned long cpu;
+
-+ if (cpu_wakeup_task)
-+ put_task_struct(cpu_wakeup_task);
-+ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = p;
-+ get_task_struct(cpu_wakeup_task);
-+ cpu_wakeup_task->preempt_timestamp_hist =
-+ ftrace_now(raw_smp_processor_id());
-+out:
-+ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
++ /* init the per-cpu posix_timer_tasklets */
++ for_each_cpu_mask(cpu, cpu_possible_map)
++ per_cpu(posix_timer_tasklist, cpu) = NULL;
++
++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_UP_PREPARE, hcpu);
++ posix_cpu_thread_call(&posix_cpu_thread_notifier, CPU_ONLINE, hcpu);
++ register_cpu_notifier(&posix_cpu_thread_notifier);
++ return 0;
++}
++early_initcall(posix_cpu_thread_init);
++#else /* CONFIG_PREEMPT_RT_BASE */
++void run_posix_cpu_timers(struct task_struct *tsk)
++{
++ __run_posix_cpu_timers(tsk);
+}
++#endif /* CONFIG_PREEMPT_RT_BASE */
+
-+static notrace void probe_wakeup_latency_hist_stop(void *v,
-+ struct task_struct *prev, struct task_struct *next)
+ /*
+ * Set one of the process-wide special case CPU timers or RLIMIT_CPU.
+ * The tsk->sighand->siglock must be held by the caller.
+diff --git a/kernel/posix-timers.c b/kernel/posix-timers.c
+index 4556182..d928a65 100644
+--- a/kernel/posix-timers.c
++++ b/kernel/posix-timers.c
+@@ -439,6 +439,7 @@ static enum hrtimer_restart posix_timer_fn(struct hrtimer *timer)
+ static struct pid *good_sigevent(sigevent_t * event)
+ {
+ struct task_struct *rtn = current->group_leader;
++ int sig = event->sigev_signo;
+
+ if ((event->sigev_notify & SIGEV_THREAD_ID ) &&
+ (!(rtn = find_task_by_vpid(event->sigev_notify_thread_id)) ||
+@@ -447,7 +448,8 @@ static struct pid *good_sigevent(sigevent_t * event)
+ return NULL;
+
+ if (((event->sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE) &&
+- ((event->sigev_signo <= 0) || (event->sigev_signo > SIGRTMAX)))
++ (sig <= 0 || sig > SIGRTMAX || sig_kernel_only(sig) ||
++ sig_kernel_coredump(sig)))
+ return NULL;
+
+ return task_pid(rtn);
+@@ -764,6 +766,20 @@ SYSCALL_DEFINE1(timer_getoverrun, timer_t, timer_id)
+ return overrun;
+ }
+
++/*
++ * Protected by RCU!
++ */
++static void timer_wait_for_callback(struct k_clock *kc, struct k_itimer *timr)
+{
-+ unsigned long flags;
-+ int cpu = task_cpu(next);
-+ unsigned long latency;
-+ cycle_t stop;
-+ struct task_struct *cpu_wakeup_task;
-+
-+ raw_spin_lock_irqsave(&wakeup_lock, flags);
++#ifdef CONFIG_PREEMPT_RT_FULL
++ if (kc->timer_set == common_timer_set)
++ hrtimer_wait_for_timer(&timr->it.real.timer);
++ else
++ /* FIXME: Whacky hack for posix-cpu-timers */
++ schedule_timeout(1);
++#endif
++}
+
-+ cpu_wakeup_task = per_cpu(wakeup_task, cpu);
+ /* Set a POSIX.1b interval timer. */
+ /* timr->it_lock is taken. */
+ static int
+@@ -841,6 +857,7 @@ retry:
+ if (!timr)
+ return -EINVAL;
+
++ rcu_read_lock();
+ kc = clockid_to_kclock(timr->it_clock);
+ if (WARN_ON_ONCE(!kc || !kc->timer_set))
+ error = -EINVAL;
+@@ -849,9 +866,12 @@ retry:
+
+ unlock_timer(timr, flag);
+ if (error == TIMER_RETRY) {
++ timer_wait_for_callback(kc, timr);
+ rtn = NULL; // We already got the old time...
++ rcu_read_unlock();
+ goto retry;
+ }
++ rcu_read_unlock();
+
+ if (old_setting && !error &&
+ copy_to_user(old_setting, &old_spec, sizeof (old_spec)))
+@@ -889,10 +909,15 @@ retry_delete:
+ if (!timer)
+ return -EINVAL;
+
++ rcu_read_lock();
+ if (timer_delete_hook(timer) == TIMER_RETRY) {
+ unlock_timer(timer, flags);
++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
++ timer);
++ rcu_read_unlock();
+ goto retry_delete;
+ }
++ rcu_read_unlock();
+
+ spin_lock(¤t->sighand->siglock);
+ list_del(&timer->list);
+@@ -918,8 +943,18 @@ static void itimer_delete(struct k_itimer *timer)
+ retry_delete:
+ spin_lock_irqsave(&timer->it_lock, flags);
+
++ /* On RT we can race with a deletion */
++ if (!timer->it_signal) {
++ unlock_timer(timer, flags);
++ return;
++ }
+
-+ if (cpu_wakeup_task == NULL)
-+ goto out;
+ if (timer_delete_hook(timer) == TIMER_RETRY) {
++ rcu_read_lock();
+ unlock_timer(timer, flags);
++ timer_wait_for_callback(clockid_to_kclock(timer->it_clock),
++ timer);
++ rcu_read_unlock();
+ goto retry_delete;
+ }
+ list_del(&timer->list);
+diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
+index 8f7b1db..2b0c52f 100644
+--- a/kernel/power/hibernate.c
++++ b/kernel/power/hibernate.c
+@@ -278,6 +278,8 @@ static int create_image(int platform_mode)
+
+ local_irq_disable();
+
++ system_state = SYSTEM_SUSPEND;
+
-+ /* Already running? */
-+ if (unlikely(current == cpu_wakeup_task))
-+ goto out_reset;
+ error = syscore_suspend();
+ if (error) {
+ printk(KERN_ERR "PM: Some system devices failed to power down, "
+@@ -305,6 +307,7 @@ static int create_image(int platform_mode)
+ syscore_resume();
+
+ Enable_irqs:
++ system_state = SYSTEM_RUNNING;
+ local_irq_enable();
+
+ Enable_cpus:
+@@ -412,6 +415,7 @@ static int resume_target_kernel(bool platform_mode)
+ goto Enable_cpus;
+
+ local_irq_disable();
++ system_state = SYSTEM_SUSPEND;
+
+ error = syscore_suspend();
+ if (error)
+@@ -445,6 +449,7 @@ static int resume_target_kernel(bool platform_mode)
+ syscore_resume();
+
+ Enable_irqs:
++ system_state = SYSTEM_RUNNING;
+ local_irq_enable();
+
+ Enable_cpus:
+@@ -524,6 +529,7 @@ int hibernation_platform_enter(void)
+ goto Platform_finish;
+
+ local_irq_disable();
++ system_state = SYSTEM_SUSPEND;
+ syscore_suspend();
+ if (pm_wakeup_pending()) {
+ error = -EAGAIN;
+@@ -536,6 +542,7 @@ int hibernation_platform_enter(void)
+
+ Power_up:
+ syscore_resume();
++ system_state = SYSTEM_RUNNING;
+ local_irq_enable();
+ enable_nonboot_cpus();
+
+diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c
+index 449ccc9..bb5dcac 100644
+--- a/kernel/power/suspend.c
++++ b/kernel/power/suspend.c
+@@ -163,6 +163,8 @@ static int suspend_enter(suspend_state_t state)
+ arch_suspend_disable_irqs();
+ BUG_ON(!irqs_disabled());
+
++ system_state = SYSTEM_SUSPEND;
+
-+ if (next != cpu_wakeup_task) {
-+ if (next->prio < cpu_wakeup_task->prio)
-+ goto out_reset;
+ error = syscore_suspend();
+ if (!error) {
+ if (!(suspend_test(TEST_CORE) || pm_wakeup_pending())) {
+@@ -172,6 +174,8 @@ static int suspend_enter(suspend_state_t state)
+ syscore_resume();
+ }
+
++ system_state = SYSTEM_RUNNING;
+
-+ if (next->prio == cpu_wakeup_task->prio)
-+ per_cpu(wakeup_sharedprio, cpu) = 1;
+ arch_suspend_enable_irqs();
+ BUG_ON(irqs_disabled());
+
+diff --git a/kernel/printk.c b/kernel/printk.c
+index 084982f..a50af4e 100644
+--- a/kernel/printk.c
++++ b/kernel/printk.c
+@@ -21,6 +21,7 @@
+ #include <linux/tty.h>
+ #include <linux/tty_driver.h>
+ #include <linux/console.h>
++#include <linux/sysrq.h>
+ #include <linux/init.h>
+ #include <linux/jiffies.h>
+ #include <linux/nmi.h>
+@@ -44,13 +45,6 @@
+
+ #include <asm/uaccess.h>
+
+-/*
+- * Architectures can override it:
+- */
+-void asmlinkage __attribute__((weak)) early_printk(const char *fmt, ...)
+-{
+-}
+-
+ #define __LOG_BUF_LEN (1 << CONFIG_LOG_BUF_SHIFT)
+
+ /* printk's without a loglevel use this.. */
+@@ -100,7 +94,7 @@ static int console_locked, console_suspended;
+ * It is also used in interesting ways to provide interlocking in
+ * console_unlock();.
+ */
+-static DEFINE_SPINLOCK(logbuf_lock);
++static DEFINE_RAW_SPINLOCK(logbuf_lock);
+
+ #define LOG_BUF_MASK (log_buf_len-1)
+ #define LOG_BUF(idx) (log_buf[(idx) & LOG_BUF_MASK])
+@@ -212,7 +206,7 @@ void __init setup_log_buf(int early)
+ return;
+ }
+
+- spin_lock_irqsave(&logbuf_lock, flags);
++ raw_spin_lock_irqsave(&logbuf_lock, flags);
+ log_buf_len = new_log_buf_len;
+ log_buf = new_log_buf;
+ new_log_buf_len = 0;
+@@ -230,7 +224,7 @@ void __init setup_log_buf(int early)
+ log_start -= offset;
+ con_start -= offset;
+ log_end -= offset;
+- spin_unlock_irqrestore(&logbuf_lock, flags);
++ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+
+ pr_info("log_buf_len: %d\n", log_buf_len);
+ pr_info("early log buf free: %d(%d%%)\n",
+@@ -363,18 +357,18 @@ int do_syslog(int type, char __user *buf, int len, bool from_file)
+ if (error)
+ goto out;
+ i = 0;
+- spin_lock_irq(&logbuf_lock);
++ raw_spin_lock_irq(&logbuf_lock);
+ while (!error && (log_start != log_end) && i < len) {
+ c = LOG_BUF(log_start);
+ log_start++;
+- spin_unlock_irq(&logbuf_lock);
++ raw_spin_unlock_irq(&logbuf_lock);
+ error = __put_user(c,buf);
+ buf++;
+ i++;
+ cond_resched();
+- spin_lock_irq(&logbuf_lock);
++ raw_spin_lock_irq(&logbuf_lock);
+ }
+- spin_unlock_irq(&logbuf_lock);
++ raw_spin_unlock_irq(&logbuf_lock);
+ if (!error)
+ error = i;
+ break;
+@@ -397,7 +391,7 @@ int do_syslog(int type, char __user *buf, int len, bool from_file)
+ count = len;
+ if (count > log_buf_len)
+ count = log_buf_len;
+- spin_lock_irq(&logbuf_lock);
++ raw_spin_lock_irq(&logbuf_lock);
+ if (count > logged_chars)
+ count = logged_chars;
+ if (do_clear)
+@@ -414,12 +408,12 @@ int do_syslog(int type, char __user *buf, int len, bool from_file)
+ if (j + log_buf_len < log_end)
+ break;
+ c = LOG_BUF(j);
+- spin_unlock_irq(&logbuf_lock);
++ raw_spin_unlock_irq(&logbuf_lock);
+ error = __put_user(c,&buf[count-1-i]);
+ cond_resched();
+- spin_lock_irq(&logbuf_lock);
++ raw_spin_lock_irq(&logbuf_lock);
+ }
+- spin_unlock_irq(&logbuf_lock);
++ raw_spin_unlock_irq(&logbuf_lock);
+ if (error)
+ break;
+ error = i;
+@@ -509,6 +503,7 @@ static void __call_console_drivers(unsigned start, unsigned end)
+ {
+ struct console *con;
+
++ migrate_disable();
+ for_each_console(con) {
+ if (exclusive_console && con != exclusive_console)
+ continue;
+@@ -517,8 +512,62 @@ static void __call_console_drivers(unsigned start, unsigned end)
+ (con->flags & CON_ANYTIME)))
+ con->write(con, &LOG_BUF(start), end - start);
+ }
++ migrate_enable();
++}
+
-+ goto out;
-+ }
++#ifdef CONFIG_EARLY_PRINTK
++struct console *early_console;
+
-+ /*
-+ * The task we are waiting for is about to be switched to.
-+ * Calculate latency and store it in histogram.
-+ */
-+ stop = ftrace_now(raw_smp_processor_id());
++static void early_vprintk(const char *fmt, va_list ap)
++{
++ char buf[512];
++ int n = vscnprintf(buf, sizeof(buf), fmt, ap);
++ if (early_console)
++ early_console->write(early_console, buf, n);
++}
+
-+ latency = nsecs_to_usecs(stop - next->preempt_timestamp_hist);
++asmlinkage void early_printk(const char *fmt, ...)
++{
++ va_list ap;
++ va_start(ap, fmt);
++ early_vprintk(fmt, ap);
++ va_end(ap);
+ }
+
++/*
++ * This is independent of any log levels - a global
++ * kill switch that turns off all of printk.
++ *
++ * Used by the NMI watchdog if early-printk is enabled.
++ */
++static int __read_mostly printk_killswitch;
+
-+ if (per_cpu(wakeup_sharedprio, cpu)) {
-+ latency_hist(WAKEUP_LATENCY_SHAREDPRIO, cpu, latency, 0, stop,
-+ next);
-+ per_cpu(wakeup_sharedprio, cpu) = 0;
-+ } else {
-+ latency_hist(WAKEUP_LATENCY, cpu, latency, 0, stop, next);
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ if (timerandwakeup_enabled_data.enabled) {
-+ latency_hist(TIMERANDWAKEUP_LATENCY, cpu,
-+ next->timer_offset + latency, next->timer_offset,
-+ stop, next);
-+ }
-+#endif
-+ }
++static int __init force_early_printk_setup(char *str)
++{
++ printk_killswitch = 1;
++ return 0;
++}
++early_param("force_early_printk", force_early_printk_setup);
+
-+out_reset:
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ next->timer_offset = 0;
-+#endif
-+ put_task_struct(cpu_wakeup_task);
-+ per_cpu(wakeup_task, cpu) = NULL;
-+out:
-+ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
++void printk_kill(void)
++{
++ printk_killswitch = 1;
+}
-+#endif
+
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+static notrace void probe_hrtimer_interrupt(void *v, int cpu,
-+ long long latency_ns, struct task_struct *curr, struct task_struct *task)
++static int forced_early_printk(const char *fmt, va_list ap)
+{
-+ if (latency_ns <= 0 && task != NULL && rt_task(task) &&
-+ (task->prio < curr->prio ||
-+ (task->prio == curr->prio &&
-+ !cpumask_test_cpu(cpu, &task->cpus_allowed)))) {
-+ unsigned long latency;
-+ cycle_t now;
++ if (!printk_killswitch)
++ return 0;
++ early_vprintk(fmt, ap);
++ return 1;
++}
++#else
++static inline int forced_early_printk(const char *fmt, va_list ap)
++{
++ return 0;
++}
++#endif
+
-+ if (missed_timer_offsets_pid) {
-+ if (likely(missed_timer_offsets_pid !=
-+ task_pid_nr(task)))
-+ return;
-+ }
+ static int __read_mostly ignore_loglevel;
+
+ static int __init ignore_loglevel_setup(char *str)
+@@ -687,7 +736,7 @@ static void zap_locks(void)
+ oops_timestamp = jiffies;
+
+ /* If a crash is occurring, make sure we can't deadlock */
+- spin_lock_init(&logbuf_lock);
++ raw_spin_lock_init(&logbuf_lock);
+ /* And make sure that we print immediately */
+ sema_init(&console_sem, 1);
+ }
+@@ -779,12 +828,18 @@ static inline int can_use_console(unsigned int cpu)
+ * interrupts disabled. It should return with 'lockbuf_lock'
+ * released but interrupts still disabled.
+ */
+-static int console_trylock_for_printk(unsigned int cpu)
++static int console_trylock_for_printk(unsigned int cpu, unsigned long flags)
+ __releases(&logbuf_lock)
+ {
++#ifdef CONFIG_PREEMPT_RT_FULL
++ int lock = (!early_boot_irqs_disabled && !irqs_disabled_flags(flags) &&
++ !preempt_count()) || sysrq_in_progress;
++#else
++ int lock = 1;
++#endif
+ int retval = 0;
+
+- if (console_trylock()) {
++ if (lock && console_trylock()) {
+ retval = 1;
+
+ /*
+@@ -800,7 +855,7 @@ static int console_trylock_for_printk(unsigned int cpu)
+ }
+ }
+ printk_cpu = UINT_MAX;
+- spin_unlock(&logbuf_lock);
++ raw_spin_unlock(&logbuf_lock);
+ return retval;
+ }
+ static const char recursion_bug_msg [] =
+@@ -833,6 +888,13 @@ asmlinkage int vprintk(const char *fmt, va_list args)
+ size_t plen;
+ char special;
+
++ /*
++ * Fall back to early_printk if a debugging subsystem has
++ * killed printk output
++ */
++ if (unlikely(forced_early_printk(fmt, args)))
++ return 1;
+
-+ now = ftrace_now(cpu);
-+ latency = (unsigned long) div_s64(-latency_ns, 1000);
-+ latency_hist(MISSED_TIMER_OFFSETS, cpu, latency, latency, now,
-+ task);
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ task->timer_offset = latency;
+ boot_delay_msec();
+ printk_delay();
+
+@@ -860,7 +922,7 @@ asmlinkage int vprintk(const char *fmt, va_list args)
+ }
+
+ lockdep_off();
+- spin_lock(&logbuf_lock);
++ raw_spin_lock(&logbuf_lock);
+ printk_cpu = this_cpu;
+
+ if (recursion_bug) {
+@@ -953,8 +1015,15 @@ asmlinkage int vprintk(const char *fmt, va_list args)
+ * will release 'logbuf_lock' regardless of whether it
+ * actually gets the semaphore or not.
+ */
+- if (console_trylock_for_printk(this_cpu))
++ if (console_trylock_for_printk(this_cpu, flags)) {
++#ifndef CONFIG_PREEMPT_RT_FULL
+ console_unlock();
++#else
++ raw_local_irq_restore(flags);
++ console_unlock();
++ raw_local_irq_save(flags);
+#endif
+ }
-+}
+
+ lockdep_on();
+ out_restore_irqs:
+@@ -1213,8 +1282,8 @@ void printk_tick(void)
+
+ int printk_needs_cpu(int cpu)
+ {
+- if (cpu_is_offline(cpu))
+- printk_tick();
++ if (unlikely(cpu_is_offline(cpu)))
++ __this_cpu_write(printk_pending, 0);
+ return __this_cpu_read(printk_pending);
+ }
+
+@@ -1252,18 +1321,23 @@ void console_unlock(void)
+ console_may_schedule = 0;
+
+ for ( ; ; ) {
+- spin_lock_irqsave(&logbuf_lock, flags);
++ raw_spin_lock_irqsave(&logbuf_lock, flags);
+ wake_klogd |= log_start - log_end;
+ if (con_start == log_end)
+ break; /* Nothing to print */
+ _con_start = con_start;
+ _log_end = log_end;
+ con_start = log_end; /* Flush */
+- spin_unlock(&logbuf_lock);
++#ifndef CONFIG_PREEMPT_RT_FULL
++ raw_spin_unlock(&logbuf_lock);
+ stop_critical_timings(); /* don't trace print latency */
+ call_console_drivers(_con_start, _log_end);
+ start_critical_timings();
+ local_irq_restore(flags);
++#else
++ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
++ call_console_drivers(_con_start, _log_end);
++#endif
+ }
+ console_locked = 0;
+
+@@ -1272,7 +1346,7 @@ void console_unlock(void)
+ exclusive_console = NULL;
+
+ up(&console_sem);
+- spin_unlock_irqrestore(&logbuf_lock, flags);
++ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+ if (wake_klogd)
+ wake_up_klogd();
+ }
+@@ -1502,9 +1576,9 @@ void register_console(struct console *newcon)
+ * console_unlock(); will print out the buffered messages
+ * for us.
+ */
+- spin_lock_irqsave(&logbuf_lock, flags);
++ raw_spin_lock_irqsave(&logbuf_lock, flags);
+ con_start = log_start;
+- spin_unlock_irqrestore(&logbuf_lock, flags);
++ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+ /*
+ * We're about to replay the log buffer. Only do this to the
+ * just-registered console to avoid excessive message spam to
+@@ -1711,10 +1785,10 @@ void kmsg_dump(enum kmsg_dump_reason reason)
+ /* Theoretically, the log could move on after we do this, but
+ there's not a lot we can do about that. The new messages
+ will overwrite the start of what we dump. */
+- spin_lock_irqsave(&logbuf_lock, flags);
++ raw_spin_lock_irqsave(&logbuf_lock, flags);
+ end = log_end & LOG_BUF_MASK;
+ chars = logged_chars;
+- spin_unlock_irqrestore(&logbuf_lock, flags);
++ raw_spin_unlock_irqrestore(&logbuf_lock, flags);
+
+ if (chars > end) {
+ s1 = log_buf + log_buf_len - chars + end;
+diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
+index 7784bd2..5e0577b 100644
+--- a/kernel/rcupdate.c
++++ b/kernel/rcupdate.c
+@@ -72,6 +72,7 @@ int debug_lockdep_rcu_enabled(void)
+ }
+ EXPORT_SYMBOL_GPL(debug_lockdep_rcu_enabled);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /**
+ * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section?
+ *
+@@ -91,6 +92,7 @@ int rcu_read_lock_bh_held(void)
+ return in_softirq() || irqs_disabled();
+ }
+ EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held);
+#endif
-+
-+static __init int latency_hist_init(void)
+
+ #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */
+
+diff --git a/kernel/rcutree.c b/kernel/rcutree.c
+index ba06207..71341ab 100644
+--- a/kernel/rcutree.c
++++ b/kernel/rcutree.c
+@@ -166,6 +166,12 @@ void rcu_sched_qs(int cpu)
+ rdp->passed_quiesc = 1;
+ }
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++void rcu_bh_qs(int cpu)
+{
-+ struct dentry *latency_hist_root = NULL;
-+ struct dentry *dentry;
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ struct dentry *dentry_sharedprio;
-+#endif
-+ struct dentry *entry;
-+ struct dentry *enable_root;
-+ int i = 0;
-+ struct hist_data *my_hist;
-+ char name[64];
-+ char *cpufmt = "CPU%d";
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ char *cpufmt_maxlatproc = "max_latency-CPU%d";
-+ struct maxlatproc_data *mp = NULL;
-+#endif
-+
-+ dentry = tracing_init_dentry();
-+ latency_hist_root = debugfs_create_dir(latency_hist_dir_root, dentry);
-+ enable_root = debugfs_create_dir("enable", latency_hist_root);
-+
-+#ifdef CONFIG_INTERRUPT_OFF_HIST
-+ dentry = debugfs_create_dir(irqsoff_hist_dir, latency_hist_root);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(irqsoff_hist, i), &latency_hist_fops);
-+ my_hist = &per_cpu(irqsoff_hist, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+ }
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)IRQSOFF_LATENCY, &latency_hist_reset_fops);
++ rcu_preempt_qs(cpu);
++}
++#else
+ void rcu_bh_qs(int cpu)
+ {
+ struct rcu_data *rdp = &per_cpu(rcu_bh_data, cpu);
+@@ -174,6 +180,7 @@ void rcu_bh_qs(int cpu)
+ barrier();
+ rdp->passed_quiesc = 1;
+ }
+#endif
-+
-+#ifdef CONFIG_PREEMPT_OFF_HIST
-+ dentry = debugfs_create_dir(preemptoff_hist_dir,
-+ latency_hist_root);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(preemptoff_hist, i), &latency_hist_fops);
-+ my_hist = &per_cpu(preemptoff_hist, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+ }
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)PREEMPTOFF_LATENCY, &latency_hist_reset_fops);
+
+ /*
+ * Note a context switch. This is a quiescent state for RCU-sched,
+@@ -216,6 +223,7 @@ long rcu_batches_completed_sched(void)
+ }
+ EXPORT_SYMBOL_GPL(rcu_batches_completed_sched);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * Return the number of RCU BH batches processed thus far for debug & stats.
+ */
+@@ -233,6 +241,7 @@ void rcu_bh_force_quiescent_state(void)
+ force_quiescent_state(&rcu_bh_state, 0);
+ }
+ EXPORT_SYMBOL_GPL(rcu_bh_force_quiescent_state);
+#endif
-+
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
-+ dentry = debugfs_create_dir(preemptirqsoff_hist_dir,
-+ latency_hist_root);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(preemptirqsoff_hist, i), &latency_hist_fops);
-+ my_hist = &per_cpu(preemptirqsoff_hist, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+ }
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)PREEMPTIRQSOFF_LATENCY, &latency_hist_reset_fops);
+
+ /*
+ * Record the number of times rcutorture tests have been initiated and
+@@ -1153,7 +1162,7 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp)
+ else
+ raw_spin_unlock_irqrestore(&rnp->lock, flags);
+ if (need_report & RCU_OFL_TASKS_EXP_GP)
+- rcu_report_exp_rnp(rsp, rnp);
++ rcu_report_exp_rnp(rsp, rnp, true);
+ rcu_node_kthread_setaffinity(rnp, -1);
+ }
+
+@@ -1579,6 +1588,7 @@ void call_rcu_sched(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
+ }
+ EXPORT_SYMBOL_GPL(call_rcu_sched);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * Queue an RCU for invocation after a quicker grace period.
+ */
+@@ -1587,6 +1597,7 @@ void call_rcu_bh(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
+ __call_rcu(head, func, &rcu_bh_state);
+ }
+ EXPORT_SYMBOL_GPL(call_rcu_bh);
+#endif
-+
-+#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
-+ entry = debugfs_create_file("preemptirqsoff", 0644,
-+ enable_root, (void *)&preemptirqsoff_enabled_data,
-+ &enable_fops);
+
+ /**
+ * synchronize_sched - wait until an rcu-sched grace period has elapsed.
+@@ -1628,6 +1639,7 @@ void synchronize_sched(void)
+ }
+ EXPORT_SYMBOL_GPL(synchronize_sched);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /**
+ * synchronize_rcu_bh - wait until an rcu_bh grace period has elapsed.
+ *
+@@ -1653,6 +1665,7 @@ void synchronize_rcu_bh(void)
+ destroy_rcu_head_on_stack(&rcu.head);
+ }
+ EXPORT_SYMBOL_GPL(synchronize_rcu_bh);
+#endif
-+
-+#ifdef CONFIG_WAKEUP_LATENCY_HIST
-+ dentry = debugfs_create_dir(wakeup_latency_hist_dir,
-+ latency_hist_root);
-+ dentry_sharedprio = debugfs_create_dir(
-+ wakeup_latency_hist_dir_sharedprio, dentry);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(wakeup_latency_hist, i),
-+ &latency_hist_fops);
-+ my_hist = &per_cpu(wakeup_latency_hist, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+
-+ entry = debugfs_create_file(name, 0444, dentry_sharedprio,
-+ &per_cpu(wakeup_latency_hist_sharedprio, i),
-+ &latency_hist_fops);
-+ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+
-+ sprintf(name, cpufmt_maxlatproc, i);
-+
-+ mp = &per_cpu(wakeup_maxlatproc, i);
-+ entry = debugfs_create_file(name, 0444, dentry, mp,
-+ &maxlatproc_fops);
-+ clear_maxlatprocdata(mp);
-+
-+ mp = &per_cpu(wakeup_maxlatproc_sharedprio, i);
-+ entry = debugfs_create_file(name, 0444, dentry_sharedprio, mp,
-+ &maxlatproc_fops);
-+ clear_maxlatprocdata(mp);
-+ }
-+ entry = debugfs_create_file("pid", 0644, dentry,
-+ (void *)&wakeup_pid, &pid_fops);
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)WAKEUP_LATENCY, &latency_hist_reset_fops);
-+ entry = debugfs_create_file("reset", 0644, dentry_sharedprio,
-+ (void *)WAKEUP_LATENCY_SHAREDPRIO, &latency_hist_reset_fops);
-+ entry = debugfs_create_file("wakeup", 0644,
-+ enable_root, (void *)&wakeup_latency_enabled_data,
-+ &enable_fops);
+
+ /*
+ * Check to see if there is any immediate RCU-related work to be done
+@@ -1806,6 +1819,7 @@ static void _rcu_barrier(struct rcu_state *rsp,
+ mutex_unlock(&rcu_barrier_mutex);
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /**
+ * rcu_barrier_bh - Wait until all in-flight call_rcu_bh() callbacks complete.
+ */
+@@ -1814,6 +1828,7 @@ void rcu_barrier_bh(void)
+ _rcu_barrier(&rcu_bh_state, call_rcu_bh);
+ }
+ EXPORT_SYMBOL_GPL(rcu_barrier_bh);
+#endif
+
+ /**
+ * rcu_barrier_sched - Wait for in-flight call_rcu_sched() callbacks.
+diff --git a/kernel/rcutree.h b/kernel/rcutree.h
+index 01b2ccd..a8faf16 100644
+--- a/kernel/rcutree.h
++++ b/kernel/rcutree.h
+@@ -422,6 +422,7 @@ DECLARE_PER_CPU(struct rcu_data, rcu_preempt_data);
+ /* Forward declarations for rcutree_plugin.h */
+ static void rcu_bootup_announce(void);
+ long rcu_batches_completed(void);
++static void rcu_preempt_qs(int cpu);
+ static void rcu_preempt_note_context_switch(int cpu);
+ static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp);
+ #ifdef CONFIG_HOTPLUG_CPU
+@@ -443,7 +444,8 @@ static void rcu_preempt_check_callbacks(int cpu);
+ static void rcu_preempt_process_callbacks(void);
+ void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu));
+ #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU)
+-static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp);
++static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
++ bool wake);
+ #endif /* #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_TREE_PREEMPT_RCU) */
+ static int rcu_preempt_pending(int cpu);
+ static int rcu_preempt_needs_cpu(int cpu);
+diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h
+index 8aafbb8..2bc9a23 100644
+--- a/kernel/rcutree_plugin.h
++++ b/kernel/rcutree_plugin.h
+@@ -299,6 +299,9 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+ int empty_exp;
+ unsigned long flags;
+ struct list_head *np;
++#ifdef CONFIG_RCU_BOOST
++ struct rt_mutex *rbmp = NULL;
++#endif /* #ifdef CONFIG_RCU_BOOST */
+ struct rcu_node *rnp;
+ int special;
+
+@@ -318,7 +321,7 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+ }
+
+ /* Hardware IRQ handlers cannot block. */
+- if (in_irq() || in_serving_softirq()) {
++ if (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_OFFSET)) {
+ local_irq_restore(flags);
+ return;
+ }
+@@ -344,6 +347,7 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+ smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
+ np = rcu_next_node_entry(t, rnp);
+ list_del_init(&t->rcu_node_entry);
++ t->rcu_blocked_node = NULL;
+ if (&t->rcu_node_entry == rnp->gp_tasks)
+ rnp->gp_tasks = np;
+ if (&t->rcu_node_entry == rnp->exp_tasks)
+@@ -351,13 +355,12 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+ #ifdef CONFIG_RCU_BOOST
+ if (&t->rcu_node_entry == rnp->boost_tasks)
+ rnp->boost_tasks = np;
+- /* Snapshot and clear ->rcu_boosted with rcu_node lock held. */
+- if (t->rcu_boosted) {
+- special |= RCU_READ_UNLOCK_BOOSTED;
+- t->rcu_boosted = 0;
++ /* Snapshot/clear ->rcu_boost_mutex with rcu_node lock held. */
++ if (t->rcu_boost_mutex) {
++ rbmp = t->rcu_boost_mutex;
++ t->rcu_boost_mutex = NULL;
+ }
+ #endif /* #ifdef CONFIG_RCU_BOOST */
+- t->rcu_blocked_node = NULL;
+
+ /*
+ * If this was the last task on the current list, and if
+@@ -371,10 +374,8 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+
+ #ifdef CONFIG_RCU_BOOST
+ /* Unboost if we were boosted. */
+- if (special & RCU_READ_UNLOCK_BOOSTED) {
+- rt_mutex_unlock(t->rcu_boost_mutex);
+- t->rcu_boost_mutex = NULL;
+- }
++ if (rbmp)
++ rt_mutex_unlock(rbmp);
+ #endif /* #ifdef CONFIG_RCU_BOOST */
+
+ /*
+@@ -382,7 +383,7 @@ static noinline void rcu_read_unlock_special(struct task_struct *t)
+ * then we need to report up the rcu_node hierarchy.
+ */
+ if (!empty_exp && !rcu_preempted_readers_exp(rnp))
+- rcu_report_exp_rnp(&rcu_preempt_state, rnp);
++ rcu_report_exp_rnp(&rcu_preempt_state, rnp, true);
+ } else {
+ local_irq_restore(flags);
+ }
+@@ -711,7 +712,8 @@ static int sync_rcu_preempt_exp_done(struct rcu_node *rnp)
+ *
+ * Caller must hold sync_rcu_preempt_exp_mutex.
+ */
+-static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
++static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
++ bool wake)
+ {
+ unsigned long flags;
+ unsigned long mask;
+@@ -724,7 +726,8 @@ static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
+ }
+ if (rnp->parent == NULL) {
+ raw_spin_unlock_irqrestore(&rnp->lock, flags);
+- wake_up(&sync_rcu_preempt_exp_wq);
++ if (wake)
++ wake_up(&sync_rcu_preempt_exp_wq);
+ break;
+ }
+ mask = rnp->grpmask;
+@@ -757,7 +760,7 @@ sync_rcu_preempt_exp_init(struct rcu_state *rsp, struct rcu_node *rnp)
+ must_wait = 1;
+ }
+ if (!must_wait)
+- rcu_report_exp_rnp(rsp, rnp);
++ rcu_report_exp_rnp(rsp, rnp, false);
+ }
+
+ /*
+@@ -1048,9 +1051,9 @@ EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);
+ * report on tasks preempted in RCU read-side critical sections during
+ * expedited RCU grace periods.
+ */
+-static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp)
++static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
++ bool wake)
+ {
+- return;
+ }
+
+ #endif /* #ifdef CONFIG_HOTPLUG_CPU */
+@@ -1199,7 +1202,6 @@ static int rcu_boost(struct rcu_node *rnp)
+ t = container_of(tb, struct task_struct, rcu_node_entry);
+ rt_mutex_init_proxy_locked(&mtx, t);
+ t->rcu_boost_mutex = &mtx;
+- t->rcu_boosted = 1;
+ raw_spin_unlock_irqrestore(&rnp->lock, flags);
+ rt_mutex_lock(&mtx); /* Side effect: boosts task t's priority. */
+ rt_mutex_unlock(&mtx); /* Keep lockdep happy. */
+@@ -1890,7 +1892,7 @@ EXPORT_SYMBOL_GPL(synchronize_sched_expedited);
+
+ #endif /* #else #ifndef CONFIG_SMP */
+
+-#if !defined(CONFIG_RCU_FAST_NO_HZ)
++#if 1 /* !defined(CONFIG_RCU_FAST_NO_HZ) */
+
+ /*
+ * Check to see if any future RCU-related work will need to be done
+diff --git a/kernel/relay.c b/kernel/relay.c
+index 859ea5a..7e66dc4 100644
+--- a/kernel/relay.c
++++ b/kernel/relay.c
+@@ -336,6 +336,10 @@ static void wakeup_readers(unsigned long data)
+ {
+ struct rchan_buf *buf = (struct rchan_buf *)data;
+ wake_up_interruptible(&buf->read_wait);
++ /*
++ * Stupid polling for now:
++ */
++ mod_timer(&buf->timer, jiffies + 1);
+ }
+
+ /**
+@@ -353,6 +357,7 @@ static void __relay_reset(struct rchan_buf *buf, unsigned int init)
+ init_waitqueue_head(&buf->read_wait);
+ kref_init(&buf->kref);
+ setup_timer(&buf->timer, wakeup_readers, (unsigned long)buf);
++ mod_timer(&buf->timer, jiffies + 1);
+ } else
+ del_timer_sync(&buf->timer);
+
+@@ -733,15 +738,6 @@ size_t relay_switch_subbuf(struct rchan_buf *buf, size_t length)
+ else
+ buf->early_bytes += buf->chan->subbuf_size -
+ buf->padding[old_subbuf];
+- smp_mb();
+- if (waitqueue_active(&buf->read_wait))
+- /*
+- * Calling wake_up_interruptible() from here
+- * will deadlock if we happen to be logging
+- * from the scheduler (trying to re-grab
+- * rq->lock), so defer it.
+- */
+- mod_timer(&buf->timer, jiffies + 1);
+ }
+
+ old = buf->data;
+diff --git a/kernel/res_counter.c b/kernel/res_counter.c
+index 34683ef..21e9ec4 100644
+--- a/kernel/res_counter.c
++++ b/kernel/res_counter.c
+@@ -43,7 +43,7 @@ int res_counter_charge(struct res_counter *counter, unsigned long val,
+ struct res_counter *c, *u;
+
+ *limit_fail_at = NULL;
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ for (c = counter; c != NULL; c = c->parent) {
+ spin_lock(&c->lock);
+ ret = res_counter_charge_locked(c, val);
+@@ -62,7 +62,7 @@ undo:
+ spin_unlock(&u->lock);
+ }
+ done:
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ return ret;
+ }
+
+@@ -79,13 +79,13 @@ void res_counter_uncharge(struct res_counter *counter, unsigned long val)
+ unsigned long flags;
+ struct res_counter *c;
+
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ for (c = counter; c != NULL; c = c->parent) {
+ spin_lock(&c->lock);
+ res_counter_uncharge_locked(c, val);
+ spin_unlock(&c->lock);
+ }
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ }
+
+
+diff --git a/kernel/rt.c b/kernel/rt.c
+new file mode 100644
+index 0000000..092d6b3
+--- /dev/null
++++ b/kernel/rt.c
+@@ -0,0 +1,442 @@
++/*
++ * kernel/rt.c
++ *
++ * Real-Time Preemption Support
++ *
++ * started by Ingo Molnar:
++ *
++ * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar <mingo at redhat.com>
++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx at timesys.com>
++ *
++ * historic credit for proving that Linux spinlocks can be implemented via
++ * RT-aware mutexes goes to many people: The Pmutex project (Dirk Grambow
++ * and others) who prototyped it on 2.4 and did lots of comparative
++ * research and analysis; TimeSys, for proving that you can implement a
++ * fully preemptible kernel via the use of IRQ threading and mutexes;
++ * Bill Huey for persuasively arguing on lkml that the mutex model is the
++ * right one; and to MontaVista, who ported pmutexes to 2.6.
++ *
++ * This code is a from-scratch implementation and is not based on pmutexes,
++ * but the idea of converting spinlocks to mutexes is used here too.
++ *
++ * lock debugging, locking tree, deadlock detection:
++ *
++ * Copyright (C) 2004, LynuxWorks, Inc., Igor Manyilov, Bill Huey
++ * Released under the General Public License (GPL).
++ *
++ * Includes portions of the generic R/W semaphore implementation from:
++ *
++ * Copyright (c) 2001 David Howells (dhowells at redhat.com).
++ * - Derived partially from idea by Andrea Arcangeli <andrea at suse.de>
++ * - Derived also from comments by Linus
++ *
++ * Pending ownership of locks and ownership stealing:
++ *
++ * Copyright (C) 2005, Kihon Technologies Inc., Steven Rostedt
++ *
++ * (also by Steven Rostedt)
++ * - Converted single pi_lock to individual task locks.
++ *
++ * By Esben Nielsen:
++ * Doing priority inheritance with help of the scheduler.
++ *
++ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx at timesys.com>
++ * - major rework based on Esben Nielsens initial patch
++ * - replaced thread_info references by task_struct refs
++ * - removed task->pending_owner dependency
++ * - BKL drop/reacquire for semaphore style locks to avoid deadlocks
++ * in the scheduler return path as discussed with Steven Rostedt
++ *
++ * Copyright (C) 2006, Kihon Technologies Inc.
++ * Steven Rostedt <rostedt at goodmis.org>
++ * - debugged and patched Thomas Gleixner's rework.
++ * - added back the cmpxchg to the rework.
++ * - turned atomic require back on for SMP.
++ */
+
-+#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
-+ dentry = debugfs_create_dir(missed_timer_offsets_dir,
-+ latency_hist_root);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(missed_timer_offsets, i), &latency_hist_fops);
-+ my_hist = &per_cpu(missed_timer_offsets, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
-+
-+ sprintf(name, cpufmt_maxlatproc, i);
-+ mp = &per_cpu(missed_timer_offsets_maxlatproc, i);
-+ entry = debugfs_create_file(name, 0444, dentry, mp,
-+ &maxlatproc_fops);
-+ clear_maxlatprocdata(mp);
-+ }
-+ entry = debugfs_create_file("pid", 0644, dentry,
-+ (void *)&missed_timer_offsets_pid, &pid_fops);
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)MISSED_TIMER_OFFSETS, &latency_hist_reset_fops);
-+ entry = debugfs_create_file("missed_timer_offsets", 0644,
-+ enable_root, (void *)&missed_timer_offsets_enabled_data,
-+ &enable_fops);
-+#endif
++#include <linux/spinlock.h>
++#include <linux/rtmutex.h>
++#include <linux/sched.h>
++#include <linux/delay.h>
++#include <linux/module.h>
++#include <linux/kallsyms.h>
++#include <linux/syscalls.h>
++#include <linux/interrupt.h>
++#include <linux/plist.h>
++#include <linux/fs.h>
++#include <linux/futex.h>
++#include <linux/hrtimer.h>
+
-+#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
-+ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
-+ dentry = debugfs_create_dir(timerandwakeup_latency_hist_dir,
-+ latency_hist_root);
-+ for_each_possible_cpu(i) {
-+ sprintf(name, cpufmt, i);
-+ entry = debugfs_create_file(name, 0444, dentry,
-+ &per_cpu(timerandwakeup_latency_hist, i),
-+ &latency_hist_fops);
-+ my_hist = &per_cpu(timerandwakeup_latency_hist, i);
-+ atomic_set(&my_hist->hist_mode, 1);
-+ my_hist->min_lat = 0xFFFFFFFFUL;
++#include "rtmutex_common.h"
+
-+ sprintf(name, cpufmt_maxlatproc, i);
-+ mp = &per_cpu(timerandwakeup_maxlatproc, i);
-+ entry = debugfs_create_file(name, 0444, dentry, mp,
-+ &maxlatproc_fops);
-+ clear_maxlatprocdata(mp);
-+ }
-+ entry = debugfs_create_file("reset", 0644, dentry,
-+ (void *)TIMERANDWAKEUP_LATENCY, &latency_hist_reset_fops);
-+ entry = debugfs_create_file("timerandwakeup", 0644,
-+ enable_root, (void *)&timerandwakeup_enabled_data,
-+ &enable_fops);
++/*
++ * struct mutex functions
++ */
++void __mutex_do_init(struct mutex *mutex, const char *name,
++ struct lock_class_key *key)
++{
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ /*
++ * Make sure we are not reinitializing a held lock:
++ */
++ debug_check_no_locks_freed((void *)mutex, sizeof(*mutex));
++ lockdep_init_map(&mutex->dep_map, name, key, 0);
+#endif
-+ return 0;
++ mutex->lock.save_state = 0;
+}
++EXPORT_SYMBOL(__mutex_do_init);
+
-+__initcall(latency_hist_init);
-Index: linux-3.0/Documentation/hwlat_detector.txt
-===================================================================
---- /dev/null
-+++ linux-3.0/Documentation/hwlat_detector.txt
-@@ -0,0 +1,64 @@
-+Introduction:
-+-------------
++void __lockfunc _mutex_lock(struct mutex *lock)
++{
++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
++ rt_mutex_lock(&lock->lock);
++}
++EXPORT_SYMBOL(_mutex_lock);
+
-+The module hwlat_detector is a special purpose kernel module that is used to
-+detect large system latencies induced by the behavior of certain underlying
-+hardware or firmware, independent of Linux itself. The code was developed
-+originally to detect SMIs (System Management Interrupts) on x86 systems,
-+however there is nothing x86 specific about this patchset. It was
-+originally written for use by the "RT" patch since the Real Time
-+kernel is highly latency sensitive.
++int __lockfunc _mutex_lock_interruptible(struct mutex *lock)
++{
++ int ret;
+
-+SMIs are usually not serviced by the Linux kernel, which typically does not
-+even know that they are occuring. SMIs are instead are set up by BIOS code
-+and are serviced by BIOS code, usually for "critical" events such as
-+management of thermal sensors and fans. Sometimes though, SMIs are used for
-+other tasks and those tasks can spend an inordinate amount of time in the
-+handler (sometimes measured in milliseconds). Obviously this is a problem if
-+you are trying to keep event service latencies down in the microsecond range.
++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
++ ret = rt_mutex_lock_interruptible(&lock->lock, 0);
++ if (ret)
++ mutex_release(&lock->dep_map, 1, _RET_IP_);
++ return ret;
++}
++EXPORT_SYMBOL(_mutex_lock_interruptible);
+
-+The hardware latency detector works by hogging all of the cpus for configurable
-+amounts of time (by calling stop_machine()), polling the CPU Time Stamp Counter
-+for some period, then looking for gaps in the TSC data. Any gap indicates a
-+time when the polling was interrupted and since the machine is stopped and
-+interrupts turned off the only thing that could do that would be an SMI.
++int __lockfunc _mutex_lock_killable(struct mutex *lock)
++{
++ int ret;
+
-+Note that the SMI detector should *NEVER* be used in a production environment.
-+It is intended to be run manually to determine if the hardware platform has a
-+problem with long system firmware service routines.
++ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
++ ret = rt_mutex_lock_killable(&lock->lock, 0);
++ if (ret)
++ mutex_release(&lock->dep_map, 1, _RET_IP_);
++ return ret;
++}
++EXPORT_SYMBOL(_mutex_lock_killable);
+
-+Usage:
-+------
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass)
++{
++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_);
++ rt_mutex_lock(&lock->lock);
++}
++EXPORT_SYMBOL(_mutex_lock_nested);
+
-+Loading the module hwlat_detector passing the parameter "enabled=1" (or by
-+setting the "enable" entry in "hwlat_detector" debugfs toggled on) is the only
-+step required to start the hwlat_detector. It is possible to redefine the
-+threshold in microseconds (us) above which latency spikes will be taken
-+into account (parameter "threshold=").
++void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest)
++{
++ mutex_acquire_nest(&lock->dep_map, 0, 0, nest, _RET_IP_);
++ rt_mutex_lock(&lock->lock);
++}
++EXPORT_SYMBOL(_mutex_lock_nest_lock);
+
-+Example:
++int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass)
++{
++ int ret;
+
-+ # modprobe hwlat_detector enabled=1 threshold=100
++ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_);
++ ret = rt_mutex_lock_interruptible(&lock->lock, 0);
++ if (ret)
++ mutex_release(&lock->dep_map, 1, _RET_IP_);
++ return ret;
++}
++EXPORT_SYMBOL(_mutex_lock_interruptible_nested);
+
-+After the module is loaded, it creates a directory named "hwlat_detector" under
-+the debugfs mountpoint, "/debug/hwlat_detector" for this text. It is necessary
-+to have debugfs mounted, which might be on /sys/debug on your system.
++int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass)
++{
++ int ret;
+
-+The /debug/hwlat_detector interface contains the following files:
++ mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
++ ret = rt_mutex_lock_killable(&lock->lock, 0);
++ if (ret)
++ mutex_release(&lock->dep_map, 1, _RET_IP_);
++ return ret;
++}
++EXPORT_SYMBOL(_mutex_lock_killable_nested);
++#endif
+
-+count - number of latency spikes observed since last reset
-+enable - a global enable/disable toggle (0/1), resets count
-+max - maximum hardware latency actually observed (usecs)
-+sample - a pipe from which to read current raw sample data
-+ in the format <timestamp> <latency observed usecs>
-+ (can be opened O_NONBLOCK for a single sample)
-+threshold - minimum latency value to be considered (usecs)
-+width - time period to sample with CPUs held (usecs)
-+ must be less than the total window size (enforced)
-+window - total period of sampling, width being inside (usecs)
++int __lockfunc _mutex_trylock(struct mutex *lock)
++{
++ int ret = rt_mutex_trylock(&lock->lock);
+
-+By default we will set width to 500,000 and window to 1,000,000, meaning that
-+we will sample every 1,000,000 usecs (1s) for 500,000 usecs (0.5s). If we
-+observe any latencies that exceed the threshold (initially 100 usecs),
-+then we write to a global sample ring buffer of 8K samples, which is
-+consumed by reading from the "sample" (pipe) debugfs file interface.
-Index: linux-3.0/MAINTAINERS
-===================================================================
---- linux-3.0.orig/MAINTAINERS
-+++ linux-3.0/MAINTAINERS
-@@ -2911,6 +2911,15 @@ L: linuxppc-dev at lists.ozlabs.org
- S: Odd Fixes
- F: drivers/tty/hvc/
-
-+HARDWARE LATENCY DETECTOR
-+P: Jon Masters
-+M: jcm at jonmasters.org
-+W: http://www.kernel.org/pub/linux/kernel/people/jcm/hwlat_detector/
-+S: Supported
-+L: linux-kernel at vger.kernel.org
-+F: Documentation/hwlat_detector.txt
-+F: drivers/misc/hwlat_detector.c
++ if (ret)
++ mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_);
++
++ return ret;
++}
++EXPORT_SYMBOL(_mutex_trylock);
++
++void __lockfunc _mutex_unlock(struct mutex *lock)
++{
++ mutex_release(&lock->dep_map, 1, _RET_IP_);
++ rt_mutex_unlock(&lock->lock);
++}
++EXPORT_SYMBOL(_mutex_unlock);
+
- HARDWARE MONITORING
- M: Jean Delvare <khali at linux-fr.org>
- M: Guenter Roeck <guenter.roeck at ericsson.com>
-Index: linux-3.0/drivers/misc/Makefile
-===================================================================
---- linux-3.0.orig/drivers/misc/Makefile
-+++ linux-3.0/drivers/misc/Makefile
-@@ -46,3 +46,4 @@ obj-y += ti-st/
- obj-$(CONFIG_AB8500_PWM) += ab8500-pwm.o
- obj-y += lis3lv02d/
- obj-y += carma/
-+obj-$(CONFIG_HWLAT_DETECTOR) += hwlat_detector.o
-Index: linux-3.0/drivers/misc/hwlat_detector.c
-===================================================================
---- /dev/null
-+++ linux-3.0/drivers/misc/hwlat_detector.c
-@@ -0,0 +1,1212 @@
+/*
-+ * hwlat_detector.c - A simple Hardware Latency detector.
-+ *
-+ * Use this module to detect large system latencies induced by the behavior of
-+ * certain underlying system hardware or firmware, independent of Linux itself.
-+ * The code was developed originally to detect the presence of SMIs on Intel
-+ * and AMD systems, although there is no dependency upon x86 herein.
-+ *
-+ * The classical example usage of this module is in detecting the presence of
-+ * SMIs or System Management Interrupts on Intel and AMD systems. An SMI is a
-+ * somewhat special form of hardware interrupt spawned from earlier CPU debug
-+ * modes in which the (BIOS/EFI/etc.) firmware arranges for the South Bridge
-+ * LPC (or other device) to generate a special interrupt under certain
-+ * circumstances, for example, upon expiration of a special SMI timer device,
-+ * due to certain external thermal readings, on certain I/O address accesses,
-+ * and other situations. An SMI hits a special CPU pin, triggers a special
-+ * SMI mode (complete with special memory map), and the OS is unaware.
-+ *
-+ * Although certain hardware-inducing latencies are necessary (for example,
-+ * a modern system often requires an SMI handler for correct thermal control
-+ * and remote management) they can wreak havoc upon any OS-level performance
-+ * guarantees toward low-latency, especially when the OS is not even made
-+ * aware of the presence of these interrupts. For this reason, we need a
-+ * somewhat brute force mechanism to detect these interrupts. In this case,
-+ * we do it by hogging all of the CPU(s) for configurable timer intervals,
-+ * sampling the built-in CPU timer, looking for discontiguous readings.
-+ *
-+ * WARNING: This implementation necessarily introduces latencies. Therefore,
-+ * you should NEVER use this module in a production environment
-+ * requiring any kind of low-latency performance guarantee(s).
-+ *
-+ * Copyright (C) 2008-2009 Jon Masters, Red Hat, Inc. <jcm at redhat.com>
-+ *
-+ * Includes useful feedback from Clark Williams <clark at redhat.com>
-+ *
-+ * This file is licensed under the terms of the GNU General Public
-+ * License version 2. This program is licensed "as is" without any
-+ * warranty of any kind, whether express or implied.
++ * rwlock_t functions
+ */
++int __lockfunc rt_write_trylock(rwlock_t *rwlock)
++{
++ int ret = rt_mutex_trylock(&rwlock->lock);
++
++ migrate_disable();
++ if (ret)
++ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_);
++ else
++ migrate_enable();
++
++ return ret;
++}
++EXPORT_SYMBOL(rt_write_trylock);
++
++int __lockfunc rt_write_trylock_irqsave(rwlock_t *rwlock, unsigned long *flags)
++{
++ int ret;
++
++ *flags = 0;
++ migrate_disable();
++ ret = rt_write_trylock(rwlock);
++ if (!ret)
++ migrate_enable();
++ return ret;
++}
++EXPORT_SYMBOL(rt_write_trylock_irqsave);
++
++int __lockfunc rt_read_trylock(rwlock_t *rwlock)
++{
++ struct rt_mutex *lock = &rwlock->lock;
++ int ret = 1;
++
++ /*
++ * recursive read locks succeed when current owns the lock,
++ * but not when read_depth == 0 which means that the lock is
++ * write locked.
++ */
++ migrate_disable();
++ if (rt_mutex_owner(lock) != current)
++ ret = rt_mutex_trylock(lock);
++ else if (!rwlock->read_depth)
++ ret = 0;
++
++ if (ret) {
++ rwlock->read_depth++;
++ rwlock_acquire_read(&rwlock->dep_map, 0, 1, _RET_IP_);
++ } else
++ migrate_enable();
++
++ return ret;
++}
++EXPORT_SYMBOL(rt_read_trylock);
+
-+#include <linux/module.h>
-+#include <linux/init.h>
-+#include <linux/ring_buffer.h>
-+#include <linux/stop_machine.h>
-+#include <linux/time.h>
-+#include <linux/hrtimer.h>
-+#include <linux/kthread.h>
-+#include <linux/debugfs.h>
-+#include <linux/seq_file.h>
-+#include <linux/uaccess.h>
-+#include <linux/version.h>
-+#include <linux/delay.h>
-+#include <linux/slab.h>
++void __lockfunc rt_write_lock(rwlock_t *rwlock)
++{
++ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_);
++ __rt_spin_lock(&rwlock->lock);
++}
++EXPORT_SYMBOL(rt_write_lock);
+
-+#define BUF_SIZE_DEFAULT 262144UL /* 8K*(sizeof(entry)) */
-+#define BUF_FLAGS (RB_FL_OVERWRITE) /* no block on full */
-+#define U64STR_SIZE 22 /* 20 digits max */
++void __lockfunc rt_read_lock(rwlock_t *rwlock)
++{
++ struct rt_mutex *lock = &rwlock->lock;
+
-+#define VERSION "1.0.0"
-+#define BANNER "hwlat_detector: "
-+#define DRVNAME "hwlat_detector"
-+#define DEFAULT_SAMPLE_WINDOW 1000000 /* 1s */
-+#define DEFAULT_SAMPLE_WIDTH 500000 /* 0.5s */
-+#define DEFAULT_LAT_THRESHOLD 10 /* 10us */
++ rwlock_acquire_read(&rwlock->dep_map, 0, 0, _RET_IP_);
+
-+/* Module metadata */
++ /*
++ * recursive read locks succeed when current owns the lock
++ */
++ if (rt_mutex_owner(lock) != current)
++ __rt_spin_lock(lock);
++ rwlock->read_depth++;
++}
+
-+MODULE_LICENSE("GPL");
-+MODULE_AUTHOR("Jon Masters <jcm at redhat.com>");
-+MODULE_DESCRIPTION("A simple hardware latency detector");
-+MODULE_VERSION(VERSION);
++EXPORT_SYMBOL(rt_read_lock);
+
-+/* Module parameters */
++void __lockfunc rt_write_unlock(rwlock_t *rwlock)
++{
++ /* NOTE: we always pass in '1' for nested, for simplicity */
++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_);
++ __rt_spin_unlock(&rwlock->lock);
++}
++EXPORT_SYMBOL(rt_write_unlock);
+
-+static int debug;
-+static int enabled;
-+static int threshold;
++void __lockfunc rt_read_unlock(rwlock_t *rwlock)
++{
++ rwlock_release(&rwlock->dep_map, 1, _RET_IP_);
+
-+module_param(debug, int, 0); /* enable debug */
-+module_param(enabled, int, 0); /* enable detector */
-+module_param(threshold, int, 0); /* latency threshold */
++ /* Release the lock only when read_depth is down to 0 */
++ if (--rwlock->read_depth == 0)
++ __rt_spin_unlock(&rwlock->lock);
++}
++EXPORT_SYMBOL(rt_read_unlock);
+
-+/* Buffering and sampling */
++unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock)
++{
++ rt_write_lock(rwlock);
+
-+static struct ring_buffer *ring_buffer; /* sample buffer */
-+static DEFINE_MUTEX(ring_buffer_mutex); /* lock changes */
-+static unsigned long buf_size = BUF_SIZE_DEFAULT;
-+static struct task_struct *kthread; /* sampling thread */
++ return 0;
++}
++EXPORT_SYMBOL(rt_write_lock_irqsave);
+
-+/* DebugFS filesystem entries */
++unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock)
++{
++ rt_read_lock(rwlock);
+
-+static struct dentry *debug_dir; /* debugfs directory */
-+static struct dentry *debug_max; /* maximum TSC delta */
-+static struct dentry *debug_count; /* total detect count */
-+static struct dentry *debug_sample_width; /* sample width us */
-+static struct dentry *debug_sample_window; /* sample window us */
-+static struct dentry *debug_sample; /* raw samples us */
-+static struct dentry *debug_threshold; /* threshold us */
-+static struct dentry *debug_enable; /* enable/disable */
++ return 0;
++}
++EXPORT_SYMBOL(rt_read_lock_irqsave);
+
-+/* Individual samples and global state */
++void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key)
++{
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ /*
++ * Make sure we are not reinitializing a held lock:
++ */
++ debug_check_no_locks_freed((void *)rwlock, sizeof(*rwlock));
++ lockdep_init_map(&rwlock->dep_map, name, key, 0);
++#endif
++ rwlock->lock.save_state = 1;
++ rwlock->read_depth = 0;
++}
++EXPORT_SYMBOL(__rt_rwlock_init);
+
-+struct sample; /* latency sample */
-+struct data; /* Global state */
++/*
++ * rw_semaphores
++ */
+
-+/* Sampling functions */
-+static int __buffer_add_sample(struct sample *sample);
-+static struct sample *buffer_get_sample(struct sample *sample);
-+static int get_sample(void *unused);
++void rt_up_write(struct rw_semaphore *rwsem)
++{
++ rwsem_release(&rwsem->dep_map, 1, _RET_IP_);
++ rt_mutex_unlock(&rwsem->lock);
++}
++EXPORT_SYMBOL(rt_up_write);
+
-+/* Threading and state */
-+static int kthread_fn(void *unused);
-+static int start_kthread(void);
-+static int stop_kthread(void);
-+static void __reset_stats(void);
-+static int init_stats(void);
++void rt_up_read(struct rw_semaphore *rwsem)
++{
++ rwsem_release(&rwsem->dep_map, 1, _RET_IP_);
++ if (--rwsem->read_depth == 0)
++ rt_mutex_unlock(&rwsem->lock);
++}
++EXPORT_SYMBOL(rt_up_read);
+
-+/* Debugfs interface */
-+static ssize_t simple_data_read(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos, const u64 *entry);
-+static ssize_t simple_data_write(struct file *filp, const char __user *ubuf,
-+ size_t cnt, loff_t *ppos, u64 *entry);
-+static int debug_sample_fopen(struct inode *inode, struct file *filp);
-+static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos);
-+static int debug_sample_release(struct inode *inode, struct file *filp);
-+static int debug_enable_fopen(struct inode *inode, struct file *filp);
-+static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos);
-+static ssize_t debug_enable_fwrite(struct file *file,
-+ const char __user *user_buffer,
-+ size_t user_size, loff_t *offset);
++/*
++ * downgrade a write lock into a read lock
++ * - just wake up any readers at the front of the queue
++ */
++void rt_downgrade_write(struct rw_semaphore *rwsem)
++{
++ BUG_ON(rt_mutex_owner(&rwsem->lock) != current);
++ rwsem->read_depth = 1;
++}
++EXPORT_SYMBOL(rt_downgrade_write);
+
-+/* Initialization functions */
-+static int init_debugfs(void);
-+static void free_debugfs(void);
-+static int detector_init(void);
-+static void detector_exit(void);
++int rt_down_write_trylock(struct rw_semaphore *rwsem)
++{
++ int ret = rt_mutex_trylock(&rwsem->lock);
+
-+/* Individual latency samples are stored here when detected and packed into
-+ * the ring_buffer circular buffer, where they are overwritten when
-+ * more than buf_size/sizeof(sample) samples are received. */
-+struct sample {
-+ u64 seqnum; /* unique sequence */
-+ u64 duration; /* ktime delta */
-+ struct timespec timestamp; /* wall time */
-+ unsigned long lost;
-+};
++ if (ret)
++ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_);
++ return ret;
++}
++EXPORT_SYMBOL(rt_down_write_trylock);
+
-+/* keep the global state somewhere. Mostly used under stop_machine. */
-+static struct data {
++void rt_down_write(struct rw_semaphore *rwsem)
++{
++ rwsem_acquire(&rwsem->dep_map, 0, 0, _RET_IP_);
++ rt_mutex_lock(&rwsem->lock);
++}
++EXPORT_SYMBOL(rt_down_write);
+
-+ struct mutex lock; /* protect changes */
++void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass)
++{
++ rwsem_acquire(&rwsem->dep_map, subclass, 0, _RET_IP_);
++ rt_mutex_lock(&rwsem->lock);
++}
++EXPORT_SYMBOL(rt_down_write_nested);
+
-+ u64 count; /* total since reset */
-+ u64 max_sample; /* max hardware latency */
-+ u64 threshold; /* sample threshold level */
++int rt_down_read_trylock(struct rw_semaphore *rwsem)
++{
++ struct rt_mutex *lock = &rwsem->lock;
++ int ret = 1;
+
-+ u64 sample_window; /* total sampling window (on+off) */
-+ u64 sample_width; /* active sampling portion of window */
++ /*
++ * recursive read locks succeed when current owns the rwsem,
++ * but not when read_depth == 0 which means that the rwsem is
++ * write locked.
++ */
++ if (rt_mutex_owner(lock) != current)
++ ret = rt_mutex_trylock(&rwsem->lock);
++ else if (!rwsem->read_depth)
++ ret = 0;
+
-+ atomic_t sample_open; /* whether the sample file is open */
++ if (ret) {
++ rwsem->read_depth++;
++ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_);
++ }
++ return ret;
++}
++EXPORT_SYMBOL(rt_down_read_trylock);
+
-+ wait_queue_head_t wq; /* waitqeue for new sample values */
++static void __rt_down_read(struct rw_semaphore *rwsem, int subclass)
++{
++ struct rt_mutex *lock = &rwsem->lock;
+
-+} data;
++ rwsem_acquire_read(&rwsem->dep_map, subclass, 0, _RET_IP_);
++
++ if (rt_mutex_owner(lock) != current)
++ rt_mutex_lock(&rwsem->lock);
++ rwsem->read_depth++;
++}
++
++void rt_down_read(struct rw_semaphore *rwsem)
++{
++ __rt_down_read(rwsem, 0);
++}
++EXPORT_SYMBOL(rt_down_read);
++
++void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass)
++{
++ __rt_down_read(rwsem, subclass);
++}
++EXPORT_SYMBOL(rt_down_read_nested);
++
++void __rt_rwsem_init(struct rw_semaphore *rwsem, char *name,
++ struct lock_class_key *key)
++{
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ /*
++ * Make sure we are not reinitializing a held lock:
++ */
++ debug_check_no_locks_freed((void *)rwsem, sizeof(*rwsem));
++ lockdep_init_map(&rwsem->dep_map, name, key, 0);
++#endif
++ rwsem->read_depth = 0;
++ rwsem->lock.save_state = 0;
++}
++EXPORT_SYMBOL(__rt_rwsem_init);
+
+/**
-+ * __buffer_add_sample - add a new latency sample recording to the ring buffer
-+ * @sample: The new latency sample value
++ * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0
++ * @cnt: the atomic which we are to dec
++ * @lock: the mutex to return holding if we dec to 0
+ *
-+ * This receives a new latency sample and records it in a global ring buffer.
-+ * No additional locking is used in this case - suited for stop_machine use.
++ * return true and hold lock if we dec to 0, return false otherwise
+ */
-+static int __buffer_add_sample(struct sample *sample)
++int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock)
+{
-+ return ring_buffer_write(ring_buffer,
-+ sizeof(struct sample), sample);
++ /* dec if we can't possibly hit 0 */
++ if (atomic_add_unless(cnt, -1, 1))
++ return 0;
++ /* we might hit 0, so take the lock */
++ mutex_lock(lock);
++ if (!atomic_dec_and_test(cnt)) {
++ /* when we actually did the dec, we didn't hit 0 */
++ mutex_unlock(lock);
++ return 0;
++ }
++ /* we hit 0, and we hold the lock */
++ return 1;
+}
-+
-+/**
-+ * buffer_get_sample - remove a hardware latency sample from the ring buffer
-+ * @sample: Pre-allocated storage for the sample
++EXPORT_SYMBOL(atomic_dec_and_mutex_lock);
+diff --git a/kernel/rtmutex-debug.c b/kernel/rtmutex-debug.c
+index 3c7cbc2..a2e7e72 100644
+--- a/kernel/rtmutex-debug.c
++++ b/kernel/rtmutex-debug.c
+@@ -29,61 +29,6 @@
+
+ #include "rtmutex_common.h"
+
+-# define TRACE_WARN_ON(x) WARN_ON(x)
+-# define TRACE_BUG_ON(x) BUG_ON(x)
+-
+-# define TRACE_OFF() \
+-do { \
+- if (rt_trace_on) { \
+- rt_trace_on = 0; \
+- console_verbose(); \
+- if (raw_spin_is_locked(¤t->pi_lock)) \
+- raw_spin_unlock(¤t->pi_lock); \
+- } \
+-} while (0)
+-
+-# define TRACE_OFF_NOLOCK() \
+-do { \
+- if (rt_trace_on) { \
+- rt_trace_on = 0; \
+- console_verbose(); \
+- } \
+-} while (0)
+-
+-# define TRACE_BUG_LOCKED() \
+-do { \
+- TRACE_OFF(); \
+- BUG(); \
+-} while (0)
+-
+-# define TRACE_WARN_ON_LOCKED(c) \
+-do { \
+- if (unlikely(c)) { \
+- TRACE_OFF(); \
+- WARN_ON(1); \
+- } \
+-} while (0)
+-
+-# define TRACE_BUG_ON_LOCKED(c) \
+-do { \
+- if (unlikely(c)) \
+- TRACE_BUG_LOCKED(); \
+-} while (0)
+-
+-#ifdef CONFIG_SMP
+-# define SMP_TRACE_BUG_ON_LOCKED(c) TRACE_BUG_ON_LOCKED(c)
+-#else
+-# define SMP_TRACE_BUG_ON_LOCKED(c) do { } while (0)
+-#endif
+-
+-/*
+- * deadlock detection flag. We turn it off when we detect
+- * the first problem because we dont want to recurse back
+- * into the tracing code when doing error printk or
+- * executing a BUG():
+- */
+-static int rt_trace_on = 1;
+-
+ static void printk_task(struct task_struct *p)
+ {
+ if (p)
+@@ -111,8 +56,8 @@ static void printk_lock(struct rt_mutex *lock, int print_owner)
+
+ void rt_mutex_debug_task_free(struct task_struct *task)
+ {
+- WARN_ON(!plist_head_empty(&task->pi_waiters));
+- WARN_ON(task->pi_blocked_on);
++ DEBUG_LOCKS_WARN_ON(!plist_head_empty(&task->pi_waiters));
++ DEBUG_LOCKS_WARN_ON(task->pi_blocked_on);
+ }
+
+ /*
+@@ -125,7 +70,7 @@ void debug_rt_mutex_deadlock(int detect, struct rt_mutex_waiter *act_waiter,
+ {
+ struct task_struct *task;
+
+- if (!rt_trace_on || detect || !act_waiter)
++ if (!debug_locks || detect || !act_waiter)
+ return;
+
+ task = rt_mutex_owner(act_waiter->lock);
+@@ -139,7 +84,7 @@ void debug_rt_mutex_print_deadlock(struct rt_mutex_waiter *waiter)
+ {
+ struct task_struct *task;
+
+- if (!waiter->deadlock_lock || !rt_trace_on)
++ if (!waiter->deadlock_lock || !debug_locks)
+ return;
+
+ rcu_read_lock();
+@@ -149,7 +94,10 @@ void debug_rt_mutex_print_deadlock(struct rt_mutex_waiter *waiter)
+ return;
+ }
+
+- TRACE_OFF_NOLOCK();
++ if (!debug_locks_off()) {
++ rcu_read_unlock();
++ return;
++ }
+
+ printk("\n============================================\n");
+ printk( "[ BUG: circular locking deadlock detected! ]\n");
+@@ -180,7 +128,6 @@ void debug_rt_mutex_print_deadlock(struct rt_mutex_waiter *waiter)
+
+ printk("[ turning off deadlock detection."
+ "Please report this trace. ]\n\n");
+- local_irq_disable();
+ }
+
+ void debug_rt_mutex_lock(struct rt_mutex *lock)
+@@ -189,7 +136,7 @@ void debug_rt_mutex_lock(struct rt_mutex *lock)
+
+ void debug_rt_mutex_unlock(struct rt_mutex *lock)
+ {
+- TRACE_WARN_ON_LOCKED(rt_mutex_owner(lock) != current);
++ DEBUG_LOCKS_WARN_ON(rt_mutex_owner(lock) != current);
+ }
+
+ void
+@@ -199,7 +146,7 @@ debug_rt_mutex_proxy_lock(struct rt_mutex *lock, struct task_struct *powner)
+
+ void debug_rt_mutex_proxy_unlock(struct rt_mutex *lock)
+ {
+- TRACE_WARN_ON_LOCKED(!rt_mutex_owner(lock));
++ DEBUG_LOCKS_WARN_ON(!rt_mutex_owner(lock));
+ }
+
+ void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter)
+@@ -213,8 +160,8 @@ void debug_rt_mutex_init_waiter(struct rt_mutex_waiter *waiter)
+ void debug_rt_mutex_free_waiter(struct rt_mutex_waiter *waiter)
+ {
+ put_pid(waiter->deadlock_task_pid);
+- TRACE_WARN_ON(!plist_node_empty(&waiter->list_entry));
+- TRACE_WARN_ON(!plist_node_empty(&waiter->pi_list_entry));
++ DEBUG_LOCKS_WARN_ON(!plist_node_empty(&waiter->list_entry));
++ DEBUG_LOCKS_WARN_ON(!plist_node_empty(&waiter->pi_list_entry));
+ memset(waiter, 0x22, sizeof(*waiter));
+ }
+
+diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c
+index ab44911..15eaf30 100644
+--- a/kernel/rtmutex.c
++++ b/kernel/rtmutex.c
+@@ -8,6 +8,12 @@
+ * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt
+ * Copyright (C) 2006 Esben Nielsen
+ *
++ * Adaptive Spinlocks:
++ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich,
++ * and Peter Morreale,
++ * Adaptive Spinlocks simplification:
++ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt <srostedt at redhat.com>
+ *
-+ * This retrieves a hardware latency sample from the global circular buffer
-+ */
-+static struct sample *buffer_get_sample(struct sample *sample)
+ * See Documentation/rt-mutex-design.txt for details.
+ */
+ #include <linux/spinlock.h>
+@@ -67,6 +73,11 @@ static void fixup_rt_mutex_waiters(struct rt_mutex *lock)
+ clear_rt_mutex_waiters(lock);
+ }
+
++static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter)
+{
-+ struct ring_buffer_event *e = NULL;
-+ struct sample *s = NULL;
-+ unsigned int cpu = 0;
++ return waiter && waiter != PI_WAKEUP_INPROGRESS;
++}
+
-+ if (!sample)
-+ return NULL;
+ /*
+ * We can speed up the acquire/release, if the architecture
+ * supports cmpxchg and if there's no debugging state to be set up
+@@ -90,6 +101,12 @@ static inline void mark_rt_mutex_waiters(struct rt_mutex *lock)
+ }
+ #endif
+
++static inline void init_lists(struct rt_mutex *lock)
++{
++ if (unlikely(!lock->wait_list.node_list.prev))
++ plist_head_init_raw(&lock->wait_list, &lock->wait_lock);
++}
+
-+ mutex_lock(&ring_buffer_mutex);
-+ for_each_online_cpu(cpu) {
-+ e = ring_buffer_consume(ring_buffer, cpu, NULL, &sample->lost);
-+ if (e)
-+ break;
-+ }
+ /*
+ * Calculate task priority from the waiter list priority
+ *
+@@ -136,6 +153,14 @@ static void rt_mutex_adjust_prio(struct task_struct *task)
+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+ }
+
++static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter)
++{
++ if (waiter->savestate)
++ wake_up_lock_sleeper(waiter->task);
++ else
++ wake_up_process(waiter->task);
++}
+
-+ if (e) {
-+ s = ring_buffer_event_data(e);
-+ memcpy(sample, s, sizeof(struct sample));
-+ } else
-+ sample = NULL;
-+ mutex_unlock(&ring_buffer_mutex);
+ /*
+ * Max number of times we'll walk the boosting chain:
+ */
+@@ -196,7 +221,7 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task,
+ * reached or the state of the chain has changed while we
+ * dropped the locks.
+ */
+- if (!waiter)
++ if (!rt_mutex_real_waiter(waiter))
+ goto out_unlock_pi;
+
+ /*
+@@ -247,13 +272,15 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task,
+ /* Release the task */
+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+ if (!rt_mutex_owner(lock)) {
++ struct rt_mutex_waiter *lock_top_waiter;
++
+ /*
+ * If the requeue above changed the top waiter, then we need
+ * to wake the new top waiter up to try to get the lock.
+ */
+-
+- if (top_waiter != rt_mutex_top_waiter(lock))
+- wake_up_process(rt_mutex_top_waiter(lock)->task);
++ lock_top_waiter = rt_mutex_top_waiter(lock);
++ if (top_waiter != lock_top_waiter)
++ rt_mutex_wake_waiter(lock_top_waiter);
+ raw_spin_unlock(&lock->wait_lock);
+ goto out_put_task;
+ }
+@@ -298,6 +325,25 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task,
+ return ret;
+ }
+
+
-+ return sample;
-+}
++#define STEAL_NORMAL 0
++#define STEAL_LATERAL 1
+
-+/**
-+ * get_sample - sample the CPU TSC and look for likely hardware latencies
-+ * @unused: This is not used but is a part of the stop_machine API
-+ *
-+ * Used to repeatedly capture the CPU TSC (or similar), looking for potential
-+ * hardware-induced latency. Called under stop_machine, with data.lock held.
++/*
++ * Note that RT tasks are excluded from lateral-steals to prevent the
++ * introduction of an unbounded latency
+ */
-+static int get_sample(void *unused)
++static inline int lock_is_stealable(struct task_struct *task,
++ struct task_struct *pendowner, int mode)
+{
-+ ktime_t start, t1, t2;
-+ s64 diff, total = 0;
-+ u64 sample = 0;
-+ int ret = 1;
-+
-+ start = ktime_get(); /* start timestamp */
-+
-+ do {
++ if (mode == STEAL_NORMAL || rt_task(task)) {
++ if (task->prio >= pendowner->prio)
++ return 0;
++ } else if (task->prio > pendowner->prio)
++ return 0;
++ return 1;
++}
+
-+ t1 = ktime_get(); /* we'll look for a discontinuity */
-+ t2 = ktime_get();
+ /*
+ * Try to take an rt-mutex
+ *
+@@ -307,8 +353,9 @@ static int rt_mutex_adjust_prio_chain(struct task_struct *task,
+ * @task: the task which wants to acquire the lock
+ * @waiter: the waiter that is queued to the lock's wait list. (could be NULL)
+ */
+-static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
+- struct rt_mutex_waiter *waiter)
++static int
++__try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
++ struct rt_mutex_waiter *waiter, int mode)
+ {
+ /*
+ * We have to be careful here if the atomic speedups are
+@@ -341,12 +388,14 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
+ * 3) it is top waiter
+ */
+ if (rt_mutex_has_waiters(lock)) {
+- if (task->prio >= rt_mutex_top_waiter(lock)->list_entry.prio) {
+- if (!waiter || waiter != rt_mutex_top_waiter(lock))
+- return 0;
+- }
++ struct task_struct *pown = rt_mutex_top_waiter(lock)->task;
+
-+ total = ktime_to_us(ktime_sub(t2, start)); /* sample width */
-+ diff = ktime_to_us(ktime_sub(t2, t1)); /* current diff */
++ if (task != pown && !lock_is_stealable(task, pown, mode))
++ return 0;
+ }
+
++ /* We got the lock. */
+
-+ /* This shouldn't happen */
-+ if (diff < 0) {
-+ printk(KERN_ERR BANNER "time running backwards\n");
-+ goto out;
-+ }
+ if (waiter || rt_mutex_has_waiters(lock)) {
+ unsigned long flags;
+ struct rt_mutex_waiter *top;
+@@ -371,7 +420,6 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+ }
+
+- /* We got the lock. */
+ debug_rt_mutex_lock(lock);
+
+ rt_mutex_set_owner(lock, task);
+@@ -381,6 +429,13 @@ static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
+ return 1;
+ }
+
++static inline int
++try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
++ struct rt_mutex_waiter *waiter)
++{
++ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL);
++}
+
-+ if (diff > sample)
-+ sample = diff; /* only want highest value */
+ /*
+ * Task blocks on lock.
+ *
+@@ -399,6 +454,23 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
+ int chain_walk = 0, res;
+
+ raw_spin_lock_irqsave(&task->pi_lock, flags);
+
-+ } while (total <= data.sample_width);
++ /*
++ * In the case of futex requeue PI, this will be a proxy
++ * lock. The task will wake unaware that it is enqueueed on
++ * this lock. Avoid blocking on two locks and corrupting
++ * pi_blocked_on via the PI_WAKEUP_INPROGRESS
++ * flag. futex_wait_requeue_pi() sets this when it wakes up
++ * before requeue (due to a signal or timeout). Do not enqueue
++ * the task if PI_WAKEUP_INPROGRESS is set.
++ */
++ if (task != current && task->pi_blocked_on == PI_WAKEUP_INPROGRESS) {
++ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
++ return -EAGAIN;
++ }
+
-+ /* If we exceed the threshold value, we have found a hardware latency */
-+ if (sample > data.threshold) {
-+ struct sample s;
++ BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on));
+
-+ data.count++;
-+ s.seqnum = data.count;
-+ s.duration = sample;
-+ s.timestamp = CURRENT_TIME;
-+ __buffer_add_sample(&s);
+ __rt_mutex_adjust_prio(task);
+ waiter->task = task;
+ waiter->lock = lock;
+@@ -423,7 +495,7 @@ static int task_blocks_on_rt_mutex(struct rt_mutex *lock,
+ plist_add(&waiter->pi_list_entry, &owner->pi_waiters);
+
+ __rt_mutex_adjust_prio(owner);
+- if (owner->pi_blocked_on)
++ if (rt_mutex_real_waiter(owner->pi_blocked_on))
+ chain_walk = 1;
+ raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
+ }
+@@ -478,7 +550,7 @@ static void wakeup_next_waiter(struct rt_mutex *lock)
+
+ raw_spin_unlock_irqrestore(¤t->pi_lock, flags);
+
+- wake_up_process(waiter->task);
++ rt_mutex_wake_waiter(waiter);
+ }
+
+ /*
+@@ -517,7 +589,7 @@ static void remove_waiter(struct rt_mutex *lock,
+ }
+ __rt_mutex_adjust_prio(owner);
+
+- if (owner->pi_blocked_on)
++ if (rt_mutex_real_waiter(owner->pi_blocked_on))
+ chain_walk = 1;
+
+ raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
+@@ -551,23 +623,316 @@ void rt_mutex_adjust_pi(struct task_struct *task)
+ raw_spin_lock_irqsave(&task->pi_lock, flags);
+
+ waiter = task->pi_blocked_on;
+- if (!waiter || waiter->list_entry.prio == task->prio) {
++ if (!rt_mutex_real_waiter(waiter) ||
++ waiter->list_entry.prio == task->prio) {
+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+ return;
+ }
+
+- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+-
+ /* gets dropped in rt_mutex_adjust_prio_chain()! */
+ get_task_struct(task);
++ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+ rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task);
+ }
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++/*
++ * preemptible spin_lock functions:
++ */
++static inline void rt_spin_lock_fastlock(struct rt_mutex *lock,
++ void (*slowfn)(struct rt_mutex *lock))
++{
++ might_sleep();
+
-+ /* Keep a running maximum ever recorded hardware latency */
-+ if (sample > data.max_sample)
-+ data.max_sample = sample;
-+ }
++ if (likely(rt_mutex_cmpxchg(lock, NULL, current)))
++ rt_mutex_deadlock_account_lock(lock, current);
++ else
++ slowfn(lock);
++}
+
-+ ret = 0;
-+out:
-+ return ret;
++static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock,
++ void (*slowfn)(struct rt_mutex *lock))
++{
++ if (likely(rt_mutex_cmpxchg(lock, current, NULL)))
++ rt_mutex_deadlock_account_unlock(current);
++ else
++ slowfn(lock);
+}
+
++#ifdef CONFIG_SMP
+/*
-+ * kthread_fn - The CPU time sampling/hardware latency detection kernel thread
-+ * @unused: A required part of the kthread API.
-+ *
-+ * Used to periodically sample the CPU TSC via a call to get_sample. We
-+ * use stop_machine, whith does (intentionally) introduce latency since we
-+ * need to ensure nothing else might be running (and thus pre-empting).
-+ * Obviously this should never be used in production environments.
-+ *
-+ * stop_machine will schedule us typically only on CPU0 which is fine for
-+ * almost every real-world hardware latency situation - but we might later
-+ * generalize this if we find there are any actualy systems with alternate
-+ * SMI delivery or other non CPU0 hardware latencies.
++ * Note that owner is a speculative pointer and dereferencing relies
++ * on rcu_read_lock() and the check against the lock owner.
+ */
-+static int kthread_fn(void *unused)
++static int adaptive_wait(struct rt_mutex *lock,
++ struct task_struct *owner)
+{
-+ int err = 0;
-+ u64 interval = 0;
-+
-+ while (!kthread_should_stop()) {
-+
-+ mutex_lock(&data.lock);
++ int res = 0;
+
-+ err = stop_machine(get_sample, unused, 0);
-+ if (err) {
-+ /* Houston, we have a problem */
-+ mutex_unlock(&data.lock);
-+ goto err_out;
++ rcu_read_lock();
++ for (;;) {
++ if (owner != rt_mutex_owner(lock))
++ break;
++ /*
++ * Ensure that owner->on_cpu is dereferenced _after_
++ * checking the above to be valid.
++ */
++ barrier();
++ if (!owner->on_cpu) {
++ res = 1;
++ break;
+ }
-+
-+ wake_up(&data.wq); /* wake up reader(s) */
-+
-+ interval = data.sample_window - data.sample_width;
-+ do_div(interval, USEC_PER_MSEC); /* modifies interval value */
-+
-+ mutex_unlock(&data.lock);
-+
-+ if (msleep_interruptible(interval))
-+ goto out;
++ cpu_relax();
+ }
-+ goto out;
-+err_out:
-+ printk(KERN_ERR BANNER "could not call stop_machine, disabling\n");
-+ enabled = 0;
-+out:
-+ return err;
-+
++ rcu_read_unlock();
++ return res;
+}
-+
-+/**
-+ * start_kthread - Kick off the hardware latency sampling/detector kthread
-+ *
-+ * This starts a kernel thread that will sit and sample the CPU timestamp
-+ * counter (TSC or similar) and look for potential hardware latencies.
-+ */
-+static int start_kthread(void)
++#else
++static int adaptive_wait(struct rt_mutex *lock,
++ struct task_struct *orig_owner)
+{
-+ kthread = kthread_run(kthread_fn, NULL,
-+ DRVNAME);
-+ if (IS_ERR(kthread)) {
-+ printk(KERN_ERR BANNER "could not start sampling thread\n");
-+ enabled = 0;
-+ return -ENOMEM;
-+ }
-+
-+ return 0;
++ return 1;
+}
++#endif
+
-+/**
-+ * stop_kthread - Inform the hardware latency samping/detector kthread to stop
++# define pi_lock(lock) raw_spin_lock_irq(lock)
++# define pi_unlock(lock) raw_spin_unlock_irq(lock)
++
++/*
++ * Slow path lock function spin_lock style: this variant is very
++ * careful not to miss any non-lock wakeups.
+ *
-+ * This kicks the running hardware latency sampling/detector kernel thread and
-+ * tells it to stop sampling now. Use this on unload and at system shutdown.
++ * We store the current state under p->pi_lock in p->saved_state and
++ * the try_to_wake_up() code handles this accordingly.
+ */
-+static int stop_kthread(void)
++static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock)
+{
++ struct task_struct *lock_owner, *self = current;
++ struct rt_mutex_waiter waiter, *top_waiter;
+ int ret;
+
-+ ret = kthread_stop(kthread);
++ rt_mutex_init_waiter(&waiter, true);
+
-+ return ret;
-+}
++ raw_spin_lock(&lock->wait_lock);
++ init_lists(lock);
+
-+/**
-+ * __reset_stats - Reset statistics for the hardware latency detector
-+ *
-+ * We use data to store various statistics and global state. We call this
-+ * function in order to reset those when "enable" is toggled on or off, and
-+ * also at initialization. Should be called with data.lock held.
-+ */
-+static void __reset_stats(void)
-+{
-+ data.count = 0;
-+ data.max_sample = 0;
-+ ring_buffer_reset(ring_buffer); /* flush out old sample entries */
-+}
++ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) {
++ raw_spin_unlock(&lock->wait_lock);
++ return;
++ }
+
-+/**
-+ * init_stats - Setup global state statistics for the hardware latency detector
-+ *
-+ * We use data to store various statistics and global state. We also use
-+ * a global ring buffer (ring_buffer) to keep raw samples of detected hardware
-+ * induced system latencies. This function initializes these structures and
-+ * allocates the global ring buffer also.
-+ */
-+static int init_stats(void)
-+{
-+ int ret = -ENOMEM;
++ BUG_ON(rt_mutex_owner(lock) == self);
+
-+ mutex_init(&data.lock);
-+ init_waitqueue_head(&data.wq);
-+ atomic_set(&data.sample_open, 0);
++ /*
++ * We save whatever state the task is in and we'll restore it
++ * after acquiring the lock taking real wakeups into account
++ * as well. We are serialized via pi_lock against wakeups. See
++ * try_to_wake_up().
++ */
++ pi_lock(&self->pi_lock);
++ self->saved_state = self->state;
++ __set_current_state(TASK_UNINTERRUPTIBLE);
++ pi_unlock(&self->pi_lock);
+
-+ ring_buffer = ring_buffer_alloc(buf_size, BUF_FLAGS);
++ ret = task_blocks_on_rt_mutex(lock, &waiter, self, 0);
++ BUG_ON(ret);
+
-+ if (WARN(!ring_buffer, KERN_ERR BANNER
-+ "failed to allocate ring buffer!\n"))
-+ goto out;
++ for (;;) {
++ /* Try to acquire the lock again. */
++ if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL))
++ break;
+
-+ __reset_stats();
-+ data.threshold = DEFAULT_LAT_THRESHOLD; /* threshold us */
-+ data.sample_window = DEFAULT_SAMPLE_WINDOW; /* window us */
-+ data.sample_width = DEFAULT_SAMPLE_WIDTH; /* width us */
++ top_waiter = rt_mutex_top_waiter(lock);
++ lock_owner = rt_mutex_owner(lock);
+
-+ ret = 0;
++ raw_spin_unlock(&lock->wait_lock);
+
-+out:
-+ return ret;
++ debug_rt_mutex_print_deadlock(&waiter);
+
-+}
++ if (top_waiter != &waiter || adaptive_wait(lock, lock_owner))
++ schedule_rt_mutex(lock);
+
-+/*
-+ * simple_data_read - Wrapper read function for global state debugfs entries
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ * @entry: The entry to read from
-+ *
-+ * This function provides a generic read implementation for the global state
-+ * "data" structure debugfs filesystem entries. It would be nice to use
-+ * simple_attr_read directly, but we need to make sure that the data.lock
-+ * spinlock is held during the actual read (even though we likely won't ever
-+ * actually race here as the updater runs under a stop_machine context).
-+ */
-+static ssize_t simple_data_read(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos, const u64 *entry)
-+{
-+ char buf[U64STR_SIZE];
-+ u64 val = 0;
-+ int len = 0;
++ raw_spin_lock(&lock->wait_lock);
+
-+ memset(buf, 0, sizeof(buf));
++ pi_lock(&self->pi_lock);
++ __set_current_state(TASK_UNINTERRUPTIBLE);
++ pi_unlock(&self->pi_lock);
++ }
+
-+ if (!entry)
-+ return -EFAULT;
++ /*
++ * Restore the task state to current->saved_state. We set it
++ * to the original state above and the try_to_wake_up() code
++ * has possibly updated it when a real (non-rtmutex) wakeup
++ * happened while we were blocked. Clear saved_state so
++ * try_to_wakeup() does not get confused.
++ */
++ pi_lock(&self->pi_lock);
++ __set_current_state(self->saved_state);
++ self->saved_state = TASK_RUNNING;
++ pi_unlock(&self->pi_lock);
+
-+ mutex_lock(&data.lock);
-+ val = *entry;
-+ mutex_unlock(&data.lock);
++ /*
++ * try_to_take_rt_mutex() sets the waiter bit
++ * unconditionally. We might have to fix that up:
++ */
++ fixup_rt_mutex_waiters(lock);
+
-+ len = snprintf(buf, sizeof(buf), "%llu\n", (unsigned long long)val);
++ BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock));
++ BUG_ON(!plist_node_empty(&waiter.list_entry));
+
-+ return simple_read_from_buffer(ubuf, cnt, ppos, buf, len);
++ raw_spin_unlock(&lock->wait_lock);
+
++ debug_rt_mutex_free_waiter(&waiter);
+}
+
+/*
-+ * simple_data_write - Wrapper write function for global state debugfs entries
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to write value from
-+ * @cnt: The maximum number of bytes to write
-+ * @ppos: The current "file" position
-+ * @entry: The entry to write to
-+ *
-+ * This function provides a generic write implementation for the global state
-+ * "data" structure debugfs filesystem entries. It would be nice to use
-+ * simple_attr_write directly, but we need to make sure that the data.lock
-+ * spinlock is held during the actual write (even though we likely won't ever
-+ * actually race here as the updater runs under a stop_machine context).
++ * Slow path to release a rt_mutex spin_lock style
+ */
-+static ssize_t simple_data_write(struct file *filp, const char __user *ubuf,
-+ size_t cnt, loff_t *ppos, u64 *entry)
++static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock)
+{
-+ char buf[U64STR_SIZE];
-+ int csize = min(cnt, sizeof(buf));
-+ u64 val = 0;
-+ int err = 0;
++ raw_spin_lock(&lock->wait_lock);
+
-+ memset(buf, '\0', sizeof(buf));
-+ if (copy_from_user(buf, ubuf, csize))
-+ return -EFAULT;
++ debug_rt_mutex_unlock(lock);
+
-+ buf[U64STR_SIZE-1] = '\0'; /* just in case */
-+ err = strict_strtoull(buf, 10, &val);
-+ if (err)
-+ return -EINVAL;
++ rt_mutex_deadlock_account_unlock(current);
+
-+ mutex_lock(&data.lock);
-+ *entry = val;
-+ mutex_unlock(&data.lock);
++ if (!rt_mutex_has_waiters(lock)) {
++ lock->owner = NULL;
++ raw_spin_unlock(&lock->wait_lock);
++ return;
++ }
+
-+ return csize;
++ wakeup_next_waiter(lock);
++
++ raw_spin_unlock(&lock->wait_lock);
++
++ /* Undo pi boosting.when necessary */
++ rt_mutex_adjust_prio(current);
+}
+
-+/**
-+ * debug_count_fopen - Open function for "count" debugfs entry
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function provides an open implementation for the "count" debugfs
-+ * interface to the hardware latency detector.
-+ */
-+static int debug_count_fopen(struct inode *inode, struct file *filp)
++void __lockfunc rt_spin_lock(spinlock_t *lock)
+{
-+ return 0;
++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
++ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_);
+}
++EXPORT_SYMBOL(rt_spin_lock);
+
-+/**
-+ * debug_count_fread - Read function for "count" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "count" debugfs
-+ * interface to the hardware latency detector. Can be used to read the
-+ * number of latency readings exceeding the configured threshold since
-+ * the detector was last reset (e.g. by writing a zero into "count").
-+ */
-+static ssize_t debug_count_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
++void __lockfunc __rt_spin_lock(struct rt_mutex *lock)
+{
-+ return simple_data_read(filp, ubuf, cnt, ppos, &data.count);
++ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock);
+}
++EXPORT_SYMBOL(__rt_spin_lock);
+
-+/**
-+ * debug_count_fwrite - Write function for "count" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "count" debugfs
-+ * interface to the hardware latency detector. Can be used to write a
-+ * desired value, especially to zero the total count.
-+ */
-+static ssize_t debug_count_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass)
+{
-+ return simple_data_write(filp, ubuf, cnt, ppos, &data.count);
++ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
++ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
+}
++EXPORT_SYMBOL(rt_spin_lock_nested);
++#endif
+
-+/**
-+ * debug_enable_fopen - Dummy open function for "enable" debugfs interface
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function provides an open implementation for the "enable" debugfs
-+ * interface to the hardware latency detector.
-+ */
-+static int debug_enable_fopen(struct inode *inode, struct file *filp)
++void __lockfunc rt_spin_unlock(spinlock_t *lock)
+{
-+ return 0;
++ /* NOTE: we always pass in '1' for nested, for simplicity */
++ spin_release(&lock->dep_map, 1, _RET_IP_);
++ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock);
+}
++EXPORT_SYMBOL(rt_spin_unlock);
+
-+/**
-+ * debug_enable_fread - Read function for "enable" debugfs interface
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "enable" debugfs
-+ * interface to the hardware latency detector. Can be used to determine
-+ * whether the detector is currently enabled ("0\n" or "1\n" returned).
-+ */
-+static ssize_t debug_enable_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
++void __lockfunc __rt_spin_unlock(struct rt_mutex *lock)
+{
-+ char buf[4];
-+
-+ if ((cnt < sizeof(buf)) || (*ppos))
-+ return 0;
-+
-+ buf[0] = enabled ? '1' : '0';
-+ buf[1] = '\n';
-+ buf[2] = '\0';
-+ if (copy_to_user(ubuf, buf, strlen(buf)))
-+ return -EFAULT;
-+ return *ppos = strlen(buf);
++ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock);
+}
++EXPORT_SYMBOL(__rt_spin_unlock);
+
-+/**
-+ * debug_enable_fwrite - Write function for "enable" debugfs interface
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "enable" debugfs
-+ * interface to the hardware latency detector. Can be used to enable or
-+ * disable the detector, which will have the side-effect of possibly
-+ * also resetting the global stats and kicking off the measuring
-+ * kthread (on an enable) or the converse (upon a disable).
++/*
++ * Wait for the lock to get unlocked: instead of polling for an unlock
++ * (like raw spinlocks do), we lock and unlock, to force the kernel to
++ * schedule if there's contention:
+ */
-+static ssize_t debug_enable_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
++void __lockfunc rt_spin_unlock_wait(spinlock_t *lock)
+{
-+ char buf[4];
-+ int csize = min(cnt, sizeof(buf));
-+ long val = 0;
-+ int err = 0;
-+
-+ memset(buf, '\0', sizeof(buf));
-+ if (copy_from_user(buf, ubuf, csize))
-+ return -EFAULT;
++ spin_lock(lock);
++ spin_unlock(lock);
++}
++EXPORT_SYMBOL(rt_spin_unlock_wait);
+
-+ buf[sizeof(buf)-1] = '\0'; /* just in case */
-+ err = strict_strtoul(buf, 10, &val);
-+ if (0 != err)
-+ return -EINVAL;
++int __lockfunc rt_spin_trylock(spinlock_t *lock)
++{
++ int ret = rt_mutex_trylock(&lock->lock);
+
-+ if (val) {
-+ if (enabled)
-+ goto unlock;
-+ enabled = 1;
-+ __reset_stats();
-+ if (start_kthread())
-+ return -EFAULT;
-+ } else {
-+ if (!enabled)
-+ goto unlock;
-+ enabled = 0;
-+ err = stop_kthread();
-+ if (err) {
-+ printk(KERN_ERR BANNER "cannot stop kthread\n");
-+ return -EFAULT;
-+ }
-+ wake_up(&data.wq); /* reader(s) should return */
-+ }
-+unlock:
-+ return csize;
++ if (ret)
++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
++ return ret;
+}
++EXPORT_SYMBOL(rt_spin_trylock);
+
-+/**
-+ * debug_max_fopen - Open function for "max" debugfs entry
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function provides an open implementation for the "max" debugfs
-+ * interface to the hardware latency detector.
-+ */
-+static int debug_max_fopen(struct inode *inode, struct file *filp)
++int __lockfunc rt_spin_trylock_bh(spinlock_t *lock)
+{
-+ return 0;
++ int ret;
++
++ local_bh_disable();
++ ret = rt_mutex_trylock(&lock->lock);
++ if (ret) {
++ migrate_disable();
++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
++ } else
++ local_bh_enable();
++ return ret;
+}
++EXPORT_SYMBOL(rt_spin_trylock_bh);
+
-+/**
-+ * debug_max_fread - Read function for "max" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "max" debugfs
-+ * interface to the hardware latency detector. Can be used to determine
-+ * the maximum latency value observed since it was last reset.
-+ */
-+static ssize_t debug_max_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
++int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags)
+{
-+ return simple_data_read(filp, ubuf, cnt, ppos, &data.max_sample);
++ int ret;
++
++ *flags = 0;
++ migrate_disable();
++ ret = rt_mutex_trylock(&lock->lock);
++ if (ret)
++ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
++ else
++ migrate_enable();
++ return ret;
+}
++EXPORT_SYMBOL(rt_spin_trylock_irqsave);
+
-+/**
-+ * debug_max_fwrite - Write function for "max" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "max" debugfs
-+ * interface to the hardware latency detector. Can be used to reset the
-+ * maximum or set it to some other desired value - if, then, subsequent
-+ * measurements exceed this value, the maximum will be updated.
-+ */
-+static ssize_t debug_max_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
++int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock)
+{
-+ return simple_data_write(filp, ubuf, cnt, ppos, &data.max_sample);
++ /* Subtract 1 from counter unless that drops it to 0 (ie. it was 1) */
++ if (atomic_add_unless(atomic, -1, 1))
++ return 0;
++ migrate_disable();
++ rt_spin_lock(lock);
++ if (atomic_dec_and_test(atomic))
++ return 1;
++ rt_spin_unlock(lock);
++ migrate_enable();
++ return 0;
+}
++EXPORT_SYMBOL(atomic_dec_and_spin_lock);
+
-+
-+/**
-+ * debug_sample_fopen - An open function for "sample" debugfs interface
-+ * @inode: The in-kernel inode representation of this debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function handles opening the "sample" file within the hardware
-+ * latency detector debugfs directory interface. This file is used to read
-+ * raw samples from the global ring_buffer and allows the user to see a
-+ * running latency history. Can be opened blocking or non-blocking,
-+ * affecting whether it behaves as a buffer read pipe, or does not.
-+ * Implements simple locking to prevent multiple simultaneous use.
-+ */
-+static int debug_sample_fopen(struct inode *inode, struct file *filp)
++void
++__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key)
+{
-+ if (!atomic_add_unless(&data.sample_open, 1, 1))
-+ return -EBUSY;
-+ else
-+ return 0;
++#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ /*
++ * Make sure we are not reinitializing a held lock:
++ */
++ debug_check_no_locks_freed((void *)lock, sizeof(*lock));
++ lockdep_init_map(&lock->dep_map, name, key, 0);
++#endif
+}
++EXPORT_SYMBOL(__rt_spin_lock_init);
+
-+/**
-+ * debug_sample_fread - A read function for "sample" debugfs interface
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that will contain the samples read
-+ * @cnt: The maximum bytes to read from the debugfs "file"
-+ * @ppos: The current position in the debugfs "file"
++#endif /* PREEMPT_RT_FULL */
++
+ /**
+ * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop
+ * @lock: the rt_mutex to take
+ * @state: the state the task should block in (TASK_INTERRUPTIBLE
+- * or TASK_UNINTERRUPTIBLE)
++ * or TASK_UNINTERRUPTIBLE)
+ * @timeout: the pre-initialized and started timer, or NULL for none
+ * @waiter: the pre-initialized rt_mutex_waiter
+ *
+@@ -623,9 +988,10 @@ rt_mutex_slowlock(struct rt_mutex *lock, int state,
+ struct rt_mutex_waiter waiter;
+ int ret = 0;
+
+- debug_rt_mutex_init_waiter(&waiter);
++ rt_mutex_init_waiter(&waiter, false);
+
+ raw_spin_lock(&lock->wait_lock);
++ init_lists(lock);
+
+ /* Try to acquire the lock again: */
+ if (try_to_take_rt_mutex(lock, current, NULL)) {
+@@ -678,6 +1044,7 @@ rt_mutex_slowtrylock(struct rt_mutex *lock)
+ int ret = 0;
+
+ raw_spin_lock(&lock->wait_lock);
++ init_lists(lock);
+
+ if (likely(rt_mutex_owner(lock) != current)) {
+
+@@ -791,12 +1158,12 @@ EXPORT_SYMBOL_GPL(rt_mutex_lock);
+ /**
+ * rt_mutex_lock_interruptible - lock a rt_mutex interruptible
+ *
+- * @lock: the rt_mutex to be locked
++ * @lock: the rt_mutex to be locked
+ * @detect_deadlock: deadlock detection on/off
+ *
+ * Returns:
+- * 0 on success
+- * -EINTR when interrupted by a signal
++ * 0 on success
++ * -EINTR when interrupted by a signal
+ * -EDEADLK when the lock would deadlock (when deadlock detection is on)
+ */
+ int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock,
+@@ -810,17 +1177,38 @@ int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock,
+ EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible);
+
+ /**
++ * rt_mutex_lock_killable - lock a rt_mutex killable
+ *
-+ * This function handles reading from the "sample" file within the hardware
-+ * latency detector debugfs directory interface. This file is used to read
-+ * raw samples from the global ring_buffer and allows the user to see a
-+ * running latency history. By default this will block pending a new
-+ * value written into the sample buffer, unless there are already a
-+ * number of value(s) waiting in the buffer, or the sample file was
-+ * previously opened in a non-blocking mode of operation.
++ * @lock: the rt_mutex to be locked
++ * @detect_deadlock: deadlock detection on/off
++ *
++ * Returns:
++ * 0 on success
++ * -EINTR when interrupted by a signal
++ * -EDEADLK when the lock would deadlock (when deadlock detection is on)
+ */
-+static ssize_t debug_sample_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
++int __sched rt_mutex_lock_killable(struct rt_mutex *lock,
++ int detect_deadlock)
+{
-+ int len = 0;
-+ char buf[64];
-+ struct sample *sample = NULL;
-+
-+ if (!enabled)
-+ return 0;
-+
-+ sample = kzalloc(sizeof(struct sample), GFP_KERNEL);
-+ if (!sample)
-+ return -ENOMEM;
-+
-+ while (!buffer_get_sample(sample)) {
-+
-+ DEFINE_WAIT(wait);
++ might_sleep();
+
-+ if (filp->f_flags & O_NONBLOCK) {
-+ len = -EAGAIN;
-+ goto out;
-+ }
++ return rt_mutex_fastlock(lock, TASK_KILLABLE,
++ detect_deadlock, rt_mutex_slowlock);
++}
++EXPORT_SYMBOL_GPL(rt_mutex_lock_killable);
+
-+ prepare_to_wait(&data.wq, &wait, TASK_INTERRUPTIBLE);
-+ schedule();
-+ finish_wait(&data.wq, &wait);
++/**
+ * rt_mutex_timed_lock - lock a rt_mutex interruptible
+ * the timeout structure is provided
+ * by the caller
+ *
+- * @lock: the rt_mutex to be locked
++ * @lock: the rt_mutex to be locked
+ * @timeout: timeout structure or NULL (no timeout)
+ * @detect_deadlock: deadlock detection on/off
+ *
+ * Returns:
+- * 0 on success
+- * -EINTR when interrupted by a signal
++ * 0 on success
++ * -EINTR when interrupted by a signal
+ * -ETIMEDOUT when the timeout expired
+ * -EDEADLK when the lock would deadlock (when deadlock detection is on)
+ */
+@@ -889,7 +1277,6 @@ EXPORT_SYMBOL_GPL(rt_mutex_destroy);
+ void __rt_mutex_init(struct rt_mutex *lock, const char *name)
+ {
+ lock->owner = NULL;
+- raw_spin_lock_init(&lock->wait_lock);
+ plist_head_init_raw(&lock->wait_list, &lock->wait_lock);
+
+ debug_rt_mutex_init(lock, name);
+@@ -909,7 +1296,7 @@ EXPORT_SYMBOL_GPL(__rt_mutex_init);
+ void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
+ struct task_struct *proxy_owner)
+ {
+- __rt_mutex_init(lock, NULL);
++ rt_mutex_init(lock);
+ debug_rt_mutex_proxy_lock(lock, proxy_owner);
+ rt_mutex_set_owner(lock, proxy_owner);
+ rt_mutex_deadlock_account_lock(lock, proxy_owner);
+diff --git a/kernel/rtmutex_common.h b/kernel/rtmutex_common.h
+index 53a66c8..a688a29 100644
+--- a/kernel/rtmutex_common.h
++++ b/kernel/rtmutex_common.h
+@@ -49,6 +49,7 @@ struct rt_mutex_waiter {
+ struct plist_node pi_list_entry;
+ struct task_struct *task;
+ struct rt_mutex *lock;
++ bool savestate;
+ #ifdef CONFIG_DEBUG_RT_MUTEXES
+ unsigned long ip;
+ struct pid *deadlock_task_pid;
+@@ -103,6 +104,8 @@ static inline struct task_struct *rt_mutex_owner(struct rt_mutex *lock)
+ /*
+ * PI-futex support (proxy locking functions, etc.):
+ */
++#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1)
+
-+ if (signal_pending(current)) {
-+ len = -EINTR;
-+ goto out;
-+ }
+ extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock);
+ extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
+ struct task_struct *proxy_owner);
+@@ -123,4 +126,12 @@ extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
+ # include "rtmutex.h"
+ #endif
+
++static inline void
++rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate)
++{
++ debug_rt_mutex_init_waiter(waiter);
++ waiter->task = NULL;
++ waiter->savestate = savestate;
++}
+
-+ if (!enabled) { /* enable was toggled */
-+ len = 0;
-+ goto out;
+ #endif
+diff --git a/kernel/rwsem.c b/kernel/rwsem.c
+index cae050b..6c6e7fa 100644
+--- a/kernel/rwsem.c
++++ b/kernel/rwsem.c
+@@ -16,20 +16,19 @@
+ /*
+ * lock for reading
+ */
+-void __sched down_read(struct rw_semaphore *sem)
++void __sched anon_down_read(struct rw_anon_semaphore *sem)
+ {
+ might_sleep();
+ rwsem_acquire_read(&sem->dep_map, 0, 0, _RET_IP_);
+
+ LOCK_CONTENDED(sem, __down_read_trylock, __down_read);
+ }
+-
+-EXPORT_SYMBOL(down_read);
++EXPORT_SYMBOL(anon_down_read);
+
+ /*
+ * trylock for reading -- returns 1 if successful, 0 if contention
+ */
+-int down_read_trylock(struct rw_semaphore *sem)
++int anon_down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ int ret = __down_read_trylock(sem);
+
+@@ -37,26 +36,24 @@ int down_read_trylock(struct rw_semaphore *sem)
+ rwsem_acquire_read(&sem->dep_map, 0, 1, _RET_IP_);
+ return ret;
+ }
+-
+-EXPORT_SYMBOL(down_read_trylock);
++EXPORT_SYMBOL(anon_down_read_trylock);
+
+ /*
+ * lock for writing
+ */
+-void __sched down_write(struct rw_semaphore *sem)
++void __sched anon_down_write(struct rw_anon_semaphore *sem)
+ {
+ might_sleep();
+ rwsem_acquire(&sem->dep_map, 0, 0, _RET_IP_);
+
+ LOCK_CONTENDED(sem, __down_write_trylock, __down_write);
+ }
+-
+-EXPORT_SYMBOL(down_write);
++EXPORT_SYMBOL(anon_down_write);
+
+ /*
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+-int down_write_trylock(struct rw_semaphore *sem)
++int anon_down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ int ret = __down_write_trylock(sem);
+
+@@ -64,37 +61,34 @@ int down_write_trylock(struct rw_semaphore *sem)
+ rwsem_acquire(&sem->dep_map, 0, 1, _RET_IP_);
+ return ret;
+ }
+-
+-EXPORT_SYMBOL(down_write_trylock);
++EXPORT_SYMBOL(anon_down_write_trylock);
+
+ /*
+ * release a read lock
+ */
+-void up_read(struct rw_semaphore *sem)
++void anon_up_read(struct rw_anon_semaphore *sem)
+ {
+ rwsem_release(&sem->dep_map, 1, _RET_IP_);
+
+ __up_read(sem);
+ }
+-
+-EXPORT_SYMBOL(up_read);
++EXPORT_SYMBOL(anon_up_read);
+
+ /*
+ * release a write lock
+ */
+-void up_write(struct rw_semaphore *sem)
++void anon_up_write(struct rw_anon_semaphore *sem)
+ {
+ rwsem_release(&sem->dep_map, 1, _RET_IP_);
+
+ __up_write(sem);
+ }
+-
+-EXPORT_SYMBOL(up_write);
++EXPORT_SYMBOL(anon_up_write);
+
+ /*
+ * downgrade write lock to read lock
+ */
+-void downgrade_write(struct rw_semaphore *sem)
++void anon_downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ /*
+ * lockdep: a downgraded write will live on as a write
+@@ -102,46 +96,41 @@ void downgrade_write(struct rw_semaphore *sem)
+ */
+ __downgrade_write(sem);
+ }
+-
+-EXPORT_SYMBOL(downgrade_write);
++EXPORT_SYMBOL(anon_downgrade_write);
+
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+
+-void down_read_nested(struct rw_semaphore *sem, int subclass)
++void anon_down_read_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ might_sleep();
+ rwsem_acquire_read(&sem->dep_map, subclass, 0, _RET_IP_);
+
+ LOCK_CONTENDED(sem, __down_read_trylock, __down_read);
+ }
++EXPORT_SYMBOL(anon_down_read_nested);
+
+-EXPORT_SYMBOL(down_read_nested);
+-
+-void down_read_non_owner(struct rw_semaphore *sem)
++void anon_down_read_non_owner(struct rw_anon_semaphore *sem)
+ {
+ might_sleep();
+
+ __down_read(sem);
+ }
++EXPORT_SYMBOL(anon_down_read_non_owner);
+
+-EXPORT_SYMBOL(down_read_non_owner);
+-
+-void down_write_nested(struct rw_semaphore *sem, int subclass)
++void anon_down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ might_sleep();
+ rwsem_acquire(&sem->dep_map, subclass, 0, _RET_IP_);
+
+ LOCK_CONTENDED(sem, __down_write_trylock, __down_write);
+ }
++EXPORT_SYMBOL(anon_down_write_nested);
+
+-EXPORT_SYMBOL(down_write_nested);
+-
+-void up_read_non_owner(struct rw_semaphore *sem)
++void anon_up_read_non_owner(struct rw_anon_semaphore *sem)
+ {
+ __up_read(sem);
+ }
+-
+-EXPORT_SYMBOL(up_read_non_owner);
++EXPORT_SYMBOL(anon_up_read_non_owner);
+
+ #endif
+
+diff --git a/kernel/sched.c b/kernel/sched.c
+index 063d7a4..640f740 100644
+--- a/kernel/sched.c
++++ b/kernel/sched.c
+@@ -71,6 +71,7 @@
+ #include <linux/ctype.h>
+ #include <linux/ftrace.h>
+ #include <linux/slab.h>
++#include <linux/init_task.h>
+
+ #include <asm/tlb.h>
+ #include <asm/irq_regs.h>
+@@ -185,6 +186,7 @@ void init_rt_bandwidth(struct rt_bandwidth *rt_b, u64 period, u64 runtime)
+
+ hrtimer_init(&rt_b->rt_period_timer,
+ CLOCK_MONOTONIC, HRTIMER_MODE_REL);
++ rt_b->rt_period_timer.irqsafe = 1;
+ rt_b->rt_period_timer.function = sched_rt_period_timer;
+ }
+
+@@ -800,7 +802,11 @@ late_initcall(sched_init_debug);
+ * Number of tasks to iterate in a single balance run.
+ * Limited because this is done with IRQs disabled.
+ */
++#ifndef CONFIG_PREEMPT_RT_FULL
+ const_debug unsigned int sysctl_sched_nr_migrate = 32;
++#else
++const_debug unsigned int sysctl_sched_nr_migrate = 8;
++#endif
+
+ /*
+ * period over which we average the RT time consumption, measured
+@@ -1136,6 +1142,7 @@ static void init_rq_hrtick(struct rq *rq)
+
+ hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ rq->hrtick_timer.function = hrtick;
++ rq->hrtick_timer.irqsafe = 1;
+ }
+ #else /* CONFIG_SCHED_HRTICK */
+ static inline void hrtick_clear(struct rq *rq)
+@@ -2378,11 +2385,11 @@ static int select_fallback_rq(int cpu, struct task_struct *p)
+
+ /* Look for allowed, online CPU in same node. */
+ for_each_cpu_and(dest_cpu, nodemask, cpu_active_mask)
+- if (cpumask_test_cpu(dest_cpu, &p->cpus_allowed))
++ if (cpumask_test_cpu(dest_cpu, tsk_cpus_allowed(p)))
+ return dest_cpu;
+
+ /* Any allowed, online CPU? */
+- dest_cpu = cpumask_any_and(&p->cpus_allowed, cpu_active_mask);
++ dest_cpu = cpumask_any_and(tsk_cpus_allowed(p), cpu_active_mask);
+ if (dest_cpu < nr_cpu_ids)
+ return dest_cpu;
+
+@@ -2397,7 +2404,12 @@ static int select_fallback_rq(int cpu, struct task_struct *p)
+ printk(KERN_INFO "process %d (%s) no longer affine to cpu%d\n",
+ task_pid_nr(p), p->comm, cpu);
+ }
+-
++ /*
++ * Clear PF_THREAD_BOUND, otherwise we wreckage
++ * migrate_disable/enable. See optimization for
++ * PF_THREAD_BOUND tasks there.
++ */
++ p->flags &= ~PF_THREAD_BOUND;
+ return dest_cpu;
+ }
+
+@@ -2419,7 +2431,7 @@ int select_task_rq(struct task_struct *p, int sd_flags, int wake_flags)
+ * [ this allows ->select_task() to simply return task_cpu(p) and
+ * not worry about this generic constraint ]
+ */
+- if (unlikely(!cpumask_test_cpu(cpu, &p->cpus_allowed) ||
++ if (unlikely(!cpumask_test_cpu(cpu, tsk_cpus_allowed(p)) ||
+ !cpu_online(cpu)))
+ cpu = select_fallback_rq(task_cpu(p), p);
+
+@@ -2477,10 +2489,6 @@ static void ttwu_activate(struct rq *rq, struct task_struct *p, int en_flags)
+ {
+ activate_task(rq, p, en_flags);
+ p->on_rq = 1;
+-
+- /* if a worker is waking up, notify workqueue */
+- if (p->flags & PF_WQ_WORKER)
+- wq_worker_waking_up(p, cpu_of(rq));
+ }
+
+ /*
+@@ -2678,8 +2686,25 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags)
+
+ smp_wmb();
+ raw_spin_lock_irqsave(&p->pi_lock, flags);
+- if (!(p->state & state))
++ if (!(p->state & state)) {
++ /*
++ * The task might be running due to a spinlock sleeper
++ * wakeup. Check the saved state and set it to running
++ * if the wakeup condition is true.
++ */
++ if (!(wake_flags & WF_LOCK_SLEEPER)) {
++ if (p->saved_state & state)
++ p->saved_state = TASK_RUNNING;
+ }
+ goto out;
+ }
+
-+ len = snprintf(buf, sizeof(buf), "%010lu.%010lu\t%llu\n",
-+ sample->timestamp.tv_sec,
-+ sample->timestamp.tv_nsec,
-+ sample->duration);
-+
-+
-+ /* handling partial reads is more trouble than it's worth */
-+ if (len > cnt)
-+ goto out;
-+
-+ if (copy_to_user(ubuf, buf, len))
-+ len = -EFAULT;
-+
-+out:
-+ kfree(sample);
-+ return len;
-+}
-+
-+/**
-+ * debug_sample_release - Release function for "sample" debugfs interface
-+ * @inode: The in-kernel inode represenation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function completes the close of the debugfs interface "sample" file.
-+ * Frees the sample_open "lock" so that other users may open the interface.
-+ */
-+static int debug_sample_release(struct inode *inode, struct file *filp)
-+{
-+ atomic_dec(&data.sample_open);
-+
-+ return 0;
-+}
-+
-+/**
-+ * debug_threshold_fopen - Open function for "threshold" debugfs entry
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function provides an open implementation for the "threshold" debugfs
-+ * interface to the hardware latency detector.
-+ */
-+static int debug_threshold_fopen(struct inode *inode, struct file *filp)
-+{
-+ return 0;
-+}
-+
-+/**
-+ * debug_threshold_fread - Read function for "threshold" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "threshold" debugfs
-+ * interface to the hardware latency detector. It can be used to determine
-+ * the current threshold level at which a latency will be recorded in the
-+ * global ring buffer, typically on the order of 10us.
-+ */
-+static ssize_t debug_threshold_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
-+{
-+ return simple_data_read(filp, ubuf, cnt, ppos, &data.threshold);
-+}
-+
-+/**
-+ * debug_threshold_fwrite - Write function for "threshold" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "threshold" debugfs
-+ * interface to the hardware latency detector. It can be used to configure
-+ * the threshold level at which any subsequently detected latencies will
-+ * be recorded into the global ring buffer.
-+ */
-+static ssize_t debug_threshold_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
-+{
-+ int ret;
-+
-+ ret = simple_data_write(filp, ubuf, cnt, ppos, &data.threshold);
-+
-+ if (enabled)
-+ wake_up_process(kthread);
-+
-+ return ret;
-+}
-+
++ /*
++ * If this is a regular wakeup, then we can unconditionally
++ * clear the saved state of a "lock sleeper".
++ */
++ if (!(wake_flags & WF_LOCK_SLEEPER))
++ p->saved_state = TASK_RUNNING;
+
+ success = 1; /* we're going to change ->state */
+ cpu = task_cpu(p);
+@@ -2735,40 +2760,6 @@ out:
+ }
+
+ /**
+- * try_to_wake_up_local - try to wake up a local task with rq lock held
+- * @p: the thread to be awakened
+- *
+- * Put @p on the run-queue if it's not already there. The caller must
+- * ensure that this_rq() is locked, @p is bound to this_rq() and not
+- * the current task.
+- */
+-static void try_to_wake_up_local(struct task_struct *p)
+-{
+- struct rq *rq = task_rq(p);
+-
+- BUG_ON(rq != this_rq());
+- BUG_ON(p == current);
+- lockdep_assert_held(&rq->lock);
+-
+- if (!raw_spin_trylock(&p->pi_lock)) {
+- raw_spin_unlock(&rq->lock);
+- raw_spin_lock(&p->pi_lock);
+- raw_spin_lock(&rq->lock);
+- }
+-
+- if (!(p->state & TASK_NORMAL))
+- goto out;
+-
+- if (!p->on_rq)
+- ttwu_activate(rq, p, ENQUEUE_WAKEUP);
+-
+- ttwu_do_wakeup(rq, p, 0);
+- ttwu_stat(p, smp_processor_id(), 0);
+-out:
+- raw_spin_unlock(&p->pi_lock);
+-}
+-
+-/**
+ * wake_up_process - Wake up a specific process
+ * @p: The process to be woken up.
+ *
+@@ -2785,6 +2776,18 @@ int wake_up_process(struct task_struct *p)
+ }
+ EXPORT_SYMBOL(wake_up_process);
+
+/**
-+ * debug_width_fopen - Open function for "width" debugfs entry
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
++ * wake_up_lock_sleeper - Wake up a specific process blocked on a "sleeping lock"
++ * @p: The process to be woken up.
+ *
-+ * This function provides an open implementation for the "width" debugfs
-+ * interface to the hardware latency detector.
++ * Same as wake_up_process() above, but wake_flags=WF_LOCK_SLEEPER to indicate
++ * the nature of the wakeup.
+ */
-+static int debug_width_fopen(struct inode *inode, struct file *filp)
++int wake_up_lock_sleeper(struct task_struct *p)
+{
-+ return 0;
++ return try_to_wake_up(p, TASK_ALL, WF_LOCK_SLEEPER);
+}
+
-+/**
-+ * debug_width_fread - Read function for "width" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "width" debugfs
-+ * interface to the hardware latency detector. It can be used to determine
-+ * for how many us of the total window us we will actively sample for any
-+ * hardware-induced latecy periods. Obviously, it is not possible to
-+ * sample constantly and have the system respond to a sample reader, or,
-+ * worse, without having the system appear to have gone out to lunch.
-+ */
-+static ssize_t debug_width_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
-+{
-+ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_width);
-+}
+ int wake_up_state(struct task_struct *p, unsigned int state)
+ {
+ return try_to_wake_up(p, state, 0);
+@@ -3060,8 +3063,12 @@ static void finish_task_switch(struct rq *rq, struct task_struct *prev)
+ finish_lock_switch(rq, prev);
+
+ fire_sched_in_preempt_notifiers(current);
++ /*
++ * We use mmdrop_delayed() here so we don't have to do the
++ * full __mmdrop() when we are the last user.
++ */
+ if (mm)
+- mmdrop(mm);
++ mmdrop_delayed(mm);
+ if (unlikely(prev_state == TASK_DEAD)) {
+ /*
+ * Remove function-return probe instances associated with this
+@@ -4182,6 +4189,126 @@ static inline void schedule_debug(struct task_struct *prev)
+ schedstat_inc(this_rq(), sched_count);
+ }
+
++#if defined(CONFIG_PREEMPT_RT_FULL) && defined(CONFIG_SMP)
++#define MIGRATE_DISABLE_SET_AFFIN (1<<30) /* Can't make a negative */
++#define migrate_disabled_updated(p) ((p)->migrate_disable & MIGRATE_DISABLE_SET_AFFIN)
++#define migrate_disable_count(p) ((p)->migrate_disable & ~MIGRATE_DISABLE_SET_AFFIN)
+
-+/**
-+ * debug_width_fwrite - Write function for "width" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "width" debugfs
-+ * interface to the hardware latency detector. It can be used to configure
-+ * for how many us of the total window us we will actively sample for any
-+ * hardware-induced latency periods. Obviously, it is not possible to
-+ * sample constantly and have the system respond to a sample reader, or,
-+ * worse, without having the system appear to have gone out to lunch. It
-+ * is enforced that width is less that the total window size.
-+ */
-+static ssize_t debug_width_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
++static inline void update_migrate_disable(struct task_struct *p)
+{
-+ char buf[U64STR_SIZE];
-+ int csize = min(cnt, sizeof(buf));
-+ u64 val = 0;
-+ int err = 0;
-+
-+ memset(buf, '\0', sizeof(buf));
-+ if (copy_from_user(buf, ubuf, csize))
-+ return -EFAULT;
-+
-+ buf[U64STR_SIZE-1] = '\0'; /* just in case */
-+ err = strict_strtoull(buf, 10, &val);
-+ if (0 != err)
-+ return -EINVAL;
++ const struct cpumask *mask;
+
-+ mutex_lock(&data.lock);
-+ if (val < data.sample_window)
-+ data.sample_width = val;
-+ else {
-+ mutex_unlock(&data.lock);
-+ return -EINVAL;
-+ }
-+ mutex_unlock(&data.lock);
++ if (likely(!p->migrate_disable))
++ return;
+
-+ if (enabled)
-+ wake_up_process(kthread);
++ /* Did we already update affinity? */
++ if (unlikely(migrate_disabled_updated(p)))
++ return;
+
-+ return csize;
-+}
++ /*
++ * Since this is always current we can get away with only locking
++ * rq->lock, the ->cpus_allowed value can normally only be changed
++ * while holding both p->pi_lock and rq->lock, but seeing that this
++ * is current, we cannot actually be waking up, so all code that
++ * relies on serialization against p->pi_lock is out of scope.
++ *
++ * Having rq->lock serializes us against things like
++ * set_cpus_allowed_ptr() that can still happen concurrently.
++ */
++ mask = tsk_cpus_allowed(p);
+
-+/**
-+ * debug_window_fopen - Open function for "window" debugfs entry
-+ * @inode: The in-kernel inode representation of the debugfs "file"
-+ * @filp: The active open file structure for the debugfs "file"
-+ *
-+ * This function provides an open implementation for the "window" debugfs
-+ * interface to the hardware latency detector. The window is the total time
-+ * in us that will be considered one sample period. Conceptually, windows
-+ * occur back-to-back and contain a sample width period during which
-+ * actual sampling occurs.
-+ */
-+static int debug_window_fopen(struct inode *inode, struct file *filp)
-+{
-+ return 0;
-+}
++ if (p->sched_class->set_cpus_allowed)
++ p->sched_class->set_cpus_allowed(p, mask);
++ p->rt.nr_cpus_allowed = cpumask_weight(mask);
+
-+/**
-+ * debug_window_fread - Read function for "window" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The userspace provided buffer to read value into
-+ * @cnt: The maximum number of bytes to read
-+ * @ppos: The current "file" position
-+ *
-+ * This function provides a read implementation for the "window" debugfs
-+ * interface to the hardware latency detector. The window is the total time
-+ * in us that will be considered one sample period. Conceptually, windows
-+ * occur back-to-back and contain a sample width period during which
-+ * actual sampling occurs. Can be used to read the total window size.
-+ */
-+static ssize_t debug_window_fread(struct file *filp, char __user *ubuf,
-+ size_t cnt, loff_t *ppos)
-+{
-+ return simple_data_read(filp, ubuf, cnt, ppos, &data.sample_window);
++ /* Let migrate_enable know to fix things back up */
++ p->migrate_disable |= MIGRATE_DISABLE_SET_AFFIN;
+}
+
-+/**
-+ * debug_window_fwrite - Write function for "window" debugfs entry
-+ * @filp: The active open file structure for the debugfs "file"
-+ * @ubuf: The user buffer that contains the value to write
-+ * @cnt: The maximum number of bytes to write to "file"
-+ * @ppos: The current position in the debugfs "file"
-+ *
-+ * This function provides a write implementation for the "window" debufds
-+ * interface to the hardware latency detetector. The window is the total time
-+ * in us that will be considered one sample period. Conceptually, windows
-+ * occur back-to-back and contain a sample width period during which
-+ * actual sampling occurs. Can be used to write a new total window size. It
-+ * is enfoced that any value written must be greater than the sample width
-+ * size, or an error results.
-+ */
-+static ssize_t debug_window_fwrite(struct file *filp,
-+ const char __user *ubuf,
-+ size_t cnt,
-+ loff_t *ppos)
++void migrate_disable(void)
+{
-+ char buf[U64STR_SIZE];
-+ int csize = min(cnt, sizeof(buf));
-+ u64 val = 0;
-+ int err = 0;
-+
-+ memset(buf, '\0', sizeof(buf));
-+ if (copy_from_user(buf, ubuf, csize))
-+ return -EFAULT;
-+
-+ buf[U64STR_SIZE-1] = '\0'; /* just in case */
-+ err = strict_strtoull(buf, 10, &val);
-+ if (0 != err)
-+ return -EINVAL;
++ struct task_struct *p = current;
+
-+ mutex_lock(&data.lock);
-+ if (data.sample_width < val)
-+ data.sample_window = val;
-+ else {
-+ mutex_unlock(&data.lock);
-+ return -EINVAL;
++ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
++#ifdef CONFIG_SCHED_DEBUG
++ p->migrate_disable_atomic++;
++#endif
++ return;
+ }
-+ mutex_unlock(&data.lock);
-+
-+ return csize;
-+}
-+
-+/*
-+ * Function pointers for the "count" debugfs file operations
-+ */
-+static const struct file_operations count_fops = {
-+ .open = debug_count_fopen,
-+ .read = debug_count_fread,
-+ .write = debug_count_fwrite,
-+ .owner = THIS_MODULE,
-+};
+
-+/*
-+ * Function pointers for the "enable" debugfs file operations
-+ */
-+static const struct file_operations enable_fops = {
-+ .open = debug_enable_fopen,
-+ .read = debug_enable_fread,
-+ .write = debug_enable_fwrite,
-+ .owner = THIS_MODULE,
-+};
++#ifdef CONFIG_SCHED_DEBUG
++ WARN_ON_ONCE(p->migrate_disable_atomic);
++#endif
+
-+/*
-+ * Function pointers for the "max" debugfs file operations
-+ */
-+static const struct file_operations max_fops = {
-+ .open = debug_max_fopen,
-+ .read = debug_max_fread,
-+ .write = debug_max_fwrite,
-+ .owner = THIS_MODULE,
-+};
++ preempt_disable();
++ if (p->migrate_disable) {
++ p->migrate_disable++;
++ preempt_enable();
++ return;
++ }
+
-+/*
-+ * Function pointers for the "sample" debugfs file operations
-+ */
-+static const struct file_operations sample_fops = {
-+ .open = debug_sample_fopen,
-+ .read = debug_sample_fread,
-+ .release = debug_sample_release,
-+ .owner = THIS_MODULE,
-+};
++ pin_current_cpu();
++ p->migrate_disable = 1;
++ preempt_enable();
++}
++EXPORT_SYMBOL_GPL(migrate_disable);
+
-+/*
-+ * Function pointers for the "threshold" debugfs file operations
-+ */
-+static const struct file_operations threshold_fops = {
-+ .open = debug_threshold_fopen,
-+ .read = debug_threshold_fread,
-+ .write = debug_threshold_fwrite,
-+ .owner = THIS_MODULE,
-+};
++void migrate_enable(void)
++{
++ struct task_struct *p = current;
++ const struct cpumask *mask;
++ unsigned long flags;
++ struct rq *rq;
+
-+/*
-+ * Function pointers for the "width" debugfs file operations
-+ */
-+static const struct file_operations width_fops = {
-+ .open = debug_width_fopen,
-+ .read = debug_width_fread,
-+ .write = debug_width_fwrite,
-+ .owner = THIS_MODULE,
-+};
++ if (in_atomic() || p->flags & PF_THREAD_BOUND) {
++#ifdef CONFIG_SCHED_DEBUG
++ p->migrate_disable_atomic--;
++#endif
++ return;
++ }
+
-+/*
-+ * Function pointers for the "window" debugfs file operations
-+ */
-+static const struct file_operations window_fops = {
-+ .open = debug_window_fopen,
-+ .read = debug_window_fread,
-+ .write = debug_window_fwrite,
-+ .owner = THIS_MODULE,
-+};
++#ifdef CONFIG_SCHED_DEBUG
++ WARN_ON_ONCE(p->migrate_disable_atomic);
++#endif
++ WARN_ON_ONCE(p->migrate_disable <= 0);
+
-+/**
-+ * init_debugfs - A function to initialize the debugfs interface files
-+ *
-+ * This function creates entries in debugfs for "hwlat_detector", including
-+ * files to read values from the detector, current samples, and the
-+ * maximum sample that has been captured since the hardware latency
-+ * dectector was started.
-+ */
-+static int init_debugfs(void)
-+{
-+ int ret = -ENOMEM;
++ preempt_disable();
++ if (migrate_disable_count(p) > 1) {
++ p->migrate_disable--;
++ preempt_enable();
++ return;
++ }
+
-+ debug_dir = debugfs_create_dir(DRVNAME, NULL);
-+ if (!debug_dir)
-+ goto err_debug_dir;
++ if (unlikely(migrate_disabled_updated(p))) {
++ /*
++ * Undo whatever update_migrate_disable() did, also see there
++ * about locking.
++ */
++ rq = this_rq();
++ raw_spin_lock_irqsave(&rq->lock, flags);
+
-+ debug_sample = debugfs_create_file("sample", 0444,
-+ debug_dir, NULL,
-+ &sample_fops);
-+ if (!debug_sample)
-+ goto err_sample;
++ /*
++ * Clearing migrate_disable causes tsk_cpus_allowed to
++ * show the tasks original cpu affinity.
++ */
++ p->migrate_disable = 0;
++ mask = tsk_cpus_allowed(p);
++ if (p->sched_class->set_cpus_allowed)
++ p->sched_class->set_cpus_allowed(p, mask);
++ p->rt.nr_cpus_allowed = cpumask_weight(mask);
++ raw_spin_unlock_irqrestore(&rq->lock, flags);
++ } else
++ p->migrate_disable = 0;
+
-+ debug_count = debugfs_create_file("count", 0444,
-+ debug_dir, NULL,
-+ &count_fops);
-+ if (!debug_count)
-+ goto err_count;
++ unpin_current_cpu();
++ preempt_enable();
++}
++EXPORT_SYMBOL_GPL(migrate_enable);
++#else
++static inline void update_migrate_disable(struct task_struct *p) { }
++#define migrate_disabled_updated(p) 0
++#endif
+
-+ debug_max = debugfs_create_file("max", 0444,
-+ debug_dir, NULL,
-+ &max_fops);
-+ if (!debug_max)
-+ goto err_max;
+ static void put_prev_task(struct rq *rq, struct task_struct *prev)
+ {
+ if (prev->on_rq || rq->skip_clock_update < 0)
+@@ -4241,6 +4368,8 @@ need_resched:
+
+ raw_spin_lock_irq(&rq->lock);
+
++ update_migrate_disable(prev);
+
-+ debug_sample_window = debugfs_create_file("window", 0644,
-+ debug_dir, NULL,
-+ &window_fops);
-+ if (!debug_sample_window)
-+ goto err_window;
+ switch_count = &prev->nivcsw;
+ if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) {
+ if (unlikely(signal_pending_state(prev->state, prev))) {
+@@ -4248,19 +4377,6 @@ need_resched:
+ } else {
+ deactivate_task(rq, prev, DEQUEUE_SLEEP);
+ prev->on_rq = 0;
+-
+- /*
+- * If a worker went to sleep, notify and ask workqueue
+- * whether it wants to wake up a task to maintain
+- * concurrency.
+- */
+- if (prev->flags & PF_WQ_WORKER) {
+- struct task_struct *to_wakeup;
+-
+- to_wakeup = wq_worker_sleeping(prev, cpu);
+- if (to_wakeup)
+- try_to_wake_up_local(to_wakeup);
+- }
+ }
+ switch_count = &prev->nvcsw;
+ }
+@@ -4294,15 +4410,23 @@ need_resched:
+
+ post_schedule(rq);
+
+- preempt_enable_no_resched();
++ __preempt_enable_no_resched();
+ if (need_resched())
+ goto need_resched;
+ }
+
+ static inline void sched_submit_work(struct task_struct *tsk)
+ {
+- if (!tsk->state)
++ if (!tsk->state || tsk_is_pi_blocked(tsk))
+ return;
+
-+ debug_sample_width = debugfs_create_file("width", 0644,
-+ debug_dir, NULL,
-+ &width_fops);
-+ if (!debug_sample_width)
-+ goto err_width;
++ /*
++ * If a worker went to sleep, notify and ask workqueue whether
++ * it wants to wake up a task to maintain concurrency.
++ */
++ if (tsk->flags & PF_WQ_WORKER)
++ wq_worker_sleeping(tsk);
+
-+ debug_threshold = debugfs_create_file("threshold", 0644,
-+ debug_dir, NULL,
-+ &threshold_fops);
-+ if (!debug_threshold)
-+ goto err_threshold;
+ /*
+ * If we are going to sleep and we have plugged IO queued,
+ * make sure to submit it to avoid deadlocks.
+@@ -4311,15 +4435,37 @@ static inline void sched_submit_work(struct task_struct *tsk)
+ blk_schedule_flush_plug(tsk);
+ }
+
++static inline void sched_update_worker(struct task_struct *tsk)
++{
++ if (tsk_is_pi_blocked(tsk))
++ return;
+
-+ debug_enable = debugfs_create_file("enable", 0644,
-+ debug_dir, &enabled,
-+ &enable_fops);
-+ if (!debug_enable)
-+ goto err_enable;
++ if (tsk->flags & PF_WQ_WORKER)
++ wq_worker_running(tsk);
++}
+
-+ else {
-+ ret = 0;
-+ goto out;
+ asmlinkage void __sched schedule(void)
+ {
+ struct task_struct *tsk = current;
+
+ sched_submit_work(tsk);
+ __schedule();
++ sched_update_worker(tsk);
+ }
+ EXPORT_SYMBOL(schedule);
+
++/**
++ * schedule_preempt_disabled - called with preemption disabled
++ *
++ * Returns with preemption disabled. Note: preempt_count must be 1
++ */
++void __sched schedule_preempt_disabled(void)
++{
++ __preempt_enable_no_resched();
++ schedule();
++ preempt_disable();
++}
++
+ #ifdef CONFIG_MUTEX_SPIN_ON_OWNER
+
+ static inline bool owner_running(struct mutex *lock, struct task_struct *owner)
+@@ -4391,7 +4537,16 @@ asmlinkage void __sched notrace preempt_schedule(void)
+
+ do {
+ add_preempt_count_notrace(PREEMPT_ACTIVE);
++ /*
++ * The add/subtract must not be traced by the function
++ * tracer. But we still want to account for the
++ * preempt off latency tracer. Since the _notrace versions
++ * of add/subtract skip the accounting for latency tracer
++ * we must force it manually.
++ */
++ start_critical_timings();
+ __schedule();
++ stop_critical_timings();
+ sub_preempt_count_notrace(PREEMPT_ACTIVE);
+
+ /*
+@@ -4814,9 +4969,8 @@ long __sched sleep_on_timeout(wait_queue_head_t *q, long timeout)
+ EXPORT_SYMBOL(sleep_on_timeout);
+
+ #ifdef CONFIG_RT_MUTEXES
+-
+ /*
+- * rt_mutex_setprio - set the current priority of a task
++ * task_setprio - set the current priority of a task
+ * @p: task
+ * @prio: prio value (kernel-internal form)
+ *
+@@ -4825,7 +4979,7 @@ EXPORT_SYMBOL(sleep_on_timeout);
+ *
+ * Used by the rt_mutex code to implement priority inheritance logic.
+ */
+-void rt_mutex_setprio(struct task_struct *p, int prio)
++void task_setprio(struct task_struct *p, int prio)
+ {
+ int oldprio, on_rq, running;
+ struct rq *rq;
+@@ -4835,6 +4989,24 @@ void rt_mutex_setprio(struct task_struct *p, int prio)
+
+ rq = __task_rq_lock(p);
+
++ /*
++ * Idle task boosting is a nono in general. There is one
++ * exception, when PREEMPT_RT and NOHZ is active:
++ *
++ * The idle task calls get_next_timer_interrupt() and holds
++ * the timer wheel base->lock on the CPU and another CPU wants
++ * to access the timer (probably to cancel it). We can safely
++ * ignore the boosting request, as the idle CPU runs this code
++ * with interrupts disabled and will complete the lock
++ * protected section without being interrupted. So there is no
++ * real need to boost.
++ */
++ if (unlikely(p == rq->idle)) {
++ WARN_ON(p != rq->curr);
++ WARN_ON(p->pi_blocked_on);
++ goto out_unlock;
+ }
+
-+err_enable:
-+ debugfs_remove(debug_threshold);
-+err_threshold:
-+ debugfs_remove(debug_sample_width);
-+err_width:
-+ debugfs_remove(debug_sample_window);
-+err_window:
-+ debugfs_remove(debug_max);
-+err_max:
-+ debugfs_remove(debug_count);
-+err_count:
-+ debugfs_remove(debug_sample);
-+err_sample:
-+ debugfs_remove(debug_dir);
-+err_debug_dir:
-+out:
-+ return ret;
-+}
+ trace_sched_pi_setprio(p, prio);
+ oldprio = p->prio;
+ prev_class = p->sched_class;
+@@ -4858,9 +5030,9 @@ void rt_mutex_setprio(struct task_struct *p, int prio)
+ enqueue_task(rq, p, oldprio < prio ? ENQUEUE_HEAD : 0);
+
+ check_class_changed(rq, p, prev_class, oldprio);
++out_unlock:
+ __task_rq_unlock(rq);
+ }
+-
+ #endif
+
+ void set_user_nice(struct task_struct *p, long nice)
+@@ -4995,7 +5167,13 @@ EXPORT_SYMBOL(task_nice);
+ */
+ int idle_cpu(int cpu)
+ {
+- return cpu_curr(cpu) == cpu_rq(cpu)->idle;
++ struct rq *rq = cpu_rq(cpu);
+
-+/**
-+ * free_debugfs - A function to cleanup the debugfs file interface
-+ */
-+static void free_debugfs(void)
-+{
-+ /* could also use a debugfs_remove_recursive */
-+ debugfs_remove(debug_enable);
-+ debugfs_remove(debug_threshold);
-+ debugfs_remove(debug_sample_width);
-+ debugfs_remove(debug_sample_window);
-+ debugfs_remove(debug_max);
-+ debugfs_remove(debug_count);
-+ debugfs_remove(debug_sample);
-+ debugfs_remove(debug_dir);
-+}
++#ifdef CONFIG_SMP
++ return rq->curr == rq->idle && !rq->nr_running && !rq->wake_list;
++#else
++ return rq->curr == rq->idle && !rq->nr_running;
++#endif
+ }
+
+ /**
+@@ -5529,7 +5707,7 @@ SYSCALL_DEFINE0(sched_yield)
+ __release(rq->lock);
+ spin_release(&rq->lock.dep_map, 1, _THIS_IP_);
+ do_raw_spin_unlock(&rq->lock);
+- preempt_enable_no_resched();
++ __preempt_enable_no_resched();
+
+ schedule();
+
+@@ -5543,9 +5721,17 @@ static inline int should_resched(void)
+
+ static void __cond_resched(void)
+ {
+- add_preempt_count(PREEMPT_ACTIVE);
+- __schedule();
+- sub_preempt_count(PREEMPT_ACTIVE);
++ do {
++ add_preempt_count(PREEMPT_ACTIVE);
++ __schedule();
++ sub_preempt_count(PREEMPT_ACTIVE);
++ /*
++ * Check again in case we missed a preemption
++ * opportunity between schedule and now.
++ */
++ barrier();
+
-+/**
-+ * detector_init - Standard module initialization code
-+ */
-+static int detector_init(void)
-+{
-+ int ret = -ENOMEM;
++ } while (need_resched());
+ }
+
+ int __sched _cond_resched(void)
+@@ -5586,6 +5772,7 @@ int __cond_resched_lock(spinlock_t *lock)
+ }
+ EXPORT_SYMBOL(__cond_resched_lock);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ int __sched __cond_resched_softirq(void)
+ {
+ BUG_ON(!in_softirq());
+@@ -5599,6 +5786,7 @@ int __sched __cond_resched_softirq(void)
+ return 0;
+ }
+ EXPORT_SYMBOL(__cond_resched_softirq);
++#endif
+
+ /**
+ * yield - yield the current processor to other threads.
+@@ -5845,7 +6033,7 @@ void show_state_filter(unsigned long state_filter)
+ printk(KERN_INFO
+ " task PC stack pid father\n");
+ #endif
+- read_lock(&tasklist_lock);
++ rcu_read_lock();
+ do_each_thread(g, p) {
+ /*
+ * reset the NMI-timeout, listing all files on a slow
+@@ -5861,7 +6049,7 @@ void show_state_filter(unsigned long state_filter)
+ #ifdef CONFIG_SCHED_DEBUG
+ sysrq_sched_debug_show();
+ #endif
+- read_unlock(&tasklist_lock);
++ rcu_read_unlock();
+ /*
+ * Only show locks if all tasks are dumped:
+ */
+@@ -5922,6 +6110,9 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu)
+ */
+ idle->sched_class = &idle_sched_class;
+ ftrace_graph_init_idle_task(idle, cpu);
++#if defined(CONFIG_SMP)
++ sprintf(idle->comm, "%s/%d", INIT_TASK_COMM, cpu);
++#endif
+ }
+
+ /*
+@@ -5983,12 +6174,12 @@ static inline void sched_init_granularity(void)
+ #ifdef CONFIG_SMP
+ void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
+ {
+- if (p->sched_class && p->sched_class->set_cpus_allowed)
+- p->sched_class->set_cpus_allowed(p, new_mask);
+- else {
+- cpumask_copy(&p->cpus_allowed, new_mask);
++ if (!migrate_disabled_updated(p)) {
++ if (p->sched_class && p->sched_class->set_cpus_allowed)
++ p->sched_class->set_cpus_allowed(p, new_mask);
+ p->rt.nr_cpus_allowed = cpumask_weight(new_mask);
+ }
++ cpumask_copy(&p->cpus_allowed, new_mask);
+ }
+
+ /*
+@@ -6039,7 +6230,7 @@ int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
+ do_set_cpus_allowed(p, new_mask);
+
+ /* Can the task run on the task's current CPU? If so, we're done */
+- if (cpumask_test_cpu(task_cpu(p), new_mask))
++ if (cpumask_test_cpu(task_cpu(p), new_mask) || __migrate_disabled(p))
+ goto out;
+
+ dest_cpu = cpumask_any_and(cpu_active_mask, new_mask);
+@@ -6086,7 +6277,7 @@ static int __migrate_task(struct task_struct *p, int src_cpu, int dest_cpu)
+ if (task_cpu(p) != src_cpu)
+ goto done;
+ /* Affinity changed (again). */
+- if (!cpumask_test_cpu(dest_cpu, &p->cpus_allowed))
++ if (!cpumask_test_cpu(dest_cpu, tsk_cpus_allowed(p)))
+ goto fail;
+
+ /*
+@@ -6128,6 +6319,8 @@ static int migration_cpu_stop(void *data)
+
+ #ifdef CONFIG_HOTPLUG_CPU
+
++static DEFINE_PER_CPU(struct mm_struct *, idle_last_mm);
+
-+ printk(KERN_INFO BANNER "version %s\n", VERSION);
+ /*
+ * Ensures that the idle task is using init_mm right before its cpu goes
+ * offline.
+@@ -6140,7 +6333,12 @@ void idle_task_exit(void)
+
+ if (mm != &init_mm)
+ switch_mm(mm, &init_mm, current);
+- mmdrop(mm);
+
-+ ret = init_stats();
-+ if (0 != ret)
-+ goto out;
++ /*
++ * Defer the cleanup to an alive cpu. On RT we can neither
++ * call mmdrop() nor mmdrop_delayed() from here.
++ */
++ per_cpu(idle_last_mm, smp_processor_id()) = mm;
+ }
+
+ /*
+@@ -6458,6 +6656,12 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu)
+ migrate_nr_uninterruptible(rq);
+ calc_global_load_remove(rq);
+ break;
++ case CPU_DEAD:
++ if (per_cpu(idle_last_mm, cpu)) {
++ mmdrop(per_cpu(idle_last_mm, cpu));
++ per_cpu(idle_last_mm, cpu) = NULL;
++ }
++ break;
+ #endif
+ }
+
+@@ -8175,7 +8379,8 @@ void __init sched_init(void)
+ #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP
+ static inline int preempt_count_equals(int preempt_offset)
+ {
+- int nested = (preempt_count() & ~PREEMPT_ACTIVE) + rcu_preempt_depth();
++ int nested = (preempt_count() & ~PREEMPT_ACTIVE) +
++ sched_rcu_preempt_depth();
+
+ return (nested == preempt_offset);
+ }
+diff --git a/kernel/sched_cpupri.c b/kernel/sched_cpupri.c
+index 2722dc1..90faffd 100644
+--- a/kernel/sched_cpupri.c
++++ b/kernel/sched_cpupri.c
+@@ -47,9 +47,6 @@ static int convert_prio(int prio)
+ return cpupri;
+ }
+
+-#define for_each_cpupri_active(array, idx) \
+- for_each_set_bit(idx, array, CPUPRI_NR_PRIORITIES)
+-
+ /**
+ * cpupri_find - find the best (lowest-pri) CPU in the system
+ * @cp: The cpupri context
+@@ -71,11 +68,38 @@ int cpupri_find(struct cpupri *cp, struct task_struct *p,
+ int idx = 0;
+ int task_pri = convert_prio(p->prio);
+
+- for_each_cpupri_active(cp->pri_active, idx) {
+- struct cpupri_vec *vec = &cp->pri_to_cpu[idx];
++ if (task_pri >= MAX_RT_PRIO)
++ return 0;
+
+- if (idx >= task_pri)
+- break;
++ for (idx = 0; idx < task_pri; idx++) {
++ struct cpupri_vec *vec = &cp->pri_to_cpu[idx];
++ int skip = 0;
+
-+ ret = init_debugfs();
-+ if (0 != ret)
-+ goto err_stats;
++ if (!atomic_read(&(vec)->count))
++ skip = 1;
++ /*
++ * When looking at the vector, we need to read the counter,
++ * do a memory barrier, then read the mask.
++ *
++ * Note: This is still all racey, but we can deal with it.
++ * Ideally, we only want to look at masks that are set.
++ *
++ * If a mask is not set, then the only thing wrong is that we
++ * did a little more work than necessary.
++ *
++ * If we read a zero count but the mask is set, because of the
++ * memory barriers, that can only happen when the highest prio
++ * task for a run queue has left the run queue, in which case,
++ * it will be followed by a pull. If the task we are processing
++ * fails to find a proper place to go, that pull request will
++ * pull this task if the run queue is running at a lower
++ * priority.
++ */
++ smp_rmb();
+
-+ if (enabled)
-+ ret = start_kthread();
++ /* Need to do the rmb for every iteration */
++ if (skip)
++ continue;
+
+ if (cpumask_any_and(&p->cpus_allowed, vec->mask) >= nr_cpu_ids)
+ continue;
+@@ -115,7 +139,7 @@ void cpupri_set(struct cpupri *cp, int cpu, int newpri)
+ {
+ int *currpri = &cp->cpu_to_pri[cpu];
+ int oldpri = *currpri;
+- unsigned long flags;
++ int do_mb = 0;
+
+ newpri = convert_prio(newpri);
+
+@@ -134,26 +158,41 @@ void cpupri_set(struct cpupri *cp, int cpu, int newpri)
+ if (likely(newpri != CPUPRI_INVALID)) {
+ struct cpupri_vec *vec = &cp->pri_to_cpu[newpri];
+
+- raw_spin_lock_irqsave(&vec->lock, flags);
+-
+ cpumask_set_cpu(cpu, vec->mask);
+- vec->count++;
+- if (vec->count == 1)
+- set_bit(newpri, cp->pri_active);
+-
+- raw_spin_unlock_irqrestore(&vec->lock, flags);
++ /*
++ * When adding a new vector, we update the mask first,
++ * do a write memory barrier, and then update the count, to
++ * make sure the vector is visible when count is set.
++ */
++ smp_mb__before_atomic_inc();
++ atomic_inc(&(vec)->count);
++ do_mb = 1;
+ }
+ if (likely(oldpri != CPUPRI_INVALID)) {
+ struct cpupri_vec *vec = &cp->pri_to_cpu[oldpri];
+
+- raw_spin_lock_irqsave(&vec->lock, flags);
+-
+- vec->count--;
+- if (!vec->count)
+- clear_bit(oldpri, cp->pri_active);
++ /*
++ * Because the order of modification of the vec->count
++ * is important, we must make sure that the update
++ * of the new prio is seen before we decrement the
++ * old prio. This makes sure that the loop sees
++ * one or the other when we raise the priority of
++ * the run queue. We don't care about when we lower the
++ * priority, as that will trigger an rt pull anyway.
++ *
++ * We only need to do a memory barrier if we updated
++ * the new priority vec.
++ */
++ if (do_mb)
++ smp_mb__after_atomic_inc();
+
-+ goto out;
++ /*
++ * When removing from the vector, we decrement the counter first
++ * do a memory barrier and then clear the mask.
++ */
++ atomic_dec(&(vec)->count);
++ smp_mb__after_atomic_inc();
+ cpumask_clear_cpu(cpu, vec->mask);
+-
+- raw_spin_unlock_irqrestore(&vec->lock, flags);
+ }
+
+ *currpri = newpri;
+@@ -175,8 +214,7 @@ int cpupri_init(struct cpupri *cp)
+ for (i = 0; i < CPUPRI_NR_PRIORITIES; i++) {
+ struct cpupri_vec *vec = &cp->pri_to_cpu[i];
+
+- raw_spin_lock_init(&vec->lock);
+- vec->count = 0;
++ atomic_set(&vec->count, 0);
+ if (!zalloc_cpumask_var(&vec->mask, GFP_KERNEL))
+ goto cleanup;
+ }
+diff --git a/kernel/sched_cpupri.h b/kernel/sched_cpupri.h
+index 9fc7d38..6b4cd17 100644
+--- a/kernel/sched_cpupri.h
++++ b/kernel/sched_cpupri.h
+@@ -12,9 +12,8 @@
+ /* values 2-101 are RT priorities 0-99 */
+
+ struct cpupri_vec {
+- raw_spinlock_t lock;
+- int count;
+- cpumask_var_t mask;
++ atomic_t count;
++ cpumask_var_t mask;
+ };
+
+ struct cpupri {
+diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c
+index a6710a1..528032b 100644
+--- a/kernel/sched_debug.c
++++ b/kernel/sched_debug.c
+@@ -235,6 +235,9 @@ void print_rt_rq(struct seq_file *m, int cpu, struct rt_rq *rt_rq)
+ P(rt_throttled);
+ PN(rt_time);
+ PN(rt_runtime);
++#ifdef CONFIG_SMP
++ P(rt_nr_migratory);
++#endif
+
+ #undef PN
+ #undef P
+@@ -484,6 +487,10 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m)
+ P(se.load.weight);
+ P(policy);
+ P(prio);
++#ifdef CONFIG_PREEMPT_RT_FULL
++ P(migrate_disable);
++#endif
++ P(rt.nr_cpus_allowed);
+ #undef PN
+ #undef __PN
+ #undef P
+diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
+index c768588..46b7855 100644
+--- a/kernel/sched_fair.c
++++ b/kernel/sched_fair.c
+@@ -1565,7 +1565,7 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p,
+
+ /* Skip over this group if it has no CPUs allowed */
+ if (!cpumask_intersects(sched_group_cpus(group),
+- &p->cpus_allowed))
++ tsk_cpus_allowed(p)))
+ continue;
+
+ local_group = cpumask_test_cpu(this_cpu,
+@@ -1611,7 +1611,7 @@ find_idlest_cpu(struct sched_group *group, struct task_struct *p, int this_cpu)
+ int i;
+
+ /* Traverse only the allowed CPUs */
+- for_each_cpu_and(i, sched_group_cpus(group), &p->cpus_allowed) {
++ for_each_cpu_and(i, sched_group_cpus(group), tsk_cpus_allowed(p)) {
+ load = weighted_cpuload(i);
+
+ if (load < min_load || (load == min_load && i == this_cpu)) {
+@@ -1655,7 +1655,7 @@ static int select_idle_sibling(struct task_struct *p, int target)
+ if (!(sd->flags & SD_SHARE_PKG_RESOURCES))
+ break;
+
+- for_each_cpu_and(i, sched_domain_span(sd), &p->cpus_allowed) {
++ for_each_cpu_and(i, sched_domain_span(sd), tsk_cpus_allowed(p)) {
+ if (idle_cpu(i)) {
+ target = i;
+ break;
+@@ -1698,7 +1698,7 @@ select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flags)
+ int sync = wake_flags & WF_SYNC;
+
+ if (sd_flag & SD_BALANCE_WAKE) {
+- if (cpumask_test_cpu(cpu, &p->cpus_allowed))
++ if (cpumask_test_cpu(cpu, tsk_cpus_allowed(p)))
+ want_affine = 1;
+ new_cpu = prev_cpu;
+ }
+@@ -2067,7 +2067,7 @@ int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu,
+ * 2) cannot be migrated to this CPU due to cpus_allowed, or
+ * 3) are cache-hot on their current CPU.
+ */
+- if (!cpumask_test_cpu(this_cpu, &p->cpus_allowed)) {
++ if (!cpumask_test_cpu(this_cpu, tsk_cpus_allowed(p))) {
+ schedstat_inc(p, se.statistics.nr_failed_migrations_affine);
+ return 0;
+ }
+@@ -2173,6 +2173,10 @@ balance_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest,
+ */
+ if (idle == CPU_NEWLY_IDLE)
+ break;
+
-+err_stats:
-+ ring_buffer_free(ring_buffer);
-+out:
-+ return ret;
++ if (raw_spin_is_contended(&this_rq->lock) ||
++ raw_spin_is_contended(&busiest->lock))
++ break;
+ #endif
+
+ /*
+@@ -2277,6 +2281,20 @@ load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest,
+ rem_load_move -= moved_load;
+ if (rem_load_move < 0)
+ break;
+
-+}
++#ifdef CONFIG_PREEMPT
++ /*
++ * NEWIDLE balancing is a source of latency, so preemptible
++ * kernels will stop after the first task is pulled to minimize
++ * the critical section.
++ */
++ if (idle == CPU_NEWLY_IDLE && this_rq->nr_running)
++ break;
+
-+/**
-+ * detector_exit - Standard module cleanup code
-+ */
-+static void detector_exit(void)
-+{
-+ int err;
++ if (raw_spin_is_contended(&this_rq->lock) ||
++ raw_spin_is_contended(&busiest->lock))
++ break;
++#endif
+ }
+ rcu_read_unlock();
+
+@@ -3418,7 +3436,7 @@ redo:
+ * moved to this_cpu
+ */
+ if (!cpumask_test_cpu(this_cpu,
+- &busiest->curr->cpus_allowed)) {
++ tsk_cpus_allowed(busiest->curr))) {
+ raw_spin_unlock_irqrestore(&busiest->lock,
+ flags);
+ all_pinned = 1;
+diff --git a/kernel/sched_features.h b/kernel/sched_features.h
+index 1e7066d..8099785 100644
+--- a/kernel/sched_features.h
++++ b/kernel/sched_features.h
+@@ -65,10 +65,15 @@ SCHED_FEAT(OWNER_SPIN, 1)
+ */
+ SCHED_FEAT(NONIRQ_POWER, 1)
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * Queue remote wakeups on the target CPU and process them
+ * using the scheduler IPI. Reduces rq->lock contention/bounces.
+ */
+ SCHED_FEAT(TTWU_QUEUE, 1)
++#else
++SCHED_FEAT(TTWU_QUEUE, 0)
++#endif
+
+ SCHED_FEAT(FORCE_SD_OVERLAP, 0)
++SCHED_FEAT(RT_RUNTIME_SHARE, 1)
+diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c
+index 17f2319..53525cb 100644
+--- a/kernel/sched_rt.c
++++ b/kernel/sched_rt.c
+@@ -536,6 +536,9 @@ static int balance_runtime(struct rt_rq *rt_rq)
+ {
+ int more = 0;
+
++ if (!sched_feat(RT_RUNTIME_SHARE))
++ return more;
+
-+ if (enabled) {
-+ enabled = 0;
-+ err = stop_kthread();
-+ if (err)
-+ printk(KERN_ERR BANNER "cannot stop kthread\n");
-+ }
+ if (rt_rq->rt_time > rt_rq->rt_runtime) {
+ raw_spin_unlock(&rt_rq->rt_runtime_lock);
+ more = do_balance_runtime(rt_rq);
+@@ -553,12 +556,9 @@ static inline int balance_runtime(struct rt_rq *rt_rq)
+
+ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
+ {
+- int i, idle = 1;
++ int i, idle = 1, throttled = 0;
+ const struct cpumask *span;
+
+- if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF)
+- return 1;
+-
+ span = sched_rt_period_mask();
+ for_each_cpu(i, span) {
+ int enqueue = 0;
+@@ -593,12 +593,17 @@ static int do_sched_rt_period_timer(struct rt_bandwidth *rt_b, int overrun)
+ if (!rt_rq_throttled(rt_rq))
+ enqueue = 1;
+ }
++ if (rt_rq->rt_throttled)
++ throttled = 1;
+
+ if (enqueue)
+ sched_rt_rq_enqueue(rt_rq);
+ raw_spin_unlock(&rq->lock);
+ }
+
++ if (!throttled && (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF))
++ return 1;
+
-+ free_debugfs();
-+ ring_buffer_free(ring_buffer); /* free up the ring buffer */
+ return idle;
+ }
+
+@@ -630,7 +635,24 @@ static int sched_rt_runtime_exceeded(struct rt_rq *rt_rq)
+ return 0;
+
+ if (rt_rq->rt_time > runtime) {
+- rt_rq->rt_throttled = 1;
++ struct rt_bandwidth *rt_b = sched_rt_bandwidth(rt_rq);
+
-+}
++ /*
++ * Don't actually throttle groups that have no runtime assigned
++ * but accrue some time due to boosting.
++ */
++ if (likely(rt_b->rt_runtime)) {
++ rt_rq->rt_throttled = 1;
++ printk_once(KERN_WARNING "sched: RT throttling activated\n");
++ } else {
++ /*
++ * In case we did anyway, make it go away,
++ * replenishment is a joke, since it will replenish us
++ * with exactly 0 ns.
++ */
++ rt_rq->rt_time = 0;
++ }
+
-+module_init(detector_init);
-+module_exit(detector_exit);
-Index: linux-3.0/arch/arm/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/arm/kernel/early_printk.c
-+++ linux-3.0/arch/arm/kernel/early_printk.c
-@@ -29,28 +29,17 @@ static void early_console_write(struct c
- early_write(s, n);
- }
+ if (rt_rq_throttled(rt_rq)) {
+ sched_rt_rq_dequeue(rt_rq);
+ return 1;
+@@ -658,7 +680,8 @@ static void update_curr_rt(struct rq *rq)
+ if (unlikely((s64)delta_exec < 0))
+ delta_exec = 0;
--static struct console early_console = {
-+static struct console early_console_dev = {
- .name = "earlycon",
- .write = early_console_write,
- .flags = CON_PRINTBUFFER | CON_BOOT,
- .index = -1,
- };
+- schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec));
++ schedstat_set(curr->se.statistics.exec_max,
++ max(curr->se.statistics.exec_max, delta_exec));
--asmlinkage void early_printk(const char *fmt, ...)
--{
-- char buf[512];
-- int n;
-- va_list ap;
--
-- va_start(ap, fmt);
-- n = vscnprintf(buf, sizeof(buf), fmt, ap);
-- early_write(buf, n);
-- va_end(ap);
--}
--
- static int __init setup_early_printk(char *buf)
+ curr->se.sum_exec_runtime += delta_exec;
+ account_group_exec_runtime(curr, delta_exec);
+@@ -1186,7 +1209,7 @@ static void deactivate_task(struct rq *rq, struct task_struct *p, int sleep);
+ static int pick_rt_task(struct rq *rq, struct task_struct *p, int cpu)
{
-- register_console(&early_console);
-+ early_console = &early_console_dev;
-+ register_console(&early_console_dev);
+ if (!task_running(rq, p) &&
+- (cpu < 0 || cpumask_test_cpu(cpu, &p->cpus_allowed)) &&
++ (cpu < 0 || cpumask_test_cpu(cpu, tsk_cpus_allowed(p))) &&
+ (p->rt.nr_cpus_allowed > 1))
+ return 1;
return 0;
- }
+@@ -1331,7 +1354,7 @@ static struct rq *find_lock_lowest_rq(struct task_struct *task, struct rq *rq)
+ */
+ if (unlikely(task_rq(task) != rq ||
+ !cpumask_test_cpu(lowest_rq->cpu,
+- &task->cpus_allowed) ||
++ tsk_cpus_allowed(task)) ||
+ task_running(rq, task) ||
+ !task->on_rq)) {
-Index: linux-3.0/arch/blackfin/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/blackfin/kernel/early_printk.c
-+++ linux-3.0/arch/blackfin/kernel/early_printk.c
-@@ -25,8 +25,6 @@ extern struct console *bfin_earlyserial_
- extern struct console *bfin_jc_early_init(void);
- #endif
+@@ -1614,9 +1637,6 @@ static void set_cpus_allowed_rt(struct task_struct *p,
--static struct console *early_console;
+ update_rt_migration(&rq->rt);
+ }
-
- /* Default console */
- #define DEFAULT_PORT 0
- #define DEFAULT_CFLAG CS8|B57600
-Index: linux-3.0/arch/microblaze/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/microblaze/kernel/early_printk.c
-+++ linux-3.0/arch/microblaze/kernel/early_printk.c
-@@ -21,7 +21,6 @@
- #include <asm/setup.h>
- #include <asm/prom.h>
+- cpumask_copy(&p->cpus_allowed, new_mask);
+- p->rt.nr_cpus_allowed = weight;
+ }
--static u32 early_console_initialized;
- static u32 base_addr;
+ /* Assumes rq->lock is held */
+diff --git a/kernel/sched_stats.h b/kernel/sched_stats.h
+index 331e01b..87f9e36 100644
+--- a/kernel/sched_stats.h
++++ b/kernel/sched_stats.h
+@@ -282,10 +282,10 @@ static inline void account_group_user_time(struct task_struct *tsk,
+ if (!cputimer->running)
+ return;
- #ifdef CONFIG_SERIAL_UARTLITE_CONSOLE
-@@ -109,31 +108,14 @@ static struct console early_serial_uart1
- };
- #endif /* CONFIG_SERIAL_8250_CONSOLE */
+- spin_lock(&cputimer->lock);
++ raw_spin_lock(&cputimer->lock);
+ cputimer->cputime.utime =
+ cputime_add(cputimer->cputime.utime, cputime);
+- spin_unlock(&cputimer->lock);
++ raw_spin_unlock(&cputimer->lock);
+ }
--static struct console *early_console;
--
--void early_printk(const char *fmt, ...)
--{
-- char buf[512];
-- int n;
-- va_list ap;
--
-- if (early_console_initialized) {
-- va_start(ap, fmt);
-- n = vscnprintf(buf, 512, fmt, ap);
-- early_console->write(early_console, buf, n);
-- va_end(ap);
-- }
--}
--
- int __init setup_early_printk(char *opt)
- {
-- if (early_console_initialized)
-+ if (early_console)
- return 1;
+ /**
+@@ -306,10 +306,10 @@ static inline void account_group_system_time(struct task_struct *tsk,
+ if (!cputimer->running)
+ return;
- #ifdef CONFIG_SERIAL_UARTLITE_CONSOLE
- base_addr = early_uartlite_console();
- if (base_addr) {
-- early_console_initialized = 1;
- #ifdef CONFIG_MMU
- early_console_reg_tlb_alloc(base_addr);
- #endif
-@@ -151,7 +133,6 @@ int __init setup_early_printk(char *opt)
- base_addr = early_uart16550_console();
- base_addr &= ~3; /* clear register offset */
- if (base_addr) {
-- early_console_initialized = 1;
- #ifdef CONFIG_MMU
- early_console_reg_tlb_alloc(base_addr);
- #endif
-@@ -171,9 +152,9 @@ int __init setup_early_printk(char *opt)
+- spin_lock(&cputimer->lock);
++ raw_spin_lock(&cputimer->lock);
+ cputimer->cputime.stime =
+ cputime_add(cputimer->cputime.stime, cputime);
+- spin_unlock(&cputimer->lock);
++ raw_spin_unlock(&cputimer->lock);
+ }
- void __init disable_early_printk(void)
- {
-- if (!early_console_initialized || !early_console)
-+ if (!early_console)
+ /**
+@@ -330,7 +330,7 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
+ if (!cputimer->running)
return;
- printk(KERN_WARNING "disabling early console\n");
- unregister_console(early_console);
-- early_console_initialized = 0;
-+ early_console = NULL;
+
+- spin_lock(&cputimer->lock);
++ raw_spin_lock(&cputimer->lock);
+ cputimer->cputime.sum_exec_runtime += ns;
+- spin_unlock(&cputimer->lock);
++ raw_spin_unlock(&cputimer->lock);
}
-Index: linux-3.0/arch/mips/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/mips/kernel/early_printk.c
-+++ linux-3.0/arch/mips/kernel/early_printk.c
-@@ -25,20 +25,18 @@ early_console_write(struct console *con,
- }
+diff --git a/kernel/semaphore.c b/kernel/semaphore.c
+index 94a62c0..d831841 100644
+--- a/kernel/semaphore.c
++++ b/kernel/semaphore.c
+@@ -54,12 +54,12 @@ void down(struct semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ if (likely(sem->count > 0))
+ sem->count--;
+ else
+ __down(sem);
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
}
+ EXPORT_SYMBOL(down);
--static struct console early_console __initdata = {
-+static struct console early_console_prom = {
- .name = "early",
- .write = early_console_write,
- .flags = CON_PRINTBUFFER | CON_BOOT,
- .index = -1
- };
+@@ -77,12 +77,12 @@ int down_interruptible(struct semaphore *sem)
+ unsigned long flags;
+ int result = 0;
+
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ if (likely(sem->count > 0))
+ sem->count--;
+ else
+ result = __down_interruptible(sem);
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
+
+ return result;
+ }
+@@ -103,12 +103,12 @@ int down_killable(struct semaphore *sem)
+ unsigned long flags;
+ int result = 0;
+
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ if (likely(sem->count > 0))
+ sem->count--;
+ else
+ result = __down_killable(sem);
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
+
+ return result;
+ }
+@@ -132,11 +132,11 @@ int down_trylock(struct semaphore *sem)
+ unsigned long flags;
+ int count;
+
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ count = sem->count - 1;
+ if (likely(count >= 0))
+ sem->count = count;
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
+
+ return (count < 0);
+ }
+@@ -157,12 +157,12 @@ int down_timeout(struct semaphore *sem, long jiffies)
+ unsigned long flags;
+ int result = 0;
+
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ if (likely(sem->count > 0))
+ sem->count--;
+ else
+ result = __down_timeout(sem, jiffies);
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
--static int early_console_initialized __initdata;
--
- void __init setup_early_printk(void)
+ return result;
+ }
+@@ -179,12 +179,12 @@ void up(struct semaphore *sem)
{
-- if (early_console_initialized)
-+ if (early_console)
- return;
-- early_console_initialized = 1;
-+ early_console = &early_console_prom;
+ unsigned long flags;
-- register_console(&early_console);
-+ register_console(&early_console_prom);
+- spin_lock_irqsave(&sem->lock, flags);
++ raw_spin_lock_irqsave(&sem->lock, flags);
+ if (likely(list_empty(&sem->wait_list)))
+ sem->count++;
+ else
+ __up(sem);
+- spin_unlock_irqrestore(&sem->lock, flags);
++ raw_spin_unlock_irqrestore(&sem->lock, flags);
}
-Index: linux-3.0/arch/powerpc/kernel/udbg.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/udbg.c
-+++ linux-3.0/arch/powerpc/kernel/udbg.c
-@@ -169,15 +169,13 @@ static struct console udbg_console = {
- .index = 0,
- };
+ EXPORT_SYMBOL(up);
--static int early_console_initialized;
--
+@@ -217,9 +217,9 @@ static inline int __sched __down_common(struct semaphore *sem, long state,
+ if (timeout <= 0)
+ goto timed_out;
+ __set_task_state(task, state);
+- spin_unlock_irq(&sem->lock);
++ raw_spin_unlock_irq(&sem->lock);
+ timeout = schedule_timeout(timeout);
+- spin_lock_irq(&sem->lock);
++ raw_spin_lock_irq(&sem->lock);
+ if (waiter.up)
+ return 0;
+ }
+diff --git a/kernel/signal.c b/kernel/signal.c
+index 43fee1c..3f15914 100644
+--- a/kernel/signal.c
++++ b/kernel/signal.c
+@@ -300,13 +300,45 @@ static bool task_participate_group_stop(struct task_struct *task)
+ return false;
+ }
+
++#ifdef __HAVE_ARCH_CMPXCHG
++static inline struct sigqueue *get_task_cache(struct task_struct *t)
++{
++ struct sigqueue *q = t->sigqueue_cache;
++
++ if (cmpxchg(&t->sigqueue_cache, q, NULL) != q)
++ return NULL;
++ return q;
++}
++
++static inline int put_task_cache(struct task_struct *t, struct sigqueue *q)
++{
++ if (cmpxchg(&t->sigqueue_cache, NULL, q) == NULL)
++ return 0;
++ return 1;
++}
++
++#else
++
++static inline struct sigqueue *get_task_cache(struct task_struct *t)
++{
++ return NULL;
++}
++
++static inline int put_task_cache(struct task_struct *t, struct sigqueue *q)
++{
++ return 1;
++}
++
++#endif
++
/*
- * Called by setup_system after ppc_md->probe and ppc_md->early_init.
- * Call it again after setting udbg_putc in ppc_md->setup_arch.
+ * allocate a new signal queue record
+ * - this may be called without locks if and only if t == current, otherwise an
+ * appropriate lock must be held to stop the target task from exiting
*/
- void __init register_early_udbg_console(void)
+ static struct sigqueue *
+-__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimit)
++__sigqueue_do_alloc(int sig, struct task_struct *t, gfp_t flags,
++ int override_rlimit, int fromslab)
{
-- if (early_console_initialized)
-+ if (early_console)
- return;
-
- if (!udbg_putc)
-@@ -187,7 +185,7 @@ void __init register_early_udbg_console(
- printk(KERN_INFO "early console immortal !\n");
- udbg_console.flags &= ~CON_BOOT;
+ struct sigqueue *q = NULL;
+ struct user_struct *user;
+@@ -323,7 +355,10 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi
+ if (override_rlimit ||
+ atomic_read(&user->sigpending) <=
+ task_rlimit(t, RLIMIT_SIGPENDING)) {
+- q = kmem_cache_alloc(sigqueue_cachep, flags);
++ if (!fromslab)
++ q = get_task_cache(t);
++ if (!q)
++ q = kmem_cache_alloc(sigqueue_cachep, flags);
+ } else {
+ print_dropped_signal(sig);
}
-- early_console_initialized = 1;
-+ early_console = &udbg_console;
- register_console(&udbg_console);
+@@ -340,6 +375,13 @@ __sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags, int override_rlimi
+ return q;
}
-Index: linux-3.0/arch/sh/kernel/sh_bios.c
-===================================================================
---- linux-3.0.orig/arch/sh/kernel/sh_bios.c
-+++ linux-3.0/arch/sh/kernel/sh_bios.c
-@@ -144,8 +144,6 @@ static struct console bios_console = {
- .index = -1,
- };
-
--static struct console *early_console;
--
- static int __init setup_early_printk(char *buf)
++static struct sigqueue *
++__sigqueue_alloc(int sig, struct task_struct *t, gfp_t flags,
++ int override_rlimit)
++{
++ return __sigqueue_do_alloc(sig, t, flags, override_rlimit, 0);
++}
++
+ static void __sigqueue_free(struct sigqueue *q)
{
- int keep_early = 0;
-Index: linux-3.0/arch/sparc/kernel/setup_32.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/setup_32.c
-+++ linux-3.0/arch/sparc/kernel/setup_32.c
-@@ -220,6 +220,7 @@ void __init setup_arch(char **cmdline_p)
-
- boot_flags_init(*cmdline_p);
-
-+ early_console = &prom_early_console;
- register_console(&prom_early_console);
-
- /* Set sparc_cpu_model */
-Index: linux-3.0/arch/sparc/kernel/setup_64.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/setup_64.c
-+++ linux-3.0/arch/sparc/kernel/setup_64.c
-@@ -469,6 +469,12 @@ static void __init init_sparc64_elf_hwca
- popc_patch();
+ if (q->flags & SIGQUEUE_PREALLOC)
+@@ -349,6 +391,21 @@ static void __sigqueue_free(struct sigqueue *q)
+ kmem_cache_free(sigqueue_cachep, q);
}
-+static inline void register_prom_console(void)
++static void sigqueue_free_current(struct sigqueue *q)
+{
-+ early_console = &prom_early_console;
-+ register_console(&prom_early_console);
++ struct user_struct *up;
++
++ if (q->flags & SIGQUEUE_PREALLOC)
++ return;
++
++ up = q->user;
++ if (rt_prio(current->normal_prio) && !put_task_cache(current, q)) {
++ atomic_dec(&up->sigpending);
++ free_uid(up);
++ } else
++ __sigqueue_free(q);
+}
+
- void __init setup_arch(char **cmdline_p)
+ void flush_sigqueue(struct sigpending *queue)
{
- /* Initialize PROM console and command line. */
-@@ -480,7 +486,7 @@ void __init setup_arch(char **cmdline_p)
- #ifdef CONFIG_EARLYFB
- if (btext_find_display())
- #endif
-- register_console(&prom_early_console);
-+ register_prom_console();
-
- if (tlb_type == hypervisor)
- printk("ARCH: SUN4V\n");
-Index: linux-3.0/arch/tile/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/tile/kernel/early_printk.c
-+++ linux-3.0/arch/tile/kernel/early_printk.c
-@@ -32,25 +32,8 @@ static struct console early_hv_console =
- };
-
- /* Direct interface for emergencies */
--static struct console *early_console = &early_hv_console;
--static int early_console_initialized;
- static int early_console_complete;
+ struct sigqueue *q;
+@@ -362,6 +419,21 @@ void flush_sigqueue(struct sigpending *queue)
+ }
--static void early_vprintk(const char *fmt, va_list ap)
--{
-- char buf[512];
-- int n = vscnprintf(buf, sizeof(buf), fmt, ap);
-- early_console->write(early_console, buf, n);
--}
--
--void early_printk(const char *fmt, ...)
--{
-- va_list ap;
-- va_start(ap, fmt);
-- early_vprintk(fmt, ap);
-- va_end(ap);
--}
--
- void early_panic(const char *fmt, ...)
+ /*
++ * Called from __exit_signal. Flush tsk->pending and
++ * tsk->sigqueue_cache
++ */
++void flush_task_sigqueue(struct task_struct *tsk)
++{
++ struct sigqueue *q;
++
++ flush_sigqueue(&tsk->pending);
++
++ q = get_task_cache(tsk);
++ if (q)
++ kmem_cache_free(sigqueue_cachep, q);
++}
++
++/*
+ * Flush all pending signals for a task.
+ */
+ void __flush_signals(struct task_struct *t)
+@@ -509,7 +581,7 @@ static void collect_signal(int sig, struct sigpending *list, siginfo_t *info)
+ still_pending:
+ list_del_init(&first->list);
+ copy_siginfo(info, &first->info);
+- __sigqueue_free(first);
++ sigqueue_free_current(first);
+ } else {
+ /*
+ * Ok, it wasn't in the queue. This must be
+@@ -555,6 +627,8 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info)
{
- va_list ap;
-@@ -68,14 +51,13 @@ static int __initdata keep_early;
+ int signr;
- static int __init setup_early_printk(char *str)
- {
-- if (early_console_initialized)
-+ if (early_console)
- return 1;
++ WARN_ON_ONCE(tsk != current);
++
+ /* We only dequeue private signals from ourselves, we don't let
+ * signalfd steal them
+ */
+@@ -637,6 +711,9 @@ void signal_wake_up(struct task_struct *t, int resume)
- if (str != NULL && strncmp(str, "keep", 4) == 0)
- keep_early = 1;
+ set_tsk_thread_flag(t, TIF_SIGPENDING);
- early_console = &early_hv_console;
-- early_console_initialized = 1;
- register_console(early_console);
++ if (unlikely(t == current))
++ return;
++
+ /*
+ * For SIGKILL, we want to wake it up in the stopped/traced/killable
+ * case. We don't check t->state here because there is a race with it
+@@ -1179,12 +1256,12 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
+ struct sighand_struct *sighand;
+
+ for (;;) {
+- local_irq_save(*flags);
++ local_irq_save_nort(*flags);
+ rcu_read_lock();
+ sighand = rcu_dereference(tsk->sighand);
+ if (unlikely(sighand == NULL)) {
+ rcu_read_unlock();
+- local_irq_restore(*flags);
++ local_irq_restore_nort(*flags);
+ break;
+ }
- return 0;
-@@ -84,12 +66,12 @@ static int __init setup_early_printk(cha
- void __init disable_early_printk(void)
- {
- early_console_complete = 1;
-- if (!early_console_initialized || !early_console)
-+ if (!early_console)
- return;
- if (!keep_early) {
- early_printk("disabling early console\n");
- unregister_console(early_console);
-- early_console_initialized = 0;
-+ early_console = NULL;
- } else {
- early_printk("keeping early console\n");
+@@ -1195,7 +1272,7 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk,
+ }
+ spin_unlock(&sighand->siglock);
+ rcu_read_unlock();
+- local_irq_restore(*flags);
++ local_irq_restore_nort(*flags);
}
-@@ -97,7 +79,7 @@ void __init disable_early_printk(void)
- void warn_early_printk(void)
+ return sighand;
+@@ -1434,7 +1511,8 @@ EXPORT_SYMBOL(kill_pid);
+ */
+ struct sigqueue *sigqueue_alloc(void)
{
-- if (early_console_complete || early_console_initialized)
-+ if (early_console_complete || early_console)
- return;
- early_printk("\
- Machine shutting down before console output is fully initialized.\n\
-Index: linux-3.0/arch/um/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/um/kernel/early_printk.c
-+++ linux-3.0/arch/um/kernel/early_printk.c
-@@ -16,7 +16,7 @@ static void early_console_write(struct c
- um_early_printk(s, n);
- }
+- struct sigqueue *q = __sigqueue_alloc(-1, current, GFP_KERNEL, 0);
++ /* Preallocated sigqueue objects always from the slabcache ! */
++ struct sigqueue *q = __sigqueue_do_alloc(-1, current, GFP_KERNEL, 0, 1);
--static struct console early_console = {
-+static struct console early_console_dev = {
- .name = "earlycon",
- .write = early_console_write,
- .flags = CON_BOOT,
-@@ -25,8 +25,10 @@ static struct console early_console = {
+ if (q)
+ q->flags |= SIGQUEUE_PREALLOC;
+@@ -1782,15 +1860,7 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info)
+ if (gstop_done && !real_parent_is_ptracer(current))
+ do_notify_parent_cldstop(current, false, why);
- static int __init setup_early_printk(char *buf)
- {
-- register_console(&early_console);
--
-+ if (!early_console) {
-+ early_console = &early_console_dev;
-+ register_console(&early_console_dev);
-+ }
- return 0;
- }
+- /*
+- * Don't want to allow preemption here, because
+- * sys_ptrace() needs this task to be inactive.
+- *
+- * XXX: implement read_unlock_no_resched().
+- */
+- preempt_disable();
+ read_unlock(&tasklist_lock);
+- preempt_enable_no_resched();
+ schedule();
+ } else {
+ /*
+diff --git a/kernel/softirq.c b/kernel/softirq.c
+index fca82c3..026a283 100644
+--- a/kernel/softirq.c
++++ b/kernel/softirq.c
+@@ -24,6 +24,7 @@
+ #include <linux/ftrace.h>
+ #include <linux/smp.h>
+ #include <linux/tick.h>
++#include <linux/locallock.h>
-Index: linux-3.0/arch/unicore32/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/unicore32/kernel/early_printk.c
-+++ linux-3.0/arch/unicore32/kernel/early_printk.c
-@@ -33,21 +33,17 @@ static struct console early_ocd_console
- .index = -1,
+ #define CREATE_TRACE_POINTS
+ #include <trace/events/irq.h>
+@@ -61,6 +62,67 @@ char *softirq_to_name[NR_SOFTIRQS] = {
+ "TASKLET", "SCHED", "HRTIMER", "RCU"
};
--/* Direct interface for emergencies */
--static struct console *early_console = &early_ocd_console;
--
--static int __initdata keep_early;
--
- static int __init setup_early_printk(char *buf)
- {
-- if (!buf)
-+ int keep_early;
++#ifdef CONFIG_NO_HZ
++# ifdef CONFIG_PREEMPT_RT_FULL
++/*
++ * On preempt-rt a softirq might be blocked on a lock. There might be
++ * no other runnable task on this CPU because the lock owner runs on
++ * some other CPU. So we have to go into idle with the pending bit
++ * set. Therefor we need to check this otherwise we warn about false
++ * positives which confuses users and defeats the whole purpose of
++ * this test.
++ *
++ * This code is called with interrupts disabled.
++ */
++void softirq_check_pending_idle(void)
++{
++ static int rate_limit;
++ u32 warnpending = 0, pending = local_softirq_pending();
+
-+ if (!buf || early_console)
- return 0;
-
- if (strstr(buf, "keep"))
- keep_early = 1;
-
-- if (!strncmp(buf, "ocd", 3))
-- early_console = &early_ocd_console;
-+ early_console = &early_ocd_console;
++ if (rate_limit >= 10)
++ return;
++
++ if (pending) {
++ struct task_struct *tsk;
++
++ tsk = __get_cpu_var(ksoftirqd);
++ /*
++ * The wakeup code in rtmutex.c wakes up the task
++ * _before_ it sets pi_blocked_on to NULL under
++ * tsk->pi_lock. So we need to check for both: state
++ * and pi_blocked_on.
++ */
++ raw_spin_lock(&tsk->pi_lock);
++
++ if (!tsk->pi_blocked_on && !(tsk->state == TASK_RUNNING))
++ warnpending = 1;
++
++ raw_spin_unlock(&tsk->pi_lock);
++ }
++
++ if (warnpending) {
++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
++ pending);
++ rate_limit++;
++ }
++}
++# else
++/*
++ * On !PREEMPT_RT we just printk rate limited:
++ */
++void softirq_check_pending_idle(void)
++{
++ static int rate_limit;
++
++ if (rate_limit < 10) {
++ printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
++ local_softirq_pending());
++ rate_limit++;
++ }
++}
++# endif
++#endif
++
+ /*
+ * we cannot loop indefinitely here to avoid userspace starvation,
+ * but we also don't want to introduce a worst case 1/HZ latency
+@@ -76,6 +138,36 @@ static void wakeup_softirqd(void)
+ wake_up_process(tsk);
+ }
- if (keep_early)
- early_console->flags &= ~CON_BOOT;
-Index: linux-3.0/arch/x86/kernel/early_printk.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/early_printk.c
-+++ linux-3.0/arch/x86/kernel/early_printk.c
-@@ -169,25 +169,9 @@ static struct console early_serial_conso
- .index = -1,
- };
++static void handle_pending_softirqs(u32 pending, int cpu, int need_rcu_bh_qs)
++{
++ struct softirq_action *h = softirq_vec;
++ unsigned int prev_count = preempt_count();
++
++ local_irq_enable();
++ for ( ; pending; h++, pending >>= 1) {
++ unsigned int vec_nr = h - softirq_vec;
++
++ if (!(pending & 1))
++ continue;
++
++ kstat_incr_softirqs_this_cpu(vec_nr);
++ trace_softirq_entry(vec_nr);
++ h->action(h);
++ trace_softirq_exit(vec_nr);
++ if (unlikely(prev_count != preempt_count())) {
++ printk(KERN_ERR
++ "huh, entered softirq %u %s %p with preempt_count %08x exited with %08x?\n",
++ vec_nr, softirq_to_name[vec_nr], h->action,
++ prev_count, (unsigned int) preempt_count());
++ preempt_count() = prev_count;
++ }
++ if (need_rcu_bh_qs)
++ rcu_bh_qs(cpu);
++ }
++ local_irq_disable();
++}
++
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * preempt_count and SOFTIRQ_OFFSET usage:
+ * - preempt_count is changed by SOFTIRQ_OFFSET on entering or leaving
+@@ -206,7 +298,6 @@ EXPORT_SYMBOL(local_bh_enable_ip);
--/* Direct interface for emergencies */
--static struct console *early_console = &early_vga_console;
--static int __initdata early_console_initialized;
--
--asmlinkage void early_printk(const char *fmt, ...)
--{
-- char buf[512];
-- int n;
-- va_list ap;
--
-- va_start(ap, fmt);
-- n = vscnprintf(buf, sizeof(buf), fmt, ap);
-- early_console->write(early_console, buf, n);
-- va_end(ap);
--}
--
- static inline void early_console_register(struct console *con, int keep_early)
+ asmlinkage void __do_softirq(void)
{
-- if (early_console->index != -1) {
-+ if (con->index != -1) {
- printk(KERN_CRIT "ERROR: earlyprintk= %s already used\n",
- con->name);
- return;
-@@ -207,9 +191,8 @@ static int __init setup_early_printk(cha
- if (!buf)
- return 0;
-
-- if (early_console_initialized)
-+ if (early_console)
- return 0;
-- early_console_initialized = 1;
+- struct softirq_action *h;
+ __u32 pending;
+ int max_restart = MAX_SOFTIRQ_RESTART;
+ int cpu;
+@@ -215,7 +306,7 @@ asmlinkage void __do_softirq(void)
+ account_system_vtime(current);
- keep = (strstr(buf, "keep") != NULL);
+ __local_bh_disable((unsigned long)__builtin_return_address(0),
+- SOFTIRQ_OFFSET);
++ SOFTIRQ_OFFSET);
+ lockdep_softirq_enter();
-Index: linux-3.0/include/linux/console.h
-===================================================================
---- linux-3.0.orig/include/linux/console.h
-+++ linux-3.0/include/linux/console.h
-@@ -133,6 +133,7 @@ struct console {
- for (con = console_drivers; con != NULL; con = con->next)
+ cpu = smp_processor_id();
+@@ -223,36 +314,7 @@ restart:
+ /* Reset the pending bitmask before enabling irqs */
+ set_softirq_pending(0);
- extern int console_set_on_cmdline;
-+extern struct console *early_console;
+- local_irq_enable();
+-
+- h = softirq_vec;
+-
+- do {
+- if (pending & 1) {
+- unsigned int vec_nr = h - softirq_vec;
+- int prev_count = preempt_count();
+-
+- kstat_incr_softirqs_this_cpu(vec_nr);
+-
+- trace_softirq_entry(vec_nr);
+- h->action(h);
+- trace_softirq_exit(vec_nr);
+- if (unlikely(prev_count != preempt_count())) {
+- printk(KERN_ERR "huh, entered softirq %u %s %p"
+- "with preempt_count %08x,"
+- " exited with %08x?\n", vec_nr,
+- softirq_to_name[vec_nr], h->action,
+- prev_count, preempt_count());
+- preempt_count() = prev_count;
+- }
+-
+- rcu_bh_qs(cpu);
+- }
+- h++;
+- pending >>= 1;
+- } while (pending);
+-
+- local_irq_disable();
++ handle_pending_softirqs(pending, cpu, 1);
- extern int add_preferred_console(char *name, int idx, char *options);
- extern int update_console_cmdline(char *name, int idx, char *name_new, int idx_new, char *options);
-Index: linux-3.0/include/linux/printk.h
-===================================================================
---- linux-3.0.orig/include/linux/printk.h
-+++ linux-3.0/include/linux/printk.h
-@@ -88,8 +88,15 @@ int no_printk(const char *fmt, ...)
- return 0;
+ pending = local_softirq_pending();
+ if (pending && --max_restart)
+@@ -267,6 +329,26 @@ restart:
+ __local_bh_enable(SOFTIRQ_OFFSET);
}
-+#ifdef CONFIG_EARLY_PRINTK
- extern asmlinkage __attribute__ ((format (printf, 1, 2)))
- void early_printk(const char *fmt, ...);
-+extern void printk_kill(void);
-+#else
-+static inline __attribute__ ((format (printf, 1, 2))) __cold
-+void early_printk(const char *s, ...) { }
-+static inline void printk_kill(void) { }
-+#endif
-
- extern int printk_needs_cpu(int cpu);
- extern void printk_tick(void);
-@@ -109,7 +116,6 @@ extern int __printk_ratelimit(const char
- #define printk_ratelimit() __printk_ratelimit(__func__)
- extern bool printk_timed_ratelimit(unsigned long *caller_jiffies,
- unsigned int interval_msec);
--
- extern int printk_delay_msec;
- extern int dmesg_restrict;
- extern int kptr_restrict;
-Index: linux-3.0/kernel/Kconfig.preempt
-===================================================================
---- linux-3.0.orig/kernel/Kconfig.preempt
-+++ linux-3.0/kernel/Kconfig.preempt
-@@ -1,3 +1,9 @@
-+config PREEMPT
-+ bool
++/*
++ * Called with preemption disabled from run_ksoftirqd()
++ */
++static int ksoftirqd_do_softirq(int cpu)
++{
++ /*
++ * Preempt disable stops cpu going offline.
++ * If already offline, we'll be on wrong CPU:
++ * don't process.
++ */
++ if (cpu_is_offline(cpu))
++ return -1;
+
-+config PREEMPT_RT_BASE
-+ bool
-+ select PREEMPT
-
- choice
- prompt "Preemption Model"
-@@ -33,8 +39,9 @@ config PREEMPT_VOLUNTARY
++ local_irq_disable();
++ if (local_softirq_pending())
++ __do_softirq();
++ local_irq_enable();
++ return 0;
++}
++
+ #ifndef __ARCH_HAS_DO_SOFTIRQ
- Select this if you are building a kernel for a desktop system.
+ asmlinkage void do_softirq(void)
+@@ -289,6 +371,184 @@ asmlinkage void do_softirq(void)
--config PREEMPT
-+config PREEMPT_LL
- bool "Preemptible Kernel (Low-Latency Desktop)"
-+ select PREEMPT
- help
- This option reduces the latency of the kernel by making
- all kernel code (that is not executing in a critical section)
-@@ -50,5 +57,20 @@ config PREEMPT
- embedded system with latency requirements in the milliseconds
- range.
+ #endif
-+config PREEMPT_RTB
-+ bool "Preemptible Kernel (Basic RT)"
-+ select PREEMPT_RT_BASE
-+ help
-+ This option is basically the same as (Low-Latency Desktop) but
-+ enables changes which are preliminary for the full preemptiple
-+ RT kernel.
++static inline void local_bh_disable_nort(void) { local_bh_disable(); }
++static inline void _local_bh_enable_nort(void) { _local_bh_enable(); }
++static inline void ksoftirqd_set_sched_params(void) { }
++static inline void ksoftirqd_clr_sched_params(void) { }
++
++#else /* !PREEMPT_RT_FULL */
++
++/*
++ * On RT we serialize softirq execution with a cpu local lock
++ */
++static DEFINE_LOCAL_IRQ_LOCK(local_softirq_lock);
++static DEFINE_PER_CPU(struct task_struct *, local_softirq_runner);
++
++static void __do_softirq_common(int need_rcu_bh_qs);
++
++void __do_softirq(void)
++{
++ __do_softirq_common(0);
++}
++
++void __init softirq_early_init(void)
++{
++ local_irq_lock_init(local_softirq_lock);
++}
++
++void local_bh_disable(void)
++{
++ migrate_disable();
++ current->softirq_nestcnt++;
++}
++EXPORT_SYMBOL(local_bh_disable);
++
++void local_bh_enable(void)
++{
++ if (WARN_ON(current->softirq_nestcnt == 0))
++ return;
++
++ if ((current->softirq_nestcnt == 1) &&
++ local_softirq_pending() &&
++ local_trylock(local_softirq_lock)) {
++
++ local_irq_disable();
++ if (local_softirq_pending())
++ __do_softirq();
++ local_irq_enable();
++ local_unlock(local_softirq_lock);
++ WARN_ON(current->softirq_nestcnt != 1);
++ }
++ current->softirq_nestcnt--;
++ migrate_enable();
++}
++EXPORT_SYMBOL(local_bh_enable);
++
++void local_bh_enable_ip(unsigned long ip)
++{
++ local_bh_enable();
++}
++EXPORT_SYMBOL(local_bh_enable_ip);
++
++/* For tracing */
++int notrace __in_softirq(void)
++{
++ if (__get_cpu_var(local_softirq_lock).owner == current)
++ return __get_cpu_var(local_softirq_lock).nestcnt;
++ return 0;
++}
++
++int in_serving_softirq(void)
++{
++ int res;
++
++ preempt_disable();
++ res = __get_cpu_var(local_softirq_runner) == current;
++ preempt_enable();
++ return res;
++}
++EXPORT_SYMBOL(in_serving_softirq);
++
++/*
++ * Called with bh and local interrupts disabled. For full RT cpu must
++ * be pinned.
++ */
++static void __do_softirq_common(int need_rcu_bh_qs)
++{
++ u32 pending = local_softirq_pending();
++ int cpu = smp_processor_id();
++
++ current->softirq_nestcnt++;
++
++ /* Reset the pending bitmask before enabling irqs */
++ set_softirq_pending(0);
++
++ __get_cpu_var(local_softirq_runner) = current;
++
++ lockdep_softirq_enter();
++
++ handle_pending_softirqs(pending, cpu, need_rcu_bh_qs);
++
++ pending = local_softirq_pending();
++ if (pending)
++ wakeup_softirqd();
++
++ lockdep_softirq_exit();
++ __get_cpu_var(local_softirq_runner) = NULL;
++
++ current->softirq_nestcnt--;
++}
++
++static int __thread_do_softirq(int cpu)
++{
++ /*
++ * Prevent the current cpu from going offline.
++ * pin_current_cpu() can reenable preemption and block on the
++ * hotplug mutex. When it returns, the current cpu is
++ * pinned. It might be the wrong one, but the offline check
++ * below catches that.
++ */
++ pin_current_cpu();
++ /*
++ * If called from ksoftirqd (cpu >= 0) we need to check
++ * whether we are on the wrong cpu due to cpu offlining. If
++ * called via thread_do_softirq() no action required.
++ */
++ if (cpu >= 0 && cpu_is_offline(cpu)) {
++ unpin_current_cpu();
++ return -1;
++ }
++ preempt_enable();
++ local_lock(local_softirq_lock);
++ local_irq_disable();
++ /*
++ * We cannot switch stacks on RT as we want to be able to
++ * schedule!
++ */
++ if (local_softirq_pending())
++ __do_softirq_common(cpu >= 0);
++ local_unlock(local_softirq_lock);
++ unpin_current_cpu();
++ preempt_disable();
++ local_irq_enable();
++ return 0;
++}
++
++/*
++ * Called from netif_rx_ni(). Preemption enabled.
++ */
++void thread_do_softirq(void)
++{
++ if (!in_serving_softirq()) {
++ preempt_disable();
++ __thread_do_softirq(-1);
++ preempt_enable();
++ }
++}
++
++static int ksoftirqd_do_softirq(int cpu)
++{
++ return __thread_do_softirq(cpu);
++}
+
-+config PREEMPT_RT_FULL
-+ bool "Fully Preemptible Kernel (RT)"
-+ depends on IRQ_FORCED_THREADING
-+ select PREEMPT_RT_BASE
-+ help
-+ All and everything
++static inline void local_bh_disable_nort(void) { }
++static inline void _local_bh_enable_nort(void) { }
+
- endchoice
-
-Index: linux-3.0/include/asm-generic/bug.h
-===================================================================
---- linux-3.0.orig/include/asm-generic/bug.h
-+++ linux-3.0/include/asm-generic/bug.h
-@@ -3,6 +3,10 @@
-
- #include <linux/compiler.h>
-
-+#ifndef __ASSEMBLY__
-+extern void __WARN_ON(const char *func, const char *file, const int line);
-+#endif /* __ASSEMBLY__ */
++static inline void ksoftirqd_set_sched_params(void)
++{
++ struct sched_param param = { .sched_priority = 1 };
+
- #ifdef CONFIG_BUG
-
- #ifdef CONFIG_GENERIC_BUG
-@@ -201,4 +205,18 @@ extern void warn_slowpath_null(const cha
- # define WARN_ON_SMP(x) ({0;})
- #endif
++ sched_setscheduler(current, SCHED_FIFO, ¶m);
++}
++
++static inline void ksoftirqd_clr_sched_params(void)
++{
++ struct sched_param param = { .sched_priority = 0 };
++
++ sched_setscheduler(current, SCHED_NORMAL, ¶m);
++}
++
++#endif /* PREEMPT_RT_FULL */
+ /*
+ * Enter an interrupt context.
+ */
+@@ -302,9 +562,9 @@ void irq_enter(void)
+ * Prevent raise_softirq from needlessly waking up ksoftirqd
+ * here, as softirq will be serviced on return from interrupt.
+ */
+- local_bh_disable();
++ local_bh_disable_nort();
+ tick_check_idle(cpu);
+- _local_bh_enable();
++ _local_bh_enable_nort();
+ }
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+# define BUG_ON_RT(c) BUG_ON(c)
-+# define BUG_ON_NONRT(c) do { } while (0)
-+# define WARN_ON_RT(condition) WARN_ON(condition)
-+# define WARN_ON_NONRT(condition) do { } while (0)
-+# define WARN_ON_ONCE_NONRT(condition) do { } while (0)
+ __irq_enter();
+@@ -313,6 +573,7 @@ void irq_enter(void)
+ #ifdef __ARCH_IRQ_EXIT_IRQS_DISABLED
+ static inline void invoke_softirq(void)
+ {
++#ifndef CONFIG_PREEMPT_RT_FULL
+ if (!force_irqthreads)
+ __do_softirq();
+ else {
+@@ -321,10 +582,14 @@ static inline void invoke_softirq(void)
+ wakeup_softirqd();
+ __local_bh_enable(SOFTIRQ_OFFSET);
+ }
+#else
-+# define BUG_ON_RT(c) do { } while (0)
-+# define BUG_ON_NONRT(c) BUG_ON(c)
-+# define WARN_ON_RT(condition) do { } while (0)
-+# define WARN_ON_NONRT(condition) WARN_ON(condition)
-+# define WARN_ON_ONCE_NONRT(condition) WARN_ON_ONCE(condition)
++ wakeup_softirqd();
+#endif
-+
- #endif
-Index: linux-3.0/include/linux/irqflags.h
-===================================================================
---- linux-3.0.orig/include/linux/irqflags.h
-+++ linux-3.0/include/linux/irqflags.h
-@@ -25,8 +25,6 @@
- # define trace_softirqs_enabled(p) ((p)->softirqs_enabled)
- # define trace_hardirq_enter() do { current->hardirq_context++; } while (0)
- # define trace_hardirq_exit() do { current->hardirq_context--; } while (0)
--# define lockdep_softirq_enter() do { current->softirq_context++; } while (0)
--# define lockdep_softirq_exit() do { current->softirq_context--; } while (0)
- # define INIT_TRACE_IRQFLAGS .softirqs_enabled = 1,
+ }
#else
- # define trace_hardirqs_on() do { } while (0)
-@@ -39,9 +37,15 @@
- # define trace_softirqs_enabled(p) 0
- # define trace_hardirq_enter() do { } while (0)
- # define trace_hardirq_exit() do { } while (0)
-+# define INIT_TRACE_IRQFLAGS
-+#endif
-+
-+#if defined(CONFIG_TRACE_IRQFLAGS) && !defined(CONFIG_PREEMPT_RT_FULL)
-+# define lockdep_softirq_enter() do { current->softirq_context++; } while (0)
-+# define lockdep_softirq_exit() do { current->softirq_context--; } while (0)
+ static inline void invoke_softirq(void)
+ {
++#ifndef CONFIG_PREEMPT_RT_FULL
+ if (!force_irqthreads)
+ do_softirq();
+ else {
+@@ -333,6 +598,9 @@ static inline void invoke_softirq(void)
+ wakeup_softirqd();
+ __local_bh_enable(SOFTIRQ_OFFSET);
+ }
+#else
- # define lockdep_softirq_enter() do { } while (0)
- # define lockdep_softirq_exit() do { } while (0)
--# define INIT_TRACE_IRQFLAGS
++ wakeup_softirqd();
++#endif
+ }
#endif
- #if defined(CONFIG_IRQSOFF_TRACER) || \
-@@ -147,4 +151,23 @@
+@@ -353,7 +621,7 @@ void irq_exit(void)
+ if (idle_cpu(smp_processor_id()) && !in_interrupt() && !need_resched())
+ tick_nohz_stop_sched_tick(0);
+ #endif
+- preempt_enable_no_resched();
++ __preempt_enable_no_resched();
+ }
- #endif /* CONFIG_TRACE_IRQFLAGS_SUPPORT */
+ /*
+@@ -739,29 +1007,21 @@ void __init softirq_init(void)
-+/*
-+ * local_irq* variants depending on RT/!RT
-+ */
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# define local_irq_disable_nort() do { } while (0)
-+# define local_irq_enable_nort() do { } while (0)
-+# define local_irq_save_nort(flags) do { local_save_flags(flags); } while (0)
-+# define local_irq_restore_nort(flags) do { (void)(flags); } while (0)
-+# define local_irq_disable_rt() local_irq_disable()
-+# define local_irq_enable_rt() local_irq_enable()
-+#else
-+# define local_irq_disable_nort() local_irq_disable()
-+# define local_irq_enable_nort() local_irq_enable()
-+# define local_irq_save_nort(flags) local_irq_save(flags)
-+# define local_irq_restore_nort(flags) local_irq_restore(flags)
-+# define local_irq_disable_rt() do { } while (0)
-+# define local_irq_enable_rt() do { } while (0)
-+#endif
+ static int run_ksoftirqd(void * __bind_cpu)
+ {
++ ksoftirqd_set_sched_params();
+
- #endif
-Index: linux-3.0/drivers/ata/libata-sff.c
-===================================================================
---- linux-3.0.orig/drivers/ata/libata-sff.c
-+++ linux-3.0/drivers/ata/libata-sff.c
-@@ -677,9 +677,9 @@ unsigned int ata_sff_data_xfer_noirq(str
- unsigned long flags;
- unsigned int consumed;
+ set_current_state(TASK_INTERRUPTIBLE);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- consumed = ata_sff_data_xfer(dev, buf, buflen, rw);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ while (!kthread_should_stop()) {
+ preempt_disable();
+- if (!local_softirq_pending()) {
+- preempt_enable_no_resched();
+- schedule();
+- preempt_disable();
+- }
++ if (!local_softirq_pending())
++ schedule_preempt_disabled();
- return consumed;
- }
-@@ -718,7 +718,7 @@ static void ata_pio_sector(struct ata_qu
- unsigned long flags;
+ __set_current_state(TASK_RUNNING);
- /* FIXME: use a bounce buffer */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- buf = kmap_atomic(page, KM_IRQ0);
+ while (local_softirq_pending()) {
+- /* Preempt disable stops cpu going offline.
+- If already offline, we'll be on wrong CPU:
+- don't process */
+- if (cpu_is_offline((long)__bind_cpu))
++ if (ksoftirqd_do_softirq((long) __bind_cpu))
+ goto wait_to_die;
+- local_irq_disable();
+- if (local_softirq_pending())
+- __do_softirq();
+- local_irq_enable();
+- preempt_enable_no_resched();
++ __preempt_enable_no_resched();
+ cond_resched();
+ preempt_disable();
+ rcu_note_context_switch((long)__bind_cpu);
+@@ -774,6 +1034,7 @@ static int run_ksoftirqd(void * __bind_cpu)
- /* do the actual data transfer */
-@@ -726,7 +726,7 @@ static void ata_pio_sector(struct ata_qu
- do_write);
+ wait_to_die:
+ preempt_enable();
++ ksoftirqd_clr_sched_params();
+ /* Wait for kthread_stop */
+ set_current_state(TASK_INTERRUPTIBLE);
+ while (!kthread_should_stop()) {
+@@ -850,9 +1111,8 @@ static int __cpuinit cpu_callback(struct notifier_block *nfb,
+ int hotcpu = (unsigned long)hcpu;
+ struct task_struct *p;
- kunmap_atomic(buf, KM_IRQ0);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- } else {
- buf = page_address(page);
- ap->ops->sff_data_xfer(qc->dev, buf + offset, qc->sect_size,
-@@ -863,7 +863,7 @@ next_sg:
- unsigned long flags;
+- switch (action) {
++ switch (action & ~CPU_TASKS_FROZEN) {
+ case CPU_UP_PREPARE:
+- case CPU_UP_PREPARE_FROZEN:
+ p = kthread_create_on_node(run_ksoftirqd,
+ hcpu,
+ cpu_to_node(hotcpu),
+@@ -865,19 +1125,16 @@ static int __cpuinit cpu_callback(struct notifier_block *nfb,
+ per_cpu(ksoftirqd, hotcpu) = p;
+ break;
+ case CPU_ONLINE:
+- case CPU_ONLINE_FROZEN:
+ wake_up_process(per_cpu(ksoftirqd, hotcpu));
+ break;
+ #ifdef CONFIG_HOTPLUG_CPU
+ case CPU_UP_CANCELED:
+- case CPU_UP_CANCELED_FROZEN:
+ if (!per_cpu(ksoftirqd, hotcpu))
+ break;
+ /* Unbind so it can run. Fall thru. */
+ kthread_bind(per_cpu(ksoftirqd, hotcpu),
+ cpumask_any(cpu_online_mask));
+- case CPU_DEAD:
+- case CPU_DEAD_FROZEN: {
++ case CPU_POST_DEAD: {
+ static const struct sched_param param = {
+ .sched_priority = MAX_RT_PRIO-1
+ };
+diff --git a/kernel/spinlock.c b/kernel/spinlock.c
+index be6517f..f2887e5 100644
+--- a/kernel/spinlock.c
++++ b/kernel/spinlock.c
+@@ -110,8 +110,11 @@ void __lockfunc __raw_##op##_lock_bh(locktype##_t *lock) \
+ * __[spin|read|write]_lock_bh()
+ */
+ BUILD_LOCK_OPS(spin, raw_spinlock);
++
++#ifndef CONFIG_PREEMPT_RT_FULL
+ BUILD_LOCK_OPS(read, rwlock);
+ BUILD_LOCK_OPS(write, rwlock);
++#endif
- /* FIXME: use bounce buffer */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- buf = kmap_atomic(page, KM_IRQ0);
+ #endif
- /* do the actual data transfer */
-@@ -871,7 +871,7 @@ next_sg:
- count, rw);
+@@ -195,6 +198,8 @@ void __lockfunc _raw_spin_unlock_bh(raw_spinlock_t *lock)
+ EXPORT_SYMBOL(_raw_spin_unlock_bh);
+ #endif
- kunmap_atomic(buf, KM_IRQ0);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- } else {
- buf = page_address(page);
- consumed = ap->ops->sff_data_xfer(dev, buf + offset,
-Index: linux-3.0/drivers/ide/alim15x3.c
-===================================================================
---- linux-3.0.orig/drivers/ide/alim15x3.c
-+++ linux-3.0/drivers/ide/alim15x3.c
-@@ -234,7 +234,7 @@ static int init_chipset_ali15x3(struct p
++#ifndef CONFIG_PREEMPT_RT_FULL
++
+ #ifndef CONFIG_INLINE_READ_TRYLOCK
+ int __lockfunc _raw_read_trylock(rwlock_t *lock)
+ {
+@@ -339,6 +344,8 @@ void __lockfunc _raw_write_unlock_bh(rwlock_t *lock)
+ EXPORT_SYMBOL(_raw_write_unlock_bh);
+ #endif
- isa_dev = pci_get_device(PCI_VENDOR_ID_AL, PCI_DEVICE_ID_AL_M1533, NULL);
++#endif /* !PREEMPT_RT_FULL */
++
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
+ void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass)
+diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c
+index 0cae1cc..08e4370 100644
+--- a/kernel/stop_machine.c
++++ b/kernel/stop_machine.c
+@@ -29,12 +29,12 @@ struct cpu_stop_done {
+ atomic_t nr_todo; /* nr left to execute */
+ bool executed; /* actually executed? */
+ int ret; /* collected return value */
+- struct completion completion; /* fired if nr_todo reaches 0 */
++ struct task_struct *waiter; /* woken when nr_todo reaches 0 */
+ };
- if (m5229_revision < 0xC2) {
- /*
-@@ -325,7 +325,7 @@ out:
+ /* the actual stopper, one per every possible cpu, enabled on online cpus */
+ struct cpu_stopper {
+- spinlock_t lock;
++ raw_spinlock_t lock;
+ bool enabled; /* is this stopper enabled? */
+ struct list_head works; /* list of pending works */
+ struct task_struct *thread; /* stopper thread */
+@@ -46,7 +46,7 @@ static void cpu_stop_init_done(struct cpu_stop_done *done, unsigned int nr_todo)
+ {
+ memset(done, 0, sizeof(*done));
+ atomic_set(&done->nr_todo, nr_todo);
+- init_completion(&done->completion);
++ done->waiter = current;
+ }
+
+ /* signal completion unless @done is NULL */
+@@ -55,8 +55,10 @@ static void cpu_stop_signal_done(struct cpu_stop_done *done, bool executed)
+ if (done) {
+ if (executed)
+ done->executed = true;
+- if (atomic_dec_and_test(&done->nr_todo))
+- complete(&done->completion);
++ if (atomic_dec_and_test(&done->nr_todo)) {
++ wake_up_process(done->waiter);
++ done->waiter = NULL;
++ }
}
- pci_dev_put(north);
- pci_dev_put(isa_dev);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- return 0;
}
-Index: linux-3.0/drivers/ide/hpt366.c
-===================================================================
---- linux-3.0.orig/drivers/ide/hpt366.c
-+++ linux-3.0/drivers/ide/hpt366.c
-@@ -1241,7 +1241,7 @@ static int __devinit init_dma_hpt366(ide
+@@ -66,7 +68,7 @@ static void cpu_stop_queue_work(struct cpu_stopper *stopper,
+ {
+ unsigned long flags;
- dma_old = inb(base + 2);
+- spin_lock_irqsave(&stopper->lock, flags);
++ raw_spin_lock_irqsave(&stopper->lock, flags);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
+ if (stopper->enabled) {
+ list_add_tail(&work->list, &stopper->works);
+@@ -74,7 +76,23 @@ static void cpu_stop_queue_work(struct cpu_stopper *stopper,
+ } else
+ cpu_stop_signal_done(work->done, false);
- dma_new = dma_old;
- pci_read_config_byte(dev, hwif->channel ? 0x4b : 0x43, &masterdma);
-@@ -1252,7 +1252,7 @@ static int __devinit init_dma_hpt366(ide
- if (dma_new != dma_old)
- outb(dma_new, base + 2);
+- spin_unlock_irqrestore(&stopper->lock, flags);
++ raw_spin_unlock_irqrestore(&stopper->lock, flags);
++}
++
++static void wait_for_stop_done(struct cpu_stop_done *done)
++{
++ set_current_state(TASK_UNINTERRUPTIBLE);
++ while (atomic_read(&done->nr_todo)) {
++ schedule();
++ set_current_state(TASK_UNINTERRUPTIBLE);
++ }
++ /*
++ * We need to wait until cpu_stop_signal_done() has cleared
++ * done->waiter.
++ */
++ while (done->waiter)
++ cpu_relax();
++ set_current_state(TASK_RUNNING);
+ }
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ /**
+@@ -108,7 +126,7 @@ int stop_one_cpu(unsigned int cpu, cpu_stop_fn_t fn, void *arg)
- printk(KERN_INFO " %s: BM-DMA at 0x%04lx-0x%04lx\n",
- hwif->name, base, base + 7);
-Index: linux-3.0/drivers/ide/ide-io-std.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide-io-std.c
-+++ linux-3.0/drivers/ide/ide-io-std.c
-@@ -174,7 +174,7 @@ void ide_input_data(ide_drive_t *drive,
- unsigned long uninitialized_var(flags);
+ cpu_stop_init_done(&done, 1);
+ cpu_stop_queue_work(&per_cpu(cpu_stopper, cpu), &work);
+- wait_for_completion(&done.completion);
++ wait_for_stop_done(&done);
+ return done.executed ? done.ret : -ENOENT;
+ }
- if ((io_32bit & 2) && !mmio) {
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- ata_vlb_sync(io_ports->nsect_addr);
- }
+@@ -133,6 +151,7 @@ void stop_one_cpu_nowait(unsigned int cpu, cpu_stop_fn_t fn, void *arg,
+ }
-@@ -185,7 +185,7 @@ void ide_input_data(ide_drive_t *drive,
- insl(data_addr, buf, words);
+ DEFINE_MUTEX(stop_cpus_mutex);
++static DEFINE_MUTEX(stopper_lock);
+ /* static data for stop_cpus */
+ static DEFINE_PER_CPU(struct cpu_stop_work, stop_cpus_work);
- if ((io_32bit & 2) && !mmio)
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+@@ -152,17 +171,16 @@ int __stop_cpus(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg)
+ cpu_stop_init_done(&done, cpumask_weight(cpumask));
- if (((len + 1) & 3) < 2)
- return;
-@@ -218,7 +218,7 @@ void ide_output_data(ide_drive_t *drive,
- unsigned long uninitialized_var(flags);
+ /*
+- * Disable preemption while queueing to avoid getting
+- * preempted by a stopper which might wait for other stoppers
+- * to enter @fn which can lead to deadlock.
++ * Make sure that all work is queued on all cpus before we
++ * any of the cpus can execute it.
+ */
+- preempt_disable();
++ mutex_lock(&stopper_lock);
+ for_each_cpu(cpu, cpumask)
+ cpu_stop_queue_work(&per_cpu(cpu_stopper, cpu),
+ &per_cpu(stop_cpus_work, cpu));
+- preempt_enable();
++ mutex_unlock(&stopper_lock);
- if ((io_32bit & 2) && !mmio) {
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- ata_vlb_sync(io_ports->nsect_addr);
- }
+- wait_for_completion(&done.completion);
++ wait_for_stop_done(&done);
+ return done.executed ? done.ret : -ENOENT;
+ }
-@@ -229,7 +229,7 @@ void ide_output_data(ide_drive_t *drive,
- outsl(data_addr, buf, words);
+@@ -250,13 +268,13 @@ repeat:
+ }
- if ((io_32bit & 2) && !mmio)
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ work = NULL;
+- spin_lock_irq(&stopper->lock);
++ raw_spin_lock_irq(&stopper->lock);
+ if (!list_empty(&stopper->works)) {
+ work = list_first_entry(&stopper->works,
+ struct cpu_stop_work, list);
+ list_del_init(&work->list);
+ }
+- spin_unlock_irq(&stopper->lock);
++ raw_spin_unlock_irq(&stopper->lock);
- if (((len + 1) & 3) < 2)
- return;
-Index: linux-3.0/drivers/ide/ide-io.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide-io.c
-+++ linux-3.0/drivers/ide/ide-io.c
-@@ -659,7 +659,7 @@ void ide_timer_expiry (unsigned long dat
- /* disable_irq_nosync ?? */
- disable_irq(hwif->irq);
- /* local CPU only, as if we were handling an interrupt */
-- local_irq_disable();
-+ local_irq_disable_nort();
- if (hwif->polling) {
- startstop = handler(drive);
- } else if (drive_is_ready(drive)) {
-Index: linux-3.0/drivers/ide/ide-iops.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide-iops.c
-+++ linux-3.0/drivers/ide/ide-iops.c
-@@ -129,12 +129,12 @@ int __ide_wait_stat(ide_drive_t *drive,
- if ((stat & ATA_BUSY) == 0)
- break;
+ if (work) {
+ cpu_stop_fn_t fn = work->fn;
+@@ -266,6 +284,16 @@ repeat:
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- *rstat = stat;
- return -EBUSY;
- }
- }
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- }
- /*
- * Allow status to settle, then read it again.
-Index: linux-3.0/drivers/ide/ide-probe.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide-probe.c
-+++ linux-3.0/drivers/ide/ide-probe.c
-@@ -196,10 +196,10 @@ static void do_identify(ide_drive_t *dri
- int bswap = 1;
+ __set_current_state(TASK_RUNNING);
- /* local CPU only; some systems need this */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- /* read 512 bytes of id info */
- hwif->tp_ops->input_data(drive, NULL, id, SECTOR_SIZE);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
++ /*
++ * Wait until the stopper finished scheduling on all
++ * cpus
++ */
++ mutex_lock(&stopper_lock);
++ /*
++ * Let other cpu threads continue as well
++ */
++ mutex_unlock(&stopper_lock);
++
+ /* cpu stop callbacks are not allowed to sleep */
+ preempt_disable();
- drive->dev_flags |= IDE_DFLAG_ID_READ;
- #ifdef DEBUG
-Index: linux-3.0/drivers/ide/ide-taskfile.c
-===================================================================
---- linux-3.0.orig/drivers/ide/ide-taskfile.c
-+++ linux-3.0/drivers/ide/ide-taskfile.c
-@@ -250,7 +250,7 @@ void ide_pio_bytes(ide_drive_t *drive, s
+@@ -280,7 +308,13 @@ repeat:
+ kallsyms_lookup((unsigned long)fn, NULL, NULL, NULL,
+ ksym_buf), arg);
- page_is_high = PageHighMem(page);
- if (page_is_high)
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
++ /*
++ * Make sure that the wakeup and setting done->waiter
++ * to NULL is atomic.
++ */
++ local_irq_disable();
+ cpu_stop_signal_done(done, true);
++ local_irq_enable();
+ } else
+ schedule();
- buf = kmap_atomic(page, KM_BIO_SRC_IRQ) + offset;
+@@ -308,6 +342,7 @@ static int __cpuinit cpu_stop_cpu_callback(struct notifier_block *nfb,
+ if (IS_ERR(p))
+ return notifier_from_errno(PTR_ERR(p));
+ get_task_struct(p);
++ p->flags |= PF_STOMPER;
+ kthread_bind(p, cpu);
+ sched_set_stop_task(cpu, p);
+ stopper->thread = p;
+@@ -317,9 +352,9 @@ static int __cpuinit cpu_stop_cpu_callback(struct notifier_block *nfb,
+ /* strictly unnecessary, as first user will wake it */
+ wake_up_process(stopper->thread);
+ /* mark enabled */
+- spin_lock_irq(&stopper->lock);
++ raw_spin_lock_irq(&stopper->lock);
+ stopper->enabled = true;
+- spin_unlock_irq(&stopper->lock);
++ raw_spin_unlock_irq(&stopper->lock);
+ break;
-@@ -271,7 +271,7 @@ void ide_pio_bytes(ide_drive_t *drive, s
- kunmap_atomic(buf, KM_BIO_SRC_IRQ);
+ #ifdef CONFIG_HOTPLUG_CPU
+@@ -332,11 +367,11 @@ static int __cpuinit cpu_stop_cpu_callback(struct notifier_block *nfb,
+ /* kill the stopper */
+ kthread_stop(stopper->thread);
+ /* drain remaining works */
+- spin_lock_irq(&stopper->lock);
++ raw_spin_lock_irq(&stopper->lock);
+ list_for_each_entry(work, &stopper->works, list)
+ cpu_stop_signal_done(work->done, false);
+ stopper->enabled = false;
+- spin_unlock_irq(&stopper->lock);
++ raw_spin_unlock_irq(&stopper->lock);
+ /* release the stopper */
+ put_task_struct(stopper->thread);
+ stopper->thread = NULL;
+@@ -367,7 +402,7 @@ static int __init cpu_stop_init(void)
+ for_each_possible_cpu(cpu) {
+ struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu);
+
+- spin_lock_init(&stopper->lock);
++ raw_spin_lock_init(&stopper->lock);
+ INIT_LIST_HEAD(&stopper->works);
+ }
- if (page_is_high)
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c
+index e0980f0..cf52fda 100644
+--- a/kernel/time/clocksource.c
++++ b/kernel/time/clocksource.c
+@@ -186,6 +186,7 @@ static struct timer_list watchdog_timer;
+ static DECLARE_WORK(watchdog_work, clocksource_watchdog_work);
+ static DEFINE_SPINLOCK(watchdog_lock);
+ static int watchdog_running;
++static atomic_t watchdog_reset_pending;
- len -= nr_bytes;
- }
-@@ -414,7 +414,7 @@ static ide_startstop_t pre_task_out_intr
- }
+ static int clocksource_watchdog_kthread(void *data);
+ static void __clocksource_change_rating(struct clocksource *cs, int rating);
+@@ -247,12 +248,14 @@ static void clocksource_watchdog(unsigned long data)
+ struct clocksource *cs;
+ cycle_t csnow, wdnow;
+ int64_t wd_nsec, cs_nsec;
+- int next_cpu;
++ int next_cpu, reset_pending;
- if ((drive->dev_flags & IDE_DFLAG_UNMASK) == 0)
-- local_irq_disable();
-+ local_irq_disable_nort();
+ spin_lock(&watchdog_lock);
+ if (!watchdog_running)
+ goto out;
- ide_set_handler(drive, &task_pio_intr, WAIT_WORSTCASE);
++ reset_pending = atomic_read(&watchdog_reset_pending);
++
+ list_for_each_entry(cs, &watchdog_list, wd_list) {
-Index: linux-3.0/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
-===================================================================
---- linux-3.0.orig/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
-+++ linux-3.0/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
-@@ -794,7 +794,7 @@ void ipoib_mcast_restart_task(struct wor
+ /* Clocksource already marked unstable? */
+@@ -268,7 +271,8 @@ static void clocksource_watchdog(unsigned long data)
+ local_irq_enable();
- ipoib_mcast_stop_thread(dev, 0);
+ /* Clocksource initialized ? */
+- if (!(cs->flags & CLOCK_SOURCE_WATCHDOG)) {
++ if (!(cs->flags & CLOCK_SOURCE_WATCHDOG) ||
++ atomic_read(&watchdog_reset_pending)) {
+ cs->flags |= CLOCK_SOURCE_WATCHDOG;
+ cs->wd_last = wdnow;
+ cs->cs_last = csnow;
+@@ -283,8 +287,11 @@ static void clocksource_watchdog(unsigned long data)
+ cs->cs_last = csnow;
+ cs->wd_last = wdnow;
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- netif_addr_lock(dev);
- spin_lock(&priv->lock);
++ if (atomic_read(&watchdog_reset_pending))
++ continue;
++
+ /* Check the deviation from the watchdog clocksource. */
+- if (abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD) {
++ if ((abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD)) {
+ clocksource_unstable(cs, cs_nsec - wd_nsec);
+ continue;
+ }
+@@ -303,6 +310,13 @@ static void clocksource_watchdog(unsigned long data)
+ }
-@@ -876,7 +876,7 @@ void ipoib_mcast_restart_task(struct wor
+ /*
++ * We only clear the watchdog_reset_pending, when we did a
++ * full cycle through all clocksources.
++ */
++ if (reset_pending)
++ atomic_dec(&watchdog_reset_pending);
++
++ /*
+ * Cycle through CPUs to check if the CPUs stay synchronized
+ * to each other.
+ */
+@@ -344,23 +358,7 @@ static inline void clocksource_reset_watchdog(void)
- spin_unlock(&priv->lock);
- netif_addr_unlock(dev);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ static void clocksource_resume_watchdog(void)
+ {
+- unsigned long flags;
+-
+- /*
+- * We use trylock here to avoid a potential dead lock when
+- * kgdb calls this code after the kernel has been stopped with
+- * watchdog_lock held. When watchdog_lock is held we just
+- * return and accept, that the watchdog might trigger and mark
+- * the monitored clock source (usually TSC) unstable.
+- *
+- * This does not affect the other caller clocksource_resume()
+- * because at this point the kernel is UP, interrupts are
+- * disabled and nothing can hold watchdog_lock.
+- */
+- if (!spin_trylock_irqsave(&watchdog_lock, flags))
+- return;
+- clocksource_reset_watchdog();
+- spin_unlock_irqrestore(&watchdog_lock, flags);
++ atomic_inc(&watchdog_reset_pending);
+ }
- /* We have to cancel outside of the spinlock */
- list_for_each_entry_safe(mcast, tmcast, &remove_list, list) {
-Index: linux-3.0/drivers/input/gameport/gameport.c
-===================================================================
---- linux-3.0.orig/drivers/input/gameport/gameport.c
-+++ linux-3.0/drivers/input/gameport/gameport.c
-@@ -87,12 +87,12 @@ static int gameport_measure_speed(struct
- tx = 1 << 30;
+ static void clocksource_enqueue_watchdog(struct clocksource *cs)
+diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c
+index f6117a4..b510ba9 100644
+--- a/kernel/time/ntp.c
++++ b/kernel/time/ntp.c
+@@ -358,7 +358,7 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer)
+ {
+ enum hrtimer_restart res = HRTIMER_NORESTART;
- for(i = 0; i < 50; i++) {
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- GET_TIME(t1);
- for (t = 0; t < 50; t++) gameport_read(gameport);
- GET_TIME(t2);
- GET_TIME(t3);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- udelay(i * 10);
- if ((t = DELTA(t2,t1) - DELTA(t3,t2)) < tx) tx = t;
- }
-@@ -111,11 +111,11 @@ static int gameport_measure_speed(struct
- tx = 1 << 30;
+- write_seqlock(&xtime_lock);
++ raw_write_seqlock(&xtime_lock);
- for(i = 0; i < 50; i++) {
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- rdtscl(t1);
- for (t = 0; t < 50; t++) gameport_read(gameport);
- rdtscl(t2);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- udelay(i * 10);
- if (t2 - t1 < tx) tx = t2 - t1;
+ switch (time_state) {
+ case TIME_OK:
+@@ -388,7 +388,7 @@ static enum hrtimer_restart ntp_leap_second(struct hrtimer *timer)
+ break;
}
-Index: linux-3.0/arch/x86/include/asm/acpi.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/acpi.h
-+++ linux-3.0/arch/x86/include/asm/acpi.h
-@@ -51,8 +51,8 @@
-
- #define ACPI_ASM_MACROS
- #define BREAKPOINT3
--#define ACPI_DISABLE_IRQS() local_irq_disable()
--#define ACPI_ENABLE_IRQS() local_irq_enable()
-+#define ACPI_DISABLE_IRQS() local_irq_disable_nort()
-+#define ACPI_ENABLE_IRQS() local_irq_enable_nort()
- #define ACPI_FLUSH_CPU_CACHE() wbinvd()
- int __acpi_acquire_global_lock(unsigned int *lock);
-Index: linux-3.0/kernel/user.c
-===================================================================
---- linux-3.0.orig/kernel/user.c
-+++ linux-3.0/kernel/user.c
-@@ -129,11 +129,11 @@ void free_uid(struct user_struct *up)
- if (!up)
- return;
+- write_sequnlock(&xtime_lock);
++ raw_write_sequnlock(&xtime_lock);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- if (atomic_dec_and_lock(&up->__count, &uidhash_lock))
- free_user(up, flags);
- else
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ return res;
}
+@@ -663,7 +663,7 @@ int do_adjtimex(struct timex *txc)
- struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid)
-Index: linux-3.0/kernel/res_counter.c
-===================================================================
---- linux-3.0.orig/kernel/res_counter.c
-+++ linux-3.0/kernel/res_counter.c
-@@ -43,7 +43,7 @@ int res_counter_charge(struct res_counte
- struct res_counter *c, *u;
+ getnstimeofday(&ts);
- *limit_fail_at = NULL;
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- for (c = counter; c != NULL; c = c->parent) {
- spin_lock(&c->lock);
- ret = res_counter_charge_locked(c, val);
-@@ -62,7 +62,7 @@ undo:
- spin_unlock(&u->lock);
- }
- done:
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- return ret;
- }
+- write_seqlock_irq(&xtime_lock);
++ raw_write_seqlock_irq(&xtime_lock);
-@@ -79,13 +79,13 @@ void res_counter_uncharge(struct res_cou
- unsigned long flags;
- struct res_counter *c;
+ if (txc->modes & ADJ_ADJTIME) {
+ long save_adjust = time_adjust;
+@@ -705,7 +705,7 @@ int do_adjtimex(struct timex *txc)
+ /* fill PPS status fields */
+ pps_fill_timex(txc);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- for (c = counter; c != NULL; c = c->parent) {
- spin_lock(&c->lock);
- res_counter_uncharge_locked(c, val);
- spin_unlock(&c->lock);
- }
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- }
+- write_sequnlock_irq(&xtime_lock);
++ raw_write_sequnlock_irq(&xtime_lock);
+ txc->time.tv_sec = ts.tv_sec;
+ txc->time.tv_usec = ts.tv_nsec;
+@@ -903,7 +903,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts)
-Index: linux-3.0/drivers/usb/core/hcd.c
-===================================================================
---- linux-3.0.orig/drivers/usb/core/hcd.c
-+++ linux-3.0/drivers/usb/core/hcd.c
-@@ -2120,7 +2120,7 @@ irqreturn_t usb_hcd_irq (int irq, void *
- * when the first handler doesn't use it. So let's just
- * assume it's never used.
- */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
+ pts_norm = pps_normalize_ts(*phase_ts);
- if (unlikely(HCD_DEAD(hcd) || !HCD_HW_ACCESSIBLE(hcd))) {
- rc = IRQ_NONE;
-@@ -2133,7 +2133,7 @@ irqreturn_t usb_hcd_irq (int irq, void *
- rc = IRQ_HANDLED;
- }
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- return rc;
- }
- EXPORT_SYMBOL_GPL(usb_hcd_irq);
-Index: linux-3.0/drivers/tty/tty_ldisc.c
-===================================================================
---- linux-3.0.orig/drivers/tty/tty_ldisc.c
-+++ linux-3.0/drivers/tty/tty_ldisc.c
-@@ -70,7 +70,7 @@ static void put_ldisc(struct tty_ldisc *
- * We really want an "atomic_dec_and_lock_irqsave()",
- * but we don't have it, so this does it by hand.
- */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- if (atomic_dec_and_lock(&ld->users, &tty_ldisc_lock)) {
- struct tty_ldisc_ops *ldo = ld->ops;
+ /* clear the error bits, they will be set again if needed */
+ time_status &= ~(STA_PPSJITTER | STA_PPSWANDER | STA_PPSERROR);
+@@ -916,7 +916,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts)
+ * just start the frequency interval */
+ if (unlikely(pps_fbase.tv_sec == 0)) {
+ pps_fbase = *raw_ts;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+ return;
+ }
-@@ -81,7 +81,7 @@ static void put_ldisc(struct tty_ldisc *
- kfree(ld);
+@@ -931,7 +931,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts)
+ time_status |= STA_PPSJITTER;
+ /* restart the frequency calibration interval */
+ pps_fbase = *raw_ts;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+ pr_err("hardpps: PPSJITTER: bad pulse\n");
return;
}
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- wake_up(&tty_ldisc_idle);
+@@ -948,7 +948,7 @@ void hardpps(const struct timespec *phase_ts, const struct timespec *raw_ts)
+
+ hardpps_update_phase(pts_norm.nsec);
+
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
}
+ EXPORT_SYMBOL(hardpps);
-Index: linux-3.0/lib/scatterlist.c
-===================================================================
---- linux-3.0.orig/lib/scatterlist.c
-+++ linux-3.0/lib/scatterlist.c
-@@ -423,7 +423,7 @@ void sg_miter_stop(struct sg_mapping_ite
- flush_kernel_dcache_page(miter->page);
+diff --git a/kernel/time/tick-common.c b/kernel/time/tick-common.c
+index 119528d..174dba1 100644
+--- a/kernel/time/tick-common.c
++++ b/kernel/time/tick-common.c
+@@ -63,13 +63,13 @@ int tick_is_oneshot_available(void)
+ static void tick_periodic(int cpu)
+ {
+ if (tick_do_timer_cpu == cpu) {
+- write_seqlock(&xtime_lock);
++ raw_write_seqlock(&xtime_lock);
- if (miter->__flags & SG_MITER_ATOMIC) {
-- WARN_ON(!irqs_disabled());
-+ WARN_ON_NONRT(!irqs_disabled());
- kunmap_atomic(miter->addr, KM_BIO_SRC_IRQ);
- } else
- kunmap(miter->page);
-@@ -463,7 +463,7 @@ static size_t sg_copy_buffer(struct scat
+ /* Keep track of the next tick event */
+ tick_next_period = ktime_add(tick_next_period, tick_period);
+
+ do_timer(1);
+- write_sequnlock(&xtime_lock);
++ raw_write_sequnlock(&xtime_lock);
+ }
- sg_miter_start(&miter, sgl, nents, sg_flags);
+ update_process_times(user_mode(get_irq_regs()));
+diff --git a/kernel/time/tick-internal.h b/kernel/time/tick-internal.h
+index 1009b06..dbda970 100644
+--- a/kernel/time/tick-internal.h
++++ b/kernel/time/tick-internal.h
+@@ -143,4 +143,4 @@ static inline int tick_device_is_functional(struct clock_event_device *dev)
+ #endif
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
+ extern void do_timer(unsigned long ticks);
+-extern seqlock_t xtime_lock;
++extern raw_seqlock_t xtime_lock;
+diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c
+index d5097c4..411fabc 100644
+--- a/kernel/time/tick-sched.c
++++ b/kernel/time/tick-sched.c
+@@ -56,7 +56,7 @@ static void tick_do_update_jiffies64(ktime_t now)
+ return;
- while (sg_miter_next(&miter) && offset < buflen) {
- unsigned int len;
-@@ -480,7 +480,7 @@ static size_t sg_copy_buffer(struct scat
+ /* Reevalute with xtime_lock held */
+- write_seqlock(&xtime_lock);
++ raw_write_seqlock(&xtime_lock);
- sg_miter_stop(&miter);
+ delta = ktime_sub(now, last_jiffies_update);
+ if (delta.tv64 >= tick_period.tv64) {
+@@ -79,7 +79,7 @@ static void tick_do_update_jiffies64(ktime_t now)
+ /* Keep the tick_next_period variable up to date */
+ tick_next_period = ktime_add(last_jiffies_update, tick_period);
+ }
+- write_sequnlock(&xtime_lock);
++ raw_write_sequnlock(&xtime_lock);
+ }
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- return offset;
+ /*
+@@ -89,12 +89,12 @@ static ktime_t tick_init_jiffy_update(void)
+ {
+ ktime_t period;
+
+- write_seqlock(&xtime_lock);
++ raw_write_seqlock(&xtime_lock);
+ /* Did we start the jiffies update yet ? */
+ if (last_jiffies_update.tv64 == 0)
+ last_jiffies_update = tick_next_period;
+ period = last_jiffies_update;
+- write_sequnlock(&xtime_lock);
++ raw_write_sequnlock(&xtime_lock);
+ return period;
}
-Index: linux-3.0/net/mac80211/rx.c
-===================================================================
---- linux-3.0.orig/net/mac80211/rx.c
-+++ linux-3.0/net/mac80211/rx.c
-@@ -2853,7 +2853,7 @@ void ieee80211_rx(struct ieee80211_hw *h
- struct ieee80211_supported_band *sband;
- struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb);
+@@ -304,13 +304,7 @@ void tick_nohz_stop_sched_tick(int inidle)
+ goto end;
-- WARN_ON_ONCE(softirq_count() == 0);
-+ WARN_ON_ONCE_NONRT(softirq_count() == 0);
+ if (unlikely(local_softirq_pending() && cpu_online(cpu))) {
+- static int ratelimit;
+-
+- if (ratelimit < 10) {
+- printk(KERN_ERR "NOHZ: local_softirq_pending %02x\n",
+- (unsigned int) local_softirq_pending());
+- ratelimit++;
+- }
++ softirq_check_pending_idle();
+ goto end;
+ }
- if (WARN_ON(status->band < 0 ||
- status->band >= IEEE80211_NUM_BANDS))
-Index: linux-3.0/include/linux/page_cgroup.h
-===================================================================
---- linux-3.0.orig/include/linux/page_cgroup.h
-+++ linux-3.0/include/linux/page_cgroup.h
-@@ -30,6 +30,10 @@ enum {
- */
- struct page_cgroup {
- unsigned long flags;
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ spinlock_t pcg_lock;
-+ spinlock_t pcm_lock;
-+#endif
- struct mem_cgroup *mem_cgroup;
- struct list_head lru; /* per cgroup LRU list */
- };
-@@ -96,30 +100,54 @@ static inline void lock_page_cgroup(stru
- * Don't take this lock in IRQ context.
- * This lock is for pc->mem_cgroup, USED, CACHE, MIGRATION
+@@ -777,6 +771,7 @@ void tick_setup_sched_timer(void)
+ * Emulate tick processing via per-CPU hrtimers:
*/
-+#ifndef CONFIG_PREEMPT_RT_BASE
- bit_spin_lock(PCG_LOCK, &pc->flags);
-+#else
-+ spin_lock(&pc->pcg_lock);
-+#endif
- }
+ hrtimer_init(&ts->sched_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
++ ts->sched_timer.irqsafe = 1;
+ ts->sched_timer.function = tick_sched_timer;
- static inline void unlock_page_cgroup(struct page_cgroup *pc)
- {
-+#ifndef CONFIG_PREEMPT_RT_BASE
- bit_spin_unlock(PCG_LOCK, &pc->flags);
-+#else
-+ spin_unlock(&pc->pcg_lock);
-+#endif
- }
+ /* Get the next period (per cpu) */
+diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c
+index 342408c..bdbf452 100644
+--- a/kernel/time/timekeeping.c
++++ b/kernel/time/timekeeping.c
+@@ -139,8 +139,7 @@ static inline s64 timekeeping_get_ns_raw(void)
+ * This read-write spinlock protects us from races in SMP while
+ * playing with xtime.
+ */
+-__cacheline_aligned_in_smp DEFINE_SEQLOCK(xtime_lock);
+-
++__cacheline_aligned_in_smp DEFINE_RAW_SEQLOCK(xtime_lock);
- static inline void move_lock_page_cgroup(struct page_cgroup *pc,
- unsigned long *flags)
- {
-+#ifndef CONFIG_PREEMPT_RT_BASE
- /*
- * We know updates to pc->flags of page cache's stats are from both of
- * usual context or IRQ context. Disable IRQ to avoid deadlock.
- */
- local_irq_save(*flags);
- bit_spin_lock(PCG_MOVE_LOCK, &pc->flags);
-+#else
-+ spin_lock_irqsave(&pc->pcm_lock, *flags);
-+#endif
- }
+ /*
+ * The current time
+@@ -361,7 +360,7 @@ int do_settimeofday(const struct timespec *tv)
+ if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC)
+ return -EINVAL;
- static inline void move_unlock_page_cgroup(struct page_cgroup *pc,
- unsigned long *flags)
- {
-+#ifndef CONFIG_PREEMPT_RT_BASE
- bit_spin_unlock(PCG_MOVE_LOCK, &pc->flags);
- local_irq_restore(*flags);
-+#else
-+ spin_unlock_irqrestore(&pc->pcm_lock, *flags);
-+#endif
-+}
-+
-+static inline void page_cgroup_lock_init(struct page_cgroup *pc)
-+{
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ spin_lock_init(&pc->pcg_lock);
-+ spin_lock_init(&pc->pcm_lock);
-+#endif
- }
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
- #ifdef CONFIG_SPARSEMEM
-Index: linux-3.0/mm/page_cgroup.c
-===================================================================
---- linux-3.0.orig/mm/page_cgroup.c
-+++ linux-3.0/mm/page_cgroup.c
-@@ -17,6 +17,7 @@ static void __meminit init_page_cgroup(s
- set_page_cgroup_array_id(pc, id);
- pc->mem_cgroup = NULL;
- INIT_LIST_HEAD(&pc->lru);
-+ page_cgroup_lock_init(pc);
- }
- static unsigned long total_usage;
+ timekeeping_forward_now();
-Index: linux-3.0/fs/buffer.c
-===================================================================
---- linux-3.0.orig/fs/buffer.c
-+++ linux-3.0/fs/buffer.c
-@@ -328,8 +328,7 @@ static void end_buffer_async_read(struct
- * decide that the page is now completely done.
- */
- first = page_buffers(page);
-- local_irq_save(flags);
-- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
-+ flags = bh_uptodate_lock_irqsave(first);
- clear_buffer_async_read(bh);
- unlock_buffer(bh);
- tmp = bh;
-@@ -342,8 +341,7 @@ static void end_buffer_async_read(struct
- }
- tmp = tmp->b_this_page;
- } while (tmp != bh);
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-+ bh_uptodate_unlock_irqrestore(first, flags);
+@@ -377,7 +376,7 @@ int do_settimeofday(const struct timespec *tv)
+ update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
+ timekeeper.mult);
- /*
- * If none of the buffers had errors and they are all
-@@ -355,9 +353,7 @@ static void end_buffer_async_read(struct
- return;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
- still_busy:
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-- return;
-+ bh_uptodate_unlock_irqrestore(first, flags);
- }
+ /* signal hrtimers about time change */
+ clock_was_set();
+@@ -401,7 +400,7 @@ int timekeeping_inject_offset(struct timespec *ts)
+ if ((unsigned long)ts->tv_nsec >= NSEC_PER_SEC)
+ return -EINVAL;
- /*
-@@ -391,8 +387,7 @@ void end_buffer_async_write(struct buffe
- }
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
- first = page_buffers(page);
-- local_irq_save(flags);
-- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
-+ flags = bh_uptodate_lock_irqsave(first);
+ timekeeping_forward_now();
- clear_buffer_async_write(bh);
- unlock_buffer(bh);
-@@ -404,15 +399,12 @@ void end_buffer_async_write(struct buffe
- }
- tmp = tmp->b_this_page;
- }
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-+ bh_uptodate_unlock_irqrestore(first, flags);
- end_page_writeback(page);
- return;
+@@ -414,7 +413,7 @@ int timekeeping_inject_offset(struct timespec *ts)
+ update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
+ timekeeper.mult);
- still_busy:
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-- return;
-+ bh_uptodate_unlock_irqrestore(first, flags);
- }
- EXPORT_SYMBOL(end_buffer_async_write);
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
-@@ -3220,6 +3212,7 @@ struct buffer_head *alloc_buffer_head(gf
- struct buffer_head *ret = kmem_cache_zalloc(bh_cachep, gfp_flags);
- if (ret) {
- INIT_LIST_HEAD(&ret->b_assoc_buffers);
-+ buffer_head_init_locks(ret);
- preempt_disable();
- __this_cpu_inc(bh_accounting.nr);
- recalc_bh_state();
-Index: linux-3.0/fs/ntfs/aops.c
-===================================================================
---- linux-3.0.orig/fs/ntfs/aops.c
-+++ linux-3.0/fs/ntfs/aops.c
-@@ -108,8 +108,7 @@ static void ntfs_end_buffer_async_read(s
- "0x%llx.", (unsigned long long)bh->b_blocknr);
- }
- first = page_buffers(page);
-- local_irq_save(flags);
-- bit_spin_lock(BH_Uptodate_Lock, &first->b_state);
-+ flags = bh_uptodate_lock_irqsave(first);
- clear_buffer_async_read(bh);
- unlock_buffer(bh);
- tmp = bh;
-@@ -124,8 +123,7 @@ static void ntfs_end_buffer_async_read(s
- }
- tmp = tmp->b_this_page;
- } while (tmp != bh);
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-+ bh_uptodate_unlock_irqrestore(first, flags);
- /*
- * If none of the buffers had errors then we can set the page uptodate,
- * but we first have to perform the post read mst fixups, if the
-@@ -146,13 +144,13 @@ static void ntfs_end_buffer_async_read(s
- recs = PAGE_CACHE_SIZE / rec_size;
- /* Should have been verified before we got here... */
- BUG_ON(!recs);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- kaddr = kmap_atomic(page, KM_BIO_SRC_IRQ);
- for (i = 0; i < recs; i++)
- post_read_mst_fixup((NTFS_RECORD*)(kaddr +
- i * rec_size), rec_size);
- kunmap_atomic(kaddr, KM_BIO_SRC_IRQ);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- flush_dcache_page(page);
- if (likely(page_uptodate && !PageError(page)))
- SetPageUptodate(page);
-@@ -160,9 +158,7 @@ static void ntfs_end_buffer_async_read(s
- unlock_page(page);
- return;
- still_busy:
-- bit_spin_unlock(BH_Uptodate_Lock, &first->b_state);
-- local_irq_restore(flags);
-- return;
-+ bh_uptodate_unlock_irqrestore(first, flags);
+ /* signal hrtimers about time change */
+ clock_was_set();
+@@ -568,7 +567,7 @@ void __init timekeeping_init(void)
+ read_persistent_clock(&now);
+ read_boot_clock(&boot);
+
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
+
+ ntp_init();
+
+@@ -589,7 +588,7 @@ void __init timekeeping_init(void)
+ -boot.tv_sec, -boot.tv_nsec);
+ total_sleep_time.tv_sec = 0;
+ total_sleep_time.tv_nsec = 0;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
}
- /**
-Index: linux-3.0/include/linux/buffer_head.h
-===================================================================
---- linux-3.0.orig/include/linux/buffer_head.h
-+++ linux-3.0/include/linux/buffer_head.h
-@@ -72,8 +72,52 @@ struct buffer_head {
- struct address_space *b_assoc_map; /* mapping this buffer is
- associated with */
- atomic_t b_count; /* users using this buffer_head */
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ spinlock_t b_uptodate_lock;
-+#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \
-+ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE)
-+ spinlock_t b_state_lock;
-+ spinlock_t b_journal_head_lock;
-+#endif
-+#endif
- };
+ /* time in seconds when suspend began */
+@@ -630,7 +629,7 @@ void timekeeping_inject_sleeptime(struct timespec *delta)
+ if (!(ts.tv_sec == 0 && ts.tv_nsec == 0))
+ return;
-+static inline unsigned long bh_uptodate_lock_irqsave(struct buffer_head *bh)
-+{
-+ unsigned long flags;
-+
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ local_irq_save(flags);
-+ bit_spin_lock(BH_Uptodate_Lock, &bh->b_state);
-+#else
-+ spin_lock_irqsave(&bh->b_uptodate_lock, flags);
-+#endif
-+ return flags;
-+}
-+
-+static inline void
-+bh_uptodate_unlock_irqrestore(struct buffer_head *bh, unsigned long flags)
-+{
-+#ifndef CONFIG_PREEMPT_RT_BASE
-+ bit_spin_unlock(BH_Uptodate_Lock, &bh->b_state);
-+ local_irq_restore(flags);
-+#else
-+ spin_unlock_irqrestore(&bh->b_uptodate_lock, flags);
-+#endif
-+}
-+
-+static inline void buffer_head_init_locks(struct buffer_head *bh)
-+{
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ spin_lock_init(&bh->b_uptodate_lock);
-+#if defined(CONFIG_JBD) || defined(CONFIG_JBD_MODULE) || \
-+ defined(CONFIG_JBD2) || defined(CONFIG_JBD2_MODULE)
-+ spin_lock_init(&bh->b_state_lock);
-+ spin_lock_init(&bh->b_journal_head_lock);
-+#endif
-+#endif
-+}
-+
- /*
- * macro tricks to expand the set_buffer_foo(), clear_buffer_foo()
- * and buffer_foo() functions.
-Index: linux-3.0/lib/Kconfig.debug
-===================================================================
---- linux-3.0.orig/lib/Kconfig.debug
-+++ linux-3.0/lib/Kconfig.debug
-@@ -62,6 +62,28 @@ config MAGIC_SYSRQ
- keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
- unless you really know what this hack does.
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
+ timekeeping_forward_now();
-+config MAGIC_SYSRQ_FORCE_PRINTK
-+ bool "Force printk from Magic SysRq"
-+ depends on MAGIC_SYSRQ && PREEMPT_RT_FULL
-+ default n
-+ help
-+ Allow the output from Magic SysRq to be output immediately, even if
-+ this causes large latencies. This can cause performance problems
-+ for real-time processes.
-+
-+ If PREEMPT_RT_FULL, printk() will not try to acquire the console lock
-+ when interrupts or preemption are disabled. If the console lock is
-+ not acquired the printk() output will be buffered, but will not be
-+ output immediately. Some drivers call into the Magic SysRq code
-+ with interrupts or preemption disabled, so the output of Magic SysRq
-+ will be buffered instead of printing immediately if this option is
-+ not selected.
-+
-+ Even with this option selected, Magic SysRq output will be delayed
-+ if the attempt to acquire the console lock fails.
-+
-+ Don't say Y unless you really know what this hack does.
-+
- config MAGIC_SYSRQ_DEFAULT_MASK
- hex "Default mask for Magic SysRq keys on the console"
- depends on MAGIC_SYSRQ
-@@ -151,7 +173,7 @@ config DEBUG_KERNEL
+ __timekeeping_inject_sleeptime(delta);
+@@ -640,7 +639,7 @@ void timekeeping_inject_sleeptime(struct timespec *delta)
+ update_vsyscall(&xtime, &wall_to_monotonic, timekeeper.clock,
+ timekeeper.mult);
- config DEBUG_SHIRQ
- bool "Debug shared IRQ handlers"
-- depends on DEBUG_KERNEL && GENERIC_HARDIRQS
-+ depends on DEBUG_KERNEL && GENERIC_HARDIRQS && !PREEMPT_RT_BASE
- help
- Enable this to generate a spurious interrupt as soon as a shared
- interrupt handler is registered, and just before one is deregistered.
-Index: linux-3.0/kernel/irq/handle.c
-===================================================================
---- linux-3.0.orig/kernel/irq/handle.c
-+++ linux-3.0/kernel/irq/handle.c
-@@ -156,8 +156,11 @@ handle_irq_event_percpu(struct irq_desc
- action = action->next;
- } while (action);
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+ /* FIXME: Can we unbreak that ? */
- if (random & IRQF_SAMPLE_RANDOM)
- add_interrupt_randomness(irq);
-+#endif
+ /* signal hrtimers about time change */
+ clock_was_set();
+@@ -663,7 +662,7 @@ static void timekeeping_resume(void)
- if (!noirqdebug)
- note_interrupt(irq, desc, retval);
-Index: linux-3.0/kernel/irq/spurious.c
-===================================================================
---- linux-3.0.orig/kernel/irq/spurious.c
-+++ linux-3.0/kernel/irq/spurious.c
-@@ -339,6 +339,11 @@ MODULE_PARM_DESC(noirqdebug, "Disable ir
+ clocksource_resume();
- static int __init irqfixup_setup(char *str)
- {
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ printk(KERN_WARNING "irqfixup boot option not supported "
-+ "w/ CONFIG_PREEMPT_RT\n");
-+ return 1;
-+#endif
- irqfixup = 1;
- printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n");
- printk(KERN_WARNING "This may impact system performance.\n");
-@@ -351,6 +356,11 @@ module_param(irqfixup, int, 0644);
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
- static int __init irqpoll_setup(char *str)
- {
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ printk(KERN_WARNING "irqpoll boot option not supported "
-+ "w/ CONFIG_PREEMPT_RT\n");
-+ return 1;
-+#endif
- irqfixup = 2;
- printk(KERN_WARNING "Misrouted IRQ fixup and polling support "
- "enabled\n");
-Index: linux-3.0/kernel/irq/manage.c
-===================================================================
---- linux-3.0.orig/kernel/irq/manage.c
-+++ linux-3.0/kernel/irq/manage.c
-@@ -18,6 +18,7 @@
- #include "internals.h"
+ if (timespec_compare(&ts, &timekeeping_suspend_time) > 0) {
+ ts = timespec_sub(ts, timekeeping_suspend_time);
+@@ -673,7 +672,7 @@ static void timekeeping_resume(void)
+ timekeeper.clock->cycle_last = timekeeper.clock->read(timekeeper.clock);
+ timekeeper.ntp_error = 0;
+ timekeeping_suspended = 0;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
- #ifdef CONFIG_IRQ_FORCED_THREADING
-+# ifndef CONFIG_PREEMPT_RT_BASE
- __read_mostly bool force_irqthreads;
+ touch_softlockup_watchdog();
- static int __init setup_forced_irqthreads(char *arg)
-@@ -26,6 +27,7 @@ static int __init setup_forced_irqthread
- return 0;
+@@ -689,10 +688,10 @@ static int timekeeping_suspend(void)
+
+ read_persistent_clock(&timekeeping_suspend_time);
+
+- write_seqlock_irqsave(&xtime_lock, flags);
++ raw_write_seqlock_irqsave(&xtime_lock, flags);
+ timekeeping_forward_now();
+ timekeeping_suspended = 1;
+- write_sequnlock_irqrestore(&xtime_lock, flags);
++ raw_write_sequnlock_irqrestore(&xtime_lock, flags);
+
+ clockevents_notify(CLOCK_EVT_NOTIFY_SUSPEND, NULL);
+ clocksource_suspend();
+@@ -1121,7 +1120,7 @@ ktime_t ktime_get_monotonic_offset(void)
+ */
+ void xtime_update(unsigned long ticks)
+ {
+- write_seqlock(&xtime_lock);
++ raw_write_seqlock(&xtime_lock);
+ do_timer(ticks);
+- write_sequnlock(&xtime_lock);
++ raw_write_sequnlock(&xtime_lock);
}
- early_param("threadirqs", setup_forced_irqthreads);
-+# endif
- #endif
+diff --git a/kernel/time/timer_stats.c b/kernel/time/timer_stats.c
+index a5d0a3a..0b537f2 100644
+--- a/kernel/time/timer_stats.c
++++ b/kernel/time/timer_stats.c
+@@ -81,7 +81,7 @@ struct entry {
+ /*
+ * Spinlock protecting the tables - not taken during lookup:
+ */
+-static DEFINE_SPINLOCK(table_lock);
++static DEFINE_RAW_SPINLOCK(table_lock);
- /**
-Index: linux-3.0/drivers/net/atl1c/atl1c_main.c
-===================================================================
---- linux-3.0.orig/drivers/net/atl1c/atl1c_main.c
-+++ linux-3.0/drivers/net/atl1c/atl1c_main.c
-@@ -2218,11 +2218,8 @@ static netdev_tx_t atl1c_xmit_frame(stru
+ /*
+ * Per-CPU lookup locks for fast hash lookup:
+@@ -188,7 +188,7 @@ static struct entry *tstat_lookup(struct entry *entry, char *comm)
+ prev = NULL;
+ curr = *head;
+
+- spin_lock(&table_lock);
++ raw_spin_lock(&table_lock);
+ /*
+ * Make sure we have not raced with another CPU:
+ */
+@@ -215,7 +215,7 @@ static struct entry *tstat_lookup(struct entry *entry, char *comm)
+ *head = curr;
}
+ out_unlock:
+- spin_unlock(&table_lock);
++ raw_spin_unlock(&table_lock);
- tpd_req = atl1c_cal_tpd_req(skb);
-- if (!spin_trylock_irqsave(&adapter->tx_lock, flags)) {
-- if (netif_msg_pktdata(adapter))
-- dev_info(&adapter->pdev->dev, "tx locked\n");
-- return NETDEV_TX_LOCKED;
-- }
-+ spin_lock_irqsave(&adapter->tx_lock, flags);
+ return curr;
+ }
+diff --git a/kernel/timer.c b/kernel/timer.c
+index 8cff361..937799f 100644
+--- a/kernel/timer.c
++++ b/kernel/timer.c
+@@ -75,6 +75,7 @@ struct tvec_root {
+ struct tvec_base {
+ spinlock_t lock;
+ struct timer_list *running_timer;
++ wait_queue_head_t wait_for_running_timer;
+ unsigned long timer_jiffies;
+ unsigned long next_timer;
+ struct tvec_root tv1;
+@@ -653,6 +654,36 @@ static struct tvec_base *lock_timer_base(struct timer_list *timer,
+ }
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
++static inline struct tvec_base *switch_timer_base(struct timer_list *timer,
++ struct tvec_base *old,
++ struct tvec_base *new)
++{
++ /* See the comment in lock_timer_base() */
++ timer_set_base(timer, NULL);
++ spin_unlock(&old->lock);
++ spin_lock(&new->lock);
++ timer_set_base(timer, new);
++ return new;
++}
++#else
++static inline struct tvec_base *switch_timer_base(struct timer_list *timer,
++ struct tvec_base *old,
++ struct tvec_base *new)
++{
++ /*
++ * We cannot do the above because we might be preempted and
++ * then the preempter would see NULL and loop forever.
++ */
++ if (spin_trylock(&new->lock)) {
++ timer_set_base(timer, new);
++ spin_unlock(&old->lock);
++ return new;
++ }
++ return old;
++}
++#endif
+
- if (skb->mark == 0x01)
- type = atl1c_trans_high;
- else
-Index: linux-3.0/drivers/net/atl1e/atl1e_main.c
-===================================================================
---- linux-3.0.orig/drivers/net/atl1e/atl1e_main.c
-+++ linux-3.0/drivers/net/atl1e/atl1e_main.c
-@@ -1800,8 +1800,7 @@ static netdev_tx_t atl1e_xmit_frame(stru
- return NETDEV_TX_OK;
- }
- tpd_req = atl1e_cal_tdp_req(skb);
-- if (!spin_trylock_irqsave(&adapter->tx_lock, flags))
-- return NETDEV_TX_LOCKED;
-+ spin_lock_irqsave(&adapter->tx_lock, flags);
-
- if (atl1e_tpd_avail(adapter) < tpd_req) {
- /* no enough descriptor, just stop queue */
-Index: linux-3.0/drivers/net/chelsio/sge.c
-===================================================================
---- linux-3.0.orig/drivers/net/chelsio/sge.c
-+++ linux-3.0/drivers/net/chelsio/sge.c
-@@ -1679,8 +1679,7 @@ static int t1_sge_tx(struct sk_buff *skb
- struct cmdQ *q = &sge->cmdQ[qid];
- unsigned int credits, pidx, genbit, count, use_sched_skb = 0;
-
-- if (!spin_trylock(&q->lock))
-- return NETDEV_TX_LOCKED;
-+ spin_lock(&q->lock);
-
- reclaim_completed_tx(sge, q);
-
-Index: linux-3.0/drivers/net/rionet.c
-===================================================================
---- linux-3.0.orig/drivers/net/rionet.c
-+++ linux-3.0/drivers/net/rionet.c
-@@ -176,11 +176,7 @@ static int rionet_start_xmit(struct sk_b
- u16 destid;
- unsigned long flags;
+ static inline int
+ __mod_timer(struct timer_list *timer, unsigned long expires,
+ bool pending_only, int pinned)
+@@ -679,12 +710,15 @@ __mod_timer(struct timer_list *timer, unsigned long expires,
-- local_irq_save(flags);
-- if (!spin_trylock(&rnet->tx_lock)) {
-- local_irq_restore(flags);
-- return NETDEV_TX_LOCKED;
-- }
-+ spin_lock_irqsave(&rnet->tx_lock, flags);
+ debug_activate(timer, expires);
- if ((rnet->tx_cnt + 1) > RIONET_TX_RING_SIZE) {
- netif_stop_queue(ndev);
-Index: linux-3.0/drivers/net/s2io.c
-===================================================================
---- linux-3.0.orig/drivers/net/s2io.c
-+++ linux-3.0/drivers/net/s2io.c
-@@ -4142,12 +4142,7 @@ static netdev_tx_t s2io_xmit(struct sk_b
- [skb->priority & (MAX_TX_FIFOS - 1)];
- fifo = &mac_control->fifos[queue];
++ preempt_disable_rt();
+ cpu = smp_processor_id();
-- if (do_spin_lock)
-- spin_lock_irqsave(&fifo->tx_lock, flags);
-- else {
-- if (unlikely(!spin_trylock_irqsave(&fifo->tx_lock, flags)))
-- return NETDEV_TX_LOCKED;
-- }
-+ spin_lock_irqsave(&fifo->tx_lock, flags);
+ #if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
+ if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu))
+ cpu = get_nohz_timer_target();
+ #endif
++ preempt_enable_rt();
++
+ new_base = per_cpu(tvec_bases, cpu);
- if (sp->config.multiq) {
- if (__netif_subqueue_stopped(dev, fifo->fifo_no)) {
-Index: linux-3.0/drivers/net/sungem.c
-===================================================================
---- linux-3.0.orig/drivers/net/sungem.c
-+++ linux-3.0/drivers/net/sungem.c
-@@ -1013,10 +1013,8 @@ static netdev_tx_t gem_start_xmit(struct
- (csum_stuff_off << 21));
+ if (base != new_base) {
+@@ -695,14 +729,8 @@ __mod_timer(struct timer_list *timer, unsigned long expires,
+ * handler yet has not finished. This also guarantees that
+ * the timer is serialized wrt itself.
+ */
+- if (likely(base->running_timer != timer)) {
+- /* See the comment in lock_timer_base() */
+- timer_set_base(timer, NULL);
+- spin_unlock(&base->lock);
+- base = new_base;
+- spin_lock(&base->lock);
+- timer_set_base(timer, base);
+- }
++ if (likely(base->running_timer != timer))
++ base = switch_timer_base(timer, base, new_base);
}
-- if (!spin_trylock_irqsave(&gp->tx_lock, flags)) {
-- /* Tell upper layer to requeue */
-- return NETDEV_TX_LOCKED;
-- }
-+ spin_lock_irqsave(&gp->tx_lock, flags);
-+
- /* We raced with gem_do_stop() */
- if (!gp->running) {
- spin_unlock_irqrestore(&gp->tx_lock, flags);
-Index: linux-3.0/drivers/net/tehuti.c
-===================================================================
---- linux-3.0.orig/drivers/net/tehuti.c
-+++ linux-3.0/drivers/net/tehuti.c
-@@ -1627,13 +1627,8 @@ static netdev_tx_t bdx_tx_transmit(struc
- unsigned long flags;
+ timer->expires = expires;
+@@ -885,6 +913,29 @@ void add_timer_on(struct timer_list *timer, int cpu)
+ }
+ EXPORT_SYMBOL_GPL(add_timer_on);
- ENTER;
-- local_irq_save(flags);
-- if (!spin_trylock(&priv->tx_lock)) {
-- local_irq_restore(flags);
-- DBG("%s[%s]: TX locked, returning NETDEV_TX_LOCKED\n",
-- BDX_DRV_NAME, ndev->name);
-- return NETDEV_TX_LOCKED;
-- }
++#ifdef CONFIG_PREEMPT_RT_FULL
++/*
++ * Wait for a running timer
++ */
++static void wait_for_running_timer(struct timer_list *timer)
++{
++ struct tvec_base *base = timer->base;
+
-+ spin_lock_irqsave(&priv->tx_lock, flags);
++ if (base->running_timer == timer)
++ wait_event(base->wait_for_running_timer,
++ base->running_timer != timer);
++}
++
++# define wakeup_timer_waiters(b) wake_up(&(b)->wait_for_tunning_timer)
++#else
++static inline void wait_for_running_timer(struct timer_list *timer)
++{
++ cpu_relax();
++}
++
++# define wakeup_timer_waiters(b) do { } while (0)
++#endif
++
+ /**
+ * del_timer - deactive a timer.
+ * @timer: the timer to be deactivated
+@@ -953,7 +1004,7 @@ out:
+ }
+ EXPORT_SYMBOL(try_to_del_timer_sync);
- /* build tx descriptor */
- BDX_ASSERT(f->m.wptr >= f->m.memsz); /* started with valid wptr */
-Index: linux-3.0/drivers/net/3c59x.c
-===================================================================
---- linux-3.0.orig/drivers/net/3c59x.c
-+++ linux-3.0/drivers/net/3c59x.c
-@@ -843,9 +843,9 @@ static void poll_vortex(struct net_devic
- {
- struct vortex_private *vp = netdev_priv(dev);
- unsigned long flags;
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- (vp->full_bus_master_rx ? boomerang_interrupt:vortex_interrupt)(dev->irq,dev);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+-#ifdef CONFIG_SMP
++#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL)
+ /**
+ * del_timer_sync - deactivate a timer and wait for the handler to finish.
+ * @timer: the timer to be deactivated
+@@ -1013,7 +1064,7 @@ int del_timer_sync(struct timer_list *timer)
+ int ret = try_to_del_timer_sync(timer);
+ if (ret >= 0)
+ return ret;
+- cpu_relax();
++ wait_for_running_timer(timer);
+ }
}
- #endif
+ EXPORT_SYMBOL(del_timer_sync);
+@@ -1124,10 +1175,11 @@ static inline void __run_timers(struct tvec_base *base)
-@@ -1921,12 +1921,12 @@ static void vortex_tx_timeout(struct net
- * Block interrupts because vortex_interrupt does a bare spin_lock()
- */
- unsigned long flags;
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- if (vp->full_bus_master_tx)
- boomerang_interrupt(dev->irq, dev);
- else
- vortex_interrupt(dev->irq, dev);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ spin_unlock_irq(&base->lock);
+ call_timer_fn(timer, fn, data);
++ base->running_timer = NULL;
+ spin_lock_irq(&base->lock);
}
}
+- base->running_timer = NULL;
++ wake_up(&base->wait_for_running_timer);
+ spin_unlock_irq(&base->lock);
+ }
-Index: linux-3.0/drivers/net/gianfar.c
-===================================================================
---- linux-3.0.orig/drivers/net/gianfar.c
-+++ linux-3.0/drivers/net/gianfar.c
-@@ -1668,7 +1668,7 @@ void stop_gfar(struct net_device *dev)
+@@ -1266,6 +1318,23 @@ unsigned long get_next_timer_interrupt(unsigned long now)
+ */
+ if (cpu_is_offline(smp_processor_id()))
+ return now + NEXT_TIMER_MAX_DELTA;
++
++#ifdef CONFIG_PREEMPT_RT_FULL
++ /*
++ * On PREEMPT_RT we cannot sleep here. If the trylock does not
++ * succeed then we return the worst-case 'expires in 1 tick'
++ * value. We use the rt functions here directly to avoid a
++ * migrate_disable() call.
++ */
++ if (spin_do_trylock(&base->lock)) {
++ if (time_before_eq(base->next_timer, base->timer_jiffies))
++ base->next_timer = __next_timer_interrupt(base);
++ expires = base->next_timer;
++ rt_spin_unlock(&base->lock);
++ } else {
++ expires = now + 1;
++ }
++#else
+ spin_lock(&base->lock);
+ if (time_before_eq(base->next_timer, base->timer_jiffies))
+ base->next_timer = __next_timer_interrupt(base);
+@@ -1274,7 +1343,7 @@ unsigned long get_next_timer_interrupt(unsigned long now)
+
+ if (time_before_eq(expires, now))
+ return now;
+-
++#endif
+ return cmp_next_hrtimer_event(now, expires);
+ }
+ #endif
+@@ -1290,14 +1359,13 @@ void update_process_times(int user_tick)
+ /* Note: this timer irq context must be accounted for as well. */
+ account_process_tick(p, user_tick);
++ scheduler_tick();
+ run_local_timers();
+ rcu_check_callbacks(cpu, user_tick);
+- printk_tick();
+-#ifdef CONFIG_IRQ_WORK
++#if defined(CONFIG_IRQ_WORK) && !defined(CONFIG_PREEMPT_RT_FULL)
+ if (in_irq())
+ irq_work_run();
+ #endif
+- scheduler_tick();
+ run_posix_cpu_timers(p);
+ }
- /* Lock it down */
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- lock_tx_qs(priv);
- lock_rx_qs(priv);
+@@ -1308,6 +1376,11 @@ static void run_timer_softirq(struct softirq_action *h)
+ {
+ struct tvec_base *base = __this_cpu_read(tvec_bases);
-@@ -1676,7 +1676,7 @@ void stop_gfar(struct net_device *dev)
++#if defined(CONFIG_IRQ_WORK)
++ irq_work_run();
++#endif
++
++ printk_tick();
+ hrtimer_run_pending();
- unlock_rx_qs(priv);
- unlock_tx_qs(priv);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
+ if (time_after_eq(jiffies, base->timer_jiffies))
+@@ -1634,6 +1707,7 @@ static int __cpuinit init_timers_cpu(int cpu)
+ }
- /* Free the IRQs */
- if (priv->device_flags & FSL_GIANFAR_DEV_HAS_MULTI_INTR) {
-@@ -2951,7 +2951,7 @@ static void adjust_link(struct net_devic
- struct phy_device *phydev = priv->phydev;
- int new_state = 0;
+ spin_lock_init(&base->lock);
++ init_waitqueue_head(&base->wait_for_running_timer);
-- local_irq_save(flags);
-+ local_irq_save_nort(flags);
- lock_tx_qs(priv);
+ for (j = 0; j < TVN_SIZE; j++) {
+ INIT_LIST_HEAD(base->tv5.vec + j);
+@@ -1669,6 +1743,7 @@ static void __cpuinit migrate_timers(int cpu)
+ {
+ struct tvec_base *old_base;
+ struct tvec_base *new_base;
++ unsigned long flags;
+ int i;
- if (phydev->link) {
-@@ -3019,7 +3019,7 @@ static void adjust_link(struct net_devic
- if (new_state && netif_msg_link(priv))
- phy_print_status(phydev);
- unlock_tx_qs(priv);
-- local_irq_restore(flags);
-+ local_irq_restore_nort(flags);
- }
+ BUG_ON(cpu_online(cpu));
+@@ -1678,8 +1753,11 @@ static void __cpuinit migrate_timers(int cpu)
+ * The caller is globally serialized and nobody else
+ * takes two locks at once, deadlock is not possible.
+ */
+- spin_lock_irq(&new_base->lock);
+- spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
++ local_irq_save(flags);
++ while (!spin_trylock(&new_base->lock))
++ cpu_relax();
++ while (!spin_trylock(&old_base->lock))
++ cpu_relax();
- /* Update the hash table based on the current list of multicast
-Index: linux-3.0/drivers/usb/host/ohci-hcd.c
-===================================================================
---- linux-3.0.orig/drivers/usb/host/ohci-hcd.c
-+++ linux-3.0/drivers/usb/host/ohci-hcd.c
-@@ -833,9 +833,13 @@ static irqreturn_t ohci_irq (struct usb_
- }
+ BUG_ON(old_base->running_timer);
- if (ints & OHCI_INTR_WDH) {
-- spin_lock (&ohci->lock);
-- dl_done_list (ohci);
-- spin_unlock (&ohci->lock);
-+ if (ohci->hcca->done_head == 0) {
-+ ints &= ~OHCI_INTR_WDH;
-+ } else {
-+ spin_lock (&ohci->lock);
-+ dl_done_list (ohci);
-+ spin_unlock (&ohci->lock);
-+ }
+@@ -1693,7 +1771,9 @@ static void __cpuinit migrate_timers(int cpu)
}
- if (quirk_zfmicro(ohci) && (ints & OHCI_INTR_SF)) {
-Index: linux-3.0/include/linux/percpu.h
-===================================================================
---- linux-3.0.orig/include/linux/percpu.h
-+++ linux-3.0/include/linux/percpu.h
-@@ -48,6 +48,31 @@
- preempt_enable(); \
- } while (0)
+ spin_unlock(&old_base->lock);
+- spin_unlock_irq(&new_base->lock);
++ spin_unlock(&new_base->lock);
++ local_irq_restore(flags);
++
+ put_cpu_var(tvec_bases);
+ }
+ #endif /* CONFIG_HOTPLUG_CPU */
+diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
+index 2ad39e5..1f1d8e0 100644
+--- a/kernel/trace/Kconfig
++++ b/kernel/trace/Kconfig
+@@ -192,6 +192,24 @@ config IRQSOFF_TRACER
+ enabled. This option and the preempt-off timing option can be
+ used together or separately.)
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+# define get_local_var(var) get_cpu_var(var)
-+# define put_local_var(var) put_cpu_var(var)
-+# define get_local_ptr(var) get_cpu_ptr(var)
-+# define put_local_ptr(var) put_cpu_ptr(var)
-+#else
-+# define get_local_var(var) (*({ \
-+ migrate_disable(); \
-+ &__get_cpu_var(var); }))
++config INTERRUPT_OFF_HIST
++ bool "Interrupts-off Latency Histogram"
++ depends on IRQSOFF_TRACER
++ help
++ This option generates continuously updated histograms (one per cpu)
++ of the duration of time periods with interrupts disabled. The
++ histograms are disabled by default. To enable them, write a non-zero
++ number to
+
-+# define put_local_var(var) do { \
-+ (void)&(var); \
-+ migrate_enable(); \
-+} while (0)
++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff
+
-+# define get_local_ptr(var) ({ \
-+ migrate_disable(); \
-+ this_cpu_ptr(var); })
++ If PREEMPT_OFF_HIST is also selected, additional histograms (one
++ per cpu) are generated that accumulate the duration of time periods
++ when both interrupts and preemption are disabled. The histogram data
++ will be located in the debug file system at
+
-+# define put_local_ptr(var) do { \
-+ (void)(var); \
-+ migrate_enable(); \
-+} while (0)
-+#endif
++ /sys/kernel/debug/tracing/latency_hist/irqsoff
+
- /* minimum unit size, also is the maximum supported allocation size */
- #define PCPU_MIN_UNIT_SIZE PFN_ALIGN(32 << 10)
+ config PREEMPT_TRACER
+ bool "Preemption-off Latency Tracer"
+ default n
+@@ -214,6 +232,24 @@ config PREEMPT_TRACER
+ enabled. This option and the irqs-off timing option can be
+ used together or separately.)
-Index: linux-3.0/include/linux/locallock.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/locallock.h
-@@ -0,0 +1,230 @@
-+#ifndef _LINUX_LOCALLOCK_H
-+#define _LINUX_LOCALLOCK_H
++config PREEMPT_OFF_HIST
++ bool "Preemption-off Latency Histogram"
++ depends on PREEMPT_TRACER
++ help
++ This option generates continuously updated histograms (one per cpu)
++ of the duration of time periods with preemption disabled. The
++ histograms are disabled by default. To enable them, write a non-zero
++ number to
+
-+#include <linux/spinlock.h>
++ /sys/kernel/debug/tracing/latency_hist/enable/preemptirqsoff
+
-+#ifdef CONFIG_PREEMPT_RT_BASE
++ If INTERRUPT_OFF_HIST is also selected, additional histograms (one
++ per cpu) are generated that accumulate the duration of time periods
++ when both interrupts and preemption are disabled. The histogram data
++ will be located in the debug file system at
+
-+#ifdef CONFIG_DEBUG_SPINLOCK
-+# define LL_WARN(cond) WARN_ON(cond)
-+#else
-+# define LL_WARN(cond) do { } while (0)
-+#endif
++ /sys/kernel/debug/tracing/latency_hist/preemptoff
+
-+/*
-+ * per cpu lock based substitute for local_irq_*()
-+ */
-+struct local_irq_lock {
-+ spinlock_t lock;
-+ struct task_struct *owner;
-+ int nestcnt;
-+ unsigned long flags;
-+};
+ config SCHED_TRACER
+ bool "Scheduling Latency Tracer"
+ select GENERIC_TRACER
+@@ -223,6 +259,74 @@ config SCHED_TRACER
+ This tracer tracks the latency of the highest priority task
+ to be scheduled in, starting from the point it has woken up.
+
++config WAKEUP_LATENCY_HIST
++ bool "Scheduling Latency Histogram"
++ depends on SCHED_TRACER
++ help
++ This option generates continuously updated histograms (one per cpu)
++ of the scheduling latency of the highest priority task.
++ The histograms are disabled by default. To enable them, write a
++ non-zero number to
+
-+#define DEFINE_LOCAL_IRQ_LOCK(lvar) \
-+ DEFINE_PER_CPU(struct local_irq_lock, lvar) = { \
-+ .lock = __SPIN_LOCK_UNLOCKED((lvar).lock) }
++ /sys/kernel/debug/tracing/latency_hist/enable/wakeup
+
-+#define local_irq_lock_init(lvar) \
-+ do { \
-+ int __cpu; \
-+ for_each_possible_cpu(__cpu) \
-+ spin_lock_init(&per_cpu(lvar, __cpu).lock); \
-+ } while (0)
++ Two different algorithms are used, one to determine the latency of
++ processes that exclusively use the highest priority of the system and
++ another one to determine the latency of processes that share the
++ highest system priority with other processes. The former is used to
++ improve hardware and system software, the latter to optimize the
++ priority design of a given system. The histogram data will be
++ located in the debug file system at
+
-+static inline void __local_lock(struct local_irq_lock *lv)
-+{
-+ if (lv->owner != current) {
-+ spin_lock(&lv->lock);
-+ LL_WARN(lv->owner);
-+ LL_WARN(lv->nestcnt);
-+ lv->owner = current;
-+ }
-+ lv->nestcnt++;
-+}
++ /sys/kernel/debug/tracing/latency_hist/wakeup
+
-+#define local_lock(lvar) \
-+ do { __local_lock(&get_local_var(lvar)); } while (0)
++ and
+
-+static inline int __local_trylock(struct local_irq_lock *lv)
-+{
-+ if (lv->owner != current && spin_trylock(&lv->lock)) {
-+ LL_WARN(lv->owner);
-+ LL_WARN(lv->nestcnt);
-+ lv->owner = current;
-+ lv->nestcnt = 1;
-+ return 1;
-+ }
-+ return 0;
-+}
++ /sys/kernel/debug/tracing/latency_hist/wakeup/sharedprio
+
-+#define local_trylock(lvar) \
-+ ({ \
-+ int __locked; \
-+ __locked = __local_trylock(&get_local_var(lvar)); \
-+ if (!__locked) \
-+ put_local_var(lvar); \
-+ __locked; \
-+ })
++ If both Scheduling Latency Histogram and Missed Timer Offsets
++ Histogram are selected, additional histogram data will be collected
++ that contain, in addition to the wakeup latency, the timer latency, in
++ case the wakeup was triggered by an expired timer. These histograms
++ are available in the
+
-+static inline void __local_unlock(struct local_irq_lock *lv)
-+{
-+ LL_WARN(lv->nestcnt == 0);
-+ LL_WARN(lv->owner != current);
-+ if (--lv->nestcnt)
-+ return;
++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup
+
-+ lv->owner = NULL;
-+ spin_unlock(&lv->lock);
-+}
++ directory. They reflect the apparent interrupt and scheduling latency
++ and are best suitable to determine the worst-case latency of a given
++ system. To enable these histograms, write a non-zero number to
+
-+#define local_unlock(lvar) \
-+ do { \
-+ __local_unlock(&__get_cpu_var(lvar)); \
-+ put_local_var(lvar); \
-+ } while (0)
++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup
+
-+static inline void __local_lock_irq(struct local_irq_lock *lv)
-+{
-+ spin_lock_irqsave(&lv->lock, lv->flags);
-+ LL_WARN(lv->owner);
-+ LL_WARN(lv->nestcnt);
-+ lv->owner = current;
-+ lv->nestcnt = 1;
-+}
++config MISSED_TIMER_OFFSETS_HIST
++ depends on HIGH_RES_TIMERS
++ select GENERIC_TRACER
++ bool "Missed Timer Offsets Histogram"
++ help
++ Generate a histogram of missed timer offsets in microseconds. The
++ histograms are disabled by default. To enable them, write a non-zero
++ number to
+
-+#define local_lock_irq(lvar) \
-+ do { __local_lock_irq(&get_local_var(lvar)); } while (0)
++ /sys/kernel/debug/tracing/latency_hist/enable/missed_timer_offsets
+
-+static inline void __local_unlock_irq(struct local_irq_lock *lv)
-+{
-+ LL_WARN(!lv->nestcnt);
-+ LL_WARN(lv->owner != current);
-+ lv->owner = NULL;
-+ lv->nestcnt = 0;
-+ spin_unlock_irq(&lv->lock);
-+}
++ The histogram data will be located in the debug file system at
+
-+#define local_unlock_irq(lvar) \
-+ do { \
-+ __local_unlock_irq(&__get_cpu_var(lvar)); \
-+ put_local_var(lvar); \
-+ } while (0)
++ /sys/kernel/debug/tracing/latency_hist/missed_timer_offsets
+
-+static inline int __local_lock_irqsave(struct local_irq_lock *lv)
-+{
-+ if (lv->owner != current) {
-+ __local_lock_irq(lv);
-+ return 0;
-+ } else {
-+ lv->nestcnt++;
-+ return 1;
-+ }
-+}
++ If both Scheduling Latency Histogram and Missed Timer Offsets
++ Histogram are selected, additional histogram data will be collected
++ that contain, in addition to the wakeup latency, the timer latency, in
++ case the wakeup was triggered by an expired timer. These histograms
++ are available in the
+
-+#define local_lock_irqsave(lvar, _flags) \
-+ do { \
-+ if (__local_lock_irqsave(&get_local_var(lvar))) \
-+ put_local_var(lvar); \
-+ _flags = __get_cpu_var(lvar).flags; \
-+ } while (0)
++ /sys/kernel/debug/tracing/latency_hist/timerandwakeup
+
-+static inline int __local_unlock_irqrestore(struct local_irq_lock *lv,
-+ unsigned long flags)
-+{
-+ LL_WARN(!lv->nestcnt);
-+ LL_WARN(lv->owner != current);
-+ if (--lv->nestcnt)
-+ return 0;
++ directory. They reflect the apparent interrupt and scheduling latency
++ and are best suitable to determine the worst-case latency of a given
++ system. To enable these histograms, write a non-zero number to
+
-+ lv->owner = NULL;
-+ spin_unlock_irqrestore(&lv->lock, lv->flags);
-+ return 1;
-+}
++ /sys/kernel/debug/tracing/latency_hist/enable/timerandwakeup
+
-+#define local_unlock_irqrestore(lvar, flags) \
-+ do { \
-+ if (__local_unlock_irqrestore(&__get_cpu_var(lvar), flags)) \
-+ put_local_var(lvar); \
-+ } while (0)
+ config ENABLE_DEFAULT_TRACERS
+ bool "Trace process context switches and events"
+ depends on !GENERIC_TRACER
+diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile
+index 761c510..1aeb262 100644
+--- a/kernel/trace/Makefile
++++ b/kernel/trace/Makefile
+@@ -34,6 +34,10 @@ obj-$(CONFIG_FUNCTION_TRACER) += trace_functions.o
+ obj-$(CONFIG_IRQSOFF_TRACER) += trace_irqsoff.o
+ obj-$(CONFIG_PREEMPT_TRACER) += trace_irqsoff.o
+ obj-$(CONFIG_SCHED_TRACER) += trace_sched_wakeup.o
++obj-$(CONFIG_INTERRUPT_OFF_HIST) += latency_hist.o
++obj-$(CONFIG_PREEMPT_OFF_HIST) += latency_hist.o
++obj-$(CONFIG_WAKEUP_LATENCY_HIST) += latency_hist.o
++obj-$(CONFIG_MISSED_TIMER_OFFSETS_HIST) += latency_hist.o
+ obj-$(CONFIG_NOP_TRACER) += trace_nop.o
+ obj-$(CONFIG_STACK_TRACER) += trace_stack.o
+ obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o
+diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
+index ef9271b..1eef6cf 100644
+--- a/kernel/trace/ftrace.c
++++ b/kernel/trace/ftrace.c
+@@ -1182,8 +1182,14 @@ alloc_and_copy_ftrace_hash(int size_bits, struct ftrace_hash *hash)
+ return NULL;
+ }
+
++static void
++ftrace_hash_rec_disable(struct ftrace_ops *ops, int filter_hash);
++static void
++ftrace_hash_rec_enable(struct ftrace_ops *ops, int filter_hash);
+
-+#define local_spin_trylock_irq(lvar, lock) \
-+ ({ \
-+ int __locked; \
-+ local_lock_irq(lvar); \
-+ __locked = spin_trylock(lock); \
-+ if (!__locked) \
-+ local_unlock_irq(lvar); \
-+ __locked; \
-+ })
+ static int
+-ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src)
++ftrace_hash_move(struct ftrace_ops *ops, int enable,
++ struct ftrace_hash **dst, struct ftrace_hash *src)
+ {
+ struct ftrace_func_entry *entry;
+ struct hlist_node *tp, *tn;
+@@ -1193,9 +1199,16 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src)
+ unsigned long key;
+ int size = src->count;
+ int bits = 0;
++ int ret;
+ int i;
+
+ /*
++ * Remove the current set, update the hash and add
++ * them back.
++ */
++ ftrace_hash_rec_disable(ops, enable);
+
-+#define local_spin_lock_irq(lvar, lock) \
-+ do { \
-+ local_lock_irq(lvar); \
-+ spin_lock(lock); \
-+ } while (0)
++ /*
+ * If the new source is empty, just free dst and assign it
+ * the empty_hash.
+ */
+@@ -1215,9 +1228,10 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src)
+ if (bits > FTRACE_HASH_MAX_BITS)
+ bits = FTRACE_HASH_MAX_BITS;
+
++ ret = -ENOMEM;
+ new_hash = alloc_ftrace_hash(bits);
+ if (!new_hash)
+- return -ENOMEM;
++ goto out;
+
+ size = 1 << src->size_bits;
+ for (i = 0; i < size; i++) {
+@@ -1236,7 +1250,16 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src)
+ rcu_assign_pointer(*dst, new_hash);
+ free_ftrace_hash_rcu(old_hash);
+
+- return 0;
++ ret = 0;
++ out:
++ /*
++ * Enable regardless of ret:
++ * On success, we enable the new hash.
++ * On failure, we re-enable the original hash.
++ */
++ ftrace_hash_rec_enable(ops, enable);
+
-+#define local_spin_unlock_irq(lvar, lock) \
-+ do { \
-+ spin_unlock(lock); \
-+ local_unlock_irq(lvar); \
-+ } while (0)
++ return ret;
+ }
+
+ /*
+@@ -2877,7 +2900,7 @@ ftrace_set_regex(struct ftrace_ops *ops, unsigned char *buf, int len,
+ ftrace_match_records(hash, buf, len);
+
+ mutex_lock(&ftrace_lock);
+- ret = ftrace_hash_move(orig_hash, hash);
++ ret = ftrace_hash_move(ops, enable, orig_hash, hash);
+ mutex_unlock(&ftrace_lock);
+
+ mutex_unlock(&ftrace_regex_lock);
+@@ -3060,18 +3083,12 @@ ftrace_regex_release(struct inode *inode, struct file *file)
+ orig_hash = &iter->ops->notrace_hash;
+
+ mutex_lock(&ftrace_lock);
+- /*
+- * Remove the current set, update the hash and add
+- * them back.
+- */
+- ftrace_hash_rec_disable(iter->ops, filter_hash);
+- ret = ftrace_hash_move(orig_hash, iter->hash);
+- if (!ret) {
+- ftrace_hash_rec_enable(iter->ops, filter_hash);
+- if (iter->ops->flags & FTRACE_OPS_FL_ENABLED
+- && ftrace_enabled)
+- ftrace_run_update_code(FTRACE_ENABLE_CALLS);
+- }
++ ret = ftrace_hash_move(iter->ops, filter_hash,
++ orig_hash, iter->hash);
++ if (!ret && (iter->ops->flags & FTRACE_OPS_FL_ENABLED)
++ && ftrace_enabled)
++ ftrace_run_update_code(FTRACE_ENABLE_CALLS);
+
-+#define local_spin_lock_irqsave(lvar, lock, flags) \
-+ do { \
-+ local_lock_irqsave(lvar, flags); \
-+ spin_lock(lock); \
-+ } while (0)
+ mutex_unlock(&ftrace_lock);
+ }
+ free_ftrace_hash(iter->hash);
+diff --git a/kernel/trace/latency_hist.c b/kernel/trace/latency_hist.c
+new file mode 100644
+index 0000000..9d49fcb
+--- /dev/null
++++ b/kernel/trace/latency_hist.c
+@@ -0,0 +1,1170 @@
++/*
++ * kernel/trace/latency_hist.c
++ *
++ * Add support for histograms of preemption-off latency and
++ * interrupt-off latency and wakeup latency, it depends on
++ * Real-Time Preemption Support.
++ *
++ * Copyright (C) 2005 MontaVista Software, Inc.
++ * Yi Yang <yyang at ch.mvista.com>
++ *
++ * Converted to work with the new latency tracer.
++ * Copyright (C) 2008 Red Hat, Inc.
++ * Steven Rostedt <srostedt at redhat.com>
++ *
++ */
++#include <linux/module.h>
++#include <linux/debugfs.h>
++#include <linux/seq_file.h>
++#include <linux/percpu.h>
++#include <linux/kallsyms.h>
++#include <linux/uaccess.h>
++#include <linux/sched.h>
++#include <linux/slab.h>
++#include <asm/atomic.h>
++#include <asm/div64.h>
+
-+#define local_spin_unlock_irqrestore(lvar, lock, flags) \
-+ do { \
-+ spin_unlock(lock); \
-+ local_unlock_irqrestore(lvar, flags); \
-+ } while (0)
++#include "trace.h"
++#include <trace/events/sched.h>
+
-+#define get_locked_var(lvar, var) \
-+ (*({ \
-+ local_lock(lvar); \
-+ &__get_cpu_var(var); \
-+ }))
++#define CREATE_TRACE_POINTS
++#include <trace/events/hist.h>
+
-+#define put_locked_var(lvar, var) local_unlock(lvar)
++enum {
++ IRQSOFF_LATENCY = 0,
++ PREEMPTOFF_LATENCY,
++ PREEMPTIRQSOFF_LATENCY,
++ WAKEUP_LATENCY,
++ WAKEUP_LATENCY_SHAREDPRIO,
++ MISSED_TIMER_OFFSETS,
++ TIMERANDWAKEUP_LATENCY,
++ MAX_LATENCY_TYPE,
++};
+
-+#define local_lock_cpu(lvar) \
-+ ({ \
-+ local_lock(lvar); \
-+ smp_processor_id(); \
-+ })
++#define MAX_ENTRY_NUM 10240
+
-+#define local_unlock_cpu(lvar) local_unlock(lvar)
++struct hist_data {
++ atomic_t hist_mode; /* 0 log, 1 don't log */
++ long offset; /* set it to MAX_ENTRY_NUM/2 for a bipolar scale */
++ unsigned long min_lat;
++ unsigned long max_lat;
++ unsigned long long below_hist_bound_samples;
++ unsigned long long above_hist_bound_samples;
++ unsigned long long accumulate_lat;
++ unsigned long long total_samples;
++ unsigned long long hist_array[MAX_ENTRY_NUM];
++};
+
-+#else /* PREEMPT_RT_BASE */
++struct enable_data {
++ int latency_type;
++ int enabled;
++};
+
-+#define DEFINE_LOCAL_IRQ_LOCK(lvar) __typeof__(const int) lvar
++static char *latency_hist_dir_root = "latency_hist";
+
-+static inline void local_irq_lock_init(int lvar) { }
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++static DEFINE_PER_CPU(struct hist_data, irqsoff_hist);
++static char *irqsoff_hist_dir = "irqsoff";
++static DEFINE_PER_CPU(cycles_t, hist_irqsoff_start);
++static DEFINE_PER_CPU(int, hist_irqsoff_counting);
++#endif
++
++#ifdef CONFIG_PREEMPT_OFF_HIST
++static DEFINE_PER_CPU(struct hist_data, preemptoff_hist);
++static char *preemptoff_hist_dir = "preemptoff";
++static DEFINE_PER_CPU(cycles_t, hist_preemptoff_start);
++static DEFINE_PER_CPU(int, hist_preemptoff_counting);
++#endif
++
++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST)
++static DEFINE_PER_CPU(struct hist_data, preemptirqsoff_hist);
++static char *preemptirqsoff_hist_dir = "preemptirqsoff";
++static DEFINE_PER_CPU(cycles_t, hist_preemptirqsoff_start);
++static DEFINE_PER_CPU(int, hist_preemptirqsoff_counting);
++#endif
++
++#if defined(CONFIG_PREEMPT_OFF_HIST) || defined(CONFIG_INTERRUPT_OFF_HIST)
++static notrace void probe_preemptirqsoff_hist(void *v, int reason, int start);
++static struct enable_data preemptirqsoff_enabled_data = {
++ .latency_type = PREEMPTIRQSOFF_LATENCY,
++ .enabled = 0,
++};
++#endif
+
-+#define local_lock(lvar) preempt_disable()
-+#define local_unlock(lvar) preempt_enable()
-+#define local_lock_irq(lvar) local_irq_disable()
-+#define local_unlock_irq(lvar) local_irq_enable()
-+#define local_lock_irqsave(lvar, flags) local_irq_save(flags)
-+#define local_unlock_irqrestore(lvar, flags) local_irq_restore(flags)
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++struct maxlatproc_data {
++ char comm[FIELD_SIZEOF(struct task_struct, comm)];
++ char current_comm[FIELD_SIZEOF(struct task_struct, comm)];
++ int pid;
++ int current_pid;
++ int prio;
++ int current_prio;
++ long latency;
++ long timeroffset;
++ cycle_t timestamp;
++};
++#endif
+
-+#define local_spin_trylock_irq(lvar, lock) spin_trylock_irq(lock)
-+#define local_spin_lock_irq(lvar, lock) spin_lock_irq(lock)
-+#define local_spin_unlock_irq(lvar, lock) spin_unlock_irq(lock)
-+#define local_spin_lock_irqsave(lvar, lock, flags) \
-+ spin_lock_irqsave(lock, flags)
-+#define local_spin_unlock_irqrestore(lvar, lock, flags) \
-+ spin_unlock_irqrestore(lock, flags)
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist);
++static DEFINE_PER_CPU(struct hist_data, wakeup_latency_hist_sharedprio);
++static char *wakeup_latency_hist_dir = "wakeup";
++static char *wakeup_latency_hist_dir_sharedprio = "sharedprio";
++static notrace void probe_wakeup_latency_hist_start(void *v,
++ struct task_struct *p, int success);
++static notrace void probe_wakeup_latency_hist_stop(void *v,
++ struct task_struct *prev, struct task_struct *next);
++static notrace void probe_sched_migrate_task(void *,
++ struct task_struct *task, int cpu);
++static struct enable_data wakeup_latency_enabled_data = {
++ .latency_type = WAKEUP_LATENCY,
++ .enabled = 0,
++};
++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc);
++static DEFINE_PER_CPU(struct maxlatproc_data, wakeup_maxlatproc_sharedprio);
++static DEFINE_PER_CPU(struct task_struct *, wakeup_task);
++static DEFINE_PER_CPU(int, wakeup_sharedprio);
++static unsigned long wakeup_pid;
++#endif
+
-+#define get_locked_var(lvar, var) get_cpu_var(var)
-+#define put_locked_var(lvar, var) put_cpu_var(var)
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++static DEFINE_PER_CPU(struct hist_data, missed_timer_offsets);
++static char *missed_timer_offsets_dir = "missed_timer_offsets";
++static notrace void probe_hrtimer_interrupt(void *v, int cpu,
++ long long offset, struct task_struct *curr, struct task_struct *task);
++static struct enable_data missed_timer_offsets_enabled_data = {
++ .latency_type = MISSED_TIMER_OFFSETS,
++ .enabled = 0,
++};
++static DEFINE_PER_CPU(struct maxlatproc_data, missed_timer_offsets_maxlatproc);
++static unsigned long missed_timer_offsets_pid;
++#endif
+
-+#define local_lock_cpu(lvar) get_cpu()
-+#define local_unlock_cpu(lvar) put_cpu()
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++static DEFINE_PER_CPU(struct hist_data, timerandwakeup_latency_hist);
++static char *timerandwakeup_latency_hist_dir = "timerandwakeup";
++static struct enable_data timerandwakeup_enabled_data = {
++ .latency_type = TIMERANDWAKEUP_LATENCY,
++ .enabled = 0,
++};
++static DEFINE_PER_CPU(struct maxlatproc_data, timerandwakeup_maxlatproc);
++#endif
+
++void notrace latency_hist(int latency_type, int cpu, unsigned long latency,
++ unsigned long timeroffset, cycle_t stop,
++ struct task_struct *p)
++{
++ struct hist_data *my_hist;
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ struct maxlatproc_data *mp = NULL;
+#endif
+
++ if (cpu < 0 || cpu >= NR_CPUS || latency_type < 0 ||
++ latency_type >= MAX_LATENCY_TYPE)
++ return;
++
++ switch (latency_type) {
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ case IRQSOFF_LATENCY:
++ my_hist = &per_cpu(irqsoff_hist, cpu);
++ break;
++#endif
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ case PREEMPTOFF_LATENCY:
++ my_hist = &per_cpu(preemptoff_hist, cpu);
++ break;
++#endif
++#if defined(CONFIG_PREEMPT_OFF_HIST) && defined(CONFIG_INTERRUPT_OFF_HIST)
++ case PREEMPTIRQSOFF_LATENCY:
++ my_hist = &per_cpu(preemptirqsoff_hist, cpu);
++ break;
++#endif
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ case WAKEUP_LATENCY:
++ my_hist = &per_cpu(wakeup_latency_hist, cpu);
++ mp = &per_cpu(wakeup_maxlatproc, cpu);
++ break;
++ case WAKEUP_LATENCY_SHAREDPRIO:
++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu);
++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu);
++ break;
++#endif
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ case MISSED_TIMER_OFFSETS:
++ my_hist = &per_cpu(missed_timer_offsets, cpu);
++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu);
++ break;
++#endif
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ case TIMERANDWAKEUP_LATENCY:
++ my_hist = &per_cpu(timerandwakeup_latency_hist, cpu);
++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu);
++ break;
+#endif
-Index: linux-3.0/include/linux/smp.h
-===================================================================
---- linux-3.0.orig/include/linux/smp.h
-+++ linux-3.0/include/linux/smp.h
-@@ -80,7 +80,6 @@ void __smp_call_function_single(int cpui
-
- int smp_call_function_any(const struct cpumask *mask,
- smp_call_func_t func, void *info, int wait);
--
- /*
- * Generic and arch helpers
- */
-@@ -173,6 +172,9 @@ smp_call_function_any(const struct cpuma
- #define get_cpu() ({ preempt_disable(); smp_processor_id(); })
- #define put_cpu() preempt_enable()
-
-+#define get_cpu_light() ({ migrate_disable(); smp_processor_id(); })
-+#define put_cpu_light() migrate_enable()
+
- /*
- * Callback to arch code if there's nosmp or maxcpus=0 on the
- * boot command line:
-Index: linux-3.0/mm/swap.c
-===================================================================
---- linux-3.0.orig/mm/swap.c
-+++ linux-3.0/mm/swap.c
-@@ -31,6 +31,7 @@
- #include <linux/backing-dev.h>
- #include <linux/memcontrol.h>
- #include <linux/gfp.h>
-+#include <linux/locallock.h>
-
- #include "internal.h"
-
-@@ -41,6 +42,9 @@ static DEFINE_PER_CPU(struct pagevec[NR_
- static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs);
- static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs);
-
-+static DEFINE_LOCAL_IRQ_LOCK(rotate_lock);
-+static DEFINE_LOCAL_IRQ_LOCK(swap_lock);
++ default:
++ return;
++ }
+
- /*
- * This path almost never happens for VM activity - pages are normally
- * freed via pagevecs. But it gets used by networking.
-@@ -267,11 +271,11 @@ void rotate_reclaimable_page(struct page
- unsigned long flags;
-
- page_cache_get(page);
-- local_irq_save(flags);
-+ local_lock_irqsave(rotate_lock, flags);
- pvec = &__get_cpu_var(lru_rotate_pvecs);
- if (!pagevec_add(pvec, page))
- pagevec_move_tail(pvec);
-- local_irq_restore(flags);
-+ local_unlock_irqrestore(rotate_lock, flags);
- }
- }
-
-@@ -327,12 +331,13 @@ static void activate_page_drain(int cpu)
- void activate_page(struct page *page)
- {
- if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) {
-- struct pagevec *pvec = &get_cpu_var(activate_page_pvecs);
-+ struct pagevec *pvec = &get_locked_var(swap_lock,
-+ activate_page_pvecs);
-
- page_cache_get(page);
- if (!pagevec_add(pvec, page))
- pagevec_lru_move_fn(pvec, __activate_page, NULL);
-- put_cpu_var(activate_page_pvecs);
-+ put_locked_var(swap_lock, activate_page_pvecs);
- }
- }
-
-@@ -373,12 +378,12 @@ EXPORT_SYMBOL(mark_page_accessed);
-
- void __lru_cache_add(struct page *page, enum lru_list lru)
- {
-- struct pagevec *pvec = &get_cpu_var(lru_add_pvecs)[lru];
-+ struct pagevec *pvec = &get_locked_var(swap_lock, lru_add_pvecs)[lru];
-
- page_cache_get(page);
- if (!pagevec_add(pvec, page))
- ____pagevec_lru_add(pvec, lru);
-- put_cpu_var(lru_add_pvecs);
-+ put_locked_var(swap_lock, lru_add_pvecs);
- }
- EXPORT_SYMBOL(__lru_cache_add);
-
-@@ -512,9 +517,9 @@ static void drain_cpu_pagevecs(int cpu)
- unsigned long flags;
-
- /* No harm done if a racing interrupt already did this */
-- local_irq_save(flags);
-+ local_lock_irqsave(rotate_lock, flags);
- pagevec_move_tail(pvec);
-- local_irq_restore(flags);
-+ local_unlock_irqrestore(rotate_lock, flags);
- }
-
- pvec = &per_cpu(lru_deactivate_pvecs, cpu);
-@@ -542,18 +547,19 @@ void deactivate_page(struct page *page)
- return;
-
- if (likely(get_page_unless_zero(page))) {
-- struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs);
-+ struct pagevec *pvec = &get_locked_var(swap_lock,
-+ lru_deactivate_pvecs);
-
- if (!pagevec_add(pvec, page))
- pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL);
-- put_cpu_var(lru_deactivate_pvecs);
-+ put_locked_var(swap_lock, lru_deactivate_pvecs);
- }
- }
-
- void lru_add_drain(void)
- {
-- drain_cpu_pagevecs(get_cpu());
-- put_cpu();
-+ drain_cpu_pagevecs(local_lock_cpu(swap_lock));
-+ local_unlock_cpu(swap_lock);
- }
-
- static void lru_add_drain_per_cpu(struct work_struct *dummy)
-@@ -783,6 +789,9 @@ void __init swap_setup(void)
- {
- unsigned long megs = totalram_pages >> (20 - PAGE_SHIFT);
-
-+ local_irq_lock_init(rotate_lock);
-+ local_irq_lock_init(swap_lock);
++ latency += my_hist->offset;
++
++ if (atomic_read(&my_hist->hist_mode) == 0)
++ return;
++
++ if (latency < 0 || latency >= MAX_ENTRY_NUM) {
++ if (latency < 0)
++ my_hist->below_hist_bound_samples++;
++ else
++ my_hist->above_hist_bound_samples++;
++ } else
++ my_hist->hist_array[latency]++;
+
- #ifdef CONFIG_SWAP
- bdi_init(swapper_space.backing_dev_info);
- #endif
-Index: linux-3.0/mm/vmscan.c
-===================================================================
---- linux-3.0.orig/mm/vmscan.c
-+++ linux-3.0/mm/vmscan.c
-@@ -1277,8 +1277,8 @@ static int too_many_isolated(struct zone
- */
- static noinline_for_stack void
- putback_lru_pages(struct zone *zone, struct scan_control *sc,
-- unsigned long nr_anon, unsigned long nr_file,
-- struct list_head *page_list)
-+ unsigned long nr_anon, unsigned long nr_file,
-+ struct list_head *page_list, unsigned long nr_reclaimed)
- {
- struct page *page;
- struct pagevec pvec;
-@@ -1289,7 +1289,12 @@ putback_lru_pages(struct zone *zone, str
- /*
- * Put back any unfreeable pages.
- */
-- spin_lock(&zone->lru_lock);
-+ spin_lock_irq(&zone->lru_lock);
++ if (unlikely(latency > my_hist->max_lat ||
++ my_hist->min_lat == ULONG_MAX)) {
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ if (latency_type == WAKEUP_LATENCY ||
++ latency_type == WAKEUP_LATENCY_SHAREDPRIO ||
++ latency_type == MISSED_TIMER_OFFSETS ||
++ latency_type == TIMERANDWAKEUP_LATENCY) {
++ strncpy(mp->comm, p->comm, sizeof(mp->comm));
++ strncpy(mp->current_comm, current->comm,
++ sizeof(mp->current_comm));
++ mp->pid = task_pid_nr(p);
++ mp->current_pid = task_pid_nr(current);
++ mp->prio = p->prio;
++ mp->current_prio = current->prio;
++ mp->latency = latency;
++ mp->timeroffset = timeroffset;
++ mp->timestamp = stop;
++ }
++#endif
++ my_hist->max_lat = latency;
++ }
++ if (unlikely(latency < my_hist->min_lat))
++ my_hist->min_lat = latency;
++ my_hist->total_samples++;
++ my_hist->accumulate_lat += latency;
++}
+
-+ if (current_is_kswapd())
-+ __count_vm_events(KSWAPD_STEAL, nr_reclaimed);
-+ __count_zone_vm_events(PGSTEAL, zone, nr_reclaimed);
++static void *l_start(struct seq_file *m, loff_t *pos)
++{
++ loff_t *index_ptr = NULL;
++ loff_t index = *pos;
++ struct hist_data *my_hist = m->private;
+
- while (!list_empty(page_list)) {
- int lru;
- page = lru_to_page(page_list);
-@@ -1464,12 +1469,7 @@ shrink_inactive_list(unsigned long nr_to
- nr_reclaimed += shrink_page_list(&page_list, zone, sc);
- }
-
-- local_irq_disable();
-- if (current_is_kswapd())
-- __count_vm_events(KSWAPD_STEAL, nr_reclaimed);
-- __count_zone_vm_events(PGSTEAL, zone, nr_reclaimed);
--
-- putback_lru_pages(zone, sc, nr_anon, nr_file, &page_list);
-+ putback_lru_pages(zone, sc, nr_anon, nr_file, &page_list, nr_reclaimed);
-
- trace_mm_vmscan_lru_shrink_inactive(zone->zone_pgdat->node_id,
- zone_idx(zone),
-Index: linux-3.0/include/linux/vmstat.h
-===================================================================
---- linux-3.0.orig/include/linux/vmstat.h
-+++ linux-3.0/include/linux/vmstat.h
-@@ -29,7 +29,9 @@ DECLARE_PER_CPU(struct vm_event_state, v
-
- static inline void __count_vm_event(enum vm_event_item item)
- {
-+ preempt_disable_rt();
- __this_cpu_inc(vm_event_states.event[item]);
-+ preempt_enable_rt();
- }
-
- static inline void count_vm_event(enum vm_event_item item)
-@@ -39,7 +41,9 @@ static inline void count_vm_event(enum v
-
- static inline void __count_vm_events(enum vm_event_item item, long delta)
- {
-+ preempt_disable_rt();
- __this_cpu_add(vm_event_states.event[item], delta);
-+ preempt_enable_rt();
- }
-
- static inline void count_vm_events(enum vm_event_item item, long delta)
-Index: linux-3.0/mm/vmstat.c
-===================================================================
---- linux-3.0.orig/mm/vmstat.c
-+++ linux-3.0/mm/vmstat.c
-@@ -216,6 +216,7 @@ void __mod_zone_page_state(struct zone *
- long x;
- long t;
-
-+ preempt_disable_rt();
- x = delta + __this_cpu_read(*p);
-
- t = __this_cpu_read(pcp->stat_threshold);
-@@ -225,6 +226,7 @@ void __mod_zone_page_state(struct zone *
- x = 0;
- }
- __this_cpu_write(*p, x);
-+ preempt_enable_rt();
- }
- EXPORT_SYMBOL(__mod_zone_page_state);
-
-@@ -257,6 +259,7 @@ void __inc_zone_state(struct zone *zone,
- s8 __percpu *p = pcp->vm_stat_diff + item;
- s8 v, t;
-
-+ preempt_disable_rt();
- v = __this_cpu_inc_return(*p);
- t = __this_cpu_read(pcp->stat_threshold);
- if (unlikely(v > t)) {
-@@ -265,6 +268,7 @@ void __inc_zone_state(struct zone *zone,
- zone_page_state_add(v + overstep, zone, item);
- __this_cpu_write(*p, -overstep);
- }
-+ preempt_enable_rt();
- }
-
- void __inc_zone_page_state(struct page *page, enum zone_stat_item item)
-@@ -279,6 +283,7 @@ void __dec_zone_state(struct zone *zone,
- s8 __percpu *p = pcp->vm_stat_diff + item;
- s8 v, t;
-
-+ preempt_disable_rt();
- v = __this_cpu_dec_return(*p);
- t = __this_cpu_read(pcp->stat_threshold);
- if (unlikely(v < - t)) {
-@@ -287,6 +292,7 @@ void __dec_zone_state(struct zone *zone,
- zone_page_state_add(v - overstep, zone, item);
- __this_cpu_write(*p, overstep);
- }
-+ preempt_enable_rt();
- }
-
- void __dec_zone_page_state(struct page *page, enum zone_stat_item item)
-Index: linux-3.0/include/linux/mm.h
-===================================================================
---- linux-3.0.orig/include/linux/mm.h
-+++ linux-3.0/include/linux/mm.h
-@@ -1229,27 +1229,59 @@ static inline pmd_t *pmd_alloc(struct mm
- * overflow into the next struct page (as it might with DEBUG_SPINLOCK).
- * When freeing, reset page->mapping so free_pages_check won't complain.
- */
-+#ifndef CONFIG_PREEMPT_RT_FULL
++ if (index == 0) {
++ char minstr[32], avgstr[32], maxstr[32];
+
- #define __pte_lockptr(page) &((page)->ptl)
--#define pte_lock_init(_page) do { \
-- spin_lock_init(__pte_lockptr(_page)); \
--} while (0)
++ atomic_dec(&my_hist->hist_mode);
+
-+static inline struct page *pte_lock_init(struct page *page)
++ if (likely(my_hist->total_samples)) {
++ unsigned long avg = (unsigned long)
++ div64_u64(my_hist->accumulate_lat,
++ my_hist->total_samples);
++ snprintf(minstr, sizeof(minstr), "%ld",
++ (long) my_hist->min_lat - my_hist->offset);
++ snprintf(avgstr, sizeof(avgstr), "%ld",
++ (long) avg - my_hist->offset);
++ snprintf(maxstr, sizeof(maxstr), "%ld",
++ (long) my_hist->max_lat - my_hist->offset);
++ } else {
++ strcpy(minstr, "<undef>");
++ strcpy(avgstr, minstr);
++ strcpy(maxstr, minstr);
++ }
++
++ seq_printf(m, "#Minimum latency: %s microseconds\n"
++ "#Average latency: %s microseconds\n"
++ "#Maximum latency: %s microseconds\n"
++ "#Total samples: %llu\n"
++ "#There are %llu samples lower than %ld"
++ " microseconds.\n"
++ "#There are %llu samples greater or equal"
++ " than %ld microseconds.\n"
++ "#usecs\t%16s\n",
++ minstr, avgstr, maxstr,
++ my_hist->total_samples,
++ my_hist->below_hist_bound_samples,
++ -my_hist->offset,
++ my_hist->above_hist_bound_samples,
++ MAX_ENTRY_NUM - my_hist->offset,
++ "samples");
++ }
++ if (index < MAX_ENTRY_NUM) {
++ index_ptr = kmalloc(sizeof(loff_t), GFP_KERNEL);
++ if (index_ptr)
++ *index_ptr = index;
++ }
++
++ return index_ptr;
++}
++
++static void *l_next(struct seq_file *m, void *p, loff_t *pos)
+{
-+ spin_lock_init(__pte_lockptr(page));
-+ return page;
++ loff_t *index_ptr = p;
++ struct hist_data *my_hist = m->private;
++
++ if (++*pos >= MAX_ENTRY_NUM) {
++ atomic_inc(&my_hist->hist_mode);
++ return NULL;
++ }
++ *index_ptr = *pos;
++ return index_ptr;
+}
+
- #define pte_lock_deinit(page) ((page)->mapping = NULL)
++static void l_stop(struct seq_file *m, void *p)
++{
++ kfree(p);
++}
+
-+#else /* !PREEMPT_RT_FULL */
++static int l_show(struct seq_file *m, void *p)
++{
++ int index = *(loff_t *) p;
++ struct hist_data *my_hist = m->private;
+
-+/*
-+ * On PREEMPT_RT_FULL the spinlock_t's are too large to embed in the
-+ * page frame, hence it only has a pointer and we need to dynamically
-+ * allocate the lock when we allocate PTE-pages.
-+ *
-+ * This is an overall win, since only a small fraction of the pages
-+ * will be PTE pages under normal circumstances.
-+ */
++ seq_printf(m, "%6ld\t%16llu\n", index - my_hist->offset,
++ my_hist->hist_array[index]);
++ return 0;
++}
++
++static struct seq_operations latency_hist_seq_op = {
++ .start = l_start,
++ .next = l_next,
++ .stop = l_stop,
++ .show = l_show
++};
++
++static int latency_hist_open(struct inode *inode, struct file *file)
++{
++ int ret;
++
++ ret = seq_open(file, &latency_hist_seq_op);
++ if (!ret) {
++ struct seq_file *seq = file->private_data;
++ seq->private = inode->i_private;
++ }
++ return ret;
++}
++
++static struct file_operations latency_hist_fops = {
++ .open = latency_hist_open,
++ .read = seq_read,
++ .llseek = seq_lseek,
++ .release = seq_release,
++};
++
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++static void clear_maxlatprocdata(struct maxlatproc_data *mp)
++{
++ mp->comm[0] = mp->current_comm[0] = '\0';
++ mp->prio = mp->current_prio = mp->pid = mp->current_pid =
++ mp->latency = mp->timeroffset = -1;
++ mp->timestamp = 0;
++}
++#endif
++
++static void hist_reset(struct hist_data *hist)
++{
++ atomic_dec(&hist->hist_mode);
+
-+#define __pte_lockptr(page) ((page)->ptl)
++ memset(hist->hist_array, 0, sizeof(hist->hist_array));
++ hist->below_hist_bound_samples = 0ULL;
++ hist->above_hist_bound_samples = 0ULL;
++ hist->min_lat = ULONG_MAX;
++ hist->max_lat = 0UL;
++ hist->total_samples = 0ULL;
++ hist->accumulate_lat = 0ULL;
+
-+extern struct page *pte_lock_init(struct page *page);
-+extern void pte_lock_deinit(struct page *page);
++ atomic_inc(&hist->hist_mode);
++}
+
-+#endif /* PREEMPT_RT_FULL */
++static ssize_t
++latency_hist_reset(struct file *file, const char __user *a,
++ size_t size, loff_t *off)
++{
++ int cpu;
++ struct hist_data *hist = NULL;
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ struct maxlatproc_data *mp = NULL;
++#endif
++ off_t latency_type = (off_t) file->private_data;
+
- #define pte_lockptr(mm, pmd) ({(void)(mm); __pte_lockptr(pmd_page(*(pmd)));})
- #else /* !USE_SPLIT_PTLOCKS */
- /*
- * We use mm->page_table_lock to guard all pagetable pages of the mm.
- */
--#define pte_lock_init(page) do {} while (0)
-+static inline struct page *pte_lock_init(struct page *page) { return page; }
- #define pte_lock_deinit(page) do {} while (0)
- #define pte_lockptr(mm, pmd) ({(void)(pmd); &(mm)->page_table_lock;})
- #endif /* USE_SPLIT_PTLOCKS */
-
--static inline void pgtable_page_ctor(struct page *page)
-+static inline struct page *__pgtable_page_ctor(struct page *page)
- {
-- pte_lock_init(page);
-- inc_zone_page_state(page, NR_PAGETABLE);
-+ page = pte_lock_init(page);
-+ if (page)
-+ inc_zone_page_state(page, NR_PAGETABLE);
-+ return page;
- }
-
-+#define pgtable_page_ctor(page) \
-+do { \
-+ page = __pgtable_page_ctor(page); \
-+} while (0)
++ for_each_online_cpu(cpu) {
+
- static inline void pgtable_page_dtor(struct page *page)
- {
- pte_lock_deinit(page);
-Index: linux-3.0/include/linux/mm_types.h
-===================================================================
---- linux-3.0.orig/include/linux/mm_types.h
-+++ linux-3.0/include/linux/mm_types.h
-@@ -12,6 +12,7 @@
- #include <linux/completion.h>
- #include <linux/cpumask.h>
- #include <linux/page-debug-flags.h>
-+#include <linux/rcupdate.h>
- #include <asm/page.h>
- #include <asm/mmu.h>
-
-@@ -77,7 +78,11 @@ struct page {
- */
- };
- #if USE_SPLIT_PTLOCKS
-+#ifndef CONFIG_PREEMPT_RT_FULL
- spinlock_t ptl;
-+#else
-+ spinlock_t *ptl;
++ switch (latency_type) {
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ case PREEMPTOFF_LATENCY:
++ hist = &per_cpu(preemptoff_hist, cpu);
++ break;
+#endif
- #endif
- struct kmem_cache *slab; /* SLUB: Pointer to slab */
- struct page *first_page; /* Compound tail pages */
-@@ -330,6 +335,9 @@ struct mm_struct {
- #ifdef CONFIG_CPUMASK_OFFSTACK
- struct cpumask cpumask_allocation;
- #endif
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ struct rcu_head delayed_drop;
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ case IRQSOFF_LATENCY:
++ hist = &per_cpu(irqsoff_hist, cpu);
++ break;
+#endif
- };
-
- static inline void mm_init_cpumask(struct mm_struct *mm)
-Index: linux-3.0/init/Kconfig
-===================================================================
---- linux-3.0.orig/init/Kconfig
-+++ linux-3.0/init/Kconfig
-@@ -401,7 +401,7 @@ config TREE_PREEMPT_RCU
-
- config TINY_RCU
- bool "UP-only small-memory-footprint RCU"
-- depends on !SMP
-+ depends on !SMP && !PREEMPT_RT_FULL
- help
- This option selects the RCU implementation that is
- designed for UP systems from which real-time response
-@@ -410,7 +410,7 @@ config TINY_RCU
-
- config TINY_PREEMPT_RCU
- bool "Preemptible UP-only small-memory-footprint RCU"
-- depends on !SMP && PREEMPT
-+ depends on !SMP && PREEMPT && !PREEMPT_RT_FULL
- help
- This option selects the RCU implementation that is designed
- for real-time UP systems. This option greatly reduces the
-@@ -1220,6 +1220,7 @@ config SLAB
-
- config SLUB
- bool "SLUB (Unqueued Allocator)"
-+ depends on !PREEMPT_RT_FULL
- help
- SLUB is a slab allocator that minimizes cache line usage
- instead of managing queues of cached objects (SLAB approach).
-@@ -1231,6 +1232,7 @@ config SLUB
- config SLOB
- depends on EXPERT
- bool "SLOB (Simple Allocator)"
-+ depends on !PREEMPT_RT_FULL
- help
- SLOB replaces the stock allocator with a drastically simpler
- allocator. SLOB is generally more space efficient but
-Index: linux-3.0/include/linux/radix-tree.h
-===================================================================
---- linux-3.0.orig/include/linux/radix-tree.h
-+++ linux-3.0/include/linux/radix-tree.h
-@@ -201,7 +201,13 @@ unsigned long radix_tree_next_hole(struc
- unsigned long index, unsigned long max_scan);
- unsigned long radix_tree_prev_hole(struct radix_tree_root *root,
- unsigned long index, unsigned long max_scan);
-+
-+#ifndef CONFIG_PREEMPT_RT_FULL
- int radix_tree_preload(gfp_t gfp_mask);
-+#else
-+static inline int radix_tree_preload(gfp_t gm) { return 0; }
++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
++ case PREEMPTIRQSOFF_LATENCY:
++ hist = &per_cpu(preemptirqsoff_hist, cpu);
++ break;
+#endif
-+
- void radix_tree_init(void);
- void *radix_tree_tag_set(struct radix_tree_root *root,
- unsigned long index, unsigned int tag);
-@@ -225,7 +231,7 @@ int radix_tree_tagged(struct radix_tree_
-
- static inline void radix_tree_preload_end(void)
- {
-- preempt_enable();
-+ preempt_enable_nort();
- }
-
- #endif /* _LINUX_RADIX_TREE_H */
-Index: linux-3.0/lib/radix-tree.c
-===================================================================
---- linux-3.0.orig/lib/radix-tree.c
-+++ linux-3.0/lib/radix-tree.c
-@@ -166,12 +166,13 @@ radix_tree_node_alloc(struct radix_tree_
- * succeed in getting a node here (and never reach
- * kmem_cache_alloc)
- */
-- rtp = &__get_cpu_var(radix_tree_preloads);
-+ rtp = &get_cpu_var(radix_tree_preloads);
- if (rtp->nr) {
- ret = rtp->nodes[rtp->nr - 1];
- rtp->nodes[rtp->nr - 1] = NULL;
- rtp->nr--;
- }
-+ put_cpu_var(radix_tree_preloads);
- }
- if (ret == NULL)
- ret = kmem_cache_alloc(radix_tree_node_cachep, gfp_mask);
-@@ -206,6 +207,7 @@ radix_tree_node_free(struct radix_tree_n
- call_rcu(&node->rcu_head, radix_tree_node_rcu_free);
- }
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /*
- * Load up this CPU's radix_tree_node buffer with sufficient objects to
- * ensure that the addition of a single element in the tree cannot fail. On
-@@ -240,6 +242,7 @@ out:
- return ret;
- }
- EXPORT_SYMBOL(radix_tree_preload);
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ case WAKEUP_LATENCY:
++ hist = &per_cpu(wakeup_latency_hist, cpu);
++ mp = &per_cpu(wakeup_maxlatproc, cpu);
++ break;
++ case WAKEUP_LATENCY_SHAREDPRIO:
++ hist = &per_cpu(wakeup_latency_hist_sharedprio, cpu);
++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, cpu);
++ break;
+#endif
-
- /*
- * Return the maximum key which can be store into a
-Index: linux-3.0/kernel/panic.c
-===================================================================
---- linux-3.0.orig/kernel/panic.c
-+++ linux-3.0/kernel/panic.c
-@@ -330,9 +330,11 @@ static u64 oops_id;
-
- static int init_oops_id(void)
- {
-+#ifndef CONFIG_PREEMPT_RT_FULL
- if (!oops_id)
- get_random_bytes(&oops_id, sizeof(oops_id));
- else
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ case MISSED_TIMER_OFFSETS:
++ hist = &per_cpu(missed_timer_offsets, cpu);
++ mp = &per_cpu(missed_timer_offsets_maxlatproc, cpu);
++ break;
+#endif
- oops_id++;
-
- return 0;
-Index: linux-3.0/ipc/mqueue.c
-===================================================================
---- linux-3.0.orig/ipc/mqueue.c
-+++ linux-3.0/ipc/mqueue.c
-@@ -820,12 +820,17 @@ static inline void pipelined_send(struct
- struct msg_msg *message,
- struct ext_wait_queue *receiver)
- {
-+ /*
-+ * Keep them in one critical section for PREEMPT_RT:
-+ */
-+ preempt_disable_rt();
- receiver->msg = message;
- list_del(&receiver->list);
- receiver->state = STATE_PENDING;
- wake_up_process(receiver->task);
- smp_wmb();
- receiver->state = STATE_READY;
-+ preempt_enable_rt();
- }
-
- /* pipelined_receive() - if there is task waiting in sys_mq_timedsend()
-@@ -839,15 +844,19 @@ static inline void pipelined_receive(str
- wake_up_interruptible(&info->wait_q);
- return;
- }
-+ /*
-+ * Keep them in one critical section for PREEMPT_RT:
-+ */
-+ preempt_disable_rt();
- msg_insert(sender->msg, info);
- list_del(&sender->list);
- sender->state = STATE_PENDING;
- wake_up_process(sender->task);
- smp_wmb();
- sender->state = STATE_READY;
-+ preempt_enable_rt();
- }
--
--SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr,
-+ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr,
- size_t, msg_len, unsigned int, msg_prio,
- const struct timespec __user *, u_abs_timeout)
- {
-Index: linux-3.0/ipc/msg.c
-===================================================================
---- linux-3.0.orig/ipc/msg.c
-+++ linux-3.0/ipc/msg.c
-@@ -259,12 +259,20 @@ static void expunge_all(struct msg_queue
- while (tmp != &msq->q_receivers) {
- struct msg_receiver *msr;
-
-+ /*
-+ * Make sure that the wakeup doesnt preempt
-+ * this CPU prematurely. (on PREEMPT_RT)
-+ */
-+ preempt_disable_rt();
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ case TIMERANDWAKEUP_LATENCY:
++ hist = &per_cpu(timerandwakeup_latency_hist, cpu);
++ mp = &per_cpu(timerandwakeup_maxlatproc, cpu);
++ break;
++#endif
++ }
+
- msr = list_entry(tmp, struct msg_receiver, r_list);
- tmp = tmp->next;
- msr->r_msg = NULL;
- wake_up_process(msr->r_tsk);
- smp_mb();
- msr->r_msg = ERR_PTR(res);
++ hist_reset(hist);
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ if (latency_type == WAKEUP_LATENCY ||
++ latency_type == WAKEUP_LATENCY_SHAREDPRIO ||
++ latency_type == MISSED_TIMER_OFFSETS ||
++ latency_type == TIMERANDWAKEUP_LATENCY)
++ clear_maxlatprocdata(mp);
++#endif
++ }
+
-+ preempt_enable_rt();
- }
- }
-
-@@ -611,6 +619,12 @@ static inline int pipelined_send(struct
- !security_msg_queue_msgrcv(msq, msg, msr->r_tsk,
- msr->r_msgtype, msr->r_mode)) {
-
-+ /*
-+ * Make sure that the wakeup doesnt preempt
-+ * this CPU prematurely. (on PREEMPT_RT)
-+ */
-+ preempt_disable_rt();
++ return size;
++}
+
- list_del(&msr->r_list);
- if (msr->r_maxsize < msg->m_ts) {
- msr->r_msg = NULL;
-@@ -624,9 +638,11 @@ static inline int pipelined_send(struct
- wake_up_process(msr->r_tsk);
- smp_mb();
- msr->r_msg = msg;
-+ preempt_enable_rt();
-
- return 1;
- }
-+ preempt_enable_rt();
- }
- }
- return 0;
-Index: linux-3.0/kernel/relay.c
-===================================================================
---- linux-3.0.orig/kernel/relay.c
-+++ linux-3.0/kernel/relay.c
-@@ -336,6 +336,10 @@ static void wakeup_readers(unsigned long
- {
- struct rchan_buf *buf = (struct rchan_buf *)data;
- wake_up_interruptible(&buf->read_wait);
-+ /*
-+ * Stupid polling for now:
-+ */
-+ mod_timer(&buf->timer, jiffies + 1);
- }
-
- /**
-@@ -353,6 +357,7 @@ static void __relay_reset(struct rchan_b
- init_waitqueue_head(&buf->read_wait);
- kref_init(&buf->kref);
- setup_timer(&buf->timer, wakeup_readers, (unsigned long)buf);
-+ mod_timer(&buf->timer, jiffies + 1);
- } else
- del_timer_sync(&buf->timer);
-
-@@ -733,15 +738,6 @@ size_t relay_switch_subbuf(struct rchan_
- else
- buf->early_bytes += buf->chan->subbuf_size -
- buf->padding[old_subbuf];
-- smp_mb();
-- if (waitqueue_active(&buf->read_wait))
-- /*
-- * Calling wake_up_interruptible() from here
-- * will deadlock if we happen to be logging
-- * from the scheduler (trying to re-grab
-- * rq->lock), so defer it.
-- */
-- mod_timer(&buf->timer, jiffies + 1);
- }
-
- old = buf->data;
-Index: linux-3.0/net/ipv4/route.c
-===================================================================
---- linux-3.0.orig/net/ipv4/route.c
-+++ linux-3.0/net/ipv4/route.c
-@@ -242,7 +242,7 @@ struct rt_hash_bucket {
- };
-
- #if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \
-- defined(CONFIG_PROVE_LOCKING)
-+ defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_PREEMPT_RT_FULL)
- /*
- * Instead of using one spinlock for each rt_hash_bucket, we use a table of spinlocks
- * The size of this table is a power of two and depends on the number of CPUS.
-Index: linux-3.0/include/linux/timer.h
-===================================================================
---- linux-3.0.orig/include/linux/timer.h
-+++ linux-3.0/include/linux/timer.h
-@@ -276,7 +276,7 @@ extern void add_timer(struct timer_list
-
- extern int try_to_del_timer_sync(struct timer_list *timer);
-
--#ifdef CONFIG_SMP
-+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL)
- extern int del_timer_sync(struct timer_list *timer);
- #else
- # define del_timer_sync(t) del_timer(t)
-Index: linux-3.0/kernel/timer.c
-===================================================================
---- linux-3.0.orig/kernel/timer.c
-+++ linux-3.0/kernel/timer.c
-@@ -75,6 +75,7 @@ struct tvec_root {
- struct tvec_base {
- spinlock_t lock;
- struct timer_list *running_timer;
-+ wait_queue_head_t wait_for_running_timer;
- unsigned long timer_jiffies;
- unsigned long next_timer;
- struct tvec_root tv1;
-@@ -653,6 +654,36 @@ static struct tvec_base *lock_timer_base
- }
- }
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+static inline struct tvec_base *switch_timer_base(struct timer_list *timer,
-+ struct tvec_base *old,
-+ struct tvec_base *new)
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++static ssize_t
++show_pid(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
+{
-+ /* See the comment in lock_timer_base() */
-+ timer_set_base(timer, NULL);
-+ spin_unlock(&old->lock);
-+ spin_lock(&new->lock);
-+ timer_set_base(timer, new);
-+ return new;
++ char buf[64];
++ int r;
++ unsigned long *this_pid = file->private_data;
++
++ r = snprintf(buf, sizeof(buf), "%lu\n", *this_pid);
++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
+}
-+#else
-+static inline struct tvec_base *switch_timer_base(struct timer_list *timer,
-+ struct tvec_base *old,
-+ struct tvec_base *new)
++
++static ssize_t do_pid(struct file *file, const char __user *ubuf,
++ size_t cnt, loff_t *ppos)
+{
-+ /*
-+ * We cannot do the above because we might be preempted and
-+ * then the preempter would see NULL and loop forever.
-+ */
-+ if (spin_trylock(&new->lock)) {
-+ timer_set_base(timer, new);
-+ spin_unlock(&old->lock);
-+ return new;
++ char buf[64];
++ unsigned long pid;
++ unsigned long *this_pid = file->private_data;
++
++ if (cnt >= sizeof(buf))
++ return -EINVAL;
++
++ if (copy_from_user(&buf, ubuf, cnt))
++ return -EFAULT;
++
++ buf[cnt] = '\0';
++
++ if (strict_strtoul(buf, 10, &pid))
++ return(-EINVAL);
++
++ *this_pid = pid;
++
++ return cnt;
++}
++#endif
++
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++static ssize_t
++show_maxlatproc(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
++{
++ int r;
++ struct maxlatproc_data *mp = file->private_data;
++ int strmaxlen = (TASK_COMM_LEN * 2) + (8 * 8);
++ unsigned long long t;
++ unsigned long usecs, secs;
++ char *buf;
++
++ if (mp->pid == -1 || mp->current_pid == -1) {
++ buf = "(none)\n";
++ return simple_read_from_buffer(ubuf, cnt, ppos, buf,
++ strlen(buf));
+ }
-+ return old;
++
++ buf = kmalloc(strmaxlen, GFP_KERNEL);
++ if (buf == NULL)
++ return -ENOMEM;
++
++ t = ns2usecs(mp->timestamp);
++ usecs = do_div(t, USEC_PER_SEC);
++ secs = (unsigned long) t;
++ r = snprintf(buf, strmaxlen,
++ "%d %d %ld (%ld) %s <- %d %d %s %lu.%06lu\n", mp->pid,
++ MAX_RT_PRIO-1 - mp->prio, mp->latency, mp->timeroffset, mp->comm,
++ mp->current_pid, MAX_RT_PRIO-1 - mp->current_prio, mp->current_comm,
++ secs, usecs);
++ r = simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
++ kfree(buf);
++ return r;
+}
+#endif
+
- static inline int
- __mod_timer(struct timer_list *timer, unsigned long expires,
- bool pending_only, int pinned)
-@@ -679,12 +710,15 @@ __mod_timer(struct timer_list *timer, un
-
- debug_activate(timer, expires);
-
-+ preempt_disable_rt();
- cpu = smp_processor_id();
-
- #if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
- if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu))
- cpu = get_nohz_timer_target();
- #endif
-+ preempt_enable_rt();
++static ssize_t
++show_enable(struct file *file, char __user *ubuf, size_t cnt, loff_t *ppos)
++{
++ char buf[64];
++ struct enable_data *ed = file->private_data;
++ int r;
+
- new_base = per_cpu(tvec_bases, cpu);
-
- if (base != new_base) {
-@@ -695,14 +729,8 @@ __mod_timer(struct timer_list *timer, un
- * handler yet has not finished. This also guarantees that
- * the timer is serialized wrt itself.
- */
-- if (likely(base->running_timer != timer)) {
-- /* See the comment in lock_timer_base() */
-- timer_set_base(timer, NULL);
-- spin_unlock(&base->lock);
-- base = new_base;
-- spin_lock(&base->lock);
-- timer_set_base(timer, base);
-- }
-+ if (likely(base->running_timer != timer))
-+ base = switch_timer_base(timer, base, new_base);
- }
-
- timer->expires = expires;
-@@ -885,6 +913,29 @@ void add_timer_on(struct timer_list *tim
- }
- EXPORT_SYMBOL_GPL(add_timer_on);
-
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+/*
-+ * Wait for a running timer
-+ */
-+static void wait_for_running_timer(struct timer_list *timer)
++ r = snprintf(buf, sizeof(buf), "%d\n", ed->enabled);
++ return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
++}
++
++static ssize_t
++do_enable(struct file *file, const char __user *ubuf, size_t cnt, loff_t *ppos)
+{
-+ struct tvec_base *base = timer->base;
++ char buf[64];
++ long enable;
++ struct enable_data *ed = file->private_data;
++
++ if (cnt >= sizeof(buf))
++ return -EINVAL;
+
-+ if (base->running_timer == timer)
-+ wait_event(base->wait_for_running_timer,
-+ base->running_timer != timer);
-+}
++ if (copy_from_user(&buf, ubuf, cnt))
++ return -EFAULT;
+
-+# define wakeup_timer_waiters(b) wake_up(&(b)->wait_for_tunning_timer)
-+#else
-+static inline void wait_for_running_timer(struct timer_list *timer)
-+{
-+ cpu_relax();
-+}
++ buf[cnt] = 0;
+
-+# define wakeup_timer_waiters(b) do { } while (0)
-+#endif
++ if (strict_strtol(buf, 10, &enable))
++ return(-EINVAL);
+
- /**
- * del_timer - deactive a timer.
- * @timer: the timer to be deactivated
-@@ -953,7 +1004,7 @@ out:
- }
- EXPORT_SYMBOL(try_to_del_timer_sync);
-
--#ifdef CONFIG_SMP
-+#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT_RT_FULL)
- /**
- * del_timer_sync - deactivate a timer and wait for the handler to finish.
- * @timer: the timer to be deactivated
-@@ -1013,7 +1064,7 @@ int del_timer_sync(struct timer_list *ti
- int ret = try_to_del_timer_sync(timer);
- if (ret >= 0)
- return ret;
-- cpu_relax();
-+ wait_for_running_timer(timer);
- }
- }
- EXPORT_SYMBOL(del_timer_sync);
-@@ -1124,10 +1175,11 @@ static inline void __run_timers(struct t
-
- spin_unlock_irq(&base->lock);
- call_timer_fn(timer, fn, data);
-+ base->running_timer = NULL;
- spin_lock_irq(&base->lock);
- }
- }
-- base->running_timer = NULL;
-+ wake_up(&base->wait_for_running_timer);
- spin_unlock_irq(&base->lock);
- }
-
-@@ -1266,6 +1318,23 @@ unsigned long get_next_timer_interrupt(u
- */
- if (cpu_is_offline(smp_processor_id()))
- return now + NEXT_TIMER_MAX_DELTA;
++ if ((enable && ed->enabled) || (!enable && !ed->enabled))
++ return cnt;
+
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ /*
-+ * On PREEMPT_RT we cannot sleep here. If the trylock does not
-+ * succeed then we return the worst-case 'expires in 1 tick'
-+ * value. We use the rt functions here directly to avoid a
-+ * migrate_disable() call.
-+ */
-+ if (spin_do_trylock(&base->lock)) {
-+ if (time_before_eq(base->next_timer, base->timer_jiffies))
-+ base->next_timer = __next_timer_interrupt(base);
-+ expires = base->next_timer;
-+ rt_spin_unlock(&base->lock);
-+ } else {
-+ expires = now + 1;
-+ }
-+#else
- spin_lock(&base->lock);
- if (time_before_eq(base->next_timer, base->timer_jiffies))
- base->next_timer = __next_timer_interrupt(base);
-@@ -1274,7 +1343,7 @@ unsigned long get_next_timer_interrupt(u
-
- if (time_before_eq(expires, now))
- return now;
--
++ if (enable) {
++ int ret;
++
++ switch (ed->latency_type) {
++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
++ case PREEMPTIRQSOFF_LATENCY:
++ ret = register_trace_preemptirqsoff_hist(
++ probe_preemptirqsoff_hist, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_preemptirqsoff_hist "
++ "to trace_preemptirqsoff_hist\n");
++ return ret;
++ }
++ break;
+#endif
- return cmp_next_hrtimer_event(now, expires);
- }
- #endif
-@@ -1290,14 +1359,13 @@ void update_process_times(int user_tick)
-
- /* Note: this timer irq context must be accounted for as well. */
- account_process_tick(p, user_tick);
-+ scheduler_tick();
- run_local_timers();
- rcu_check_callbacks(cpu, user_tick);
-- printk_tick();
--#ifdef CONFIG_IRQ_WORK
-+#if defined(CONFIG_IRQ_WORK) && !defined(CONFIG_PREEMPT_RT_FULL)
- if (in_irq())
- irq_work_run();
- #endif
-- scheduler_tick();
- run_posix_cpu_timers(p);
- }
-
-@@ -1308,6 +1376,11 @@ static void run_timer_softirq(struct sof
- {
- struct tvec_base *base = __this_cpu_read(tvec_bases);
-
-+#if defined(CONFIG_IRQ_WORK)
-+ irq_work_run();
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ case WAKEUP_LATENCY:
++ ret = register_trace_sched_wakeup(
++ probe_wakeup_latency_hist_start, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_wakeup_latency_hist_start "
++ "to trace_sched_wakeup\n");
++ return ret;
++ }
++ ret = register_trace_sched_wakeup_new(
++ probe_wakeup_latency_hist_start, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_wakeup_latency_hist_start "
++ "to trace_sched_wakeup_new\n");
++ unregister_trace_sched_wakeup(
++ probe_wakeup_latency_hist_start, NULL);
++ return ret;
++ }
++ ret = register_trace_sched_switch(
++ probe_wakeup_latency_hist_stop, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_wakeup_latency_hist_stop "
++ "to trace_sched_switch\n");
++ unregister_trace_sched_wakeup(
++ probe_wakeup_latency_hist_start, NULL);
++ unregister_trace_sched_wakeup_new(
++ probe_wakeup_latency_hist_start, NULL);
++ return ret;
++ }
++ ret = register_trace_sched_migrate_task(
++ probe_sched_migrate_task, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_sched_migrate_task "
++ "to trace_sched_migrate_task\n");
++ unregister_trace_sched_wakeup(
++ probe_wakeup_latency_hist_start, NULL);
++ unregister_trace_sched_wakeup_new(
++ probe_wakeup_latency_hist_start, NULL);
++ unregister_trace_sched_switch(
++ probe_wakeup_latency_hist_stop, NULL);
++ return ret;
++ }
++ break;
+#endif
-+
-+ printk_tick();
- hrtimer_run_pending();
-
- if (time_after_eq(jiffies, base->timer_jiffies))
-@@ -1634,6 +1707,7 @@ static int __cpuinit init_timers_cpu(int
- }
-
- spin_lock_init(&base->lock);
-+ init_waitqueue_head(&base->wait_for_running_timer);
-
- for (j = 0; j < TVN_SIZE; j++) {
- INIT_LIST_HEAD(base->tv5.vec + j);
-@@ -1669,6 +1743,7 @@ static void __cpuinit migrate_timers(int
- {
- struct tvec_base *old_base;
- struct tvec_base *new_base;
-+ unsigned long flags;
- int i;
-
- BUG_ON(cpu_online(cpu));
-@@ -1678,8 +1753,11 @@ static void __cpuinit migrate_timers(int
- * The caller is globally serialized and nobody else
- * takes two locks at once, deadlock is not possible.
- */
-- spin_lock_irq(&new_base->lock);
-- spin_lock_nested(&old_base->lock, SINGLE_DEPTH_NESTING);
-+ local_irq_save(flags);
-+ while (!spin_trylock(&new_base->lock))
-+ cpu_relax();
-+ while (!spin_trylock(&old_base->lock))
-+ cpu_relax();
-
- BUG_ON(old_base->running_timer);
-
-@@ -1693,7 +1771,9 @@ static void __cpuinit migrate_timers(int
- }
-
- spin_unlock(&old_base->lock);
-- spin_unlock_irq(&new_base->lock);
-+ spin_unlock(&new_base->lock);
-+ local_irq_restore(flags);
-+
- put_cpu_var(tvec_bases);
- }
- #endif /* CONFIG_HOTPLUG_CPU */
-Index: linux-3.0/include/linux/hrtimer.h
-===================================================================
---- linux-3.0.orig/include/linux/hrtimer.h
-+++ linux-3.0/include/linux/hrtimer.h
-@@ -111,6 +111,8 @@ struct hrtimer {
- enum hrtimer_restart (*function)(struct hrtimer *);
- struct hrtimer_clock_base *base;
- unsigned long state;
-+ struct list_head cb_entry;
-+ int irqsafe;
- #ifdef CONFIG_TIMER_STATS
- int start_pid;
- void *start_site;
-@@ -147,6 +149,7 @@ struct hrtimer_clock_base {
- int index;
- clockid_t clockid;
- struct timerqueue_head active;
-+ struct list_head expired;
- ktime_t resolution;
- ktime_t (*get_time)(void);
- ktime_t softirq_time;
-@@ -187,6 +190,9 @@ struct hrtimer_cpu_base {
- unsigned long nr_hangs;
- ktime_t max_hang_time;
- #endif
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ wait_queue_head_t wait;
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ case MISSED_TIMER_OFFSETS:
++ ret = register_trace_hrtimer_interrupt(
++ probe_hrtimer_interrupt, NULL);
++ if (ret) {
++ pr_info("wakeup trace: Couldn't assign "
++ "probe_hrtimer_interrupt "
++ "to trace_hrtimer_interrupt\n");
++ return ret;
++ }
++ break;
++#endif
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ case TIMERANDWAKEUP_LATENCY:
++ if (!wakeup_latency_enabled_data.enabled ||
++ !missed_timer_offsets_enabled_data.enabled)
++ return -EINVAL;
++ break;
++#endif
++ default:
++ break;
++ }
++ } else {
++ switch (ed->latency_type) {
++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
++ case PREEMPTIRQSOFF_LATENCY:
++ {
++ int cpu;
++
++ unregister_trace_preemptirqsoff_hist(
++ probe_preemptirqsoff_hist, NULL);
++ for_each_online_cpu(cpu) {
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ per_cpu(hist_irqsoff_counting,
++ cpu) = 0;
+#endif
- struct hrtimer_clock_base clock_base[HRTIMER_MAX_CLOCK_BASES];
- };
-
-@@ -374,6 +380,13 @@ static inline int hrtimer_restart(struct
- return hrtimer_start_expires(timer, HRTIMER_MODE_ABS);
- }
-
-+/* Softirq preemption could deadlock timer removal */
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ extern void hrtimer_wait_for_timer(const struct hrtimer *timer);
-+#else
-+# define hrtimer_wait_for_timer(timer) do { cpu_relax(); } while (0)
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ per_cpu(hist_preemptoff_counting,
++ cpu) = 0;
++#endif
++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
++ per_cpu(hist_preemptirqsoff_counting,
++ cpu) = 0;
++#endif
++ }
++ }
++ break;
+#endif
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ case WAKEUP_LATENCY:
++ {
++ int cpu;
+
- /* Query timers: */
- extern ktime_t hrtimer_get_remaining(const struct hrtimer *timer);
- extern int hrtimer_get_res(const clockid_t which_clock, struct timespec *tp);
-Index: linux-3.0/kernel/itimer.c
-===================================================================
---- linux-3.0.orig/kernel/itimer.c
-+++ linux-3.0/kernel/itimer.c
-@@ -214,6 +214,7 @@ again:
- /* We are sharing ->siglock with it_real_fn() */
- if (hrtimer_try_to_cancel(timer) < 0) {
- spin_unlock_irq(&tsk->sighand->siglock);
-+ hrtimer_wait_for_timer(&tsk->signal->real_timer);
- goto again;
- }
- expires = timeval_to_ktime(value->it_value);
-Index: linux-3.0/include/linux/rcupdate.h
-===================================================================
---- linux-3.0.orig/include/linux/rcupdate.h
-+++ linux-3.0/include/linux/rcupdate.h
-@@ -78,7 +78,13 @@ struct rcu_head {
- extern void call_rcu_sched(struct rcu_head *head,
- void (*func)(struct rcu_head *rcu));
- extern void synchronize_sched(void);
++ unregister_trace_sched_wakeup(
++ probe_wakeup_latency_hist_start, NULL);
++ unregister_trace_sched_wakeup_new(
++ probe_wakeup_latency_hist_start, NULL);
++ unregister_trace_sched_switch(
++ probe_wakeup_latency_hist_stop, NULL);
++ unregister_trace_sched_migrate_task(
++ probe_sched_migrate_task, NULL);
+
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# define rcu_barrier_bh rcu_barrier
-+#else
- extern void rcu_barrier_bh(void);
++ for_each_online_cpu(cpu) {
++ per_cpu(wakeup_task, cpu) = NULL;
++ per_cpu(wakeup_sharedprio, cpu) = 0;
++ }
++ }
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ timerandwakeup_enabled_data.enabled = 0;
++#endif
++ break;
++#endif
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ case MISSED_TIMER_OFFSETS:
++ unregister_trace_hrtimer_interrupt(
++ probe_hrtimer_interrupt, NULL);
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ timerandwakeup_enabled_data.enabled = 0;
++#endif
++ break;
+#endif
++ default:
++ break;
++ }
++ }
++ ed->enabled = enable;
++ return cnt;
++}
+
- extern void rcu_barrier_sched(void);
-
- static inline void __rcu_read_lock_bh(void)
-@@ -104,6 +110,11 @@ void synchronize_rcu(void);
- * types of kernel builds, the rcu_read_lock() nesting depth is unknowable.
- */
- #define rcu_preempt_depth() (current->rcu_read_lock_nesting)
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+#define sched_rcu_preempt_depth() rcu_preempt_depth()
-+#else
-+static inline int sched_rcu_preempt_depth(void) { return 0; }
++static const struct file_operations latency_hist_reset_fops = {
++ .open = tracing_open_generic,
++ .write = latency_hist_reset,
++};
++
++static const struct file_operations enable_fops = {
++ .open = tracing_open_generic,
++ .read = show_enable,
++ .write = do_enable,
++};
++
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++static const struct file_operations pid_fops = {
++ .open = tracing_open_generic,
++ .read = show_pid,
++ .write = do_pid,
++};
++
++static const struct file_operations maxlatproc_fops = {
++ .open = tracing_open_generic,
++ .read = show_maxlatproc,
++};
+#endif
-
- #else /* #ifdef CONFIG_PREEMPT_RCU */
-
-@@ -127,6 +138,8 @@ static inline int rcu_preempt_depth(void
- return 0;
- }
-
-+#define sched_rcu_preempt_depth() rcu_preempt_depth()
+
- #endif /* #else #ifdef CONFIG_PREEMPT_RCU */
-
- /* Internal to kernel */
-@@ -222,7 +235,14 @@ static inline int rcu_read_lock_held(voi
- * rcu_read_lock_bh_held() is defined out of line to avoid #include-file
- * hell.
- */
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+static inline int rcu_read_lock_bh_held(void)
++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
++static notrace void probe_preemptirqsoff_hist(void *v, int reason,
++ int starthist)
+{
-+ return rcu_read_lock_held();
-+}
-+#else
- extern int rcu_read_lock_bh_held(void);
-+#endif
-
- /**
- * rcu_read_lock_sched_held() - might we be in RCU-sched read-side critical section?
-@@ -631,8 +651,13 @@ static inline void rcu_read_unlock(void)
- static inline void rcu_read_lock_bh(void)
- {
- __rcu_read_lock_bh();
++ int cpu = raw_smp_processor_id();
++ int time_set = 0;
+
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ rcu_read_lock();
-+#else
- __acquire(RCU_BH);
- rcu_read_acquire_bh();
++ if (starthist) {
++ cycle_t uninitialized_var(start);
++
++ if (!preempt_count() && !irqs_disabled())
++ return;
++
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ if ((reason == IRQS_OFF || reason == TRACE_START) &&
++ !per_cpu(hist_irqsoff_counting, cpu)) {
++ per_cpu(hist_irqsoff_counting, cpu) = 1;
++ start = ftrace_now(cpu);
++ time_set++;
++ per_cpu(hist_irqsoff_start, cpu) = start;
++ }
+#endif
- }
-
- /*
-@@ -642,8 +667,12 @@ static inline void rcu_read_lock_bh(void
- */
- static inline void rcu_read_unlock_bh(void)
- {
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ rcu_read_unlock();
-+#else
- rcu_read_release_bh();
- __release(RCU_BH);
++
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ if ((reason == PREEMPT_OFF || reason == TRACE_START) &&
++ !per_cpu(hist_preemptoff_counting, cpu)) {
++ per_cpu(hist_preemptoff_counting, cpu) = 1;
++ if (!(time_set++))
++ start = ftrace_now(cpu);
++ per_cpu(hist_preemptoff_start, cpu) = start;
++ }
+#endif
- __rcu_read_unlock_bh();
- }
-
-@@ -750,6 +779,9 @@ extern void call_rcu(struct rcu_head *he
-
- #endif /* #else #ifdef CONFIG_PREEMPT_RCU */
-
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+#define call_rcu_bh call_rcu
-+#else
- /**
- * call_rcu_bh() - Queue an RCU for invocation after a quicker grace period.
- * @head: structure to be used for queueing the RCU updates.
-@@ -770,6 +802,7 @@ extern void call_rcu(struct rcu_head *he
- */
- extern void call_rcu_bh(struct rcu_head *head,
- void (*func)(struct rcu_head *head));
++
++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
++ if (per_cpu(hist_irqsoff_counting, cpu) &&
++ per_cpu(hist_preemptoff_counting, cpu) &&
++ !per_cpu(hist_preemptirqsoff_counting, cpu)) {
++ per_cpu(hist_preemptirqsoff_counting, cpu) = 1;
++ if (!time_set)
++ start = ftrace_now(cpu);
++ per_cpu(hist_preemptirqsoff_start, cpu) = start;
++ }
+#endif
-
- /*
- * debug_rcu_head_queue()/debug_rcu_head_unqueue() are used internally
-Index: linux-3.0/kernel/stop_machine.c
-===================================================================
---- linux-3.0.orig/kernel/stop_machine.c
-+++ linux-3.0/kernel/stop_machine.c
-@@ -29,12 +29,12 @@ struct cpu_stop_done {
- atomic_t nr_todo; /* nr left to execute */
- bool executed; /* actually executed? */
- int ret; /* collected return value */
-- struct completion completion; /* fired if nr_todo reaches 0 */
-+ struct task_struct *waiter; /* woken when nr_todo reaches 0 */
- };
-
- /* the actual stopper, one per every possible cpu, enabled on online cpus */
- struct cpu_stopper {
-- spinlock_t lock;
-+ raw_spinlock_t lock;
- bool enabled; /* is this stopper enabled? */
- struct list_head works; /* list of pending works */
- struct task_struct *thread; /* stopper thread */
-@@ -46,7 +46,7 @@ static void cpu_stop_init_done(struct cp
- {
- memset(done, 0, sizeof(*done));
- atomic_set(&done->nr_todo, nr_todo);
-- init_completion(&done->completion);
-+ done->waiter = current;
- }
-
- /* signal completion unless @done is NULL */
-@@ -55,8 +55,10 @@ static void cpu_stop_signal_done(struct
- if (done) {
- if (executed)
- done->executed = true;
-- if (atomic_dec_and_test(&done->nr_todo))
-- complete(&done->completion);
-+ if (atomic_dec_and_test(&done->nr_todo)) {
-+ wake_up_process(done->waiter);
-+ done->waiter = NULL;
++ } else {
++ cycle_t uninitialized_var(stop);
++
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ if ((reason == IRQS_ON || reason == TRACE_STOP) &&
++ per_cpu(hist_irqsoff_counting, cpu)) {
++ cycle_t start = per_cpu(hist_irqsoff_start, cpu);
++
++ stop = ftrace_now(cpu);
++ time_set++;
++ if (start && stop >= start) {
++ unsigned long latency =
++ nsecs_to_usecs(stop - start);
++
++ latency_hist(IRQSOFF_LATENCY, cpu, latency, 0,
++ stop, NULL);
++ }
++ per_cpu(hist_irqsoff_counting, cpu) = 0;
+ }
- }
- }
-
-@@ -66,7 +68,7 @@ static void cpu_stop_queue_work(struct c
- {
- unsigned long flags;
-
-- spin_lock_irqsave(&stopper->lock, flags);
-+ raw_spin_lock_irqsave(&stopper->lock, flags);
-
- if (stopper->enabled) {
- list_add_tail(&work->list, &stopper->works);
-@@ -74,7 +76,23 @@ static void cpu_stop_queue_work(struct c
- } else
- cpu_stop_signal_done(work->done, false);
-
-- spin_unlock_irqrestore(&stopper->lock, flags);
-+ raw_spin_unlock_irqrestore(&stopper->lock, flags);
-+}
++#endif
+
-+static void wait_for_stop_done(struct cpu_stop_done *done)
-+{
-+ set_current_state(TASK_UNINTERRUPTIBLE);
-+ while (atomic_read(&done->nr_todo)) {
-+ schedule();
-+ set_current_state(TASK_UNINTERRUPTIBLE);
-+ }
-+ /*
-+ * We need to wait until cpu_stop_signal_done() has cleared
-+ * done->waiter.
-+ */
-+ while (done->waiter)
-+ cpu_relax();
-+ set_current_state(TASK_RUNNING);
- }
-
- /**
-@@ -108,7 +126,7 @@ int stop_one_cpu(unsigned int cpu, cpu_s
-
- cpu_stop_init_done(&done, 1);
- cpu_stop_queue_work(&per_cpu(cpu_stopper, cpu), &work);
-- wait_for_completion(&done.completion);
-+ wait_for_stop_done(&done);
- return done.executed ? done.ret : -ENOENT;
- }
-
-@@ -133,6 +151,7 @@ void stop_one_cpu_nowait(unsigned int cp
- }
-
- DEFINE_MUTEX(stop_cpus_mutex);
-+static DEFINE_MUTEX(stopper_lock);
- /* static data for stop_cpus */
- static DEFINE_PER_CPU(struct cpu_stop_work, stop_cpus_work);
-
-@@ -152,17 +171,16 @@ int __stop_cpus(const struct cpumask *cp
- cpu_stop_init_done(&done, cpumask_weight(cpumask));
-
- /*
-- * Disable preemption while queueing to avoid getting
-- * preempted by a stopper which might wait for other stoppers
-- * to enter @fn which can lead to deadlock.
-+ * Make sure that all work is queued on all cpus before we
-+ * any of the cpus can execute it.
- */
-- preempt_disable();
-+ mutex_lock(&stopper_lock);
- for_each_cpu(cpu, cpumask)
- cpu_stop_queue_work(&per_cpu(cpu_stopper, cpu),
- &per_cpu(stop_cpus_work, cpu));
-- preempt_enable();
-+ mutex_unlock(&stopper_lock);
-
-- wait_for_completion(&done.completion);
-+ wait_for_stop_done(&done);
- return done.executed ? done.ret : -ENOENT;
- }
-
-@@ -250,13 +268,13 @@ repeat:
- }
-
- work = NULL;
-- spin_lock_irq(&stopper->lock);
-+ raw_spin_lock_irq(&stopper->lock);
- if (!list_empty(&stopper->works)) {
- work = list_first_entry(&stopper->works,
- struct cpu_stop_work, list);
- list_del_init(&work->list);
- }
-- spin_unlock_irq(&stopper->lock);
-+ raw_spin_unlock_irq(&stopper->lock);
-
- if (work) {
- cpu_stop_fn_t fn = work->fn;
-@@ -266,6 +284,16 @@ repeat:
-
- __set_current_state(TASK_RUNNING);
-
-+ /*
-+ * Wait until the stopper finished scheduling on all
-+ * cpus
-+ */
-+ mutex_lock(&stopper_lock);
-+ /*
-+ * Let other cpu threads continue as well
-+ */
-+ mutex_unlock(&stopper_lock);
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ if ((reason == PREEMPT_ON || reason == TRACE_STOP) &&
++ per_cpu(hist_preemptoff_counting, cpu)) {
++ cycle_t start = per_cpu(hist_preemptoff_start, cpu);
+
- /* cpu stop callbacks are not allowed to sleep */
- preempt_disable();
-
-@@ -280,7 +308,13 @@ repeat:
- kallsyms_lookup((unsigned long)fn, NULL, NULL, NULL,
- ksym_buf), arg);
-
-+ /*
-+ * Make sure that the wakeup and setting done->waiter
-+ * to NULL is atomic.
-+ */
-+ local_irq_disable();
- cpu_stop_signal_done(done, true);
-+ local_irq_enable();
- } else
- schedule();
-
-@@ -308,6 +342,7 @@ static int __cpuinit cpu_stop_cpu_callba
- if (IS_ERR(p))
- return notifier_from_errno(PTR_ERR(p));
- get_task_struct(p);
-+ p->flags |= PF_STOMPER;
- kthread_bind(p, cpu);
- sched_set_stop_task(cpu, p);
- stopper->thread = p;
-@@ -317,9 +352,9 @@ static int __cpuinit cpu_stop_cpu_callba
- /* strictly unnecessary, as first user will wake it */
- wake_up_process(stopper->thread);
- /* mark enabled */
-- spin_lock_irq(&stopper->lock);
-+ raw_spin_lock_irq(&stopper->lock);
- stopper->enabled = true;
-- spin_unlock_irq(&stopper->lock);
-+ raw_spin_unlock_irq(&stopper->lock);
- break;
-
- #ifdef CONFIG_HOTPLUG_CPU
-@@ -332,11 +367,11 @@ static int __cpuinit cpu_stop_cpu_callba
- /* kill the stopper */
- kthread_stop(stopper->thread);
- /* drain remaining works */
-- spin_lock_irq(&stopper->lock);
-+ raw_spin_lock_irq(&stopper->lock);
- list_for_each_entry(work, &stopper->works, list)
- cpu_stop_signal_done(work->done, false);
- stopper->enabled = false;
-- spin_unlock_irq(&stopper->lock);
-+ raw_spin_unlock_irq(&stopper->lock);
- /* release the stopper */
- put_task_struct(stopper->thread);
- stopper->thread = NULL;
-@@ -367,7 +402,7 @@ static int __init cpu_stop_init(void)
- for_each_possible_cpu(cpu) {
- struct cpu_stopper *stopper = &per_cpu(cpu_stopper, cpu);
-
-- spin_lock_init(&stopper->lock);
-+ raw_spin_lock_init(&stopper->lock);
- INIT_LIST_HEAD(&stopper->works);
- }
-
-Index: linux-3.0/include/linux/cpu.h
-===================================================================
---- linux-3.0.orig/include/linux/cpu.h
-+++ linux-3.0/include/linux/cpu.h
-@@ -60,14 +60,16 @@ enum {
- */
- CPU_PRI_SCHED_ACTIVE = INT_MAX,
- CPU_PRI_CPUSET_ACTIVE = INT_MAX - 1,
-- CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
-- CPU_PRI_CPUSET_INACTIVE = INT_MIN,
-
- /* migration should happen before other stuff but after perf */
-- CPU_PRI_PERF = 20,
-- CPU_PRI_MIGRATION = 10,
-- /* prepare workqueues for other notifiers */
-- CPU_PRI_WORKQUEUE = 5,
-+ CPU_PRI_PERF = 20,
-+ CPU_PRI_MIGRATION = 10,
-+ CPU_PRI_WORKQUEUE_ACTIVE = 5, /* prepare workqueues for others */
-+ CPU_PRI_NORMAL = 0,
-+ CPU_PRI_WORKQUEUE_INACTIVE = -5, /* flush workqueues after others */
++ if (!(time_set++))
++ stop = ftrace_now(cpu);
++ if (start && stop >= start) {
++ unsigned long latency =
++ nsecs_to_usecs(stop - start);
+
-+ CPU_PRI_SCHED_INACTIVE = INT_MIN + 1,
-+ CPU_PRI_CPUSET_INACTIVE = INT_MIN,
- };
-
- #ifdef CONFIG_SMP
-@@ -134,6 +136,8 @@ extern struct sysdev_class cpu_sysdev_cl
-
- extern void get_online_cpus(void);
- extern void put_online_cpus(void);
-+extern void pin_current_cpu(void);
-+extern void unpin_current_cpu(void);
- #define hotcpu_notifier(fn, pri) cpu_notifier(fn, pri)
- #define register_hotcpu_notifier(nb) register_cpu_notifier(nb)
- #define unregister_hotcpu_notifier(nb) unregister_cpu_notifier(nb)
-@@ -156,6 +160,8 @@ static inline void cpu_hotplug_driver_un
-
- #define get_online_cpus() do { } while (0)
- #define put_online_cpus() do { } while (0)
-+static inline void pin_current_cpu(void) { }
-+static inline void unpin_current_cpu(void) { }
- #define hotcpu_notifier(fn, pri) do { (void)(fn); } while (0)
- /* These aren't inline functions due to a GCC bug. */
- #define register_hotcpu_notifier(nb) ({ (void)(nb); 0; })
-Index: linux-3.0/kernel/cpu.c
-===================================================================
---- linux-3.0.orig/kernel/cpu.c
-+++ linux-3.0/kernel/cpu.c
-@@ -57,6 +57,104 @@ static struct {
- .refcount = 0,
- };
-
-+struct hotplug_pcp {
-+ struct task_struct *unplug;
-+ int refcount;
-+ struct completion synced;
-+};
++ latency_hist(PREEMPTOFF_LATENCY, cpu, latency,
++ 0, stop, NULL);
++ }
++ per_cpu(hist_preemptoff_counting, cpu) = 0;
++ }
++#endif
++
++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
++ if ((!per_cpu(hist_irqsoff_counting, cpu) ||
++ !per_cpu(hist_preemptoff_counting, cpu)) &&
++ per_cpu(hist_preemptirqsoff_counting, cpu)) {
++ cycle_t start = per_cpu(hist_preemptirqsoff_start, cpu);
++
++ if (!time_set)
++ stop = ftrace_now(cpu);
++ if (start && stop >= start) {
++ unsigned long latency =
++ nsecs_to_usecs(stop - start);
++ latency_hist(PREEMPTIRQSOFF_LATENCY, cpu,
++ latency, 0, stop, NULL);
++ }
++ per_cpu(hist_preemptirqsoff_counting, cpu) = 0;
++ }
++#endif
++ }
++}
++#endif
++
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++static DEFINE_RAW_SPINLOCK(wakeup_lock);
++static notrace void probe_sched_migrate_task(void *v, struct task_struct *task,
++ int cpu)
++{
++ int old_cpu = task_cpu(task);
+
-+static DEFINE_PER_CPU(struct hotplug_pcp, hotplug_pcp);
++ if (cpu != old_cpu) {
++ unsigned long flags;
++ struct task_struct *cpu_wakeup_task;
+
-+/**
-+ * pin_current_cpu - Prevent the current cpu from being unplugged
-+ *
-+ * Lightweight version of get_online_cpus() to prevent cpu from being
-+ * unplugged when code runs in a migration disabled region.
-+ *
-+ * Must be called with preemption disabled (preempt_count = 1)!
-+ */
-+void pin_current_cpu(void)
++ raw_spin_lock_irqsave(&wakeup_lock, flags);
++
++ cpu_wakeup_task = per_cpu(wakeup_task, old_cpu);
++ if (task == cpu_wakeup_task) {
++ put_task_struct(cpu_wakeup_task);
++ per_cpu(wakeup_task, old_cpu) = NULL;
++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = task;
++ get_task_struct(cpu_wakeup_task);
++ }
++
++ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
++ }
++}
++
++static notrace void probe_wakeup_latency_hist_start(void *v,
++ struct task_struct *p, int success)
+{
-+ struct hotplug_pcp *hp;
++ unsigned long flags;
++ struct task_struct *curr = current;
++ int cpu = task_cpu(p);
++ struct task_struct *cpu_wakeup_task;
+
-+retry:
-+ hp = &__get_cpu_var(hotplug_pcp);
++ raw_spin_lock_irqsave(&wakeup_lock, flags);
+
-+ if (!hp->unplug || hp->refcount || preempt_count() > 1 ||
-+ hp->unplug == current || (current->flags & PF_STOMPER)) {
-+ hp->refcount++;
-+ return;
++ cpu_wakeup_task = per_cpu(wakeup_task, cpu);
++
++ if (wakeup_pid) {
++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) ||
++ p->prio == curr->prio)
++ per_cpu(wakeup_sharedprio, cpu) = 1;
++ if (likely(wakeup_pid != task_pid_nr(p)))
++ goto out;
++ } else {
++ if (likely(!rt_task(p)) ||
++ (cpu_wakeup_task && p->prio > cpu_wakeup_task->prio) ||
++ p->prio > curr->prio)
++ goto out;
++ if ((cpu_wakeup_task && p->prio == cpu_wakeup_task->prio) ||
++ p->prio == curr->prio)
++ per_cpu(wakeup_sharedprio, cpu) = 1;
+ }
-+ preempt_enable();
-+ mutex_lock(&cpu_hotplug.lock);
-+ mutex_unlock(&cpu_hotplug.lock);
-+ preempt_disable();
-+ goto retry;
++
++ if (cpu_wakeup_task)
++ put_task_struct(cpu_wakeup_task);
++ cpu_wakeup_task = per_cpu(wakeup_task, cpu) = p;
++ get_task_struct(cpu_wakeup_task);
++ cpu_wakeup_task->preempt_timestamp_hist =
++ ftrace_now(raw_smp_processor_id());
++out:
++ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
+}
+
-+/**
-+ * unpin_current_cpu - Allow unplug of current cpu
-+ *
-+ * Must be called with preemption or interrupts disabled!
-+ */
-+void unpin_current_cpu(void)
++static notrace void probe_wakeup_latency_hist_stop(void *v,
++ struct task_struct *prev, struct task_struct *next)
+{
-+ struct hotplug_pcp *hp = &__get_cpu_var(hotplug_pcp);
++ unsigned long flags;
++ int cpu = task_cpu(next);
++ unsigned long latency;
++ cycle_t stop;
++ struct task_struct *cpu_wakeup_task;
+
-+ WARN_ON(hp->refcount <= 0);
++ raw_spin_lock_irqsave(&wakeup_lock, flags);
+
-+ /* This is safe. sync_unplug_thread is pinned to this cpu */
-+ if (!--hp->refcount && hp->unplug && hp->unplug != current &&
-+ !(current->flags & PF_STOMPER))
-+ wake_up_process(hp->unplug);
++ cpu_wakeup_task = per_cpu(wakeup_task, cpu);
++
++ if (cpu_wakeup_task == NULL)
++ goto out;
++
++ /* Already running? */
++ if (unlikely(current == cpu_wakeup_task))
++ goto out_reset;
++
++ if (next != cpu_wakeup_task) {
++ if (next->prio < cpu_wakeup_task->prio)
++ goto out_reset;
++
++ if (next->prio == cpu_wakeup_task->prio)
++ per_cpu(wakeup_sharedprio, cpu) = 1;
++
++ goto out;
++ }
++
++ /*
++ * The task we are waiting for is about to be switched to.
++ * Calculate latency and store it in histogram.
++ */
++ stop = ftrace_now(raw_smp_processor_id());
++
++ latency = nsecs_to_usecs(stop - next->preempt_timestamp_hist);
++
++ if (per_cpu(wakeup_sharedprio, cpu)) {
++ latency_hist(WAKEUP_LATENCY_SHAREDPRIO, cpu, latency, 0, stop,
++ next);
++ per_cpu(wakeup_sharedprio, cpu) = 0;
++ } else {
++ latency_hist(WAKEUP_LATENCY, cpu, latency, 0, stop, next);
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ if (timerandwakeup_enabled_data.enabled) {
++ latency_hist(TIMERANDWAKEUP_LATENCY, cpu,
++ next->timer_offset + latency, next->timer_offset,
++ stop, next);
++ }
++#endif
++ }
++
++out_reset:
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ next->timer_offset = 0;
++#endif
++ put_task_struct(cpu_wakeup_task);
++ per_cpu(wakeup_task, cpu) = NULL;
++out:
++ raw_spin_unlock_irqrestore(&wakeup_lock, flags);
+}
++#endif
+
-+/*
-+ * FIXME: Is this really correct under all circumstances ?
-+ */
-+static int sync_unplug_thread(void *data)
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++static notrace void probe_hrtimer_interrupt(void *v, int cpu,
++ long long latency_ns, struct task_struct *curr, struct task_struct *task)
+{
-+ struct hotplug_pcp *hp = data;
++ if (latency_ns <= 0 && task != NULL && rt_task(task) &&
++ (task->prio < curr->prio ||
++ (task->prio == curr->prio &&
++ !cpumask_test_cpu(cpu, &task->cpus_allowed)))) {
++ unsigned long latency;
++ cycle_t now;
+
-+ preempt_disable();
-+ hp->unplug = current;
-+ set_current_state(TASK_UNINTERRUPTIBLE);
-+ while (hp->refcount) {
-+ schedule_preempt_disabled();
-+ set_current_state(TASK_UNINTERRUPTIBLE);
++ if (missed_timer_offsets_pid) {
++ if (likely(missed_timer_offsets_pid !=
++ task_pid_nr(task)))
++ return;
++ }
++
++ now = ftrace_now(cpu);
++ latency = (unsigned long) div_s64(-latency_ns, 1000);
++ latency_hist(MISSED_TIMER_OFFSETS, cpu, latency, latency, now,
++ task);
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ task->timer_offset = latency;
++#endif
+ }
-+ set_current_state(TASK_RUNNING);
-+ preempt_enable();
-+ complete(&hp->synced);
-+ return 0;
+}
++#endif
++
++static __init int latency_hist_init(void)
++{
++ struct dentry *latency_hist_root = NULL;
++ struct dentry *dentry;
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ struct dentry *dentry_sharedprio;
++#endif
++ struct dentry *entry;
++ struct dentry *enable_root;
++ int i = 0;
++ struct hist_data *my_hist;
++ char name[64];
++ char *cpufmt = "CPU%d";
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) || \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ char *cpufmt_maxlatproc = "max_latency-CPU%d";
++ struct maxlatproc_data *mp = NULL;
++#endif
++
++ dentry = tracing_init_dentry();
++ latency_hist_root = debugfs_create_dir(latency_hist_dir_root, dentry);
++ enable_root = debugfs_create_dir("enable", latency_hist_root);
++
++#ifdef CONFIG_INTERRUPT_OFF_HIST
++ dentry = debugfs_create_dir(irqsoff_hist_dir, latency_hist_root);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(irqsoff_hist, i), &latency_hist_fops);
++ my_hist = &per_cpu(irqsoff_hist, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
++ }
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)IRQSOFF_LATENCY, &latency_hist_reset_fops);
++#endif
++
++#ifdef CONFIG_PREEMPT_OFF_HIST
++ dentry = debugfs_create_dir(preemptoff_hist_dir,
++ latency_hist_root);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(preemptoff_hist, i), &latency_hist_fops);
++ my_hist = &per_cpu(preemptoff_hist, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
++ }
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)PREEMPTOFF_LATENCY, &latency_hist_reset_fops);
++#endif
++
++#if defined(CONFIG_INTERRUPT_OFF_HIST) && defined(CONFIG_PREEMPT_OFF_HIST)
++ dentry = debugfs_create_dir(preemptirqsoff_hist_dir,
++ latency_hist_root);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(preemptirqsoff_hist, i), &latency_hist_fops);
++ my_hist = &per_cpu(preemptirqsoff_hist, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
++ }
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)PREEMPTIRQSOFF_LATENCY, &latency_hist_reset_fops);
++#endif
++
++#if defined(CONFIG_INTERRUPT_OFF_HIST) || defined(CONFIG_PREEMPT_OFF_HIST)
++ entry = debugfs_create_file("preemptirqsoff", 0644,
++ enable_root, (void *)&preemptirqsoff_enabled_data,
++ &enable_fops);
++#endif
+
-+/*
-+ * Start the sync_unplug_thread on the target cpu and wait for it to
-+ * complete.
-+ */
-+static int cpu_unplug_begin(unsigned int cpu)
-+{
-+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu);
-+ struct task_struct *tsk;
++#ifdef CONFIG_WAKEUP_LATENCY_HIST
++ dentry = debugfs_create_dir(wakeup_latency_hist_dir,
++ latency_hist_root);
++ dentry_sharedprio = debugfs_create_dir(
++ wakeup_latency_hist_dir_sharedprio, dentry);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
+
-+ init_completion(&hp->synced);
-+ tsk = kthread_create(sync_unplug_thread, hp, "sync_unplug/%d", cpu);
-+ if (IS_ERR(tsk))
-+ return (PTR_ERR(tsk));
-+ kthread_bind(tsk, cpu);
-+ wake_up_process(tsk);
-+ wait_for_completion(&hp->synced);
-+ return 0;
-+}
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(wakeup_latency_hist, i),
++ &latency_hist_fops);
++ my_hist = &per_cpu(wakeup_latency_hist, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
+
-+static void cpu_unplug_done(unsigned int cpu)
-+{
-+ struct hotplug_pcp *hp = &per_cpu(hotplug_pcp, cpu);
++ entry = debugfs_create_file(name, 0444, dentry_sharedprio,
++ &per_cpu(wakeup_latency_hist_sharedprio, i),
++ &latency_hist_fops);
++ my_hist = &per_cpu(wakeup_latency_hist_sharedprio, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
+
-+ hp->unplug = NULL;
-+}
++ sprintf(name, cpufmt_maxlatproc, i);
+
- void get_online_cpus(void)
- {
- might_sleep();
-@@ -210,13 +308,14 @@ static int __ref take_cpu_down(void *_pa
- /* Requires cpu_add_remove_lock to be held */
- static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
- {
-- int err, nr_calls = 0;
-+ int mycpu, err, nr_calls = 0;
- void *hcpu = (void *)(long)cpu;
- unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0;
- struct take_cpu_down_param tcd_param = {
- .mod = mod,
- .hcpu = hcpu,
- };
-+ cpumask_var_t cpumask;
-
- if (num_online_cpus() == 1)
- return -EBUSY;
-@@ -224,7 +323,26 @@ static int __ref _cpu_down(unsigned int
- if (!cpu_online(cpu))
- return -EINVAL;
-
-+ /* Move the downtaker off the unplug cpu */
-+ if (!alloc_cpumask_var(&cpumask, GFP_KERNEL))
-+ return -ENOMEM;
-+ cpumask_andnot(cpumask, cpu_online_mask, cpumask_of(cpu));
-+ set_cpus_allowed_ptr(current, cpumask);
-+ free_cpumask_var(cpumask);
-+ migrate_disable();
-+ mycpu = smp_processor_id();
-+ if (mycpu == cpu) {
-+ printk(KERN_ERR "Yuck! Still on unplug CPU\n!");
-+ migrate_enable();
-+ return -EBUSY;
-+ }
++ mp = &per_cpu(wakeup_maxlatproc, i);
++ entry = debugfs_create_file(name, 0444, dentry, mp,
++ &maxlatproc_fops);
++ clear_maxlatprocdata(mp);
+
- cpu_hotplug_begin();
-+ err = cpu_unplug_begin(cpu);
-+ if (err) {
-+ printk("cpu_unplug_begin(%d) failed\n", cpu);
-+ goto out_cancel;
++ mp = &per_cpu(wakeup_maxlatproc_sharedprio, i);
++ entry = debugfs_create_file(name, 0444, dentry_sharedprio, mp,
++ &maxlatproc_fops);
++ clear_maxlatprocdata(mp);
+ }
-
- err = __cpu_notify(CPU_DOWN_PREPARE | mod, hcpu, -1, &nr_calls);
- if (err) {
-@@ -263,6 +381,9 @@ static int __ref _cpu_down(unsigned int
- check_for_tasks(cpu);
-
- out_release:
-+ cpu_unplug_done(cpu);
-+out_cancel:
-+ migrate_enable();
- cpu_hotplug_done();
- if (!err)
- cpu_notify_nofail(CPU_POST_DEAD | mod, hcpu);
-Index: linux-3.0/include/linux/ftrace_event.h
-===================================================================
---- linux-3.0.orig/include/linux/ftrace_event.h
-+++ linux-3.0/include/linux/ftrace_event.h
-@@ -49,7 +49,8 @@ struct trace_entry {
- unsigned char flags;
- unsigned char preempt_count;
- int pid;
-- int padding;
-+ unsigned short migrate_disable;
-+ unsigned short padding;
- };
-
- #define FTRACE_MAX_EVENT \
-Index: linux-3.0/kernel/trace/trace_events.c
-===================================================================
---- linux-3.0.orig/kernel/trace/trace_events.c
-+++ linux-3.0/kernel/trace/trace_events.c
-@@ -116,6 +116,7 @@ static int trace_define_common_fields(vo
- __common_field(unsigned char, flags);
- __common_field(unsigned char, preempt_count);
- __common_field(int, pid);
-+ __common_field(unsigned short, migrate_disable);
- __common_field(int, padding);
-
- return ret;
-Index: linux-3.0/kernel/trace/trace_output.c
-===================================================================
---- linux-3.0.orig/kernel/trace/trace_output.c
-+++ linux-3.0/kernel/trace/trace_output.c
-@@ -591,6 +591,11 @@ int trace_print_lat_fmt(struct trace_seq
- else
- ret = trace_seq_putc(s, '.');
-
-+ if (entry->migrate_disable)
-+ ret = trace_seq_printf(s, "%x", entry->migrate_disable);
-+ else
-+ ret = trace_seq_putc(s, '.');
-+
- return ret;
- }
-
-Index: linux-3.0/kernel/sched_debug.c
-===================================================================
---- linux-3.0.orig/kernel/sched_debug.c
-+++ linux-3.0/kernel/sched_debug.c
-@@ -235,6 +235,9 @@ void print_rt_rq(struct seq_file *m, int
- P(rt_throttled);
- PN(rt_time);
- PN(rt_runtime);
-+#ifdef CONFIG_SMP
-+ P(rt_nr_migratory);
-+#endif
-
- #undef PN
- #undef P
-@@ -484,6 +487,10 @@ void proc_sched_show_task(struct task_st
- P(se.load.weight);
- P(policy);
- P(prio);
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ P(migrate_disable);
-+#endif
-+ P(rt.nr_cpus_allowed);
- #undef PN
- #undef __PN
- #undef P
-Index: linux-3.0/kernel/trace/trace.h
-===================================================================
---- linux-3.0.orig/kernel/trace/trace.h
-+++ linux-3.0/kernel/trace/trace.h
-@@ -322,7 +322,6 @@ void trace_init_global_iter(struct trace
-
- void tracing_iter_reset(struct trace_iterator *iter, int cpu);
-
--void default_wait_pipe(struct trace_iterator *iter);
- void poll_wait_pipe(struct trace_iterator *iter);
-
- void ftrace(struct trace_array *tr,
-Index: linux-3.0/kernel/Kconfig.locks
-===================================================================
---- linux-3.0.orig/kernel/Kconfig.locks
-+++ linux-3.0/kernel/Kconfig.locks
-@@ -199,4 +199,4 @@ config INLINE_WRITE_UNLOCK_IRQRESTORE
- def_bool !DEBUG_SPINLOCK && ARCH_INLINE_WRITE_UNLOCK_IRQRESTORE
-
- config MUTEX_SPIN_ON_OWNER
-- def_bool SMP && !DEBUG_MUTEXES
-+ def_bool SMP && !DEBUG_MUTEXES && !PREEMPT_RT_FULL
-Index: linux-3.0/include/linux/hardirq.h
-===================================================================
---- linux-3.0.orig/include/linux/hardirq.h
-+++ linux-3.0/include/linux/hardirq.h
-@@ -60,7 +60,11 @@
- #define HARDIRQ_OFFSET (1UL << HARDIRQ_SHIFT)
- #define NMI_OFFSET (1UL << NMI_SHIFT)
-
--#define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET)
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+# define SOFTIRQ_DISABLE_OFFSET (2 * SOFTIRQ_OFFSET)
-+#else
-+# define SOFTIRQ_DISABLE_OFFSET (0)
-+#endif
-
- #ifndef PREEMPT_ACTIVE
- #define PREEMPT_ACTIVE_BITS 1
-@@ -73,10 +77,17 @@
- #endif
-
- #define hardirq_count() (preempt_count() & HARDIRQ_MASK)
--#define softirq_count() (preempt_count() & SOFTIRQ_MASK)
- #define irq_count() (preempt_count() & (HARDIRQ_MASK | SOFTIRQ_MASK \
- | NMI_MASK))
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+# define softirq_count() (preempt_count() & SOFTIRQ_MASK)
-+# define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET)
-+#else
-+# define softirq_count() (0UL)
-+extern int in_serving_softirq(void);
++ entry = debugfs_create_file("pid", 0644, dentry,
++ (void *)&wakeup_pid, &pid_fops);
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)WAKEUP_LATENCY, &latency_hist_reset_fops);
++ entry = debugfs_create_file("reset", 0644, dentry_sharedprio,
++ (void *)WAKEUP_LATENCY_SHAREDPRIO, &latency_hist_reset_fops);
++ entry = debugfs_create_file("wakeup", 0644,
++ enable_root, (void *)&wakeup_latency_enabled_data,
++ &enable_fops);
+#endif
+
- /*
- * Are we doing bottom half or hardware interrupt processing?
- * Are we in a softirq context? Interrupt context?
-@@ -86,7 +97,6 @@
- #define in_irq() (hardirq_count())
- #define in_softirq() (softirq_count())
- #define in_interrupt() (irq_count())
--#define in_serving_softirq() (softirq_count() & SOFTIRQ_OFFSET)
-
- /*
- * Are we in NMI context?
-Index: linux-3.0/arch/powerpc/kernel/irq.c
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/irq.c
-+++ linux-3.0/arch/powerpc/kernel/irq.c
-@@ -449,6 +449,7 @@ void irq_ctx_init(void)
- }
- }
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- static inline void do_softirq_onstack(void)
- {
- struct thread_info *curtp, *irqtp;
-@@ -478,7 +479,7 @@ void do_softirq(void)
-
- local_irq_restore(flags);
- }
--
-+#endif
-
- /*
- * IRQ controller and virtual interrupts
-Index: linux-3.0/arch/powerpc/kernel/misc_32.S
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/misc_32.S
-+++ linux-3.0/arch/powerpc/kernel/misc_32.S
-@@ -34,6 +34,7 @@
-
- .text
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- _GLOBAL(call_do_softirq)
- mflr r0
- stw r0,4(r1)
-@@ -44,6 +45,7 @@ _GLOBAL(call_do_softirq)
- lwz r0,4(r1)
- mtlr r0
- blr
++#ifdef CONFIG_MISSED_TIMER_OFFSETS_HIST
++ dentry = debugfs_create_dir(missed_timer_offsets_dir,
++ latency_hist_root);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(missed_timer_offsets, i), &latency_hist_fops);
++ my_hist = &per_cpu(missed_timer_offsets, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
++
++ sprintf(name, cpufmt_maxlatproc, i);
++ mp = &per_cpu(missed_timer_offsets_maxlatproc, i);
++ entry = debugfs_create_file(name, 0444, dentry, mp,
++ &maxlatproc_fops);
++ clear_maxlatprocdata(mp);
++ }
++ entry = debugfs_create_file("pid", 0644, dentry,
++ (void *)&missed_timer_offsets_pid, &pid_fops);
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)MISSED_TIMER_OFFSETS, &latency_hist_reset_fops);
++ entry = debugfs_create_file("missed_timer_offsets", 0644,
++ enable_root, (void *)&missed_timer_offsets_enabled_data,
++ &enable_fops);
+#endif
-
- _GLOBAL(call_handle_irq)
- mflr r0
-Index: linux-3.0/arch/powerpc/kernel/misc_64.S
-===================================================================
---- linux-3.0.orig/arch/powerpc/kernel/misc_64.S
-+++ linux-3.0/arch/powerpc/kernel/misc_64.S
-@@ -29,6 +29,7 @@
-
- .text
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- _GLOBAL(call_do_softirq)
- mflr r0
- std r0,16(r1)
-@@ -39,6 +40,7 @@ _GLOBAL(call_do_softirq)
- ld r0,16(r1)
- mtlr r0
- blr
++
++#if defined(CONFIG_WAKEUP_LATENCY_HIST) && \
++ defined(CONFIG_MISSED_TIMER_OFFSETS_HIST)
++ dentry = debugfs_create_dir(timerandwakeup_latency_hist_dir,
++ latency_hist_root);
++ for_each_possible_cpu(i) {
++ sprintf(name, cpufmt, i);
++ entry = debugfs_create_file(name, 0444, dentry,
++ &per_cpu(timerandwakeup_latency_hist, i),
++ &latency_hist_fops);
++ my_hist = &per_cpu(timerandwakeup_latency_hist, i);
++ atomic_set(&my_hist->hist_mode, 1);
++ my_hist->min_lat = 0xFFFFFFFFUL;
++
++ sprintf(name, cpufmt_maxlatproc, i);
++ mp = &per_cpu(timerandwakeup_maxlatproc, i);
++ entry = debugfs_create_file(name, 0444, dentry, mp,
++ &maxlatproc_fops);
++ clear_maxlatprocdata(mp);
++ }
++ entry = debugfs_create_file("reset", 0644, dentry,
++ (void *)TIMERANDWAKEUP_LATENCY, &latency_hist_reset_fops);
++ entry = debugfs_create_file("timerandwakeup", 0644,
++ enable_root, (void *)&timerandwakeup_enabled_data,
++ &enable_fops);
+#endif
-
- _GLOBAL(call_handle_irq)
- ld r8,0(r6)
-Index: linux-3.0/arch/sh/kernel/irq.c
-===================================================================
---- linux-3.0.orig/arch/sh/kernel/irq.c
-+++ linux-3.0/arch/sh/kernel/irq.c
-@@ -149,6 +149,7 @@ void irq_ctx_exit(int cpu)
- hardirq_ctx[cpu] = NULL;
++ return 0;
++}
++
++__initcall(latency_hist_init);
+diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
+index b0c7aa4..2467714 100644
+--- a/kernel/trace/ring_buffer.c
++++ b/kernel/trace/ring_buffer.c
+@@ -1040,6 +1040,44 @@ static int rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer,
+ return -ENOMEM;
}
-+#ifndef CONFIG_PREEMPT_RT_FULL
- asmlinkage void do_softirq(void)
- {
- unsigned long flags;
-@@ -191,6 +192,7 @@ asmlinkage void do_softirq(void)
-
- local_irq_restore(flags);
- }
++static inline int ok_to_lock(void)
++{
++ if (in_nmi())
++ return 0;
++#ifdef CONFIG_PREEMPT_RT_FULL
++ if (in_atomic())
++ return 0;
+#endif
- #else
- static inline void handle_one_irq(unsigned int irq)
++ return 1;
++}
++
++static int
++read_buffer_lock(struct ring_buffer_per_cpu *cpu_buffer,
++ unsigned long *flags)
++{
++ /*
++ * If an NMI die dumps out the content of the ring buffer
++ * do not grab locks. We also permanently disable the ring
++ * buffer too. A one time deal is all you get from reading
++ * the ring buffer from an NMI.
++ */
++ if (!ok_to_lock()) {
++ if (spin_trylock_irqsave(&cpu_buffer->reader_lock, *flags))
++ return 1;
++ tracing_off_permanent();
++ return 0;
++ }
++ spin_lock_irqsave(&cpu_buffer->reader_lock, *flags);
++ return 1;
++}
++
++static void
++read_buffer_unlock(struct ring_buffer_per_cpu *cpu_buffer,
++ unsigned long flags, int locked)
++{
++ if (locked)
++ spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++}
+ static struct ring_buffer_per_cpu *
+ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu)
{
-Index: linux-3.0/arch/sparc/kernel/irq_64.c
-===================================================================
---- linux-3.0.orig/arch/sparc/kernel/irq_64.c
-+++ linux-3.0/arch/sparc/kernel/irq_64.c
-@@ -700,6 +700,7 @@ void __irq_entry handler_irq(int pil, st
- set_irq_regs(old_regs);
- }
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- void do_softirq(void)
+@@ -1250,9 +1288,11 @@ rb_remove_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned nr_pages)
{
- unsigned long flags;
-@@ -725,6 +726,7 @@ void do_softirq(void)
-
- local_irq_restore(flags);
- }
-+#endif
-
- #ifdef CONFIG_HOTPLUG_CPU
- void fixup_irqs(void)
-Index: linux-3.0/arch/x86/kernel/entry_64.S
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/entry_64.S
-+++ linux-3.0/arch/x86/kernel/entry_64.S
-@@ -1197,6 +1197,7 @@ ENTRY(kernel_execve)
- CFI_ENDPROC
- END(kernel_execve)
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /* Call softirq on interrupt stack. Interrupts are off. */
- ENTRY(call_softirq)
- CFI_STARTPROC
-@@ -1216,6 +1217,7 @@ ENTRY(call_softirq)
- ret
- CFI_ENDPROC
- END(call_softirq)
-+#endif
+ struct buffer_page *bpage;
+ struct list_head *p;
++ unsigned long flags;
+ unsigned i;
++ int locked;
- #ifdef CONFIG_XEN
- zeroentry xen_hypervisor_callback xen_do_hypervisor_callback
-Index: linux-3.0/arch/x86/kernel/irq_32.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/irq_32.c
-+++ linux-3.0/arch/x86/kernel/irq_32.c
-@@ -149,6 +149,7 @@ void __cpuinit irq_ctx_init(int cpu)
- cpu, per_cpu(hardirq_ctx, cpu), per_cpu(softirq_ctx, cpu));
- }
+- spin_lock_irq(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ rb_head_page_deactivate(cpu_buffer);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- asmlinkage void do_softirq(void)
- {
- unsigned long flags;
-@@ -179,6 +180,7 @@ asmlinkage void do_softirq(void)
+ for (i = 0; i < nr_pages; i++) {
+@@ -1270,7 +1310,7 @@ rb_remove_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned nr_pages)
+ rb_check_pages(cpu_buffer);
- local_irq_restore(flags);
+ out:
+- spin_unlock_irq(&cpu_buffer->reader_lock);
++ read_buffer_unlock(cpu_buffer, flags, locked);
}
-+#endif
- bool handle_irq(unsigned irq, struct pt_regs *regs)
+ static void
+@@ -1279,9 +1319,11 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer,
{
-Index: linux-3.0/arch/x86/kernel/irq_64.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/irq_64.c
-+++ linux-3.0/arch/x86/kernel/irq_64.c
-@@ -62,7 +62,7 @@ bool handle_irq(unsigned irq, struct pt_
- return true;
- }
-
--
-+#ifndef CONFIG_PREEMPT_RT_FULL
- extern void call_softirq(void);
-
- asmlinkage void do_softirq(void)
-@@ -82,3 +82,4 @@ asmlinkage void do_softirq(void)
- }
- local_irq_restore(flags);
- }
-+#endif
-Index: linux-3.0/drivers/md/raid5.c
-===================================================================
---- linux-3.0.orig/drivers/md/raid5.c
-+++ linux-3.0/drivers/md/raid5.c
-@@ -1226,8 +1226,9 @@ static void __raid_run_ops(struct stripe
- struct raid5_percpu *percpu;
- unsigned long cpu;
-
-- cpu = get_cpu();
-+ cpu = get_cpu_light();
- percpu = per_cpu_ptr(conf->percpu, cpu);
-+ spin_lock(&percpu->lock);
- if (test_bit(STRIPE_OP_BIOFILL, &ops_request)) {
- ops_run_biofill(sh);
- overlap_clear++;
-@@ -1279,7 +1280,8 @@ static void __raid_run_ops(struct stripe
- if (test_and_clear_bit(R5_Overlap, &dev->flags))
- wake_up(&sh->raid_conf->wait_for_overlap);
- }
-- put_cpu();
-+ spin_unlock(&percpu->lock);
-+ put_cpu_light();
- }
-
- #ifdef CONFIG_MULTICORE_RAID456
-@@ -4756,6 +4758,7 @@ static int raid5_alloc_percpu(raid5_conf
- break;
- }
- per_cpu_ptr(conf->percpu, cpu)->scribble = scribble;
-+ spin_lock_init(&per_cpu_ptr(conf->percpu, cpu)->lock);
- }
- #ifdef CONFIG_HOTPLUG_CPU
- conf->cpu_notify.notifier_call = raid456_cpu_notify;
-Index: linux-3.0/drivers/md/raid5.h
-===================================================================
---- linux-3.0.orig/drivers/md/raid5.h
-+++ linux-3.0/drivers/md/raid5.h
-@@ -402,6 +402,7 @@ struct raid5_private_data {
-
- /* per cpu variables */
- struct raid5_percpu {
-+ spinlock_t lock; /* Protection for -RT */
- struct page *spare_page; /* Used when checking P/Q in raid6 */
- void *scribble; /* space for constructing buffer
- * lists and performing address
-Index: linux-3.0/include/linux/rtmutex.h
-===================================================================
---- linux-3.0.orig/include/linux/rtmutex.h
-+++ linux-3.0/include/linux/rtmutex.h
-@@ -14,7 +14,7 @@
-
- #include <linux/linkage.h>
- #include <linux/plist.h>
--#include <linux/spinlock_types.h>
-+#include <linux/spinlock_types_raw.h>
-
- extern int max_lock_depth; /* for sysctl */
-
-@@ -29,9 +29,10 @@ struct rt_mutex {
- raw_spinlock_t wait_lock;
- struct plist_head wait_list;
- struct task_struct *owner;
--#ifdef CONFIG_DEBUG_RT_MUTEXES
- int save_state;
-- const char *name, *file;
-+#ifdef CONFIG_DEBUG_RT_MUTEXES
-+ const char *file;
-+ const char *name;
- int line;
- void *magic;
- #endif
-@@ -56,19 +57,39 @@ struct hrtimer_sleeper;
- #ifdef CONFIG_DEBUG_RT_MUTEXES
- # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname) \
- , .name = #mutexname, .file = __FILE__, .line = __LINE__
--# define rt_mutex_init(mutex) __rt_mutex_init(mutex, __func__)
-+
-+# define rt_mutex_init(mutex) \
-+ do { \
-+ raw_spin_lock_init(&(mutex)->wait_lock); \
-+ __rt_mutex_init(mutex, #mutex); \
-+ } while (0)
-+
- extern void rt_mutex_debug_task_free(struct task_struct *tsk);
- #else
- # define __DEBUG_RT_MUTEX_INITIALIZER(mutexname)
--# define rt_mutex_init(mutex) __rt_mutex_init(mutex, NULL)
-+
-+# define rt_mutex_init(mutex) \
-+ do { \
-+ raw_spin_lock_init(&(mutex)->wait_lock); \
-+ __rt_mutex_init(mutex, #mutex); \
-+ } while (0)
-+
- # define rt_mutex_debug_task_free(t) do { } while (0)
- #endif
+ struct buffer_page *bpage;
+ struct list_head *p;
++ unsigned long flags;
+ unsigned i;
++ int locked;
--#define __RT_MUTEX_INITIALIZER(mutexname) \
-- { .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
-+#define __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
-+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(mutexname.wait_lock) \
- , .wait_list = PLIST_HEAD_INIT_RAW(mutexname.wait_list, mutexname.wait_lock) \
- , .owner = NULL \
-- __DEBUG_RT_MUTEX_INITIALIZER(mutexname)}
-+ __DEBUG_RT_MUTEX_INITIALIZER(mutexname)
-+
-+
-+#define __RT_MUTEX_INITIALIZER(mutexname) \
-+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) }
-+
-+#define __RT_MUTEX_INITIALIZER_SAVE_STATE(mutexname) \
-+ { __RT_MUTEX_INITIALIZER_PLAIN(mutexname) \
-+ , .save_state = 1 }
+- spin_lock_irq(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ rb_head_page_deactivate(cpu_buffer);
- #define DEFINE_RT_MUTEX(mutexname) \
- struct rt_mutex mutexname = __RT_MUTEX_INITIALIZER(mutexname)
-@@ -90,6 +111,7 @@ extern void rt_mutex_destroy(struct rt_m
- extern void rt_mutex_lock(struct rt_mutex *lock);
- extern int rt_mutex_lock_interruptible(struct rt_mutex *lock,
- int detect_deadlock);
-+extern int rt_mutex_lock_killable(struct rt_mutex *lock, int detect_deadlock);
- extern int rt_mutex_timed_lock(struct rt_mutex *lock,
- struct hrtimer_sleeper *timeout,
- int detect_deadlock);
-Index: linux-3.0/kernel/rtmutex.c
-===================================================================
---- linux-3.0.orig/kernel/rtmutex.c
-+++ linux-3.0/kernel/rtmutex.c
-@@ -8,6 +8,12 @@
- * Copyright (C) 2005 Kihon Technologies Inc., Steven Rostedt
- * Copyright (C) 2006 Esben Nielsen
- *
-+ * Adaptive Spinlocks:
-+ * Copyright (C) 2008 Novell, Inc., Gregory Haskins, Sven Dietrich,
-+ * and Peter Morreale,
-+ * Adaptive Spinlocks simplification:
-+ * Copyright (C) 2008 Red Hat, Inc., Steven Rostedt <srostedt at redhat.com>
-+ *
- * See Documentation/rt-mutex-design.txt for details.
- */
- #include <linux/spinlock.h>
-@@ -67,6 +73,11 @@ static void fixup_rt_mutex_waiters(struc
- clear_rt_mutex_waiters(lock);
+ for (i = 0; i < nr_pages; i++) {
+@@ -1296,7 +1338,7 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer,
+ rb_check_pages(cpu_buffer);
+
+ out:
+- spin_unlock_irq(&cpu_buffer->reader_lock);
++ read_buffer_unlock(cpu_buffer, flags, locked);
}
-+static int rt_mutex_real_waiter(struct rt_mutex_waiter *waiter)
-+{
-+ return waiter && waiter != PI_WAKEUP_INPROGRESS;
-+}
-+
- /*
- * We can speed up the acquire/release, if the architecture
- * supports cmpxchg and if there's no debugging state to be set up
-@@ -90,6 +101,12 @@ static inline void mark_rt_mutex_waiters
+ /**
+@@ -2784,15 +2826,16 @@ void ring_buffer_iter_reset(struct ring_buffer_iter *iter)
+ {
+ struct ring_buffer_per_cpu *cpu_buffer;
+ unsigned long flags;
++ int locked;
+
+ if (!iter)
+ return;
+
+ cpu_buffer = iter->cpu_buffer;
+
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ rb_iter_reset(iter);
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
}
- #endif
+ EXPORT_SYMBOL_GPL(ring_buffer_iter_reset);
-+static inline void init_lists(struct rt_mutex *lock)
-+{
-+ if (unlikely(!lock->wait_list.node_list.prev))
-+ plist_head_init_raw(&lock->wait_list, &lock->wait_lock);
-+}
-+
- /*
- * Calculate task priority from the waiter list priority
- *
-@@ -136,6 +153,14 @@ static void rt_mutex_adjust_prio(struct
- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+@@ -3210,21 +3253,6 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
}
+ EXPORT_SYMBOL_GPL(ring_buffer_iter_peek);
-+static void rt_mutex_wake_waiter(struct rt_mutex_waiter *waiter)
-+{
-+ if (waiter->savestate)
-+ wake_up_lock_sleeper(waiter->task);
-+ else
-+ wake_up_process(waiter->task);
-+}
-+
- /*
- * Max number of times we'll walk the boosting chain:
- */
-@@ -196,7 +221,7 @@ static int rt_mutex_adjust_prio_chain(st
- * reached or the state of the chain has changed while we
- * dropped the locks.
- */
-- if (!waiter)
-+ if (!rt_mutex_real_waiter(waiter))
- goto out_unlock_pi;
+-static inline int rb_ok_to_lock(void)
+-{
+- /*
+- * If an NMI die dumps out the content of the ring buffer
+- * do not grab locks. We also permanently disable the ring
+- * buffer too. A one time deal is all you get from reading
+- * the ring buffer from an NMI.
+- */
+- if (likely(!in_nmi()))
+- return 1;
+-
+- tracing_off_permanent();
+- return 0;
+-}
+-
+ /**
+ * ring_buffer_peek - peek at the next event to be read
+ * @buffer: The ring buffer to read
+@@ -3242,22 +3270,17 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts,
+ struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
+ struct ring_buffer_event *event;
+ unsigned long flags;
+- int dolock;
++ int locked;
- /*
-@@ -247,13 +272,15 @@ static int rt_mutex_adjust_prio_chain(st
- /* Release the task */
- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
- if (!rt_mutex_owner(lock)) {
-+ struct rt_mutex_waiter *lock_top_waiter;
-+
- /*
- * If the requeue above changed the top waiter, then we need
- * to wake the new top waiter up to try to get the lock.
- */
+ if (!cpumask_test_cpu(cpu, buffer->cpumask))
+ return NULL;
+
+- dolock = rb_ok_to_lock();
+ again:
+- local_irq_save(flags);
+- if (dolock)
+- spin_lock(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ event = rb_buffer_peek(cpu_buffer, ts, lost_events);
+ if (event && event->type_len == RINGBUF_TYPE_PADDING)
+ rb_advance_reader(cpu_buffer);
+- if (dolock)
+- spin_unlock(&cpu_buffer->reader_lock);
+- local_irq_restore(flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
+ if (event && event->type_len == RINGBUF_TYPE_PADDING)
+ goto again;
+@@ -3279,11 +3302,12 @@ ring_buffer_iter_peek(struct ring_buffer_iter *iter, u64 *ts)
+ struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
+ struct ring_buffer_event *event;
+ unsigned long flags;
++ int locked;
+
+ again:
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ event = rb_iter_peek(iter, ts);
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
+ if (event && event->type_len == RINGBUF_TYPE_PADDING)
+ goto again;
+@@ -3309,9 +3333,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
+ struct ring_buffer_per_cpu *cpu_buffer;
+ struct ring_buffer_event *event = NULL;
+ unsigned long flags;
+- int dolock;
-
-- if (top_waiter != rt_mutex_top_waiter(lock))
-- wake_up_process(rt_mutex_top_waiter(lock)->task);
-+ lock_top_waiter = rt_mutex_top_waiter(lock);
-+ if (top_waiter != lock_top_waiter)
-+ rt_mutex_wake_waiter(lock_top_waiter);
- raw_spin_unlock(&lock->wait_lock);
- goto out_put_task;
+- dolock = rb_ok_to_lock();
++ int locked;
+
+ again:
+ /* might be called in atomic */
+@@ -3321,9 +3343,7 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
+ goto out;
+
+ cpu_buffer = buffer->buffers[cpu];
+- local_irq_save(flags);
+- if (dolock)
+- spin_lock(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+
+ event = rb_buffer_peek(cpu_buffer, ts, lost_events);
+ if (event) {
+@@ -3331,9 +3351,8 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts,
+ rb_advance_reader(cpu_buffer);
}
-@@ -298,6 +325,25 @@ static int rt_mutex_adjust_prio_chain(st
- return ret;
- }
+- if (dolock)
+- spin_unlock(&cpu_buffer->reader_lock);
+- local_irq_restore(flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
-+#define STEAL_NORMAL 0
-+#define STEAL_LATERAL 1
-+
-+/*
-+ * Note that RT tasks are excluded from lateral-steals to prevent the
-+ * introduction of an unbounded latency
-+ */
-+static inline int lock_is_stealable(struct task_struct *task,
-+ struct task_struct *pendowner, int mode)
-+{
-+ if (mode == STEAL_NORMAL || rt_task(task)) {
-+ if (task->prio >= pendowner->prio)
-+ return 0;
-+ } else if (task->prio > pendowner->prio)
-+ return 0;
-+ return 1;
-+}
-+
- /*
- * Try to take an rt-mutex
- *
-@@ -307,8 +353,9 @@ static int rt_mutex_adjust_prio_chain(st
- * @task: the task which wants to acquire the lock
- * @waiter: the waiter that is queued to the lock's wait list. (could be NULL)
- */
--static int try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
-- struct rt_mutex_waiter *waiter)
-+static int
-+__try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
-+ struct rt_mutex_waiter *waiter, int mode)
+
+ out:
+ preempt_enable();
+@@ -3418,17 +3437,18 @@ ring_buffer_read_start(struct ring_buffer_iter *iter)
{
- /*
- * We have to be careful here if the atomic speedups are
-@@ -341,12 +388,14 @@ static int try_to_take_rt_mutex(struct r
- * 3) it is top waiter
- */
- if (rt_mutex_has_waiters(lock)) {
-- if (task->prio >= rt_mutex_top_waiter(lock)->list_entry.prio) {
-- if (!waiter || waiter != rt_mutex_top_waiter(lock))
-- return 0;
-- }
-+ struct task_struct *pown = rt_mutex_top_waiter(lock)->task;
-+
-+ if (task != pown && !lock_is_stealable(task, pown, mode))
-+ return 0;
- }
+ struct ring_buffer_per_cpu *cpu_buffer;
+ unsigned long flags;
++ int locked;
+
+ if (!iter)
+ return;
+
+ cpu_buffer = iter->cpu_buffer;
-+ /* We got the lock. */
-+
- if (waiter || rt_mutex_has_waiters(lock)) {
- unsigned long flags;
- struct rt_mutex_waiter *top;
-@@ -371,7 +420,6 @@ static int try_to_take_rt_mutex(struct r
- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
- }
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ arch_spin_lock(&cpu_buffer->lock);
+ rb_iter_reset(iter);
+ arch_spin_unlock(&cpu_buffer->lock);
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+ }
+ EXPORT_SYMBOL_GPL(ring_buffer_read_start);
-- /* We got the lock. */
- debug_rt_mutex_lock(lock);
+@@ -3462,8 +3482,9 @@ ring_buffer_read(struct ring_buffer_iter *iter, u64 *ts)
+ struct ring_buffer_event *event;
+ struct ring_buffer_per_cpu *cpu_buffer = iter->cpu_buffer;
+ unsigned long flags;
++ int locked;
- rt_mutex_set_owner(lock, task);
-@@ -381,6 +429,13 @@ static int try_to_take_rt_mutex(struct r
- return 1;
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ again:
+ event = rb_iter_peek(iter, ts);
+ if (!event)
+@@ -3474,7 +3495,7 @@ ring_buffer_read(struct ring_buffer_iter *iter, u64 *ts)
+
+ rb_advance_iter(iter);
+ out:
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
+ return event;
}
+@@ -3537,13 +3558,14 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu)
+ {
+ struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu];
+ unsigned long flags;
++ int locked;
-+static inline int
-+try_to_take_rt_mutex(struct rt_mutex *lock, struct task_struct *task,
-+ struct rt_mutex_waiter *waiter)
-+{
-+ return __try_to_take_rt_mutex(lock, task, waiter, STEAL_NORMAL);
-+}
-+
- /*
- * Task blocks on lock.
- *
-@@ -399,6 +454,23 @@ static int task_blocks_on_rt_mutex(struc
- int chain_walk = 0, res;
+ if (!cpumask_test_cpu(cpu, buffer->cpumask))
+ return;
- raw_spin_lock_irqsave(&task->pi_lock, flags);
-+
-+ /*
-+ * In the case of futex requeue PI, this will be a proxy
-+ * lock. The task will wake unaware that it is enqueueed on
-+ * this lock. Avoid blocking on two locks and corrupting
-+ * pi_blocked_on via the PI_WAKEUP_INPROGRESS
-+ * flag. futex_wait_requeue_pi() sets this when it wakes up
-+ * before requeue (due to a signal or timeout). Do not enqueue
-+ * the task if PI_WAKEUP_INPROGRESS is set.
-+ */
-+ if (task != current && task->pi_blocked_on == PI_WAKEUP_INPROGRESS) {
-+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
-+ return -EAGAIN;
-+ }
-+
-+ BUG_ON(rt_mutex_real_waiter(task->pi_blocked_on));
-+
- __rt_mutex_adjust_prio(task);
- waiter->task = task;
- waiter->lock = lock;
-@@ -423,7 +495,7 @@ static int task_blocks_on_rt_mutex(struc
- plist_add(&waiter->pi_list_entry, &owner->pi_waiters);
+ atomic_inc(&cpu_buffer->record_disabled);
- __rt_mutex_adjust_prio(owner);
-- if (owner->pi_blocked_on)
-+ if (rt_mutex_real_waiter(owner->pi_blocked_on))
- chain_walk = 1;
- raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
- }
-@@ -478,7 +550,7 @@ static void wakeup_next_waiter(struct rt
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
- raw_spin_unlock_irqrestore(¤t->pi_lock, flags);
+ if (RB_WARN_ON(cpu_buffer, local_read(&cpu_buffer->committing)))
+ goto out;
+@@ -3555,7 +3577,7 @@ void ring_buffer_reset_cpu(struct ring_buffer *buffer, int cpu)
+ arch_spin_unlock(&cpu_buffer->lock);
-- wake_up_process(waiter->task);
-+ rt_mutex_wake_waiter(waiter);
- }
+ out:
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
- /*
-@@ -517,7 +589,7 @@ static void remove_waiter(struct rt_mute
- }
- __rt_mutex_adjust_prio(owner);
+ atomic_dec(&cpu_buffer->record_disabled);
+ }
+@@ -3582,22 +3604,16 @@ int ring_buffer_empty(struct ring_buffer *buffer)
+ {
+ struct ring_buffer_per_cpu *cpu_buffer;
+ unsigned long flags;
+- int dolock;
++ int locked;
+ int cpu;
+ int ret;
-- if (owner->pi_blocked_on)
-+ if (rt_mutex_real_waiter(owner->pi_blocked_on))
- chain_walk = 1;
+- dolock = rb_ok_to_lock();
+-
+ /* yes this is racy, but if you don't like the race, lock the buffer */
+ for_each_buffer_cpu(buffer, cpu) {
+ cpu_buffer = buffer->buffers[cpu];
+- local_irq_save(flags);
+- if (dolock)
+- spin_lock(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ ret = rb_per_cpu_empty(cpu_buffer);
+- if (dolock)
+- spin_unlock(&cpu_buffer->reader_lock);
+- local_irq_restore(flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
- raw_spin_unlock_irqrestore(&owner->pi_lock, flags);
-@@ -551,23 +623,316 @@ void rt_mutex_adjust_pi(struct task_stru
- raw_spin_lock_irqsave(&task->pi_lock, flags);
+ if (!ret)
+ return 0;
+@@ -3616,22 +3632,16 @@ int ring_buffer_empty_cpu(struct ring_buffer *buffer, int cpu)
+ {
+ struct ring_buffer_per_cpu *cpu_buffer;
+ unsigned long flags;
+- int dolock;
++ int locked;
+ int ret;
- waiter = task->pi_blocked_on;
-- if (!waiter || waiter->list_entry.prio == task->prio) {
-+ if (!rt_mutex_real_waiter(waiter) ||
-+ waiter->list_entry.prio == task->prio) {
- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
- return;
- }
+ if (!cpumask_test_cpu(cpu, buffer->cpumask))
+ return 1;
-- raw_spin_unlock_irqrestore(&task->pi_lock, flags);
+- dolock = rb_ok_to_lock();
-
- /* gets dropped in rt_mutex_adjust_prio_chain()! */
- get_task_struct(task);
-+ raw_spin_unlock_irqrestore(&task->pi_lock, flags);
- rt_mutex_adjust_prio_chain(task, 0, NULL, NULL, task);
+ cpu_buffer = buffer->buffers[cpu];
+- local_irq_save(flags);
+- if (dolock)
+- spin_lock(&cpu_buffer->reader_lock);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+ ret = rb_per_cpu_empty(cpu_buffer);
+- if (dolock)
+- spin_unlock(&cpu_buffer->reader_lock);
+- local_irq_restore(flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
+ return ret;
}
+@@ -3805,6 +3815,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer,
+ unsigned int commit;
+ unsigned int read;
+ u64 save_timestamp;
++ int locked;
+ int ret = -1;
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+/*
-+ * preemptible spin_lock functions:
-+ */
-+static inline void rt_spin_lock_fastlock(struct rt_mutex *lock,
-+ void (*slowfn)(struct rt_mutex *lock))
-+{
-+ might_sleep();
-+
-+ if (likely(rt_mutex_cmpxchg(lock, NULL, current)))
-+ rt_mutex_deadlock_account_lock(lock, current);
-+ else
-+ slowfn(lock);
-+}
-+
-+static inline void rt_spin_lock_fastunlock(struct rt_mutex *lock,
-+ void (*slowfn)(struct rt_mutex *lock))
-+{
-+ if (likely(rt_mutex_cmpxchg(lock, current, NULL)))
-+ rt_mutex_deadlock_account_unlock(current);
-+ else
-+ slowfn(lock);
-+}
-+
-+#ifdef CONFIG_SMP
-+/*
-+ * Note that owner is a speculative pointer and dereferencing relies
-+ * on rcu_read_lock() and the check against the lock owner.
-+ */
-+static int adaptive_wait(struct rt_mutex *lock,
-+ struct task_struct *owner)
-+{
-+ int res = 0;
-+
-+ rcu_read_lock();
-+ for (;;) {
-+ if (owner != rt_mutex_owner(lock))
-+ break;
-+ /*
-+ * Ensure that owner->on_cpu is dereferenced _after_
-+ * checking the above to be valid.
-+ */
-+ barrier();
-+ if (!owner->on_cpu) {
-+ res = 1;
-+ break;
-+ }
-+ cpu_relax();
-+ }
-+ rcu_read_unlock();
-+ return res;
-+}
-+#else
-+static int adaptive_wait(struct rt_mutex *lock,
-+ struct task_struct *orig_owner)
-+{
-+ return 1;
-+}
+ if (!cpumask_test_cpu(cpu, buffer->cpumask))
+@@ -3826,7 +3837,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer,
+ if (!bpage)
+ goto out;
+
+- spin_lock_irqsave(&cpu_buffer->reader_lock, flags);
++ locked = read_buffer_lock(cpu_buffer, &flags);
+
+ reader = rb_get_reader_page(cpu_buffer);
+ if (!reader)
+@@ -3949,7 +3960,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer,
+ memset(&bpage->data[commit], 0, BUF_PAGE_SIZE - commit);
+
+ out_unlock:
+- spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags);
++ read_buffer_unlock(cpu_buffer, flags, locked);
+
+ out:
+ return ret;
+diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
+index 0731e81a..0a42c28 100644
+--- a/kernel/trace/trace.c
++++ b/kernel/trace/trace.c
+@@ -341,7 +341,7 @@ unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK |
+ TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE;
+
+ static int trace_stop_count;
+-static DEFINE_SPINLOCK(tracing_start_lock);
++static DEFINE_RAW_SPINLOCK(tracing_start_lock);
+
+ /**
+ * trace_wake_up - wake up tasks waiting for trace input
+@@ -351,6 +351,7 @@ static DEFINE_SPINLOCK(tracing_start_lock);
+ */
+ void trace_wake_up(void)
+ {
++#ifndef CONFIG_PREEMPT_RT_FULL
+ int cpu;
+
+ if (trace_flags & TRACE_ITER_BLOCK)
+@@ -363,6 +364,7 @@ void trace_wake_up(void)
+ if (!runqueue_is_locked(cpu))
+ wake_up(&trace_wait);
+ put_cpu();
+#endif
-+
-+# define pi_lock(lock) raw_spin_lock_irq(lock)
-+# define pi_unlock(lock) raw_spin_unlock_irq(lock)
-+
-+/*
-+ * Slow path lock function spin_lock style: this variant is very
-+ * careful not to miss any non-lock wakeups.
-+ *
-+ * We store the current state under p->pi_lock in p->saved_state and
-+ * the try_to_wake_up() code handles this accordingly.
-+ */
-+static void noinline __sched rt_spin_lock_slowlock(struct rt_mutex *lock)
-+{
-+ struct task_struct *lock_owner, *self = current;
-+ struct rt_mutex_waiter waiter, *top_waiter;
-+ int ret;
-+
-+ rt_mutex_init_waiter(&waiter, true);
-+
-+ raw_spin_lock(&lock->wait_lock);
-+ init_lists(lock);
-+
-+ if (__try_to_take_rt_mutex(lock, self, NULL, STEAL_LATERAL)) {
-+ raw_spin_unlock(&lock->wait_lock);
-+ return;
-+ }
-+
-+ BUG_ON(rt_mutex_owner(lock) == self);
-+
-+ /*
-+ * We save whatever state the task is in and we'll restore it
-+ * after acquiring the lock taking real wakeups into account
-+ * as well. We are serialized via pi_lock against wakeups. See
-+ * try_to_wake_up().
-+ */
-+ pi_lock(&self->pi_lock);
-+ self->saved_state = self->state;
-+ __set_current_state(TASK_UNINTERRUPTIBLE);
-+ pi_unlock(&self->pi_lock);
-+
-+ ret = task_blocks_on_rt_mutex(lock, &waiter, self, 0);
-+ BUG_ON(ret);
-+
-+ for (;;) {
-+ /* Try to acquire the lock again. */
-+ if (__try_to_take_rt_mutex(lock, self, &waiter, STEAL_LATERAL))
-+ break;
-+
-+ top_waiter = rt_mutex_top_waiter(lock);
-+ lock_owner = rt_mutex_owner(lock);
-+
-+ raw_spin_unlock(&lock->wait_lock);
-+
-+ debug_rt_mutex_print_deadlock(&waiter);
-+
-+ if (top_waiter != &waiter || adaptive_wait(lock, lock_owner))
-+ schedule_rt_mutex(lock);
-+
-+ raw_spin_lock(&lock->wait_lock);
-+
-+ pi_lock(&self->pi_lock);
-+ __set_current_state(TASK_UNINTERRUPTIBLE);
-+ pi_unlock(&self->pi_lock);
-+ }
-+
-+ /*
-+ * Restore the task state to current->saved_state. We set it
-+ * to the original state above and the try_to_wake_up() code
-+ * has possibly updated it when a real (non-rtmutex) wakeup
-+ * happened while we were blocked. Clear saved_state so
-+ * try_to_wakeup() does not get confused.
-+ */
-+ pi_lock(&self->pi_lock);
-+ __set_current_state(self->saved_state);
-+ self->saved_state = TASK_RUNNING;
-+ pi_unlock(&self->pi_lock);
-+
-+ /*
-+ * try_to_take_rt_mutex() sets the waiter bit
-+ * unconditionally. We might have to fix that up:
-+ */
-+ fixup_rt_mutex_waiters(lock);
-+
-+ BUG_ON(rt_mutex_has_waiters(lock) && &waiter == rt_mutex_top_waiter(lock));
-+ BUG_ON(!plist_node_empty(&waiter.list_entry));
-+
-+ raw_spin_unlock(&lock->wait_lock);
-+
-+ debug_rt_mutex_free_waiter(&waiter);
-+}
-+
-+/*
-+ * Slow path to release a rt_mutex spin_lock style
-+ */
-+static void noinline __sched rt_spin_lock_slowunlock(struct rt_mutex *lock)
-+{
-+ raw_spin_lock(&lock->wait_lock);
-+
-+ debug_rt_mutex_unlock(lock);
-+
-+ rt_mutex_deadlock_account_unlock(current);
-+
-+ if (!rt_mutex_has_waiters(lock)) {
-+ lock->owner = NULL;
-+ raw_spin_unlock(&lock->wait_lock);
-+ return;
-+ }
-+
-+ wakeup_next_waiter(lock);
-+
-+ raw_spin_unlock(&lock->wait_lock);
-+
-+ /* Undo pi boosting.when necessary */
-+ rt_mutex_adjust_prio(current);
-+}
-+
-+void __lockfunc rt_spin_lock(spinlock_t *lock)
-+{
-+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
-+ spin_acquire(&lock->dep_map, 0, 0, _RET_IP_);
-+}
-+EXPORT_SYMBOL(rt_spin_lock);
-+
-+void __lockfunc __rt_spin_lock(struct rt_mutex *lock)
-+{
-+ rt_spin_lock_fastlock(lock, rt_spin_lock_slowlock);
-+}
-+EXPORT_SYMBOL(__rt_spin_lock);
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass)
-+{
-+ rt_spin_lock_fastlock(&lock->lock, rt_spin_lock_slowlock);
-+ spin_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
-+}
-+EXPORT_SYMBOL(rt_spin_lock_nested);
+ }
+
+ static int __init set_buf_size(char *str)
+@@ -716,6 +718,12 @@ update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
+ }
+ #endif /* CONFIG_TRACER_MAX_TRACE */
+
++#ifndef CONFIG_PREEMPT_RT_FULL
++static void default_wait_pipe(struct trace_iterator *iter);
++#else
++#define default_wait_pipe poll_wait_pipe
+#endif
+
-+void __lockfunc rt_spin_unlock(spinlock_t *lock)
-+{
-+ /* NOTE: we always pass in '1' for nested, for simplicity */
-+ spin_release(&lock->dep_map, 1, _RET_IP_);
-+ rt_spin_lock_fastunlock(&lock->lock, rt_spin_lock_slowunlock);
-+}
-+EXPORT_SYMBOL(rt_spin_unlock);
-+
-+void __lockfunc __rt_spin_unlock(struct rt_mutex *lock)
-+{
-+ rt_spin_lock_fastunlock(lock, rt_spin_lock_slowunlock);
-+}
-+EXPORT_SYMBOL(__rt_spin_unlock);
-+
-+/*
-+ * Wait for the lock to get unlocked: instead of polling for an unlock
-+ * (like raw spinlocks do), we lock and unlock, to force the kernel to
-+ * schedule if there's contention:
-+ */
-+void __lockfunc rt_spin_unlock_wait(spinlock_t *lock)
-+{
-+ spin_lock(lock);
-+ spin_unlock(lock);
-+}
-+EXPORT_SYMBOL(rt_spin_unlock_wait);
-+
-+int __lockfunc rt_spin_trylock(spinlock_t *lock)
-+{
-+ int ret = rt_mutex_trylock(&lock->lock);
-+
-+ if (ret)
-+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
-+ return ret;
-+}
-+EXPORT_SYMBOL(rt_spin_trylock);
-+
-+int __lockfunc rt_spin_trylock_bh(spinlock_t *lock)
-+{
-+ int ret;
-+
-+ local_bh_disable();
-+ ret = rt_mutex_trylock(&lock->lock);
-+ if (ret) {
-+ migrate_disable();
-+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
-+ } else
-+ local_bh_enable();
-+ return ret;
-+}
-+EXPORT_SYMBOL(rt_spin_trylock_bh);
+ /**
+ * register_tracer - register a tracer with the ftrace system.
+ * @type - the plugin for the tracer
+@@ -958,7 +966,7 @@ void tracing_start(void)
+ if (tracing_disabled)
+ return;
+
+- spin_lock_irqsave(&tracing_start_lock, flags);
++ raw_spin_lock_irqsave(&tracing_start_lock, flags);
+ if (--trace_stop_count) {
+ if (trace_stop_count < 0) {
+ /* Someone screwed up their debugging */
+@@ -983,7 +991,7 @@ void tracing_start(void)
+
+ ftrace_start();
+ out:
+- spin_unlock_irqrestore(&tracing_start_lock, flags);
++ raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
+ }
+
+ /**
+@@ -998,7 +1006,7 @@ void tracing_stop(void)
+ unsigned long flags;
+
+ ftrace_stop();
+- spin_lock_irqsave(&tracing_start_lock, flags);
++ raw_spin_lock_irqsave(&tracing_start_lock, flags);
+ if (trace_stop_count++)
+ goto out;
+
+@@ -1016,7 +1024,7 @@ void tracing_stop(void)
+ arch_spin_unlock(&ftrace_max_lock);
+
+ out:
+- spin_unlock_irqrestore(&tracing_start_lock, flags);
++ raw_spin_unlock_irqrestore(&tracing_start_lock, flags);
+ }
+
+ void trace_stop_cmdline_recording(void);
+@@ -1120,6 +1128,8 @@ tracing_generic_entry_update(struct trace_entry *entry, unsigned long flags,
+ ((pc & HARDIRQ_MASK) ? TRACE_FLAG_HARDIRQ : 0) |
+ ((pc & SOFTIRQ_MASK) ? TRACE_FLAG_SOFTIRQ : 0) |
+ (need_resched() ? TRACE_FLAG_NEED_RESCHED : 0);
+
-+int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags)
++ entry->migrate_disable = (tsk) ? __migrate_disabled(tsk) & 0xFF : 0;
+ }
+ EXPORT_SYMBOL_GPL(tracing_generic_entry_update);
+
+@@ -1757,9 +1767,10 @@ static void print_lat_help_header(struct seq_file *m)
+ seq_puts(m, "# | / _----=> need-resched \n");
+ seq_puts(m, "# || / _---=> hardirq/softirq \n");
+ seq_puts(m, "# ||| / _--=> preempt-depth \n");
+- seq_puts(m, "# |||| / delay \n");
+- seq_puts(m, "# cmd pid ||||| time | caller \n");
+- seq_puts(m, "# \\ / ||||| \\ | / \n");
++ seq_puts(m, "# |||| / _--=> migrate-disable\n");
++ seq_puts(m, "# ||||| / delay \n");
++ seq_puts(m, "# cmd pid |||||| time | caller \n");
++ seq_puts(m, "# \\ / ||||| \\ | / \n");
+ }
+
+ static void print_func_help_header(struct seq_file *m)
+@@ -3067,6 +3078,7 @@ static int tracing_release_pipe(struct inode *inode, struct file *file)
+ return 0;
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ static unsigned int
+ tracing_poll_pipe(struct file *filp, poll_table *poll_table)
+ {
+@@ -3088,8 +3100,7 @@ tracing_poll_pipe(struct file *filp, poll_table *poll_table)
+ }
+ }
+
+-
+-void default_wait_pipe(struct trace_iterator *iter)
++static void default_wait_pipe(struct trace_iterator *iter)
+ {
+ DEFINE_WAIT(wait);
+
+@@ -3100,6 +3111,20 @@ void default_wait_pipe(struct trace_iterator *iter)
+
+ finish_wait(&trace_wait, &wait);
+ }
++#else
++static unsigned int
++tracing_poll_pipe(struct file *filp, poll_table *poll_table)
+{
-+ int ret;
-+
-+ *flags = 0;
-+ migrate_disable();
-+ ret = rt_mutex_trylock(&lock->lock);
-+ if (ret)
-+ spin_acquire(&lock->dep_map, 0, 1, _RET_IP_);
-+ else
-+ migrate_enable();
-+ return ret;
-+}
-+EXPORT_SYMBOL(rt_spin_trylock_irqsave);
++ struct trace_iterator *iter = filp->private_data;
+
-+int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock)
-+{
-+ /* Subtract 1 from counter unless that drops it to 0 (ie. it was 1) */
-+ if (atomic_add_unless(atomic, -1, 1))
-+ return 0;
-+ migrate_disable();
-+ rt_spin_lock(lock);
-+ if (atomic_dec_and_test(atomic))
-+ return 1;
-+ rt_spin_unlock(lock);
-+ migrate_enable();
++ if ((trace_flags & TRACE_ITER_BLOCK) || !trace_empty(iter))
++ return POLLIN | POLLRDNORM;
++ poll_wait_pipe(iter);
++ if (!trace_empty(iter))
++ return POLLIN | POLLRDNORM;
+ return 0;
+}
-+EXPORT_SYMBOL(atomic_dec_and_spin_lock);
-+
-+void
-+__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key)
-+{
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ /*
-+ * Make sure we are not reinitializing a held lock:
-+ */
-+ debug_check_no_locks_freed((void *)lock, sizeof(*lock));
-+ lockdep_init_map(&lock->dep_map, name, key, 0);
+#endif
-+}
-+EXPORT_SYMBOL(__rt_spin_lock_init);
-+
-+#endif /* PREEMPT_RT_FULL */
-+
- /**
- * __rt_mutex_slowlock() - Perform the wait-wake-try-to-take loop
- * @lock: the rt_mutex to take
- * @state: the state the task should block in (TASK_INTERRUPTIBLE
-- * or TASK_UNINTERRUPTIBLE)
-+ * or TASK_UNINTERRUPTIBLE)
- * @timeout: the pre-initialized and started timer, or NULL for none
- * @waiter: the pre-initialized rt_mutex_waiter
- *
-@@ -623,9 +988,10 @@ rt_mutex_slowlock(struct rt_mutex *lock,
- struct rt_mutex_waiter waiter;
- int ret = 0;
-- debug_rt_mutex_init_waiter(&waiter);
-+ rt_mutex_init_waiter(&waiter, false);
+ /*
+ * This is a make-shift waitqueue.
+diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
+index f807407..a25df79 100644
+--- a/kernel/trace/trace.h
++++ b/kernel/trace/trace.h
+@@ -322,7 +322,6 @@ void trace_init_global_iter(struct trace_iterator *iter);
- raw_spin_lock(&lock->wait_lock);
-+ init_lists(lock);
+ void tracing_iter_reset(struct trace_iterator *iter, int cpu);
- /* Try to acquire the lock again: */
- if (try_to_take_rt_mutex(lock, current, NULL)) {
-@@ -678,6 +1044,7 @@ rt_mutex_slowtrylock(struct rt_mutex *lo
- int ret = 0;
+-void default_wait_pipe(struct trace_iterator *iter);
+ void poll_wait_pipe(struct trace_iterator *iter);
- raw_spin_lock(&lock->wait_lock);
-+ init_lists(lock);
+ void ftrace(struct trace_array *tr,
+diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
+index 3e2a7c9..c71844c 100644
+--- a/kernel/trace/trace_events.c
++++ b/kernel/trace/trace_events.c
+@@ -116,7 +116,8 @@ static int trace_define_common_fields(void)
+ __common_field(unsigned char, flags);
+ __common_field(unsigned char, preempt_count);
+ __common_field(int, pid);
+- __common_field(int, padding);
++ __common_field(unsigned short, migrate_disable);
++ __common_field(unsigned short, padding);
- if (likely(rt_mutex_owner(lock) != current)) {
+ return ret;
+ }
+diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c
+index c77424b..456629b 100644
+--- a/kernel/trace/trace_irqsoff.c
++++ b/kernel/trace/trace_irqsoff.c
+@@ -17,13 +17,14 @@
+ #include <linux/fs.h>
-@@ -791,12 +1158,12 @@ EXPORT_SYMBOL_GPL(rt_mutex_lock);
- /**
- * rt_mutex_lock_interruptible - lock a rt_mutex interruptible
- *
-- * @lock: the rt_mutex to be locked
-+ * @lock: the rt_mutex to be locked
- * @detect_deadlock: deadlock detection on/off
- *
- * Returns:
-- * 0 on success
-- * -EINTR when interrupted by a signal
-+ * 0 on success
-+ * -EINTR when interrupted by a signal
- * -EDEADLK when the lock would deadlock (when deadlock detection is on)
- */
- int __sched rt_mutex_lock_interruptible(struct rt_mutex *lock,
-@@ -810,17 +1177,38 @@ int __sched rt_mutex_lock_interruptible(
- EXPORT_SYMBOL_GPL(rt_mutex_lock_interruptible);
+ #include "trace.h"
++#include <trace/events/hist.h>
- /**
-+ * rt_mutex_lock_killable - lock a rt_mutex killable
-+ *
-+ * @lock: the rt_mutex to be locked
-+ * @detect_deadlock: deadlock detection on/off
-+ *
-+ * Returns:
-+ * 0 on success
-+ * -EINTR when interrupted by a signal
-+ * -EDEADLK when the lock would deadlock (when deadlock detection is on)
-+ */
-+int __sched rt_mutex_lock_killable(struct rt_mutex *lock,
-+ int detect_deadlock)
-+{
-+ might_sleep();
-+
-+ return rt_mutex_fastlock(lock, TASK_KILLABLE,
-+ detect_deadlock, rt_mutex_slowlock);
-+}
-+EXPORT_SYMBOL_GPL(rt_mutex_lock_killable);
-+
-+/**
- * rt_mutex_timed_lock - lock a rt_mutex interruptible
- * the timeout structure is provided
- * by the caller
- *
-- * @lock: the rt_mutex to be locked
-+ * @lock: the rt_mutex to be locked
- * @timeout: timeout structure or NULL (no timeout)
- * @detect_deadlock: deadlock detection on/off
- *
- * Returns:
-- * 0 on success
-- * -EINTR when interrupted by a signal
-+ * 0 on success
-+ * -EINTR when interrupted by a signal
- * -ETIMEDOUT when the timeout expired
- * -EDEADLK when the lock would deadlock (when deadlock detection is on)
+ static struct trace_array *irqsoff_trace __read_mostly;
+ static int tracer_enabled __read_mostly;
+
+ static DEFINE_PER_CPU(int, tracing_cpu);
+
+-static DEFINE_SPINLOCK(max_trace_lock);
++static DEFINE_RAW_SPINLOCK(max_trace_lock);
+
+ enum {
+ TRACER_IRQS_OFF = (1 << 1),
+@@ -319,7 +320,7 @@ check_critical_timing(struct trace_array *tr,
+ if (!report_latency(delta))
+ goto out;
+
+- spin_lock_irqsave(&max_trace_lock, flags);
++ raw_spin_lock_irqsave(&max_trace_lock, flags);
+
+ /* check if we are still the max latency */
+ if (!report_latency(delta))
+@@ -342,7 +343,7 @@ check_critical_timing(struct trace_array *tr,
+ max_sequence++;
+
+ out_unlock:
+- spin_unlock_irqrestore(&max_trace_lock, flags);
++ raw_spin_unlock_irqrestore(&max_trace_lock, flags);
+
+ out:
+ data->critical_sequence = max_sequence;
+@@ -424,11 +425,13 @@ void start_critical_timings(void)
+ {
+ if (preempt_trace() || irq_trace())
+ start_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
++ trace_preemptirqsoff_hist(TRACE_START, 1);
+ }
+ EXPORT_SYMBOL_GPL(start_critical_timings);
+
+ void stop_critical_timings(void)
+ {
++ trace_preemptirqsoff_hist(TRACE_STOP, 0);
+ if (preempt_trace() || irq_trace())
+ stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
+ }
+@@ -438,6 +441,7 @@ EXPORT_SYMBOL_GPL(stop_critical_timings);
+ #ifdef CONFIG_PROVE_LOCKING
+ void time_hardirqs_on(unsigned long a0, unsigned long a1)
+ {
++ trace_preemptirqsoff_hist(IRQS_ON, 0);
+ if (!preempt_trace() && irq_trace())
+ stop_critical_timing(a0, a1);
+ }
+@@ -446,6 +450,7 @@ void time_hardirqs_off(unsigned long a0, unsigned long a1)
+ {
+ if (!preempt_trace() && irq_trace())
+ start_critical_timing(a0, a1);
++ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+ }
+
+ #else /* !CONFIG_PROVE_LOCKING */
+@@ -471,6 +476,7 @@ inline void print_irqtrace_events(struct task_struct *curr)
*/
-@@ -889,7 +1277,6 @@ EXPORT_SYMBOL_GPL(rt_mutex_destroy);
- void __rt_mutex_init(struct rt_mutex *lock, const char *name)
+ void trace_hardirqs_on(void)
{
- lock->owner = NULL;
-- raw_spin_lock_init(&lock->wait_lock);
- plist_head_init_raw(&lock->wait_list, &lock->wait_lock);
++ trace_preemptirqsoff_hist(IRQS_ON, 0);
+ if (!preempt_trace() && irq_trace())
+ stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
+ }
+@@ -480,11 +486,13 @@ void trace_hardirqs_off(void)
+ {
+ if (!preempt_trace() && irq_trace())
+ start_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
++ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+ }
+ EXPORT_SYMBOL(trace_hardirqs_off);
+
+ void trace_hardirqs_on_caller(unsigned long caller_addr)
+ {
++ trace_preemptirqsoff_hist(IRQS_ON, 0);
+ if (!preempt_trace() && irq_trace())
+ stop_critical_timing(CALLER_ADDR0, caller_addr);
+ }
+@@ -494,6 +502,7 @@ void trace_hardirqs_off_caller(unsigned long caller_addr)
+ {
+ if (!preempt_trace() && irq_trace())
+ start_critical_timing(CALLER_ADDR0, caller_addr);
++ trace_preemptirqsoff_hist(IRQS_OFF, 1);
+ }
+ EXPORT_SYMBOL(trace_hardirqs_off_caller);
+
+@@ -503,13 +512,15 @@ EXPORT_SYMBOL(trace_hardirqs_off_caller);
+ #ifdef CONFIG_PREEMPT_TRACER
+ void trace_preempt_on(unsigned long a0, unsigned long a1)
+ {
+- if (preempt_trace())
++ trace_preemptirqsoff_hist(PREEMPT_ON, 0);
++ if (preempt_trace() && !irq_trace())
+ stop_critical_timing(a0, a1);
+ }
- debug_rt_mutex_init(lock, name);
-@@ -909,7 +1296,7 @@ EXPORT_SYMBOL_GPL(__rt_mutex_init);
- void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
- struct task_struct *proxy_owner)
+ void trace_preempt_off(unsigned long a0, unsigned long a1)
{
-- __rt_mutex_init(lock, NULL);
-+ rt_mutex_init(lock);
- debug_rt_mutex_proxy_lock(lock, proxy_owner);
- rt_mutex_set_owner(lock, proxy_owner);
- rt_mutex_deadlock_account_lock(lock, proxy_owner);
-Index: linux-3.0/kernel/futex.c
-===================================================================
---- linux-3.0.orig/kernel/futex.c
-+++ linux-3.0/kernel/futex.c
-@@ -1410,6 +1410,16 @@ retry_private:
- requeue_pi_wake_futex(this, &key2, hb2);
- drop_count++;
- continue;
-+ } else if (ret == -EAGAIN) {
-+ /*
-+ * Waiter was woken by timeout or
-+ * signal and has set pi_blocked_on to
-+ * PI_WAKEUP_INPROGRESS before we
-+ * tried to enqueue it on the rtmutex.
-+ */
-+ this->pi_state = NULL;
-+ free_pi_state(pi_state);
-+ continue;
- } else if (ret) {
- /* -EDEADLK */
- this->pi_state = NULL;
-@@ -2254,7 +2264,7 @@ static int futex_wait_requeue_pi(u32 __u
- struct hrtimer_sleeper timeout, *to = NULL;
- struct rt_mutex_waiter rt_waiter;
- struct rt_mutex *pi_mutex = NULL;
-- struct futex_hash_bucket *hb;
-+ struct futex_hash_bucket *hb, *hb2;
- union futex_key key2 = FUTEX_KEY_INIT;
- struct futex_q q = futex_q_init;
- int res, ret;
-@@ -2276,8 +2286,7 @@ static int futex_wait_requeue_pi(u32 __u
- * The waiter is allocated on our stack, manipulated by the requeue
- * code while we sleep on uaddr.
- */
-- debug_rt_mutex_init_waiter(&rt_waiter);
-- rt_waiter.task = NULL;
-+ rt_mutex_init_waiter(&rt_waiter, false);
-
- ret = get_futex_key(uaddr2, flags & FLAGS_SHARED, &key2, VERIFY_WRITE);
- if (unlikely(ret != 0))
-@@ -2298,20 +2307,55 @@ static int futex_wait_requeue_pi(u32 __u
- /* Queue the futex_q, drop the hb lock, wait for wakeup. */
- futex_wait_queue_me(hb, &q, to);
+- if (preempt_trace())
++ trace_preemptirqsoff_hist(PREEMPT_ON, 1);
++ if (preempt_trace() && !irq_trace())
+ start_critical_timing(a0, a1);
+ }
+ #endif /* CONFIG_PREEMPT_TRACER */
+diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c
+index e37de49..50bef75 100644
+--- a/kernel/trace/trace_output.c
++++ b/kernel/trace/trace_output.c
+@@ -591,6 +591,11 @@ int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry)
+ else
+ ret = trace_seq_putc(s, '.');
-- spin_lock(&hb->lock);
-- ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to);
-- spin_unlock(&hb->lock);
-- if (ret)
-- goto out_put_keys;
-+ /*
-+ * On RT we must avoid races with requeue and trying to block
-+ * on two mutexes (hb->lock and uaddr2's rtmutex) by
-+ * serializing access to pi_blocked_on with pi_lock.
-+ */
-+ raw_spin_lock_irq(¤t->pi_lock);
-+ if (current->pi_blocked_on) {
-+ /*
-+ * We have been requeued or are in the process of
-+ * being requeued.
-+ */
-+ raw_spin_unlock_irq(¤t->pi_lock);
-+ } else {
-+ /*
-+ * Setting pi_blocked_on to PI_WAKEUP_INPROGRESS
-+ * prevents a concurrent requeue from moving us to the
-+ * uaddr2 rtmutex. After that we can safely acquire
-+ * (and possibly block on) hb->lock.
-+ */
-+ current->pi_blocked_on = PI_WAKEUP_INPROGRESS;
-+ raw_spin_unlock_irq(¤t->pi_lock);
-+
-+ spin_lock(&hb->lock);
-+
-+ /*
-+ * Clean up pi_blocked_on. We might leak it otherwise
-+ * when we succeeded with the hb->lock in the fast
-+ * path.
-+ */
-+ raw_spin_lock_irq(¤t->pi_lock);
-+ current->pi_blocked_on = NULL;
-+ raw_spin_unlock_irq(¤t->pi_lock);
++ if (entry->migrate_disable)
++ ret = trace_seq_printf(s, "%x", entry->migrate_disable);
++ else
++ ret = trace_seq_putc(s, '.');
+
-+ ret = handle_early_requeue_pi_wakeup(hb, &q, &key2, to);
-+ spin_unlock(&hb->lock);
-+ if (ret)
-+ goto out_put_keys;
-+ }
+ return ret;
+ }
- /*
-- * In order for us to be here, we know our q.key == key2, and since
-- * we took the hb->lock above, we also know that futex_requeue() has
-- * completed and we no longer have to concern ourselves with a wakeup
-- * race with the atomic proxy lock acquisition by the requeue code. The
-- * futex_requeue dropped our key1 reference and incremented our key2
-- * reference count.
-+ * In order to be here, we have either been requeued, are in
-+ * the process of being requeued, or requeue successfully
-+ * acquired uaddr2 on our behalf. If pi_blocked_on was
-+ * non-null above, we may be racing with a requeue. Do not
-+ * rely on q->lock_ptr to be hb2->lock until after blocking on
-+ * hb->lock or hb2->lock. The futex_requeue dropped our key1
-+ * reference and incremented our key2 reference count.
- */
-+ hb2 = hash_futex(&key2);
+diff --git a/kernel/user.c b/kernel/user.c
+index 9e03e9c..1b5b986 100644
+--- a/kernel/user.c
++++ b/kernel/user.c
+@@ -129,11 +129,11 @@ void free_uid(struct user_struct *up)
+ if (!up)
+ return;
- /* Check if the requeue code acquired the second futex for us. */
- if (!q.rt_waiter) {
-@@ -2320,9 +2364,10 @@ static int futex_wait_requeue_pi(u32 __u
- * did a lock-steal - fix up the PI-state in that case.
- */
- if (q.pi_state && (q.pi_state->owner != current)) {
-- spin_lock(q.lock_ptr);
-+ spin_lock(&hb2->lock);
-+ BUG_ON(&hb2->lock != q.lock_ptr);
- ret = fixup_pi_state_owner(uaddr2, &q, current);
-- spin_unlock(q.lock_ptr);
-+ spin_unlock(&hb2->lock);
- }
- } else {
- /*
-@@ -2335,7 +2380,8 @@ static int futex_wait_requeue_pi(u32 __u
- ret = rt_mutex_finish_proxy_lock(pi_mutex, to, &rt_waiter, 1);
- debug_rt_mutex_free_waiter(&rt_waiter);
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+ if (atomic_dec_and_lock(&up->__count, &uidhash_lock))
+ free_user(up, flags);
+ else
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ }
-- spin_lock(q.lock_ptr);
-+ spin_lock(&hb2->lock);
-+ BUG_ON(&hb2->lock != q.lock_ptr);
- /*
- * Fixup the pi_state owner and possibly acquire the lock if we
- * haven't already.
-Index: linux-3.0/kernel/rtmutex_common.h
-===================================================================
---- linux-3.0.orig/kernel/rtmutex_common.h
-+++ linux-3.0/kernel/rtmutex_common.h
-@@ -49,6 +49,7 @@ struct rt_mutex_waiter {
- struct plist_node pi_list_entry;
- struct task_struct *task;
- struct rt_mutex *lock;
-+ bool savestate;
- #ifdef CONFIG_DEBUG_RT_MUTEXES
- unsigned long ip;
- struct pid *deadlock_task_pid;
-@@ -103,6 +104,8 @@ static inline struct task_struct *rt_mut
- /*
- * PI-futex support (proxy locking functions, etc.):
- */
-+#define PI_WAKEUP_INPROGRESS ((struct rt_mutex_waiter *) 1)
-+
- extern struct task_struct *rt_mutex_next_owner(struct rt_mutex *lock);
- extern void rt_mutex_init_proxy_locked(struct rt_mutex *lock,
- struct task_struct *proxy_owner);
-@@ -123,4 +126,12 @@ extern int rt_mutex_finish_proxy_lock(st
- # include "rtmutex.h"
- #endif
+ struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid)
+diff --git a/kernel/watchdog.c b/kernel/watchdog.c
+index 3d0c56a..84e3cf1 100644
+--- a/kernel/watchdog.c
++++ b/kernel/watchdog.c
+@@ -208,6 +208,8 @@ static struct perf_event_attr wd_hw_attr = {
+ .disabled = 1,
+ };
-+static inline void
-+rt_mutex_init_waiter(struct rt_mutex_waiter *waiter, bool savestate)
-+{
-+ debug_rt_mutex_init_waiter(waiter);
-+ waiter->task = NULL;
-+ waiter->savestate = savestate;
-+}
++static DEFINE_RAW_SPINLOCK(watchdog_output_lock);
+
- #endif
-Index: linux-3.0/include/linux/plist.h
-===================================================================
---- linux-3.0.orig/include/linux/plist.h
-+++ linux-3.0/include/linux/plist.h
-@@ -77,13 +77,15 @@
+ /* Callback function for perf event subsystem */
+ static void watchdog_overflow_callback(struct perf_event *event, int nmi,
+ struct perf_sample_data *data,
+@@ -234,10 +236,19 @@ static void watchdog_overflow_callback(struct perf_event *event, int nmi,
+ if (__this_cpu_read(hard_watchdog_warn) == true)
+ return;
- #include <linux/kernel.h>
- #include <linux/list.h>
--#include <linux/spinlock_types.h>
+- if (hardlockup_panic)
++ /*
++ * If early-printk is enabled then make sure we do not
++ * lock up in printk() and kill console logging:
++ */
++ printk_kill();
+
-+struct raw_spinlock;
-+struct spinlock;
-
- struct plist_head {
- struct list_head node_list;
- #ifdef CONFIG_DEBUG_PI_LIST
-- raw_spinlock_t *rawlock;
-- spinlock_t *spinlock;
-+ struct raw_spinlock *rawlock;
-+ struct spinlock *spinlock;
- #endif
- };
++ if (hardlockup_panic) {
+ panic("Watchdog detected hard LOCKUP on cpu %d", this_cpu);
+- else
++ } else {
++ raw_spin_lock(&watchdog_output_lock);
+ WARN(1, "Watchdog detected hard LOCKUP on cpu %d", this_cpu);
++ raw_spin_unlock(&watchdog_output_lock);
++ }
-@@ -144,7 +146,7 @@ struct plist_node {
- * @lock: spinlock protecting the list (debugging)
- */
- static inline void
--plist_head_init(struct plist_head *head, spinlock_t *lock)
-+plist_head_init(struct plist_head *head, struct spinlock *lock)
- {
- INIT_LIST_HEAD(&head->node_list);
- #ifdef CONFIG_DEBUG_PI_LIST
-@@ -159,7 +161,7 @@ plist_head_init(struct plist_head *head,
- * @lock: raw_spinlock protecting the list (debugging)
+ __this_cpu_write(hard_watchdog_warn, true);
+ return;
+@@ -320,7 +331,7 @@ static enum hrtimer_restart watchdog_timer_fn(struct hrtimer *hrtimer)
*/
- static inline void
--plist_head_init_raw(struct plist_head *head, raw_spinlock_t *lock)
-+plist_head_init_raw(struct plist_head *head, struct raw_spinlock *lock)
+ static int watchdog(void *unused)
{
- INIT_LIST_HEAD(&head->node_list);
- #ifdef CONFIG_DEBUG_PI_LIST
-Index: linux-3.0/include/linux/rwlock_types.h
-===================================================================
---- linux-3.0.orig/include/linux/rwlock_types.h
-+++ linux-3.0/include/linux/rwlock_types.h
-@@ -1,6 +1,10 @@
- #ifndef __LINUX_RWLOCK_TYPES_H
- #define __LINUX_RWLOCK_TYPES_H
+- static struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
++ struct sched_param param = { .sched_priority = MAX_RT_PRIO-1 };
+ struct hrtimer *hrtimer = &__raw_get_cpu_var(watchdog_hrtimer);
+
+ sched_setscheduler(current, SCHED_FIFO, ¶m);
+@@ -349,7 +360,8 @@ static int watchdog(void *unused)
+ set_current_state(TASK_INTERRUPTIBLE);
+ }
+ __set_current_state(TASK_RUNNING);
+-
++ param.sched_priority = 0;
++ sched_setscheduler(current, SCHED_NORMAL, ¶m);
+ return 0;
+ }
+
+@@ -422,6 +434,7 @@ static void watchdog_prepare_cpu(int cpu)
+ WARN_ON(per_cpu(softlockup_watchdog, cpu));
+ hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ hrtimer->function = watchdog_timer_fn;
++ hrtimer->irqsafe = 1;
+ }
-+#if !defined(__LINUX_SPINLOCK_TYPES_H)
-+# error "Do not include directly, include spinlock_types.h"
-+#endif
-+
- /*
- * include/linux/rwlock_types.h - generic rwlock type definitions
- * and initializers
-@@ -43,6 +47,7 @@ typedef struct {
- RW_DEP_MAP_INIT(lockname) }
- #endif
+ static int watchdog_enable(int cpu)
+diff --git a/kernel/workqueue.c b/kernel/workqueue.c
+index aec02b6..3a9e52a 100644
+--- a/kernel/workqueue.c
++++ b/kernel/workqueue.c
+@@ -41,6 +41,7 @@
+ #include <linux/debug_locks.h>
+ #include <linux/lockdep.h>
+ #include <linux/idr.h>
++#include <linux/delay.h>
--#define DEFINE_RWLOCK(x) rwlock_t x = __RW_LOCK_UNLOCKED(x)
-+#define DEFINE_RWLOCK(name) \
-+ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name)
+ #include "workqueue_sched.h"
- #endif /* __LINUX_RWLOCK_TYPES_H */
-Index: linux-3.0/include/linux/spinlock_types.h
-===================================================================
---- linux-3.0.orig/include/linux/spinlock_types.h
-+++ linux-3.0/include/linux/spinlock_types.h
-@@ -9,80 +9,15 @@
- * Released under the General Public License (GPL).
- */
+@@ -57,20 +58,10 @@ enum {
+ WORKER_DIE = 1 << 1, /* die die die */
+ WORKER_IDLE = 1 << 2, /* is idle */
+ WORKER_PREP = 1 << 3, /* preparing to run works */
+- WORKER_ROGUE = 1 << 4, /* not bound to any cpu */
+- WORKER_REBIND = 1 << 5, /* mom is home, come back */
+- WORKER_CPU_INTENSIVE = 1 << 6, /* cpu intensive */
+- WORKER_UNBOUND = 1 << 7, /* worker is unbound */
++ WORKER_CPU_INTENSIVE = 1 << 4, /* cpu intensive */
++ WORKER_UNBOUND = 1 << 5, /* worker is unbound */
--#if defined(CONFIG_SMP)
--# include <asm/spinlock_types.h>
--#else
--# include <linux/spinlock_types_up.h>
--#endif
--
--#include <linux/lockdep.h>
--
--typedef struct raw_spinlock {
-- arch_spinlock_t raw_lock;
--#ifdef CONFIG_GENERIC_LOCKBREAK
-- unsigned int break_lock;
--#endif
--#ifdef CONFIG_DEBUG_SPINLOCK
-- unsigned int magic, owner_cpu;
-- void *owner;
--#endif
--#ifdef CONFIG_DEBUG_LOCK_ALLOC
-- struct lockdep_map dep_map;
--#endif
--} raw_spinlock_t;
--
--#define SPINLOCK_MAGIC 0xdead4ead
--
--#define SPINLOCK_OWNER_INIT ((void *)-1L)
+- WORKER_NOT_RUNNING = WORKER_PREP | WORKER_ROGUE | WORKER_REBIND |
+- WORKER_CPU_INTENSIVE | WORKER_UNBOUND,
-
--#ifdef CONFIG_DEBUG_LOCK_ALLOC
--# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
--#else
--# define SPIN_DEP_MAP_INIT(lockname)
--#endif
-+#include <linux/spinlock_types_raw.h>
+- /* gcwq->trustee_state */
+- TRUSTEE_START = 0, /* start */
+- TRUSTEE_IN_CHARGE = 1, /* trustee in charge of gcwq */
+- TRUSTEE_BUTCHER = 2, /* butcher workers */
+- TRUSTEE_RELEASE = 3, /* release workers */
+- TRUSTEE_DONE = 4, /* trustee is done */
++ WORKER_NOT_RUNNING = WORKER_PREP | WORKER_CPU_INTENSIVE | WORKER_UNBOUND,
--#ifdef CONFIG_DEBUG_SPINLOCK
--# define SPIN_DEBUG_INIT(lockname) \
-- .magic = SPINLOCK_MAGIC, \
-- .owner_cpu = -1, \
-- .owner = SPINLOCK_OWNER_INIT,
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+# include <linux/spinlock_types_nort.h>
-+# include <linux/rwlock_types.h>
- #else
--# define SPIN_DEBUG_INIT(lockname)
-+# include <linux/rtmutex.h>
-+# include <linux/spinlock_types_rt.h>
-+# include <linux/rwlock_types_rt.h>
- #endif
+ BUSY_WORKER_HASH_ORDER = 6, /* 64 pointers */
+ BUSY_WORKER_HASH_SIZE = 1 << BUSY_WORKER_HASH_ORDER,
+@@ -84,7 +75,6 @@ enum {
+ (min two ticks) */
+ MAYDAY_INTERVAL = HZ / 10, /* and then every 100ms */
+ CREATE_COOLDOWN = HZ, /* time to breath after fail */
+- TRUSTEE_COOLDOWN = HZ / 10, /* for trustee draining */
--#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \
-- { \
-- .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \
-- SPIN_DEBUG_INIT(lockname) \
-- SPIN_DEP_MAP_INIT(lockname) }
--
--#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \
-- (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname)
--
--#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x)
--
--typedef struct spinlock {
-- union {
-- struct raw_spinlock rlock;
--
--#ifdef CONFIG_DEBUG_LOCK_ALLOC
--# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map))
-- struct {
-- u8 __padding[LOCK_PADSIZE];
-- struct lockdep_map dep_map;
-- };
--#endif
-- };
--} spinlock_t;
--
--#define __SPIN_LOCK_INITIALIZER(lockname) \
-- { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } }
--
--#define __SPIN_LOCK_UNLOCKED(lockname) \
-- (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname)
--
--#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x)
--
--#include <linux/rwlock_types.h>
--
- #endif /* __LINUX_SPINLOCK_TYPES_H */
-Index: linux-3.0/include/linux/spinlock_types_nort.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/spinlock_types_nort.h
-@@ -0,0 +1,33 @@
-+#ifndef __LINUX_SPINLOCK_TYPES_NORT_H
-+#define __LINUX_SPINLOCK_TYPES_NORT_H
-+
-+#ifndef __LINUX_SPINLOCK_TYPES_H
-+#error "Do not include directly. Include spinlock_types.h instead"
-+#endif
-+
-+/*
-+ * The non RT version maps spinlocks to raw_spinlocks
-+ */
-+typedef struct spinlock {
-+ union {
-+ struct raw_spinlock rlock;
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+# define LOCK_PADSIZE (offsetof(struct raw_spinlock, dep_map))
-+ struct {
-+ u8 __padding[LOCK_PADSIZE];
-+ struct lockdep_map dep_map;
-+ };
-+#endif
-+ };
-+} spinlock_t;
-+
-+#define __SPIN_LOCK_INITIALIZER(lockname) \
-+ { { .rlock = __RAW_SPIN_LOCK_INITIALIZER(lockname) } }
-+
-+#define __SPIN_LOCK_UNLOCKED(lockname) \
-+ (spinlock_t ) __SPIN_LOCK_INITIALIZER(lockname)
-+
-+#define DEFINE_SPINLOCK(x) spinlock_t x = __SPIN_LOCK_UNLOCKED(x)
-+
-+#endif
-Index: linux-3.0/include/linux/spinlock_types_raw.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/spinlock_types_raw.h
-@@ -0,0 +1,56 @@
-+#ifndef __LINUX_SPINLOCK_TYPES_RAW_H
-+#define __LINUX_SPINLOCK_TYPES_RAW_H
-+
-+#if defined(CONFIG_SMP)
-+# include <asm/spinlock_types.h>
-+#else
-+# include <linux/spinlock_types_up.h>
-+#endif
-+
-+#include <linux/lockdep.h>
-+
-+typedef struct raw_spinlock {
-+ arch_spinlock_t raw_lock;
-+#ifdef CONFIG_GENERIC_LOCKBREAK
-+ unsigned int break_lock;
-+#endif
-+#ifdef CONFIG_DEBUG_SPINLOCK
-+ unsigned int magic, owner_cpu;
-+ void *owner;
-+#endif
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+} raw_spinlock_t;
-+
-+#define SPINLOCK_MAGIC 0xdead4ead
-+
-+#define SPINLOCK_OWNER_INIT ((void *)-1L)
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+# define SPIN_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
-+#else
-+# define SPIN_DEP_MAP_INIT(lockname)
-+#endif
-+
-+#ifdef CONFIG_DEBUG_SPINLOCK
-+# define SPIN_DEBUG_INIT(lockname) \
-+ .magic = SPINLOCK_MAGIC, \
-+ .owner_cpu = -1, \
-+ .owner = SPINLOCK_OWNER_INIT,
-+#else
-+# define SPIN_DEBUG_INIT(lockname)
-+#endif
-+
-+#define __RAW_SPIN_LOCK_INITIALIZER(lockname) \
-+ { \
-+ .raw_lock = __ARCH_SPIN_LOCK_UNLOCKED, \
-+ SPIN_DEBUG_INIT(lockname) \
-+ SPIN_DEP_MAP_INIT(lockname) }
-+
-+#define __RAW_SPIN_LOCK_UNLOCKED(lockname) \
-+ (raw_spinlock_t) __RAW_SPIN_LOCK_INITIALIZER(lockname)
-+
-+#define DEFINE_RAW_SPINLOCK(x) raw_spinlock_t x = __RAW_SPIN_LOCK_UNLOCKED(x)
-+
-+#endif
-Index: linux-3.0/include/linux/rwlock_types_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/rwlock_types_rt.h
-@@ -0,0 +1,33 @@
-+#ifndef __LINUX_RWLOCK_TYPES_RT_H
-+#define __LINUX_RWLOCK_TYPES_RT_H
-+
-+#ifndef __LINUX_SPINLOCK_TYPES_H
-+#error "Do not include directly. Include spinlock_types.h instead"
-+#endif
-+
-+/*
-+ * rwlocks - rtmutex which allows single reader recursion
-+ */
-+typedef struct {
-+ struct rt_mutex lock;
-+ int read_depth;
-+ unsigned int break_lock;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+} rwlock_t;
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+# define RW_DEP_MAP_INIT(lockname) .dep_map = { .name = #lockname }
-+#else
-+# define RW_DEP_MAP_INIT(lockname)
-+#endif
-+
-+#define __RW_LOCK_UNLOCKED(name) \
-+ { .lock = __RT_MUTEX_INITIALIZER_SAVE_STATE(name.lock), \
-+ RW_DEP_MAP_INIT(name) }
+ /*
+ * Rescue workers are used only on emergencies and shared by
+@@ -136,7 +126,7 @@ struct worker {
+ unsigned long last_active; /* L: last active timestamp */
+ unsigned int flags; /* X: flags */
+ int id; /* I: worker id */
+- struct work_struct rebind_work; /* L: rebind worker to cpu */
++ int sleeping; /* None */
+ };
+
+ /*
+@@ -163,10 +153,8 @@ struct global_cwq {
+
+ struct ida worker_ida; /* L: for worker IDs */
+
+- struct task_struct *trustee; /* L: for gcwq shutdown */
+- unsigned int trustee_state; /* L: trustee state */
+- wait_queue_head_t trustee_wait; /* trustee wait */
+ struct worker *first_idle; /* L: first idle worker */
++ wait_queue_head_t idle_wait;
+ } ____cacheline_aligned_in_smp;
+
+ /*
+@@ -657,66 +645,58 @@ static void wake_up_worker(struct global_cwq *gcwq)
+ }
+
+ /**
+- * wq_worker_waking_up - a worker is waking up
+- * @task: task waking up
+- * @cpu: CPU @task is waking up to
+- *
+- * This function is called during try_to_wake_up() when a worker is
+- * being awoken.
++ * wq_worker_running - a worker is running again
++ * @task: task returning from sleep
+ *
+- * CONTEXT:
+- * spin_lock_irq(rq->lock)
++ * This function is called when a worker returns from schedule()
+ */
+-void wq_worker_waking_up(struct task_struct *task, unsigned int cpu)
++void wq_worker_running(struct task_struct *task)
+ {
+ struct worker *worker = kthread_data(task);
+
++ if (!worker->sleeping)
++ return;
+ if (!(worker->flags & WORKER_NOT_RUNNING))
+- atomic_inc(get_gcwq_nr_running(cpu));
++ atomic_inc(get_gcwq_nr_running(smp_processor_id()));
++ worker->sleeping = 0;
+ }
+
+ /**
+ * wq_worker_sleeping - a worker is going to sleep
+ * @task: task going to sleep
+- * @cpu: CPU in question, must be the current CPU number
+- *
+- * This function is called during schedule() when a busy worker is
+- * going to sleep. Worker on the same cpu can be woken up by
+- * returning pointer to its task.
+ *
+- * CONTEXT:
+- * spin_lock_irq(rq->lock)
+- *
+- * RETURNS:
+- * Worker task on @cpu to wake up, %NULL if none.
++ * This function is called from schedule() when a busy worker is
++ * going to sleep.
+ */
+-struct task_struct *wq_worker_sleeping(struct task_struct *task,
+- unsigned int cpu)
++void wq_worker_sleeping(struct task_struct *task)
+ {
+- struct worker *worker = kthread_data(task), *to_wakeup = NULL;
+- struct global_cwq *gcwq = get_gcwq(cpu);
+- atomic_t *nr_running = get_gcwq_nr_running(cpu);
++ struct worker *worker = kthread_data(task);
++ struct global_cwq *gcwq;
++ int cpu;
+
+ if (worker->flags & WORKER_NOT_RUNNING)
+- return NULL;
++ return;
+
+- /* this can only happen on the local cpu */
+- BUG_ON(cpu != raw_smp_processor_id());
++ if (WARN_ON_ONCE(worker->sleeping))
++ return;
+
-+#define DEFINE_RWLOCK(name) \
-+ rwlock_t name __cacheline_aligned_in_smp = __RW_LOCK_UNLOCKED(name)
++ worker->sleeping = 1;
+
++ cpu = smp_processor_id();
++ gcwq = get_gcwq(cpu);
++ spin_lock_irq(&gcwq->lock);
+ /*
+ * The counterpart of the following dec_and_test, implied mb,
+ * worklist not empty test sequence is in insert_work().
+ * Please read comment there.
+- *
+- * NOT_RUNNING is clear. This means that trustee is not in
+- * charge and we're running on the local cpu w/ rq lock held
+- * and preemption disabled, which in turn means that none else
+- * could be manipulating idle_list, so dereferencing idle_list
+- * without gcwq lock is safe.
+ */
+- if (atomic_dec_and_test(nr_running) && !list_empty(&gcwq->worklist))
+- to_wakeup = first_worker(gcwq);
+- return to_wakeup ? to_wakeup->task : NULL;
++ if (atomic_dec_and_test(get_gcwq_nr_running(cpu)) &&
++ !list_empty(&gcwq->worklist)) {
++ worker = first_worker(gcwq);
++ if (worker)
++ wake_up_process(worker->task);
++ }
++ spin_unlock_irq(&gcwq->lock);
+ }
+
+ /**
+@@ -978,13 +958,38 @@ static bool is_chained_work(struct workqueue_struct *wq)
+ return false;
+ }
+
+-static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
+- struct work_struct *work)
++static void ___queue_work(struct workqueue_struct *wq, struct global_cwq *gcwq,
++ struct work_struct *work)
+ {
+- struct global_cwq *gcwq;
+ struct cpu_workqueue_struct *cwq;
+ struct list_head *worklist;
+ unsigned int work_flags;
+
-+#endif
-Index: linux-3.0/include/linux/spinlock_types_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/spinlock_types_rt.h
-@@ -0,0 +1,49 @@
-+#ifndef __LINUX_SPINLOCK_TYPES_RT_H
-+#define __LINUX_SPINLOCK_TYPES_RT_H
++ /* gcwq determined, get cwq and queue */
++ cwq = get_cwq(gcwq->cpu, wq);
++ trace_workqueue_queue_work(gcwq->cpu, cwq, work);
+
-+#ifndef __LINUX_SPINLOCK_TYPES_H
-+#error "Do not include directly. Include spinlock_types.h instead"
-+#endif
++ BUG_ON(!list_empty(&work->entry));
+
-+/*
-+ * PREEMPT_RT: spinlocks - an RT mutex plus lock-break field:
-+ */
-+typedef struct spinlock {
-+ struct rt_mutex lock;
-+ unsigned int break_lock;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+} spinlock_t;
++ cwq->nr_in_flight[cwq->work_color]++;
++ work_flags = work_color_to_flags(cwq->work_color);
+
-+#ifdef CONFIG_DEBUG_RT_MUTEXES
-+# define __RT_SPIN_INITIALIZER(name) \
-+ { \
-+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
-+ .save_state = 1, \
-+ .file = __FILE__, \
-+ .line = __LINE__ , \
-+ }
-+#else
-+# define __RT_SPIN_INITIALIZER(name) \
-+ { \
-+ .wait_lock = __RAW_SPIN_LOCK_UNLOCKED(name.wait_lock), \
-+ .save_state = 1, \
++ if (likely(cwq->nr_active < cwq->max_active)) {
++ trace_workqueue_activate_work(work);
++ cwq->nr_active++;
++ worklist = gcwq_determine_ins_pos(gcwq, cwq);
++ } else {
++ work_flags |= WORK_STRUCT_DELAYED;
++ worklist = &cwq->delayed_works;
+ }
-+#endif
-+
-+/*
-+.wait_list = PLIST_HEAD_INIT_RAW((name).lock.wait_list, (name).lock.wait_lock)
-+*/
+
-+#define __SPIN_LOCK_UNLOCKED(name) \
-+ { .lock = __RT_SPIN_INITIALIZER(name.lock), \
-+ SPIN_DEP_MAP_INIT(name) }
++ insert_work(cwq, work, worklist, work_flags);
++}
+
-+#define __DEFINE_SPINLOCK(name) \
-+ spinlock_t name = __SPIN_LOCK_UNLOCKED(name)
++static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
++ struct work_struct *work)
++{
++ struct global_cwq *gcwq;
+ unsigned long flags;
+
+ debug_work_activate(work);
+@@ -1030,27 +1035,32 @@ static void __queue_work(unsigned int cpu, struct workqueue_struct *wq,
+ spin_lock_irqsave(&gcwq->lock, flags);
+ }
+
+- /* gcwq determined, get cwq and queue */
+- cwq = get_cwq(gcwq->cpu, wq);
+- trace_workqueue_queue_work(cpu, cwq, work);
++ ___queue_work(wq, gcwq, work);
+
+- BUG_ON(!list_empty(&work->entry));
++ spin_unlock_irqrestore(&gcwq->lock, flags);
++}
+
+- cwq->nr_in_flight[cwq->work_color]++;
+- work_flags = work_color_to_flags(cwq->work_color);
++/**
++ * queue_work_on - queue work on specific cpu
++ * @cpu: CPU number to execute work on
++ * @wq: workqueue to use
++ * @work: work to queue
++ *
++ * Returns 0 if @work was already on a queue, non-zero otherwise.
++ *
++ * We queue the work to a specific CPU, the caller must ensure it
++ * can't go away.
++ */
++static int
++__queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
++{
++ int ret = 0;
+
+- if (likely(cwq->nr_active < cwq->max_active)) {
+- trace_workqueue_activate_work(work);
+- cwq->nr_active++;
+- worklist = gcwq_determine_ins_pos(gcwq, cwq);
+- } else {
+- work_flags |= WORK_STRUCT_DELAYED;
+- worklist = &cwq->delayed_works;
++ if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
++ __queue_work(cpu, wq, work);
++ ret = 1;
+ }
+-
+- insert_work(cwq, work, worklist, work_flags);
+-
+- spin_unlock_irqrestore(&gcwq->lock, flags);
++ return ret;
+ }
+
+ /**
+@@ -1067,34 +1077,19 @@ int queue_work(struct workqueue_struct *wq, struct work_struct *work)
+ {
+ int ret;
+
+- ret = queue_work_on(get_cpu(), wq, work);
+- put_cpu();
++ ret = __queue_work_on(get_cpu_light(), wq, work);
++ put_cpu_light();
+
+ return ret;
+ }
+ EXPORT_SYMBOL_GPL(queue_work);
+
+-/**
+- * queue_work_on - queue work on specific cpu
+- * @cpu: CPU number to execute work on
+- * @wq: workqueue to use
+- * @work: work to queue
+- *
+- * Returns 0 if @work was already on a queue, non-zero otherwise.
+- *
+- * We queue the work to a specific CPU, the caller must ensure it
+- * can't go away.
+- */
+ int
+ queue_work_on(int cpu, struct workqueue_struct *wq, struct work_struct *work)
+ {
+- int ret = 0;
++ WARN_ON(wq->flags & WQ_NON_AFFINE);
+
+- if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
+- __queue_work(cpu, wq, work);
+- ret = 1;
+- }
+- return ret;
++ return __queue_work_on(cpu, wq, work);
+ }
+ EXPORT_SYMBOL_GPL(queue_work_on);
+
+@@ -1140,6 +1135,8 @@ int queue_delayed_work_on(int cpu, struct workqueue_struct *wq,
+ struct timer_list *timer = &dwork->timer;
+ struct work_struct *work = &dwork->work;
+
++ WARN_ON((wq->flags & WQ_NON_AFFINE) && cpu != -1);
+
-+#define DEFINE_SPINLOCK(name) \
-+ spinlock_t name __cacheline_aligned_in_smp = __SPIN_LOCK_UNLOCKED(name)
+ if (!test_and_set_bit(WORK_STRUCT_PENDING_BIT, work_data_bits(work))) {
+ unsigned int lcpu;
+
+@@ -1205,12 +1202,13 @@ static void worker_enter_idle(struct worker *worker)
+ /* idle_list is LIFO */
+ list_add(&worker->entry, &gcwq->idle_list);
+
+- if (likely(!(worker->flags & WORKER_ROGUE))) {
+- if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer))
+- mod_timer(&gcwq->idle_timer,
+- jiffies + IDLE_WORKER_TIMEOUT);
+- } else
+- wake_up_all(&gcwq->trustee_wait);
++ if (gcwq->nr_idle == gcwq->nr_workers)
++ wake_up_all(&gcwq->idle_wait);
+
-+#endif
-Index: linux-3.0/include/linux/mutex.h
-===================================================================
---- linux-3.0.orig/include/linux/mutex.h
-+++ linux-3.0/include/linux/mutex.h
-@@ -17,6 +17,17 @@
++ if (too_many_workers(gcwq) && !timer_pending(&gcwq->idle_timer)) {
++ mod_timer(&gcwq->idle_timer,
++ jiffies + IDLE_WORKER_TIMEOUT);
++ }
- #include <asm/atomic.h>
+ /* sanity check nr_running */
+ WARN_ON_ONCE(gcwq->nr_workers == gcwq->nr_idle &&
+@@ -1287,8 +1285,14 @@ __acquires(&gcwq->lock)
+ return false;
+ if (task_cpu(task) == gcwq->cpu &&
+ cpumask_equal(¤t->cpus_allowed,
+- get_cpu_mask(gcwq->cpu)))
++ get_cpu_mask(gcwq->cpu))) {
++ /*
++ * Since we're binding to a particular cpu and need to
++ * stay there for correctness, mark us PF_THREAD_BOUND.
++ */
++ task->flags |= PF_THREAD_BOUND;
+ return true;
++ }
+ spin_unlock_irq(&gcwq->lock);
+
+ /*
+@@ -1302,20 +1306,15 @@ __acquires(&gcwq->lock)
+ }
+ }
+
+-/*
+- * Function for worker->rebind_work used to rebind rogue busy workers
+- * to the associated cpu which is coming back online. This is
+- * scheduled by cpu up but can race with other cpu hotplug operations
+- * and may be executed twice without intervening cpu down.
+- */
+-static void worker_rebind_fn(struct work_struct *work)
++static void worker_unbind_and_unlock(struct worker *worker)
+ {
+- struct worker *worker = container_of(work, struct worker, rebind_work);
+ struct global_cwq *gcwq = worker->gcwq;
++ struct task_struct *task = worker->task;
+
+- if (worker_maybe_bind_and_lock(worker))
+- worker_clr_flags(worker, WORKER_REBIND);
+-
++ /*
++ * Its no longer required we're PF_THREAD_BOUND, the work is done.
++ */
++ task->flags &= ~PF_THREAD_BOUND;
+ spin_unlock_irq(&gcwq->lock);
+ }
+
+@@ -1327,7 +1326,6 @@ static struct worker *alloc_worker(void)
+ if (worker) {
+ INIT_LIST_HEAD(&worker->entry);
+ INIT_LIST_HEAD(&worker->scheduled);
+- INIT_WORK(&worker->rebind_work, worker_rebind_fn);
+ /* on creation a worker is in !idle && prep state */
+ worker->flags = WORKER_PREP;
+ }
+@@ -1382,15 +1380,9 @@ static struct worker *create_worker(struct global_cwq *gcwq, bool bind)
+ if (IS_ERR(worker->task))
+ goto fail;
+
+- /*
+- * A rogue worker will become a regular one if CPU comes
+- * online later on. Make sure every worker has
+- * PF_THREAD_BOUND set.
+- */
+ if (bind && !on_unbound_cpu)
+ kthread_bind(worker->task, gcwq->cpu);
+ else {
+- worker->task->flags |= PF_THREAD_BOUND;
+ if (on_unbound_cpu)
+ worker->flags |= WORKER_UNBOUND;
+ }
+@@ -1667,13 +1659,6 @@ static bool manage_workers(struct worker *worker)
+
+ gcwq->flags &= ~GCWQ_MANAGING_WORKERS;
+
+- /*
+- * The trustee might be waiting to take over the manager
+- * position, tell it we're done.
+- */
+- if (unlikely(gcwq->trustee))
+- wake_up_all(&gcwq->trustee_wait);
+-
+ return ret;
+ }
+
+@@ -2074,7 +2059,7 @@ repeat:
+ if (keep_working(gcwq))
+ wake_up_worker(gcwq);
+
+- spin_unlock_irq(&gcwq->lock);
++ worker_unbind_and_unlock(rescuer);
+ }
+
+ schedule();
+@@ -2970,7 +2955,6 @@ struct workqueue_struct *__alloc_workqueue_key(const char *name,
+ if (IS_ERR(rescuer->task))
+ goto err;
+
+- rescuer->task->flags |= PF_THREAD_BOUND;
+ wake_up_process(rescuer->task);
+ }
+
+@@ -3189,171 +3173,76 @@ EXPORT_SYMBOL_GPL(work_busy);
+ * gcwqs serve mix of short, long and very long running works making
+ * blocked draining impractical.
+ *
+- * This is solved by allowing a gcwq to be detached from CPU, running
+- * it with unbound (rogue) workers and allowing it to be reattached
+- * later if the cpu comes back online. A separate thread is created
+- * to govern a gcwq in such state and is called the trustee of the
+- * gcwq.
+- *
+- * Trustee states and their descriptions.
+- *
+- * START Command state used on startup. On CPU_DOWN_PREPARE, a
+- * new trustee is started with this state.
+- *
+- * IN_CHARGE Once started, trustee will enter this state after
+- * assuming the manager role and making all existing
+- * workers rogue. DOWN_PREPARE waits for trustee to
+- * enter this state. After reaching IN_CHARGE, trustee
+- * tries to execute the pending worklist until it's empty
+- * and the state is set to BUTCHER, or the state is set
+- * to RELEASE.
+- *
+- * BUTCHER Command state which is set by the cpu callback after
+- * the cpu has went down. Once this state is set trustee
+- * knows that there will be no new works on the worklist
+- * and once the worklist is empty it can proceed to
+- * killing idle workers.
+- *
+- * RELEASE Command state which is set by the cpu callback if the
+- * cpu down has been canceled or it has come online
+- * again. After recognizing this state, trustee stops
+- * trying to drain or butcher and clears ROGUE, rebinds
+- * all remaining workers back to the cpu and releases
+- * manager role.
+- *
+- * DONE Trustee will enter this state after BUTCHER or RELEASE
+- * is complete.
+- *
+- * trustee CPU draining
+- * took over down complete
+- * START -----------> IN_CHARGE -----------> BUTCHER -----------> DONE
+- * | | ^
+- * | CPU is back online v return workers |
+- * ----------------> RELEASE --------------
+ */
+
+-/**
+- * trustee_wait_event_timeout - timed event wait for trustee
+- * @cond: condition to wait for
+- * @timeout: timeout in jiffies
+- *
+- * wait_event_timeout() for trustee to use. Handles locking and
+- * checks for RELEASE request.
+- *
+- * CONTEXT:
+- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
+- * multiple times. To be used by trustee.
+- *
+- * RETURNS:
+- * Positive indicating left time if @cond is satisfied, 0 if timed
+- * out, -1 if canceled.
+- */
+-#define trustee_wait_event_timeout(cond, timeout) ({ \
+- long __ret = (timeout); \
+- while (!((cond) || (gcwq->trustee_state == TRUSTEE_RELEASE)) && \
+- __ret) { \
+- spin_unlock_irq(&gcwq->lock); \
+- __wait_event_timeout(gcwq->trustee_wait, (cond) || \
+- (gcwq->trustee_state == TRUSTEE_RELEASE), \
+- __ret); \
+- spin_lock_irq(&gcwq->lock); \
+- } \
+- gcwq->trustee_state == TRUSTEE_RELEASE ? -1 : (__ret); \
+-})
++static int __devinit workqueue_cpu_up_callback(struct notifier_block *nfb,
++ unsigned long action,
++ void *hcpu)
++{
++ unsigned int cpu = (unsigned long)hcpu;
++ struct global_cwq *gcwq = get_gcwq(cpu);
++ struct worker *uninitialized_var(new_worker);
++ unsigned long flags;
+
+-/**
+- * trustee_wait_event - event wait for trustee
+- * @cond: condition to wait for
+- *
+- * wait_event() for trustee to use. Automatically handles locking and
+- * checks for CANCEL request.
+- *
+- * CONTEXT:
+- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
+- * multiple times. To be used by trustee.
+- *
+- * RETURNS:
+- * 0 if @cond is satisfied, -1 if canceled.
+- */
+-#define trustee_wait_event(cond) ({ \
+- long __ret1; \
+- __ret1 = trustee_wait_event_timeout(cond, MAX_SCHEDULE_TIMEOUT);\
+- __ret1 < 0 ? -1 : 0; \
+-})
++ action &= ~CPU_TASKS_FROZEN;
+
+-static int __cpuinit trustee_thread(void *__gcwq)
+-{
+- struct global_cwq *gcwq = __gcwq;
+- struct worker *worker;
+- struct work_struct *work;
+- struct hlist_node *pos;
+- long rc;
+- int i;
++ switch (action) {
++ case CPU_UP_PREPARE:
++ BUG_ON(gcwq->first_idle);
++ new_worker = create_worker(gcwq, false);
++ if (!new_worker)
++ return NOTIFY_BAD;
++ case CPU_UP_CANCELED:
++ case CPU_ONLINE:
++ break;
++ default:
++ return notifier_from_errno(0);
++ }
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \
-+ , .dep_map = { .name = #lockname }
-+#else
-+# define __DEP_MAP_MUTEX_INITIALIZER(lockname)
-+#endif
-+
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+#include <linux/mutex_rt.h>
-+#else
-+
- /*
- * Simple, straightforward mutexes with strict semantics:
- *
-@@ -95,13 +106,6 @@ do { \
- # define mutex_destroy(mutex) do { } while (0)
- #endif
+- BUG_ON(gcwq->cpu != smp_processor_id());
++ /* some are called w/ irq disabled, don't disturb irq status */
++ spin_lock_irqsave(&gcwq->lock, flags);
--#ifdef CONFIG_DEBUG_LOCK_ALLOC
--# define __DEP_MAP_MUTEX_INITIALIZER(lockname) \
-- , .dep_map = { .name = #lockname }
--#else
--# define __DEP_MAP_MUTEX_INITIALIZER(lockname)
--#endif
--
- #define __MUTEX_INITIALIZER(lockname) \
- { .count = ATOMIC_INIT(1) \
- , .wait_lock = __SPIN_LOCK_UNLOCKED(lockname.wait_lock) \
-@@ -167,6 +171,9 @@ extern int __must_check mutex_lock_killa
- */
- extern int mutex_trylock(struct mutex *lock);
- extern void mutex_unlock(struct mutex *lock);
-+
-+#endif /* !PREEMPT_RT_FULL */
-+
- extern int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock);
+- spin_lock_irq(&gcwq->lock);
+- /*
+- * Claim the manager position and make all workers rogue.
+- * Trustee must be bound to the target cpu and can't be
+- * cancelled.
+- */
+- BUG_ON(gcwq->cpu != smp_processor_id());
+- rc = trustee_wait_event(!(gcwq->flags & GCWQ_MANAGING_WORKERS));
+- BUG_ON(rc < 0);
++ switch (action) {
++ case CPU_UP_PREPARE:
++ BUG_ON(gcwq->first_idle);
++ gcwq->first_idle = new_worker;
++ break;
- #ifndef CONFIG_HAVE_ARCH_MUTEX_CPU_RELAX
-Index: linux-3.0/include/linux/mutex_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/mutex_rt.h
-@@ -0,0 +1,84 @@
-+#ifndef __LINUX_MUTEX_RT_H
-+#define __LINUX_MUTEX_RT_H
-+
-+#ifndef __LINUX_MUTEX_H
-+#error "Please include mutex.h"
-+#endif
-+
-+#include <linux/rtmutex.h>
-+
-+/* FIXME: Just for __lockfunc */
-+#include <linux/spinlock.h>
-+
-+struct mutex {
-+ struct rt_mutex lock;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+};
-+
-+#define __MUTEX_INITIALIZER(mutexname) \
-+ { \
-+ .lock = __RT_MUTEX_INITIALIZER(mutexname.lock) \
-+ __DEP_MAP_MUTEX_INITIALIZER(mutexname) \
-+ }
-+
-+#define DEFINE_MUTEX(mutexname) \
-+ struct mutex mutexname = __MUTEX_INITIALIZER(mutexname)
-+
-+extern void __mutex_do_init(struct mutex *lock, const char *name, struct lock_class_key *key);
-+extern void __lockfunc _mutex_lock(struct mutex *lock);
-+extern int __lockfunc _mutex_lock_interruptible(struct mutex *lock);
-+extern int __lockfunc _mutex_lock_killable(struct mutex *lock);
-+extern void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass);
-+extern void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest_lock);
-+extern int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass);
-+extern int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass);
-+extern int __lockfunc _mutex_trylock(struct mutex *lock);
-+extern void __lockfunc _mutex_unlock(struct mutex *lock);
-+
-+#define mutex_is_locked(l) rt_mutex_is_locked(&(l)->lock)
-+#define mutex_lock(l) _mutex_lock(l)
-+#define mutex_lock_interruptible(l) _mutex_lock_interruptible(l)
-+#define mutex_lock_killable(l) _mutex_lock_killable(l)
-+#define mutex_trylock(l) _mutex_trylock(l)
-+#define mutex_unlock(l) _mutex_unlock(l)
-+#define mutex_destroy(l) rt_mutex_destroy(&(l)->lock)
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+# define mutex_lock_nested(l, s) _mutex_lock_nested(l, s)
-+# define mutex_lock_interruptible_nested(l, s) \
-+ _mutex_lock_interruptible_nested(l, s)
-+# define mutex_lock_killable_nested(l, s) \
-+ _mutex_lock_killable_nested(l, s)
-+
-+# define mutex_lock_nest_lock(lock, nest_lock) \
-+do { \
-+ typecheck(struct lockdep_map *, &(nest_lock)->dep_map); \
-+ _mutex_lock_nest_lock(lock, &(nest_lock)->dep_map); \
-+} while (0)
-+
-+#else
-+# define mutex_lock_nested(l, s) _mutex_lock(l)
-+# define mutex_lock_interruptible_nested(l, s) \
-+ _mutex_lock_interruptible(l)
-+# define mutex_lock_killable_nested(l, s) \
-+ _mutex_lock_killable(l)
-+# define mutex_lock_nest_lock(lock, nest_lock) mutex_lock(lock)
-+#endif
-+
-+# define mutex_init(mutex) \
-+do { \
-+ static struct lock_class_key __key; \
-+ \
-+ rt_mutex_init(&(mutex)->lock); \
-+ __mutex_do_init((mutex), #mutex, &__key); \
-+} while (0)
-+
-+# define __mutex_init(mutex, name, key) \
-+do { \
-+ rt_mutex_init(&(mutex)->lock); \
-+ __mutex_do_init((mutex), name, key); \
-+} while (0)
-+
-+#endif
-Index: linux-3.0/include/linux/rwsem_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/rwsem_rt.h
-@@ -0,0 +1,105 @@
-+#ifndef _LINUX_RWSEM_RT_H
-+#define _LINUX_RWSEM_RT_H
-+
-+#ifndef _LINUX_RWSEM_H
-+#error "Include rwsem.h"
-+#endif
-+
-+/*
-+ * RW-semaphores are a spinlock plus a reader-depth count.
-+ *
-+ * Note that the semantics are different from the usual
-+ * Linux rw-sems, in PREEMPT_RT mode we do not allow
-+ * multiple readers to hold the lock at once, we only allow
-+ * a read-lock owner to read-lock recursively. This is
-+ * better for latency, makes the implementation inherently
-+ * fair and makes it simpler as well.
-+ */
-+
-+#include <linux/rtmutex.h>
-+
-+struct rw_semaphore {
-+ struct rt_mutex lock;
-+ int read_depth;
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ struct lockdep_map dep_map;
-+#endif
-+};
-+
-+#define __RWSEM_INITIALIZER(name) \
-+ { .lock = __RT_MUTEX_INITIALIZER(name.lock), \
-+ RW_DEP_MAP_INIT(name) }
-+
-+#define DECLARE_RWSEM(lockname) \
-+ struct rw_semaphore lockname = __RWSEM_INITIALIZER(lockname)
-+
-+extern void __rt_rwsem_init(struct rw_semaphore *rwsem, char *name,
-+ struct lock_class_key *key);
-+
-+# define rt_init_rwsem(sem) \
-+do { \
-+ static struct lock_class_key __key; \
-+ \
-+ rt_mutex_init(&(sem)->lock); \
-+ __rt_rwsem_init((sem), #sem, &__key); \
-+} while (0)
-+
-+extern void rt_down_write(struct rw_semaphore *rwsem);
-+extern void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass);
-+extern void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass);
-+extern void rt_down_read(struct rw_semaphore *rwsem);
-+extern int rt_down_write_trylock(struct rw_semaphore *rwsem);
-+extern int rt_down_read_trylock(struct rw_semaphore *rwsem);
-+extern void rt_up_read(struct rw_semaphore *rwsem);
-+extern void rt_up_write(struct rw_semaphore *rwsem);
-+extern void rt_downgrade_write(struct rw_semaphore *rwsem);
-+
-+#define init_rwsem(sem) rt_init_rwsem(sem)
-+#define rwsem_is_locked(s) rt_mutex_is_locked(&(s)->lock)
-+
-+static inline void down_read(struct rw_semaphore *sem)
-+{
-+ rt_down_read(sem);
-+}
-+
-+static inline int down_read_trylock(struct rw_semaphore *sem)
-+{
-+ return rt_down_read_trylock(sem);
-+}
-+
-+static inline void down_write(struct rw_semaphore *sem)
-+{
-+ rt_down_write(sem);
-+}
-+
-+static inline int down_write_trylock(struct rw_semaphore *sem)
-+{
-+ return rt_down_write_trylock(sem);
-+}
-+
-+static inline void up_read(struct rw_semaphore *sem)
-+{
-+ rt_up_read(sem);
-+}
-+
-+static inline void up_write(struct rw_semaphore *sem)
-+{
-+ rt_up_write(sem);
-+}
-+
-+static inline void downgrade_write(struct rw_semaphore *sem)
-+{
-+ rt_downgrade_write(sem);
-+}
-+
-+static inline void down_read_nested(struct rw_semaphore *sem, int subclass)
-+{
-+ return rt_down_read_nested(sem, subclass);
+- gcwq->flags |= GCWQ_MANAGING_WORKERS;
++ case CPU_UP_CANCELED:
++ destroy_worker(gcwq->first_idle);
++ gcwq->first_idle = NULL;
++ break;
+
+- list_for_each_entry(worker, &gcwq->idle_list, entry)
+- worker->flags |= WORKER_ROGUE;
++ case CPU_ONLINE:
++ spin_unlock_irq(&gcwq->lock);
++ kthread_bind(gcwq->first_idle->task, cpu);
++ spin_lock_irq(&gcwq->lock);
++ gcwq->flags |= GCWQ_MANAGE_WORKERS;
++ start_worker(gcwq->first_idle);
++ gcwq->first_idle = NULL;
++ break;
++ }
+
+- for_each_busy_worker(worker, i, pos, gcwq)
+- worker->flags |= WORKER_ROGUE;
++ spin_unlock_irqrestore(&gcwq->lock, flags);
+
+- /*
+- * Call schedule() so that we cross rq->lock and thus can
+- * guarantee sched callbacks see the rogue flag. This is
+- * necessary as scheduler callbacks may be invoked from other
+- * cpus.
+- */
+- spin_unlock_irq(&gcwq->lock);
+- schedule();
+- spin_lock_irq(&gcwq->lock);
++ return notifier_from_errno(0);
+}
-+
-+static inline void down_write_nested(struct rw_semaphore *sem, int subclass)
+
+- /*
+- * Sched callbacks are disabled now. Zap nr_running. After
+- * this, nr_running stays zero and need_more_worker() and
+- * keep_working() are always true as long as the worklist is
+- * not empty.
+- */
+- atomic_set(get_gcwq_nr_running(gcwq->cpu), 0);
++static void flush_gcwq(struct global_cwq *gcwq)
+{
-+ rt_down_write_nested(sem, subclass);
-+}
-+
-+#endif
-Index: linux-3.0/include/linux/rwlock_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/rwlock_rt.h
-@@ -0,0 +1,123 @@
-+#ifndef __LINUX_RWLOCK_RT_H
-+#define __LINUX_RWLOCK_RT_H
-+
-+#ifndef __LINUX_SPINLOCK_H
-+#error Do not include directly. Use spinlock.h
-+#endif
-+
-+#define rwlock_init(rwl) \
-+do { \
-+ static struct lock_class_key __key; \
-+ \
-+ rt_mutex_init(&(rwl)->lock); \
-+ __rt_rwlock_init(rwl, #rwl, &__key); \
-+} while (0)
-+
-+extern void __lockfunc rt_write_lock(rwlock_t *rwlock);
-+extern void __lockfunc rt_read_lock(rwlock_t *rwlock);
-+extern int __lockfunc rt_write_trylock(rwlock_t *rwlock);
-+extern int __lockfunc rt_write_trylock_irqsave(rwlock_t *trylock, unsigned long *flags);
-+extern int __lockfunc rt_read_trylock(rwlock_t *rwlock);
-+extern void __lockfunc rt_write_unlock(rwlock_t *rwlock);
-+extern void __lockfunc rt_read_unlock(rwlock_t *rwlock);
-+extern unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock);
-+extern unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock);
-+extern void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key);
-+
-+#define read_trylock(lock) __cond_lock(lock, rt_read_trylock(lock))
-+#define write_trylock(lock) __cond_lock(lock, rt_write_trylock(lock))
-+
-+#define write_trylock_irqsave(lock, flags) \
-+ __cond_lock(lock, rt_write_trylock_irqsave(lock, &flags))
-+
-+#define read_lock_irqsave(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ migrate_disable(); \
-+ flags = rt_read_lock_irqsave(lock); \
-+ } while (0)
-+
-+#define write_lock_irqsave(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ migrate_disable(); \
-+ flags = rt_write_lock_irqsave(lock); \
-+ } while (0)
-+
-+#define read_lock(lock) \
-+ do { \
-+ migrate_disable(); \
-+ rt_read_lock(lock); \
-+ } while (0)
-+
-+#define read_lock_bh(lock) \
-+ do { \
-+ local_bh_disable(); \
-+ migrate_disable(); \
-+ rt_read_lock(lock); \
-+ } while (0)
-+
-+#define read_lock_irq(lock) read_lock(lock)
-+
-+#define write_lock(lock) \
-+ do { \
-+ migrate_disable(); \
-+ rt_write_lock(lock); \
-+ } while (0)
-+
-+#define write_lock_bh(lock) \
-+ do { \
-+ local_bh_disable(); \
-+ migrate_disable(); \
-+ rt_write_lock(lock); \
-+ } while (0)
-+
-+#define write_lock_irq(lock) write_lock(lock)
-+
-+#define read_unlock(lock) \
-+ do { \
-+ rt_read_unlock(lock); \
-+ migrate_enable(); \
-+ } while (0)
-+
-+#define read_unlock_bh(lock) \
-+ do { \
-+ rt_read_unlock(lock); \
-+ migrate_enable(); \
-+ local_bh_enable(); \
-+ } while (0)
-+
-+#define read_unlock_irq(lock) read_unlock(lock)
-+
-+#define write_unlock(lock) \
-+ do { \
-+ rt_write_unlock(lock); \
-+ migrate_enable(); \
-+ } while (0)
++ struct work_struct *work, *nw;
++ struct worker *worker, *n;
++ LIST_HEAD(non_affine_works);
+
+- spin_unlock_irq(&gcwq->lock);
+- del_timer_sync(&gcwq->idle_timer);
+ spin_lock_irq(&gcwq->lock);
++ list_for_each_entry_safe(work, nw, &gcwq->worklist, entry) {
++ struct workqueue_struct *wq = get_work_cwq(work)->wq;
+
+- /*
+- * We're now in charge. Notify and proceed to drain. We need
+- * to keep the gcwq running during the whole CPU down
+- * procedure as other cpu hotunplug callbacks may need to
+- * flush currently running tasks.
+- */
+- gcwq->trustee_state = TRUSTEE_IN_CHARGE;
+- wake_up_all(&gcwq->trustee_wait);
++ if (wq->flags & WQ_NON_AFFINE)
++ list_move(&work->entry, &non_affine_works);
++ }
+
+- /*
+- * The original cpu is in the process of dying and may go away
+- * anytime now. When that happens, we and all workers would
+- * be migrated to other cpus. Try draining any left work. We
+- * want to get it over with ASAP - spam rescuers, wake up as
+- * many idlers as necessary and create new ones till the
+- * worklist is empty. Note that if the gcwq is frozen, there
+- * may be frozen works in freezable cwqs. Don't declare
+- * completion while frozen.
+- */
+- while (gcwq->nr_workers != gcwq->nr_idle ||
+- gcwq->flags & GCWQ_FREEZING ||
+- gcwq->trustee_state == TRUSTEE_IN_CHARGE) {
++ while (!list_empty(&gcwq->worklist)) {
+ int nr_works = 0;
+
+ list_for_each_entry(work, &gcwq->worklist, entry) {
+@@ -3367,189 +3256,54 @@ static int __cpuinit trustee_thread(void *__gcwq)
+ wake_up_process(worker->task);
+ }
+
++ spin_unlock_irq(&gcwq->lock);
+
-+#define write_unlock_bh(lock) \
-+ do { \
-+ rt_write_unlock(lock); \
-+ migrate_enable(); \
-+ local_bh_enable(); \
-+ } while (0)
+ if (need_to_create_worker(gcwq)) {
+- spin_unlock_irq(&gcwq->lock);
+- worker = create_worker(gcwq, false);
+- spin_lock_irq(&gcwq->lock);
+- if (worker) {
+- worker->flags |= WORKER_ROGUE;
++ worker = create_worker(gcwq, true);
++ if (worker)
+ start_worker(worker);
+- }
+ }
+
+- /* give a breather */
+- if (trustee_wait_event_timeout(false, TRUSTEE_COOLDOWN) < 0)
+- break;
+- }
+-
+- /*
+- * Either all works have been scheduled and cpu is down, or
+- * cpu down has already been canceled. Wait for and butcher
+- * all workers till we're canceled.
+- */
+- do {
+- rc = trustee_wait_event(!list_empty(&gcwq->idle_list));
+- while (!list_empty(&gcwq->idle_list))
+- destroy_worker(list_first_entry(&gcwq->idle_list,
+- struct worker, entry));
+- } while (gcwq->nr_workers && rc >= 0);
++ wait_event_timeout(gcwq->idle_wait,
++ gcwq->nr_idle == gcwq->nr_workers, HZ/10);
+
+- /*
+- * At this point, either draining has completed and no worker
+- * is left, or cpu down has been canceled or the cpu is being
+- * brought back up. There shouldn't be any idle one left.
+- * Tell the remaining busy ones to rebind once it finishes the
+- * currently scheduled works by scheduling the rebind_work.
+- */
+- WARN_ON(!list_empty(&gcwq->idle_list));
+-
+- for_each_busy_worker(worker, i, pos, gcwq) {
+- struct work_struct *rebind_work = &worker->rebind_work;
+-
+- /*
+- * Rebind_work may race with future cpu hotplug
+- * operations. Use a separate flag to mark that
+- * rebinding is scheduled.
+- */
+- worker->flags |= WORKER_REBIND;
+- worker->flags &= ~WORKER_ROGUE;
++ spin_lock_irq(&gcwq->lock);
++ }
+
+- /* queue rebind_work, wq doesn't matter, use the default one */
+- if (test_and_set_bit(WORK_STRUCT_PENDING_BIT,
+- work_data_bits(rebind_work)))
+- continue;
++ WARN_ON(gcwq->nr_workers != gcwq->nr_idle);
+
+- debug_work_activate(rebind_work);
+- insert_work(get_cwq(gcwq->cpu, system_wq), rebind_work,
+- worker->scheduled.next,
+- work_color_to_flags(WORK_NO_COLOR));
+- }
++ list_for_each_entry_safe(worker, n, &gcwq->idle_list, entry)
++ destroy_worker(worker);
+
+- /* relinquish manager role */
+- gcwq->flags &= ~GCWQ_MANAGING_WORKERS;
++ WARN_ON(gcwq->nr_workers || gcwq->nr_idle);
+
+- /* notify completion */
+- gcwq->trustee = NULL;
+- gcwq->trustee_state = TRUSTEE_DONE;
+- wake_up_all(&gcwq->trustee_wait);
+ spin_unlock_irq(&gcwq->lock);
+- return 0;
+-}
+
+-/**
+- * wait_trustee_state - wait for trustee to enter the specified state
+- * @gcwq: gcwq the trustee of interest belongs to
+- * @state: target state to wait for
+- *
+- * Wait for the trustee to reach @state. DONE is already matched.
+- *
+- * CONTEXT:
+- * spin_lock_irq(gcwq->lock) which may be released and regrabbed
+- * multiple times. To be used by cpu_callback.
+- */
+-static void __cpuinit wait_trustee_state(struct global_cwq *gcwq, int state)
+-__releases(&gcwq->lock)
+-__acquires(&gcwq->lock)
+-{
+- if (!(gcwq->trustee_state == state ||
+- gcwq->trustee_state == TRUSTEE_DONE)) {
+- spin_unlock_irq(&gcwq->lock);
+- __wait_event(gcwq->trustee_wait,
+- gcwq->trustee_state == state ||
+- gcwq->trustee_state == TRUSTEE_DONE);
+- spin_lock_irq(&gcwq->lock);
++ gcwq = get_gcwq(get_cpu_light());
++ spin_lock_irq(&gcwq->lock);
++ list_for_each_entry_safe(work, nw, &non_affine_works, entry) {
++ list_del_init(&work->entry);
++ ___queue_work(get_work_cwq(work)->wq, gcwq, work);
+ }
++ spin_unlock_irq(&gcwq->lock);
++ put_cpu_light();
+ }
+
+-static int __devinit workqueue_cpu_callback(struct notifier_block *nfb,
++static int __devinit workqueue_cpu_down_callback(struct notifier_block *nfb,
+ unsigned long action,
+ void *hcpu)
+ {
+ unsigned int cpu = (unsigned long)hcpu;
+ struct global_cwq *gcwq = get_gcwq(cpu);
+- struct task_struct *new_trustee = NULL;
+- struct worker *uninitialized_var(new_worker);
+- unsigned long flags;
+
+ action &= ~CPU_TASKS_FROZEN;
+
+- switch (action) {
+- case CPU_DOWN_PREPARE:
+- new_trustee = kthread_create(trustee_thread, gcwq,
+- "workqueue_trustee/%d\n", cpu);
+- if (IS_ERR(new_trustee))
+- return notifier_from_errno(PTR_ERR(new_trustee));
+- kthread_bind(new_trustee, cpu);
+- /* fall through */
+- case CPU_UP_PREPARE:
+- BUG_ON(gcwq->first_idle);
+- new_worker = create_worker(gcwq, false);
+- if (!new_worker) {
+- if (new_trustee)
+- kthread_stop(new_trustee);
+- return NOTIFY_BAD;
+- }
+- }
++ switch (action) {
++ case CPU_DOWN_PREPARE:
++ flush_gcwq(gcwq);
++ break;
++ }
+
+- /* some are called w/ irq disabled, don't disturb irq status */
+- spin_lock_irqsave(&gcwq->lock, flags);
+-
+- switch (action) {
+- case CPU_DOWN_PREPARE:
+- /* initialize trustee and tell it to acquire the gcwq */
+- BUG_ON(gcwq->trustee || gcwq->trustee_state != TRUSTEE_DONE);
+- gcwq->trustee = new_trustee;
+- gcwq->trustee_state = TRUSTEE_START;
+- wake_up_process(gcwq->trustee);
+- wait_trustee_state(gcwq, TRUSTEE_IN_CHARGE);
+- /* fall through */
+- case CPU_UP_PREPARE:
+- BUG_ON(gcwq->first_idle);
+- gcwq->first_idle = new_worker;
+- break;
+-
+- case CPU_DYING:
+- /*
+- * Before this, the trustee and all workers except for
+- * the ones which are still executing works from
+- * before the last CPU down must be on the cpu. After
+- * this, they'll all be diasporas.
+- */
+- gcwq->flags |= GCWQ_DISASSOCIATED;
+- break;
+-
+- case CPU_POST_DEAD:
+- gcwq->trustee_state = TRUSTEE_BUTCHER;
+- /* fall through */
+- case CPU_UP_CANCELED:
+- destroy_worker(gcwq->first_idle);
+- gcwq->first_idle = NULL;
+- break;
+-
+- case CPU_DOWN_FAILED:
+- case CPU_ONLINE:
+- gcwq->flags &= ~GCWQ_DISASSOCIATED;
+- if (gcwq->trustee_state != TRUSTEE_DONE) {
+- gcwq->trustee_state = TRUSTEE_RELEASE;
+- wake_up_process(gcwq->trustee);
+- wait_trustee_state(gcwq, TRUSTEE_DONE);
+- }
+-
+- /*
+- * Trustee is done and there might be no worker left.
+- * Put the first_idle in and request a real manager to
+- * take a look.
+- */
+- spin_unlock_irq(&gcwq->lock);
+- kthread_bind(gcwq->first_idle->task, cpu);
+- spin_lock_irq(&gcwq->lock);
+- gcwq->flags |= GCWQ_MANAGE_WORKERS;
+- start_worker(gcwq->first_idle);
+- gcwq->first_idle = NULL;
+- break;
+- }
+-
+- spin_unlock_irqrestore(&gcwq->lock, flags);
+
+ return notifier_from_errno(0);
+ }
+@@ -3747,7 +3501,8 @@ static int __init init_workqueues(void)
+ unsigned int cpu;
+ int i;
+
+- cpu_notifier(workqueue_cpu_callback, CPU_PRI_WORKQUEUE);
++ cpu_notifier(workqueue_cpu_up_callback, CPU_PRI_WORKQUEUE_ACTIVE);
++ hotcpu_notifier(workqueue_cpu_down_callback, CPU_PRI_WORKQUEUE_INACTIVE);
+
+ /* initialize gcwqs */
+ for_each_gcwq_cpu(cpu) {
+@@ -3770,9 +3525,7 @@ static int __init init_workqueues(void)
+ (unsigned long)gcwq);
+
+ ida_init(&gcwq->worker_ida);
+-
+- gcwq->trustee_state = TRUSTEE_DONE;
+- init_waitqueue_head(&gcwq->trustee_wait);
++ init_waitqueue_head(&gcwq->idle_wait);
+ }
+
+ /* create the initial worker */
+diff --git a/kernel/workqueue_sched.h b/kernel/workqueue_sched.h
+index 2d10fc9..3bf73e2 100644
+--- a/kernel/workqueue_sched.h
++++ b/kernel/workqueue_sched.h
+@@ -4,6 +4,5 @@
+ * Scheduler hooks for concurrency managed workqueue. Only to be
+ * included from sched.c and workqueue.c.
+ */
+-void wq_worker_waking_up(struct task_struct *task, unsigned int cpu);
+-struct task_struct *wq_worker_sleeping(struct task_struct *task,
+- unsigned int cpu);
++void wq_worker_running(struct task_struct *task);
++void wq_worker_sleeping(struct task_struct *task);
+diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
+--- a/lib/Kconfig.debug
++++ b/lib/Kconfig.debug
+@@ -62,6 +62,28 @@ config MAGIC_SYSRQ
+ keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
+ unless you really know what this hack does.
+
++config MAGIC_SYSRQ_FORCE_PRINTK
++ bool "Force printk from Magic SysRq"
++ depends on MAGIC_SYSRQ && PREEMPT_RT_FULL
++ default n
++ help
++ Allow the output from Magic SysRq to be output immediately, even if
++ this causes large latencies. This can cause performance problems
++ for real-time processes.
+
-+#define write_unlock_irq(lock) write_unlock(lock)
++ If PREEMPT_RT_FULL, printk() will not try to acquire the console lock
++ when interrupts or preemption are disabled. If the console lock is
++ not acquired the printk() output will be buffered, but will not be
++ output immediately. Some drivers call into the Magic SysRq code
++ with interrupts or preemption disabled, so the output of Magic SysRq
++ will be buffered instead of printing immediately if this option is
++ not selected.
+
-+#define read_unlock_irqrestore(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ (void) flags; \
-+ rt_read_unlock(lock); \
-+ migrate_enable(); \
-+ } while (0)
++ Even with this option selected, Magic SysRq output will be delayed
++ if the attempt to acquire the console lock fails.
+
-+#define write_unlock_irqrestore(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ (void) flags; \
-+ rt_write_unlock(lock); \
-+ migrate_enable(); \
-+ } while (0)
++ Don't say Y unless you really know what this hack does.
+
-+#endif
-Index: linux-3.0/include/linux/spinlock.h
-===================================================================
---- linux-3.0.orig/include/linux/spinlock.h
-+++ linux-3.0/include/linux/spinlock.h
-@@ -254,7 +254,11 @@ static inline void do_raw_spin_unlock(ra
- #define raw_spin_can_lock(lock) (!raw_spin_is_locked(lock))
+ config MAGIC_SYSRQ_DEFAULT_MASK
+ hex "Default mask for Magic SysRq keys on the console"
+ depends on MAGIC_SYSRQ
+@@ -151,7 +173,7 @@ config DEBUG_KERNEL
+
+ config DEBUG_SHIRQ
+ bool "Debug shared IRQ handlers"
+- depends on DEBUG_KERNEL && GENERIC_HARDIRQS
++ depends on DEBUG_KERNEL && GENERIC_HARDIRQS && !PREEMPT_RT_BASE
+ help
+ Enable this to generate a spurious interrupt as soon as a shared
+ interrupt handler is registered, and just before one is deregistered.
+diff --git a/lib/atomic64.c b/lib/atomic64.c
+index a21c12b..f9c5b29 100644
+--- a/lib/atomic64.c
++++ b/lib/atomic64.c
+@@ -29,11 +29,11 @@
+ * Ensure each lock is in a separate cacheline.
+ */
+ static union {
+- spinlock_t lock;
++ raw_spinlock_t lock;
+ char pad[L1_CACHE_BYTES];
+ } atomic64_lock[NR_LOCKS] __cacheline_aligned_in_smp;
+
+-static inline spinlock_t *lock_addr(const atomic64_t *v)
++static inline raw_spinlock_t *lock_addr(const atomic64_t *v)
+ {
+ unsigned long addr = (unsigned long) v;
+
+@@ -45,12 +45,12 @@ static inline spinlock_t *lock_addr(const atomic64_t *v)
+ long long atomic64_read(const atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
+ }
+ EXPORT_SYMBOL(atomic64_read);
+@@ -58,34 +58,34 @@ EXPORT_SYMBOL(atomic64_read);
+ void atomic64_set(atomic64_t *v, long long i)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
- /* Include rwlock functions */
--#include <linux/rwlock.h>
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# include <linux/rwlock_rt.h>
-+#else
-+# include <linux/rwlock.h>
-+#endif
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ v->counter = i;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ }
+ EXPORT_SYMBOL(atomic64_set);
- /*
- * Pull the _spin_*()/_read_*()/_write_*() functions/declarations:
-@@ -265,6 +269,10 @@ static inline void do_raw_spin_unlock(ra
- # include <linux/spinlock_api_up.h>
- #endif
+ void atomic64_add(long long a, atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# include <linux/spinlock_rt.h>
-+#else /* PREEMPT_RT_FULL */
-+
- /*
- * Map the spin_lock functions to the raw variants for PREEMPT_RT=n
- */
-@@ -397,4 +405,6 @@ extern int _atomic_dec_and_lock(atomic_t
- #define atomic_dec_and_lock(atomic, lock) \
- __cond_lock(lock, _atomic_dec_and_lock(atomic, lock))
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ v->counter += a;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ }
+ EXPORT_SYMBOL(atomic64_add);
-+#endif /* !PREEMPT_RT_FULL */
-+
- #endif /* __LINUX_SPINLOCK_H */
-Index: linux-3.0/include/linux/spinlock_api_smp.h
-===================================================================
---- linux-3.0.orig/include/linux/spinlock_api_smp.h
-+++ linux-3.0/include/linux/spinlock_api_smp.h
-@@ -191,6 +191,8 @@ static inline int __raw_spin_trylock_bh(
- return 0;
+ long long atomic64_add_return(long long a, atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter += a;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
}
+ EXPORT_SYMBOL(atomic64_add_return);
+@@ -93,23 +93,23 @@ EXPORT_SYMBOL(atomic64_add_return);
+ void atomic64_sub(long long a, atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
--#include <linux/rwlock_api_smp.h>
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+# include <linux/rwlock_api_smp.h>
-+#endif
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ v->counter -= a;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ }
+ EXPORT_SYMBOL(atomic64_sub);
- #endif /* __LINUX_SPINLOCK_API_SMP_H */
-Index: linux-3.0/include/linux/spinlock_rt.h
-===================================================================
---- /dev/null
-+++ linux-3.0/include/linux/spinlock_rt.h
-@@ -0,0 +1,166 @@
-+#ifndef __LINUX_SPINLOCK_RT_H
-+#define __LINUX_SPINLOCK_RT_H
-+
-+#ifndef __LINUX_SPINLOCK_H
-+#error Do not include directly. Use spinlock.h
-+#endif
-+
-+extern void
-+__rt_spin_lock_init(spinlock_t *lock, char *name, struct lock_class_key *key);
-+
-+#define spin_lock_init(slock) \
-+do { \
-+ static struct lock_class_key __key; \
-+ \
-+ rt_mutex_init(&(slock)->lock); \
-+ __rt_spin_lock_init(slock, #slock, &__key); \
-+} while (0)
-+
-+extern void __lockfunc rt_spin_lock(spinlock_t *lock);
-+extern unsigned long __lockfunc rt_spin_lock_trace_flags(spinlock_t *lock);
-+extern void __lockfunc rt_spin_lock_nested(spinlock_t *lock, int subclass);
-+extern void __lockfunc rt_spin_unlock(spinlock_t *lock);
-+extern void __lockfunc rt_spin_unlock_wait(spinlock_t *lock);
-+extern int __lockfunc rt_spin_trylock_irqsave(spinlock_t *lock, unsigned long *flags);
-+extern int __lockfunc rt_spin_trylock_bh(spinlock_t *lock);
-+extern int __lockfunc rt_spin_trylock(spinlock_t *lock);
-+extern int atomic_dec_and_spin_lock(atomic_t *atomic, spinlock_t *lock);
-+
-+/*
-+ * lockdep-less calls, for derived types like rwlock:
-+ * (for trylock they can use rt_mutex_trylock() directly.
-+ */
-+extern void __lockfunc __rt_spin_lock(struct rt_mutex *lock);
-+extern void __lockfunc __rt_spin_unlock(struct rt_mutex *lock);
-+
-+#define spin_lock_local(lock) rt_spin_lock(lock)
-+#define spin_unlock_local(lock) rt_spin_unlock(lock)
-+
-+#define spin_lock(lock) \
-+ do { \
-+ migrate_disable(); \
-+ rt_spin_lock(lock); \
-+ } while (0)
-+
-+#define spin_lock_bh(lock) \
-+ do { \
-+ local_bh_disable(); \
-+ migrate_disable(); \
-+ rt_spin_lock(lock); \
-+ } while (0)
-+
-+#define spin_lock_irq(lock) spin_lock(lock)
-+
-+#define spin_do_trylock(lock) __cond_lock(lock, rt_spin_trylock(lock))
-+
-+#define spin_trylock(lock) \
-+({ \
-+ int __locked; \
-+ migrate_disable(); \
-+ __locked = spin_do_trylock(lock); \
-+ if (!__locked) \
-+ migrate_enable(); \
-+ __locked; \
-+})
-+
-+#ifdef CONFIG_LOCKDEP
-+# define spin_lock_nested(lock, subclass) \
-+ do { \
-+ migrate_disable(); \
-+ rt_spin_lock_nested(lock, subclass); \
-+ } while (0)
-+
-+# define spin_lock_irqsave_nested(lock, flags, subclass) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ flags = 0; \
-+ migrate_disable(); \
-+ rt_spin_lock_nested(lock, subclass); \
-+ } while (0)
-+#else
-+# define spin_lock_nested(lock, subclass) spin_lock(lock)
-+
-+# define spin_lock_irqsave_nested(lock, flags, subclass) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ flags = 0; \
-+ spin_lock(lock); \
-+ } while (0)
-+#endif
-+
-+#define spin_lock_irqsave(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ flags = 0; \
-+ spin_lock(lock); \
-+ } while (0)
-+
-+static inline unsigned long spin_lock_trace_flags(spinlock_t *lock)
-+{
-+ unsigned long flags = 0;
-+#ifdef CONFIG_TRACE_IRQFLAGS
-+ flags = rt_spin_lock_trace_flags(lock);
-+#else
-+ spin_lock(lock); /* lock_local */
-+#endif
-+ return flags;
-+}
-+
-+/* FIXME: we need rt_spin_lock_nest_lock */
-+#define spin_lock_nest_lock(lock, nest_lock) spin_lock_nested(lock, 0)
-+
-+#define spin_unlock(lock) \
-+ do { \
-+ rt_spin_unlock(lock); \
-+ migrate_enable(); \
-+ } while (0)
-+
-+#define spin_unlock_bh(lock) \
-+ do { \
-+ rt_spin_unlock(lock); \
-+ migrate_enable(); \
-+ local_bh_enable(); \
-+ } while (0)
-+
-+#define spin_unlock_irq(lock) spin_unlock(lock)
-+
-+#define spin_unlock_irqrestore(lock, flags) \
-+ do { \
-+ typecheck(unsigned long, flags); \
-+ (void) flags; \
-+ spin_unlock(lock); \
-+ } while (0)
-+
-+#define spin_trylock_bh(lock) __cond_lock(lock, rt_spin_trylock_bh(lock))
-+#define spin_trylock_irq(lock) spin_trylock(lock)
-+
-+#define spin_trylock_irqsave(lock, flags) \
-+ rt_spin_trylock_irqsave(lock, &(flags))
-+
-+#define spin_unlock_wait(lock) rt_spin_unlock_wait(lock)
-+
-+#ifdef CONFIG_GENERIC_LOCKBREAK
-+# define spin_is_contended(lock) ((lock)->break_lock)
-+#else
-+# define spin_is_contended(lock) (((void)(lock), 0))
-+#endif
-+
-+static inline int spin_can_lock(spinlock_t *lock)
-+{
-+ return !rt_mutex_is_locked(&lock->lock);
-+}
-+
-+static inline int spin_is_locked(spinlock_t *lock)
-+{
-+ return rt_mutex_is_locked(&lock->lock);
-+}
-+
-+static inline void assert_spin_locked(spinlock_t *lock)
-+{
-+ BUG_ON(!spin_is_locked(lock));
-+}
-+
-+#define atomic_dec_and_lock(atomic, lock) \
-+ atomic_dec_and_spin_lock(atomic, lock)
-+
+ long long atomic64_sub_return(long long a, atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter -= a;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
+ }
+ EXPORT_SYMBOL(atomic64_sub_return);
+@@ -117,14 +117,14 @@ EXPORT_SYMBOL(atomic64_sub_return);
+ long long atomic64_dec_if_positive(atomic64_t *v)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter - 1;
+ if (val >= 0)
+ v->counter = val;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
+ }
+ EXPORT_SYMBOL(atomic64_dec_if_positive);
+@@ -132,14 +132,14 @@ EXPORT_SYMBOL(atomic64_dec_if_positive);
+ long long atomic64_cmpxchg(atomic64_t *v, long long o, long long n)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter;
+ if (val == o)
+ v->counter = n;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
+ }
+ EXPORT_SYMBOL(atomic64_cmpxchg);
+@@ -147,13 +147,13 @@ EXPORT_SYMBOL(atomic64_cmpxchg);
+ long long atomic64_xchg(atomic64_t *v, long long new)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ long long val;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ val = v->counter;
+ v->counter = new;
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return val;
+ }
+ EXPORT_SYMBOL(atomic64_xchg);
+@@ -161,15 +161,15 @@ EXPORT_SYMBOL(atomic64_xchg);
+ int atomic64_add_unless(atomic64_t *v, long long a, long long u)
+ {
+ unsigned long flags;
+- spinlock_t *lock = lock_addr(v);
++ raw_spinlock_t *lock = lock_addr(v);
+ int ret = 0;
+
+- spin_lock_irqsave(lock, flags);
++ raw_spin_lock_irqsave(lock, flags);
+ if (v->counter != u) {
+ v->counter += a;
+ ret = 1;
+ }
+- spin_unlock_irqrestore(lock, flags);
++ raw_spin_unlock_irqrestore(lock, flags);
+ return ret;
+ }
+ EXPORT_SYMBOL(atomic64_add_unless);
+@@ -179,7 +179,7 @@ static int init_atomic64_lock(void)
+ int i;
+
+ for (i = 0; i < NR_LOCKS; ++i)
+- spin_lock_init(&atomic64_lock[i].lock);
++ raw_spin_lock_init(&atomic64_lock[i].lock);
+ return 0;
+ }
+
+diff --git a/lib/debugobjects.c b/lib/debugobjects.c
+index a78b7c6..9b622c9 100644
+--- a/lib/debugobjects.c
++++ b/lib/debugobjects.c
+@@ -306,7 +306,10 @@ __debug_object_init(void *addr, struct debug_obj_descr *descr, int onstack)
+ struct debug_obj *obj;
+ unsigned long flags;
+
+- fill_pool();
++#ifdef CONFIG_PREEMPT_RT_FULL
++ if (preempt_count() == 0 && !irqs_disabled())
+#endif
-Index: linux-3.0/kernel/Makefile
-===================================================================
---- linux-3.0.orig/kernel/Makefile
-+++ linux-3.0/kernel/Makefile
-@@ -7,7 +7,7 @@ obj-y = sched.o fork.o exec_domain.o
- sysctl.o sysctl_binary.o capability.o ptrace.o timer.o user.o \
- signal.o sys.o kmod.o workqueue.o pid.o \
- rcupdate.o extable.o params.o posix-timers.o \
-- kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \
-+ kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o \
- hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \
- notifier.o ksysfs.o pm_qos_params.o sched_clock.o cred.o \
- async.o range.o jump_label.o
-@@ -29,7 +29,10 @@ obj-$(CONFIG_PROFILING) += profile.o
- obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o
- obj-$(CONFIG_STACKTRACE) += stacktrace.o
- obj-y += time/
-+ifneq ($(CONFIG_PREEMPT_RT_FULL),y)
-+obj-y += mutex.o
- obj-$(CONFIG_DEBUG_MUTEXES) += mutex-debug.o
-+endif
- obj-$(CONFIG_LOCKDEP) += lockdep.o
- ifeq ($(CONFIG_PROC_FS),y)
- obj-$(CONFIG_LOCKDEP) += lockdep_proc.o
-@@ -41,6 +44,7 @@ endif
- obj-$(CONFIG_RT_MUTEXES) += rtmutex.o
- obj-$(CONFIG_DEBUG_RT_MUTEXES) += rtmutex-debug.o
- obj-$(CONFIG_RT_MUTEX_TESTER) += rtmutex-tester.o
-+obj-$(CONFIG_PREEMPT_RT_FULL) += rt.o
- obj-$(CONFIG_GENERIC_ISA_DMA) += dma.o
- obj-$(CONFIG_SMP) += smp.o
- ifneq ($(CONFIG_SMP),y)
-Index: linux-3.0/kernel/rt.c
-===================================================================
---- /dev/null
-+++ linux-3.0/kernel/rt.c
-@@ -0,0 +1,442 @@
-+/*
-+ * kernel/rt.c
-+ *
-+ * Real-Time Preemption Support
-+ *
-+ * started by Ingo Molnar:
-+ *
-+ * Copyright (C) 2004-2006 Red Hat, Inc., Ingo Molnar <mingo at redhat.com>
-+ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx at timesys.com>
-+ *
-+ * historic credit for proving that Linux spinlocks can be implemented via
-+ * RT-aware mutexes goes to many people: The Pmutex project (Dirk Grambow
-+ * and others) who prototyped it on 2.4 and did lots of comparative
-+ * research and analysis; TimeSys, for proving that you can implement a
-+ * fully preemptible kernel via the use of IRQ threading and mutexes;
-+ * Bill Huey for persuasively arguing on lkml that the mutex model is the
-+ * right one; and to MontaVista, who ported pmutexes to 2.6.
-+ *
-+ * This code is a from-scratch implementation and is not based on pmutexes,
-+ * but the idea of converting spinlocks to mutexes is used here too.
-+ *
-+ * lock debugging, locking tree, deadlock detection:
-+ *
-+ * Copyright (C) 2004, LynuxWorks, Inc., Igor Manyilov, Bill Huey
-+ * Released under the General Public License (GPL).
-+ *
-+ * Includes portions of the generic R/W semaphore implementation from:
-+ *
-+ * Copyright (c) 2001 David Howells (dhowells at redhat.com).
-+ * - Derived partially from idea by Andrea Arcangeli <andrea at suse.de>
-+ * - Derived also from comments by Linus
-+ *
-+ * Pending ownership of locks and ownership stealing:
-+ *
-+ * Copyright (C) 2005, Kihon Technologies Inc., Steven Rostedt
-+ *
-+ * (also by Steven Rostedt)
-+ * - Converted single pi_lock to individual task locks.
-+ *
-+ * By Esben Nielsen:
-+ * Doing priority inheritance with help of the scheduler.
-+ *
-+ * Copyright (C) 2006, Timesys Corp., Thomas Gleixner <tglx at timesys.com>
-+ * - major rework based on Esben Nielsens initial patch
-+ * - replaced thread_info references by task_struct refs
-+ * - removed task->pending_owner dependency
-+ * - BKL drop/reacquire for semaphore style locks to avoid deadlocks
-+ * in the scheduler return path as discussed with Steven Rostedt
-+ *
-+ * Copyright (C) 2006, Kihon Technologies Inc.
-+ * Steven Rostedt <rostedt at goodmis.org>
-+ * - debugged and patched Thomas Gleixner's rework.
-+ * - added back the cmpxchg to the rework.
-+ * - turned atomic require back on for SMP.
-+ */
-+
-+#include <linux/spinlock.h>
-+#include <linux/rtmutex.h>
-+#include <linux/sched.h>
-+#include <linux/delay.h>
-+#include <linux/module.h>
-+#include <linux/kallsyms.h>
-+#include <linux/syscalls.h>
-+#include <linux/interrupt.h>
-+#include <linux/plist.h>
-+#include <linux/fs.h>
-+#include <linux/futex.h>
-+#include <linux/hrtimer.h>
-+
-+#include "rtmutex_common.h"
++ fill_pool();
+
+ db = get_bucket((unsigned long) addr);
+
+@@ -1015,9 +1018,9 @@ static int __init debug_objects_replace_static_objects(void)
+ }
+ }
+
++ local_irq_enable();
+ printk(KERN_DEBUG "ODEBUG: %d of %d active objects replaced\n", cnt,
+ obj_pool_used);
+- local_irq_enable();
+ return 0;
+ free:
+ hlist_for_each_entry_safe(obj, node, tmp, &objects, node) {
+diff --git a/lib/percpu_counter.c b/lib/percpu_counter.c
+index 28f2c33..f087105 100644
+--- a/lib/percpu_counter.c
++++ b/lib/percpu_counter.c
+@@ -59,13 +59,13 @@ void percpu_counter_set(struct percpu_counter *fbc, s64 amount)
+ {
+ int cpu;
+
+- spin_lock(&fbc->lock);
++ raw_spin_lock(&fbc->lock);
+ for_each_possible_cpu(cpu) {
+ s32 *pcount = per_cpu_ptr(fbc->counters, cpu);
+ *pcount = 0;
+ }
+ fbc->count = amount;
+- spin_unlock(&fbc->lock);
++ raw_spin_unlock(&fbc->lock);
+ }
+ EXPORT_SYMBOL(percpu_counter_set);
+
+@@ -76,10 +76,10 @@ void __percpu_counter_add(struct percpu_counter *fbc, s64 amount, s32 batch)
+ preempt_disable();
+ count = __this_cpu_read(*fbc->counters) + amount;
+ if (count >= batch || count <= -batch) {
+- spin_lock(&fbc->lock);
++ raw_spin_lock(&fbc->lock);
+ fbc->count += count;
+ __this_cpu_write(*fbc->counters, 0);
+- spin_unlock(&fbc->lock);
++ raw_spin_unlock(&fbc->lock);
+ } else {
+ __this_cpu_write(*fbc->counters, count);
+ }
+@@ -96,13 +96,13 @@ s64 __percpu_counter_sum(struct percpu_counter *fbc)
+ s64 ret;
+ int cpu;
+
+- spin_lock(&fbc->lock);
++ raw_spin_lock(&fbc->lock);
+ ret = fbc->count;
+ for_each_online_cpu(cpu) {
+ s32 *pcount = per_cpu_ptr(fbc->counters, cpu);
+ ret += *pcount;
+ }
+- spin_unlock(&fbc->lock);
++ raw_spin_unlock(&fbc->lock);
+ return ret;
+ }
+ EXPORT_SYMBOL(__percpu_counter_sum);
+@@ -110,7 +110,7 @@ EXPORT_SYMBOL(__percpu_counter_sum);
+ int __percpu_counter_init(struct percpu_counter *fbc, s64 amount,
+ struct lock_class_key *key)
+ {
+- spin_lock_init(&fbc->lock);
++ raw_spin_lock_init(&fbc->lock);
+ lockdep_set_class(&fbc->lock, key);
+ fbc->count = amount;
+ fbc->counters = alloc_percpu(s32);
+@@ -173,11 +173,11 @@ static int __cpuinit percpu_counter_hotcpu_callback(struct notifier_block *nb,
+ s32 *pcount;
+ unsigned long flags;
+
+- spin_lock_irqsave(&fbc->lock, flags);
++ raw_spin_lock_irqsave(&fbc->lock, flags);
+ pcount = per_cpu_ptr(fbc->counters, cpu);
+ fbc->count += *pcount;
+ *pcount = 0;
+- spin_unlock_irqrestore(&fbc->lock, flags);
++ raw_spin_unlock_irqrestore(&fbc->lock, flags);
+ }
+ mutex_unlock(&percpu_counters_lock);
+ #endif
+diff --git a/lib/proportions.c b/lib/proportions.c
+index d50746a..05df848 100644
+--- a/lib/proportions.c
++++ b/lib/proportions.c
+@@ -190,7 +190,7 @@ prop_adjust_shift(int *pl_shift, unsigned long *pl_period, int new_shift)
+
+ int prop_local_init_percpu(struct prop_local_percpu *pl)
+ {
+- spin_lock_init(&pl->lock);
++ raw_spin_lock_init(&pl->lock);
+ pl->shift = 0;
+ pl->period = 0;
+ return percpu_counter_init(&pl->events, 0);
+@@ -226,7 +226,7 @@ void prop_norm_percpu(struct prop_global *pg, struct prop_local_percpu *pl)
+ if (pl->period == global_period)
+ return;
+
+- spin_lock_irqsave(&pl->lock, flags);
++ raw_spin_lock_irqsave(&pl->lock, flags);
+ prop_adjust_shift(&pl->shift, &pl->period, pg->shift);
+
+ /*
+@@ -247,7 +247,7 @@ void prop_norm_percpu(struct prop_global *pg, struct prop_local_percpu *pl)
+ percpu_counter_set(&pl->events, 0);
+
+ pl->period = global_period;
+- spin_unlock_irqrestore(&pl->lock, flags);
++ raw_spin_unlock_irqrestore(&pl->lock, flags);
+ }
+
+ /*
+@@ -324,7 +324,7 @@ void prop_fraction_percpu(struct prop_descriptor *pd,
+
+ int prop_local_init_single(struct prop_local_single *pl)
+ {
+- spin_lock_init(&pl->lock);
++ raw_spin_lock_init(&pl->lock);
+ pl->shift = 0;
+ pl->period = 0;
+ pl->events = 0;
+@@ -356,7 +356,7 @@ void prop_norm_single(struct prop_global *pg, struct prop_local_single *pl)
+ if (pl->period == global_period)
+ return;
+
+- spin_lock_irqsave(&pl->lock, flags);
++ raw_spin_lock_irqsave(&pl->lock, flags);
+ prop_adjust_shift(&pl->shift, &pl->period, pg->shift);
+ /*
+ * For each missed period, we half the local counter.
+@@ -367,7 +367,7 @@ void prop_norm_single(struct prop_global *pg, struct prop_local_single *pl)
+ else
+ pl->events = 0;
+ pl->period = global_period;
+- spin_unlock_irqrestore(&pl->lock, flags);
++ raw_spin_unlock_irqrestore(&pl->lock, flags);
+ }
+
+ /*
+diff --git a/lib/radix-tree.c b/lib/radix-tree.c
+index 7ea2e03..205adb8 100644
+--- a/lib/radix-tree.c
++++ b/lib/radix-tree.c
+@@ -166,12 +166,13 @@ radix_tree_node_alloc(struct radix_tree_root *root)
+ * succeed in getting a node here (and never reach
+ * kmem_cache_alloc)
+ */
+- rtp = &__get_cpu_var(radix_tree_preloads);
++ rtp = &get_cpu_var(radix_tree_preloads);
+ if (rtp->nr) {
+ ret = rtp->nodes[rtp->nr - 1];
+ rtp->nodes[rtp->nr - 1] = NULL;
+ rtp->nr--;
+ }
++ put_cpu_var(radix_tree_preloads);
+ }
+ if (ret == NULL)
+ ret = kmem_cache_alloc(radix_tree_node_cachep, gfp_mask);
+@@ -206,6 +207,7 @@ radix_tree_node_free(struct radix_tree_node *node)
+ call_rcu(&node->rcu_head, radix_tree_node_rcu_free);
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ /*
+ * Load up this CPU's radix_tree_node buffer with sufficient objects to
+ * ensure that the addition of a single element in the tree cannot fail. On
+@@ -240,6 +242,7 @@ out:
+ return ret;
+ }
+ EXPORT_SYMBOL(radix_tree_preload);
++#endif
+
+ /*
+ * Return the maximum key which can be store into a
+diff --git a/lib/ratelimit.c b/lib/ratelimit.c
+index 027a03f..c96d500 100644
+--- a/lib/ratelimit.c
++++ b/lib/ratelimit.c
+@@ -39,7 +39,7 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func)
+ * in addition to the one that will be printed by
+ * the entity that is holding the lock already:
+ */
+- if (!spin_trylock_irqsave(&rs->lock, flags))
++ if (!raw_spin_trylock_irqsave(&rs->lock, flags))
+ return 0;
+
+ if (!rs->begin)
+@@ -60,7 +60,7 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func)
+ rs->missed++;
+ ret = 0;
+ }
+- spin_unlock_irqrestore(&rs->lock, flags);
++ raw_spin_unlock_irqrestore(&rs->lock, flags);
+
+ return ret;
+ }
+diff --git a/lib/rwsem-spinlock.c b/lib/rwsem-spinlock.c
+index ffc9fc7..26b7766 100644
+--- a/lib/rwsem-spinlock.c
++++ b/lib/rwsem-spinlock.c
+@@ -17,24 +17,24 @@ struct rwsem_waiter {
+ #define RWSEM_WAITING_FOR_WRITE 0x00000002
+ };
+
+-int rwsem_is_locked(struct rw_semaphore *sem)
++int anon_rwsem_is_locked(struct rw_anon_semaphore *sem)
+ {
+ int ret = 1;
+ unsigned long flags;
+
+- if (spin_trylock_irqsave(&sem->wait_lock, flags)) {
++ if (raw_spin_trylock_irqsave(&sem->wait_lock, flags)) {
+ ret = (sem->activity != 0);
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ }
+ return ret;
+ }
+-EXPORT_SYMBOL(rwsem_is_locked);
++EXPORT_SYMBOL(anon_rwsem_is_locked);
+
+ /*
+ * initialise the semaphore
+ */
+-void __init_rwsem(struct rw_semaphore *sem, const char *name,
+- struct lock_class_key *key)
++void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
++ struct lock_class_key *key)
+ {
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ /*
+@@ -44,10 +44,10 @@ void __init_rwsem(struct rw_semaphore *sem, const char *name,
+ lockdep_init_map(&sem->dep_map, name, key, 0);
+ #endif
+ sem->activity = 0;
+- spin_lock_init(&sem->wait_lock);
++ raw_spin_lock_init(&sem->wait_lock);
+ INIT_LIST_HEAD(&sem->wait_list);
+ }
+-EXPORT_SYMBOL(__init_rwsem);
++EXPORT_SYMBOL(__init_anon_rwsem);
+
+ /*
+ * handle the lock release when processes blocked on it that can now run
+@@ -58,8 +58,8 @@ EXPORT_SYMBOL(__init_rwsem);
+ * - woken process blocks are discarded from the list after having task zeroed
+ * - writers are only woken if wakewrite is non-zero
+ */
+-static inline struct rw_semaphore *
+-__rwsem_do_wake(struct rw_semaphore *sem, int wakewrite)
++static inline struct rw_anon_semaphore *
++__rwsem_do_wake(struct rw_anon_semaphore *sem, int wakewrite)
+ {
+ struct rwsem_waiter *waiter;
+ struct task_struct *tsk;
+@@ -117,8 +117,8 @@ __rwsem_do_wake(struct rw_semaphore *sem, int wakewrite)
+ /*
+ * wake a single writer
+ */
+-static inline struct rw_semaphore *
+-__rwsem_wake_one_writer(struct rw_semaphore *sem)
++static inline struct rw_anon_semaphore *
++__rwsem_wake_one_writer(struct rw_anon_semaphore *sem)
+ {
+ struct rwsem_waiter *waiter;
+ struct task_struct *tsk;
+@@ -139,18 +139,18 @@ __rwsem_wake_one_writer(struct rw_semaphore *sem)
+ /*
+ * get a read lock on the semaphore
+ */
+-void __sched __down_read(struct rw_semaphore *sem)
++void __sched __down_read(struct rw_anon_semaphore *sem)
+ {
+ struct rwsem_waiter waiter;
+ struct task_struct *tsk;
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ if (sem->activity >= 0 && list_empty(&sem->wait_list)) {
+ /* granted */
+ sem->activity++;
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ goto out;
+ }
+
+@@ -165,7 +165,7 @@ void __sched __down_read(struct rw_semaphore *sem)
+ list_add_tail(&waiter.list, &sem->wait_list);
+
+ /* we don't need to touch the semaphore struct anymore */
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ /* wait to be given the lock */
+ for (;;) {
+@@ -183,13 +183,13 @@ void __sched __down_read(struct rw_semaphore *sem)
+ /*
+ * trylock for reading -- returns 1 if successful, 0 if contention
+ */
+-int __down_read_trylock(struct rw_semaphore *sem)
++int __down_read_trylock(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+ int ret = 0;
+
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ if (sem->activity >= 0 && list_empty(&sem->wait_list)) {
+ /* granted */
+@@ -197,7 +197,7 @@ int __down_read_trylock(struct rw_semaphore *sem)
+ ret = 1;
+ }
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ return ret;
+ }
+@@ -206,18 +206,18 @@ int __down_read_trylock(struct rw_semaphore *sem)
+ * get a write lock on the semaphore
+ * - we increment the waiting count anyway to indicate an exclusive lock
+ */
+-void __sched __down_write_nested(struct rw_semaphore *sem, int subclass)
++void __sched __down_write_nested(struct rw_anon_semaphore *sem, int subclass)
+ {
+ struct rwsem_waiter waiter;
+ struct task_struct *tsk;
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ if (sem->activity == 0 && list_empty(&sem->wait_list)) {
+ /* granted */
+ sem->activity = -1;
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ goto out;
+ }
+
+@@ -232,7 +232,7 @@ void __sched __down_write_nested(struct rw_semaphore *sem, int subclass)
+ list_add_tail(&waiter.list, &sem->wait_list);
+
+ /* we don't need to touch the semaphore struct anymore */
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ /* wait to be given the lock */
+ for (;;) {
+@@ -247,7 +247,7 @@ void __sched __down_write_nested(struct rw_semaphore *sem, int subclass)
+ ;
+ }
+
+-void __sched __down_write(struct rw_semaphore *sem)
++void __sched __down_write(struct rw_anon_semaphore *sem)
+ {
+ __down_write_nested(sem, 0);
+ }
+@@ -255,12 +255,12 @@ void __sched __down_write(struct rw_semaphore *sem)
+ /*
+ * trylock for writing -- returns 1 if successful, 0 if contention
+ */
+-int __down_write_trylock(struct rw_semaphore *sem)
++int __down_write_trylock(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+ int ret = 0;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ if (sem->activity == 0 && list_empty(&sem->wait_list)) {
+ /* granted */
+@@ -268,7 +268,7 @@ int __down_write_trylock(struct rw_semaphore *sem)
+ ret = 1;
+ }
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ return ret;
+ }
+@@ -276,48 +276,48 @@ int __down_write_trylock(struct rw_semaphore *sem)
+ /*
+ * release a read lock on the semaphore
+ */
+-void __up_read(struct rw_semaphore *sem)
++void __up_read(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ if (--sem->activity == 0 && !list_empty(&sem->wait_list))
+ sem = __rwsem_wake_one_writer(sem);
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ }
+
+ /*
+ * release a write lock on the semaphore
+ */
+-void __up_write(struct rw_semaphore *sem)
++void __up_write(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ sem->activity = 0;
+ if (!list_empty(&sem->wait_list))
+ sem = __rwsem_do_wake(sem, 1);
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ }
+
+ /*
+ * downgrade a write lock into a read lock
+ * - just wake up any readers at the front of the queue
+ */
+-void __downgrade_write(struct rw_semaphore *sem)
++void __downgrade_write(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ sem->activity = 1;
+ if (!list_empty(&sem->wait_list))
+ sem = __rwsem_do_wake(sem, 0);
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+ }
+
+diff --git a/lib/rwsem.c b/lib/rwsem.c
+index aa7c305..5cb598c 100644
+--- a/lib/rwsem.c
++++ b/lib/rwsem.c
+@@ -11,8 +11,8 @@
+ /*
+ * Initialize an rwsem:
+ */
+-void __init_rwsem(struct rw_semaphore *sem, const char *name,
+- struct lock_class_key *key)
++void __init_anon_rwsem(struct rw_anon_semaphore *sem, const char *name,
++ struct lock_class_key *key)
+ {
+ #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ /*
+@@ -22,11 +22,11 @@ void __init_rwsem(struct rw_semaphore *sem, const char *name,
+ lockdep_init_map(&sem->dep_map, name, key, 0);
+ #endif
+ sem->count = RWSEM_UNLOCKED_VALUE;
+- spin_lock_init(&sem->wait_lock);
++ raw_spin_lock_init(&sem->wait_lock);
+ INIT_LIST_HEAD(&sem->wait_list);
+ }
+
+-EXPORT_SYMBOL(__init_rwsem);
++EXPORT_SYMBOL(__init_anon_rwsem);
+
+ struct rwsem_waiter {
+ struct list_head list;
+@@ -54,8 +54,8 @@ struct rwsem_waiter {
+ * - woken process blocks are discarded from the list after having task zeroed
+ * - writers are only woken if downgrading is false
+ */
+-static struct rw_semaphore *
+-__rwsem_do_wake(struct rw_semaphore *sem, int wake_type)
++static struct rw_anon_semaphore *
++__rwsem_do_wake(struct rw_anon_semaphore *sem, int wake_type)
+ {
+ struct rwsem_waiter *waiter;
+ struct task_struct *tsk;
+@@ -169,8 +169,8 @@ __rwsem_do_wake(struct rw_semaphore *sem, int wake_type)
+ /*
+ * wait for a lock to be granted
+ */
+-static struct rw_semaphore __sched *
+-rwsem_down_failed_common(struct rw_semaphore *sem,
++static struct rw_anon_semaphore __sched *
++rwsem_down_failed_common(struct rw_anon_semaphore *sem,
+ unsigned int flags, signed long adjustment)
+ {
+ struct rwsem_waiter waiter;
+@@ -180,7 +180,7 @@ rwsem_down_failed_common(struct rw_semaphore *sem,
+ set_task_state(tsk, TASK_UNINTERRUPTIBLE);
+
+ /* set up my own style of waitqueue */
+- spin_lock_irq(&sem->wait_lock);
++ raw_spin_lock_irq(&sem->wait_lock);
+ waiter.task = tsk;
+ waiter.flags = flags;
+ get_task_struct(tsk);
+@@ -204,7 +204,7 @@ rwsem_down_failed_common(struct rw_semaphore *sem,
+ adjustment == -RWSEM_ACTIVE_WRITE_BIAS)
+ sem = __rwsem_do_wake(sem, RWSEM_WAKE_READ_OWNED);
+
+- spin_unlock_irq(&sem->wait_lock);
++ raw_spin_unlock_irq(&sem->wait_lock);
+
+ /* wait to be given the lock */
+ for (;;) {
+@@ -222,7 +222,8 @@ rwsem_down_failed_common(struct rw_semaphore *sem,
+ /*
+ * wait for the read lock to be granted
+ */
+-struct rw_semaphore __sched *rwsem_down_read_failed(struct rw_semaphore *sem)
++struct rw_anon_semaphore __sched *
++rwsem_down_read_failed(struct rw_anon_semaphore *sem)
+ {
+ return rwsem_down_failed_common(sem, RWSEM_WAITING_FOR_READ,
+ -RWSEM_ACTIVE_READ_BIAS);
+@@ -231,7 +232,8 @@ struct rw_semaphore __sched *rwsem_down_read_failed(struct rw_semaphore *sem)
+ /*
+ * wait for the write lock to be granted
+ */
+-struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem)
++struct rw_anon_semaphore __sched *
++rwsem_down_write_failed(struct rw_anon_semaphore *sem)
+ {
+ return rwsem_down_failed_common(sem, RWSEM_WAITING_FOR_WRITE,
+ -RWSEM_ACTIVE_WRITE_BIAS);
+@@ -241,17 +243,17 @@ struct rw_semaphore __sched *rwsem_down_write_failed(struct rw_semaphore *sem)
+ * handle waking up a waiter on the semaphore
+ * - up_read/up_write has decremented the active part of count if we come here
+ */
+-struct rw_semaphore *rwsem_wake(struct rw_semaphore *sem)
++struct rw_anon_semaphore *rwsem_wake(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ /* do nothing if list empty */
+ if (!list_empty(&sem->wait_list))
+ sem = __rwsem_do_wake(sem, RWSEM_WAKE_ANY);
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ return sem;
+ }
+@@ -261,17 +263,17 @@ struct rw_semaphore *rwsem_wake(struct rw_semaphore *sem)
+ * - caller incremented waiting part of count and discovered it still negative
+ * - just wake up any readers at the front of the queue
+ */
+-struct rw_semaphore *rwsem_downgrade_wake(struct rw_semaphore *sem)
++struct rw_anon_semaphore *rwsem_downgrade_wake(struct rw_anon_semaphore *sem)
+ {
+ unsigned long flags;
+
+- spin_lock_irqsave(&sem->wait_lock, flags);
++ raw_spin_lock_irqsave(&sem->wait_lock, flags);
+
+ /* do nothing if list empty */
+ if (!list_empty(&sem->wait_list))
+ sem = __rwsem_do_wake(sem, RWSEM_WAKE_READ_OWNED);
+
+- spin_unlock_irqrestore(&sem->wait_lock, flags);
++ raw_spin_unlock_irqrestore(&sem->wait_lock, flags);
+
+ return sem;
+ }
+diff --git a/lib/scatterlist.c b/lib/scatterlist.c
+index 4ceb05d..181f364 100644
+--- a/lib/scatterlist.c
++++ b/lib/scatterlist.c
+@@ -423,7 +423,7 @@ void sg_miter_stop(struct sg_mapping_iter *miter)
+ flush_kernel_dcache_page(miter->page);
+
+ if (miter->__flags & SG_MITER_ATOMIC) {
+- WARN_ON(!irqs_disabled());
++ WARN_ON_NONRT(!irqs_disabled());
+ kunmap_atomic(miter->addr, KM_BIO_SRC_IRQ);
+ } else
+ kunmap(miter->page);
+@@ -463,7 +463,7 @@ static size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents,
+
+ sg_miter_start(&miter, sgl, nents, sg_flags);
+
+- local_irq_save(flags);
++ local_irq_save_nort(flags);
+
+ while (sg_miter_next(&miter) && offset < buflen) {
+ unsigned int len;
+@@ -480,7 +480,7 @@ static size_t sg_copy_buffer(struct scatterlist *sgl, unsigned int nents,
+
+ sg_miter_stop(&miter);
+
+- local_irq_restore(flags);
++ local_irq_restore_nort(flags);
+ return offset;
+ }
+
+diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c
+index 4689cb0..b5e9241 100644
+--- a/lib/smp_processor_id.c
++++ b/lib/smp_processor_id.c
+@@ -22,7 +22,7 @@ notrace unsigned int debug_smp_processor_id(void)
+ * Kernel threads bound to a single CPU can safely use
+ * smp_processor_id():
+ */
+- if (cpumask_equal(¤t->cpus_allowed, cpumask_of(this_cpu)))
++ if (cpumask_equal(tsk_cpus_allowed(current), cpumask_of(this_cpu)))
+ goto out;
+
+ /*
+@@ -39,9 +39,9 @@ notrace unsigned int debug_smp_processor_id(void)
+ if (!printk_ratelimit())
+ goto out_enable;
+
+- printk(KERN_ERR "BUG: using smp_processor_id() in preemptible [%08x] "
+- "code: %s/%d\n",
+- preempt_count() - 1, current->comm, current->pid);
++ printk(KERN_ERR "BUG: using smp_processor_id() in preemptible [%08x %08x] "
++ "code: %s/%d\n", preempt_count() - 1,
++ __migrate_disabled(current), current->comm, current->pid);
+ print_symbol("caller is %s\n", (long)__builtin_return_address(0));
+ dump_stack();
+
+diff --git a/lib/spinlock_debug.c b/lib/spinlock_debug.c
+index 4755b98..4566b69b 100644
+--- a/lib/spinlock_debug.c
++++ b/lib/spinlock_debug.c
+@@ -31,6 +31,7 @@ void __raw_spin_lock_init(raw_spinlock_t *lock, const char *name,
+
+ EXPORT_SYMBOL(__raw_spin_lock_init);
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ void __rwlock_init(rwlock_t *lock, const char *name,
+ struct lock_class_key *key)
+ {
+@@ -46,8 +47,8 @@ void __rwlock_init(rwlock_t *lock, const char *name,
+ lock->owner = SPINLOCK_OWNER_INIT;
+ lock->owner_cpu = -1;
+ }
+-
+ EXPORT_SYMBOL(__rwlock_init);
++#endif
+
+ static void spin_bug(raw_spinlock_t *lock, const char *msg)
+ {
+@@ -154,6 +155,7 @@ void do_raw_spin_unlock(raw_spinlock_t *lock)
+ arch_spin_unlock(&lock->raw_lock);
+ }
+
++#ifndef CONFIG_PREEMPT_RT_FULL
+ static void rwlock_bug(rwlock_t *lock, const char *msg)
+ {
+ if (!debug_locks_off())
+@@ -295,3 +297,5 @@ void do_raw_write_unlock(rwlock_t *lock)
+ debug_write_unlock(lock);
+ arch_write_unlock(&lock->raw_lock);
+ }
+
-+/*
-+ * struct mutex functions
-+ */
-+void __mutex_do_init(struct mutex *mutex, const char *name,
-+ struct lock_class_key *key)
++#endif
+diff --git a/mm/Kconfig b/mm/Kconfig
+index 8ca47a5..3c36545 100644
+--- a/mm/Kconfig
++++ b/mm/Kconfig
+@@ -304,7 +304,7 @@ config NOMMU_INITIAL_TRIM_EXCESS
+
+ config TRANSPARENT_HUGEPAGE
+ bool "Transparent Hugepage Support"
+- depends on X86 && MMU
++ depends on X86 && MMU && !PREEMPT_RT_FULL
+ select COMPACTION
+ help
+ Transparent Hugepages allows the kernel to use huge pages and
+diff --git a/mm/filemap.c b/mm/filemap.c
+index a8251a8..be1b637 100644
+--- a/mm/filemap.c
++++ b/mm/filemap.c
+@@ -2040,7 +2040,7 @@ size_t iov_iter_copy_from_user_atomic(struct page *page,
+ char *kaddr;
+ size_t copied;
+
+- BUG_ON(!in_atomic());
++ BUG_ON(!pagefault_disabled());
+ kaddr = kmap_atomic(page, KM_USER0);
+ if (likely(i->nr_segs == 1)) {
+ int left;
+diff --git a/mm/memory.c b/mm/memory.c
+index 95a7799..52f6771 100644
+--- a/mm/memory.c
++++ b/mm/memory.c
+@@ -1290,13 +1290,6 @@ static unsigned long unmap_page_range(struct mmu_gather *tlb,
+ return addr;
+ }
+
+-#ifdef CONFIG_PREEMPT
+-# define ZAP_BLOCK_SIZE (8 * PAGE_SIZE)
+-#else
+-/* No preempt: go for improved straight-line efficiency */
+-# define ZAP_BLOCK_SIZE (1024 * PAGE_SIZE)
+-#endif
+-
+ /**
+ * unmap_vmas - unmap a range of memory covered by a list of vma's
+ * @tlb: address of the caller's struct mmu_gather
+@@ -3435,6 +3428,32 @@ unlock:
+ return 0;
+ }
+
++#ifdef CONFIG_PREEMPT_RT_FULL
++void pagefault_disable(void)
+{
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ migrate_disable();
++ current->pagefault_disabled++;
+ /*
-+ * Make sure we are not reinitializing a held lock:
++ * make sure to have issued the store before a pagefault
++ * can hit.
+ */
-+ debug_check_no_locks_freed((void *)mutex, sizeof(*mutex));
-+ lockdep_init_map(&mutex->dep_map, name, key, 0);
-+#endif
-+ mutex->lock.save_state = 0;
-+}
-+EXPORT_SYMBOL(__mutex_do_init);
-+
-+void __lockfunc _mutex_lock(struct mutex *lock)
-+{
-+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
-+ rt_mutex_lock(&lock->lock);
-+}
-+EXPORT_SYMBOL(_mutex_lock);
-+
-+int __lockfunc _mutex_lock_interruptible(struct mutex *lock)
-+{
-+ int ret;
-+
-+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
-+ ret = rt_mutex_lock_interruptible(&lock->lock, 0);
-+ if (ret)
-+ mutex_release(&lock->dep_map, 1, _RET_IP_);
-+ return ret;
-+}
-+EXPORT_SYMBOL(_mutex_lock_interruptible);
-+
-+int __lockfunc _mutex_lock_killable(struct mutex *lock)
-+{
-+ int ret;
-+
-+ mutex_acquire(&lock->dep_map, 0, 0, _RET_IP_);
-+ ret = rt_mutex_lock_killable(&lock->lock, 0);
-+ if (ret)
-+ mutex_release(&lock->dep_map, 1, _RET_IP_);
-+ return ret;
-+}
-+EXPORT_SYMBOL(_mutex_lock_killable);
-+
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+void __lockfunc _mutex_lock_nested(struct mutex *lock, int subclass)
-+{
-+ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_);
-+ rt_mutex_lock(&lock->lock);
-+}
-+EXPORT_SYMBOL(_mutex_lock_nested);
-+
-+void __lockfunc _mutex_lock_nest_lock(struct mutex *lock, struct lockdep_map *nest)
-+{
-+ mutex_acquire_nest(&lock->dep_map, 0, 0, nest, _RET_IP_);
-+ rt_mutex_lock(&lock->lock);
-+}
-+EXPORT_SYMBOL(_mutex_lock_nest_lock);
-+
-+int __lockfunc _mutex_lock_interruptible_nested(struct mutex *lock, int subclass)
-+{
-+ int ret;
-+
-+ mutex_acquire_nest(&lock->dep_map, subclass, 0, NULL, _RET_IP_);
-+ ret = rt_mutex_lock_interruptible(&lock->lock, 0);
-+ if (ret)
-+ mutex_release(&lock->dep_map, 1, _RET_IP_);
-+ return ret;
++ barrier();
+}
-+EXPORT_SYMBOL(_mutex_lock_interruptible_nested);
++EXPORT_SYMBOL_GPL(pagefault_disable);
+
-+int __lockfunc _mutex_lock_killable_nested(struct mutex *lock, int subclass)
++void pagefault_enable(void)
+{
-+ int ret;
-+
-+ mutex_acquire(&lock->dep_map, subclass, 0, _RET_IP_);
-+ ret = rt_mutex_lock_killable(&lock->lock, 0);
-+ if (ret)
-+ mutex_release(&lock->dep_map, 1, _RET_IP_);
-+ return ret;
++ /*
++ * make sure to issue those last loads/stores before enabling
++ * the pagefault handler again.
++ */
++ barrier();
++ current->pagefault_disabled--;
++ migrate_enable();
+}
-+EXPORT_SYMBOL(_mutex_lock_killable_nested);
++EXPORT_SYMBOL_GPL(pagefault_enable);
+#endif
+
-+int __lockfunc _mutex_trylock(struct mutex *lock)
-+{
-+ int ret = rt_mutex_trylock(&lock->lock);
-+
-+ if (ret)
-+ mutex_acquire(&lock->dep_map, 0, 1, _RET_IP_);
-+
-+ return ret;
-+}
-+EXPORT_SYMBOL(_mutex_trylock);
-+
-+void __lockfunc _mutex_unlock(struct mutex *lock)
-+{
-+ mutex_release(&lock->dep_map, 1, _RET_IP_);
-+ rt_mutex_unlock(&lock->lock);
-+}
-+EXPORT_SYMBOL(_mutex_unlock);
+ /*
+ * By the time we get here, we already hold the mm semaphore
+ */
+@@ -3983,3 +4002,35 @@ void copy_user_huge_page(struct page *dst, struct page *src,
+ }
+ }
+ #endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_HUGETLBFS */
+
++#if defined(CONFIG_PREEMPT_RT_FULL) && (USE_SPLIT_PTLOCKS > 0)
+/*
-+ * rwlock_t functions
++ * Heinous hack, relies on the caller doing something like:
++ *
++ * pte = alloc_pages(PGALLOC_GFP, 0);
++ * if (pte)
++ * pgtable_page_ctor(pte);
++ * return pte;
++ *
++ * This ensures we release the page and return NULL when the
++ * lock allocation fails.
+ */
-+int __lockfunc rt_write_trylock(rwlock_t *rwlock)
-+{
-+ int ret = rt_mutex_trylock(&rwlock->lock);
-+
-+ migrate_disable();
-+ if (ret)
-+ rwlock_acquire(&rwlock->dep_map, 0, 1, _RET_IP_);
-+ else
-+ migrate_enable();
-+
-+ return ret;
-+}
-+EXPORT_SYMBOL(rt_write_trylock);
-+
-+int __lockfunc rt_write_trylock_irqsave(rwlock_t *rwlock, unsigned long *flags)
++struct page *pte_lock_init(struct page *page)
+{
-+ int ret;
-+
-+ *flags = 0;
-+ migrate_disable();
-+ ret = rt_write_trylock(rwlock);
-+ if (!ret)
-+ migrate_enable();
-+ return ret;
++ page->ptl = kmalloc(sizeof(spinlock_t), GFP_KERNEL);
++ if (page->ptl) {
++ spin_lock_init(__pte_lockptr(page));
++ } else {
++ __free_page(page);
++ page = NULL;
++ }
++ return page;
+}
-+EXPORT_SYMBOL(rt_write_trylock_irqsave);
+
-+int __lockfunc rt_read_trylock(rwlock_t *rwlock)
++void pte_lock_deinit(struct page *page)
+{
-+ struct rt_mutex *lock = &rwlock->lock;
-+ int ret = 1;
-+
-+ /*
-+ * recursive read locks succeed when current owns the lock,
-+ * but not when read_depth == 0 which means that the lock is
-+ * write locked.
-+ */
-+ migrate_disable();
-+ if (rt_mutex_owner(lock) != current)
-+ ret = rt_mutex_trylock(lock);
-+ else if (!rwlock->read_depth)
-+ ret = 0;
-+
-+ if (ret) {
-+ rwlock->read_depth++;
-+ rwlock_acquire_read(&rwlock->dep_map, 0, 1, _RET_IP_);
-+ } else
-+ migrate_enable();
-+
-+ return ret;
++ kfree(page->ptl);
++ page->mapping = NULL;
+}
-+EXPORT_SYMBOL(rt_read_trylock);
+
-+void __lockfunc rt_write_lock(rwlock_t *rwlock)
-+{
-+ rwlock_acquire(&rwlock->dep_map, 0, 0, _RET_IP_);
-+ __rt_spin_lock(&rwlock->lock);
-+}
-+EXPORT_SYMBOL(rt_write_lock);
++#endif
+diff --git a/mm/mmu_context.c b/mm/mmu_context.c
+index 9e82e93..a9d4ddc 100644
+--- a/mm/mmu_context.c
++++ b/mm/mmu_context.c
+@@ -26,6 +26,7 @@ void use_mm(struct mm_struct *mm)
+ struct task_struct *tsk = current;
+
+ task_lock(tsk);
++ local_irq_disable_rt();
+ active_mm = tsk->active_mm;
+ if (active_mm != mm) {
+ atomic_inc(&mm->mm_count);
+@@ -33,6 +34,7 @@ void use_mm(struct mm_struct *mm)
+ }
+ tsk->mm = mm;
+ switch_mm(active_mm, mm, tsk);
++ local_irq_enable_rt();
+ task_unlock(tsk);
+
+ if (active_mm != mm)
+diff --git a/mm/page_alloc.c b/mm/page_alloc.c
+index 0f50cdb..d892db7 100644
+--- a/mm/page_alloc.c
++++ b/mm/page_alloc.c
+@@ -57,6 +57,7 @@
+ #include <linux/ftrace_event.h>
+ #include <linux/memcontrol.h>
+ #include <linux/prefetch.h>
++#include <linux/locallock.h>
+
+ #include <asm/tlbflush.h>
+ #include <asm/div64.h>
+@@ -222,6 +223,18 @@ EXPORT_SYMBOL(nr_node_ids);
+ EXPORT_SYMBOL(nr_online_nodes);
+ #endif
+
++static DEFINE_LOCAL_IRQ_LOCK(pa_lock);
+
-+void __lockfunc rt_read_lock(rwlock_t *rwlock)
-+{
-+ struct rt_mutex *lock = &rwlock->lock;
++#ifdef CONFIG_PREEMPT_RT_BASE
++# define cpu_lock_irqsave(cpu, flags) \
++ spin_lock_irqsave(&per_cpu(pa_lock, cpu).lock, flags)
++# define cpu_unlock_irqrestore(cpu, flags) \
++ spin_unlock_irqrestore(&per_cpu(pa_lock, cpu).lock, flags)
++#else
++# define cpu_lock_irqsave(cpu, flags) local_irq_save(flags)
++# define cpu_unlock_irqrestore(cpu, flags) local_irq_restore(flags)
++#endif
+
-+ rwlock_acquire_read(&rwlock->dep_map, 0, 0, _RET_IP_);
+ int page_group_by_mobility_disabled __read_mostly;
+
+ static void set_pageblock_migratetype(struct page *page, int migratetype)
+@@ -580,7 +593,7 @@ static inline int free_pages_check(struct page *page)
+ }
+
+ /*
+- * Frees a number of pages from the PCP lists
++ * Frees a number of pages which have been collected from the pcp lists.
+ * Assumes all pages on list are in same zone, and of same order.
+ * count is the number of pages to free.
+ *
+@@ -591,16 +604,42 @@ static inline int free_pages_check(struct page *page)
+ * pinned" detection logic.
+ */
+ static void free_pcppages_bulk(struct zone *zone, int count,
+- struct per_cpu_pages *pcp)
++ struct list_head *list)
+ {
+- int migratetype = 0;
+- int batch_free = 0;
+ int to_free = count;
++ unsigned long flags;
+
+- spin_lock(&zone->lock);
++ spin_lock_irqsave(&zone->lock, flags);
+ zone->all_unreclaimable = 0;
+ zone->pages_scanned = 0;
+
++ while (!list_empty(list)) {
++ struct page *page = list_first_entry(list, struct page, lru);
+
-+ /*
-+ * recursive read locks succeed when current owns the lock
-+ */
-+ if (rt_mutex_owner(lock) != current)
-+ __rt_spin_lock(lock);
-+ rwlock->read_depth++;
++ /* must delete as __free_one_page list manipulates */
++ list_del(&page->lru);
++ /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
++ __free_one_page(page, zone, 0, page_private(page));
++ trace_mm_page_pcpu_drain(page, 0, page_private(page));
++ to_free--;
++ }
++ WARN_ON(to_free != 0);
++ __mod_zone_page_state(zone, NR_FREE_PAGES, count);
++ spin_unlock_irqrestore(&zone->lock, flags);
+}
+
-+EXPORT_SYMBOL(rt_read_lock);
-+
-+void __lockfunc rt_write_unlock(rwlock_t *rwlock)
++/*
++ * Moves a number of pages from the PCP lists to free list which
++ * is freed outside of the locked region.
++ *
++ * Assumes all pages on list are in same zone, and of same order.
++ * count is the number of pages to free.
++ */
++static void isolate_pcp_pages(int to_free, struct per_cpu_pages *src,
++ struct list_head *dst)
+{
-+ /* NOTE: we always pass in '1' for nested, for simplicity */
-+ rwlock_release(&rwlock->dep_map, 1, _RET_IP_);
-+ __rt_spin_unlock(&rwlock->lock);
-+}
-+EXPORT_SYMBOL(rt_write_unlock);
++ int migratetype = 0, batch_free = 0;
+
-+void __lockfunc rt_read_unlock(rwlock_t *rwlock)
-+{
-+ rwlock_release(&rwlock->dep_map, 1, _RET_IP_);
+ while (to_free) {
+ struct page *page;
+ struct list_head *list;
+@@ -616,7 +655,7 @@ static void free_pcppages_bulk(struct zone *zone, int count,
+ batch_free++;
+ if (++migratetype == MIGRATE_PCPTYPES)
+ migratetype = 0;
+- list = &pcp->lists[migratetype];
++ list = &src->lists[migratetype];
+ } while (list_empty(list));
+
+ /* This is the only non-empty list. Free them all. */
+@@ -624,28 +663,25 @@ static void free_pcppages_bulk(struct zone *zone, int count,
+ batch_free = to_free;
+
+ do {
+- page = list_entry(list->prev, struct page, lru);
+- /* must delete as __free_one_page list manipulates */
++ page = list_last_entry(list, struct page, lru);
+ list_del(&page->lru);
+- /* MIGRATE_MOVABLE list may include MIGRATE_RESERVEs */
+- __free_one_page(page, zone, 0, page_private(page));
+- trace_mm_page_pcpu_drain(page, 0, page_private(page));
++ list_add(&page->lru, dst);
+ } while (--to_free && --batch_free && !list_empty(list));
+ }
+- __mod_zone_page_state(zone, NR_FREE_PAGES, count);
+- spin_unlock(&zone->lock);
+ }
+
+ static void free_one_page(struct zone *zone, struct page *page, int order,
+ int migratetype)
+ {
+- spin_lock(&zone->lock);
++ unsigned long flags;
+
-+ /* Release the lock only when read_depth is down to 0 */
-+ if (--rwlock->read_depth == 0)
-+ __rt_spin_unlock(&rwlock->lock);
-+}
-+EXPORT_SYMBOL(rt_read_unlock);
++ spin_lock_irqsave(&zone->lock, flags);
+ zone->all_unreclaimable = 0;
+ zone->pages_scanned = 0;
+
+ __free_one_page(page, zone, order, migratetype);
+ __mod_zone_page_state(zone, NR_FREE_PAGES, 1 << order);
+- spin_unlock(&zone->lock);
++ spin_unlock_irqrestore(&zone->lock, flags);
+ }
+
+ static bool free_pages_prepare(struct page *page, unsigned int order)
+@@ -682,13 +718,13 @@ static void __free_pages_ok(struct page *page, unsigned int order)
+ if (!free_pages_prepare(page, order))
+ return;
+
+- local_irq_save(flags);
++ local_lock_irqsave(pa_lock, flags);
+ if (unlikely(wasMlocked))
+ free_page_mlock(page);
+ __count_vm_events(PGFREE, 1 << order);
+ free_one_page(page_zone(page), page, order,
+ get_pageblock_migratetype(page));
+- local_irq_restore(flags);
++ local_unlock_irqrestore(pa_lock, flags);
+ }
+
+ /*
+@@ -1064,16 +1100,18 @@ static int rmqueue_bulk(struct zone *zone, unsigned int order,
+ void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp)
+ {
+ unsigned long flags;
++ LIST_HEAD(dst);
+ int to_drain;
+
+- local_irq_save(flags);
++ local_lock_irqsave(pa_lock, flags);
+ if (pcp->count >= pcp->batch)
+ to_drain = pcp->batch;
+ else
+ to_drain = pcp->count;
+- free_pcppages_bulk(zone, to_drain, pcp);
++ isolate_pcp_pages(to_drain, pcp, &dst);
+ pcp->count -= to_drain;
+- local_irq_restore(flags);
++ local_unlock_irqrestore(pa_lock, flags);
++ free_pcppages_bulk(zone, to_drain, &dst);
+ }
+ #endif
+
+@@ -1092,16 +1130,21 @@ static void drain_pages(unsigned int cpu)
+ for_each_populated_zone(zone) {
+ struct per_cpu_pageset *pset;
+ struct per_cpu_pages *pcp;
++ LIST_HEAD(dst);
++ int count;
+
+- local_irq_save(flags);
++ cpu_lock_irqsave(cpu, flags);
+ pset = per_cpu_ptr(zone->pageset, cpu);
+
+ pcp = &pset->pcp;
+- if (pcp->count) {
+- free_pcppages_bulk(zone, pcp->count, pcp);
++ count = pcp->count;
++ if (count) {
++ isolate_pcp_pages(count, pcp, &dst);
+ pcp->count = 0;
+ }
+- local_irq_restore(flags);
++ cpu_unlock_irqrestore(cpu, flags);
++ if (count)
++ free_pcppages_bulk(zone, count, &dst);
+ }
+ }
+
+@@ -1118,7 +1161,14 @@ void drain_local_pages(void *arg)
+ */
+ void drain_all_pages(void)
+ {
++#ifndef CONFIG_PREEMPT_RT_BASE
+ on_each_cpu(drain_local_pages, NULL, 1);
++#else
++ int i;
+
-+unsigned long __lockfunc rt_write_lock_irqsave(rwlock_t *rwlock)
-+{
-+ rt_write_lock(rwlock);
++ for_each_online_cpu(i)
++ drain_pages(i);
++#endif
+ }
+
+ #ifdef CONFIG_HIBERNATION
+@@ -1174,7 +1224,7 @@ void free_hot_cold_page(struct page *page, int cold)
+
+ migratetype = get_pageblock_migratetype(page);
+ set_page_private(page, migratetype);
+- local_irq_save(flags);
++ local_lock_irqsave(pa_lock, flags);
+ if (unlikely(wasMlocked))
+ free_page_mlock(page);
+ __count_vm_event(PGFREE);
+@@ -1201,12 +1251,19 @@ void free_hot_cold_page(struct page *page, int cold)
+ list_add(&page->lru, &pcp->lists[migratetype]);
+ pcp->count++;
+ if (pcp->count >= pcp->high) {
+- free_pcppages_bulk(zone, pcp->batch, pcp);
++ LIST_HEAD(dst);
++ int count;
+
-+ return 0;
-+}
-+EXPORT_SYMBOL(rt_write_lock_irqsave);
++ isolate_pcp_pages(pcp->batch, pcp, &dst);
+ pcp->count -= pcp->batch;
++ count = pcp->batch;
++ local_unlock_irqrestore(pa_lock, flags);
++ free_pcppages_bulk(zone, count, &dst);
++ return;
+ }
+
+ out:
+- local_irq_restore(flags);
++ local_unlock_irqrestore(pa_lock, flags);
+ }
+
+ /*
+@@ -1301,7 +1358,7 @@ again:
+ struct per_cpu_pages *pcp;
+ struct list_head *list;
+
+- local_irq_save(flags);
++ local_lock_irqsave(pa_lock, flags);
+ pcp = &this_cpu_ptr(zone->pageset)->pcp;
+ list = &pcp->lists[migratetype];
+ if (list_empty(list)) {
+@@ -1333,17 +1390,19 @@ again:
+ */
+ WARN_ON_ONCE(order > 1);
+ }
+- spin_lock_irqsave(&zone->lock, flags);
++ local_spin_lock_irqsave(pa_lock, &zone->lock, flags);
+ page = __rmqueue(zone, order, migratetype);
+- spin_unlock(&zone->lock);
+- if (!page)
++ if (!page) {
++ spin_unlock(&zone->lock);
+ goto failed;
++ }
+ __mod_zone_page_state(zone, NR_FREE_PAGES, -(1 << order));
++ spin_unlock(&zone->lock);
+ }
+
+ __count_zone_vm_events(PGALLOC, zone, 1 << order);
+ zone_statistics(preferred_zone, zone, gfp_flags);
+- local_irq_restore(flags);
++ local_unlock_irqrestore(pa_lock, flags);
+
+ VM_BUG_ON(bad_range(zone, page));
+ if (prep_new_page(page, order, gfp_flags))
+@@ -1351,7 +1410,7 @@ again:
+ return page;
+
+ failed:
+- local_irq_restore(flags);
++ local_unlock_irqrestore(pa_lock, flags);
+ return NULL;
+ }
+
+@@ -1912,8 +1971,8 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
+ if (*did_some_progress != COMPACT_SKIPPED) {
+
+ /* Page migration frees to the PCP lists but we want merging */
+- drain_pages(get_cpu());
+- put_cpu();
++ drain_pages(get_cpu_light());
++ put_cpu_light();
+
+ page = get_page_from_freelist(gfp_mask, nodemask,
+ order, zonelist, high_zoneidx,
+@@ -3685,14 +3744,16 @@ static int __zone_pcp_update(void *data)
+ for_each_possible_cpu(cpu) {
+ struct per_cpu_pageset *pset;
+ struct per_cpu_pages *pcp;
++ LIST_HEAD(dst);
+
+ pset = per_cpu_ptr(zone->pageset, cpu);
+ pcp = &pset->pcp;
+
+- local_irq_save(flags);
+- free_pcppages_bulk(zone, pcp->count, pcp);
++ cpu_lock_irqsave(cpu, flags);
++ isolate_pcp_pages(pcp->count, pcp, &dst);
++ free_pcppages_bulk(zone, pcp->count, &dst);
+ setup_pageset(pset, batch);
+- local_irq_restore(flags);
++ cpu_unlock_irqrestore(cpu, flags);
+ }
+ return 0;
+ }
+@@ -5004,6 +5065,7 @@ static int page_alloc_cpu_notify(struct notifier_block *self,
+ void __init page_alloc_init(void)
+ {
+ hotcpu_notifier(page_alloc_cpu_notify, 0);
++ local_irq_lock_init(pa_lock);
+ }
+
+ /*
+diff --git a/mm/page_cgroup.c b/mm/page_cgroup.c
+index 53bffc6..d61df5d 100644
+--- a/mm/page_cgroup.c
++++ b/mm/page_cgroup.c
+@@ -17,6 +17,7 @@ static void __meminit init_page_cgroup(struct page_cgroup *pc, unsigned long id)
+ set_page_cgroup_array_id(pc, id);
+ pc->mem_cgroup = NULL;
+ INIT_LIST_HEAD(&pc->lru);
++ page_cgroup_lock_init(pc);
+ }
+ static unsigned long total_usage;
+
+diff --git a/mm/slab.c b/mm/slab.c
+index d96e223..015cd76 100644
+--- a/mm/slab.c
++++ b/mm/slab.c
+@@ -116,6 +116,7 @@
+ #include <linux/kmemcheck.h>
+ #include <linux/memory.h>
+ #include <linux/prefetch.h>
++#include <linux/locallock.h>
+
+ #include <asm/cacheflush.h>
+ #include <asm/tlbflush.h>
+@@ -620,6 +621,51 @@ int slab_is_available(void)
+ static struct lock_class_key on_slab_l3_key;
+ static struct lock_class_key on_slab_alc_key;
+
++static struct lock_class_key debugobj_l3_key;
++static struct lock_class_key debugobj_alc_key;
+
-+unsigned long __lockfunc rt_read_lock_irqsave(rwlock_t *rwlock)
++static void slab_set_lock_classes(struct kmem_cache *cachep,
++ struct lock_class_key *l3_key, struct lock_class_key *alc_key,
++ int q)
+{
-+ rt_read_lock(rwlock);
++ struct array_cache **alc;
++ struct kmem_list3 *l3;
++ int r;
+
-+ return 0;
-+}
-+EXPORT_SYMBOL(rt_read_lock_irqsave);
++ l3 = cachep->nodelists[q];
++ if (!l3)
++ return;
+
-+void __rt_rwlock_init(rwlock_t *rwlock, char *name, struct lock_class_key *key)
-+{
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
++ lockdep_set_class(&l3->list_lock, l3_key);
++ alc = l3->alien;
+ /*
-+ * Make sure we are not reinitializing a held lock:
++ * FIXME: This check for BAD_ALIEN_MAGIC
++ * should go away when common slab code is taught to
++ * work even without alien caches.
++ * Currently, non NUMA code returns BAD_ALIEN_MAGIC
++ * for alloc_alien_cache,
+ */
-+ debug_check_no_locks_freed((void *)rwlock, sizeof(*rwlock));
-+ lockdep_init_map(&rwlock->dep_map, name, key, 0);
-+#endif
-+ rwlock->lock.save_state = 1;
-+ rwlock->read_depth = 0;
++ if (!alc || (unsigned long)alc == BAD_ALIEN_MAGIC)
++ return;
++ for_each_node(r) {
++ if (alc[r])
++ lockdep_set_class(&alc[r]->lock, alc_key);
++ }
+}
-+EXPORT_SYMBOL(__rt_rwlock_init);
-+
-+/*
-+ * rw_semaphores
-+ */
+
-+void rt_up_write(struct rw_semaphore *rwsem)
++static void slab_set_debugobj_lock_classes_node(struct kmem_cache *cachep, int node)
+{
-+ rwsem_release(&rwsem->dep_map, 1, _RET_IP_);
-+ rt_mutex_unlock(&rwsem->lock);
++ slab_set_lock_classes(cachep, &debugobj_l3_key, &debugobj_alc_key, node);
+}
-+EXPORT_SYMBOL(rt_up_write);
+
-+void rt_up_read(struct rw_semaphore *rwsem)
++static void slab_set_debugobj_lock_classes(struct kmem_cache *cachep)
+{
-+ rwsem_release(&rwsem->dep_map, 1, _RET_IP_);
-+ if (--rwsem->read_depth == 0)
-+ rt_mutex_unlock(&rwsem->lock);
-+}
-+EXPORT_SYMBOL(rt_up_read);
++ int node;
+
-+/*
-+ * downgrade a write lock into a read lock
-+ * - just wake up any readers at the front of the queue
-+ */
-+void rt_downgrade_write(struct rw_semaphore *rwsem)
-+{
-+ BUG_ON(rt_mutex_owner(&rwsem->lock) != current);
-+ rwsem->read_depth = 1;
++ for_each_online_node(node)
++ slab_set_debugobj_lock_classes_node(cachep, node);
+}
-+EXPORT_SYMBOL(rt_downgrade_write);
+
-+int rt_down_write_trylock(struct rw_semaphore *rwsem)
-+{
-+ int ret = rt_mutex_trylock(&rwsem->lock);
-+
-+ if (ret)
-+ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_);
-+ return ret;
-+}
-+EXPORT_SYMBOL(rt_down_write_trylock);
+ static void init_node_lock_keys(int q)
+ {
+ struct cache_sizes *s = malloc_sizes;
+@@ -628,29 +674,14 @@ static void init_node_lock_keys(int q)
+ return;
+
+ for (s = malloc_sizes; s->cs_size != ULONG_MAX; s++) {
+- struct array_cache **alc;
+ struct kmem_list3 *l3;
+- int r;
+
+ l3 = s->cs_cachep->nodelists[q];
+ if (!l3 || OFF_SLAB(s->cs_cachep))
+ continue;
+- lockdep_set_class(&l3->list_lock, &on_slab_l3_key);
+- alc = l3->alien;
+- /*
+- * FIXME: This check for BAD_ALIEN_MAGIC
+- * should go away when common slab code is taught to
+- * work even without alien caches.
+- * Currently, non NUMA code returns BAD_ALIEN_MAGIC
+- * for alloc_alien_cache,
+- */
+- if (!alc || (unsigned long)alc == BAD_ALIEN_MAGIC)
+- continue;
+- for_each_node(r) {
+- if (alc[r])
+- lockdep_set_class(&alc[r]->lock,
+- &on_slab_alc_key);
+- }
+
-+void rt_down_write(struct rw_semaphore *rwsem)
-+{
-+ rwsem_acquire(&rwsem->dep_map, 0, 0, _RET_IP_);
-+ rt_mutex_lock(&rwsem->lock);
-+}
-+EXPORT_SYMBOL(rt_down_write);
++ slab_set_lock_classes(s->cs_cachep, &on_slab_l3_key,
++ &on_slab_alc_key, q);
+ }
+ }
+
+@@ -669,6 +700,14 @@ static void init_node_lock_keys(int q)
+ static inline void init_lock_keys(void)
+ {
+ }
+
-+void rt_down_write_nested(struct rw_semaphore *rwsem, int subclass)
++static void slab_set_debugobj_lock_classes_node(struct kmem_cache *cachep, int node)
+{
-+ rwsem_acquire(&rwsem->dep_map, subclass, 0, _RET_IP_);
-+ rt_mutex_lock(&rwsem->lock);
+}
-+EXPORT_SYMBOL(rt_down_write_nested);
+
-+int rt_down_read_trylock(struct rw_semaphore *rwsem)
++static void slab_set_debugobj_lock_classes(struct kmem_cache *cachep)
+{
-+ struct rt_mutex *lock = &rwsem->lock;
-+ int ret = 1;
-+
-+ /*
-+ * recursive read locks succeed when current owns the rwsem,
-+ * but not when read_depth == 0 which means that the rwsem is
-+ * write locked.
-+ */
-+ if (rt_mutex_owner(lock) != current)
-+ ret = rt_mutex_trylock(&rwsem->lock);
-+ else if (!rwsem->read_depth)
-+ ret = 0;
-+
-+ if (ret) {
-+ rwsem->read_depth++;
-+ rwsem_acquire(&rwsem->dep_map, 0, 1, _RET_IP_);
-+ }
-+ return ret;
+}
-+EXPORT_SYMBOL(rt_down_read_trylock);
+ #endif
+
+ /*
+@@ -678,12 +717,66 @@ static DEFINE_MUTEX(cache_chain_mutex);
+ static struct list_head cache_chain;
+
+ static DEFINE_PER_CPU(struct delayed_work, slab_reap_work);
++static DEFINE_PER_CPU(struct list_head, slab_free_list);
++static DEFINE_LOCAL_IRQ_LOCK(slab_lock);
+
-+static void __rt_down_read(struct rw_semaphore *rwsem, int subclass)
++#ifndef CONFIG_PREEMPT_RT_BASE
++# define slab_on_each_cpu(func, cp) on_each_cpu(func, cp, 1)
++#else
++/*
++ * execute func() for all CPUs. On PREEMPT_RT we dont actually have
++ * to run on the remote CPUs - we only have to take their CPU-locks.
++ * (This is a rare operation, so cacheline bouncing is not an issue.)
++ */
++static void
++slab_on_each_cpu(void (*func)(void *arg, int this_cpu), void *arg)
+{
-+ struct rt_mutex *lock = &rwsem->lock;
-+
-+ rwsem_acquire_read(&rwsem->dep_map, subclass, 0, _RET_IP_);
++ unsigned int i;
+
-+ if (rt_mutex_owner(lock) != current)
-+ rt_mutex_lock(&rwsem->lock);
-+ rwsem->read_depth++;
++ for_each_online_cpu(i)
++ func(arg, i);
+}
++#endif
+
-+void rt_down_read(struct rw_semaphore *rwsem)
++static void free_delayed(struct list_head *h)
+{
-+ __rt_down_read(rwsem, 0);
++ while(!list_empty(h)) {
++ struct page *page = list_first_entry(h, struct page, lru);
++
++ list_del(&page->lru);
++ __free_pages(page, page->index);
++ }
+}
-+EXPORT_SYMBOL(rt_down_read);
+
-+void rt_down_read_nested(struct rw_semaphore *rwsem, int subclass)
++static void unlock_l3_and_free_delayed(spinlock_t *list_lock)
+{
-+ __rt_down_read(rwsem, subclass);
++ LIST_HEAD(tmp);
++
++ list_splice_init(&__get_cpu_var(slab_free_list), &tmp);
++ local_spin_unlock_irq(slab_lock, list_lock);
++ free_delayed(&tmp);
+}
-+EXPORT_SYMBOL(rt_down_read_nested);
+
-+void __rt_rwsem_init(struct rw_semaphore *rwsem, char *name,
-+ struct lock_class_key *key)
++static void unlock_slab_and_free_delayed(unsigned long flags)
+{
-+#ifdef CONFIG_DEBUG_LOCK_ALLOC
-+ /*
-+ * Make sure we are not reinitializing a held lock:
-+ */
-+ debug_check_no_locks_freed((void *)rwsem, sizeof(*rwsem));
-+ lockdep_init_map(&rwsem->dep_map, name, key, 0);
-+#endif
-+ rwsem->read_depth = 0;
-+ rwsem->lock.save_state = 0;
-+}
-+EXPORT_SYMBOL(__rt_rwsem_init);
++ LIST_HEAD(tmp);
+
-+/**
-+ * atomic_dec_and_mutex_lock - return holding mutex if we dec to 0
-+ * @cnt: the atomic which we are to dec
-+ * @lock: the mutex to return holding if we dec to 0
-+ *
-+ * return true and hold lock if we dec to 0, return false otherwise
-+ */
-+int atomic_dec_and_mutex_lock(atomic_t *cnt, struct mutex *lock)
++ list_splice_init(&__get_cpu_var(slab_free_list), &tmp);
++ local_unlock_irqrestore(slab_lock, flags);
++ free_delayed(&tmp);
++}
+
+ static inline struct array_cache *cpu_cache_get(struct kmem_cache *cachep)
+ {
+ return cachep->array[smp_processor_id()];
+ }
+
++static inline struct array_cache *cpu_cache_get_on_cpu(struct kmem_cache *cachep,
++ int cpu)
+{
-+ /* dec if we can't possibly hit 0 */
-+ if (atomic_add_unless(cnt, -1, 1))
-+ return 0;
-+ /* we might hit 0, so take the lock */
-+ mutex_lock(lock);
-+ if (!atomic_dec_and_test(cnt)) {
-+ /* when we actually did the dec, we didn't hit 0 */
-+ mutex_unlock(lock);
-+ return 0;
-+ }
-+ /* we hit 0, and we hold the lock */
-+ return 1;
++ return cachep->array[cpu];
+}
-+EXPORT_SYMBOL(atomic_dec_and_mutex_lock);
-Index: linux-3.0/kernel/spinlock.c
-===================================================================
---- linux-3.0.orig/kernel/spinlock.c
-+++ linux-3.0/kernel/spinlock.c
-@@ -110,8 +110,11 @@ void __lockfunc __raw_##op##_lock_bh(loc
- * __[spin|read|write]_lock_bh()
- */
- BUILD_LOCK_OPS(spin, raw_spinlock);
+
-+#ifndef CONFIG_PREEMPT_RT_FULL
- BUILD_LOCK_OPS(read, rwlock);
- BUILD_LOCK_OPS(write, rwlock);
-+#endif
+ static inline struct kmem_cache *__find_general_cachep(size_t size,
+ gfp_t gfpflags)
+ {
+@@ -1021,9 +1114,10 @@ static void reap_alien(struct kmem_cache *cachep, struct kmem_list3 *l3)
+ if (l3->alien) {
+ struct array_cache *ac = l3->alien[node];
- #endif
+- if (ac && ac->avail && spin_trylock_irq(&ac->lock)) {
++ if (ac && ac->avail &&
++ local_spin_trylock_irq(slab_lock, &ac->lock)) {
+ __drain_alien_cache(cachep, ac, node);
+- spin_unlock_irq(&ac->lock);
++ local_spin_unlock_irq(slab_lock, &ac->lock);
+ }
+ }
+ }
+@@ -1038,9 +1132,9 @@ static void drain_alien_cache(struct kmem_cache *cachep,
+ for_each_online_node(i) {
+ ac = alien[i];
+ if (ac) {
+- spin_lock_irqsave(&ac->lock, flags);
++ local_spin_lock_irqsave(slab_lock, &ac->lock, flags);
+ __drain_alien_cache(cachep, ac, i);
+- spin_unlock_irqrestore(&ac->lock, flags);
++ local_spin_unlock_irqrestore(slab_lock, &ac->lock, flags);
+ }
+ }
+ }
+@@ -1119,11 +1213,11 @@ static int init_cache_nodelists_node(int node)
+ cachep->nodelists[node] = l3;
+ }
-@@ -195,6 +198,8 @@ void __lockfunc _raw_spin_unlock_bh(raw_
- EXPORT_SYMBOL(_raw_spin_unlock_bh);
+- spin_lock_irq(&cachep->nodelists[node]->list_lock);
++ local_spin_lock_irq(slab_lock, &cachep->nodelists[node]->list_lock);
+ cachep->nodelists[node]->free_limit =
+ (1 + nr_cpus_node(node)) *
+ cachep->batchcount + cachep->num;
+- spin_unlock_irq(&cachep->nodelists[node]->list_lock);
++ local_spin_unlock_irq(slab_lock, &cachep->nodelists[node]->list_lock);
+ }
+ return 0;
+ }
+@@ -1148,7 +1242,7 @@ static void __cpuinit cpuup_canceled(long cpu)
+ if (!l3)
+ goto free_array_cache;
+
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
+
+ /* Free limit for this kmem_list3 */
+ l3->free_limit -= cachep->batchcount;
+@@ -1156,7 +1250,7 @@ static void __cpuinit cpuup_canceled(long cpu)
+ free_block(cachep, nc->entry, nc->avail, node);
+
+ if (!cpumask_empty(mask)) {
+- spin_unlock_irq(&l3->list_lock);
++ unlock_l3_and_free_delayed(&l3->list_lock);
+ goto free_array_cache;
+ }
+
+@@ -1170,7 +1264,7 @@ static void __cpuinit cpuup_canceled(long cpu)
+ alien = l3->alien;
+ l3->alien = NULL;
+
+- spin_unlock_irq(&l3->list_lock);
++ unlock_l3_and_free_delayed(&l3->list_lock);
+
+ kfree(shared);
+ if (alien) {
+@@ -1244,7 +1338,7 @@ static int __cpuinit cpuup_prepare(long cpu)
+ l3 = cachep->nodelists[node];
+ BUG_ON(!l3);
+
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
+ if (!l3->shared) {
+ /*
+ * We are serialised from CPU_DEAD or
+@@ -1259,9 +1353,11 @@ static int __cpuinit cpuup_prepare(long cpu)
+ alien = NULL;
+ }
#endif
+- spin_unlock_irq(&l3->list_lock);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
+ kfree(shared);
+ free_alien_cache(alien);
++ if (cachep->flags & SLAB_DEBUG_OBJECTS)
++ slab_set_debugobj_lock_classes_node(cachep, node);
+ }
+ init_node_lock_keys(node);
-+#ifndef CONFIG_PREEMPT_RT_FULL
+@@ -1448,6 +1544,10 @@ void __init kmem_cache_init(void)
+ if (num_possible_nodes() == 1)
+ use_alien_caches = 0;
+
++ local_irq_lock_init(slab_lock);
++ for_each_possible_cpu(i)
++ INIT_LIST_HEAD(&per_cpu(slab_free_list, i));
+
- #ifndef CONFIG_INLINE_READ_TRYLOCK
- int __lockfunc _raw_read_trylock(rwlock_t *lock)
+ for (i = 0; i < NUM_INIT_LISTS; i++) {
+ kmem_list3_init(&initkmem_list3[i]);
+ if (i < MAX_NUMNODES)
+@@ -1625,6 +1725,9 @@ void __init kmem_cache_init_late(void)
{
-@@ -339,6 +344,8 @@ void __lockfunc _raw_write_unlock_bh(rwl
- EXPORT_SYMBOL(_raw_write_unlock_bh);
- #endif
+ struct kmem_cache *cachep;
-+#endif /* !PREEMPT_RT_FULL */
++ /* Annotate slab for lockdep -- annotate the malloc caches */
++ init_lock_keys();
+
- #ifdef CONFIG_DEBUG_LOCK_ALLOC
+ /* 6) resize the head arrays to their final sizes */
+ mutex_lock(&cache_chain_mutex);
+ list_for_each_entry(cachep, &cache_chain, next)
+@@ -1635,9 +1738,6 @@ void __init kmem_cache_init_late(void)
+ /* Done! */
+ g_cpucache_up = FULL;
- void __lockfunc _raw_spin_lock_nested(raw_spinlock_t *lock, int subclass)
-Index: linux-3.0/lib/spinlock_debug.c
-===================================================================
---- linux-3.0.orig/lib/spinlock_debug.c
-+++ linux-3.0/lib/spinlock_debug.c
-@@ -31,6 +31,7 @@ void __raw_spin_lock_init(raw_spinlock_t
+- /* Annotate slab for lockdep -- annotate the malloc caches */
+- init_lock_keys();
+-
+ /*
+ * Register a cpu startup notifier callback that initializes
+ * cpu_cache_get for all new cpus
+@@ -1725,12 +1825,14 @@ static void *kmem_getpages(struct kmem_cache *cachep, gfp_t flags, int nodeid)
+ /*
+ * Interface to system's page release.
+ */
+-static void kmem_freepages(struct kmem_cache *cachep, void *addr)
++static void kmem_freepages(struct kmem_cache *cachep, void *addr, bool delayed)
+ {
+ unsigned long i = (1 << cachep->gfporder);
+- struct page *page = virt_to_page(addr);
++ struct page *page, *basepage = virt_to_page(addr);
+ const unsigned long nr_freed = i;
- EXPORT_SYMBOL(__raw_spin_lock_init);
++ page = basepage;
++
+ kmemcheck_free_shadow(page, cachep->gfporder);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- void __rwlock_init(rwlock_t *lock, const char *name,
- struct lock_class_key *key)
+ if (cachep->flags & SLAB_RECLAIM_ACCOUNT)
+@@ -1746,7 +1848,13 @@ static void kmem_freepages(struct kmem_cache *cachep, void *addr)
+ }
+ if (current->reclaim_state)
+ current->reclaim_state->reclaimed_slab += nr_freed;
+- free_pages((unsigned long)addr, cachep->gfporder);
++
++ if (!delayed) {
++ free_pages((unsigned long)addr, cachep->gfporder);
++ } else {
++ basepage->index = cachep->gfporder;
++ list_add(&basepage->lru, &__get_cpu_var(slab_free_list));
++ }
+ }
+
+ static void kmem_rcu_free(struct rcu_head *head)
+@@ -1754,7 +1862,7 @@ static void kmem_rcu_free(struct rcu_head *head)
+ struct slab_rcu *slab_rcu = (struct slab_rcu *)head;
+ struct kmem_cache *cachep = slab_rcu->cachep;
+
+- kmem_freepages(cachep, slab_rcu->addr);
++ kmem_freepages(cachep, slab_rcu->addr, false);
+ if (OFF_SLAB(cachep))
+ kmem_cache_free(cachep->slabp_cache, slab_rcu);
+ }
+@@ -1973,7 +2081,8 @@ static void slab_destroy_debugcheck(struct kmem_cache *cachep, struct slab *slab
+ * Before calling the slab must have been unlinked from the cache. The
+ * cache-lock is not held/needed.
+ */
+-static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp)
++static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp,
++ bool delayed)
{
-@@ -46,8 +47,8 @@ void __rwlock_init(rwlock_t *lock, const
- lock->owner = SPINLOCK_OWNER_INIT;
- lock->owner_cpu = -1;
- }
--
- EXPORT_SYMBOL(__rwlock_init);
-+#endif
+ void *addr = slabp->s_mem - slabp->colouroff;
- static void spin_bug(raw_spinlock_t *lock, const char *msg)
- {
-@@ -154,6 +155,7 @@ void do_raw_spin_unlock(raw_spinlock_t *
- arch_spin_unlock(&lock->raw_lock);
- }
+@@ -1986,7 +2095,7 @@ static void slab_destroy(struct kmem_cache *cachep, struct slab *slabp)
+ slab_rcu->addr = addr;
+ call_rcu(&slab_rcu->head, kmem_rcu_free);
+ } else {
+- kmem_freepages(cachep, addr);
++ kmem_freepages(cachep, addr, delayed);
+ if (OFF_SLAB(cachep))
+ kmem_cache_free(cachep->slabp_cache, slabp);
+ }
+@@ -2424,6 +2533,16 @@ kmem_cache_create (const char *name, size_t size, size_t align,
+ goto oops;
+ }
-+#ifndef CONFIG_PREEMPT_RT_FULL
- static void rwlock_bug(rwlock_t *lock, const char *msg)
++ if (flags & SLAB_DEBUG_OBJECTS) {
++ /*
++ * Would deadlock through slab_destroy()->call_rcu()->
++ * debug_object_activate()->kmem_cache_alloc().
++ */
++ WARN_ON_ONCE(flags & SLAB_DESTROY_BY_RCU);
++
++ slab_set_debugobj_lock_classes(cachep);
++ }
++
+ /* cache setup completed, link it into the list */
+ list_add(&cachep->next, &cache_chain);
+ oops:
+@@ -2441,7 +2560,7 @@ EXPORT_SYMBOL(kmem_cache_create);
+ #if DEBUG
+ static void check_irq_off(void)
{
- if (!debug_locks_off())
-@@ -295,3 +297,5 @@ void do_raw_write_unlock(rwlock_t *lock)
- debug_write_unlock(lock);
- arch_write_unlock(&lock->raw_lock);
+- BUG_ON(!irqs_disabled());
++ BUG_ON_NONRT(!irqs_disabled());
}
-+
-+#endif
-Index: linux-3.0/include/linux/rcutree.h
-===================================================================
---- linux-3.0.orig/include/linux/rcutree.h
-+++ linux-3.0/include/linux/rcutree.h
-@@ -57,7 +57,11 @@ static inline void exit_rcu(void)
-
- #endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
-+#ifndef CONFIG_PREEMPT_RT_FULL
- extern void synchronize_rcu_bh(void);
-+#else
-+# define synchronize_rcu_bh() synchronize_rcu()
-+#endif
- extern void synchronize_sched_expedited(void);
- extern void synchronize_rcu_expedited(void);
+ static void check_irq_on(void)
+@@ -2476,26 +2595,43 @@ static void drain_array(struct kmem_cache *cachep, struct kmem_list3 *l3,
+ struct array_cache *ac,
+ int force, int node);
-@@ -71,13 +75,19 @@ extern void rcu_barrier(void);
- extern unsigned long rcutorture_testseq;
- extern unsigned long rcutorture_vernum;
- extern long rcu_batches_completed(void);
--extern long rcu_batches_completed_bh(void);
- extern long rcu_batches_completed_sched(void);
+-static void do_drain(void *arg)
++static void __do_drain(void *arg, unsigned int cpu)
+ {
+ struct kmem_cache *cachep = arg;
+ struct array_cache *ac;
+- int node = numa_mem_id();
++ int node = cpu_to_mem(cpu);
- extern void rcu_force_quiescent_state(void);
--extern void rcu_bh_force_quiescent_state(void);
- extern void rcu_sched_force_quiescent_state(void);
+- check_irq_off();
+- ac = cpu_cache_get(cachep);
++ ac = cpu_cache_get_on_cpu(cachep, cpu);
+ spin_lock(&cachep->nodelists[node]->list_lock);
+ free_block(cachep, ac->entry, ac->avail, node);
+ spin_unlock(&cachep->nodelists[node]->list_lock);
+ ac->avail = 0;
+ }
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+extern void rcu_bh_force_quiescent_state(void);
-+extern long rcu_batches_completed_bh(void);
++#ifndef CONFIG_PREEMPT_RT_BASE
++static void do_drain(void *arg)
++{
++ __do_drain(arg, smp_processor_id());
++}
+#else
-+# define rcu_bh_force_quiescent_state rcu_force_quiescent_state
-+# define rcu_batches_completed_bh rcu_batches_completed
++static void do_drain(void *arg, int cpu)
++{
++ LIST_HEAD(tmp);
++
++ spin_lock_irq(&per_cpu(slab_lock, cpu).lock);
++ __do_drain(arg, cpu);
++ list_splice_init(&per_cpu(slab_free_list, cpu), &tmp);
++ spin_unlock_irq(&per_cpu(slab_lock, cpu).lock);
++ free_delayed(&tmp);
++}
+#endif
+
- /* A context switch is a grace period for RCU-sched and RCU-bh. */
- static inline int rcu_blocking_is_gp(void)
+ static void drain_cpu_caches(struct kmem_cache *cachep)
{
-Index: linux-3.0/kernel/rcupdate.c
-===================================================================
---- linux-3.0.orig/kernel/rcupdate.c
-+++ linux-3.0/kernel/rcupdate.c
-@@ -72,6 +72,7 @@ int debug_lockdep_rcu_enabled(void)
- }
- EXPORT_SYMBOL_GPL(debug_lockdep_rcu_enabled);
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- /**
- * rcu_read_lock_bh_held() - might we be in RCU-bh read-side critical section?
- *
-@@ -91,6 +92,7 @@ int rcu_read_lock_bh_held(void)
- return in_softirq() || irqs_disabled();
- }
- EXPORT_SYMBOL_GPL(rcu_read_lock_bh_held);
-+#endif
+ struct kmem_list3 *l3;
+ int node;
- #endif /* #ifdef CONFIG_DEBUG_LOCK_ALLOC */
+- on_each_cpu(do_drain, cachep, 1);
++ slab_on_each_cpu(do_drain, cachep);
+ check_irq_on();
+ for_each_online_node(node) {
+ l3 = cachep->nodelists[node];
+@@ -2526,10 +2662,10 @@ static int drain_freelist(struct kmem_cache *cache,
+ nr_freed = 0;
+ while (nr_freed < tofree && !list_empty(&l3->slabs_free)) {
-Index: linux-3.0/include/linux/lglock.h
-===================================================================
---- linux-3.0.orig/include/linux/lglock.h
-+++ linux-3.0/include/linux/lglock.h
-@@ -70,6 +70,9 @@
- extern void name##_global_lock_online(void); \
- extern void name##_global_unlock_online(void); \
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
+ p = l3->slabs_free.prev;
+ if (p == &l3->slabs_free) {
+- spin_unlock_irq(&l3->list_lock);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
+ goto out;
+ }
-+
-+#ifndef CONFIG_PREEMPT_RT_FULL
-+
- #define DEFINE_LGLOCK(name) \
- \
- DEFINE_PER_CPU(arch_spinlock_t, name##_lock); \
-@@ -169,4 +172,101 @@
- preempt_enable(); \
- } \
- EXPORT_SYMBOL(name##_global_unlock);
-+
-+#else /* !PREEMPT_RT_FULL */
-+#define DEFINE_LGLOCK(name) \
-+ \
-+ DEFINE_PER_CPU(struct rt_mutex, name##_lock); \
-+ DEFINE_LGLOCK_LOCKDEP(name); \
-+ \
-+ void name##_lock_init(void) { \
-+ int i; \
-+ LOCKDEP_INIT_MAP(&name##_lock_dep_map, #name, &name##_lock_key, 0); \
-+ for_each_possible_cpu(i) { \
-+ struct rt_mutex *lock; \
-+ lock = &per_cpu(name##_lock, i); \
-+ rt_mutex_init(lock); \
-+ } \
-+ } \
-+ EXPORT_SYMBOL(name##_lock_init); \
-+ \
-+ void name##_local_lock(void) { \
-+ struct rt_mutex *lock; \
-+ migrate_disable(); \
-+ rwlock_acquire_read(&name##_lock_dep_map, 0, 0, _THIS_IP_); \
-+ lock = &__get_cpu_var(name##_lock); \
-+ __rt_spin_lock(lock); \
-+ } \
-+ EXPORT_SYMBOL(name##_local_lock); \
-+ \
-+ void name##_local_unlock(void) { \
-+ struct rt_mutex *lock; \
-+ rwlock_release(&name##_lock_dep_map, 1, _THIS_IP_); \
-+ lock = &__get_cpu_var(name##_lock); \
-+ __rt_spin_unlock(lock); \
-+ migrate_enable(); \
-+ } \
-+ EXPORT_SYMBOL(name##_local_unlock); \
-+ \
-+ void name##_local_lock_cpu(int cpu) { \
-+ struct rt_mutex *lock; \
-+ rwlock_acquire_read(&name##_lock_dep_map, 0, 0, _THIS_IP_); \
-+ lock = &per_cpu(name##_lock, cpu); \
-+ __rt_spin_lock(lock); \
-+ } \
-+ EXPORT_SYMBOL(name##_local_lock_cpu); \
-+ \
-+ void name##_local_unlock_cpu(int cpu) { \
-+ struct rt_mutex *lock; \
-+ rwlock_release(&name##_lock_dep_map, 1, _THIS_IP_); \
-+ lock = &per_cpu(name##_lock, cpu); \
-+ __rt_spin_unlock(lock); \
-+ } \
-+ EXPORT_SYMBOL(name##_local_unlock_cpu); \
-+ \
-+ void name##_global_lock_online(void) { \
-+ int i; \
-+ rwlock_acquire(&name##_lock_dep_map, 0, 0, _RET_IP_); \
-+ for_each_online_cpu(i) { \
-+ struct rt_mutex *lock; \
-+ lock = &per_cpu(name##_lock, i); \
-+ __rt_spin_lock(lock); \
-+ } \
-+ } \
-+ EXPORT_SYMBOL(name##_global_lock_online); \
-+ \
-+ void name##_global_unlock_online(void) { \
-+ int i; \
-+ rwlock_release(&name##_lock_dep_map, 1, _RET_IP_); \
-+ for_each_online_cpu(i) { \
-+ struct rt_mutex *lock; \
-+ lock = &per_cpu(name##_lock, i); \
-+ __rt_spin_unlock(lock); \
-+ } \
-+ } \
-+ EXPORT_SYMBOL(name##_global_unlock_online); \
-+ \
-+ void name##_global_lock(void) { \
-+ int i; \
-+ rwlock_acquire(&name##_lock_dep_map, 0, 0, _RET_IP_); \
-+ for_each_possible_cpu(i) { \
-+ struct rt_mutex *lock; \
-+ lock = &per_cpu(name##_lock, i); \
-+ __rt_spin_lock(lock); \
-+ } \
-+ } \
-+ EXPORT_SYMBOL(name##_global_lock); \
-+ \
-+ void name##_global_unlock(void) { \
-+ int i; \
-+ rwlock_release(&name##_lock_dep_map, 1, _RET_IP_); \
-+ for_each_possible_cpu(i) { \
-+ struct rt_mutex *lock; \
-+ lock = &per_cpu(name##_lock, i); \
-+ __rt_spin_unlock(lock); \
-+ } \
-+ } \
-+ EXPORT_SYMBOL(name##_global_unlock);
-+#endif /* PRREMPT_RT_FULL */
-+
- #endif
-Index: linux-3.0/drivers/tty/serial/8250.c
-===================================================================
---- linux-3.0.orig/drivers/tty/serial/8250.c
-+++ linux-3.0/drivers/tty/serial/8250.c
-@@ -38,6 +38,7 @@
- #include <linux/nmi.h>
- #include <linux/mutex.h>
- #include <linux/slab.h>
-+#include <linux/kdb.h>
+@@ -2543,8 +2679,8 @@ static int drain_freelist(struct kmem_cache *cache,
+ * to the cache.
+ */
+ l3->free_objects -= cache->num;
+- spin_unlock_irq(&l3->list_lock);
+- slab_destroy(cache, slabp);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
++ slab_destroy(cache, slabp, false);
+ nr_freed++;
+ }
+ out:
+@@ -2838,7 +2974,7 @@ static int cache_grow(struct kmem_cache *cachep,
+ offset *= cachep->colour_off;
- #include <asm/io.h>
- #include <asm/irq.h>
-@@ -1678,12 +1679,14 @@ static irqreturn_t serial8250_interrupt(
+ if (local_flags & __GFP_WAIT)
+- local_irq_enable();
++ local_unlock_irq(slab_lock);
- l = l->next;
+ /*
+ * The test for missing atomic flag is performed here, rather than
+@@ -2868,7 +3004,7 @@ static int cache_grow(struct kmem_cache *cachep,
+ cache_init_objs(cachep, slabp);
-+#ifndef CONFIG_PREEMPT_RT_FULL
- if (l == i->head && pass_counter++ > PASS_LIMIT) {
- /* If we hit this, we're dead. */
- printk_ratelimited(KERN_ERR
- "serial8250: too much work for irq%d\n", irq);
- break;
- }
-+#endif
- } while (l != end);
+ if (local_flags & __GFP_WAIT)
+- local_irq_disable();
++ local_lock_irq(slab_lock);
+ check_irq_off();
+ spin_lock(&l3->list_lock);
- spin_unlock(&i->lock);
-@@ -2894,14 +2897,14 @@ serial8250_console_write(struct console
+@@ -2879,10 +3015,10 @@ static int cache_grow(struct kmem_cache *cachep,
+ spin_unlock(&l3->list_lock);
+ return 1;
+ opps1:
+- kmem_freepages(cachep, objp);
++ kmem_freepages(cachep, objp, false);
+ failed:
+ if (local_flags & __GFP_WAIT)
+- local_irq_disable();
++ local_lock_irq(slab_lock);
+ return 0;
+ }
- touch_nmi_watchdog();
+@@ -3280,11 +3416,11 @@ retry:
+ * set and go into memory reserves if necessary.
+ */
+ if (local_flags & __GFP_WAIT)
+- local_irq_enable();
++ local_unlock_irq(slab_lock);
+ kmem_flagcheck(cache, flags);
+ obj = kmem_getpages(cache, local_flags, numa_mem_id());
+ if (local_flags & __GFP_WAIT)
+- local_irq_disable();
++ local_lock_irq(slab_lock);
+ if (obj) {
+ /*
+ * Insert into the appropriate per node queues
+@@ -3400,7 +3536,7 @@ __cache_alloc_node(struct kmem_cache *cachep, gfp_t flags, int nodeid,
+ return NULL;
+
+ cache_alloc_debugcheck_before(cachep, flags);
+- local_irq_save(save_flags);
++ local_lock_irqsave(slab_lock, save_flags);
+
+ if (nodeid == -1)
+ nodeid = slab_node;
+@@ -3425,7 +3561,7 @@ __cache_alloc_node(struct kmem_cache *cachep, gfp_t flags, int nodeid,
+ /* ___cache_alloc_node can fall back to other nodes */
+ ptr = ____cache_alloc_node(cachep, flags, nodeid);
+ out:
+- local_irq_restore(save_flags);
++ local_unlock_irqrestore(slab_lock, save_flags);
+ ptr = cache_alloc_debugcheck_after(cachep, flags, ptr, caller);
+ kmemleak_alloc_recursive(ptr, obj_size(cachep), 1, cachep->flags,
+ flags);
+@@ -3485,9 +3621,9 @@ __cache_alloc(struct kmem_cache *cachep, gfp_t flags, void *caller)
+ return NULL;
+
+ cache_alloc_debugcheck_before(cachep, flags);
+- local_irq_save(save_flags);
++ local_lock_irqsave(slab_lock, save_flags);
+ objp = __do_cache_alloc(cachep, flags);
+- local_irq_restore(save_flags);
++ local_unlock_irqrestore(slab_lock, save_flags);
+ objp = cache_alloc_debugcheck_after(cachep, flags, objp, caller);
+ kmemleak_alloc_recursive(objp, obj_size(cachep), 1, cachep->flags,
+ flags);
+@@ -3535,7 +3671,7 @@ static void free_block(struct kmem_cache *cachep, void **objpp, int nr_objects,
+ * a different cache, refer to comments before
+ * alloc_slabmgmt.
+ */
+- slab_destroy(cachep, slabp);
++ slab_destroy(cachep, slabp, true);
+ } else {
+ list_add(&slabp->list, &l3->slabs_free);
+ }
+@@ -3798,12 +3934,12 @@ void kmem_cache_free(struct kmem_cache *cachep, void *objp)
+ {
+ unsigned long flags;
- local_irq_save(flags);
-- if (up->port.sysrq) {
-- /* serial8250_handle_port() already took the lock */
-+ if (unlikely(in_kdb_printk())) {
- locked = 0;
-- } else if (oops_in_progress) {
-- locked = spin_trylock(&up->port.lock);
-- } else
-- spin_lock(&up->port.lock);
-+ } else {
-+ if (up->port.sysrq || oops_in_progress)
-+ locked = spin_trylock_irqsave(&up->port.lock, flags);
-+ else
-+ spin_lock_irqsave(&up->port.lock, flags);
-+ }
+ debug_check_no_locks_freed(objp, obj_size(cachep));
+ if (!(cachep->flags & SLAB_DEBUG_OBJECTS))
+ debug_check_no_obj_freed(objp, obj_size(cachep));
++ local_lock_irqsave(slab_lock, flags);
+ __cache_free(cachep, objp, __builtin_return_address(0));
+- local_irq_restore(flags);
++ unlock_slab_and_free_delayed(flags);
- /*
- * First save the IER then disable the interrupts
-@@ -2933,8 +2936,7 @@ serial8250_console_write(struct console
- check_modem_status(up);
+ trace_kmem_cache_free(_RET_IP_, objp);
+ }
+@@ -3827,13 +3963,13 @@ void kfree(const void *objp)
- if (locked)
-- spin_unlock(&up->port.lock);
+ if (unlikely(ZERO_OR_NULL_PTR(objp)))
+ return;
+- local_irq_save(flags);
+ kfree_debugcheck(objp);
+ c = virt_to_cache(objp);
+ debug_check_no_locks_freed(objp, obj_size(c));
+ debug_check_no_obj_freed(objp, obj_size(c));
++ local_lock_irqsave(slab_lock, flags);
+ __cache_free(c, (void *)objp, __builtin_return_address(0));
- local_irq_restore(flags);
-+ spin_unlock_irqrestore(&up->port.lock, flags);
++ unlock_slab_and_free_delayed(flags);
}
+ EXPORT_SYMBOL(kfree);
- static int __init serial8250_console_setup(struct console *co, char *options)
-Index: linux-3.0/drivers/tty/tty_buffer.c
-===================================================================
---- linux-3.0.orig/drivers/tty/tty_buffer.c
-+++ linux-3.0/drivers/tty/tty_buffer.c
-@@ -493,10 +493,14 @@ void tty_flip_buffer_push(struct tty_str
- tty->buf.tail->commit = tty->buf.tail->used;
- spin_unlock_irqrestore(&tty->buf.lock, flags);
+@@ -3876,7 +4012,7 @@ static int alloc_kmemlist(struct kmem_cache *cachep, gfp_t gfp)
+ if (l3) {
+ struct array_cache *shared = l3->shared;
-+#ifndef CONFIG_PREEMPT_RT_FULL
- if (tty->low_latency)
- flush_to_ldisc(&tty->buf.work);
- else
- schedule_work(&tty->buf.work);
-+#else
-+ flush_to_ldisc(&tty->buf.work);
-+#endif
- }
- EXPORT_SYMBOL(tty_flip_buffer_push);
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
-Index: linux-3.0/drivers/tty/serial/omap-serial.c
-===================================================================
---- linux-3.0.orig/drivers/tty/serial/omap-serial.c
-+++ linux-3.0/drivers/tty/serial/omap-serial.c
-@@ -946,13 +946,12 @@ serial_omap_console_write(struct console
- unsigned int ier;
- int locked = 1;
+ if (shared)
+ free_block(cachep, shared->entry,
+@@ -3889,7 +4025,8 @@ static int alloc_kmemlist(struct kmem_cache *cachep, gfp_t gfp)
+ }
+ l3->free_limit = (1 + nr_cpus_node(node)) *
+ cachep->batchcount + cachep->num;
+- spin_unlock_irq(&l3->list_lock);
++ unlock_l3_and_free_delayed(&l3->list_lock);
++
+ kfree(shared);
+ free_alien_cache(new_alien);
+ continue;
+@@ -3936,17 +4073,30 @@ struct ccupdate_struct {
+ struct array_cache *new[NR_CPUS];
+ };
-- local_irq_save(flags);
- if (up->port.sysrq)
- locked = 0;
- else if (oops_in_progress)
-- locked = spin_trylock(&up->port.lock);
-+ locked = spin_trylock_irqsave(&up->port.lock, flags);
- else
-- spin_lock(&up->port.lock);
-+ spin_lock_irqsave(&up->port.lock, flags);
+-static void do_ccupdate_local(void *info)
++static void __do_ccupdate_local(void *info, int cpu)
+ {
+ struct ccupdate_struct *new = info;
+ struct array_cache *old;
- /*
- * First save the IER then disable the interrupts
-@@ -979,8 +978,7 @@ serial_omap_console_write(struct console
- check_modem_status(up);
+- check_irq_off();
+- old = cpu_cache_get(new->cachep);
++ old = cpu_cache_get_on_cpu(new->cachep, cpu);
- if (locked)
-- spin_unlock(&up->port.lock);
-- local_irq_restore(flags);
-+ spin_unlock_irqrestore(&up->port.lock, flags);
+- new->cachep->array[smp_processor_id()] = new->new[smp_processor_id()];
+- new->new[smp_processor_id()] = old;
++ new->cachep->array[cpu] = new->new[cpu];
++ new->new[cpu] = old;
++}
++
++#ifndef CONFIG_PREEMPT_RT_BASE
++static void do_ccupdate_local(void *info)
++{
++ __do_ccupdate_local(info, smp_processor_id());
++}
++#else
++static void do_ccupdate_local(void *info, int cpu)
++{
++ spin_lock_irq(&per_cpu(slab_lock, cpu).lock);
++ __do_ccupdate_local(info, cpu);
++ spin_unlock_irq(&per_cpu(slab_lock, cpu).lock);
}
++#endif
- static int __init
-Index: linux-3.0/fs/namespace.c
-===================================================================
---- linux-3.0.orig/fs/namespace.c
-+++ linux-3.0/fs/namespace.c
-@@ -341,8 +341,14 @@ int mnt_want_write(struct vfsmount *mnt)
- * incremented count after it has set MNT_WRITE_HOLD.
- */
- smp_mb();
-- while (mnt->mnt_flags & MNT_WRITE_HOLD)
-+ /*
-+ * No need to keep preemption disabled accross the spin loop.
-+ */
-+ while (mnt->mnt_flags & MNT_WRITE_HOLD) {
-+ preempt_enable();
- cpu_relax();
-+ preempt_disable();
-+ }
- /*
- * After the slowpath clears MNT_WRITE_HOLD, mnt_is_readonly will
- * be set to match its requirements. So we must not load that until
-@@ -352,9 +358,7 @@ int mnt_want_write(struct vfsmount *mnt)
- if (__mnt_is_readonly(mnt)) {
- mnt_dec_writers(mnt);
- ret = -EROFS;
-- goto out;
+ /* Always called with the cache_chain_mutex held */
+ static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
+@@ -3971,7 +4121,7 @@ static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
+ }
+ new->cachep = cachep;
+
+- on_each_cpu(do_ccupdate_local, (void *)new, 1);
++ slab_on_each_cpu(do_ccupdate_local, (void *)new);
+
+ check_irq_on();
+ cachep->batchcount = batchcount;
+@@ -3982,9 +4132,11 @@ static int do_tune_cpucache(struct kmem_cache *cachep, int limit,
+ struct array_cache *ccold = new->new[i];
+ if (!ccold)
+ continue;
+- spin_lock_irq(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
++ local_spin_lock_irq(slab_lock,
++ &cachep->nodelists[cpu_to_mem(i)]->list_lock);
+ free_block(cachep, ccold->entry, ccold->avail, cpu_to_mem(i));
+- spin_unlock_irq(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
++
++ unlock_l3_and_free_delayed(&cachep->nodelists[cpu_to_mem(i)]->list_lock);
+ kfree(ccold);
}
--out:
- preempt_enable();
- return ret;
- }
-Index: linux-3.0/fs/btrfs/locking.c
-===================================================================
---- linux-3.0.orig/fs/btrfs/locking.c
-+++ linux-3.0/fs/btrfs/locking.c
-@@ -93,6 +93,7 @@ static int btrfs_spin_on_block(struct ex
- */
- int btrfs_try_spin_lock(struct extent_buffer *eb)
- {
-+#ifndef CONFIG_PREEMPT_RT_FULL
- int i;
-
- if (btrfs_spin_on_block(eb)) {
-@@ -112,6 +113,7 @@ int btrfs_try_spin_lock(struct extent_bu
- return 1;
- spin_unlock(&eb->lock);
+ kfree(new);
+@@ -4060,7 +4212,7 @@ static void drain_array(struct kmem_cache *cachep, struct kmem_list3 *l3,
+ if (ac->touched && !force) {
+ ac->touched = 0;
+ } else {
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
+ if (ac->avail) {
+ tofree = force ? ac->avail : (ac->limit + 4) / 5;
+ if (tofree > ac->avail)
+@@ -4070,7 +4222,7 @@ static void drain_array(struct kmem_cache *cachep, struct kmem_list3 *l3,
+ memmove(ac->entry, &(ac->entry[tofree]),
+ sizeof(void *) * ac->avail);
+ }
+- spin_unlock_irq(&l3->list_lock);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
}
-+#endif
- return 0;
}
-Index: linux-3.0/fs/exec.c
-===================================================================
---- linux-3.0.orig/fs/exec.c
-+++ linux-3.0/fs/exec.c
-@@ -844,6 +844,7 @@ static int exec_mmap(struct mm_struct *m
- }
- }
- task_lock(tsk);
-+ local_irq_disable_rt();
- active_mm = tsk->active_mm;
- tsk->mm = mm;
- tsk->active_mm = mm;
-@@ -852,6 +853,7 @@ static int exec_mmap(struct mm_struct *m
- atomic_dec(&old_mm->oom_disable_count);
- atomic_inc(&tsk->mm->oom_disable_count);
- }
-+ local_irq_enable_rt();
- task_unlock(tsk);
- arch_pick_mmap_layout(mm);
- if (old_mm) {
-Index: linux-3.0/mm/mmu_context.c
-===================================================================
---- linux-3.0.orig/mm/mmu_context.c
-+++ linux-3.0/mm/mmu_context.c
-@@ -26,6 +26,7 @@ void use_mm(struct mm_struct *mm)
- struct task_struct *tsk = current;
+@@ -4209,7 +4361,7 @@ static int s_show(struct seq_file *m, void *p)
+ continue;
- task_lock(tsk);
-+ local_irq_disable_rt();
- active_mm = tsk->active_mm;
- if (active_mm != mm) {
- atomic_inc(&mm->mm_count);
-@@ -33,6 +34,7 @@ void use_mm(struct mm_struct *mm)
- }
- tsk->mm = mm;
- switch_mm(active_mm, mm, tsk);
-+ local_irq_enable_rt();
- task_unlock(tsk);
+ check_irq_on();
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
- if (active_mm != mm)
-Index: linux-3.0/arch/x86/include/asm/stackprotector.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/stackprotector.h
-+++ linux-3.0/arch/x86/include/asm/stackprotector.h
-@@ -58,7 +58,7 @@
- */
- static __always_inline void boot_init_stack_canary(void)
- {
-- u64 canary;
-+ u64 uninitialized_var(canary);
- u64 tsc;
+ list_for_each_entry(slabp, &l3->slabs_full, list) {
+ if (slabp->inuse != cachep->num && !error)
+@@ -4234,7 +4386,7 @@ static int s_show(struct seq_file *m, void *p)
+ if (l3->shared)
+ shared_avail += l3->shared->avail;
- #ifdef CONFIG_X86_64
-@@ -69,8 +69,16 @@ static __always_inline void boot_init_st
- * of randomness. The TSC only matters for very early init,
- * there it already has some randomness on most systems. Later
- * on during the bootup the random pool has true entropy too.
-+ *
-+ * For preempt-rt we need to weaken the randomness a bit, as
-+ * we can't call into the random generator from atomic context
-+ * due to locking constraints. We just leave canary
-+ * uninitialized and use the TSC based randomness on top of
-+ * it.
- */
-+#ifndef CONFIG_PREEMPT_RT_FULL
- get_random_bytes(&canary, sizeof(canary));
-+#endif
- tsc = __native_read_tsc();
- canary += tsc + (tsc << 32UL);
+- spin_unlock_irq(&l3->list_lock);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
+ }
+ num_slabs += active_slabs;
+ num_objs = num_slabs * cachep->num;
+@@ -4463,13 +4615,13 @@ static int leaks_show(struct seq_file *m, void *p)
+ continue;
-Index: linux-3.0/arch/x86/Kconfig
-===================================================================
---- linux-3.0.orig/arch/x86/Kconfig
-+++ linux-3.0/arch/x86/Kconfig
-@@ -152,10 +152,10 @@ config ARCH_MAY_HAVE_PC_FDC
- def_bool ISA_DMA_API
+ check_irq_on();
+- spin_lock_irq(&l3->list_lock);
++ local_spin_lock_irq(slab_lock, &l3->list_lock);
- config RWSEM_GENERIC_SPINLOCK
-- def_bool !X86_XADD
-+ def_bool !X86_XADD || PREEMPT_RT_FULL
+ list_for_each_entry(slabp, &l3->slabs_full, list)
+ handle_slab(n, cachep, slabp);
+ list_for_each_entry(slabp, &l3->slabs_partial, list)
+ handle_slab(n, cachep, slabp);
+- spin_unlock_irq(&l3->list_lock);
++ local_spin_unlock_irq(slab_lock, &l3->list_lock);
+ }
+ name = cachep->name;
+ if (n[0] == n[1]) {
+diff --git a/mm/swap.c b/mm/swap.c
+index 87627f1..f0cc321 100644
+--- a/mm/swap.c
++++ b/mm/swap.c
+@@ -31,6 +31,7 @@
+ #include <linux/backing-dev.h>
+ #include <linux/memcontrol.h>
+ #include <linux/gfp.h>
++#include <linux/locallock.h>
- config RWSEM_XCHGADD_ALGORITHM
-- def_bool X86_XADD
-+ def_bool X86_XADD && !RWSEM_GENERIC_SPINLOCK && !PREEMPT_RT_FULL
+ #include "internal.h"
- config ARCH_HAS_CPU_IDLE_WAIT
- def_bool y
-Index: linux-3.0/arch/x86/include/asm/page_64_types.h
-===================================================================
---- linux-3.0.orig/arch/x86/include/asm/page_64_types.h
-+++ linux-3.0/arch/x86/include/asm/page_64_types.h
-@@ -14,12 +14,21 @@
- #define IRQ_STACK_ORDER 2
- #define IRQ_STACK_SIZE (PAGE_SIZE << IRQ_STACK_ORDER)
+@@ -41,6 +42,9 @@ static DEFINE_PER_CPU(struct pagevec[NR_LRU_LISTS], lru_add_pvecs);
+ static DEFINE_PER_CPU(struct pagevec, lru_rotate_pvecs);
+ static DEFINE_PER_CPU(struct pagevec, lru_deactivate_pvecs);
--#define STACKFAULT_STACK 1
--#define DOUBLEFAULT_STACK 2
--#define NMI_STACK 3
--#define DEBUG_STACK 4
--#define MCE_STACK 5
--#define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+# define STACKFAULT_STACK 0
-+# define DOUBLEFAULT_STACK 1
-+# define NMI_STACK 2
-+# define DEBUG_STACK 0
-+# define MCE_STACK 3
-+# define N_EXCEPTION_STACKS 3 /* hw limit: 7 */
-+#else
-+# define STACKFAULT_STACK 1
-+# define DOUBLEFAULT_STACK 2
-+# define NMI_STACK 3
-+# define DEBUG_STACK 4
-+# define MCE_STACK 5
-+# define N_EXCEPTION_STACKS 5 /* hw limit: 7 */
-+#endif
++static DEFINE_LOCAL_IRQ_LOCK(rotate_lock);
++static DEFINE_LOCAL_IRQ_LOCK(swap_lock);
++
+ /*
+ * This path almost never happens for VM activity - pages are normally
+ * freed via pagevecs. But it gets used by networking.
+@@ -267,11 +271,11 @@ void rotate_reclaimable_page(struct page *page)
+ unsigned long flags;
- #define PUD_PAGE_SIZE (_AC(1, UL) << PUD_SHIFT)
- #define PUD_PAGE_MASK (~(PUD_PAGE_SIZE-1))
-Index: linux-3.0/arch/x86/kernel/cpu/common.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/common.c
-+++ linux-3.0/arch/x86/kernel/cpu/common.c
-@@ -1050,7 +1050,9 @@ DEFINE_PER_CPU(unsigned int, irq_count)
- */
- static const unsigned int exception_stack_sizes[N_EXCEPTION_STACKS] = {
- [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STKSZ,
-+#if DEBUG_STACK > 0
- [DEBUG_STACK - 1] = DEBUG_STKSZ
-+#endif
- };
+ page_cache_get(page);
+- local_irq_save(flags);
++ local_lock_irqsave(rotate_lock, flags);
+ pvec = &__get_cpu_var(lru_rotate_pvecs);
+ if (!pagevec_add(pvec, page))
+ pagevec_move_tail(pvec);
+- local_irq_restore(flags);
++ local_unlock_irqrestore(rotate_lock, flags);
+ }
+ }
- static DEFINE_PER_CPU_PAGE_ALIGNED(char, exception_stacks
-Index: linux-3.0/arch/x86/kernel/dumpstack_64.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/dumpstack_64.c
-+++ linux-3.0/arch/x86/kernel/dumpstack_64.c
-@@ -21,10 +21,14 @@
- (N_EXCEPTION_STACKS + DEBUG_STKSZ/EXCEPTION_STKSZ - 2)
+@@ -327,12 +331,13 @@ static void activate_page_drain(int cpu)
+ void activate_page(struct page *page)
+ {
+ if (PageLRU(page) && !PageActive(page) && !PageUnevictable(page)) {
+- struct pagevec *pvec = &get_cpu_var(activate_page_pvecs);
++ struct pagevec *pvec = &get_locked_var(swap_lock,
++ activate_page_pvecs);
- static char x86_stack_ids[][8] = {
-+#if DEBUG_STACK > 0
- [ DEBUG_STACK-1 ] = "#DB",
-+#endif
- [ NMI_STACK-1 ] = "NMI",
- [ DOUBLEFAULT_STACK-1 ] = "#DF",
-+#if STACKFAULT_STACK > 0
- [ STACKFAULT_STACK-1 ] = "#SS",
-+#endif
- [ MCE_STACK-1 ] = "#MC",
- #if DEBUG_STKSZ > EXCEPTION_STKSZ
- [ N_EXCEPTION_STACKS ...
-Index: linux-3.0/arch/x86/kernel/cpu/perf_event.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/perf_event.c
-+++ linux-3.0/arch/x86/kernel/cpu/perf_event.c
-@@ -142,6 +142,8 @@ struct cpu_hw_events {
- * AMD specific bits
- */
- struct amd_nb *amd_nb;
-+
-+ void *kfree_on_online;
- };
+ page_cache_get(page);
+ if (!pagevec_add(pvec, page))
+ pagevec_lru_move_fn(pvec, __activate_page, NULL);
+- put_cpu_var(activate_page_pvecs);
++ put_locked_var(swap_lock, activate_page_pvecs);
+ }
+ }
- #define __EVENT_CONSTRAINT(c, n, m, w) {\
-@@ -1452,10 +1454,12 @@ static int __cpuinit
- x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu)
+@@ -373,12 +378,12 @@ EXPORT_SYMBOL(mark_page_accessed);
+
+ void __lru_cache_add(struct page *page, enum lru_list lru)
{
- unsigned int cpu = (long)hcpu;
-+ struct cpu_hw_events *cpuc = &per_cpu(cpu_hw_events, cpu);
- int ret = NOTIFY_OK;
+- struct pagevec *pvec = &get_cpu_var(lru_add_pvecs)[lru];
++ struct pagevec *pvec = &get_locked_var(swap_lock, lru_add_pvecs)[lru];
- switch (action & ~CPU_TASKS_FROZEN) {
- case CPU_UP_PREPARE:
-+ cpuc->kfree_on_online = NULL;
- if (x86_pmu.cpu_prepare)
- ret = x86_pmu.cpu_prepare(cpu);
- break;
-@@ -1465,6 +1469,10 @@ x86_pmu_notifier(struct notifier_block *
- x86_pmu.cpu_starting(cpu);
- break;
+ page_cache_get(page);
+ if (!pagevec_add(pvec, page))
+ ____pagevec_lru_add(pvec, lru);
+- put_cpu_var(lru_add_pvecs);
++ put_locked_var(swap_lock, lru_add_pvecs);
+ }
+ EXPORT_SYMBOL(__lru_cache_add);
-+ case CPU_ONLINE:
-+ kfree(cpuc->kfree_on_online);
-+ break;
-+
- case CPU_DYING:
- if (x86_pmu.cpu_dying)
- x86_pmu.cpu_dying(cpu);
-Index: linux-3.0/arch/x86/kernel/cpu/perf_event_amd.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/perf_event_amd.c
-+++ linux-3.0/arch/x86/kernel/cpu/perf_event_amd.c
-@@ -336,7 +336,7 @@ static void amd_pmu_cpu_starting(int cpu
- continue;
+@@ -512,9 +517,9 @@ static void drain_cpu_pagevecs(int cpu)
+ unsigned long flags;
- if (nb->nb_id == nb_id) {
-- kfree(cpuc->amd_nb);
-+ cpuc->kfree_on_online = cpuc->amd_nb;
- cpuc->amd_nb = nb;
- break;
- }
-Index: linux-3.0/arch/x86/kernel/cpu/perf_event_intel.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/cpu/perf_event_intel.c
-+++ linux-3.0/arch/x86/kernel/cpu/perf_event_intel.c
-@@ -1267,7 +1267,7 @@ static void intel_pmu_cpu_starting(int c
- struct intel_percore *pc = per_cpu(cpu_hw_events, i).per_core;
+ /* No harm done if a racing interrupt already did this */
+- local_irq_save(flags);
++ local_lock_irqsave(rotate_lock, flags);
+ pagevec_move_tail(pvec);
+- local_irq_restore(flags);
++ local_unlock_irqrestore(rotate_lock, flags);
+ }
- if (pc && pc->core_id == core_id) {
-- kfree(cpuc->per_core);
-+ cpuc->kfree_on_online = cpuc->per_core;
- cpuc->per_core = pc;
- break;
- }
-Index: linux-3.0/fs/eventpoll.c
-===================================================================
---- linux-3.0.orig/fs/eventpoll.c
-+++ linux-3.0/fs/eventpoll.c
-@@ -438,12 +438,12 @@ static int ep_poll_wakeup_proc(void *pri
- */
- static void ep_poll_safewake(wait_queue_head_t *wq)
- {
-- int this_cpu = get_cpu();
-+ int this_cpu = get_cpu_light();
+ pvec = &per_cpu(lru_deactivate_pvecs, cpu);
+@@ -542,18 +547,19 @@ void deactivate_page(struct page *page)
+ return;
- ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS,
- ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu);
+ if (likely(get_page_unless_zero(page))) {
+- struct pagevec *pvec = &get_cpu_var(lru_deactivate_pvecs);
++ struct pagevec *pvec = &get_locked_var(swap_lock,
++ lru_deactivate_pvecs);
+ if (!pagevec_add(pvec, page))
+ pagevec_lru_move_fn(pvec, lru_deactivate_fn, NULL);
+- put_cpu_var(lru_deactivate_pvecs);
++ put_locked_var(swap_lock, lru_deactivate_pvecs);
+ }
+ }
+
+ void lru_add_drain(void)
+ {
+- drain_cpu_pagevecs(get_cpu());
- put_cpu();
-+ put_cpu_light();
++ drain_cpu_pagevecs(local_lock_cpu(swap_lock));
++ local_unlock_cpu(swap_lock);
}
- /*
-Index: linux-3.0/mm/vmalloc.c
-===================================================================
---- linux-3.0.orig/mm/vmalloc.c
-+++ linux-3.0/mm/vmalloc.c
-@@ -789,7 +789,7 @@ static struct vmap_block *new_vmap_block
+ static void lru_add_drain_per_cpu(struct work_struct *dummy)
+@@ -783,6 +789,9 @@ void __init swap_setup(void)
+ {
+ unsigned long megs = totalram_pages >> (20 - PAGE_SHIFT);
+
++ local_irq_lock_init(rotate_lock);
++ local_irq_lock_init(swap_lock);
++
+ #ifdef CONFIG_SWAP
+ bdi_init(swapper_space.backing_dev_info);
+ #endif
+diff --git a/mm/truncate.c b/mm/truncate.c
+index e13f22e..3ee8d53 100644
+--- a/mm/truncate.c
++++ b/mm/truncate.c
+@@ -622,12 +622,12 @@ int vmtruncate_range(struct inode *inode, loff_t offset, loff_t end)
+ return -ENOSYS;
+
+ mutex_lock(&inode->i_mutex);
+- down_write(&inode->i_alloc_sem);
++ anon_down_write(&inode->i_alloc_sem);
+ unmap_mapping_range(mapping, offset, (end - offset), 1);
+ inode->i_op->truncate_range(inode, offset, end);
+ /* unmap again to remove racily COWed private pages */
+ unmap_mapping_range(mapping, offset, (end - offset), 1);
+- up_write(&inode->i_alloc_sem);
++ anon_up_write(&inode->i_alloc_sem);
+ mutex_unlock(&inode->i_mutex);
+
+ return 0;
+diff --git a/mm/vmalloc.c b/mm/vmalloc.c
+index 65d5fd2..95c34be 100644
+--- a/mm/vmalloc.c
++++ b/mm/vmalloc.c
+@@ -789,7 +789,7 @@ static struct vmap_block *new_vmap_block(gfp_t gfp_mask)
struct vmap_block *vb;
struct vmap_area *va;
unsigned long vb_idx;
@@ -26190,7 +26590,7 @@
node = numa_node_id();
-@@ -828,12 +828,13 @@ static struct vmap_block *new_vmap_block
+@@ -828,12 +828,13 @@ static struct vmap_block *new_vmap_block(gfp_t gfp_mask)
BUG_ON(err);
radix_tree_preload_end();
@@ -26206,7 +26606,7 @@
return vb;
}
-@@ -914,7 +915,7 @@ static void *vb_alloc(unsigned long size
+@@ -914,7 +915,7 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask)
struct vmap_block *vb;
unsigned long addr = 0;
unsigned int order;
@@ -26215,7 +26615,7 @@
BUG_ON(size & ~PAGE_MASK);
BUG_ON(size > PAGE_SIZE*VMAP_MAX_ALLOC);
-@@ -922,7 +923,8 @@ static void *vb_alloc(unsigned long size
+@@ -922,7 +923,8 @@ static void *vb_alloc(unsigned long size, gfp_t gfp_mask)
again:
rcu_read_lock();
@@ -26234,649 +26634,379 @@
rcu_read_unlock();
if (!addr) {
-Index: linux-3.0/include/linux/workqueue.h
-===================================================================
---- linux-3.0.orig/include/linux/workqueue.h
-+++ linux-3.0/include/linux/workqueue.h
-@@ -254,9 +254,10 @@ enum {
- WQ_MEM_RECLAIM = 1 << 3, /* may be used for memory reclaim */
- WQ_HIGHPRI = 1 << 4, /* high priority */
- WQ_CPU_INTENSIVE = 1 << 5, /* cpu instensive workqueue */
-+ WQ_NON_AFFINE = 1 << 6, /* free to move works around cpus */
-
-- WQ_DYING = 1 << 6, /* internal: workqueue is dying */
-- WQ_RESCUER = 1 << 7, /* internal: workqueue has rescuer */
-+ WQ_DYING = 1 << 7, /* internal: workqueue is dying */
-+ WQ_RESCUER = 1 << 8, /* internal: workqueue has rescuer */
-
- WQ_MAX_ACTIVE = 512, /* I like 512, better ideas? */
- WQ_MAX_UNBOUND_PER_CPU = 4, /* 4 * #cpus for unbound wq */
-Index: linux-3.0/lib/debugobjects.c
-===================================================================
---- linux-3.0.orig/lib/debugobjects.c
-+++ linux-3.0/lib/debugobjects.c
-@@ -306,7 +306,10 @@ __debug_object_init(void *addr, struct d
- struct debug_obj *obj;
- unsigned long flags;
-
-- fill_pool();
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ if (preempt_count() == 0 && !irqs_disabled())
-+#endif
-+ fill_pool();
-
- db = get_bucket((unsigned long) addr);
-
-@@ -1015,9 +1018,9 @@ static int __init debug_objects_replace_
- }
- }
-
-+ local_irq_enable();
- printk(KERN_DEBUG "ODEBUG: %d of %d active objects replaced\n", cnt,
- obj_pool_used);
-- local_irq_enable();
- return 0;
- free:
- hlist_for_each_entry_safe(obj, node, tmp, &objects, node) {
-Index: linux-3.0/include/linux/jump_label.h
-===================================================================
---- linux-3.0.orig/include/linux/jump_label.h
-+++ linux-3.0/include/linux/jump_label.h
-@@ -4,7 +4,7 @@
- #include <linux/types.h>
- #include <linux/compiler.h>
-
--#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
-+#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL) && !defined(CONFIG_PREEMPT_BASE)
-
- struct jump_label_key {
- atomic_t enabled;
-Index: linux-3.0/include/linux/netdevice.h
-===================================================================
---- linux-3.0.orig/include/linux/netdevice.h
-+++ linux-3.0/include/linux/netdevice.h
-@@ -1764,6 +1764,7 @@ struct softnet_data {
- unsigned dropped;
- struct sk_buff_head input_pkt_queue;
- struct napi_struct backlog;
-+ struct sk_buff_head tofree_queue;
- };
-
- static inline void input_queue_head_incr(struct softnet_data *sd)
-Index: linux-3.0/include/linux/skbuff.h
-===================================================================
---- linux-3.0.orig/include/linux/skbuff.h
-+++ linux-3.0/include/linux/skbuff.h
-@@ -118,6 +118,7 @@ struct sk_buff_head {
-
- __u32 qlen;
- spinlock_t lock;
-+ raw_spinlock_t raw_lock;
- };
-
- struct sk_buff;
-@@ -875,6 +876,12 @@ static inline void skb_queue_head_init(s
- __skb_queue_head_init(list);
- }
-
-+static inline void skb_queue_head_init_raw(struct sk_buff_head *list)
-+{
-+ raw_spin_lock_init(&list->raw_lock);
-+ __skb_queue_head_init(list);
-+}
-+
- static inline void skb_queue_head_init_class(struct sk_buff_head *list,
- struct lock_class_key *class)
- {
-Index: linux-3.0/arch/x86/kernel/irq_work.c
-===================================================================
---- linux-3.0.orig/arch/x86/kernel/irq_work.c
-+++ linux-3.0/arch/x86/kernel/irq_work.c
-@@ -18,6 +18,7 @@ void smp_irq_work_interrupt(struct pt_re
- irq_exit();
- }
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- void arch_irq_work_raise(void)
- {
- #ifdef CONFIG_X86_LOCAL_APIC
-@@ -28,3 +29,4 @@ void arch_irq_work_raise(void)
- apic_wait_icr_idle();
- #endif
- }
-+#endif
-Index: linux-3.0/kernel/events/core.c
-===================================================================
---- linux-3.0.orig/kernel/events/core.c
-+++ linux-3.0/kernel/events/core.c
-@@ -3916,9 +3916,15 @@ static bool perf_output_space(struct per
-
- static void perf_output_wakeup(struct perf_output_handle *handle)
+diff --git a/mm/vmscan.c b/mm/vmscan.c
+index 6072d74..b74883e 100644
+--- a/mm/vmscan.c
++++ b/mm/vmscan.c
+@@ -1277,8 +1277,8 @@ static int too_many_isolated(struct zone *zone, int file,
+ */
+ static noinline_for_stack void
+ putback_lru_pages(struct zone *zone, struct scan_control *sc,
+- unsigned long nr_anon, unsigned long nr_file,
+- struct list_head *page_list)
++ unsigned long nr_anon, unsigned long nr_file,
++ struct list_head *page_list, unsigned long nr_reclaimed)
{
-+ int queue = handle->nmi;
-+
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ queue |= irqs_disabled() || preempt_count();
-+#endif
-+
- atomic_set(&handle->buffer->poll, POLL_IN);
-
-- if (handle->nmi) {
-+ if (queue) {
- handle->event->pending_wakeup = 1;
- irq_work_queue(&handle->event->pending);
- } else
-Index: linux-3.0/kernel/irq_work.c
-===================================================================
---- linux-3.0.orig/kernel/irq_work.c
-+++ linux-3.0/kernel/irq_work.c
-@@ -125,8 +125,10 @@ void irq_work_run(void)
- if (this_cpu_read(irq_work_list) == NULL)
- return;
-
-+#ifndef CONFIG_PREEMPT_RT_FULL
- BUG_ON(!in_irq());
- BUG_ON(!irqs_disabled());
-+#endif
-
- list = this_cpu_xchg(irq_work_list, NULL);
-
-Index: linux-3.0/Documentation/sysrq.txt
-===================================================================
---- linux-3.0.orig/Documentation/sysrq.txt
-+++ linux-3.0/Documentation/sysrq.txt
-@@ -57,10 +57,17 @@ On PowerPC - Press 'ALT - Print Screen (
- On other - If you know of the key combos for other architectures, please
- let me know so I can add them to this section.
-
--On all - write a character to /proc/sysrq-trigger. e.g.:
--
-+On all - write a character to /proc/sysrq-trigger, e.g.:
- echo t > /proc/sysrq-trigger
-
-+On all - Enable network SysRq by writing a cookie to icmp_echo_sysrq, e.g.
-+ echo 0x01020304 >/proc/sys/net/ipv4/icmp_echo_sysrq
-+ Send an ICMP echo request with this pattern plus the particular
-+ SysRq command key. Example:
-+ # ping -c1 -s57 -p0102030468
-+ will trigger the SysRq-H (help) command.
-+
-+
- * What are the 'command' keys?
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- 'b' - Will immediately reboot the system without syncing or unmounting
-Index: linux-3.0/include/net/netns/ipv4.h
-===================================================================
---- linux-3.0.orig/include/net/netns/ipv4.h
-+++ linux-3.0/include/net/netns/ipv4.h
-@@ -47,6 +47,7 @@ struct netns_ipv4 {
-
- int sysctl_icmp_echo_ignore_all;
- int sysctl_icmp_echo_ignore_broadcasts;
-+ int sysctl_icmp_echo_sysrq;
- int sysctl_icmp_ignore_bogus_error_responses;
- int sysctl_icmp_ratelimit;
- int sysctl_icmp_ratemask;
-Index: linux-3.0/net/ipv4/icmp.c
-===================================================================
---- linux-3.0.orig/net/ipv4/icmp.c
-+++ linux-3.0/net/ipv4/icmp.c
-@@ -67,6 +67,7 @@
- #include <linux/jiffies.h>
- #include <linux/kernel.h>
- #include <linux/fcntl.h>
-+#include <linux/sysrq.h>
- #include <linux/socket.h>
- #include <linux/in.h>
- #include <linux/inet.h>
-@@ -801,6 +802,30 @@ out_err:
- }
-
- /*
-+ * 32bit and 64bit have different timestamp length, so we check for
-+ * the cookie at offset 20 and verify it is repeated at offset 50
-+ */
-+#define CO_POS0 20
-+#define CO_POS1 50
-+#define CO_SIZE sizeof(int)
-+#define ICMP_SYSRQ_SIZE 57
-+
-+/*
-+ * We got a ICMP_SYSRQ_SIZE sized ping request. Check for the cookie
-+ * pattern and if it matches send the next byte as a trigger to sysrq.
-+ */
-+static void icmp_check_sysrq(struct net *net, struct sk_buff *skb)
-+{
-+ int cookie = htonl(net->ipv4.sysctl_icmp_echo_sysrq);
-+ char *p = skb->data;
-+
-+ if (!memcmp(&cookie, p + CO_POS0, CO_SIZE) &&
-+ !memcmp(&cookie, p + CO_POS1, CO_SIZE) &&
-+ p[CO_POS0 + CO_SIZE] == p[CO_POS1 + CO_SIZE])
-+ handle_sysrq(p[CO_POS0 + CO_SIZE]);
-+}
+ struct page *page;
+ struct pagevec pvec;
+@@ -1289,7 +1289,12 @@ putback_lru_pages(struct zone *zone, struct scan_control *sc,
+ /*
+ * Put back any unfreeable pages.
+ */
+- spin_lock(&zone->lru_lock);
++ spin_lock_irq(&zone->lru_lock);
+
-+/*
- * Handle ICMP_ECHO ("ping") requests.
- *
- * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo
-@@ -827,6 +852,11 @@ static void icmp_echo(struct sk_buff *sk
- icmp_param.data_len = skb->len;
- icmp_param.head_len = sizeof(struct icmphdr);
- icmp_reply(&icmp_param, skb);
++ if (current_is_kswapd())
++ __count_vm_events(KSWAPD_STEAL, nr_reclaimed);
++ __count_zone_vm_events(PGSTEAL, zone, nr_reclaimed);
+
-+ if (skb->len == ICMP_SYSRQ_SIZE &&
-+ net->ipv4.sysctl_icmp_echo_sysrq) {
-+ icmp_check_sysrq(net, skb);
-+ }
+ while (!list_empty(page_list)) {
+ int lru;
+ page = lru_to_page(page_list);
+@@ -1464,12 +1469,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct zone *zone,
+ nr_reclaimed += shrink_page_list(&page_list, zone, sc);
}
- }
-Index: linux-3.0/net/ipv4/sysctl_net_ipv4.c
-===================================================================
---- linux-3.0.orig/net/ipv4/sysctl_net_ipv4.c
-+++ linux-3.0/net/ipv4/sysctl_net_ipv4.c
-@@ -694,6 +694,13 @@ static struct ctl_table ipv4_net_table[]
- .proc_handler = proc_dointvec
- },
- {
-+ .procname = "icmp_echo_sysrq",
-+ .data = &init_net.ipv4.sysctl_icmp_echo_sysrq,
-+ .maxlen = sizeof(int),
-+ .mode = 0644,
-+ .proc_handler = proc_dointvec
-+ },
-+ {
- .procname = "icmp_ignore_bogus_error_responses",
- .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses,
- .maxlen = sizeof(int),
-Index: linux-3.0/include/linux/kdb.h
-===================================================================
---- linux-3.0.orig/include/linux/kdb.h
-+++ linux-3.0/include/linux/kdb.h
-@@ -153,12 +153,14 @@ extern int kdb_register(char *, kdb_func
- extern int kdb_register_repeat(char *, kdb_func_t, char *, char *,
- short, kdb_repeat_t);
- extern int kdb_unregister(char *);
-+#define in_kdb_printk() (kdb_trap_printk)
- #else /* ! CONFIG_KGDB_KDB */
- #define kdb_printf(...)
- #define kdb_init(x)
- #define kdb_register(...)
- #define kdb_register_repeat(...)
- #define kdb_uregister(x)
-+#define in_kdb_printk() (0)
- #endif /* CONFIG_KGDB_KDB */
- enum {
- KDB_NOT_INITIALIZED,
-Index: linux-3.0/kernel/debug/kdb/kdb_io.c
-===================================================================
---- linux-3.0.orig/kernel/debug/kdb/kdb_io.c
-+++ linux-3.0/kernel/debug/kdb/kdb_io.c
-@@ -539,7 +539,6 @@ int vkdb_printf(const char *fmt, va_list
- int diag;
- int linecount;
- int logging, saved_loglevel = 0;
-- int saved_trap_printk;
- int got_printf_lock = 0;
- int retlen = 0;
- int fnd, len;
-@@ -550,8 +549,6 @@ int vkdb_printf(const char *fmt, va_list
- unsigned long uninitialized_var(flags);
+- local_irq_disable();
+- if (current_is_kswapd())
+- __count_vm_events(KSWAPD_STEAL, nr_reclaimed);
+- __count_zone_vm_events(PGSTEAL, zone, nr_reclaimed);
+-
+- putback_lru_pages(zone, sc, nr_anon, nr_file, &page_list);
++ putback_lru_pages(zone, sc, nr_anon, nr_file, &page_list, nr_reclaimed);
- preempt_disable();
-- saved_trap_printk = kdb_trap_printk;
-- kdb_trap_printk = 0;
+ trace_mm_vmscan_lru_shrink_inactive(zone->zone_pgdat->node_id,
+ zone_idx(zone),
+diff --git a/mm/vmstat.c b/mm/vmstat.c
+index 20c18b7..8da80e9 100644
+--- a/mm/vmstat.c
++++ b/mm/vmstat.c
+@@ -216,6 +216,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item,
+ long x;
+ long t;
- /* Serialize kdb_printf if multiple cpus try to write at once.
- * But if any cpu goes recursive in kdb, just print the output,
-@@ -807,7 +804,6 @@ kdb_print_out:
- } else {
- __release(kdb_printf_lock);
++ preempt_disable_rt();
+ x = delta + __this_cpu_read(*p);
+
+ t = __this_cpu_read(pcp->stat_threshold);
+@@ -225,6 +226,7 @@ void __mod_zone_page_state(struct zone *zone, enum zone_stat_item item,
+ x = 0;
}
-- kdb_trap_printk = saved_trap_printk;
- preempt_enable();
- return retlen;
+ __this_cpu_write(*p, x);
++ preempt_enable_rt();
}
-@@ -817,9 +813,11 @@ int kdb_printf(const char *fmt, ...)
- va_list ap;
- int r;
+ EXPORT_SYMBOL(__mod_zone_page_state);
-+ kdb_trap_printk++;
- va_start(ap, fmt);
- r = vkdb_printf(fmt, ap);
- va_end(ap);
-+ kdb_trap_printk--;
+@@ -257,6 +259,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item)
+ s8 __percpu *p = pcp->vm_stat_diff + item;
+ s8 v, t;
- return r;
++ preempt_disable_rt();
+ v = __this_cpu_inc_return(*p);
+ t = __this_cpu_read(pcp->stat_threshold);
+ if (unlikely(v > t)) {
+@@ -265,6 +268,7 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item)
+ zone_page_state_add(v + overstep, zone, item);
+ __this_cpu_write(*p, -overstep);
+ }
++ preempt_enable_rt();
}
-Index: linux-3.0/kernel/ksysfs.c
-===================================================================
---- linux-3.0.orig/kernel/ksysfs.c
-+++ linux-3.0/kernel/ksysfs.c
-@@ -132,6 +132,15 @@ KERNEL_ATTR_RO(vmcoreinfo);
- #endif /* CONFIG_KEXEC */
+ void __inc_zone_page_state(struct page *page, enum zone_stat_item item)
+@@ -279,6 +283,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item)
+ s8 __percpu *p = pcp->vm_stat_diff + item;
+ s8 v, t;
-+#if defined(CONFIG_PREEMPT_RT_FULL)
-+static ssize_t realtime_show(struct kobject *kobj,
-+ struct kobj_attribute *attr, char *buf)
-+{
-+ return sprintf(buf, "%d\n", 1);
-+}
-+KERNEL_ATTR_RO(realtime);
-+#endif
-+
- /* whether file capabilities are enabled */
- static ssize_t fscaps_show(struct kobject *kobj,
- struct kobj_attribute *attr, char *buf)
-@@ -181,6 +190,9 @@ static struct attribute * kernel_attrs[]
- &kexec_crash_size_attr.attr,
- &vmcoreinfo_attr.attr,
- #endif
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ &realtime_attr.attr,
-+#endif
- NULL
- };
++ preempt_disable_rt();
+ v = __this_cpu_dec_return(*p);
+ t = __this_cpu_read(pcp->stat_threshold);
+ if (unlikely(v < - t)) {
+@@ -287,6 +292,7 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item)
+ zone_page_state_add(v - overstep, zone, item);
+ __this_cpu_write(*p, overstep);
+ }
++ preempt_enable_rt();
+ }
-Index: linux-3.0/ipc/sem.c
-===================================================================
---- linux-3.0.orig/ipc/sem.c
-+++ linux-3.0/ipc/sem.c
-@@ -415,6 +415,13 @@ undo:
- static void wake_up_sem_queue_prepare(struct list_head *pt,
- struct sem_queue *q, int error)
+ void __dec_zone_page_state(struct page *page, enum zone_stat_item item)
+diff --git a/net/core/dev.c b/net/core/dev.c
+index f14f601..c22b0e4 100644
+--- a/net/core/dev.c
++++ b/net/core/dev.c
+@@ -213,14 +213,14 @@ static inline struct hlist_head *dev_index_hash(struct net *net, int ifindex)
+ static inline void rps_lock(struct softnet_data *sd)
{
-+#ifdef CONFIG_PREEMPT_RT_BASE
-+ struct task_struct *p = q->sleeper;
-+ get_task_struct(p);
-+ q->status = error;
-+ wake_up_process(p);
-+ put_task_struct(p);
-+#else
- if (list_empty(pt)) {
- /*
- * Hold preempt off so that we don't get preempted and have the
-@@ -426,6 +433,7 @@ static void wake_up_sem_queue_prepare(st
- q->pid = error;
-
- list_add_tail(&q->simple_list, pt);
-+#endif
+ #ifdef CONFIG_RPS
+- spin_lock(&sd->input_pkt_queue.lock);
++ raw_spin_lock(&sd->input_pkt_queue.raw_lock);
+ #endif
}
- /**
-@@ -439,6 +447,7 @@ static void wake_up_sem_queue_prepare(st
- */
- static void wake_up_sem_queue_do(struct list_head *pt)
+ static inline void rps_unlock(struct softnet_data *sd)
{
-+#ifndef CONFIG_PREEMPT_RT_BASE
- struct sem_queue *q, *t;
- int did_something;
-
-@@ -451,6 +460,7 @@ static void wake_up_sem_queue_do(struct
- }
- if (did_something)
- preempt_enable();
-+#endif
+ #ifdef CONFIG_RPS
+- spin_unlock(&sd->input_pkt_queue.lock);
++ raw_spin_unlock(&sd->input_pkt_queue.raw_lock);
+ #endif
}
- static void unlink_queue(struct sem_array *sma, struct sem_queue *q)
-Index: linux-3.0/drivers/tty/serial/cpm_uart/cpm_uart_core.c
-===================================================================
---- linux-3.0.orig/drivers/tty/serial/cpm_uart/cpm_uart_core.c
-+++ linux-3.0/drivers/tty/serial/cpm_uart/cpm_uart_core.c
-@@ -1225,7 +1225,7 @@ static void cpm_uart_console_write(struc
+@@ -2902,16 +2902,46 @@ int netif_rx_ni(struct sk_buff *skb)
{
- struct uart_cpm_port *pinfo = &cpm_uart_ports[co->index];
- unsigned long flags;
-- int nolock = oops_in_progress;
-+ int nolock = oops_in_progress || sysrq_in_progress;
+ int err;
- if (unlikely(nolock)) {
- local_irq_save(flags);
-Index: linux-3.0/drivers/tty/sysrq.c
-===================================================================
---- linux-3.0.orig/drivers/tty/sysrq.c
-+++ linux-3.0/drivers/tty/sysrq.c
-@@ -492,6 +492,23 @@ static void __sysrq_put_key_op(int key,
- sysrq_key_table[i] = op_p;
+- preempt_disable();
++ migrate_disable();
+ err = netif_rx(skb);
+ if (local_softirq_pending())
+- do_softirq();
+- preempt_enable();
++ thread_do_softirq();
++ migrate_enable();
+
+ return err;
}
+ EXPORT_SYMBOL(netif_rx_ni);
-+#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
-+
-+int sysrq_in_progress;
-+
-+static void set_sysrq_in_progress(int value)
++#ifdef CONFIG_PREEMPT_RT_FULL
++/*
++ * RT runs ksoftirqd as a real time thread and the root_lock is a
++ * "sleeping spinlock". If the trylock fails then we can go into an
++ * infinite loop when ksoftirqd preempted the task which actually
++ * holds the lock, because we requeue q and raise NET_TX softirq
++ * causing ksoftirqd to loop forever.
++ *
++ * It's safe to use spin_lock on RT here as softirqs run in thread
++ * context and cannot deadlock against the thread which is holding
++ * root_lock.
++ *
++ * On !RT the trylock might fail, but there we bail out from the
++ * softirq loop after 10 attempts which we can't do on RT. And the
++ * task holding root_lock cannot be preempted, so the only downside of
++ * that trylock is that we need 10 loops to decide that we should have
++ * given up in the first one :)
++ */
++static inline int take_root_lock(spinlock_t *lock)
+{
-+ sysrq_in_progress = value;
++ spin_lock(lock);
++ return 1;
+}
-+
+#else
-+
-+static void set_sysrq_in_progress(int value)
++static inline int take_root_lock(spinlock_t *lock)
+{
++ return spin_trylock(lock);
+}
-+
+#endif
+
- void __handle_sysrq(int key, bool check_mask)
+ static void net_tx_action(struct softirq_action *h)
{
- struct sysrq_key_op *op_p;
-@@ -500,6 +517,9 @@ void __handle_sysrq(int key, bool check_
- unsigned long flags;
+ struct softnet_data *sd = &__get_cpu_var(softnet_data);
+@@ -2950,7 +2980,7 @@ static void net_tx_action(struct softirq_action *h)
+ head = head->next_sched;
- spin_lock_irqsave(&sysrq_key_table_lock, flags);
-+
-+ set_sysrq_in_progress(1);
-+
- /*
- * Raise the apparent loglevel to maximum so that the sysrq header
- * is shown to provide the user with positive feedback. We do not
-@@ -541,6 +561,9 @@ void __handle_sysrq(int key, bool check_
- printk("\n");
- console_loglevel = orig_log_level;
+ root_lock = qdisc_lock(q);
+- if (spin_trylock(root_lock)) {
++ if (take_root_lock(root_lock)) {
+ smp_mb__before_clear_bit();
+ clear_bit(__QDISC_STATE_SCHED,
+ &q->state);
+@@ -3276,7 +3306,7 @@ static void flush_backlog(void *arg)
+ skb_queue_walk_safe(&sd->input_pkt_queue, skb, tmp) {
+ if (skb->dev == dev) {
+ __skb_unlink(skb, &sd->input_pkt_queue);
+- kfree_skb(skb);
++ __skb_queue_tail(&sd->tofree_queue, skb);
+ input_queue_head_incr(sd);
+ }
+ }
+@@ -3285,10 +3315,13 @@ static void flush_backlog(void *arg)
+ skb_queue_walk_safe(&sd->process_queue, skb, tmp) {
+ if (skb->dev == dev) {
+ __skb_unlink(skb, &sd->process_queue);
+- kfree_skb(skb);
++ __skb_queue_tail(&sd->tofree_queue, skb);
+ input_queue_head_incr(sd);
+ }
}
+
-+ set_sysrq_in_progress(0);
-+
- spin_unlock_irqrestore(&sysrq_key_table_lock, flags);
++ if (!skb_queue_empty(&sd->tofree_queue))
++ raise_softirq_irqoff(NET_RX_SOFTIRQ);
}
-Index: linux-3.0/include/linux/sysrq.h
-===================================================================
---- linux-3.0.orig/include/linux/sysrq.h
-+++ linux-3.0/include/linux/sysrq.h
-@@ -38,6 +38,11 @@ struct sysrq_key_op {
- int enable_mask;
- };
-
-+#ifdef CONFIG_MAGIC_SYSRQ_FORCE_PRINTK
-+extern int sysrq_in_progress;
-+#else
-+#define sysrq_in_progress 0
-+#endif
- #ifdef CONFIG_MAGIC_SYSRQ
+ static int napi_gro_complete(struct sk_buff *skb)
+@@ -3766,10 +3799,17 @@ static void net_rx_action(struct softirq_action *h)
+ struct softnet_data *sd = &__get_cpu_var(softnet_data);
+ unsigned long time_limit = jiffies + 2;
+ int budget = netdev_budget;
++ struct sk_buff *skb;
+ void *have;
- /* Generic SysRq interface -- you may call it from any device driver, supplying
-Index: linux-3.0/arch/x86/kvm/x86.c
-===================================================================
---- linux-3.0.orig/arch/x86/kvm/x86.c
-+++ linux-3.0/arch/x86/kvm/x86.c
-@@ -4900,6 +4900,13 @@ int kvm_arch_init(void *opaque)
- goto out;
- }
+ local_irq_disable();
-+#ifdef CONFIG_PREEMPT_RT_FULL
-+ if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) {
-+ printk(KERN_ERR "RT requires X86_FEATURE_CONSTANT_TSC\n");
-+ return -EOPNOTSUPP;
++ while ((skb = __skb_dequeue(&sd->tofree_queue))) {
++ local_irq_enable();
++ kfree_skb(skb);
++ local_irq_disable();
+ }
-+#endif
+
- r = kvm_mmu_module_init();
- if (r)
- goto out;
-Index: linux-3.0/drivers/scsi/fcoe/fcoe.c
-===================================================================
---- linux-3.0.orig/drivers/scsi/fcoe/fcoe.c
-+++ linux-3.0/drivers/scsi/fcoe/fcoe.c
-@@ -1113,7 +1113,7 @@ static void fcoe_percpu_thread_destroy(u
- struct sk_buff *skb;
- #ifdef CONFIG_SMP
- struct fcoe_percpu_s *p0;
-- unsigned targ_cpu = get_cpu();
-+ unsigned targ_cpu = get_cpu_light();
- #endif /* CONFIG_SMP */
-
- FCOE_DBG("Destroying receive thread for CPU %d\n", cpu);
-@@ -1169,7 +1169,7 @@ static void fcoe_percpu_thread_destroy(u
- kfree_skb(skb);
- spin_unlock_bh(&p->fcoe_rx_list.lock);
+ while (!list_empty(&sd->poll_list)) {
+ struct napi_struct *n;
+ int work, weight;
+@@ -6197,6 +6237,9 @@ static int dev_cpu_callback(struct notifier_block *nfb,
+ netif_rx(skb);
+ input_queue_head_incr(oldsd);
}
-- put_cpu();
-+ put_cpu_light();
- #else
- /*
- * This a non-SMP scenario where the singular Rx thread is
-@@ -1365,11 +1365,11 @@ err2:
- static int fcoe_alloc_paged_crc_eof(struct sk_buff *skb, int tlen)
- {
- struct fcoe_percpu_s *fps;
-- int rc;
-+ int rc, cpu = get_cpu_light();
-
-- fps = &get_cpu_var(fcoe_percpu);
-+ fps = &per_cpu(fcoe_percpu, cpu);
- rc = fcoe_get_paged_crc_eof(skb, tlen, fps);
-- put_cpu_var(fcoe_percpu);
-+ put_cpu_light();
++ while ((skb = __skb_dequeue(&oldsd->tofree_queue))) {
++ kfree_skb(skb);
++ }
- return rc;
- }
-@@ -1561,6 +1561,7 @@ static inline int fcoe_filter_frames(str
- stats->InvalidCRCCount++;
- if (stats->InvalidCRCCount < 5)
- printk(KERN_WARNING "fcoe: dropping frame with CRC error\n");
-+ put_cpu();
- return -EINVAL;
+ return NOTIFY_OK;
}
+@@ -6462,8 +6505,9 @@ static int __init net_dev_init(void)
+ struct softnet_data *sd = &per_cpu(softnet_data, i);
-@@ -1605,7 +1606,7 @@ static void fcoe_recv_frame(struct sk_bu
+ memset(sd, 0, sizeof(*sd));
+- skb_queue_head_init(&sd->input_pkt_queue);
+- skb_queue_head_init(&sd->process_queue);
++ skb_queue_head_init_raw(&sd->input_pkt_queue);
++ skb_queue_head_init_raw(&sd->process_queue);
++ skb_queue_head_init_raw(&sd->tofree_queue);
+ sd->completion_queue = NULL;
+ INIT_LIST_HEAD(&sd->poll_list);
+ sd->output_queue = NULL;
+diff --git a/net/core/sock.c b/net/core/sock.c
+index aebb419..ebbd3e8 100644
+--- a/net/core/sock.c
++++ b/net/core/sock.c
+@@ -2035,12 +2035,11 @@ void lock_sock_nested(struct sock *sk, int subclass)
+ if (sk->sk_lock.owned)
+ __lock_sock(sk);
+ sk->sk_lock.owned = 1;
+- spin_unlock(&sk->sk_lock.slock);
++ spin_unlock_bh(&sk->sk_lock.slock);
+ /*
+ * The sk_lock has mutex_lock() semantics here:
*/
- hp = (struct fcoe_hdr *) skb_network_header(skb);
+ mutex_acquire(&sk->sk_lock.dep_map, subclass, 0, _RET_IP_);
+- local_bh_enable();
+ }
+ EXPORT_SYMBOL(lock_sock_nested);
-- stats = per_cpu_ptr(lport->dev_stats, get_cpu());
-+ stats = per_cpu_ptr(lport->dev_stats, get_cpu_light());
- if (unlikely(FC_FCOE_DECAPS_VER(hp) != FC_FCOE_VER)) {
- if (stats->ErrorFrames < 5)
- printk(KERN_WARNING "fcoe: FCoE version "
-@@ -1637,13 +1638,13 @@ static void fcoe_recv_frame(struct sk_bu
- goto drop;
+diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
+index 23ef31b..f4bdc7a 100644
+--- a/net/ipv4/icmp.c
++++ b/net/ipv4/icmp.c
+@@ -67,6 +67,7 @@
+ #include <linux/jiffies.h>
+ #include <linux/kernel.h>
+ #include <linux/fcntl.h>
++#include <linux/sysrq.h>
+ #include <linux/socket.h>
+ #include <linux/in.h>
+ #include <linux/inet.h>
+@@ -801,6 +802,30 @@ out_err:
+ }
- if (!fcoe_filter_frames(lport, fp)) {
-- put_cpu();
-+ put_cpu_light();
- fc_exch_recv(lport, fp);
- return;
+ /*
++ * 32bit and 64bit have different timestamp length, so we check for
++ * the cookie at offset 20 and verify it is repeated at offset 50
++ */
++#define CO_POS0 20
++#define CO_POS1 50
++#define CO_SIZE sizeof(int)
++#define ICMP_SYSRQ_SIZE 57
++
++/*
++ * We got a ICMP_SYSRQ_SIZE sized ping request. Check for the cookie
++ * pattern and if it matches send the next byte as a trigger to sysrq.
++ */
++static void icmp_check_sysrq(struct net *net, struct sk_buff *skb)
++{
++ int cookie = htonl(net->ipv4.sysctl_icmp_echo_sysrq);
++ char *p = skb->data;
++
++ if (!memcmp(&cookie, p + CO_POS0, CO_SIZE) &&
++ !memcmp(&cookie, p + CO_POS1, CO_SIZE) &&
++ p[CO_POS0 + CO_SIZE] == p[CO_POS1 + CO_SIZE])
++ handle_sysrq(p[CO_POS0 + CO_SIZE]);
++}
++
++/*
+ * Handle ICMP_ECHO ("ping") requests.
+ *
+ * RFC 1122: 3.2.2.6 MUST have an echo server that answers ICMP echo
+@@ -827,6 +852,11 @@ static void icmp_echo(struct sk_buff *skb)
+ icmp_param.data_len = skb->len;
+ icmp_param.head_len = sizeof(struct icmphdr);
+ icmp_reply(&icmp_param, skb);
++
++ if (skb->len == ICMP_SYSRQ_SIZE &&
++ net->ipv4.sysctl_icmp_echo_sysrq) {
++ icmp_check_sysrq(net, skb);
++ }
}
- drop:
- stats->ErrorFrames++;
-- put_cpu();
-+ put_cpu_light();
- kfree_skb(skb);
}
-Index: linux-3.0/drivers/scsi/fcoe/fcoe_ctlr.c
-===================================================================
---- linux-3.0.orig/drivers/scsi/fcoe/fcoe_ctlr.c
-+++ linux-3.0/drivers/scsi/fcoe/fcoe_ctlr.c
-@@ -716,7 +716,7 @@ static unsigned long fcoe_ctlr_age_fcfs(
- unsigned long sel_time = 0;
- struct fcoe_dev_stats *stats;
-
-- stats = per_cpu_ptr(fip->lp->dev_stats, get_cpu());
-+ stats = per_cpu_ptr(fip->lp->dev_stats, get_cpu_light());
+diff --git a/net/ipv4/route.c b/net/ipv4/route.c
+index 75ef66f..7c6c3a8 100644
+--- a/net/ipv4/route.c
++++ b/net/ipv4/route.c
+@@ -242,7 +242,7 @@ struct rt_hash_bucket {
+ };
- list_for_each_entry_safe(fcf, next, &fip->fcfs, list) {
- deadline = fcf->time + fcf->fka_period + fcf->fka_period / 2;
-@@ -749,7 +749,7 @@ static unsigned long fcoe_ctlr_age_fcfs(
- sel_time = fcf->time;
- }
- }
-- put_cpu();
-+ put_cpu_light();
- if (sel_time && !fip->sel_fcf && !fip->sel_time) {
- sel_time += msecs_to_jiffies(FCOE_CTLR_START_DELAY);
- fip->sel_time = sel_time;
-Index: linux-3.0/drivers/scsi/libfc/fc_exch.c
-===================================================================
---- linux-3.0.orig/drivers/scsi/libfc/fc_exch.c
-+++ linux-3.0/drivers/scsi/libfc/fc_exch.c
-@@ -705,10 +705,10 @@ static struct fc_exch *fc_exch_em_alloc(
- }
- memset(ep, 0, sizeof(*ep));
+ #if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \
+- defined(CONFIG_PROVE_LOCKING)
++ defined(CONFIG_PROVE_LOCKING) || defined(CONFIG_PREEMPT_RT_FULL)
+ /*
+ * Instead of using one spinlock for each rt_hash_bucket, we use a table of spinlocks
+ * The size of this table is a power of two and depends on the number of CPUS.
+diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c
+index 57d0752..326162e 100644
+--- a/net/ipv4/sysctl_net_ipv4.c
++++ b/net/ipv4/sysctl_net_ipv4.c
+@@ -694,6 +694,13 @@ static struct ctl_table ipv4_net_table[] = {
+ .proc_handler = proc_dointvec
+ },
+ {
++ .procname = "icmp_echo_sysrq",
++ .data = &init_net.ipv4.sysctl_icmp_echo_sysrq,
++ .maxlen = sizeof(int),
++ .mode = 0644,
++ .proc_handler = proc_dointvec
++ },
++ {
+ .procname = "icmp_ignore_bogus_error_responses",
+ .data = &init_net.ipv4.sysctl_icmp_ignore_bogus_error_responses,
+ .maxlen = sizeof(int),
+diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c
+index 7fa8c6b..ea6b707 100644
+--- a/net/mac80211/rx.c
++++ b/net/mac80211/rx.c
+@@ -2853,7 +2853,7 @@ void ieee80211_rx(struct ieee80211_hw *hw, struct sk_buff *skb)
+ struct ieee80211_supported_band *sband;
+ struct ieee80211_rx_status *status = IEEE80211_SKB_RXCB(skb);
-- cpu = get_cpu();
-+ cpu = get_cpu_light();
- pool = per_cpu_ptr(mp->pool, cpu);
- spin_lock_bh(&pool->lock);
-- put_cpu();
-+ put_cpu_light();
+- WARN_ON_ONCE(softirq_count() == 0);
++ WARN_ON_ONCE_NONRT(softirq_count() == 0);
- /* peek cache of free slot */
- if (pool->left != FC_XID_UNKNOWN) {
-Index: linux-3.0/arch/Kconfig
-===================================================================
---- linux-3.0.orig/arch/Kconfig
-+++ linux-3.0/arch/Kconfig
-@@ -6,6 +6,7 @@ config OPROFILE
- tristate "OProfile system profiling"
- depends on PROFILING
- depends on HAVE_OPROFILE
-+ depends on !PREEMPT_RT_FULL
- select RING_BUFFER
- select RING_BUFFER_ALLOW_SWAP
- help
-Index: linux-3.0/drivers/net/Kconfig
-===================================================================
---- linux-3.0.orig/drivers/net/Kconfig
-+++ linux-3.0/drivers/net/Kconfig
-@@ -3410,6 +3410,7 @@ config NET_FC
+ if (WARN_ON(status->band < 0 ||
+ status->band >= IEEE80211_NUM_BANDS))
+diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
+index a178cb3..2ebce3d 100644
+--- a/net/netfilter/ipvs/ip_vs_ctl.c
++++ b/net/netfilter/ipvs/ip_vs_ctl.c
+@@ -3679,7 +3679,7 @@ int __net_init __ip_vs_control_init(struct net *net)
+ int idx;
+ struct netns_ipvs *ipvs = net_ipvs(net);
- config NETCONSOLE
- tristate "Network console logging support"
-+ depends on !PREEMPT_RT_FULL
- ---help---
- If you want to log kernel messages over the network, enable this.
- See <file:Documentation/networking/netconsole.txt> for details.
-Index: linux-3.0/mm/Kconfig
-===================================================================
---- linux-3.0.orig/mm/Kconfig
-+++ linux-3.0/mm/Kconfig
-@@ -304,7 +304,7 @@ config NOMMU_INITIAL_TRIM_EXCESS
+- ipvs->rs_lock = __RW_LOCK_UNLOCKED(ipvs->rs_lock);
++ rwlock_init(&ipvs->rs_lock);
- config TRANSPARENT_HUGEPAGE
- bool "Transparent Hugepage Support"
-- depends on X86 && MMU
-+ depends on X86 && MMU && !PREEMPT_RT_FULL
- select COMPACTION
- help
- Transparent Hugepages allows the kernel to use huge pages and
-Index: linux-3.0/init/Makefile
-===================================================================
---- linux-3.0.orig/init/Makefile
-+++ linux-3.0/init/Makefile
-@@ -29,4 +29,4 @@ silent_chk_compile.h = :
- include/generated/compile.h: FORCE
- @$($(quiet)chk_compile.h)
- $(Q)$(CONFIG_SHELL) $(srctree)/scripts/mkcompile_h $@ \
-- "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CC) $(KBUILD_CFLAGS)"
-+ "$(UTS_MACHINE)" "$(CONFIG_SMP)" "$(CONFIG_PREEMPT)" "$(CONFIG_PREEMPT_RT_FULL)" "$(CC) $(KBUILD_CFLAGS)"
-Index: linux-3.0/scripts/mkcompile_h
-===================================================================
---- linux-3.0.orig/scripts/mkcompile_h
-+++ linux-3.0/scripts/mkcompile_h
+ /* Initialize rs_table */
+ for (idx = 0; idx < IP_VS_RTAB_SIZE; idx++)
+diff --git a/scripts/mkcompile_h b/scripts/mkcompile_h
+index f221ddf..5f44009 100755
+--- a/scripts/mkcompile_h
++++ b/scripts/mkcompile_h
@@ -4,7 +4,8 @@ TARGET=$1
ARCH=$2
SMP=$3
Added: people/ukleinek/3.0-rt/linux-2.6/debian/patches/series/6ptx2-extra
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ people/ukleinek/3.0-rt/linux-2.6/debian/patches/series/6ptx2-extra Tue Nov 22 20:00:09 2011 (r18312)
@@ -0,0 +1 @@
++ features/all/rt/patch-3.0.9-rt26.patch featureset=rt
More information about the Kernel-svn-changes
mailing list