[linux] 01/01: Merge tag 'debian/4.2.6-1'

debian-kernel at lists.debian.org debian-kernel at lists.debian.org
Tue Nov 10 16:17:30 UTC 2015


This is an automated email from the git hooks/post-receive script.

benh pushed a commit to branch master
in repository linux.

commit b531af692925f1f2f905b16b3b26381b4bb15d75
Merge: 868f3e2 2d9a6bc
Author: Ben Hutchings <ben at decadent.org.uk>
Date:   Tue Nov 10 16:12:32 2015 +0000

    Merge tag 'debian/4.2.6-1'
    
    Refresh some patches.

 debian/changelog                                   | 90 ++++++++++++++++++++++
 ...ia-media-vivid-osd-fix-info-leak-in-ioctl.patch | 31 ++++++++
 ...-when-sending-a-message-on-unbound-socket.patch | 69 +++++++++++++++++
 ...sbvision-fix-overflow-of-interfaces-array.patch | 31 ++++++++
 ...-intercept-ac-to-avoid-guest-host-exploit.patch | 38 +++++++++
 ...x-avoid-guest-host-dos-by-intercepting-ac.patch | 34 ++++++++
 debian/patches/series                              |  5 ++
 7 files changed, 298 insertions(+)

diff --cc debian/changelog
index beb93c2,44ffe2b..438e035
--- a/debian/changelog
+++ b/debian/changelog
@@@ -1,76 -1,93 +1,166 @@@
 +linux (4.3-1~exp2) UNRELEASED; urgency=medium
 +
 +  [ Ben Hutchings ]
 +  * qxl: Enable by default (Closes: #779515)
 +  * mv643xx_eth: Re-enable TSO, fixed upstream in 4.3
 +  * debian/control: Move patchutils from Build-Depends to Build-Depends-Indep,
 +    as we only use filterdiff when building linux-source-<version>
 +  * debian/control,debian/rules: Support a 'stage1' build profile which
 +    builds only linux-libc-dev (Closes: #695243)
 +  * debian/control: Add ':any' to Build-Depends on python3, to support cross-
 +    bootstrap
 +  * [s390*] Update linux-compiler metapackage to gcc-4.9
 +
 +  [ Ian Campbell ]
 +  * [armel/orion5x] Enable Device Tree for orion5x. Patch from Roger Shimizu
 +    (Closes: #803159)
 +  * [armel/orion5x] Enable CONFIG_DEBUG_LL_UART_8250.
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Sat, 07 Nov 2015 14:18:48 +0000
 +
 +linux (4.3-1~exp1) experimental; urgency=medium
 +
 +  * New upstream release
 +
 +  [ Ben Hutchings ]
 +  * netfilter: Enable NFT_DUP_IPV4, NFT_DUP_IPV6 as modules (Closes: #803370)
 +  * tests: Add autopkgtest support
 +  * [x86] Compile with gcc-5
 +  * [x86] Enable PINCTRL_BAYTRAIL (Closes: #797949)
 +
 +  [ Salvatore Bonaccorso ]
 +  * Fix typo in image.plain.postinst template.
 +    Add missing space in warn message causing typo "dangling linkto".
 +    Thanks to Jakub Wilk <jwilk at debian.org> (Closes: #803323)
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Wed, 04 Nov 2015 07:45:13 +0000
 +
 +linux (4.3~rc7-1~exp1) experimental; urgency=medium
 +
 +  * New upstream release candidate
 +    - [x86] smpboot: Fix CPU #1 boot timeout (Closes: #802464)
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Wed, 28 Oct 2015 11:04:27 +0900
 +
 +linux (4.3~rc5-1~exp1) experimental; urgency=medium
 +
 +  * New upstream release candidate
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Wed, 14 Oct 2015 00:48:41 +0100
 +
 +linux (4.3~rc4-1~exp1) experimental; urgency=medium
 +
 +  * New upstream release candidate
 +
 +  [ Ben Hutchings ]
 +  * [armhf] dts: Fix Makefile target for sun4i-a10-itead-iteaduino-plus
 +    (fixes FTBFS)
 +  * [mips*] io: Define ioremap_uc (fixes FTBFS)
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Tue, 06 Oct 2015 23:27:45 +0100
 +
 +linux (4.3~rc3-1~exp1) experimental; urgency=medium
 +
 +  * New upstream release candidate
 +
 +  [ Ben Hutchings ]
 +  * Disable CRAMFS; it was obsoleted by squashfs and initramfs
 +  * [i386] Replace 586 flavour with 686
 +    - Enable support for OLPC and other Geode-based systems in the 686 flavour
 +    - udeb: Update kernel-versions
 +
 + -- Ben Hutchings <ben at decadent.org.uk>  Sun, 27 Sep 2015 21:02:54 +0100
 +
+ linux (4.2.6-1) unstable; urgency=medium
+ 
+   * New upstream stable update:
+     https://www.kernel.org/pub/linux/kernel/v4.x/ChangeLog-4.2.6
+     - mmc: core: Fix init_card in 52Mhz (regression in 4.2)
+     - rtlwifi: rtl8821ae: Fix system lockups on boot (regression in 4.2)
+     - iwlwifi: mvm: init card correctly on ctkill exit check
+       (regression in 3.18)
+     - iwlwifi: mvm: flush fw_dump_wk when mvm fails to start
+       (regression in 3.18)
+     - [x86] iommu/vt-d: fix range computation when making room for large pages
+     - [x86] iommu/amd: Fix BUG when faulting a PROT_NONE VMA
+     - [x86] iommu/amd: Don't clear DTE flags when modifying it
+     - drm: fix mutex leak in drm_dp_get_mst_branch_device
+     - drm: Correct arguments to list_tail_add in create blob ioctl
+     - drm: crtc: integer overflow in drm_property_create_blob()
+     - rtl28xxu: fix control message flaws (regression in 4.0)
+     - ALSA: hda - Fix deadlock at error in building PCM
+     - [x86] ioapic: Prevent NULL pointer dereference in setup_ioapic_dest()
+       (regression in 4.2.4)
+     - mm: make sendfile(2) killable
+     - drm/radeon/dpm: don't add pwm attributes if DPM is disabled
+       (regression in 4.0)
+     - [x86] drm/i915: Restore lost DPLL register write on gen2-4
+       (regression in 3.18)
+     - [x86] drm/i915: Deny wrapping an userptr into a framebuffer
+     - drm/radeon: don't try to recreate sysfs entries on resume
+       (regression in 4.2.5)
+     - drm/radeon: fix dpms when driver backlight control is disabled
+       (regression in 4.2.4)
+     - drm/radeon: move bl encoder assignment into bl init
+     - rbd: require stable pages if message data CRCs are enabled
+     - rbd: don't leak parent_spec in rbd_dev_probe_parent()
+     - rbd: prevent kernel stack blow up on rbd map
+     - [armhf] EXYNOS: Fix double of_node_put() when parsing child power domains
+       (regression in 4.2)
+     - [armhf] dts: Fix audio card detection on Peach boards (regression in 4.1)
+     - [arm64] Revert "ARM64: unwind: Fix PC calculation"
+     - block: don't release bdi while request_queue has live references
+       (regression in 4.2)
+     - dm btree remove: fix a bug when rebalancing nodes after removal
+     - dm cache: the CLEAN_SHUTDOWN flag was not being set
+     - dm btree: fix leak of bufio-backed block in btree_split_beneath error path
+     - Revert "serial: 8250_dma: don't bother DMA with small transfers"
+       (regression in 4.0)
+     - [armel] i2c: mv64xxx: really allow I2C offloading (regression in 3.19)
+     - clkdev: fix clk_add_alias() with a NULL alias device name
+       (regression in 4.2)
+     - fbcon: initialize blink interval before calling fb_set_par
+       (regression in 4.2)
+     - PCI: Prevent out of bounds access in numa_node override
+     - ovl: free stack of paths in ovl_fill_super (regression in 4.0)
+     - ovl: free lower_mnt array in ovl_put_super (regression in 4.0)
+     - ovl: fix dentry reference leak
+     - ovl: fix open in stacked overlay (regression in 4.2)
+     - [x86] Input: alps - only the Dell Latitude D420/430/620/630 have separate
+       stick button bits (regression in 4.1)
+     - crypto: api - Only abort operations on fatal signal
+     - md/raid1: submit_bio_wait() returns 0 on success (regression in 3.10)
+     - md/raid10: submit_bio_wait() returns 0 on success (regression in 3.10)
+     - md/raid5: fix locking in handle_stripe_clean_event() (regression in 3.13)
+     - Revert "md: allow a partially recovered device to be hot-added to an
+       array." (regression in 3.14)
+     - [amd64] EDAC, sb_edac: Fix TAD presence check for sbridge_mci_bind_devs()
+       (regression in 4.2)
+     - mvsas: Fix NULL pointer dereference in mvs_slot_task_free
+     - netfilter: ipset: Fix sleeping memory allocation in atomic context
+       (regression in 4.2)
+     - btrfs: fix possible leak in btrfs_ioctl_balance() (regression in 4.2.5)
+     - kvm: irqchip: fix memory leak (regression in 4.2)
+     - [armhf] thermal: exynos: Fix register read in TMU (regression in 4.2)
+     - blk-mq: fix use-after-free in blk_mq_free_tag_set() (regression in 4.2)
+     - IB/cm: Fix rb-tree duplicate free and use-after-free
+     - sched/deadline: Fix migration of SCHED_DEADLINE tasks (regression in 4.2)
+     - [arm64] compat: fix stxr failure case in SWP emulation
+     - NVMe: Fix memory leak on retried commands
+     - [x86] drm/vmwgfx: Fix up user_dmabuf refcounting
+     - thp: use is_zero_pfn() only after pte_present() check (regression in 4.1)
+     - xen: fix backport of previous kexec patch
+ 
+   [ Ben Hutchings ]
+   * usbvision: fix overflow of interfaces array (CVE-2015-7833)
+   * RDS: fix race condition when sending a message on unbound socket
+     (CVE-2015-7990)
+   * media/vivid-osd: fix info leak in ioctl (CVE-2015-7884)
+   * [x86] KVM: Intercept #AC to avoid guest->host denial-of-service
+     (CVE-2015-5307)
+ 
+  -- Ben Hutchings <ben at decadent.org.uk>  Tue, 10 Nov 2015 14:35:05 +0000
+ 
  linux (4.2.5-1) unstable; urgency=medium
  
    * New upstream stable update:
diff --cc debian/patches/bugfix/all/rds-fix-race-condition-when-sending-a-message-on-unbound-socket.patch
index 0000000,a6c0f6a..299242e
mode 000000,100644..100644
--- a/debian/patches/bugfix/all/rds-fix-race-condition-when-sending-a-message-on-unbound-socket.patch
+++ b/debian/patches/bugfix/all/rds-fix-race-condition-when-sending-a-message-on-unbound-socket.patch
@@@ -1,0 -1,69 +1,69 @@@
+ From: Quentin Casasnovas <quentin.casasnovas at oracle.com>
+ Subject: RDS: fix race condition when sending a message on unbound socket.
+ Date: Fri, 16 Oct 2015 17:11:42 +0200
+ Origin: https://lkml.org/lkml/2015/10/16/530
+ 
+ Sasha's found a NULL pointer dereference in the RDS connection code when
+ sending a message to an apparently unbound socket.  The problem is caused
+ by the code checking if the socket is bound in rds_sendmsg(), which checks
+ the rs_bound_addr field without taking a lock on the socket.  This opens a
+ race where rs_bound_addr is temporarily set but where the transport is not
+ in rds_bind(), leading to a NULL pointer dereference when trying to
+ dereference 'trans' in __rds_conn_create().
+ 
+ Vegard wrote a reproducer for this issue, so kindly ask him to share if
+ you're interested.
+ 
+ I cannot reproduce the NULL pointer dereference using Vegard's reproducer
+ with this patch, whereas I could without.
+ 
+ Complete earlier incomplete fix to CVE-2015-6937:
+ 
+   74e98eb08588 ("RDS: verify the underlying transport exists before creating a connection")
+ 
+ Signed-off-by: Quentin Casasnovas <quentin.casasnovas at oracle.com>
+ Reviewed-by: Vegard Nossum <vegard.nossum at oracle.com>
+ Reviewed-by: Sasha Levin <sasha.levin at oracle.com>
+ Cc: Vegard Nossum <vegard.nossum at oracle.com>
+ Cc: Sasha Levin <sasha.levin at oracle.com>
+ Cc: Chien Yen <chien.yen at oracle.com>
+ Cc: Santosh Shilimkar <santosh.shilimkar at oracle.com>
+ Cc: David S. Miller <davem at davemloft.net>
+ Cc: stable at vger.kernel.org
+ ---
+  net/rds/connection.c | 6 ------
+  net/rds/send.c       | 4 +++-
+  2 files changed, 3 insertions(+), 7 deletions(-)
+ 
+ --- a/net/rds/connection.c
+ +++ b/net/rds/connection.c
 -@@ -187,12 +187,6 @@ new_conn:
++@@ -190,12 +190,6 @@ new_conn:
+  		}
+  	}
+  
+ -	if (trans == NULL) {
+ -		kmem_cache_free(rds_conn_slab, conn);
+ -		conn = ERR_PTR(-ENODEV);
+ -		goto out;
+ -	}
+ -
+  	conn->c_trans = trans;
+  
+  	ret = trans->conn_alloc(conn, gfp);
+ --- a/net/rds/send.c
+ +++ b/net/rds/send.c
 -@@ -986,11 +986,13 @@ int rds_sendmsg(struct socket *sock, str
++@@ -1009,11 +1009,13 @@ int rds_sendmsg(struct socket *sock, str
+  		release_sock(sk);
+  	}
+  
+ -	/* racing with another thread binding seems ok here */
+ +	lock_sock(sk);
+  	if (daddr == 0 || rs->rs_bound_addr == 0) {
+ +		release_sock(sk);
+  		ret = -ENOTCONN; /* XXX not a great errno */
+  		goto out;
+  	}
+ +	release_sock(sk);
+  
 - 	/* size of rm including all sgs */
 - 	ret = rds_rm_size(msg, payload_len);
++ 	if (payload_len > rds_sk_sndbuf(rs)) {
++ 		ret = -EMSGSIZE;
diff --cc debian/patches/series
index d590375,14939bf..13d3234
--- a/debian/patches/series
+++ b/debian/patches/series
@@@ -77,11 -86,25 +77,16 @@@ features/all/grsecurity/grsecurity-kcon
  #features/all/grsecurity/grsecurity-kbuild.patch
  features/all/grsecurity/grkernsec_perf_harden.patch
  
 -bugfix/all/gfs2-make-statistics-unsigned-suitable-for-use-with-.patch
 -bugfix/all/dcache-reduce-the-scope-of-i_lock-in-d_splice_alias.patch
 -bugfix/all/rds-verify-the-underlying-transport-exists-before-cr.patch
  bugfix/all/media-uvcvideo-disable-hardware-timestamps-by-defaul.patch
 -bugfix/mips/mips-pgtable-bits.h-correct-_page_global_shift-build.patch
 -features/all/ath10k-add-qca6164-support.patch
 -debian/block-fix-abi-change-in-4.2.2.patch
 -bugfix/all/nbd-fix-timeout-detection.patch
 -bugfix/all/nbd-remove-variable-pid.patch
 -bugfix/all/nbd-add-locking-for-tasks.patch
  
 -bugfix/all/KEYS-Fix-race-between-key-destruction-and-finding-a-.patch
 -bugfix/all/KEYS-Fix-crash-when-attempt-to-garbage-collect-an-un.patch
 -bugfix/all/KEYS-Don-t-permit-request_key-to-construct-a-new-key.patch
 -
 -debian/target-fix-abi-change-in-4.2.4.patch
 -debian/signal-fix-abi-change-in-4.2.4.patch
 +bugfix/all/selftests-add-missing-include-directives.patch
 +bugfix/all/selftests-memfd-stop-unnecessary-rebuilds.patch
 +bugfix/all/selftests-kprobe-choose-an-always-defined-function-t.patch
 +bugfix/all/selftests-make-scripts-executable.patch
 +bugfix/all/selftests-vm-try-harder-to-allocate-huge-pages.patch
 +bugfix/all/selftests-breakpoints-actually-build-it.patch
+ bugfix/all/usbvision-fix-overflow-of-interfaces-array.patch
+ bugfix/all/rds-fix-race-condition-when-sending-a-message-on-unbound-socket.patch
+ bugfix/all/media-media-vivid-osd-fix-info-leak-in-ioctl.patch
+ bugfix/x86/kvm-x86-vmx-avoid-guest-host-dos-by-intercepting-ac.patch
+ bugfix/x86/kvm-x86-svm-intercept-ac-to-avoid-guest-host-exploit.patch

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/kernel/linux.git



More information about the Kernel-svn-changes mailing list