[Glibc-bsd-commits] r3553 - trunk/web/patches

Robert Millan rmh at alioth.debian.org
Sun Jul 10 11:51:39 UTC 2011


Author: rmh
Date: 2011-07-10 11:51:39 +0000 (Sun, 10 Jul 2011)
New Revision: 3553

Modified:
   trunk/web/patches/clisp.txt
Log:
Remove obsolete information about clisp

Modified: trunk/web/patches/clisp.txt
===================================================================
--- trunk/web/patches/clisp.txt	2011-07-10 10:39:01 UTC (rev 3552)
+++ trunk/web/patches/clisp.txt	2011-07-10 11:51:39 UTC (rev 3553)
@@ -1,77 +1,6 @@
 
-clisp currently fails on both kfreebsd-i386 and kfreebsd-amd64
+clisp currently fails on kfreebsd-amd64
 
-Even clisp 2.38-7 (available from gnuab) is not buildable on 
-current kfreebsd-i386 (neither with current libsigsegv 2.4-4, 
-neither with libsigsegv 2.3-3). 
-This is long standing indirect regression, cause still unknown :-(
-
-
-There is also issue with libsigsegv in our glibc (fixed in r2025).
-The pthread wrapper around signal handler does not pass 4th argument.   
-As a result, loading of libpthread.so breaks functionality of libsigsegv
-
-
-libsigsegv 2.4-4 needs small fixup for amd64,
-upcoming libsigsegv 2.5 also for configure.ac/configure:
-
-diff -ur libsigsegv-2.5/src/fault-freebsd-i386.h libsigsegv-2.5/src/fault-freebsd-i386.h
---- libsigsegv-2.5/src/fault-freebsd-i386.h	2005-05-20 21:47:21.000000000 +0200
-+++ libsigsegv-2.5/src/fault-freebsd-i386.h	2007-11-15 16:38:38.000000000 +0100
-@@ -18,4 +18,8 @@
- #define SIGSEGV_FAULT_HANDLER_ARGLIST  int sig, int code, struct sigcontext *scp, void *addr
- #define SIGSEGV_FAULT_ADDRESS  addr
- #define SIGSEGV_FAULT_CONTEXT  scp
-+#ifdef __x86_64__
-+#define SIGSEGV_FAULT_STACKPOINTER  scp->sc_rsp
-+#else
- #define SIGSEGV_FAULT_STACKPOINTER  scp->sc_esp
-+#endif
-diff -ur libsigsegv-2.5/configure.ac libsigsegv-2.5/configure.ac
---- libsigsegv-2.5/configure.ac	2007-11-11 19:35:23.000000000 +0100
-+++ libsigsegv-2.5/configure.ac	2007-11-15 16:34:43.000000000 +0100
-@@ -74,7 +74,7 @@
- dnl List of signals that are sent when an invalid virtual memory address
- dnl is accessed, or when the stack overflows.
- case "$host_os" in
--  sunos4* | freebsd* | openbsd* | netbsd*)
-+  sunos4* | freebsd* | openbsd* | netbsd* | k*bsd*)
-     CFG_SIGNALS=signals-bsd.h ;;
-   hpux*)
-     CFG_SIGNALS=signals-hpux.h ;;
-@@ -298,7 +298,7 @@
- fi
- if test -z "$CFG_FAULT" && test "$sv_cv_fault_bsd" = yes; then
-   case "$host_os" in
--    freebsd*)
-+    freebsd*|kfreebsd*)
-       case "$host_cpu" in
-         i?86 | x86_64)
-           CFG_FAULT=fault-freebsd-i386.h
-diff -ur libsigsegv-2.5/configure libsigsegv-2.5/configure
---- libsigsegv-2.5/configure	2007-11-11 19:39:31.000000000 +0100
-+++ libsigsegv-2.5/configure	2007-11-15 16:36:26.000000000 +0100
-@@ -19468,7 +19468,7 @@
- 
- 
- case "$host_os" in
--  sunos4* | freebsd* | openbsd* | netbsd*)
-+  sunos4* | freebsd* | openbsd* | netbsd* | k*bsd*)
-     CFG_SIGNALS=signals-bsd.h ;;
-   hpux*)
-     CFG_SIGNALS=signals-hpux.h ;;
-@@ -23351,7 +23351,7 @@
- fi
- if test -z "$CFG_FAULT" && test "$sv_cv_fault_bsd" = yes; then
-   case "$host_os" in
--    freebsd*)
-+    freebsd*|kfreebsd*)
-       case "$host_cpu" in
-         i?86 | x86_64)
-           CFG_FAULT=fault-freebsd-i386.h
-
-
-
 ffcall does not work correctly, see #451356
 
 also main clisp needs fix up for kfreebsd-amd64:
@@ -79,21 +8,6 @@
 
 only in patch2:
 unchanged:
---- clisp-2.42.orig/configure
-+++ clisp-2.42/configure
-@@ -726,8 +726,10 @@
- stacksizelimit=`ulimit -s 2>/dev/null || :` # cygwin /bin/sh ulimit is broken
- # need 3 separate test calls because of "integer expression expected" errors
- # when $stacksizelimit is "" or "unlimited" (no short-circuiting!)
-+set +e
- test -z "$stacksizelimit" || { test "$stacksizelimit" != unlimited && test "$stacksizelimit" -lt ${STACK_LIMIT}; }
- STACK_TOO_SMALL=$?      # 0=true => need to reset; 1=false => big enough
-+set -e
- 
- if test -z "$do_build"; then
- 
-only in patch2:
-unchanged:
 --- clisp-2.42.orig/src/lispbibl.d
 +++ clisp-2.42/src/lispbibl.d
 @@ -3000,7 +3000,7 @@




More information about the Glibc-bsd-commits mailing list