]> git.ipfire.org Git - thirdparty/valgrind.git/log
thirdparty/valgrind.git
9 years agoRemove non-ASCII characters from this file. No functional change.
Julian Seward [Fri, 21 Aug 2015 11:04:48 +0000 (11:04 +0000)] 
Remove non-ASCII characters from this file.  No functional change.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15576

9 years agoUpdate.
Julian Seward [Fri, 21 Aug 2015 10:53:32 +0000 (10:53 +0000)] 
Update.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15575

9 years agoWhen scanning the statement list for bogus literals we
Florian Krohm [Thu, 20 Aug 2015 22:45:41 +0000 (22:45 +0000)] 
When scanning the statement list for bogus literals we
can stop once the first such literal was found. No need
to scan further.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15574

9 years agoFix incorrect header guard. Reported by giacomopoz@gmail.com
Florian Krohm [Thu, 20 Aug 2015 21:41:28 +0000 (21:41 +0000)] 
Fix incorrect header guard. Reported by giacomopoz@gmail.com
Fixes BZ #351531

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15573

9 years agoInitial data (brk) segment is setup on demand, when a first brk() syscall
Ivo Raisr [Thu, 20 Aug 2015 20:25:19 +0000 (20:25 +0000)] 
Initial data (brk) segment is setup on demand, when a first brk() syscall
is made. It cannot be established during client image initialization because
that would conflict with a temporary stack which ld.so.1 (when executed directly)
uses for loading the target dynamic executable.
See PRE(sys_brk) in syswrap-solaris.c.

Preparatory work for ldsoexec support.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15572

9 years agoFollow-up fix for r15565.
Ivo Raisr [Thu, 20 Aug 2015 05:50:49 +0000 (05:50 +0000)] 
Follow-up fix for r15565.
Expected output of some helgrind tests slightly differed on Solaris.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15571

9 years agoAlso install sigsegv handler in safe-pthread tests wrapper.
Mark Wielaard [Wed, 19 Aug 2015 13:26:28 +0000 (13:26 +0000)] 
Also install sigsegv handler in safe-pthread tests wrapper.

In case we do recognize the xend, but detect it is invalid
(used outside a transaction) we generate a segsegv instead
of a sigill. Handle that in the same way in the test case.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15570

9 years agoFollow-up fix for r15565: sa_restorer should not be used. n-i-bz.
Rhys Kidd [Wed, 19 Aug 2015 12:18:31 +0000 (12:18 +0000)] 
Follow-up fix for r15565: sa_restorer should not be used. n-i-bz.

It is obsolete and not specified by POSIX. See man sigaction on Linux.
No regressions reported.

The following error may be seen on platforms that don't implement this extension:

depbase=`echo tc12_rwl_trivial.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`;\
    gcc -DHAVE_CONFIG_H -I. -I../..  -I../.. -I../../include -I../../coregrind -I../../include -I../../VEX/pub -I../../VEX/pub -DVGA_amd64=1 -DVGO_darwin=1 -DVGP_amd64_darwin=1 -DVGPV_amd64_darwin_vanilla=1 -DVGA_SEC_x86=1 -DVGP_SEC_amd64_darwin=1  -Winline -Wall -Wshadow -Wno-long-long -g -fno-stack-protector  -Wno-format-extra-args -Wno-literal-range -Wno-tautological-constant-out-of-range-compare -Wno-self-assign -Wno-string-plus-int -Wno-uninitialized -Wno-unused-value  -arch x86_64  -MT tc12_rwl_trivial.o -MD -MP -MF $depbase.Tpo -c -o tc12_rwl_trivial.o tc12_rwl_trivial.c &&\
    mv -f $depbase.Tpo $depbase.Po
In file included from tc12_rwl_trivial.c:8:
./safe-pthread.h:37:7: error: no member named 'sa_restorer' in 'struct sigaction'
   sa.sa_restorer = NULL;
   ~~ ^
1 error generated.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15569

9 years agoFix comparison in function VG_(iseqsigset).
Florian Krohm [Wed, 19 Aug 2015 10:09:08 +0000 (10:09 +0000)] 
Fix comparison in function VG_(iseqsigset).
Patch by dcb314@hotmail.com
Fixes BZ #351474.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15568

9 years agoInstall the SIGILL handler everywhere so we get consistent
Tom Hughes [Wed, 19 Aug 2015 08:27:06 +0000 (08:27 +0000)] 
Install the SIGILL handler everywhere so we get consistent
stacks and don't have to worry about __GLIBC_PREREQ not being
defined on all platforms.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15567

9 years agoHandle (unwanted) line number info in filter_defcfaexpr.
Mark Wielaard [Tue, 18 Aug 2015 14:34:49 +0000 (14:34 +0000)] 
Handle (unwanted) line number info in filter_defcfaexpr.

This is just a testsuite cleanup. In the case defcfaexpr was compiled
with line number information (which might happen in some distro builds)
just throw it away because the expected output just wants to see the
function names.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15566

9 years agoAttempt to work around issues with xend being executed unconditionally
Tom Hughes [Tue, 18 Aug 2015 10:29:20 +0000 (10:29 +0000)] 
Attempt to work around issues with xend being executed unconditionally
when a pthread_rwlock is used in an invalid way.

Recent glibcs use transactional memory instructions to do lock ellision
but will sometimes, when locks are used in an invalid way, may calls to
xend on systems which don't support it, on the grounds that the program
is invalid anyway.

So we try and catch and ignore the resulting SIGILL in our tests that
deliberately work with invalid locks.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15565

9 years agoFix a second format string error observed whilst building on OSX 10.10.
Julian Seward [Tue, 18 Aug 2015 07:54:22 +0000 (07:54 +0000)] 
Fix a second format string error observed whilst building on OSX 10.10.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15564

9 years agoFix a format string error observed whilst building on OSX 10.10.
Julian Seward [Mon, 17 Aug 2015 21:19:19 +0000 (21:19 +0000)] 
Fix a format string error observed whilst building on OSX 10.10.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15563

9 years agoRuntime linker ld.so.1 can be now run under Valgrind on Solaris.
Ivo Raisr [Mon, 17 Aug 2015 19:13:12 +0000 (19:13 +0000)] 
Runtime linker ld.so.1 can be now run under Valgrind on Solaris.
Fixes BZ#351386.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15562

9 years agoAdd bug338606 testfile to none/tests/scripts/Makefile.am EXTRA_DIST.
Mark Wielaard [Mon, 17 Aug 2015 15:18:02 +0000 (15:18 +0000)] 
Add bug338606 testfile to none/tests/scripts/Makefile.am EXTRA_DIST.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15561

9 years agoUpdate.
Julian Seward [Mon, 17 Aug 2015 14:03:16 +0000 (14:03 +0000)] 
Update.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15560

9 years agoAdd test cases for VCVT.{S,U}32.F32, S[n], S[n], #imm.
Julian Seward [Mon, 17 Aug 2015 08:23:31 +0000 (08:23 +0000)] 
Add test cases for VCVT.{S,U}32.F32, S[n], S[n], #imm.
Pertains to #342783.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15559

9 years agoarm32: make sure we actually generate real 64-bit loads/stores for the
Julian Seward [Sun, 16 Aug 2015 17:23:50 +0000 (17:23 +0000)] 
arm32: make sure we actually generate real 64-bit loads/stores for the
64 bit tests, rather than two 32 bit transactions.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15558

9 years agoEnable building of ARMv8 crypto instruction tests, having finally
Julian Seward [Sun, 16 Aug 2015 12:18:26 +0000 (12:18 +0000)] 
Enable building of ARMv8 crypto instruction tests, having finally
figured out what the relevant gcc flag is.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15557

9 years agoFix format string confusion that was somehow related to r15510.
Julian Seward [Sun, 16 Aug 2015 11:54:38 +0000 (11:54 +0000)] 
Fix format string confusion that was somehow related to r15510.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15556

9 years agoAdd test cases for PRFM (register). Pertains to #345177.
Julian Seward [Sun, 16 Aug 2015 11:46:16 +0000 (11:46 +0000)] 
Add test cases for PRFM (register).  Pertains to #345177.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15555

9 years agoUpdate script solaris/build_solaris_package to take into account
Ivo Raisr [Sat, 15 Aug 2015 19:51:35 +0000 (19:51 +0000)] 
Update script solaris/build_solaris_package to take into account
SVN version control software used by Valgrind.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15554

9 years agoIncrease test coverage on OS X, by re-enabling the none/tests/amd64/avx2-1 regression...
Rhys Kidd [Sat, 15 Aug 2015 12:21:42 +0000 (12:21 +0000)] 
Increase test coverage on OS X, by re-enabling the none/tests/amd64/avx2-1 regression test.
n-i-bz

(Unfortunately I don’t have right here the hw support, but build environment works)

$ perl tests/vg_regtest none/tests/amd64/avx2-1
avx2-1:          (skipping, prereq failed: test -x avx2-1 && ../../../tests/x86_amd64_features amd64-avx)

== 0 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0 stdoutB failures, 0 post failures ==

On OS X 10.10

Before:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

After:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15553

9 years agoPreliminary OS X memalign() shim work, although this specific regression test remains...
Rhys Kidd [Sat, 15 Aug 2015 11:40:27 +0000 (11:40 +0000)] 
Preliminary OS X memalign() shim work, although this specific regression test remains guarded due to clang build issues with _randArray.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15552

9 years agoIncrease test coverage on OS X, by re-enabling the memcheck/tests/amd64/xsave-avx...
Rhys Kidd [Sat, 15 Aug 2015 11:16:35 +0000 (11:16 +0000)] 
Increase test coverage on OS X, by re-enabling the memcheck/tests/amd64/xsave-avx regression tests with a handy memalign() shim.
n-i-bz

(Unfortunately I don’t have right here the hw support, but build environment works)

$ perl tests/vg_regtest memcheck/tests/amd64/xsave-avx
xsave-avx:       (skipping, prereq failed: test -x xsave-avx && ../../../tests/x86_amd64_features amd64-avx)

== 0 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0 stdoutB failures, 0 post failures ==

On OS X 10.10

Before:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

After:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15551

9 years agoIncrease test coverage on OS X, by re-enabling the none/tests/amd64/avx-1 regression...
Rhys Kidd [Sat, 15 Aug 2015 11:04:02 +0000 (11:04 +0000)] 
Increase test coverage on OS X, by re-enabling the none/tests/amd64/avx-1 regression tests with a handy memalign() shim.
n-i-bz

(Unfortunately I don’t have right here the hw support, but build environment works)

$ perl tests/vg_regtest none/tests/amd64/avx-1
avx-1:           (skipping, prereq failed: test -x avx-1 && ../../../tests/x86_amd64_features amd64-avx)

== 0 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0 stdoutB failures, 0 post failures ==

On OS X 10.10

Before:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

After:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15550

9 years agoIncrease test coverage on OS X, by re-enabling the none/tests/*/movbe regression...
Rhys Kidd [Sat, 15 Aug 2015 10:47:55 +0000 (10:47 +0000)] 
Increase test coverage on OS X, by re-enabling the none/tests/*/movbe regression tests with a handy memalign() shim.
n-i-bz

$ perl tests/vg_regtest none/tests/x86/movbe none/tests/amd64/movbe
movbe:           valgrind   -q ./movbe
movbe:           valgrind   -q ./movbe

== 2 tests, 0 stderr failures, 0 stdout failures, 0 stderrB failures, 0 stdoutB failures, 0 post failures ==

On OS X 10.10

Before:

== 592 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

After:

== 594 tests, 215 stderr failures, 9 stdout failures, 0 stderrB failures, 0 stdoutB failures, 30 post failures ==

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15549

9 years agoPlumb through memalign16() rather than having four implementations of the same code...
Rhys Kidd [Sat, 15 Aug 2015 08:55:14 +0000 (08:55 +0000)] 
Plumb through memalign16() rather than having four implementations of the same code. n-i-bz.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15548

9 years agoUpdate for VEX r3173.
Rhys Kidd [Sat, 15 Aug 2015 07:40:06 +0000 (07:40 +0000)] 
Update for VEX r3173.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15547

9 years agoAdd regression test for bz#350062 - vex x86->IR: unhandled instruction bytes: 0x66...
Rhys Kidd [Sat, 15 Aug 2015 07:06:57 +0000 (07:06 +0000)] 
Add regression test for bz#350062 - vex x86->IR: unhandled instruction bytes: 0x66 0xF 0x3A 0xB (ROUNDSD) on OS X.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15546

9 years agoNewer Solaris has added /proc/self/cmdline and /proc/<pid>/cmdline.
Ivo Raisr [Fri, 14 Aug 2015 20:50:11 +0000 (20:50 +0000)] 
Newer Solaris has added /proc/self/cmdline and /proc/<pid>/cmdline.
Add support for it in the syswrap machinery and test it more thoroughly.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15545

9 years agoBug 348345 - Assertion fails for negative lineno
Julian Seward [Fri, 14 Aug 2015 09:02:46 +0000 (09:02 +0000)] 
Bug 348345 - Assertion fails for negative lineno
(in ML_(addLineInfo))
Patch from Jim Garrison (jim@garrison.cc)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15544

9 years agoUpdate.
Julian Seward [Fri, 14 Aug 2015 08:29:49 +0000 (08:29 +0000)] 
Update.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15543

9 years ago345338 - TIOCGSERIAL and TIOCSSERIAL ioctl support on Linux
Julian Seward [Fri, 14 Aug 2015 08:17:37 +0000 (08:17 +0000)] 
345338 - TIOCGSERIAL and TIOCSSERIAL ioctl support on Linux
Patch from Martin Ling (martin-kdebugs@earth.li)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15542

9 years agoBug 349034 - Add Lustre ioctls LL_IOC_GROUP_LOCK and LL_IOC_GROUP_UNLOCK
Julian Seward [Fri, 14 Aug 2015 08:05:35 +0000 (08:05 +0000)] 
Bug 349034 - Add Lustre ioctls LL_IOC_GROUP_LOCK and LL_IOC_GROUP_UNLOCK
Patch from Frank Zago (fzago@cray.com)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15541

9 years agoFix 350202 - Add limited param to 'monitor block_list'
Philippe Waroquiers [Thu, 13 Aug 2015 22:49:32 +0000 (22:49 +0000)] 
Fix 350202 - Add limited param to 'monitor block_list'

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15540

9 years agoRemove 343173 helgrind crash during stack unwind from 3_10_BUGSTATUS.txt
Philippe Waroquiers [Thu, 13 Aug 2015 22:42:34 +0000 (22:42 +0000)] 
Remove 343173  helgrind crash during stack unwind from 3_10_BUGSTATUS.txt
as it was fixed some time ago (and announced in NEWS)

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15539

9 years agoImprove mmap MAP_HUGETLB support.
Florian Krohm [Thu, 13 Aug 2015 20:35:03 +0000 (20:35 +0000)] 
Improve mmap MAP_HUGETLB support.
This is a follow up to r14682:

When an mmap retry is done without any constraints, the kernel can
place it into free or reservation segments (i.e. anywhere there is no
mapping yet).
In r14682 a sanity check made the hypothesis that the new mapping was
in a free segment, but it does not hold at least on Linux 3.12 and 3.16
on amd64 (tested under Debian).
There is no risk in allowing the mapping to end up in (what was
previously) a reservation at this point, because it is also allowed.

Patch by Guillaume Knispel <xilun0@gmail.com>. Fixes BZ #348269.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15538

9 years agoBug 339563 - The DVB demux DMX_STOP ioctl doesn't have a wrapper.
Julian Seward [Thu, 13 Aug 2015 14:53:17 +0000 (14:53 +0000)] 
Bug 339563 - The DVB demux DMX_STOP ioctl doesn't have a wrapper.
Patch from mchehab@infradead.org (Mauro Carvalho Chehab).

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15537

9 years agodo_cmpxchg8b(): fix incorrect inline assembly (wrong register name)
Julian Seward [Thu, 13 Aug 2015 14:46:10 +0000 (14:46 +0000)] 
do_cmpxchg8b(): fix incorrect inline assembly (wrong register name)
and updated expected result.  n-i-bz.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15536

9 years agoAdd test cases for PCMPxSTRx case 0x18.
Julian Seward [Thu, 13 Aug 2015 14:19:48 +0000 (14:19 +0000)] 
Add test cases for PCMPxSTRx case 0x18.
Pertains to #348574, vex r3172.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15535

9 years agoAdd test cases for PCMPxSTRx case 0x42.
Julian Seward [Thu, 13 Aug 2015 13:47:02 +0000 (13:47 +0000)] 
Add test cases for PCMPxSTRx case 0x42.
Pertains to #339820, vex r3171.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15534

9 years agoUpdate following amd64 AVX front end fixes.
Julian Seward [Thu, 13 Aug 2015 12:55:09 +0000 (12:55 +0000)] 
Update following amd64 AVX front end fixes.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15533

9 years agoAdd test cases for missing AVX VCMPPD/VCMPPS cases.
Julian Seward [Thu, 13 Aug 2015 12:42:28 +0000 (12:42 +0000)] 
Add test cases for missing AVX VCMPPD/VCMPPS cases.
Pertains to vex r3170.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15532

9 years agoWrite a SysRes value as an unsigned number.
Florian Krohm [Thu, 13 Aug 2015 12:27:48 +0000 (12:27 +0000)] 
Write a SysRes value as an unsigned number.
Previously, whether the bitpattern was interpreted as a
signed or unsigned value depended on the platform.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15531

10 years agoFix the regression from r15518 which broke builds
Ivo Raisr [Wed, 12 Aug 2015 16:03:16 +0000 (16:03 +0000)] 
Fix the regression from r15518 which broke builds
on older Solaris releases.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15530

10 years agoUpdate.
Julian Seward [Wed, 12 Aug 2015 14:47:26 +0000 (14:47 +0000)] 
Update.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15529

10 years agoProperly guard memcheck/tests/amd64/xsave-avx. This test relies on memalign(), which...
Rhys Kidd [Wed, 12 Aug 2015 14:19:09 +0000 (14:19 +0000)] 
Properly guard memcheck/tests/amd64/xsave-avx. This test relies on memalign(), which is not present on OS X. Along with a few other avx tests needs refactoring to use posix_memalign() down the track. n-i-bz.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15528

10 years agoFix svn propset for memcheck/tests/amd64/xsave-avx since r15524. n-i-bz.
Rhys Kidd [Wed, 12 Aug 2015 13:42:42 +0000 (13:42 +0000)] 
Fix svn propset for memcheck/tests/amd64/xsave-avx since r15524. n-i-bz.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15527

10 years agovalgrind.h: Suppress client request code generation on x32
Bart Van Assche [Wed, 12 Aug 2015 13:29:10 +0000 (13:29 +0000)] 
valgrind.h: Suppress client request code generation on x32

Ensure that no code is generated for client requests on x32.

From: Matthias Schwarzott <zzam@gentoo.org>

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15526

10 years agoFix clobber.
Florian Krohm [Wed, 12 Aug 2015 13:24:33 +0000 (13:24 +0000)] 
Fix clobber.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15525

10 years agoAdd test cases for AVX XSAVE and XRSTOR instructions.
Julian Seward [Wed, 12 Aug 2015 11:35:27 +0000 (11:35 +0000)] 
Add test cases for AVX XSAVE and XRSTOR instructions.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15524

10 years agoEnhance VG_(machine_get_hwcaps) to check enough stuff to ensure
Julian Seward [Wed, 12 Aug 2015 11:16:47 +0000 (11:16 +0000)] 
Enhance VG_(machine_get_hwcaps) to check enough stuff to ensure
that VEX's XSAVE/XRSTOR implementation will work correctly.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15523

10 years agoHandle new ILGop_IdentV128 introduced by vex r3169.
Julian Seward [Wed, 12 Aug 2015 11:16:23 +0000 (11:16 +0000)] 
Handle new ILGop_IdentV128 introduced by vex r3169.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15522

10 years agoRemove extraneous "-x" which I suspect is debugging support that got
Julian Seward [Wed, 12 Aug 2015 10:59:52 +0000 (10:59 +0000)] 
Remove extraneous "-x" which I suspect is debugging support that got
acidentally left in by r15291.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15521

10 years agoFor those testcases that crash one way or another
Florian Krohm [Tue, 11 Aug 2015 20:58:15 +0000 (20:58 +0000)] 
For those testcases that crash one way or another
remove any core files after running the test.
Patch by Matthias Schwarzott <zzam@gentoo.org>

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15520

10 years agoIn filter_stderr_basic do not remove lines beginning with
Florian Krohm [Tue, 11 Aug 2015 20:37:28 +0000 (20:37 +0000)] 
In filter_stderr_basic do not remove lines beginning with
--pid:
i.e. debug output. Asserts originating from the address space
manager look exactly like that and we do not want to filter those
out by default.
Remove some specialised filter scripts that are no longer needed
and update a few .vgtest files accordingly.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15519

10 years agoAdd support for repository cache protocol version 26.
Ivo Raisr [Tue, 11 Aug 2015 09:12:40 +0000 (09:12 +0000)] 
Add support for repository cache protocol version 26.
Patch contributed by Tomas Jedlicka <jedlickat@gmail.com>.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15518

10 years agoBug #351140 arm64 syscalls setuid (146) and setresgid (149) not implemented.
Mark Wielaard [Sun, 9 Aug 2015 22:37:52 +0000 (22:37 +0000)] 
Bug #351140 arm64 syscalls setuid (146) and setresgid (149) not implemented.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15517

10 years agoWhen profiling memory events in memcheck, event counts larget
Florian Krohm [Sun, 9 Aug 2015 20:55:39 +0000 (20:55 +0000)] 
When profiling memory events in memcheck, event counts larget
than INT_MAX have been observed in practice. Make counters
64 bit entities.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15516

10 years agoConsistently use exit code 1 (== POSIX's EXIT_FAILURE) if things
Florian Krohm [Sun, 9 Aug 2015 20:29:18 +0000 (20:29 +0000)] 
Consistently use exit code 1 (== POSIX's EXIT_FAILURE) if things
go wrong. As we can tell from the error messages what whent wrong
there is no need to have different exit codes to distinguish.
Spotted by Matthias Schwarzott.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15515

10 years agoFix a few function declarations.
Florian Krohm [Sun, 9 Aug 2015 16:27:44 +0000 (16:27 +0000)] 
Fix a few function declarations.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15514

10 years agoRemove duplicate definition ofVALGRIND_HG_ENABLE_CHECKING , wrongly introduced in...
Philippe Waroquiers [Sun, 9 Aug 2015 14:43:33 +0000 (14:43 +0000)] 
Remove duplicate definition ofVALGRIND_HG_ENABLE_CHECKING , wrongly introduced in rev 15207

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15513

10 years agoFix another GCC 5.1 warning.
Florian Krohm [Sun, 9 Aug 2015 08:54:27 +0000 (08:54 +0000)] 
Fix another GCC 5.1 warning.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15512

10 years agoFix build breakage introduced in r15510.
Florian Krohm [Sun, 9 Aug 2015 08:16:54 +0000 (08:16 +0000)] 
Fix build breakage introduced in r15510.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15511

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Sat, 8 Aug 2015 21:45:33 +0000 (21:45 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15510

10 years agoSimplify the build script a lot by using the download_prerequisites
Florian Krohm [Sat, 8 Aug 2015 21:22:55 +0000 (21:22 +0000)] 
Simplify the build script a lot by using the download_prerequisites
script and build prerequisites in tree. Suggested by Dimitry <dimhen@gmail.com>
Also add --disable-bootstrap and some verbiage as to how to speed
up the GCC build even further.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15509

10 years agoFix two more format inconsistencies found on s390. The one
Florian Krohm [Sat, 8 Aug 2015 21:08:31 +0000 (21:08 +0000)] 
Fix two more format inconsistencies found on s390. The one
in m_stacktrace.c also existed for other architectures. Fixed
as well.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15508

10 years agoInitial checkin
Florian Krohm [Sat, 8 Aug 2015 14:07:07 +0000 (14:07 +0000)] 
Initial checkin

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15507

10 years agoDocument another limitation of the Solaris port
Ivo Raisr [Sat, 8 Aug 2015 00:21:26 +0000 (00:21 +0000)] 
Document another limitation of the Solaris port
regarding to AT_SUN_SYSSTAT_ADDR and AT_SUN_SYSSTAT_ZONE_ADDR.
n-i-bz

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15506

10 years agoDisable gdb XML file for TileGX
Zhi-Gang Liu [Fri, 7 Aug 2015 13:25:24 +0000 (13:25 +0000)] 
Disable gdb XML file for TileGX
There will be no support --vgdb-shadow-registers=yes
for TileGX platform.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15505

10 years agoEnsure gdbserver_tests/mcbreak.vgtest works with gcc 4.9
Philippe Waroquiers [Thu, 6 Aug 2015 16:42:12 +0000 (16:42 +0000)] 
Ensure gdbserver_tests/mcbreak.vgtest works with gcc 4.9
At least on x86+gcc4.9, the test fails because step+finish
makes the program stop on the breakme call line (after the call)
instead of reaching line 120.
So; change the gdb commands so as to ensure we reach line 120

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15504

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Thu, 6 Aug 2015 12:11:33 +0000 (12:11 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15503

10 years agoAs per discussion with Julian comment out some dead assignments.
Florian Krohm [Thu, 6 Aug 2015 09:34:20 +0000 (09:34 +0000)] 
As per discussion with Julian comment out some dead assignments.
Now matches what is done for arm64.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15502

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Thu, 6 Aug 2015 09:00:48 +0000 (09:00 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15501

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Wed, 5 Aug 2015 22:27:24 +0000 (22:27 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15500

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Wed, 5 Aug 2015 21:16:09 +0000 (21:16 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15499

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Wed, 5 Aug 2015 19:04:51 +0000 (19:04 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15498

10 years ago(try to) avoid tc09_bad_unlock random failure
Philippe Waroquiers [Wed, 5 Aug 2015 17:43:11 +0000 (17:43 +0000)] 
(try to) avoid tc09_bad_unlock random failure

tc09_bad_unlock fails randomly for the following reason:
Thread 1 is creating a lock in a stack variable and locks it.
It then clones a Thread 2 that will unlock this lock.
The test fails if the Thread 2 unlocks the lock while the
main thread is still just after the clone syscall:
There is no unwind info in this area, and so doing a stacktrace
implies a nasty hack (see hg_main.c evh__pre_thread_ll_create).

There is no such hack when describing the address of the lock
(as there is no logic in the 'normal' stack trace to detect we are
in the clone syscall code).
In such a case, the unwind fail, and the lock address description
lacks the frame nr derived from the captured stack trace.

Adding --fair-sched=yes seems to make a more reproducible test.

Note that the proper solution to all these 'racy helgrind regtests'
would be to add some synchronisations operations between threads
that helgrind does not observe (e.g. using a technique similar to
the pipe big lock) and have correct (but invisible to helgrind) synchronisation
between the threads actions needed for a reproducible regtest.

Not very cheap to develop, --fair-sched=try is cheap and easy
so use that till someone courageous implements non visible synchronisation

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15497

10 years agoFix typo: add missing space between $(AM_CFLAGS) and @FLAG_FSANITIZE@
Philippe Waroquiers [Wed, 5 Aug 2015 16:34:41 +0000 (16:34 +0000)] 
Fix typo: add missing space between $(AM_CFLAGS) and @FLAG_FSANITIZE@
(spotted by Bart and Florian).

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15496

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Wed, 5 Aug 2015 14:38:57 +0000 (14:38 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15495

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Wed, 5 Aug 2015 13:46:58 +0000 (13:46 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15494

10 years agoAdd a comment.
Florian Krohm [Wed, 5 Aug 2015 13:25:58 +0000 (13:25 +0000)] 
Add a comment.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15493

10 years agoThe number of elements in a RangeMap cannot be negative.
Florian Krohm [Wed, 5 Aug 2015 13:23:11 +0000 (13:23 +0000)] 
The number of elements in a RangeMap cannot be negative.
Let the return type of VG_(sizeRangeMap) reflect that.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15492

10 years agoThe number of elements in an OSet cannot be negative.
Florian Krohm [Wed, 5 Aug 2015 12:09:55 +0000 (12:09 +0000)] 
The number of elements in an OSet cannot be negative.
Let the return type of VG_(OSetGen_Size) reflect that.
Also fix a few casts.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15491

10 years agoThe number of elements in a hash table cannot be negative.
Florian Krohm [Wed, 5 Aug 2015 11:26:10 +0000 (11:26 +0000)] 
The number of elements in a hash table cannot be negative.
Let the return type of VG_(HT_count_nodes) reflect that.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15490

10 years agoAdd missing header files.
Florian Krohm [Wed, 5 Aug 2015 07:48:32 +0000 (07:48 +0000)] 
Add missing header files.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15489

10 years agoLet "make distclean" remove generated tilegx instruction test files
Zhi-Gang Liu [Wed, 5 Aug 2015 02:38:10 +0000 (02:38 +0000)] 
Let "make distclean" remove generated tilegx instruction test files

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15488

10 years agoFix issue of "make dist BUILD_ALL_DOCS=no"
Zhi-Gang Liu [Wed, 5 Aug 2015 01:50:59 +0000 (01:50 +0000)] 
Fix issue of "make dist BUILD_ALL_DOCS=no"

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15487

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Tue, 4 Aug 2015 21:26:10 +0000 (21:26 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15486

10 years agoEnsure libvex tests in none are compiling on amd64, when configuring for 32 bits...
Philippe Waroquiers [Tue, 4 Aug 2015 19:25:53 +0000 (19:25 +0000)] 
Ensure libvex tests in none are compiling on amd64, when configuring for 32 bits only
Patch from Matthias Schwarzott

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15485

10 years agoRefine NEWS entry for memcheck xb monitor command
Philippe Waroquiers [Tue, 4 Aug 2015 19:12:05 +0000 (19:12 +0000)] 
Refine NEWS entry for memcheck xb monitor command

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15484

10 years ago* Memcheck:
Philippe Waroquiers [Tue, 4 Aug 2015 19:11:03 +0000 (19:11 +0000)] 
* Memcheck:
  - A new monitor command 'xb <addr> <len>' shows the validity bits
    of <len> bytes at <addr>. Below the validity bits, the byte
    values are shown using a layout similar to the GDB command
    'x /<len>xb <addr>'. The monitor command 'xb' is easier to use
    (in particular on little endian computers) when you need to associate
    byte data value with their corresponding validity bits.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15483

10 years agoRewrite memcheck's event counter machinery.
Florian Krohm [Tue, 4 Aug 2015 15:58:41 +0000 (15:58 +0000)] 
Rewrite memcheck's event counter machinery.
- unique event counters
- simplify PROF_EVENT
- do not allocate more event counters than needed
- compile cleanly

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15482

10 years agoFix the "make regtest" error caused by the new tilegx instruction tests
Zhi-Gang Liu [Tue, 4 Aug 2015 04:59:41 +0000 (04:59 +0000)] 
Fix the "make regtest" error caused by the new tilegx instruction tests

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15481

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Mon, 3 Aug 2015 21:21:42 +0000 (21:21 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15480

10 years agoFix printf format inconsistencies as pointed out by gcc -Wformat-signedness.
Florian Krohm [Mon, 3 Aug 2015 21:05:20 +0000 (21:05 +0000)] 
Fix printf format inconsistencies as pointed out by gcc -Wformat-signedness.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15479

10 years agoImprove the script to also show failures in the post-regtest checks
Florian Krohm [Mon, 3 Aug 2015 20:03:41 +0000 (20:03 +0000)] 
Improve the script to also show failures in the post-regtest checks
in the regtest log.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15478

10 years agoFix a typo in README.solaris.
Ivo Raisr [Mon, 3 Aug 2015 10:48:16 +0000 (10:48 +0000)] 
Fix a typo in README.solaris.
Spotted by Florian Krohm <florian@eich-krohm.de>.

git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15477