]>
git.ipfire.org Git - thirdparty/valgrind.git/log
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
Julian Seward [Fri, 21 Aug 2015 10:53:32 +0000 (10:53 +0000)]
Update.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15575
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
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
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
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
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
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It is obsolete and not specified by POSIX.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
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
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
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
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
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
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
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
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
Julian Seward [Mon, 17 Aug 2015 14:03:16 +0000 (14:03 +0000)]
Update.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15560
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
Julian Seward [Fri, 14 Aug 2015 08:29:49 +0000 (08:29 +0000)]
Update.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15543
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
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
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
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
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
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
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
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
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
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
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
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
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
Julian Seward [Wed, 12 Aug 2015 14:47:26 +0000 (14:47 +0000)]
Update.
git-svn-id: svn://svn.valgrind.org/valgrind/trunk@15529
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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
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