27ef175ef0
[lsan] Fix bug when discovering indirectly leaked objects.
...
If an object contains pointers to itself, that doesn't make it indirectly
leaked. D'oh!
llvm-svn: 192716
2013-10-15 16:00:11 +00:00
81f76f6ff7
tsan: introduce __tsan_default_options() function
...
The function allows to statically setup default values for flags.
The interafece matches what asan/msan provide.
llvm-svn: 192715
2013-10-15 15:58:11 +00:00
21e99319f4
tsan: be silent if verbosity=0
...
Currently tests fail with:
Check failed: 'AAA' == 'AAA ==26017==Could not detach from thread 361395 (errno 3).'
llvm-svn: 192711
2013-10-15 15:35:56 +00:00
7502a3a90c
tsan: use verbosity flag in sanitizer_common code directly
...
now it's available from common_flags()
llvm-svn: 192705
2013-10-15 14:12:26 +00:00
45baa00dcb
tsan: actually parse verbofity flag
...
llvm-svn: 192704
2013-10-15 14:00:28 +00:00
52ca74ec61
tsan: move verbosity flag to CommonFlags
...
llvm-svn: 192701
2013-10-15 13:28:51 +00:00
a55db0a90f
tsan: do not leave trash source files in source dir
...
llvm-svn: 192700
2013-10-15 13:23:29 +00:00
347f1a5ee0
tsan: refactor overly-complex logical condition
...
llvm-svn: 192698
2013-10-15 13:04:42 +00:00
a72f6a57e9
tsan: fix Go build
...
llvm-svn: 192696
2013-10-15 13:03:06 +00:00
f54835f768
tsan: move kernel struct definition from sanitizer_linux.h to sanitizer_platform_limits_posix.h
...
llvm-svn: 192695
2013-10-15 12:57:59 +00:00
d534fff6af
tsan: do not build llvm from scratch on each presubmit
...
llvm-svn: 192694
2013-10-15 12:54:50 +00:00
7ac0b2b0e1
tsan: use sanitizer::CommonFlags in tsan
...
llvm-svn: 192692
2013-10-15 12:25:29 +00:00
7a2bbc30a2
[msan] Remove CallocOverflow test.
...
This behaviour depends on MSAN_OPTIONS.
All interesting combinations are covered by lit_tests/allocator_returns_null.cc.
llvm-svn: 192691
2013-10-15 12:20:16 +00:00
14b9924c7b
Clean up StopTheWorld code after r192686.
...
Remove outdated comments. Also remove code that handled an issue in libc's
sigaction(), which we don't use anymore.
llvm-svn: 192689
2013-10-15 11:54:38 +00:00
6f61206f55
tsan: resolve symlinks for called_from_lib suppressions
...
llvm-svn: 192688
2013-10-15 11:34:59 +00:00
113c646c56
[msan] Implement allocator_may_return_null=1 in MemorySanitizer.
...
llvm-svn: 192687
2013-10-15 11:33:48 +00:00
59c850de6d
tsan: implement internal syscall-based versions of sigaction/sigprocmask
...
use them in stoptheworld
fixes applications that intercept sigaction/sigprocmask
llvm-svn: 192686
2013-10-15 11:31:51 +00:00
3feb458a56
Fix a typo.
...
llvm-svn: 192685
2013-10-15 11:22:33 +00:00
d2ddcf575d
[ASan] Make test from r192581 Linux-only, as wild memory contents is less predictable on Mac
...
llvm-svn: 192677
2013-10-15 07:54:27 +00:00
251d1e5d21
[msan] Test for r192599.
...
llvm-svn: 192600
2013-10-14 15:17:05 +00:00
43d90cbd86
[lsan] Support ASan's stack-use-after-return mode in LSan.
...
Treat the fake stack as live memory.
llvm-svn: 192593
2013-10-14 14:04:50 +00:00
7d7768e032
[msan] Remove a long-outdated comment.
...
llvm-svn: 192592
2013-10-14 13:30:40 +00:00
da9f5e7e7f
[asan] Improve thread lifetime tracking on POSIX systems.
...
Call AsanThread::Destroy() from a late-running TSD destructor.
Previously we called it before any user-registered TSD destructors, which caused
false positives in LeakSanitizer.
llvm-svn: 192585
2013-10-14 12:01:05 +00:00
7aacd9c172
[msan] Intercept strto(d|f|ld)_l and glibc-specific __strto(d|f|ld)_l.
...
llvm-svn: 192583
2013-10-14 11:52:40 +00:00
df6e6569c0
[ASan] Don't die with internal ASan error on large buffer overflows
...
Summary:
Out-of-bound access may touch not-yet allocated or already freed
and recycled from quarantine chunks. We should treat this situation as
a "free-range memory access" and avoid printing any data about that
irrelevant chunk (which may be inconsistent).
This should fix https://code.google.com/p/address-sanitizer/issues/detail?id=183
Reviewers: kcc
Reviewed By: kcc
CC: timurrrr, llvm-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D1893
llvm-svn: 192581
2013-10-14 11:13:54 +00:00
054d533d8d
[msan] Regression test for r192575.
...
llvm-svn: 192577
2013-10-14 09:53:39 +00:00
4314e319ab
tsan: allow to override OnPrint() callback in Go runtime
...
llvm-svn: 192576
2013-10-14 09:52:40 +00:00
c7ae7529e7
[Sanitizer] Return implementation of StackTrace::CopyFrom to header
...
llvm-svn: 192574
2013-10-14 07:36:10 +00:00
abc416f326
tsan: allow to override OnFinalize() in front-ends that does not support weak functions (Go)
...
llvm-svn: 192569
2013-10-14 06:31:03 +00:00
7f5823844a
[Sanitizer] Remove StackTrace::max_depth field
...
llvm-svn: 192534
2013-10-12 12:40:47 +00:00
c129e65661
[Sanitizer] Turn GetStackTrace() into StackTrace::Unwind()
...
llvm-svn: 192533
2013-10-12 12:23:00 +00:00
53e61b1acf
[sanitizer] Define kPthreadDestructorIterations on POSIX systems.
...
llvm-svn: 192453
2013-10-11 14:19:14 +00:00
6dc0775f1f
tsan: catch more races on file descriptors
...
llvm-svn: 192452
2013-10-11 14:13:11 +00:00
c9f9ada48d
tsan: fix epoll_ctl interceptor
...
Currently data-race-test unittests fail with the following false positive:
WARNING: ThreadSanitizer: data race (pid=20365)
Write of size 8 at 0x7da000008050 by thread T54:
#0 close tsan_interceptors.cc:1483 (racecheck_unittest-linux-amd64-O0+0x0000000eb34a)
#1 NegativeTests_epoll::Worker2() unittest/posix_tests.cc:1148 (racecheck_unittest-linux-amd64-O0+0x0000000cc6b1)
#2 MyThread::ThreadBody(MyThread*) unittest/./thread_wrappers_pthread.h:367 (racecheck_unittest-linux-amd64-O0+0x000000097500)
Previous read of size 8 at 0x7da000008050 by thread T49:
#0 epoll_ctl tsan_interceptors.cc:1646 (racecheck_unittest-linux-amd64-O0+0x0000000e9fee)
#1 NegativeTests_epoll::Worker1() unittest/posix_tests.cc:1140 (racecheck_unittest-linux-amd64-O0+0x0000000cc5b5)
#2 MyThread::ThreadBody(MyThread*) unittest/./thread_wrappers_pthread.h:367 (racecheck_unittest-linux-amd64-O0+0x000000097500)
llvm-svn: 192448
2013-10-11 13:33:22 +00:00
6c5e943953
[sanitizer] Ninja style fix to r192442.
...
llvm-svn: 192443
2013-10-11 12:15:45 +00:00
dc75cf3368
[sanitizer] Move the PTHREAD_DESTRUCTOR_ITERATIONS constant to sanitizer_linux.h.
...
Add a test.
llvm-svn: 192442
2013-10-11 12:09:49 +00:00
0cd96165f1
[Sanitizer] StackTrace: kill CopyTo and reimplement CopyFrom
...
llvm-svn: 192441
2013-10-11 11:14:20 +00:00
9d8385453b
[Sanitizer] Simplify StackTrace::FastUnwindStack interface and fix a bug with one-frame stack traces
...
llvm-svn: 192428
2013-10-11 09:58:30 +00:00
3431b38067
tsan: minor refactoring
...
Replace duplicate code snippet with function.
llvm-svn: 192357
2013-10-10 16:03:24 +00:00
fbb194ff34
tsan: add annotations to ignore synchronization operations
...
The annotations are AnnotateIgnoreSyncBegin/End,
may be useful to ignore some infrastructure synchronization
that introduces lots of false negatives.
llvm-svn: 192355
2013-10-10 15:58:12 +00:00
6f7fb43ad3
[sanitizer] Fix the parent liveness check in StopTheWorld.
...
Comparing the parent PID with 1 isn't sufficient to ensure the parent is alive,
because of prctl(PR_SET_CHILD_SUBREAPER, ...). Compare with the real parent's
recorded PID instead.
llvm-svn: 192295
2013-10-09 13:36:20 +00:00
adef754297
[sanitizer] Fix deadlock in StopTheWorld.
...
The tracer thread in StopTheWorld could wait on a mutex forever if the parent
process died before unlocking it. Use PR_SET_PDEATHSIG so that the parent would
kill the child in this scenario.
llvm-svn: 192210
2013-10-08 18:01:03 +00:00
398fb700a2
[sanitizer] Do not break the CFI unwind chain in clone().
...
Fixes build on clang 3.2, which doesn't support cfi_undefined. This change
doesn't seem to break the slow unwinder. Even if it does, the worst thing that
could happen is that we would not see a backtrace when a fatal error happens
under StopTheWorld.
llvm-svn: 192196
2013-10-08 16:38:39 +00:00
50ec07c196
[TSan] Automatically pick up llvm-symbolizer from PATH, as we do in another sanitizers
...
llvm-svn: 191954
2013-10-04 13:38:35 +00:00
12f5e639d6
[TSan] Add ValgrindSlowdown to the list of exported symbols
...
llvm-svn: 191952
2013-10-04 13:12:05 +00:00
513a3a6c0b
tsan: fix a bug in pthread_once when called from an ignored library
...
It must set thr->in_rtl = 0 before executing user callback.
llvm-svn: 191946
2013-10-04 10:27:21 +00:00
f2b811a618
Refactor the usage of strip_path_prefix option and make it more consistent across sanitizers
...
llvm-svn: 191943
2013-10-04 08:55:03 +00:00
302ec7b9bc
tsan: add memory_limit_mb flag
...
The flag allows to bound maximum process memory consumption (best effort).
If RSS reaches memory_limit_mb, tsan flushes all shadow memory.
llvm-svn: 191913
2013-10-03 17:14:35 +00:00
c967d2f49c
asan/msan/tsan: fix broken linux syscalls
...
Currently the following source code:
u64 NanoTime() {
kernel_timeval tv = {};
internal_syscall(__NR_gettimeofday, &tv, 0);
return (u64)tv.tv_sec * 1000*1000*1000 + tv.tv_usec * 1000;
}
generates the following assembly:
Dump of assembler code for function __sanitizer::NanoTime():
=> 0x00007ff30657d380 <+0>: lea -0x10(%rsp),%rdi
0x00007ff30657d385 <+5>: xor %esi,%esi
0x00007ff30657d387 <+7>: mov $0x60,%eax
0x00007ff30657d38c <+12>: syscall
0x00007ff30657d38e <+14>: mov $0,%eax
0x00007ff30657d393 <+19>: retq
with this change:
Dump of assembler code for function __sanitizer::NanoTime():
=> 0x00007faab834f380 <+0>: movq $0x0,-0x10(%rsp)
0x00007faab834f389 <+9>: movq $0x0,-0x8(%rsp)
0x00007faab834f392 <+18>: lea -0x10(%rsp),%rdi
0x00007faab834f397 <+23>: xor %esi,%esi
0x00007faab834f399 <+25>: mov $0x60,%eax
0x00007faab834f39e <+30>: syscall
0x00007faab834f3a0 <+32>: mov -0x8(%rsp),%rax
0x00007faab834f3a5 <+37>: mov -0x10(%rsp),%rcx
0x00007faab834f3aa <+42>: imul $0x3e8,%rax,%rdx
0x00007faab834f3b1 <+49>: imul $0x3b9aca00,%rcx,%rax
0x00007faab834f3b8 <+56>: add %rdx,%rax
0x00007faab834f3bb <+59>: retq
llvm-svn: 191908
2013-10-03 16:53:50 +00:00
19a1302046
asan/msan: fix "unused function 'OnExit'" warning
...
llvm-svn: 191904
2013-10-03 15:43:59 +00:00
Sergey Matveev