| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
llvm-svn: 210465
|
| |
|
|
|
|
|
| |
This should fix strtoimax/strtoumax on newer glibc.
https://code.google.com/p/memory-sanitizer/issues/detail?id=36
llvm-svn: 210463
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Instructions from __nodebug__ functions don't have file:line
information even when inlined into no-nodebug functions. As a result,
intrinsics (SSE and other) from <*intrin.h> clang headers _never_
have file:line information.
With this change, an instruction without !dbg metadata gets one from
the call instruction when inlined.
Fixes PR19001.
llvm-svn: 210459
|
| |
|
|
| |
llvm-svn: 210458
|
| |
|
|
| |
llvm-svn: 210457
|
| |
|
|
| |
llvm-svn: 210452
|
| |
|
|
| |
llvm-svn: 210370
|
| |
|
|
|
|
| |
Go calling convention does not support bools (it probably works, but uptrs are safer)
llvm-svn: 210365
|
| |
|
|
| |
llvm-svn: 210360
|
| |
|
|
|
|
| |
FuncEnter adds FuncEnter entry to trace that nobody removes later
llvm-svn: 210359
|
| |
|
|
|
|
| |
In Go it's legal to unlock from a different goroutine.
llvm-svn: 210358
|
| |
|
|
| |
llvm-svn: 210353
|
| |
|
|
| |
llvm-svn: 210351
|
| |
|
|
|
|
| |
FuncEntry can resize the shadow stack, while "thr->shadow_stack_pos[0] = pc" writes out-of-bounds.
llvm-svn: 210349
|
| |
|
|
|
|
|
| |
Go maps heap and data+bss, these regions are not adjacent.
data+bss is mapped first.
llvm-svn: 210348
|
| |
|
|
| |
llvm-svn: 210347
|
| |
|
|
|
|
| |
comparison of unsigned expression >= 0 is always true
llvm-svn: 210346
|
| |
|
|
|
|
| |
Go does not have cur_thread function.
llvm-svn: 210345
|
| |
|
|
| |
llvm-svn: 210344
|
| |
|
|
| |
llvm-svn: 210335
|
| |
|
|
|
|
| |
Bots did not like it.
llvm-svn: 210332
|
| |
|
|
| |
llvm-svn: 210331
|
| |
|
|
| |
llvm-svn: 210329
|
| |
|
|
|
|
|
| |
Don't fail if __libc_malloc_dispatch is missing; continue running
with normal linux interceptors instead.
llvm-svn: 210322
|
| |
|
|
|
|
| |
because that may actually try to use fake_stack_ (still don't know how). Yet another case where we desperately want https://sourceware.org/glibc/wiki/ThreadPropertiesAPI
llvm-svn: 210315
|
| |
|
|
| |
llvm-svn: 210301
|
| |
|
|
|
|
|
|
| |
Use caller pc of __sanitizer_cov_module_init to figure out
when 2 sequential calls are from the same module; skip
.sancov.map file update in this case.
llvm-svn: 210267
|
| |
|
|
|
|
| |
In standalone build asan-rt is stored in yet another path.
llvm-svn: 210259
|
| |
|
|
|
|
|
|
| |
This sound like a good idea in general.
Also, without this on Android we get add_lit_testsuite() with empty DEPENDS
list, and it does not work well.
llvm-svn: 210257
|
| |
|
|
|
|
| |
Pass cflags in a way that's compatible with standalone compiler-rt build.
llvm-svn: 210256
|
| |
|
|
| |
llvm-svn: 210255
|
| |
|
|
|
|
| |
Android is removing sys/timeb.h because it was removed in POSIX 2008.
llvm-svn: 210235
|
| |
|
|
| |
llvm-svn: 210232
|
| |
|
|
| |
llvm-svn: 210216
|
| |
|
|
| |
llvm-svn: 210202
|
| |
|
|
|
|
| |
Based on a patch by Stephan Bergmann.
llvm-svn: 210199
|
| |
|
|
|
|
|
|
|
|
| |
Reset coverage data on fork().
For memory-mapped mode (coverage_direct=1) this helps avoid loss of data
(before this change two processes would write to the same file simultaneously).
For normal mode, this reduces coverage dump size, because PCs from the parent
process are no longer inherited by the child.
llvm-svn: 210180
|
| |
|
|
| |
llvm-svn: 210140
|
| |
|
|
| |
llvm-svn: 210109
|
| |
|
|
| |
llvm-svn: 210108
|
| |
|
|
|
|
| |
https://code.google.com/p/address-sanitizer/wiki/OneDefinitionRuleViolation . This feature has been working quite well so far, found a few bugs and zero false positives. The main problem is that there could be lots of true positives and users will have to disable this checking until the bugs are fixed.
llvm-svn: 210106
|
| |
|
|
|
|
|
| |
dlopen()/dlclose() are not interceptable on Android,
so we update .sancov.map in module constructor callbacks.
llvm-svn: 210098
|
| |
|
|
| |
llvm-svn: 210069
|
| |
|
|
|
|
|
|
|
|
| |
No longer need to set ANDROID if COMPILER_RT_TEST_TARGET_TRIPLE is
arm-linux-androideabi.
No need to set ANDROID_COMMON_FLAGS. These flags are already in
CMAKE_CXX_FLAGS which are used in try_compile().
llvm-svn: 210053
|
| |
|
|
|
|
| |
to avoid a null function call in cout/cerr constructors
llvm-svn: 210030
|
| |
|
|
| |
llvm-svn: 210028
|
| |
|
|
| |
llvm-svn: 210027
|
| |
|
|
|
|
| |
The alternative (writing coverage at process exit) is nearly useless there.
llvm-svn: 210024
|
| |
|
|
| |
llvm-svn: 210020
|
| |
|
|
| |
llvm-svn: 210012
|