shaka-packager/tools/valgrind/tsan/ignores_win32.txt

65 lines
1.8 KiB
Plaintext
Raw Normal View History

# This file lists the functions, object files and source files
# which should be ignored (i.e. not instrumented) by ThreadSanitizer on Windows.
# We ignore security libraries for now since their instrumentation is very slow.
# TODO(timurrrr): investigate whether we need to instrument them
obj:*CRYPT32.dll
obj:*RPCRT4.dll
fun_r:*SHA256*
fun_r:*BCryptGenerateSymmetricKey*
fun_r:*CryptAcquireContext*
obj:*WINHTTP.dll
obj:*imagehlp.dll
# Instrumenting IP Helper API causes crashes.
# TODO(szym): investigate http://crbug.com/146119
obj:*IPHLPAPI.dll
# Use less detailed instrumentation of STL
fun_hist:*std::*<*
# Don't instrument some stl internals - they shouldn't be useful
fun_r:*std::_Debug*
fun_r:*std::_Lockit*
# Benign race on mutex unlock
fun:_Mtxunlock
# Benign race during clock initialization
fun_r:*InitializeClock*
# Some unknown Windows guts
fun_r:Ordinal_*
fun:unnamedImageEntryPoint
fun_r:RtlDestroyQueryDebugBuffer
fun:_updatetlocinfoEx_nolock
# Strange reports on net_unittests, maybe related to raising
# a debug exception by PlatformThread
# TODO(timurrrr): investigate
fun_r:*PlatformThread*SetName*
# Recursively ignore Histrogram::Add and friends, see http://crbug.com/62694.
fun_r:base::Histogram::Add
fun_r:base::HistogramSamples::Add
fun_r:base::HistogramBase::AddTime
# ffmpegsumo.dll appears to read a few bytes beyond the end of the buffer.
fun:_ff_prefetch_mmxext
# Shows up as a race in SHELL32.dll when deleting a directory while opening an
# unrelated file in another thread. Revealed by DiskCacheBackendTest.DeleteOld.
# See: https://code.google.com/p/data-race-test/issues/detail?id=114
fun_r:SHFileOperationW
# Ignore internal file I/O synchronization: crbug.com/146724
fun_r:_lock_file
fun_r:_lock_file2
fun_r:_lock
fun_r:_flsbuf
fun_r:_unlock_file
fun_r:_getstream
# http://crbug.com/272065
obj:*NLAapi.dll