e023491600
Change-Id: Idf0a98869e46de08165aacb7f3836b0d6c3dffcf Reviewed-on: https://gerrit.libreoffice.org/c/core/+/170335 Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk> Tested-by: Jenkins
123 lines
4.2 KiB
Text
123 lines
4.2 KiB
Text
#
|
|
# This file is part of the LibreOffice project.
|
|
#
|
|
# This Source Code Form is subject to the terms of the Mozilla Public
|
|
# License, v. 2.0. If a copy of the MPL was not distributed with this
|
|
# file, You can obtain one at http://mozilla.org/MPL/2.0/.
|
|
#
|
|
|
|
# Use via environment variable TSAN_OPTIONS=suppressions=.../tsan-suppress.txt
|
|
|
|
# There looks to be a lock ordering problem here, but I can't see how it could
|
|
# actually be hit in practice.
|
|
deadlock:cppuhelper::ServiceManager::disposing()
|
|
deadlock:cppuhelper::ServiceManager::loadImplementation
|
|
deadlock:AffineBridge::v_callInto_v
|
|
|
|
# Ignore stuff in external DBUS library.
|
|
# Some kind of dbus lock issue when we call it from psp::CUPSManager.
|
|
deadlock:_dbus_lock
|
|
race:g_dbus_connection_signal_unsubscribe
|
|
race:g_path_get_basename
|
|
race:g_date_time_format_iso8601
|
|
race:g_realloc
|
|
|
|
# inside an assert so I don't care.
|
|
race:cppu::OWeakConnectionPoint::acquire
|
|
race:AffineBridge::v_enter
|
|
race:__vsnprintf_chk
|
|
|
|
# right now, I'm not interested in deadlocks at all, too many false+
|
|
deadlock:
|
|
|
|
# This is checking SAL_STRING_IS_STATIC, which is safe because that is written at compile time.
|
|
# Depending on the optimisation/debug level, the call stack may have different leaves, hence
|
|
# specifying this in more than one way.
|
|
race:rtl::str::acquire<_rtl_uString>
|
|
race:rtl::str::release<_rtl_uString>
|
|
race:rtl_uString_acquire
|
|
race:rtl_uString_release
|
|
race:rtl_uString_assign
|
|
|
|
# I've convinced myself this is a false+, caused by ping-ponging the buffer between two
|
|
# threads, but I might be wrong.
|
|
race:XBufferedThreadedStream::getNextBlock
|
|
|
|
# I am not interested in stuff the embedded JVM does.
|
|
race:libjvm.so
|
|
|
|
# I think this is OK, because at this point we are doing
|
|
# if (nRefCount > 1)
|
|
# and we know from our callers that the refcount must be at least one
|
|
# so there is no failure mode.
|
|
race:ireallocSequence
|
|
|
|
# TODO There appears to be a race here, initialising the
|
|
# ::com::sun::star::uno::Sequence< T >::s_pType
|
|
# field. But no idea at all how to fix it.
|
|
race:cppu::getTypeFavourUnsigned
|
|
|
|
|
|
# This is all inside GIO/Glib, no idea what it is doing.
|
|
#
|
|
race:slab_allocator_alloc_chunk
|
|
race:g_source_destroy_internal
|
|
race:g_source_unref_internal
|
|
race:g_task_finalize
|
|
race:g_socket_send_message_with_timeout
|
|
race:g_idle_source_new
|
|
race:g_slice_alloc0
|
|
race:g_str_has_suffix
|
|
race:g_str_equal
|
|
race:g_strsplit
|
|
race:g_variant_store
|
|
|
|
# Python build, lock order inversion
|
|
deadlock:take_gil
|
|
|
|
# Not interested in deadlock issues when they involve the shutdown path
|
|
# - the shutdown path is special, so if we see a lock-ordering here that conflicts with
|
|
# a "normal" lock-ordering path, that is highly unlikely to result in a real issue.
|
|
deadlock:DeInitVCL
|
|
|
|
# Sometimes tsan will report data-races in these, who knows why, they should be thread-safe
|
|
race:malloc
|
|
race:free
|
|
race:close
|
|
|
|
# There is a lock-cycle here, together with the mutexes acquired in
|
|
# SvtSysLocaleOptions::SvtSysLocaleOptions
|
|
# SvtSysLocale::SvtSysLocale
|
|
# but I can't see how it could lead to a problem in practice
|
|
deadlock:ItemHolder1::impl_addItem
|
|
|
|
# ignore warning about nRefCount in sal_Sequence.
|
|
# <sberg> noelgrandin, we generally assume that sal_Int32 is small enough, and hardware sane enough,
|
|
# that reading it will always give a sane value, i.e., if the read happens in parallel with a write,
|
|
# the read produces either the old, original value, or the newly written one.
|
|
race:cppu::createEmptySequence
|
|
race:uno_type_sequence_reference2One
|
|
|
|
# ignore harmless data race on m_nTimeoutMS
|
|
race:GtkSalTimer::Start
|
|
|
|
# TODO data-race with fields on salhelper::Timer, but I see no way of fixing this because this class
|
|
# is in URE
|
|
race:salhelper::Timer
|
|
|
|
# TODO LanguageTag is copying data around without any locking at all, but it seems to be touched
|
|
# from multiple threads. I am nervous to randomly add locks because it is performance sensitive.
|
|
race:LanguageTag::
|
|
|
|
# harmless race on m_storedTypeRefs
|
|
race:cppu::getTypeEntries
|
|
|
|
# double-checked locking in our
|
|
# inline ::css::uno::Type const & cppu_detail_getUnoType(css:uno:Foo const *)
|
|
# generated code
|
|
race:typelib_typedescription_newEmpty
|
|
race:typelib_typedescriptionreference_getByName
|
|
|
|
# harmless race where two fields are sharing an address slot/word
|
|
race:ScFormulaCell::GetMatrixFlag
|
|
race:ScFormulaCell::InterpretTail
|