# # 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 # Ignore stuff in external DBUS library # Some kind of dbus lock issue when we call it from psp::CUPSManager deadlock:_dbus_lock # 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 race:rtl::str::acquire<_rtl_uString> race:rtl::str::release<_rtl_uString> # 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 # not introduced 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