Created attachment 318085 [details] Crash Log The test imported/w3c/web-platform-tests/IndexedDB/idbcursor-key.htm is a flaky assertion crash on Sierra Debug WK2 according to the dashboard. the first crash was with this build https://build.webkit.org/builders/Apple%20Sierra%20Debug%20WK2%20(Tests)/builds/2279 ASSERTION FAILED: !m_owningPointerForClose /Volumes/Data/slave/sierra-debug/build/Source/WebCore/Modules/indexeddb/server/UniqueIDBDatabase.cpp(277) : void WebCore::IDBServer::UniqueIDBDatabase::scheduleShutdownForClose() 1 0x11b48bc5d WTFCrash 2 0x111b2e730 WebCore::IDBServer::UniqueIDBDatabase::scheduleShutdownForClose() 3 0x111b29990 WebCore::IDBServer::UniqueIDBDatabase::operationAndTransactionTimerFired() 4 0x111b48d7d WTF::Function<void ()>::CallableWrapper<std::__1::__bind<void (WebCore::IDBServer::UniqueIDBDatabase::*&)(), WebCore::IDBServer::UniqueIDBDatabase*> >::call() 5 0x10f02accb WTF::Function<void ()>::operator()() const 6 0x10f02ab29 WebCore::Timer::fired() 7 0x111aec540 WebCore::ThreadTimers::sharedTimerFiredInternal() 8 0x111aecca1 WebCore::ThreadTimers::setSharedTimer(WebCore::SharedTimer*)::$_0::operator()() const 9 0x111aecc59 WTF::Function<void ()>::CallableWrapper<WebCore::ThreadTimers::setSharedTimer(WebCore::SharedTimer*)::$_0>::call() 10 0x10f02accb WTF::Function<void ()>::operator()() const 11 0x110dce985 WebCore::MainThreadSharedTimer::fired() 12 0x110dced19 WebCore::timerFired(__CFRunLoopTimer*, void*) 13 0x7fffcdce6e04 __CFRUNLOOP_IS_CALLING_OUT_TO_A_TIMER_CALLBACK_FUNCTION__ 14 0x7fffcdce6a93 __CFRunLoopDoTimer 15 0x7fffcdce65ea __CFRunLoopDoTimers 16 0x7fffcdcddfc1 __CFRunLoopRun 17 0x7fffcdcdd544 CFRunLoopRunSpecific 18 0x7fffcf70e4e2 -[NSRunLoop(NSRunLoop) runMode:beforeDate:] 19 0x7fffcf70e3ba -[NSRunLoop(NSRunLoop) run] 20 0x7fffe3c6e89b _xpc_objc_main 21 0x7fffe3c6d2e4 xpc_main 22 0x108ae8105 main 23 0x7fffe3a15235 start 24 0x1
<rdar://problem/33886906>
There's no reason to believe this isn't the same root cause as all of the other tests that hit the same ASSERT
A similar issue has been fixed in Bug 174354. Can anyone confirm this still happens?