No callstack from layout test crashes on trybots |
|||
Issue descriptionThere is no human-readable callstack in crash reports from layout test run on trybots. Example: https://test-results.appspot.com/data/layout_results/linux_chromium_rel_ng/213415/site_per_process_webkit_layout_tests%20%28with%20patch%29/layout-test-results/results.html (this is from https://ci.chromium.org/p/chromium/builders/luci.chromium.try/linux_chromium_rel_ng/213415)
,
Oct 19
I don't have much context/familiarity with the test-results url you linked, but if you go to the build, click on one of the "shard" links for the failed site_per_process_webkit_layout_tests test, you can see what appears to be symbolized stack traces. eg: https://chromium-swarm.appspot.com/task?id=40a336f8c803a510 has a bunch of 20:46:22.478 13008 [15130:15139:1018/204622.262706:FATAL:session_storage_namespace_impl_mojo.cc(136)] Check failed: ::content::BrowserThread::CurrentlyOn(BrowserThread::UI). Must be called on Chrome_UIThread; actually called on Chrome_IOThread. 20:46:22.478 13008 #0 0x0000041e962f base::debug::StackTrace::StackTrace() 20:46:22.478 13008 #1 0x00000414027b logging::LogMessage::~LogMessage() 20:46:22.478 13008 #2 0x000003980e3d content::SessionStorageNamespaceImplMojo::OpenArea() 20:46:22.478 13008 #3 0x00000219c30f blink::mojom::SessionStorageNamespaceStubDispatch::Accept() 20:46:22.478 13008 #4 0x000004228ca2 mojo::InterfaceEndpointClient::HandleValidatedMessage() 20:46:22.478 13008 #5 0x000004228826 mojo::FilterChain::Accept() 20:46:22.478 13008 #6 0x00000422a1a2 mojo::InterfaceEndpointClient::HandleIncomingMessage() 20:46:22.478 13008 #7 0x00000423170c mojo::internal::MultiplexRouter::ProcessIncomingMessage() 20:46:22.478 13008 #8 0x000004230ad0 mojo::internal::MultiplexRouter::Accept() 20:46:22.478 13008 #9 0x000004228826 mojo::FilterChain::Accept() 20:46:22.478 13008 #10 0x000004226003 mojo::Connector::ReadSingleMessage() 20:46:22.478 13008 #11 0x000004226bf4 mojo::Connector::ReadAllAvailableMessages() 20:46:22.478 13008 #12 0x000004226a56 mojo::Connector::OnHandleReadyInternal() 20:46:22.478 13008 #13 0x0000022ccc54 mojo::SimpleWatcher::DiscardReadyState() 20:46:22.478 13008 #14 0x00000421a9a3 mojo::SimpleWatcher::OnHandleReady() 20:46:22.478 13008 #15 0x0000024bba8c _ZN4base8internal7InvokerINS0_9BindStateIMN3viz14GpuServiceImplEFvN3gfx21GenericSharedMemoryIdEiRKN3gpu9SyncTokenEEJNS_7WeakPtrIS4_EES6_iS8_EEEFvvEE7RunImplISC_NSt3__15tupleIJSE_S6_iS8_EEEJLm0ELm1ELm2ELm3EEEEvOT_OT0_NSJ_16integer_sequenceImJXspT1_EEEE 20:46:22.478 13008 #16 0x000004149792 base::debug::TaskAnnotator::RunTask() 20:46:22.478 13008 #17 0x0000041486df base::MessageLoop::RunTask() 20:46:22.478 13008 #18 0x000004148ab2 base::MessageLoop::DoWork() 20:46:22.478 13008 #19 0x000004209f69 base::MessagePumpLibevent::Run() 20:46:22.478 13008 #20 0x0000041481b1 base::MessageLoop::Run() 20:46:22.478 13008 #21 0x00000416db76 base::RunLoop::Run() 20:46:22.478 13008 #22 0x0000041baf4a base::Thread::Run() 20:46:22.478 13008 #23 0x000003866684 content::BrowserProcessSubThread::IOThreadRun() 20:46:22.478 13008 #24 0x0000038665cf content::BrowserProcessSubThread::Run() 20:46:22.478 13008 #25 0x0000041bb2c5 base::Thread::ThreadMain() 20:46:22.478 13008 #26 0x0000041fc98f base::(anonymous namespace)::ThreadFunc() 20:46:22.478 13008 #27 0x7fe05fe88184 start_thread 20:46:22.479 13008 #28 0x7fe05b43903d clone Does that help at all?
,
Oct 19
,
Oct 19
RE: #c3 Thanks - the callstack you posted does help! RE: I don't have much context/familiarity with the test-results url linked This was "archive results for webkit_layout_tests (with patch)" step (#114) and "layout_test_results" link. I think this is generated while running layout tests?
,
Oct 19
Ahh I see. What if you: - open up https://test-results.appspot.com/data/layout_results/linux_chromium_rel_ng/213415/site_per_process_webkit_layout_tests%20%28with%20patch%29/layout-test-results/results.html - click on one of the failed tests - click on the "stderr" link It looks like the crash is unintuitively in the "stderr" tab, not the "crash log" tab. I guess you just have to know where to look. |
|||
►
Sign in to add a comment |
|||
Comment 1 by lukasza@chromium.org
, Oct 19