DCHECK failure in memory_instrumentation |
||
Issue description[8746:8769:0919/162750.107899:FATAL:interface_endpoint_client.cc(32)] Check failed: !is_valid. The callback passed to Coordinator::RequestGlobalMemoryDump() was never run. https://luci-logdog.appspot.com/v/?s=chromium%2Fbb%2Ftryserver.chromium.linux%2Flinux_chromium_asan_rel_ng%2F454401%2F%2B%2Frecipes%2Fsteps%2Fbrowser_tests__with_patch_%2F0%2Flogs%2FTaskManagerBrowserTest.NavigateAwayFromHungRenderer%2F0 Caused by this CL: https://chromium-review.googlesource.com/c/chromium/src/+/646047 It looks like if the renderer main thread is hung, we hit this DCHECK. This seems less than ideal. It may also account for some of the missing UMA stats we were seeing? Digging a little bit, ClientProcessImpl binds to the interface on the renderer main thread...we should probably clone the connector, and then bind to a background thread? maybe io thread?
,
Sep 21 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/6995ce043e1835be5d5fb7785112d8f1ff1fe423 commit 6995ce043e1835be5d5fb7785112d8f1ff1fe423 Author: erikchen <erikchen@chromium.org> Date: Thu Sep 21 06:49:43 2017 Fix member destruction order for memory_instrumentation::CoordinatorImpl. The bindings must be destroyed before extant callbacks. Bug: 767128 Change-Id: I5670829ec677f929a6fed74761f319a1c4486035 Reviewed-on: https://chromium-review.googlesource.com/675985 Commit-Queue: Primiano Tucci <primiano@chromium.org> Reviewed-by: Primiano Tucci <primiano@chromium.org> Cr-Commit-Position: refs/heads/master@{#503371} [modify] https://crrev.com/6995ce043e1835be5d5fb7785112d8f1ff1fe423/services/resource_coordinator/memory_instrumentation/coordinator_impl.h
,
Sep 25 2017
|
||
►
Sign in to add a comment |
||
Comment 1 by erikc...@chromium.org
, Sep 20 2017Status: Assigned (was: Untriaged)