Direct-leak in content::WebServiceWorkerRegistrationImpl::CreateForServiceWorkerGlobalScope |
||||
Issue descriptionDetailed report: https://clusterfuzz.com/testcase?key=5152276105396224 Fuzzer: inferno_twister Job Type: linux_lsan_chrome_mp Platform Id: linux Crash Type: Direct-leak Crash Address: Crash State: content::WebServiceWorkerRegistrationImpl::CreateForServiceWorkerGlobalScope content::ServiceWorkerProviderContext::TakeRegistrationForServiceWorkerGlobalSco content::ServiceWorkerContextClient::WorkerContextStarted Sanitizer: address (ASAN) Regressed: https://clusterfuzz.com/revisions?job=linux_lsan_chrome_mp&range=514498:517698 Reproducer Testcase: https://clusterfuzz.com/download?testcase_id=5152276105396224 Additional requirements: Requires HTTP Issue filed automatically. See https://github.com/google/clusterfuzz-tools for more information.
,
Feb 26 2018
Automatically assigning owner based on suspected regression changelist https://chromium.googlesource.com/chromium/src/+/86029dbfbf1c4927e12c8f8b09de04d706553472 ([ServiceWorker] Eliminate blink.mojom.ServiceWorkerRegistrationObjectInfo.handle_id). If this is incorrect, please let us know why and apply the Test-Predator-Wrong-CLs label. If you aren't the correct owner for this issue, please unassign yourself as soon as possible so it can be re-triaged.
,
Feb 28 2018
,
Feb 28 2018
One possible scenario for this leak: The service worker thread has already stopped, then WebServiceWorkerRegistrationImpl::OnConnectionError() is invoked on IO thread, which tries to post a task to our stopped service worker thread to destroy WebServiceWorkerRegistrationImpl, the PostTask failed, so the WebServiceWorkerRegistrationImpl instance has no any other chance to be destroyed .
,
Mar 2 2018
Hm, is this expected to go away once Mojo migration is done and we don't need to do the channel-associated interface anymore? It's also possible the leak detector is just not waiting long enough for the worker-related stuff to be fully cleaned up.
,
Mar 2 2018
Exactly, c#4 scenario is because we're using Channel-associated interface. So I think we can fix this easily after we decouple our interfaces from Channel-associated world later.
,
Mar 2 2018
Thanks. Since we're getting close to full Mojo (right?), it's probably fine to ignore this bug until then.
,
Mar 2 2018
Got it. #Yeah I assume our full Mojo would be achieved within Q1.
,
Mar 12 2018
ClusterFuzz testcase 5152276105396224 is flaky and no longer crashes, so closing issue. If this is incorrect, please add ClusterFuzz-Wrong label and re-open the issue. |
||||
►
Sign in to add a comment |
||||
Comment 1 by ClusterFuzz
, Feb 26 2018Labels: Test-Predator-Auto-Components