Crash in [@ mozilla::dom::(anonymous namespace)::QuotaClient::ShutdownTimedOut]
Categories
(Core :: Storage: localStorage & sessionStorage, defect, P1)
Tracking
()
Tracking | Status | |
---|---|---|
firefox-esr60 | --- | unaffected |
firefox-esr68 | --- | disabled |
firefox67 | --- | unaffected |
firefox67.0.1 | --- | unaffected |
firefox68 | + | disabled |
firefox69 | + | disabled |
firefox70 | + | fixed |
firefox71 | --- | fixed |
People
(Reporter: marcia, Assigned: janv)
References
(Blocks 1 open bug)
Details
(Keywords: crash, regression)
Crash Data
This bug is for crash report bp-a96fa52f-3190-4f3e-99a0-efc020190404.
Two crashes (One Mac and one Linux) in Build 20190404063228 - Moz crash reason MOZ_CRASH(LocalStorage shutdown timed out). Maybe expected due to Bug 1539750.
Top 10 frames of crashing thread:
0 XUL mozilla::dom:: dom/localstorage/ActorsParent.cpp:8114
1 XUL mozilla::dom:: const dom/localstorage/ActorsParent.cpp:8061
2 XUL mozilla::dom:: dom/localstorage/ActorsParent.cpp:8061
3 XUL nsTimerEvent::Run xpcom/threads/nsTimerImpl.cpp:559
4 XUL nsThread::ProcessNextEvent xpcom/threads/nsThread.cpp:1180
5 XUL NS_ProcessNextEvent xpcom/threads/nsThreadUtils.cpp:486
6 XUL mozilla::dom:: xpcom/threads/nsThreadUtils.h:348
7 XUL mozilla::dom::quota::QuotaManager::Shutdown dom/quota/ActorsParent.cpp:3344
8 XUL mozilla::dom::quota::RecvShutdownQuotaManager dom/quota/ActorsParent.cpp:2958
9 XUL mozilla::ipc::BackgroundParentImpl::RecvShutdownQuotaManager ipc/glue/BackgroundParentImpl.cpp:973
Assignee | ||
Updated•6 years ago
|
Assignee | ||
Comment 1•6 years ago
|
||
P2, given it can take long time to get rid of this completely.
Comment 2•6 years ago
|
||
Could this be a dupe of bug 1541776?
Assignee | ||
Comment 3•6 years ago
|
||
No, that one tracks shutdown hangs caused by IndexedDB. The one here, tracks new LocalStorage.
Comment 4•6 years ago
|
||
Next step here is to add more debugging information to crash annotations.
Updated•5 years ago
|
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Reporter | ||
Comment 5•5 years ago
|
||
This crash is pretty visible in 68 beta - 170 crashes in one of the signatures in b4.
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 6•5 years ago
|
||
According to:
https://crash-stats.mozilla.org/report/index/85c4699d-f196-4cc6-b7f1-0b79d0190609#allthreads
the LS thread has WaitForUnlockNotify in the stack trace. That can cause content process hangs like in bug 1554321.
It seems it can also block shutdown.
:asuth already investigated this in bug 1554321 and I was able to simulate it with my local build.
I'm going to file a new bug (with a patch) for that, so it should be fixed very soon.
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 7•5 years ago
|
||
Ok, we have zero crashes (LocalStorage shutdown hangs) for last 3 nightly builds. It seems bug 1559029 was causing most of the problems.
Reporter | ||
Comment 8•5 years ago
|
||
(In reply to Jan Varga [:janv] from comment #7)
Ok, we have zero crashes (LocalStorage shutdown hangs) for last 3 nightly builds. It seems bug 1559029 was causing most of the problems.
Jan - There are still a handful of crashes recently on 69 nightly, but I cannot tell whether they are the LocalStorage shutdown hangs or not. For instance, https://crash-stats.mozilla.org/report/index/fa86228d-f955-4cde-93cc-d1cb10190626 seems to have some of that at the top of the stack.
Assignee | ||
Comment 9•5 years ago
|
||
Yes, there are still some, I'll be adding more debugging info for that which should help with fixing remaining crashes/shutdown hangs.
Comment 10•5 years ago
|
||
This should go away for 68 when bug 1562942 lands in RC2.
Updated•5 years ago
|
Comment 11•5 years ago
|
||
Hi Jan, this signature is still sticking around in the 69 betas. Any thoughts on how to proceed here given that we're midway through the Beta cycle?
Updated•5 years ago
|
Comment 13•5 years ago
|
||
Jan is working on another high volume crash. This is in the queue after that.
Assignee | ||
Comment 14•5 years ago
|
||
It seems bug 1581127 helped with reducing LS shutdown hangs.
Assignee | ||
Comment 15•5 years ago
|
||
Yeah, the statistics look better, but there are still some unexplained shutdown blockers. I'm going to add more debugging info for that.
Assignee | ||
Updated•5 years ago
|
Crash volume is still very high for beta, is there anything else you can investigate here?
Assignee | ||
Comment 17•5 years ago
|
||
I'm working on a fix. We had to add more debugging information to crash stats to see what remaining problem is. The fix should be ready this week.
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 18•5 years ago
|
||
Marking as fixed given low frequency. Remaining issues are tracked in bug 1588498.
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Description
•