Closed Bug 1599987 Opened 4 years ago Closed 4 years ago

Crash in [@ mozilla::dom::ScreenOrientation::QueryInterface]

Categories

(Core :: DOM: Core & HTML, defect, P3)

Unspecified
Windows 10
defect

Tracking

()

RESOLVED INCOMPLETE
Tracking Status
firefox-esr68 --- unaffected
firefox70 --- unaffected
firefox71 --- unaffected
firefox72 --- fix-optional

People

(Reporter: calixte, Unassigned)

References

(Blocks 1 open bug, Regression)

Details

(Keywords: crash, regression)

Crash Data

This bug is for crash report bp-5f5e0ffc-19c4-4c25-9b93-bf1d90191128.

Top 10 frames of crashing thread:

0 xul.dll nsresult mozilla::dom::ScreenOrientation::QueryInterface dom/base/ScreenOrientation.cpp:29
1  @0x3c1500ffff 
2 xul.dll void nsPurpleBuffer::RemoveSkippable xpcom/base/nsCycleCollector.cpp:2588
3 xul.dll void nsCycleCollector::ForgetSkippable xpcom/base/nsCycleCollector.cpp:2659
4 xul.dll nsCycleCollector_forgetSkippable xpcom/base/nsCycleCollector.cpp:3863
5 xul.dll static void FireForgetSkippable dom/base/nsJSEnvironment.cpp:1258
6 xul.dll static bool CCRunnerFired dom/base/nsJSEnvironment.cpp:1877
7 xul.dll std::_Func_impl_no_alloc<bool  
8 xul.dll nsresult mozilla::IdleTaskRunner::Run xpcom/threads/IdleTaskRunner.cpp:58
9 xul.dll static void mozilla::TimedOut xpcom/threads/IdleTaskRunner.cpp:78

There is 1 crash in nightly 72 with buildid 20191127215655. In analyzing the backtrace, the regression may have been introduced by patch [1] to fix bug 1599445.

[1] https://hg.mozilla.org/mozilla-central/rev?node=f8533b22dc9b

Flags: needinfo?(jcoppeard)

I don't think that change should be responsible for a crash here; it doesn't change anything about the way nsCycleCollector_forgetSkippable works.

Flags: needinfo?(jcoppeard)

I agree. I tried to look at ScreenOrientation implementation and couldn't see any obvious problems.
Of course the stack has that weird @0x3c1500ffff, so perhaps the crash isn't about ScreenOrientation at all.

I'll try to keep an eye on this. And Calixte if you see this crash more often, please tell me about it.

But for now this isn't really actionable, when we don't even have a proper stack.

Priority: -- → P3

This is very low volume (1 crash in June, 1 in July, 1 in November).

Yeah, I think this is just the crash regression detector getting some false signal from the vast amounts of stuff that ends up with FireForgetSkippable in the stack.

Status: NEW → RESOLVED
Closed: 4 years ago
Resolution: --- → INCOMPLETE
Has Regression Range: --- → yes
You need to log in before you can comment on or make changes to this bug.