Closed Bug 1422600 Opened 7 years ago Closed 6 years ago

WebRender: Crashes not reported?

Categories

(Core :: Graphics: WebRender, defect)

defect
Not set
normal

Tracking

()

RESOLVED INCOMPLETE

People

(Reporter: emilio, Assigned: kats)

References

(Blocks 1 open bug)

Details

I got a whole-browser crash today, and my stderr looked like:

thread 'WRRenderBackend#7' panicked at 'called `Option::unwrap()` on a `None` value', /checkout/src/libcore/option.rs:335:20
note: Run with `RUST_BACKTRACE=1` for a backtrace.
Redirecting call to abort() to mozalloc_abort

I was curious about the crash, but there's no crash report created for it at all...

The rest of the stderr after that is:

ExceptionHandler::GenerateDump cloned child ExceptionHandler::WaitForContinueSignal waiting for continue signal...
14255
ExceptionHandler::SendContinueSignalToChild sent continue signal to child
[Child 2947, Chrome_ChildThread] WARNING: pipe error: Broken pipe: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 709
[Child 2808, Chrome_ChildThread] WARNING: pipe error (40): Connection reset by peer: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 353
[Child 2788, Chrome_ChildThread] WARNING: pipe error (34): Connection reset by peer: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 353
[Child 2808, Chrome_ChildThread] WARNING: pipe error: Broken pipe: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 709
[Child 2788, Chrome_ChildThread] WARNING: pipe error: Broken pipe: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 709
[Child 2713, Chrome_ChildThread] WARNING: pipe error: Broken pipe: file /builds/worker/workspace/build/src/ipc/chromium/src/chrome/common/ipc_channel_posix.cc, line 709
[GFX1-]: Receive IPC close with reason=AbnormalShutdown

I'm not sure what's going on or if it's expected for those crashes not to generate a crash report, but I thought it was worth reporting.
> called `Option::unwrap()` on a `None` value

Seen in:
* bug 1413571
* bug 1416602
* bug 1418201 comment 0 [@ @0x4084a1 ] bp-6370bffa-c622-4ee6-a4c5-a929d0171117
Blocks: wr-stability
Whiteboard: [wr-mvp] [triage]
Assignee: nobody → bugmail
Priority: -- → P1
Status: NEW → ASSIGNED
Whiteboard: [wr-mvp] [triage] → [wr-reserve]
See Also: → 1423311
We have STR to reproduce a crash in bug 1423311, so I did that on an OS X nightly build. The crash reporter popped up as expected and I was able to report the crash - https://crash-stats.mozilla.com/report/index/eba05327-2926-4c0b-8276-802070171207 is one example. One thing to note is that the crash reporter dialog was a background window, hidden by other things, which was kind of surprising. But the crash showed up in about:crashes regardless and I could submit it from there.

So basically, I'm unable to reproduce the reported scenario where the crash doesn't get reported. Emilio, if you load the attachment in bug 1423311 in a nightly build (make sure webrender and gfx.webrender.blob-images are both on) you should be able to trigger crashes, can you check again to see if you can reproduce the problem where the crashes don't get reported?
Flags: needinfo?(emilio)
Yeah, it was properly reported here too... I'm pretty sure I saw no crash report window anywhere when this crash happened, but as I don't have any STR I don't think there's anything actionable here unfortunately... I'll keep an eye out if this happens again.
Status: ASSIGNED → RESOLVED
Closed: 6 years ago
Flags: needinfo?(emilio)
Resolution: --- → INCOMPLETE
Priority: P1 → --
Whiteboard: [wr-reserve]
Thanks for checking!
You need to log in before you can comment on or make changes to this bug.