Closed Bug 1140916 Opened 9 years ago Closed 8 years ago

Intermittent test_nsITextInputProcessor.xul,browser_rules_search-filter_04.js,browser_scratchpad_restore.js | application crashed [@ js::PCToLineNumber(unsigned int,unsigned char *,unsigned char *,unsigned char *,unsigned int *)]

Categories

(Core :: JavaScript Engine, defect, P3)

x86_64
Windows 8
defect

Tracking

()

RESOLVED DUPLICATE of bug 1302914

People

(Reporter: philor, Unassigned)

References

Details

(Keywords: intermittent-failure)

      No description provided.
[Mass Closure] Closing Intermittent as a one off
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → WORKSFORME
Status: RESOLVED → REOPENED
Resolution: WORKSFORME → ---
jandem, see the incoming See Alsos. Can you please help find an owner for this widespread failure?
Component: DOM → JavaScript Engine
Flags: needinfo?(jdemooij)
See Also: → 1242067
See Also: → 1264415
See Also: → 1267483
See Also: → 1268504
See Also: → 1273317
See Also: → 1273318
(In reply to Ryan VanderMeulen [:RyanVM] from comment #10)
> jandem, see the incoming See Alsos. Can you please help find an owner for
> this widespread failure?

These are all Win64, looks like. The BackgroundHangMonitor crashes while it tries to get a JS stack trace.

Not sure why we're crashing - maybe a stale pseudo stack entry? Does the background hang monitor enable the profiler? Jim, can you please take a look?
Flags: needinfo?(jdemooij) → needinfo?(nchen)
See Also: → 1274851
(In reply to Jan de Mooij [:jandem] from comment #11)
> (In reply to Ryan VanderMeulen [:RyanVM] from comment #10)
> > jandem, see the incoming See Alsos. Can you please help find an owner for
> > this widespread failure?
> 
> These are all Win64, looks like. The BackgroundHangMonitor crashes while it
> tries to get a JS stack trace.
> 
> Not sure why we're crashing - maybe a stale pseudo stack entry? Does the
> background hang monitor enable the profiler? Jim, can you please take a look?

It doesn't use the profiler, just the pseudo-stack. It does use js::PCToLineNumber on a background thread, so maybe something changed about the way PCToLineNumber works that made it prone to crashing on a background thread?
Flags: needinfo?(nchen)
Just wondering how many dupes need to get filed before we can prioritize someone investigating this issue.
Flags: needinfo?(nihsanullah)
See Also: → 1279083, 1280562, 1281301
@Jan: 
This current orange spike started May 18thish (https://brasstacks.mozilla.com/orangefactor/?display=Bug&bugid=1140916&startday=2015-07-04&endday=2016-07-10&tree=all). The last time the function was touched looks to be back in December (https://hg.mozilla.org/mozilla-central/annotate/6b329e5cac86/js/src/jsscript.cpp#l3244). 

If the failure volume is steadily > 1 a day can we instrument the function in Nightly to try and work backwards and figure out the source of the crash? What other strats could we try?
Flags: needinfo?(nihsanullah) → needinfo?(jdemooij)
(In reply to Naveed Ihsanullah [:naveed] from comment #19)
> If the failure volume is steadily > 1 a day can we instrument the function
> in Nightly to try and work backwards and figure out the source of the crash?
> What other strats could we try?

I will try to reproduce this with rr later this week. If that doesn't work we can try instrumentation.
See Also: → 1287689
I'm working on this cluster of intermittents now.
(In reply to Jan de Mooij [:jandem] from comment #22)
> I'm working on this cluster of intermittents now.

I haven't been able to reproduce this yet, but this could be related to bug 1295775. I'll check next week if we still see these signatures after that landed.
Bulk assigning P3 to all open intermittent bugs without a priority set in Firefox components per bug 1298978.
Priority: -- → P3
Summary: Intermittent test_nsITextInputProcessor.xul | application crashed [@ js::PCToLineNumber(unsigned int,unsigned char *,unsigned char *,unsigned char *,unsigned int *)] → Intermittent test_nsITextInputProcessor.xul,browser_rules_search-filter_04.js | application crashed [@ js::PCToLineNumber(unsigned int,unsigned char *,unsigned char *,unsigned char *,unsigned int *)]
Summary: Intermittent test_nsITextInputProcessor.xul,browser_rules_search-filter_04.js | application crashed [@ js::PCToLineNumber(unsigned int,unsigned char *,unsigned char *,unsigned char *,unsigned int *)] → Intermittent test_nsITextInputProcessor.xul,browser_rules_search-filter_04.js,browser_scratchpad_restore.js | application crashed [@ js::PCToLineNumber(unsigned int,unsigned char *,unsigned char *,unsigned char *,unsigned int *)]
This whole cluster was likely fixed by bug 1302914.
Status: REOPENED → RESOLVED
Closed: 9 years ago8 years ago
Flags: needinfo?(jdemooij)
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.