Closed Bug 1016923 Opened 11 years ago Closed 11 years ago

crash in JS_NewStringCopyZ(JSContext*, char const*)

Categories

(Toolkit :: Telemetry, defect)

Product:
Toolkit
Component:
Telemetry
Platform:
x86
Windows 7
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

()

Status:
RESOLVED WORKSFORME

People

(Reporter: ted, Assigned: jchen)

References

Details

(Keywords: crash)

Crash Data

This bug was filed from the Socorro interface and is report bp-2ab6468f-3e33-4b9f-9e6e-4de112140526. ============================================================= I've crashed while restarting to apply updates the past few days: https://crash-stats.mozilla.com/report/index/ac62f98f-9728-43df-9606-ce2792140528 https://crash-stats.mozilla.com/report/index/8fdf5c67-86d4-47a7-aed9-321272140528 https://crash-stats.mozilla.com/report/index/2ab6468f-3e33-4b9f-9e6e-4de112140526 It's crashing under CreateJSHangHistogram.
The stacks on those crashes are all over the place. I wonder if this frame ought to be in the skiplist as an append to break out the callers.
Depends on: 1017044
Ted's were all from `anonymous namespace'::CreateJSHangHistogram referencing hangstack[i][0] from http://hg.mozilla.org/mozilla-central/annotate/e86a0d92d174/toolkit/components/telemetry/Telemetry.cpp#l2076 I think you're right, we should add this to the append list. Files bug 1017044.
There was a regression in the profiler pseudostack code (which the hang monitor depends on) from 5-22 nightly to 5-27 nightly (bug 1004726). I'd wait to see if more reports come in.
Blocks: 1004726
Flags: needinfo?(nchen)
Still happening :( There was no crash from when bug 1004726 was backed out (5-28 nightly) but not sure if that's the cause.
Assignee: nobody → nchen
Status: NEW → ASSIGNED
Flags: needinfo?(nchen)
I can reliably reproduce this on basically every nightly update. Also, all of my plugin-container processes end up crashing as well. https://crash-stats.mozilla.com/report/index/0ab4b725-a39d-4ba4-a0d1-b38ac2140601 etc. Would attaching a debugger before I do the next update help gather any useful info?
(In reply to Vladimir Vukicevic [:vlad] [:vladv] from comment #7) > I can reliably reproduce this on basically every nightly update. Also, all > of my plugin-container processes end up crashing as well. > > https://crash-stats.mozilla.com/report/index/0ab4b725-a39d-4ba4-a0d1- > b38ac2140601 > etc. > > Would attaching a debugger before I do the next update help gather any > useful info? That'd be great! I want to see what |i| is and what the other entries of |hangStack| are in this line, http://hg.mozilla.org/mozilla-central/annotate/cbe4f69c2e9c/toolkit/components/telemetry/Telemetry.cpp#l2076
Flags: needinfo?(vladimir)
I think I know what's going on. If you have the profiler running at some point before the restart, then the crash is most likely caused by bug 1021277.
I have the profiler extension installed, so that's definitely true.
I don't see any more crashes after the 6/6 nightly.
Status: ASSIGNED → RESOLVED
Closed: 11 years ago
Flags: needinfo?(vladimir)
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.