Closed Bug 1151643 Opened 9 years ago Closed 2 years ago

[meta] Crashes in NS_CycleCollectorSuspect3

Categories

(Core :: XPCOM, defect)

defect
Not set
normal

Tracking

()

RESOLVED WORKSFORME
Tracking Status
firefox40 --- affected
firefox42 --- affected
firefox48 --- affected
firefox49 --- affected
firefox-esr45 --- affected
firefox50 --- affected
firefox51 --- affected
firefox52 --- wontfix

People

(Reporter: mccr8, Unassigned)

References

Details

(Keywords: crash, meta)

Crash Data

Attachments

(1 file)

Crashes with this signature indicate that a cycle collected object is either being used after the cycle collector is shut down, or on a thread that does not have a cycle collector.  There are a number of these, so I thought I'd file a meta-bug.
Depends on: 1137343
Depends on: 1151646
Crash Signature: [@ NS_CycleCollectorSuspect3]
Summary: Crashes in NS_CycleCollectorSuspect3 → [meta] Crashes in NS_CycleCollectorSuspect3
Depends on: 1146416
Depends on: 1151650
Depends on: 1133892
Depends on: 1153563
Depends on: 1155328
Depends on: 1185132
Depends on: 1185144
Depends on: 1183651
Depends on: 970767
[Tracking Requested - why for this release]:

This particular crash signature is about to become #1 on Nightly. It's currently at #4 with 1493 crashes in the past 7 days.  However, the number 1, 2 and 3 signatures each have fixes landed.  This signature is already #2 on the past 3 day and past 1 day reports.

On Fx40beta, it is #21 with 851 crashes in the past 7 days.

Oddly, it doesn't appear to be affecting Fx41 in DevEd.
Keywords: crash
Just realized this is a meta bug, Andrew do you want to track the work on this particular signature here or would you rather have a dependency bug filed?
This signature is pretty generic. We'll likely get it again in the future. Maybe it makes sense to track each particular crash spike with a different bug. I'm not sure. I'll take a look at the beta crashes. I usually just look at the Nightly crashes for this. Bug 1185144 is a frequent instance of this I see on Nightly. I have a patch ready waiting for review. Bug 1183651 is another frequent instance I was seeing, though that had a patch that landed about 5 days ago.
Depends on: 1186160
OK, for  this one, let's track the individual bugs as they become important. Please nominate them for tracking if it looks like they need it.   Or, if you change your mind and gather up info in this particular bug, re-nom it. Thanks!
Depends on: 1220667
Crash volume for signature 'NS_CycleCollectorSuspect3':
 - nightly (version 51): 5 crashes from 2016-08-01.
 - aurora  (version 50): 7 crashes from 2016-08-01.
 - beta    (version 49): 122 crashes from 2016-08-02.
 - release (version 48): 189 crashes from 2016-07-25.
 - esr     (version 45): 51 crashes from 2016-05-02.

Crash volume on the last weeks (Week N is from 08-22 to 08-28):
            W. N-1  W. N-2  W. N-3
 - nightly       2       2       0
 - aurora        3       3       0
 - beta         47      52       8
 - release      58      59      38
 - esr           6       6       6

Affected platforms: Windows, Mac OS X, Linux

Crash rank on the last 7 days:
           Browser   Content     Plugin
 - nightly #284
 - aurora  #981      #494
 - beta    #539      #954
 - release #386
 - esr     #1102
Attached file cc_crashes.zip
Here is a test case that deterministically crashes in https://crash-stats.mozilla.com/report/index/8d3d9d11-6750-4be8-80cc-d8eea2160827 on OS X with current latest FF Nightly. Opening cc_crashes.html via file:// is enough to trigger it for me.
Err sorry, open the STR via file://, and then refresh the page - the page refresh is actually what brings the browser down and not the initial load.
that crash is a worker crash. Investigating how we have some DETH object alive after worker has been shutdown.
the crashing object is mozilla::dom::IDBTransaction. Something is keeping it alive way too late during worker shutdown.
bug 1298661 filed for the issue which can be reproduce using Jukka's testcase.
(and thanks for the testcase!)
Crash volume for signature 'NS_CycleCollectorSuspect3':
 - nightly (version 52): 21 crashes from 2016-09-19.
 - aurora  (version 51): 8 crashes from 2016-09-19.
 - beta    (version 50): 118 crashes from 2016-09-20.
 - release (version 49): 276 crashes from 2016-09-05.
 - esr     (version 45): 58 crashes from 2016-07-25.

Crash volume on the last weeks (Week N is from 10-17 to 10-23):
            W. N-1  W. N-2  W. N-3  W. N-4
 - nightly      14       1       1       0
 - aurora        5       1       0       2
 - beta         29      40      34      12
 - release      79      85      62      22
 - esr           3       2       7       8

Affected platforms: Windows, Mac OS X, Linux

Crash rank on the last 7 days:
           Browser   Content     Plugin
 - nightly #66
 - aurora  #407      #595
 - beta    #715      #612
 - release #703      #1113
 - esr     #1847
Too late for firefox 52, mass-wontfix.
Depends on: 1395104

Andrew should we put NS_CycleCollectorSuspect3 in Socorro's prefix list so that the underlying crashes can "clump" under different signatures? There's still quite a bit of volume here and I'm seeing new crashes coming from my WER interceptor so we might want to make these more actionable.

Flags: needinfo?(continuation)

Sure.

Flags: needinfo?(continuation)
Depends on: 1705027
See Also: → 1606347

Closing because no crashes reported for 12 weeks.

Status: NEW → RESOLVED
Closed: 2 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.