527785 - 3.5.5 crashes [@ CompositeDataSourceImpl::cycleCollection::UnmarkPurple(nsISupports*)]

Status: RESOLVED WORKSFORME

(Core :: XPCOM, defect)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

A new topcrash appeared in Firefox 3.5.5 (probably related to the work in bug 521750).

http://crash-stats.mozilla.com/report/list?product=Firefox&query_search=signature&query_type=exact&query=CompositeDataSourceImpl%3A%3AcycleCollection%3A%3AUnmarkPurple%28nsISupports*%29&date=&range_value=1&range_unit=weeks&do_query=1&signature=CompositeDataSourceImpl%3A%3AcycleCollection%3A%3AUnmarkPurple%28nsISupports*%29

I took a look at three minidumps:
bp-0d28958e-bb6e-48b2-9273-581842091106
bp-1bb6ed70-3dc5-4bd5-a006-fe4102091106
bp-5dfc452b-3481-43a8-bb84-11af82091106

When I load them in the debugger:
 * the crashing frame shows up as nsParser::cycleCollection::UnmarkPurple instead of CompositeDataSourceImpl::... .
 * The crash is effectively on the same instruction as the crash in bug 527339
 * However, unlike bug 527339, the value of mRefCnt.mTagged is 0 (rather than a small odd integer)

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

jst points out that these could be the 3.5.5 version of these 3.5.4 crashes:
http://crash-stats.mozilla.com/report/list?product=Firefox&version=Firefox%3A3.5.4&query_search=signature&query_type=exact&query=nsXULPrototypeDocument%3A%3AcycleCollection%3A%3AUnmarkPurple%28nsISupports*%29&date=&range_value=1&range_unit=weeks&do_query=1&signature=nsXULPrototypeDocument%3A%3AcycleCollection%3A%3AUnmarkPurple%28nsISupports*%29
which look pretty much the same except for a different class

Comment 2

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Note that these appear *not* to be correlated with threadsafety problems.  (Nor do the nsXULPrototypeDocument ones.)

Comment 3

[chris hofmann]

this is oddly showing up in 3.6.12 a few days after its release at #20 ranking.  no reports of this on 3.6.11.

might be signature shifting.

Comment 4

[bugs-mozilla]

I encountered this crash with Firefox 3.6.12 on Windows 7 64bit twice today? Can I do anything to help it get fixed? Both times Firefox crashed while I wasn't using it at all (another program hat focus).

Comment 5

[Jack Yan]

Most of my crashes that cannot be traced to Java and Flash are of this variety. Firefox crashes a dozen times a day now, many of which are these.

Comment 6

[Jack Yan]

PS.: This is also showing up in 3.6.12—sorry I didn’t make the version number clear.

Comment 7

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Between 30% and 55% of these crashes look like they're triggered by some AVG product.

Comment 8

[Jack Yan]

I’ll be in the 45 to 70 per cent bracket: no AVG installed here.

A few more IDs:

bp-6271d23b-ae8b-41de-a060-e815b2101220
918e4fa1-806e-4724-a82b-9c5502101219
bp-9a2efd3b-9355-4f1f-82dc-2551b2101128

Comment 9

[Sheila Mooney]

Still a valid crash but low volume. Only 22 crashes on 8.0 in the past 4 weeks and 1 crash on Thunderbird 8.0 in the past 4 weeks. Removing top crash keyword.

Comment 10

[Andrew McCreight [:mccr8]]

The only crashes I see for this on crash-stats are for 3.6.