Closed Bug 588205 Opened 15 years ago Closed 13 years ago

Crash @ [@ mozalloc_abort ]

Categories

(Core :: General, defect)

Product:
Core
Component:
General
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

()

Status:
RESOLVED INCOMPLETE

People

(Reporter: cbook, Unassigned)

References

(
URL
)

Details

(Keywords: crash, regression, Whiteboard: [crashkill], [mobile-crash])

Crash Data

new crash for Firefox 4 from chofmanns crash-stats list. So far this are only Mac crashes. Comments show no pattern so far. http://crash-stats.mozilla.com/report/list?signature=mozalloc_abort
These look like out-of-memory crashes. But otherwise I can't discern any pattern from the few stacks I looked at.
this one is 8 minutes after installation and 4 seconds after start up. http://crash-stats.mozilla.com/report/index/9a518f80-926c-4a2a-9d9e-29eb82100811
but generally that doesn't seem to be the case. Correlation to startup or time of session 27 total crashes for mozalloc_abort on 20100816-crashdata.csv 2 startup crashes inside 30 sec. 3 startup crashes inside 3 min. might also be cross platform with these other signatures signature list 20 mozalloc_abort 2 mozalloc_abort(char const* const) 1 mozalloc_abort(char const*) checking --- mozalloc_abort 20100816-crashdata.csv found in: 4.0b3 4.0b2 4.0b4pre 4.0b1 release total-crashes mozalloc_abort crashes pct. all 290649 23 7.91333e-05 4.0b3 14631 15 0.00102522 4.0b2 3878 6 0.00154719 4.0b4pre 1186 5 0.00421585 4.0b1 1753 1 0.000570451 os breakdown mozalloc_abortTotal 23 Win5.1 0.00 Win6.0 0.00 Win6.1 0.09 Mac10.4 0.00 Mac10.5 0.13 Mac10.6 0.78 Lin2.4 0.00
According to trunk crash data, a test case in Bug 586561 can generate this stack. The most recent reports on the Mac nightly are here: http://tinyurl.com/2czl6q5
In addition to bug 586561, there are a few bugs whose signatures include mozalloc_abort; among those bugs, Marcia reported bug 586439 and bz reported bug 573369. Both of those seem related to startup/restart (though neither sound like they're Mac-specific).
We probably need to skiplist mozalloc_abort, since it's just a general "someone did something wrong" frame. marcia just linked this crash report: http://crash-stats.mozilla.com/report/index/bp-30f0aed3-b49a-45c2-bb9a-4a2492100817 where we should probably be skipping the first two frames and using [@ mozilla::plugins::PPluginInstanceChild::OnCallReceived ] as the signature.
Indeed, mozalloc_abort just means "we're intentionally aborting".
Filed bug 588438 on updating the skip list.
My crash report pushed to this bug (report: http://crash-stats.mozilla.com/report/index/6921923f-7fb4-4d1d-8e25-884a82100822) Happened as I was opening a new tab with cmd+t while watching a YouTube video. Took several seconds for the tab to open and I got the spinning beach ball at the same time.
That stack looks like bug 588445. (Once bug 588438 is fixed the signatures should change to match.)
Note I still see mozalloc_abort in Firefox 4 even though bug 588438 is "fixed". bp-37459dba-95bd-4279-98e5-71e282110208 running the test case http://murderdeathkitty.net/fuck/killall.html in bug 413842
Crash Signature: [@ mozalloc_abort ]
Whiteboard: [crashkill] → [crashkill], [mobile-crash]
I don't see base mozalloc_abort around at all in either Firefox or Fennec in the last 4 weeks, though mozalloc_abort(char const* const) is around with low volume. Does this bug still makes sense or can we resolve it?
(In reply to Robert Kaiser (:kairo@mozilla.com) from comment #12) > I don't see base mozalloc_abort around at all in either Firefox or Fennec in > the last 4 weeks, though mozalloc_abort(char const* const) is around with > low volume. Does this bug still makes sense or can we resolve it? mozalloc_abort is now in the Socorro skiplist. I close it as incomplete.
Status: NEW → RESOLVED
Closed: 13 years ago
OS: Mac OS X → All
Hardware: x86 → All
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.