Closed Bug 583591 Opened 14 years ago Closed 14 years ago

test_crash_purevirtual.js and test_crashreporter_crash.js perma-fail on linux64 Xo

Categories

(Toolkit :: Crash Reporting, defect)

x86
Linux
defect
Not set
normal

Tracking

()

RESOLVED FIXED
Tracking Status
blocking2.0 --- -

People

(Reporter: jdm, Assigned: ted)

References

Details

(Keywords: intermittent-failure)

Since bug 522770 landed, bug 572055 and bug 573290 have become permanently failing tests on linux64 Xo builds.
blocking2.0: --- → ?
FWIW, this appears to have started on the TM Merge.
Removing the dependency on bug 522770 because it is almost impossible for that to be the cause.
No longer depends on: 522770
If they're perma-orange, then disable them or back something out. Don't just leave them on and repeatedly star them. It's certainly possible that this is just bug 573290 happening again.
This is not permanently orange, though it does fail more often than not. I don't know how to disable this test on just Linux64 though :-(
I pushed a change that should fix this: http://hg.mozilla.org/mozilla-central/rev/43e26f5fad26 It fixed it on my local machine.
Assignee: nobody → ted.mielczarek
Status: NEW → RESOLVED
Closed: 14 years ago
Resolution: --- → FIXED
Status: RESOLVED → REOPENED
Resolution: FIXED → ---
I disabled this test to get rid of the perma-orange: http://hg.mozilla.org/mozilla-central/rev/899fb16a8fff
Also disabled the plugin submission tests: http://hg.mozilla.org/mozilla-central/rev/7da39af0206f
From a conversation in IRC it seems this is a test only problem and doesn't actually affect Firefox. Would like to get it fixed for the test coverage but not a blocker.
blocking2.0: ? → -
This sems to be happening on windows Xo, too: http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1281010957.1281012154.21613.gz WINNT 5.2 mozilla-central opt test xpcshell on 2010/08/05 05:22:37 s: win32-slave51 test_crash_purevirtual.js | test failed (with xpcshell return code: 0) http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1281017371.1281019685.20274.gz WINNT 5.2 mozilla-central opt test xpcshell on 2010/08/05 07:09:31 s: win32-slave19 test_crashreporter_crash.js | test failed (with xpcshell return code: 0) Perhaps the failure is timing-related, and the timing was more to our disadvantage on linux (making it perma-orange there) than on Windows (where it's randomorange)?
Blocks: 438871
OS: Linux → All
Hardware: x86 → All
Version: unspecified → Trunk
I don't believe that's the same issue. There's bug 572055 filed on another intermittent failure. The Linux failure mode was known (we disabled it on Linux x86 in bug 573290), but the Windows failure mode is not.
Filed Bug 584852 on the logs in comment 16. (it looks distinct from bug 572055). Apologies for the bugspam here.
No longer blocks: 438871
OS: All → Linux
Hardware: All → x86
(In reply to comment #18) > Filed Bug 584852 on the logs in comment 16. (d'oh, looks like that's actually a dupe of bug 584720. So that's what comment 16 is. Sorry again for the bugspam.)
Depends on: 584582
*sigh* this was all green on Try when I pushed the patch from bug 584582 there. I guess we'll have to disable it again.
Blocks: 438871
My patch in bug 620974 re-enables these and it went green on try. I filed bug 625342 to get a few more Try server xpcshell runs to get some confidence.
Depends on: 620974, 625342
These tests ran multiple times on the try server without any orange, so I'm going to mark this FIXED for now.
Status: REOPENED → RESOLVED
Closed: 14 years ago14 years ago
Resolution: --- → FIXED
Whiteboard: [orange]
You need to log in before you can comment on or make changes to this bug.