Closed Bug 587767 Opened 10 years ago Closed 7 years ago

Random hang (command timed out) in test_cookies_read.js

Categories

(Core :: Networking: Cookies, defect)

All
Linux
defect
Not set
normal

Tracking

()

RESOLVED DUPLICATE of bug 595925

People

(Reporter: standard8, Unassigned)

References

Details

(Keywords: intermittent-failure)

On the Thunderbird Linux tinderboxes we're seeing a fairly frequent xpcshell-test hang in test_cookies_read.js:

TEST-INFO | /buildbot/linux-comm-central-check/build/objdir/mozilla/_tests/xpcshell/test_cookies/unit/test_cookies_read.js | running test ...

command timed out: 300 seconds without output, killing pid 11417
process killed by signal 9
program finished with exit code -1
TinderboxPrint: xpcshell-tests<br/><em class="testfail">T-FAIL</em>
buildbot.slave.commands.TimeoutError: command timed out: 300 seconds without output, killing pid 11417
TinderboxPrint: xpcshell-tests<br/><em class="testfail">timeout</em>

http://tinderbox.mozilla.org/showlog.cgi?log=Thunderbird/1281983292.1281987477.22382.gz

Unfortunately that's all the information I have at this stage.
We haven't seen anything like this on Firefox. Not sure what could be different here. What network.cookie prefs are your tests using?
We use the default prefs. We do have some mailnews specific code, but not sure why that would cause an issue.

I've tried locally on my mac and not been able to reproduce, will try on a Linux VM soon.
I've tried reproducing this today, all I've been able to see so far is that the close after removing some cookies can take up to 40 seconds or so - which seems like a potential deadlock area.

This is the line where it was temporarily hanging:

http://hg.mozilla.org/mozilla-central/annotate/c1bb86ae655a/extensions/cookie/test/unit/test_cookies_read.js#l73
Pretty sure this will be fixed by bug 595925. Reopen if you're still seeing this after that lands?
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → DUPLICATE
Duplicate of bug: 595925
Blocks: 438871
Duplicate of this bug: 601508
Whiteboard: [orange]
seems the problem is back https://tbpl.mozilla.org/php/getParsedLog.php?id=25758311&tree=Mozilla-Inbound
Status: RESOLVED → REOPENED
Resolution: DUPLICATE → ---
Oh, no wonder the summary of this seems a little off for what we're starring it as, it's an entirely different thing from years ago that shouldn't have been reopened. Filed bug 966713.
Status: REOPENED → RESOLVED
Closed: 10 years ago7 years ago
Resolution: --- → DUPLICATE
Duplicate of bug: 595925
You need to log in before you can comment on or make changes to this bug.