460843 - tracemonkey unittest machine burning (moz2-win32-slave14)

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P2)

Description

[Robert Sayre]

http://tinderbox.mozilla.org/showbuilds.cgi?tree=TraceMonkey

It is stuck removing a file, and called "WINNT 5.2 mozilla-central moz2-win32-slave14 dep unit test", but I think that is a misnomer.

Comment 1

[Nick Thomas [:nthomas] (UTC+12)]

The original failure was an xpcshell crash:

TEST-PASS | ../../_tests/xpcshell-simple/test_storage/unit/test_bug-365166.js | all tests passed
TEST-PASS | ../../_tests/xpcshell-simple/test_storage/unit/test_bug-393952.js | all tests passed
TEST-PASS | ../../_tests/xpcshell-simple/test_storage/unit/test_bug-429521.js | all tests passed
TEST-PASS | ../../_tests/xpcshell-simple/test_storage/unit/test_bug-444233.js | all tests passed

command timed out: 300 seconds without output

I think you need to look at a normal log to figure out what the next test is. I've killed xpcshell.exe and triggered another build.

We need to get the patch from bug 445578 on this master.cfg, and maybe deploy bug 420216.

Comment 2

[Nick Thomas [:nthomas] (UTC+12)]

The next build timed out in mochitest following:

*** 38262 INFO Running /tests/layout/generic/test/test_bug460532.html...
*** 38263 INFO TEST-PASS | /tests/layout/generic/test/test_bug460532.html | pass
*** 38265 INFO Running /tests/layout/generic/test/test_character_movement.html...

http://tinderbox.mozilla.org/showlog.cgi?tree=TraceMonkey&errorparser=unix&logfile=1224538033.1224546757.7711.gz&buildtime=1224538033&buildname=WINNT%205.2%20mozilla-central%20moz2-win32-slave14%20dep%20unit%20test&fulltext=1

bm-win2k3-unittest-02-hw is not exhibiting this problem (it has something wrong in make check, and mochichrome). Do you want a clobber or to let it cycle ?

Comment 3

[Nick Thomas [:nthomas] (UTC+12)]

The win32 boxes on mozilla-central are also having this problem.

Comment 4

[Nick Thomas [:nthomas] (UTC+12)]

Ok, the failure in comment #2 is me triggering a mozilla-central pull instead of a tracemonkey one. The tests went green once the human got out of the way and the machine started pulling from tracemonkey again.

I think we should still enact bug 420216 on this box, but I'll get that tomorrow.

Comment 5

[Nick Thomas [:nthomas] (UTC+12)]

(In reply to comment #4)
> I think we should still enact bug 420216 on this box, but I'll get that
> tomorrow.

Done.