725872 - [Thunderbird] Permanent "TEST-UNEXPECTED-FAIL | test-phish-simple ==" from test_partial.js | Update gcc 4.5 to include all patches from Firefox builders

Status: RESOLVED FIXED

(Mozilla Messaging Graveyard :: Release Engineering, defect)

Description

[Mark Banner (:standard8)]

Since bug 673470 landed on mozilla-central, we've been seeing this test failure on Thunderbird's linux 32 builders (VMs).

AFAICT this has been permanent on the 32 bit builders.

TEST-UNEXPECTED-FAIL | /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/components/url-classifier/tests/unit/test_partial.js | test failed (with xpcshell return code: 0), see following log:
>>>>>>>

TEST-INFO | (xpcshell/head.js) | test 1 pending

TEST-INFO | (xpcshell/head.js) | test 2 pending
running testPartialAdds

TEST-INFO | (xpcshell/head.js) | test 2 finished

TEST-INFO | (xpcshell/head.js) | running event loop

TEST-PASS | /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/components/url-classifier/tests/unit/head_urlclassifier.js | [null : 184] test-phish-simple;a:1 == test-phish-simple;a:1

TEST-UNEXPECTED-FAIL | /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/components/url-classifier/tests/unit/head_urlclassifier.js | test-phish-simple ==  - See following stack:
JS frame :: /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/head.js :: do_throw :: line 462
JS frame :: /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/head.js :: _do_check_eq :: line 556
JS frame :: /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/head.js :: do_check_eq :: line 577
JS frame :: /buildbot/comm-central-linux-opt-unittest-xpcshell/build/xpcshell/tests/toolkit/components/url-classifier/tests/unit/head_urlclassifier.js :: <TOP_LEVEL> :: line 198

TEST-INFO | (xpcshell/head.js) | exiting test

Comment 1

[Justin Wood (:Callek)]

Any chance you can run this on a debug build and post the output from the test there?

Its quite hard to trace where the problem is for me on a Windows machine, and I don't have a linux machine to test (its also not failing on SeaMonkey Debug afaict)

Comment 2

[Gian-Carlo Pascutto [:gcp]]

Might want to log this while you're at it:

NSPR_LOG_MODULES="nsChannelClassifier:5, UrlClassifierDbService:5, UrlClassifierHashCompleter:5, timestamp:5, nsUrlClassifierPrefixSet:5, UrlClassifierStreamUpdater:5"

Comment 3

[Serge Gautherie (:sgautherie)]

Fwiw:
http://mxr.mozilla.org/mozilla-central/source/toolkit/components/url-classifier/tests/unit/test_partial.js

Comment 4

[Gian-Carlo Pascutto [:gcp]]

This might be a compiler bug:

https://bugzilla.mozilla.org/show_bug.cgi?id=669410#c67

Comment 5

[Mark Banner (:standard8)]

Ok, I did a try build with:

MOZ_OPTIMIZE=

in toolkit/components/url-classifier/Makefile.in

and xpcshell-tests passed. Therefore this is a compiler bug that's been exposed.


gozer/jhopkins: we need to update our gcc on the Linux 32 builders to 4.5 (which is already on there) but include all 3 patches currently on the Firefox builders. RelEng have a rpm to do this apparently.

Comment 6

[Philippe M. Chiasson (:gozer)]

gcc-4.5 rpm being rebuilt with the patches right now.

Comment 7

[John Hopkins (:jhopkins)]

Updated gcc-4.5 RPM deployed to linux32 build machines.

Comment 8

[Serge Gautherie (:sgautherie)]

Ftr,

I checked from
"Tue Feb 14 00:04:10 2012 PST".

Perma-orange, until
"Wed Feb 15 13:34:41 2012 PST": 1st build fails, 2nd build succeeds.

Since then, a few random oranges:
"Wed Feb 15 18:05:06 2012 PST": 2 on momo-vm-fedora12-05
"Thu Feb 16 14:23:56 2012 PST": 1 on momo-vm-fedora12-04
"Fri Feb 17 09:25:02 2012 PST": 1 on momo-vm-fedora12-04
"Fri Feb 17 20:28:28 2012 PST": 1 on momo-vm-fedora12-05
"Mon Feb 20 14:22:53 2012 PST": 1 on momo-vm-fedora12-03

Comment 9

[Mark Banner (:standard8)]

(In reply to Serge Gautherie (:sgautherie) from comment #8)
> Since then, a few random oranges:
> "Wed Feb 15 18:05:06 2012 PST": 2 on momo-vm-fedora12-05
> "Thu Feb 16 14:23:56 2012 PST": 1 on momo-vm-fedora12-04
> "Fri Feb 17 09:25:02 2012 PST": 1 on momo-vm-fedora12-04
> "Fri Feb 17 20:28:28 2012 PST": 1 on momo-vm-fedora12-05
> "Mon Feb 20 14:22:53 2012 PST": 1 on momo-vm-fedora12-03

I suggest a new bug for those, as there seems to be a semi-random issue here.

> Depends on: 712540

Sorry, but that's got nothing to do with this bug, and doesn't stop this one, or any other orange bug being fixed.

Comment 10

[Serge Gautherie (:sgautherie)]

(In reply to Serge Gautherie (:sgautherie) from comment #8)
> Since then, a few random oranges:

"Tue Feb 21 07:35:06 2012 PST": 1 on s: momo-vm-fedora12-04

Comment 11

[Lukas Blakk [:lsblakk] use ?needinfo]

[Triage Comment]
untracking this for firefox13, please re-nominate if there's a case here.

Comment 12

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=ThunderbirdTrunk/1334100491.1334101211.31478.gz
Linux comm-central test xpcshell on 2012/04/10 16:28:11