558610 - mochitests-2: test_Prototype.html fails testElementReplace, testGetValue and testSetValue. And these failures are not _counted_

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Serge Gautherie (:sgautherie)]

This could be intermittent, but it (started to) fail on the 3 platforms at the same time...

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1270941321.1270945345.22362.gz
Linux comm-central-trunk debug test mochitests-2/5 on 2010/04/10 16:15:21
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1270947405.1270950187.31638.gz
OS X 10.5 comm-central-trunk debug test mochitests-2/5 on 2010/04/10 17:56:45
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1270944164.1270946304.24412.gz
WINNT 5.2 comm-central-trunk debug test mochitests-2/5 on 2010/04/10 17:02:44
{
3041 ERROR TEST-UNEXPECTED-FAIL | /tests/dom/tests/mochitest/ajax/prototype/test_Prototype.html | testElementReplace - 6 assertions, 1 failures, 0 errors
3164 ERROR TEST-UNEXPECTED-FAIL | /tests/dom/tests/mochitest/ajax/prototype/test_Prototype.html | testGetValue - 3 assertions, 1 failures, 0 errors
3165 ERROR TEST-UNEXPECTED-FAIL | /tests/dom/tests/mochitest/ajax/prototype/test_Prototype.html | testSetValue - 12 assertions, 1 failures, 0 errors
}

The other odd part is that these errors (only) are not counted as failures:
{
8280 INFO Passed: 3567
8281 INFO Failed: 0
8282 INFO Todo:   435
}

Regression timeframe:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=aa360768a319&tochange=1a6aac2558e4

Comment 1

[Serge Gautherie (:sgautherie)]

Optimized Firefox seems to have this bug too:

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1271031669.1271032176.17793.gz&fulltext=1
WINNT 5.2 mozilla-central opt test mochitests-2/5 on 2010/04/11 17:21:09
s: win32-slave25

Comment 2

[:Ms2ger (he/him; ⌚ UTC+1/+2)]

Attachment: Count them [See comment 3]

It looks like these failures aren't counted because the test uses its own copy of the test files.

Comment 3

[Serge Gautherie (:sgautherie)]

Comment on attachment 438495 [details] [diff] [review]
Count them
[See comment 3]


That's indeed the counting problem, which depends on bug 427500 and blocks bug 451287.
I may have not filed it yet as I'm desperately waiting for review of the former.

This is also not the only affected file: see at least
http://mxr.mozilla.org/mozilla-central/search?string=lib%2FSimpleTest.js&case=1&find=%2Fdom%2Ftests%2Fmochitest%2Fajax%2F

Could you file a separate bug on that issue?
Thanks.

Comment 4

[Timothy Nikkel (:tnikkel)]

Bisected the range from comment 0:

The first bad revision is:
changeset:   40657:2b560e956063
user:        Geoff Lankow <[e-mail]>
date:        Sat Apr 10 16:05:58 2010 -0400
summary:     Bug 166235: moz-user-select = 'none' should prevent copying to the clipboard. r=mats

Comment 5

[Mats Palmgren (inactive)]

The regression from bug 166235 has been fixed now:
http://hg.mozilla.org/mozilla-central/rev/81c83a98798d