492821 - [test disabled] TEST-UNEXPECTED-FAIL | test_videoDocumentTitle.html | Doc title incorrect - got "320x240.ogv", expected "bug461281.ogg"

Status: RESOLVED WORKSFORME

(Core :: Audio/Video, defect)

Description

[Jeff Muizelaar [:jrmuizel]]

This test seems to be failing consistently on linux:
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242232957.1242236034.2073.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242233871.1242239737.9176.gz
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242227885.1242235196.710.gz

It started happening after 5a70e386a50c which doesn't make very much sense.

Comment 1

[Jeff Muizelaar [:jrmuizel]]

Looks like it happened before too (da613c9fae8c):
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242208481.1242215875.29624.gz

Comment 2

[Robert O'Callahan (:roc) (email my personal email if necessary)]

It's happening almost all the time.

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242256987.1242264121.16712.gz
Linux mozilla-central unit test on 2009/05/13 16:23:07

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242256475.1242262017.14201.gz
Linux mozilla-central unit test on 2009/05/13 16:14:35

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242255008.1242259088.10535.gz#err3
Linux mozilla-central unit test on 2009/05/13 15:50:08

Comment 3

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Naturally I can't reproduce it in my Linux VM, even so :-(

Comment 4

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242303436.1242308748.16754.gz
Linux mozilla-central unit test on 2009/05/14 05:17:16

Comment 5

[Brad Lassey [:blassey] (use needinfo?)]

I believe this showed up with changeset 397dee5a493a

Comment 6

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242310957.1242318558.4118.gz
Linux mozilla-central unit test on 2009/05/14 07:22:37

Comment 7

[Joe Drew (not getting mail)]

http://tinderbox.mozilla.org/showlog.cgi?log=MozillaTry/1242341045.1242346579.19939.gz

Comment 8

[Joe Drew (not getting mail)]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242343960.1242352377.27517.gz

Comment 9

[(not currently active) Ted Mielczarek]

Linux mozilla-central unit test on 2009/05/15 03:44:09  
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242384249.1242391345.31240.gz

Comment 10

[Johnathan Nightingale [:johnath]]

Roc - Can we disable this test while we debug?  It's awfully noisy.  Or do we need access to a machine to understand what's failing?

Comment 11

[Jeff Muizelaar [:jrmuizel]]

Linux
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242404399.1242407596.3462.gz

Comment 12

[Robert O'Callahan (:roc) (email my personal email if necessary)]

I suppose we could disable it.

Comment 13

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242429504.1242437873.28857.gz
Linux mozilla-central unit test on 2009/05/15 16:18:24

Comment 14

[Josh Aas]

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1242432224.1242436009.25941.gz
Linux mozilla-central unit test on 2009/05/15 17:03:44

Comment 15

[Josh Aas]

I disabled this test, fails too often.

http://hg.mozilla.org/mozilla-central/rev/b7bbafb114e2

Comment 16

[Jim Mathies [:jimm]]

5/20:

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox3.5/1242829388.1242837598.28324.gz&fulltext=1

Comment 17

[Jim Mathies [:jimm]]

This was disabled today due to numerous failures on 1.9.1.

http://hg.mozilla.org/releases/mozilla-1.9.1/rev/be346d384ff1

Comment 18

[Boris Zbarsky [:bzbarsky]]

So I think we should reenable this, but add a URI test in addition to the title.  If/when it fails, it'll at least tell us whether we're even dealing with the right document, right?

The most likely thing going on here, imo, is that we're getting some random load event from _somewhere_.  Could that be happening?  If so, can we adjust the test to detect multiple loads for the same contentDocument and also error on that so we'll see whether that happens?

Comment 19

[Robert O'Callahan (:roc) (email my personal email if necessary)]

I can't see how that would happen, but that sounds like a reasonable thing to do.

Comment 20

[Boris Zbarsky [:bzbarsky]]

See, I spent a bit of time reading this code, and I don't see how this bug could happen at all...  Hence the question.  We've eliminated all the possible things.  ;)

Comment 21

[David Humphrey (:humph)]

Attachment: Add more checks to test

Boris, reading what you wrote in comment 18, would these changes do what you want?  If I've misunderstood, let me know and I'll fix.  I also can't reproduce this.

Comment 22

[Boris Zbarsky [:bzbarsky]]

The gDocuments thing is probably a good addition, but I'm not quite sure why comparing window.locations to cwl makes sense, and it's not really what I was after in comment 18.  I was thinking something along the lines of assigning the expected URI whenever we assign |title| and then testing the actual URI against it.

Also, we'll need to actually reenable the test.

Comment 23

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 525686 [details] [diff] [review]
Add more checks to test

Per comment 22.

Comment 24

[Ryan VanderMeulen [:RyanVM]]

Runs without issue now. Re-enabled.