461281 - CPU usage goes to 100% after playing some Ogg files

Status: RESOLVED FIXED

(Core :: Audio/Video, defect)

Description

[cajbir (:cajbir)]

Attachment: testcase

Steps to reproduce:

1) Create an html file with the following:
   <video src='http://www.double.co.nz/video_test/ufohit.ogg' controls></video>
2) Load html file and use the built in controls to play.

What happens:

Sound plays, then the cpu usage goes to 100% and stays there.

What should happen:

Sounds plays, cpu usage drops to zero.

See testcase for html file.

Comment 1

[cajbir (:cajbir)]

According to oggz-validate the .ogg file has no end of stream marker. I tested a number of ogg files without an eos and they all have this issue.

Comment 2

[cajbir (:cajbir)]

Attachment: Fixes bug

nsChannelReader was not correctly implementing the liboggplay reader interface in cases of the read returning zero bytes.

Comment 3

[cajbir (:cajbir)]

Attachment: Binary file required for test

Comment 4

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

Comment on attachment 344424 [details] [diff] [review]
Binary file required for test

reviewing binary data blobs is silly :-)

Comment 5

[cajbir (:cajbir)]

http://hg.mozilla.org/mozilla-central/rev/8493c800cc20
http://hg.mozilla.org/mozilla-central/rev/f291cb19c48b

Comment 6

[Olli Pettay [:smaug][[e-mail]]]

Linux mozilla-central moz2-linux-slave08 dep unit test on 2008/10/23 02:34:27
*** 27381 INFO Running /tests/content/media/video/test/test_bug461281.html...
ALSA lib confmisc.c:670:(snd_func_card_driver) cannot find card '0'
NEXT ERROR ALSA lib conf.c:3479:(_snd_config_evaluate) function snd_func_card_driver returned error: No such device
ALSA lib confmisc.c:391:(snd_func_concat) error evaluating strings
ALSA lib conf.c:3479:(_snd_config_evaluate) function snd_func_concat returned error: No such device
ALSA lib confmisc.c:1070:(snd_func_refer) error evaluating name
ALSA lib conf.c:3479:(_snd_config_evaluate) function snd_func_refer returned error: No such device
ALSA lib conf.c:3947:(snd_config_expand) Evaluate error: No such device
ALSA lib pcm.c:2143:(snd_pcm_open_noupdate) Unknown PCM default

Comment 7

[cajbir (:cajbir)]

Backed out due to following error on unit test box:
http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1224754467.1224758383.17269.gz

Backouts:
http://hg.mozilla.org/mozilla-central/rev/26b35872bdf8
http://hg.mozilla.org/mozilla-central/rev/143306ec72ec
http://hg.mozilla.org/mozilla-central/rev/96dc7b9715a7
http://hg.mozilla.org/mozilla-central/rev/443a4674d69f

This bug added a test that played an audio file. Need to look into how lack of an audio device is handled.

Comment 8

[cajbir (:cajbir)]

It looks like the code handled the lack of an audio device ok, and in fact the test passed. Unfortunately alsa is noisy and printed to the stderr/stdout that it failed to find the device and the unit test box sees the word 'error' there - or sees it as unexpected output.

Comment 9

[cajbir (:cajbir)]

I found out how to disable the noisy alsa output. I'll disuss with the libsydneyaudio maintainers to see what's the best way to include this, and when I get a fix for it, re-commit this patch.

Comment 10

[cajbir (:cajbir)]

Attachment: libsydneyaudio update to stop stderr output

Comment 11

[cajbir (:cajbir)]

http://hg.mozilla.org/mozilla-central/rev/135a2999f0dc
http://hg.mozilla.org/mozilla-central/rev/909342c1efe0

Comment 12

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 344423 [details] [diff] [review]
Fixes bug

+  if (!NS_SUCCEEDED(rv)) {

fwiw, that should be if (NS_FAILED(rv))