Bug 73692

Summary: Layout tests media/event-attributes.html and media/video-playbackrate.html failing
Product: WebKit Reporter: Vincent Scheib <scheib>
Component: Tools / TestsAssignee: Andrew Scherkus <scherkus>
Status: RESOLVED FIXED    
Severity: Normal CC: antonm, enne, eric.carlson, fischman, jknotten, kbr, scherkus, vrk, vsevik, webkit.review.bot
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Patch none

Description Vincent Scheib 2011-12-02 13:57:01 PST
The following layout test is failing on Chromium

media/event-attributes.html

http://test-results.appspot.com/dashboards/flakiness_dashboard.html#showExpectations=true&tests=media%2Fevent-attributes.html

It is turning very flaky, but not consistently failing except on Webkit Win (dbg)(2).

Probable cause:
http://trac.webkit.org/log/?verbose=on&rev=101842&stop_rev=101822
Comment 1 Vincent Scheib 2011-12-02 14:03:35 PST
http://trac.webkit.org/changeset/101850
Comment 2 Vsevolod Vlasov 2011-12-05 03:20:51 PST
Marked flaky on leopard also.
Committed r101984: <http://trac.webkit.org/changeset/101984>
Comment 3 Adrienne Walker 2011-12-16 12:24:16 PST
Committed r103092: <http://trac.webkit.org/changeset/103092>
Comment 4 Ami Fischman 2011-12-16 15:24:01 PST
scherkus: canplaythrough fires, but seeked doesn't.  (about 30% flaky when running 10 parallel DRTs running this one test) 
Any chance this is related to bug 73681?  The two tests started failing at the same time (judging from bug-filing times).
Comment 5 John Knottenbelt 2012-01-10 06:04:23 PST
This test is also crashing on Win dbg (2) after http://trac.webkit.org/changeset/104550
Comment 6 Andrew Scherkus 2012-01-17 09:57:05 PST
vrk it looks like we're getting flaky progress event counts -- do you think any of your canplaythrough changes in chromium might have caused this?
Comment 7 Victoria Kirst 2012-01-17 17:18:09 PST
(In reply to comment #6)
> vrk it looks like we're getting flaky progress event counts -- do you think any of your canplaythrough changes in chromium might have caused this?

Shouldn't be me -- I haven't landed my changes yet! What's currently checked in is the same canplaythrough behavior as always.

Actually I haven't landed my changes yet because I was running into progress event problems as well.... curious indeed!
Comment 8 Kenneth Russell 2012-06-22 13:27:13 PDT
I've seen this test assert a couple of times during my WebKit gardening shift:

10:31:03.711 1286 worker/3 media/event-attributes.html crashed, stack trace:
10:31:03.711 1286   [1970:-989777920:1098324960206:FATAL:null_audio_sink.cc(97)] Check failed: playback_rate_ > 0.0f (0 vs. 0)
10:31:03.711 1286   Backtrace:
10:31:03.711 1286   	0   DumpRenderTree                      0x11d5d99f start + 8662631
10:31:03.711 1286   	1   DumpRenderTree                      0x11d5d93b start + 8662531
10:31:03.711 1286   	2   DumpRenderTree                      0x11daceac start + 8987508
10:31:03.711 1286   	3   DumpRenderTree                      0x11dab99b start + 8982115
10:31:03.711 1286   	4   DumpRenderTree                      0x152a11f4 start + 64513724
10:31:03.711 1286   	5   DumpRenderTree                      0x152a33f4 start + 64522428
10:31:03.711 1286   	6   DumpRenderTree                      0x152a32f3 start + 64522171
10:31:03.712 1286   	7   DumpRenderTree                      0x152a323c start + 64521988
10:31:03.712 1286   	8   DumpRenderTree                      0x11dc686b start + 9092403
10:31:03.712 1286   	9   DumpRenderTree                      0x11dc3da8 start + 9081456
10:31:03.712 1286   	10  DumpRenderTree                      0x11dc4151 start + 9082393
10:31:03.712 1286   	11  DumpRenderTree                      0x11dc4708 start + 9083856
10:31:03.712 1286   	12  DumpRenderTree                      0x11dd5e75 start + 9155389
10:31:03.712 1286   	13  DumpRenderTree                      0x11dc3702 start + 9079754
10:31:03.712 1286   	14  DumpRenderTree                      0x11dc2a1b start + 9076451
10:31:03.712 1286   	15  DumpRenderTree                      0x11dc2960 start + 9076264
10:31:03.712 1286   	16  DumpRenderTree                      0x11e74d92 start + 9806426
10:31:03.712 1286   	17  DumpRenderTree                      0x11e74f43 start + 9806859
10:31:03.712 1286   	18  DumpRenderTree                      0x11e73066 start + 9798958
10:31:03.712 1286   	19  libSystem.B.dylib                   0x94e1a055 _pthread_start + 321
10:31:03.712 1286   	20  libSystem.B.dylib                   0x94e19f12 thread_start + 34
10:31:03.712 1286   
10:31:03.717 1268   media/event-attributes.html -> unexpected crash


http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Mac10.5%20%28dbg%29%282%29/builds/8202

I'm going to mark the test as a flaky crasher as well.

This bug's been open for a while...would be good to get it reproduced / investigated.
Comment 9 Kenneth Russell 2012-06-22 13:38:08 PDT
Committed r121054: <http://trac.webkit.org/changeset/121054>
Comment 10 Kenneth Russell 2012-06-22 13:38:22 PDT
Oops, accidentally closed bug. Reopening.
Comment 11 Kenneth Russell 2012-06-22 16:51:18 PDT
media/video-playbackrate.html is asserting similarly:

http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Mac10.5%20%28dbg%29%282%29/builds/8209


15:59:05.797 1285 worker/3 media/video-playbackrate.html crashed, stack trace:
15:59:05.797 1285   [1972:-1014968320:1198821482313:FATAL:null_audio_sink.cc(97)] Check failed: playback_rate_ > 0.0f (0 vs. 0)
15:59:05.797 1285   Backtrace:
15:59:05.797 1285   	0   DumpRenderTree                      0x2ac3c9df start + 8671319
15:59:05.798 1285   	1   DumpRenderTree                      0x2ac3c97b start + 8671219
15:59:05.798 1285   	2   DumpRenderTree                      0x2ac8beec start + 8996196
15:59:05.798 1285   	3   DumpRenderTree                      0x2ac8a9db start + 8990803
15:59:05.798 1285   	4   DumpRenderTree                      0x2e18a9e4 start + 64565340
15:59:05.798 1285   	5   DumpRenderTree                      0x2e18cbe4 start + 64574044
15:59:05.798 1285   	6   DumpRenderTree                      0x2e18cae3 start + 64573787
15:59:05.798 1285   	7   DumpRenderTree                      0x2e18ca2c start + 64573604
15:59:05.798 1285   	8   DumpRenderTree                      0x2aca58ab start + 9101091
15:59:05.798 1285   	9   DumpRenderTree                      0x2aca2de8 start + 9090144
15:59:05.798 1285   	10  DumpRenderTree                      0x2aca3191 start + 9091081
15:59:05.798 1285   	11  DumpRenderTree                      0x2aca3748 start + 9092544
15:59:05.798 1285   	12  DumpRenderTree                      0x2acb4eb5 start + 9164077
15:59:05.798 1285   	13  DumpRenderTree                      0x2aca2742 start + 9088442
15:59:05.798 1285   	14  DumpRenderTree                      0x2aca1a5b start + 9085139
15:59:05.798 1285   	15  DumpRenderTree                      0x2aca19a0 start + 9084952
15:59:05.798 1285   	16  DumpRenderTree                      0x2ad53dd2 start + 9815114
15:59:05.798 1285   	17  DumpRenderTree                      0x2ad53f83 start + 9815547
15:59:05.798 1285   	18  DumpRenderTree                      0x2ad520a6 start + 9807646
15:59:05.798 1285   	19  libSystem.B.dylib                   0x94e1a055 _pthread_start + 321
15:59:05.798 1285   	20  libSystem.B.dylib                   0x94e19f12 thread_start + 34
15:59:05.798 1285   
15:59:05.805 1269   media/video-playbackrate.html -> unexpected crash


I'm marking that one similarly. We need an owner for this; Andrew, assigning to you.
Comment 12 Ami Fischman 2012-07-31 17:38:23 PDT
Created attachment 155690 [details]
Patch
Comment 13 WebKit Review Bot 2012-07-31 20:50:29 PDT
Comment on attachment 155690 [details]
Patch

Clearing flags on attachment: 155690

Committed r124293: <http://trac.webkit.org/changeset/124293>
Comment 14 WebKit Review Bot 2012-07-31 20:50:33 PDT
All reviewed patches have been landed.  Closing bug.