Intermittent dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html | maxDifference: 0.9999571442604065, first bad index: 1792 with test-data offset 0 and expected-data offset 0; corresponding values 0 and -0.6874497532844543 --- differences
Categories
(Core :: Web Audio, defect, P5)
Tracking
()
People
(Reporter: intermittent-bug-filer, Unassigned)
References
(Regressed 1 open bug)
Details
(Keywords: intermittent-failure, regression, Whiteboard: [stockwell disabled])
Attachments
(1 file)
#[markdown(off)]
Filed by: ccoroiu [at] mozilla.com
https://treeherder.mozilla.org/logviewer.html#?job_id=239425792&repo=autoland
19:23:46 INFO - TEST-START | dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html
19:23:46 INFO - GECKO(3472) | NOW: enabled
19:23:46 INFO - TEST-INFO | started process screenshot
19:23:46 INFO - TEST-INFO | screenshot: exit 0
19:23:46 INFO - Buffered messages logged at 19:23:46
19:23:46 INFO - TEST-PASS | dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html | Correct number of channels for expected buffer 0
19:23:46 INFO - TEST-PASS | dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html | Correct number of expected frames
19:23:46 INFO - Buffered messages finished
19:23:46 INFO - TEST-UNEXPECTED-FAIL | dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html | maxDifference: 0.9999571442604065, first bad index: 1792 with test-data offset 0 and expected-data offset 0; corresponding values 0 and -0.6874497532844543 --- differences - got 255, expected +0
19:23:46 INFO - SimpleTest.is@SimpleTest/SimpleTest.js:320:16
19:23:46 INFO - compareChannels@dom/media/webaudio/test/webaudio.js:78:3
19:23:46 INFO - compareBuffers@dom/media/webaudio/test/webaudio.js:103:5
19:23:46 INFO - testOutput/sp.onaudioprocess@dom/media/webaudio/test/webaudio.js:214:11
19:23:46 INFO - EventHandlerNonNull*testOutput@dom/media/webaudio/test/webaudio.js:211:9
19:23:46 INFO - runTestOnContext@dom/media/webaudio/test/webaudio.js:202:9
19:23:46 INFO - testOnNormalContext@dom/media/webaudio/test/webaudio.js:222:7
19:23:46 INFO - runTestFunction@dom/media/webaudio/test/webaudio.js:253:5
19:23:46 INFO - rval@SimpleTest/SimpleTest.js:146:30
19:23:46 INFO - EventHandlerNonNull*this.addLoadEvent@SimpleTest/SimpleTest.js:171:13
19:23:46 INFO - @SimpleTest/SimpleTest.js:1453:5
19:23:46 INFO - Not taking screenshot here: see the one that was previously logged
19:23:46 INFO - TEST-UNEXPECTED-FAIL | dom/media/webaudio/test/test_mediaStreamTrackAudioSourceNode.html | maxDifference: 0.9999571442604065, first bad index: 1792 with test-data offset 0 and expected-data offset 0; corresponding values 0 and 0.6874497532844543 --- differences - got 255, expected +0
19:23:46 INFO - SimpleTest.is@SimpleTest/SimpleTest.js:320:16
19:23:46 INFO - compareChannels@dom/media/webaudio/test/webaudio.js:78:3
19:23:46 INFO - compareBuffers@dom/media/webaudio/test/webaudio.js:103:5
19:23:46 INFO - testOutput/sp.onaudioprocess@dom/media/webaudio/test/webaudio.js:214:11
19:23:46 INFO - EventHandlerNonNull*testOutput@dom/media/webaudio/test/webaudio.js:211:9
19:23:46 INFO - runTestOnContext@dom/media/webaudio/test/webaudio.js:202:9
19:23:46 INFO - testOnNormalContext@dom/media/webaudio/test/webaudio.js:222:7
19:23:46 INFO - runTestFunction@dom/media/webaudio/test/webaudio.js:253:5
19:23:46 INFO - rval@SimpleTest/SimpleTest.js:146:30
19:23:46 INFO - EventHandlerNonNull*this.addLoadEvent@SimpleTest/SimpleTest.js:171:13
19:23:46 INFO - @SimpleTest/SimpleTest.js:1453:5
Comment 1•7 years ago
|
||
This likely started when bug 1324548 landed here:
https://hg.mozilla.org/integration/autoland/rev/397a9577d91b77a4d2bfb46f09375345fb022df1
Paul: Could you please take a look at this new failure?
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
Comment 5•7 years ago
|
||
Retriggered:
Based on the above, Bug 1324548 seems to be the root for this.
https://treeherder.mozilla.org/#/jobs?repo=autoland&selectedJob=239455141&searchStr=android%2C8.0%2Cpixel2%2Cpgo%2Cmochitests%2Ctest-android-hw-p2-8-0-arm7-api-16%2Fpgo-mochitest-media-2%2Cm%28mda2%29&revision=2bb8fcac7edfed5ffad76f0b2df33682b830f48b
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
Comment 9•7 years ago
|
||
Is there still a way to buffer some data internally in a track or somewhere else? I think this is a phase shift caused by buffering something somewhere.
Comment 10•7 years ago
|
||
There are some buffers.
In the MSG we have the track buffers in StreamTracks, i.e., MediaStream::mTracks, [1]. In a SourceMediaStream this buffer could outgrow or underrun an iteration. A TrackUnionStream that copies from an input track will insert silence if the input track underran (i.e., is blocked). For a microphone source there are some conditions early on after starting where underrunning (i.e., pushing silence instead of real samples) is OK, [2]. If we're using a fake microphone (MediaEngineDefaultAudio) we should just be pulled instead, and no underruns can occur here. Any other streams than SourceMediaStream can't underrun either, unless they're not doing ProcessInput() properly.
This test doesn't seem to be using getUserMedia, so the StreamTracks buffers should just follow the iterations. No guarantees for this are in place really (we must still support SourceMediaStream's pushing), so possibly there's a bug somewhere.
Other buffers I can think of would be in AudioInputProcessing::PacketizeAndProcess so that would again not be used in this test.
I can't speak to any other buffers in WebAudio, where I suppose something could also be happening.
[1] https://searchfox.org/mozilla-central/rev/1b2636e8517aa48422ed516affe4d28cb7fa220a/dom/media/MediaStreamGraph.h#545
[2] https://searchfox.org/mozilla-central/rev/1b2636e8517aa48422ed516affe4d28cb7fa220a/dom/media/webrtc/MediaEngineWebRTCAudio.cpp#821-822
Comment 11•7 years ago
|
||
The frequency here is very high and this is occuring on the beta sims:
There are 398 total failures in the last 7 days on
pgo: android-hw-p2-8-0-arm7-api-16
opt: linux32-shippable, linux64-shippable, osx-10-10-shippable, windows10-64-shippable, windows7-32, windows7-32-shippable
Paul, can we disable this?
Updated•7 years ago
|
Comment 13•7 years ago
|
||
Joel is there a special condition to disable just shippables?
Comment 15•7 years ago
|
||
Comment 16•7 years ago
|
||
Updated•7 years ago
|
Comment 17•7 years ago
|
||
| bugherder | ||
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
| Comment hidden (Intermittent Failures Robot) |
Comment 25•4 years ago
|
||
Green on Try now and being re-enabled by https://phabricator.services.mozilla.com/D133090.
Updated•4 years ago
|
Description
•