1143971 - Failing Test endOfStream completes previous seek to truncated duration

Status: RESOLVED FIXED

(Core :: Audio/Video, defect, P2)

Description

[Jean-Yves Avenard [:jya]]

Will fail the webref media-source/mediasource-duration.html with:
TIMEOUT expected FAIL  Test endOfStream completes previous seek to truncated duration

This bug tracks the cause of it.

Comment 1

[Jean-Yves Avenard [:jya]]

Attachment: Part2. Update webref tests

update test. We don't pass on linux and webm as we fail an assert to calculate the duration of the video (6.042 vs 6.045)

Comment 2

[Jean-Yves Avenard [:jya]]

Comment on attachment 8579130 [details] [diff] [review]
Part2. Update webref tests

wrong bug #

Comment 3

[Jean-Yves Avenard [:jya]]

Attachment: Part1. Come out of waiting mode when mediasource is ended

Comes out of waiting mode when mediasource is ended. Behaviour was racey. If we decoded all content before EndOfStream() was called we would be stuck in waiting mode forever.

Comment 4

[Jean-Yves Avenard [:jya]]

Attachment: Part2. Allow seek to end of mediasource.duration

When seeking to the end of the mediasource duration, we would have stalled as the mPendingSeekTime would never be in the buffered range. So instead, we seek to the last of the audio sample and the last video sample. The MDSM seek logic will hit an end of stream and 
automatically complete the seek as required.

Comment 5

[Jean-Yves Avenard [:jya]]

Attachment: Part2. Re-enable mediasource-duration.html tests

with bug 1144617 applied, we now pass the test reliably.

Comment 6

[Jean-Yves Avenard [:jya]]

Comment on attachment 8580488 [details] [diff] [review]
Part2. Allow seek to end of mediasource.duration

belonged in bug 1144617

Comment 7

[Jean-Yves Avenard [:jya]]

remote:   https://hg.mozilla.org/integration/mozilla-inbound/rev/1fadcc2735db
remote:   https://hg.mozilla.org/integration/mozilla-inbound/rev/43525f7e2fd9

Comment 8

[Jean-Yves Avenard [:jya]]

bad rebase before push :(

remote:   https://hg.mozilla.org/integration/mozilla-inbound/rev/ab5aa8850e25

Comment 9

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/2ee1fb53c59c
https://hg.mozilla.org/mozilla-central/rev/1fadcc2735db
https://hg.mozilla.org/mozilla-central/rev/ab5aa8850e25

Comment 10

[Jean-Yves Avenard [:jya]]

This prevent stalls. And remove the intermittent failure that cause tests to be disabled

Comment 11

[Ralph Giles (:rillian)]

Comment on attachment 8580483 [details] [diff] [review]
Part1. Come out of waiting mode when mediasource is ended

Approval Request Comment
[Feature/regressing bug #]: MSE
[User impact if declined]: Spec compliance, less consistent testing.
[Describe test coverage new/current, TreeHerder]: Landed on m-c; web platform test.
[Risks and why]: Change is specific to the MSE code. Not trivial, but looks reasonable to me.
[String/UUID change made/needed]: None.

Comment 12

[Ralph Giles (:rillian)]

https://hg.mozilla.org/releases/mozilla-aurora/rev/c29c4393b55e
https://hg.mozilla.org/releases/mozilla-aurora/rev/52d27124efe7

Bustage fixed rolled into the first commit.

Comment 13

[Ryan VanderMeulen [:RyanVM]]

Re-disabled on Windows because it still permafails.
https://hg.mozilla.org/releases/mozilla-aurora/rev/0b7762fb7639

Comment 14

[Jean-Yves Avenard [:jya]]

the test re-enabled will only pass with bug 1143586 and bug 1142433