Bug 167749 – LayoutTest http/tests/media/track-in-band-hls-metadata-crash.html is a flaky timeout

• Component: Media
• Status: RESOLVED FIXED

[Ryan Haddad]

LayoutTest http/tests/media/track-in-band-hls-metadata-crash.html is a flaky timeout

https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=http%2Ftests%2Fmedia%2Ftrack-in-band-hls-metadata-crash.html

[Ryan Haddad]

This test was added with https://trac.webkit.org/changeset/211495

[Ryan Haddad]

This appears to be limited to El Capitan.

[Ryan Haddad]

Jer, this test continues to be very flaky. Can you take a look?

[Radar WebKit Bug Importer]

<rdar://problem/30434020>

[Ryan Haddad]

Marked test as flaky in https://trac.webkit.org/changeset/211985

[Per Arne Vollan]

Created attachment 309593 [details]
Patch

[Per Arne Vollan]

Comment on attachment 309593 [details]
Patch

Thanks for reviewing!

[WebKit Commit Bot]

Comment on attachment 309593 [details]
Patch

Clearing flags on attachment: 309593

Committed r216588: <http://trac.webkit.org/changeset/216588>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Per Arne Vollan]

Reopening to attach new patch.

[Per Arne Vollan]

Created attachment 309700 [details]
Patch

[Eric Carlson]

Comment on attachment 309700 [details]
Patch

This is fine if this works, but another way to remove the flakiness of the current version is to not log the 'canplaythrough' event: document.addEventListener('canplaythrough', () => { video.play() });

[Per Arne Vollan]

(In reply to Eric Carlson from comment #12)
> Comment on attachment 309700 [details]
> Patch
> 
> This is fine if this works, but another way to remove the flakiness of the
> current version is to not log the 'canplaythrough' event:
> document.addEventListener('canplaythrough', () => { video.play() });

Ah, I see, thanks! I also noticed that the 'canplaythrough' event was dispatched twice on one of the bots, causing 'video.play()' to be called two times.

[Eric Carlson]

(In reply to Per Arne Vollan from comment #13)
> (In reply to Eric Carlson from comment #12)
> > Comment on attachment 309700 [details]
> > Patch
> > 
> > This is fine if this works, but another way to remove the flakiness of the
> > current version is to not log the 'canplaythrough' event:
> > document.addEventListener('canplaythrough', () => { video.play() });
> 
> Ah, I see, thanks! I also noticed that the 'canplaythrough' event was
> dispatched twice on one of the bots, causing 'video.play()' to be called two
> times.

document.oncanplaythrough = () => { document.removeAttribute('oncanplaythrough'); video.play() });

[WebKit Commit Bot]

Comment on attachment 309700 [details]
Patch

Clearing flags on attachment: 309700

Committed r216680: <http://trac.webkit.org/changeset/216680>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Per Arne Vollan]

Reopening to attach new patch.

[Per Arne Vollan]

Created attachment 309890 [details]
Patch

[Per Arne Vollan]

(In reply to Eric Carlson from comment #14)
> (In reply to Per Arne Vollan from comment #13)
> > (In reply to Eric Carlson from comment #12)
> > > Comment on attachment 309700 [details]
> > > Patch
> > > 
> > > This is fine if this works, but another way to remove the flakiness of the
> > > current version is to not log the 'canplaythrough' event:
> > > document.addEventListener('canplaythrough', () => { video.play() });
> > 
> > Ah, I see, thanks! I also noticed that the 'canplaythrough' event was
> > dispatched twice on one of the bots, causing 'video.play()' to be called two
> > times.
> 
> document.oncanplaythrough = () => {
> document.removeAttribute('oncanplaythrough'); video.play() });

Thanks! I have updated the patch :)

[WebKit Commit Bot]

Comment on attachment 309890 [details]
Patch

Clearing flags on attachment: 309890

Committed r216885: <http://trac.webkit.org/changeset/216885>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.