WebKit Bugzilla
New
Browse
Search+
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
NEW
237799
Unable to resume HLS stream in Safari after backgrounding the app
https://bugs.webkit.org/show_bug.cgi?id=237799
Summary
Unable to resume HLS stream in Safari after backgrounding the app
Davis Gossage
Reported
2022-03-11 15:51:46 PST
Created
attachment 454520
[details]
WebKit.GPU process logs I'm not able to resume an HLS stream in mobile Safari after backgrounding the app Issue is 100% reproducible on Simulator running iOS 13.4 RC. It also reproduces on my iPhone 12 running iOS 13.4 RC Issue occurred at 15:35:39 Steps to reproduce - Open
https://codepen.io/dgoss/full/eYyOXJX
in mobile Safari - Play 10 seconds of the HLS stream, pause - Background the app for 10 seconds - Open mobile Safari, tap Play WebKit logs that seem relevant App enters background, WebKit tells MediaPlayer to purge resources default 15:35:13.302978-0800 71785 com.apple.WebKit.GPU MediaPlayerPrivateAVFoundationObjC::setBufferingPolicy(4B3AC1364A25C7FC) PurgeResources default 15:35:13.303014-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfsi_performCurrentResourceConservationAction: [0x125d07280|P/MV] Setting resourceConservationLevelWhenPaused to 3 CoreMedia deletes some tracks default 15:35:13.306446-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_DeleteTrack: <0x126040a00|I/TKA.01> (0) track 3 default 15:35:13.306461-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_StopFeedingTrack: track 3 stopped feeding without setting EndTimer default 15:35:13.306510-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_DeleteTrack: <0x126040a00|I/TKA.01> (0) track 5 default 15:35:13.306575-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_StopFeedingTrack: track 5 stopped feeding without setting EndTimer More purging default 15:35:16.594082-0800 71785 com.apple.WebKit.GPU 0x104fe3430 - GPUProcess::prepareToSuspend(), isSuspensionImminent: 0 default 15:35:38.120907-0800 71785 com.apple.WebKit.GPU MediaPlayerPrivateAVFoundationObjC::setBufferingPolicy(4B3AC1364A25C7FC) MakeResourcesPurgeable App enters foreground and FigStreamPlayer is not ready to play, FigStreamPlayer never recovers default 15:35:39.340332-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_CheckIfLikelyToKeepUpAndNotify: tried to force playback but NOT ready for playback (0 buffered) default 15:35:39.340410-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_PostNotificationFromDispatch: posting Remote_RateWillChangeTo on 0x125d07280 (0x6000027d5340) default 15:35:39.340435-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_PostNotificationFromDispatch: posting RateDidChange on 0x125d07280 (0x6000027d5100) default 15:35:39.340451-0800 71785 com.apple.WebKit.GPU <<<< FigStreamPlayer >>>> fpfs_PostNotificationFromDispatch: posting PlaybackStateDidChange on 0x125d07280 (0x6000027d5020)
Attachments
WebKit.GPU process logs
(144.02 KB, text/plain)
2022-03-11 15:51 PST
,
Davis Gossage
no flags
Details
simulator sysdiagnose
(
deleted
)
2022-03-11 15:53 PST
,
Davis Gossage
no flags
Details
video of issue
(10.04 MB, video/mp4)
2022-03-11 15:54 PST
,
Davis Gossage
no flags
Details
Show Obsolete
(1)
View All
Add attachment
proposed patch, testcase, etc.
Davis Gossage
Comment 1
2022-03-11 15:53:18 PST
Created
attachment 454521
[details]
simulator sysdiagnose
Davis Gossage
Comment 2
2022-03-11 15:54:45 PST
Created
attachment 454522
[details]
video of issue
Davis Gossage
Comment 3
2022-03-11 15:57:04 PST
Also filed feedback
https://feedbackassistant.apple.com/feedback/9954511
Alexey Proskuryakov
Comment 4
2022-03-11 16:40:01 PST
rdar://90184838
Alexey Proskuryakov
Comment 5
2022-03-11 16:42:01 PST
The content of
attachment 454521
[details]
has been deleted for the following reason: Let's not keep the sysdiagnose here, it's sufficient to have internally at Apple.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug