User Profile

In the latest versions of iOS (at least since 13.1.3), background video rendering has been broken for our app. Sometimes the video is rendered with black frames, other times it crashes.It seems to be related to the following error we’re now seeing in the console:Execution of the command buffer was aborted due to an error during execution. Insufficient Permission (to submit GPU work from background) (IOAF code 6) GLDRendererMetal command buffer completion error: Error Domain=MTLCommandBufferErrorDomain Code=7 "Insufficient PermiOur current working theory is that some underlying framework has changed to use Metal under the hood, which does not have permission to run in the background (because it uses the GPU).Any ideas what we can do as a workaround, and in general what causes this issue?

We're getting the following crash when playing audio through AVAudioPlayerNode (via AudioKit AKPlayer):Fatal Exception: com.apple.coreaudio.avfaudio player did not see an IO cycle. Fatal Exception: com.apple.coreaudio.avfaudio 0 CoreFoundation 0x1a585c98c __exceptionPreprocess 1 libobjc.A.dylib 0x1a55850a4 objc_exception_throw 2 CoreFoundation 0x1a575ed1c +[_CFXNotificationTokenRegistration keyCallbacks] 3 AVFAudio 0x1b21d1e24 AVAE_RaiseException(NSString*, ...) 4 AVFAudio 0x1b227356c AVAudioPlayerNodeImpl::StartImpl(AVAudioTime*) 5 AVFAudio 0x1b2270910 -[AVAudioPlayerNode play] 6 OurApp 0x1049920f0 AKPlayer.play(from:to:at:hostTime:) 7 OurApp 0x1049af9fc AKPlayer.play(from:to:when:hostTime:) 8 OurApp 0x1049af840 AKPlayer.play(when:hostTime:)We're getting these crash logs through Crashlytics.We play with a 0.2s delay.I haven't been able to reliably reproduce the crash, nor find any documentation about what the error message means. We've seen another crash in the same area if the audio engine is not running, but according to our logs, the audio engine is running when this particular crash accurs.The code is not doing anything fancy, just downloading and playing a small (<1 minute) mp3 audio file. The only complicating factor is that the audio session is also configured for recording (that's the next step after picking and previewing the audio file).Any suggestions are welcome.