[Bug 743758] osxaudiosrc supports only 44100 sample rate on iOS

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=743758

--- Comment #35 from Ilya Konstantinov <ilya.konstantinov at gmail.com> ---
(In reply to Arun Raghavan from comment #31)
> Review of attachment 298928 [details] [review]:
> 
> ::: sys/osxaudio/gstosxcoreaudio.c
> @@ +72,3 @@
> +
> +  /* Assume AudioUnitRemovePropertyListenerWithUserData will block
> +   * until the last listener finishes, so core_audio remains valid. */
> 
> Can we confirm this assumption (which sounds reasonable), and if not, deal
> with it however appropriate?

I don't have access to Apple's sources :(

I've asked on the coreaudio-api list but I've yet to receive a response
http://lists.apple.com/archives/coreaudio-api/2015/Feb/msg00028.html

I can't think of an appropriate way to deal with it. If you have zero
guarantees as to when a function can be called, there's little you can do.
However, I think we can just assume it doesn't get called after
AudioUnitUninitialize and remove this alarming comment.

> @@ +75,3 @@
> +  core_audio = GST_CORE_AUDIO (inRefCon);
> +  g_assert (inUnit == core_audio->audiounit);
> +  parent = core_audio->osxbuf->parent;  /* osxbuf is our owner, no need to
> ref it */
> 
> What is the function of the parent here?

Oops, remains of old code. Will remove.

> @@ +151,3 @@
> +      kAudioUnitProperty_AudioChannelLayout, _audio_unit_property_listener,
> +      core_audio);
> +  if (status) {
> 
> Not too important, but we check status != noErr in the rest of the code.

OK.
BTW, we have both plenty of "(status)" and "(status != noErr)" in the code.
Personally I prefer "(status != noErr)" for being more verbose, so I'll change
this particular instance.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.