[Bug 769278] aacparse: a few fixes and improvements for LOAS parsing

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=769278

--- Comment #14 from Vincent Penquerc'h <vincent.penquerch at collabora.co.uk> ---
(In reply to Sebastian Dröge (slomo) from comment #12)

> +    if (audio_object_type == 22) {
> 
> How can we have audio_object_type == 5 || 29 above, and then here suddenly
> 22? Does get_audio_object_type() read it from a different place than where
> we got it before?

It reads from the bitstream every time, so it's a different place, yes. The
spec for all that format here is fascinatingly complex (hence why a lot of it
is left out).

> @@ +586,3 @@
> +  (void) extension_audio_object_type;
> +  (void) ps;
> +  (void) sbr;
> 
> I'd rather comment these out. It's relatively common to do these things, but
> it's actually undefined C behaviour AFAIU. The compiler would be correct to
> just optimize away the whole function, call system("rm -rf /") or anything
> else here ;)

Casting a rvalue to void is undefined behavior ?

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