Re: [Qemu-devel] [PATCH v3 15/50] audio: reduce glob_audio_state usage

[Gerd Hoffmann]

On Wed, Jan 23, 2019 at 09:16:07PM +0100, Zoltán Kővágó wrote:
> On 2019-01-17 10:22, Gerd Hoffmann wrote:
> > On Thu, Jan 17, 2019 at 12:36:48AM +0100, Kővágó, Zoltán wrote:
> >> Remove glob_audio_state from functions, where possible without breaking
> >> the API.  This means that most static functions in audio.c now take an
> >> AudioState pointer instead of implicitly using glob_audio_state.  Also
> >> included a pointer in SWVoice*, HWVoice* structs, so that functions
> >> dealing them can know the audio state without having to pass it around
> >> separately.
> >>
> >> This is required in order to support multiple simultaneous audio
> >> backends (added in a later commit).
> > 
> > I think it makes sense to split the series into smaller pieces, here
> > would be a good place for the split (i.e. patches 1-14 which introduce
> > -audiodev will be the first batch, this patch starts the next batch).
> 
> Alright, I think this series is pretty useless without at least the
> multiple simultaneous backends part, but if it's easier to review in
> smaller chunks, let's do it.

Also start merging stuff even though later in the series there are still
open issues to be tackled (coreaudio breaking for example).

So, I think with the first part the only open issue is whenever we go
with the nested types (i.e. patch #1 as-is) or not.  Given that the
one-element-structs added in that patch will get additional fields I
think the nesting makes sense.

cheers,
  Gerd