[alsa-devel] [PATCH] ALSA: hda - delay resume haswell hdmi codec in system resume

[Takashi Iwai]

At Fri, 5 Apr 2013 16:58:54 +0000,
Lin, Mengdong wrote:
> 
> > -----Original Message-----
> > From: Takashi Iwai [mailto:tiwai at suse.de]
> > Sent: Tuesday, April 02, 2013 5:53 PM
> > To: Lin, Mengdong
> > Cc: alsa-devel at alsa-project.org
> > Subject: Re: [alsa-devel] [PATCH] ALSA: hda - delay resume haswell hdmi codec
> > in system resume
> > 
> > At Tue, 26 Mar 2013 07:02:54 +0000,
> > Lin, Mengdong wrote:
> > >
> > > Is it okay to add a new parameter to snd_hda_resume(), to indicate whether
> > codec resume should be delayed or not?
> > > Is because snd_hda_resume() is only only called by azx_resume() but also
> > called by azx_bus_reset().
> > > The latter can be called on a fatal verb execution failure and codec resume
> > delay is not suitable for this case.
> > 
> > Well, I prefer creating a new function (e.g. snd_hda_bus_reset()) called from
> > azx_bus_reset().
> > 
> > 
> > Takashi
> 
> Hi Takashi,
> 
> Do you mean that in azx_bus_reset(), we can merge the snd_hda_suspend() and snd_hda_resume() to a new function like snd_hda_bus_reset()?

Yes.  Calling snd_hda_suspend() following snd_hda_resume() immediately
is just an ad hoc way of reset.  It should have been a dedicated reset
function from the beginning.


Takashi

> snd_hda_suspend() and snd_hda_resume() look symmetrical here and so I don't want to destroy the symmetry.
> 
> static void azx_bus_reset(struct hda_bus *bus)
> {
> 	...
> #ifdef CONFIG_PM
> 	if (chip->initialized) {
> 		struct azx_pcm *p;
> 		list_for_each_entry(p, &chip->pcm_list, list)
> 			snd_pcm_suspend_all(p->pcm);
> 		snd_hda_suspend(chip->bus);
> 		snd_hda_resume(chip->bus);
> 	}
> #endif
> 	bus->in_reset = 0;
> }
> 
> Thanks
> Mengdong
>