[alsa-devel] [RFC v3 06/11] ASoC: hdac_hda: add ASoC based HDA codec driver

[Takashi Iwai]

On Thu, 21 Dec 2017 17:39:10 +0100,
Ughreja, Rakesh A wrote:
> 
> 
> 
> >-----Original Message-----
> >From: Takashi Iwai [mailto:tiwai at suse.de]
> >Sent: Thursday, December 21, 2017 9:18 PM
> >To: Ughreja, Rakesh A <rakesh.a.ughreja at intel.com>
> >Cc: Mark Brown <broonie at kernel.org>; alsa-devel at alsa-project.org; Koul, Vinod
> ><vinod.koul at intel.com>; pierre-louis.bossart at linux.intel.com;
> >liam.r.girdwood at linux.intel.com; Patches Audio <patches.audio at intel.com>
> >Subject: Re: [alsa-devel] [RFC v3 06/11] ASoC: hdac_hda: add ASoC based HDA
> >codec driver
> 
> >>
> >> Let me know if this is the right way to go. I can send you the full series
> >> once you are okay with the direction.
> >
> >Yes, this is a sort of idea I had.
> >My original thought was to have an extra ops in hdac_bus, and refer
> >directly there instead of setting in each hda_codec object.  Not coded
> >yet, so I can't judge which one is better.  Maybe you can actually
> >quick-try and test differences.
> 
> Let me try that out and will move to hdac_bus if I don't run into any
> practical issues.
> 
> >
> >But still an open question is where to hook.  You've put the branch at
> >the very beginning of each probe/remove, that is, completely replacing
> >the whole probe/remove callbacks.  Meanwhile, the legacy codec driver
> >still expects the legacy hda_codec object handling, so keeping more
> >common stuff would make sense.  That is, if we do switching at the
> >beginning, the rest should just compose the same helper functions in
> >slightly different ways.  Or we share most parts of probe/remove in
> >both legacy and asoc but branch off after some later point in the
> >probe/remove functions.
> 
> I think I am wrong in doing the above. I was thinking that I cannot do
> all those steps done in the probe before the snd_soc_codec probe gets
> called.
> 
> After looking at the code again after your comments, It looks like 
> I can call all these before the card is registered by the machine driver.
> 
> Is that right understanding ? same thing is happening in legacy
> driver also. The card is registered at the end. So I will just skip that step.

Well, I'm not 100% sure about the ordering, so we just need trying.


> >In anyway, a whole patchset would be helpful so that I can give it a
> >try, too.  But maybe I'll have little time tomorrow and a few days
> >thereafter due to vacation.
> 
> Sure, will send you the series tomorrow by correcting few more things.

OK, thanks.


Takashi