[PATCH] ASoC: core: restore dpcm flags semantics

[Jerome Brunet]

On Thu 30 Jul 2020 at 20:52, Mark Brown <broonie at kernel.org> wrote:

> On Thu, Jul 30, 2020 at 11:06:23AM -0500, Pierre-Louis Bossart wrote:
>> On 7/30/20 4:04 AM, Jerome Brunet wrote:
>> > On Wed 29 Jul 2020 at 17:56, Pierre-Louis Bossart <pierre-louis.bossart at linux.intel.com> wrote:
>> > > On 7/29/20 10:46 AM, Jerome Brunet wrote:
>
>> > > > The flag previously allowed card drivers to disable a stream direction on
>> > > > a link (whether or not such feature is deemed useful).
>
> Right, and I can see a use case for this if someone has a board that
> for some reason didn't physically connect one of the directions for some
> reason - perhaps they were running out of pins or something.  It's not
> clear if anyone's actually doing that though.
>
>> > > > Forcing the flags to be aligned with DAI caps just make the information
>> > > > the flag carry redundant with DAI caps, breaking a few cards along the way.
>
>> > > > This change drops the added error conditions and restore the initial flag
>> > > > semantics.
>
> I'm not 100% clear, have we actually found cases where the flags are
> used or is this something found through inspection and review?

One last thing I'd like to understand. Is this behavior of throwing an
error going to applied to the non-DPCM case as well ? so at least thing
are consistent between both cases ?

IOW:
 * An error is now throw if dpcm_capture is set on the link and the CPU
 DAI support playback_only
 * on non-DPCM links, will an error be thrown as well if playback_only
 is not set and the CPU on the link happen to not support capture ?
 
>
>> >   * It worked for every user of DPCM so a far.
>
>> Not completely true, when Morimoto-san added snd_soc_dai_stream_valid() it
>> exposed tons of cases where the information on direction was not provided in
>> a reliable at the DAI level. I will assert that we are still finding out
>> cases with broken DAI configurations, and as a result we will also find
>> broken dailink configurations. Your picture of DPCM as a perfectly
>> functional system that I broke is a distortion of reality.
>
>> The reality is that we have to work in steps, first make sure all DAIs are
>> properly described, then work on the dailinks and optimize at a later point.
>> we will need warnings to find out what the problem cases are, and move
>> slowly.
>
> This was all triggered by Morimoto-san's changes like you say.  DPCM has
> quite a lot of problems in general, here IIRC the issues were that we
> had multiple different ways of doing similar things which it wasn't
> quite clear if people were even using.  The intention with the warnings
> was to remove them one way or another, they're mainly intended to flush
> out actual active usage of the flags as opposed to redundant usage of
> them which could be confused/broken.
>
> This could definitely have been clearer in the changelogs though.