[alsa-devel] [PATCH v1 8/9] ASoC: hdac_hda: add asoc extension for legacy HDA codec drivers

[Pierre-Louis Bossart]

On 2/26/18 1:44 AM, Ughreja, Rakesh A wrote:
> 
> 
>> -----Original Message-----
>> From: Pierre-Louis Bossart [mailto:pierre-louis.bossart at linux.intel.com]
>> Sent: Friday, February 23, 2018 10:25 PM
>> To: Ughreja, Rakesh A <rakesh.a.ughreja at intel.com>; alsa-devel at alsa-
>> project.org; broonie at kernel.org; tiwai at suse.de;
>> liam.r.girdwood at linux.intel.com
>> Cc: Koul, Vinod <vinod.koul at intel.com>; Patches Audio
>> <patches.audio at intel.com>
>> Subject: Re: [PATCH v1 8/9] ASoC: hdac_hda: add asoc extension for legacy HDA
>> codec drivers
>>
>> On 2/23/18 2:12 AM, Rakesh Ughreja wrote:
>>> This patch adds a kernel module which is used by the legacy HDA
>>> codec drivers as library. This implements hdac_ext_bus_ops to enable
>>> the reuse of legacy HDA codec drivers with ASoC platform drivers.
>>>
>>> Signed-off-by: Rakesh Ughreja <rakesh.a.ughreja at intel.com>
>>> ---
>>>    sound/pci/hda/hda_bind.c    |   6 +
>>>    sound/soc/codecs/Kconfig    |   5 +
>>>    sound/soc/codecs/Makefile   |   2 +
>>>    sound/soc/codecs/hdac_hda.c | 448
>> ++++++++++++++++++++++++++++++++++++++++++++
>>>    sound/soc/codecs/hdac_hda.h |  23 +++
>>>    5 files changed, 484 insertions(+)
>>>    create mode 100644 sound/soc/codecs/hdac_hda.c
>>>    create mode 100644 sound/soc/codecs/hdac_hda.h
>>
>> so now we have both hdac_hdmi and hdac_hda?
>> Not sure I get it.
> 
> hdac_hdmi is the ASoC HDMI driver which exists today. All the
> intel ASoC driver which are primarily used for I2S codecs uses it.
> I am not deleting or removing the support for that.
> 
> hdac_hda is the ASoC wrapper around the legacy HDA drivers.
> 
> Now with this patch series, we have two choices for HDMI/iDisp
> codec driver. Either to use the legacy HDMI codec driver by using
> the ASoC wrapper or use the existing ASoC hdac_hdmi driver.
> 
> Since Intel ASoC platform driver is already proven and tested
> with ASoC hdac_hdmi driver, I am using that in this patch series.

I get your point, but I will assert that the legacy HDMI codec has been 
tested a lot more than the ASoC one (only for Chromebooks) so I wonder 
if we shouldn't deprecate hdac_hdmi moving forward. Having two codec 
implementations which both talk to the i915 driver makes no sense for 
long term support. We don't need to do this now but it should be on the 
TODO list along with topology support in machine drivers.