[alsa-devel] [Sound-open-firmware] [PATCH v4 16/20] ASoC: SOF: Add PCI device support

[Pierre-Louis Bossart]

On 4/1/19 3:05 PM, Takashi Iwai wrote:
> On Mon, 01 Apr 2019 19:26:13 +0200,
> Pierre-Louis Bossart wrote:
>>
>> On 3/29/19 8:30 PM, Pierre-Louis Bossart wrote:
>>> On 3/28/19 6:08 PM, Andy Shevchenko wrote:
>>>> On Thu, Mar 28, 2019 at 02:21:47PM -0400, Pierre-Louis Bossart wrote:
>>>>>>>> +#if IS_ENABLED(CONFIG_SND_SOC_SOF_EDISON)
>>>>>>>
>>>>>>> Can we use Merrifield / mrfld instead of EDISON in entire series?
>>>>>
>>>>> we could, but I don't know of any other platform than Edison to run the
>>>>> code. I know it's less accurate from an architecture perspective
>>>>> but felt
>>>>> Merrifield was confusing for non-Intel folks.
>>>>
>>>> We use Merrifield across the entire kernel. It would be confusing
>>>> other way around.
>>>> So, please, change it to be consistent with the rest of the kernel.
>>>
>>> Since this is the only comment so far, I'll add a follow-up patch.
>>
>> Discard this reply, I'll address this in an update to address
>> Takashi's comments as well.
>>
>>
>> Actually I need feedback from reviewers/maintainers: I could either
>> provide an update addressing just comments from Andy and Takashi, or
>> provide a larger update that would include known fixes and
>> simplifications from SOF contributors on github, e.g. on the IPC. We
>> have about 15-20 delta patches that were accepted on github, not sure
>> what the preference is, just addressing comments so far or getting the
>> latest and greatest patches squashed?
> 
> IMO, one or two more whole patchset refresh would be still worth.
> Then we can merge the base, and go for refinement with each small
> change, hopefully in this merge cycle.
> 
> For the resubmission, please give me a bit more time.  I had no time
> for further review in the last week due to completely other
> businesses (the office room movement, etc)...

Sure, I am understand how time consuming this can be and how 'reviewer 
fatigue' can occur. We can add a couple of additional changes, e.g. for 
the IPC and some resource management while you (and others) go over the 
current patchset and tentatively resubmit next week, would that work?