Re: [RESEND x3][PATCH v4] arm64: dts: hi6220: Add k3-dma and i2s/hdmi audio support

[John Stultz]

On Tue, Jun 13, 2017 at 2:46 PM, Mark Brown <broonie@xxxxxxxxxx> wrote:
> On Tue, Jun 13, 2017 at 01:42:15PM -0700, John Stultz wrote:
>
>> Its a trivial bit of code, but its in the drm directory and I suspect
>> at this stage I'm not going to be able to get maintainer attention for
>> it in time for 4.13 (plus since it depends on things in the ASoC tree,
>> so it needs cross-maintainer acks and probably to go through the ASoC
>> tree).
>
> Can I apply it directly?  Given that it's all ASoC code it might be the
> most expedient way to handle this.  It doesn't look like there's any DRM
> changes to meaningfully conflict with.

I just sent it out, and I'll leave it to your discretion as to if it
can be picked up and go through your tree or if we need to wait for
acks or the next merge window.

>> Plus the new dts changes with the graph need review (where as the
>> existing ones have been sent out 5 or so times w/o comment) before
>> they are merged.  Its just a ton of extra friction that doesn't gain
>> much and just keeps functionality from working upstream.
>
> What's happened here is that because you've been sending it for so long
> the code around it changed so there's new review comments.
>
>> As for your concern about follow up, I've got the patches lined up here:
>>   https://git.linaro.org/people/john.stultz/android-dev.git/commit/?h=dev/hikey-mainline-WIP&id=31e8b980d93087cf30cd708fb0b2cc48e906d003
>
>>   https://git.linaro.org/people/john.stultz/android-dev.git/commit/?h=dev/hikey-mainline-WIP&id=f24a0ccdeeb46ce3eb8864fd43f7a34c673ff9da
>
>> And will revise and submit the adv7511 one for initial review today.
>
>> So again, I'm happy to migrate over. But I just think, as the graph
>> stuff doesn't actually benefit us much in this specific case (there is
>> only one supported audio out path right now), its not worth blocking
>> getting the functionality working upstream for another 3-6 months.
>
> If it might take two kernel releases to get such a trivial change in
> then that's not giving me a warm fuzzy feeling...   there's something
> not working very well with the HiKey support overall with frequent
> process problems (only today I had someone resubmitting already applied
> code).

Well, I can't speak to other things, but we do seem to have trouble
getting dts changes picked up via Wei, and we probably need to add
some extra maintainers to make sure things don't drop through. I think
Ulf was considering this?

thanks
-john