Re: [PATCH v1 1/5] watchdog: intel_scu_watchdog: Remove driver for deprecated platform

[Hans de Goede <hdegoede@redhat.com>]

Hi,

On 1/25/21 12:15 PM, Andy Shevchenko wrote:
> On Sat, Jan 23, 2021 at 2:08 PM Hans de Goede <hdegoede@redhat.com> wrote:
>> On 1/23/21 1:27 AM, Guenter Roeck wrote:
>>> On 1/22/21 1:04 PM, Hans de Goede wrote:
>>>> On 1/22/21 4:45 PM, Andy Shevchenko wrote:
>>>>> On Fri, Jan 22, 2021 at 05:22:25PM +0200, Andy Shevchenko wrote:
>>>>>> On Fri, Jan 22, 2021 at 03:59:34PM +0100, Hans de Goede wrote:
> 
>>>>>>> What is the plan for merging this series ?
>>>>>>>
>>>>>>> It touches files under:
>>>>>>> arch/x86
>>>>>>> drivers/watchdog
>>>>>>> drivers/platform/x86
>>>>>>>
>>>>>>> It is probably best if the entire series is merged through a single
>>>>>>> tree. I don't expect this to cause any conflicts with current / upcoming
>>>>>>> changes under drivers/platform/x86, so I'm fine with this being merged
>>>>>>> through another tree.
>>>>>>>
>>>>>>> Or if I can get an ack for that from the x86 and watchdog maintainers
>>>>>>> I can merge the entire series through the pdx86 tree.
>>>>>>
>>>>>> For time being I'm collecting tags and comments.
>>>>>> As of today this series is independent per se, but other (later) clean up
>>>>>> patches may rely on this.
>>>>>>
>>>>>> One strategy can be providing an immutable branch for anybody who wants it and
>>>>>> merge thru PDx86 or watchdog tree (I guess PDx86 is preferable because there
>>>>>> are more patches touching other stuff, though independently, in this matters).
>>>>>
>>>>> Yes, I think one more time and guess the best is:
>>>>>  - get tags from x86 / watchdog
>>>>>  - get tag from RTC
>>>>>  - collect everything in PDx86 immutable branch
>>>>
>>>> Ok, that works for me. I assume that it is best for the other subsystems
>>>> if I create an immutable branch for this based on 5.11-rc1
>>>> (and then merge that into pdx86/for-next) ?
>>>>
>>>> Guenter, I believe you have already reviewed all the watchdog changes,
>>>> are you ok with me creating an immutable branch for the entire series
>>>> and then sending you a pull-req for that ?
>>>
>>> Wim handles watchdog pull requests,
> 
> Wim, can you Ack the watchdog changes?
> 
>> Ah, sorry I was under the impression you would do that because you were
>> doing the reviews (which I'm sure Wim appreciates).
>>
>>> and he usually creates a branch
>>> for Linus to pull from either during or shortly before a commit window.
>>> Also, I don't immediately see why the watchdog tree should carry the other
>>> patches of this series, immutable or not; I would not expect a dependency
>>> against those.
>>>
>>> Anyway, I am fine if the the wdt driver removal is submitted through some
>>> other branch.
>>
>> Hmm, looking at the code again it indeed seems that patch 1/5 + 2/5 could
>> be simply merged through the watchdog tree, as they don't appear to
>> have any interdependencies with the arch/x86 and drivers/platform/x86
>> changes.  Andy ?
> 
> The first patch is a dependency to the later (that are not published yet).
> I was considering gathering all dependent patches in immutable
> branches and doing in three parts / phases.
> 
> I can prepare a branch with a first batch (with Acks from respective
> maintainers) and send a PR to you.
> Would it work?

I guess you would then publish this as an immutable branch yourself and
send a PR to all involved subsys-maintainers ?

You have commit rights to the pdx86/platform-drivers-x86 git repo if
I'm not mistaken, so you could publish the immutable branch there if
you want.

That actually sounds like a good plan to me, assuming the other maintainers
are ok with it too.

Regards,

Hans