2015-11-10 00:30:37

by Olof Johansson

[permalink] [raw]
Subject: Re: [PATCH v2 4/7] ARM: at91/dt: sama5d2: add pio controller node

Hi Linus,

On Mon, Sep 21, 2015 at 11:24 AM, Linus Walleij
<[email protected]> wrote:
> On Wed, Sep 16, 2015 at 8:37 AM, Ludovic Desroches
> <[email protected]> wrote:
>
>> Add pio4 controller node to enable pinmux and gpio.
>>
>> Signed-off-by: Ludovic Desroches <[email protected]>
>
> Patch applied.

Please don't merge DT changes through driver trees unless there's a
very specific reason to do so, since it introduces random conflicts.


Thanks!

-Olof


2015-11-17 11:06:08

by Linus Walleij

[permalink] [raw]
Subject: Re: [PATCH v2 4/7] ARM: at91/dt: sama5d2: add pio controller node

On Tue, Nov 10, 2015 at 1:30 AM, Olof Johansson <[email protected]> wrote:
> On Mon, Sep 21, 2015 at 11:24 AM, Linus Walleij
> <[email protected]> wrote:
>> On Wed, Sep 16, 2015 at 8:37 AM, Ludovic Desroches
>> <[email protected]> wrote:
>>
>>> Add pio4 controller node to enable pinmux and gpio.
>>>
>>> Signed-off-by: Ludovic Desroches <[email protected]>
>>
>> Patch applied.
>
> Please don't merge DT changes through driver trees unless there's a
> very specific reason to do so, since it introduces random conflicts.

Sorry :(

Even noted this in the pull request to Torvalds, it was in the bottom
of my patch stack so had been in -next for ages, I was afraid it
would create more problem than it solves if I reverted the patch,
but I guess I should have done so anyways.

Yours,
Linus Walleij

2015-11-17 19:15:16

by Olof Johansson

[permalink] [raw]
Subject: Re: [PATCH v2 4/7] ARM: at91/dt: sama5d2: add pio controller node

On Tue, Nov 17, 2015 at 3:06 AM, Linus Walleij <[email protected]> wrote:
> On Tue, Nov 10, 2015 at 1:30 AM, Olof Johansson <[email protected]> wrote:
>> On Mon, Sep 21, 2015 at 11:24 AM, Linus Walleij
>> <[email protected]> wrote:
>>> On Wed, Sep 16, 2015 at 8:37 AM, Ludovic Desroches
>>> <[email protected]> wrote:
>>>
>>>> Add pio4 controller node to enable pinmux and gpio.
>>>>
>>>> Signed-off-by: Ludovic Desroches <[email protected]>
>>>
>>> Patch applied.
>>
>> Please don't merge DT changes through driver trees unless there's a
>> very specific reason to do so, since it introduces random conflicts.
>
> Sorry :(
>
> Even noted this in the pull request to Torvalds, it was in the bottom
> of my patch stack so had been in -next for ages, I was afraid it
> would create more problem than it solves if I reverted the patch,
> but I guess I should have done so anyways.

Yeah, it's OK -- I spotted that pull request later as well.

It's not a big deal in most specific instances, I'd say. It's just in
aggregate it becomes a bother.

So, just see this as a public reminder since we've seen it creep into
other driver trees a bit more lately. Mistakes will still happen but
try to keep it down. And for those who submit patches, feel free to
point out in the patch that you don't expect the driver/subsystem
maintainer to apply it to help them out.


-Olof

2015-11-18 07:25:15

by Ludovic Desroches

[permalink] [raw]
Subject: Re: [PATCH v2 4/7] ARM: at91/dt: sama5d2: add pio controller node

On Tue, Nov 17, 2015 at 11:15:11AM -0800, Olof Johansson wrote:
> On Tue, Nov 17, 2015 at 3:06 AM, Linus Walleij <[email protected]> wrote:
> > On Tue, Nov 10, 2015 at 1:30 AM, Olof Johansson <[email protected]> wrote:
> >> On Mon, Sep 21, 2015 at 11:24 AM, Linus Walleij
> >> <[email protected]> wrote:
> >>> On Wed, Sep 16, 2015 at 8:37 AM, Ludovic Desroches
> >>> <[email protected]> wrote:
> >>>
> >>>> Add pio4 controller node to enable pinmux and gpio.
> >>>>
> >>>> Signed-off-by: Ludovic Desroches <[email protected]>
> >>>
> >>> Patch applied.
> >>
> >> Please don't merge DT changes through driver trees unless there's a
> >> very specific reason to do so, since it introduces random conflicts.
> >
> > Sorry :(
> >
> > Even noted this in the pull request to Torvalds, it was in the bottom
> > of my patch stack so had been in -next for ages, I was afraid it
> > would create more problem than it solves if I reverted the patch,
> > but I guess I should have done so anyways.
>
> Yeah, it's OK -- I spotted that pull request later as well.
>
> It's not a big deal in most specific instances, I'd say. It's just in
> aggregate it becomes a bother.
>
> So, just see this as a public reminder since we've seen it creep into
> other driver trees a bit more lately. Mistakes will still happen but
> try to keep it down. And for those who submit patches, feel free to
> point out in the patch that you don't expect the driver/subsystem
> maintainer to apply it to help them out.
>

Ok. Sorry, I didn't think it could cause so many conflicts if not taken
by Nicolas.


Ludovic