[PATCH] MAINTAINERS: add arm at kernel.org as the list for arm64 defconfig changes

[Olof Johansson]

On Wed, May 3, 2017 at 10:33 AM, Timur Tabi <timur at codeaurora.org> wrote:
> On 05/03/2017 12:26 PM, Olof Johansson wrote:
>> arm at kernel.org as an email alias will break down if it starts being
>> cc:d on large number of patches. Ideally we want the platform
>> maintainers as filters/reviewers in front to keep the volume down. So
>> far it's been kept to nearly only material from maintainers, and
>> mostly pull requests.
>>
>> Defconfig changes usually come in through maintainers. Said
>> maintainers already know to send code to us at arm at kernel.org for
>> their platforms. So this should probably go in through Andy Gross? You
>> didn't even cc him, so added now.
>
> How would I know to CC: him, or anyone else, without a MAINTAINERS entry?
> I've already posted these patches 5 times, and every time I do, someone
> tells me that I'm CCing the wrong person.

There's a handy tool called get_maintainer:

olof at brutus:~/work/linux ((52aa3ec...)) $ ./scripts/get_maintainer.pl
-f arch/arm64/configs/defconfig
Catalin Marinas <catalin.marinas at arm.com> (maintainer:ARM64 PORT
(AARCH64 ARCHITECTURE),commit_signer:6/61=10%)
Will Deacon <will.deacon at arm.com> (maintainer:ARM64 PORT (AARCH64 ARCHITECTURE))
Arnd Bergmann <arnd at arndb.de> (commit_signer:10/61=16%)
Kevin Hilman <khilman at baylibre.com> (commit_signer:9/61=15%,authored:6/61=10%)
Olof Johansson <olof at lixom.net> (commit_signer:6/61=10%)
Wei Xu <xuwei5 at hisilicon.com> (commit_signer:6/61=10%)

If you read the output, you can see that the people who have applied
most of these patches have been Will, Arnd, Kevin and myself. Adding
those on Cc could have been a good first step.


> Are you saying that I should send the 9 patches to 9 different people, and
> just pray that they all get merged in properly?  Each of the maintainers is
> going to ask for an ACK from an ARM maintainer anyway, whoever he is (I
> don't know any more).
>
> There's got to be a better way.

No, I'm not saying that, and there is a better way: Send it through
Andy if in doubt -- it should be his job as your platform maintainer
to help guide you through the system if needed.

>> We won't consider any patches at this very time, since we're in the
>> middle of the merge window. Main priority right now is to prepare the
>> material we have for pull requests and merge.
>>
>> This is not the time to send us new material, but you probably already
>> know that.
>
> Fair enough, but the patches aren't new.  They've been been floating on the
> list for three weeks now.

While I understand your frustration, this is also not how it works at our end.

Sometimes we tweak defconfigs late in the merge window and sometimes
even during early -rc, so let us get back to this after we've handled
the merges.


-Olof