Re: [PATCH 1/3] Add patman patch automation script

[Simon Glass]

Hi Richard,

On 3 May 2015 at 14:13, Richard Weinberger <richard@xxxxxx> wrote:
> Simon,
>
> Am 03.05.2015 um 21:54 schrieb Simon Glass:
>> Hi Richard,
>>
>> On 3 May 2015 at 13:16, Richard Weinberger <richard.weinberger@xxxxxxxxx> wrote:
>>> On Sun, May 3, 2015 at 8:29 PM, Simon Glass <sjg@xxxxxxxxxxxx> wrote:
>>>> This tool is a Python script which:
>>>> - Creates patch directly from your branch
>>>> - Cleans them up by removing unwanted tags
>>>> - Inserts a cover letter with change lists
>>>> - Runs the patches through checkpatch.pl and its own checks
>>>> - Optionally emails them out to selected people
>>>
>>> Don't get me wrong but is this really worth 3000+ lines of python?
>>> The tasks you describe can be done using a few lines bash.
>>
>> #!/bin/bash
>> patman $@
>>
>> I obviously failed in my attempt to briefly explain what it does.
>> Please check out the cover letter [1], README [2], or perhaps use it
>> on a series. With respect to the length, it could be slimmed down a
>> bit if that is important.
>
> the README file did the trick. ;)
> Sounds like a useful tool to manage patch series.
>
> But I don't think it makes much sense to carry it with the Linux kernel tree.
> Other projects can also use it and it does not seem to be very Linux kernel
> specific.
> git, quilt and other great tools also have their own repositories.

My reasoning is that:

- more will find it / use it if it is in-tree
- it avoids installation and old-version problems (e.g. I suppose this
is why the device tree compiler is built-in)
- it is somewhat Linux-specific (e.g. uses get_maintainers,
checkpatch.pl) and can break if checkpatch.pl if the wrong version
(e.g. you check out and send patches from an older tree)
- it could be built into the Linux workflow [1] and might thereby
reduce the amount of confusion and errors (did you run checkpatch?,
your change log is in the wrong place, you forgot to add your
sign-off, etc.)

That said, I could see this having a repo of its own, with the Linux
version a downstream copy, a bit like dtc (and maybe Kbuild/Kconfig -
I don't know). The feature set is probably mature enough to support
that now.

Regards,
Simon

[1] https://www.kernel.org/doc/Documentation/SubmittingPatches section 14
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/