[Linux-kernel-mentees] [RFC PATCH v2] checkpatch: add shebang check to EXECUTE_PERMISSIONS

Lukas Bulwahn lukas.bulwahn at gmail.com
Fri Oct 16 10:33:37 UTC 2020


On Fri, Oct 16, 2020 at 12:24 PM Dwaipayan Ray <dwaipayanray1 at gmail.com> wrote:
>
> On Fri, Oct 16, 2020 at 3:41 PM Lukas Bulwahn <lukas.bulwahn at gmail.com> wrote:
> >
> > >
> > > Lukas, Ayush didn't follow up with next iteration of patches for
> > > GIT_COMMIT_ID related issues. And I had discovered a bug in his
> > > implementation. Is it okay if I look into the issue?
> > >
> >
> > Dwaipayan already wanted to pick up the work on the GIT_COMMIT_ID issues.
> > So, he is working on that already.
> >
> > > >
> > > > Just one bigger project example: Comparing clang-format suggestions on
> > > > patches against checkpatch.pl suggestions are fine-tuning both of them to fit to
> > > > the actual kernel style.
> > >
> > > I'm not familiar with how clang-format tool is used in the kernel
> > > repository. But I would like to know about it more and how I can
> > > contribute here. If there are any tasks that would help me get a
> > > direction please mention.
> > >
> >
> > This is probably much more interesting to get involved in...
> >
> > Some background information is in the kernel documentation; some
> > further information can be found on this presentation,
> > https://www.youtube.com/watch?v=FFjV9f_Ub9o, starting at 2:09.
> >
> >
> > You can start with running clang-format on the kernel sources just to
> > get a first feeling on what the tool does and where it completely
> > changes the code, and especially those cases where it changes code
> > that was perfectly okay before, we should do something to
> > adjust/configure clang-format.
> >
> > Mid term goal: If we find a directory that needs no changes with
> > clang-format (after configuring), then it would be worth informing the
> > maintainers that clang-format works for them.
> >
> > Lukas
>
> Hi Lukas, Hi Ujjwal,
> This seems pretty interesting to me. I could help out with the
> evaluation work as well considering the kernel source is very
> vast.
>
> So currently if we run both checkpatch and clang-format on the
> kernel sources and notice substantial differences between the two,
> which one is preferable? Or do we just go with intuition and take
> whatever the kernel programming style generally suggests?
>

Basic assumption probably needs to be:

Both tools are wrong ;) and first intuition is maybe the best guess,
and still wrong ;) Only the maintainer really knows the preferred
style...

But when we start looking at it, we will adjust tools and ask
maintainers about their preferences and then we simply document those
preferences and create patches for kernel documentation. This
contributes directly to the maintainer entry profiles (something that
is still not very common, though...).

Getting a good overview on what both tools report will give some
interesting initial insights into this topic.

Let us get all the ongoing work on kernel patches done and then start
looking into this...

Lukas


More information about the Linux-kernel-mentees mailing list