Re: Re: [PATCH v3] RDMA/siw: Pass a pointer to virt_to_page()

[Leon Romanovsky <leon@kernel.org>]

On Mon, Sep 05, 2022 at 02:07:34PM +0200, Linus Walleij wrote:
> On Mon, Sep 5, 2022 at 2:02 PM Bernard Metzler <BMT@zurich.ibm.com> wrote:
> 
> > Can we easily fix the two line wraps introduced by this
> > patch? Without sending an explicit patch on top --
> 
> Yeah Lean can just augment it when applying.

I already promoted that patch to non-rebasable for-next.

> 
> > I'd
> > suggest adding just two line breaks to it. I'd be happy
> > to see siw code continues to adhere to the 80 char's
> > per line style.
> 
> You will be fighting an uphill battle since checkpatch (which is
> what we use to check syntax) now accepts 100 chars/line.
> commit bdc48fa11e46f867ea4d75fa59ee87a7f48be144
> "checkpatch/coding-style: deprecate 80-column warning"
> 
> If there is infiniband consensus to stay with 80 chars per
> line, you should send a patch to checkpatch so that it
> warns for this for patches to drivers/rdma.

It is not infiniband specific, many other subsystems and reviewers
continue to use 80-char limit.

The change to checkpatch came after Linus said that authors should
use their best judgment while dealing with line lengths. Unfortunately,
it was vague enough to apply it to checkpatch.

We continue to use 80 char limit, because clang formatter continues
to wrap everything to 80 chars.

Thanks

> 
> Yours,
> Linus Walleij