Re: [GIT PULL] final round of SCSI updates for the 5.15+ merge window

[Linus Torvalds <torvalds@linux-foundation.org>]

On Fri, Nov 12, 2021 at 5:43 AM James Bottomley
<James.Bottomley@hansenpartnership.com> wrote:
>
> This series is all the stragglers that didn't quite make the first
> merge window pull.  It's mostly minor updates and bug fixes of merge
> window code but it also has two driver updates: ufs and qla2xxx.

Hmm? No diffstat?

I suspect it's because there's a merge in there, and thus multiple
merge base commits, and the autogenerated diffstat ends up being
worthless.

In that situation, the nice thing to do is to at least tell me why
there's no diffstat, but optimally you can do a temporary throw-away
merge in a temporary branch just to get the diffstat.

And yes, "git request-pull" could do that, automating this all and
reporting any conflicts at the same time.

But git historically did *not* do that just because it requires a
working tree and can be messy, and because the "just do the diff from
the merge base" works fine for maintainers that don't do merges
themselves, and so the only maintainers that can hit this issue are
the maintainers that also should be able to do that temporary merge
thing on their own.

Anyway, I don't require that temporary merge, but I _do_ really want
to get notified of "look, I did the diffstat, and it was useless
garbage, so I'm not including it here".

Because as-is, this just looks like an incomplete pull request.

I've done the pull, verified the shortlog, and checked that the
(proper) diffstat all looks sane. But I am writing this email just to
say "you could have done so much better".

                 Linus