[Linux-kernel-mentees] comment on checkpatch

[Lukas Bulwahn]

On Tue, Jan 12, 2021 at 7:02 AM Joe Perches <joe at perches.com> wrote:
>
> On Tue, 2021-01-12 at 11:05 +0530, Dwaipayan Ray wrote:
> > On Tue, Jan 12, 2021 at 10:38 AM Lukas Bulwahn <lukas.bulwahn at gmail.com> wrote:
> > >
> > > https://lore.kernel.org/linux-next/20210111230029.GM4728@sirena.org.uk/T/#m355576186555c4b6e85dbb9a1b3fe3d5e80f6e9f
> > >
> > > Aditya, Dwaipayan, Have a look. checkpatch already handles this, right?
> > >
> > Hi
> > The following Fixes line in commit log emits no warning:
> >    Fixes: Use packed mode for 32 bits per word transfers to increase
> >
> > Neither does a fixes with mismatched commit description:
> >    Fixes: e4da63cda51f ("networking: packet_mmap: fix old config reference")
> >
> > which should have been infact:
> >    Fixes: e4da63cda51f ("docs: networking: packet_mmap: fix old config
> > reference")
> >
> > I could work on this if Lukas and Joe are okay with it.
>
> It's not really a solvable problem as the SHA1 probably exists in an
> individual local tree.
>
>

Joe, so did you deactivate checking Fixes: at some point?

We can always check if it makes locally sense; that still does not
avoid the situation when a Fixes to an individual local tree and later
during integration hashes change, but it does sanity-checking on Fixes
tags for the hashes and descriptions that are stable.

Lukas