RE: [RFC PATCH v4 00/13] module: core code clean up

[Christophe Leroy]

> -----Message d'origine-----
> De : Aaron Tomlin <atomlin@xxxxxxxxxx>
> Envoyé : lundi 7 février 2022 19:02
> À : Christophe Leroy <christophe.leroy@xxxxxxxxxx>
> Cc : Aaron Tomlin <atomlin@xxxxxxxxxxx>; Luis Chamberlain
> <mcgrof@xxxxxxxxxx>; Michal Suchanek <msuchanek@xxxxxxx>; cl@xxxxxxxxx;
> pmladek@xxxxxxxx; mbenes@xxxxxxx; akpm@xxxxxxxxxxxxxxxxxxxx;
> jeyu@xxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; linux-modules@xxxxxxxxxxxxxxx;
> live-patching@xxxxxxxxxxxxxxx; ghalat@xxxxxxxxxx; allen.lkml@xxxxxxxxx;
> void@xxxxxxxxxxxxx; joe@xxxxxxxxxxx
> Objet : Re: [RFC PATCH v4 00/13] module: core code clean up
> 
> On Mon 2022-02-07 17:17 +0000, Christophe Leroy wrote:
> > Yes and that's the purpose of the patch I proposed at
> > https://patchwork.kernel.org/project/linux-
> modules/patch/203348805c9ac9851d8939d15cb9802ef047b5e2.1643919758.gi
> t.christophe.leroy@xxxxxxxxxx/
> 
> I see.
> 
> > Allthough I need to find out what's the problem reported by the robot.
> 
> I'll have a look too.
> 
> > As suggested by Luis, this fix should go once all ongoing work is done.
> > But it would be nice if you could just remove patch 5 from you series,
> > otherwise we would have to revert it later.
> 
> Perhaps it might be easier if I keep the patch within the series; once
> merged into module-next, by Luis, you can rebase and then add the "Fixes:"
> tag to resolve the issue, no?

I don't think it is easier.

If we do that it means we'll move some code from main.c to arch_strict_rwx.c with your series, then move it back to main.c and remove arch_strict_rwx.c when we do the fix. That's not good for history tracking because the code we move back and forth will appear as new code in main.c whereas it's code that has been there for years.

As we know the code will be back in main.c at the end, I looks easier to me to not move it at all by not applying your patch 5.

Thanks
Christophe