Re: [PATCH] MAINTAINERS: hardening: Add __counted_by regex

[Justin Stitt]

On Tue, Sep 26, 2023 at 5:35 PM Justin Stitt <justinstitt@xxxxxxxxxx> wrote:
>
> On Tue, Sep 26, 2023 at 1:57 PM Eric Biggers <ebiggers@xxxxxxxxxx> wrote:
> >
> > Hi Kees,
> >
> > On Mon, Sep 25, 2023 at 10:20:41AM -0700, Kees Cook wrote:
> > > Since __counted_by annotations may also require that code be changed to
> > > get initialization ordering correct, let's get an extra group of eyes on
> > > code that is working on these annotations.
> > >
> > > Signed-off-by: Kees Cook <keescook@xxxxxxxxxxxx>
> > > ---
> > >  MAINTAINERS | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/MAINTAINERS b/MAINTAINERS
> > > index 737dcc7a2155..741285b8246e 100644
> > > --- a/MAINTAINERS
> > > +++ b/MAINTAINERS
> > > @@ -11405,6 +11405,7 @@ F:    kernel/configs/hardening.config
> > >  F:   mm/usercopy.c
> > >  K:   \b(add|choose)_random_kstack_offset\b
> > >  K:   \b__check_(object_size|heap_object)\b
> > > +K:   \b__counted_by\b
> > >
> >
> > Are you sure you want to volunteer to maintain every file that contains
> > "__counted_by"?  That's what "K" does; get_maintainer.pl will list you (and
> > linux-hardening@xxxxxxxxxxxxxxx) for every such file.
>
> Do people call get_maintainer.pl on specific tree files as opposed to
> invoking it against a .patch file? In the event of the .patch file
> "K:" should only pick-up what's in the patch and not read into the
> files outside of the context that the diff provides.

FWIW, b4 just uses the patches and not entire files:

   ...
    try:
        tos, ccs, tag_msg, patches = get_prep_branch_as_patches()
    except RuntimeError:
        logger.info('No commits in branch')
        return

    logger.info('Collecting To/Cc addresses')
    # Go through the messages to make to/cc headers
    for commit, msg in patches:
        if not msg or not commit:
            continue

        logger.debug('Collecting from: %s', msg.get('subject'))
        msgbytes = msg.as_bytes()
        ...


>
> If needed, I could send a patch adding a "D:" which would only
> consider patches and not tree files -- reducing noise.
>
> >
> > Other users of "K" have been surprised by this behavior.  It seems that most
> > people expect it to only apply to patches, not to files.  Given that you're
> > interested in using this functionality, have you considered updating
> > checkpatch.pl to handle it in the way that you probably expect that it works?
> >
> > - Eric
> >
>
> Thanks
> Justin