Re: [PATCH] checkpatch: Fix check for embedded filename

From: Dwaipayan Ray
Date: Sat Jun 05 2021 - 15:01:38 EST


On Sun, Jun 6, 2021 at 12:03 AM Joe Perches <joe@xxxxxxxxxxx> wrote:
>
> On Sat, 2021-06-05 at 23:03 +0530, Dwaipayan Ray wrote:
> > On Sat, Jun 5, 2021 at 10:51 PM Joe Perches <joe@xxxxxxxxxxx> wrote:
> > >
> > > On Sat, 2021-06-05 at 21:42 +0530, Dwaipayan Ray wrote:
> > > > When checkpatch is run on file contents piped through stdin,
> > > > it may generate false EMBEDDED_FILENAME warnings if the
> > > > --file flag is used.
> > >
> > > I think this is a "don't do that" scenario and this change
> > > is not necessary.
> > >
> >
> > Okay then. I will drop this.
> > Sorry for the noise.
>
> I think you want something like this:
> ---
> scripts/checkpatch.pl | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> index dad87c3686326..582f8e07d32d5 100755
> --- a/scripts/checkpatch.pl
> +++ b/scripts/checkpatch.pl
> @@ -331,6 +331,7 @@ help(0) if ($help);
>
> die "$P: --git cannot be used with --file or --fix\n" if ($git && ($file || $fix));
> die "$P: --verbose cannot be used with --terse\n" if ($verbose && $terse);
> +die "$P: -f/--file requires at least one filename\n" if ($file && $#ARGV < 0);
>
> if ($color =~ /^[01]$/) {
> $color = !$color;
>

Yes, that's nice.
Most of the checks don't work with piped input when --file
is specified. So disabling it will avoid any confusion.

I can send in a patch if it's okay with you.

Thanks,
Dwaipayan.