linux-kernel - Re: [PATCH v1] checkpatch: Warn about sign offs if there's an accidental patch separator

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAP-5=fWR+Mwme6ogL6=WQG8gJ2J+wHkva8wBHdCrfiX57nijfw@mail.gmail.com>
Date: Wed, 14 Jan 2026 11:01:24 -0800
From: Ian Rogers <irogers@...gle.com>
To: Joe Perches <joe@...ches.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>, Kuan-Wei Chiu <visitorckw@...il.com>, 
	Dwaipayan Ray <dwaipayanray1@...il.com>, Lukas Bulwahn <lukas.bulwahn@...il.com>, 
	Stephen Rothwell <sfr@...b.auug.org.au>, Andy Whitcroft <apw@...onical.com>, linux-kernel@...r.kernel.org, 
	Namhyung Kim <namhyung@...nel.org>
Subject: Re: [PATCH v1] checkpatch: Warn about sign offs if there's an
 accidental patch separator

On Wed, Jan 14, 2026 at 10:20 AM Joe Perches <joe@...ches.com> wrote:
>
> On Tue, 2026-01-13 at 19:33 -0800, Andrew Morton wrote:
> > I was bitten by this recently.  Someone's changelog had
> >
> > ----------- stuff here
> >
> > and git-quiltimport remove half the changelog and all the metadata.
> >
> > git is wrong.  I say so coz I invented the --- convention.
> > submitting-patches.rst says
> >
> >         - A marker line containing simply ``---``.
> >
> > The astonishingly old https://www.ozlabs.org/~akpm/stuff/tpp.txt says
> > "...  scripts will treat a ^--- string as ...".  I regret not
> > explicitly using "^---$".
> >
> > I'd like checkpatch to emit a warning in this case.  If a line starts
> > with --- then please let's warn the user that downstream tooling will
> > screw this up.
>
> Maybe something like this:
> ---
>  scripts/checkpatch.pl | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
> index e0f1d6a6bc636..8dd58710302bc 100755
> --- a/scripts/checkpatch.pl
> +++ b/scripts/checkpatch.pl
> @@ -3035,6 +3035,14 @@ sub process {
>                         }
>                 }
>
> +# Check for invalid patch separator
> +               if ($in_commit_log &&
> +                   $line =~ /^---.+/) {
> +                       ERROR("BAD_PATCH_SEPARATOR",
> +                             "Invalid patch separator - some tools may have problems applying this\n" . $herecurr)
> +               }
> +
> +
>  # Check for patch separator
>                 if ($line =~ /^---$/) {
>                         $has_patch_separator = 1;

Sgtm, it still seems potentially error prone that sign offs are
counted even when not in the commit message. ie this bit of this
patch:
```
 # Check the patch for a signoff:
-               if ($line =~ /^\s*signed-off-by:\s*(.*)/i) {
+               if (!$is_patch && $line =~ /^\s*signed-off-by:\s*(.*)/i) {
                        $signoff++;
                        $in_commit_log = 0;
                        if ($author ne ''  && $authorsignoff != 1) {
```

Thanks,
Ian