Re: [PATCH v6 00/16] Infrastructure to allow fixing exec deadlocks

[Kees Cook <keescook@chromium.org>]

On Mon, Mar 30, 2020 at 01:14:59PM -0700, Matthew Wilcox wrote:
> On Mon, Mar 30, 2020 at 10:12:02PM +0200, Bernd Edlinger wrote:
> > On 3/29/20 5:44 AM, Kees Cook wrote:
> > > On Sat, Mar 28, 2020 at 11:32:35PM +0100, Bernd Edlinger wrote:
> > >> Oh, do I understand you right, that I can add a From: in the
> > >> *body* of the mail, and then the From: in the MIME header part
> > >> which I cannot change is ignored, so I can make you the author?
> > > 
> > > Correct. (If you use "git send-email" it'll do this automatically.)
> > > 
> > > e.g., trimmed from my workflow:
> > > 
> > > git format-patch -n --to "$to" --cover-letter -o outgoing/ \
> > > 	--subject-prefix "PATCH v$version" "$SHA"
> > > edit outgoing/0000-*
> > > git send-email --transfer-encoding=8bit --8bit-encoding=UTF-8 \
> > > 	--from="$ME" --to="$to" --cc="$ME" --cc="...more..." outgoing/*
> > > 
> > > 
> > 
> > Okay, thanks, I see that is very helpful information for me, and in
> > this case I had also fixed a small bug in one of Eric's patches, which
> > was initially overlooked (aquiring mutexes in wrong order,
> > releasing an unlocked mutex in some error paths).
> > I am completely unexperienced, and something that complex was not
> > expected to happen :-) so this is just to make sure I can handle it
> > correctly if something like this happens again.
> > 
> > In the case of PATCH v6 05/16 I removed the Reviewd-by: Bernd Edlinger
> > since it is now somehow two authors and reviewing own code is obviously
> > not ok, instead I added a Signed-off-by: Bernd Edlinger (and posted the
> > whole series on Eric's behalf (after asking Eric's permissing per off-list
> > e-mail, which probably ended in his spam folder)
> > 
> > Is this having two Signed-off-by: for mutliple authors the
> > correct way to handle a shared authorship?
> 
> If the patch comes through you, then Reviewed-by: is inappropriate.
> Instead, you should use Signed-off-by: in the second sense of
> Documentation/process/submitting-patches.rst
> 
> This also documents how to handle "minor changes" that you make.

And in the true case of multiple authors, have both SoBs, but also add a
Co-developed-by: for the non-"git author" author. Specific details:
https://www.kernel.org/doc/html/latest/process/submitting-patches.html#when-to-use-acked-by-cc-and-co-developed-by

-- 
Kees Cook