Re: [PATCH v3] Documentation/process: Add text to indicate supporters should be mailed

[Joe Perches <joe@perches.com>]

On Mon, 2022-10-03 at 10:17 +0900, Akira Yokosawa wrote:
> On Sun, 02 Oct 2022 16:55:05 -0700, Joe Perches wrote:
> > On Mon, 2022-10-03 at 08:04 +0900, Akira Yokosawa wrote:
> > > Hello Joe,
> > > 
> > > Thank you for chiming in.
> > > 
> > > On 2022/10/03 0:49, Joe Perches wrote:
> > > > On Sun, 2022-10-02 at 09:58 +0200, Krzysztof Kozlowski wrote:
> > > > > The easiest to achieve it is to run with --no-git-fallback and CC entire
> > > > > output. However it does not mean submitter must run with
> > > > > --no-git-fallback. It is only for this generic rule - CC entire output
> > > > > of get_maintainers.pl.
> > > > > 
> > > > > If you add such rule "CC entire output of get_maintainers.pl" and do not
> > > > > mention no-git-fallback, some folks will think they need to CC all these
> > > > > people who made one commit to your file...
> > > > 
> > > > false.
> > > > 
> > > > git-fallback is _not_ used when there is a listed maintainer for a
> > > > specific file.
> > > > 
> > > > If there is a use of git-fallback, it's because there is _no_
> > > > specified maintainer for a specific file.
> > > > 
> > > >    --git-fallback => use git when no exact MAINTAINERS pattern (default: 1)
> > > > 
> > > > i.e.: It's not "your file" if you don't maintain it.
> > > 
> > > Joe, I sometimes see unexpected output WRT --git-fallback.
> > > 
> > > Example:
> > > 
> > >     $ ./get_maintainer.pl -f Documentation/doc-guide/sphinx.rst
> > >     Jonathan Corbet <corbet@lwn.net> (maintainer:DOCUMENTATION,commit_signer:1/1=100%)
> > >   <-- ???
> > >     Lukas Bulwahn <lukas.bulwahn@gmail.com> (commit_signer:1/1=100%,authored:1/1=100%,added_lines:2/2=100%,removed_lines:2/2=100%)
> > >     <-- ???
> > >     linux-doc@vger.kernel.org (open list:DOCUMENTATION)
> > > 
> > >     linux-kernel@vger.kernel.org (open list)
> > > 
> > > As you see, --git-fallback is used in this case. Why?
> > > It looks strange to me as Jon is listed as a "maintainer".
> > > 
> > > Having "F:	Documentation/" in MAINTAINERS does not suffice?
> > 
> > No.  It's not an exact pattern match as the files below the
> > top level of Documentation are not specifically matched by
> > "F:	Documentation/".
> For me, calling this is "not an exact pattern match" sounds
> inconsistent with the explanation (quoted below) near the top of
> MAINTAINERS:
> 
> 	F: *Files* and directories wildcard patterns.
> 	   A trailing slash includes all files and subdirectory files.
> What am I missing?
> Does this explanation needs update?

Maybe.  Suggest some text.

Read about the pattern-depth entries (basically, it's the
count of forward slashes '/' in a maintained file pattern)

Look for MAINTAINER entries where there are <foo>/*/ entries too.

For instance:

MAINTAINERS-INTEL ETHERNET DRIVERS
MAINTAINERS-M:  Jesse Brandeburg <jesse.brandeburg@intel.com>
MAINTAINERS-M:  Tony Nguyen <anthony.l.nguyen@intel.com>
MAINTAINERS-L:  intel-wired-lan@lists.osuosl.org (moderated for non-subscribers)
MAINTAINERS-S:  Supported
MAINTAINERS-W:  http://www.intel.com/support/feedback.htm
MAINTAINERS-W:  http://e1000.sourceforge.net/
MAINTAINERS-Q:  http://patchwork.ozlabs.org/project/intel-wired-lan/list/
MAINTAINERS-T:  git git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue.git
MAINTAINERS-T:  git git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue.git
MAINTAINERS-F:  Documentation/networking/device_drivers/ethernet/intel/
MAINTAINERS-F:  drivers/net/ethernet/intel/
MAINTAINERS:F:  drivers/net/ethernet/intel/*/     <<< Here >>>
MAINTAINERS-F:  include/linux/avf/virtchnl.h
MAINTAINERS-F:  include/linux/net/intel/iidc.h

So this entry is show that all of drivers/net/ethernet/intel/<foo>/
are directly maintained.