Re: [bug] get_maintainer.pl incomplete output

[Joe Perches]

On Wed, 2013-10-16 at 21:19 -0700, David Rientjes wrote:
> On Wed, 16 Oct 2013, Joe Perches wrote:
> 
> > > I haven't looked closely at scripts/get_maintainer.pl, but I recently 
> > > wrote a patch touching mm/vmpressure.c and it doesn't list the file's 
> > > author, Anton Vorontsov <anton.vorontsov@xxxxxxxxxx>.
> > > 
> > > Even when I do scripts/get_maintainer.pl -f mm/vmpressure.c, his entry is 
> > > missing and git blame attributs >90% of the lines to his authorship.
> > > 
> > > $ ./scripts/get_maintainer.pl -f mm/vmpressure.c 
> > > Tejun Heo <tj@xxxxxxxxxx> (commit_signer:6/7=86%)
> > > Michal Hocko <mhocko@xxxxxxx> (commit_signer:5/7=71%)
> > > Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> (commit_signer:4/7=57%)
> > > Li Zefan <lizefan@xxxxxxxxxx> (commit_signer:3/7=43%)
> > > "Kirill A. Shutemov" <kirill@xxxxxxxxxxxxx> (commit_signer:1/7=14%)
> > > linux-mm@xxxxxxxxx (open list:MEMORY MANAGEMENT)
> > > linux-kernel@xxxxxxxxxxxxxxx (open list)
> > > 
> > > Any ideas?
> > 
> > get_maintainer has a lot of options.
> > 
> > get_maintainer tries to find people that are either
> > listed in the MAINTAINERS file or that have recently
> > (in the last year by default) worked on the file.
> > 
> > If you want to find all authors, use the --git-blame option
> > 
> > It's not the default because it can take quite awhile to run.
> > 
> 
> Hmm, it's a little strange to only consider recent activity when >90% of 
> the lines were written by someone not listed.  Isn't there any faster way 
> to determine that besides using the expensive git blame?

Not so far as I know.

> Something like 
> weighing the output of "git show --shortstat" for all commits in "git log 
> mm/vmpressure.c" to determine the most important recent changes?  That 
> should be fairly cheap.

Important can be hard to determine.

git blame effectively does a --follow

get_maintainers already does a git log <file>
and accumulates all the signatures for the
time selected by --git-since

It doesn't weight each commit by +/- count.

I think the most significant negative to
the current get_maintainer is that only the
"signature" lines are considered.

The "Author:" line isn't.

The default ris for get_maintainer to only list
a maximum of 5 "maintainers" ordered by signature
count.

Anton has only signed/acked 1.


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/