Re: [PATCH tip/core/rcu 2/3] rcu: Add designated reviewers for RCU

[Peter Zijlstra <peterz@infradead.org>]

On Thu, Jul 10, 2014 at 04:16:23PM -0700, josh@joshtriplett.org wrote:
> On Thu, Jul 10, 2014 at 03:00:12AM -0700, Joe Perches wrote:
> > On Thu, 2014-07-10 at 11:39 +0200, Peter Zijlstra wrote:
> > > On Wed, Jul 09, 2014 at 06:23:22AM -0700, Joe Perches wrote:
> > > > > SCHEDULER:
> > > > > ...
> > > > > R:   Steven Rostedt <rostedt@goodmis.org> (kernel/sched/rt.c)
> > > > > R:   Juri Lelli <jundri.lelli@gmail.com>    (kernel/sched/deadline.c)
> > > > 
> > > > Maybe a better syntax might be something like:
> > > > R:	Steven Rostedt
> > > > 	F:	kernel/sched/rt.c
> > > > 
> > > > where optional F:/X: lines override the default
> > > > assumption of all F:/X: from the section.
> > > 
> > > Would RF: make sense? Instead of the indenting.
> > 
> > Maybe.
> > 
> > As a preface:
> > 
> > I doubt the need for associating a subset of the files
> > patterns for a subsystem with a particular reviewer.
> > 
> > If a reviewer is interested enough in a subsystem to
> > volunteer to read patches then that reviewer likely won't
> > be overburdened by getting a few more emailed patches
> > that may be outside a scope of interest.
> 
> I agree.  And if a subset of files needs a separate set of maintainers
> or reviewers, it doesn't seem excessive to split it into a separate
> MAINTAINERS entry.  For instance, if you want kernel/sched/rt.c to have
> an additional set of maintainers/reviewers, just add a MAINTAINERS entry
> for "SCHEDULER - REALTIME" with an appropriate "F:" line.

Ideally I'd want semantic boundaries, but given this all needs to be
robot parsed that's going to be a massive pain.

I guess I'm just not going to use it then. Too much hassle.