lkml.org 
[lkml]   [2014]   [Feb]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [RFC PATCH] rcu: move SRCU grace period work to power efficient workqueue
From
Date
On Thu, 2014-02-27 at 15:43 +0100, Frederic Weisbecker wrote: 
> On Mon, Feb 17, 2014 at 06:26:41AM +0100, Mike Galbraith wrote:
> > On Wed, 2014-02-12 at 19:23 +0100, Frederic Weisbecker wrote:
> > > On Mon, Feb 10, 2014 at 10:47:29AM -0800, Paul E. McKenney wrote:
> > > > On Mon, Feb 10, 2014 at 06:08:31PM +0800, Lai Jiangshan wrote:
> > > > > Acked-by: Lai Jiangshan <laijs@cn.fujitsu.com>
> > > >
> > > > Thank you all, queued for 3.15.
> > > >
> > > > We should also have some facility for moving the SRCU workqueues to
> > > > housekeeping/timekeeping kthreads in the NO_HZ_FULL case. Or does
> > > > this patch already have that effect?
> > >
> > > Kevin Hilman and me plan to try to bring a new Kconfig option that could let
> > > us control the unbound workqueues affinity through sysfs.
> >
> > Handing control to the user seemed like a fine thing, so I started
> > making a boot option to enable it. Forcing WQ_SYSFS on at sysfs
> > decision spot doesn't go well, init order matters :) Post init frobbing
> > required if you want to see/frob all unbound.
>
> I'm curious about the details. Is that because some workqueues are registered
> before sysfs is even initialized?

Yeah. I put in a test, and told it if not ready, go get ready and flag
yourself as having BTDT (one registration being plenty), but that only
made a different explosion. Post-init rescan is definitely a better
plan, it can't be worse :)

-Mike



\
 
 \ /
  Last update: 2014-02-27 17:01    [W:0.142 / U:1.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site