lkml.org 
[lkml]   [2011]   [Sep]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 1/4] nohz: Split extended quiescent state handling from nohz switch
On Tue, Sep 13, 2011 at 02:06:45AM +0200, Frederic Weisbecker wrote:
> On Wed, Sep 07, 2011 at 07:22:33AM -0700, Paul E. McKenney wrote:
> > On Sun, Sep 04, 2011 at 04:36:43PM -0700, Paul E. McKenney wrote:
> > > On Sat, Aug 20, 2011 at 07:30:49PM +0200, Frederic Weisbecker wrote:
> > > > It is assumed that rcu won't be used once we switch to tickless
> > > > mode and until we restart the tick. However this is not always
> > > > true, as in x86-64 where we dereference the idle notifiers after
> > > > the tick is stopped.
> > > >
> > > > To prepare for fixing this, split the tickless mode switching and
> > > > RCU extended quiescent state logics.
> > > > Make tick_nohz_stop/restart_sched_tick() RCU agnostic but provide
> > > > a new pair of APIs tick_nohz_enter/exit_idle() that keep the
> > > > old behaviour by handling both the nohz mode and RCU extended
> > > > quiescent states, then convert every archs to use these.
> > > >
> > > > Archs that want to switch to RCU extended QS to some custom points
> > > > can do it later by changing the parameter in tick_nohz_enter,exit_idle()
> > > > to false and call rcu_enter,exit() separately.
> > >
> > > This approach looks quite good to me! A few comments below.
> >
> > But I get RCU stall warnings when running it on powerpc on top of
> > the patch set at:
> >
> > https://lkml.org/lkml/2011/9/7/64
> >
> > At first glance, it appears that CPUs are entering dyntick-idle
> > mode without RCU being informed. Any thoughts on diagnostics?
> >
> > Thanx, Paul
>
> I've just tested with your rcu/next branch from github and applied
> my patches on top of it but did not experience a stall in x86.

Thank you for checking this!

> That may be related to powerpc more precisely, or may be any arch
> but on some particular condition.

OK, I guess my next step is to put a WARN_ON_ONCE() in the idle loop
to check to see if there might be some other idle loop buried in
PowerPC somewhere...

Thanx, Paul


\
 
 \ /
  Last update: 2011-09-13 18:51    [W:0.282 / U:0.044 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site