Messages in this thread | | | From | "Rafael J. Wysocki" <> | Subject | Re: [PATCH] ledtrig-cpu: use spin_lock to replace mutex lock | Date | Tue, 23 Oct 2012 02:27:08 +0200 |
| |
On Monday 22 of October 2012 16:15:58 Bryan Wu wrote: > On Mon, Oct 22, 2012 at 3:38 PM, Rafael J. Wysocki <rjw@sisk.pl> wrote: > > On Monday 22 of October 2012 11:59:19 Bryan Wu wrote: > >> Hiya, > >> > >> Can I get some Acked or Tested-by from Rafael or Miles before I put it > >> in my linux-leds tree? > > > > Well, I just explained why the current code didn't work. :-) > > > > Anyway, please feel free to add > > > > Acked-by: Rafael J. Wysocki <rjw@sisk.pl> > > > > I was just reminded by Nathan Lynch <ntl@pobox.com> most of > led-trigger API like register/unregister/event are potentially sleep. > They should not be used in atomic context. Any suggestion to fix it?
Well, design ledtrig_cpu_syscore_suspend() so that it can be called from atomic context? You can use the observation that no locking is necessary in that routine to simplify it, perhaps.
Thanks, Rafael
-- I speak only for myself. Rafael J. Wysocki, Intel Open Source Technology Center.
| |