* Re: [PATCH] irqchip: gic: call gic_cpu_init() as well in CPU_STARTING_FROZEN case
[not found] ` <20130619070253.GA3405-rvtDTF3kK1ictlrPMvKcciBecyulp+rMXqFh9Ls21Oc@public.gmane.org>
@ 2013-06-20 2:09 ` Joseph Lo
[not found] ` <1371694169.2155.12.camel-yx3yKKdKkHfc7b1ADBJPm0n48jw8i0AO@public.gmane.org>
0 siblings, 1 reply; 2+ messages in thread
From: Joseph Lo @ 2013-06-20 2:09 UTC (permalink / raw)
To: Shawn Guo
Cc: arm-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org, Catalin Marinas,
Thomas Gleixner,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org,
linux-tegra-u79uwXL29TY76Z2rM5mHXA
On Wed, 2013-06-19 at 15:02 +0800, Shawn Guo wrote:
> On Fri, Jun 14, 2013 at 08:33:59PM +0800, Shawn Guo wrote:
> > On Wed, Jun 12, 2013 at 04:17:49PM +0100, Catalin Marinas wrote:
> > > On Wed, Jun 12, 2013 at 12:30:27PM +0100, Shawn Guo wrote:
> > > > diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
> > > > index 1760ceb..19ceaa6 100644
> > > > --- a/drivers/irqchip/irq-gic.c
> > > > +++ b/drivers/irqchip/irq-gic.c
> > > > @@ -705,7 +705,7 @@ static int gic_irq_domain_xlate(struct irq_domain *d,
> > > > static int __cpuinit gic_secondary_init(struct notifier_block *nfb,
> > > > unsigned long action, void *hcpu)
> > > > {
> > > > - if (action == CPU_STARTING)
> > > > + if (action == CPU_STARTING || action == CPU_STARTING_FROZEN)
> > > > gic_cpu_init(&gic_data[0]);
> > > > return NOTIFY_OK;
> > > > }
> > >
> > > Acked-by: Catalin Marinas <catalin.marinas-5wv7dgnIgG8@public.gmane.org>
> >
> > Thomas,
> >
> > Can you please send the fix for 3.10? Or are you fine with that I ask
> > arm-soc folks to send it?
>
> Arnd, Olof,
>
> Can you help send the fix for 3.10?
>
The commit c011470 also break suspend/resume for Tegra platform on all 3.10-rcX.
> Commit c011470 (irqchip: gic: Perform the gic_secondary_init() call via
> CPU notifier) moves gic_secondary_init() that used to be called in
> .smp_secondary_init hook into a notifier call. But it changes the
> system behavior a little bit. Before the commit, gic_cpu_init()
> is called not only when kernel brings up the secondary cores but also
> when system resuming procedure hot-plugs the cores back to kernel.
> While after the commit, the function will not be called in the latter
> case, where the 'action' will not be CPU_STARTING but
> CPU_STARTING_FROZEN. This behavior difference at least causes the
> following suspend/resume regression on imx6q.
>
Shawn's patch fix the suspend/resume regression for Tegra also.
Tested-by: Joseph Lo <josephl-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
Thanks,
Joseph
^ permalink raw reply [flat|nested] 2+ messages in thread