Re: [PATCH] PM / sleep / irq: Do not suspend wakeup interrupts

[Rafael J. Wysocki]

On Tuesday, July 15, 2014 02:36:46 PM Alexander Stein wrote:
> On Tuesday 15 July 2014 14:50:28, Rafael J. Wysocki wrote:
> > On Tuesday, July 15, 2014 02:22:25 PM Alexander Stein wrote:
> > > On Thursday 10 July 2014 23:37:54, Rafael J. Wysocki wrote:
> > > > From: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > > > 
> > > > If an IRQ has been configured for wakeup via enable_irq_wake(), the
> > > > driver who has done that must be prepared for receiving interrupts
> > > > after suspend_device_irqs() has returned, so there is no need to
> > > > "suspend" such IRQs.  Moreover, if drivers using enable_irq_wake()
> > > > actually want to receive interrupts after suspend_device_irqs() has
> > > > returned, they need to add IRQF_NO_SUSPEND to the IRQ flags while
> > > > requesting the IRQs, which shouldn't be necessary (it also goes a bit
> > > > too far, as IRQF_NO_SUSPEND causes the IRQ to be ignored by
> > > > suspend_device_irqs() all the time regardless of whether or not it
> > > > has been configured for signaling wakeup).
> > > > 
> > > > For the above reasons, make __disable_irq() ignore IRQ descriptors
> > > > with IRQD_WAKEUP_STATE set when its suspend argument is true which
> > > > effectively causes them to behave like IRQs with IRQF_NO_SUSPEND
> > > > set.
> > > > 
> > > > This also allows IRQs configured for wakeup via enable_irq_wake()
> > > > to work as wakeup interrupts for the "freeze" (suspend-to-idle)
> > > > sleep mode automatically just like for any other sleep states.
> > > > 
> > > > Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > > > ---
> > > >  kernel/irq/manage.c |    3 ++-
> > > >  1 file changed, 2 insertions(+), 1 deletion(-)
> > > > 
> > > > Index: linux-pm/kernel/irq/manage.c
> > > > ===================================================================
> > > > --- linux-pm.orig/kernel/irq/manage.c
> > > > +++ linux-pm/kernel/irq/manage.c
> > > > @@ -385,7 +385,8 @@ setup_affinity(unsigned int irq, struct
> > > >  void __disable_irq(struct irq_desc *desc, unsigned int irq, bool suspend)
> > > >  {
> > > >  	if (suspend) {
> > > > -		if (!desc->action || (desc->action->flags & IRQF_NO_SUSPEND))
> > > > +		if (!desc->action || (desc->action->flags & IRQF_NO_SUSPEND)
> > > > +		    || irqd_has_set(&desc->irq_data, IRQD_WAKEUP_STATE))
> > > >  			return;
> > > >  		desc->istate |= IRQS_SUSPENDED;
> > > >  	}
> > > 
> > > Nice, this fixes my wakeup problem from freeze using gpio-keys. Unfortunately my SPI-attached touchscreen controller cannot be used for wakeup from freeze. Using it to wakeup from mem does work instead. Any ideas what might be wrong in this case?
> > 
> > Not without looking at the code in question.
> > 
> > One guess would be a missing call to enable_irq_wake().
> 
> No, ads7846.c does call 'enable_irq_wake(ts->spi->irq)'. So is this a platform or driver specific problem?

Well, both code paths are identical from the driver's perspective, so I'd bet
on the platform, but that's only a guess at this point.

Rafael

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/