Re: [RFC][PATCH 2/3] PM / sleep: Mechanism to avoid resuming runtime-suspended devices unnecessarily

[Rafael J. Wysocki]

On Wednesday, May 14, 2014 10:53:16 AM Alan Stern wrote:
> On Tue, 13 May 2014, Rafael J. Wysocki wrote:
> 
> > > It would be surprising if ->prepare() needed to make any difficult
> > > checks.  This would imply that the device could have multiple
> > > runtime-suspend states, some of which are appropriate for system
> > > suspend while others aren't.  Not impossible, but I wouldn't expect it
> > > to come up often.
> > 
> > That is the case for every device with ACPI power management in principle. :-)
> > 
> > Please see patch [3/3] for details.
> 
> I don't understand enough about the ACPI subsystem to follow the
> details of that patch.
> 
> > OK, I've updated the $subject patch in the meantime and the result is appended
> > Former patch [1/3] is not necessary any more now and patch [3/3] is still valid.
> > 
> > Rafael
> > 
> > ---
> > From: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> > Subject: PM / sleep: Mechanism to avoid resuming runtime-suspended devices unnecessarily
> > 
> > Currently, some subsystems (e.g. PCI and the ACPI PM domain) have to
> > resume all runtime-suspended devices during system suspend, mostly
> > because those devices may need to be reprogrammed due to different
> > wakeup settings for system sleep and for runtime PM.
> 
> ...
> 
> > Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
> 
> This is looking quite good.  I have one suggestion for a small 
> improvement...
> 
> > @@ -1332,6 +1338,16 @@ static int __device_suspend(struct devic
> >  	if (dev->power.syscore)
> >  		goto Complete;
> >  
> > +	if (dev->power.direct_complete) {
> > +		pm_runtime_disable(dev);
> > +		if (dev->power.disable_depth == 1
> > +		    && pm_runtime_status_suspended(dev))
> > +			goto Complete;
> > +
> > +		dev->power.direct_complete = false;
> > +		pm_runtime_enable(dev);
> > +	}
> 
> Do we want to allow ->prepare() to return > 0 if the device isn't
> runtime suspended?  If we do then non-suspended devices may be a common
> case.  We should then avoid the extra overhead of disable + enable.  
> So I would write:
> 
> 	if (dev->power.direct_complete) {
> 		if (pm_runtime_status_suspended(dev)) {
> 			pm_runtime_disable(dev);
> 			if (dev->power.disable_depth == 1
> 			    && pm_runtime_status_suspended(dev))
> 				goto Complete;
> 			pm_runtime_enable(dev);
> 		}
> 		dev->power.direct_complete = false;
> 	}

That is a good idea, thanks!

> Also, now that we have finally settled on the appropriate API, there
> needs to ba a patch updating the PM documentation.

Absolutely.  I thought about updating the documentation in the same patch
(at least the comments in pm.h), but I guess a separate patch for files
under Documentation/ may be better.

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/