Re: [RFC][PATCH 1/10] PM: Add flag for devices capable of generatingrun-time wake-up events

[Alan Stern]

On Mon, 16 Nov 2009, Rafael J. Wysocki wrote:

> On Monday 16 November 2009, Alan Stern wrote:
> > On Mon, 16 Nov 2009, Rafael J. Wysocki wrote:
> > 
> > > From: Rafael J. Wysocki <rjw@xxxxxxx>
> > > 
> > > Apparently, there are devices that can wake up the system from sleep
> > > states and yet are incapable of generating wake-up events at run
> > > time.  Thus, introduce a flag indicating if given device is capable
> > > of generating run-time wake-up events.
> > 
> > This raises the question: Who is responsible for setting the new 
> > flag?  The code that registers the device?
> 
> Yes, in general.  The platform.

And for non-platform devices (hot-pluggable, for example)?  Presumably 
you would want the driver that detects and registers the device to set 
this flag.

> Actually, I needed it for PCI, but I thought it would be better to put it at
> the core level.
> 
> > What if the kernel can't tell whether or not the device can generate 
> > runtime wake-up events?
> 
> Do you have any specific examples in mind?

What about Matthew's example of an ACPI GPE which might or might not 
cause a runtime wake-up event, depending on the AML code in the BIOS?

> > What if the user wants to override the kernel's setting?  Should there 
> > be a sysfs attribute controlling the flag?
> 
> I have no plans for adding anything like that.

So if the kernel makes a mistake here, the user won't be able to 
correct it.

Alan Stern

--
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/