Re: subtle pm_runtime_put_sync race and sdio functions

[Alan Stern <stern@rowland.harvard.edu>]

On Thu, 23 Dec 2010, Ohad Ben-Cohen wrote:

> On Sun, Dec 19, 2010 at 12:22 PM, Rafael J. Wysocki <rjw@sisk.pl> wrote:
> > That said, I think we may do something different that perhaps will make your
> > life somewhat easier.
> ...
> > So, I think we can add a "runtime only" flag working as described above.
> 
> That will definitely solve the suspend/resume issue we're having.
> 
> But due to /sys/devices/.../power/control, we will still have to
> bypass runtime PM for this particular device.
> 
> We need a way to unconditionally power down the device, and as long as
> runtime PM can be disabled by the user, we can't use it.

I'm still not aware of the details of your situation.  Nevertheless, it
shouldn't be hard to set up a suspend() routine that would do whatever
was needed to power down the device, whether this means calling the 
runtime_suspend() routine directly or something else.  That's basically 
what every other driver does.

By definition, system sleep transitions require bypassing runtime PM.  
There's nothing odd or unusual about it.

Alan Stern