Re: [PATCH] add firmware disk state and clean up

[David Brownell <david-b@pacbell.net>]

On Tuesday 20 March 2007 3:09 pm, Pavel Machek wrote:
> Hi!
> 
> > > +static int at91_pm_valid(suspend_state_t state)
> > > +{
> > > +	return state == PM_SUSPEND_MEM;
> > > +}
> > >  
> > >  static struct pm_ops at91_pm_ops ={
> > >  	.valid		= at91_pm_valid_state,
         ^^^^^

> > >  	.prepare	= at91_pm_prepare,
> > >  	.enter		= at91_pm_enter,
> > > +	.valid		= at91_pm_valid,
         ^^^^^

> > >  };
> > >  
> > >  static int __init at91_pm_init(void)
> > 
> > Clearly bogus.  The STR support isn't upstream yet, true, but
> 
> No... look again. 

No, YOU look again.  If that compiles, it's a GCC bug.


> This tries to disable suspend-to-disk on ARM, so 
> that only suspend-to-RAM is enabled. I'd not call it bogus.
> 
> The series is needed to stop pm from reporting "platform" on ARM;
> "platform" only makes sense on ACPI systems.
> 
> 								Pavel