Re: Next step of smp support & fix device suspending

[Pavel Machek <pavel@suse.cz>]

> > This introduces functions for stopping all-but-boot-cpus, which will
> > be needed for smp suspend, and fixes level for calling driver model:
> > there's no D4 power level, only D3 (means device off), and tg3 driver
> > actually cares. Ugh and one useless mdelay killed, and
> > freeze_processes() now BUGS() if its not compiled in. [We can probably
> > just remove it for non-CONFIG_PM case in future]. It is bad idea to
> > pretend success, and nobody should ever call it in !CONFIG_PM case
> > anyway. Please apply,
> 
> Nice, just a couple of questions...

Thanks.

> > -	device_power_down(4);
> > +	device_power_down(3);
> 
> There are defined values in include/linux/device.h. You should be using
> those, instead of the magic constants (even if the magic constants
> actually make sense as the power states :).

Ugh, I have not figured which magic constants should I use. Should
that be SUSPEND_POWER_DOWN?

> >  	PRINTK( "Waiting for DMAs to settle down...\n");
> >  	mdelay(1000);	/* We do not want some readahead with DMA to corrupt our memory, right?
> >  			   Do it with disabled interrupts for best effect. That way, if some
> 
> On a related note, can we kill this piece of code? It's not clear that
> it's necessary. If it is, it begs for a more systematic way of achieving
> the goal.

Ok, killed, I'll propagate it if there are no problems.
								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!