Re: [PATCH 2/2] acpi_pm.c: check for monotonicity

[Andrew Morton <akpm@linux-foundation.org>]

On Mon, 18 Aug 2008 22:18:44 +0200
Andreas Mohr <andi@lisas.de> wrote:

> Hi,
> 
> On Mon, Aug 18, 2008 at 10:11:15PM +0200, Dominik Brodowski wrote:
> > +	if (good != 10) {
> > +		printk(KERN_INFO "PM-Timer had no reasonable result:"
> > +		       " 0x%#llx - aborting.\n", value1);
> > +		return -ENODEV;
> >  	}
> > -	printk(KERN_INFO "PM-Timer had no reasonable result:"
> > -			" 0x%#llx - aborting.\n", value1);
> > -	return -ENODEV;
> 
> Technically spoken this log message could now be considered partially
> outdated... (we're doing 10 evaluations after all, not one with a
> precise end result).
> 
> 
> Seeing a define for those several open-coded 10 loops values would be nice.
> 

Also it's a bit dodgy printing a cycle_t with %llx.  We don't _know_
that cycle_t was implemented with `long long' - if this was always
true, we wouldn't (or shouldn't) have a cycle_t at all.

But it seems that it happens to work for all architectures which
implement acpi.