Re: [PATCH] only irq-safe atomic ops

On Sat, Feb 23, 2002 at 02:10:25PM -0800, Andrew Morton wrote:
> Roman Zippel wrote:
> > 
> > Hi,
> > 
> > [e-mail] wrote:
> > 
> > > Right. Without preemption it is safe to do
> > >         c = smp_get_cpuid();
> > >        ...
> > >         x = ++local_cache[c]
> > >        ..
> > >
> > >        y = ++different_local_cache[c];
> > >       ..
> > 
> > Just add:
> >         smp_put_cpuid();
> > 
> > Is that so much worse?
> > 
> 
> ooh.  me likee.

Cool. 
Me likee code with unmatched smp_get_cpuid/smp_put_cpuid.
Much nicer to write
	x = ++local_cache[smp_getcpuid()];
	smp_put_cuid();
than boring old
	x = ++ local_cache[c];

Is this part of some scheme to make the GPL support model actually
pay?

	c = smp_get_cpuid(); // disables preemption

	...
	f(); // oops, me forgotee, this function also references cpuid
	..
	x = ++local_cache[c]; // live dangerously
	smp_put_cpuid(); // G_d knows what that does now.

Oh, wait, I know - reference counts for get_cpuid! How hard can that
be? See how simple it is?  One simple step at a time.

-- 
---------------------------------------------------------
Victor Yodaiken 
Finite State Machine Labs: The RTLinux Company.
 www.fsmlabs.com  www.rtlinux.com