Re: mmotm 2010-06-03-16-36 uploaded

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

On Fri, 4 Jun 2010 09:41:58 +0800 Dave Young <hidave.darkstar@gmail.com> wrote:

> On Fri, Jun 4, 2010 at 7:36 AM,  <akpm@linux-foundation.org> wrote:
> > The mm-of-the-moment snapshot 2010-06-03-16-36 has been uploaded to
> >
> > __ http://userweb.kernel.org/~akpm/mmotm/
> >
> 
> Hi, intel_idle build fails:
> 
> drivers/idle/intel_idle.c: In function ___intel_idle___:
> drivers/idle/intel_idle.c:234: error: too few arguments to function ___trace_power
> _start___
> make[2]: *** [drivers/idle/intel_idle.o] Error 1
> make[1]: *** [drivers/idle] Error 2
> make[1]: *** Waiting for unfinished jobs....
> 

Caused by
x86-cpufreq-make-trace_power_frequency-cpufreq-driver-independent.patch
which changed trace_power_start().

drivers/idle/intel_idle.c wasn't there when Thomas wrote that patch.

this, I guess:

--- a/drivers/idle/intel_idle.c~x86-cpufreq-make-trace_power_frequency-cpufreq-driver-independent-fix
+++ a/drivers/idle/intel_idle.c
@@ -231,7 +231,7 @@ static int intel_idle(struct cpuidle_dev
 
 	stop_critical_timings();
 #ifndef MODULE
-	trace_power_start(POWER_CSTATE, (eax >> 4) + 1);
+	trace_power_start(POWER_CSTATE, (eax >> 4) + 1, cpu);
 #endif
 	if (!need_resched()) {
 
_


it's a bit odd that all trace_power_start() callers just pass in
smp_processor_id().  Why not do it within trace_power_start() itself?