Re: [patch] x86_64: fix sched_clock()

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

On Fri, 25 May 2007 10:45:47 +0200 Andi Kleen <andi@firstfloor.org> wrote:

> On Fri, May 25, 2007 at 10:44:26AM +0200, Ingo Molnar wrote:
> > 
> > * Andi Kleen <andi@firstfloor.org> wrote:
> > 
> > > > arch/i386/kernel/sched-clock.c (now built on x86_64 too with the 
> > > > patch i sent) uses the tsc_disable global flag which is non-existent 
> > > > on x86_64. So my tsc.h change adds that global flag, always-defined 
> > > > to 0.
> > > 
> > > My version of sched_clock.c doesn't have any reference to tsc_disable.
> > 
> > must be an -mm fix. I used -mm as a basis of my work. Please apply my 
> > patch.
> 
> I would prefer to find out why the mm patch was added and then hopefully
> remove it.  IMNSHO it should not be needed.
> 

This?    I sent it to you earlier this week:

From: Rusty Russell <rusty@rustcorp.com.au>

If you set tsc_disable (eg "notsc" on cmdline), sched-clock.c gives a
divide by zero on boot.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 arch/i386/kernel/sched-clock.c |    2 +-
 1 files changed, 1 insertion(+), 1 deletion(-)

diff -puN arch/i386/kernel/sched-clock.c~fix-x86_64-mm-sched-clock-share arch/i386/kernel/sched-clock.c
--- a/arch/i386/kernel/sched-clock.c~fix-x86_64-mm-sched-clock-share
+++ a/arch/i386/kernel/sched-clock.c
@@ -115,7 +115,7 @@ unsigned long long printk_clock(void)
 static void resync_sc_freq(struct sc_data *sc, unsigned int newfreq)
 {
 	sc->sync_base = jiffies;
-	if (!cpu_has_tsc) {
+	if (!cpu_has_tsc || tsc_disable) {
 		sc->unstable = 1;
 		return;
 	}
_