Re: 2.6.19-rc5-mm1

From: Benoit Boissinot
Date: Fri Nov 10 2006 - 15:20:47 EST


On 11/10/06, Benoit Boissinot <bboissin@xxxxxxxxx> wrote:
On 11/10/06, Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:
> On Fri, 2006-11-10 at 17:29 +0100, Benoit Boissinot wrote:
> > On 11/8/06, Andrew Morton <akpm@xxxxxxxx> wrote:
> > > [snip]
> > > - The hrtimer+dynticks code still doesn't work right for machines which halt
> > > their TSC in low-power states.
> > >
> >
> > With CONFIG_NO_HZ=y, xmoto (xmoto.sf.net, a 3d game) is sluggish, the
> > movement is not fluid (it is "bursty").
> >
> > .config is at http://perso.ens-lyon.fr/benoit.boissinot/kernel/config-2.6.19-rc5-mm1
> > lspci -vv: http://perso.ens-lyon.fr/benoit.boissinot/kernel/docked_lspci
> > dmesg: http://perso.ens-lyon.fr/benoit.boissinot/kernel/dmesg-2.6.19-rc5-mm1
>
> I'm confused about that one:
>
> [ 8.966364] Disabling NO_HZ and high resolution timers due to timer broadcasting (C3 stops local apic)
>
> This message is nowhere in rc5-mm1. It was in rc4-mmX, but got removed
> in the updates.
>
I forget to mention I reverted the following patches from -mm:
i386-apic-timer-use-clockevents-broadcast.patch
acpi-verify-lapic-timer.patch
acpi-verify-lapic-timer-exports.patch
acpi-verify-lapic-timer-fix.patch

since it did not boot with them.

> > I can test any patch or provide any needed information.
>
> http://tglx.de/private/tglx/2.6.19-rc5-mm1-dyntick.diff
>
> That's the rework I did yesterday.
>

I'll undo the reverts I did and try it on top of -mm

It works fine with the following additional patch.

Thanks,

Benoit

Index: linux-mm/arch/i386/kernel/apic.c
===================================================================
--- a/arch/i386/kernel/apic.c 2006-11-10 20:42:30.000000000 +0100
+++ b/arch/i386/kernel/apic.c 2006-11-10 20:42:41.000000000 +0100
@@ -610,6 +610,7 @@
if (evt->event_handler)
clockevents_set_broadcast(evt, broadcast);
}
+EXPORT_SYMBOL_GPL(lapic_timer_idle_broadcast);

int setup_profiling_timer(unsigned int multiplier)
{
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/