Re: 2.6.19-rc2-mm1

From: Andrew Morton
Date: Thu Oct 19 2006 - 12:18:41 EST


On 19 Oct 2006 14:32:11 +0200
Andi Kleen <ak@xxxxxxx> wrote:

> Andrew Morton <akpm@xxxxxxxx> writes:
>
> > On Wed, 18 Oct 2006 16:01:05 -0700
> > Badari Pulavarty <pbadari@xxxxxxxxxx> wrote:
> >
> > > > Is the NMI watchdog ticking over?
> > >
> > > I think so.
> > >
> > > # dmesg | grep NMI
> > > ACPI: LAPIC_NMI (acpi_id[0x00] high edge lint[0x1])
> > > ACPI: LAPIC_NMI (acpi_id[0x01] high edge lint[0x1])
> > > ACPI: LAPIC_NMI (acpi_id[0x02] high edge lint[0x1])
> > > ACPI: LAPIC_NMI (acpi_id[0x03] high edge lint[0x1])
> > > testing NMI watchdog ... OK.
> >
> >
> > What does it say in /proc/interrupts?
> >
> > The x86_64 nmi watchdog handling looks rather complex.
> >
> > <checks a couple of x86-64 machines>
> >
> > The /proc/interrutps NMI count seems to be going up by about
> > one-per-minute. How odd. Maybe you just need to wait longer.
>
> That's consistent with a idle machine. The perfctr used by the nmi
> watchdog only increases when the CPU isn't halted and when it's idle
> it's not doing very much. When something actually loops it should
> increase much faster though.
>

ok..

What is the reason for not using an apic/lapic NMI source for the watchdog?
-
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/