Re: libata/sata_nv latency on NVIDIA CK804 [was Re: AMD64 X2 lost ticks on PM timer]

["Bill Rugolsky Jr." <brugolsky@telemetry-investments.com>]

On Wed, Mar 22, 2006 at 05:09:08PM +0100, Andi Kleen wrote:
> perfctr0 is already programmed. You can just use rdpmc on it
> 
> Also my latest patchkit has a debugging patch for lost tries
> 
> ftp://ftp.firstfloor.org/pub/ak/x86_64/quilt/patches/lost-cli-debug

Excellent.

> Can you test it with this patch? 
 
I'll give it a try this evening; the test box is an
FC5 upgrade guinea pig at the moment, and only half-migrated.

> I'm still not quite convinced you're barking at the right tree
> with these latency traces because it doesn't match the symptoms.

Yes, I agree that that the latency issue is not due to the inb(),
as I get lost ticks almost continuously at 1 KHZ, and the inb()-related
latency occurs with much lower frequency.

However, a 17ms inb() latency is rather dismal, and I need to
address that hardware issue, perhaps by purchasing a 4-port card.

I have not been chasing the other issue because the test box appears to
keep perfect time with the -rt patch (on SMP) and NTP doesn't complain at
all; I haven't yet had the chance to determine whether that is entirely
due to John Stultz's new code, or if the preemption is responsible.

	-Bill

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