Subject: Re: Poor timekeeping on x86 smp branch?
To: John Franklin <franklin@elfie.org>
From: Rafal Boni <rafal@attbi.com>
List: tech-smp
Date: 04/23/2002 20:59:02
In message <20020423185856.GC7106@deathmitten.example.org>, you write: 

-> On Tue, Apr 23, 2002 at 02:45:22PM -0400, Rafal Boni wrote:
-> > In message <20020423182102.C51A47A1@ab.use.net>, you write: 
-> > -> | 	I figured I'd send a heads-up if this wasn't a known problem; i
-> n
-> > -> | 	case it *isn't* I can do some poking around, but unfortunately
-> > -> | 	have too many other things on my plate to do any thorough diggi
-> ng.
-> > -> 
-> > -> How horrible is horrible?
-> > 
-> > *Really* horrible; lost 45 minutes over the course of 9 hours and ntpd
-> > seemed to be primarily responsible for that badness... The clock in that
-> > box is nowhere near as bad by itself.
-> 
-> I've seen this before.  On your dmesg output, look for lines like:
-> 
-> cpu0: calibrating local timer
-> cpu0: apic clock running at 66 MHz
-> 
-> Is the apic running at a reasonable rate?

Nope.  Thanks for finding this... Here's what the apic said for that boot:

    cpu0: apic clock running at 160 MHz

Every other boot has/should have:

    cpu0: apic clock running at 100 MHz

--rafal

----
Rafal Boni                                                     rafal@attbi.com
  We are all worms.  But I do believe I am a glowworm.  -- Winston Churchill