From owner-freebsd-current  Wed Nov  1 07:38:24 1995
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id HAA18772
          for current-outgoing; Wed, 1 Nov 1995 07:38:24 -0800
Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id HAA18765
          for <current@FreeBSD.org>; Wed, 1 Nov 1995 07:38:10 -0800
Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.6.12/8.6.9) id RAA19641; Wed, 1 Nov 1995 17:33:13 +0200
From: John Hay <jhay@mikom.csir.co.za>
Message-Id: <199511011533.RAA19641@zibbi.mikom.csir.co.za>
Subject: Re: Time problems
To: bde@zeta.org.au (Bruce Evans)
Date: Wed, 1 Nov 1995 17:33:13 +0200 (SAT)
Cc: wollman@lcs.mit.edu, current@FreeBSD.org
In-Reply-To: <199511010903.UAA24620@godzilla.zeta.org.au> from "Bruce Evans" at Nov 1, 95 08:03:05 pm
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1824      
Sender: owner-current@FreeBSD.org
Precedence: bulk

> 
> >OK, I changed the values in calibrate_cyclecounter() like suggested and now
> >my time stay much closer to reality.
> 
> >My CPU even probe as a 90MHz processor!
> >> CPU: 90-MHz Pentium 735\\90 (Pentium-class CPU)
> >>   Origin = "GenuineIntel"  Id = 0x525  Stepping=5
> >>   Features=0x3bf<FPU,VME,PSE,MCE,CX8,APIC>
> 
> >Can we get some kernel config option to do this or will I have to change
> >clock.c after every ctm? Or will this get to be the default?
> 
> No.  Reducing the calibration time should just decrease the accuracy of
> the calibration (and speed up the boot :-).  Find out which clock is
> inaccurate or which part of the calibration routine is buggy.  Define
> DELAYDEBUG to get some debugging code in clock.c.  It currently
> generates delays of powers of 10 usec up to 10 seconds.  You could
> increase the delats to 100, 1000, ... seconds and time them with a
> stopwatch to see how accurate DELAY() is.
> 
> Bruce
> 
OK, I did that and the 10 second delay takes about 8.5 seconds and the 100
second delay takes about 88 seconds. Here is the first part of dmesg:

CPU: DELAY(1)... 1 calls to getit() at -14 usec each
DELAY(10)... 1 calls to getit() at -5 usec each
DELAY(100)... 22 calls to getit() at 3 usec each
DELAY(1000)... 261 calls to getit() at 3 usec each
DELAY(10000)... 2648 calls to getit() at 3 usec each
DELAY(100000)... 26511 calls to getit() at 3 usec each
DELAY(1000000)... 238632 calls to getit() at 4 usec each
DELAY(10000000)... 2351886 calls to getit() at 4 usec each
DELAY(100000000)... 23667381 calls to getit() at 4 usec each
892492-MHz Pentium 735\\90 or 815\\100 (Pentium-class CPU)

>From about 100000 the values that is being printed vary quite a lot between
boots.

Now how do I go further to find out where my problem is?

John
-- 
John Hay -- John.Hay@csir.co.za