X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f
From: RayeR <glaux AT centrum DOT cz>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: jitter between 2 hardware interrupt
Date: Mon, 8 Mar 2010 17:04:26 -0800 (PST)
Organization: http://groups.google.com
Lines: 18
Message-ID: <b2059afb-1817-4113-8832-b365bee69678@z35g2000yqd.googlegroups.com>
References: <201003041907 DOT o24J7K5f029877 AT delorie DOT com> <607b5402-a0d4-4151-bf50-9ff3dbf0cd4b AT i25g2000yqm DOT googlegroups DOT com> 
	<d33c0554-f8eb-4af4-a7b0-024b8cb4e8fa AT 33g2000yqj DOT googlegroups DOT com>
NNTP-Posting-Host: 213.220.252.124
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
X-Trace: posting.google.com 1268096666 29968 127.0.0.1 (9 Mar 2010 01:04:26 GMT)
X-Complaints-To: groups-abuse AT google DOT com
NNTP-Posting-Date: Tue, 9 Mar 2010 01:04:26 +0000 (UTC)
Complaints-To: groups-abuse AT google DOT com
Injection-Info: z35g2000yqd.googlegroups.com; posting-host=213.220.252.124; 
	posting-account=Q0wMHAoAAADjYrghh94FTf6YnbpTqZgp
User-Agent: G2/1.0
X-HTTP-UserAgent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.23) 
	Gecko/20090825 SeaMonkey/1.1.18,gzip(gfe),gzip(gfe)
Bytes: 2203
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

> yes, but I test on 2 or 3 minutes and I have the jitter. I don't think
> that CPU clock change during this testing. But for a long test, I
> should be careful!

The TSC clock change problem may occur in some caces, e.g.:
- you start your program and perform TSC calibration but CPU was in
low power mode before. Then you make high CPU load and it will switch
clock to higher speed during your TSC calibration.
- everytime during the run may occur some condition that changes power
state, e.g. low battery or presence of AC power.
My NTB Compaq EVO 600 with PIII Tualatin mobile runs at 1,2GHz maximum
from AC power and 800MHz from battery. When I connect/disconnect AC
adapter the CPU clock changes.
- there may be some issues on multiprocessor/multicore systems where
every core has its separate TSC. I had managed to desynchronize TSC on
2 cores of my desktop E8400.

So IMHO TSC IS NO LONGER reliable source for timing...