Message-ID: <3B5E055C.388EDB30@cyberoptics.com>
From: Eric Rudd <rudd AT cyberoptics DOT com>
Organization: CyberOptics
X-Mailer: Mozilla 4.72 [en] (Win95; U)
X-Accept-Language: en,pdf
MIME-Version: 1.0
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Cpu speed detection using DJGPP?
References: <458df9bf DOT 0107231326 DOT 14df82e1 AT posting DOT google DOT com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Lines: 22
Date: Tue, 24 Jul 2001 18:31:40 -0500
NNTP-Posting-Host: 38.196.93.9
X-Trace: client 996017503 38.196.93.9 (Tue, 24 Jul 2001 19:31:43 EDT)
NNTP-Posting-Date: Tue, 24 Jul 2001 19:31:43 EDT
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

Daniel Friberg wrote:

> How do I go about doing this?

I am in the process of getting an IP release from my company for some
code to do just that.  When that happens, I'll post the code somehow.

To satisfy your curiosity in the meantime, you measure the time elapsed
between two successive increments of the rawclock() function,  using the
rdtsc instruction to get a timing in processor clocks, and multiply the
rdtsc timing by the rawclock() frequency, which is nominally
13.125E6/(65536.*11.) Hz.  If you are willing to time many rawclock()
increments, you can get the processor clock rate accurate to a small
fraction of a percent.

On non-Intel processors that lack the rdtsc instruction, you have to use
tricks like measuring the timing of a div instruction for various
arguments.  That's much less accurate.

-Eric Rudd
rudd AT cyberoptics DOT com