Date: Wed, 6 Jan 1999 11:40:40 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Samir Barjoud <samir AT mindspring DOT com>
cc: djgpp AT delorie DOT com
Subject: Re: timing in u sec range
In-Reply-To: <MPG.10fcab3b31b1166f989697@news.miami.mindspring.com>
Message-ID: <Pine.SUN.3.91.990106114025.8103S-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com


On Tue, 5 Jan 1999, Samir Barjoud wrote:

> I've written code that lets you get millisecond accurate timing
> under DOS and win9x. If RDTSC is available, it will use it, otherwise it 
> uses the PIT (increases the PIT frequency).

If millisecond resolution is good enough, why can't you use the CMOS
Real-Time clock?  It has an accuracy of 1/1024 second, and should work
equally well on DOS, Windows, and elsewhere, without making
unnecessary intrusion on the system timer tick.

`uclock' wants to supply MICROsecond resolution, so it cannot use the
CMOS clock.