BOUNCE rtl@rtlinux.org: Approval required: Non-member submissionfrom ["gzeljko" <gzeljko@sezampro.yu>] (fwd)

[Der Herr Hofrat <der.herr@hofr.at>]

>From owner-rtl Sun May 20 12:24:50 2001
Received: from mail1.sezampro.yu (mail1.sezampro.yu [194.106.188.3])
	by hq.fsmlabs.com (8.11.2/8.11.2) with SMTP id f4KIOnr13267
	for <rtl@fsmlabs.com>; Sun, 20 May 2001 12:24:50 -0600
Received: from gzeljko@b103.ppp-bg.sezampro.yu by SezamPro with SMTP on Sun, 20 May 2001 20:20:00 +0200
Message-ID: <002d01c0e159$69685cc0$67bd6ac2@zko>
From: "gzeljko" <gzeljko@sezampro.yu>
To: <rtl@fsmlabs.com>
References: <OF92BE28D3.68C1072A-ON88256A45.00780215@aero.org> <3AFA954A.D4BE75E0@opersys.com> <001401c0e08b$070e5670$aebd6ac2@zko>
Subject: Re: [rtl] do_gettimeofday().
Date: Sun, 20 May 2001 20:17:38 +0200
MIME-Version: 1.0
Content-Type: text/plain;
	charset="ISO-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook Express 5.00.2314.1300
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300

From: gzeljko <gzeljko@sezampro.yu>
> I'm just testing RTLinux (and hardware).
> Motivated with this thread I write small test program  (appended)
> for uniprocessor pentium machine to know what we are talking about.
> 
> On my test machine (K6/200) it givs stable (2-3 ppm) drift cca. 160 ppm.
> 

Exploreing farther I found it's easy solvable, by patching RTLinux to make 
scaler_pentium_to_hrtime public and writeing own calibration routine.
It's works nice, drift is 0 (as it's depend of software).

Now, I'm wondering why scaler_pentium_to_hrtime is private? Probably
there is reson. (given that RTLinux is really nice work).

But, time calibration is definite trade-off: spent time vs accuracy, so it's
worth IMHO to allow customisation (calibration module?).

gzeljko