From owner-freebsd-current Tue Nov 21 12: 8:17 2000 Delivered-To: freebsd-current@freebsd.org Received: from granger.mail.mindspring.net (granger.mail.mindspring.net [207.69.200.148]) by hub.freebsd.org (Postfix) with ESMTP id E376737B4D7 for ; Tue, 21 Nov 2000 12:08:13 -0800 (PST) Received: from ix.netcom.com (sil-wa15-43.ix.netcom.com [207.93.148.43]) by granger.mail.mindspring.net (8.9.3/8.8.5) with ESMTP id PAA14814 for ; Tue, 21 Nov 2000 15:08:08 -0500 (EST) Received: (from tomdean@localhost) by ix.netcom.com (8.11.1/8.9.3) id eALK83E00524; Tue, 21 Nov 2000 12:08:03 -0800 (PST) (envelope-from tomdean@ix.netcom.com) Date: Tue, 21 Nov 2000 12:08:03 -0800 (PST) Message-Id: <200011212008.eALK83E00524@ix.netcom.com> From: "Thomas D. Dean" To: current@FreeBSD.ORG In-reply-to: <200011200212.eAK2Cj500371@ix.netcom.com> (tomdean@ix.netcom.com) Subject: Re: Clock Strangeness in UP Kernel References: <200011200212.eAK2Cj500371@ix.netcom.com> Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG I have a high clock drift rate. The problem is in selecting the timecounter, at least on my machine. Both the TSC and the i8254 timecounters are checked and, since, I believe, TSC is last, TSC is the timecounter the kernel uses. TSC is a horrible timer, at least on my machine. i8254 is not perfect, but several orders of magnitude better, at least with FreeBSD. # sysctl -a ... kern.timecounter.hardware: TSC ... # sysctl -w kern.timecounter.hardware=i8264 fixes the clock drift. Now, it is less than 2 seconds per 4 hours. And, that is well within the range ntp can satisfactorily correct for. tomdean To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message