From owner-freebsd-stable Wed Jun 13 1:14: 9 2001 Delivered-To: freebsd-stable@freebsd.org Received: from mass.dis.org (mass.dis.org [216.240.45.41]) by hub.freebsd.org (Postfix) with ESMTP id 5F3F037B405 for ; Wed, 13 Jun 2001 01:13:58 -0700 (PDT) (envelope-from msmith@mass.dis.org) Received: from mass.dis.org (localhost [127.0.0.1]) by mass.dis.org (8.11.3/8.11.3) with ESMTP id f5D8OSn08506; Wed, 13 Jun 2001 01:24:29 -0700 (PDT) (envelope-from msmith@mass.dis.org) Message-Id: <200106130824.f5D8OSn08506@mass.dis.org> X-Mailer: exmh version 2.1.1 10/15/1999 To: Pete French Cc: stable@FreeBSD.ORG Subject: Re: Very odd clock problem In-reply-to: Your message of "Mon, 11 Jun 2001 10:47:54 BST." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 13 Jun 2001 01:24:28 -0700 From: Mike Smith Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG > > Er, the system clock is not capable of representing the year 1933 (in any > > valid fashion). > > Well it surpised me too ! Can the system clock not take negative values > like time_t then ? (whatever type time_t happens to be this week :-) ) > I was assuming it was a signed 32 bit number and was getting set to > something like 0xBB000000 or thereabouts. Yes, it can *take* a negative value, but because it's typically manipulated through arithemetic with other time_t's, it'd be fairly hard for it to suddenly become large and negative by any legitimate means. > > This sounds like either memory corruption or an in-kernel sniper bug of > > some sort. You don't help the diagnosis any by saying "losing time" and > > then complaining about a "jump". The two are very different things, and > > you need to be much more specific about what is actually happening. > > Fair comment. > > What happens specificly is this: I get a user (and its always the same user) > who complains that the dates on his outgoing emails are being timestamped in > 1933. I login, type 'date' and sure enough it says it is a date in > 1933. I then reset the date. There is nothing in the log files to indicate > what has happened - other than the date on other messages jumping back > to sometime in May (the year isnt logged in /var/log/messages I assume, > but the month change is very obvious and I am assuming this co-incides > with the year date change). Why is it only this user? Are they your only user that cares? Or does the time step only happen when they are sending mail? > We were running ntpd, and stopped in case it was a corrupt ntp server > somewhere. That didnt help. We have also changed ntp servers to a different > set of machines. Once again to no effect. Could this be a clock > battery problem at all ? I wouldnt have thought the battery was used > with the machine powered up, but it is the only thing > I can think of off the top of my head. It happened again this morning, > there is no pattern to when the jumps occur that I can see. It's unlikely to be the clock battery; the clock is only read at startup; the system maintains time itself after that (keeping the clock in sync). I'm really at a loss here; as I said, it sounds like a very specific kernel sniper bug of some sort. 8( -- ... every activity meets with opposition, everyone who acts has his rivals and unfortunately opponents also. But not because people want to be opponents, rather because the tasks and relationships force people to take different points of view. [Dr. Fritz Todt] V I C T O R Y N O T V E N G E A N C E To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message