From owner-freebsd-stable Tue Aug 13 13:39:30 2002 Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 8212137B400 for ; Tue, 13 Aug 2002 13:39:28 -0700 (PDT) Received: from frostbytes.com (dsl092-065-149.bos1.dsl.speakeasy.net [66.92.65.149]) by mx1.FreeBSD.org (Postfix) with ESMTP id 81B4E43E65 for ; Tue, 13 Aug 2002 13:39:27 -0700 (PDT) (envelope-from jimf@frostbytes.com) Received: from [10.1.6.20] (cmb2-nip1.atg.com [63.116.205.150]) by frostbytes.com (8.9.3/8.9.2) with ESMTP id QAA01703; Tue, 13 Aug 2002 16:39:15 -0400 Subject: Re: FreeBSD 4.6 rl0 and xl0 watchdog timeout problems (and solution) From: Jim Frost To: Bosko Milekic Cc: freebsd-stable@FreeBSD.ORG In-Reply-To: <20020813162742.B2869@unixdaemons.com> References: <1029102290.9472.188.camel@snowball.frostbytes.com> <20020813162742.B2869@unixdaemons.com> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.3 (1.0.3-6) Date: 13 Aug 2002 16:36:35 -0400 Message-Id: <1029270995.6144.127.camel@icehouse> Mime-Version: 1.0 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 On Tue, 2002-08-13 at 16:27, Bosko Milekic wrote: > > I thought maybe the thing was incorrectly sensing the media; I still run > > 10baseT because it's here and it works and I don't see why I should > > spend money on a new hub. ifconfig said it autoconfigured to > > 10baseT/UTP but just to be sure I forced the config. Same problem. > > What was sharing the card's IRQ? When you have devices sharing IRQs, > it obviously takes longer before the handler gets to run. The > watchdog is getting fired off before the handler gets to run. Was the > interface working at all? I've realized that I did not clearly state the nature of the problem. No other cards exist in the system to share an IRQ with; this card was assigned its own (IRQ12 FWIW). What appears to be special is that the slot that the card in shares a single PCI interrupt assignment with another slot (which was empty). I.e. if I go into the BIOS config it lists a number of things that can be assigned interrupts, all of which are currently set to "auto". The PCI slot assignments are 1/3, 2/6, 4, and 5 (ie only four interrupts can be assigned between six slots). So long as the card is in slot 4 or slot 5 it works; if it's in 1, 2, 3 or 6 it does not. > Sounds good. You know, it's entirely possible that other operating > systems silently ignore the watchdog timeouts and you may just think > that FreeBSD is the problem because it's telling you that maybe you > should think about changing your setup. Could be; certainly that was the case back in the old days with cheap IDE interfaces that didn't deliver interrupts. Personally I don't much care if it's just BSD being more picky, but "watchdog timeout" did not seem to indicate "card is in the wrong PCI slot" to me. jim To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message