From owner-freebsd-current Wed Jan 29 14:16:47 2003 Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C7C1237B401 for ; Wed, 29 Jan 2003 14:16:45 -0800 (PST) Received: from heron.mail.pas.earthlink.net (heron.mail.pas.earthlink.net [207.217.120.189]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5FFD243F43 for ; Wed, 29 Jan 2003 14:16:45 -0800 (PST) (envelope-from tlambert2@mindspring.com) Received: from pool0131.cvx22-bradley.dialup.earthlink.net ([209.179.198.131] helo=mindspring.com) by heron.mail.pas.earthlink.net with asmtp (SSLv3:RC4-MD5:128) (Exim 3.33 #1) id 18e0V4-0000DV-00; Wed, 29 Jan 2003 14:16:07 -0800 Message-ID: <3E385236.29718458@mindspring.com> Date: Wed, 29 Jan 2003 14:14:14 -0800 From: Terry Lambert X-Mailer: Mozilla 4.79 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: Andrea Campi Cc: "M. Warner Losh" , Hidetoshi Shimokawa , current@FreeBSD.ORG Subject: Re: firewire hangs on Thinkpad References: <20030124144823.GA600@webcom.it> <20030125.115501.13766238.imp@bsdimp.com> <20030129114951.GA3635@webcom.it> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-ELNK-Trace: b1a02af9316fbb217a47c185c03b154d40683398e744b8a47e512cc7f6a22f6b26de22ec1049771a350badd9bab72f9c350badd9bab72f9c350badd9bab72f9c Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Andrea Campi wrote: > On Sat, Jan 25, 2003 at 11:55:01AM -0700, M. Warner Losh wrote: > > This sounds like it might be an interrupt storm. I'm not sure if the > > fwohci driver is failing to clear an interrupt source, or if the > > cardbus bridge is failing. Have you connected a fw device to the > > firewire card? > > I've been able to run a few more tests, even though I've not done abused > it in every way I have in my mind yet... > > The evidence I currently have is: > - if I load the modules at loader time everything is fine, with or without > a device attached > - if I load the modules later on, the kldload doesn't return and the system > stops responding; I can still enter DDB. The only way to recover from that is > to eject the card; at that point, the system is usable BUT as soon as there > is network activity, the system freezes hard (can't get to DDB). > > IMHO this is 100% an interrupt problem. Does this ring a bell with one of you, > or should I provide more info? I expect that the attach of the device creates an interrupt if the system is already up. This would indicate that it was an order of operations problem in the driver registration for a "live" piece of hardware. Probably, it needs to attach the interrupt handler first, and then handle the result of an active probe as a special case (if it can). This should let you set up the hardware in the right order that you don't start getting interrupts because you've enabled them before establishing a handler. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message