From owner-freebsd-ia64 Sat Apr 14 3:55:24 2001 Delivered-To: freebsd-ia64@freebsd.org Received: from anchor-post-31.mail.demon.net (anchor-post-31.mail.demon.net [194.217.242.89]) by hub.freebsd.org (Postfix) with ESMTP id 4C21A37B449; Sat, 14 Apr 2001 03:55:20 -0700 (PDT) (envelope-from dfr@nlsystems.com) Received: from [62.49.251.130] (helo=herring.nlsystems.com) by anchor-post-31.mail.demon.net with esmtp (Exim 2.12 #1) id 14oNi3-000Omh-0V; Sat, 14 Apr 2001 11:55:19 +0100 Received: from herring (herring [10.0.0.2]) by herring.nlsystems.com (8.11.2/8.11.2) with ESMTP id f3EAs3765348; Sat, 14 Apr 2001 11:54:03 +0100 (BST) (envelope-from dfr@nlsystems.com) Date: Sat, 14 Apr 2001 11:54:03 +0100 (BST) From: Doug Rabson To: John Baldwin Cc: , Subject: Re: Need some help w/ia64 :) In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-ia64@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Fri, 13 Apr 2001, John Baldwin wrote: > At the moment ia64 kernels on -current panic in the pmap code. I'd like to get > a backtrace in ddb, but I can't seem to be able to send input to the console. > I was curious if that worked at all or is known to be broken. I also recall > Dough saying that ski engaged in some Linux stupidness with respect to the > tty's that he had to work around. Unfortunately, I can't seem to find an > e-mail with the actual workaround in it, so I'd appreciate any pointers people > have. Also, I've noticed that once I start the kernel going, ski is basically > unresponsive after the kernel panics. None of the windows update if I move > them around, no mouse events, etc. Also, if anyone knows how to fix this panic > (dmesg below) that would be great as well. :) Somehow, ski manages to set VMIN and VTIME on the consoly pty to 255 which stops its console input polling from working. A workaround is to figure out which pty its using with 'stty -a < /dev/ttypX' and when you find it, fix it by 'stty min 0 time 0 < /dev/ttypX'. Unfortunately, I never got around to implementing backtraces with DDB. In general for modern versions of gcc, we will need to parse the unwind records to figure this out. For the toolchain we have now though, gcc always uses a predictable set of registers to store rp and ar.pfs. One could use this to implement a simple backtrace without too much work. -- Doug Rabson Mail: dfr@nlsystems.com Phone: +44 20 8348 6160 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-ia64" in the body of the message