From owner-freebsd-alpha Wed Oct 27 17:31:30 1999 Delivered-To: freebsd-alpha@freebsd.org Received: from alfheim.satanic.net (alfheim.satanic.net [194.200.106.33]) by hub.freebsd.org (Postfix) with SMTP id D0F0A14F51 for ; Wed, 27 Oct 1999 17:31:05 -0700 (PDT) (envelope-from nicolai@satanic.net) Received: from nicolai by alfheim.satanic.net with local (Exim) id 11gdSw-0004fi-00; Thu, 28 Oct 1999 01:30:54 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <14359.39229.647906.434227@alfheim.satanic.net> Date: Thu, 28 Oct 1999 01:30:53 +0100 (BST) From: Nicolai E M Plum To: peter.jeremy@alcatel.com.au Cc: freebsd-alpha@FreeBSD.ORG, multia-users@explode.unsw.edu.au Subject: halting back to Multia SRM In-Reply-To: <99Oct28.091659est.40344@border.alcanet.com.au> References: <99Oct28.091659est.40344@border.alcanet.com.au> X-Mailer: VM 6.67 under Emacs 19.34.1 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Peter Jeremy writes: > I notice that when I halt FreeBSD 3.2-RELEASE, my multia > reports: > > halted CPU 0 > > halt code = 5 > HALT instruction executed > PC = fffffc0000482cc0 > > access violation fault > PCB = 03F542E0 (entry) > PC = 0004FED0 > VA = 00000028 > > exception context saved starting at 03F56300 > ... > Brk 0 at 0012FF08 > > 0012FF08! BPT > > Does anyone know how to make the `halt' return directly to the SRM > console? > > Alternatively, does anyone have documentation for the `ROM monitor' > mode that I wind up in. So far, I've experimentally determined that > TAB is some sort of exception, trying hex characters lets me change > addresses and `o' single-steps. I haven't found a command to exit > back to the SRM (or even perform a reset). I believe it is the XDELTA debugger, for which I have to date found only poor documentation on Digital/Compaq's website. Typing ;P will cause more register dumps &c, then ;P again will after some seconds drop you to the >>> SRM prompt. I would also dearly like to know how to avoid this; OpenBSD does the right thing on one of my Multias if I reboot the operating system, although it emits some of the above sort of messages before it boots. FreeBSD just sits there in the debugger if either 'reboot'ed or 'halt'ed. I can't actually test the OpenBSD machine at the moment because it's got a burnt-out logic chip that I cannot find a UK source for yet (74FCT623, if anyone knows?). I can see no PROM configuration variables that are different and they are running the same SRM version. Nicolai To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message