From owner-freebsd-current Tue May 5 22:50:38 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id WAA13845 for freebsd-current-outgoing; Tue, 5 May 1998 22:50:38 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from cerebus.nectar.com (cerebus.nectar.com [204.27.67.90]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id WAA13823 for ; Tue, 5 May 1998 22:50:32 -0700 (PDT) (envelope-from nectar@cerebus.nectar.com) Received: from cerebus.nectar.com (localhost.communique.net [127.0.0.1]) by cerebus.nectar.com (8.8.8/8.8.8) with ESMTP id AAA00768 for ; Wed, 6 May 1998 00:50:02 -0500 (CDT) Message-Id: <199805060550.AAA00768@cerebus.nectar.com> X-PGP-RSAfprint: 00 F9 E6 A2 C5 4D 0A 76 26 8B 8B 57 73 D0 DE EE X-PGP-RSAkey: http://pgp.ai.mit.edu:11371/pks/lookup?op=get&search=0x094724A9 From: Jacques Vidrine To: freebsd-current@FreeBSD.ORG Subject: DEVFS/SLICE fatal trap 12 when reprobing scsi devices Date: Wed, 06 May 1998 00:50:02 -0500 Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Hello, I routinely use ``scsi -f /dev/ssc -p'' to reprobe my SCSI bus when I've switched cartridges in my Iomega jaz drive. This has been working fine, but I decided to try out the SLICE stuff today. It didn't work out, so I thought I'd report. Using -current as of about 11am CDT 1998/5/5, I enabled SLICE and got the following fatal trap when using ``scsi -f /dev/ssc -p'' (beware, copy-and-paste via actual paper): *-----* Fatal trap 12: page fault while in kernel mode fault virtual address = 0x0 fault code = supervisor read, page not present instruction pointer = 0x8:0xf0170811 stack pointer = 0x10:0xf440ed70 frame pointer = 0x10:0xf440eda0 code segment = base 0x0, limit 0xffffff, type 0x1b = DPL 0, pres 1, def32 1, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 405 (scsi) interrupt mask = kernel: type 12 trap, code = 0 Stopped at _suopen+0xa5 movl 0(%eax),%eax *-----* ``trace'' gave: *-----* _suopen(40400002, 3, 2000, f43bbc00, 3100) at _suopen+0xa5 _sscioctl(3100, 800c5108, f440eed4, 3, f43bbc00) at _sscioctl+0x69 _spec_ioctl(f440ee10, f40eea0, f01450ed, f440ee10, 0) at _spec_ioctl+0x44 _spec_vnoperate(f440ee10, 0, f0766900, 800c5108, 3) at _spec_vnoperate+0x15 _vn_ioctl(f0766900, 800c5108, f440eed4, f43bbc00, f43bbc00) at _vn_ioctl+0x111 _ioctl(f43bbc00, f440ef84) at _ioctl+0x2cf _syscall(27, 27, efbfd928, efbfd948) at _syscall+0x127 _Xsyscall() at _Xsyscall+0x35 --- syscall 0x36, eip=0x10f25, esp=0xefbfd8e8, ebp=0xefbfd948--- *-----* The trap happens only with ``option SLICE'' ... accessing ssc through a devfs filesytem without ``option SLICE'' works just as well as without devfs. The jaz drive is the only SCSI device on the test system. Jacques Vidrine To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message