From owner-freebsd-hackers Sat Jan 7 01:04:58 1995 Return-Path: hackers-owner Received: (from root@localhost) by freefall.cdrom.com (8.6.9/8.6.6) id BAA17842 for hackers-outgoing; Sat, 7 Jan 1995 01:04:58 -0800 Received: from gagme.wwa.com (root@gagme.wwa.com [198.49.174.33]) by freefall.cdrom.com (8.6.9/8.6.6) with SMTP id BAA17836 for ; Sat, 7 Jan 1995 01:04:55 -0800 Received: by gagme.wwa.com (Smail3.1.28.1 #8) id m0rQXEz-000FDmC; Sat, 7 Jan 95 03:15 CST Message-Id: From: joeg@wwa.com (joe grosch) Subject: page fault while in kernel mode To: hackers@freebsd.org Date: Sat, 7 Jan 1995 03:15:16 -0600 (CST) X-Mailer: ELM [version 2.4 PL23] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1273 Sender: hackers-owner@freebsd.org Precedence: bulk Well, I've recompiled the kernel with kernel debug options and now have the problem to the point that I can consistently reproduce this problem. I start a tar listing of a tape and push it into the background ie. tar tvf /dev/rwt0 > tape.list & I then start up emacs and crank up gdb under emacs. gdb starts up fine altho I can hear that the disk is VERY busy. As soon as try to run the program inside gdb I get a page fault. I get dropped into the kernel debugger. I am not very good with the kernel debugger but I do a trace and the debugger tells me the following; _pmap_enter( ... ) _vm_fault( ... ) _pwrite( ... ) _ptrace( ... ) _syscall( ... ) Has anyone else gotten to this point? Have I stumbled over an old bug? or is this a new bug. My source is WC's CDROM (FreeBSD2.0). My system is a 386SX40 with 4 meg of memory, a one gig SCSI disk with 50 meg of swap. I'll write up a more complete discription of the results of the kernel debug session tomorrow (It's 3am). Josef -- __/\__ | Josef Grosch | "It's been a quiet week in Lake \ / | (708) 291-9076 | Wobegon, my home town..." /_ _\ | | \/ | joeg@wwa.com | -Prairie Home Companion