From owner-freebsd-current Wed Sep 6 07:44:17 1995 Return-Path: current-owner Received: (from majordom@localhost) by freefall.freebsd.org (8.6.11/8.6.6) id HAA18863 for current-outgoing; Wed, 6 Sep 1995 07:44:17 -0700 Received: from Sysiphos (Sysiphos.MI.Uni-Koeln.DE [134.95.212.10]) by freefall.freebsd.org (8.6.11/8.6.6) with SMTP id HAA18852 for ; Wed, 6 Sep 1995 07:44:12 -0700 Received: by Sysiphos id AA21562 (5.67b/IDA-1.5 for current@freebsd.org); Wed, 6 Sep 1995 16:42:07 +0200 Message-Id: <199509061442.AA21562@Sysiphos> From: se@zpr.uni-koeln.de (Stefan Esser) Date: Wed, 6 Sep 1995 16:42:06 +0200 In-Reply-To: Jean-Marc Zucconi "crash (ncr problem?)" (Sep 6, 15:43) X-Mailer: Mail User's Shell (7.2.6 alpha(2) 7/9/95) To: Jean-Marc Zucconi Subject: Re: crash (ncr problem?) Cc: current@freebsd.org Sender: current-owner@freebsd.org Precedence: bulk On Sep 6, 15:43, Jean-Marc Zucconi wrote: } Subject: crash (ncr problem?) } Hi, } } I had a crash on a very current system (make world and kernel from } yesterday), in addition to the sig11's: } Sep 6 08:21:15 qix /kernel: t virtual address = 0x0 } Sep 6 08:21:15 qix /kernel: fault code = supervisor read, page not present } Sep 6 08:21:16 qix /kernel: instruction pointer = 0x8:0xf017f45f } Sep 6 08:21:16 qix /kernel: code segment = base 0x0, limit 0xfffff, type 0x1b } Sep 6 08:21:16 qix /kernel: = DPL 0, pres 1, def32 1, gran 1 } Sep 6 08:21:16 qix /kernel: processor eflags = interrupt enabled, resume, IOPL = 0 } Sep 6 08:21:16 qix /kernel: current process = Idle } Sep 6 08:21:16 qix /kernel: interrupt mask = bio } Sep 6 08:21:17 qix /kernel: panic: page fault } Sep 6 08:21:17 qix /kernel: } Sep 6 08:21:17 qix /kernel: syncing disks... } Sep 6 08:21:20 qix /kernel: panic: page fault } Sep 6 08:21:20 qix /kernel: } Sep 6 08:21:20 qix /kernel: dumping to dev 1, offset 0 } Sep 6 08:21:20 qix /kernel: dump 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 } } 0xf017f45f is in ncr_exception() and 0xf0110781 is in tsleep() Hmm, I had added some debug output code to the ncr_exception handler. There might be a NULL pointer dereference in some special situation somewhere, though I wouldn't expect this ... I'll disable that code again, as soon as I find time to upload another set of patches (later today). You can try doing so yourself. Locate the line containing a string of 'ERROR (' and remove the code below the printing of the 'reg:' line (script instructions and scatter gather table). (Sorry don't have access to my FreeBSD system, currently.) Regards, STefan -- Stefan Esser, Zentrum fuer Paralleles Rechnen Tel: +49 221 4706021 Universitaet zu Koeln, Weyertal 80, 50931 Koeln FAX: +49 221 4705160 ============================================================================== http://www.zpr.uni-koeln.de/staff/esser/esser.html