Date: Fri, 2 Oct 1998 20:19:34 +0930 From: Greg Lehey <grog@lemis.com> To: Josef Karthauser <joe@pavilion.net>, freebsd-current@FreeBSD.ORG Subject: Re: FreeBSD3.0-BETA crash! Message-ID: <19981002201934.C2176@freebie.lemis.com> In-Reply-To: <19981002110648.A17581@pavilion.net>; from Josef Karthauser on Fri, Oct 02, 1998 at 11:06:48AM %2B0100 References: <19981001193321.C12010@pavilion.net> <19981002104659.M24146@freebie.lemis.com> <19981002110648.A17581@pavilion.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday, 2 October 1998 at 11:06:48 +0100, Josef Karthauser wrote: > On Fri, Oct 02, 1998 at 10:46:59AM +0930, Greg Lehey wrote: >> On Thursday, 1 October 1998 at 19:33:21 +0100, Josef Karthauser wrote: >>> The crashes were page faults. >> >> At the very least, we need a stack trace. Check out >> http://www.FreeBSD.org/handbook/handbook294.html for details of how to >> do this. > > I did a complete make world last night, followed by a clean build of the > kernel this morning. > > I'm still getting crashes though: > > Fatal trap 12: page fault while in kernel mode > fault virtual address:= 0xc > fault code = supervisor read, page not present > instruction pointer = 0x8:0xf018ab82 > stack pointer = 0x10:0xf7addf1c > frame pointer = 0x10:0xf7addf28 > code segment = base 0x0, limit 0xfffff, type 0x1b > = DPL 0, pres 1, def32 1, gran 1 > = interrupt enabled, resume, IOPL = 0 > processor eflags = 91 (syslogd) > interrupt mask = > trap number = 12 > panic: page fault > > syncing disks... 10 10 5 1 done > > I've followed the instructions in the handbook but with some problems. > The savecore worked and I've got a kernel.0 and vmcore.0. > The handbook refers to 'strip -d', and the -d flag doesn't exist. The -d flag definitely exists. What happened when you tried it? It only works on debug kernels, and you haven't said that you built one. Did you? > (kgdb) exec-file /var/crash/kernel.0 > "/var/crash/kernel.0": not in executable format: File format not > recognised. Funny. This could be due to the a.out to elf conversions. > # diff -s /var/crash/kernel.0 /kernel > Files /var/crash/kernel.0 and /kernel are identical > > Am I doing something wrong or is something else broken. (For information > the machine has undergone an aout-to-ELF translation.) Well, the way I do this is typically: # cd /var/crash # gdb -k kernel.gdb vmcore.<foo> kernel.gdb is not the saved kernel (which will normally have no symbols), but the symbolized version, which will come from /sys/compile/KERNELLNAME/kernel. I don't think I've tried this since the change to ELF, so maybe there's a problem there. I'll try it Real Soon Now. We need more documentation on this area. If this doesn't make sense to you, let me know what you don't understand, and I'll expand. Greg -- See complete headers for address, home page and phone numbers finger grog@lemis.com for PGP public key To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981002201934.C2176>