Date: Wed, 13 Oct 1999 09:46:46 -0400 (EDT) From: Andrew Gallatin <gallatin@cs.duke.edu> To: "Erik H. Bakke" <erik@habatech.no> Cc: freebsd-alpha@FreeBSD.ORG Subject: Re: Timer problems with current on 164SX Message-ID: <14340.35212.744118.800649@grasshopper.cs.duke.edu> In-Reply-To: <XFMail.991013141910.erik@habatech.no> References: <XFMail.991013135700.erik@habatech.no> <XFMail.991013141910.erik@habatech.no>
next in thread | previous in thread | raw e-mail | index | archive | help
Erik H. Bakke writes: > > Any advice on this? (A log is attached) Since you're not going to get a crashdump because the disk is wedged, please configure a debug kernel with ddb & get a backtrace from ddb. Add options DDB To your config file & reconfigure using config -g. When it panics, you should be dropped into the ddb debugger. At this point, get a backtrace. (eg use the ddb command tr). After the machine comes back up, you can use gdb on compile/NAME/kernel.debug to get an idea what the interesting address is. Eg, the back trace will look something like: db> tr Debugger() at Debugger+0x2c panic() at panic+0xf4 trap() at trap+0x5c8 XentMM() at XentMM+0x20 ata_reinit() at ata_reinit+0xa8 (null)() at 0x1 In gdb, you'd do: (gdb) l *ata_reinit+0xa8 This will give you a line number. I suspect you'll a function like cia_bwx_BLAH before the fault. I'm interested in the caller of that cia function (probably some ata function). Good luck! Drew PS: I'd do this myself, but I have an ide disk on an xp1000. This machine just gives you back garbage on accesses which generate machine checks on older hardware like yours. This means when I use my ide disk, the disk just hangs, but I have no clue where the bad access is. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14340.35212.744118.800649>