On Tue, Jul 16, 2002 at 10:20:01PM -0500, Peter Seebach wrote:
> In message <200207161242.g6GCgj313106@guild.plethora.net>, Peter Seebach writes
> :
> >In message <200207160718.g6G7Icg01654@s102-n054.tele2.cz>, Jaromir Dolecek wri
> >t
> >es:
> >>Peter Seebach wrote:
> >>> My mildly busy system just crashed under load.  I was ripping two CD's at
> >>> once.  It had been doing that on and off for a couple of hours without
> >>> incident.
> >>> 
> >>> Should I be worried?
> >>
> >>Sure.
> >>
> >>What was the traceback?
> >
> >Didn't look this time.  If it happens again, I'll worry more.
> 
> I'm worrying.
> panic: todo (-56448) <=0 ; minphys broken
> 
> physio
> scsipi_do_ioctl
> cdioctl
> spec_ioctl
> VOP_IOCTL
> vn_ioctl
> sys_ioctl
> sysctl_plain(1f, 1f, 1f, 1f, 0)
> 
> (I don't have this set up for an accessible console.)
> 
> This is on a freshly rebuilt -current system as of this morning.

Hum, this seems to be related to userland scsi commands. I'd like to know
what the parameters of this one was.
Can you recompile your kernel with '-g' and the next time this happens,
get a core dump ?

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
--