> Date: Fri, 28 Jul 1995 13:10:50 -0500 (CDT)
> From: James Jegers <jimj@miller.cs.uwm.edu>
> Sender: owner-netbsd-bugs@netbsd.org

> >Number:         1294
> >Category:       kern
> >Synopsis:       GDB almost works on kernels
> >Confidential:   yes
> >Severity:       non-critical
> >Priority:       medium
> >Responsible:    kern-bug-people (Kernel Bug People)
> >State:          open
> >Class:          change-request
> >Submitter-Id:   net
> >Arrival-Date:   Fri Jul 28 14:20:00 1995
> >Last-Modified:
> >Originator:     James Jegers
> >Organization:
> 
> >Release:        Current as of Jul 5<NetBSD-current source date>
> >Environment:
> 
> System: NetBSD jimj.home.edu 1.0A NetBSD 1.0A (JIMJ) #4: Fri Jul 28 09:42:32 CDT 1995 jimj@jimj.home.edu:/usr/src/sys/arch/i386/compile/JIMJ i386
> 
> >Description:
> 	After wanting a better debugger for the kernel for some time.  I really
> 	started playing with it and kinda got it to work.
> 	There are two sets of patches here, one is for the gdb to make it
> 	read in a kernel panic/crash file or the live kernel.  I'm still having
> 	problems getting the stack output to look ok.  sometimes it works and
> 	sometimes it doesn't.  I'm not sure if we ever can get the correct
> 	information..???
> 	The other patch is to the kernel to allow user programs to read the
> 	kernel stack from /dev/kmem.  currently access is denied.  I'm not
> 	sure if this fix is the correct way to do it.  It looks like the i386
> 	port is in flux.
> 
> >How-To-Repeat:
> 
> >Fix:
> 
> diff -c2 ../../gdb/gdb/kcorelow.c ./kcorelow.c

Gak!  This fix would stick i386 stuff in machine independent code!
This clearly needs more thought...

Gordon Ross