Subject: Re: Memory management fault w/NetBSD/alpha 1.6_RC1
To: Dave Huang <khym@azeotrope.org>
From: Jason R Thorpe <thorpej@wasabisystems.com>
List: current-users
Date: 09/03/2002 15:45:28
On Tue, Sep 03, 2002 at 05:31:54PM -0501, Dave Huang wrote:

 > I don't suppose anyone knows what might have cause this:
 > 
 > CPU 0: fatal kernel trap:
 > 
 > CPU 0    trap entry = 0x2 (memory management fault)
 > CPU 0    a0         = 0x0
 > CPU 0    a1         = 0x1
 > CPU 0    a2         = 0xffffffffffffffff
 > CPU 0    pc         = 0x0
 > CPU 0    ra         = 0x0
 > CPU 0    pv         = 0xe3af7
 > CPU 0    curproc    = 0xfffffc00076c08b8
 > CPU 0        pid = 20307, comm = sshd

It jumped through a bad function pointer, maybe.  The return address is
also 0, means it was probably returning to what the RA pointed to.

My guess is the stack got smashed somewhere, and a bogus RA was fetched
from the stack before a function return.

What was the system doing when this happened?

-- 
        -- Jason R. Thorpe <thorpej@wasabisystems.com>