Subject: Re: Processor correctavke error?
To: Jason Thorpe <>
From: Tim Rightnour <>
List: port-alpha
Date: 06/10/1998 18:05:20
On 10-Jun-98 Jason Thorpe spoke unto us all:
# On Wed, 10 Jun 1998 16:45:45 -0700 (MST) 
#  Tim Rightnour <> wrote:
#  > fatal user trap:
# No... if it had been an error in user space, it would have been SIGSEGV.  But
# the address (a0) is in kernel space.

Well.. cc said "fatal signal 11"  I just assumed the mmu was saying "nothing
there fred"

# Let's take a look at what we have here:
#       a0 == 0xffffffffc8000004.  That is in K1SEG, i.e. a mapped kernel
#       segment.

Curious, how did you know this?

# How easily can you reproduce this?  What NetBSD version are you running?

I'm running the snapshot right now..  I just compiled and booted a new kernel
from current about 4 days ago..  I got this while doing a make build of the
world.  I typed make again, and it happily continued on it's merry way.  I'll
keep building and let you know if I can get it to happen again.  

I used to get fatal signal 11's all the time under i386.. It's incredibly
annoying..  I usually assume it means it hit the ulimit somewhere, and bombed. 
Normally if I just make again, it picks up where it left off and goes on it's

Tim Rightnour    -