In article <43E9DD5B.5040708@earthlink.net>,
Kenneth Freidank  <kennethcf@earthlink.net> wrote:
>Any ideas on what happened?
>
>With 3.0 release as the base, I performed a
>    cvs update -A -dP
>on either Feb 6 or Feb 7 for /usr/src and /usr/xsrc.
>
>I tarred up the files and moved them to my IBM 1161 laptop, already 
>running 3.0 release.
>
>I rebuilt the tools, with no errors.
>    ./build.sh -U tools
>
>I rebuilt GENERIC, then enabled PCI BIOS support with a couple of the 
>FIXUP flags, and ath* cardbus support.
>(I tried GENERIC first, just to ensure it would work, and it did.)
>The kernel worked, and so did my DWL-G650 version B5 wireless card!!
>    ./build.sh -U kernel=MYKERNEL
>
>I then followed up with...
>    ./build.sh -U -u -x distribution
>and got the following after several hours..
>
>After the compile line...
>    compile dix/events.o
>the kernel crashed...
>    uvm_fault(0xc0876dc0, 0xc70bc000,0,1)->0xe
>    kernel:supervisor trap page fault, code=0
>    stopped in pid 17294.1(sh) at netbsd:uvm_unmap_remove to x4d:cmpl 
>%edi,0x20($eax)
>   db>
>
>I tried dmesg...
>    db>dmesg
>and after the last line of...
>    wsdisplay0: screen....
>I got...
>    uvm_fault(0xc0876dc0, 0xc70bc000,0,1)->0xe
>
>After a reboot (power down), the df -m says
>    /dev/wd0a   3444   2867   404   87%   /
>    kernfs            0            0      0   100%   /kern
>
>Is this a hw failure, or something else? 

If it happens again, try getting a stack trace.

christos