[PLUG] Redhat 9 install on Asus P4GE-V

Geoffrey Herteg gherteg at herteg.net
Fri Apr 11 13:28:02 UTC 2003


This is a followup to a private email I sent to Launi last night...

After getting to work this morning I consulted the Word of Intel (a.k.a.
Intel SMD, Vol. 3).  Turns out the MCE Status register might be indicating
a bad microcode patch checksum, rather than just a bad processor state and
fatal error (the IA32_MCx_STATUS register seems to indicate a simple status
error 0x4, which is the bad uCode patch error).

This could explain why NT loaded, but not some of the newer distributions.
The newer kernels may be trying to load a bad microcode patch into the
processor during boot.

Tried something older than RH8, or compiling the newest kernel on a
different box and booting it off a floppy?

- Geoff

On Thu, Apr 10, 2003 at 08:12:11PM -0700, Launi wrote:
> First off here is the error:
> CPU 0: machine check exception 0000000000000004
> Bank 0: A200000084010400
>
> Launi,
>
> I think the details are in the Intel SDM, volume 3, chapter 14 
> (Machine Check Arch):
>	http://developer.intel.com/design/pentium4/manuals/245472.htm
> Look for the IA32_MCx_Status register description, it
> gives a description of what all the bit fields mean...




More information about the PLUG mailing list