Subject: Memory faults and protection from them.
To: None <>
From: Andrew Sporner <>
List: tech-kern
Date: 07/20/2000 11:56:00

A thought had occured to me a week ago.  A few
years ago I was listening to a guy at HP talk
about making memory more reliable in the HP/UX 
by "fencing" out bad pages.  In other words 
any page of memory that generated an ECC fault
would be put away somewhere and would become

Has this been talked about here before?