On Fri, 9 Jun 2000, Curt Sampson wrote:

# Date: Fri, 9 Jun 2000 13:54:35 -0400 (EDT)
# From: Curt Sampson <cjs@cynic.net>
# To: port-sparc@netbsd.org
# Subject: New Boot Blocks
# 
# On Fri, 9 Jun 2000 Havard.Eidnes@runit.sintef.no wrote:
# 
# > Would it perhaps make sense (at least as an interim measure) to
# > use a boot loader which should load at 0x380000?
# 
# Is there nobody out there capable of writing a boot loader that can
# relocate itself after loading? It strikes me as a bit...unpleasant that
# changing a SIMM can make your machine not boot any more. Not to mention
# that most users are going to have no clue about the SPARC boot process
# and which boot blocks they should install and so on.

It strikes me as odd that some memory smarts cannot be stuck into the boot-
loader!  Is there no way to do a memory probe without causing a fatal
fault?

I rather suspect that either the code is too large for a memory probe,
or the fault handler would exceed the size limitation for a bootloader...?

# cjs
# 


				--*greywolf;
--
BSD: If your server could choose, it would choose BSD.