Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!udel!rochester!kodak!sisd!jeh
From: jeh@sisd.kodak.com (Ed Hanway)
Newsgroups: comp.sys.amiga.tech
Subject: Re: further A2500 speedup?
Message-ID: <1990Sep6.203955.25236@sisd.kodak.com>
Date: 6 Sep 90 20:39:55 GMT
References: <1990Sep5.171708.8069@athena.mit.edu> <14241@cbmvax.commodore.com>
Sender: news@sisd.kodak.com
Organization: Printer Products Division Eastman Kodak
Lines: 15

In article <14241@cbmvax.commodore.com> daveh@cbmvax (Dave Haynie) writes:
>SetCPU uses the MMU to relocate the system stack, if it finds the system
>stack in Chip memory.  It doesn't try to change VBR (Vector Base Register,
>which determines where the vector table used for interrupts and exceptions
>is located) or any Exec stuff.  If MOVESSP works without building an MMU
>table, it might be useful, and if you run it before SetCPU, you will avoid
>moving the system stack twice.  Assuming it works OK in the first place; I
>don't anything about MOVESSP except what you've told me.

MOVESSP doesn't use the MMU and should work even on a vanilla 68000.
It does free the Chip RAM used by the original stack (6K, wow) for re-use,
which doesn't sound possible when using the MMU.

Ed Hanway
uunet!sisd!jeh