[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: mem.h
- To: "Julian F. Reschke" <fjr010@email.mot.com>
- Subject: Re: mem.h
- From: Michael Schwingen <rincewind@tubul.dascon.de>
- Date: Wed, 3 Jun 1998 19:21:20 +0200
- Cc: MiNT mailing list <mint@atari.archive.umich.edu>
- In-reply-to: <356FC264.AC5FF762@email.mot.com>; from Julian F. Reschke on Sat, May 30, 1998 at 10:25:08AM +0200
- References: <356FC264.AC5FF762@email.mot.com>
On Sat, May 30, 1998 at 10:25:08AM +0200, Julian F. Reschke wrote:
> MagiC is using a bit in the program header to specify that for this
> process, a TPA size as specified in the program header is big
> enough. This is a very useful feature, as it allows the kernel to
> load programs into smaller memory blocks, thus reducing another
> reason for memory fragmentation.
A related note:
It seems there are some programs which do not work on 68040 if the cache is
in copyback mode - some even require the cache to be turned off totally.
Now, as soon as we have memory protection, we could also set the cache mode
for all memory pages belonging to a special process instead of slowing down
the whole system to make these programs work.
This would require 1-2 bits in the program header, with the default value
(as on old programs) meaning 'full cache speed', so that you only have to
twiddle with these bits for bad programs.
Does this sound reasonable? Are there still unused bits left?
cu
Michael
--
Michael Schwingen, Ahornstrasse 36, 52074 Aachen