home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.emulations
- Path: news.uni-bielefeld.de!chammer
- From: chammer@hrz.uni-bielefeld.de (Carsten Hammer)
- Subject: Re: GoldenGate and Cyberstorm
- Sender: news@hermes.hrz.uni-bielefeld.de (News Administrator)
- Message-ID: <DLyI77.F8x@hermes.hrz.uni-bielefeld.de>
- Date: Mon, 29 Jan 1996 19:01:54 GMT
- References: <1173.6597T827T59@hamburg.netsurf.de> <836.6598T713T286@kenwyn.demon.co.uk> <971.6600T874T731@hamburg.netsurf.de>
- Nntp-Posting-Host: dozy.hrz.uni-bielefeld.de
- Organization: Universitaet Bielefeld, Rechenzentrum
-
-
- **just a guess on
- As far as i know GG2 needs enforcer or something similar to
- switch off caching of the io addresses. If this caching
- is not switched off io register might not be read and
- address contents are not valid.
- It might be that enforcer or whatever is used to switch off
- caching doesnt work on a 060. They are at a position
- in the memory map where they not should lay.
- (This was for a ram extension and is cached)
- GG1 and Crosslink use the normal io space for their io registers.
- This is possible because they dont map in isa cards ram linear but
- bank switched. This makes it possible that they only use
- 64k autoconfig space.
- If the missing working switching tool is the reason for the
- problem with gg2, then gg1,crosslink and the gg2(or however it is named)
- for 64k autoconfig space should still work. Anybody tested this?
- **just a guess off
-
- --
- Carsten Hammer Schwindstr. 7 33615 Bielefeld
- chammer@POST.uni-bielefeld.de
- chammer@phyd2hammer.uni-bielefeld.de
-