On Thu, Dec 06, 2001 at 07:17:54PM +0000, Richard Atkinson wrote: > On Wed, 5 Dec 2001, Eric Boon wrote: > > > > This would appear to be the register at memory location #FFFF in slot 3. > > > > Almost right - It *is* the register at #FFFF, but since > > it's used to select a sub-slot, you can't really say that it's > > in slot 3. It's not memory, it's omni-present ;-) > > It's accessed using MREQ# rather than IORQ#. Are you saying you don't have > to select slot 3 in order to change it? You do, he just meant that the slot can be rom or even emtpy but the #FFFF is still useable. If you would draw a picture of it you would probably have the address #FFFF float on top of everything in the slot, so that why it isn't in the slot, hence the omnipresent thing since it isn't tied to a subslot itself.
>In which case, does it affect > other expanded slots also? no > ie. is it possible to have more than one > secondary slot register, on a per slot basis, yes >or does writing to the same > register affect all slots? no > > Presumably when the RAM at #C000 - #FFFF is enabled, writing to this > register will also "write through" to the RAM at location #FFFF, so that > reading from this location will return the last value written to the > register. (?) no, reading from #FFFF wil return an inversed state of the register, even if there isn't any device behind it (or a rom). The ram at #FFFF will never see the write(or read) to this address. (actually I give myself 99% on this, need to check this on real machine with memmorymapper so that I can have the #FFFF address also on #BFFF) David -- For info, see http://www.stack.nl/~wynke/MSX/listinfo.html