On Fri, 5 Mar 1999, Alex Wulms wrote:
> The main problem with memman is that you can get memory from different slots.
> And you do not know in advance in which order the memory will be allocated.
> So, in the worst case scenario, each and every memory switch is a slotswitch
> between memory in different subslots. Not just a memorymapper switch.
> Switching memory between subslots is even too slow for a copier. The gap
> between two consecutive sectors will be passed completely by the time that
> the memoryswitch is done. Hence, you have to wait an entire track rotation
> when that happens.
This is high level programming, and is, by definition, much slower!
> The only thing you can do about this, is first allocating all memory via
> memman and then examining how the memory is divided over the different slots.
> Then you can make sure that you keep the slotswitching to a bare minimum.
Isn't easier to program the Mapper directly?
BTW, I think this slowness happens only when you have more than one
mapper. So, in the most part of the systems this slow down won't happen.
In SDC, first I selected the memory blocks and after I read the entire
track. Of course 2 tracks have 18kb, and 2 blocks of Megaram (or 1 block
of Mapper) have 16kb, but this can be solved with some LDIRs.
Greetings from Brazil!
-----------------------------------------------------------------
Marco Antonio Simon Dal Poz http://www.lsi.usp.br/~mdalpoz
[EMAIL PROTECTED] "Apple" (c) Copyright 1767, Sir Isaac Newton
/"\
\ / CAMPANHA DA FITA ASCII - CONTRA MAIL HTML
X ASCII RIBBON CAMPAIGN - AGAINST HTML MAIL
/ \
****
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED] and put
in the body (not subject) "unsubscribe msx [EMAIL PROTECTED]" (without the
quotes :-) Problems? contact [EMAIL PROTECTED] (www.stack.nl/~wiebe/mailinglist/)
****
