Re: How hardware accelerators deals with $0/$01

From: Mia Magnusson <mia_at_plea.se>
Date: Thu, 19 Oct 2017 22:03:33 +0200
Message-ID: <20171019220333.00003eb6@plea.se>
Den Wed, 18 Oct 2017 21:16:28 +0200 skrev groepaz@gmx.net:
> Am Mittwoch, 18. Oktober 2017, 06:00:09 CEST schrieben Sie:
> > > El 18 oct 2017, a las 5:45, Jesus Cea <jcea@jcea.es> escribió:
> > > 
> > > The only option I see is that the 6510 is running a tight loop
> > > reading commands to read/write those addresses and that the card
> > > is releasing the DMA briefly when the fast CPU access those
> > > location for the 6510 to continue execution and simulate the
> > > access, but this seems complicated and fragile.
> > The only other option I see is that when the machine boots, the
> > 6510 runs a ROM in the card that copies basic, kernal and chargen
> > to the card. After this the card takes control, disable the 6510
> > via DMA and simulates the bank switching internally snooping
> > accesses to $0/$1.
> 
> thats exactly what the existing accelerators do. additionally, when
> the external CPU writes to its (external) RAM, some DMA mechanism
> also writes into the C64s RAM, since the VICII can only "see" the
> internal RAM and else you'd get no picture.
 
How does this distinguish between I/O and RAM in the CPU adress space
area where I/O overlaps RAM?

Are we sure that the 6510 I/O port isn't accessible from external
hardware when the CPU is in DMA/three state mode?


-- 
(\_/) Copy the bunny to your mails to help
(O.o) him achieve world domination.
(> <) Come join the dark side.
/_|_\ We have cookies.

       Message was sent through the cbm-hackers mailing list
Received on 2017-10-19 21:00:03

Archive generated by hypermail 2.2.0.