On May 23, 3:23 pm, mdj <mdj....@gmail.com> wrote:
Since the Zips cache is write-
through, it seems like an oversight that they didn't simply flag the
cache lines as invalid when DMA was activated.
Probably that's because it's not so easy to tell from the 6502 socket
that the /DMA line is being pulled low.
In the case of the Transwarp I, it actually *is* a DMA peripheral.
But the DMA daisy chain supossedly serves the purpose of arbitrating
peripheral cards' DMA requests...
Isn't the transwarp I a well-behaved DMAer, then ?
Or is it that the DMA daisy chain mechanism fails short in
accomplishing it's intended purpose ?