The main problem is one of this chip market is embedded system, not only
x86. And some architecture, like in embedded, sometime doesn't support
DMA, they do support PIO and basic interfacing mode, but DMA is alien.
To be able to use DMA you need hardware and software supporting it and
you don't find it everywhere, even if it seem smarter to have it.
Andr�
Daniel Phillips wrote:
On Sunday 20 March 2005 17:25, Timothy Miller wrote:
You haven't yet given me a clear argument as to what I would replace
direct register writes WITH, or how we would get by without them in
cases where we CANNOT do DMA.
I'm not sure that there are any cases where we can't do DMA, but then
that may be because I just don't understand all the problems. Could
you please spell this out in words of one syllable?
Even if it turns out DMA is impossible in some situations, we could fall
back on the autoincrementing data register idea, and just poke the
words of what would normally be DMA into the data register one at a
time, which the chip will treat exactly as if it were real DMA. This
has to be dead simple to implement, both in hardware and drivers.
Once that is out of the way, I can go back to muttering about how direct
command parsing ought to work, according to me. Who may of course be
completely wide of the mark.
And a major priveleged command that we'd like to be able to put into
indirect is "memory move".
But direct DMA mode is perfectly adequate for that. You can't have
multiple tasks all trying to move video memory around without knowing
about each other anyway.
Regards,
Daniel
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
