On Fri, 18 Mar 2005 14:24:22 -0500
Daniel Phillips <[EMAIL PROTECTED]> wrote:
> The register model however is an obsolete throwback that I would like to
> eradicate from our design.
What's bad about it ?
> We have a command stream, commands have command
> fields, there may be no simple mapping between command fields and
> registers.
Do you mean something like filling the ring buffer with
commands like:
draw line x1,y1,x2,y2
do dma from,to,count
blit from,to
...
?
> We certainly do not need register numbers in the command
> fields. In fact, we don't need registers at all, except for DMA control
> and similar. We don't even need registers for reading GL state: the driver
> knows the GL state, and as a bonus, it knows the current values, not state
> as of some time in the past, which is what the rendering pipeline knows.
How does the driver know everything w/o simulating all stages
of the GPU ? How are errors signaled ?
> Getting rid of the requirement for reading GL state from the card gets rid
> of a whole class of messy pipeline synchronization issues.
I don't really see how you get rid of it.
Maybe it's time that you or Timothy make some kind of document
out of the ideas here with some nice graphics in it, so that
even people like me who have no clue about graphics cards understand
the discussion ^^'
Attila Kinali
--
éãåããéãåã
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
