On 9/8/07, Paul Brook <[EMAIL PROTECTED]> wrote: > > > In practice it may be easier to do everything with a fire and forget DMA > > > > What do you mean "a fire"? > > "fire and forget". > i.e. you put your commands in a DMA buffer somewere, and have the card process > them asynchronously. You use properties of the command processor (explicit > fences or implicitly processing in submission order) to queue up dependent > commands. The main CPU can then go do something else (e.g. starting to submit > the next frame/object) without having to wait for previous commands to > complete.
Heh. I guess I wasn't fully awake. Yes, this is exactly what we have in mind. -- Timothy Normand Miller http://www.cse.ohio-state.edu/~millerti Open Graphics Project _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
