Marek, does adding multi-threading to Mesa will improve things much? I think returning sooner in gl* calls(in which that makes sense ofcourse) in the 3D app, and do driver work in separate thread(s) will improve speed of command stream flow. I am by no means 3D developer(driver or app), but my threading background from other areas tells me that. Is that right?
No announcement yet.
The Open-Source ATI Driver Is Becoming A Lot Faster