From: Ian R. <id...@us...> - 2004-08-24 20:40:30
|
Ronny V. Vindenes wrote: > tir, 24,.08.2004 kl. 16.08 +0100, skrev Keith Whitwell: > >>Has anyone had a chance to try out Philipp's changes? I haven't & don't want >>to break the driver by commiting them without a little testing of my own. > > I haven't tested it yet either, but it struck me that if we're going to > add support for "non-essential" software-only extensions then we > probably should add a driconf option to enable/disable such extensions > since their existence can affect which code paths clients take? I think this is a good idea for vertex programs *for now*. Once we actually generate machine code that runs well from the vertex programs, the vertex programs should be enabled by default in all drivers. We're just in a funky situation right now where vertex program performance is much worse than regular software TCL performance. Of course, once we get to that point, I think we should get the existing TCL paths and run everything as a vertex program. I guess we can cross (and debate) that bridge when we come to it... |