On 03/11/2011 12:03 AM, Marek Olšák wrote:
On Thu, Mar 10, 2011 at 11:45 PM, Thomas Hellstrom
<thellst...@vmware.com <mailto:thellst...@vmware.com>> wrote:
Hi,
I've been working a bit on optimizing uploads of non-vbo vertex
arrays on the svga device.
Since we have no knowledge of the vertex array size before a draw
call and the vertex array may change values in between subsequent
draw calls it's pretty pointless to upload the whole array to gpu
memory. I basically want to upload only the vertices affected by
each draw call.
This is the same optimization r300g and r600g already have.
Here's where redefine_user_buffer() comes in, but to be able to
use it for this optimization we must require that a subsequent
draw call referencing the user buffer must not access any data
outside [offset, offset+size-1]. A stronger restriction than the
comment in the p_context.h header file.
The problem is we don't have any other info about user buffers in
st/mesa than the minimum and maximum index. We pass that info to both
redefine_user_buffer and draw_vbo. Whether you use
redefine_user_buffer or pipe_draw_info is up to you. The Radeon
drivers chose the latter.
Marek
Hi,
The problem I see is that we can't really use redefine_user_buffers()
for this unless we put a restriction on subsequent draw calls to not use
data outside the bounds given by redefine_user_buffers. I'm thinking in
terms of possible other users of that interface than the mesa state tracker.
I'll take a look into pipe_draw_info, though.
Thomas
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
