| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
the later fails for some projection matrices (ala Vis5D with Chromium).
|
|
|
|
|
|
|
|
|
|
|
|
generalizes the R/texture comparision operators to include all eight of
the depth test comparisons.
|
|
|
|
|
|
|
|
|
|
conformance problem
|
|
memory. These can be overridden by applications which need to manage this
memory specially. Contributed by Gerk Huisma.
Also, new code for 8-bit -> 16-bit/channel texture image storage which
fills in the least-significant bits properly.
|
|
|
|
|
|
result in _very_ blurry textures. Still need to do some optimization
of the new code in s_span.c
|
|
|
|
instead of passing a GLcontext* to ResizeBuffers(), pass a GLframebuffer*.
The idea is that a window can be resized without it being bound to a rendering
context. This makes for a nice clean-up in the XFree86 server-side GLX code.
Renamed ctx->Driver.ResizeBuffersMESA() to ctx->Driver.ResizeBuffers().
|
|
|
|
|
|
|
|
(Michael Fitzpatrick)
|
|
|
|
|
|
|
|
|
|
from rendering into one window in parallel and wasn't really needed anyway.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Renamed 'packing' to 'unpacking' since we're moving data from client -> GL.
Rnamed DST_ROW_WIDTH to DST_ROW_BYTES.
|
|
|
|
|
|
|
|
|
|
and no more switches inside loops. To do: give the 1D/3D/cube routines the
same treatment.
|
|
Use fast approximation to log().
Check for dq==0 to avoid a per-pixel divide.
|
|
|
