Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
Fix bug inserted in commit c9e723e5013443df984cb3987ffa3a9ba3384b89.
Discovered by Oliver McFadden (z3ro).
|
|
These changes are based on patch review comments from Brian Paul, Alan
Hourihane, and vehemens.
|
|
Common code was pulled out of the per-target if-statment and put at the end
of the for-loop. The common code is guarded by a new variable, curr, that
is set to point to the unit's current target in each if-statement.
|
|
Most switch-statements that have cases for these enums already use code like:
const GLuint idx = pname - GL_SOURCE0_RGB;
... texUnit->Combine.SourceRGB[idx] ...
This patch just brings the remaining bits up to speed.
|
|
|
|
|
|
|
|
|
|
such as #10687.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
found.
The later incorrectly encourages filing a bug (10703, for example).
|
|
|
|
|
|
GLushort->GLuint conversion.
This improves performance and avoids int/float/int conversion problems that
can introduce errors during glCopyTexImage(). Another fix for the depth peeling
algorithm.
|
|
|
|
This is a bit of a hack for now because the tnl module is using the swrast
module to fetch texels. The texture fetch/filter code should probably be
moved into the main/ module since it doesn't really depend upon other
swrast code.
|
|
generates an error
|
|
constants.
Previously, this limited us to 12 temp regs for vertex programs. Many vertex
shaders could exceed that. This forces us to stop using t_vb_arbprogram.c
for now because of its particular register indexing scheme. Need to increase
bits allocated for register indexing, etc.
|
|
|
|
|
|
|
|
|
|
This is a remnant from the libcwrapper days that can be removed now. It is the
last XFree86 glitch in Mesa core.
Reverts the following commit (thanks to Brian Paul for pointing to it):
commit 8b1dc68662f2a18ea4e02c6d3247c5d1da741aa8
Author: Alan Hourihane <alanh@tungstengraphics.com>
Date: Tue Jan 24 21:40:53 2006 +0000
pull in fix from 6.4 branch for XFree86Server definition
commit cc7b6810e2fe6ab485fd96d5f17bf5d9329bd153
Author: Ian Romanick <idr@us.ibm.com>
Date: Fri Oct 21 18:09:24 2005 +0000
Make execmem.c compile in the server (libGLcore). Since xf86mmap
doesn't know about MAP_ANONYMOUS, this is the only viable fix. This
issue will likely have to be revisited at some point.
|
|
|
|
|
|
|
|
|
|
|
|
The calculation of _MaxElement was wrong if the stride was larger than
elementSize, which lead to rejection of every DrawElements call which accessed
the maximum element if CheckArrayBounds was enabled.
|
|
Was removed during glsl-compiler work. Still need to go back and revisit this
because of the interaction with fragment shaders...
|
|
|
|
|
|
|
|
|
|
Nicolai writes:
When the pixmap pixel format has no alpha channel, the x11 driver
(software rendering) adds a wrapped alpha channel on request.
During SwapBuffers, this alpha channel is not copied from back to
front, which means that the front buffer doesn't really contain the
contents that the back buffer previously contained.
A subsequent glReadPixels from the front buffer will return an
incorrect result. The following patch attempts to fix this.
|
|
ctx->Shader.EmitCondCodes determines if we use condition codes.
If not, IF statement uses first operand's X component as the condition.
Added OPCODE_BRK0, OPCODE_BRK1, OPCODE_CONT0, OPCODE_CONT1 to handle
the common cases of conditional break/continue.
|
|
|
|
if not bail accordingly.
Previously we'd only do this test on compressed textures.
|
|
|
|
to fix #10232
Table6.1(in gl2.1) has been applied for glGetTexImage
before calling into _mesa_pack_rgba_span_float.
|
|
or TexImage(DEPTH_COMPONENT)
|
|
If possible, let minCol != maxCol
|