| Age | Commit message (Collapse) | Author |
|---|
|
Add a "max complexity" heuristic to allow unrolling long loops with small
bodies and short loops with large bodies.
The loop unroll limits may need further tweaking...
|
|
Loops such as this will be unrolled:
for (i = 0; i < 4; ++i) {
body;
}
where 'body' isn't too large.
This also helps to fix the issue reported in bug #19190. The problem there
is indexing vector types with a variable index. For example:
vec4 v;
v[2] = 1.0; // equivalent to v.z = 1.0
v[i] = 2.0; // variable index into vector!!
Since the for-i loop can be unrolled, we can avoid the problems associated
with variable indexing into a vector (at least in this case).
|
|
(cherry picked from commit 3740a06e28f4cd09e2a3dce2da60320aa9304df1)
|
|
Again..
|
|
|
|
uint64 is not (so?) standard, and often redefined by third parties,
causing name clashes.
|
|
|
|
Conflicts:
src/gallium/auxiliary/tgsi/tgsi_exec.c
src/gallium/drivers/cell/spu/spu_command.h
src/gallium/include/pipe/p_shader_tokens.h
src/mesa/main/config.h
src/mesa/main/mtypes.h
src/mesa/shader/prog_execute.c
src/mesa/shader/slang/slang_emit.c
src/mesa/state_tracker/st_program.c
src/mesa/state_tracker/wgl/stw_wgl.h
src/mesa/state_tracker/wgl/stw_wgl_pixelformat.h
|
|
On Windows snprintf is renamed as _snprintf.
|
|
|
|
(cherry picked from commit 8124faf89d638285d8e9aa93adc3ca7f4ee729f3)
Conflicts:
src/mesa/main/get.c
|
|
|
|
MS CRT libraries already provide a default DllMain entrypoint, and
MS Linker will complain if it finds the same symbol in two different
libraries. Therefore the DllMain has to be in (each) winsys.
|
|
|
|
|
|
|
|
|
|
|
|
Silences MSVC.
|
|
|
|
On Windows snprintf is renamed as _snprintf.
|
|
(cherry picked from commit 816fbeaa813e5cdca314a39677c74c8dc700d35a)
Conflicts:
src/mesa/main/get.c
|
|
|
|
(cherry picked from commit 1e3a44fab068f00378613456036716d0c3772969)
|
|
(cherry picked from commit 8124faf89d638285d8e9aa93adc3ca7f4ee729f3)
Conflicts:
src/mesa/main/get.c
|
|
|
|
|
|
MS CRT libraries already provide a default DllMain entrypoint, and
MS Linker will complain if it finds the same symbol in two different
libraries. Therefore the DllMain has to be in (each) winsys.
|
|
|
|
|
|
|
|
|
|
|
|
Silences MSVC.
|
|
This fixes cases such as:
vec4 v4;
vec2 v2;
v4.xz.yx = v2;
The last line now correctly compiles into MOV TEMP[1].xz, TEMP[0].yyxw;
Helps to fix the Humus Domino demo. See bug 19189.
(cherry picked from commit 9736d8f03364068c9ca786f88a4c2881d98d5768)
|
|
(cherry picked from commit 724ed5b84bc2491da244d091de889d5b35503b5d)
|
|
Now only the samplers that are actually used by texture() functions are
saved in the uniform variable list. Before, we could run out of samplers
if too many were declared while only some of them were actually used.
(cherry picked from commit 1fad6ccb756ae33ca3115f59c99ca8abbeb0321e)
Conflicts:
src/mesa/shader/slang/slang_link.c
|
|
(cherry picked from commit dfada459aabf1a02dc2dbdb6644283fc95df2586)
|
|
Fixes incorrect size information. See bug 19273.
(cherry picked from commit e8d7db31e2a6784c765911233cb3d888f612837f)
|
|
This allows code such as "vec4 a = gl_LightSource[i].ambient;" to work.
When a built-in uniform array is indexed with a variable index we need to
"unroll" the whole array into the parameter list (aka constant buffer) because
we don't know which elements may be accessed at compile-time. In the case of
the gl_LightSource array of size [8], we emit 64 state references into the
parameter array (8 elements times 8 vec4s per gl_LightSourceParameters
struct).
Previously, we only allowed constant-indexed references to uniform arrays
(such as gl_LightSource[2].position) which resulted in a single state reference
being added to the parameter array, not 64. We still optimize this case.
Users should be aware that using "gl_LightSource[i].ambient" in their shaders
is a bit expensive since state validation will involve updating all 64
light source entries in the parameter list.
(cherry picked from commit c6537ac8b8130cf2271c8d1e51137a575073c762)
|
|
(cherry picked from commit ca0540e25c86b8095511868b0cbe96d7e85f7437)
|
|
(cherry picked from commit 3be8d6db9e8bfbd1b3ebf9ac382857ad1e6ef753)
|
|
array.length() wasn't working.
Swizzle mask for accessing elements of float arrays was incorrect.
(cherry picked from commit 368df1615ef65afed96a44e1f43ade8cc703696f)
|
|
(cherry picked from commit c4341fe80acf75bf5417ffeb85fe6902218b752a)
|
|
And update some copyrights.
(cherry picked from commit 4561307a27e07e37319fdf993c37f2c1b85ee3e9)
|
|
(cherry picked from commit 49543d7177fabc848822431891266e33bc13c818)
|
|
(cherry picked from commit a1229cc9e7c78c59cad8d0df30b1f28d9bc81faf)
|
|
(cherry picked from commit a66ff046cc169c6479b00b1e7fc5d87b93ad60ed)
|
|
This allows uniform declarations with scalar/array initializers.
The code is rough though, and will be cleaned up.
(cherry picked from commit aa1b90463676ea71f24c1956dcf51deec1bf664c)
|
|
For example, a declaration like
const float[3] xxx = float[3](1.1, 2.2, 3.3);
will place the array in the constant buffer whereas a regular, non-const array
would be placed in the temporary register file.
Next up: do the same thing for uniform arrays.
|
