0bef5b97a9
glsl2: Add support for non-float constants in Mesa IR.
...
Fixes glsl-vs-vec4-indexing-4.
2010-06-25 18:01:20 -07:00
6152fa16a1
glsl2: Take out the spamming of debug flags I'd added.
2010-06-25 17:23:31 -07:00
b2d7ed2aba
glsl2: Don't clear swizzles for Mesa IR constants after fetching them.
...
Missed this while hacking in constants support. Fixes:
glsl-algebraic-mul-*
glsl-algebraic-rcp-*
glsl-vs-swizzle-swizzle-lhs
glsl-vs-vec4-indexing-6
2010-06-25 17:20:46 -07:00
78062273de
ir_reader: Free memory for S-Expressions earlier.
...
There's no point in keeping it around once we've read the IR.
Also, remove an unnecessary talloc_parent call.
2010-06-25 17:07:42 -07:00
85c978f38c
glsl2: Start trying to hook up uniforms.
...
This should be resolved with linker.cpp's location assignment, as
currently we drop that location assignment on the ground. However,
this gets basic programs using uniforms working for now.
2010-06-25 16:35:36 -07:00
0fd97db8b0
glsl2: Associate the GLenum for the type with builtin GLSL types.
2010-06-25 16:35:22 -07:00
bda27424cf
glsl2: Use the parser state as the talloc context for dead code elimination.
...
This cuts runtime by around 20% from talloc_parent() lookups.
2010-06-25 13:38:38 -07:00
abc4e52992
glsl2: Emit OPCODE_END at the end of the Mesa program.
...
The 965 driver can now run a glsl2-generated shader!
2010-06-25 13:28:16 -07:00
582b73fe69
glsl2: Hook up constant parameters in ir_to_mesa.
2010-06-25 13:27:57 -07:00
ffc845a50a
glsl2: Set InputsRead and OutputsWritten on the generated programs.
2010-06-25 12:52:01 -07:00
364fcd8ee1
glsl2: Start integrating ir_to_mesa.cpp into shader_api.h
...
The compiler is now called by the driver, and generates program
instructions. Parameter lists are still not set up, so the driver
chokes on it shortly thereafter.
2010-06-25 12:37:21 -07:00
7f2bf62d25
glsl2: Use Mesa types instead of duping them into our program.h.
2010-06-25 12:25:19 -07:00
2a3d46dd46
glsl2: Fix dependencies. (at least partially)
2010-06-25 12:23:20 -07:00
0a1b54df7a
glsl2: Replace the GLSL compiler with the glsl2 project.
2010-06-24 17:23:21 -07:00
3d6012303c
glsl2: Wrap includes of C interfaces with extern "C".
2010-06-24 17:23:19 -07:00
f4869f3326
glsl2: Remove files that had been imported for standalone.
2010-06-24 17:14:00 -07:00
e82ddb781a
glsl2: Stop .gitignoring the old standalone build system.
2010-06-24 15:52:56 -07:00
9aa0b6d728
glsl2: Move the Mesa IR codegen into mesa/shader/
2010-06-24 15:49:18 -07:00
bcc13b7444
Merge branch 'glsl2-head' into glsl2
...
This brings in the standalone GLSL compiler that we are planning on
replacing the existing Mesa GLSL compiler. It currently targets GLSL
1.20 and the Mesa IR.
2010-06-24 15:47:40 -07:00
e5cf3aadb8
glsl2: Add a README file for the new compiler.
2010-06-24 15:46:53 -07:00
2928588267
glsl2: Move the compiler to the subdirectory it will live in in Mesa.
2010-06-24 15:36:00 -07:00
0ee7d80269
Merge branch 'mesa'
...
This brings in the ir_to_mesa.cpp code I've been developing to codegen
to the Mesa IR. It does not actually generate a complete Mesa
fragment/vertex program yet.
2010-06-24 15:27:10 -07:00
6b01b50888
Move the talloc_parent lookup down in a few hot paths.
...
talloc_parent is still 80% of our runtime, but likely talloc_parent
lookups will be reduced as we improve the handling of memory
ownership.
2010-06-24 15:18:39 -07:00
e33c10328c
Merge remote branch 'cworth/master'
...
Conflicts:
ast_to_hir.cpp
ir.cpp
This brings in the talloc-based memory management work, so that the
compiler (almost) no longer leaks memory.
2010-06-24 15:13:03 -07:00
8364fc85b8
ir_to_mesa: Handle a limited subset of matrix multiplication.
...
glsl-mvp.vert now generates believable code, and mesa mode fails only
5 tests that master doesn't. I must have left out some asserts...
2010-06-24 15:05:21 -07:00
7672064756
ir_to_mesa: Handle constant matrices.
...
There's not much to it since we're not actually storing constant data yet.
2010-06-24 15:05:21 -07:00
f4bd7f262e
ir_to_mesa: Fix copy-and-wasted second argument to compare expresssion ops.
...
Fixes CorrectParse2.vert assertion due to uninitialized values.
2010-06-24 15:05:21 -07:00
4006424f5b
ir_to_mesa: Don't allocate temps for swizzles.
...
We do them in place by actually, you know, swizzling.
2010-06-24 15:05:21 -07:00
a9b619bb3b
ir_to_mesa: Set up storage for uniform vars.
2010-06-24 15:05:21 -07:00
554dbcce77
ir_to_mesa: Move the classes into the file now that we don't have the burg.
...
At 1kloc, it doesn't look like I'll want to split the ir_to_mesa file
up even once it's feature-complete. Move definitions closer to usage,
and prevent rebuilding the world when changing the definitions.
2010-06-24 15:05:21 -07:00
2401338ef8
ir_to_mesa: Remove old monoburg structure.
2010-06-24 15:05:21 -07:00
9cd8cad9f3
ir_to_mesa: Restrict dst writemasks like we did in the monoburg setup.
2010-06-24 15:05:21 -07:00
9d2b8e0b70
ir_to_mesa: Fix copy-and-wasted DIV instruction sequence.
2010-06-24 15:05:21 -07:00
0161515c39
ir_to_mesa: Remove the BURG code.
...
The promise of the BURG was to recognize multi-instruction sequences
and emit reduced sequences for them. It would have worked well for
recognizing MUL+ADD -> MAD and possibly even MIN(MAX(val, 0), 1) ->
MOV_SAT with some grammar changes. However, that potential benefit in
making those optimizations easy is outweighed by the fragility of
monoburg, the amount of (incorrect, as I wrote it) code for using it,
and the burden it was going to cause for handling operations on
aggregate types.
2010-06-24 15:05:21 -07:00
224f712950
ir_to_mesa: Fix mapping of FS texcoord inputs and color output.
2010-06-24 15:05:21 -07:00
ab386f18b0
ir_to_mesa: Try to fix up the dereference handling for the visitor rework.
...
One of the gstreamer shaders I play with now compiles, but input
mappings are wrong.
2010-06-24 15:05:21 -07:00
c23c6c773a
ir_to_mesa: Implement min and max expressions.
...
fixes glsl-orangebook-ch06-bump.frag.
2010-06-24 15:05:21 -07:00
346daeca07
ir_to_mesa: Don't assert over assignments with a constant-true condition.
2010-06-24 15:05:21 -07:00
c2014f03e8
ir_to_mesa: Add support for trunc/ceil/floor.
2010-06-24 15:05:20 -07:00
c45b615a37
ir_to_mesa: Implement neg expression.
2010-06-24 15:05:20 -07:00
3c5979565f
ir_to_mesa: Add sin/cos.
2010-06-24 15:05:20 -07:00
2c432637d0
ir_to_mesa: Start trying to support struct storage.
2010-06-24 15:05:20 -07:00
4a1bd916e7
ir_to_mesa: Fix up array indexing.
...
The grammar for array_reference_vec4_vec4 was set up wrong, so we
weren't generating instructions if necessary for the array index.
2010-06-24 15:05:20 -07:00
2ee85e20c8
ir_to_mesa: Remove stale comment about monoburg.
2010-06-24 15:05:20 -07:00
bdbd9f112e
ir_to_mesa: Add support for variable indexing of temporary arrays.
...
Fixes loop-01.vert, loop-02.vert.
2010-06-24 15:05:20 -07:00
4e5e0f018b
ir_to_mesa: Clean up some handling of builtins and arrays.
...
Constant-index dereferences of arrays should work now. One test is
regressed, but it should have been failing before this commit, too.
2010-06-24 15:05:20 -07:00
64fcbbca9c
ir_to_mesa: Add support for loops.
...
Fixes CorrectParse1 and the glsl2 loop tests that don't use arrays.
2010-06-24 15:05:20 -07:00
0c005bd773
Make loop jump mode public so I can switch on it.
2010-06-24 15:05:20 -07:00
4380099c98
ir_to_mesa: Add logic_or and logic_and to get CorrectFunction1.vert working.
2010-06-24 15:05:20 -07:00
763cd75863
ir_to_mesa: add logic_xor to get CorrectParse2.vert working.
2010-06-24 15:05:20 -07:00
Eric Anholt