summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* glsl2: Fix reversed value of step().Eric Anholt2010-06-302-38/+38
| | | | It's 0.0 if x < edge, not 1.0. Partial fix for glsl-fs-step.
* ir_to_mesa: Note which of our expr ops are unsupported 1.30 features.Eric Anholt2010-06-301-0/+11
|
* glsl2: Fix up the implementation of fract() for vector types.Eric Anholt2010-06-302-30/+6
| | | | There's no need to split each vector component out, just do vector ops.
* ir_to_mesa: Initialize the (we never use it) abs field of Mesa src regs.Eric Anholt2010-06-301-0/+1
|
* ir_to_mesa: Send the negate field on to Mesa IR.Eric Anholt2010-06-301-0/+2
| | | | Fixes glsl-fs-neg.
* ir_to_mesa: When generating a swizzle, respect the reg's current swizzle.Eric Anholt2010-06-301-5/+5
| | | | Fixes depth-tex-modes-glsl.
* glsl2: Implement AST->HIR support for the "discard" instruction.Kenneth Graunke2010-06-301-1/+1
|
* glsl2: Define new ir_discard instruction.Kenneth Graunke2010-06-3011-0/+115
|
* ir_to_mesa: Support gl_FragDepth.Eric Anholt2010-06-301-1/+1
| | | | Fixes glsl-bug-22603.
* glsl2: Allow a fragment shader to not write a color.Eric Anholt2010-06-301-6/+0
| | | | | | I can't find any text justifying this check, and it caused a reasonable-looking shader in glsl-bug-22603 (which writes only gl_FragDepth) to fail.
* mesa: Don't look in unallocatd param slots for parameter values.Eric Anholt2010-06-301-1/+1
| | | | | | | | | | glsl-derivs would add 40.0, 0.0, and 1.0 in order. When we went looking for 0.0, we'd find it in the second slot of the param, and use it, but param->Size would still be 1. When we went to add 1.0 and didn't find it, we'd put allocate it to that second slot and the 0.0 would actualy end up being 1.0. Fixes glsl-derivs, glsl-deriv-varyings.
* ir_to_mesa: Fix the indexing of attributes in the program's Attributes.Eric Anholt2010-06-301-1/+1
| | | | | This fixes GetAttribLocation returning VERT_ATTRIB_GENERIC1 instead of 1, caught by glsl-dlist-getattriblocation.
* glsl2: Fix storing of dead memory in the symbol table.Kenneth Graunke2010-06-301-2/+2
| | | | | decl->identifier is part of the AST, so it doesn't live very long. Instead, add var->name which is owned by var.
* glsl2: Preprocessed source doesn't need to live past compile time.Kenneth Graunke2010-06-302-8/+2
|
* glsl2: Create new talloc contexts the "right" way.Kenneth Graunke2010-06-302-2/+2
|
* glsl2: Use _mesa_glsl_parse_state as the talloc parent, not glsl_shader.Kenneth Graunke2010-06-306-120/+120
| | | | | | | | | | | | | _mesa_glsl_parse_state should be the parent for all temporary allocation done while compiling a shader. glsl_shader should only be used as the parent for the shader's final IR---the _result_ of compilation. Since many IR instructions may be added or discarded during optimization passes, IR should not ever be allocated to glsl_shader directly. Done via sed -i s/talloc_parent(state)/state/g and s/talloc_parent(st)/st/g. This also removes a ton of talloc_parent calls, which may help performance.
* glsl2: Steal the live IR and free the rest of the junk.Kenneth Graunke2010-06-302-0/+22
|
* linker: Don't dynamically allocate slots for linked shadersIan Romanick2010-06-302-3/+6
| | | | | The can be at most one shader per stage. There are currently only two stages. There is zero reason to dynamically size this array.
* ir_to_mesa: Tell Mesa about our choices for vertex attribute locations.Eric Anholt2010-06-301-0/+9
|
* linker: Don't automatically allocate VERT_ATTRIB_GENERIC0Ian Romanick2010-06-301-0/+6
|
* glsl2: Use Mesa's gl_shader_program instead of our own struct glsl_program.Eric Anholt2010-06-306-66/+36
| | | | This avoids more allocation and shuffling of data around.
* glsl2: Make function names and variable names be children of the node.Eric Anholt2010-06-301-3/+2
| | | | This avoids losing their memory when the parser state is freed.
* glsl2: Move our data from a glsl_shader* on the side to the main gl_shader *.Eric Anholt2010-06-307-125/+95
| | | | | This saves recompiling at link time. gl_shader->ir is made a pointer so that we don't have to bring exec_list into mtypes.h.
* ir_to_mesa: Add support for ir_unop_abs.Eric Anholt2010-06-291-0/+4
|
* ir_to_mesa: Add support for dFdx, dFdy.Eric Anholt2010-06-291-0/+8
|
* ir_to_mesa: Start adding support for texture instructions.Eric Anholt2010-06-292-2/+142
| | | | | | | Fixes: glsl-fs-bug25902 glsl-fs-sampler-numbering glsl-lod-bias
* glsl2: Keep the same number of components in implicit conversions.Kenneth Graunke2010-06-291-0/+6
| | | | Fixes piglit test glsl-implicit-conversion-01.
* glsl2: Make gl_MaxDrawBuffers available in the vertex shaderIan Romanick2010-06-291-32/+42
|
* glsl2: Make gl_MaxDrawBuffers available in the fragment shaderIan Romanick2010-06-291-0/+15
|
* glsl2: Make gl_FragData be available in GLSL 1.10 tooIan Romanick2010-06-291-12/+5
|
* glsl2: Pass MaxDrawBuffers from core Mesa into the GLSL compilerIan Romanick2010-06-294-24/+35
|
* glsl_type: Add _mesa_glsl_release_types to release all type related storageIan Romanick2010-06-293-2/+33
|
* glsl_type: All glsl_type objects live in their own talloc contextIan Romanick2010-06-292-4/+17
|
* glsl_type: Record type constructors are privateIan Romanick2010-06-292-12/+16
|
* glsl_type: Add get_record_instance methodIan Romanick2010-06-293-2/+86
|
* glsl_type: Vector, matrix, and sampler type constructors are privateIan Romanick2010-06-292-34/+42
|
* glsl_type: Make all static objects be class privateIan Romanick2010-06-293-39/+72
|
* glsl2: Use talloc_strdup when generating constructor temporary namesIan Romanick2010-06-291-8/+13
|
* glsl_type: Remove vector and matrix constructor generatorsIan Romanick2010-06-293-370/+0
| | | | | | All scalar, vector, and matrix constructors are generated in-line during AST-to-HIR translation. There is no longer any need to generate function versions of the constructors.
* glsl2: Don't flatten constructor parameters to scalarsIan Romanick2010-06-291-129/+59
| | | | | | | | Now that all scalar, vector, and matrix constructors are emitted in-line, the parameters to these constructors should not be flattened to a pile of scalars. Instead, the functions that emit the in-line constructor bodies can directly write the parameters to the correct locations in the objects being constructed.
* glsl2: Always emit matrix constructors inlineIan Romanick2010-06-291-1/+321
|
* glsl2: Always emit vector constructors inlineIan Romanick2010-06-291-1/+100
|
* ir_swizzle: Add new constructor, refactor constructorsIan Romanick2010-06-292-18/+56
| | | | | Adds a new constructor that takes an array of component values. Refactors the meat of the two constructors to an init_mask method.
* glsl2: Update TODO.Kenneth Graunke2010-06-291-4/+1
|
* glsl2: Check for non-void functions that don't have a return statement.Kenneth Graunke2010-06-292-0/+13
| | | | | | | This doesn't do any control flow analysis to ensure that the return statements are actually reached. Fixes piglit tests function5.frag and function-07.vert.
* glsl2: Reject return types with qualifiers.Kenneth Graunke2010-06-291-1/+10
| | | | Fixes piglit test return-qualifier.frag.
* glsl2: Add a method for querying if an AST type has any qualifiers.Kenneth Graunke2010-06-292-0/+11
|
* glsl2: Check that returned expressions match the function return type.Kenneth Graunke2010-06-291-3/+11
| | | | | From my reading of the specification, implicit conversions are not allowed. ATI seems to agree, though nVidia allows it without warning.
* glsl2: Invoke preprocessor before calling the compiler properIan Romanick2010-06-291-3/+11
|
* ir_to_mesa: Actually initialize the undef register for scalar_op1.Eric Anholt2010-06-281-1/+1
| | | | | Fixes glsl-sin, glsl-cos on 965, where we rely on unused src arguments in the VS having a file of PROGRAM_UNDEFINED.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-22 12:26:32 +0000