summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* r600g: fix warning introduced by last commit.Dave Airlie2010-08-301-1/+0
|
* r600g: add initial relative support to assemblerDave Airlie2010-08-304-6/+45
| | | | | | passes another ~20 piglits. /me starts to run out low hanging fruit around now.
* linker: Treat sized and unsized array types as the sameIan Romanick2010-08-291-6/+20
| | | | | | | | | | If two shaders contain variables declared with array types that have the same base type but one is sized and the other is not, linking should succeed. I'm not super pleased with the way this is implemented, and I am more convinced than ever that we need more linker tests. We especially need "negative" tests. Fixes bugzilla #29697 and piglit test glsl-link-array-01.
* r600g: add KILP supportDave Airlie2010-08-303-5/+23
| | | | passes glsl1-discard tests
* r600g: fix SSG and op3 neg writingDave Airlie2010-08-302-14/+15
| | | | 8 more piglits, mainly the two SSG tests.
* r600g : add basic loop support.Dave Airlie2010-08-303-23/+247
| | | | | | Adds BGNLOOP, BRK, CONT, ENDLOOP support, ported from r600c. 17 piglits more on r300g.tests.
* r600g: use the values from the correct literalsBas Nieuwenhuizen2010-08-301-4/+23
| | | | | | Created an array for literals as we should not always use the last declared literal. Signed-off-by: Dave Airlie <[email protected]>
* r600g: added literals where needed for POW instructionBas Nieuwenhuizen2010-08-301-0/+9
| | | | | | Fixes size calculation for the bytecode buffer. Signed-off-by: Dave Airlie <[email protected]>
* evergreen : fix shader const allocation and instruction bugs.Richard Li2010-08-296-20/+48
|
* glsl: Initialize data in ast_function_expression::hir.Vinson Lee2010-08-291-1/+1
| | | | | | | Completely initialize data that is passed to ir_constant constructor. Fixes piglit glsl-orangebook-ch06-bump valgrind uninitialized variable error on softpipe and llvmpipe.
* glsl: Completely initialize value member in ir_constant constructor.Vinson Lee2010-08-291-0/+3
| | | | | | | | | The ir_constant::ir_constant(const struct glsl_type *type, exec_list *value_list) did not completely initialize the entire value member. Fixes piglit glsl-fs-sampler-numbering-2 valgrind uninitialized value error in softpipe and llvmpipe.
* glsl: Initialize variable in ir_swizzle::constant_expression_value.Vinson Lee2010-08-291-1/+1
| | | | | | | Complete initialize data passed to ir_constant constructor. Fixes piglit glsl-mat-from-int-ctor-02 valgrind unintialized variable error with softpipe and llvmpipe.
* glsl: Initialize data in read_constant.Vinson Lee2010-08-291-1/+1
| | | | | | | Completely initialize data that is passed into a ir_constant constructor. Fixes piglit glsl-fs-mix valgrind uninitialized variable error on softpipe and llvmpipe.
* ir_to_mesa: Initialize variable in ir_to_mesa_visitor::visit.Vinson Lee2010-08-291-1/+1
| | | | | Fixes piglit glsl-fs-loop valgrind uninitialized value error on softpipe and llvmpipe.
* r600g: drop file I added by mistake in a previous commitDave Airlie2010-08-291-186/+0
|
* glsl: Initialize the rest of values of ir_constant::value.Vinson Lee2010-08-281-0/+12
| | | | | Fixes valgrind uninitialized value errors in the piglit shader tests for softpipe and llvmpipe.
* gallium: Remove unnecessary header from p_state.h.Vinson Lee2010-08-281-1/+0
| | | | Remove p_screen.h.
* nvfx: Remove util_is_pot in favor of util_is_power_of_two.Vinson Lee2010-08-282-5/+5
| | | | This is a follow up to commit 89b2897220acfacdc431f138377fbcec9f0ea812.
* util: Include missing header in u_draw.h.Vinson Lee2010-08-281-0/+1
| | | | Include p_state.h for complete type to pipe_draw_info.
* util: Add forward declaration in u_transfer.h.Vinson Lee2010-08-281-0/+1
|
* r300g,u_blitter: use u_framebufferMarek Olšák2010-08-294-38/+6
| | | | Removing another function duplication in u_blitter.
* util: remove util_is_pot in favor of util_is_power_of_twoMarek Olšák2010-08-296-18/+8
| | | | The function was duplicated.
* i965: Add initial support for texturing to the new FS backend.Eric Anholt2010-08-281-15/+128
| | | | Fixes 11 piglit tests.
* i965: Add disasm for gen5 sampler messages.Eric Anholt2010-08-281-6/+19
|
* i965: Move libdrm/C++ hack introduced in fa2deb3d to intel_context.hCedric Vivier2010-08-281-5/+0
| | | | | | | Fixes build on Linux/GCC 4.4 as libdrm includes are also used by other brw_fs_*.cpp files. Bug #29855
* st/mesa: Include missing header in st_mesa_to_tgsi.c.Vinson Lee2010-08-281-1/+2
| | | | Include p_screen.h for complete type to pipe_screen.
* softpipe: Include missing header in sp_flush.c.Vinson Lee2010-08-281-0/+1
| | | | Include p_screen.h for complete type to pipe_screen.
* draw: Include missing header in draw_vs_llvm.c.Vinson Lee2010-08-281-0/+1
| | | | Include p_screen.h for completely type to pipe_screen.
* llvmpipe: Include missing header in lp_flush.c.Vinson Lee2010-08-281-0/+1
| | | | Include p_screen.h for complete type to pipe_screen.
* r300g: fix blitting between 2D NPOT mipmapsMarek Olšák2010-08-284-23/+51
| | | | | | | | | Even though MIP filtering is not supported, we can bind an arbitrary mipmap as the zero mipmap level. NPOT textures now follow GL_TEXTURE_BASE_LEVEL and GL_TEXTURE_MIN_LOD. This fixes piglit/fbo-copyteximage.
* r300g: fix min/max lod computationMathias Fröhlich2010-08-281-2/+2
|
* r300g: set the correct value in PVS_NUM_CNTLRSMarek Olšák2010-08-281-1/+1
| | | | As per docs.
* llvmpipe: Remove unnecessary header.Vinson Lee2010-08-281-1/+0
|
* glsl2: Decompose matrix comparison into vector operationsIan Romanick2010-08-271-9/+97
|
* trace: Don't try to dump the rgba array if nullJakob Bornecrantz2010-08-281-1/+4
|
* glsl: Rename a couple of common variable names in mat_op_to_vec.Eric Anholt2010-08-271-16/+18
| | | | | It was easy while typing implementations to accidentally overwrite the original expression or assignment variables.
* glsl: Make mat_op_to_vec allocate out of the IR's parent.Eric Anholt2010-08-271-32/+36
| | | | | This will reduce memory consumption of compiled shaders by not dragging optimized-out children around.
* i965: Set the pop count on BRK/CONT inside of an if statement in the FS.Eric Anholt2010-08-271-4/+11
| | | | | | | This is the same as 8de8c97275e9555183a7e8f2238143657bbe60b2 for the VS, and fixes glsl-fs-if-nested-loop and the mandelbrot demo. Bug #29498
* glsl: Regenerate for double destroy fix.Eric Anholt2010-08-271-0/+1
|
* glsl: Protect against double compiler-destroy.Eric Anholt2010-08-271-0/+1
| | | | | DRI was doing teardown when we close the last screen, then an atexit() was added to call it as well.
* dri: Get prototype for _mesa_destroy_compiler().Eric Anholt2010-08-271-0/+1
| | | | Bug #29665.
* i965: Don't strip negate/abs flags when assigning uniform locations.Eric Anholt2010-08-271-5/+5
| | | | Fixes glsl-algebraic-sub-zero-4.
* i965: Add missing handling for BRW_OPCODE_SEL.Eric Anholt2010-08-271-0/+4
| | | | Fixes 4 piglit tests about min, max, and clamp.
* i965: Mask out higher bits of the result of BRW_CMP producing a boolean.Eric Anholt2010-08-271-0/+6
| | | | | When it says it sets the LSB, that's not just a hint as to where the result goes. Only the LSB is modified. Fixes 20 piglit cases.
* i965: Fix the types of immediate integer values.Eric Anholt2010-08-271-2/+2
| | | | When we're trying to do integer ops, handing a float in doesn't help.
* i965: Add translation for RNDD and RNDZ.Eric Anholt2010-08-271-0/+7
| | | | | | Fixes: glsl-fs-any. glsl1-integer division with uniform var
* i965: Add support for ir_binop_mod using do_mod_to_fract.Eric Anholt2010-08-271-0/+4
| | | | Fixes glsl-fs-mod.
* r600g: fix vbo sizeJerome Glisse2010-08-271-1/+1
| | | | | | Silence the kernel, vbo size is size - 1. Signed-off-by: Jerome Glisse <[email protected]>
* mesa: move null ptr check (fdo bug 29843)Brian Paul2010-08-271-2/+2
|
* mesa: Use the format info to identify packed depth/stencil formats.Nick Bowler2010-08-271-6/+2
| | | | | | | | | | | Due to a misunderstanding of the Z24_X8 and X8_Z24 formats, the earlier patch created depth/stencil wrappers for them. This broke swrast. Use the format info instead, which only identifies Z24_S8 and S8_Z24 as packed depth/stencil. It also has the advantage of being nicer code. Signed-off-by: Nick Bowler <[email protected]> Signed-off-by: Brian Paul <[email protected]>