summaryrefslogtreecommitdiffstats
path: root/src/mesa
Commit message (Collapse)AuthorAgeFilesLines
* swrast: Print out format on unexpected failure in _swrast_DrawPixels.Vinson Lee2010-10-201-1/+1
|
* mesa: Remove FEATURE_ARB_shading_language_120 macro.Kenneth Graunke2010-10-202-5/+0
| | | | Everything should be able to support 1.20 at this point.
* osmesa: link against libtallocOrion Poplawski2010-10-201-1/+1
| | | | | | Otherwise consumers have to, and that's lame. Signed-off-by: Adam Jackson <[email protected]>
* i965: Use the new style of IF statement with embedded comparison on gen6.Eric Anholt2010-10-192-4/+113
| | | | | "Everyone else" does it this way, so follow suit. It's fewer instructions, anyway.
* i965: Set the source operand types for gen6 if/else/endif to integer.Eric Anholt2010-10-191-6/+6
| | | | | I don't think this should matter, but I'm not sure, and it's recommended by a kernel checker in fulsim.
* i965: Add EU emit support for gen6's new IF instruction with comparison.Eric Anholt2010-10-192-1/+31
|
* i965: Disable thread dispatch when the FS doesn't do any work.Eric Anholt2010-10-193-4/+49
| | | | | | This should reduce the cost of generating shadow maps, for example. No performance difference measured in nexuiz, though it does trigger this path.
* i965: Remove the gen6 emit_mi_flushes I sprinkled around the driver.Eric Anholt2010-10-1910-26/+0
| | | | | These were for debugging in bringup. Now that relatively complicated apps are working, they haven't helped debug anything in quite a while.
* i965: Tell the shader compiler when we expect depth writes for gen6.Eric Anholt2010-10-191-0/+6
| | | | | | | | This fixes hangs in some Z-writes-in-shaders tests, though other pieces don't come out correctly. Bug #30392: hang in fbo-fblit-d24s8. (still fails with bad color drawn to some targets)
* mesa: fix mesa version string constructionBrian Paul2010-10-191-1/+1
| | | | | Now that MESA_MINOR=10, we no longer need the extra '0' in the version string.
* mesa: Make sure we have the talloc cflags when using the talloc headersThomas Hellstrom2010-10-191-1/+1
| | | | Signed-off-by: Thomas Hellstrom <[email protected]>
* r300/compiler: Remove unused variable.Vinson Lee2010-10-191-1/+1
|
* r300g: Add new debug option for logging vertex/fragment program statsTom Stellard2010-10-185-9/+70
|
* r300/compiler: Add a new function for more efficient dataflow analysisTom Stellard2010-10-187-132/+397
| | | | | | | rc_get_readers_normal() supplies a list of readers for a given instruction. This function is now being used by the copy propagate optimization and will eventually be used by most other optimization passes as well.
* r300/compiler: Clear empty registers after constant foldingTom Stellard2010-10-181-0/+8
|
* r300/compiler: Fix incorrect assumptionTom Stellard2010-10-183-32/+17
| | | | | It is possible for a single pair instruction arg to select from both an RGB and an Alpha source.
* r300/compiler: Create a helper function for merging presubtract sourcesTom Stellard2010-10-181-117/+102
|
* i965: Remove unused variable.Kenneth Graunke2010-10-181-1/+0
|
* i965: Avoid blits in BufferCopySubdata on gen6.Eric Anholt2010-10-181-1/+2
| | | | Fixes glean/bufferObject.
* i965: Fix scissor-offscreen on gen6 like we did pre-gen6.Eric Anholt2010-10-181-1/+13
|
* i965: Assert out on gen6 VS constant buffer reads that hang the GPU for now.Eric Anholt2010-10-181-0/+4
|
* i965: Fix assertion failure on gen6 BufferSubData to busy BO.Eric Anholt2010-10-181-10/+17
| | | | Fixes fbo-blit and probably several other tests.
* i965: Fix a weirdness in NOT handling.Eric Anholt2010-10-181-1/+1
| | | | | XOR makes much more sense. Note that the previous code would have failed for not(not(x)), but that gets optimized out.
* i965: Disable the debug printf I added for FS disasm.Eric Anholt2010-10-181-5/+7
|
* i965: Add missing "break" statement.Kenneth Graunke2010-10-181-0/+1
| | | | | Otherwise, it would try to handle arrays as structures, use uninitialized memory, and crash.
* mesa: Add missing else in do_row_3DBrian Rogers2010-10-181-1/+1
| | | | | | This fixes erroneous "bad format in do_row()" messages Signed-off-by: Brian Paul <[email protected]>
* st/mesa: update function name, commentsBrian Paul2010-10-151-6/+14
|
* st/mesa: use GLuint to avoid problem w/ uint not defined on mingw32Brian Paul2010-10-151-1/+1
|
* st/mesa: reformatting in st_cb_drawpixels.cBrian Paul2010-10-151-20/+32
|
* st/mesa: fix regressions in glDrawPixels(GL_STENCIL_INDEX)Brian Paul2010-10-152-7/+23
| | | | | | We need to keep track of three different fragment shaders: Z-only, stencil- only, and Z+stencil. Before, we were only keeping track of the first one we encountered.
* i965: Set the type of the null register to fix gen6 FS comparisons.Eric Anholt2010-10-152-24/+41
| | | | | | | | | | We often use reg_null as the destination when setting up the flag regs. However, on gen6 there aren't general implicit conversions to destination types from src types, so the comparison to produce the flag regs would be done on the integer result interpreted as a float. Hilarity ensued. Fixes 20 piglit cases.
* i965: Fix indentation after commit 3322fbafIan Romanick2010-10-151-45/+45
|
* mesa: Add missing header to shaderobj.h.Vinson Lee2010-10-151-0/+1
| | | | Include compiler.h for ASSERT symbol.
* glsl: Slightly change the semantic of _LinkedShadersIan Romanick2010-10-147-52/+45
| | | | | | | | | | | | | | | Previously _LinkedShaders was a compact array of the linked shaders for each shader stage. Now it is arranged such that each slot, indexed by the MESA_SHADER_* defines, refers to a specific shader stage. As a result, some slots will be NULL. This makes things a little more complex in the linker, but it simplifies things in other places. As a side effect _NumLinkedShaders is removed. NOTE: This may be a candidate for the 7.9 branch. If there are other patches that get backported to 7.9 that use _LinkedShader, this patch should be cherry picked also.
* i965: Fix texturing on pre-gen5.Eric Anholt2010-10-141-5/+5
| | | | | I broke it in 06fd639c519214b6ebcbf29127b6d9ed429f8641 by only testing 2 generations of hardware :(
* mesa: remove post-convolution width/height varsBrian Paul2010-10-141-35/+19
| | | | These were left-over bits from when convolution was removed.
* i965: Add support for ir_unop_round_even via the RNDE instruction.Kenneth Graunke2010-10-143-0/+8
|
* glsl: Add a new ir_unop_round_even opcode for GLSL 1.30's roundEven.Kenneth Graunke2010-10-141-0/+1
| | | | Also, update ir_to_mesa's "1.30 is unsupported" case to "handle" it.
* i965: Clean up a warning in the old fragment backend.Kenneth Graunke2010-10-141-1/+5
| | | | Hopefully this code can just go away soon.
* i965: Enable the new FS backend on pre-gen6 as well.Eric Anholt2010-10-141-12/+1
| | | | | | | | | | | It is now to the point where we have no regressing piglit tests. It also fixes Yo Frankie! and Humus DynamicBranching, probably due to the piglit bias tests that work that didn't on the Mesa IR backend. As a downside, performance takes about a 5-10% performance hit at the moment (e.g. nexuiz 19.8fps -> 18.8fps), which I plan to resolve by reintroducing 16-wide fragment shaders where possible. It is a win, though, for fragment shaders using flow control.
* i965: Correctly emit the RNDZ instruction.Kenneth Graunke2010-10-142-3/+31
| | | | | | | | | | | | | | Simply using RNDU, RNDZ, or RNDE does not produce the desired result. Rather, the RND* instructions place a value in the destination register that may be 1 less than the correct answer. They can also set per-channel "increment bits" in a flag register, which, if set, mean dest needs to be incremented by 1. A second instruction - a predicated add - completes the job. Notably, RNDD always produces the correct answer in a single instruction. Fixes piglit test glsl-fs-trunc.
* i965: Use RNDZ for ir_unop_trunc in the new FS.Kenneth Graunke2010-10-141-1/+1
| | | | The existing code used RNDD, which rounds down, rather than toward zero.
* i965: Use logical-not when emitting ir_unop_ceil.Kenneth Graunke2010-10-141-1/+1
| | | | Fixes piglit test glsl-fs-ceil.
* i965: Add peepholing of conditional mod generation from expressions.Eric Anholt2010-10-141-0/+88
| | | | | | | | | | This cuts usually 2 out of 3 instructions for flag reg generation (if statements, conditional assignment) by producing the conditional mod in the expression representing the boolean value. Fixes glsl-fs-vec4-indexing-temp-dst-in-nested-loop-combined (register allocation no longer fails for the conditional generation proliferation)
* i965: Add a function for handling the move of boolean values to flag regs.Eric Anholt2010-10-142-8/+19
| | | | | | This will be a place to peephole comparisions directly to the flag regs, and for now avoids using MOV with conditional mod on gen6, which is now illegal.
* Only install vtxfmt tables for OpenGLKristian Høgsberg2010-10-141-2/+4
| | | | | | GLES1 and GLES2 install their own exec pointers and don't need the Save table. Also, the SET_* macros use different indices for the different APIs so the offsets used in vtxfmt.c are actually wrong for the ES APIs.
* i965: Add a pass to the FS to split virtual GRFs to float channels.Eric Anholt2010-10-142-2/+116
| | | | Improves nexuiz performance 0.91% (+/- 0.54%, n=8)
* i965: Update the live interval when coalescing regs.Eric Anholt2010-10-141-0/+4
|
* i965: Set class_sizes[] for the aligned reg pair class.Eric Anholt2010-10-141-0/+1
| | | | So far, I've only seen this be a valgrind warning and not a real failure.
* Drop the "neutral" tnl moduleKristian Høgsberg2010-10-145-120/+9
| | | | | | | Just always check for FLUSH_UPDATE_CURRENT and call Driver.BeginVertices when necessary. By using the unlikely() macros, this ends up as a 10% performance improvement (for isosurf, anyway) over the old, complicated function pointer swapping.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-26 08:18:34 +0000