summaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* r600g/llvm: Only emit an instruction's explicit operandsTom Stellard2012-04-231-2/+2
|
* r600g/llvm: Handle copies between vector registersTom Stellard2012-04-232-2/+21
|
* r600g/llvm: Remove debugging hack from R600InstrInfo::copyPhysReg()Tom Stellard2012-04-231-4/+0
|
* r600g/llvm: Tell the code emitter to ignore KILL and BUNDLETom Stellard2012-04-231-1/+3
|
* r600/llvm: Add LOAD_VTX instructionTom Stellard2012-04-231-0/+13
|
* r600g: Add hooks for the LLVM shader compilerTom Stellard2012-04-233-2/+304
| | | | | | | | | The LLVM backend can now be enabled for r600g by using the --enable-r600-llvm-compiler configure flag. If you configure with this flag, you can still use the default compiler by setting the envrionment variable R600_USE_LLVM=0 Reviewed-by: Alex Deucher <[email protected]>
* r600g: Add TGSI->LLVM implementation v2Tom Stellard2012-04-232-0/+330
| | | | | | v2: Add case for ARUBA in r600_llvm_gpu_string() Reviewed-by: Alex Deucher <[email protected]>
* radeon: Move radeon_llvm_emit.cpp declarations into their own headerTom Stellard2012-04-235-19/+58
| | | | Reviewed-by: Alex Deucher <[email protected]>
* radeon: Remove HAVE_LLVM ifdefsTom Stellard2012-04-231-10/+0
| | | | | | Only LLVM 3.1 is supported, so these are not necessary. Reviewed-by: Alex Deucher <[email protected]>
* r600g: fix gpr number calculationVadim Girlin2012-04-231-0/+3
| | | | | Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Tom Stellard <[email protected]>
* r600g: split add_reg into add_reg and add_reg_bo variantsDave Airlie2012-04-234-215/+193
| | | | | | | | | This shaves 2k off the final dri.so, and removes lots of pointless NULL, 0 passing. most like pointless - but it looked nicer to me. Signed-off-by: Dave Airlie <[email protected]>
* glapi: Correct size of allocated _glapi_table structJonas Maebe2012-04-221-1/+1
| | | | | | | | | The __glapi_gentable_set_remaining_noop() routine treats the _glapi_struct as an array of _glapi_get_dispatch_table_size() pointers, so we have to allocate _glapi_get_dispatch_table_size()*sizeof(void*) bytes rather than sizeof(struct _glapi_struct) bytes. Reviewed-by: Jeremy Huddleston <[email protected]>
* r600g: enable GLSL130 on all cardsDave Airlie2012-04-221-3/+2
| | | | | | | | Alexandre Demers sent me some cayman results with no major problems. I'll rip out the env var in a week or so. Signed-off-by: Dave Airlie <[email protected]>
* r600g: enable dual src blending on r600 cardsDave Airlie2012-04-211-1/+1
| | | | | | tested on my rv610 and it passes the tests with no hangs. Signed-off-by: Dave Airlie <[email protected]>
* r600g: enable GLSL 1.30 for r600 classDave Airlie2012-04-211-1/+1
| | | | | | | | | Full piglit run on my rv610 with no regressions. This only leaves cayman, however my cayman is resisting my attempt to get through a full piglit run. Signed-off-by: Dave Airlie <[email protected]>
* r600: enable glsl 1.30 on r700Dave Airlie2012-04-211-1/+1
| | | | | | | | I've done a piglit run on rv740 and confirmed no regressions. We don't get GL3 on r700 due to transform feedback being busted still. Signed-off-by: Dave Airlie <[email protected]>
* r600g: report INTEGERS cap if glsl130 is on.Dave Airlie2012-04-201-1/+2
| | | | | | | | | This cap is used by u_blitter to decide if it can use integers in vertex data. fixes some crashes with glsl130 in piglit Signed-off-by: Dave Airlie <[email protected]>
* r600g: enable glsl 130 on evergreen.Dave Airlie2012-04-201-1/+3
| | | | | | | | | | | | I've done a piglit run on my SUMO machine and I see no regressions. Lots of things to fix (skip->fail), but hey maybe we can fix them if we can see them. I'll try and work my way across r600,700,cayman sometime if nobody else gets to them. Signed-off-by: Dave Airlie <[email protected]>
* r600g: disable I2F conversion for InstanceID if integers are supportedVadim Girlin2012-04-201-11/+16
| | | | | | Signed-off-by: Vadim Girlin <[email protected]> Reviewed-by: Michel Dänzer <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* r600g: store glsl_feature_level in the r600_screenVadim Girlin2012-04-202-1/+3
| | | | | | Signed-off-by: Vadim Girlin <[email protected]> Reviewed-by: Michel Dänzer <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* mesa: move gl_array_attrib::NewState to gl_array_object::NewArraysBrian Paul2012-04-206-18/+10
| | | | | | | | The field wasn't actually used before and it's not used now either. But this is a more logical place for it and will hopefully allow doing smarter draw/array validation (per array object) in the future. Reviewed-by: Mathias Fröhlich <[email protected]>
* i965: Convert live interval computation to using live variable analysis.Eric Anholt2012-04-191-39/+26
| | | | | | | | | | | | | | | | | | | | Our previous live interval analysis just said that anything in a loop was live for the whole loop. If you had to spill a reg in a loop, then we would consider the unspilled value live across the loop too, so you never made progress by spilling. Eventually it would consider everything in the loop unspillable and fail out. With the new analysis, things completely deffed and used inside the loop won't be marked live across the loop, so even if you spill/unspill something that used to be live across the loop, you reduce register pressure. But you usually don't even have to spill any more, since our intervals are smaller than before. This fixes assertion failure trying to compile the shader for the "glyphy" text rasterier and piglit glsl-fs-unroll-explosion. Improves Unigine Tropics performance 1.3% +/- 0.2% (n=5), by allowing more shaders to be compiled in 16-wide mode.
* i965: Move the old live interval analysis code next to the new live vars code.Eric Anholt2012-04-192-122/+122
| | | | I'm about to replace the insides of this using the new analysis.
* i965: Add support for live variable analysis using dataflow analysis.Eric Anholt2012-04-193-0/+245
|
* i965: Add basic block generator.Eric Anholt2012-04-195-0/+392
| | | | | | This takes the fs_inst list generated by the visitor, and generates a list of basic blocks with edges between them. This is a building block for data-flow analysis.
* i965/fs: Suppress printing the whole loop in BRW_OPCODE_DO annotation.Eric Anholt2012-04-191-0/+2
|
* glsl: Properly throw compile-time errors for conflicting FS output writes.Eric Anholt2012-04-191-0/+66
| | | | | | | | | | | We were checking for these at link time previously, which is not as early as mandated, and would actually fail to detect conflicting writes if dead code removal removed some writes. Fixes failures in piglit glsl-*/compiler/fragment-outputs/write-gl_Frag* Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Track in each ir_variable whether it was ever assigned.Eric Anholt2012-04-193-13/+33
| | | | | | | This will be used for some compile-and-link-time error checking, where currently we've been doing error checking only at link time. Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Convert the tests directory to automake.Eric Anholt2012-04-192-0/+3
| | | | | | | | | | | This runs optimization-test and produces the usual automake test output, which may be interesting to automated build systems. This doesn't convert the tests to be individually exposed to the automake runner, because automake doesn't like wildcards (due to being nonportable in make, not that we care). Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Fix up some comments in structure splitting, based on array-split work.Eric Anholt2012-04-191-3/+8
| | | | Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Don't allow array splitting on function arguments.Eric Anholt2012-04-191-1/+17
| | | | | | | | | | | This is the reason the declaration member existed in the reference visitor, but I didn't copy the code from structure splitting that avoided setting it. This wasn't currently a problem, because we don't allow splitting of in/out variables. But that would be nice to change some day. Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Rename the "whole_array_access" member in array splitting.Eric Anholt2012-04-191-11/+15
| | | | | | | This was carried over from structure splitting, without thinking about whether the name still made sense in this context. Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Fix up a comment explaining what a visitor class does.Eric Anholt2012-04-191-1/+4
| | | | | | | Ken noted that some of the "actual work" was happening in the caller of this class. Reviewed-by: Kenneth Graunke <[email protected]>
* glsl: Use the column_types() helper method.Eric Anholt2012-04-191-2/+1
| | | | | | Ken noted this in a review of this patch that I pushed early. Reviewed-by: Kenneth Graunke <[email protected]>
* Revert "mesa: only update _MaxElement when we actually need it"Marek Olšák2012-04-191-3/+1
| | | | | | | This reverts commit 0de5a21470b3bff9b7c8714e5d960d5ed9d01b9c. I was wrong, we use it in the vbo module too. This fixes a performance regression in Nexuiz.
* radeonsi: Fix VGPR_BIT() definition.Michel Dänzer2012-04-191-1/+1
| | | | | | | | Fixes encoding of VOP3 shader instructions. The shift was wrong for source registers 2 and 3, and the resulting value was only 32 bits, so the shift in SICodeEmitter::VOPPostEncode() didn't work as intended.
* radeonsi: Replace magic numbers for vertex buffer resource.Michel Dänzer2012-04-191-4/+8
|
* radeonsi: (User) SGPR related cleanups.Michel Dänzer2012-04-193-16/+33
| | | | | | Use the same user SGPRs for the same purpose in vertex and pixel shaders. Better calculation of the number of SGPRs to reserve.
* radeonsi: Fix sampler offsets for shader intrinsic.Michel Dänzer2012-04-191-2/+2
| | | | | The sampler number is in TGSI source register 1, and the S_LOAD_DWORD* instructions take offsets in DWORDs, not bytes.
* mesa: move unrefs of array objects earlier in _mesa_free_context_data()Brian Paul2012-04-191-3/+3
| | | | | | | | | | If a non-default array object was bound at context destruction time we'd try to unreference the array object after it was already deleted in _mesa_free_varray_data(). Now do the unref first. Fixes a regression from commit 86f53e6d6bd07e2bc3ffcadeb9a4418fbae06e0b. Reviewed-by: José Fonseca <[email protected]>
* nv50,nvc0: prevent multiple flushes when user spins on get_query_resultChristoph Bumiller2012-04-192-19/+31
|
* nv50/ir/opt: swap VP inputs to first source where possibleChristoph Bumiller2012-04-191-0/+17
|
* vbo: remove vbo_context::mat_currvalMarek Olšák2012-04-195-12/+8
| | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* vbo: remove vbo_context::generic_currvalMarek Olšák2012-04-197-12/+16
| | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* vbo: remove vbo_context::legacy_currvalMarek Olšák2012-04-197-16/+14
| | | | | | | | | | | | It's not nice when you have several variables pointing to the same array and you wanna ask your editor "where is this used" and you only get an answer for one of the four currval, legacy_currval, generic_currval, mat_currval, which is quite useless, because you never see the whole picture. Let's get rid of the additional pointers. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* vbo: remove unused variables in vbo_contextMarek Olšák2012-04-191-4/+0
| | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* vbo: don't check twice whether it's valid to renderMarek Olšák2012-04-191-20/+0
| | | | | | | | It's already done in _mesa_validate_Draw* and it's not needed to do it again unless I am missing something. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* mesa: set _NEW_VARYING_VP_INPUTS flag only if fixed-func VP is enabled & validMarek Olšák2012-04-191-1/+10
| | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* mesa: add _NEW_VARYING_VP_INPUTS for gl_context::varying_vp_inputsMarek Olšák2012-04-195-9/+11
| | | | | | | | | This is a frequently-updated state and _NEW_ARRAY already causes revalidation of the vbo module. It's kinda counter-productive to recompute arrays in the vbo module if _NEW_ARRAY is set and then set _NEW_ARRAY again. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
* mesa,vbo: properly detect when vertex arrays need to be recalculatedMarek Olšák2012-04-196-45/+25
| | | | | | | | | | | | | | | | | | | This moves the RebindArrays flag into the vbo module, consolidates the code, and adds missing vbo_draw_method calls. Also with this change, the vertex arrays are not needlessly recalculated twice. The issue with the old code was: - If recalculate_input_bindings updates vp_varying_inputs, _NEW_ARRAY is set. - _mesa_update_state is called and the vp_varying_inputs change causes regeneration of the fixed-function shaders, which also sets _NEW_PROGRAM. - The occurence of either _NEW_ARRAY or _NEW_PROGRAM sets the recalculate_inputs flag to TRUE again. - The new code sets the flag to FALSE after the second _mesa_update_state, because there can't possibly be any change which would require recalculating the arrays. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Mathias Fröhlich <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-19 17:14:38 +0000