summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* r600g/sb: don't propagate dead values in GVN passVadim Girlin2013-04-301-0/+3
| | | | | | | | | | | | | | In some cases we use value::gvn_source field to link values that are known to be equal before gvn pass (e.g. results of DOT4 in different slots of the same alu group), but then source value may become dead later and this confuses further passes. This patch resets value::gvn_source to NULL in the dce_cleanup pass if it points to dead value. Fixes segfault during shader optimization with ETQW. Signed-off-by: Vadim Girlin <[email protected]>
* r600g/sb: use simple heuristic to limit register pressureVadim Girlin2013-04-302-3/+33
| | | | | | | | | | | | | | | | | | | | | It's not a complete register pressure tracking, yet it helps to prevent register allocation problems in some cases where they were observed. The problems are uncovered by false dependencies between fetch instructions introduced by some recent changes in TGSI and/or default backend. Sometimes we have code like this: ... SAMPLE R5.xyzw, R5.xyzw ... store R5.xyzw somewhere MOV R5.x, <next x coord> MOV R5.y, <next y coord> SAMPLE R5.xyzw, R5.xyzw ... <may be repeated a lot of times> With 2D resources, z and w in SAMPLE src reg aren't used and can be simply masked, but shader backend doesn't have this information, so it's considered as data dependency by optimization algorithms.
* r600g/sb: improve error checking in ra_coalesce passVadim Girlin2013-04-302-14/+27
|
* r600g/sb: use source bytecode in case of optimization errorsVadim Girlin2013-04-305-11/+25
|
* r600g: plug in optimizing backendVadim Girlin2013-04-308-3/+155
| | | | | | Optimization is enabled with "R600_DEBUG=sb". Signed-off-by: Vadim Girlin <[email protected]>
* r600g/sb: initial commit of the optimizing shader backendVadim Girlin2013-04-3035-0/+17498
|
* r600g: use enum type for domains field in struct r600_resourceVadim Girlin2013-04-301-1/+1
| | | | This prevents the problems when the header is included in C++ code.
* r600g: add new flags to isa instruction tablesVadim Girlin2013-04-301-116/+127
|
* r600g: always create reverse lookup isa tablesVadim Girlin2013-04-301-10/+2
|
* r600g: mask unused source components for SAMPLEVadim Girlin2013-04-301-0/+20
| | | | | | | | This results in more clean shader code and may improve the quality of optimized code produced by r600-sb due to eliminated false dependencies in some cases. Signed-off-by: Vadim Girlin <[email protected]>
* intel: Remove the last spans code!Eric Anholt2013-04-3011-148/+6
| | | | | | | | The remaining bits happen to do nothing that _swrast_span_render_start()/finish() don't do. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* intel: Move the S8 offset calc function near its remaining usage.Eric Anholt2013-04-303-58/+56
| | | | | | | It's not really span code ever since we stopped using spans for S8. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* intel: Ensure renderbuffers are current when mapping them.Eric Anholt2013-04-301-6/+1
| | | | | | | | | | | | In the case of renering to windows in X, we would render to stale buffers (or not render at all!) if you hit a MapRenderbuffer as the first thing done to your window after new buffers are ready to be collected in DRI2. I think this also covers the weird comment about irb->mt being missing sometimes. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* mesa: Add a clarifying comment about rowStride of compressed textures.Eric Anholt2013-04-301-1/+3
| | | | | | | I always forget how we do this for compressed textures. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* mesa: Remove the Map field from texture images.Eric Anholt2013-04-301-1/+0
| | | | | Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Always use MapTextureImage for mapping textures for swrast.Eric Anholt2013-04-3011-300/+64
| | | | | | | | | | | | | | | Now that everything goes through ImageSlices[], we can rely on the driver's existing texture mapping function. A big block of code goes away on Radeon that looks like it was to deal with the validate that happened at SpanRenderStart, which no longer occurs since we don't need validation for the MapTextureImage hook. v2: Rewrite comment about ImageSlices, fix duplicated swImages, touch up unmap loop. Reviewed-by: Kenneth Graunke <[email protected]> (v1) Reviewed-by: Brian Paul <[email protected]>
* nouveau: Replace swrast_texture_image->Map usage with ->Buffer.Eric Anholt2013-04-301-3/+1
| | | | | | | | This code is trying to deal with providing a map in the case that AllocTexImageBuffer was called, which is hooked up to the swrast variant. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* nouveau: Just use MapTextureImage instead of duplicating the logic.Eric Anholt2013-04-301-81/+18
| | | | | | | | MapTextureImage has the exact same logic, except it can also handle swrast-allocated buffers. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Make a teximage's stored RowStride be in terms of bytes per row.Eric Anholt2013-04-306-9/+30
| | | | | | | | | | | | | For hardware drivers with pitch alignment requirements, a non-power-of-two-sized texture format won't end up being an integer number of pixels per row. Also, avoids having to change our units between MapTextureImage's rowStride and swrast's RowStride. This doesn't fully convert the compressed texel fetch path, but does make sure we don't drop any bits (not that we'd expect to). Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Replace use of teximage Map in 1D/2D paths with ImageSlices[0].Eric Anholt2013-04-303-8/+8
| | | | | | | This gets us ready for the Map field to die. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Replace ImageOffsets with an ImageSlices pointer.Eric Anholt2013-04-3013-185/+123
| | | | | | | | | | | | | This is a step toward allowing drivers to use their normal mapping paths, instead of requiring that all slice mappings come from an aligned offset from the first slice's map. This incidentally fixes missing slice handling in FXT1 swrast. v2: Use slice height helper function. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Reuse _swrast_free_texture_image_buffer from drivers.Eric Anholt2013-04-302-15/+2
| | | | | Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Move ImageOffsets allocation to shared code.Eric Anholt2013-04-304-44/+17
| | | | | Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* swrast: Clean up and explain the mapping process.Eric Anholt2013-04-302-10/+17
| | | | | | | v2: Move slice height calculation to a helper function (recommeded by Brian). Reviewed-by: Kenneth Graunke <[email protected]> (v1) Reviewed-by: Brian Paul <[email protected]>
* swrast: Factor out texture slice counting.Eric Anholt2013-04-301-4/+12
| | | | | | | This function going to get used a lot more in upcoming patches. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* radeon: Remove some dead teximage mapping code.Eric Anholt2013-04-302-52/+0
| | | | | Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* radeon: Add missing swrast field initialization.Eric Anholt2013-04-301-0/+3
| | | | | | | | This is the equivalent of intel's 80513ec8b4c812b9c6249cc5824337a5f04ab34c. Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* r600g/llvm: Fix opencl buildVincent Lejeune2013-04-301-1/+1
|
* Gallium: Use mmap on Haiku for executable memory vs mallocAlexander von Gluck IV2013-04-291-1/+1
| | | | * Haiku now has DEP enabled by default.
* Mapi: Use mmap on Haiku for executable memory vs mallocAlexander von Gluck IV2013-04-291-1/+1
| | | | * Haiku now has DEP enabled by default.
* Mesa: Use mmap on Haiku for executable memory vs mallocAlexander von Gluck IV2013-04-291-1/+1
| | | | * Haiku now has DEP enabled by default.
* r600g/llvm: get use_kill from compiler shaderVincent Lejeune2013-04-304-2/+9
|
* i965/fs: Print out the estimated cycle count in INTEL_DEBUG=wmEric Anholt2013-04-291-0/+5
| | | | | | | This could be used by shader-db for hopefully more accurate regression testing. Reviewed-by: Kenneth Graunke <[email protected]>
* i965/fs: Allow LRPs with uniform registers.Eric Anholt2013-04-293-1/+11
| | | | | | | | | Improves GLB2.7 performance on my HSW by 0.671455% +/- 0.225037% (n=62). v2: Make is_valid_3src() a method of the fs_reg. (recommended by Ken) Reviewed-by: Matt Turner <[email protected]> (v1) Reviewed-by: Kenneth Graunke <[email protected]> (v1)
* intel: Be more conservative in disabling tiling to save memory.Eric Anholt2013-04-291-3/+5
| | | | | | | | Improves GLB2.7 trex performance 1.01985% +/- 0.721366% on my IVB (n=10) and by 3.38771% +/- 0.584241% (n=15) on my HSW, due to a 32x32 ARGB8888 cubemap going from untiled to tiled. Reviewed-by: Daniel Vetter <[email protected]>
* i965: Disable Z16 on contexts that don't require it.Eric Anholt2013-04-291-1/+14
| | | | | | | | | | | | | | It appears that Z16 on Intel hardware is in fact slower than Z24, so people are getting surprisingly hurt when trying to use Z16 as a performance-versus-precision tradeoff, or when they're targeting GLES2 and that's all you get. GL 3.0+ have Z16 on the list of required exact format sizes, but GLES doesn't, so choose the better-performing layout in that case. Improves GLB 2.7 trex performance at 1920x1080 by 10.7% +/- 1.1% (n=3) on my IVB system. Reviewed-by: Kenneth Graunke <[email protected]>
* intel: Report FBO incompleteness causes through GL_ARB_debug_output.Eric Anholt2013-04-291-22/+34
| | | | Reviewed-by: Kenneth Graunke <[email protected]>
* intel: Fold the one last function intel_tex_format.c into the caller.Eric Anholt2013-04-297-27/+10
| | | | Reviewed-by: Kenneth Graunke <[email protected]>
* mesa: Fix error checking for GS UBO getters.Eric Anholt2013-04-291-2/+7
| | | | | These are supposed to be present if both things are available, but we were enabling them if either one was.
* mesa: Add a clarifying comment about EXTRA_ error checking.Eric Anholt2013-04-291-1/+7
|
* mesa: Add an extra clarifying set of braces to getter checking.Eric Anholt2013-04-291-1/+2
| | | | | | For this multi-page single statement, my thought the end was to that the next block was mis-indented, rather than that the dropped indentation actually indicated the end of the loop.
* mesa: Fix error checking for getters consisting of only API versions.Eric Anholt2013-04-291-32/+24
| | | | | | | | | In almost all of our cases, getters that are turned on for only some API variants will have an extension listed as one of the things that can enable it, and thus api_check gets set. For extra_gl30_es3 (used for NUM_EXTENSIONS, MAJOR_VERSION, MINOR_VERSION) on a GL 2.1 context, though, we would check twice, not find either one, but never actually throw the error.
* mesa: Clarify the names of error checking variables for glGet.Eric Anholt2013-04-291-22/+21
| | | | | There's no reason to actually count these things, so the integer ++ behavior was just confusing.
* i915: Add support for GL_EXT_texture_sRGB and GL_EXT_texture_sRGB_decode.Eric Anholt2013-04-293-2/+15
| | | | This brings the driver up to GL 2.1.
* i915: Always enable GL 2.0 support.Eric Anholt2013-04-292-30/+5
| | | | There's no point in shipping a non-GL2 driver today.
* i915: Correctly set the OQ counter bits.Eric Anholt2013-04-292-0/+2
| | | | | | | | | | | | While we may provide the extension, we need to tell applications that they can't actually use it: An implementation can either set QUERY_COUNTER_BITS_ARB to the value 0, or to some number greater than or equal to n. If an implementation returns 0 for QUERY_COUNTER_BITS_ARB, then the occlusion queries will always return that zero samples passed the occlusion test, and so an application should not use occlusion queries on that implementation.
* i965: Move is_math/is_tex/is_control_flow() to backend_instruction.Kenneth Graunke2013-04-296-76/+49
| | | | | | | | | | | | | | | These are entirely based on the opcode, which is available in backend_instruction. It makes sense to only implement them in one place. This changes the VS implementation of is_tex() slightly, which now accepts FS_OPCODE_TXB and SHADER_OPCODE_LOD. However, since those aren't generated in the VS anyway, it should be fine. This also makes is_control_flow() available in the VS. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Eric Anholt <[email protected]>
* draw/so: fix overflow calculationZack Rusin2013-04-271-8/+18
| | | | | | | | | | only report overflow for missing targets if they're actually being used. if the targets are missing but are not being used by any slot in the stream output declaration we should correctly just ignore them. Signed-off-by: Zack Rusin <[email protected]> Reviewed-by: José Fonseca <[email protected]>
* llvmpipe: Fix queries when screen->num_threads == 0.José Fonseca2013-04-291-2/+3
| | | | | | | | | | That is, when llvmpipe is run in single-threaded mode. Trivial. Tested with LP_NUM_THREADS=0 glean --run results --overwrite --quick --tests occluQry
* Revert "st/mesa: add a simple path to BufferData if it only discards buffer ↵José Fonseca2013-04-291-14/+0
| | | | | | | | contents" This reverts commit 5649f886f76023532538b8792605a3578cec1ed1. It causes segfaults when size is zero.
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-24 03:07:48 +0000