summaryrefslogtreecommitdiff
path: root/src/nouveau
Commit message (Expand)AuthorAgeFilesLines
* treewide: Stop lowering legacy atomicsAlyssa Rosenzweig2023-05-161-1/+0
* nv50/ir: Use unified atomicsKarol Herbst2023-05-121-208/+74
* nouveau/nir: Set isSigned on all atomic_imax/iminM Henning2023-05-081-0/+6
* nouveau/nir: image_samples/size don't have coordinatesFaith Ekstrand2023-05-031-0/+2
* nv50/codegen: Set lower_uniforms_to_uboM Henning2023-05-031-11/+2
* nouveau/codegen: Check nir_dest_num_componentsM Henning2023-04-251-1/+1
* nv50/ir: ignore CL system valuesKarol Herbst2023-04-141-3/+9
* nouveau: Enable frexp lowering in the backend.Emma Anholt2023-04-061-0/+2
* nouveau: Add missing nir_opt_algebraic_late.Emma Anholt2023-04-061-0/+2
* nouveau: Make getSize return unsigned intMichel Dänzer2023-03-174-6/+6
* nir: add assertions that loops don't have a Continue ConstructDaniel Schürmann2023-02-211-0/+1
* nir: Eliminate nir_op_f2bIan Romanick2023-02-031-6/+0
* nir: Eliminate nir_op_i2bIan Romanick2022-12-141-10/+2
* drm-shim/nouveau: fix the shim to work with nvif ioctl.Dave Airlie2022-11-092-2/+204
* gv100/ir: noop OP_BAR for nowBen Skeggs2022-11-091-0/+6
* nv50/ir: Support fmulz and ffmazYusuf Khan2022-11-081-9/+16
* nv50/ir: add prefer_nir flag for getting compiler optionsYusuf Khan2022-11-082-27/+72
* nv50/ir/nir: ignore sampler for TXF/TXQ ops.Joan Bruguera2022-11-021-0/+2
* nouveau/codegen: Support bindless texture queriesJason Ekstrand2022-11-011-3/+8
* nir/lower_idiv: Remove imprecise_32bit_loweringAlyssa Rosenzweig2022-10-271-1/+0
* nv50/ir: nir_op_b2i8 and nir_op_b2i16Yusuf Khan2022-10-271-0/+2
* nv50: call nir_lower_flrpThomas Debesse2022-10-101-0/+6
* nouveau: const cleanupAdam Jackson2022-09-164-8/+8
* nouveau: Fix compiler warnings about silly address checks in ir_print.Emma Anholt2022-09-131-11/+5
* nv50/ir: handle U8/U16 integers converting to U64Danilo Krummrich2022-09-091-1/+3
* nv50/ir: handle S8/S16 integers converting to S64Danilo Krummrich2022-09-091-5/+25
* nv50/ir: split and cvt 64bit integers for {i,u}2{i,u}{8,16}Danilo Krummrich2022-09-091-4/+22
* nv50/ir: add intermediate conversion for f2{i,u}{8,16}Danilo Krummrich2022-09-091-1/+16
* nv50/ir/nir: always round towards zero for f2i/f2uDanilo Krummrich2022-09-091-2/+3
* nv50/ir/nir: convert 8/16 bit src to 32 bit for {i,u}2f64Danilo Krummrich2022-09-091-0/+9
* nv50/ir: add isUnsignedIntType() and isIntType() helpersDanilo Krummrich2022-09-091-0/+10
* nv50/ir/nir: avoid 8/16 bit dest regs for OP_MOVDanilo Krummrich2022-09-091-2/+2
* nv50/ir/nir: convert to 32 bit for all OP_SET opcodesDanilo Krummrich2022-09-091-0/+60
* nv50/ir/nir: add conversion ops for bit width < 32Danilo Krummrich2022-09-091-0/+52
* nv50/ir: fix OP_UNION resolving when used for vector valuesKarol Herbst2022-09-081-1/+1
* nv/nir: Set ssbo CacheMode from intrinsic accessM Henning2022-09-041-5/+14
* nv50/nir: A group barrier is CTA-level not global-levelPierre Moreau2022-08-231-1/+1
* nv50/ir: Avoid generating splits of splitsPierre Moreau2022-08-231-0/+8
* nv50/ra: Fix the offset computation for compoundsPierre Moreau2022-08-231-1/+3
* nv50/peephole: Disallow combining sub 4-byte ld/st for nowPierre Moreau2022-08-231-0/+6
* nv50/ir: Handle non-32-bit values when cst folding SPLITPierre Moreau2022-08-231-1/+2
* nir: Split usub_sat lowering flag from uadd_sat.Emma Anholt2022-07-221-0/+1
* nouveau: Drop C++03 compat codeM Henning2022-07-058-64/+15
* nv50/ir/ra: Fix copying compound for movesConnor Abbott2022-06-301-21/+39
* nouveau/nir: Implement mul_zero_wins behavior for use_legacy_math_rules.Emma Anholt2022-06-101-0/+12
* gallium: Rename MUL_ZERO_WINS to LEGACY_MATH_RULES.Emma Anholt2022-06-101-1/+1
* nouveau/nvc0: disable GLSL IR loop unrollingTimothy Arceri2022-06-041-8/+35
* nouveau/nv50: disable GLSL IR loop unrollingTimothy Arceri2022-06-042-8/+16
* nouveau/codegen: drop gallium headers from the interface.Dave Airlie2022-06-033-13/+10
* nouveau/codegen: drop all ubytes from codegen.Dave Airlie2022-06-033-6/+6
View Lorry Controller Status