Lines Matching full:aco

166 - [RADV/ACO] Star Citizen Flickeriing Lights
251 - radv/aco: xfb bug
253 - Occasional flicker corruption in Rage 2, e.g. after loading, with ACO on RX 5700 XT
259 - radv,aco: Regression with global atomics with negative offsets
1398 - aco: remove 'max_waves' and use 'num_waves' to adjust for LDS and workgroup size
1399 - aco: split num_waves adjustment into separate function
1400 - aco: adjust num_waves for LDS before scheduling
1401 - aco/optimizer: prevent any overflow between SGPR and const offset on MUBUF
1402 - aco/ra: fix live-range splits of phi definitions
1403 - aco: don't re-create vectors for load_barycentric_* intrinsics
1404 - aco/ra: improve split_vector register assignment if the operand is not killed
1405 - aco/optimizer: re-combine and copy-propagate p_create_vector(p_split_vector)
1408 - aco: fix spilling of phis without temp operands
1412 - aco: correctly validate v_fma_mixhi_f16 register assignment
1421 - aco: fix packed 16bit fneg/fsat optimization
1422 - aco: avoid unnecessary copies in emit_wqm()
1423 - aco: Avoid live-range splits in Exact mode
1424 - aco: fix assertion in insert_exec_mask
1526 - aco: move info pointer to a copy.
1527 - aco: move to a minimal aco shader info struct.
1528 - aco/info: reduce the gs ring info to what is needed.
1529 - aco: remove radv specific streamout info
1530 - aco: convert vs and so info over to aco structs.
1531 - aco/info: add some more fields.
1532 - aco/radv: convert to aco shader info at the radv level.
1533 - aco: remove radv vs prolog key from aco internals.
1534 - aco: drop unused radv include
1536 - aco/radv: drop radv_nir_compiler_options from aco.
1554 - aco: refactor the radv binary builder out of the core aco fn.
1555 - aco/radv: provide a callback from aco shader building to build binary
1556 - aco/radv: provide a vs prolog callback from aco to radv.
1557 - aco: drop radv_shader.h include
2241 - aco: Remove some old comments in aco_opcodes.py.
2242 - aco: Reuse previous -1 result in find_msb to avoid using VOP3.
2244 - aco: Implement image_store d16.
2245 - aco: Implement image_load d16.
2248 - aco: Remove D16 zero components from image stores.
2256 - aco: Implement uclz.
2257 - aco: Implement usub_sat.
2258 - aco: Implement isub_sat.
2266 - radv, aco: Packed usub_sat/isub_sat.
2268 - radv, aco: Lower txf offset in NIR.
2270 - radv, aco: Round texture array layer in NIR.
2273 - aco: Print r128/a16 MIMG bits separately.
2274 - aco: Remove r128_a16 MIMG builder option.
2279 - aco/ra: Static assert that changing instruction type to VOP2 is valid.
2280 - aco/ra: Move mac encoding optimization to its own function.
2281 - aco/ir: Pad SOP2 and SOPC to the same size as SOPK.
2282 - aco/ra: Optimize some SOP2 instructions with literal to SOPK.
2283 - aco/optimizer: Convert s_add_u32 with literals to s_add_i32 if carry is not used.
2284 - aco/optimizer: Optimize SOPC with literal to SOPK.
2286 - aco: Fix swapping sources in SOPC -> SOPK optimization.
2287 - aco/assembler: Fix s_bitreplicate_b64_b32 on GFX9.
2294 - radv,aco: Don't lower and vectorize 16bit iabs.
2296 - radv, aco: Don't lower 16bit isign.
2302 - aco: Add G16 opcodes.
2303 - aco: Support 16bit sources for texture ops.
2305 - aco: Add a note about G16 without A16 to README-ISA.md.
2306 - aco/ir: Generalize (un)ordered_swapped.
2307 - aco/ir: Add integer get_cmp_info.
2308 - aco/ir: Fix swapped nle.
2309 - aco/ir: Add vcmpx opcode to get_cmp_info.
2310 - aco/ir: Add v_cmp_class to get_cmp_info.
2311 - aco/ir: Add swapped opcode for v_cmp_u/v_cmp_o.
2314 - aco: Force tex operand to have the correct sub dword size before packing.
3276 - aco: Use movk for AddressHi bits in vertex prolog
3461 - radv,aco: Use ray_launch_size_addr
3513 - aco: Remove dead nir_intrinsic_load_vertex_id case
5165 - aco: fix disassembly of SMEM with both SGPR and constant offset
5179 - aco/ra: fix vgpr_limit
5180 - aco: ensure that definitions fixed to operands have matching regclasses
5185 - aco: split load_sbt_amd result
5186 - aco: skip s_barrier if TCS patches are within subgroup
5188 - aco: fix cmpswap global atomic definition on GFX6
5197 - aco: fix p_constaddr with a non-zero offset
5198 - aco/tests: add test for p_constaddr with a non-zero offset
5199 - aco: only add/subtract low bits of program addresses
5200 - aco/ra: fix usage of invalidated iterator
5201 - aco: use p_parallelcopy for uniform reduction with zero source
5202 - aco: clarify a portion of do_pack_2x16
5203 - aco/tests: disable regalloc.subdword_alloc.reuse_16bit_operands on GFX11
5204 - aco/tests: fix gfx11 variants printed as gfx12
5205 - aco: disable sdwa on gfx11
5206 - aco: use v_perm_b32 to copy 0xff00/0x00ff/0xff/0x00
5207 - aco/tests: update for GFX11's removal of SDWA
5208 - radv,aco: disable shader cache when ACO debug options are used
5209 - aco: fix SMEM load_global_amd with non-zero offset
5210 - aco: fix SMEM load_global with VGPR address and non-zero offset
5211 - aco: consider fma with multiplication by power-of-two unfused
5216 - aco: update SMEM offset workaround for LLVM 15
5217 - aco: don't skip VS->TCS barrier if TCS output vertices doesn't match input
5218 - aco: cleanup force-waitcnt output
5219 - aco/ra: update register file when updating phi definition
5221 - aco: add and use constantValue16()
5222 - aco: don't accept med3 opcodes in get_minmax_info()
5223 - aco: include _e64 variants of 16-bit min/max in minmax optimizations
5224 - aco: ignore precise flag when optimizing integer clamps
5225 - aco: fix single-alignbyte do_pack_2x16() path with fp inline constants
5226 - aco: fix propagate_constants_vop3p with integer vop3p and 16-bit constants
5227 - aco: don't use opsel to fold constants into dot accumulation sources
5228 - aco: fix redirect combine in propagate_constants_vop3p() with negatives
5229 - aco: try sign-extending or shifting constants in propagate_constants_vop3p
5230 - aco: don't use 32-bit fp inline constants for fp16 vop3p literals
5231 - aco/tests: add vop3p constant combine tests
5235 - radv,aco,ac/llvm: use nir_op_f{sin,cos}_amd
5236 - aco: make flat access latency match mtbuf/mubuf/mimg
5237 - aco: include flat-like in vmem clause statistics
5238 - aco: make FLAT_instruction::offset signed
5239 - aco: improve support for scratch_* instructions
5240 - aco: combine additions and constants into scratch load/store
5241 - aco: handle subtractions in parse_base_offset
5242 - aco: refactor VGPR spill/reload lowering
5243 - aco: avoid WAW hazard with BVH MIMG and other VMEM
5244 - aco: include scratch/global in VMEM WAW optimization
5245 - aco: treat flat-like as vmem in some scheduling heuristics
5246 - aco: initialize scratch base registers on GFX9-GFX10.3
5247 - aco: use scratch_* for scratch load/store on GFX9+
5248 - aco: use scratch_* for VGPR spill/reload on GFX9+
5251 - aco: remove unnecessary exp_cnt increments
5252 - aco: fix LdsBranchVmemWARHazard with 2+ branch chains
5253 - aco: fix hash statistic
5256 - aco: fix 16-bit VS inputs
5257 - aco: don't expand vec3 VS input load to vec4 on GFX6
5258 - aco: add SCC clobber in build_cube_select
5385 - aco: fix load_barycentric_at_{sample,offset} on GFX6-7
5409 - aco: recognize GFX11 in few places
5410 - aco: do not align VGPRS to 8 or 16 on GFX11
5411 - radv,aco: add support for packed threadID VGPRs on GFX11
5412 - radv,aco: use the new TCS WaveID SGPR to compute vs_rel_patch_id on GFX11
5413 - aco: export MRT0 instead of NULL on GFX11
5414 - aco: do not set DLC for loads on GFX11
5415 - aco: do not set GLC stores on GFX11
5416 - aco: update LDS allocation granularity for PS on GFX11
5417 - aco: update waitcnt on GFX11
5418 - aco: do not set RESOURCE_LEVEL for buffer descriptors on GFX11
5419 - aco: do not set COMPR for exports but use 0x3 channel mask on GFX11
5420 - radv,aco: export alpha-to-coverage via MRTZ on GFX11
5421 - aco: only retrieve the scratch offset when it's declared
5422 - radv,aco: do not implicitly export the primitive ID for mesh shaders
5423 - aco: remove unreachable code about viewport index/layer and mesh shaders
5424 - radv,aco: add a workaround for binding 2D views of a 3D image on GFX9
5428 - aco: do not emit the primitive ID twice for NGG VS or TES with GS
5445 - aco: use ac_is_llvm_processor_supported() for checking LLVM asm support
5452 - aco: remove unnecessary intrinsics that are lowered at the ABI level
5534 - aco: fix validation of SOP1 instructions without definitions
5535 - radv,aco: rename radv_prolog_binary/radv_shader_prolog to shader_part
5551 - aco: fix load_barycentric_at_sample without MSAA
5558 - radv,aco: track if a fragment shader needs an epilog
5559 - radv,aco: introduce {radv,aco}_ps_epilog_key
5562 - aco: add new pseudo instruction p_jump_to_epilog
5563 - aco: emit p_jump_to_epilog if the main fragment shader has an epilog
5564 - aco: do not abort if the FS doesn't export anything but has an epilog
5565 - aco: prevent adding DONE/VM to the last export if the FS has an epilog
5566 - aco: add aco_postprocess_shader() helper
5567 - radv,aco: rename radv_aco_build_prolog to radv_aco_build_shader_part
5568 - aco: refactor export_fs_mrt_color() for PS epilogs preparation
5569 - aco: add support for compiling PS epilogs
5572 - aco: requires Exact for p_jump_to_epilog
5816 - aco: Remove now-superfluous intrinsics.
5840 - aco: Fix scratch with task shaders.
5849 - aco: Remove trailing whitespace.
5905 - aco: Remove hack for primitive ID export.
5916 - aco: Fix p_init_scratch for task shaders.
5965 - aco: Initialize spill_ctx members in constructor.