Searched refs:aco (Results 1 – 25 of 77) sorted by relevance
1234
33 [aco::statistic_hash] = {"Hash", "CRC32 hash of code and constant data"},34 [aco::statistic_instructions] = {"Instructions", "Instruction count"},35 [aco::statistic_copies] = {"Copies", "Copy instructions created for pseudo-instructions"},36 [aco::statistic_branches] = {"Branches", "Branch instructions"},37 [aco::statistic_cycles] = {"Busy Cycles", "Estimate of busy cycles"},38 …[aco::statistic_vmem_clauses] = {"VMEM Clause", "Number of VMEM clauses (includes 1-sized clauses)…39 …[aco::statistic_smem_clauses] = {"SMEM Clause", "Number of SMEM clauses (includes 1-sized clauses)…40 [aco::statistic_vmem_score] = {"VMEM Score", "Average VMEM def-use distances"},41 [aco::statistic_smem_score] = {"SMEM Score", "Average SMEM def-use distances"},42 [aco::statistic_sgpr_presched] = {"Pre-Sched SGPRs", "SGPR usage before scheduling"},[all …]
27 namespace aco { namespace85 program->statistics[aco::statistic_hash] = util_hash_crc32(code.data(), code.size() * 4); in collect_postasm_stats()
92 'aco',109 # Also link with aco
42 namespace aco { namespace
65 extern std::unique_ptr<aco::Program> program;66 extern aco::Builder bld;67 extern aco::Temp exec_input;68 extern aco::Temp inputs[16];71 void create_program(enum chip_class chip_class, aco::Stage stage,76 void finish_program(aco::Program *program);82 void writeout(unsigned i, aco::Temp tmp=aco::Temp(0, aco::s1));
32 using namespace aco;83 aco::init_program(program.get(), stage, &info, chip_class, family, &config); in create_program()149 if (aco::validate_ir(program.get())) in finish_validator_test()158 if (!aco::validate_ir(program.get())) { in finish_opt_test()162 aco::optimize(program.get()); in finish_opt_test()163 if (!aco::validate_ir(program.get())) { in finish_opt_test()173 aco::lower_to_hw_instr(program.get()); in finish_to_hw_instr_test()
27 using namespace aco;
34 - aco: Dead Rising 4 crashes in lower_to_hw_instr() on GFX6-GFX772 - aco: fix unconditional demote_to_helper73 - aco: rework lower_to_cssa()173 - aco: set vm for pos0 exports on GFX10174 - aco: fix imageSize()/textureSize() with large buffers on GFX8175 - aco: fix uninitialized data in the binary176 - aco: set exec_potentially_empty for demotes177 - aco: disable add combining for ds_swizzle_b32178 - aco: don't DCE atomics with return values179 - aco: check if multiplication/clamp is live when applying output[all …]
62 - aco: sun flickering with Assassins Creeds Origins64 - aco: wrong geometry with Assassins Creed Origins on GFX678 - aco: implement GFX6 support85 - aco: Dead Rising 4 crashes in lower_to_hw_instr() on GFX6-GFX792 - [Navi/aco] Guild Wars 2 - ring gfx timeout with commit 3bca0af293 - [radv/aco] Regression is causing a soft crash in The Witcher 3158 - radv/aco Jedi Fallen Order hair rendering buggy693 - aco: Constify radv_nir_compiler_options in isel694 - aco: Use radv_shader_args in aco_compile_shader()695 - aco: Split vector arguments at the beginning[all …]
230 - aco: sun flickering with Assassins Creeds Origins232 - aco: wrong geometry with Assassins Creed Origins on GFX6265 - aco: Minor optimization in spill_ctx constructor266 - aco: pass vars by const &1185 - aco: fix image_atomic_cmp_swap1195 - aco: add comparison operators for PhysReg1196 - aco: add sub-dword regclasses1197 - aco: refactor regClass setup for subdword VGPRs1198 - aco: validate p_create_vector with subdword elements properly1199 - aco: validate register alignment of subdword operands and definitions[all …]
1233 - aco: either copy-propagate or inline create_vector operands1234 - aco: coalesce parallelcopies during register allocation1240 - aco: fix WQM coalescing1241 - aco: restrict copying of create_vector operands to GFX9+1242 - aco: don't move create_vector subdword operands to unsupported register offsets1243 - aco: fix corner case in register allocation1244 - aco: don't allow unaligned subdword accesses on GFX6/71245 - aco: fix register assignment for p_create_vector on GFX6/71246 - aco: simplify statistics collection for copies1247 - aco: use full-register instructions to implement subdword packing on GFX6/7[all …]
139 - aco: set has_divergent_branch for discards in loops140 - aco: handle missing second predecessors at merge block phis141 - aco: skip NIR in unreachable merge blocks142 - aco: improve check for unreachable loop continue blocks143 - aco: emit IR in IF's merge block instead if the other side ends in a145 - aco: fix boolean undef regclass147 - aco: implement 64-bit VGPR constant copies in handle_operands()
36 - radv/aco: Vertex explosion on RPCS3116 - aco: add missing SCC clobber in get_buffer_size117 - aco: update phi_map in add_subdword_operand()118 - aco: ignore the ACO-inserted continue in create_continue_phis()130 - aco: fix determining if LOD is zero for nir_texop_txf/nir_texop_txs147 - aco/isel: Always export position data from VS/NGG
48 - New compiler backend "ACO" for RADV (RADV_PERFTEST=aco)58 - radv/aco Jedi Fallen Order hair rendering buggy484 - amd: Build aco only if radv is enabled782 - aco: Initial commit of independent AMD compiler783 - radv/aco: Setup alternate path in RADV to support the experimental786 - radv/aco: enable VK_EXT_shader_demote_to_helper_invocation788 - aco: only emit waitcnt on loop continues if we there was some load or793 - radv/aco: Don't lower subtractions794 - aco: call nir_opt_algebraic_late() exhaustively796 - aco: re-use existing phi instruction when lowering boolean phis[all …]
77 - \[aco\] problem compiling compute pipeline95 - radv/aco: Vertex explosion on RPCS3125 - radv, aco: dEQP-VK.glsl.atomic_operations.*_fragment_reference regressed874 - aco: Add VK_KHR_shader_terminate_invocation support.1178 - aco: execute branch instructions in WQM if necessary1180 - aco/isel: refactor code and remove unnecessary v_mov1181 - aco/isel: refactor emit_vop3a_instruction() to handle 2 operand instructions1183 - aco: propagate SGPRs into VOP1 instructions early.1184 - aco: expand create_vector more carefully w.r.t. subdword operands1185 - aco: use p_create_vector for nir_op_pack_half_2x16[all …]
171 - aco: fix operand to scc when selecting SGPR ufind_msb/ifind_msb172 - aco: ensure predecessors' p_logical_end is in WQM when a p_phi is in174 - aco: run p_wqm instructions in WQM175 - aco: don't consider loop header blocks branch blocks in177 - aco: don't always add logical edges from continue_break blocks to179 - aco: fix target calculation when vgpr spilling introduces sgpr
57 - aco/ra: use get_reg_specified() for p_extract_vector90 - aco: use UINT64_C on 64 bit constant arguments132 - aco: don't assume src=lower when splitting self-intersecting copies143 - aco: fix combining max(-min(a, b), c) if a or b uses the neg modifier160 - aco: Use program->num_waves as maximum in scheduler.
117 - aco: fix incorrect address calculation for load_barycentric_at_sample120 - aco: fix unreachable() for uniform 8/16-bit nir_op_mov from VGPR128 - aco: fix creating the dest vector when 16-bit vertex fetches are splitted129 - radv/llvm,aco: always split typed vertex buffer loads on GFX6 and GFX10+
55 - aco: fix DCE of rematerializable phi operands56 - aco/spill: only prevent rematerializable vars from being DCE'd if they haven't been renamed57 - aco/ra: fix phi operand renaming118 - aco: add block to worklist in mark_block_wqm()
156 - aco: fix interaction with 3f branch workaround and p_constaddr157 - aco: consider SDWA during value numbering158 - aco: check instruction format before waiting for a previous SMEM160 - aco: preserve more fields when combining additions into SMEM169 - aco: fix register allocation for subdword instructions on GFX10
115 - aco: fix byte_align_scalar for 3 dword vectors120 - aco: Fix emit_boolean_exclusive_scan in wave32 mode.125 - aco: Fix integer overflows when emitting parallel copies during RA
66 - aco/optimizer: don't propagate subdword temps of different size67 - aco/optimizer: don't copy-prop logical phis156 - aco: fix convert_to_SDWA() check in add_subdword_definition()157 - aco: don't consider a phi trivial if same's register doesn't match the def
59 - aco: fix scratch loads which cross element_size boundaries60 - aco: don't split store data if it was already split into more elements61 - aco: prevent infinite recursion in RA for subdword variables
92 - aco: Fix signed-vs-unsigned warning.159 - aco: keep track of which events are used in a barrier160 - aco: fix carry-out size for wave32 v_add_co_u32_e64
16 de_AT{"alemán austríaco"}50 syr{"siríaco"}