aco/gfx11: fix get_gfx11_true16_mask with v_cmp_class_f16
[platform/upstream/mesa.git] / src / amd / compiler / aco_ir.cpp
2023-07-24 Georg Lehmannaco/gfx11: fix get_gfx11_true16_mask with v_cmp_class_f16
2023-06-26 Vitaliy Triang3l... aco: Add Primitive Ordered Pixel Shading scheduling...
2023-06-20 Rhys Perryaco: don't try to form load+store clauses
2023-06-16 Eric Engestromaco: reformat according to its .clang-format
2023-06-01 Georg Lehmannaco: p_start_linear_vgpr doesn't always need exec mask
2023-05-25 Rhys Perryaco: add MIMG_instruction::strict_wqm
2023-05-24 Rhys Perryaco: add get_op_fixed_to_def() helper
2023-05-18 Georg Lehmannaco: use get_operand_size for dpp opt
2023-05-18 Georg Lehmannaco: add helper function for can_use_input_modifiers
2023-05-18 Georg Lehmannaco: add withoutVOP3 helper
2023-05-12 Georg Lehmannaco: use VOP3+DPP
2023-05-10 Timur Kristófaco: Initialize vcmpx field in get_cmp_info.
2023-05-10 Timur Kristófaco: Don't allow any VALU instruction to write m0.
2023-05-04 Rhys Perryaco/ra: create M0-affinities for s_sendmsg
2023-04-12 Timur Kristófac, aco, radv: Clarify LDS size on GFX6, and NGG shaders.
2023-03-30 Georg Lehmannaco/gfx11: allow opsel for VOP12C
2023-03-30 Georg Lehmannaco/ir: copy opsel when converting to DPP
2023-03-28 Georg Lehmannaco/ir: fix copy paste bug in convert_to_SDWA
2023-03-15 Timur Kristófaco: Remove vtx_binding from MUBUF/MTBUF instructions.
2023-03-09 Georg Lehmannaco: copy abs/neg with assignment
2023-03-07 Georg Lehmannaco: use bitfield array helpers for valu modifiers
2023-03-07 Georg Lehmannaco: remove VOP[123C]P? structs
2023-03-07 Georg Lehmannaco: treat VINTERP_INREG as VALU
2023-02-18 Timur Kristófaco/optimizer: Change v_cmp with subgroup invocation...
2023-02-14 Georg Lehmannaco: new 16bit VOP3 opcodes can use opsel
2023-02-10 Georg Lehmannaco: remove stale TODOs about v_interp opsel
2023-01-10 Rhys Perryaco/gfx11: increase vgpr_limit to 256
2023-01-10 Rhys Perryaco/gfx11: allow true 16-bit instructions to access...
2023-01-10 Rhys Perryaco: add more opcodes to can_use_DPP()
2023-01-09 Rhys Perryaco/gfx11: reduce scratch allocation alignment
2023-01-09 Georg Lehmannaco: Mark more instructions as 16bit on GFX10.
2022-12-07 Bas Nieuwenhuizenaco: Add isTrans helper.
2022-11-06 Marek Olšákamd: rename enums ARCTURUS -> MI100, ALDEBARAN -> MI200
2022-11-02 Rhys Perryaco/gfx11: increase gfx1100/gfx1101 physical vgprs
2022-11-02 Illia Abernikhinutils: Merge util/debug.* into util/u_debug.* and remov...
2022-10-25 Daniel Schürmannaco: change thread_local memory resource to pointer
2022-10-24 Timur Kristófaco: Add ACO_DEBUG=novalidateir option.
2022-10-24 Georg Lehmannaco: fmaak/fmamk can't use SDWA.
2022-10-19 Rhys Perryaco: add ACO_DEBUG=force-waitdeps
2022-09-30 Rhys Perryaco/gfx11: deallocate VGPRs at the end of the shader
2022-09-30 Rhys Perryaco: improve wait_imm unpack
2022-09-28 Daniel Schürmannaco: use monotonic_buffer_resource for instructions
2022-09-26 Rhys Perryaco/tests: add GFX11 assembly tests
2022-09-26 Rhys Perryaco: add VINTERP instruction format
2022-09-26 Rhys Perryaco: add LDSDIR instruction format
2022-09-26 Rhys Perryaco: limit GFX11 to 128 VGPRs for now
2022-09-26 Rhys Perryaco: add GFX11 opcode numbers
2022-08-16 Rhys Perryaco: fix assembly of vopc_sdwa writing exec
2022-08-16 Rhys Perryaco: rename is_cmp to is_fp_cmp
2022-07-27 Georg Lehmannaco/ir: Add swapped opcode for v_cmp_u/v_cmp_o.
2022-07-27 Georg Lehmannaco/ir: Add v_cmp_class to get_cmp_info.
2022-07-27 Georg Lehmannaco/ir: Add vcmpx opcode to get_cmp_info.
2022-07-27 Georg Lehmannaco/ir: Fix swapped nle.
2022-07-27 Georg Lehmannaco/ir: Add integer get_cmp_info.
2022-07-27 Georg Lehmannaco/ir: Generalize (un)ordered_swapped.
2022-07-08 Rhys Perryaco: initialize scratch base registers on GFX9-GFX10.3
2022-07-08 Rhys Perryaco: make FLAT_instruction::offset signed
2022-06-27 Daniel Schürmannaco: correctly validate v_fma_mixhi_f16 register assignment
2022-05-31 Rhys Perryaco: disable sdwa on gfx11
2022-05-13 Marek Olšákamd: change chip_class naming to "enum amd_gfx_level...
2022-05-12 Samuel Pitoisetaco: update waitcnt on GFX11
2022-05-12 Samuel Pitoisetaco: update LDS allocation granularity for PS on GFX11
2022-05-12 Samuel Pitoisetaco: do not align VGPRS to 8 or 16 on GFX11
2022-05-11 Dave Airlieaco/radv: convert to aco shader info at the radv level.
2022-05-11 Dave Airlieaco: move to a minimal aco shader info struct.
2022-05-11 Dave Airlieaco: move info pointer to a copy.
2022-03-25 Daniel Schürmannaco: remove 'high' parameter from can_use_opsel()
2022-03-17 Rhys Perryaco: improve support for v_fma_mix
2022-02-11 Daniel Schürmannaco/insert_exec_mask: stay in WQM while helper lanes...
2022-01-31 Rhys Perryaco: preserve pass_flags during format conversions
2022-01-20 Rhys Perryaco: create v_mac_legacy_f32/v_fmac_legacy_f32
2021-12-31 Tatsuyuki Ishiaco: support DPP8
2021-12-31 Daniel Schürmannaco: don't allow SDWA on VOP3P instructions
2021-11-30 Rhys Perryaco: don't create DPP instructions with SGPR operands
2021-11-15 Rhys Perryaco: consider pseudo-instructions reading exec in needs...
2021-10-08 Samuel Pitoisetaco: constify radv_shader_{info,args}
2021-09-02 Daniel Schürmannaco: remove explicit dst_preserve flag
2021-09-02 Daniel Schürmannaco: rewrite SDWA selector
2021-08-23 Daniel Schürmannaco: add instr_is_16bit() helper function
2021-08-23 Daniel Schürmannaco: use VOPC_SDWA on GFX9+
2021-08-19 Rhys Perryaco: move a bunch of helpers into aco_ir.h/aco_ir.cpp
2021-08-19 Rhys Perryaco: add can_use_DPP() and convert_to_DPP()
2021-07-12 Daniel Schürmannaco: Format.
2021-07-12 Daniel Schürmannaco: reorder and cleanup #includes
2021-06-08 Rhys Perryaco: optimize 32-bit extracts and inserts using SDWA
2021-06-08 Rhys Perryaco: disallow SDWA for instructions with 64-bit definit...
2021-05-18 Timur Kristófaco: Determine whether a few more instructions need...
2021-05-10 Rhys Perryaco: group loads from the same vertex binding into...
2021-04-21 Rhys Perryaco: add and use Program::progress
2021-03-19 Rhys Perryaco: add ACO_DEBUG=liveinfo
2021-03-11 Rhys Perryaco: add ACO_DEBUG=perfinfo
2021-03-11 Rhys Perryaco: move wait_imm to aco_ir.h
2021-03-02 Rhys PerryRevert "radv,aco: allow unaligned LDS access on GFX9+"
2021-02-17 Rhys Perryradv,aco: allow unaligned LDS access on GFX9+
2021-02-15 Rhys Perryaco: add DeviceInfo
2021-02-15 Rhys Perryaco: consider that GFX10.3 allocates LDS in 1024 byte...
2021-02-15 Rhys Perryradv,aco: add radv_nir_compiler_options::wgp_mode
2021-02-15 Rhys Perryaco: add Program::wgp_mode
2021-02-15 Rhys Perryaco: fix waves calculation for wave32
2021-02-12 Daniel Schürmannaco: use VCC as regular SGPR pair on GFX10
next