summaryrefslogtreecommitdiff
path: root/src/amd/compiler/aco_opcodes.py
AgeCommit message (Expand)AuthorFilesLines
2021-06-09aco: Implement nir_op_sad_u8x4.Timur Kristóf1-2/+6
2021-06-09aco: use ds_read_{u8,u16}_d16Rhys Perry1-8/+8
2021-06-09aco: use v1b/v2b for ds_read_u8/ds_read_u16Rhys Perry1-0/+5
2021-06-08aco/tests: add SDWA testsRhys Perry1-1/+1
2021-06-08aco: add p_extract/p_insertRhys Perry1-0/+8
2021-05-18aco: Implement bvh64_intersect_ray_amd intrinsic.Bas Nieuwenhuizen1-0/+1
2021-03-11aco: add instruction classesRhys Perry1-155/+185
2021-03-11aco: lower p_constaddr into separate instructions earlierRhys Perry1-1/+5
2021-02-17aco: remove special handling of load_helper_invocationDaniel Schürmann1-1/+0
2021-01-22aco: remove Format::{VOP3A,VOP3B}Rhys Perry1-3/+2
2021-01-13aco: create helpers to emit vop3p instructionsDaniel Schürmann1-0/+3
2021-01-13aco: fix VOP3P assembly, VN and validationDaniel Schürmann1-24/+25
2021-01-12aco: simplify and fix operand/definition sizesDaniel Schürmann1-17/+16
2020-11-16aco: don't use SMEM for SSBO storesRhys Perry1-2/+0
2020-11-16aco: remove v_{add,sub,subrev}_u32 on GFX8Samuel Pitoiset1-3/+3
2020-10-28aco: default to a definition size of 32Rhys Perry1-4/+4
2020-10-14aco: use VOP2 for v_cvt_pkrtz_f16_f32 if possibleDaniel Schürmann1-1/+1
2020-10-14aco: use VOP2 version of v_cvt_pkrtz_f16_f32 on GFX_6_7_10Daniel Schürmann1-1/+2
2020-08-04aco: update aco_opcodes.py for GFX10.3Rhys Perry1-2/+13
2020-07-30aco: add framework for unit testingRhys Perry1-0/+1
2020-07-28aco: rework barriers and replace can_reorderRhys Perry1-9/+6
2020-07-21aco: be more careful combining additions that could wrap into loads/storesRhys Perry1-0/+1
2020-07-17aco: fix nir_op_f2f16_rtne with non-default rounding modesRhys Perry1-0/+1
2020-07-13aco: read 0 from inactive lanes when using dppRhys Perry1-1/+1
2020-07-07aco: use VOP2 version of v_mbcnt_hi_u32_b32 on GFX6/7Rhys Perry1-1/+2
2020-06-15aco: use 32-bit inline constants for 16-bit integer instructionsRhys Perry1-7/+13
2020-06-10aco: add Info::{operand_size,definition_size}Rhys Perry1-0/+31
2020-06-02aco/gfx10: Refactor of GFX10 wave64 bpermute.Timur Kristóf1-2/+2
2020-05-02aco: remove use of f-stringsRhys Perry1-1/+1
2020-04-28aco: add various GFX10 int16 opcodesRhys Perry1-2/+11
2020-04-22aco: implement sub-dword swapsRhys Perry1-0/+9
2020-04-22aco: add VOP3P_instructionRhys Perry1-4/+4
2020-01-29aco: Fix the meaning of is_atomic.Timur Kristóf1-6/+6
2020-01-28aco: rework vertex fetching a bitRhys Perry1-3/+1
2020-01-24radv/aco,aco: implement GS on GFX9+Rhys Perry1-1/+3
2020-01-20aco: add new addr64 bit to MUBUF instructions on GFX6-GFX7Samuel Pitoiset1-0/+1
2020-01-14aco: allow input modifiers on v_cndmask_b32Rhys Perry1-1/+5
2020-01-13aco: don't DCE atomics with return valuesRhys Perry1-8/+9
2019-12-07aco: split read/writelane opcode into VOP2/VOP3 version for SI/CIDaniel Schürmann1-2/+4
2019-12-07aco: Initial GFX7 SupportDaniel Schürmann1-53/+52
2019-11-29aco: improve FLAT/GLOBAL schedulingRhys Perry1-0/+1
2019-11-29aco: fix GFX10 opcodes for some global/flat atomicsRhys Perry1-6/+6
2019-10-28aco: Implement subgroup shuffle in GFX10 wave64 mode.Timur Kristóf1-0/+2
2019-10-09aco: move s_andn2_b64 instructions out of the p_discard_ifRhys Perry1-0/+1
2019-09-25aco: check for duplicate opcode numbersRhys Perry1-0/+25
2019-09-25aco: fix opcode for s_mul_hi_i32Rhys Perry1-1/+1
2019-09-25aco: fix v_subrev_co_u32_e64 opcodeRhys Perry1-1/+1
2019-09-25aco: fix GFX9 opcode for v_xad_u32Rhys Perry1-1/+1
2019-09-19aco: Initial commit of independent AMD compilerDaniel Schürmann1-0/+1552
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-13 05:06:39 +0000