Public Git Hosting - llvm-project.git/search

2023-03-08	Stanislav Mekhanoshin	[AMDGPU] Skip buffer_wbl2 before atomic fence acquire	commit \| commitdiff \| tree
2023-02-14	Stanislav Mekhanoshin	[AMDGPU] Do not apply schedule metric for regions with...	commit \| commitdiff \| tree
2023-02-08	Stanislav Mekhanoshin	[AMDGPU] Do not exapnd fp atomics on gfx940	commit \| commitdiff \| tree
2023-02-08	Stanislav Mekhanoshin	[AMDGPU] Update atomic tests. NFC.	commit \| commitdiff \| tree
2023-02-05	Stanislav Mekhanoshin	[AMDGPU] Fix liveness in the SIOptimizeExecMaskingPreRA.cpp	commit \| commitdiff \| tree
2023-01-26	Stanislav Mekhanoshin	[AMDGPU] Split dot7 feature	commit \| commitdiff \| tree
2023-01-25	Stanislav Mekhanoshin	[AMDGPU] Remove predicates from real dot instructions...	commit \| commitdiff \| tree
2023-01-24	Stanislav Mekhanoshin	[AMDGPU] Split dot8 feature	commit \| commitdiff \| tree
2023-01-24	Stanislav Mekhanoshin	[AMDGPU] Remove dot1 and dot6 features from clang for...	commit \| commitdiff \| tree
2023-01-24	Stanislav Mekhanoshin	[AMDGPU] Add missing gfx11 tests in the directive-amdgcn...	commit \| commitdiff \| tree
2023-01-23	Stanislav Mekhanoshin	[AMDGPU] Use more consistemt way to avoid overflow...	commit \| commitdiff \| tree
2023-01-23	Stanislav Mekhanoshin	[AMDGPU] Tune scheduler on GFX10 and GFX11 for regions...	commit \| commitdiff \| tree
2023-01-19	Stanislav Mekhanoshin	[AMDGPU] Treat WMMA the same as MFMA for sched_barrier	commit \| commitdiff \| tree
2023-01-19	Stanislav Mekhanoshin	[AMDGPU] Introduce separate register limit bias in...	commit \| commitdiff \| tree
2023-01-11	Stanislav Mekhanoshin	[AMDGPU] Mark wmma intrinsics as source of divergence	commit \| commitdiff \| tree
2023-01-09	Stanislav Mekhanoshin	[AMDGPU] Cast sub-dword elements to i32 in concat_vectors	commit \| commitdiff \| tree
2023-01-09	Stanislav Mekhanoshin	[AMDGPU] More tests for vector_shuffle.packed.ll. NFC.	commit \| commitdiff \| tree
2022-11-29	Stanislav Mekhanoshin	[AMDGPU] Fine tune LDS misaligned access speed	commit \| commitdiff \| tree
2022-11-28	Stanislav Mekhanoshin	[LoadStoreVectorizer] Consider if operation is faster...	commit \| commitdiff \| tree
2022-11-17	Stanislav Mekhanoshin	[AMDGPU] Allow finer grain control of an unaligned...	commit \| commitdiff \| tree
2022-09-28	Stanislav Mekhanoshin	[AMDGPU] Move SIModeRegisterDefaults to SI MFI	commit \| commitdiff \| tree
2022-09-26	Stanislav Mekhanoshin	[AMDGPU] Move SIModeRegisterDefaults to SI MFI	commit \| commitdiff \| tree
2022-09-16	Stanislav Mekhanoshin	[AMDGPU] Fix runline for windows in sdag-print-divergence...	commit \| commitdiff \| tree
2022-09-16	Stanislav Mekhanoshin	[SDAG] Print divergence in SDNode::dump	commit \| commitdiff \| tree
2022-09-16	Stanislav Mekhanoshin	[AMDGPU] Added __builtin_amdgcn_ds_bvh_stack_rtn	commit \| commitdiff \| tree
2022-09-15	Stanislav Mekhanoshin	Fix crash while printing MMO target flags	commit \| commitdiff \| tree
2022-09-07	Stanislav Mekhanoshin	[AMDGPU] Fix liveness verifier error in hazard recognizer	commit \| commitdiff \| tree
2022-09-07	Stanislav Mekhanoshin	[AMDGPU] W/a hazard if 64 bit shift amount is a highest...	commit \| commitdiff \| tree
2022-08-30	Stanislav Mekhanoshin	[AMDGPU] Limit TID / wavefrontsize uniformness to 1D...	commit \| commitdiff \| tree
2022-08-27	Stanislav Mekhanoshin	[AMDGPU] Detect uniformness of TID / wavefrontsize	commit \| commitdiff \| tree
2022-08-16	Nicolas Miller	Fix subrange liveness checking at rematerialization	commit \| commitdiff \| tree
2022-07-27	Stanislav Mekhanoshin	[AMDGPU] Consider S_SETPRIO a scheduling boundary	commit \| commitdiff \| tree
2022-07-27	Stanislav Mekhanoshin	[AMDGPU] Precommit s_setprio scheduling test. NFC.	commit \| commitdiff \| tree
2022-07-27	Stanislav Mekhanoshin	Allow data prefetch into non-default address space	commit \| commitdiff \| tree
2022-07-18	Stanislav Mekhanoshin	[AMDGPU] Support for gfx940 fp8 smfmac	commit \| commitdiff \| tree
2022-07-18	Stanislav Mekhanoshin	[AMDGPU] Support for gfx940 fp8 mfma	commit \| commitdiff \| tree
2022-07-18	Stanislav Mekhanoshin	[AMDGPU] Support for gfx940 fp8 conversions	commit \| commitdiff \| tree
2022-06-28	Stanislav Mekhanoshin	[AMDGPU] Relax verification of soffset in scalar stores	commit \| commitdiff \| tree
2022-06-13	Stanislav Mekhanoshin	[AMDGPU] Use null for dead sdst operand	commit \| commitdiff \| tree
2022-06-13	Stanislav Mekhanoshin	[AMDGPU] Define SGPR_NULL64 register. NFCI.	commit \| commitdiff \| tree
2022-06-13	Stanislav Mekhanoshin	[AMDGPU] Make temp vgpr selection stable in indirectCopyToAGPR	commit \| commitdiff \| tree
2022-06-09	Stanislav Mekhanoshin	[AMDGPU] Use v_mad_u64_u32 for IMAD32	commit \| commitdiff \| tree
2022-06-09	Stanislav Mekhanoshin	[AMDGPU] Fix hazard handling of v_cmpx to permlane	commit \| commitdiff \| tree
2022-06-01	Stanislav Mekhanoshin	[AMDGPU] Change GISel error handling for TFE on GFX90A	commit \| commitdiff \| tree
2022-05-31	Stanislav Mekhanoshin	[AMDGPU] Fix image opcodes GlobalISel on gfx90a.	commit \| commitdiff \| tree
2022-05-24	Stanislav Mekhanoshin	[AMDGPU] Disable newly added gfx90a global isel image...	commit \| commitdiff \| tree
2022-05-24	Stanislav Mekhanoshin	[AMDGPU] Enforce alignment of image vaddr on gfx90a	commit \| commitdiff \| tree
2022-05-17	Stanislav Mekhanoshin	[AMDGPU] Add llvm.amdgcn.global.load.lds intrinsic	commit \| commitdiff \| tree
2022-05-17	Stanislav Mekhanoshin	[AMDGPU] Enable FLAT LDS DMA on gfx9/10 before gfx940	commit \| commitdiff \| tree
2022-05-17	Stanislav Mekhanoshin	[AMDGPU] Add test for no waitcnt before issuing LDS...	commit \| commitdiff \| tree
2022-05-17	Stanislav Mekhanoshin	[AMDGPU] Add intrinsics llvm.amdgcn.{raw\|struct}.buffer...	commit \| commitdiff \| tree
2022-05-16	Stanislav Mekhanoshin	[AMDGPU] Revert wide LDS DMA support.	commit \| commitdiff \| tree
2022-05-04	Stanislav Mekhanoshin	[AMDGPU] Handle LDS DMA and LDS_DIRECT hazards	commit \| commitdiff \| tree
2022-04-29	Stanislav Mekhanoshin	[AMDGPU] Produce waitcounts for LDS DMA	commit \| commitdiff \| tree
2022-04-27	Stanislav Mekhanoshin	[AMDGPU] Remove now unused variable HasLdsModifier...	commit \| commitdiff \| tree
2022-04-27	Stanislav Mekhanoshin	[AMDGPU] Add both mayLoad and mayStore to MUBUF LDS...	commit \| commitdiff \| tree
2022-04-27	Stanislav Mekhanoshin	[AMDGPU] Remove vdata from buffer to lds load	commit \| commitdiff \| tree
2022-04-26	Stanislav Mekhanoshin	[AMDGPU] Properly mark MUBUF and FLAT LDS DMA instructions...	commit \| commitdiff \| tree
2022-04-21	Stanislav Mekhanoshin	[AMDGPU] Refine 64 bit misaligned LDS ops selection	commit \| commitdiff \| tree
2022-04-20	Stanislav Mekhanoshin	[AMDGPU] Remove obsolete hack from allowsMisalignedMemoryAcc...	commit \| commitdiff \| tree
2022-04-18	Stanislav Mekhanoshin	[AMDGPU] Fix comment type in the DSInstructions.td...	commit \| commitdiff \| tree
2022-04-14	Stanislav Mekhanoshin	[AMDGPU] Remove redundand RequiredAlignment assignment...	commit \| commitdiff \| tree
2022-04-13	Stanislav Mekhanoshin	[AMDGPU] Increate hazard for store dwordx3/4 to 2 waitstates...	commit \| commitdiff \| tree
2022-04-12	Stanislav Mekhanoshin	[AMDGPU] Split unaligned 4 DWORD DS operations	commit \| commitdiff \| tree
2022-04-12	Stanislav Mekhanoshin	[AMDGPU] Update ds-alignment.ll test checks. NFC.	commit \| commitdiff \| tree
2022-04-12	Stanislav Mekhanoshin	[AMDGPU] Split unaligned 3 DWORD DS operations	commit \| commitdiff \| tree
2022-04-12	Stanislav Mekhanoshin	[AMDGPU] Refactor LDS alignment checks.	commit \| commitdiff \| tree
2022-04-09	Stanislav Mekhanoshin	[AMDGPU] Fix regression with vectorization limiting	commit \| commitdiff \| tree
2022-04-07	Stanislav Mekhanoshin	[AMDGPU] Fix handling of gfx10 LDS misaligned access bug	commit \| commitdiff \| tree
2022-04-07	Stanislav Mekhanoshin	[AMDGPU] Split unaligned LDS access instead of scalarizing	commit \| commitdiff \| tree
2022-04-07	Stanislav Mekhanoshin	[AMDGPU] Fix test difference in debug and release....	commit \| commitdiff \| tree
2022-04-07	Stanislav Mekhanoshin	[AMDGPU] Check SI LDS offset bug in the allowsMisalignedMemo...	commit \| commitdiff \| tree
2022-04-06	Stanislav Mekhanoshin	[AMDGPU] Regenerate global isel lds ops test checks...	commit \| commitdiff \| tree
2022-03-30	Stanislav Mekhanoshin	[AMDGPU] Remove empty comments in 2 tests. NFC.	commit \| commitdiff \| tree
2022-03-29	Stanislav Mekhanoshin	[AMDGPU] gfx940 VALU hazard recognizer	commit \| commitdiff \| tree
2022-03-24	Stanislav Mekhanoshin	[AMDGPU] Use GenericTable to classify DGEMM	commit \| commitdiff \| tree
2022-03-24	Stanislav Mekhanoshin	[AMDGPU] gfx940 MAI hazard recognizer	commit \| commitdiff \| tree
2022-03-24	Stanislav Mekhanoshin	[AMDGPU] Support gfx940 smfmac instructions	commit \| commitdiff \| tree
2022-03-24	Stanislav Mekhanoshin	[AMDGPU] New gfx940 mfma instructions	commit \| commitdiff \| tree
2022-03-22	Stanislav Mekhanoshin	[AMDGPU] Allow v_accvgpr_write to use SGPR on gfx90a	commit \| commitdiff \| tree
2022-03-22	Stanislav Mekhanoshin	[AMDGPU] Update mfma test to run gfx940 checks. NFC.	commit \| commitdiff \| tree
2022-03-21	Stanislav Mekhanoshin	[AMDGPU] Fix AV classes VTs. NFCI.	commit \| commitdiff \| tree
2022-03-18	Stanislav Mekhanoshin	[AMDGPU] Add gfx940 run line to gfx90a mfma test. NFC.	commit \| commitdiff \| tree
2022-03-18	Stanislav Mekhanoshin	[AMDGPU] Added gfx940 mfma dst constraint test. NFC.	commit \| commitdiff \| tree
2022-03-18	Stanislav Mekhanoshin	[AMDGPU] gfx940 basic speed model	commit \| commitdiff \| tree
2022-03-18	Stanislav Mekhanoshin	[AMDGPU] Disable some MFMA instructions on gfx940	commit \| commitdiff \| tree
2022-03-18	Stanislav Mekhanoshin	[AMDGPU] reuse blgp as neg in 2 mfma operations on...	commit \| commitdiff \| tree
2022-03-17	Stanislav Mekhanoshin	[AMDGPU] Add 2 gfx940 mfma tests. NFC.	commit \| commitdiff \| tree
2022-03-17	Stanislav Mekhanoshin	[AMDGPU] New MFMA names for existing instructions	commit \| commitdiff \| tree
2022-03-17	Stanislav Mekhanoshin	[AMDGPU] Add gfx90a and gfx940 to get_elf_mach_gfx_name.cpp	commit \| commitdiff \| tree
2022-03-17	Stanislav Mekhanoshin	[AMDGPU] Allow v_accvgpr_write to use SGPR src on gfx940	commit \| commitdiff \| tree
2022-03-16	Stanislav Mekhanoshin	[AMDGPU] Correct gfx940 memory model documentation.	commit \| commitdiff \| tree
2022-03-15	Stanislav Mekhanoshin	[AMDGPU] gfx940: disable OP_SEL on V_DOT instructions	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] Add symbolic names for gfx940 HWREGs	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] Support for gfx940 flat lds opcodes	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] Support gfx940 v_lshl_add_u64 instruction	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] flat scratch SVS addressing mode for gfx940	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] gfx940 memory model	commit \| commitdiff \| tree
2022-03-14	Stanislav Mekhanoshin	[AMDGPU] Restrict machine copy propagation from creating...	commit \| commitdiff \| tree
2022-03-11	Stanislav Mekhanoshin	[AMDGPU] Support v_mov_b64 in dpp combine	commit \| commitdiff \| tree
next

LLVM monorepo

RSS Atom