Commits · a8b4b9104c8f7ab5edb8651a900e61279e8bf931 · Roger Ferrer / llvm-epi · GitLab

Feb 24, 2022
- [AArch64][AsmParser] Arch directives should set implied features. · ff3f3a54
  Sander de Smalen authored Feb 24, 2022
  ff3f3a54
- [RISCV] Add schedule class for Zbm and Zbe extension · e1d4d1c2
  lian wang authored Feb 15, 2022
  e1d4d1c2
- [InstCombine] Canonicalize SPF to min/max intrinsics · a266af72
  Nikita Popov authored Feb 14, 2022
  a266af72
- Revert "[InstCombine] Remove one-use limitation from X-Y==0 fold" · aa551ad1
  Nikita Popov authored Feb 24, 2022
  aa551ad1
- [RISCV][NFC] Add helper function isVectorConfigInstr to reduce Repeated code. · 2ae92e19
  Chenbing.Zheng authored Feb 24, 2022
  2ae92e19
- [RISCV] Use SelectionDAG::getFreeze to simplify some code. NFC · 5b7ac107
  Craig Topper authored Feb 23, 2022
  
  5b7ac107
- [ScalarEvolution] Control flag for nonstrict inequalities in finite loops · 7e3606f4
  Mircea Trofin authored Feb 23, 2022
  7e3606f4
- PGOInstrumentation, GCOVProfiling: Split indirectbr critical edges regardless of PHIs · 6a383369
  Matthias Braun authored Feb 16, 2022
  6a383369
- Teach the AArch64 backend to instruction select the BCAX instruction. · d7105e76
  Owen Anderson authored Feb 18, 2022
  d7105e76
Feb 23, 2022
- [SampleProf][Inliner] Add an option to turn off inliner in sample-profile pass. · 142cedc2
  minglotus-6 authored Feb 22, 2022
  142cedc2
- [NVPTX] Add ex2.approx.f16/f16x2 support · 69a8350c
  Nicolas Miller authored Feb 22, 2022
  69a8350c
- [NVPTX] Add more FMA intriniscs/builtins · be672934
  Jakub Chlanda authored Feb 22, 2022
  be672934
- [NVPTX] Expose float tys min, max, abs, neg as builtins · e0dc4ac2
  Jakub Chlanda authored Feb 22, 2022
  e0dc4ac2
- [SLP] Fastpath instructions not in block being scheduled [nfc] · ed54296e
  Philip Reames authored Feb 23, 2022
  
  ed54296e
- [SLP] Replace a impossible branch condition with an assert [NFC] · a4541fdf
  Philip Reames authored Feb 23, 2022
  a4541fdf
- {SLP] Make it clear ScheduleDataMap is keyed by instructions [NFC] · 9a40f9f6
  Philip Reames authored Feb 23, 2022
  
  9a40f9f6
- [instrprof] Rename the profile kind types to be more descriptive. · b6817999
  Snehasish Kumar authored Feb 17, 2022
  b6817999
- Revert "[SLP] Remove cap on schedule window size" · 9392c0d4
  Philip Reames authored Feb 23, 2022
  9392c0d4
- Revert "[SLP] Simplify extendSchedulingRegion" · a83441e8
  Philip Reames authored Feb 23, 2022
  a83441e8
- [DAGCombiner][TargetLowering] Pass SDValue by value to isMulAddWithConstProfitable. · c7d6448d
  Craig Topper authored Feb 23, 2022
  c7d6448d
- [DAGCombine] Extend combineCarryDiamond() · afdaa86b
  Pawe Bylica authored Feb 23, 2022
  afdaa86b
- [SLP] Rearrange fields in ScheduleData for density [NFC] · 222e8610
  Philip Reames authored Feb 23, 2022
  
  222e8610
- [SLP] Remove SchedulingPriority from ScheduleData [NFC] · a3e9b32c
  Philip Reames authored Feb 23, 2022
  a3e9b32c
- [SLP] Simplify extendSchedulingRegion · 8c85f3a0
  Philip Reames authored Feb 23, 2022
  8c85f3a0
- AttributorAttributes: avoid a crashing on bad alignments · 95f3cc22
  Augie Fackler authored Feb 11, 2022
  95f3cc22
- [X86] combineX86ShufflesRecursively - pull out repeated getValueType/getSimpleValueType calls. · 427d9f60
  Simon Pilgrim authored Feb 23, 2022
  
  427d9f60
- Revert "[MachineOutliner][AArch64] NFC: Split MBBs into "outlinable ranges"" · 68c718c8
  Jessica Paquette authored Feb 23, 2022
  68c718c8
- [X86] combineX86ShufflesRecursively - don't both widening inputs before... · 87753ceb
  Simon Pilgrim authored Feb 23, 2022
  87753ceb
- [DAG] try to convert multiply to shift via demanded bits · 21d7c3bc
  Sanjay Patel authored Feb 23, 2022
  21d7c3bc
- Revert "AttributorAttributes: avoid a crashing on bad alignments" · 1fd980de
  Arthur Eubanks authored Feb 23, 2022
  1fd980de
- AttributorAttributes: avoid a crashing on bad alignments · 70ff6fbe
  Augie Fackler authored Feb 11, 2022
  70ff6fbe
- [SLP] Remove cap on schedule window size · 6adf4b03
  Philip Reames authored Feb 22, 2022
  6adf4b03
- [LAA] Add remarks for unbounded array access · 9f1c6fbf
  Malhar Jajoo authored Feb 02, 2022
  9f1c6fbf
- [X86] combineX86ShuffleChainWithExtract - don't both widening inputs after... · 22d04531
  Simon Pilgrim authored Feb 23, 2022
  22d04531
- [Bitcode] Store function type IDs rather than function types · 4d37bbc4
  Nikita Popov authored Feb 23, 2022
  4d37bbc4
- [InstCombine] Support min/max intrinsics in udiv->lshr fold · 587c7ff1
  Nikita Popov authored Feb 23, 2022
  587c7ff1
- [PowerPC] Add the Power10 LXVKQ instrution. · eb1c5a98
  Stefan Pintilie authored Feb 22, 2022
  eb1c5a98
- [MC][ELF] Use SHF_SUNW_NODISCARD instead of SHF_GNU_RETAIN on Solaris · 365be7ac
  Rainer Orth authored Feb 23, 2022
  365be7ac
- [ELF] Use SHF_SUNW_NODISCARD instead of SHF_GNU_RETAIN on Solaris · 42e391e4
  Rainer Orth authored Feb 23, 2022
  42e391e4
- [InstCombine] Further simplify udiv -> lshr folding · 03e6efb8
  Nikita Popov authored Feb 23, 2022
  03e6efb8