Commits · f0bc7d24882ab84f6398a1ea614dd0bf6e2a1085 · Roger Ferrer / llvm-epi · GitLab

Nov 15, 2021
- [analyzer] Fix region cast between the same types with different qualifiers. · f0bc7d24
  Denys Petrov authored Nov 09, 2021
  f0bc7d24
- [X86] LowerFunnelShift - pull out repeated EltSizeInBits variable. NFC. · fc7c1ceb
  Simon Pilgrim authored Nov 15, 2021
  
  fc7c1ceb
- [NFC][X86][Costmodel] Add i1 replication shuffle costmodel test coverage · bc35d5fe
  Roman Lebedev authored Nov 15, 2021
  
  bc35d5fe
- [PatternMatch] Add a new m_Any that binds a value. · a3ee67a6
  Chris Lattner authored Nov 15, 2021
  a3ee67a6
- [llvm][ubsan] Inclusive language: replace use of blacklist HandleLLVMOptions.cmake · 44a64afd
  Zarko Todorovski authored Nov 15, 2021
  44a64afd
- [x86] fold vector (X > -1) & Y to shift+andn (2nd try) · 3d01507c
  Sanjay Patel authored Nov 15, 2021
  3d01507c
- [x86] add test for vector signbit mask fold (PR52504); NFC · 6efe64cf
  Sanjay Patel authored Nov 15, 2021
  6efe64cf
- [X86][Costmodel] `getReplicationShuffleCost()`: promote 8 bit-wide elements to... · 5c7255fe
  Roman Lebedev authored Nov 15, 2021
  5c7255fe
- [X86][Costmodel] `trunc v32i16 to v64i8` can appear after legalization, cost... · a468c39c
  Roman Lebedev authored Nov 15, 2021
  a468c39c
- [X86][Costmodel] `trunc v8i64 to v16i8/v32i8/v64i8` can appear after... · 9e57d9b0
  Roman Lebedev authored Nov 15, 2021
  9e57d9b0
- [X86][Costmodel] `trunc v16i32 to v32i8/v64i8` can appear after legalization,... · 0116c708
  Roman Lebedev authored Nov 15, 2021
  0116c708
- [Flang] Add the FIR LLVMPointer Type · 49c08a22
  Kiran Chandramohan authored Nov 13, 2021
  49c08a22
- [openmp][amdgpu] Add comment warning that libm may be broken · 0e738323
  Jon Chesterfield authored Nov 15, 2021
  0e738323
- [X86] getAVX512Node() - find constant broadcasts to encourage load-folding · ea9e6aa4
  Simon Pilgrim authored Nov 15, 2021
  ea9e6aa4
- [SLP]Improve splat detection. · 036207d5
  Alexey Bataev authored Nov 12, 2021
  036207d5
- [clang] Allow clang-check to customize analyzer output file or dir name · da168dd8
  Ella Ma authored Nov 15, 2021
  da168dd8
- [fir] Add fir.global_len conversion placeholder · 677df8c7
  Valentin Clement authored Nov 15, 2021
  677df8c7
- [flang][CodeGen] Transform `fir.unboxchar` to a sequence of LLVM MLIR · 14867ffc
  Andrzej Warzynski authored Nov 10, 2021
  14867ffc
- [fir] Remove extra return in SelectTypeOpConversion · 37c7211f
  Valentin Clement authored Nov 15, 2021
  37c7211f
- [libc++] Add missing _LIBCPP_HIDE_FROM_ABI to __rewrap_iter · 855a419b
  Louis Dionne authored Nov 15, 2021
  
  855a419b
- Regenerate acle_st1*.c tests · 485c193a
  Matt Devereau authored Nov 15, 2021
  485c193a
- [NFC][InstSimplify] add test cases with base results for or-xor fold · a7f7cf11
  Mehrnoosh Heidarpour authored Nov 15, 2021
  a7f7cf11
- [SLP][NFC]Use `isa_and_nonnull` and fix comment, NFC. · b85152f8
  Alexey Bataev authored Nov 15, 2021
  
  b85152f8
- Fix an unused variable warning · d896c9f4
  Kirstóf Umann authored Nov 15, 2021
  
  d896c9f4
- Revert "[GVN][NFC] Remove redundant check" · 72b5138d
  ksyx authored Nov 15, 2021
  72b5138d
- [SLP]Do not create unused gather nodes for scalar arguments of vector intrinsics. · 6fb5bed7
  Alexey Bataev authored Nov 12, 2021
  6fb5bed7
- [CMake] Allow passing extra options to extract_symbols.py. · 00ff774f
  Simon Tatham authored Nov 15, 2021
  00ff774f
- [lldb] Unwrap the type when dereferencing the value · 95102b7d
  Andy Yankovsky authored Nov 11, 2021
  95102b7d
- [fir] Add fir.select_type conversion placeholder · 2a299e4f
  Valentin Clement authored Nov 15, 2021
  2a299e4f
- [IVDescriptor] Make sure the sign is included for negative extension. · 112c1c34
  Florian Hahn authored Nov 15, 2021
  112c1c34
- [libcxx] Fix enable_if condition of std::reverse_iterator::operator= · 6938270f
  Mikhail Maltsev authored Nov 15, 2021
  6938270f
- [llvm-nm][test] Move X86 lit.local.cfg into the X86 subfolder · 254aa65d
  James Henderson authored Nov 15, 2021
  254aa65d
- [mlir][Linalg] Fix and improve vectorization of depthwise convolutions. · 641fe707
  Nicolas Vasilache authored Nov 15, 2021
  641fe707
- [mlir][Linalg] Add bounded recursion declaration to FMAOp -> LLVM conversion. · ee80ffbf
  Nicolas Vasilache authored Nov 15, 2021
  ee80ffbf
- [mlir] Move min/max ops from Std to Arith. · 9b1d90e8
  Alexander Belyaev authored Nov 15, 2021
  9b1d90e8
- [clang-tidy] Fix a crash in modernize-loop-convert around conversion operators · 29a8d45c
  Kristóf Umann authored Oct 14, 2021
  29a8d45c
- [mlir] DialectConversion: fix OperationLegalizer::isIllegal result when... · 2a3878ea
  Butygin authored Oct 28, 2021
  2a3878ea
- [mlir][Linalg] Fix off-by-one error in conv vector size computation. · f1c86b83
  Nicolas Vasilache authored Nov 15, 2021
  f1c86b83
- Revert "[x86] fold vector (X > -1) & Y to shift+andn" · 5be64d41
  Hans Wennborg authored Nov 15, 2021
  5be64d41
- [cmake] use project relative paths when generating ASTNodeAPI.json · 9fec50f0
  Matheus Izvekov authored Nov 11, 2021
  9fec50f0