Commits · ea05ee90596ceaa361950f818980d4997e4437b9 · Roger Ferrer / llvm-epi · GitLab

Jan 28, 2022
- [RISCV] Preserve VL when truncating i64 gather/scatter indices on RV32. · ea05ee90
  Craig Topper authored Jan 28, 2022
  ea05ee90
- [InstrProf][NFC] Move function out of InstrProf.h · eea002a9
  Ellis Hoag authored Jan 27, 2022
  eea002a9
- [RISCV] Use tablegen size for getInstSizeInBytes. · de0c2d75
  Craig Topper authored Jan 28, 2022
  de0c2d75
- [SLP]Do not reorder top nodes if they do not require reordering. · cec8b614
  Alexey Bataev authored Jan 06, 2022
  cec8b614
- [DAGCombiner] Fix invalid size request in combineRepeatedFPDivisors · 5d089d9a
  Cullen Rhodes authored Jan 28, 2022
  5d089d9a
- [RISCV] Use __extendhfsf2/__truncsfhf2 for fp16 <-> fp32 · a9d5bb92
  Kito Cheng authored Apr 06, 2021
  a9d5bb92
- [CodeExtractor] Fix warning in assert (NFC) · 7d176844
  Nikita Popov authored Jan 28, 2022
  
  7d176844
- [BasicBlockUtils] Fix typo in API name (NFC) · cf0357a5
  Nikita Popov authored Jan 28, 2022
  cf0357a5
- [Loads] Require Align in isDereferenceableAndAlignedPointer() (NFC) · 8a4293f3
  Nikita Popov authored Jan 28, 2022
  8a4293f3
- [x86] try harder to scalarize a vector load with extracted integer op uses · b4b97ec8
  Sanjay Patel authored Jan 28, 2022
  b4b97ec8
- [RISCV][NFC] Make Zb* instruction naming match the convention used elsewhere in the RISC-V backend · 588f121a
  Alex Bradbury authored Jan 28, 2022
  588f121a
- [NFC][RISCV] Bundle up ISAInfo updates and checks · 5f856c5b
  eopXD authored Jan 27, 2022
  5f856c5b
- [ArgPromotion] Don't assume all entry block instrs are executed · 0ebbf343
  Nikita Popov authored Jan 28, 2022
  0ebbf343
- [ArgPromotion] Add test for non-willreturn load hoisting (NFC) · 2dc45bf4
  Nikita Popov authored Jan 28, 2022
  2dc45bf4
- [MVerifier] Don't check liveness of any debug instruction operands · 76fd78b4
  Jeremy Morse authored Jan 28, 2022
  76fd78b4
- [update_test_checks] Fix option name in warning message · de37912f
  Jay Foad authored Jan 28, 2022
  
  de37912f
- [X86] SimplifyDemandedBitsForTargetNode - fold MOVMSK(YMM) -> MOVMSK(XMM) · c7bb3665
  Simon Pilgrim authored Jan 28, 2022
  c7bb3665
- [ArgPromotion] Make areFunctionArgsABICompatible() static (NFC) · 8b36c437
  Nikita Popov authored Jan 28, 2022
  8b36c437
- [PowerPC] Update handling of splat loads for v4i32/v4f32/v2i64 to require non-extending loads. · 9cc5b064
  Amy Kwan authored Jan 27, 2022
  9cc5b064
- [AArch64][SVE][VLS] Move extends into arguments of comparisons · 81bd67e1
  David Truby authored Nov 16, 2021
  81bd67e1
- [x86] add test for possible load scalarization fold; NFC · e9768a2a
  Sanjay Patel authored Jan 28, 2022
  e9768a2a
- [X86] combineSetCCMOVMSK - don't fold MOVMSK(BITCAST(PCMPEQ(X,0))) ->... · 2a13beaa
  Simon Pilgrim authored Jan 28, 2022
  2a13beaa
- [X86] Add test showing incorrect movmsk->ptest fold · 835a7378
  Simon Pilgrim authored Jan 28, 2022
  835a7378
- [SVE] Extend isel pattern coverage for BIC. · 49178a2c
  Paul Walker authored Jan 27, 2022
  49178a2c
- [AArch64][SVE] NFC: Add test for scaled stepvector. · cc8e3dad
  Sander de Smalen authored Jan 28, 2022
  cc8e3dad
- [test][NewGVN] Use '-passes=newgvn' instead of '-basic-aa -newgvn' · bcdcf984
  Bjorn Pettersson authored Jan 26, 2022
  bcdcf984
- [test][NewGVN] Use -passes=newgvn instead of -newgvn · 4f735284
  Bjorn Pettersson authored Jan 26, 2022
  4f735284
- [RISCV] Add tests for possible splat optimizations · 10879c26
  Fraser Cormack authored Jan 28, 2022
  10879c26
- [RISCV] Fix FileCheck prefixes in RVV test · 6297f929
  Fraser Cormack authored Jan 28, 2022
  6297f929
- [llvm-mca] Plot as result of comparing multiple files · 402f91bc
  Milica Matic authored Jan 28, 2022
  402f91bc
- [X86] combineSetCCMOVMSK - match all_of patterns with X86ISD::CMP as well as X86ISD::SUB · cce6490e
  Simon Pilgrim authored Jan 28, 2022
  cce6490e
- [ConstraintElimination] Add sub nuw test with signed predicates. · cd79ca61
  Florian Hahn authored Jan 28, 2022
  cd79ca61
- [BasicAA] Add support for memmove intrinsic · d7424939
  Evgeniy Brevnov authored Jan 24, 2022
  d7424939
- Revert "[InstCombine] Fold and-reduce idiom" · fabaca10
  Hans Wennborg authored Jan 28, 2022
  fabaca10
- [CodeGen] Emit COFF symbol type for function aliases · f7d2afba
  Martin Storsjö authored Jan 27, 2022
  f7d2afba
- [X86] Fix MOVMSK(CONCAT(X,Y)) -> MOVMSK(AND/OR(X,Y)) fold for float types and demanded elements · 93c9b39d
  Simon Pilgrim authored Jan 28, 2022
  93c9b39d
- Cleanup includes for LLVMTableGenGlobalISel · bd059b3b
  serge-sans-paille authored Jan 28, 2022
  bd059b3b
- Cleanup include dependencies for LLVMTableGen · 92f49b89
  serge-sans-paille authored Jan 28, 2022
  92f49b89
- Cleanup includes for LLVMBitstream · b2e0a6d9
  serge-sans-paille authored Jan 28, 2022
  b2e0a6d9
- Move llvm::array_lenghtof to llvm/ADT/STLArrayExtras.h · 5b3b008c
  serge-sans-paille authored Jan 28, 2022
  5b3b008c