Commits · 8242509a49e019c0279305d152c7ab2b9cdc2d0d · Raul Torres / llvm-target-spread · GitLab

Mar 30, 2020
- [AMDGPU] Add Relocation Constant Support · 77ce2e21
  Jakub Kuderski authored Mar 30, 2020
  77ce2e21
- Introduce unify-loop-exits pass. · 3cbbded6
  Sameer Sahasrabuddhe authored Mar 28, 2020
  3cbbded6
- [LoopVectorize] Fix crash on "getNoopOrZeroExtend cannot truncate!" (PR45259) · dcc410b5
  Vedant Kumar authored Mar 30, 2020
  dcc410b5
- [X86] make sure POP has implicit def/use of stack pointer when materializing... · ece79f47
  Yuanfang Chen authored Mar 29, 2020
  ece79f47
- [Alignment][NFC] Use Align version of getMachineMemOperand · bdf77209
  Guillaume Chatelet authored Mar 30, 2020
  bdf77209
- GlobalISel: Minor cleanups · cc3b5590
  Matt Arsenault authored Feb 24, 2020
  
  cc3b5590
- AMDGPU/GlobalISel: Hack to fix i24 argument lowering · bb009498
  Matt Arsenault authored Mar 21, 2020
  bb009498
- AMDGPU/GlobalISel: Legalize 64-bit G_UDIV/G_UREM · 90a36bbd
  Matt Arsenault authored Feb 26, 2020
  90a36bbd
- [DebugInfo] Ensure that a demanded bits optimisation in · f6b2c003
  Chris Jackson authored Mar 30, 2020
  f6b2c003
- Revert "[Darwin] Respect -fno-unroll-loops during LTO." · 7899a111
  Florian Hahn authored Mar 30, 2020
  7899a111
- [DebugInfo] Ensure dead store elimination can mark an operand · 135709aa
  Chris Jackson authored Mar 30, 2020
  135709aa
- [Alignment][NFC] Provide tightened up functions in SelectionDAG,... · 01ba2ad9
  Guillaume Chatelet authored Mar 30, 2020
  01ba2ad9
- [X86][AVX] lowerV4X128Shuffle - attempt to widen to 2x256 to simplify shuffles · e95d04f4
  Simon Pilgrim authored Mar 30, 2020
  e95d04f4
- [Matrix] Rename emitChainedMatrixMultiply to emitMatrixMultiply (NFC). · 9e81249d
  Florian Hahn authored Mar 30, 2020
  9e81249d
- [AMDGPU] Drop const for value that is copied (NFC). · c3b03f3d
  Florian Hahn authored Mar 30, 2020
  c3b03f3d
- [ARM][LowOverheadLoops] Add horizontal reduction support · 94b195ff
  Sam Parker authored Mar 30, 2020
  94b195ff
- [Alignment][NFC] Return Align for SelectionDAGNodes::getOriginalAlignment/getAlignment · b91535f6
  Guillaume Chatelet authored Mar 27, 2020
  b91535f6
- [ARM] MVE VMOV.i64 · c9eaed51
  David Green authored Mar 28, 2020
  c9eaed51
- [NFC] Remove obsolete checks followed by fix of isGuaranteedToTransferExecutionToSuccessor · 4e0d9925
  Max Kazantsev authored Mar 30, 2020
  4e0d9925
- [Coroutines 2/2] Improve symmetric control transfer feature · 31a1d85c
  Jun Ma authored Mar 27, 2020
  31a1d85c
- [Coroutines 1/2] Improve symmetric control transfer feature · a94fa2c0
  Jun Ma authored Mar 27, 2020
  a94fa2c0
Mar 29, 2020
- [MC] Move deprecation infos from MCTargetDesc to MCInstrInfo · 854f268c
  Benjamin Kramer authored Mar 29, 2020
  854f268c
- [X86][AVX] Combine 128/256-bit lane shuffles with zeroable upper subvectors to... · 9c8ec99c
  Simon Pilgrim authored Mar 29, 2020
  9c8ec99c
- [X86] Add isAnyZero shuffle mask helper · 8206c50c
  Simon Pilgrim authored Mar 29, 2020
  
  8206c50c
- [InstCombine] Erase old mul when creating umulo · 8253a86b
  Nikita Popov authored Mar 29, 2020
  8253a86b
- [InstCombine] Use replaceOperand() in demanded elements simplification · 53d20907
  Nikita Popov authored Mar 29, 2020
  53d20907
- [InstCombine] Use replaceOperand() in assoc cast simplification · 0c871400
  Nikita Popov authored Mar 29, 2020
  0c871400
- [InstCombine] Erase old add when optimizing add overflow · a9ddcd64
  Nikita Popov authored Mar 29, 2020
  a9ddcd64
- Introduce support for lib function aligned_alloc in TLI / memory builtins · c0955edf
  Uday Bondhugula authored Mar 28, 2020
  c0955edf
- AMDGPU/GlobalISel: Remove redundant virtual · d15723ef
  Matt Arsenault authored Mar 28, 2020
  
  d15723ef
- AMDGPU: Fix using wrong instruction for FP conversion · ab7a4106
  Matt Arsenault authored Mar 29, 2020
  ab7a4106
- AMDGPU: Fix typo · 97bbe7ad
  Matt Arsenault authored Mar 29, 2020
  
  97bbe7ad
- [VectorCombine] skip debug intrinsics first for efficiency · fc3cc8a4
  Sanjay Patel authored Mar 29, 2020
  
  fc3cc8a4
- [InstCombine] Simplify select of cmpxchg transform · 26fa3375
  Nikita Popov authored Mar 29, 2020
  26fa3375
- [OpenMP] set_bits iterator yields unsigned elements, no reference (NFC). · 99913ef3
  Florian Hahn authored Mar 29, 2020
  99913ef3
- [InstCombine] Fix worklist management in varargs transform · 28f67bd5
  Nikita Popov authored Mar 29, 2020
  28f67bd5
- [InstCombine] Erase original add when creating saddo · 6f07a9e8
  Nikita Popov authored Mar 29, 2020
  6f07a9e8
- [InstCombine] Use replaceOperand() in a few more places · 1e363023
  Nikita Popov authored Mar 29, 2020
  1e363023
- [X86][AVX] Combine 128-bit lane shuffles with a zeroable upper half to EXTRACT_SUBVECTOR (PR40720) · 7734e4b3
  Simon Pilgrim authored Mar 29, 2020
  7734e4b3
- [X86] Rename matchShuffleAsByteRotate to matchShuffleAsElementRotate. NFC. · da4c7db7
  Simon Pilgrim authored Mar 29, 2020
  da4c7db7