Commits · e8df879948f9da611d88224d87af177ee67be2a7 · Roger Ferrer / llvm-epi · GitLab

Aug 22, 2015
- AMDGPU: Improve accuracy of instruction rates for some FP instructions · e8df8799
  Matt Arsenault authored Aug 22, 2015
  e8df8799
- AMDGPU: Use DFS to avoid second loop over function · 33010103
  Matt Arsenault authored Aug 22, 2015
  33010103
- AMDGPU: Make sure to run verifier after SIFixSGPRLiveRanges · c8d8e4ed
  Matt Arsenault authored Aug 22, 2015
  c8d8e4ed
- AMDGPU: Improve debug printing in SIFixSGPRLiveRanges · aba29d6a
  Matt Arsenault authored Aug 22, 2015
  aba29d6a
- AMDGPU: Move CI instructions into CIInstructions.td · 6adf07a9
  Matt Arsenault authored Aug 22, 2015
  6adf07a9
- AMDGPU: Minor cleanups to help with f16 support · f56872dc
  Matt Arsenault authored Aug 21, 2015
  f56872dc
- Improve the determinism of MergeFunctions · 057292a7
  JF Bastien authored Aug 21, 2015
  057292a7
- [LAA] Hold bounds via ValueHandles during SCEV expansion · 4e533ef7
  Adam Nemet authored Aug 21, 2015
  4e533ef7
- Standardized 'failed' to 'Failed' in LoopVectorizationRequirements. · 552a62fa
  Tyler Nowicki authored Aug 21, 2015
  552a62fa
- MIRLangRef: Add 'MIR Testing Guide' section. · ea788c4b
  Alex Lorenz authored Aug 21, 2015
  ea788c4b
- LTO: Change signature of LTOCodeGenerator::setCodePICModel() to take a Reloc::Model. · 44ee84ee
  Peter Collingbourne authored Aug 21, 2015
  44ee84ee
- AMDGPU/SI: Better handle s_wait insertion · bd8a0856
  Tom Stellard authored Aug 21, 2015
  bd8a0856
- Re-apply r245635, "[InstCombine] Transform A & (L - 1) u< L --> L != 0" · c86c162a
  Sanjoy Das authored Aug 21, 2015
  c86c162a
- Range-for-ify some things in GlobalMerge · 47bf5c01
  David Blaikie authored Aug 21, 2015
  47bf5c01
- [opaque pointer types] Fix a few easy places in GlobalMerge that were... · 9ed57a9e
  David Blaikie authored Aug 21, 2015
  9ed57a9e
Aug 21, 2015
- MIR Serialization: Serialize the pointer IR expression values in the machine · c1136ef3
  Alex Lorenz authored Aug 21, 2015
  c1136ef3
- [ARM] Fix MachO CPU Subtype selection · 366dd9fd
  Vedant Kumar authored Aug 21, 2015
  366dd9fd
- MIRParser: Split the 'parseIRConstant' method into two methods. NFC. · 5d8b0bd9
  Alex Lorenz authored Aug 21, 2015
  5d8b0bd9
- [opaque pointer types] Push the passing of value types up from... · d583b195
  David Blaikie authored Aug 21, 2015
  d583b195
- [PowerPC] PPCVSXFMAMutate should not segfault on undef input registers · ff9639d6
  Hal Finkel authored Aug 21, 2015
  ff9639d6
- AsmParser: Save and restore the parsing state for types using SlotMapping. · 1de2acd3
  Alex Lorenz authored Aug 21, 2015
  1de2acd3
- [LVI] Use a SmallVector instead of SmallPtrSet. NFC · 7a1483e7
  Bruno Cardoso Lopes authored Aug 21, 2015
  7a1483e7
- MIRLangRef: Describe the syntax for the immediate operands, register values, · b981d372
  Alex Lorenz authored Aug 21, 2015
  b981d372
- MIR Serialization: Print MCSymbol operands. · f22ca8ad
  Alex Lorenz authored Aug 21, 2015
  f22ca8ad
- Line endings fix. · 76b91d90
  Simon Pilgrim authored Aug 21, 2015
  76b91d90
- [x86] enable machine combiner reassociations for 256-bit vector min/max · f0bc07f7
  Sanjay Patel authored Aug 21, 2015
  f0bc07f7
- remove 'FeatureSlowUAMem' from AMD CPUs based on 10H micro-arch or later · dddad102
  Sanjay Patel authored Aug 21, 2015
  dddad102
- [llvm-readobj] Add support for MachO DataInCodeDataCommand. · 07e7acb6
  Davide Italiano authored Aug 21, 2015
  07e7acb6
- Add comment as follow up to r245712 · 51973e10
  David Blaikie authored Aug 21, 2015
  51973e10
- [x86] invert logic for attribute 'FeatureFastUAMem' · 9e916dc4
  Sanjay Patel authored Aug 21, 2015
  9e916dc4
- [opaque pointer type]: Pass explicit pointee type when building a constant GEP. · 88208840
  David Blaikie authored Aug 21, 2015
  88208840
- llvm-lto: Re-order code. · 55217439
  Peter Collingbourne authored Aug 21, 2015
  55217439
- Filter libraries that are not installed out of CMake exports (currently · 1e3dc527
  Dan Liew authored Aug 21, 2015
  1e3dc527
- llvm-build: Adopt generation of LLVM_LIBS_TO_EXPORT. Patch by · ef1fa56c
  Dan Liew authored Aug 21, 2015
  ef1fa56c
- llvm-build: Factor out duplicate cmake export listing. Patch by · a70fb6e0
  Dan Liew authored Aug 21, 2015
  a70fb6e0
- [x86] enable machine combiner reassociations for 128-bit vector min/max · cf942fa9
  Sanjay Patel authored Aug 21, 2015
  cf942fa9
- Remove an unnecessary use of pointee types introduced in r194220 · 401bb64b
  David Blaikie authored Aug 21, 2015
  401bb64b
- Disable Visual C++ 2013 Debug mode assert on null pointer in some STL algorithms, · 528d8d60
  Yaron Keren authored Aug 21, 2015
  528d8d60
- MIRLangRef: Describe the syntax for machine instruction names and flags. · 8eadc3f3
  Alex Lorenz authored Aug 21, 2015
  8eadc3f3
- save some testing time; get rid of the non-SSE chips in this test · c9f4aa6b
  Sanjay Patel authored Aug 21, 2015
  c9f4aa6b