Commits · ea0237205990c6652929331cf07702add46f377d · Roger Ferrer / llvm-epi · GitLab

Aug 12, 2016
- [BranchFolding] Restrict tail merging loop blocks after MBP · ea023720
  Haicheng Wu authored Aug 12, 2016
  ea023720
- WholeProgramDevirt: initialize WasDevirt in all constructors. · 89439a79
  Ivan Krasin authored Aug 12, 2016
  89439a79
- [DSE] Don't remove stores made live by a call which unwinds. · a6707f56
  Eli Friedman authored Aug 12, 2016
  a6707f56
- Refactor isValidAssumeForContext to reduce duplication and indentation. NFC. · 54a02556
  Pete Cooper authored Aug 12, 2016
  54a02556
- Use the range variant of find_if instead of unpacking begin/end · 562e8294
  David Majnemer authored Aug 12, 2016
  562e8294
- WholeProgramDevirt: fix access to a non-initialized field. · 067cc24a
  Ivan Krasin authored Aug 12, 2016
  067cc24a
- Add comment /NFC · 1ce88fa0
  Xinliang David Li authored Aug 11, 2016
  1ce88fa0
- [ADT] Migrate DepthFirstIterator to use NodeRef · 6aaeb9b1
  Tim Shen authored Aug 11, 2016
  6aaeb9b1
- Fix typos /NFC · cbb5e02f
  Xinliang David Li authored Aug 11, 2016
  cbb5e02f
- Remove unnecessary extra version of isValidAssumeForContext. NFC. · fa7ae4f3
  Pete Cooper authored Aug 11, 2016
  fa7ae4f3
- Use the range variant of find instead of unpacking begin/end · 0d955d0b
  David Majnemer authored Aug 11, 2016
  0d955d0b
- Don't import variadic functions · 332b3b22
  Piotr Padlewski authored Aug 11, 2016
  332b3b22
- X86-FMA3: Implemented commute transformation for EVEX/AVX512 FMA3 opcodes. · 6daefcf6
  Vyacheslav Klochkov authored Aug 11, 2016
  6daefcf6
- Re-commit r278066: Do not ignore SizeOfOptionalHeader in COFF header even if... · 8950a538
  Rui Ueyama authored Aug 11, 2016
  8950a538
Aug 11, 2016
- GlobalISel: support 'null' constant in translation. · 8e0c53a0
  Tim Northover authored Aug 11, 2016
  8e0c53a0
- Extend trip count instead of truncating IV in LFTR, when legal · dbcfea98
  Ehsan Amiri authored Aug 11, 2016
  dbcfea98
- [MSSA] Use is_contained · da2f38e0
  Daniel Berlin authored Aug 11, 2016
  da2f38e0
- Use range algorithms instead of unpacking begin/end · 0a16c228
  David Majnemer authored Aug 11, 2016
  0a16c228
- [Hexagon] Allow non-returning calls in hardware loops · 1b689da0
  Krzysztof Parzyszek authored Aug 11, 2016
  1b689da0
- [vim] Add more attributes to llvm.vim · 3124181e
  David Majnemer authored Aug 11, 2016
  3124181e
- AMDGPU: Remove unused tablegen utilities · 18da70dd
  Matt Arsenault authored Aug 11, 2016
  18da70dd
- [SCEV] Update interface to handle SCEVExpander insert point motion. · d0182809
  Geoff Berry authored Aug 11, 2016
  d0182809
- Remove empty file left by partial reversion. · da6f5f2d
  Tim Northover authored Aug 11, 2016
  da6f5f2d
- GlobalISel: add translation support for shift operations. · 30e67ce7
  Tim Northover authored Aug 11, 2016
  30e67ce7
- GlobalISel: support zext & sext during translation phase. · f1f7bf12
  Tim Northover authored Aug 11, 2016
  f1f7bf12
- Fix type truncation warnings · faa7506f
  Teresa Johnson authored Aug 11, 2016
  faa7506f
- Move GVNHoist tests into their own directory since it is a separate pass · 2698cbb4
  Daniel Berlin authored Aug 11, 2016
  2698cbb4
- AMDGPU : Add intrinsic for instruction v_cvt_pk_u8_f32 · 70cda075
  Wei Ding authored Aug 11, 2016
  70cda075
- Revert rL278384 which caused several buildbot failures (like check failures in CodeGen/X86/clz.ll). · 3ab58160
  Wei Mi authored Aug 11, 2016
  3ab58160
- Fix PR 28933 · f75fd1b5
  Daniel Berlin authored Aug 11, 2016
  f75fd1b5
- [ADT] Add relation operators for Optional · 0fdb2daa
  Tim Shen authored Aug 11, 2016
  0fdb2daa
- CodeGen: Avoid dereferencing end() in MachineScheduler · 38eea4a7
  Duncan P. N. Exon Smith authored Aug 11, 2016
  38eea4a7
- AMDGPU: Prune includes · 2ffe8fd2
  Matt Arsenault authored Aug 11, 2016
  2ffe8fd2
- [Hexagon] Standardize "select" pseudo-instructions · 258af19d
  Krzysztof Parzyszek authored Aug 11, 2016
  258af19d
- WholeProgramDevirt: generate more detailed and accurate remarks. · f3403fd2
  Ivan Krasin authored Aug 11, 2016
  f3403fd2
- Remove the restriction that MachineSinking is now stopped by "insert_subreg, · ec19b351
  Wei Mi authored Aug 11, 2016
  ec19b351
- If-conversion incorrectly calculates liveness of redefined registers · a003b763
  Krzysztof Parzyszek authored Aug 11, 2016
  a003b763
- Test commit · 35fef211
  Barnabas Bittner authored Aug 11, 2016
  35fef211
- Target independent codesize heuristics for Loop Idiom Recognition · 7cdf01ef
  Andrew Kaylor authored Aug 11, 2016
  7cdf01ef
- Add a new method to create SimpleInliner instance and make pre-inliner use this. · 61edc107
  Easwaran Raman authored Aug 11, 2016
  61edc107