Commits · f529b507020868ded01947aa5eba38d7ddce516c · Roger Ferrer / llvm-epi · GitLab

Jan 17, 2019
- [HotColdSplit] Allow outlining with live outputs · f529b507
  Vedant Kumar authored Jan 17, 2019
  f529b507
- [HotColdSplit] Consider resume instructions to be cold · b70e20db
  Vedant Kumar authored Jan 17, 2019
  b70e20db
- [HotColdSplit] Relax requirement that the cold sink block be extractable · 4541be06
  Vedant Kumar authored Jan 17, 2019
  4541be06
- [HotColdSplit] Simplify tests by lowering their splitting thresholds · 32a014d0
  Vedant Kumar authored Jan 17, 2019
  32a014d0
- [SampleFDO] Skip profile reading when flattened profile used in ThinLTO postlink · 3bcccdfe
  Wei Mi authored Jan 17, 2019
  3bcccdfe
- [InstCombine] Don't sink dynamic allocas · edd653bc
  Reid Kleckner authored Jan 17, 2019
  edd653bc
- [LoopSimplifyCFG] Form LCSSA when a parent loop becomes a sibling · 61a8d3fb
  Max Kazantsev authored Jan 17, 2019
  61a8d3fb
- [LoopSimplifyCFG] Fix order of deletion of complex dead subloops · 8b134169
  Max Kazantsev authored Jan 17, 2019
  8b134169
- Allow FP types for atomicrmw xchg · 0cb08e44
  Matt Arsenault authored Jan 17, 2019
  0cb08e44
- [MergeFunc] Prevent silent miscompile of vararg functions · a9906c1e
  Vedant Kumar authored Jan 17, 2019
  a9906c1e
- [FunctionComparator] Consider tail call kinds · e21ab221
  Vedant Kumar authored Jan 17, 2019
  e21ab221
- [PGO] Make pgo related options in opt more consistent. · c876e3d4
  Wei Mi authored Jan 16, 2019
  c876e3d4
Jan 16, 2019
- [SLP] Fix PR40310: The reduction nodes should stay scalar. · 18809a6b
  Alexey Bataev authored Jan 16, 2019
  18809a6b
- Assertion in isAllocaPromotable due to extra bitcast goes into lifetime marker · 3ec170c8
  Gabor Buella authored Jan 16, 2019
  3ec170c8
- Only promote args when function attributes are compatible · 3d36e5c3
  Tom Stellard authored Jan 16, 2019
  3d36e5c3
- [InstCombine]Avoid introduction of unaligned mem access · a5b0e558
  Serguei Katkov authored Jan 16, 2019
  a5b0e558
- [EH] Rename llvm.x86.seh.recoverfp intrinsic to llvm.eh.recoverfp · 436735c3
  Mandeep Singh Grang authored Jan 16, 2019
  436735c3
Jan 15, 2019
- treat invoke like call · d129d3e9
  David Callahan authored Jan 15, 2019
  d129d3e9
- [SLP] Added test for PR40310, NFC. · 9514b1c6
  Alexey Bataev authored Jan 15, 2019
  9514b1c6
- Reapply "[CodeGen][X86] Expand USUBSAT to UMAX+SUB, also for vectors" · d3b86b79
  Nikita Popov authored Jan 15, 2019
  d3b86b79
- Remove irrelevant references to legacy git repositories from · 693d39dd
  James Y Knight authored Jan 15, 2019
  693d39dd
- [SLP][X86] Split prefer-256-bit 'AVX256BW' tests from AVX2 checks · 4e38b8f8
  Simon Pilgrim authored Jan 15, 2019
  4e38b8f8
- [InstCombine] Don't undo 0 - (X * Y) canonicalization when combining subs. · 4094f34f
  Florian Hahn authored Jan 15, 2019
  4094f34f
- [NFC] Remove obsolete enum RangeCheckKind · 80242ee8
  Max Kazantsev authored Jan 15, 2019
  80242ee8
- AMDGPU: Add a fast path for icmp.i1(src, false, NE) · 33eb4d94
  Marek Olsak authored Jan 15, 2019
  33eb4d94
Jan 14, 2019
- Update GettingStarted guide to recommend that people use the new · 78655888
  James Y Knight authored Jan 14, 2019
  78655888
- [AMDGPU] Add support for TFE/LWE in image intrinsics. 2nd try · f77079f8
  David Stuttard authored Jan 14, 2019
  f77079f8
Jan 12, 2019
- [LoopVectorizer] give more advice in remark about failure to vectorize call · 7d65fe5c
  Sanjay Patel authored Jan 12, 2019
  7d65fe5c
Jan 11, 2019
- [ConstantFolding] Fold undef for integer intrinsics · 9f6e9cf7
  Nikita Popov authored Jan 11, 2019
  9f6e9cf7
- [SLP]Moved NVPTX test under NVPTX directory, NFC. · 8da9a753
  Alexey Bataev authored Jan 11, 2019
  8da9a753
- [SLP]Update test checks for the SPL vectorizer, NFC. · ce2c8b33
  Alexey Bataev authored Jan 11, 2019
  ce2c8b33
- [LTO] Record whether LTOUnit splitting is enabled in index · 290a8398
  Teresa Johnson authored Jan 11, 2019
  290a8398
- [MergeFunc] Erase unused duplicate functions if they are discardable · ee10ef73
  Vedant Kumar authored Jan 11, 2019
  ee10ef73
- [Jump Threading] Unfold a select insn that feeds a switch via a phi node · f452f116
  Ehsan Amiri authored Jan 11, 2019
  f452f116
Jan 10, 2019
- Use MemorySSA in LICM to do sinking and hoisting. · cae12eda
  Alina Sbirlea authored Jan 10, 2019
  cae12eda
- Fix RUN line in test/Transforms/LoopDeletion/crashbc.ll · 15313a37
  Bjorn Pettersson authored Jan 10, 2019
  15313a37
- [SimplifyLibCalls] Fix memchr expansion for constant strings. · d4e7a0d8
  Eli Friedman authored Jan 09, 2019
  d4e7a0d8
Jan 09, 2019
- [Inliner] Assert that the computed inline threshold is non-negative. · ed279752
  Easwaran Raman authored Jan 09, 2019
  ed279752
Jan 08, 2019
- [InstCombine] remove stale comments; NFC · bad24064
  Sanjay Patel authored Jan 08, 2019
  bad24064
- [InstCombine] canonicalize another raw IR rotate pattern to funnel shift · d023dd60
  Sanjay Patel authored Jan 08, 2019
  d023dd60