Commits · f1da33e4d3fc9797c7fb4f9ea681a2bc2254c281 · Roger Ferrer / llvm-epi · GitLab

Jun 11, 2016
- [LICM] Make isGuaranteedToExecute more accurate. · f1da33e4
  Eli Friedman authored Jun 11, 2016
  f1da33e4
- Delay dominator updation while cloning loop. · c702b8b3
  Vikram TV authored Jun 11, 2016
  c702b8b3
- [esan|cfrag] Handle complex GEP instr in the cfrag tool · bc8fbeac
  Qin Zhao authored Jun 10, 2016
  bc8fbeac
- Don't try to rotate a loop more than once - we never do this anyway. · b98294d0
  Michael Zolotukhin authored Jun 10, 2016
  b98294d0
Jun 10, 2016
- MemorySSA: fix memory access local dominance function for live on entry · e1f60b1f
  Sebastian Pop authored Jun 10, 2016
  e1f60b1f
- Remove a few gendered pronouns. · 2cf5e89e
  Nico Weber authored Jun 10, 2016
  2cf5e89e
- Disable MSan-hostile loop unswitching. · eaea297d
  Evgeniy Stepanov authored Jun 10, 2016
  eaea297d
- Move isGuaranteedToExecute out of LICM. · 122f984a
  Evgeniy Stepanov authored Jun 10, 2016
  122f984a
- Add a period. NFC. · 840b3efe
  Chad Rosier authored Jun 10, 2016
  840b3efe
- Fix whitespace. NFC. · a8bc512b
  Chad Rosier authored Jun 10, 2016
  a8bc512b
- [esan|cfrag] Add the struct field offset array in StructInfo · 0b96aa71
  Qin Zhao authored Jun 10, 2016
  0b96aa71
- [esan|cfrag] Disable load/store instrumentation for cfrag · d677d888
  Qin Zhao authored Jun 10, 2016
  d677d888
- Make sure that not interesting allocas are not instrumented. · b451f1bd
  Vitaly Buka authored Jun 09, 2016
  b451f1bd
- Unpoison stack memory in use-after-return + use-after-scope mode · 79b75d3d
  Vitaly Buka authored Jun 09, 2016
  79b75d3d
- Use ProfileSummaryInfo in inline cost analysis. · 71069cf6
  Easwaran Raman authored Jun 09, 2016
  71069cf6
Jun 09, 2016
- [PM] Port LCSSA to the new PM. · e12c487b
  Easwaran Raman authored Jun 09, 2016
  e12c487b
- [LV] Use vector phis for some secondary induction variables · c5edcdeb
  Michael Kuperstein authored Jun 09, 2016
  c5edcdeb
- Revert r272194 No need for it if loop Analysis Manager is used · ecde1c7f
  Xinliang David Li authored Jun 09, 2016
  ecde1c7f
- [ThinLTO/gold] Enable summary-based internalization · 7ab1f692
  Teresa Johnson authored Jun 09, 2016
  7ab1f692
- [LoopSimplify] Preserve LCSSA when merging exit blocks. · 8e7e7672
  Michael Zolotukhin authored Jun 08, 2016
  8e7e7672
- [LoopUnroll] Check that DT is available before trying to verify it. · aa547616
  Michael Zolotukhin authored Jun 08, 2016
  aa547616
Jun 08, 2016
- [SLPVectorizer] Handle GEP with differing constant index types · 987ab631
  Michael Zolotukhin authored Jun 08, 2016
  987ab631
- [PM] Add missing caching of GlobalsAA to EarlyCSE. · 02861d86
  Davide Italiano authored Jun 08, 2016
  02861d86
- [InstCombine] move fold of select of add/sub to helper function; NFCI · 39293138
  Sanjay Patel authored Jun 08, 2016
  39293138
- [InstCombine] fix outdated comment, simplify logic; NFCI · 384d0f21
  Sanjay Patel authored Jun 08, 2016
  384d0f21
- The patch set unroll disable pragma when unroll · 3e2f389a
  Evgeny Stupachenko authored Jun 08, 2016
  3e2f389a
- [PM] Refector LoopAccessInfo analysis code · 572135f7
  Xinliang David Li authored Jun 08, 2016
  572135f7
- [InstCombine] reduce indent; NFC · 10a2c38d
  Sanjay Patel authored Jun 08, 2016
  10a2c38d
- [MemCpyOpt] Do not exchange llvm.lifetime.start and llvm.memcpy · 7aa0ad65
  Tim Shen authored Jun 08, 2016
  7aa0ad65
- [InstCombine] use copyIRFlags() ; NFCI · 916f8a0c
  Sanjay Patel authored Jun 08, 2016
  916f8a0c
- Apply most suggestions of clang-tidy's performance-unnecessary-value-param · c321e534
  Benjamin Kramer authored Jun 08, 2016
  c321e534
- [PM] LoopSimplify. Remove unneeded pass dependencies. NFCI. · 2d5ab0a5
  Davide Italiano authored Jun 08, 2016
  2d5ab0a5
- [PM/SimplifyCFG] Preserve GlobalsAA even if the IR is mutated. · d8d83f47
  Davide Italiano authored Jun 08, 2016
  d8d83f47
- Avoid copies of std::strings and APInt/APFloats where we only read from it · 46e38f36
  Benjamin Kramer authored Jun 08, 2016
  46e38f36
Jun 07, 2016
- [PM] Preserve GlobalsAA for SROA. · 16e96d4b
  Davide Italiano authored Jun 07, 2016
  16e96d4b
- [InstCombine][AVX2] Add support for simplifying AVX2 per-element shifts to native shifts · db9893fb
  Simon Pilgrim authored Jun 07, 2016
  db9893fb
- [InstCombine][SSE] Add MOVMSK constant folding (PR27982) · 91e3ac82
  Simon Pilgrim authored Jun 07, 2016
  91e3ac82
- [InstCombine] scalarizePHI should not assume the code it sees has been CSE'd · a0c6ae02
  Michael Kuperstein authored Jun 06, 2016
  a0c6ae02
Jun 06, 2016
- [PM] Preserve the correct set of analyses for GVN. · fea0a4c5
  Davide Italiano authored Jun 06, 2016
  fea0a4c5
- [GVN] Switch dump() definition over to LLVM_DUMP_METHOD. · 82c44782
  Davide Italiano authored Jun 06, 2016
  82c44782