Commits · 4a652227dd766c2910a62e4b6c12dccb48ea9b1a · Roger Ferrer / llvm-epi-0.8 · GitLab

This is an archived project. Repository and other project resources are read-only.

Feb 14, 2017
- Reapply "[LV] Extend trunc optimization to all IVs with constant integer steps" · f09d13e5
  Matthew Simpson authored Feb 14, 2017
  f09d13e5
- Revert "[LoopVectorize] Added address space check when analysing interleaved accesses" · ec21b769
  Karl-Johan Karlsson authored Feb 14, 2017
  ec21b769
- [LoopVectorize] Added address space check when analysing interleaved accesses · 2ec409cc
  Karl-Johan Karlsson authored Feb 14, 2017
  2ec409cc
Feb 13, 2017
- Revert "[LV] Extend trunc optimization to all IVs with constant integer steps" · 659f92e2
  Matthew Simpson authored Feb 13, 2017
  659f92e2
- [LV] Extend trunc optimization to all IVs with constant integer steps · 7b7f4029
  Matthew Simpson authored Feb 13, 2017
  7b7f4029
Feb 12, 2017
- [LV/LoopAccess] Check statically if an unknown dependence distance can be · eac89d73
  Dorit Nuzman authored Feb 12, 2017
  eac89d73
Feb 11, 2017
- [ARM] Make f16 interleaved accesses expensive. · 8425f453
  Ahmed Bougacha authored Feb 11, 2017
  8425f453
Feb 10, 2017
- Encode duplication factor from loop vectorization and loop unrolling to discriminator. · fb02f714
  Dehao Chen authored Feb 10, 2017
  fb02f714
- [LV] Remove type restriction for vector phi creation · df124a75
  Matthew Simpson authored Feb 10, 2017
  df124a75
Feb 08, 2017
- [Loop Vectorizer] Cost-based decision for vectorization form of memory instruction. · 5267edd3
  Elena Demikhovsky authored Feb 08, 2017
  5267edd3
- [X86] Remove PCOMMIT instruction support since Intel has deprecated this... · e0ac7f3b
  Craig Topper authored Feb 08, 2017
  e0ac7f3b
Feb 07, 2017
- [LV] Add new ARM/AArch64 interleaved access cost model tests (NFC) · 3877f397
  Matthew Simpson authored Feb 07, 2017
  3877f397
- [LV] Simplify ARM/AArch64 interleaved access cost model tests (NFC) · 1cd02f13
  Matthew Simpson authored Feb 07, 2017
  1cd02f13
Feb 02, 2017
- [LV] Also port failure remarks to new OptimizationRemarkEmitter API · 0bf1b863
  Adam Nemet authored Feb 02, 2017
  0bf1b863
Jan 26, 2017
- [LV] Fix an issue where forming LCSSA in the place that we did would · 6f4ed077
  Chandler Carruth authored Jan 26, 2017
  6f4ed077
Jan 25, 2017
- [X86] enable memory interleaving for X86\SLM arch. · 20caee95
  Mohammed Agabaria authored Jan 25, 2017
  20caee95
Jan 19, 2017
- [LV] Run loop-simplify and LCSSA explicitly instead of "requiring" them · 230867e5
  Michael Kuperstein authored Jan 19, 2017
  230867e5
Jan 18, 2017
- [LV] Allow reductions that have several uses outside the loop · 7cefb409
  Michael Kuperstein authored Jan 18, 2017
  7cefb409
Jan 17, 2017
- [LV] Add requires asserts to test case · e2c9ad94
  Matthew Simpson authored Jan 17, 2017
  e2c9ad94
- [LV] Mark non-consecutive-like pointers non-uniform · 3fbdaa59
  Matthew Simpson authored Jan 17, 2017
  3fbdaa59
Jan 11, 2017
- [X86] fixing failed test in commit: r291657 · 81d0f170
  Mohammed Agabaria authored Jan 11, 2017
  81d0f170
- [X86] updating TTI costs for arithmetic instructions on X86\SLM arch. · 2c96c433
  Mohammed Agabaria authored Jan 11, 2017
  2c96c433
Jan 10, 2017
- [LV] Don't panic when encountering the IV of an outer loop. · ee31cbe3
  Michael Kuperstein authored Jan 10, 2017
  ee31cbe3
Jan 09, 2017
- [LV] Fix-up external IV users after updating dominator tree · cf796478
  Matthew Simpson authored Jan 09, 2017
  cf796478
Jan 05, 2017
- Currently isLikelyComplexAddressComputation tries to figure out if the given... · 23599ba7
  Mohammed Agabaria authored Jan 05, 2017
  23599ba7
Dec 22, 2016
- Renumber testcase metadata nodes after r290153. · 1eadba1c
  Adrian Prantl authored Dec 22, 2016
  1eadba1c
Dec 20, 2016
- [IR] Remove the DIExpression field from DIGlobalVariable. · bceaaa96
  Adrian Prantl authored Dec 20, 2016
  bceaaa96
Dec 19, 2016
- [LoopVersioning] Require loop-simplify form for loop versioning. · 2e03213f
  Florian Hahn authored Dec 19, 2016
  2e03213f
Dec 16, 2016
- Revert "[IR] Remove the DIExpression field from DIGlobalVariable." · 73ec0656
  Adrian Prantl authored Dec 16, 2016
  73ec0656
- Reapply "[LV] Enable vectorization of loops with conditional stores by default" · a4964f29
  Matthew Simpson authored Dec 16, 2016
  a4964f29
- [LV] Don't attempt to type-shrink scalarized instructions · 099af810
  Matthew Simpson authored Dec 16, 2016
  099af810
- Revert r289863: [LV] Enable vectorization of loops with conditional · 48b4e614
  Chandler Carruth authored Dec 16, 2016
  48b4e614
- [IR] Remove the DIExpression field from DIGlobalVariable. · 74a835cd
  Adrian Prantl authored Dec 16, 2016
  74a835cd
- Revert "[IR] Remove the DIExpression field from DIGlobalVariable." · 03c6d31a
  Adrian Prantl authored Dec 16, 2016
  03c6d31a
- [IR] Remove the DIExpression field from DIGlobalVariable. · ce139357
  Adrian Prantl authored Dec 16, 2016
  ce139357
Dec 15, 2016
- [LV] Enable vectorization of loops with conditional stores by default · 6a98bcfe
  Matthew Simpson authored Dec 15, 2016
  6a98bcfe
Dec 13, 2016
- [LV] Don't vectorize when we have a small static bound on trip count · 3d23d4a2
  Michael Kuperstein authored Dec 13, 2016
  3d23d4a2
Dec 11, 2016
- [Verifier] Add verification for TBAA metadata · 3336f681
  Sanjoy Das authored Dec 11, 2016
  3336f681
Dec 07, 2016
- [LV] Scalarize operands of predicated instructions · 364da7e5
  Matthew Simpson authored Dec 07, 2016
  364da7e5
Dec 03, 2016
- [ppc] Correctly compute the cost of loading 32/64 bit memory into VSR · 835de1f3
  Guozhi Wei authored Dec 03, 2016
  835de1f3