Commits · affb40e9d23f70fad605fe9264229b5a705f915c · Roger Ferrer / llvm-epi · GitLab

Aug 28, 2017
- [LoopUnroll] Properly update loop structure in case of successful peeling. · 20cb7e88
  Davide Italiano authored Aug 28, 2017
  20cb7e88
- Fix ARMv4 support · 0f76a35c
  Joerg Sonnenberger authored Aug 28, 2017
  0f76a35c
- Try to fix compilation problem with libstdc++ · 3923b319
  Matthias Braun authored Aug 28, 2017
  3923b319
- Address r311914 review comments · abbc4a7f
  Matthias Braun authored Aug 28, 2017
  abbc4a7f
- [LoopUnroll] Add a cl::opt to force peeling, for testing purposes. · 9a09ae44
  Davide Italiano authored Aug 28, 2017
  9a09ae44
- TableGen: Fix subreg composition/concatenation · afcff2d0
  Matthias Braun authored Aug 28, 2017
  afcff2d0
- TableGen: Add -gen-register-info-debug-dump · 3fdc099a
  Matthias Braun authored Aug 28, 2017
  3fdc099a
- [ARM] Fix bug in ARMLoadStoreOptimizer when kill flags are missing. · 75c4ae30
  Geoff Berry authored Aug 28, 2017
  75c4ae30
- Create PHI node for the return value only when the return value has uses. · 572f45a3
  Taewook Oh authored Aug 28, 2017
  572f45a3
- [CodeView] Don't output S_UDT symbols for forward decls. · a7b04174
  Zachary Turner authored Aug 28, 2017
  a7b04174
- [Power9] Add new instructions for floating point status and control registers. · c35e4de3
  Stefan Pintilie authored Aug 28, 2017
  c35e4de3
- [InstCombine] Call hasNoSignedWrap instead of hasNoUnsignedWrap to get the NSW... · 3763f0e0
  Craig Topper authored Aug 28, 2017
  3763f0e0
- [Hexagon] Check for potential bank conflicts in post-RA scheduling · 2164a271
  Krzysztof Parzyszek authored Aug 28, 2017
  2164a271
- [AMDGPU] Fix regression in AMDGPULibCalls allowing native for doubles · 312c557b
  Stanislav Mekhanoshin authored Aug 28, 2017
  312c557b
- [AMDGPU] computeKnownBitsForTargetNode for 24 bit mul · dad7cf62
  Stanislav Mekhanoshin authored Aug 28, 2017
  dad7cf62
- [Hexagon] Break up DAG mutations into separate classes, move to subtarget · 95da97ec
  Krzysztof Parzyszek authored Aug 28, 2017
  95da97ec
- [Hexagon] Move pre-RA DAG mutations to scheduler constructor · 697297af
  Krzysztof Parzyszek authored Aug 28, 2017
  697297af
- [X86] Make 128/256-bit extract_subvector Legal instead of Custom. Move... · fa86fd92
  Craig Topper authored Aug 28, 2017
  fa86fd92
- [DAGCombiner] Teach visitEXTRACT_SUBVECTOR to turn extracts of BUILD_VECTOR... · 029a21df
  Craig Topper authored Aug 28, 2017
  029a21df
- Changed Dockerfiles to install LLVM into /usr/local · 867e36de
  Ilya Biryukov authored Aug 28, 2017
  867e36de
- [Support][CommandLine] Add cl::Option::setDefault() · 1587086f
  Evgeny Mankov authored Aug 28, 2017
  1587086f
- The current version of LLVM X86 disassembler incorrectly interprets some... · 574962a3
  Andrew V. Tischenko authored Aug 28, 2017
  574962a3
- [X86][Haswell] Updating HSW instruction scheduling information · d76f7b82
  Gadi Haber authored Aug 28, 2017
  d76f7b82
- Prune whitespaces in blank lines. · ac08375a
  NAKAMURA Takumi authored Aug 28, 2017
  ac08375a
- Untabify. · a1e97a77
  NAKAMURA Takumi authored Aug 28, 2017
  a1e97a77
- [X86] Use getUnpackl helper to create an ISD::VECTOR_SHUFFLE instead of using... · 33681161
  Craig Topper authored Aug 28, 2017
  33681161
- [X86] Add an early out to combineLoopMAddPattern and combineLoopSADPattern when SSE2 is disabled. · 2c77011d
  Craig Topper authored Aug 28, 2017
  2c77011d
- [Error] Add a handleExpected utility. · 5d06c23d
  Lang Hames authored Aug 28, 2017
  5d06c23d
- revert r310985 which breaks for the following case: · 191b24d3
  Dehao Chen authored Aug 27, 2017
  191b24d3
Aug 27, 2017
- [mips] Generate NMADD and NMSUB instructions when fneg node is present · f11daad1
  Petar Jovanovic authored Aug 27, 2017
  f11daad1
- [ARM] Tidy-up condition-code support functions · b81fa993
  Javed Absar authored Aug 27, 2017
  b81fa993
- [AVX512] Add more patterns for using masked moves for subvector extracts of... · 80075a5f
  Craig Topper authored Aug 27, 2017
  80075a5f
- [DAGCombiner] allow undef shuffle operands when eliminating bitcasts (PR34111) · a7a61d97
  Sanjay Patel authored Aug 27, 2017
  a7a61d97
- [x86] add haddps test for PR34111; NFC · 4e4ba615
  Sanjay Patel authored Aug 27, 2017
  4e4ba615
- [ARM] Tidy-up ARMAsmParser. NFC. · 17ee7c09
  Javed Absar authored Aug 27, 2017
  17ee7c09
- [LV] Fix PR34248 - recommit D32871 after revert r311304 · 1f58dda4
  Ayal Zaks authored Aug 27, 2017
  1f58dda4
- [X86] Adding more tests for horizontal [F]HADD/[F]SUB for AVX512 vectors types · 23eaf52d
  Jatin Bhateja authored Aug 27, 2017
  23eaf52d
- [X86] Add a target-specific DAG combine to combine extract_subvector from all... · 36bd247f
  Craig Topper authored Aug 27, 2017
  36bd247f
- [X86] Use getOnesVector instead of using DAG.getConstant(-1). · 71dab64a
  Craig Topper authored Aug 27, 2017
  71dab64a
- [NewGVN] Use `auto` when the type is obvious NFCI. · 9bdccb37
  Davide Italiano authored Aug 26, 2017
  9bdccb37