Commits · 0c958fba14bbe65cba7a0dbf5b63a2be2ee75213 · Roger Ferrer / llvm-epi · GitLab

Dec 30, 2017
- [X86] Remove isel patterns for kshifts with types that don't support kshift natively. · 97cc7b03
  Craig Topper authored Dec 30, 2017
  97cc7b03
- [X86] Custom legalize vXi1 extract_subvector with KSHIFTR. · c5fd31a8
  Craig Topper authored Dec 30, 2017
  c5fd31a8
- [instsimplify] consistently handle undef and out of bound indices for... · e499bc30
  Philip Reames authored Dec 30, 2017
  e499bc30
Dec 29, 2017
- [MachineOperand] Fix LiveDebugVariables code after isRenamable change. · b3d126d6
  Geoff Berry authored Dec 29, 2017
  b3d126d6
- StructurizeCFG: Use phi iterator range · 8dcfa137
  Matt Arsenault authored Dec 29, 2017
  8dcfa137
- IR: Fix BasicBlock::phis for empty blocks · 344c0920
  Matt Arsenault authored Dec 29, 2017
  344c0920
- [mips] Provide correct descriptions of asm constraints in the comments. NFC · d41feef4
  Simon Atanasyan authored Dec 29, 2017
  d41feef4
- [mips] Replace assert by an error message · 970f686f
  Simon Atanasyan authored Dec 29, 2017
  970f686f
- AMDGPU: Use unique PSVs for buffer resources · e19bc2ee
  Matt Arsenault authored Dec 29, 2017
  e19bc2ee
- AMDGPU: Remove mayLoad/hasSideEffects from MIMG stores · d94b63d7
  Matt Arsenault authored Dec 29, 2017
  d94b63d7
- AMDGPU: Implement getTgtMemIntrinsic for images · 905f3518
  Matt Arsenault authored Dec 29, 2017
  905f3518
- [X86][SSE] Match PSHUFLW/PSHUFHW + PSHUFD vXi16 shuffle patterns (PR34686) · c701596e
  Simon Pilgrim authored Dec 29, 2017
  c701596e
- [AMDGPU][MC] Incorrect parsing of flat/global atomic modifiers · 414e0538
  Dmitry Preobrazhensky authored Dec 29, 2017
  414e0538
- [PowerPC] Fix for PR35688 - handle out-of-range values for r+r to r+i conversion · 4e1f5e07
  Nemanja Ivanovic authored Dec 29, 2017
  4e1f5e07
- [SCEV] Missing depth propagation in recursive call · 2c2b1f6d
  Max Kazantsev authored Dec 29, 2017
  2c2b1f6d
- Fix incorrect operand sizes for some MMX instructions: punpcklwd, punpcklbw and punpckldq. · 03ddad85
  Andrew V. Tischenko authored Dec 29, 2017
  03ddad85
- [PM] pass -debug-pass-manager flag into FunctionToLoopPassAdaptor's canonicalization PM · 02e7f024
  Fedor Sergeev authored Dec 29, 2017
  02e7f024
- Avoid modifying DbgInfo while looping in salvageDebuginfo · 2fb61343
  Dimitry Andric authored Dec 28, 2017
  2fb61343
Dec 28, 2017
- [SelectionDAG] Add creating new node debug messages for load, store, gather, and scatter. · c077ae68
  Craig Topper authored Dec 28, 2017
  c077ae68
- [X86] When lowering extending loads from v2i1/v4i1, if we have VLX, use a narrower extend. · 55cf8809
  Craig Topper authored Dec 28, 2017
  55cf8809
- [X86] Use ISD::CONCAT_VECTORS when splitting 256-bit loads in combineLoad. · c0b6cb1e
  Craig Topper authored Dec 28, 2017
  c0b6cb1e
- [X86] Fix inconsistencies in different places where we split loads/stores. · 4b311da3
  Craig Topper authored Dec 28, 2017
  4b311da3
- [SelectionDAG] Add some debug print messages to LegalizeVectorOps. · d5fed997
  Craig Topper authored Dec 28, 2017
  d5fed997
- [X86] Emit ISD::TRUNCATE instead of X86ISD::VTRUNC from... · 05cf1f33
  Craig Topper authored Dec 28, 2017
  05cf1f33
- [X86] Remove unnecessary patterns for sign extending vXi1 without VLX. · 88e26a99
  Craig Topper authored Dec 28, 2017
  88e26a99
- [WinEH] Don't emit state stores or EH thunks for available_externally functions · a2d119a0
  Reid Kleckner authored Dec 28, 2017
  a2d119a0
- [LTO] Simplify code. No functionality change intended. · 3afb1736
  Benjamin Kramer authored Dec 28, 2017
  3afb1736
- Remove superfluous copies in sample profiling. · 24cb28bb
  Benjamin Kramer authored Dec 28, 2017
  24cb28bb
- Revert r321377, it causes regression to https://reviews.llvm.org/P8055 . · 29697c13
  Guozhi Wei authored Dec 28, 2017
  29697c13
- Avoid int to string conversion in Twine or raw_ostream contexts. · 3a13ed60
  Benjamin Kramer authored Dec 28, 2017
  3a13ed60
- [RewriteStatepoints] Fix incorrect assertion · a13e163a
  Max Kazantsev authored Dec 28, 2017
  a13e163a
- [X86][SSE] Use PMADDWD for v4i32 multiplies with 17 or more leading zeros · 62411e4d
  Simon Pilgrim authored Dec 28, 2017
  62411e4d
- [InstCombine] Check for isa<Instruction> before using cast<> · 472689a1
  Simon Pilgrim authored Dec 28, 2017
  472689a1
- Revert "[memcpyopt] Teach memcpyopt to optimize across basic blocks" · 6d31001c
  Reid Kleckner authored Dec 28, 2017
  6d31001c
Dec 27, 2017
- [X86] Add CLWB to icelake. · 55cfa89f
  Craig Topper authored Dec 27, 2017
  55cfa89f
- [X86] Reimplement r321437 using custom lowering instead of as a DAG combine. · 72bbbeb2
  Craig Topper authored Dec 27, 2017
  72bbbeb2
- [AArch64] Change order of candidate FMLS patterns · 9439f549
  Matthew Simpson authored Dec 27, 2017
  9439f549
- [X86] Fix vmul combine for AVX1 targets. · 293f3430
  Benjamin Kramer authored Dec 27, 2017
  293f3430
- [InstCombine] Gracefully handle out of range extractelement indices · e7d032f1
  Simon Pilgrim authored Dec 27, 2017
  e7d032f1
- [DAGCombine] foldBinOpIntoSelect can fail to constant fold in some cases. · 6fad3cbc
  Simon Pilgrim authored Dec 27, 2017
  6fad3cbc