Commits · fc501a9223e2e8ea8202fca06a58d6750f9a73a8 · Roger Ferrer / llvm-epi · GitLab

Jul 25, 2018
- [X86] Use a shift plus an lea for multiplying by a constant that is a power of 2 plus 2/4/8. · fc501a92
  Craig Topper authored Jul 25, 2018
  fc501a92
- [X86] Expand mul by pow2 + 2 using a shift and two adds similar to what we do for pow2 - 2. · 5be253d9
  Craig Topper authored Jul 25, 2018
  5be253d9
- [X86] Use a two lea sequence for multiply by 37, 41, and 73. · 56c104f1
  Craig Topper authored Jul 24, 2018
  56c104f1
- [X86] Change multiply by 26 to use two multiplies by 5 and an add instead of... · f8fcee70
  Craig Topper authored Jul 24, 2018
  f8fcee70
Jul 24, 2018
- [X86] When expanding a multiply by a negative of one less than a power of 2,... · 5ddc0a2b
  Craig Topper authored Jul 24, 2018
  5ddc0a2b
- [X86] Generalize the multiply by 30 lowering to generic multipy by power 2 minus 2. · 6d29891b
  Craig Topper authored Jul 24, 2018
  6d29891b
- [X86] Change multiply by 19 to use (9 * X) * 2 + X instead of (5 * X) * 4 - 1. · 86d6320b
  Craig Topper authored Jul 24, 2018
  86d6320b
- [MachineOutliner][NFC] Move target frame info into OutlinedFunction · 69f517df
  Jessica Paquette authored Jul 24, 2018
  69f517df
- Put "built-in" function definitions in global Used list, for LTO. (fix bug 34169) · e06bac47
  Peter Collingbourne authored Jul 24, 2018
  e06bac47
- [x86] Teach the x86 backend that it can fold between TCRETURNm* and TCRETURNr*... · c9313a9e
  Chandler Carruth authored Jul 24, 2018
  c9313a9e
- [MachineOutliner][NFC] Make Candidates own their call information · fca55129
  Jessica Paquette authored Jul 24, 2018
  fca55129
- [mips] Fix local dynamic TLS with Sym64 · 28ded4ee
  Simon Atanasyan authored Jul 24, 2018
  28ded4ee
- [x86/SLH] Extract the core register hardening logic to a low-level · 54529146
  Chandler Carruth authored Jul 24, 2018
  54529146
- [x86/SLH] Tidy up a comment, using doxygen structure and wording it to · 376113da
  Chandler Carruth authored Jul 24, 2018
  376113da
- [ARM] Disable ARMCodeGenPrepare by default · 8b93e82c
  Sam Parker authored Jul 24, 2018
  8b93e82c
- AMDGPU/GlobalISel: Legalize G_INSERT · b7f19e6d
  Tom Stellard authored Jul 24, 2018
  b7f19e6d
- AMDGPU/GlobalISel: Remove unnecessary legality constraint for G_EXTRACT · 2d37929c
  Tom Stellard authored Jul 24, 2018
  2d37929c
- [x86/SLH] Simplify the code for hardening a loaded value. NFC. · 66fbbbca
  Chandler Carruth authored Jul 24, 2018
  66fbbbca
- [x86/SLH] Remove complex SHRX-based post-load hardening. · b46c22de
  Chandler Carruth authored Jul 24, 2018
  b46c22de
- [AArch64] Use MCAsmInfoMicrosoft and MCAsmInfoGNUCOFF as base classes · c2b70140
  Martin Storsjö authored Jul 23, 2018
  c2b70140
Jul 23, 2018
- Re-land r335297 "[X86] Implement more of x86-64 large and medium PIC code models" · 980c4df0
  Reid Kleckner authored Jul 23, 2018
  980c4df0
- [Hexagon] Handle unnamed globals in HexagonConstExpr · 9500a24f
  Krzysztof Parzyszek authored Jul 23, 2018
  9500a24f
- [ARM] Use unique_ptr to fix memory leak introduced in r337701 · 58407ca0
  Fangrui Song authored Jul 23, 2018
  58407ca0
- OpChain has subclasses, so add a virtual destructor. · e5daf612
  Jordan Rupprecht authored Jul 23, 2018
  e5daf612
- [ARM] Follow-up to r337709. · d773cb99
  Matt Morehouse authored Jul 23, 2018
  d773cb99
- [ARM] Add doFinalization() to ARMCodeGenPrepare pass. · a70685f6
  Matt Morehouse authored Jul 23, 2018
  a70685f6
- [ARM][NFC] ParallelDSP reorganisation · 89a3799a
  Sam Parker authored Jul 23, 2018
  89a3799a
- [SystemZ] Fix dumpSU() method in SystemZHazardRecognizer. · 59c94bec
  Jonas Paulsson authored Jul 23, 2018
  59c94bec
- [ARM] ARMCodeGenPrepare backend pass · 3828c6ff
  Sam Parker authored Jul 23, 2018
  3828c6ff
- [NFC][MCA] ZnVer1: Update RegisterFile to identify false dependencies on... · 52b85377
  Roman Lebedev authored Jul 23, 2018
  52b85377
- [x86/SLH] Fix a bug where we would harden tail calls twice -- once as · 1d926fb9
  Chandler Carruth authored Jul 23, 2018
  1d926fb9
- [x86/SLH] Rename and comment the main hardening function. NFC. · 0477b401
  Chandler Carruth authored Jul 23, 2018
  0477b401
Jul 22, 2018
- [X86] Remove the max vector width restriction from combineLoopMAddPattern and... · b2a626b5
  Craig Topper authored Jul 22, 2018
  b2a626b5
Jul 21, 2018
- [mips] Factor out register class selection for global base register. NFC · 2c5b18f7
  Simon Atanasyan authored Jul 21, 2018
  2c5b18f7
- [mips] Move out the WrapperPat declaration from the NotInMicroMips predicate · ecd1e0af
  Simon Atanasyan authored Jul 21, 2018
  ecd1e0af
Jul 20, 2018
- AMDGPU: Use existing function to check for VGPRs · 5ae765e6
  Matt Arsenault authored Jul 20, 2018
  5ae765e6
- Revert "[X86][AVX] Convert X86ISD::VBROADCAST demanded elts combine to use... · 64c7fa32
  Benjamin Kramer authored Jul 20, 2018
  64c7fa32
- [X86] Remove isel patterns for MOVSS/MOVSD ISD opcodes with integer types. · 28ac623f
  Craig Topper authored Jul 20, 2018
  28ac623f
- [X86] Remove what appear to be unnecessary uses of DCI.CombineTo · 6194ccf8
  Craig Topper authored Jul 20, 2018
  6194ccf8
- [X86][XOP] Fix SUB constant folding for VPSHA/VPSHL shift lowering · 70fcd0f4
  Simon Pilgrim authored Jul 20, 2018
  70fcd0f4