Commits · dd175bc6c42dc41f9d9e78e1b1e0e43a040e402d · Roger Ferrer / llvm-epi · GitLab

Aug 15, 2015
- [AArch64] Fix FMLS scalar-indexed-from-2s-after-neg patterns. · cd357872
  Ahmed Bougacha authored Aug 14, 2015
  cd357872
Aug 14, 2015
- AMDGPU/SI: Add missing spill class · bef1094e
  Tom Stellard authored Aug 14, 2015
  bef1094e
- Revert "[ARM] Fix MachO CPU Subtype selection" · 980b6cc4
  Renato Golin authored Aug 14, 2015
  980b6cc4
- [ARM] Fix MachO CPU Subtype selection · 2f079be7
  Vedant Kumar authored Aug 14, 2015
  2f079be7
- [x86] fix allowsMisalignedMemoryAccess() implementation · ed502905
  Sanjay Patel authored Aug 14, 2015
  ed502905
- Revert "Centralize the information about which object format we are using." · dbaf0498
  Rafael Espindola authored Aug 14, 2015
  dbaf0498
- don't repeaat function names in comments; NFC · 2e75341b
  Sanjay Patel authored Aug 14, 2015
  2e75341b
- Centralize the information about which object format we are using. · 90eb70c8
  Rafael Espindola authored Aug 14, 2015
  90eb70c8
- [AArch64] FMINNAN/FMAXNAN on f16 is not legal. · 63be1987
  James Molloy authored Aug 14, 2015
  63be1987
- [IR] Add token types · b611e3f5
  David Majnemer authored Aug 14, 2015
  b611e3f5
- PowerPC: remove dead initialization (NFC) · 3e190cb0
  Saleem Abdulrasool authored Aug 14, 2015
  3e190cb0
Aug 13, 2015
- [AMDGPU] Use the general SMAX/SMIN/UMAX/UMIN pattern matching and remove the AMDGPU implementation · 72182518
  Simon Pilgrim authored Aug 13, 2015
  72182518
- [AArch64] Provide "too few operands" diags on short-form NEON also. · 80e4ac80
  Ahmed Bougacha authored Aug 13, 2015
  80e4ac80
- [X86][SSE] Use the general SMAX/SMIN/UMAX/UMIN pattern matching and remove the X86 implementation · 4a8d6b3b
  Simon Pilgrim authored Aug 13, 2015
  4a8d6b3b
- Remove and forbid raw_svector_ostream::flush() calls. · 556b21aa
  Yaron Keren authored Aug 13, 2015
  556b21aa
- Scalar to vector conversions using direct moves · 1c39ca65
  Nemanja Ivanovic authored Aug 13, 2015
  1c39ca65
- [ARM] FMINNAN/FMAXNAN of f64 are not legal. · 31117875
  James Molloy authored Aug 13, 2015
  31117875
- [ARM] Allow vmin/vmax of scalars to be emitted without UseNEONForFP. · c71f78f4
  James Molloy authored Aug 13, 2015
  c71f78f4
- [SystemZ] Support large LLVM IR struct return values · a887f062
  Ulrich Weigand authored Aug 13, 2015
  a887f062
- [ARM] Reorganise and simplify thumb-1 load/store selection · 68acdcb4
  John Brawn authored Aug 13, 2015
  68acdcb4
- [AArch64] Also custom-lowering mismatched vector/f16 FCOPYSIGN. · 2a97b1bc
  Ahmed Bougacha authored Aug 13, 2015
  2a97b1bc
Aug 12, 2015
- WebAssembly: floating-point comparisons · 71d29ace
  JF Bastien authored Aug 12, 2015
  71d29ace
- 80-cols; NFC · 2366168b
  Sanjay Patel authored Aug 12, 2015
  2366168b
- fix typo; NFC · dc87d144
  Sanjay Patel authored Aug 12, 2015
  dc87d144
- [mips][microMIPS] Create microMIPS64r6 subtarget and implement DALIGN, DAUI,... · 366783e1
  Zoran Jovanovic authored Aug 12, 2015
  366783e1
- [X86] Disable mul -> shl + lea combine when compiling for minsize · fe0d9bb6
  Michael Kuperstein authored Aug 12, 2015
  fe0d9bb6
- [X86] Allow x86 call frame optimization to fold more loads into pushes · bc7f99a3
  Michael Kuperstein authored Aug 12, 2015
  bc7f99a3
- AMDGPU: Fix assert on dbg_value instructions · c5746865
  Matt Arsenault authored Aug 12, 2015
  c5746865
- [InstCombine] Move SSE/AVX vector blend folding to instcombiner · 8c049d5c
  Simon Pilgrim authored Aug 12, 2015
  8c049d5c
- X86: hoist a condition into a variable (NFC) · 9e5f2a96
  Saleem Abdulrasool authored Aug 12, 2015
  9e5f2a96
- [x86] enable machine combiner reassociations for 256-bit vector FP mul/add · 260b6d36
  Sanjay Patel authored Aug 12, 2015
  260b6d36
- PseudoSourceValue: Transform the mips subclass to target independent subclasses · 5659a2f9
  Alex Lorenz authored Aug 11, 2015
  5659a2f9
- PseudoSourceValue: Replace global manager with a manager in a machine function. · e40c8a2b
  Alex Lorenz authored Aug 11, 2015
  e40c8a2b
- PseudoSourceValue: Introduce a 'PSVKind' enumerator. · c49e4fe9
  Alex Lorenz authored Aug 11, 2015
  c49e4fe9
- Use 32-bit divides instead of 64-bit divides where possible. · 438ffe5e
  Mark Heffernan authored Aug 11, 2015
  438ffe5e
Aug 11, 2015
- WebAssembly: implement comparison. · da06bce8
  JF Bastien authored Aug 11, 2015
  da06bce8
- [x86] enable machine combiner reassociations for 128-bit vector single/double multiplies · 2c6a0157
  Sanjay Patel authored Aug 11, 2015
  2c6a0157
- WebAssembly: implement WebAssemblyTargetLowering::getTargetNodeName · 480c8408
  JF Bastien authored Aug 11, 2015
  480c8408
- Use llvm::make_unique to fix the MSVC build. · 3adc7ce9
  Rafael Espindola authored Aug 11, 2015
  3adc7ce9
- [X86] Allow merging of immediates within a basic block for code size savings · 243c073a
  Michael Kuperstein authored Aug 11, 2015
  243c073a