Commits · edb6f0385219f997d2aeb0eeb1317504dad2d28f · Roger Ferrer / llvm-epi · GitLab

Jan 14, 2015
- R600/SI: Remove some redudant load testcases. · edb6f038
  Matt Arsenault authored Jan 14, 2015
  edb6f038
- R600/SI: Fix bad code with unaligned byte vector loads · e6986636
  Matt Arsenault authored Jan 14, 2015
  e6986636
- Implement new way of expanding extloads. · bd223423
  Matt Arsenault authored Jan 14, 2015
  bd223423
- Revert "r225811 - Revert "r225808 - [PowerPC] Add StackMap/PatchPoint support"" · 934361a4
  Hal Finkel authored Jan 14, 2015
  934361a4
- Insert random noops to increase security against ROP attacks (llvm) · dcdd5ad2
  JF Bastien authored Jan 14, 2015
  dcdd5ad2
- CodeGen support for x86_64 SEH catch handlers in LLVM · 0a57f655
  Reid Kleckner authored Jan 14, 2015
  0a57f655
- [AVX512] Add 16x32 unpck tests as well · d23c88db
  Adam Nemet authored Jan 13, 2015
  d23c88db
Jan 13, 2015
- Fix function names in tests from r225838. · f07a0ba9
  Adam Nemet authored Jan 13, 2015
  f07a0ba9
- [AVX512] Unpack support in new shuffle lowering · e5dbcb7f
  Adam Nemet authored Jan 13, 2015
  e5dbcb7f
- R600: Implement getRsqrtEstimate · e93d06a5
  Matt Arsenault authored Jan 13, 2015
  e93d06a5
- R600: Make cttz / ctlz cheap to speculate · b56d8433
  Matt Arsenault authored Jan 13, 2015
  b56d8433
- Use the integrated assembler as default on SystemZ · bd039299
  Ulrich Weigand authored Jan 13, 2015
  bd039299
- Use the integrated assembler as default on PowerPC · 6b577e26
  Ulrich Weigand authored Jan 13, 2015
  6b577e26
- Revert "r225808 - [PowerPC] Add StackMap/PatchPoint support" · 63fb9281
  Hal Finkel authored Jan 13, 2015
  63fb9281
- [PowerPC] Add StackMap/PatchPoint support · 821befd5
  Hal Finkel authored Jan 13, 2015
  821befd5
- [mips][microMIPS] Fix issue with 16b instructions in jr instruction delay slot · e7cad7a1
  Jozef Kolek authored Jan 13, 2015
  e7cad7a1
- Rename llvm.recoverframeallocation to llvm.framerecover · 3542ace6
  Reid Kleckner authored Jan 13, 2015
  3542ace6
- Add the llvm.frameallocate and llvm.recoverframeallocation intrinsics · e9b89318
  Reid Kleckner authored Jan 13, 2015
  e9b89318
- Combine fcmp + select to fminnum / fmaxnum if no nans and legal · a982e4f8
  Matt Arsenault authored Jan 13, 2015
  a982e4f8
- musttail: Only set the inreg flag for fastcall and vectorcall · bba20f06
  Reid Kleckner authored Jan 12, 2015
  bba20f06
Jan 12, 2015
- Debug info: Factor out the creation of DWARF expressions from AsmPrinter · b16d9ebb
  Adrian Prantl authored Jan 12, 2015
  b16d9ebb
- [X86] Also create+widen FMIN/FMAX nodes for v2f32. · 291833b9
  Ahmed Bougacha authored Jan 12, 2015
  291833b9
- [X86] Make SSE min/max testcases more explicit. NFC. · 66fde538
  Ahmed Bougacha authored Jan 12, 2015
  66fde538
- R600/SI: Use RegisterOperands to specify which operands can accept immediates · b6550529
  Tom Stellard authored Jan 12, 2015
  b6550529
- [PowerPC] Fix calls to non-function objects · 87deb0b8
  Hal Finkel authored Jan 12, 2015
  87deb0b8
Jan 11, 2015
- Revert most of r225597 · 14141f94
  David Majnemer authored Jan 11, 2015
  14141f94
- X86: Properly decode shuffle masks when the constant pool type is weird · 292d0c79
  David Majnemer authored Jan 11, 2015
  292d0c79
- X86: teach X86TargetLowering about L,M,O constraints · 9cf2679d
  Saleem Abdulrasool authored Jan 11, 2015
  9cf2679d
- [x86] Remove some windows line endings that snuck into the tests here. · c491f72e
  Chandler Carruth authored Jan 11, 2015
  c491f72e
- Fix PR22179. · 81401d4b
  Sanjoy Das authored Jan 10, 2015
  81401d4b
Jan 10, 2015
- [X86][SSE] Improved (v)insertps shuffle matching · 94a4cc02
  Simon Pilgrim authored Jan 10, 2015
  94a4cc02
- [PowerPC] Mark zext of a small scalar load as free · 5d5d1539
  Hal Finkel authored Jan 10, 2015
  5d5d1539
- Fully fix Bug #22115. · 654346e6
  Justin Hibbits authored Jan 10, 2015
  654346e6
Jan 09, 2015
- [X86][SSE] Avoid vector byte shuffles with zero by using pshufb to create zeros · ec1f2c2c
  Simon Pilgrim authored Jan 09, 2015
  ec1f2c2c
- [mips] Add support for accessing $gp as a named register. · 1440bb2a
  Daniel Sanders authored Jan 09, 2015
  1440bb2a
- RegisterCoalescer: Fix removeCopyByCommutingDef with subreg liveness · 7e873845
  Matthias Braun authored Jan 09, 2015
  7e873845
- [PowerPC] Fold [sz]ext with fp_to_int lowering where possible · 6c39269a
  Hal Finkel authored Jan 09, 2015
  6c39269a
Jan 08, 2015
- [PowerPC] Mark all instructions as non-cheap for MachineLICM · 3c0952b0
  Hal Finkel authored Jan 08, 2015
  3c0952b0
- [ARM] Fix a bug in constant island pass that was triggering an assertion. · 442b40c2
  Akira Hatanaka authored Jan 08, 2015
  442b40c2
- Add saving and restoring of r30 to the prologue and epilogue, respectively · 98a532dd
  Justin Hibbits authored Jan 08, 2015
  98a532dd