Commits · 6d675f4e35cc50746f888f9bed5ba22a8a4aed34 · Roger Ferrer / llvm-epi-0.8 · GitLab

This project is archived. Its data is read-only.

Nov 15, 2014
- [X86][SSE] Improve legal SHUFP and PSHUFD shuffle matching · 6d675f4e
  Simon Pilgrim authored Nov 15, 2014
  6d675f4e
- R600: Permute operands when selecting legacy min/max · 36094d78
  Matt Arsenault authored Nov 15, 2014
  36094d78
Nov 14, 2014
- ARM: refactor .cfi_def_cfa_offset emission. · 603d3165
  Tim Northover authored Nov 14, 2014
  603d3165
- ARM: correctly calculate the offset of FP in its push. · 9d2d218f
  Tim Northover authored Nov 14, 2014
  9d2d218f
- ARM: simplify test. · a0691c89
  Tim Northover authored Nov 14, 2014
  a0691c89
- R600/SI: Fix spilling of m0 register · bdd567d8
  Tom Stellard authored Nov 14, 2014
  bdd567d8
- R600/SI: Combine min3/max3 instructions · cc3c2b39
  Matt Arsenault authored Nov 14, 2014
  cc3c2b39
- R600/SI: Fix verifier error from a branch on IMPLICIT_DEF · 72858935
  Matt Arsenault authored Nov 14, 2014
  72858935
- R600/SI: Match integer min / max instructions · d28a7fde
  Matt Arsenault authored Nov 14, 2014
  d28a7fde
- R600/SI: Use S_BFE_I64 for 64-bit sext_inreg · 94812216
  Matt Arsenault authored Nov 14, 2014
  94812216
- [AVX512] Add 512b masked integer shift by immediate patterns. · 04400449
  Cameron McInally authored Nov 14, 2014
  04400449
- [PowerPC] Add VSX builtins for vec_div · 76746929
  Bill Schmidt authored Nov 14, 2014
  76746929
- X86: use getConstant rather than getTargetConstant behind BUILD_VECTOR. · d3be12a6
  Tim Northover authored Nov 14, 2014
  d3be12a6
- Allow the use of functions as typeinfo in landingpad clauses · 283bc2ed
  Reid Kleckner authored Nov 14, 2014
  283bc2ed
- First stage of call lowering for Mips fast-isel · d5c4196c
  Reed Kotler authored Nov 13, 2014
  d5c4196c
- R600/SI: Fix fmin_legacy / fmax_legacy matching for SI · da59f3de
  Matt Arsenault authored Nov 13, 2014
  da59f3de
Nov 13, 2014
- [x86] Add some tests for specific patterns of lane-flips combined with · 99b261ce
  Chandler Carruth authored Nov 13, 2014
  99b261ce
- [FastISel][AArch64] Don't bail during simple GEP instruction selection. · 0af310d0
  Juergen Ributzka authored Nov 13, 2014
  0af310d0
- R600/SI: Use s_movk_i32 · 77849929
  Matt Arsenault authored Nov 13, 2014
  77849929
- R600: Fix assert on empty function · 6ef66144
  Matt Arsenault authored Nov 13, 2014
  6ef66144
- R600: Error on initializer for LDS. · cc8d3b87
  Matt Arsenault authored Nov 13, 2014
  cc8d3b87
- R600/SI: Get rid of FCLAMP_SI pseudo · 1cffa4c1
  Matt Arsenault authored Nov 13, 2014
  1cffa4c1
- R600/SI: Allow commuting with src2_modifiers · 581a7a69
  Matt Arsenault authored Nov 13, 2014
  581a7a69
- R600/SI: Allow commuting some 3 op instructions · 95e48668
  Matt Arsenault authored Nov 13, 2014
  95e48668
- ARM: allow constpool entry to be moved to the user's block in all cases. · 631cc9ce
  Tim Northover authored Nov 13, 2014
  631cc9ce
- ARM: avoid duplicating branches during constant islands. · ab85dcc7
  Tim Northover authored Nov 13, 2014
  ab85dcc7
- ARM: add @llvm.arm.space intrinsic for testing ConstantIslands. · 650b0ee5
  Tim Northover authored Nov 13, 2014
  650b0ee5
- AVX-512: SINT_TO_FP cost model and some bugfixes · d5e95b57
  Elena Demikhovsky authored Nov 13, 2014
  d5e95b57
- [x86] Teach the vector shuffle lowering to make a more nuanced decision · fee91883
  Chandler Carruth authored Nov 13, 2014
  fee91883
- [x86] Don't form overly fragmented blends when splitting and · 253dd39a
  Chandler Carruth authored Nov 13, 2014
  253dd39a
- [CodeGenPrepare] Handle zero extensions in the TypePromotionHelper. · f5485bb0
  Quentin Colombet authored Nov 13, 2014
  f5485bb0
- [FastISel][AArch64] Optimize select when one of the operands is a 'true' or 'false' value. · 957a1454
  Juergen Ributzka authored Nov 13, 2014
  957a1454
- [FastISel][AArch64] Fold the cmp into the select when possible. · 424c5fd1
  Juergen Ributzka authored Nov 13, 2014
  424c5fd1
- [FastISel][AArch64] Extend 'select' lowering to support also i1 to i16. · d1a042ab
  Juergen Ributzka authored Nov 13, 2014
  d1a042ab
Nov 12, 2014
- Expose the number of Newton-Raphson iterations applied to the hardware's... · f6f7d5d1
  Sanjay Patel authored Nov 12, 2014
  f6f7d5d1
- [AVX512] Add integer shift by immediate intrinsics. · 73a6bca3
  Cameron McInally authored Nov 12, 2014
  73a6bca3
- Revert part of the PIC tests (TLS part) · 21c5353f
  Justin Hibbits authored Nov 12, 2014
  21c5353f
- Fix thet tests. · b296c973
  Justin Hibbits authored Nov 12, 2014
  b296c973
- Add support for small-model PIC for PowerPC. · a88b6057
  Justin Hibbits authored Nov 12, 2014
  a88b6057
- [mips][micromips] Add predicate 'InMicroMips' at CodeGen patterns for microMIPS instructions · fd888630
  Zoran Jovanovic authored Nov 12, 2014
  fd888630