Commits · e6b4ba1c41f35fccdd9dac4967cdd9bbc2074fb9 · Roger Ferrer / llvm-epi · GitLab

Jul 23, 2014
- In unroll pragma syntax and loop hint metadata, change "enable" forms to a new... · e6b4ba1c
  Mark Heffernan authored Jul 23, 2014
  e6b4ba1c
- [AArch64] Lower sdiv x, pow2 using add + select + shift. · 17020f96
  Chad Rosier authored Jul 23, 2014
  17020f96
- [SKX] Enabling mask instructions: encoding, lowering · 74acbb77
  Robert Khasanov authored Jul 23, 2014
  74acbb77
- ARM: spot SBFX-compatbile code expressed with sign_extend_inreg · 14ff2df0
  Tim Northover authored Jul 23, 2014
  14ff2df0
- ARM: add patterns for [su]xta[bh] from just a shift. · 7ad2a0e0
  Tim Northover authored Jul 23, 2014
  7ad2a0e0
- [ARM] Make the assembler reject unpredictable pre/post-indexed ARM STRB instructions. · 27272791
  Tilmann Scheller authored Jul 23, 2014
  27272791
- AArch64: remove "arm64_be" support in favour of "aarch64_be". · 35910d7f
  Tim Northover authored Jul 23, 2014
  35910d7f
- [ARM] Make the assembler reject unpredictable pre/post-indexed ARM STR instructions. · 3352a58d
  Tilmann Scheller authored Jul 23, 2014
  3352a58d
- Revert r211771. It was: "[X86] Improve the selection of SSE3/AVX addsub instructions". · 842355e9
  Andrea Di Biagio authored Jul 23, 2014
  842355e9
- [x86] Clean up a test case to use check labels and spell out the exact · 30253f32
  Chandler Carruth authored Jul 23, 2014
  30253f32
- [ARM] Add regression test for the earlyclobber constraint of ARM STRB. · bd0c4576
  Tilmann Scheller authored Jul 23, 2014
  bd0c4576
- [ARM] Add earlyclobber constraint to pre/post-indexed ARM STRH instructions. · c28f0d58
  Tilmann Scheller authored Jul 23, 2014
  c28f0d58
- [SDAG] Make the DAGCombine worklist not grow endlessly due to duplicate · 9a0051cd
  Chandler Carruth authored Jul 23, 2014
  9a0051cd
- We may visit a call that uses an alloca multiple times in callUsesLocalStack,... · aba900c2
  Nick Lewycky authored Jul 23, 2014
  aba900c2
- Rework to let RuntimeDyld/X86/MachO_x86-64_PIC_relocations.s pass on win32. · 0a6af439
  NAKAMURA Takumi authored Jul 23, 2014
  0a6af439
- Suppress a test on win32 for now,... · 24c0b46e
  NAKAMURA Takumi authored Jul 23, 2014
  24c0b46e
- RuntimeDyld/X86/MachO_x86-64_PIC_relocations.s: Use %/T here, or sed(1) would... · 7d793879
  NAKAMURA Takumi authored Jul 23, 2014
  7d793879
- XFAIL the test on MIPS · 7325ac39
  Juergen Ributzka authored Jul 22, 2014
  7325ac39
- [FastIsel][AArch64] Add support for the FastLowerCall and FastLowerIntrinsicCall target-hooks. · 2581fa50
  Juergen Ributzka authored Jul 22, 2014
  2581fa50
- [AArch64] Use CHECK-LABEL in ARM64 ABI unit tests. · 88f6faf1
  Juergen Ributzka authored Jul 22, 2014
  88f6faf1
- [MCJIT] Refactor and add stub inspection to the RuntimeDyldChecker framework. · f7acddde
  Lang Hames authored Jul 22, 2014
  f7acddde
Jul 22, 2014
- [RuntimeDyld][MachO][AArch64] Add a helper function for encoding addends in instructions. · f5609288
  Juergen Ributzka authored Jul 22, 2014
  f5609288
- This patch implements optimization as mentioned in PR19753: Optimize... · 3a8c2c1e
  Suyog Sarda authored Jul 22, 2014
  3a8c2c1e
- Added InstCombine transform for pattern "(A & B) ^ (A ^ B) -> (A | B)" · b60ec909
  Suyog Sarda authored Jul 22, 2014
  b60ec909
- Added InstCombine Transform for patterns: · d64faf6c
  Suyog Sarda authored Jul 22, 2014
  d64faf6c
- Make use of the align parameter attribute for all pointer arguments · ccc70906
  Hal Finkel authored Jul 22, 2014
  ccc70906
- X86: drop relocations on __eh_frame sections globally. · 0942e390
  Tim Northover authored Jul 22, 2014
  0942e390
- This patch implements transform for pattern "(A | B) ^ (~A) -> (A | ~B)". · 521237ca
  Suyog Sarda authored Jul 22, 2014
  521237ca
- [OCaml] Don't truncate constants over 32 bits in Llvm.const_int. · daacd614
  Peter Zotov authored Jul 22, 2014
  daacd614
- [mips] Fix two patterns that select i32's (for MIPS32r6) / i64's (for MIPS64r6) · 319f0ff3
  Sasa Stankovic authored Jul 22, 2014
  319f0ff3
- AVX-512: Fixed intrinsic of VSQRTPS/PD instructions. · f164859e
  Elena Demikhovsky authored Jul 22, 2014
  f164859e
- Revert of r213521. This change introduced a non-hermetic test (depending on a · a0cc1654
  Richard Smith authored Jul 22, 2014
  a0cc1654
- Rename metadata llvm.loop.vectorize.unroll to llvm.loop.vectorize.interleave. · 9d20e427
  Mark Heffernan authored Jul 21, 2014
  9d20e427
- Add some tests for NVPTX lowering of cmpxchg · 24d8aacd
  Eli Bendersky authored Jul 21, 2014
  24d8aacd
Jul 21, 2014
- Revert "Recommit r212203: Don't try to construct debug LexicalScopes hierarchy... · 26f2268c
  David Blaikie authored Jul 21, 2014
  26f2268c
- Replace the result usages while legalizing cmpxchg. · 63bee2a2
  Logan Chien authored Jul 21, 2014
  63bee2a2
- R600/SI: Add instruction shrinking pass · 1aaad697
  Tom Stellard authored Jul 21, 2014
  1aaad697
- R600/SI: Clean up some of the unused REGISTER_{LOAD,STORE} code · e812f2fd
  Tom Stellard authored Jul 21, 2014
  e812f2fd
- R600/SI: Use scratch memory for large private arrays · b02094e1
  Tom Stellard authored Jul 21, 2014
  b02094e1
- [mips] Do not emit '.module fp=...' unless we really need to. · e22244b7
  Daniel Sanders authored Jul 21, 2014
  e22244b7