Commits · e25a6fd390e77d953001e2087fc54c2c1169cbe4 · Roger Ferrer / llvm-epi · GitLab

Nov 07, 2017
- [SLP] Fix PR35047: Fix default cost model for cast op in X86. · e25a6fd3
  Alexey Bataev authored Nov 07, 2017
  e25a6fd3
- Add a -D flag to FileCheck to define variables · 46e1fd61
  Alexander Richardson authored Nov 07, 2017
  46e1fd61
- [X86] Regenerate select tests · 9a6b720f
  Simon Pilgrim authored Nov 07, 2017
  9a6b720f
- [GlobalISel] Enable legalizing non-power-of-2 sized types. · af9814a1
  Kristof Beyls authored Nov 07, 2017
  af9814a1
- [CGP] Disable Select instruction handling in optimizeMemoryInst. NFC · 36520029
  Serguei Katkov authored Nov 07, 2017
  36520029
- [X86] Don't clobber reserved registers with stack adjustments · c02b237e
  Bjorn Steinbrink authored Nov 07, 2017
  c02b237e
- [X86] Add patterns to fold a 64-bit load into the EVEX vcvtph2ps instructions. · e7fb3002
  Craig Topper authored Nov 07, 2017
  e7fb3002
- [X86] Add patterns for folding a v16i8 with the VEX vcvtph2ps intrinsics. · 0231b1d4
  Craig Topper authored Nov 07, 2017
  0231b1d4
- [X86] Add a test for a 128-bit vector load feeding a cvtph2ps intrinsic. · 65fc5332
  Craig Topper authored Nov 07, 2017
  65fc5332
- [X86] Remove alignment from a load in the f16c intrinsic test. The alignment... · 8942b33f
  Craig Topper authored Nov 07, 2017
  8942b33f
- [X86] Add support for using EVEX instructions for the legacy vcvtph2ps intrinsics. · cf8e6d0a
  Craig Topper authored Nov 07, 2017
  cf8e6d0a
- [X86] Add AVX512VL command line to f16c intrinsic test to show missed EVEX... · 75510dd6
  Craig Topper authored Nov 07, 2017
  75510dd6
- [X86] Use IMPLICIT_DEF in VEX/EVEX vcvtss2sd/vcvtsd2ss patterns instead of a COPY_TO_REGCLASS. · afc3c820
  Craig Topper authored Nov 07, 2017
  afc3c820
- Make DIExpression::createFragmentExpression() return an Optional. · 25a09dd4
  Adrian Prantl authored Nov 07, 2017
  25a09dd4
- [IPO/LowerTypesTest] Skip blockaddress(es) when replacing uses. · 1a46affb
  Davide Italiano authored Nov 07, 2017
  1a46affb
Nov 06, 2017
- InstCombine: salvage the debug info of DCE'ed add instructions. · 182f9fea
  Adrian Prantl authored Nov 06, 2017
  182f9fea
- [X86] Make FeatureAVX512 imply FeatureF16C. · 428a4e63
  Craig Topper authored Nov 06, 2017
  428a4e63
- [ValueTracking] readonly (const) is a requirement for converting sqrt to llvm.sqrt; nnan is not · 86d24f16
  Sanjay Patel authored Nov 06, 2017
  86d24f16
- Revert r317510 "[InstCombine] Pull shifts through a select plus binop with constant" · 8c4b10e8
  Hans Wennborg authored Nov 06, 2017
  8c4b10e8
- [MIRPrinter] Use %subreg.xxx syntax for subregister index operands · a42ed3e3
  Bjorn Pettersson authored Nov 06, 2017
  a42ed3e3
- [InstCombine] Pull shifts through a select plus binop with constant · 89176473
  Craig Topper authored Nov 06, 2017
  89176473
- Adds code to PPC ISEL lowering to recognize byte inserts from vector_shuffles,... · 030621bb
  Graham Yiu authored Nov 06, 2017
  030621bb
- Include already promoted counts when computing SUM for VP. · 5d2a1a50
  Dehao Chen authored Nov 06, 2017
  5d2a1a50
- [PPC] Use xxbrd to speed up bswap64 · e3b8d9a3
  Guozhi Wei authored Nov 06, 2017
  e3b8d9a3
- AMDGPU: Select v_mad_u64_u32 and v_mad_i64_i32 · 4f6318fe
  Matt Arsenault authored Nov 06, 2017
  4f6318fe
- [IR] redefine 'UnsafeAlgebra' / 'reassoc' fast-math-flags and add 'trans' fast-math-flag · 629c4115
  Sanjay Patel authored Nov 06, 2017
  629c4115
- [SLP] Test for PR35047, NFC. · 676350c1
  Alexey Bataev authored Nov 06, 2017
  676350c1
- [AMDGPU] Change alloca addr space of r600 to 5 for amdgiz environment · cc56a8b1
  Yaxun Liu authored Nov 06, 2017
  cc56a8b1
- [AMDGPU] Fix assertion due to assuming pointer in default addr space is 32 bit · 1ac16619
  Yaxun Liu authored Nov 06, 2017
  1ac16619
- [mips] Add movep for microMIPS32R6 and fix microMIPS32r3 version · 169df4e2
  Simon Dardis authored Nov 06, 2017
  169df4e2
- [LV][X86] update the cost of interleaving mem. access of floats · 66917583
  Mohammed Agabaria authored Nov 06, 2017
  66917583
- [mips] Fix PR35140 · e5779538
  Simon Dardis authored Nov 06, 2017
  e5779538
- [X86][AVX512] Improve lowering of AVX512 test intrinsics · bb86686a
  Uriel Korach authored Nov 06, 2017
  bb86686a
- X86 ISel: Basic support for variable-index vector permutations · 31226980
  Zvi Rackover authored Nov 06, 2017
  31226980
- [test] Add test files that were missed from SVN r317459 · 568c18cd
  Martin Storsjö authored Nov 06, 2017
  568c18cd
- [ObjectYAML] Map relocation types for COFF ARMNT and ARM64 · bed0c519
  Martin Storsjö authored Nov 06, 2017
  bed0c519
- [x86][AVX512] Lowering Broadcastm intrinsics to LLVM IR · 7b705f1f
  Jina Nahias authored Nov 06, 2017
  7b705f1f
- [X86] Use EVEX encoded intrinsics for legacy FMA intrinsics when possible. · 70eaeae7
  Craig Topper authored Nov 06, 2017
  70eaeae7
- [X86] Add avx512vl command line to fma-instrinsics-x86.ll · 25cfa4cb
  Craig Topper authored Nov 06, 2017
  25cfa4cb
- [X86] Simplify command lines on the fma-instrinsics-x86.ll test and add -show-mc-encoding. · 7e48aa89
  Craig Topper authored Nov 06, 2017
  7e48aa89