Commits · f116e4acc7f2a3187ffabb3e3286db416121ff0c · Roger Ferrer / llvm-epi · GitLab

Feb 07, 2016
- [X86][SSE] Resolve target shuffle inputs to sentinels to permit more combines · f116e4ac
  Simon Pilgrim authored Feb 07, 2016
  f116e4ac
- [X86][SSE] Regenerate PSHUFB shuffle mask comments tests · a8d76d87
  Simon Pilgrim authored Feb 07, 2016
  a8d76d87
- [X86][SSE] Added support for MOVHPD/MOVLPD + MOVHPS/MOVLPS shuffle decoding. · a3d67447
  Simon Pilgrim authored Feb 07, 2016
  a3d67447
- [X86][AVX512] add intrinsics of Scalar FP to integer conversion with rounding mode · ad5c3fc4
  Asaf Badouh authored Feb 07, 2016
  ad5c3fc4
- AVX512: VPBROADCASTB/W/D/Q from GPR intrinsics implementation. · 0aeda374
  Igor Breger authored Feb 07, 2016
  0aeda374
Feb 06, 2016
- [X86][AVX2] Regenerated broadcast domain tests · 4108368a
  Simon Pilgrim authored Feb 06, 2016
  4108368a
- [X86][SSE] Add tests for MOVHLPS/MOVLHPS shuffle lowering. · 672808a8
  Simon Pilgrim authored Feb 06, 2016
  672808a8
- [X86][AVX512] Added support for VPMOVZX shuffle decoding. · 0acc32a3
  Simon Pilgrim authored Feb 06, 2016
  0acc32a3
- [X86][AVX512] Fixed prefix ordering for lzcnt tests. · 83e04913
  Simon Pilgrim authored Feb 06, 2016
  83e04913
- [X86][SSE] Regenerate vector shift tests · a09a154c
  Simon Pilgrim authored Feb 06, 2016
  a09a154c
- line endings fix · 63b1ecab
  Simon Pilgrim authored Feb 06, 2016
  63b1ecab
- [X86][SSE] Don't replace an existing 32-bit load with its duplicate · 9e369f2a
  Simon Pilgrim authored Feb 06, 2016
  9e369f2a
Feb 05, 2016
- AMDGPU: Account for LDS alignment · 7f83397d
  Matt Arsenault authored Feb 05, 2016
  7f83397d
- AMDGPU: Preserve alignments on new created globals · cf84e26f
  Matt Arsenault authored Feb 05, 2016
  cf84e26f
- Some stackslots are allocated to vregs which have no real reference. · a62f0589
  Wei Mi authored Feb 05, 2016
  a62f0589
- [WebAssembly] Update the select instructions' operand orders to match the spec. · d46b0926
  Dan Gohman authored Feb 05, 2016
  d46b0926
- Add the missing test case for PR26193 · d05e072b
  Nemanja Ivanovic authored Feb 05, 2016
  d05e072b
- Revert "[AArch64] Improve load/store optimizer to handle LDUR + LDR (take 3)." · 6274e522
  Renato Golin authored Feb 05, 2016
  6274e522
- Fix for PR 26356 · b6fdce4c
  Nemanja Ivanovic authored Feb 04, 2016
  b6fdce4c
Feb 04, 2016
- Provide a test case for rl259798 · 220b4fe4
  Nemanja Ivanovic authored Feb 04, 2016
  220b4fe4
- [X86][SSE] Select domain for 32/64-bit partial loads for EltsFromConsecutiveLoads · 7823fd25
  Simon Pilgrim authored Feb 04, 2016
  7823fd25
- [AArch64] Improve load/store optimizer to handle LDUR + LDR (take 3). · 05f8020c
  Chad Rosier authored Feb 04, 2016
  05f8020c
- [AArch64] Multiply extended 32-bit ints with `[U|S]MADDL' · 33b3bd17
  Silviu Baranga authored Feb 04, 2016
  33b3bd17
- The canonical way to XFAIL a test for all targets is XFAIL: *, not XFAIL: · e4dff62f
  Benjamin Kramer authored Feb 04, 2016
  e4dff62f
- [PPC] Move PPC test to a PPC-specific dir · c455e2f4
  Renato Golin authored Feb 04, 2016
  c455e2f4
- [X86][SSE] Add general 32-bit LOAD + VZEXT_MOVL support to EltsFromConsecutiveLoads · 6788f33c
  Simon Pilgrim authored Feb 04, 2016
  6788f33c
- Revert "[AArch64] Improve load/store optimizer to handle LDUR + LDR." · 18896c0f
  Chad Rosier authored Feb 04, 2016
  18896c0f
- [X86][SSE] Added i686 target tests to make sure we are correctly loading... · 528e94e9
  Simon Pilgrim authored Feb 04, 2016
  528e94e9
- AVX-512: Fixed a bug in FMA instruction selection on KNL · 86528270
  Elena Demikhovsky authored Feb 04, 2016
  86528270
- [Power PC] softening long double type · 23e44f5e
  Petar Jovanovic authored Feb 04, 2016
  23e44f5e
- [AArch64] Improve load/store optimizer to handle LDUR + LDR. · feec2aeb
  Chad Rosier authored Feb 04, 2016
  feec2aeb
- [AVX512] add vfmadd132ss and vfmadd132sd Intrinsic · 7d733604
  Michael Zuckerman authored Feb 04, 2016
  7d733604
- [X86] Add AVX512 vector zext tests · 8159cf11
  Simon Pilgrim authored Feb 04, 2016
  8159cf11
- [NVPTX] Disable performance optimizations when OptLevel==None · f650441b
  Jingyue Wu authored Feb 04, 2016
  f650441b
- Test case for PR 26381 · 155402c9
  Nemanja Ivanovic authored Feb 04, 2016
  155402c9
- [SCEV] Try to reuse existing value during SCEV expansion · a49559be
  Wei Mi authored Feb 04, 2016
  a49559be
Feb 03, 2016
- [SelectionDAG] Fix CombineToPreIndexedLoadStore O(n^2) behavior · f99f0d5a
  Tim Shen authored Feb 03, 2016
  f99f0d5a
- ARM: support TLS for WoA · f36005a3
  Saleem Abdulrasool authored Feb 03, 2016
  f36005a3
- Revert r259662, which caused regressions on polly tests. · 97de3858
  Wei Mi authored Feb 03, 2016
  97de3858
- [ScheduleDAGInstrs::buildSchedGraph()] Handling of memory dependecies rewritten. · ac29f017
  Jonas Paulsson authored Feb 03, 2016
  ac29f017