Commits · 24c3a2395f5a3899bf8c8582388b51a589f9edaa · Roger Ferrer / llvm-epi-0.8 · GitLab

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

Jan 12, 2017
- [AVX-512] Improve lowering of zero_extend of v4i1 to v4i32 and v2i1 to v2i64... · 24c3a239
  Craig Topper authored Jan 12, 2017
  24c3a239
- [AVX-512] Improve lowering of sign_extend of v4i1 to v4i32 and v2i1 to v2i64... · 69ab67b2
  Craig Topper authored Jan 12, 2017
  69ab67b2
- [X86][AVX512] Fix PR31515 - Do not flip vselect condition if it's not a vXi1 mask · c5ba925e
  Elad Cohen authored Jan 12, 2017
  c5ba925e
- [AVX-512] Add more varied avx512 feature command lines to the avx512-cvt.ll... · 56f9610b
  Craig Topper authored Jan 12, 2017
  56f9610b
- Make a test actually test what it set out to test. · b4d9a310
  Chandler Carruth authored Jan 12, 2017
  b4d9a310
- AMDGPU: Fold fneg into fma or fmad · 63f95379
  Matt Arsenault authored Jan 12, 2017
  63f95379
- AMDGPU: Fold fneg into fmul · 4103a81d
  Matt Arsenault authored Jan 12, 2017
  4103a81d
- AMDGPU: Fold fneg into fadd · 2529fba9
  Matt Arsenault authored Jan 12, 2017
  2529fba9
- AMDGPU: Pull fneg/fabs out of a select · 2a04ff97
  Matt Arsenault authored Jan 11, 2017
  2a04ff97
Jan 11, 2017
- AMDGPU: Fix shrinking of addc/subb. · 24a1273a
  Matt Arsenault authored Jan 11, 2017
  24a1273a
- AMDGPU: Fix sext_inreg for i1 in i16 · 682eb439
  Matt Arsenault authored Jan 11, 2017
  682eb439
- AMDGPU: Fix breaking VOP3 v_add_i32s · 28bd4cbe
  Matt Arsenault authored Jan 11, 2017
  28bd4cbe
- [asan] Set alignment of __asan_global_* globals to sizeof(GlobalStruct) · 503162b4
  Kuba Mracek authored Jan 11, 2017
  503162b4
- AMDGPU: Fix folding immediates into mac src2 · 69e3001b
  Matt Arsenault authored Jan 11, 2017
  69e3001b
- Add test that verifies we don't peel loops in optsize functions. NFC. · 991c2e0e
  Michael Kuperstein authored Jan 11, 2017
  991c2e0e
- LowerTypeTests: Represent the memory region size with the constant size-1. · 7636532c
  Peter Collingbourne authored Jan 11, 2017
  7636532c
- [SCEV] Make howFarToZero max backedge-taken count check for precondition. · bd6dedaa
  Eli Friedman authored Jan 11, 2017
  bd6dedaa
- [SCEV] Make howFarToZero use a simpler formula for max backedge-taken count. · 83962656
  Eli Friedman authored Jan 11, 2017
  83962656
- Re-apply r291205, "LowerTypeTests: Split the pass in two: a resolution phase... · 6bca5a0d
  Peter Collingbourne authored Jan 11, 2017
  6bca5a0d
- NewGVN: Fix PR31594, by tracking the store count of congruence · f6eba4be
  Daniel Berlin authored Jan 11, 2017
  f6eba4be
- Resubmit "[PGO] Turn off comdat renaming in IR PGO by default" · 20f5df1d
  Rong Xu authored Jan 11, 2017
  20f5df1d
- Revert "CodeGen: Allow small copyable blocks to "break" the CFG." · efe56fed
  Kyle Butt authored Jan 11, 2017
  efe56fed
- [ARM] More aggressive matching for vpadd and vpaddl. · 3a03742c
  Eli Friedman authored Jan 11, 2017
  3a03742c
- [SLP] Remove bogus assert. · f69e6466
  Michael Kuperstein authored Jan 11, 2017
  f69e6466
- [X86][XOP] Add vpermil2ps target shuffle -> insertps combine test · fd627bf8
  Simon Pilgrim authored Jan 11, 2017
  fd627bf8
- Revert rL291205 because it breaks Chrome tests under CFI. · 42e6b4fd
  Ivan Krasin authored Jan 11, 2017
  42e6b4fd
- [ARM] Fix test CodeGen/ARM/fpcmp_ueq.ll broken by rL290616 · 56f5b893
  Evgeny Astigeevich authored Jan 11, 2017
  56f5b893
- X86 CodeGen: Optimized pattern for truncate with unsigned saturation. · 9d0e7c33
  Elena Demikhovsky authored Jan 11, 2017
  9d0e7c33
- [AMDGPU] Assembler: SDWA/DPP should not accept scalar registers and immediate operands · 9772eb39
  Sam Kolton authored Jan 11, 2017
  9772eb39
- [X86][AVX512BW] Vectorize v64i8 vector shifts · 5a81fefa
  Simon Pilgrim authored Jan 11, 2017
  5a81fefa
- Fix line endings · c22c889f
  Simon Pilgrim authored Jan 11, 2017
  c22c889f
- [X86] Fix PR30926 - Add patterns for (v)cvtsi2s{s,d} and (v)cvtsd2s{s,d} · 0c260107
  Elad Cohen authored Jan 11, 2017
  0c260107
- [X86] fixing failed test in commit: r291657 · 81d0f170
  Mohammed Agabaria authored Jan 11, 2017
  81d0f170
- [X86] updating TTI costs for arithmetic instructions on X86\SLM arch. · 2c96c433
  Mohammed Agabaria authored Jan 11, 2017
  2c96c433
- [XRay] Define the library for XRay trace logs · d6c18657
  Dean Michael Berris authored Jan 11, 2017
  d6c18657
- [PM] Rewrite the loop pass manager to use a worklist and augmented run · 410eaeb0
  Chandler Carruth authored Jan 11, 2017
  410eaeb0
- Revert r291645 "[DAGCombiner] Teach DAG combiner to fold (vselect (N0 xor... · 7af39837
  Craig Topper authored Jan 11, 2017
  7af39837
- [LICM] Report failing to hoist conditionally-executed loads · e2aaf3a3
  Adam Nemet authored Jan 11, 2017
  e2aaf3a3
- [LICM] Report failing to hoist a load with an invariant address · 81941b31
  Adam Nemet authored Jan 11, 2017
  81941b31
- [LICM] Report successful hoist/sink/promotion · 358433ce
  Adam Nemet authored Jan 11, 2017
  358433ce