Commits · 949f9b9f641720804832e460b5e04df47ec6b8a5 · Lorenzo Albano / LLVM bpEVL · GitLab

Nov 10, 2016
- Fix type ambiguity with std::max · 805d43a0
  Zachary Turner authored Nov 10, 2016
  805d43a0
- [Support] Improve flexibility of binary blob formatter. · 4a86af07
  Zachary Turner authored Nov 10, 2016
  4a86af07
- [Target] Rename X86/ARM Assembly printer to reflect reality. · a22ddddf
  Davide Italiano authored Nov 10, 2016
  a22ddddf
- Revert r286437 r286438, they caused PR30976 · 85740f6b
  Nico Weber authored Nov 10, 2016
  85740f6b
- [OptDiag] Remove non-printable chars from function name · 7da20c39
  Adam Nemet authored Nov 10, 2016
  7da20c39
- [SelectionDAG] Add support for vector demandedelts in TRUNCATE opcodes · d67af68f
  Simon Pilgrim authored Nov 10, 2016
  d67af68f
- Add comments about why we put LoopSink pass at the very late stage. · 5492f864
  Dehao Chen authored Nov 10, 2016
  5492f864
- Restore part of "[ThinLTO] Prevent exporting of locals used/defined in module level asm" · a081145e
  Teresa Johnson authored Nov 10, 2016
  a081145e
- Use common SDLoc. NFCI. · 33fef8e8
  Simon Pilgrim authored Nov 10, 2016
  33fef8e8
- [SelectionDAG] Add support for vector demandedelts in MUL opcodes · ee187fd6
  Simon Pilgrim authored Nov 10, 2016
  ee187fd6
- AMDGPU: Add VI i16 support · 115a6156
  Tom Stellard authored Nov 10, 2016
  115a6156
- [SelectionDAG] Add support for vector demandedelts in SRA opcodes · ca57e53d
  Simon Pilgrim authored Nov 10, 2016
  ca57e53d
- [DAGCombiner] Correctly extract the ConstOrConstSplat shift value for SHL nodes · 37c9034b
  Simon Pilgrim authored Nov 10, 2016
  37c9034b
- [SelectionDAG] Add support for vector demandedelts in SHL/SRL opcodes · 3bf99c05
  Simon Pilgrim authored Nov 10, 2016
  3bf99c05
- [ARM] Thumb2 LDR (literal) should accept PC as the destination · 18ca2adf
  Oliver Stannard authored Nov 10, 2016
  18ca2adf
- [SCEVExpander] Hoist unsigned divisons when safe · 3d75b62f
  Sanjoy Das authored Nov 10, 2016
  3d75b62f
- [SCEVExpander] Don't hoist divisions · e30a2814
  Sanjoy Das authored Nov 10, 2016
  e30a2814
- [AVX-512] Allow legacy cvtpd2dq intrinsics to select EVEX encoded instruction when available. · bd298c37
  Craig Topper authored Nov 10, 2016
  bd298c37
- [AVX-512][X86] Convert avx_cvtt_ps2dq_256 and sse2_cvttps2dq intrinsics to... · e0845d8e
  Craig Topper authored Nov 10, 2016
  e0845d8e
- [X86] Convert int_x86_avx_cvtt_pd2dq_256 to fp_to_sint using the intrinsics... · f37b9b9b
  Craig Topper authored Nov 10, 2016
  f37b9b9b
- [X86] Move some custom patterns into the currently empty pattern of their... · 2afed2c7
  Craig Topper authored Nov 10, 2016
  2afed2c7
- [X86] Remove some patterns still referencing int_x86_sse2_cvttpd2dq that... · 1d2e74f0
  Craig Topper authored Nov 10, 2016
  1d2e74f0
- [SCEV] Eta reduce some lambdas; NFC · 0ae390ab
  Sanjoy Das authored Nov 10, 2016
  0ae390ab
- [InstCombine] avoid infinite loop from shuffle-extract-insert sequence (PR30923) · 4e1b5a53
  Sanjay Patel authored Nov 10, 2016
  4e1b5a53
- Re-apply r286384, "X86: Introduce the "relocImm" ComplexPattern, which... · 32ab3a81
  Peter Collingbourne authored Nov 09, 2016
  32ab3a81
- Add isHotBB helper function to ProfileSummaryInfo · 38a666d6
  Dehao Chen authored Nov 09, 2016
  38a666d6
- Preserve assumption cache in loop-rotate. · ddbf83ea
  Eli Friedman authored Nov 09, 2016
  ddbf83ea
Nov 09, 2016
- GlobalISel: fix typo. NFC · 09dd2496
  Tim Northover authored Nov 09, 2016
  09dd2496
- GlobalISel: translate invoke and landingpad instructions · a9105be4
  Tim Northover authored Nov 09, 2016
  a9105be4
- Make BitcodeReader::parseIdentificationBlock() robust to EOF · 67d1a412
  Mehdi Amini authored Nov 09, 2016
  67d1a412
- Minor unroll pass refacoring. · c2698cd9
  Evgeny Stupachenko authored Nov 09, 2016
  c2698cd9
- [Verifier] clang-format a section; NFC · 26f28a28
  Sanjoy Das authored Nov 09, 2016
  26f28a28
- [SCEV] Refactor out a useful pattern; NFC · 6b46a0d1
  Sanjoy Das authored Nov 09, 2016
  6b46a0d1
- Revert r286384, "X86: Introduce the "relocImm" ComplexPattern, which... · a9cadedd
  Peter Collingbourne authored Nov 09, 2016
  a9cadedd
- X86: Introduce the "relocImm" ComplexPattern, which represents a relocatable immediate. · 4c15db45
  Peter Collingbourne authored Nov 09, 2016
  4c15db45
- [Hexagon] Silence "sometimes uninitialized" warning in HexagonCopyToCombine · f817efbb
  Krzysztof Parzyszek authored Nov 09, 2016
  f817efbb
- Bitcode: Change the materializer interface to return llvm::Error. · 7f00d0a1
  Peter Collingbourne authored Nov 09, 2016
  7f00d0a1
- [Hexagon] Separate Hexagon subreg indices for different register classes · a540997c
  Krzysztof Parzyszek authored Nov 09, 2016
  a540997c
- [Hexagon] Eliminate Insert4 pseudo-instruction, use combines instead · 601d7eb1
  Krzysztof Parzyszek authored Nov 09, 2016
  601d7eb1
- [SystemZ] A few fixes in scheduler files. · e127fe70
  Jonas Paulsson authored Nov 09, 2016
  e127fe70