Ignore:
Timestamp:
May 6, 2018, 7:28:51 AM (13 months ago)
Author:
cameron
Message:

mvmd_srl, mvmd_sll for AVX2 to improve StreamCompressKernel? performance

File:
1 edited

Legend:

Unmodified
Added
Removed
  • icGREP/icgrep-devel/icgrep/IR_Gen/idisa_avx_builder.h

    r6016 r6017  
    4848    std::pair<llvm::Value *, llvm::Value *> bitblock_indexed_advance(llvm::Value * a, llvm::Value * index_strm, llvm::Value * shiftin, unsigned shift) override;
    4949    llvm::Value * hsimd_signmask(unsigned fw, llvm::Value * a) override;
     50    llvm::Value * mvmd_srl(unsigned fw, llvm::Value * a, llvm::Value * shift) override;
     51    llvm::Value * mvmd_sll(unsigned fw, llvm::Value * a, llvm::Value * shift) override;
    5052    llvm::Value * mvmd_compress(unsigned fw, llvm::Value * a, llvm::Value * select_mask) override;
    5153    llvm::Value * simd_pext(unsigned fw, llvm::Value * v, llvm::Value * extract_mask) override;
Note: See TracChangeset for help on using the changeset viewer.