imgproc:simd Enable VSX and wide universal intrinsics for accumulate operations
authorSayed Adel <seiko@imavr.com>
Tue, 2 Oct 2018 18:44:03 +0000 (20:44 +0200)
committerSayed Adel <seiko@imavr.com>
Thu, 11 Oct 2018 02:37:12 +0000 (04:37 +0200)
commit8965f3ae0640f1ec2e389a2fad19f48fc45e1728
tree4a336493cca56b376d176ae880649e2e2dcb7919
parent5771fd693d559c25975153337aa5e5aa09fdca3d
imgproc:simd Enable VSX and wide universal intrinsics for accumulate operations

  - improve cpu dispatching calls to allow more SIMD extentions
    (SSE4.1, AVX2, VSX)
  - wide universal intrinsics
  - replace dummy v_expand with v_expand_low
  - replace v_expand + v_mul_wrap with v_mul_expand for product accumulate operations
  - use FMA for accumulate operations
  - add mask and more types to accumulate's performance tests
modules/imgproc/CMakeLists.txt
modules/imgproc/perf/perf_accumulate.cpp
modules/imgproc/src/accum.simd.hpp