[libcxx-commits] [PATCH] D59925: The optimized parallel versions of sort, stable_sort algorithms.

[Mikhail Dvorskiy via Phabricator via libcxx-commits]

MikeDvorskiy created this revision.
MikeDvorskiy added reviewers: rodgert, ldionne.
Herald added subscribers: libcxx-commits, jdoerfert, mgrang.

Optimization of parallel sort, stable_sort). Result for my benchmarks(4e6 elements, int, double ,44 cores CPU): speed up 3x-10x (depended on type and data – random, sorted or partial sorted).


Repository:
  rPSTL pstl

https://reviews.llvm.org/D59925

Files:
  include/pstl/internal/algorithm_impl.h
  include/pstl/internal/parallel_backend_tbb.h
  include/pstl/internal/parallel_backend_utils.h

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D59925.192596.patch
Type: text/x-patch
Size: 38703 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/libcxx-commits/attachments/20190328/474c065e/attachment-0001.bin>