[libcxx-dev] Parallel stable sort

[Christopher Nelson via libcxx-dev]

Hello friends,

I am now working on the parallel stable sort backend. I've looked at the
sort implemented by the TBB backend and it seems to depend on TBB in
complex ways. I'm not really sure that I could translate that into
reasonable OpenMP.

However, I have found a parallel stable three-way quicksort
<https://software.intel.com/content/www/us/en/develop/articles/an-efficient-parallel-three-way-quicksort-using-intel-c-compiler-and-openmp-45-library.html>
algorithm that has sample code
<https://software.intel.com/content/www/us/en/develop/articles/how-to-implement-a-parallel-stable-three-way-quicksort-using-intel-c-compiler-and-openmp-45.html>
from
Intel. This algorithm looks reasonably straightforward, and it appears to
have good performance.

If there are no objections I would like to go forward with this
implementation. Or, if there is another algorithm that is preferable,
please let me know.

Thank you!
-={C}=-
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/libcxx-dev/attachments/20201015/04592434/attachment.html>