Adds support for AVX2 for 32-bit types for quicksort and quickselect#60

Merged

r-devulap merged 19 commits intonumpy:mainfrom

sterrettm2:avx2-32bit

Oct 24, 2023

Commits on Oct 20, 2023

Core AVX2 code logic

sterrettm2
authored and
Raghuveer Devulapalli
committed
Bug fix in runtime dispatch and variadic args
Raghuveer Devulapalli
committed
Bugfixes in AVX2 code

sterrettm2
authored and
Raghuveer Devulapalli
committed
Removed some unused code, changed how single compresstore works for AVX2

sterrettm2
authored and
Raghuveer Devulapalli
committed
Some small bug fixes

sterrettm2
authored and
Raghuveer Devulapalli
committed
Removes unnecessary forward includes

sterrettm2
authored and
Raghuveer Devulapalli
committed
Fixed types in a few places

sterrettm2
authored and
Raghuveer Devulapalli
committed
Make pivot selection simpler
Raghuveer Devulapalli
committed
Fixes/changes many small things

sterrettm2
authored and
Raghuveer Devulapalli
committed
Fix rebase merge bugs
Raghuveer Devulapalli
committed
Move avx2_vector declaration to common includes
Raghuveer Devulapalli
committed
Add avx2 to examples
Raghuveer Devulapalli
committed
Changed partition code
sterrettm2
committed

Commits on Oct 23, 2023

Get rid of the avx2_mask_helper
Raghuveer Devulapalli
committed
format files
Raghuveer Devulapalli
committed
Wrapper function for qsort, qselect and partialsort
Raghuveer Devulapalli
committed
Header include directives
Raghuveer Devulapalli
committed
Remove need for avx2-32bit-common.h
Raghuveer Devulapalli
committed
Explicit instantiation of avx512_partial_qsort for _Float16
Raghuveer Devulapalli
committed