Merge pull request #169 from r-devulap/v6.0rc

Update README

README.md

@@ -182,6 +182,7 @@ gets more complicated. Sorting by Euclidean distance can be up-to 10x faster.
 ## Downstream projects using x86-simd-sort
 
 - NumPy uses this as a [submodule](https://github.com/numpy/numpy/pull/22315) to accelerate `np.sort, np.argsort, np.partition and np.argpartition`.
+- PyTorch uses this as a [submodule](https://github.com/pytorch/pytorch/pull/127936) to accelerate `torch.sort, torch.argsort`.
 - A slightly modifed version this library has been integrated into [openJDK](https://github.com/openjdk/jdk/pull/14227).
 - [GRAPE](https://github.com/alibaba/libgrape-lite.git): C++ library for parallel graph processing.
 - AVX-512 version of the key-value sort has been submitted to [Oceanbase](https://github.com/oceanbase/oceanbase/pull/1325).

meson.build

@@ -1,5 +1,5 @@
 project('x86-simd-sort', 'cpp',
-        version : '5.0.x',
+        version : '6.0.x',
         license : 'BSD 3-clause',
         default_options : ['cpp_std=c++17'])
 fs = import('fs')

src/avx2-64bit-qsort.hpp

@@ -395,8 +395,7 @@ struct avx2_vector<uint64_t> {
 static_assert(sizeof(size_t) == sizeof(uint64_t),
               "Size of size_t and uint64_t are not the same");
 template <>
-struct avx2_vector<size_t> : public avx2_vector<uint64_t> {
-};
+struct avx2_vector<size_t> : public avx2_vector<uint64_t> {};
 #endif
 
 template <>

src/avx512-64bit-common.h

@@ -968,8 +968,7 @@ struct zmm_vector<uint64_t> {
 static_assert(sizeof(size_t) == sizeof(uint64_t),
               "Size of size_t and uint64_t are not the same");
 template <>
-struct zmm_vector<size_t> : public zmm_vector<uint64_t> {
-};
+struct zmm_vector<size_t> : public zmm_vector<uint64_t> {};
 #endif
 
 template <>