disable task-based parallel on windows to avoid /openmp:llvm (#50)

CMakeLists.txt

@@ -97,7 +97,7 @@ endif()
 
 if(MSVC)
   add_compile_definitions(_USE_MATH_DEFINES)
-  add_compile_options(/openmp:llvm)
+  # add_compile_options(/openmp:llvm)
 endif()
 
 ##############

include/small_gicp/util/sort_omp.hpp

@@ -4,6 +4,11 @@
 
 #include <algorithm>
 
+#ifdef _MSC_VER
+#pragma message("warning: Task-based OpenMP parallelism is not well supported on windows.")
+#pragma message("warning: Thus, OpenMP-based downsampling is only partially parallelized on windows.")
+#endif
+
 namespace small_gicp {
 
 template <typename RandomAccessIterator, typename Compare>
@@ -28,11 +33,15 @@ void merge_sort_omp_impl(RandomAccessIterator first, RandomAccessIterator last,
 
 template <typename RandomAccessIterator, typename Compare>
 void merge_sort_omp(RandomAccessIterator first, RandomAccessIterator last, const Compare& comp, int num_threads) {
+#ifndef _MSC_VER
 #pragma omp parallel num_threads(num_threads)
   {
 #pragma omp single nowait
     { merge_sort_omp_impl(first, last, comp); }
   }
+#else
+  std::stable_sort(first, last, comp);
+#endif
 }
 
 template <typename RandomAccessIterator, typename Compare>
@@ -65,11 +74,15 @@ void quick_sort_omp_impl(RandomAccessIterator first, RandomAccessIterator last,
 
 template <typename RandomAccessIterator, typename Compare>
 void quick_sort_omp(RandomAccessIterator first, RandomAccessIterator last, const Compare& comp, int num_threads) {
+#ifndef _MSC_VER
 #pragma omp parallel num_threads(num_threads)
   {
 #pragma omp single nowait
     { quick_sort_omp_impl(first, last, comp); }
   }
+#else
+  std::sort(first, last, comp);
+#endif
 }
 
 }  // namespace small_gicp