Merge pull request #1883 from IntelPython/change-descending-from-template-parameter-to-an-argument

Change descending from template parameter to an argument

Author: oleksandr-pavlyk

dpctl/tensor/CMakeLists.txt

@@ -114,9 +114,11 @@ set(_reduction_sources
 set(_sorting_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/sort.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/argsort.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/searchsorted.cpp
+)
+set(_sorting_radix_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/radix_sort.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/radix_argsort.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/searchsorted.cpp
 )
 set(_static_lib_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/simplify_iteration_space.cpp
@@ -153,6 +155,10 @@ set(_tensor_sorting_impl_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/tensor_sorting.cpp
     ${_sorting_sources}
 )
+set(_tensor_sorting_radix_impl_sources
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/tensor_sorting_radix.cpp
+    ${_sorting_radix_sources}
+)
 set(_linalg_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/elementwise_functions/elementwise_functions_type_utils.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/linalg_functions/dot.cpp
@@ -162,10 +168,10 @@ set(_tensor_linalg_impl_sources
     ${_linalg_sources}
 )
 set(_accumulator_sources
-${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/accumulators_common.cpp
-${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_logsumexp.cpp
-${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_prod.cpp
-${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_sum.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/accumulators_common.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_logsumexp.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_prod.cpp
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/accumulators/cumulative_sum.cpp
 )
 set(_tensor_accumulation_impl_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/tensor_accumulation.cpp
@@ -207,6 +213,12 @@ add_sycl_to_target(TARGET ${python_module_name} SOURCES ${_tensor_sorting_impl_s
 target_link_libraries(${python_module_name} PRIVATE ${_static_lib_trgt})
 list(APPEND _py_trgts ${python_module_name})
 
+set(python_module_name _tensor_sorting_radix_impl)
+pybind11_add_module(${python_module_name} MODULE ${_tensor_sorting_radix_impl_sources})
+add_sycl_to_target(TARGET ${python_module_name} SOURCES ${_tensor_sorting_radix_impl_sources})
+target_link_libraries(${python_module_name} PRIVATE ${_static_lib_trgt})
+list(APPEND _py_trgts ${python_module_name})
+
 set(python_module_name _tensor_linalg_impl)
 pybind11_add_module(${python_module_name} MODULE ${_tensor_linalg_impl_sources})
 add_sycl_to_target(TARGET ${python_module_name} SOURCES ${_tensor_linalg_impl_sources})

dpctl/tensor/_sorting.py

@@ -22,13 +22,15 @@
 from ._tensor_sorting_impl import (
     _argsort_ascending,
     _argsort_descending,
+    _sort_ascending,
+    _sort_descending,
+)
+from ._tensor_sorting_radix_impl import (
     _radix_argsort_ascending,
     _radix_argsort_descending,
     _radix_sort_ascending,
     _radix_sort_descending,
     _radix_sort_dtype_supported,
-    _sort_ascending,
-    _sort_descending,
 )
 
 __all__ = ["sort", "argsort"]

dpctl/tensor/libtensor/include/kernels/sorting/merge_sort.hpp

@@ -807,8 +807,7 @@ sycl::event stable_argsort_axis1_contig_impl(
     const IndexComp<IndexTy, argTy, ValueComp> index_comp{arg_tp, ValueComp{}};
 
     static constexpr size_t determine_automatically = 0;
-    size_t sorted_block_size =
-        (sort_nelems >= 512) ? 512 : determine_automatically;
+    size_t sorted_block_size = determine_automatically;
 
     const size_t total_nelems = iter_nelems * sort_nelems;