Merge _sorting_impl and _sorting_radix_impl modules

Doing so reduces the binary size. Previously, '_tensor_sorting_impl'
module has size 22'448'920 bytes, and '_tensor_sorting_radix_impl'
has size 31'927'256 bytes.

Total size was 54'376'176 bytes.

After this change, the total size of the new '_tensor_sorting_impl'
is 49'790'872, which is about 4Mb of savings.

Author: oleksandr-pavlyk

dpctl/tensor/CMakeLists.txt

@@ -114,12 +114,10 @@ set(_reduction_sources
 set(_sorting_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/merge_sort.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/merge_argsort.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/searchsorted.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/topk.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
+    ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/sorting/topk.cpp
 )
 set(_static_lib_sources
     ${CMAKE_CURRENT_SOURCE_DIR}/libtensor/source/simplify_iteration_space.cpp
@@ -156,10 +154,6 @@ 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
@@ -214,12 +208,6 @@ 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

@@ -25,16 +25,14 @@
 from ._tensor_sorting_impl import (
     _argsort_ascending,
     _argsort_descending,
-    _sort_ascending,
-    _sort_descending,
-    _topk,
-)
-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,
+    _topk,
 )
 
 __all__ = ["sort", "argsort"]

dpctl/tensor/libtensor/source/sorting/topk.cpp

@@ -31,18 +31,19 @@
 #include <utility>
 #include <vector>
 
+#include <sycl/sycl.hpp>
+
 #include "dpctl4pybind11.hpp"
 #include <pybind11/pybind11.h>
 #include <pybind11/stl.h>
-#include <sycl/sycl.hpp>
 
+#include "kernels/sorting/topk.hpp"
 #include "utils/math_utils.hpp"
 #include "utils/memory_overlap.hpp"
 #include "utils/output_validation.hpp"
 #include "utils/type_dispatch.hpp"
 #include "utils/type_utils.hpp"
 
-#include "kernels/sorting/topk.hpp"
 #include "rich_comparisons.hpp"
 #include "topk.hpp"
 

dpctl/tensor/libtensor/source/tensor_sorting.cpp

@@ -27,6 +27,8 @@
 
 #include "sorting/merge_argsort.hpp"
 #include "sorting/merge_sort.hpp"
+#include "sorting/radix_argsort.hpp"
+#include "sorting/radix_sort.hpp"
 #include "sorting/searchsorted.hpp"
 #include "sorting/topk.hpp"
 
@@ -37,5 +39,7 @@ PYBIND11_MODULE(_tensor_sorting_impl, m)
     dpctl::tensor::py_internal::init_merge_sort_functions(m);
     dpctl::tensor::py_internal::init_merge_argsort_functions(m);
     dpctl::tensor::py_internal::init_searchsorted_functions(m);
+    dpctl::tensor::py_internal::init_radix_sort_functions(m);
+    dpctl::tensor::py_internal::init_radix_argsort_functions(m);
     dpctl::tensor::py_internal::init_topk_functions(m);
 }