CUTLASS 3.2.1 (#1113)

* Updates for 3.2.1 release.

* Minor fix in gemm op profiler for raster order.

* Add scheduler mapping for raster order in the kernels.

cmake/NvidiaCutlassConfig.cmake

@@ -2,6 +2,11 @@ get_filename_component(NvidiaCutlass_CMAKE_DIR "${CMAKE_CURRENT_LIST_FILE}" PATH
 
 include(CMakeFindDependencyMacro)
 
-if(NOT TARGET nvidia::cutlass::CUTLASS)
-    include("${NvidiaCutlass_CMAKE_DIR}/NvidiaCutlassTargets.cmake")
+if(TARGET nvidia::cutlass::CUTLASS)
+  return()
 endif()
+
+include("${NvidiaCutlass_CMAKE_DIR}/NvidiaCutlassTargets.cmake")
+
+# For backward compatibility with the old name
+add_library(cutlass_lib ALIAS cutlass_library)