CUTLASS 3.6.0 (#1850)

* v3.6

* update changelog

* update readme

* fix typo

* fixing typos

* hopper gemm with weight prefetch

---------

Co-authored-by: yuzhai <yuzhai@nvidia.com>
Co-authored-by: Haicheng Wu <haichengw@nvidia.com>

include/cute/swizzle.hpp

@@ -30,13 +30,11 @@
  **************************************************************************************************/
 #pragma once
 
-#include <cute/config.hpp>
-
-#include <cute/container/tuple.hpp>
-#include <cute/algorithm/tuple_algorithms.hpp>
-#include <cute/numeric/integer_sequence.hpp>
-#include <cute/numeric/integral_constant.hpp>
-#include <cute/numeric/math.hpp>
+#include <cute/config.hpp>                      // CUTE_HOST_DEVICE
+#include <cute/container/tuple.hpp>             // cute::is_tuple
+#include <cute/numeric/integral_constant.hpp>   // cute::constant
+#include <cute/numeric/math.hpp>                // cute::max, cute::min
+#include <cute/algorithm/tuple_algorithms.hpp>  // cute::transform_apply
 
 namespace cute
 {
@@ -488,4 +486,13 @@ CUTE_HOST std::ostream& operator<<(std::ostream& os, MixedBits<S,F> const& m)
 }
 #endif // !defined(__CUDACC_RTC__)
 
+//
+// Helper Function
+//
+template <class T, class = void>                      // Default No-Swizzle
+struct get_swizzle { using type = Swizzle<0,4,3>; };
+
+template <class T>
+using get_swizzle_t = typename get_swizzle<T>::type;
+
 } // end namespace cute