[SYCL] Delay inclusion of sycl/types.hpp in core.hpp (#13902)

Can't eliminate it completely until
#13760 is merged, but this PR brings
us closer to that goal.

Author: aelovikov-intel

sycl/include/sycl/ext/intel/experimental/task_sequence.hpp

@@ -10,6 +10,7 @@
 
 #include <CL/__spirv/spirv_ops.hpp>
 #include <CL/__spirv/spirv_types.hpp>
+#include <sycl/aspects.hpp>
 #include <sycl/exception.hpp>
 #include <sycl/ext/intel/experimental/fpga_kernel_properties.hpp>
 #include <sycl/ext/intel/experimental/task_sequence_properties.hpp>

sycl/include/sycl/ext/oneapi/experimental/prefetch.hpp

@@ -10,6 +10,7 @@
 
 #include <CL/__spirv/spirv_ops.hpp>
 #include <sycl/ext/oneapi/properties/properties.hpp>
+#include <sycl/types.hpp>
 
 namespace sycl {
 inline namespace _V1 {

sycl/include/sycl/ext/oneapi/properties/properties.hpp

@@ -8,10 +8,10 @@
 
 #pragma once
 
+#include <sycl/detail/is_device_copyable.hpp>
 #include <sycl/ext/oneapi/properties/property.hpp>       // for IsRuntimePr...
 #include <sycl/ext/oneapi/properties/property_utils.hpp> // for Sorted, Mer...
 #include <sycl/ext/oneapi/properties/property_value.hpp> // for property_value
-#include <sycl/types.hpp>                                // for is_device_c...
 
 #include <tuple>       // for tuple, tupl...
 #include <type_traits> // for enable_if_t

sycl/include/sycl/sub_group.hpp

@@ -18,7 +18,6 @@
 #include <sycl/memory_enums.hpp>               // for memory_scope
 #include <sycl/multi_ptr.hpp>                  // for multi_ptr
 #include <sycl/range.hpp>                      // for range
-#include <sycl/types.hpp>                      // for vec
 
 #ifdef __SYCL_DEVICE_ONLY__
 #include <sycl/ext/oneapi/functional.hpp>
@@ -33,7 +32,7 @@ inline namespace _V1 {
 template <typename T, access::address_space Space,
           access::decorated DecorateAddress>
 class multi_ptr;
-
+template <typename Type, int NumElements> class vec;
 namespace detail {
 
 namespace sub_group {

sycl/test-e2e/Basic/boolean.cpp

@@ -3,9 +3,9 @@
 #include <cassert>
 #include <iostream>
 
-#include <sycl/detail/core.hpp>
-
 #include <sycl/detail/boolean.hpp>
+#include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 namespace s = sycl;

sycl/test-e2e/Basic/buffer/buffer.cpp

@@ -13,6 +13,7 @@
 #include <iostream>
 #include <memory>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/Basic/group_async_copy.cpp

@@ -6,6 +6,7 @@
 
 #include <iostream>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 #include <typeinfo>
 
 using namespace sycl;

sycl/test-e2e/Basic/group_async_copy_legacy.cpp

@@ -6,6 +6,7 @@
 
 #include <iostream>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 #include <typeinfo>
 
 using namespace sycl;

sycl/test-e2e/Basic/half_type.cpp

@@ -13,6 +13,7 @@
 #include <cmath>
 #include <iostream>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 #include <unordered_set>
 
 using namespace sycl;

sycl/test-e2e/Basic/parallel_for_indexers.cpp

@@ -2,6 +2,7 @@
 // RUN: %{run} %t2.out
 
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 #include <cassert>
 #include <memory>

sycl/test-e2e/Basic/parallel_for_range_roundup.cpp

@@ -24,6 +24,7 @@
 //
 #include <iostream>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/Basic/swizzle_op.cpp

@@ -11,6 +11,7 @@
 
 #include <cassert>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/Basic/vector/bool.cpp

@@ -13,6 +13,7 @@
 //===----------------------------------------------------------------------===//
 
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 constexpr int size = 2;
 

sycl/test-e2e/Basic/vector/byte.cpp

@@ -13,6 +13,7 @@
 //===----------------------------------------------------------------------===//
 
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 #include <cstddef> // std::byte
 #include <tuple>   // std::ignore

sycl/test-e2e/Basic/vector/int-convert.cpp

@@ -11,6 +11,7 @@
 // RUN: %if preview-breaking-changes-supported %{ %{run} %t2.out %}
 
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 #include <cstdint>
 #include <iostream>

sycl/test-e2e/Basic/vector/operators.cpp

@@ -11,6 +11,7 @@
 
 #define SYCL_SIMPLE_SWIZZLES
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 namespace s = sycl;
 
 template <typename ResultVecT>

sycl/test-e2e/Basic/vector/vec_binary_scalar_order.hpp

@@ -3,6 +3,7 @@
 #pragma once
 
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 template <typename T>
 using rel_t = std::conditional_t<

sycl/test-e2e/KernelFusion/internalize_non_unit_localsize.cpp

@@ -15,6 +15,7 @@
 #include <sycl/detail/core.hpp>
 #include <sycl/ext/codeplay/experimental/fusion_wrapper.hpp>
 #include <sycl/properties/all_properties.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/KernelFusion/internalize_vfunc.cpp

@@ -9,6 +9,7 @@
 #include <sycl/detail/core.hpp>
 #include <sycl/ext/codeplay/experimental/fusion_wrapper.hpp>
 #include <sycl/properties/all_properties.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/Regression/get_spec_const_vec16.cpp

@@ -5,8 +5,8 @@
 // of sycl::vec<char, 16> type is correct.
 
 #include <sycl/detail/core.hpp>
-
 #include <sycl/specialization_id.hpp>
+#include <sycl/types.hpp>
 
 #include <algorithm>
 

sycl/test-e2e/Regression/group.cpp

@@ -11,6 +11,7 @@
 
 #include <iostream>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 using namespace sycl;
 

sycl/test-e2e/Regression/vec_rel_swizzle_ops.cpp

@@ -6,6 +6,7 @@
 
 #include <cstdlib>
 #include <sycl/detail/core.hpp>
+#include <sycl/types.hpp>
 
 template <typename T, typename ResultT>
 bool testAndOperator(const std::string &typeName) {

sycl/test/include_deps/sycl_detail_core.hpp.cpp

@@ -133,12 +133,6 @@
 // CHECK-NEXT: nd_item.hpp
 // CHECK-NEXT: nd_range.hpp
 // CHECK-NEXT: sub_group.hpp
-// CHECK-NEXT: types.hpp
-// CHECK-NEXT: marray.hpp
-// CHECK-NEXT: vector.hpp
-// CHECK-NEXT: detail/memcpy.hpp
-// CHECK-NEXT: detail/vector_convert.hpp
-// CHECK-NEXT: swizzles.def
 // CHECK-NEXT: ext/oneapi/functional.hpp
 // CHECK-NEXT: detail/spirv.hpp
 // CHECK-NEXT: ext/oneapi/experimental/non_uniform_groups.hpp
@@ -148,6 +142,12 @@
 // CHECK-NEXT: builtins_utils_vec.hpp
 // CHECK-NEXT: builtins_utils_scalar.hpp
 // CHECK-NEXT: detail/boolean.hpp
+// CHECK-NEXT: marray.hpp
+// CHECK-NEXT: types.hpp
+// CHECK-NEXT: vector.hpp
+// CHECK-NEXT: detail/memcpy.hpp
+// CHECK-NEXT: detail/vector_convert.hpp
+// CHECK-NEXT: swizzles.def
 // CHECK-NEXT: detail/builtins/common_functions.inc
 // CHECK-NEXT: detail/builtins/helper_macros.hpp
 // CHECK-NEXT: detail/builtins/geometric_functions.inc