[libc++][NFC] Simplify the implementation of __promote

[philnik777]

This depends on enabling the use of extensions.

[llvmbot]

@llvm/pr-subscribers-libcxx

Author: Nikolas Klauser (philnik777)

Changes

This depends on enabling the use of extensions.

---

Full diff: https://github.com/llvm/llvm-project/pull/81379.diff

1 Files Affected:

• (modified) libcxx/include/__type_traits/promote.h (+6-53)

diff --git a/libcxx/include/__type_traits/promote.h b/libcxx/include/__type_traits/promote.h
index e22b4a422c2c80..65eb7b37403edf 100644
--- a/libcxx/include/__type_traits/promote.h
+++ b/libcxx/include/__type_traits/promote.h
@@ -11,8 +11,7 @@
 
 #include <__config>
 #include <__type_traits/integral_constant.h>
-#include <__type_traits/is_same.h>
-#include <__utility/declval.h>
+#include <__type_traits/is_arithmetic.h>
 
 #if !defined(_LIBCPP_HAS_NO_PRAGMA_SYSTEM_HEADER)
 #  pragma GCC system_header
@@ -20,9 +19,10 @@
 
 _LIBCPP_BEGIN_NAMESPACE_STD
 
-template <class _Tp>
-struct __numeric_type {
-  static void __test(...);
+template <class... _Args>
+class __promote {
+  static_assert((is_arithmetic<_Args>::value && ...));
+
   static float __test(float);
   static double __test(char);
   static double __test(int);
@@ -38,57 +38,10 @@ struct __numeric_type {
   static double __test(double);
   static long double __test(long double);
 
-  typedef decltype(__test(std::declval<_Tp>())) type;
-  static const bool value = _IsNotSame<type, void>::value;
-};
-
-template <>
-struct __numeric_type<void> {
-  static const bool value = true;
-};
-
-template <class _A1,
-          class _A2 = void,
-          class _A3 = void,
-          bool      = __numeric_type<_A1>::value && __numeric_type<_A2>::value && __numeric_type<_A3>::value>
-class __promote_imp {
-public:
-  static const bool value = false;
-};
-
-template <class _A1, class _A2, class _A3>
-class __promote_imp<_A1, _A2, _A3, true> {
-private:
-  typedef typename __promote_imp<_A1>::type __type1;
-  typedef typename __promote_imp<_A2>::type __type2;
-  typedef typename __promote_imp<_A3>::type __type3;
-
 public:
-  typedef decltype(__type1() + __type2() + __type3()) type;
-  static const bool value = true;
+  using type = decltype((__test(_Args()) + ...));
 };
 
-template <class _A1, class _A2>
-class __promote_imp<_A1, _A2, void, true> {
-private:
-  typedef typename __promote_imp<_A1>::type __type1;
-  typedef typename __promote_imp<_A2>::type __type2;
-
-public:
-  typedef decltype(__type1() + __type2()) type;
-  static const bool value = true;
-};
-
-template <class _A1>
-class __promote_imp<_A1, void, void, true> {
-public:
-  typedef typename __numeric_type<_A1>::type type;
-  static const bool value = true;
-};
-
-template <class _A1, class _A2 = void, class _A3 = void>
-class __promote : public __promote_imp<_A1, _A2, _A3> {};
-
 _LIBCPP_END_NAMESPACE_STD
 
 #endif // _LIBCPP___TYPE_TRAITS_PROMOTE_H

[ldionne]

I think we should #if the old code as a whole for Clang 17 just to get this patch unblocked, and we can remove that "workaround" when we drop support for Clang 17. It's so easy to clean these up that I wouldn't stall this patch just for that reason.

[EricWF]

I think we should #if the old code as a whole for Clang 17 just to get this patch unblocked, and we can remove that "workaround" when we drop support for Clang 17. It's so easy to clean these up that I wouldn't stall this patch just for that reason.

This is supposed to be a simplification, meaning it only provides value if it simplifies the code. We've just added a second implementation, which is not a simplification, it's a complication.

There's no need to land code like this while it doesn't provided the stated value.

This should have waited until all supported compilers allowed us to do it.