This PR was initially accepts-invalid, but I think it's actually valid
C++20 code. My reasoning is that in C++20 we no longer require the
declaration of operator== (#if-defed in the test), because C++20's
[temp.names]/2 says "A name is also considered to refer to a template
if it is an unqualified-id followed by a < and name lookup either finds
one or more functions or finds nothing." so when we're parsing
constexpr friend bool operator==<T>(T lhs, const Foo& rhs);
we treat "operator==" as a template name, because name lookup of
"operator==" found nothing and we have an operator-function-id, which is
an unqualified-id, and it's followed by a <. So the declaration isn't
needed to treat "operator==<T>" as a template-id.
Tested x86_64-pc-linux-gnu, ok for trunk?
PR c++/93807
* g++.dg/cpp2a/fn-template20.C: New test.
---
gcc/testsuite/g++.dg/cpp2a/fn-template20.C | 34 ++++++++++++++++++++++
1 file changed, 34 insertions(+)
create mode 100644 gcc/testsuite/g++.dg/cpp2a/fn-template20.C
diff --git a/gcc/testsuite/g++.dg/cpp2a/fn-template20.C
b/gcc/testsuite/g++.dg/cpp2a/fn-template20.C
new file mode 100644
index 00000000000..c558ad1f2b6
--- /dev/null
+++ b/gcc/testsuite/g++.dg/cpp2a/fn-template20.C
@@ -0,0 +1,34 @@
+// PR c++/93807
+// { dg-do compile { target c++11 } }
+
+// In C++17, we need the following declaration to treat operator== as
+// a template name. In C++20, this is handled by [temp.names]/2.
+#if __cplusplus <= 201703L
+template <typename T>
+class Foo;
+template <typename T>
+constexpr bool operator==(T lhs, const Foo<T>& rhs);
+#endif
+
+template <typename T>
+class Foo {
+public:
+ constexpr Foo(T k) : mK(k) {}
+
+ constexpr friend bool operator==<T>(T lhs, const Foo& rhs);
+private:
+ T mK;
+};
+
+template <typename T>
+constexpr bool
+operator==(T lhs, const Foo<T>& rhs)
+{
+ return lhs == rhs.mK;
+}
+
+int
+main ()
+{
+ return 1 == Foo<int>(1) ? 0 : 1;
+}
base-commit: e76100ced607218a3bf26344fd57d7384a7c18b5
--
Marek Polacek • Red Hat, Inc. • 300 A St, Boston, MA
