Outdated test for CWG2917 "Disallow multiple friend-type-specifiers for a friend template"

[Endilll]

This Core issue saw substantial updates since the test was written, including the part that resolution should not be applied retroactively ("This is not a DR"). We need to revisit the test and make sure our test corresponds with the latest resolution and incorporates additional use cases that Core considered (if there are any).

@Sirraide both Core issue and the test are yours. Mind picking this up?

[llvmbot]

@llvm/issue-subscribers-c-1

Author: Vlad Serebrennikov (Endilll)

[This](https://cplusplus.github.io/CWG/issues/2917.html) Core issue saw substantial updates since the test was written, including the part that resolution should not be applied retroactively ("This is not a DR"). We need to revisit the test and make sure our test corresponds with the latest resolution and incorporates additional use cases that Core considered (if there are any).

@Sirraide both Core issue and the test are yours. Mind picking this up?

[llvmbot]

@llvm/issue-subscribers-clang-frontend

Author: Vlad Serebrennikov (Endilll)

[This](https://cplusplus.github.io/CWG/issues/2917.html) Core issue saw substantial updates since the test was written, including the part that resolution should not be applied retroactively ("This is not a DR"). We need to revisit the test and make sure our test corresponds with the latest resolution and incorporates additional use cases that Core considered (if there are any).

@Sirraide both Core issue and the test are yours. Mind picking this up?

[Sirraide]

This Core issue saw substantial updates since the test was written, including the part that resolution should not be applied retroactively ("This is not a DR"). We need to revisit the test and make sure our test corresponds with the latest resolution and incorporates additional use cases that Core considered (if there are any).

@Sirraide both Core issue and the test are yours. Mind picking this up?

I don’t think anything actually changed? The wording did, but the semantics are the same as the first resolution I believe (and the examples are identical too).

[Sirraide]

the part that resolution should not be applied retroactively ("This is not a DR")

Also, from what I can tell, I don’t think it’s possible for the wording added by this issue to apply to anything before C++26.