Clang incorrectly considers a class with an anonymous union member to not be const-default-constructible even if a union member has a default member initializer

[zygoloid]

Testcase:

struct A {
  union {
    int n = 0;
    int m;
  };
};
const A a;

This is valid, but Clang rejects:

<source>:7:9: error: default initialization of an object of const type 'const A' without a user-provided default constructor

(Clang accepts if m is removed.) See https://eel.is/c++draft/dcl.init#general-8.3 -- because n has an initializer, the entire anonymous union is treated as initialized.

[llvmbot]

@llvm/issue-subscribers-clang-frontend

Author: Richard Smith (zygoloid)

Testcase: ```c++ struct A { union { int n = 0; int m; }; }; const A a; ``` This is valid, but Clang rejects: ```console <source>:7:9: error: default initialization of an object of const type 'const A' without a user-provided default constructor ``` (Clang accepts if `m` is removed.) See https://eel.is/c++draft/dcl.init#general-8.3 -- because `n` has an initializer, the entire anonymous union is treated as initialized.

[llvmbot]

@llvm/issue-subscribers-c-11

Author: Richard Smith (zygoloid)

Testcase: ```c++ struct A { union { int n = 0; int m; }; }; const A a; ``` This is valid, but Clang rejects: ```console <source>:7:9: error: default initialization of an object of const type 'const A' without a user-provided default constructor ``` (Clang accepts if `m` is removed.) See https://eel.is/c++draft/dcl.init#general-8.3 -- because `n` has an initializer, the entire anonymous union is treated as initialized.

[efriedma-quic]

I think CXXRecordDecl::allowConstDefaultInit() needs to special-case unions to handle the rule (specifically, it needs to check for HasInClassInitializer).

[llvmbot]

Hi!

This issue may be a good introductory issue for people new to working on LLVM. If you would like to work on this issue, your first steps are:

1. Check that no other contributor has already been assigned to this issue. If you believe that no one is actually working on it despite an assignment, ping the person. After one week without a response, the assignee may be changed.
2. In the comments of this issue, request for it to be assigned to you, or just create a pull request after following the steps below. Mention this issue in the description of the pull request.
3. Fix the issue locally.
4. Run the test suite locally. Remember that the subdirectories under test/ create fine-grained testing targets, so you can e.g. use make check-clang-ast to only run Clang's AST tests.
5. Create a Git commit.
6. Run git clang-format HEAD~1 to format your changes.
7. Open a pull request to the upstream repository on GitHub. Detailed instructions can be found in GitHub's documentation. Mention this issue in the description of the pull request.

If you have any further questions about this issue, don't hesitate to ask via a comment in the thread below.

[llvmbot]

@llvm/issue-subscribers-good-first-issue

Author: Richard Smith (zygoloid)

Testcase: ```c++ struct A { union { int n = 0; int m; }; }; const A a; ``` This is valid, but Clang rejects: ```console <source>:7:9: error: default initialization of an object of const type 'const A' without a user-provided default constructor ``` (Clang accepts if `m` is removed.) See https://eel.is/c++draft/dcl.init#general-8.3 -- because `n` has an initializer, the entire anonymous union is treated as initialized.

[ofAlpaca]

This is my first time contributing to this project. I would like to give it a try. Could you assign it to me?

[ofAlpaca]

Hi @zygoloid and @efriedma-quic,
I'm new here, so I don't quite understand the purpose of the 'good first issue' label.
I've noticed @Rajveer100 has been resolving many issues labeled as such.
I thought these issues were meant to help newcomers like me become more familiar with the LLVM project.
However, it looks like any contributor can take over this issue and fix it on their own without notifying assignee.
If senior LLVM contributors can solve any issues they want, what's the point of labeling them as 'good first issue'?
Once @Rajveer100 fix it, you can close this issue any time.
I think beginner is not welcome here.

[AaronBallman]

Hi @zygoloid and @efriedma-quic, I'm new here, so I don't quite understand the purpose of the 'good first issue' label.

We put that label on issues when we think it's something that a new contributor might be able to tackle as a way to get introduced to the code base.

I've noticed @Rajveer100 has been resolving many issues labeled as such. I thought these issues were meant to help newcomers like me become more familiar with the LLVM project. However, it looks like any contributor can take over this issue and fix it on their own without notifying assignee. If senior LLVM contributors can solve any issues they want, what's the point of labeling them as 'good first issue'? Once @Rajveer100 fix it, you can close this issue any time. I think beginner is not welcome here.

Any contributor can tackle any issue they'd like; we don't typically use issue assignments because of how easy it is for the assignment to go stale and then nobody works on it (but we have started assigning people when they ask to be assigned, sort of as a soft trial to see how well it goes). FWIW, @Rajveer100 is also a relative newcomer to the community and they may not have noticed that this issue was already assigned to you.

[ofAlpaca]

Hi @AaronBallman
Thank you for your explanation.
I agree with you about the problem of assignment getting stalled.
However, @Rajveer100's behavior is still very unfriendly.