Statement attribute on a case label suppresses 'bypassing variable initialization' error

[dwblaikie]

https://godbolt.org/z/j4b69q1dq

void Func(int x) {
    switch (x) {
        [[likely]] case 0:
        case 1:
            int i = 3;
        case 2:
            break;
    }
    switch (x) {
        case 0:
        case 1:
            int i = 3;
        case 2:
            break;
    }
}

<source>:14:9: error: cannot jump from switch statement to this case label
   14 |         case 2:
      |         ^
<source>:13:17: note: jump bypasses variable initialization
   13 |             int i = 3;
      |              

[shafik]

So this changes the AST: https://godbolt.org/z/nev6hxT19 adding:

|-AttributedStmt <line:4:9, line:6:22>
    |   | |-LikelyAttr <line:4:11>

before the CaseStmt I wonder if we are just not accounting for that.

[dwblaikie]

So this changes the AST: https://godbolt.org/z/nev6hxT19 adding:

|-AttributedStmt <line:4:9, line:6:22>
    |   | |-LikelyAttr <line:4:11>

before the CaseStmt I wonder if we are just not accounting for that.

Yep. Seems likely - I'd expect it's not a terribly difficult fix.

[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. In the comments of the issue, request for it to be assigned to you.
2. Fix the issue locally.
3. 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.
4. Create a Git commit.
5. Run git clang-format HEAD~1 to format your changes.
6. Open a pull request to the upstream repository on GitHub. Detailed instructions can be found in GitHub's documentation.

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: David Blaikie (dwblaikie)

https://godbolt.org/z/j4b69q1dq ``` void Func(int x) { switch (x) { [[likely]] case 0: case 1: int i = 3; case 2: break; } switch (x) { case 0: case 1: int i = 3; case 2: break; } } ``` ``` <source>:14:9: error: cannot jump from switch statement to this case label 14 | case 2: | ^ <source>:13:17: note: jump bypasses variable initialization 13 | int i = 3; | ```

[shashu8660]

hii My name is shashank s I'm this year outreachy applicant . As im beginner at open source i want to make my first contribution at good first issue so i would like @dwblaikie to assign this issue to me

[dwblaikie]

Sure thing - done

[shashu8660]

im facing little difficulty to take my first step in this issue @dwblaikie i would like you to guide me take the first step to solve this issue . I took so much time to install and learn the llvm work flow. plz i would like some assist to take first step to contribution.

[Rajveer100]

@dwblaikie @shafik
Due to the [[likely]] attribute a nested structure is being formed in the AST, so as per the jump scope checker, we are skipping it:

// It's okay to jump out from a nested scope.
  if (CommonScope == ToScope) return;

[dwblaikie]

im facing little difficulty to take my first step in this issue @dwblaikie i would like you to guide me take the first step to solve this issue . I took so much time to install and learn the llvm work flow. plz i would like some assist to take first step to contribution.

My usual approach when trying to figure this sort of thing out is to break at the diagnostic that's causing the problem (in this case the positive case has the diagnostic) - and have a new debugger session debugging the other case (non-diagnostic) and try to figure out where those things diverge.

Looks like @Rajveer100 might've pointed to a good/more precise place to start, though.