[Bug c++/105524] New: -Wmaybe-uninitialized false-positive with switch of enum with more than 2 elements

[Bug c++/105524] New: -Wmaybe-uninitialized false-positive with switch of enum with more than 2 elements

[nunoplopes at sapo dot pt]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105524

            Bug ID: 105524
           Summary: -Wmaybe-uninitialized false-positive with switch of
                    enum with more than 2 elements
           Product: gcc
           Version: 12.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: nunoplopes at sapo dot pt
  Target Milestone: ---

The following program gets a false-positive when compiled with `-Wall -O3
-fstrict-enums`:


// switch to 0 to make false-positive go away
#define DEF_C 1

enum foo { A, B
#if DEF_C
, C
#endif
 };
void g(const char*);

void f(foo x) {
  const char *str;
  switch (x) {
    case A: str = "a"; break;
    case B: str = "b"; break;
#if DEF_C
    case C: str = "c"; break;
#endif
  }
  g(str);
}


https://gcc.godbolt.org/z/h1rWqezY3


<source>: In function 'void f(foo)':
<source>:19:4: error: 'str' may be used uninitialized
[-Werror=maybe-uninitialized]
   19 |   g(str);
      |   ~^~~~~
<source>:11:15: note: 'str' was declared here
   11 |   const char *str;
      |               ^~~
cc1plus: all warnings being treated as errors
Compiler returned: 1


The error only repros with enums with more than 2 elements.

[Bug c++/105524] -Wmaybe-uninitialized false-positive with switch of enum with more than 2 elements

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105524

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
You misunderstand the semantics of strict enum.

That is here a value of 3 is still valid under strict enum and all.