https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121438

            Bug ID: 121438
           Summary: Improve mergability of array's initializer_list
           Product: gcc
           Version: 14.2.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: enhancement
          Priority: P3
         Component: c++
          Assignee: pinskia at gcc dot gnu.org
          Reporter: pinskia at gcc dot gnu.org
  Target Milestone: ---
            Target: aarch64

I noticed that
```
#include <initializer_list>
std::initializer_list<int> a = {1, 2, 3, 0};
```

is mergeable on x86_64 but not aarch64. This is because the alignment of the
array decl is smaller than size.
I implemented something similar for CSWITCH in r16-2595-gf1c80147641783 and we
should do something similar in the gimplifier when it comes to the alignment.

Reply via email to