https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124226
--- Comment #4 from Liam Powell <liam at liampwll dot com> --- Even if this were the result of fuzzing the compiler, which it is not, then I still think it's unreasonable for this to be marked as a WONTFIX as it's a silent memory corruption bug. The incorrect address that's used here is always going to be pointing to valid memory so it's very likely that usage of the access type object is going to result in memory corruption which will be very difficult to track down. I only happened to see this one because I tested that the fallback handler works immediately after changing the above line, I think there's a lot of cases where this would go unnoticed and then be very difficult to track down later.
