================
@@ -17389,6 +17389,19 @@ Sema::ActOnTag(Scope *S, unsigned TagSpec, TagUseKind
TUK, SourceLocation KWLoc,
Previous.clear();
}
+ // I think DC check should be DC->isStdNamespace()?
+ // Also these guards are questionable - it's possible to get incorrect
+ // codegen when the declared type does not match the expected type.
+ // e.g. something like
+ // namespace std { struct align_val_t { explicit align_val_t(size_t); } };
+ // can result in an operator new/delete decl picking up this specified
+ // align_val_t struct, but the align_val_t constructed implicitly has the
+ // mismatching internal definition.
+ //
+ // The correct course of action is probably to perform this logic at the
point
----------------
erichkeane wrote:
This is perhaps not a bad idea, though it could go in a different patch.
Typically we take the fact that defining that in user code is UB, so we can
count on its structure.
https://github.com/llvm/llvm-project/pull/113510
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
