https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91423
--- Comment #5 from Anders Schau Knatten <anders at knatten dot org> --- (In reply to Andrew Pinski from comment #4) > Vec size = s.size; > > you are invoking the copy constructor here ... > Which means you are taking the address (implicitly). Good point. This should be safe though, as `Vec` itself is packed, so its copy constructor would have to handle unaligned arguments. (In fact, if `Vec` wasn't packed, the compiler would not allow `S` to be packed either.) I'm however not sure if we should expect gcc to take this information into account when deciding whether to emit the warning, what do you think?
