http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50594
--- Comment #10 from Daniel Krügler <daniel.kruegler at googlemail dot com> 2011-10-03 13:49:01 UTC --- (In reply to comment #8) I agree that given the "make static" contract of -fwhole-program (which I was not aware about) the compiler behaves accordingly. I wonder whether it would be OK to consider the magic functions as special even for -fwhole-program, because they do have a "magic-kind-of" specification in the standard.
