------- Comment #19 from pcarlini at suse dot de 2005-11-22 16:59 ------- (In reply to comment #18) > Well, yes. That would at least silence the warning (which would be annoying, > if we apply the C++ warning patch).
It would be more than annoying, it would be awful, because would be also emitted to end users of the library, not only at library build time. So... As I see the issue, either we come up with an actual miscompilation caused by the punning, or we have to apply the hack, for this library ABI. I'll try to think a bit more about the issue, then, in a few days we should make a decision... -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24975
