------- Comment #36 from paolo dot carlini at oracle dot com 2009-10-30 02:51 ------- First, send it to me and Benjamin (b...@redhat.com), because we want to be 100% sure everything is ok. Then, start thinking about the fix ;) Really, please-please, keep it minimal, we don't want for now the super-dupe algorithm, we want something using only swaps and not regressing performance-wise wrt what we have now. And clear, easy to understand. Additional testcases are always welcome.
-- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41351
