http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56935
--- Comment #6 from Yuri Rumyantsev <ysrumyan at gmail dot com> 2013-04-22 14:46:16 UTC --- Richard, Sorry for troubles since we found out the real cause of performance degradation - code layout was changed after your fix and it caused ~5% slowdown on 253.perlbmk bench on IVB. So you can close this bug as 'user error'.