[Bug tree-optimization/49079] [4.6/4.7 Regression] Bogus constant folding

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49079

--- Comment #1 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-05-20 
12:34:27 UTC ---
Hm, no, it's not constant-folding from s2, I misread.