[Bug c/57080] Invalid optimization (-O2) when doing double -> int conversion (on big endian archs(?))

[ondrej at sury dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57080

--- Comment #1 from Ondřej Surý <ondrej at sury dot org> 2013-04-26 10:27:30 
UTC ---
Created attachment 29945
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=29945
Generated from gd.c, affected code is in clip_1d function