http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40210
--- Comment #11 from Pawel Sikora <pluto at agmk dot net> 2011-10-07 18:45:57 UTC --- (In reply to comment #10) > (In reply to comment #9) > > Created attachment 25442 [details] > > testcase > > I think those are hard to optimize really since those are inline-asm really. > And the unsigned short one gets optimized correctly. ahhh, glibc uses a generic i386 implementation (ror+xchg) in <byteorder.h> while it has an optimized <byteswap.h> for i486+ :(
