https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126299
--- Comment #4 from Mikael Pettersson <mikpelinux at gmail dot com> --- (In reply to pkoning from comment #3) > Thanks Mikael. But that "better looking" output only considers the upper > half of the SImode value, which is clearly wrong. It's wrong for arbitrary comparisons, but this test case only does "< 0" and for that you only need to look at the high word. I haven't looked at what it does for the other test cases.
