https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124819
--- Comment #18 from Jerry DeLisle <jvdelisle at gcc dot gnu.org> --- Created attachment 64206 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=64206&action=edit A modified test case that is less sensitive to variations around 475.78 I played around a bit to see if the test could be de-senitized. This tests fine here with quite a bit of variation in epsilon and the value around 475. Seems to work OK here, but one can experiment with it.
