https://gcc.gnu.org/bugzilla/show_bug.cgi?id=127063
--- Comment #20 from Harald Anlauf <anlauf at gcc dot gnu.org> --- (In reply to Harald Anlauf from comment #19) > (In reply to Mikael Morin from comment #18) > > I'm seeing the failure on aarch64-unknown-linux-gnu at revision > > r17-5755-g78cfe216a8e042: > > STOP 16 > > > > the difference amounts to 1.0462256988369260E-015, so the test fails. > > With 1e-14 as threshold, the test passes. > > Just update the threshold to what you think it should be. > The 1.e-15 was a little tight for some of the platforms we support. I just tested: spacing(atand(y)) gives 3.5527136788005009E-015 So 4.e-15 should be safe, or any larger number. Will push this, if there are no objections.
