https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69573
--- Comment #12 from Martin Sebor <msebor at gcc dot gnu.org> ---
(In reply to Dominique d'Humieres from comment #11)
Thanks for the update. That is what I would expect as well. But removing the
"{ target { ! { ia32 } } }" directives without doing anything for targets where
the alignment of long double is less than 16 (such as IA32 where it's 4) will
cause failures there. The patch I referenced in comment #2 is designed to
avoid those failures. Does the patch let the test pass on
x86_64-apple-darwin15, both with -m32 and -m64?
