http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47648
Tobias Burnus <burnus at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |burnus at gcc dot gnu.org
--- Comment #1 from Tobias Burnus <burnus at gcc dot gnu.org> 2011-02-09
07:05:13 UTC ---
See PR 47649 for configure logs.
The issue is that __float128 is supported, but due to missing functionality
(unclear whether FreeBSD's libc or libgcc should provide them) libquadmath does
not build - cf. bug 47649 comment 5.
The question is how one solves this. A work around would be disable libquadmath
via libgfortran's configure if it is not available. As REAL(16) will be still
available, there will be test-suite failures.
I think a proper workaround is to build with --disable-libquadmath-support
until the _U_Q* are supported. The question is only whether this is a GCC bug
(libgcc) or a FreeBSD bug (libc).
