https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68665
--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> --- (In reply to Robert Shuler from comment #5) > I just did a from-scratch installation on a different computer (at the > office) with Win7 of cygwin-64 GCC and the quad math package. The results > are exactly the same. Built-in _float128 works, but identical error > messages trying to use the functions. In this case, however THE DLLs ARE > MISSING. > > Here is result of trying -lquadmath on the office installation. Try: gcc -o test128a test128a.c -lquadmath
