FRIDAY, JULY 24, 2026
libgfortran
libgfortran: Add long double math support for hppa64-hpux
libgfortran on hppa64-hpux now supports C99 long double math routines using libquadmath.
This change adds support for C99 math routines with long double precision on hppa64-hpux systems. Although the platform lacks native support for these routines, its long double type conforms to the IEEE 60559 specification, enabling the use of quad-precision routines from libquadmath.
In Details
The hppa64-hpux target now leverages libquadmath for long double math functions in libgfortran. This addresses PR libfortran/126305 by providing C99 math support where native implementations are unavailable, aligning with the platform's 16-byte IEEE 754 long double.