re PR fortran/46625 (libquadmath: Mangle internal symbols; rename __float128 <->...
authorJakub Jelinek <jakub@redhat.com>
Sun, 16 Jan 2011 16:55:27 +0000 (17:55 +0100)
committerJakub Jelinek <jakub@gcc.gnu.org>
Sun, 16 Jan 2011 16:55:27 +0000 (17:55 +0100)
commite313241f91a24365102a657ff8bb04a647b7963b
tree252065baee12c4b6545a47616cc3a36114d3be22
parent327f73c19ef091381ecc8a57dd0bb696732c3f0f
re PR fortran/46625 (libquadmath: Mangle internal symbols; rename __float128 <-> string functions)

PR fortran/46625
* quadmath.map (QUADMATH_1.0): Remove quadmath_strtopQ
and quadmath_dtoaq.  Add strtoflt128 and quadmath_flt128tostr.
* quadmath_weak.h (quadmath_strtopQ, quadmath_dtoaq): Remove.
(strtoflt128, quadmath_flt128tostr): Add.
* gdtoa/strtopQ.c (quadmath_strtopQ): Rename to...
(strtoflt128): ... this.  Return __float128, instead of writing
to memory pointed by last argument.
* quadmath.h: Use C style comments instead of C++ style.
(quadmath_strtopQ, quadmath_dtoaq): Remove prototypes.
(strtoflt128, quadmath_flt128tostr): Add prototypes.
* libquadmath.texi (quadmath_dtoaq): Rename to quadmath_flt128tostr.
(quadmath_strtopQ): Rename to strtoflt128.  Adjust prototype,
adjust examples.
* quadmath_io.c (quadmath_dtoaq): Rename to...
(quadmath_flt128tostr): ... this.
libgfortran/
* io/write_float.def (DTOAQ): Use quadmath_flt128tostr
instead of quadmath_dtoa.
* io/transfer128.c (tmp1, tmp2): New variables, bring in
strtoflt128 and quadmath_flt128tostr.
(transfer_real128, transfer_real128_write, transfer_complex128,
transfer_complex128_write): Remove tmp1/tmp2 variables.
* io/read.c (convert_real): Use strtoflt128 instead of
quadmath_strtopQ, adjust for the changed arguments and return
value.

From-SVN: r168856
libgfortran/ChangeLog
libgfortran/io/read.c
libgfortran/io/transfer128.c
libgfortran/io/write_float.def
libquadmath/ChangeLog
libquadmath/gdtoa/strtopQ.c
libquadmath/libquadmath.texi
libquadmath/quadmath.h
libquadmath/quadmath.map
libquadmath/quadmath_io.c
libquadmath/quadmath_weak.h