This module wraps the MPFR library functions.
The MPFR library is a C library for multiple-precision floating-point
computations with exact rounding (also called correct rounding). It
is based on the GMP multiple-precision C library.
The main goal of MPFR is to provide a library for multiple-precision
floating-point computation which is both efficient and has a
well-defined semantics. It copies the good ideas from the
ANSI/IEEE-754 standard for double-precision floating-point
arithmetic (53-bit mantissa).
It also offers a wide range of trig, log, and other functions, and
constants.
MPFR, like GMP, is free. It is distributed under the
GNU Lesser General Public License (GNU Lesser GPL).
Because this module wraps the MPFR functions, it requires that both
the GMP C library and the MPFR C library have been installed.
For the GMP library see:
http://gmplib.org
For the MPFR library see:
http://www.mpfr.org
Minimum required version of gmp is gmp-4.2.0.
Minimum required version of mpfr is mpfr-2.4.0.
Some features of Math::MPFR require later versions
of the gmp and mpfr libraries.
You'll get errors if you try to build Math::MPFR against an
insufficiently recent version of the mpfr library.
To build this module you need perl 5.6.0 or later. I'm not sure of all
the requirements to get it to build with earlier versions of perl, but
it's definite that the XS code relating to operator overloading will
not compile with perl versions prior to 5.6.
Build in the usual way:
perl Makefile.PL
make
make test
make install
When building this module, the GMP and MPFR libraries will need to be
accessible. If those files are in a location where your compiler does
not find them by default, then instead of running 'perl Makefile.PL',
you'll need to run:
perl Makefile.pl INC="-I/path/to/gmp_includes -I/path/to/mpfr_includes"
LIBS="-L/path/to/mpfr_lib -lmpfr -L/path/to/gmp_lib -lgmp"
Other commandline args you may want to provide to 'perl Makefile.PL' are
D64=1 (for _Decimal64 support) and F128=1 (for __float128 support).
See the Decimal64 and Float128 sections (below).
==================================
64-bit-int and long double support
==================================
If your perl's Config reports that 'ivsize' is greater than or
equal to 8 && 'ivtype' is not 'long', then Math::MPFR will, by default,
be built with access to the mpfr_*_uj and mpfr_*_sj functions. Else,
access to those functions is, by default, denied. You can override the
default by opening up the Makefile.PL and uncommenting the appropriate
line (just a few lines down from the top of the file ... the comments
in that file should make it clear).
Similarly, if your perl's Config reports that 'nvsize' is greater than
8 then Math::MPFR will, be built with access to the mpfr_*_ld functions.
Else, access to those functions is, by default, denied. Again, you can
override the default by opening up the Makefile.PL and uncommenting the
appropriate line.
I believe it to be both unnecessary and inadvisable to override the
default - but if you do find that you need to override the default,
please let me know. That you have found it necessary to override the
default is something I probably ought to know about.
I am contactable by email at sisyphus at(@) cpan dot (.) org.
=========================================================
Conversion between Math::Decimal64 and Math::MPFR objects
=========================================================
The functions Rmpfr_set_DECIMAL64 and Rmpfr_get_DECIMAL64 allow
conversion between Math::Decimal64 (_Decimal64) and Math::MPFR (mpfr_t)
objects.
For this feature to be available you firstly need a compiler that has
the _Decimal64 data type.
Then you need to build and install Math::Decimal64.
You'll also need to have the mpfr library (version 3.2.0 at least)
configured to handle the _Decimal64-mpfr_t conversions. See the mpfr
documentation for details of that aspect.
If those pre-conditions are met, you also need to provide D64=1
as a command line argument to 'perl Makefile.PL'. (Altenatively you
could edit the Makefile.PL before building Math::MPFR and set its
$have_decimal64 variable to 1.)
Similarly, ensuring that Decimal64 support is *not* built in can be
achieved by leaving $have_decimal64 set to zero.
=========================================================
Conversion between Math::Float128 and Math::MPFR objects
=========================================================
The functions Rmpfr_set_FLOAT128 and Rmpfr_get_FLOAT128 allow
conversion between Math::Float128 (__float128) and Math::MPFR (mpfr_t)
objects.
For this feature to be available you firstly need a compiler that has
the __float128 data type.
Then you need to build and install Math::Float128.
You'll also need to have the mpfr library (version 3.2.0 at least)
configured to handle the __float128-mpfr_t conversions. See the mpfr
documentation for details of that aspect. (It's just a matter of
providing Configure with the --enable-float128 option when building
mpfr-3.2.0 or later.)
If those pre-conditions are met, you also need to provide F128=1
as a command line argument to 'perl Makefile.PL'. (Altenatively you
could edit the Makefile.PL before building Math::MPFR and set its
$have_float128 variable to 1.)
Similarly, ensuring that __float128 support is *not* built in can be
achieved by leaving $have_float128 set to zero.
==========================================================
Conversion between Math::LongDouble and Math::MPFR objects
==========================================================
There are 2 functions provided for conversion between Math::LongDouble
and Math::MPFR objects:
Rmpfr_set_LD (assign the value of a Math::LongDouble object to a
Math::MPFR object)
Rmpfr_get_LD (assign the value of a Math::MPFR object to a
Math::Long Double object)
If your perl's nvtype is long double, just use Rmpfr_set_ld and
Rmpfr_get_ld instead.
Math::LongDouble provides 'long double' support for perls whose
nvtype is other than 'long double'.
=======================================================
Conversion between __float128 NV and Math::MPFR objects
=======================================================
As of perl-5.21.4 it is possible to build perl with an nvtype of
__float128. This is achieved by specifying -Dusequadmath as a
Configure arg .... but, it will not work for all architectures
and compilers.
If $Config{nvtype} specifies '__float128' && if your mpfr library
(need 3.2.0 or later) was built with --enable-float128 then you
can pass your __float128 NV values to and from Math::MPFR using
the Rmpfr_set_float128() and Rmpfr_get_float128() functions.
And the overloaded operations will also accept the __float128 NV
arguments as one would want.
However, for that to work, you also need to have built Math::MPFR
by starting with:
perl Makefile.PL F128=1
(Altenatively you could edit the Makefile.PL before building
Math::MPFR and set its $have_float128 variable to 1.)