gmp, The GNU Multiple Precision Arithmetic Library
Free library for arbitrary precision arithmetic, operating on signed integers, rational numbers, and floating point numbers.
Website: https://www.gmplib.org
Platforms supported: Win32, Win64, Linux, FreeBSD, Haiku
Headers to include: gmp.bi
Header version: 6.0.0
Haiku R1 beta 6 amd64: `examples/manual/libraries/gmp.bas` compiled and ran
with FreeBASIC 1.20.4-3 against HaikuPorts GMP 6.3.0, computing `2^65536`.
Native types and ownership
Website: https://www.gmplib.org
Platforms supported: Win32, Win64, Linux, FreeBSD, Haiku
Headers to include: gmp.bi
Header version: 6.0.0
Haiku R1 beta 6 amd64: `examples/manual/libraries/gmp.bas` compiled and ran
with FreeBASIC 1.20.4-3 against HaikuPorts GMP 6.3.0, computing `2^65536`.
Native types and ownership
The Win64 SDK uses 64-bit mp_limb_t values while C long remains 32-bit. mp_bitcnt_t, mp_size_t and mp_exp_t keep their C widths. Use the binding's named types for limb pointers and low-level calls; FreeBASIC Integer is not a general replacement for a C type.
mpz_get_str(0, ...) allocates through GMP's current allocator. Release such a string through the corresponding registered GMP free function, obtained with mp_get_memory_functions. The example supplies a FreeBASIC-owned buffer instead, sized by mpz_sizeinbase plus two bytes for a sign and terminating zero, and releases it with Deallocate. Release initialized number internals with mpz_clear before freeing the caller's number buffer.
The installed GMP 6.3.0 SDK has been checked against native C layouts and actual high-bit limb arguments and carry returns on Win32 and Win64, with both FreeBASIC backends. The decimal example's complete 19,729-digit result is checked independently.
mpz_get_str(0, ...) allocates through GMP's current allocator. Release such a string through the corresponding registered GMP free function, obtained with mp_get_memory_functions. The example supplies a FreeBASIC-owned buffer instead, sized by mpz_sizeinbase plus two bytes for a sign and terminating zero, and releases it with Deallocate. Release initialized number internals with mpz_clear before freeing the caller's number buffer.
The installed GMP 6.3.0 SDK has been checked against native C layouts and actual high-bit limb arguments and carry returns on Win32 and Win64, with both FreeBASIC backends. The decimal example's complete 19,729-digit result is checked independently.
Example
'' examples/manual/libraries/gmp.bas
''
'' Example extracted from the FreeBASIC Manual
'' from topic 'gmp, The GNU Multiple Precision Arithmetic Library'
'' Purpose: Compute and print 2^65536 through the native GMP library.
'' Responsibilities: Own the number buffer and decimal output storage.
'' This example contains no custom GMP allocator or limb arithmetic.
''
'' See Also: https://www.freebasic.net/wiki/wikka.php?wakka=ExtLibgmp
'' --------
''
'' Ownership:
''
'' The example owns the __mpz_struct buffer allocated for bignum and the
'' output string supplied to mpz_get_str(). Each is released on its matching
'' successful allocation path.
#include Once "gmp.bi"
Dim As Integer status = 0
Dim As mpz_ptr bignum = Allocate(SizeOf(__mpz_struct))
If bignum = 0 Then
Print "Unable to allocate the GMP number."
status = 1
Else
mpz_init_set_si(bignum, 2)
mpz_pow_ui(bignum, bignum, 65536)
Print "2^65536 = ";
' GMP can use a custom allocator. Supply our own output buffer so that
' Deallocate() never receives memory owned by the library's allocator.
' Two extra bytes cover a possible sign and the terminating zero.
Dim As UInteger digits = mpz_sizeinbase(bignum, 10)
Dim As ZString Ptr s
If digits > 0 AndAlso digits <= CUInt(-1) - 2 Then s = Allocate(digits + 2)
If s <> 0 Then
If mpz_get_str(s, 10, bignum) <> 0 Then
Print *s;
Else
Print "<conversion failed>";
status = 1
End If
Deallocate(s)
s = 0
Else
Print "<output allocation failed>";
status = 1
End If
Print
mpz_clear(bignum)
Deallocate(bignum)
bignum = 0
End If
End status
'' end of examples/manual/libraries/gmp.bas
''
'' Example extracted from the FreeBASIC Manual
'' from topic 'gmp, The GNU Multiple Precision Arithmetic Library'
'' Purpose: Compute and print 2^65536 through the native GMP library.
'' Responsibilities: Own the number buffer and decimal output storage.
'' This example contains no custom GMP allocator or limb arithmetic.
''
'' See Also: https://www.freebasic.net/wiki/wikka.php?wakka=ExtLibgmp
'' --------
''
'' Ownership:
''
'' The example owns the __mpz_struct buffer allocated for bignum and the
'' output string supplied to mpz_get_str(). Each is released on its matching
'' successful allocation path.
#include Once "gmp.bi"
Dim As Integer status = 0
Dim As mpz_ptr bignum = Allocate(SizeOf(__mpz_struct))
If bignum = 0 Then
Print "Unable to allocate the GMP number."
status = 1
Else
mpz_init_set_si(bignum, 2)
mpz_pow_ui(bignum, bignum, 65536)
Print "2^65536 = ";
' GMP can use a custom allocator. Supply our own output buffer so that
' Deallocate() never receives memory owned by the library's allocator.
' Two extra bytes cover a possible sign and the terminating zero.
Dim As UInteger digits = mpz_sizeinbase(bignum, 10)
Dim As ZString Ptr s
If digits > 0 AndAlso digits <= CUInt(-1) - 2 Then s = Allocate(digits + 2)
If s <> 0 Then
If mpz_get_str(s, 10, bignum) <> 0 Then
Print *s;
Else
Print "<conversion failed>";
status = 1
End If
Deallocate(s)
s = 0
Else
Print "<output allocation failed>";
status = 1
End If
mpz_clear(bignum)
Deallocate(bignum)
bignum = 0
End If
End status
'' end of examples/manual/libraries/gmp.bas