flint.h – global definitions

Macros

The file flint.h contains various useful macros.

The macro constant FLINT_BITS is set at compile time to be the number of bits per limb on the machine. FLINT requires it to be either 32 or 64 bits. Other architectures are not currently supported.

The macro constant FLINT_D_BITS is set at compile time to be the number of bits per double on the machine or one less than the number of bits per limb, whichever is smaller. This will have the value \(53\) or \(31\) on currently supported architectures. Numerous internal functions using precomputed inverses only support operands up to FLINT_D_BITS bits, hence the macro.

The macro FLINT_ABS(x) returns the absolute value of \(x\) for primitive signed numerical types. It might fail for least negative values such as INT_MIN and WORD_MIN.

The macro FLINT_MIN(x, y) returns the minimum of \(x\) and \(y\) for primitive signed or unsigned numerical types. This macro is only safe to use when \(x\) and \(y\) are of the same type, to avoid problems with integer promotion.

Similar to the previous macro, FLINT_MAX(x, y) returns the maximum of \(x\) and \(y\).

mp_limb_t FLINT_BIT_COUNT(mp_limb_t x)

Returns the number of binary bits required to represent an ulong x. If \(x\) is zero, returns \(0\).

Derived from this there are the two macros FLINT_FLOG2(x) and FLINT_CLOG2(x) which, for any \(x \geq 1\), compute \(\lfloor \log_2 x \rfloor\) and \(\lceil \log_2 x \rceil\).

To determine the current FLINT version a number of macros are available. For example, if the current FLINT version is 2.4.0 then __FLINT_VERSION will have the value \(2\), __FLINT_MINOR will have the value \(4\) and __FLINT_PATCHLEVEL will have the value \(0\).

The __FLINT_RELEASE macro gives a single number representing the FLINT version. For example, it will have the value 20400 for version 2.4.0.

The FLINT_VERSION macro is a static text string giving the version number, e.g. “2.4” or “2.4.1”. Note that if the final digit is a zero it is suppressed.

Integer types

The char, short and int types are assumed to be two’s complement types with exactly 8, 16 and 32 bits. This is not technically guaranteed by the C standard, but it is true on mainstream platforms. ; Since the C types long and unsigned long do not have a standardized size in practice, FLINT defines slong and ulong types which are guaranteed to be 32 bits on a 32-bit system and 64 bits on a 64-bit system. They are also guaranteed to have the same size as GMP’s mp_limb_t. GMP builds with a different limb size configuration are not supported at all. For convenience, the macro FLINT_BITS specifies the word length (32 or 64) of the system.

type slong

The slong type is used for precisions, bit counts, loop indices, array sizes, and the like, even when those values are known to be nonnegative. It is also used for small integer-valued coefficients. In method names, an slong parameter is denoted by si, for example arb_add_si().

The constants WORD_MIN and WORD_MAX give the range of this type. This type can be printed with flint_printf using the format string %wd.

type ulong

The ulong type is used for integer-valued coefficients that are known to be unsigned, and for values that require the full 32-bit or 64-bit range. In method names, a ulong parameter is denoted by ui, for example arb_add_ui().

The constant UWORD_MAX gives the range of this type. This type can be printed with flint_printf using the format string %wu.

The following GMP-defined types are used in methods that manipulate the internal representation of numbers (using limb arrays).

type mp_limb_t

A single limb.

type mp_ptr

Pointer to a writable array of limbs.

type mp_srcptr

Pointer to a read-only array of limbs.

type mp_size_t

A limb count (always nonnegative).

type flint_bitcnt_t

A bit offset within an array of limbs (always nonnegative).

Allocation Functions

void *flint_malloc(size_t size)

Allocate size bytes of memory.

void *flint_realloc(void *ptr, size_t size)

Reallocate an area of memory previously allocated by flint_malloc(), flint_realloc(), or flint_calloc().

void *flint_calloc(size_t num, size_t size)

Allocate num objects of size bytes each, and zero the allocated memory.

void flint_free(void *ptr)

Free a section of memory allocated by flint_malloc(), flint_realloc(), or flint_calloc().

Random Numbers

type flint_rand_s

A structure holding the state of a flint pseudo random number generator.

type flint_rand_t

An array of length 1 of flint_rand_s.

flint_rand_s *flint_rand_alloc()

Allocates a flint_rand_t object to be used like a heap-allocated flint_rand_t in external libraries. The random state is not initialised.

void flint_rand_free(flint_rand_s *state)

Frees a random state object as allocated using flint_rand_alloc().

void flint_randinit(flint_rand_t state)

Initialize a flint_rand_t.

void flint_randclear(flint_rand_t state)

Free all memory allocated by flint_rand_init().

Thread functions

void flint_set_num_threads(int num_threads)

Set up a thread pool of num_threads - 1 worker threads (in addition to the master thread) and set the maximum number of worker threads the master thread can start to num_threads - 1.

This function may only be called globally from the master thread. It can also be called at a global level to change the size of the thread pool, but an exception is raised if the thread pool is in use (threads have been woken but not given back). The function cannot be called from inside worker threads.

int flint_get_num_threads(void)

When called at the global level, this function returns one more than the number of worker threads in the Flint thread pool, i.e. it returns the number of workers in the thread pool plus one for the master thread.

In general, this function returns one more than the number of additional worker threads that can be started by the current thread.

Use thread_pool_wake() to set this number for a given worker thread.

See also: flint_get_num_available_threads().

int flint_set_num_workers(int num_workers)

Restricts the number of worker threads that can be started by the current thread to num_workers. This function can be called from any thread.

Assumes that the Flint thread pool is already set up.

The function returns the old number of worker threads that can be started.

The function can only be used to reduce the number of workers that can be started from a thread. It cannot be used to increase the number. If a higher number is passed, the function has no effect.

The number of workers must be restored to the original value by a call to flint_reset_num_workers() before the thread is returned to the thread pool.

The main use of this function and flint_reset_num_workers() is to cheaply and temporarily restrict the number of workers that can be started, e.g. by a function that one wishes to call from a thread, and cheaply restore the number of workers to its original value before exiting the current thread.

void flint_reset_num_workers(int num_workers)

After a call to flint_set_num_workers() this function must be called to set the number of workers that may be started by the current thread back to its original value.

Input/Output

int flint_printf(const char *str, ...)

int flint_vprintf(const char *str, va_list ap)

int flint_fprintf(FILE *f, const char *str, ...)

int flint_sprintf(char *s, const char *str, ...)

These are equivalent to the standard library functions printf, vprintf, fprintf, and sprintf with an additional length modifier “w” for use with an mp_limb_t type. This modifier can be used with format specifiers “d”, “x”, or “u”, thereby outputting the limb as a signed decimal, hexadecimal, or unsigned decimal integer.

int flint_scanf(const char *str, ...)

int flint_fscanf(FILE *f, const char *str, ...)

int flint_sscanf(const char *s, const char *str, ...)

These are equivalent to the standard library functions scanf, fscanf, and sscanf with an additional length modifier “w” for reading an mp_limb_t type.

Exceptions

When FLINT encounters a problem, mostly illegal input, it currently aborts. There is an experimental interface for generating proper exceptions flint_throw, but this is currently rarely used and experimental - you should expect this to change.

At the end, all of FLINT’s exceptions call abort() to terminate the program. Using flint_set_abort(void (*abort_func)(void)), the user can install a function that will be called instead. Similar to the exceptions, this should be regarded as experimental.