package mopsa

  1. Overview
  2. Docs
MOPSA: A Modular and Open Platform for Static Analysis using Abstract Interpretation

Install 

dune-project
 Dependency

mopsa.lip6.fr Readme Changelog Edit opam file Versions (3)

Authors

  1. A
    Antoine Miné
  2. A
    Abdelraouf Ouadjaout
  3. M
    Matthieu Journault
  4. A
    Aymeric Fromherz
  5. R
    Raphaël Monat
  6. F
    Francesco Parolini
  7. M
    Marco Milanese
  8. J
    Jérôme Boillot

Maintainers

  1. A
    Antoine Miné <antoine.mine@lip6.fr>
  2. A
    Abdelraouf Ouadjaout <ouadjaout@gmail.com>
  3. R
    Raphaël Monat <raphael.monat@inria.fr>

Sources

mopsa-analyzer-v1.2.tar.gz
md5=37966e98ffeebcedc09bd6e9b2b81f69
sha512=40d4d826c25f680766c07eccbabdf5e8a4fa023016e8a164e4e4f6b3781c8484dc4df437055721dfd19b9db8fb7fe3b61236c4833186d346fc7204a68d01eaaa

doc/mopsa.mopsa_utils/Mopsa_utils/ItvUtils/Float/Single/index.html

Module Float.Single

Single precision operations.

Single precision numbers are stored inside OCaml's floats, that are actually double precion, but rounding is done to single precision.

Operations

val add_near : t -> t -> t
val add_up : t -> t -> t
val add_down : t -> t -> t
val add_zero : t -> t -> t

Addition

val sub_near : t -> t -> t
val sub_up : t -> t -> t
val sub_down : t -> t -> t
val sub_zero : t -> t -> t

Subtraction

val mul_near : t -> t -> t
val mul_up : t -> t -> t
val mul_down : t -> t -> t
val mul_zero : t -> t -> t

Multiplication

val mulz_near : t -> t -> t
val mulz_up : t -> t -> t
val mulz_down : t -> t -> t
val mulz_zero : t -> t -> t

Special multiplication where 0 times an infinity is 0, not NaN. This is particularly useful for interal bounds.

val div_near : t -> t -> t
val div_up : t -> t -> t
val div_down : t -> t -> t
val div_zero : t -> t -> t

Division

val divz_near : t -> t -> t
val divz_up : t -> t -> t
val divz_down : t -> t -> t
val divz_zero : t -> t -> t

Special division where 0 / 0 is 0, not NaN. This is particularly useful for interal bounds.

val mod_near : t -> t -> t
val mod_up : t -> t -> t
val mod_down : t -> t -> t
val mod_zero : t -> t -> t

Remainder.

val square_near : t -> t
val square_up : t -> t
val square_down : t -> t
val square_zero : t -> t

Square.

val sqrt_near : t -> t
val sqrt_up : t -> t
val sqrt_down : t -> t
val sqrt_zero : t -> t

Square root.

val round_int_near : t -> t
val round_int_up : t -> t
val round_int_down : t -> t
val round_int_zero : t -> t

Rounding to an integer.

val of_double_near : t -> t
val of_double_up : t -> t
val of_double_down : t -> t
val of_double_zero : t -> t

Rounding from double to single precision.

val of_int_near : int -> t
val of_int_up : int -> t
val of_int_down : int -> t
val of_int_zero : int -> t
val of_int_cur : int -> t

Conversion from int with rounding.

val of_int64_near : int64 -> t
val of_int64_up : int64 -> t
val of_int64_down : int64 -> t
val of_int64_zero : int64 -> t
val of_int64_cur : int64 -> t

Conversion from int64 with rounding.

Code from Zarith to convert from Z.t to float using the current rounding mode. We add a version for single precision rounding.

val round_z_to_single : Z.t -> bool -> t
val z_to_single : Z.t -> t
val of_z_near : Z.t -> t
val of_z_up : Z.t -> t
val of_z_down : Z.t -> t
val of_z_zero : Z.t -> t

Conversion from Zarith with rounding.

Floating-point format characteristics

val mantissa_bits : int
val exponent_bits : int
val exponent_bias : int
val min_exponent : int
val max_exponent : int
val nan_infinity_exponent : int
val min_denormal : t
val min_normal : t
val max_normal : t
val max_exact : t
val ulp : t

Units in the last place (relative precision).

val rep_of_bits : int32 -> bit_float
val bits_of_rep : bit_float -> int32
val to_bits : t -> int32
val of_bits : int32 -> t
val succ : t -> t
val pred : t -> t

Returns the float immediately follownig or preceeding the argument.

val succ_zero : t -> t
val pred_zero : t -> t

As succ and pred, but does not cross zero.

val of_string_up : string -> t
val of_string_down : string -> t

Conversion from string.