package mirage-crypto-rng-miou-unix

  1. Overview
  2. Docs
Legend:
Library
Module
Module type
Parameter
Class
Class type

RNG seeding on Miou_unix.

This module initializes a RNG with getrandom(), and CPU RNG. On BSD system (FreeBSD, OpenBSD, MacOS) getentropy() is used instead of getrandom(). On Windows 10 or higher, BCryptGenRandom() is used with the default RNG. Windows 8 or lower are not supported by this library.

module Pfortuna : Mirage_crypto_rng.Generator

Pfortuna, a domain-safe CSPRNG proposed by Schneier.

type rng

Type of tasks seeding the RNG.

val initialize : ?g:'a -> ?sleep:int64 -> 'a Mirage_crypto_rng.generator -> rng

initialize ?g ?sleep (module Generator) will allow the RNG to operate in a returned task. This task periodically launches sub-tasks that seed the engine (using getrandom(), getentropy() or BCryptGenRandom() depending on the system). These sub-tasks must be cleaned periodically (in seconds) according to the sleep parameter given (defaults to 1 second).

The user must then kill the returned task at the end of the program to be sure to clean everything. Otherwise, Miou will complain with the exception Still_has_children.

We strongly recommend using Pfortuna as an RNG engine rather than Mirage_crypto_rng.Fortuna. The engine is launched in parallel with the other tasks if at least one domain is available. To ensure that there is no compromise in the values generated by a data-race, Pfortuna is an domain-safe implementation of Fortuna.

The user cannot make any subsequent calls to initialize. In other words, you can only initialise a single rng task. You must kill the returned rng if you want to re-initialise the RNG.

A basic usage of mirage-crypto-rng-miou-unix is:

let () = Miou_unix.run @@ fun () ->
  let rng = Mirage_crypto_rng_miou_unix.(initialize (module Pfortuna)) in
  let str = Mirage_crypto_rng.generate 16 in
  Format.printf "random: %S\n%!" str;
  Mirage_crypto_rng_miou_unix.kill rng
val kill : rng -> unit

kill rng terminates the background task which seeds the RNG.

OCaml

Innovation. Community. Security.