RDRAND

A module to use Intel's hardware RNG with python's random class

Installation

pip install . after cloning the repository to a local folder (preferably, don't use - python setup.py install - as it is deprecated since Python 3.11) 

USAGE

#easy_install rdrand
#python

>>>from rdrand import RdRandom
>>>r = RdRandom()

>>>from rdrand import RdSeedom
>>>s = RdSeedom()

At this point, ``r`` and ``s`` will behave just like ``random``

``RdRandom`` is a subclass of ``random.Random``, and behaves like ``random.Random``,
 but it uses inline assembly to access the hardware RNG using the RdRand instruction. This should be
a cryptographically secure drop in replacement for ``random`` with a prediction complexity bound of O(             2^128), if the Intel random number
generator is valid. No mitigation is done to modify the output of the hardware to prevent problems wit             h Intel's implementation. Caveat Emptor.

``RdSeedom`` is a subclass of ``random.Random``, and behaves like ``random.Random``,
 but it uses inline assembly to access the hardware RNG using the RdSeed instruction. This should be
a cryptographically secure drop in replacement for ``random`` returning full entropy bits, if the Inte             l random number
generator is valid. No mitigation is done to modify the output of the hardware to prevent problems wit             h Intel's implementation. Caveat Emptor.

Also, both RdRandom and RdSeedom include the function ``r.getrandbytes(i)`` where ``i`` is a positive              int. This returns a string
of length ``i`` filled with random bytes, which is ideal for generating a key or using directly in a p             rotocol.

Please note, as with any security solution, it is possible to subvert this. Please understand the full              context before deploying. I am not liable for misuse or clever hackers.

Certain changes made in C code to make it compatible with mingw64 gcc
tool-chain for seamless installation on windows. Tested on Windows 11 using
msys2 shell.