[cryptography] Kernel space vs userspace RNG

Thierry Moreau thierry.moreau at connotech.com
Thu May 5 09:54:52 EDT 2016


On 05/05/16 09:40 AM, shawn wilson wrote:
> Just reflecting on the Linux RNG thread a bit ago, is there any
> technical reason to have RNG in kernel space?

The procurement of an RNG source for crypto is always a *system* design 
issue.

The expectation that a kernel offering (intended for a wide range of CPU 
architectures, each of which being deployed in its own range of systems) 
can solve this system issue is IMHO naive.

Thus, kernel space vs user space makes little difference.

This being said, the kernel developers appear to make good faith efforts 
to adapt to the ever evolving digital electronics paradigms prevailing 
in a few mainstream system architectures. Is this effective versus some 
criteria for RNG quality? Is this good enough for you?

It's your duty to figure out, I guess.

Regards,

- Thierry Moreau



More information about the cryptography mailing list