[Cryptography] Apple's Early Random PRNG

Peter Gutmann pgut001 at cs.auckland.ac.nz
Tue Mar 18 22:42:17 EDT 2014


Theodore Ts'o <tytso at mit.edu> writes:

>If anyone has any suggestions about how to influence ARM SOC vendors to
>provide something liek RDRAND, short of compromising photos from web cams of
>company execs provided courtesy of GCHQ :-), I'm sure lots of people would
>appreciate any ideas....

Some of them already do, but they're so badly done that they're effectively
useless: No way to check for their presence a la Intel's CPUID (you just have
to know that the SoC you're using supports it), no way to access it from user
space, weird restrictions on how it can be used (you need to set up various
hardware registers in specific ways, there can be access conflicts if
different tasks try and get at things at the same time, etc).  So it's there
(at least in some cases), but you can't have it.

Peter.


More information about the cryptography mailing list