Messages in this thread | | | Date | Tue, 25 Jan 2000 00:26:20 -0500 | From | Sandy Harris <> | Subject | Re: Intel 810 Random Number Generator |
| |
David Whysong wrote: > > On Mon, 24 Jan 2000 nathan.zook@amd.com wrote: > > > ... As I recall from the press, the RNG calculates an > >index into a 2^16 byte array of "true random data", ... > > If that is really what the RNG does, then it's useless. What you describe > is equivalent to a normal algorithmic pseudo-random number generator.
Only if the index is generated algorithmically.
> If the index is somehow "randomly" (not algorithmically) generated, then > there is no point in having an array of 2^16 bytes! You could just return > the index as your random number.
There is a point. If your 16 bit values have only around 8 bits of entropy, then running them through such an array is a fast and (with good values in the array) effective method of generating outputs whose size matches the entropy.
(Methinks you could do as well or better with several smaller arrays, but that is not germane here.)
> So a large array implies an algorithmic approach,
By no means.
> which means the RNG will return little if any real entropy. > > Entropy has to come from somewhere something unguessable. Thermal > fluctuations or quantum spin states (Stern-Gerlach device, anyone) work > well in theory. But there is a difference between theory and practice...
A good reference for the practice is RFC 1750.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |