Re: [PATCH] (0/4) Entropy accounting fixes

[Linus Torvalds]

On 17 Aug 2002, Robert Love wrote:
> 
> [1] this is why I wrote my netdev-random patches.  some machines just
>     have to take the entropy from the network card... there is nothing
>     else.

I suspect that Oliver is 100% correct in that the current code is just
_too_ trusting. And parts of his patches seem to be in the "obviously
good" category (ie the xor'ing of the buffers instead of overwriting).

So I think that if we just made the code be much less trusting (say, 
consider the TSC information per interrupt to give only a single bit of 
entropy, for example), and coupled that with making network devices always 
be considered sources of entropy, we'd have a reasonable balance. 

		Linus