Could this be applied to random(4)?

Could this be applied to random(4)?

[Sandy Harris]

A theoretical paper on getting provably excellent randomness from two
relatively weak input sources.
https://www.sciencenews.org/article/new-technique-produces-real-randomness

Re: Could this be applied to random(4)?

[Stephan Mueller]

Am Freitag, 27. Mai 2016, 13:38:05 schrieb Sandy Harris:

Hi Sandy,

> A theoretical paper on getting provably excellent randomness from two
> relatively weak input sources.
> https://www.sciencenews.org/article/new-technique-produces-real-randomness

This document describes extractors. Those extractors are intended to combine 
*independent* sources with weak entropy.

None of our sources we have in add_*_randomness are independent.

In addition, I am not sure why this research is so hyped. I think that the 3-
source extractor described in [1] is more efficient and easier to implement

[1] "Extracting Randomness Using Few Independent Sources" by Boaz Barak, 
Russell Impagliazzo, Avi Wigderson

Ciao
Stephan

Re: Could this be applied to random(4)?

[Sandy Harris]

On Fri, May 27, 2016 at 2:30 PM, Stephan Mueller <smueller@chronox.de> wrote:

> This document describes extractors. Those extractors are intended to combine
> *independent* sources with weak entropy.
>
> None of our sources we have in add_*_randomness are independent.

No, but it would be easy to get two independent sources, interrupts
and some sort of timer jitter thing as in my maxwell, your jitter
driver, havege, ...

Re: Could this be applied to random(4)?

[Stephan Mueller]

Am Freitag, 27. Mai 2016, 16:09:53 schrieb Sandy Harris:

Hi Sandy,

> On Fri, May 27, 2016 at 2:30 PM, Stephan Mueller <smueller@chronox.de> 
wrote:
> > This document describes extractors. Those extractors are intended to
> > combine *independent* sources with weak entropy.
> > 
> > None of our sources we have in add_*_randomness are independent.
> 
> No, but it would be easy to get two independent sources, interrupts
> and some sort of timer jitter thing as in my maxwell, your jitter
> driver, havege, ...

I think I would concur with you here. From my LRNG code:

static int lrng_pdrbg_seed_internal(u8 *outbuf, u32 outbuflen, bool 
fullentropy,
                                    bool drain)
{
...
        /*
         * Concatenate the output of the noise sources. This would be the
         * spot to add an entropy extractor logic if desired.
...

> --
> To unsubscribe from this list: send the line "unsubscribe linux-crypto" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


Ciao
Stephan