x.crypto.ascon: improves the core of Ascon permutation routine

[blackshirt]

This PR improves the core of x.crypto.ascon permutation routine in the mean of internal routine ascon_pnr function.
After some profiling, the most widely called routines was 3 function of the core permutation routines, ie, ascon_pnr, ascon_perm and utility ascon_rotate_right used in permutation.
The number before patch using sum256 as a test case:

== Before patch ==
Benchmarking ascon.sum256 before patch ...
Average time ascon.sum256 before patch: 11 µs

And here the number after patch

== After patch ==
Benchmarking ascon.sum256 after patch ...
Average time ascon.sum256 after patch: 6 µs

Nearly 50% improvement in performance was a good addition. The numbers i talk on Discord confirm this approach.

This patch fundamentally unifies ascon_pnr and ascon_perm into single one routine. Its also embed directly bits rotate right logic onto this single routine and improves all performances.
This patch also contains some small cleans up, updates readme file. and other small thing.

[huly-for-github]

Connected to Huly®: V_0.6-25145

[spytheman]

You can add the code that you run for the benches in the bench/crypto folder or in vlib/x/crypto/ascon/bench/ . That way it can be run by everyone, interested in changing the module in the future.

[blackshirt]

You can add the code that you run for the benches in the bench/crypto folder or in vlib/x/crypto/ascon/bench/ . That way it can be run by everyone, interested in changing the module in the future.

Okey .. i would add it in next time of cycles.Its on plan

[spytheman]

Okey .. i would add it in next time of cycles.Its on plan

Thank you 🙇🏻 .