fix(errors): Handle randomness generation and invalid random values as errors in cryptographic code

[oxarbitrage]

Motivation

We want to avoid some panics in some cryptography code. There are some specific problems that were pointed by the audit.

This should close #6338

The code modified here is currently not used but we expect to use it in the future and we will want to avoid panics.

Solution

Most related code use a string as an error with this pattern.
However, this is not always possible, for example if you want to do that here you will get the following compiler error:

inherent associated types are unstable
see issue #8995 <https://github.com/rust-lang/rust/issues/8995> for more information

So i think that limitation might be the cause of not using try_fill_bytes() in the first place.

What is needed is our own errors for handling this and that is what this PR tries to do.

I think all the string errors should be eventually refactored to this but i didnt wanted to do that here as it will be too intrusive so i just fixed what was pointed by the audit ticket.

Review

I think anyone can review, the names chosen could be improved maybe.

Reviewer Checklist

• Will the PR name make sense to users?
  • Does it need extra CHANGELOG info? (new features, breaking changes, large changes)
• Are the PR labels correct?
• Does the code do what the ticket and PR says?
  • Does it change concurrent code, unsafe code, or consensus rules?
• How do you know it works? Does it have tests?

Follow Up Work

Add errors for all string errors used inside zebra-chain ?

[teor2345]

Looks great, thanks for making minimal changes!

I'd just like to check a few cryptographic details.

[codecov]

Codecov Report

Merging #6385 (47a3adb) into main (f2133c9) will increase coverage by 0.04%.
The diff coverage is 0.00%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #6385      +/-   ##
==========================================
+ Coverage   77.92%   77.97%   +0.04%     
==========================================
  Files         306      307       +1     
  Lines       40236    40249      +13     
==========================================
+ Hits        31355    31383      +28     
+ Misses       8881     8866      -15     

[mpguerra]

Looks great, thanks for making minimal changes!

I'd just like to check a few cryptographic details.

I'm not sure we're going to be able to get a cryptography review until after RWC and/or Easter so this may have to move to Sprint 8.

EDIT: Unless @upbqdn is around today or tomorrow?

[upbqdn]

I'll review this.

[upbqdn]

The error names make sense to me, but I'd rather let @dconnolly give it the final approval.

[teor2345]

This is waiting for a final review from Deirdre when she gets back.

[dconnolly]

@oxarbitrage @upbqdn @teor2345 I made some suggestions in ed6596a, lmk what you think?

[teor2345]

Looks good to me, just needs to resolve a merge conflict?

[daira]

I question the idea that we want to avoid a panic when a cryptographic RNG fails. #6338 (comment)