Crossprop tests for bitwise and blockinfo

[csgui]

Initial pull request with some crossprop tests:

• bitwise
• blockinfo

I would like to validate the approach in this PR to make it the base for other implementations.

[codecov]

Codecov Report

Attention: 8 lines in your changes are missing coverage. Please review.

Comparison is base (c0d3f3c) 91.38% compared to head (97b971e) 91.45%.
Report is 8 commits behind head on main.

Files	Patch %	Lines
clar2wasm/src/tools.rs	55.55%	4 Missing and 4 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #309      +/-   ##
==========================================
+ Coverage   91.38%   91.45%   +0.06%     
==========================================
  Files          40       40              
  Lines       13803    13936     +133     
  Branches    13803    13936     +133     
==========================================
+ Hits        12614    12745     +131     
+ Misses        558      545      -13     
- Partials      631      646      +15     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Acaccia]

I don't really have anything else to add. The other issue with crosscheck with env is already noted.

[csgui]

Tests are failing randomly when checking the expression (get-block-info? time uXX). The difference in the results are always by 1. Seems a false positive.

Any take on that @Acaccia @obycode ?

Test failed: assertion `left == right` failed: Compiled and interpreted results diverge! (get-block-info? time u5)
compiled: Ok(Some(Optional(OptionalData { data: Some(UInt(1706712683)) })))
interpreted: Ok(Some(Optional(OptionalData { data: Some(UInt(1706712684)) })))

[Acaccia]

@csgui I don't know why this test gives you this result, but it would be nice to get the explanation before merging this PR. Because for now, I have no idea if it's the test that is wrong, or if there is a bug in the code.

[csgui]

@Acaccia Yep, that is the reason I have opened this discussion. Maybe more pair of 👀 can bring some light to this issue.

[obycode]

In the simulation, when the datastore is created, the time is saved, then each block is assigned a time that is this genesis time plus height * some constant. This means that if the creation times of the two datastores (one for the interpreter and one for clarity-wasm) are different, then this value will be different, in this case by 1 second.

[Acaccia]

two datastores (one for the interpreter and one for clarity-wasm) are different

That seems like a serious bug to me. Not for the code concerned by this PR but for all the crosscheck tests in general. I'm not sure we can rely on the compared results if both have different initial values for the datastore.

Can't we create only one at the beginning and clone it?

[csgui]

🤔 ... fair point @Acaccia . We are relying on different TestEnvironment instances for both interpreted and compiled results.

Seems that we should tackle that.

[Acaccia]

Since this is supposed to be the introduction to crossprop, would it be possible to add documentation to each crosscheck functions?
That would be helpful for newcomers.

[Acaccia]

LGTM!

[obycode]

Thanks @csgui!