fix: push definition trait constraints after trait item constraint

[asterite]

Description

Problem

Resolves #9349

Summary

I still don't fully understand how everything is tied together, but this comment from @jfecher made me wonder if it's indeed an ordering issue.

Inside type_check_variable_with_bindings we potentially push these constraints:

1. a constraint for the trait method itself
2. constraints for any where clause on the method

Before this PR we were pushing 2 before 1. There was also a comment about why 1 needs to come before 2, though since that comment was written the select_impl argument was added so I think that comment became obsolete at that point (since the constraints for the where clause won't lead to selecting a different impl).

Specifically, for the concrete example in the test we got that the type of the BigCurve::new call is:

fn() -> '7

Then we have this constraint for point 1:

'7: BigCurve<'6>

and this one for point 2:

'6: BigNum

We were checking point 2 before point 1, and it was failing because '6 was unbound.

When we flip the order of operations, when we solve '7: BigCurve<'6> we get these type bindings applied:

{
    '6: (
        '6,
        Normal,
        B'4,
    ),
}

At that point '6 is B'4 (which I guess has an assumed impl for BigNum, according to the function signature), then the constraint '6: BigNum will pass.

Additional Context

Thank you @jfecher ! I merely tried flipping the order here, the debugged things a bit more for the above explanation (it was hard to see that the bindings produced by verify_trait_constraint would make the second check work because when it's flipped we don't get to check the other one). The conclusion is that you solved it, I merely sent the code change 😅

(well, we still have to see if CI passes... I do wonder if flipping the order would cause a problem in other scenarios)

Documentation

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Test Suite Duration'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: 22ae2aa	Previous: ccbea58	Ratio
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_blob	3 s	2 s	1.50
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_private-kernel-lib	2 s	1 s	2

This comment was automatically generated by workflow using github-action-benchmark.

CC: @TomAFrench

[jfecher]

Looks good! Let's just add a comment to help protect this against further reorderings