fix(acir): Do not read empty array under inactive predicates for various slice ops

[vezenovm]

Description

Problem

Resolves https://github.com/noir-lang/noir/security/advisories/GHSA-mvh4-wv69-m9xh

Summary

I decided to implement a zero length check in ACIR gen for SlicePopFront, SlicePopBack, SliceRemove, and SliceInsert.

As per the comments under ssa_gen/mod.rs:

In ACIR we do have protection against reading empty slices (it returns "Index Out of Bounds"), so we don't get invalid reads. 
The memory operations in ACIR ignore the side effect variables, so even if we added a constraint here, it could still fail when it inevitably tries to read from an empty slice anyway. 
We have to handle that by removing operations which are known to fail and replace them with conditional constraints that do take the side effect into account.

In order to have a defense in depth we should defend against empty slices in ACIR gen. ACIR gen does not want to rely on SSA gen checks to prevent ACIR memory read failures. Even so it has already been noted that ACIR memory ops ignore side effects variables by design so we need to prevent those operations which we know will fail in ACIR gen.

Changes:

• Use existing has_zero_length to check if the slice block is empty after initializing the dynamic slice contents array
• This checks the actual ACIR value representation
• Works even when dynamic length is unknown
• If zero length we return dummy values and assert that the predicate is zero to ensure safety
• Removed the is_constant_zero_length check from fix(acir-gen): Use the side effect variable in slice_pop_back #10455. The assertion was incorrect and needed to use the predicate as well (we just use the check included in this PR now). It is unlikely, but if we had a slice with a length known to be a constant zero under an invalid predicate this would have triggered unnecessary failures
• Tests under execution_success and execution_failure as well as a couple ACIR gen unit test updates

Additional Context

User Documentation

Check one:

• No user documentation needed.
• Changes in docs/ included in this PR.
• [For Experimental Features] Changes in docs/ 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 'Compilation Time'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.20.

Benchmark suite	Current: b822cd0	Previous: 1eeaba1	Ratio
rollup-block-root	1.97 s	1.46 s	1.35

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

CC: @TomAFrench