fix(ssa): Do not deduplicate division by a zero constant

[vezenovm]

Description

Problem*

No issue just a bug I found while working on other tasks.

Summary*

If the rhs of a div/mod was zero we would return false from requires_acir_gen_predicate. This would result in deduplicating a div/mod by a zero constant.
e.g. this SSA:

        acir(inline) fn main f0 {
          b0(v0: u32, v1: u32, v2: u1):
            enable_side_effects v2
            v4 = div v1, u32 0
            v5 = not v2
            enable_side_effects v5
            v6 = div v1, u32 0
            return
        }

would turn into this SSA:

        acir(inline) fn main f0 {
          b0(v0: u32, v1: u32, v2: u1):
            enable_side_effects v2
            v4 = div v1, u32 0
            v5 = not v2
            enable_side_effects v5
            return
        }

If v2 was inactive and v5 was active, we would miss needing to fail when processing the second division.

Additional Context

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]

Changes to circuit sizes

Generated at commit: 57e9776b31b7278500bf6457c3ac4cb8151fb192, compared to commit: 2391a8ef05498bac0d7d601c4db79b0621ca3339

🧾 Summary (10% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
regression	-12 ✅	-4.11%	-16 ✅	-0.42%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
regression	280 (-12)	-4.11%	3,811 (-16)	-0.42%

[TomAFrench]

Nice catch

[TomAFrench]

Looks like we're doing a lot of duplicate divisions by constants!