fix(SSA): validate MakeArray instruction

[asterite]

Blocked by #8918 as currently the databus puts values of the wrong type in MakeArray.

Description

Problem

Resolves #9158

Summary

With the validation in some multi_scalar_mul simplification tests started to fail. It turned out that some MakeArray types were incorrect there so this PR fixes that too.

Additional Context

While doing this I realized call argument types aren't checked against the target function so I might do that in a separate PR.

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]

ACVM Benchmarks

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
purely_sequential_opcodes	254903 ns/iter (± 1268)	250319 ns/iter (± 888)	1.02
perfectly_parallel_opcodes	223661 ns/iter (± 4685)	219732 ns/iter (± 1535)	1.02
perfectly_parallel_batch_inversion_opcodes	2781475 ns/iter (± 8723)	2781014 ns/iter (± 15137)	1.00

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

[github-actions]

Artifact Size

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	716.6 KB	716.6 KB	1
private-kernel-reset	1972.3 KB	1972.3 KB	1
private-kernel-tail	543 KB	543 KB	1
rollup-base-private	4943 KB	4943 KB	1
rollup-base-public	3958.4 KB	3958.4 KB	1
rollup-block-root-empty	3881.2 KB	3881.2 KB	1
rollup-block-root-single-tx	32668.1 KB	32668.1 KB	1
rollup-block-root	32702.3 KB	32702.3 KB	1
rollup-merge	187.1 KB	187.1 KB	1
rollup-root	398.6 KB	398.6 KB	1
semaphore-depth-10	632.4 KB	632.4 KB	1
sha512-100-bytes	622.9 KB	622.9 KB	1

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

[github-actions]

Compilation Time

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	2.074 s	1.964 s	1.06
private-kernel-reset	8.304 s	8.156 s	1.02
private-kernel-tail	1.63 s	1.612 s	1.01
rollup-base-private	17.56 s	17.86 s	0.98
rollup-base-public	14.76 s	15.02 s	0.98
rollup-block-root-empty	21.26 s	20.84 s	1.02
rollup-block-root-single-tx	198 s	213 s	0.93
rollup-block-root	206 s	199 s	1.04
rollup-merge	1.814 s	1.666 s	1.09
rollup-root	1.816 s	1.928 s	0.94
semaphore-depth-10	0.768 s	0.76 s	1.01
sha512-100-bytes	1.885 s	1.98 s	0.95

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

[github-actions]

Execution Time

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	0.015 s	0.015 s	1
private-kernel-reset	0.154 s	0.152 s	1.01
private-kernel-tail	0.01 s	0.01 s	1
rollup-base-private	0.294 s	0.299 s	0.98
rollup-base-public	0.184 s	0.188 s	0.98
rollup-block-root	13.4 s	13.1 s	1.02
rollup-merge	0.002 s	0.002 s	1
rollup-root	0.007 s	0.004 s	1.75
semaphore-depth-10	0.019 s	0.019 s	1
sha512-100-bytes	0.086 s	0.055 s	1.56

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

[github-actions]

Compilation Memory

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	235.34 MB	235.38 MB	1.00
private-kernel-reset	543.37 MB	543.37 MB	1
private-kernel-tail	209.04 MB	209.04 MB	1
rollup-base-private	1370 MB	1370 MB	1
rollup-base-public	1460 MB	1460 MB	1
rollup-block-root-empty	1080 MB	1080 MB	1
rollup-block-root-single-tx	9380 MB	9380 MB	1
rollup-block-root	9380 MB	9380 MB	1
rollup-merge	326.97 MB	326.97 MB	1
rollup-root	337.38 MB	337.38 MB	1
semaphore_depth_10	104.57 MB	104.57 MB	1
sha512_100_bytes	255.68 MB	255.76 MB	1.00

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

[github-actions]

Execution Memory

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	202.08 MB	202.08 MB	1
private-kernel-reset	238.11 MB	238.11 MB	1
private-kernel-tail	191.3 MB	191.3 MB	1
rollup-base-private	500.49 MB	500.49 MB	1
rollup-base-public	433.35 MB	433.35 MB	1
rollup-block-root	1510 MB	1510 MB	1
rollup-merge	322.9 MB	322.9 MB	1
rollup-root	325.54 MB	325.54 MB	1
semaphore_depth_10	69.3 MB	69.3 MB	1
sha512_100_bytes	55.56 MB	55.56 MB	1

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

[github-actions]

Test Suite Duration

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
test_report_AztecProtocol_aztec-packages_noir-projects_aztec-nr	131 s	134 s	0.98
test_report_AztecProtocol_aztec-packages_noir-projects_noir-contracts	153 s	149 s	1.03
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_blob	177 s	178 s	0.99
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_private-kernel-lib	268 s	267 s	1.00
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_reset-kernel-lib	34 s	34 s	1
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_rollup-lib	692 s	702 s	0.99
test_report_AztecProtocol_aztec-packages_noir-projects_noir-protocol-circuits_crates_types	128 s	128 s	1
test_report_noir-lang_noir-bignum_	379 s	403 s	0.94
test_report_noir-lang_noir_bigcurve_	363 s	320 s	1.13
test_report_noir-lang_sha512_	16 s	16 s	1
test_report_zkpassport_noir-ecdsa_	117 s	108 s	1.08
test_report_zkpassport_noir_rsa_	1 s	1 s	1

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

[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: dd3ef76	Previous: 87aeccf	Ratio
test_report_zkpassport_noir_rsa_	2 s	1 s	2

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

CC: @TomAFrench

[github-actions]

Opcode count

Details

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
private-kernel-inner	15080 opcodes	15080 opcodes	1
private-kernel-reset	68865 opcodes	68865 opcodes	1
private-kernel-tail	11347 opcodes	11347 opcodes	1
rollup-base-private	261916 opcodes	261916 opcodes	1
rollup-base-public	201205 opcodes	201205 opcodes	1
rollup-block-root-empty	70748 opcodes	70748 opcodes	1
rollup-block-root-single-tx	1079444 opcodes	1079444 opcodes	1
rollup-block-root	1080739 opcodes	1080739 opcodes	1
rollup-merge	1419 opcodes	1419 opcodes	1
rollup-root	3136 opcodes	3136 opcodes	1
semaphore-depth-10	5763 opcodes	5763 opcodes	1
sha512-100-bytes	22169 opcodes	22169 opcodes	1

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

[github-actions]

Changes to number of Brillig opcodes executed

Generated at commit: 129f157eec725f857d0d0808aa844ac676192a69, compared to commit: 39a837906d40a728c0413f7ccc703df8e40f9716

🧾 Summary (10% most significant diffs)

Program	Brillig opcodes (+/-)	%
embedded_curve_ops_inliner_max	-28 ✅	-12.56%
embedded_curve_ops_inliner_zero	-28 ✅	-12.56%

---

Full diff report 👇

Program	Brillig opcodes (+/-)	%
embedded_curve_ops_inliner_min	291 (-28)	-8.78%
embedded_curve_ops_inliner_max	195 (-28)	-12.56%
embedded_curve_ops_inliner_zero	195 (-28)	-12.56%

[github-actions]

Changes to Brillig bytecode sizes

Generated at commit: 129f157eec725f857d0d0808aa844ac676192a69, compared to commit: 39a837906d40a728c0413f7ccc703df8e40f9716

🧾 Summary (10% most significant diffs)

Program	Brillig opcodes (+/-)	%
embedded_curve_ops_inliner_max	-142 ✅	-40.00%
embedded_curve_ops_inliner_zero	-142 ✅	-40.00%

---

Full diff report 👇

Program	Brillig opcodes (+/-)	%
embedded_curve_ops_inliner_min	280 (-154)	-35.48%
embedded_curve_ops_inliner_max	213 (-142)	-40.00%
embedded_curve_ops_inliner_zero	213 (-142)	-40.00%

[github-actions]

Changes to circuit sizes

Generated at commit: 129f157eec725f857d0d0808aa844ac676192a69, compared to commit: 39a837906d40a728c0413f7ccc703df8e40f9716

🧾 Summary (10% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
embedded_curve_ops	-38 ✅	-74.51%	-924 ✅	-14.60%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
embedded_curve_ops	13 (-38)	-74.51%	5,404 (-924)	-14.60%

[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: dd3ef76	Previous: 8797651	Ratio
private-kernel-inner	2.458 s	2.024 s	1.21

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

CC: @TomAFrench

[github-actions]

⚠️ Performance Alert ⚠️

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

Benchmark suite	Current: e3dee5c	Previous: 39a8379	Ratio
rollup-root	0.007 s	0.004 s	1.75
sha512-100-bytes	0.086 s	0.055 s	1.56

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

CC: @TomAFrench

[aakoshh]

Great stuff 👏

[aakoshh]

Are the circuit size changes expected?

[asterite]

Are the circuit size changes expected?

Good question. I hope so! 😄

[asterite]

Are the circuit size changes expected?

The multi_scalar_multiplication simplification was returning incorrect types. For example instead of [(Field, Field, u1); N] it was returning [Field; 6] which I guess later on has some different implications compared to the correct type. But... I'll compare the SSAs to see what changed to understand this more.

[asterite]

I just checked the SSA difference and it seems now we can reuse existing arrays. Previously they would get an incorrect type after simplification so they didn't match previous arrays. Or at least that's one difference...

[asterite]

@TomAFrench Could you check if this change is okay? Mainly to make sure the performance improvements are correct and that I didn't break anything 😅

[vezenovm]

I looked over the improvements and as you mentioned they look to be happening as simplifications revealed themselves once we accurately marked the array types. When looking at things to optimize in the past I was always thinking the embedded_curve_add was more expensive than expected. I didn't investigate too deeply where the exact simplification occurs that has knockdown benefits, but the types you marked look correct.