feat: Adding autogenerated variants for the reset circuit

[sirasistant]

This PR adds:

• Autogeneration of kernel reset variants by declaring them in a json file
• Updating of the nargo workspace so the new variants get compiled
• Autogeneration from the same JSON of the circuits.js typings for the variants, and validation of the sizes
• Updates across code to support multiple variants for the reset. However, if you add variants in the json descriptor, you'll still have to manually update noir-protocol-circuits-types/index.ts to add imports for the jsons of the variants (however typescript will cry in the places that you need to modify, thanks to the circuits.js autogeneration)
• Updates the kernel prover to choose the appropiate variant for the hints size and trims the hints

[github-actions]

Changes to circuit sizes

Generated at commit: f1b3cf5704c8d9607755a42c6e1ccc2719b960cf, compared to commit: 07dc0726501bc78d691e1d2360dda84d1a93b9c5

🧾 Summary (100% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
private_kernel_reset	+7,474 ❌	+2.99%	+32,472 ❌	+1.30%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
private_kernel_reset	257,186 (+7,474)	+2.99%	2,527,722 (+32,472)	+1.30%

[AztecBot]

Benchmark results

Metrics with a significant change:

• tx_pxe_processing_time_ms (0): 28,731 (-30%)
• tx_pxe_processing_time_ms (2): 98,693 (-21%)
• app_circuit_proving_time_in_ms (Test:emit_nullifier): 2,864 (+71%)
• app_circuit_proving_time_in_ms (FPC:fee_entrypoint_public): 16,599 (+69%)
• app_circuit_proving_time_in_ms (FPC:fee_entrypoint_private): 8,914 (-43%)
• app_circuit_proving_time_in_ms (Token:unshield): 50,443 (-18%)
• app_circuit_proving_time_in_ms (SchnorrAccount:spend_private_authwit): 2,691 (-34%)
• protocol_circuit_witness_generation_time_in_ms (base-parity): 862 (+250%)
• protocol_circuit_witness_generation_time_in_ms (public-kernel-app-logic): 313 (-20%)
• protocol_circuit_witness_generation_time_in_ms (base-rollup): 5,190 (+17%)
• protocol_circuit_witness_generation_time_in_ms (merge-rollup): 84.8 (-41%)
• protocol_circuit_proving_time_in_ms (private-kernel-inner): 40,489 (-16%)
• protocol_circuit_proving_time_in_ms (public-kernel-setup): 929 (+20%)

Detailed results

All benchmarks are run on txs on the Benchmarking contract on the repository. Each tx consists of a batch call to create_note and increment_balance, which guarantees that each tx has a private call, a nested private call, a public call, and a nested public call, as well as an emitted private note, an unencrypted log, and public storage read and write.

This benchmark source data is available in JSON format on S3 here.

L2 block published to L1

Each column represents the number of txs on an L2 block published to L1.

Metric	8 txs	32 txs	64 txs
l1_rollup_calldata_size_in_bytes	740	740	740
l1_rollup_calldata_gas	6,692	6,680	6,668
l1_rollup_execution_gas	586,877	586,865	586,853
l2_block_processing_time_in_ms	1,622	6,019	11,964
l2_block_building_time_in_ms	35,435 (-1%)	139,286	276,424 (-1%)
l2_block_rollup_simulation_time_in_ms	35,094 (-1%)	138,503	274,921 (-1%)
l2_block_public_tx_process_time_in_ms	16,540 (-2%)	71,104	142,426 (-2%)

L2 chain processing

Each column represents the number of blocks on the L2 chain where each block has 16 txs.

Metric	5 blocks	10 blocks
node_history_sync_time_in_ms	18,081	33,664 (+1%)
node_database_size_in_bytes	21,418,064	38,092,880
pxe_database_size_in_bytes	29,868	59,425

Circuits stats

Stats on running time and I/O sizes collected for every kernel circuit run across all benchmarks.

Circuit	protocol_circuit_simulation_time_in_ms	protocol_circuit_witness_generation_time_in_ms	protocol_circuit_proving_time_in_ms	protocol_circuit_input_size_in_bytes	protocol_circuit_output_size_in_bytes	protocol_circuit_proof_size_in_bytes	protocol_circuit_num_public_inputs	protocol_circuit_size_in_gates
private-kernel-init	163 (+1%)	3,662 (+5%)	22,997 (-8%)	19,955	61,967	86,688	2,642	1,048,576
private-kernel-inner	551	4,804 (+11%)	⚠️ 40,489 (-16%)	88,990	61,967	86,688	2,642	2,097,152
private-kernel-reset-small	569	2,619	23,194	117,865	61,967	86,688	2,642	1,048,576
private-kernel-tail	541	3,974 (+1%)	37,800 (-12%)	86,817	79,422	10,656	266	2,097,152
base-parity	8.23 (+2%)	⚠️ 862 (+250%)	4,286 (+11%)	128	64.0	2,208	2.00	131,072
root-parity	51.1	128 (-16%)	39,831 (-9%)	27,064	64.0	2,720	18.0	2,097,152
base-rollup	921 (+1%)	⚠️ 5,190 (+17%)	43,079 (-10%)	111,126	925	3,104	30.0	2,097,152
root-rollup	72.4 (+2%)	178 (-5%)	8,368 (-7%)	11,454	789	3,424	40.0	524,288
public-kernel-app-logic	197 (-1%)	⚠️ 313 (-20%)	772 (+4%)	96,915	85,063	116,416	3,571	4,096
public-kernel-tail	893	870 (+6%)	1,110 (+11%)	388,175	7,723	10,144	250	512
public-kernel-setup	170 (+2%)	367 (-5%)	⚠️ 929 (+20%)	138,235	85,063	116,416	3,571	4,096
public-kernel-teardown	167 (-3%)	411 (+4%)	930 (+2%)	143,244	85,063	116,416	3,571	4,096
merge-rollup	6.75 (-1%)	⚠️ 84.8 (-41%)	2,204 (+6%)	2,696	925	3,104	30.0	65,536
private-kernel-tail-to-public	N/A	9,524 (+4%)	72,593 (-8%)	N/A	N/A	116,928	3,587	4,194,304

Stats on running time collected for app circuits

Function	app_circuit_proof_size_in_bytes	app_circuit_proving_time_in_ms	app_circuit_size_in_gates	app_circuit_num_public_inputs
SchnorrAccount:entrypoint	16,096	48,009 (-8%)	2,097,152	436
Test:emit_nullifier	16,096	⚠️ 2,864 (+71%)	65,536	436
FPC:fee_entrypoint_public	16,096	⚠️ 16,599 (+69%)	524,288	436
FPC:fee_entrypoint_private	16,096	⚠️ 8,914 (-43%)	524,288	436
Token:unshield	16,096	⚠️ 50,443 (-18%)	2,097,152	436
SchnorrAccount:spend_private_authwit	16,096	⚠️ 2,691 (-34%)	131,072	436
Token:transfer	16,096	34,714 (-14%)	2,097,152	436

Tree insertion stats

The duration to insert a fixed batch of leaves into each tree type.

Metric	1 leaves	16 leaves	64 leaves	128 leaves	512 leaves	1024 leaves	2048 leaves	4096 leaves	32 leaves
batch_insert_into_append_only_tree_16_depth_ms	13.1 (+2%)	21.5 (+2%)	N/A	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_count	16.7	31.8	N/A	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_ms	0.770 (+2%)	0.662 (+2%)	N/A	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	61.6	97.4	317 (+1%)	611	1,189	2,360	N/A
batch_insert_into_append_only_tree_32_depth_hash_count	N/A	N/A	95.9	159	543	1,055	2,079	4,127	N/A
batch_insert_into_append_only_tree_32_depth_hash_ms	N/A	N/A	0.632	0.602	0.576 (+1%)	0.570	0.565	0.565	N/A
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	73.8 (+1%)	144 (+1%)	456 (+1%)	897	1,769	3,515 (-1%)	N/A
batch_insert_into_indexed_tree_20_depth_hash_count	N/A	N/A	106	208	692	1,363	2,707	5,395	N/A
batch_insert_into_indexed_tree_20_depth_hash_ms	N/A	N/A	0.645 (+1%)	0.647 (+1%)	0.621 (+1%)	0.617	0.616	0.614	N/A
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	80.6
batch_insert_into_indexed_tree_40_depth_hash_count	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	108
batch_insert_into_indexed_tree_40_depth_hash_ms	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	0.710

Miscellaneous

Transaction sizes based on how many contract classes are registered in the tx.

Metric	0 registered classes	1 registered classes
tx_size_in_bytes	84,711	673,189

Transaction size based on fee payment method

| Metric | |
| - | |

Transaction processing duration by data writes.

Metric	0 new note hashes	1 new note hashes	2 new note hashes
tx_pxe_processing_time_ms	⚠️ 28,731 (-30%)	3,306 (-12%)	⚠️ 98,693 (-21%)

Metric	0 public data writes	1 public data writes	2 public data writes	4 public data writes	8 public data writes
tx_sequencer_processing_time_ms	1,323 (-11%)	2,291	1,909 (+2%)	2,838 (+2%)	1,817 (+2%)

[LeilaWang]

Amazing! 🚀