feat(avm): more no fake rows + virtual dyn gas (part 1)

[fcarreiro]

• Fixed more things to have 1 row per opcode
• Introduce separate virtual gas gadget with columns for dynamic gas

Some things still have to be constrained/fixed but I want to get this PR in because it was getting big, and this was the first point where it compiled and tests passed.

[fcarreiro]

• feat(avm): more no fake rows + virtual dyn gas (part 1) #7942 👈
• master

This stack of pull requests is managed by Graphite. Learn more about stacking.

Join @fcarreiro and the rest of your teammates on Graphite

[AztecBot]

Benchmark results

Metrics with a significant change:

• avm_simulation_time_ms (Token:transfer_public): 22.9 (-37%)

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.

Proof generation

Each column represents the number of threads used in proof generation.

Metric	1 threads	4 threads	16 threads	32 threads	64 threads
proof_construction_time_sha256_ms	5,802 (+1%)	1,558	713 (+3%)	745 (-1%)	772
proof_construction_time_sha256_30_ms	11,953 (+1%)	3,216 (+2%)	1,431 (+2%)	1,460 (+2%)	1,471
proof_construction_time_sha256_100_ms	45,758 (+3%)	12,138 (-2%)	5,515 (-3%)	5,442 (-2%)	5,399 (-2%)
proof_construction_time_poseidon_hash_ms	78.0 (-1%)	34.0	34.0	59.0 (+2%)	89.0 (+1%)
proof_construction_time_poseidon_hash_30_ms	1,530	421 (+1%)	203 (+1%)	223 (-2%)	267 (+2%)
proof_construction_time_poseidon_hash_100_ms	5,648	1,520 (-2%)	680	715 (-2%)	753

L2 block published to L1

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

Metric	4 txs	8 txs	16 txs
l1_rollup_calldata_size_in_bytes	4,324	7,844	14,852
l1_rollup_calldata_gas	49,744	92,482	177,740
l1_rollup_execution_gas	1,294,407	2,042,113	3,869,410
l2_block_processing_time_in_ms	257	437 (-4%)	857 (+5%)
l2_block_building_time_in_ms	9,289 (-1%)	18,113 (-1%)	36,383 (-1%)
l2_block_rollup_simulation_time_in_ms	9,288 (-1%)	18,112 (-1%)	36,383 (-1%)
l2_block_public_tx_process_time_in_ms	7,857 (-1%)	16,552 (-1%)	34,750 (-1%)

L2 chain processing

Each column represents the number of blocks on the L2 chain where each block has 8 txs.

Metric	3 blocks	5 blocks
node_history_sync_time_in_ms	2,919 (-4%)	3,764 (-5%)
node_database_size_in_bytes	12,722,256	16,826,448
pxe_database_size_in_bytes	16,254	26,813

Circuits stats

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

Circuit	simulation_time_in_ms	witness_generation_time_in_ms	input_size_in_bytes	output_size_in_bytes	proving_time_in_ms
private-kernel-init	101 (+5%)	394 (+1%)	21,846	44,858	N/A
private-kernel-inner	169 (-2%)	702 (+1%)	72,545	45,005	N/A
private-kernel-reset-tiny	511 (+2%)	1,194	69,614	44,844	N/A
private-kernel-tail	272	196 (+2%)	50,760	52,256	N/A
base-parity	5.61	N/A	160	96.0	N/A
root-parity	33.5 (+1%)	N/A	69,084	96.0	N/A
base-rollup	2,858 (-1%)	N/A	187,817	664	N/A
root-rollup	38.4 (-1%)	N/A	54,525	716	N/A
public-kernel-setup	101 (+2%)	N/A	103,760	71,222	N/A
public-kernel-app-logic	106	N/A	103,599	71,222	N/A
public-kernel-tail	579 (+1%)	N/A	409,190	16,414	N/A
private-kernel-reset-small	482	N/A	66,533	45,629	N/A
private-kernel-tail-to-public	1,491 (+13%)	712 (+1%)	507,554	1,697	N/A
public-kernel-teardown	98.1 (+4%)	N/A	104,005	71,222	N/A
merge-rollup	19.1 (-1%)	N/A	35,742	664	N/A
undefined	N/A	N/A	N/A	N/A	67,609 (-1%)

Stats on running time collected for app circuits

Function	input_size_in_bytes	output_size_in_bytes	witness_generation_time_in_ms
ContractClassRegisterer:register	1,344	11,731	347 (+1%)
ContractInstanceDeployer:deploy	1,408	11,731	18.1 (-1%)
MultiCallEntrypoint:entrypoint	1,920	11,731	458 (-1%)
FeeJuice:deploy	1,376	11,731	412 (+6%)
SchnorrAccount:constructor	1,312	11,731	171
SchnorrAccount:entrypoint	2,304	11,731	512 (+1%)
Token:privately_mint_private_note	1,280	11,731	236 (+9%)
FPC:fee_entrypoint_public	1,344	11,731	22.3 (-3%)
Token:transfer	1,312	11,731	404
Benchmarking:create_note	1,344	11,731	167
SchnorrAccount:verify_private_authwit	1,280	11,731	27.6
Token:unshield	1,376	11,731	628 (-1%)
FPC:fee_entrypoint_private	1,376	11,731	842 (-1%)

AVM Simulation

Time to simulate various public functions in the AVM.

Function	time_ms	bytecode_size_in_bytes
FeeJuice:_increase_public_balance	53.9	8,139
FeeJuice:set_portal	9.61 (+4%)	2,362
Token:constructor	81.7 (+1%)	31,107
FPC:constructor	53.3 (-2%)	22,380
FeeJuice:mint_public	38.1 (+1%)	6,150
Token:mint_public	73.9 (+3%)	11,720
Token:assert_minter_and_mint	51.6 (+23%)	8,028
AuthRegistry:set_authorized	40.0 (-8%)	4,537
FPC:prepare_fee	240 (+2%)	8,812
Token:transfer_public	⚠️ 22.9 (-37%)	47,374
FPC:pay_refund	56.4 (-21%)	12,114
Benchmarking:increment_balance	971 (-1%)	7,450
Token:_increase_public_balance	39.2 (-6%)	8,960
FPC:pay_refund_with_shielded_rebate	67.0	12,663

Public DB Access

Time to access various public DBs.

Function	time_ms
get-nullifier-index	0.163 (-1%)

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	256 leaves	512 leaves	1024 leaves
batch_insert_into_append_only_tree_16_depth_ms	2.18 (-1%)	3.85	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_count	16.8	31.7	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_ms	0.113	0.109	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	11.4 (-2%)	17.7 (-3%)	31.2 (-4%)	58.1 (-3%)	115 (+1%)
batch_insert_into_append_only_tree_32_depth_hash_count	N/A	N/A	95.9	159	287	543	1,055
batch_insert_into_append_only_tree_32_depth_hash_ms	N/A	N/A	0.110 (-2%)	0.103 (-3%)	0.102 (-3%)	0.100 (-3%)	0.103 (+3%)
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	14.2 (-2%)	25.9 (-1%)	44.0	81.5 (-4%)	164 (+3%)
batch_insert_into_indexed_tree_20_depth_hash_count	N/A	N/A	109	207	355	691	1,363
batch_insert_into_indexed_tree_20_depth_hash_ms	N/A	N/A	0.108 (-2%)	0.105 (-1%)	0.107	0.101 (-4%)	0.103 (+2%)
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	16.3 (-1%)	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_count	N/A	N/A	132	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_ms	N/A	N/A	0.105 (-1%)	N/A	N/A	N/A	N/A

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	64,779	668,997

Transaction size based on fee payment method

| Metric | |
| - | |

[jeanmon]

Good job @fcarreiro ! Please address my comments before merging.

[jeanmon]

We are excluding the extra last row to constrain l2_gas_remaining. By definition, l2_gas_remaining is the gas quantity remaining before the execution of the pertaining row. This means that to get the gas_remaining quantity after the whole execution, we need access to gas_remaining set in the "extra row". However, the current "CUR_AND_NEXT_ARE_MAIN" create an exception and the prover can put whatever value in the extra row. I might be missing something, but please analyze whether the above remark is valid.

[fcarreiro]

You are right! I added some comments, and changed to use sel_execution_row instead. This was making the gas test fail, but after further inspection it was because we were running out of gas.

[fcarreiro]

Merge activity

• Aug 14, 7:19 AM EDT: @fcarreiro merged this pull request with Graphite.