feat: removing redundant key fetching

[benesjan]

Fixes #7954

I replaced the use of encode_and_encrypt_note with encode_and_encrypt_note_with_keys in most of the places as it allowed for reusing the obtained keys. Note that there is only 1 legimate place remaining where it made sense to keep on using encode_and_encrypt_note and that is here. All the other places are either test contracts or the token blacklist contract which is very outdated by now and hence it didn't seem to be worth it to update it.

Given this I think we should nuke encode_and_encrypt_note to keep the API simpler and to make devs write efficient code. Does the reviewer agree? (possibly also encode_and_encrypt_event)

Token::transfer(...) gates before were 49296 and after 38903. Diff of 10393 gates.

[benesjan]

• feat: removing redundant key fetching #8043 👈
• chore: merging TokenWithRefunds with Token #8042
• master

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

Join @benesjan and the rest of your teammates on Graphite

[AztecBot]

Benchmark results

Metrics with a significant change:

• protocol_circuit_simulation_time_in_ms (private-kernel-tail-to-public): 651 (-37%)
• avm_simulation_time_ms (Token:transfer_public): 19.0 (-30%)

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	5,756	1,553	709	762 (+1%)	781 (+1%)
proof_construction_time_sha256_30_ms	11,949 (+1%)	3,182	1,421	1,471 (+2%)	1,505 (+2%)
proof_construction_time_sha256_100_ms	45,489	12,053 (+1%)	5,480 (-1%)	5,453	5,408 (-1%)
proof_construction_time_poseidon_hash_ms	78.0	34.0	34.0	57.0 (-3%)	89.0 (+2%)
proof_construction_time_poseidon_hash_30_ms	1,529	420 (-1%)	202	230 (-1%)	268
proof_construction_time_poseidon_hash_100_ms	5,637	1,510	678	745 (+3%)	757 (+1%)

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	49,720	92,488	177,440
l1_rollup_execution_gas	1,383,214	2,130,950	3,957,941
l2_block_processing_time_in_ms	250 (+1%)	448 (+3%)	827 (+1%)
l2_block_building_time_in_ms	9,033 (+1%)	17,492	34,916 (+1%)
l2_block_rollup_simulation_time_in_ms	9,033 (+1%)	17,491	34,916 (+1%)
l2_block_public_tx_process_time_in_ms	7,590 (+1%)	15,995	33,387 (+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	2,982 (+5%)	3,768 (-2%)
node_database_size_in_bytes	12,640,336	16,732,240
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	83.8 (-6%)	400 (+1%)	21,502	44,858	N/A
private-kernel-inner	163 (+5%)	713	72,180	45,005	N/A
private-kernel-reset-tiny	476 (+1%)	875	65,502	44,844	N/A
private-kernel-tail	209 (+5%)	162	50,606	52,256	N/A
base-parity	5.65 (+2%)	N/A	160	96.0	N/A
root-parity	33.3	N/A	69,084	96.0	N/A
base-rollup	2,749	N/A	187,817	664	N/A
root-rollup	38.9	N/A	54,525	716	N/A
public-kernel-setup	83.8	N/A	103,760	71,222	N/A
public-kernel-app-logic	95.8	N/A	103,599	71,222	N/A
public-kernel-tail	554	N/A	409,190	16,414	N/A
private-kernel-reset-small	462 (-2%)	N/A	66,085	45,629	N/A
private-kernel-tail-to-public	⚠️ 651 (-37%)	650	473,529	1,697	N/A
public-kernel-teardown	82.9	N/A	104,005	71,222	N/A
merge-rollup	19.0 (-1%)	N/A	35,742	664	N/A
undefined	N/A	N/A	N/A	N/A	65,027

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	343
ContractInstanceDeployer:deploy	1,408	11,731	18.2 (-1%)
MultiCallEntrypoint:entrypoint	1,920	11,731	404 (-4%)
FeeJuice:deploy	1,376	11,731	390
SchnorrAccount:constructor	1,312	11,731	74.6 (-29%)
SchnorrAccount:entrypoint	2,304	11,731	413 (-4%)
Token:privately_mint_private_note	1,280	11,731	104 (-26%)
FPC:fee_entrypoint_public	1,344	11,731	29.3 (+9%)
Token:transfer	1,312	11,731	238 (-12%)
Benchmarking:create_note	1,344	11,731	88.2 (-14%)
SchnorrAccount:verify_private_authwit	1,280	11,731	27.3 (-1%)
Token:unshield	1,376	11,731	526 (-6%)
FPC:fee_entrypoint_private	1,376	11,731	716 (-4%)

AVM Simulation

Time to simulate various public functions in the AVM.

Function	time_ms	bytecode_size_in_bytes
FeeJuice:_increase_public_balance	54.1 (+2%)	7,739
FeeJuice:set_portal	11.7 (-1%)	2,354
Token:constructor	83.0 (+4%)	26,525
FPC:constructor	56.3 (+7%)	18,383
FeeJuice:mint_public	38.4 (+2%)	5,877
Token:mint_public	44.1 (+2%)	10,917
Token:assert_minter_and_mint	65.2 (-2%)	7,512
AuthRegistry:set_authorized	35.8 (-21%)	4,391
FPC:prepare_fee	238 (+3%)	7,712
Token:transfer_public	⚠️ 19.0 (-30%)	39,946
FPC:pay_refund	57.3 (-1%)	10,811
Benchmarking:increment_balance	934	6,929
Token:_increase_public_balance	40.1 (+1%)	8,433
FPC:pay_refund_with_shielded_rebate	65.1	11,360

Public DB Access

Time to access various public DBs.

Function	time_ms
get-nullifier-index	0.158 (-4%)

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	2.17 (-1%)	3.87 (-1%)	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 (-1%)	0.109 (-1%)	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	11.7	18.0 (+1%)	31.3 (+2%)	60.1 (+2%)	113 (-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	0.112 (-1%)	0.104 (+1%)	0.101 (+1%)	0.103 (+1%)	0.101
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	14.6 (+1%)	26.0 (+2%)	44.3 (+3%)	82.3	161 (-1%)
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	0.111 (+1%)	0.105 (+2%)	0.107 (+3%)	0.101 (-1%)	0.101 (-2%)
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	16.5 (-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 | |
| - | |

[nventuro]

Token::transfer(...) gates before were 49296 and after 38903. Diff of 10393 gates.

This is aligned with what I'd expect - we're performing two fewer key reads (to and from's npk, which we already have), and each of these is at least one merkle read (4k gates) plus some hashing of all 4 keys.

Given this I think we should nuke encode_and_encrypt_note to keep the API simpler and to make devs write efficient code. Does the reviewer agree? (possibly also encode_and_encrypt_event)

Yeah, I was thinking the same thing. Though I'd keep the short rename by also renaming _with_keys instead of just nuking the short variant - there's no point in saying 'with keys' if we always need the keys. API breakage but that's what we live for.

[nventuro]

Looks great, thanks!