feat: refactor shared mutable to include private and unconstrained getters

[sklppy88]

Please read contributing guidelines and remove this line.

[sklppy88]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• feat: refactoring key getters #7487
• feat: add unconstrained context to key getters #7486
• feat: refactor shared mutable to include private and unconstrained getters #7465 👈
• feat: add unconstrained getters to sharedmutable #7429
• feat: add unconstrained context to txe #7448
• master

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

Join @sklppy88 and the rest of your teammates on Graphite

[AztecBot]

Benchmark results

Metrics with a significant change:

• protocol_circuit_proving_time_in_ms (root-parity): 30,102 (-40%)
• protocol_circuit_proving_time_in_ms (public-kernel-setup): 17,517 (-21%)
• protocol_circuit_proving_time_in_ms (public-kernel-teardown): 17,994 (-21%)
• protocol_circuit_proving_time_in_ms (root-rollup): 28,104 (-31%)

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	5,771 (+7%)	1,554 (+5%)	708 (+6%)	748 (+3%)	770 (+6%)
proof_construction_time_sha256_30_ms	11,876 (+9%)	3,184 (+7%)	1,410 (+8%)	1,449 (+6%)	1,473 (+6%)
proof_construction_time_sha256_100_ms	44,146 (+6%)	11,926 (+5%)	5,433 (+7%)	5,397 (-5%)	5,377 (+6%)
proof_construction_time_poseidon_hash_ms	78.0 (+4%)	34.0	34.0	57.0	88.0 (-1%)
proof_construction_time_poseidon_hash_30_ms	1,528 (+6%)	417 (+4%)	201 (+4%)	219 (+2%)	264 (+1%)
proof_construction_time_poseidon_hash_100_ms	5,659 (+6%)	1,525 (+4%)	674 (+5%)	714 (+4%)	750 (+6%)

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	708	708	708
l1_rollup_calldata_gas	6,588	6,594	6,600
l1_rollup_execution_gas	583,050	583,207	583,364
l2_block_processing_time_in_ms	758 (-6%)	1,427 (-5%)	2,689 (-7%)
l2_block_building_time_in_ms	13,157 (-5%)	26,073 (-3%)	50,642 (-5%)
l2_block_rollup_simulation_time_in_ms	13,156 (-5%)	26,072 (-3%)	50,641 (-5%)
l2_block_public_tx_process_time_in_ms	11,026 (-5%)	23,720 (-3%)	48,269 (-5%)

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	7,103 (-5%)	10,068 (-4%)
node_database_size_in_bytes	12,468,304	16,441,424
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	proof_size_in_bytes	num_public_inputs	size_in_gates
private-kernel-init	116 (-4%)	403 (+2%)	21,732	54,085	N/A	N/A	N/A	N/A
private-kernel-inner	229 (-3%)	701 (-1%)	81,199	54,264	N/A	N/A	N/A	N/A
private-kernel-tail	1,127 (-1%)	2,488 (-1%)	61,808	61,554	N/A	N/A	N/A	N/A
base-parity	6.42 (-5%)	811 (-6%)	160	96.0	1,965 (-2%)	12,676	3.00	131,072
root-parity	107 (-4%)	116 (-1%)	64,860	96.0	⚠️ 30,102 (-40%)	12,676	3.00	4,194,304
base-rollup	4,222 (-3%)	4,765 (-3%)	180,315	632	42,911 (-12%)	13,476	28.0	4,194,304
root-rollup	151 (-6%)	122 (+1%)	51,261	652	⚠️ 28,104 (-31%)	13,412	26.0	4,194,304
public-kernel-setup	199 (-2%)	2,328 (-4%)	112,536	80,310	⚠️ 17,517 (-21%)	118,916	3,323	2,097,152
public-kernel-app-logic	150 (-3%)	3,206 (-2%)	112,536	80,310	10,131 (+1%)	118,916	3,323	1,048,576
public-kernel-tail	894 (-3%)	25,360 (-7%)	409,413	10,046	64,903 (-11%)	24,868	384	8,388,608
private-kernel-reset-tiny	298 (-4%)	987 (-1%)	77,007	53,954	N/A	N/A	N/A	N/A
private-kernel-tail-to-public	6,715 (-1%)	2,831	777,415	1,632	N/A	N/A	N/A	N/A
public-kernel-teardown	139 (-4%)	3,207 (-2%)	112,536	80,310	⚠️ 17,994 (-21%)	118,916	3,323	2,097,152
merge-rollup	57.4 (-6%)	N/A	33,566	632	N/A	N/A	N/A	N/A
undefined	N/A	N/A	N/A	N/A	164,222 (+2%)	N/A	N/A	N/A

Stats on running time collected for app circuits

Function	input_size_in_bytes	output_size_in_bytes	witness_generation_time_in_ms	proof_size_in_bytes	proving_time_in_ms
ContractClassRegisterer:register	1,312	9,344	396 (-3%)	N/A	N/A
ContractInstanceDeployer:deploy	1,376	9,344	24.8 (-5%)	N/A	N/A
MultiCallEntrypoint:entrypoint	1,888	9,344	624 (-5%)	N/A	N/A
GasToken:deploy	1,344	9,344	570 (-3%)	N/A	N/A
SchnorrAccount:constructor	1,280	9,344	424 (-8%)	N/A	N/A
SchnorrAccount:entrypoint	2,272	9,344	757 (-7%)	N/A	N/A
Token:privately_mint_private_note	1,248	9,344	490 (-7%)	N/A	N/A
FPC:fee_entrypoint_public	1,312	9,344	100 (-3%)	N/A	N/A
Token:transfer	1,280	9,344	1,454 (-6%)	N/A	N/A
AuthRegistry:set_authorized (avm)	19,222	N/A	N/A	95,392	2,041 (-4%)
FPC:prepare_fee (avm)	26,664	N/A	N/A	95,456	2,763 (-2%)
Token:transfer_public (avm)	42,914	N/A	N/A	95,456	3,132 (-2%)
AuthRegistry:consume (avm)	33,100	N/A	N/A	95,424	2,691 (-4%)
FPC:pay_refund (avm)	36,829	N/A	N/A	95,424	10,707 (-3%)
Benchmarking:create_note	1,312	9,344	422 (-10%)	N/A	N/A
SchnorrAccount:verify_private_authwit	1,248	9,344	41.5 (-10%)	N/A	N/A
Token:unshield	1,344	9,344	1,192 (-7%)	N/A	N/A
FPC:fee_entrypoint_private	1,344	9,344	1,548 (-7%)	N/A	N/A

AVM Simulation

Time to simulate various public functions in the AVM.

Function	time_ms	bytecode_size_in_bytes
GasToken:_increase_public_balance	97.6 (-2%)	13,790
GasToken:set_portal	13.2 (+2%)	3,339
Token:constructor	120 (-7%)	23,692
FPC:constructor	91.7 (+1%)	13,592
GasToken:mint_public	80.2 (-3%)	10,158
Token:mint_public	330 (-5%)	19,034
Token:assert_minter_and_mint	48.5 (-1%)	12,925
AuthRegistry:set_authorized	45.0 (-7%)	7,812
FPC:prepare_fee	107 (-12%)	15,062
Token:transfer_public	45.9 (+6%)	31,218
FPC:pay_refund	130 (+2%)	25,260
Benchmarking:increment_balance	1,343 (-4%)	15,267
Token:_increase_public_balance	14.9 (-19%)	15,006
FPC:pay_refund_with_shielded_rebate	183 (-9%)	26,347

Public DB Access

Time to access various public DBs.

Function	time_ms
get-nullifier-index	0.159 (+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	10.3 (-4%)	16.7 (-4%)	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	0.597 (-5%)	0.512 (-5%)	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	47.9 (-5%)	75.8 (-6%)	130 (-7%)	247 (-5%)	469 (-7%)
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	0.490 (-5%)	0.466 (-6%)	0.448 (-7%)	0.447 (-6%)	0.438 (-7%)
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	59.3 (-5%)	113 (-5%)	182 (-7%)	356 (-5%)	690 (-7%)
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	0.501 (-5%)	0.506 (-5%)	0.477 (-7%)	0.481 (-5%)	0.474 (-7%)
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	72.7 (-5%)	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_count	N/A	N/A	133	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_ms	N/A	N/A	0.517 (-5%)	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	76,837	667,838

Transaction size based on fee payment method

| Metric | |
| - | |

[nventuro]

I'm not sure I'm a fan of the approach of methods for external reads, since it deviates quite a bit from other state variables, and will likely be quite dangerous. Raw acess to external contract state is not something we want to encourage: key rotation is an exception due to the massive performance impact, plus clear semantics around keys. But those kind of usecases are I think better served by helper functions, not core methods.

In my mind the largest issue with methods is the redundancy in the context object: we've modeled storage access in some execution, but then break that model by not using some of the context's values (in this case the address). The notion of 'my storage includes a shared mutable state variable, which I can read and write, but I can also use to read/write external variables as long as they look exactly like it' seems bizarre.

For the concrete SharedMutablePrivateGetter issue, I think I'd much rather have the struct be removed and replaced with a free-standing function, instead of trying to bundle it in the basic SharedMutable. But even then we're likely not going to use SharedMutable for key rotation (the major use case for the private getter), so we might as well not worry about this issue for the time being, and come back to it once our plans have better solidified.