feat: TXE 2: Electric boogaloo

[Thunkar]

Adds new capabilities to TXE, and a proper TokenContract suite to showcase them:

• Private -> Public calls
• Public -> Public calls
• Actual account contract deployment for complex interactions
• Auhtwits, both public and private
• Note cache storage
• Speed improvements

The full test suite is comparable to the e2e one (with some caveats, some things just cannot be tested yet using TXE), but runs significantly faster. Interestingly, runs faster on an M3 macbook than on mainframe (120s vs. 240s).

e2e tests take more than 600s (without stored snapshots), but the gap increases as we reduce the amount of tests launched simultaneously. An unconstrained private transfer runs in less than 10s using TXE!

If launching many tests in parallel, it is recommended to increase the nargo timeout for foreign calls to avoid timeout errors: NARGO_FOREIGN_CALL_TIMEOUT=300000

Closes #7086

[AztecBot]

Benchmark results

Metrics with a significant change:

• avm_simulation_time_ms (Token:mint_public): 70.6 (-88%)
• avm_simulation_time_ms (Token:assert_minter_and_mint): 211 (+187%)
• protocol_circuit_witness_generation_time_in_ms (private-kernel-inner): 1,123 (+17%)

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_30_ms	11,689	3,137 (-2%)	1,695 (-4%)	1,461 (-11%)	1,572 (+8%)
proof_construction_time_sha256_100_ms	44,758	11,743	5,439	5,428	5,367
proof_construction_time_poseidon_hash_ms	78.0	34.0	34.0	57.0 (-2%)	89.0
proof_construction_time_poseidon_hash_30_ms	1,513	415	199	221 (-3%)	270 (+1%)
proof_construction_time_poseidon_hash_100_ms	5,725	1,574 (-1%)	735 (+1%)	770 (+1%)	792

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	1,412	1,412	1,412
l1_rollup_calldata_gas	9,476	9,472	9,476
l1_rollup_execution_gas	610,297	610,293	610,297
l2_block_processing_time_in_ms	751	1,400	2,673 (-1%)
l2_block_building_time_in_ms	25,398 (-1%)	51,420	99,087 (-2%)
l2_block_rollup_simulation_time_in_ms	25,302 (-2%)	51,111	98,714 (-2%)
l2_block_public_tx_process_time_in_ms	21,734 (-2%)	47,314	95,021 (-2%)

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	6,900 (-2%)	9,818
node_database_size_in_bytes	12,202,064	16,085,072
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	proving_time_in_ms	input_size_in_bytes	output_size_in_bytes	proof_size_in_bytes	num_public_inputs	size_in_gates
private-kernel-init	131 (+2%)	470	13,202 (-3%)	20,634	67,190	92,352	2,819	524,288
private-kernel-inner	396 (+1%)	⚠️ 1,123 (+17%)	50,174 (-2%)	94,902	67,190	92,352	2,819	2,097,152
private-kernel-tail	372	1,824 (-2%)	51,546 (-4%)	99,121	71,733	14,912	399	2,097,152
base-parity	6.25	1,894 (-3%)	2,729	128	64.0	2,208	2.00	131,072
root-parity	49.1	72.9 (-5%)	39,909 (-2%)	27,100	64.0	2,720	18.0	2,097,152
base-rollup	7,915 (-1%)	4,974 (+1%)	85,491 (-2%)	170,330	756	3,648	47.0	4,194,304
root-rollup	111 (+2%)	85.1 (-5%)	23,411 (+4%)	25,309	620	3,456	41.0	1,048,576
public-kernel-setup	718 (+1%)	3,756	45,316	116,905	93,334	125,344	3,850	2,097,152
public-kernel-app-logic	611	4,777	47,095 (-1%)	116,905	93,334	125,344	3,850	2,097,152
public-kernel-tail	1,417	37,856 (+8%)	192,473	511,910	10,014	14,912	399	8,388,608
private-kernel-reset-small	593	2,009 (-1%)	46,612 (-1%)	123,313	67,190	92,352	2,819	2,097,152
public-kernel-teardown	616 (+1%)	4,784 (+2%)	47,266	116,905	93,334	125,344	3,850	2,097,152
merge-rollup	29.2 (+1%)	N/A	N/A	16,542	756	N/A	N/A	N/A
private-kernel-tail-to-public	N/A	6,893 (-1%)	100,088 (-2%)	N/A	N/A	125,344	3,850	4,194,304

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	size_in_gates	num_public_inputs
ContractClassRegisterer:register	1,344	9,944	424	N/A	N/A	N/A	N/A
ContractInstanceDeployer:deploy	1,408	9,944	40.1 (+2%)	N/A	N/A	N/A	N/A
MultiCallEntrypoint:entrypoint	1,920	9,944	1,786	N/A	N/A	N/A	N/A
GasToken:deploy	1,376	9,944	988 (+2%)	N/A	N/A	N/A	N/A
SchnorrAccount:constructor	1,312	9,944	1,415	N/A	N/A	N/A	N/A
SchnorrAccount:entrypoint	2,304	9,944	2,848 (+1%)	16,768	57,530 (+4%)	2,097,152	457
Token:privately_mint_private_note	1,280	9,944	1,639	N/A	N/A	N/A	N/A
FPC:fee_entrypoint_public	1,344	9,944	358 (+4%)	16,768	10,985 (-2%)	524,288	457
Token:transfer	1,312	9,944	4,522 (+3%)	16,768	46,765	2,097,152	457
AuthRegistry:set_authorized (avm)	21,043	N/A	N/A	87,200	1,385 (+2%)	N/A	N/A
FPC:prepare_fee (avm)	28,841	N/A	N/A	88,032	5,871	N/A	N/A
Token:transfer_public (avm)	44,971	N/A	N/A	87,865	4,782 (-1%)	N/A	N/A
AuthRegistry:consume (avm)	34,973	N/A	N/A	87,616	3,202 (+7%)	N/A	N/A
FPC:pay_refund (avm)	33,573	N/A	N/A	88,448	11,405	N/A	N/A
Benchmarking:create_note	1,344	9,944	1,422 (+1%)	N/A	N/A	N/A	N/A
SchnorrAccount:verify_private_authwit	1,280	9,944	77.7 (+7%)	N/A	N/A	N/A	N/A
Token:unshield	1,376	9,944	3,773 (+6%)	N/A	N/A	N/A	N/A
FPC:fee_entrypoint_private	1,376	9,944	4,737 (+2%)	N/A	N/A	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	66.5 (+2%)	13,873
GasToken:set_portal	17.2 (-6%)	3,495
Token:constructor	91.0 (-1%)	24,207
FPC:constructor	63.4	13,893
GasToken:mint_public	52.2 (+2%)	10,241
Token:mint_public	⚠️ 70.6 (-88%)	19,216
Token:assert_minter_and_mint	⚠️ 211 (+187%)	13,034
AuthRegistry:set_authorized	32.5 (+7%)	7,869
FPC:prepare_fee	149 (-2%)	15,187
Token:transfer_public	50.1 (-4%)	31,425
FPC:pay_refund	147 (+1%)	20,080
Benchmarking:increment_balance	2,680 (-1%)	15,465
Token:_increase_public_balance	56.1	15,089
FPC:pay_refund_with_shielded_rebate	146 (-1%)	21,167

Public DB Access

Time to access various public DBs.

Function	time_ms
get-nullifier-index	0.157 (+2%)

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	16.7 (+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	0.596	0.513	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	48.9 (+1%)	75.9	131 (-1%)	244	466 (-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	0.499 (+2%)	0.466	0.450	0.442	0.435 (-1%)
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	59.9 (+1%)	111 (-1%)	180 (-1%)	350 (-1%)	689
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	0.506 (+1%)	0.499	0.476 (-1%)	0.475	0.473
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	72.9	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	0.519	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	85,672	670,983

Transaction size based on fee payment method

| Metric | |
| - | |

[LHerskind]

I think this is really awesome work! For some of the test, I think we can be more explicit and check storage before and after updates to ensure that they were updated correctly. That was what the token simulator was mostly there for in the E2E tests, but some of the admin stuff etc I think we can make it slightly nicer and more explicit that it is really as expected etc (this was not done for the E2E, so more an improvement generally to both).

Tiny comments on naming, think this is mergable, but would prefer if you can add a few extra storage checks in the tests, rest can be dealt with in separate PR's 👍

[LHerskind]

I find some of the cheatcode structure slightly odd, e.g., having it a bit all over the place instead of having a folder just with cheatcodes. Ideally think it would be nice to have it outside of the others, to not pollute 🤷. But it is something I think we could deal with in a separate PR.

Might also just be that I am not thinking about something and we need it this way 👍.

[Thunkar]

Most of it is in aztec-nr, but for authwits it was necessary to do it this way due to circular dependencies: aztec-nr -> auth -> aztec-nr.

We could have a separate crate, but I though it wouldn't be so bad if our packages exported their tests interfaces themselves.

[LHerskind]

Was this the one that we could also get rid of?

[Thunkar]

No, in some places it makes sense. Naming could be improved though, because we use time traveling to "consolidate" the state and thus being able to read from historical data (for example, we cannot read contract public bytecode until it is deployed and accesible)

[LHerskind]

Do you have any ideas around how we could assert that it fails with a specific error? 👀

[Thunkar]

It's tricky. Dealing with strings in noir is pain, but will take a look.