refactor: Meld flavor and and circuit builder modules

[codygunton]

The proof_system module has become a pile of things, and the name is useless. The PR resolves that and moves us toward a model where flavor and circuit builders live side-by-side since this is how we implement check_circuit.

Split proof_system module.


• Ultra & GoblinUltra go into into stdlib_circuit_builders
• VM trace builders go into single folder alongside other stuff (no stdlib dependency, so little benefit to building separately)
• Library stuff goes into plonk_honk_shared (library stuff); will refactor when Plonk moves out.
• Polynomials module gets the polynomial store.

We also simplify a little bit of the linking structure.
New module structure:

[AztecBot]

Benchmark results

No metrics with a significant change found.

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.

Values are compared against data from master at commit 2136a66c and shown if the difference exceeds 1%.

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	676	676	676
l1_rollup_calldata_gas	6,424	6,424	6,424
l1_rollup_execution_gas	585,757	585,757	585,757
l2_block_processing_time_in_ms	1,260 (-3%)	4,594 (-3%)	9,078 (-1%)
note_successful_decrypting_time_in_ms	178 (+7%)	538 (+4%)	926 (-6%)
note_trial_decrypting_time_in_ms	82.9 (+20%)	55.7 (+68%)	54.8 (-55%)
l2_block_building_time_in_ms	16,733 (-4%)	62,291 (-3%)	123,334 (-6%)
l2_block_rollup_simulation_time_in_ms	7,826 (-1%)	27,129 (-4%)	53,626 (-2%)
l2_block_public_tx_process_time_in_ms	8,886 (-7%)	35,110 (-3%)	69,615 (-9%)

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	13,310 (-3%)	28,045
note_history_successful_decrypting_time_in_ms	1,204 (-3%)	2,679 (+10%)
note_history_trial_decrypting_time_in_ms	90.7 (+9%)	134 (-12%)
node_database_size_in_bytes	18,640,976	35,086,416
pxe_database_size_in_bytes	29,859	59,414

Circuits stats

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

Circuit	circuit_simulation_time_in_ms	circuit_input_size_in_bytes	circuit_output_size_in_bytes
private-kernel-init	233 (-2%)	44,377	28,214
private-kernel-ordering	207 (-2%)	52,880	14,296
base-parity	4,463	128	311
root-parity	1,620 (-3%)	1,244	311
base-rollup	17,502 (-2%)	165,760	861
root-rollup	50.2	4,359	725
private-kernel-inner	590 (-2%)	73,794	28,214
public-kernel-app-logic	388 (-3%)	35,251	28,217
public-kernel-tail	170 (-3%)	40,928	28,217
merge-rollup	8.45 (-2%)	2,568	861

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	10.0	16.0	N/A	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_count	16.8	31.6	N/A	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_ms	0.583	0.495	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	45.3 (-2%)	70.6 (-2%)	223 (-3%)	439 (-1%)	834 (-4%)	1,659 (-4%)	N/A
batch_insert_into_append_only_tree_32_depth_hash_count	N/A	N/A	96.0	159	543	1,055	2,079	4,127	N/A
batch_insert_into_append_only_tree_32_depth_hash_ms	N/A	N/A	0.465 (-2%)	0.436 (-2%)	0.408 (-3%)	0.411 (-1%)	0.398 (-4%)	0.397 (-4%)	N/A
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	54.2 (+1%)	104 (-4%)	324 (-5%)	665 (+1%)	1,260 (-5%)	2,511 (-5%)	N/A
batch_insert_into_indexed_tree_20_depth_hash_count	N/A	N/A	104	207	691	1,363	2,707	5,395	N/A
batch_insert_into_indexed_tree_20_depth_hash_ms	N/A	N/A	0.481	0.467 (-4%)	0.442 (-5%)	0.458 (+1%)	0.437 (-5%)	0.436 (-5%)	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	60.1 (-3%)
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	109
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	0.525 (-3%)

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	14,966	494,914

Transaction processing duration by data writes.

Metric	0 new note hashes	1 new note hashes
tx_pxe_processing_time_ms	2,929 (-5%)	1,602 (-1%)

Metric	0 public data writes	1 public data writes
tx_sequencer_processing_time_ms	16.5 (+14%)	1,116 (-2%)

[jeanmon]

LGTM.

[maramihali]

I dont quite get why the flavor should be in the stdlib_circuit_builders since the circuit builders are not templated by flavor but maybe I am missing sth

[codygunton]

Flavor and circuit builder go side-by-side now--they have to be designed together, anyway.

[maramihali]

where have these tests disappeared?

[codygunton]

Discussed ofline, they moved to another file having the same name but in the translator folder.