feat: bb uses goblin

[codygunton]

"vertical slice" of goblin integrating with ACIR and BB.

Beginning of goblin ultra honk interface for bb. Can only run a basic test (assert_statement) for now and is not intended for use outside of the CI check that it's still working, for now.

• adds bb and bb.js command prove_and_verify_goblin. Adds bb.js bindings for goblin prove/verify
• modifies ACIR dsl folder to be able to take a goblin builder
• adds CI calls to bb.js and bb that use prove_and_verify_goblin
• allowing ability to load grumpkin SRS through memory, needed for bb and (especially) bb.js. This allows an alternate source of points to be used other than the default file-based grumpkin loader, which mostly only works in dev (though revisit: could this work for native bb?)

[kevaundray]

We allow two proving systems to be created, once of them is "snark friendly" meaning that it uses PedersenBlake in the transcript and the other is solidity friendly meaning that it uses keccak in the transcript.

When creating a proof that we will want to verify in a another circuit, we use the is_recursive flag.

See https://hackmd.io/@6iQDuIePQjyYBqDChYw_jg/SyLHcKsSa for a few more details

[codygunton]

Thanks for explaining. Our plans is to do violence to the code to get something working in the non-recursive case this week then check in with you and Maxim (prob next year?) to look things over together and rework to something non-disgusting :)

[kevaundray]

Oh god :)

[kevaundray]

Bear in mind that if a release is made, then this code will make its way into Noir user's hands -- haven't looked over the changes, though it seems that the solidity verifier contract may no longer work for them if this were to be merged?

[codygunton]

Ok don't worry, we just want to get it working in this branch, won't merge

[ludamad]

can be removed

[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 a60b71a5 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	128 txs
l1_rollup_calldata_size_in_bytes	45,444	179,588	716,132
l1_rollup_calldata_gas	222,984	867,776	3,449,420
l1_rollup_execution_gas	842,071	3,594,884	22,205,082
l2_block_processing_time_in_ms	1,357 (+1%)	5,338 (+4%)	21,175
note_successful_decrypting_time_in_ms	344	1,093 (-2%)	3,916 (+2%)
note_trial_decrypting_time_in_ms	55.4 (+4%)	37.3 (-1%)	138
l2_block_building_time_in_ms	13,731 (+1%)	53,698 (-1%)	216,870
l2_block_rollup_simulation_time_in_ms	10,182 (+2%)	39,716 (-1%)	160,510
l2_block_public_tx_process_time_in_ms	3,515	13,914	56,113

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	15,706 (-2%)	29,393
note_history_successful_decrypting_time_in_ms	2,553 (+1%)	5,170 (+2%)
note_history_trial_decrypting_time_in_ms	99.3 (+17%)	151 (+8%)
node_database_size_in_bytes	3,910,865	4,256,099
pxe_database_size_in_bytes	29,940	59,499

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	199	43,109	20,441
private-kernel-ordering	114 (-1%)	25,833	9,689
base-rollup	2,055	244,144	881
root-rollup	84.8 (-4%)	4,088	889
private-kernel-inner	260	64,516	20,441
public-kernel-private-input	171 (-1%)	25,203	20,441
public-kernel-non-first-iteration	170 (-1%)	25,245	20,441
merge-rollup	11.1 (-2%)	2,608	881

Miscellaneous

Transaction sizes based on how many contracts are deployed in the tx.

Metric	0 deployed contracts	1 deployed contracts
tx_size_in_bytes	10,323	26,359

[ludamad]

Should this be deleted?

[ledwards2225]

deleted

[ludamad]

nit: usually there's a better name than X_utility. But to be clear, it's holidays, get'er in

[ledwards2225]

changed to instance_inspector - useful for now but may delete in the near future

[ludamad]

the old path actually never found the file, always downloaded

[kevaundray]

Wait why has this now been merged?

see: #3636 (comment)

[ledwards2225]

Wait why has this now been merged?

see: #3636 (comment)

Not to worry! This still isn't hooked up to anything real, it just runs a few new tests using Goblin. Sorry - poor communication on our part. When Cody said "this wont be merged" we were still thinking it wouldn't be. Since then we decided to do something more isolated very much for the purpose of getting it merged