feat: avm executor - layout and groundwork

[Maddiaa0]

Base work for avm - heavily a wip -> making a pr now for eyeballs

Key ideas:
Executor:

• There is an Avm Executor that has a reference to the current node state
• Any avm calls are triggered by interacting with the executor
  • fetching bytecode
  • triggering the interpreter

Interpreter:

• The interpreter contains:
  • The bytecode to execute, and a statemanager

AvmStateManager:

• The interpreter uses this as its source of truth when requesting data. It is a thin layer between the dbs that handles caching and rolling back state of failed calls.

Journal
Rather than reading and writing to the db, we introduce a journal.
All intermediate storage operations are read from here.
Journal structure:

• When a call ends; it is merged with the caller's journal
• When a transaction succeeds the journal is merged with the blocks journal
• When a block succeeds a journal is commited to db.

Opcode structure
Each opcode takes in a view of the state manager and the current call context.

• storage operations impact the current journal through the state manager
• jump instructions can impact the current opcode via changing the pc in the context

This diagram loosely shows the structure

fixes: #3317

[AztecBot]

Benchmark results

Metrics with a significant change:

• batch_insert_into_indexed_tree_ms (512): 529 (+47%)

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 9f61ed1f 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	221,040	865,868	3,445,916
l1_rollup_execution_gas	934,184	3,968,342	24,789,452
l2_block_processing_time_in_ms	1,515 (+14%)	5,158 (+2%)	21,527 (+2%)
note_successful_decrypting_time_in_ms	312 (+1%)	945 (+3%)	3,414 (+2%)
note_trial_decrypting_time_in_ms	21.3	70.6 (+55%)	146 (+12%)
l2_block_building_time_in_ms	14,877	59,295	238,639 (+1%)
l2_block_rollup_simulation_time_in_ms	10,499	41,962 (+1%)	169,479 (+1%)
l2_block_public_tx_process_time_in_ms	4,343	17,256	68,886 (+1%)

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	14,942 (-1%)	28,718
note_history_successful_decrypting_time_in_ms	2,207	4,487 (+4%)
note_history_trial_decrypting_time_in_ms	116 (-10%)	186 (+16%)
node_database_size_in_bytes	3,627,127	3,723,872
pxe_database_size_in_bytes	29,923	59,478

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	213	44,681	25,561
private-kernel-ordering	165	43,753	14,809
base-rollup	1,064	128,338	881
root-rollup	82.1 (+2%)	4,088	709
private-kernel-inner	291	71,208	25,561
public-kernel-private-input	205	31,863	25,561
public-kernel-non-first-iteration	203	31,905	25,561
merge-rollup	8.28 (+1%)	2,608	881

Tree insertion stats

The duration to insert a fixed batch of leaves into each tree type.

Metric	1 leaves	2 leaves	8 leaves	16 leaves	32 leaves	64 leaves	128 leaves	512 leaves	1024 leaves	2048 leaves	8192 leaves
batch_insert_into_append_only_tree_ms	11.2 (+1%)	16.9 (+7%)	13.6	18.0 (+1%)	24.2 (+3%)	50.1 (+1%)	84.2 (+9%)	243	469	916 (+1%)	3,704 (+4%)
batch_insert_into_indexed_tree_ms	N/A	N/A	N/A	64.5	N/A	60.0 (+1%)	110 (-4%)	⚠️ 529 (+47%)	732 (-1%)	1,412 (+1%)	5,596 (+1%)

Miscellaneous

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

Metric	0 deployed contracts	1 deployed contracts
tx_size_in_bytes	15,455	32,224

Transaction processing duration by data writes.

Metric	0 new commitments	1 new commitments
tx_pxe_processing_time_ms	374	1,083

Metric	1 public data writes
tx_sequencer_processing_time_ms	535

[Maddiaa0]

Current dependencies on/for this PR:

• master
  • PR feat: avm executor - layout and groundwork #3928 👈
    • PR feat(public-vm): avm journal #3945

This stack of pull requests is managed by Graphite.

[dbanks12]

Very nice work! Left some comments (mostly minor).

[dbanks12]

LGTM, but we should still resolve the remaining comments :)