Skip to content

Latest commit

 

History

History
160 lines (124 loc) · 4.53 KB

README.md

File metadata and controls

160 lines (124 loc) · 4.53 KB

Mollusk

SVM program test harness.

Harness

The harness is designed to directly invoke the loaded executable program using the rBPF VM, bypassing any transaction sanitization and runtime checks, and instead directly processing the instruction with the VM.

let program_id = Pubkey::new_unique();
let key1 = Pubkey::new_unique();
let key2 = Pubkey::new_unique();

let instruction = Instruction::new_with_bytes(
    program_id,
    &[],
    vec![
        AccountMeta::new(key1, false),
        AccountMeta::new_readonly(key2, false),
    ],
);

let accounts = vec![
    (key1, AccountSharedData::default()),
    (key2, AccountSharedData::default()),
];

let mollusk = Mollusk::new(program_id, "my_program");

let result = mollusk.process_instruction(&instruction, &accounts);

You can also use the Check API provided by Mollusk for easy post-execution checks, rather than writing them manually. The API method process_and_validate_instruction will still return the result, allowing you to perform further checks if you desire.

let sender = Pubkey::new_unique();
let recipient = Pubkey::new_unique();

let base_lamports = 100_000_000u64;
let transfer_amount = 42_000u64;

let instruction = system_instruction::transfer(&sender, &recipient, transfer_amount);
let accounts = [
    (
        sender,
        AccountSharedData::new(base_lamports, 0, &system_program::id()),
    ),
    (
        recipient,
        AccountSharedData::new(base_lamports, 0, &system_program::id()),
    ),
];
let checks = vec![
    Check::success(),
    Check::compute_units(system_processor::DEFAULT_COMPUTE_UNITS),
    Check::account(&sender)
        .lamports(base_lamports - transfer_amount)
        .build(),
    Check::account(&recipient)
        .lamports(base_lamports + transfer_amount)
        .build(),
];

Mollusk::default().process_and_validate_instruction(
    &instruction,
    &accounts,
    &checks,
);

Compute Unit Bencher

Mollusk also offers a compute unit usage bencher for profiling a program's compute unit usage.

Example:

// If using with `cargo bench`, tell Mollusk where to find the program.
std::env::set_var("SBF_OUT_DIR", "../target/deploy");

// Optionally disable logging.
solana_logger::setup_with("");

/* Instruction & accounts setup ... */

let mollusk = Mollusk::new(&program_id, "my_program");

MolluskComputeUnitBencher::new(mollusk)
    .bench(("bench0", &instruction0, &accounts0))
    .bench(("bench1", &instruction1, &accounts1))
    .bench(("bench2", &instruction2, &accounts2))
    .bench(("bench3", &instruction3, &accounts3))
    .must_pass(true)
    .out_dir("../target/benches")
    .execute();

You can invoke this benchmark test with cargo bench. Don't forget to add a bench to your project's Cargo.toml.

[[bench]]
name = "compute_units"
harness = false

Mollusk will output bench details to the output directory in Markdown.

Note: Delta is the change since the last time the bench was run.

Name CUs Delta
bench0 450 --
bench1 579 -129
bench2 1,204 +754
bench3 2,811 +2,361

Fuzz Fixture Support

Mollusk also has first-class support for generating fixtures from tests, which can be used for things like fuzzing.

There are two protobuf layouts supported by Mollusk:

  • org.mollusk.svm: The protobuf layouts defined by the Mollusk library, which map directly to the structure of a Mollusk unit test.
  • org.solana.sealevel.v1: The protobuf layouts defined by Firedancer and used to test program instructions between targets on Firedancer and Agave.

Each protobuf layout has its own corresponding crate for all of the binding support: mollusk-svm-fuzz-fixture and mollusk-svm-fuzz-fixture-firedancer respectively.

The base library itself (mollusk-svm) provides support for working with fixtures directly from a Mollusk instance, via the fuzz and fuzz-fd feature flags, which can be used standalone or together.

When either fuzz-fixture feature flag is enabled, Mollusk can do the following:

  • Generate a fixture from a unit test.
  • Process a given fixture as a unit test.
  • Convert to and from fixtures to Mollusk tests and results.

To generate a fuzz fixture from a Mollusk unit test, provide the necessary environment variables alongside your call to cargo test-sbf, like so:

EJECT_FUZZ_FIXTURES="my/fixtures/dir" cargo test-sbf ...

JSON versions of fixtures are also supported.

See the documentation in harness/src/lib.rs for more information.