runtimeverification / evm-semantics

K Semantics of the Ethereum Virtual Machine (EVM)
BSD 3-Clause "New" or "Revised" License
503 stars 135 forks source link

Parallelize symbolic execution of tests #2009

Open tothtamas28 opened 11 months ago

tothtamas28 commented 11 months ago

This allows better utilization in case the number of tests remaining is smaller than the number of workers. (In particular if there's a single test.)

Design constraints

  1. Although kore-rpc is able to handle requests in parallel, each proof will probably need its own KoreServer instance. This is because pyk.proof.reachability.APRProver calls add_dependencies_module in its initializer. This limitation need further investigation though, maybe there's a way to prevent interference between proofs on the same server.
  2. Due to the Global Interpreter Lock in CPython, we either need to ensure that Python code has minimal overhead, or use process-level parallelism instead of threads. This means that we need to ensure pickleability of objects that are sent between processes.

Draft

This is based on pyk.kore.pool.KoreServerPool (design constraint 1) that currently uses concurrent.futures.ThreadPoolExecutor internally. As mentioned, this will probably have be replaced by ProcessPoolExecutor (design constraint 2).

Main thread initializes the pool with some initial tasks, and then processes results as they become available in a loop. Processing a result involves:

# pending is a collection of futures
while pending:
    # pick a done future to process
    current_future = next((future for future in pending if future.done()), None)

    # if there is none, sleep for a while then retry
    if current_future is None:
        time.sleep(0.1)
        continue

    # get the result - if an exception occurred in in the thread, it will be raised
    result = current_future.result()
    # update the corresponding proof on disk
    write_data_to_disk(result)

    # schedule new tasks
    next_tasks = get_next_tasks(result)
    for fn, args in next_tasks:
        next_future = pool.submit(fn, *args)
        pending.append(next_future)

    # this task is completed
    pending.remove(current_future)

Ideally, granularity of a task should be a single proof step (i.e. an RPC call and as much of the necessary pre- and post-processing as makes sense). This is the most challenging part of the issue: it requires countrparts of KCFGExplore, APRPRover, kevm_prove, etc. that 1) enable fine-grained control of the proof process 2) do not write to disk. This suggests a prover interface like:

class Prover:
    def advance_proof(self, proof_state: ProofState) -> ProofStep:
        # Does a single RPC call based on proof state and prover configuration. Does not write to disk.
        ...

Further considerations

Task scheduling

It would be good to have control on task scheduling. For example, in the simplest case, tests are assigned a unique priority, and each task inherits the priority of the test it belongs to. This means that a test with a lower priority can only run if tests with a higher priority don't have enough tasks scheduled to utilize all workers. In a good design, such scheduling strategies can be provided by the caller (e.g. in form of a callable that takes a task and returns its priority).

Unfortunately Executor implementations take tasks from Queue, so tasks are processed in FIFO order. But maybe there's a way to wrap the executor and use an additional PriorityQueue:

Graceful shutdown

On an exception or user request, the prover should shut down gracefully. In particular:

yale-vinson commented 7 months ago

Issue blocked by Backend Booster #383

yale-vinson commented 3 months ago

@nwatson22 As Backend Booster #383 is closed, should we unblock this and close it out as well?