18574276475

Committed 16 Oct 2025 08:43PM UTC coverage: 73.757% (+0.1%) from 73.624%

Build # 18574276475

Build Type

Pull #726

github

Committed by

web-flow

Commit Message

Merge 70117bfed into aa3485c08

Pull Request Pull Request #726: feat: HTTP-JSON RPC framework (server-side)

Run Details

265 of 301 new or added lines in 13 files covered. (88.04%)

14 existing lines in 3 files now uncovered.

25250 of 34234 relevant lines covered (73.76%)

662959.28 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

68.7

/neptune-core/src/lib.rs

// recursion limit for macros (e.g. triton_asm!)
#![recursion_limit = "2048"]
#![deny(clippy::shadow_unrelated)]
//
// enables nightly feature async_fn_track_caller for crate feature log-slow-write-lock.
// log-slow-write-lock logs warning when a write-lock is held longer than 100 millis.
// to enable: cargo +nightly build --features log-slow-write-lock
#![cfg_attr(feature = "track-lock-location", feature(async_fn_track_caller))]
//
// If code coverage tool `cargo-llvm-cov` is running with the nightly toolchain,
// enable the unstable “coverage” attribute. This allows using the annotation
// `#[coverage(off)]` to explicitly exclude certain parts of the code from
// being considered as “code under test.” Most prominently, the annotation
// should be added to every `#[cfg(test)]` module. Since the “coverage”
// feature is enable only conditionally, the annotation to use is:
// `#[cfg_attr(coverage_nightly, coverage(off))]`.
//
// See also:
// - https://github.com/Neptune-Crypto/neptune-core/issues/570
// - https://github.com/taiki-e/cargo-llvm-cov#exclude-code-from-coverage
// - https://github.com/rust-lang/rust/issues/84605
#![cfg_attr(coverage_nightly, feature(coverage_attribute))]

// danda: making all of these pub for now, so docs are generated.
// later maybe we ought to split some stuff out into re-usable crate(s)...?
pub mod api;
pub mod application;
pub mod macros;
pub mod prelude;
pub mod protocol;
pub mod state;
pub mod util_types;

#[cfg(test)]
#[cfg_attr(coverage_nightly, coverage(off))]
pub mod tests;

#[cfg_attr(coverage_nightly, coverage(off))]
pub mod bench_helpers;

use std::env;
use std::path::PathBuf;

use anyhow::Context;
use anyhow::Result;
use application::config::cli_args;
use chrono::DateTime;
use chrono::Local;
use chrono::Utc;
use futures::future;
use futures::Future;
use futures::StreamExt;
use itertools::Itertools;
use prelude::tasm_lib;
use prelude::triton_vm;
use prelude::twenty_first;
use protocol::consensus::block::Block;
use protocol::peer::handshake_data::HandshakeData;
use state::wallet::wallet_file::WalletFileContext;
use state::GlobalState;
use tarpc::server;
use tarpc::server::incoming::Incoming;
use tarpc::server::Channel;
use tarpc::tokio_serde::formats::*;
use tokio::net::TcpListener;
use tokio::sync::broadcast;
use tokio::sync::mpsc;
use tokio::time::Instant;
use tracing::debug;
use tracing::info;
use triton_vm::prelude::BFieldElement;

use crate::application::config::data_directory::DataDirectory;
use crate::application::json_rpc::server::http::RpcServer;
use crate::application::locks::tokio as sync_tokio;
use crate::application::loops::channel::MainToMiner;
use crate::application::loops::channel::MainToPeerTask;
use crate::application::loops::channel::MinerToMain;
use crate::application::loops::channel::PeerTaskToMain;
use crate::application::loops::channel::RPCServerToMain;
use crate::application::loops::connect_to_peers::call_peer;
use crate::application::loops::main_loop::MainLoopHandler;
use crate::application::rpc::server::RPC;
use crate::state::archival_state::ArchivalState;
use crate::state::wallet::wallet_state::WalletState;
use crate::state::GlobalStateLock;

pub const SUCCESS_EXIT_CODE: i32 = 0;
pub const COMPOSITION_FAILED_EXIT_CODE: i32 = 159;

/// Magic string to ensure other program is Neptune Core
pub const MAGIC_STRING_REQUEST: &[u8; 15] = b"7B8AB7FC438F411";
pub const MAGIC_STRING_RESPONSE: &[u8; 15] = b"Hello Neptune!\n";
const PEER_CHANNEL_CAPACITY: usize = 1000;
const MINER_CHANNEL_CAPACITY: usize = 10;
const RPC_CHANNEL_CAPACITY: usize = 1000;
const VERSION: &str = env!("CARGO_PKG_VERSION");

/// Causes compilation failures on targets where `u32` does not fit within a
/// `usize`.
const _MIN_PTR_WIDTH: () = {
    #[cfg(target_pointer_width = "16")]
    compile_error!("This crate requires a target pointer width of at least 32 bits.");
};

pub async fn initialize(cli_args: cli_args::Args) -> Result<MainLoopHandler> {
    async fn spawn(fut: impl Future<Output = ()> + Send + 'static) {
        tokio::spawn(fut);
    }

    // see comment for Network::performs_automated_mining()
    if cli_args.mine() && !cli_args.network.performs_automated_mining() {
        anyhow::bail!("Automatic mining is not supported for network {}.  Try again without --compose or --guess flags.", cli_args.network);
    }

    info!("Starting neptune-core node on {}.", cli_args.network);

    // Get data directory (wallet, block database), create one if none exists
    let data_directory = DataDirectory::get(cli_args.data_dir.clone(), cli_args.network)?;
    DataDirectory::create_dir_if_not_exists(&data_directory.root_dir_path()).await?;
    info!("Data directory is {}", data_directory);

    let (rpc_server_to_main_tx, rpc_server_to_main_rx) =
        mpsc::channel::<RPCServerToMain>(RPC_CHANNEL_CAPACITY);
    let genesis = Block::genesis(cli_args.network);
    let global_state =
        GlobalState::try_new(data_directory.clone(), genesis, cli_args.clone()).await?;
    let mut global_state_lock =
        GlobalStateLock::from_global_state(global_state, rpc_server_to_main_tx.clone());

    // Construct the broadcast channel to communicate from the main task to peer tasks
    let (main_to_peer_broadcast_tx, _main_to_peer_broadcast_rx) =
        broadcast::channel::<MainToPeerTask>(PEER_CHANNEL_CAPACITY);

    // Add the MPSC (multi-producer, single consumer) channel for peer-task-to-main communication
    let (peer_task_to_main_tx, peer_task_to_main_rx) =
        mpsc::channel::<PeerTaskToMain>(PEER_CHANNEL_CAPACITY);

    if let Some(block_import_directory) =
        global_state_lock.cli().import_blocks_from_directory.clone()
    {
        info!(
            "Importing blocks from directory \"{}\"",
            block_import_directory.to_string_lossy()
        );

        let flush_period = global_state_lock.cli().import_block_flush_period;
        let validate_blocks = !global_state_lock.cli().disable_validation_in_block_import;
        let num_blocks_read = global_state_lock
            .lock_guard_mut()
            .await
            .import_blocks_from_directory(&block_import_directory, flush_period, validate_blocks)
            .await?;
        info!("Successfully imported {num_blocks_read} blocks.");
    }

    if !cli_args.triton_vm_env_vars.is_empty() {
        info!(
            "Triton VM environment variables set to: {}",
            cli_args.triton_vm_env_vars
        );
    }

    if !cli_args.whitelisted_composers.is_empty() {
        info!(
            "Whitelisted composers:\n{}",
            cli_args.whitelisted_composers.iter().join("\n")
        );
    }

    // Check if we need to restore the wallet database, and if so, do it.
    info!("Checking if we need to restore UTXOs");
    global_state_lock
        .lock_guard_mut()
        .await
        .restore_monitored_utxos_from_recovery_data()
        .await?;
    info!("UTXO restoration check complete");

    // Bind socket to port on this machine, to handle incoming connections from peers
    let incoming_peer_listener = if let Some(incoming_peer_listener) = cli_args.own_listen_port() {
        let ret = TcpListener::bind((cli_args.peer_listen_addr, incoming_peer_listener))
           .await
           .with_context(|| format!("Failed to bind to local TCP port {}:{}. Is an instance of this program already running?", cli_args.peer_listen_addr, incoming_peer_listener))?;
        info!("Now listening for incoming peer-connections");
        ret
    } else {
        info!("Not accepting incoming peer-connections");
        TcpListener::bind("127.0.0.1:0").await?
    };

    // Connect to peers, and provide each peer task with a thread-safe copy of the state
    let own_handshake_data: HandshakeData =
        global_state_lock.lock_guard().await.get_own_handshakedata();
    info!(
        "Most known canonical block has height {}",
        own_handshake_data.tip_header.height
    );
    let mut task_join_handles = vec![];
    for peer_address in global_state_lock.cli().peers.clone() {
        let peer_state_var = global_state_lock.clone(); // bump arc refcount
        let main_to_peer_broadcast_rx_clone: broadcast::Receiver<MainToPeerTask> =
            main_to_peer_broadcast_tx.subscribe();
        let peer_task_to_main_tx_clone: mpsc::Sender<PeerTaskToMain> = peer_task_to_main_tx.clone();
        let peer_join_handle = tokio::task::spawn(async move {
            call_peer(
                peer_address,
                peer_state_var.clone(),
                main_to_peer_broadcast_rx_clone,
                peer_task_to_main_tx_clone,
                own_handshake_data,
                1, // All outgoing connections have distance 1
            )
            .await;
        });
        task_join_handles.push(peer_join_handle);
    }
    debug!("Made outgoing connections to peers");

    // Start mining tasks if requested
    let (miner_to_main_tx, miner_to_main_rx) = mpsc::channel::<MinerToMain>(MINER_CHANNEL_CAPACITY);
    let (main_to_miner_tx, main_to_miner_rx) = mpsc::channel::<MainToMiner>(MINER_CHANNEL_CAPACITY);
    let miner_state_lock = global_state_lock.clone(); // bump arc refcount.
    if global_state_lock.cli().mine() {
        let miner_join_handle = tokio::task::spawn(async move {
            application::loops::mine_loop::mine(
                main_to_miner_rx,
                miner_to_main_tx,
                miner_state_lock,
            )
            .await
            .expect("Error in mining task");
        });
        task_join_handles.push(miner_join_handle);
        info!("Started mining task");
    }

    // Start RPC server for CLI request and more. It's important that this is done as late
    // as possible, so requests do not hang while initialization code runs.
    let mut rpc_listener = tarpc::serde_transport::tcp::listen(
        format!("127.0.0.1:{}", global_state_lock.cli().rpc_port),
        Json::default,
    )
    .await?;
    rpc_listener.config_mut().max_frame_length(usize::MAX);

    let rpc_state_lock = global_state_lock.clone();

    // each time we start neptune-core a new RPC cookie is generated.
    let valid_tokens: Vec<application::rpc::auth::Token> = vec![
        crate::application::rpc::auth::Cookie::try_new(&data_directory)
            .await?
            .into(),
    ];

    let rpc_join_handle = tokio::spawn(async move {
        rpc_listener
            // Ignore accept errors.
            .filter_map(|r| future::ready(r.ok()))
            .map(server::BaseChannel::with_defaults)
            // Limit channels to 5 per IP. 1 for dashboard and a few more for CLI interactions
            .max_channels_per_key(5, |t| t.transport().peer_addr().unwrap().ip())
            // serve is generated by the service attribute. It takes as input any type implementing
            // the generated RPC trait.
            .map(move |channel| {
                let server = application::rpc::server::NeptuneRPCServer::new(
                    rpc_state_lock.clone(),
                    rpc_server_to_main_tx.clone(),
                    data_directory.clone(),
                    valid_tokens.clone(),
                );

                channel.execute(server.serve()).for_each(spawn)
            })
            // Max 10 channels.
            .buffer_unordered(10)
            .for_each(|_| async {})
            .await;
    });
    task_join_handles.push(rpc_join_handle);
    info!("Started RPC server");

    if let Some(addr) = global_state_lock.cli().listen_rpc {
        let listener = TcpListener::bind(addr).await.unwrap();
        let json_rpc_state_lock = global_state_lock.clone();

        let json_rpc_join_handle = tokio::spawn(async move {
            let rpc_server = RpcServer::new(json_rpc_state_lock);
            rpc_server.serve(listener).await;
        });
        task_join_handles.push(json_rpc_join_handle);

        info!("Started HTTP-JSON RPC server on {}.", addr);
    }

    // Handle incoming connections, messages from peer tasks, and messages from the mining task
    Ok(MainLoopHandler::new(
        incoming_peer_listener,
        global_state_lock,
        main_to_peer_broadcast_tx,
        peer_task_to_main_tx,
        main_to_miner_tx,
        peer_task_to_main_rx,
        miner_to_main_rx,
        rpc_server_to_main_rx,
        task_join_handles,
    ))
}

/// Time a fn call.  Duration is returned as a float in seconds.
pub fn time_fn_call<O>(f: impl FnOnce() -> O) -> (O, f64) {
    let start = Instant::now();
    let output = f();
    let elapsed = start.elapsed();
    let total_time = elapsed.as_secs() as f64 + f64::from(elapsed.subsec_nanos()) / 1e9;
    (output, total_time)
}

/// Time an async fn call.  Duration is returned as a float in seconds.
pub async fn time_fn_call_async<F, O>(f: F) -> (O, f64)
where
    F: std::future::Future<Output = O>,
{
    let start = Instant::now();
    let output = f.await;
    let elapsed = start.elapsed();
    let total_time = elapsed.as_secs() as f64 + f64::from(elapsed.subsec_nanos()) / 1e9;
    (output, total_time)
}

/// Converts a UTC millisecond timestamp (millis since 1970 UTC) into
/// a `DateTime<Local>`, ie local-time.
///
/// # Return Value
///
/// Returns `None` if
///  - the given argument cannot be converted to an `i64`, or
///  - the given argument, after conversion to `i64`, is out of range (whatever
///    that means).
pub fn utc_timestamp_to_localtime<T>(timestamp: T) -> Option<DateTime<Local>>
where
    T: TryInto<i64>,
    <T as TryInto<i64>>::Error: std::fmt::Debug,
{
    let millis: i64 = timestamp.try_into().ok()?;
    let utc = DateTime::<Utc>::from_timestamp_millis(millis)?;
    Some(utc.with_timezone(&Local))
}

#[cfg(feature = "log-lock_events")]
pub(crate) fn current_thread_id() -> u64 {
    // workaround: parse thread_id debug output into a u64.
    // (because ThreadId::as_u64() is unstable)
    let thread_id_dbg: String = format!("{:?}", std::thread::current().id());
    let nums_u8 = &thread_id_dbg
        .chars()
        .filter_map(|c| {
            if c.is_ascii_digit() {
                Some(c as u8)
            } else {
                None
            }
        })
        .collect::<Vec<u8>>();
    let nums = String::from_utf8_lossy(nums_u8).to_string();

    nums.parse::<u64>().unwrap()
}

// This is a callback fn passed to AtomicRw, AtomicMutex
// and called when a lock event occurs.  This way
// we can track which threads+tasks are acquiring
// which locks for reads and/or mutations.
pub(crate) fn log_tokio_lock_event_cb(lock_event: sync_tokio::LockEvent) {
    #[cfg(feature = "log-lock_events")]
    log_tokio_lock_event(&lock_event);

    match lock_event.acquisition() {
        #[cfg(feature = "log-slow-read-lock")]
        sync_tokio::LockAcquisition::Read => log_slow_locks(&lock_event, "read"),
        #[cfg(feature = "log-slow-write-lock")]
        sync_tokio::LockAcquisition::Write => log_slow_locks(&lock_event, "write"),

        _ => {}
    }
}

// notes:
//   1. this feature is very verbose in the logs.
//   2. It's not really needed except when debugging lock acquisitions
//   3. tracing-tests causes a big mem-leak for tests with this.
#[cfg(feature = "log-lock_events")]
pub(crate) fn log_tokio_lock_event(lock_event: &sync_tokio::LockEvent) {
    use std::ops::Sub;

    let tokio_id = match tokio::task::try_id() {
        Some(id) => format!("{}", id),
        None => "?".to_string(),
    };

    let location_str = match lock_event.location() {
        Some(l) => format!("\n\t|-- acquirer: {}", l),
        None => String::default(),
    };
    let waited_for_acquire_str = match (lock_event.try_acquire_at(), lock_event.acquire_at()) {
        (Some(t), Some(a)) => format!(
            "\n\t|-- waited for acquire: {} secs",
            a.sub(t).as_secs_f32()
        ),
        _ => String::default(),
    };
    let held_str = match lock_event.acquire_at() {
        Some(t) if matches!(lock_event, sync_tokio::LockEvent::Release { .. }) => {
            format!("\n\t|-- held: {} secs", t.elapsed().as_secs_f32())
        }
        _ => String::default(),
    };

    let info = lock_event.info();

    tracing::trace!(
            ?lock_event,
            "{} tokio lock `{}` of type `{}` for `{}` by\n\t|-- thread {}, (`{}`)\n\t|-- tokio task {}{}{}{}\n\t|--",
            lock_event.event_type_name(),
            info.name().unwrap_or("?"),
            info.lock_type(),
            lock_event.acquisition(),
            current_thread_id(),
            std::thread::current().name().unwrap_or("?"),
            tokio_id,
            location_str,
            waited_for_acquire_str,
            held_str,
    );
}

#[cfg(any(feature = "log-slow-read-lock", feature = "log-slow-write-lock"))]
pub(crate) fn log_slow_locks(event: &sync_tokio::LockEvent, read_or_write: &str) {
    use std::ops::Sub;
    if matches!(event, sync_tokio::LockEvent::Acquire { .. }) {
        if let (Some(try_acquire_at), Some(acquire_at), Some(location)) =
            (event.try_acquire_at(), event.acquire_at(), event.location())
        {
            let duration = acquire_at.sub(try_acquire_at);
            let env_var = format!(
                "LOG_SLOW_{}_LOCK_ACQUIRE_THRESHOLD",
                read_or_write.to_uppercase()
            );
            let max_duration_secs = match std::env::var(env_var) {
                Ok(t) => t.parse().unwrap(),
                Err(_) => 0.1,
            };

            if duration.as_secs_f32() > max_duration_secs {
                tracing::warn!(
                    "{}-lock held for {} seconds. (exceeds max: {} secs)  location: {}",
                    read_or_write,
                    duration.as_secs_f32(),
                    max_duration_secs,
                    location
                );
            }
        }
    }

    if let (Some(acquired_at), Some(location)) = (event.acquire_at(), event.location()) {
        let duration = acquired_at.elapsed();
        let env_var = format!("LOG_SLOW_{}_LOCK_THRESHOLD", read_or_write.to_uppercase());
        let max_duration_secs = match std::env::var(env_var) {
            Ok(t) => t.parse().unwrap(),
            Err(_) => 0.1,
        };

        if duration.as_secs_f32() > max_duration_secs {
            tracing::warn!(
                "{}-lock held for {} seconds. (exceeds max: {} secs)  location: {}",
                read_or_write,
                duration.as_secs_f32(),
                max_duration_secs,
                location
            );
        }
    }
}

const LOG_TOKIO_LOCK_EVENT_CB: sync_tokio::LockCallbackFn = log_tokio_lock_event_cb;

/// for logging how long a scope takes to execute.
///
/// If an optional threshold value is provided then nothing will be
/// logged unless execution duration exceeds the threshold.
/// In that case a tracing::warn!() is logged.
///
/// If no threshold value is provided then a tracing::debug!()
/// is always logged with the duration.
///
/// for convenience see macros:
///  crate::macros::log_slow_scope,
///  crate::macros::log_scope_duration,
#[derive(Debug, Clone)]
pub struct ScopeDurationLogger<'a> {
    start: Instant,
    description: &'a str,
    log_slow_fn_threshold: Option<f64>,
    location: &'static std::panic::Location<'static>,
}
impl<'a> ScopeDurationLogger<'a> {
    #[track_caller]
    pub fn new(description: &'a str, log_slow_fn_threshold: Option<f64>) -> Self {
        Self {
            start: Instant::now(),
            description,
            log_slow_fn_threshold,
            location: std::panic::Location::caller(),
        }
    }

    #[track_caller]
    pub fn new_with_threshold(description: &'a str, log_slow_fn_threshold: f64) -> Self {
        Self::new(description, Some(log_slow_fn_threshold))
    }

    #[track_caller]
    pub fn new_default_threshold(description: &'a str) -> Self {
        Self::new_with_threshold(
            description,
            match env::var("LOG_SLOW_SCOPE_THRESHOLD") {
                Ok(t) => t.parse().unwrap(),
                Err(_) => 0.001,
            },
        )
    }

    #[track_caller]
    pub fn new_without_threshold(description: &'a str) -> Self {
        Self::new(description, None)
    }
}

impl Drop for ScopeDurationLogger<'_> {
    fn drop(&mut self) {
        let elapsed = self.start.elapsed();
        let duration = elapsed.as_secs_f64();

        if let Some(threshold) = self.log_slow_fn_threshold {
            if duration >= threshold {
                let msg = format!(
                    "executed {} in {} secs.  exceeds slow fn threshold of {} secs.  location: {}",
                    self.description, duration, threshold, self.location,
                );

                tracing::debug!("{}", msg);
            }
        } else {
            let msg = format!(
                "executed {} in {} secs.  location: {}",
                self.description, duration, self.location,
            );

            tracing::debug!("{}", msg);
        }
    }
}

/// recursively copy source dir to destination
pub(crate) fn copy_dir_recursive(source: &PathBuf, destination: &PathBuf) -> std::io::Result<()> {
    if !source.is_dir() {
        return Err(std::io::Error::new(
            std::io::ErrorKind::NotADirectory,
            format!("not a directory: {}", source.display()),
        ));
    }
    std::fs::create_dir_all(destination)?;
    for entry in std::fs::read_dir(source)? {
        let entry = entry?;
        let dest_path = &destination.join(entry.file_name());
        if entry.path().is_dir() {
            copy_dir_recursive(&entry.path(), dest_path)?;
        } else {
            std::fs::copy(entry.path(), dest_path)?;
        }
    }
    Ok(())
}

Neptune-Crypto / neptune-core / 18574276475

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous