Add recorder: raw CAN and GNSS sessions with failsafes and a storage bound

Reads SocketCAN and gpsd directly. Sessions start on the first frame or on
'ctl start', end after 15 s of bus idle or on 'ctl stop'; blips under 3 s are
discarded. candump -L chunks fdatasync'd every second and zstd'd once closed;
oldest sessions deleted to stay under 50 GB / above 20 GB free. Kernel and
queue drops, link changes and clock steps are recorded as events.

Bench-tested on vcan with the first car capture (byte-identical) and installed
on the car Orin as yukart-recorder.service.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01DDwfd6Ki8UfG4hK6W5YpJz
This commit is contained in:
2026-10-09 20:15:13 +09:00
commit 1f678fe87f
19 changed files with 1784 additions and 0 deletions
+149
View File
@@ -0,0 +1,149 @@
//! Producer threads. Each owns one input, reconnects on its own, and never blocks
//! on the writer: CAN frames that do not fit the queue are counted, not waited for.
use std::io::{BufRead, BufReader, Read, Write};
use std::net::{TcpStream, ToSocketAddrs};
use std::os::unix::fs::PermissionsExt;
use std::os::unix::net::UnixListener;
use std::sync::atomic::{AtomicBool, AtomicU64, Ordering};
use std::sync::mpsc::{self, SyncSender, TrySendError};
use std::sync::Arc;
use std::thread;
use std::time::Duration;
use crate::can::{CanSocket, Frame, Recv};
use crate::clock;
pub enum Event {
Can(Frame),
/// Frames the kernel dropped before we read them (SO_RXQ_OVFL).
KernelDrops(u32),
CanLink(bool, String),
Gnss(f64, String),
GnssLink(bool, String),
Ctl(String, mpsc::Sender<String>),
}
pub fn spawn_can(iface: String, tx: SyncSender<Event>, queue_drops: Arc<AtomicU64>, stop: Arc<AtomicBool>) {
thread::Builder::new().name("can".into()).spawn(move || {
let mut reported: Option<String> = None;
while !stop.load(Ordering::Relaxed) {
let mut sock = match CanSocket::open(&iface, Duration::from_millis(200)) {
Ok(s) => s,
Err(e) => {
// An unplugged adapter is a normal state; say so once, then poll quietly.
let msg = format!("{iface}: {e}");
if reported.as_deref() != Some(&msg) {
let _ = tx.send(Event::CanLink(false, msg.clone()));
reported = Some(msg);
}
thread::sleep(Duration::from_secs(1));
continue;
}
};
let _ = tx.send(Event::CanLink(true, format!("{iface}: socket open")));
reported = None;
while !stop.load(Ordering::Relaxed) {
match sock.recv() {
Ok(Recv::Timeout) => {}
Ok(Recv::Frame(f, drops)) => {
if drops > 0 {
let _ = tx.try_send(Event::KernelDrops(drops));
}
match tx.try_send(Event::Can(f)) {
Ok(()) => {}
Err(TrySendError::Full(_)) => {
queue_drops.fetch_add(1, Ordering::Relaxed);
}
Err(TrySendError::Disconnected(_)) => return,
}
}
Err(e) => {
// ENETDOWN / ENODEV: the link went down or the adapter left.
let _ = tx.send(Event::CanLink(false, format!("{iface}: {e}")));
reported = Some(String::new());
thread::sleep(Duration::from_millis(500));
break;
}
}
}
}
}).expect("spawn can thread");
}
pub fn spawn_gnss(addr: String, tx: SyncSender<Event>, stop: Arc<AtomicBool>) {
thread::Builder::new().name("gnss".into()).spawn(move || {
// Report a failing gpsd once, not every retry.
let mut reported_down = false;
while !stop.load(Ordering::Relaxed) {
let conn = addr
.to_socket_addrs()
.ok()
.and_then(|mut a| a.next())
.ok_or_else(|| std::io::Error::other("unresolvable"))
.and_then(|a| TcpStream::connect_timeout(&a, Duration::from_secs(2)));
let mut stream = match conn {
Ok(s) => s,
Err(e) => {
if !reported_down {
let _ = tx.send(Event::GnssLink(false, format!("gpsd {addr}: {e}")));
reported_down = true;
}
thread::sleep(Duration::from_secs(3));
continue;
}
};
let _ = stream.set_read_timeout(Some(Duration::from_secs(1)));
if stream.write_all(b"?WATCH={\"enable\":true,\"json\":true};\n").is_err() {
continue;
}
let _ = tx.send(Event::GnssLink(true, format!("gpsd {addr}: watching")));
let mut reader = BufReader::new(stream);
let mut line = String::new();
loop {
if stop.load(Ordering::Relaxed) {
return;
}
match reader.read_line(&mut line) {
Ok(0) => break,
Ok(_) => {
if line.ends_with('\n') {
let rx = clock::realtime();
let _ = tx.send(Event::Gnss(rx, line.trim_end().to_string()));
line.clear();
}
}
Err(e) if matches!(e.kind(), std::io::ErrorKind::WouldBlock | std::io::ErrorKind::TimedOut) => {}
Err(_) => break,
}
}
let _ = tx.send(Event::GnssLink(false, format!("gpsd {addr}: connection lost")));
reported_down = true;
thread::sleep(Duration::from_secs(1));
}
}).expect("spawn gnss thread");
}
/// One command per connection, one reply line back.
pub fn spawn_ctl(path: String, tx: SyncSender<Event>) -> std::io::Result<()> {
let _ = std::fs::remove_file(&path);
let listener = UnixListener::bind(&path)?;
std::fs::set_permissions(&path, std::fs::Permissions::from_mode(0o660))?;
thread::Builder::new().name("ctl".into()).spawn(move || {
for conn in listener.incoming() {
let Ok(mut conn) = conn else { continue };
let _ = conn.set_read_timeout(Some(Duration::from_secs(2)));
let mut buf = Vec::new();
let _ = (&mut conn).take(4096).read_to_end(&mut buf);
let cmd = String::from_utf8_lossy(&buf).trim().to_string();
let (rtx, rrx) = mpsc::channel();
if tx.send(Event::Ctl(cmd, rtx)).is_err() {
return;
}
let reply = rrx.recv_timeout(Duration::from_secs(5)).unwrap_or_else(|_| "error: no reply".into());
let _ = conn.write_all(reply.as_bytes());
let _ = conn.write_all(b"\n");
}
})?;
Ok(())
}