Files
punktfunk/crates/pf-vdisplay/src/vdisplay/proc.rs
T
enricobuehlerandClaude Fable 5 ac3dc4323f fix(native/session): make the stop flag enforceable so a session can't outlive its client
Native sessions could survive long after the client was gone, in two
independent ways.

HOST: `stop` was only advisory. The one thing that ends a session is the
stream thread returning — every teardown (conn.close, the joins, and the
RAII drops of the session permit, admission entry and stream marker)
sits after that await, and nothing forced the issue. The encode loop
checks `stop` between iterations, so any unbounded call INSIDE one never
reaches the check, and one stuck syscall became a permanent zombie: it
held its semaphore slot (four of those and the host stops accepting
QUIC entirely), its admission entry (a later client gets "host busy"
forever), and not even the console's Stop button could clear it — that
button sets this same flag.

  * Bound the wait: once the session has been told to stop, the thread
    gets STREAM_STOP_GRACE (90 s, well past the 40 s capture-rebuild
    budget) to return, then teardown runs anyway. The thread is detached,
    not killed — Rust can't cancel a blocking thread — so it keeps its
    capturer/encoder until the stuck call returns, but the session's slot
    and admission entry come back and the host keeps serving. It logs at
    ERROR as the host wedge it is.
  * Bound the audio/input joins too — the last unbounded await in
    teardown.
  * Take the session permit AFTER the QUIC handshake instead of before
    `accept()`, so a host at its concurrency cap still accepts and the
    waiting client sees a live path instead of a silent dial timeout.
  * Bound the compositor helpers that caused the wedge in the first
    place: new pf-vdisplay `proc::{status_within, output_within}` kill a
    child that outlives its budget. `kscreen-doctor` is a Wayland client
    of the very compositor it configures, so against a wedged KWin it
    never returned; same for systemctl/dbus against a stuck session bus.

CLIENTS: the connection was never closed, so the host was right to keep
the session — it still had a live, keep-alive-answering peer.

  * Android: backgrounding did no teardown at all, and Android doesn't
    suspend the process, so the worker kept answering keep-alives until
    the OS reclaimed it (on a TV box, never). End the session on ON_STOP,
    via the existing onDispose path; a plain close, not a quit, so the
    host lingers the display for a fast return.
  * Apple: the .background arm was iOS-only AND gated on an opt-in that
    defaults off, so backgrounding did nothing — while the `audio`
    background mode kept the app (and its connection) alive indefinitely.
    Act unconditionally, and cover tvOS.
  * Core: `conn.close()` only queues the frame, and run_pump is the body
    of a block_on whose runtime is dropped the instant it returns, so the
    driver could never put it on the wire — a deliberate quit reached the
    host as silence (8 s idle timeout, no quit code, and the linger meant
    for an unwanted disconnect). Carry the endpoint out of the handshake
    and flush with wait_idle(), the same discipline the pairing and probe
    paths already use.

Linux check/clippy/tests green: 262 host, 71 pf-vdisplay (incl. new
bounded-process tests), 231 core.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-24 02:11:49 +02:00

115 lines
4.8 KiB
Rust

//! Time-bounded child-process helpers.
//!
//! Every compositor query this crate makes shells out to a helper (`kscreen-doctor`, `systemctl`,
//! `pw-dump`, …), and most of them are *clients of the very thing being diagnosed*: `kscreen-doctor`
//! is a Wayland client, so against a wedged KWin it blocks in its own connect and **never returns**.
//! `Command::status()` / `Command::output()` have no timeout, so one hung helper pinned the calling
//! thread forever — and on the host that thread is the session's stream thread, whose only way to
//! end a session is to return. A stuck query therefore became a permanently stuck session.
//!
//! These wrappers bound the wait: poll for exit until the budget runs out, then kill the child and
//! report [`std::io::ErrorKind::TimedOut`], so callers see a plain "the helper failed" error and
//! take their existing failure path instead of hanging.
use std::io::{Error, ErrorKind, Result};
use std::process::{Command, ExitStatus, Output};
use std::time::{Duration, Instant};
/// Poll interval while waiting for a child to exit. Short enough that a fast helper (the normal
/// case — `kscreen-doctor` answers in tens of ms) isn't measurably delayed.
const POLL: Duration = Duration::from_millis(20);
/// Run `cmd` to completion, killing it if it outlives `budget`.
///
/// Stdout/stderr are left as the caller configured them (inherited by default), so this is for
/// commands run for their exit status alone — see [`output_within`] when the output is read.
pub(crate) fn status_within(cmd: &mut Command, budget: Duration) -> Result<ExitStatus> {
let mut child = cmd.spawn()?;
let deadline = Instant::now() + budget;
loop {
match child.try_wait()? {
Some(status) => return Ok(status),
None if Instant::now() >= deadline => {
let _ = child.kill();
let _ = child.wait(); // reap it — never leave a zombie behind
return Err(timed_out(cmd, budget));
}
None => std::thread::sleep(POLL),
}
}
}
/// Run `cmd` to completion and capture its stdout/stderr, killing it if it outlives `budget`.
///
/// The output is read only after the child has exited, so a helper that fills the pipe buffer and
/// stalls is caught by the budget rather than deadlocking the reader (these helpers emit at most a
/// few hundred KiB, well under any real pipe pressure).
pub(crate) fn output_within(cmd: &mut Command, budget: Duration) -> Result<Output> {
let mut child = cmd
.stdout(std::process::Stdio::piped())
.stderr(std::process::Stdio::piped())
.spawn()?;
let deadline = Instant::now() + budget;
loop {
match child.try_wait()? {
// Exited: `wait_with_output` now only drains already-buffered pipes.
Some(_) => return child.wait_with_output(),
None if Instant::now() >= deadline => {
let _ = child.kill();
let _ = child.wait();
return Err(timed_out(cmd, budget));
}
None => std::thread::sleep(POLL),
}
}
}
fn timed_out(cmd: &Command, budget: Duration) -> Error {
let program = cmd.get_program().to_string_lossy().to_string();
tracing::warn!(
program,
budget_ms = budget.as_millis() as u64,
"helper did not exit within its budget — killed it (a wedged compositor/session bus is the \
usual cause); treating it as a failed query"
);
Error::new(
ErrorKind::TimedOut,
format!("`{program}` did not exit within {budget:?}"),
)
}
#[cfg(test)]
mod tests {
use super::*;
/// A helper that never exits must be killed at the budget and reported as `TimedOut` — the
/// whole point of the module (an unbounded `status()` here is what wedged a whole session).
#[test]
fn a_hung_child_is_killed_at_the_budget() {
let started = Instant::now();
let err = status_within(Command::new("sleep").arg("30"), Duration::from_millis(150))
.expect_err("must time out");
assert_eq!(err.kind(), ErrorKind::TimedOut);
assert!(
started.elapsed() < Duration::from_secs(5),
"must return at its budget, not the child's lifetime (took {:?})",
started.elapsed()
);
}
/// The normal path is unaffected: a quick command still yields its status and its output.
#[test]
fn a_quick_child_returns_normally() {
let st = status_within(&mut Command::new("true"), Duration::from_secs(5)).expect("ran");
assert!(st.success());
let out = output_within(
Command::new("echo").arg("punktfunk"),
Duration::from_secs(5),
)
.expect("ran");
assert!(out.status.success());
assert_eq!(String::from_utf8_lossy(&out.stdout).trim(), "punktfunk");
}
}