feat(host): KDE-reliability phase 2 — pipeline retry, graceful capture teardown, refresh reconcile

Hardens the virtual-display → capture → encode bring-up against the transient failures that surfaced as black screens / wrong refresh on cold KDE sessions. - m3: build_pipeline_with_retry wraps the initial vd.create() + first-frame with bounded exponential backoff (4 attempts, 500ms→2s). is_permanent_build_error classifies config/version/missing-tool failures so they fail fast instead of burning the retry budget. Encoder + frame clock now pace to the *achieved* refresh reported in VirtualOutput::preferred_mode, not the requested rate. - capture/linux: PortalCapturer::Drop sends a pipewire channel quit and joins the thread, so a dropped/failed/retried capturer releases its PipeWire thread + EGL/ CUDA context promptly instead of leaking it to process exit. First-frame timeout now reports the node id and distinguishes "format never negotiated" from "negotiated but no buffers arrived" via a negotiated flag set in param_changed. - vdisplay/kwin: set_custom_refresh reads back the active mode from kscreen-doctor and returns the refresh KWin actually gave us (a rejected custom mode silently leaves the output at 60Hz); create() carries it into preferred_mode. - vdisplay/gamescope: find_gamescope_node requires the Video/Source object (the node.name=gamescope tag is on two objects; the other wedges the link); a version check warns on <3.16.22 (the PipeWire-1.6 capture-deadlock signature). Live-validated against headless KWin: 720p120 build with requested=120 achieved=120, zero-copy CUDA frames, and no per-session thread accumulation across back-to-back sessions. Tests: +3 unit (retry classifier, gamescope version parse); 49 host tests green, clippy/fmt clean. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 20:03:11 +00:00
parent a4eacabecd
commit 75eb8fa0d6
4 changed files with 427 additions and 53 deletions
@@ -11,9 +11,11 @@
 //! frames leave the pipewire thread over a bounded channel. The authoritative frame size
 //! comes from the negotiated PipeWire format, not the portal's size hint.
 //!
-//! Cleanup note (M0): the two threads are detached and torn down at process exit. A
-//! graceful stop (pipewire `channel` quit + Session close) belongs with the M2 session
-//! lifecycle.
+//! Cleanup: the pipewire thread is stopped deterministically — [`PortalCapturer`]'s `Drop`
+//! sends a pipewire `channel` quit and joins the thread, so dropping a capturer (session end,
+//! or a retried/failed pipeline build) releases its EGL importer / CUDA context promptly
+//! instead of leaking it to process exit. The portal thread (when used) still parks on its zbus
+//! connection until process exit.

 use super::{CapturedFrame, Capturer, FramePayload, PixelFormat};
 use anyhow::{anyhow, Context, Result};
@@ -31,6 +33,19 @@ use std::time::Duration;
 pub struct PortalCapturer {
    frames: Receiver<CapturedFrame>,
    active: Arc<AtomicBool>,
+    /// Set true once the PipeWire stream agrees a video format. Read in [`next_frame`]'s timeout
+    /// branch to tell "format never negotiated" (modifier/format mismatch) apart from "negotiated
+    /// but no buffers arrived" (compositor idle/unmapped) — the two black-screen root causes.
+    negotiated: Arc<AtomicBool>,
+    /// The PipeWire node this capturer consumes — surfaced in error messages for diagnosis.
+    node_id: u32,
+    /// Stops the PipeWire loop on teardown (sent in `Drop`). Without it a dropped or failed
+    /// capturer leaks its PipeWire thread — and its EGL importer / CUDA context — because
+    /// `mainloop.run()` otherwise blocks until process exit. `Option` so `Drop` can take it.
+    quit: Option<::pipewire::channel::Sender<()>>,
+    /// Joined in `Drop` (after `quit`) so teardown is synchronous: the importer/CUDA context is
+    /// released before the next pipeline builds, not left racing it.
+    join: Option<thread::JoinHandle<()>>,
    /// Owns the virtual output (if this capturer was built from one) — dropped when the capturer
    /// is, releasing the compositor-side output via the keepalive's own `Drop`. `None` for the
    /// portal source (its session ends with the portal thread's zbus connection).
@@ -64,12 +79,7 @@ impl PortalCapturer {
            node_id,
            "ScreenCast portal session started; connecting PipeWire"
        );
-        let (frames, active) = spawn_pipewire(Some(fd), node_id, None)?;
-        Ok(PortalCapturer {
-            frames,
-            active,
-            _keepalive: None,
-        })
+        Ok(spawn_pipewire(Some(fd), node_id, None)?.into_capturer(node_id, None))
    }

    /// Build a capturer from an already-created virtual output ([`crate::vdisplay::VirtualOutput`]):
@@ -81,40 +91,84 @@ impl PortalCapturer {
            node_id = vout.node_id,
            "connecting PipeWire to virtual output"
        );
-        let (frames, active) = spawn_pipewire(vout.remote_fd, vout.node_id, vout.preferred_mode)?;
-        Ok(PortalCapturer {
-            frames,
-            active,
-            _keepalive: Some(vout.keepalive),
-        })
+        let node_id = vout.node_id;
+        Ok(
+            spawn_pipewire(vout.remote_fd, node_id, vout.preferred_mode)?
+                .into_capturer(node_id, Some(vout.keepalive)),
+        )
+    }
+}
+
+/// Live PipeWire-thread handles returned by [`spawn_pipewire`]: the frame channel, the
+/// activation flag the per-frame copy gates on, a "format negotiated" flag (timeout diagnostics),
+/// a quit sender that stops the loop, and the thread's join handle (synchronous teardown).
+struct PwHandles {
+    frames: Receiver<CapturedFrame>,
+    active: Arc<AtomicBool>,
+    negotiated: Arc<AtomicBool>,
+    quit: ::pipewire::channel::Sender<()>,
+    join: thread::JoinHandle<()>,
+}
+
+impl PwHandles {
+    /// Assemble a [`PortalCapturer`] around these handles. `node_id` is carried for diagnostics;
+    /// `keepalive` owns the virtual output (drops after the PipeWire thread is joined).
+    fn into_capturer(self, node_id: u32, keepalive: Option<Box<dyn Send>>) -> PortalCapturer {
+        PortalCapturer {
+            frames: self.frames,
+            active: self.active,
+            negotiated: self.negotiated,
+            node_id,
+            quit: Some(self.quit),
+            join: Some(self.join),
+            _keepalive: keepalive,
+        }
    }
 }

 /// Spawn the PipeWire consumer thread for `node_id` (fd `Some` = portal remote, `None` =
-/// default daemon) and return the frame channel + the activation flag it gates on.
-/// `preferred` seeds the format negotiation's default size/framerate — for Mutter virtual
-/// monitors this is what actually sizes the monitor.
+/// default daemon) and return its [`PwHandles`]. `preferred` seeds the format negotiation's
+/// default size/framerate — for Mutter virtual monitors this is what actually sizes the monitor.
 fn spawn_pipewire(
    fd: Option<OwnedFd>,
    node_id: u32,
    preferred: Option<(u32, u32, u32)>,
-) -> Result<(Receiver<CapturedFrame>, Arc<AtomicBool>)> {
+) -> Result<PwHandles> {
    // Frames flow from the pipewire thread over a small bounded channel.
    let (frame_tx, frame_rx) = sync_channel::<CapturedFrame>(8);
    let active = Arc::new(AtomicBool::new(false));
    let active_cb = active.clone();
+    let negotiated = Arc::new(AtomicBool::new(false));
+    let negotiated_cb = negotiated.clone();
+    // pipewire's own cross-thread channel: the receiver attaches to the loop and quits it; the
+    // sender lives on the capturer and fires in its `Drop`. Absolute `::pipewire` path — the
+    // inner `mod pipewire` shadows the crate name at this scope.
+    let (quit_tx, quit_rx) = ::pipewire::channel::channel::<()>();
    let zerocopy = crate::zerocopy::enabled();
-    thread::Builder::new()
+    let join = thread::Builder::new()
        .name("punktfunk-pipewire".into())
        .spawn(move || {
-            if let Err(e) =
-                pipewire::pipewire_thread(fd, node_id, frame_tx, active_cb, zerocopy, preferred)
-            {
+            if let Err(e) = pipewire::pipewire_thread(
+                fd,
+                node_id,
+                frame_tx,
+                active_cb,
+                negotiated_cb,
+                zerocopy,
+                preferred,
+                quit_rx,
+            ) {
                tracing::error!(error = %format!("{e:#}"), "pipewire capture thread failed");
            }
        })
        .context("spawn pipewire thread")?;
-    Ok((frame_rx, active))
+    Ok(PwHandles {
+        frames: frame_rx,
+        active,
+        negotiated,
+        quit: quit_tx,
+        join,
+    })
 }

 impl Capturer for PortalCapturer {
@@ -122,8 +176,30 @@ impl Capturer for PortalCapturer {
        // First frame can lag behind format negotiation; later frames arrive at ~fps.
        match self.frames.recv_timeout(Duration::from_secs(10)) {
            Ok(frame) => Ok(frame),
-            Err(RecvTimeoutError::Timeout) => Err(anyhow!("no PipeWire frame within 10s")),
-            Err(RecvTimeoutError::Disconnected) => Err(anyhow!("PipeWire capture thread ended")),
+            Err(RecvTimeoutError::Timeout) => {
+                // Split the two black-screen root causes apart so the operator gets a cause, not
+                // just a symptom: did the format negotiate (compositor produced no buffers) or
+                // not (no acceptable format / node never emitted a param)?
+                if self.negotiated.load(Ordering::Relaxed) {
+                    Err(anyhow!(
+                        "no PipeWire frame within 10s (node {}): format negotiated but no buffers \
+                         arrived — the compositor produced no frames (virtual output idle/unmapped, \
+                         or capture never started)",
+                        self.node_id
+                    ))
+                } else {
+                    Err(anyhow!(
+                        "no PipeWire frame within 10s (node {}): format negotiation never \
+                         completed — the compositor offered no format this consumer accepts \
+                         (pixel-format/modifier mismatch) or the node never emitted a Format param",
+                        self.node_id
+                    ))
+                }
+            }
+            Err(RecvTimeoutError::Disconnected) => Err(anyhow!(
+                "PipeWire capture thread ended before a frame (node {})",
+                self.node_id
+            )),
        }
    }

@@ -148,6 +224,22 @@ impl Capturer for PortalCapturer {
    }
 }

+impl Drop for PortalCapturer {
+    fn drop(&mut self) {
+        // Stop the PipeWire loop and wait for the thread to unwind BEFORE the keepalive (virtual
+        // output) drops: quit → join releases the EGL importer / CUDA context, then field-drop
+        // order releases the output. Without this, `mainloop.run()` blocks until process exit, so
+        // every dropped/failed capturer (e.g. a retried first-frame attempt) leaks a thread + GPU
+        // context. `send` errors only if the thread already exited — then `join` returns at once.
+        if let Some(quit) = self.quit.take() {
+            let _ = quit.send(());
+        }
+        if let Some(join) = self.join.take() {
+            let _ = join.join();
+        }
+    }
+}
+
 /// The portal handshake: connect ScreenCast, select a single monitor, start, open the
 /// PipeWire remote, hand the fd + node id back, then keep the session alive.
 fn portal_thread(setup_tx: std::sync::mpsc::Sender<Result<(OwnedFd, u32), String>>) {
@@ -369,6 +461,9 @@ mod pipewire {
        tx: SyncSender<CapturedFrame>,
        /// When false (no active stream), skip the de-pad copy — the buffer is just released.
        active: Arc<AtomicBool>,
+        /// Set once a video format is agreed (`param_changed`), so a first-frame timeout can tell
+        /// "format never negotiated" apart from "negotiated but no buffers arrived".
+        negotiated: Arc<AtomicBool>,
        /// Present when zero-copy is enabled: imports a dmabuf → CUDA device buffer.
        importer: Option<crate::zerocopy::EglImporter>,
    }
@@ -583,17 +678,28 @@ mod pipewire {
        })
    }

+    #[allow(clippy::too_many_arguments)]
    pub fn pipewire_thread(
        fd: Option<OwnedFd>,
        node_id: u32,
        tx: SyncSender<CapturedFrame>,
        active: Arc<AtomicBool>,
+        negotiated: Arc<AtomicBool>,
        zerocopy: bool,
        preferred: Option<(u32, u32, u32)>,
+        quit_rx: pw::channel::Receiver<()>,
    ) -> Result<()> {
        crate::pwinit::ensure_init();

        let mainloop = pw::main_loop::MainLoopRc::new(None).context("pw MainLoop")?;
+        // A quit signal (capturer `Drop`) lands here on the loop thread and stops `run()` so the
+        // thread unwinds instead of blocking to process exit. Hold the attachment for the loop's
+        // life; the cloned loop handle is the one the callback quits.
+        let quit_loop = mainloop.clone();
+        let _quit_attach = quit_rx.attach(mainloop.loop_(), move |()| {
+            tracing::debug!("pipewire: quit signal received — stopping capture loop");
+            quit_loop.quit();
+        });
        let context = pw::context::ContextRc::new(&mainloop, None).context("pw Context")?;
        // A portal source hands us an fd to a (sandboxed) PipeWire remote; the KWin
        // virtual-output source has no fd — its node lives on the user's default daemon.
@@ -647,6 +753,7 @@ mod pipewire {
            modifier: 0,
            tx,
            active,
+            negotiated,
            importer,
        };

@@ -687,6 +794,7 @@ mod pipewire {
                    return;
                }
                if ud.info.parse(param).is_ok() {
+                    ud.negotiated.store(true, Ordering::Relaxed);
                    let sz = ud.info.size();
                    ud.format = map_format(ud.info.format());
                    ud.modifier = ud.info.modifier();
@@ -822,7 +822,8 @@ fn virtual_stream(
    let compositor = crate::vdisplay::detect().context("detect compositor")?;
    tracing::info!(?compositor, ?mode, "punktfunk/1 virtual display");
    let mut vd = crate::vdisplay::open(compositor)?;
-    let (mut capturer, mut enc, mut frame, mut interval) = build_pipeline(&mut vd, mode)?;
+    let (mut capturer, mut enc, mut frame, mut interval) =
+        build_pipeline_with_retry(&mut vd, mode)?;

    let deadline = std::time::Instant::now() + std::time::Duration::from_secs(seconds as u64);
    let mut next = std::time::Instant::now();
@@ -885,11 +886,98 @@ type Pipeline = (
    std::time::Duration,
 );

+/// Build the pipeline, retrying *transient* failures with bounded exponential backoff.
+///
+/// Bringing a virtual output to first-frame races several async steps — the compositor parenting
+/// the output, the portal/RemoteDesktop grant, PipeWire format negotiation — any of which can
+/// momentarily time out on a cold session. A single timed-out attempt shouldn't abort the whole
+/// punktfunk/1 session. But a *permanent* failure (unsupported compositor/mode, a KWin too old to
+/// create virtual outputs, a missing tool) must fail fast instead of burning the budget — so the
+/// error chain is classified and permanent ones short-circuit. Each failed attempt drops its
+/// capturer, which (via `PortalCapturer::Drop`) tears the PipeWire thread + virtual output down
+/// before the next attempt — no leak across retries.
+fn build_pipeline_with_retry(
+    vd: &mut Box<dyn crate::vdisplay::VirtualDisplay>,
+    mode: punktfunk_core::Mode,
+) -> Result<Pipeline> {
+    const MAX_ATTEMPTS: u32 = 4;
+    let mut backoff = std::time::Duration::from_millis(500);
+    for attempt in 1..=MAX_ATTEMPTS {
+        match build_pipeline(vd, mode) {
+            Ok(pipe) => {
+                if attempt > 1 {
+                    tracing::info!(attempt, "pipeline up after retry");
+                }
+                return Ok(pipe);
+            }
+            Err(e) => {
+                let chain = format!("{e:#}");
+                let permanent = is_permanent_build_error(&chain);
+                if permanent || attempt == MAX_ATTEMPTS {
+                    let why = if permanent {
+                        "permanent"
+                    } else {
+                        "out of retries"
+                    };
+                    return Err(e).with_context(|| {
+                        format!("pipeline build failed ({why}) after {attempt} attempt(s)")
+                    });
+                }
+                tracing::warn!(
+                    attempt,
+                    max = MAX_ATTEMPTS,
+                    backoff_ms = backoff.as_millis() as u64,
+                    error = %chain,
+                    "pipeline build failed — retrying"
+                );
+                std::thread::sleep(backoff);
+                backoff = (backoff * 2).min(std::time::Duration::from_secs(2));
+            }
+        }
+    }
+    unreachable!("the final attempt returns inside the loop")
+}
+
+/// Is a pipeline-build error permanent (retrying won't help within this session)? Matches the
+/// error chain against signatures that don't change between attempts: unsupported compositor or
+/// mode, a KWin too old to expose virtual outputs, a missing/unparseable config, a tool that
+/// isn't installed. Everything else — portal/PipeWire negotiation timeouts, "no frame within
+/// 10s", transient node races — is treated as transient and retried. Biased toward "transient":
+/// a misjudged permanent error only costs a few seconds before it fails anyway.
+fn is_permanent_build_error(chain: &str) -> bool {
+    const PERMANENT: &[&str] = &[
+        "virtual displays require linux",
+        "unknown punktfunk_compositor",
+        "could not detect compositor",
+        "could not find output", // KWin < 6.5.6: createVirtualOutput unsupported
+        "must be a node id",     // PUNKTFUNK_GAMESCOPE_NODE not an integer
+        "is it installed",       // gamescope / kscreen-doctor not on PATH
+    ];
+    let lower = chain.to_ascii_lowercase();
+    PERMANENT.iter().any(|p| lower.contains(p))
+}
+
 fn build_pipeline(
    vd: &mut Box<dyn crate::vdisplay::VirtualDisplay>,
    mode: punktfunk_core::Mode,
 ) -> Result<Pipeline> {
    let vout = vd.create(mode).context("create virtual output")?;
+    // The backend reports the refresh it actually achieved in `preferred_mode.2` (KWin may cap a
+    // virtual output at 60 Hz if the custom-mode install was rejected). Pace the encoder + frame
+    // clock to that, not the requested rate, so we don't emit phantom duplicate frames over a
+    // slower source. Falls back to the requested rate when a backend reports nothing.
+    let effective_hz = vout
+        .preferred_mode
+        .map(|(_, _, hz)| hz)
+        .filter(|&hz| hz > 0)
+        .unwrap_or(mode.refresh_hz);
+    if effective_hz != mode.refresh_hz {
+        tracing::warn!(
+            requested = mode.refresh_hz,
+            effective = effective_hz,
+            "compositor did not honor the requested refresh — encoding at the achieved rate"
+        );
+    }
    let mut capturer =
        crate::capture::capture_virtual_output(vout).context("capture virtual output")?;
    capturer.set_active(true);
@@ -899,12 +987,12 @@ fn build_pipeline(
        frame.format,
        frame.width,
        frame.height,
-        mode.refresh_hz,
+        effective_hz,
        20_000_000,
        frame.is_cuda(),
    )
    .context("open NVENC")?;
-    let interval = std::time::Duration::from_secs_f64(1.0 / mode.refresh_hz.max(1) as f64);
+    let interval = std::time::Duration::from_secs_f64(1.0 / effective_hz.max(1) as f64);
    Ok((capturer, enc, frame, interval))
 }

@@ -912,6 +1000,29 @@ fn build_pipeline(
 mod tests {
    use super::*;

+    #[test]
+    fn permanent_errors_short_circuit_retry() {
+        // Permanent: config / version / missing-tool — retrying within a session can't fix these.
+        assert!(is_permanent_build_error(
+            "create virtual output: KWin virtual output failed: Could not find output"
+        ));
+        assert!(is_permanent_build_error(
+            "unknown PUNKTFUNK_COMPOSITOR 'foo' (kwin|wlroots|mutter|gamescope)"
+        ));
+        assert!(is_permanent_build_error(
+            "spawn gamescope (is it installed? `apt install gamescope`)"
+        ));
+        assert!(is_permanent_build_error("virtual displays require Linux"));
+        // Transient: negotiation/timeout races — exactly what backoff is for.
+        assert!(!is_permanent_build_error(
+            "first frame: no PipeWire frame within 10s (node 42): format negotiation never completed"
+        ));
+        assert!(!is_permanent_build_error(
+            "create virtual output: timed out creating the KWin virtual output"
+        ));
+        assert!(!is_permanent_build_error("open NVENC: device busy"));
+    }
+
    fn gp(kind: InputKind, code: u32, x: i32, pad: u32) -> InputEvent {
        InputEvent {
            kind,
@@ -48,6 +48,7 @@ impl VirtualDisplay for GamescopeDisplay {
                keepalive: Box::new(()),
            });
        }
+        check_gamescope_version(); // diagnostic only — warns on known-deadlock-prone versions
        let proc = GamescopeProc(spawn(mode.width, mode.height, mode.refresh_hz.max(1))?);
        // gamescope creates its PipeWire node a moment after start; poll for it (the proc is held
        // alive meanwhile, and killed if we give up).
@@ -147,24 +148,92 @@ fn node_from_log() -> Option<u32> {
 }

 /// Find the `gamescope` `Video/Source` node id in a `pw-dump` snapshot of the default daemon.
+///
+/// `node.name=gamescope` appears on TWO objects (the adapter *and* the inner stream node); only
+/// the one whose `media.class` is `Video/Source` is a valid capture target — connecting to the
+/// other wedges the link. So we require `Video/Source` first and fall back to a bare name match
+/// only if no class-tagged node is present (older gamescope that doesn't set media.class).
 fn find_gamescope_node() -> Option<u32> {
    let out = Command::new("pw-dump").output().ok()?;
    let dump: serde_json::Value = serde_json::from_slice(&out.stdout).ok()?;
-    for obj in dump.as_array()? {
+    let nodes = dump.as_array()?;
+    let node_props = |obj: &serde_json::Value| -> Option<(u32, String, String)> {
        if obj.get("type").and_then(|t| t.as_str()) != Some("PipeWire:Interface:Node") {
-            continue;
+            return None;
        }
+        let id = obj.get("id").and_then(|i| i.as_u64())? as u32;
        let props = obj.get("info").and_then(|i| i.get("props"));
        let name = props
            .and_then(|p| p.get("node.name"))
            .and_then(|n| n.as_str())
-            .unwrap_or("");
+            .unwrap_or("")
+            .to_string();
        let class = props
            .and_then(|p| p.get("media.class"))
            .and_then(|n| n.as_str())
-            .unwrap_or("");
-        if name == "gamescope" || (class == "Video/Source" && name.contains("gamescope")) {
-            return obj.get("id").and_then(|i| i.as_u64()).map(|x| x as u32);
+            .unwrap_or("")
+            .to_string();
+        Some((id, name, class))
+    };
+    // Preferred: a Video/Source node named (or containing) "gamescope".
+    for obj in nodes {
+        if let Some((id, name, class)) = node_props(obj) {
+            if class == "Video/Source" && (name == "gamescope" || name.contains("gamescope")) {
+                return Some(id);
+            }
+        }
+    }
+    // Fallback: a node literally named "gamescope" with no usable class tag.
+    for obj in nodes {
+        if let Some((id, name, _)) = node_props(obj) {
+            if name == "gamescope" {
+                tracing::warn!(
+                    node_id = id,
+                    "gamescope node has no media.class=Video/Source tag — capturing it anyway"
+                );
+                return Some(id);
+            }
+        }
+    }
+    None
+}
+
+/// Minimum gamescope that captures reliably: below 3.16.22, headless PipeWire capture deadlocks
+/// against PipeWire ≥ 1.6 (a loop-lock bug) and a stuck link head-blocks the whole daemon.
+const MIN_GAMESCOPE: (u32, u32, u32) = (3, 16, 22);
+
+/// Best-effort: warn loudly if the installed gamescope is older than [`MIN_GAMESCOPE`]. Parsing
+/// failures are silent (don't block a possibly-fine custom build) — this is a diagnostic, not a
+/// gate. Returns the parsed version when it could read one.
+fn check_gamescope_version() -> Option<(u32, u32, u32)> {
+    let out = Command::new("gamescope").arg("--version").output().ok()?;
+    // gamescope prints the version banner to stderr on some builds, stdout on others.
+    let text = format!(
+        "{}{}",
+        String::from_utf8_lossy(&out.stdout),
+        String::from_utf8_lossy(&out.stderr)
+    );
+    let ver = parse_version(&text)?;
+    if ver < MIN_GAMESCOPE {
+        tracing::warn!(
+            found = %format!("{}.{}.{}", ver.0, ver.1, ver.2),
+            min = %format!("{}.{}.{}", MIN_GAMESCOPE.0, MIN_GAMESCOPE.1, MIN_GAMESCOPE.2),
+            "gamescope is older than the minimum for reliable headless capture — expect a \
+             capture deadlock against PipeWire ≥ 1.6 (a wedged link head-blocks the daemon); \
+             upgrade gamescope or use PUNKTFUNK_COMPOSITOR=kwin|mutter"
+        );
+    }
+    Some(ver)
+}
+
+/// Extract the first `X.Y.Z` version triple from arbitrary text (e.g. `gamescope version 3.16.22`).
+fn parse_version(text: &str) -> Option<(u32, u32, u32)> {
+    for token in text.split(|c: char| !(c.is_ascii_digit() || c == '.')) {
+        let mut parts = token.split('.');
+        let (a, b, c) = (parts.next()?, parts.next(), parts.next());
+        let (Some(b), Some(c)) = (b, c) else { continue };
+        if let (Ok(a), Ok(b), Ok(c)) = (a.parse(), b.parse(), c.parse()) {
+            return Some((a, b, c));
        }
    }
    None
@@ -179,3 +248,28 @@ impl Drop for GamescopeProc {
        let _ = self.0.wait();
    }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::{parse_version, MIN_GAMESCOPE};
+
+    #[test]
+    fn parses_version_banner() {
+        assert_eq!(parse_version("gamescope version 3.16.22"), Some((3, 16, 22)));
+        assert_eq!(
+            parse_version("gamescope: version v3.15.9 (no PipeWire)"),
+            Some((3, 15, 9))
+        );
+        assert_eq!(parse_version("3.16.20-1.fc41"), Some((3, 16, 20)));
+        assert_eq!(parse_version("no version here"), None);
+        assert_eq!(parse_version("only 3.16 here"), None); // needs a full triple
+    }
+
+    #[test]
+    fn flags_known_bad_versions() {
+        // The 26.04-shipped 3.16.20 is below the minimum (PipeWire 1.6 deadlock).
+        assert!(parse_version("gamescope version 3.16.20").unwrap() < MIN_GAMESCOPE);
+        assert!(parse_version("gamescope version 3.16.22").unwrap() >= MIN_GAMESCOPE);
+        assert!(parse_version("gamescope version 3.17.0").unwrap() >= MIN_GAMESCOPE);
+    }
+}
@@ -91,17 +91,22 @@ impl VirtualDisplay for KwinDisplay {
        };
        tracing::info!(node_id, width, height, "KWin virtual output ready");
        // KWin creates virtual outputs at a hardcoded 60 Hz and `stream_virtual_output` has no
-        // refresh argument, so when the client wants more we install + select a custom mode
-        // (supported on virtual outputs since KWin 6.6). Done before capture connects PipeWire so
-        // the stream negotiates at the higher rate. First cut shells out to kscreen-doctor; the
-        // in-process kde_output_management_v2 client is a follow-up.
-        if mode.refresh_hz > 60 {
-            set_custom_refresh(width, height, mode.refresh_hz);
-        }
+        // refresh argument, so above 60 Hz we install + select a custom mode (supported on virtual
+        // outputs since KWin 6.6) before capture connects PipeWire, so the stream negotiates at the
+        // higher rate. First cut shells out to kscreen-doctor; the in-process
+        // kde_output_management_v2 client is a follow-up. `set_custom_refresh` reads back and
+        // returns what KWin *actually* achieved so the encoder paces to the real source rate (a
+        // rejected custom mode leaves the output at 60 Hz). At ≤60 Hz there's nothing to install —
+        // the source runs 60 Hz and the encoder downsamples — so carry the requested rate through.
+        let achieved_hz = if mode.refresh_hz > 60 {
+            set_custom_refresh(width, height, mode.refresh_hz)
+        } else {
+            mode.refresh_hz
+        };
        Ok(VirtualOutput {
            node_id,
            remote_fd: None,
-            preferred_mode: Some((mode.width, mode.height, mode.refresh_hz)),
+            preferred_mode: Some((mode.width, mode.height, achieved_hz)),
            keepalive: Box::new(StopGuard(stop)),
        })
    }
@@ -109,8 +114,11 @@ impl VirtualDisplay for KwinDisplay {

 /// Best-effort: raise the just-created virtual output's refresh above KWin's default 60 Hz by
 /// installing + selecting a custom mode via `kscreen-doctor` (the output is `Virtual-<VOUT_NAME>`,
-/// refresh given in mHz). Failure leaves the source at 60 Hz — the stream still works, just capped.
-fn set_custom_refresh(width: u32, height: u32, hz: u32) {
+/// refresh given in mHz), then **read back the active mode** and return the refresh KWin actually
+/// gave us. The apply command can report success yet leave the output at 60 Hz (mode rejected),
+/// and a silent rate mismatch surfaces downstream as judder / duplicated frames — so the caller
+/// paces the encoder to the *achieved* rate, not the requested one.
+fn set_custom_refresh(width: u32, height: u32, hz: u32) -> u32 {
    let output = format!("Virtual-{VOUT_NAME}");
    let mhz = hz.saturating_mul(1000);
    let run = |arg: String| {
@@ -124,17 +132,70 @@ fn set_custom_refresh(width: u32, height: u32, hz: u32) {
    let _ = run(format!(
        "output.{output}.addCustomMode.{width}.{height}.{mhz}.full"
    ));
-    if run(format!("output.{output}.mode.{width}x{height}@{hz}")) {
-        tracing::info!(output, hz, "KWin virtual output: custom refresh applied");
-    } else {
-        tracing::warn!(
-            output,
-            hz,
-            "kscreen-doctor refresh set failed — source stays 60 Hz (is kscreen-doctor installed?)"
-        );
+    let applied = run(format!("output.{output}.mode.{width}x{height}@{hz}"));
+    match read_active_refresh(&output) {
+        Some(achieved) if achieved >= hz => {
+            tracing::info!(
+                output,
+                requested = hz,
+                achieved,
+                "KWin virtual output: custom refresh applied"
+            );
+            achieved
+        }
+        Some(achieved) => {
+            tracing::warn!(
+                output,
+                requested = hz,
+                achieved,
+                applied,
+                "KWin virtual output refresh below requested — pacing the encoder to the achieved \
+                 rate (custom-mode install rejected? is kscreen-doctor up to date?)"
+            );
+            achieved.max(1)
+        }
+        None => {
+            tracing::warn!(
+                output,
+                requested = hz,
+                applied,
+                "could not read back KWin virtual output refresh — assuming 60 Hz (is \
+                 kscreen-doctor installed?)"
+            );
+            60
+        }
    }
 }

+/// Read the active refresh (Hz, rounded) of `output` from `kscreen-doctor -j`. `None` if the
+/// tool, the output, or its current mode can't be found. Mode/output ids come through as either
+/// JSON strings or numbers depending on the KWin version, so both are accepted.
+fn read_active_refresh(output: &str) -> Option<u32> {
+    let out = std::process::Command::new("kscreen-doctor")
+        .arg("-j")
+        .output()
+        .ok()?;
+    let doc: serde_json::Value = serde_json::from_slice(&out.stdout).ok()?;
+    let as_id = |v: &serde_json::Value| -> Option<String> {
+        v.as_str()
+            .map(|s| s.to_string())
+            .or_else(|| v.as_u64().map(|n| n.to_string()))
+    };
+    let o = doc
+        .get("outputs")?
+        .as_array()?
+        .iter()
+        .find(|o| o.get("name").and_then(|n| n.as_str()) == Some(output))?;
+    let current = o.get("currentModeId").and_then(as_id)?;
+    let mode = o
+        .get("modes")?
+        .as_array()?
+        .iter()
+        .find(|m| m.get("id").and_then(as_id).as_deref() == Some(current.as_str()))?;
+    let hz = mode.get("refreshRate").and_then(|r| r.as_f64())?;
+    Some(hz.round() as u32)
+}
+
 /// Dropping this releases the KWin virtual output: it flips the keepalive thread's `stop`, which
 /// drops the Wayland connection and makes KWin reclaim the output.
 struct StopGuard(Arc<AtomicBool>);