fix(sway/ipc): reconnect on disconnect instead of breaking + CPU-spinning

When sway's event-subscription send buffer overflows during an event
flood, sway closes the client connection. The sway IPC event worker
(SleeperThread running handleEvent -> recv) then threw on every
iteration and the SleeperThread immediately re-invoked it, leaving the
sway modules broken while busy-looping on a dead socket and pegging a
CPU.

Mirror the niri backend's reconnect loop: on a read/EOF/parse error from
the event socket, close the old connection, back off for a couple of
seconds (so we don't busy-spin), re-open the socket and replay the same
subscriptions, then resume. A running_ flag set at the start of teardown
makes the worker bail out cleanly instead of reconnecting to a socket
that is being closed on purpose. The IPC message protocol and event
parsing are unchanged.

Fixes #3166.

Author: Alexays

include/modules/sway/ipc/client.hpp

@@ -2,10 +2,12 @@
 
 #include <sigc++/sigc++.h>
 
+#include <atomic>
 #include <cstdint>
 #include <functional>
 #include <mutex>
 #include <string>
+#include <vector>
 
 #include "ipc.hpp"
 #include "util/SafeSignal.hpp"
@@ -43,6 +45,14 @@ class Ipc {
   struct ipc_response send(int fd, uint32_t type, const std::string& payload = "");
   struct ipc_response recv(int fd);
 
+  // Re-establish the event socket and re-subscribe after sway drops us, backing
+  // off between attempts so we don't busy-loop while sway is unavailable.
+  void reconnectEvent();
+
+  std::string socketPath_;
+  std::vector<std::string> subscribed_events_;
+  std::atomic<bool> running_{true};
+
   util::ScopedFd fd_;
   util::ScopedFd fd_event_;
   std::mutex mutex_;

src/modules/sway/ipc/client.cpp

@@ -8,12 +8,14 @@
 #include <unistd.h>
 
 #include <cerrno>
+#include <chrono>
 #include <cstdio>
 #include <cstdlib>
 #include <cstring>
 #include <limits>
 #include <stdexcept>
 #include <string_view>
+#include <thread>
 #include <utility>
 
 #include "modules/sway/ipc/ipc.hpp"
@@ -41,12 +43,15 @@ void sendAll(int fd, const char* data, size_t size, const char* what) {
 }  // namespace
 
 Ipc::Ipc() {
-  const std::string socketPath = getSocketPath();
-  fd_ = util::ScopedFd(open(socketPath));
-  fd_event_ = util::ScopedFd(open(socketPath));
+  socketPath_ = getSocketPath();
+  fd_ = util::ScopedFd(open(socketPath_));
+  fd_event_ = util::ScopedFd(open(socketPath_));
 }
 
 Ipc::~Ipc() {
+  // Signal the worker before stopping it so an in-flight recv/reconnect bails
+  // out instead of trying to reconnect to a socket we're tearing down.
+  running_ = false;
   thread_.stop();
 
   if (fd_ > 0) {
@@ -191,11 +196,48 @@ void Ipc::subscribe(const std::string& payload) {
   if (res.payload != "{\"success\": true}") {
     throw std::runtime_error("Unable to subscribe ipc event");
   }
+  // Remember the subscription so we can replay it if we have to reconnect.
+  subscribed_events_.push_back(payload);
+}
+
+void Ipc::reconnectEvent() {
+  // Sway closed our event connection (typically because its send buffer filled
+  // up during an event flood). Re-establish the socket and re-subscribe to the
+  // same events, backing off between attempts so we don't busy-loop and peg a
+  // CPU while sway is unavailable or keeps dropping us.
+  while (running_) {
+    std::this_thread::sleep_for(std::chrono::seconds(2));
+    if (!running_) {
+      return;
+    }
+    try {
+      fd_event_.reset(open(socketPath_));
+      for (const auto& payload : subscribed_events_) {
+        const auto res = Ipc::send(fd_event_, IPC_SUBSCRIBE, payload);
+        if (res.payload != "{\"success\": true}") {
+          throw std::runtime_error("Unable to re-subscribe ipc event");
+        }
+      }
+      spdlog::info("Reconnected to sway IPC event socket");
+      return;
+    } catch (const std::exception& e) {
+      spdlog::warn("Failed to reconnect to sway IPC ({}), retrying", e.what());
+    }
+  }
 }
 
 void Ipc::handleEvent() {
-  const auto res = Ipc::recv(fd_event_);
-  signal_event.emit(res);
+  try {
+    const auto res = Ipc::recv(fd_event_);
+    signal_event.emit(res);
+  } catch (const std::exception& e) {
+    if (!running_) {
+      // The Ipc is being torn down; the socket was closed on purpose.
+      return;
+    }
+    spdlog::warn("Lost sway IPC event connection ({}), reconnecting", e.what());
+    reconnectEvent();
+  }
 }
 
 }  // namespace waybar::modules::sway