Fix TUI /fork shutdown routing

Co-authored-by: Codex <noreply@openai.com>
This commit is contained in:
Charles Cunningham
2026-03-24 18:48:17 -07:00
co-authored by Codex
parent 3f638d4ac9
commit d85f889894
+73 -1
View File
@@ -4416,8 +4416,14 @@ impl App {
// This guard is only for intentional thread-switch shutdowns.
// App-exit shutdowns are tracked by `pending_shutdown_exit_thread_id`
// and resolved in `handle_active_thread_event`.
//
// For primary-thread swaps such as `/fork` and `/resume`, the old
// thread's `ShutdownComplete` can arrive after we've already replaced
// the chat widget. Those events do not carry a source thread id, so the
// thread id stored in this guard is advisory only; once armed, consume
// the next shutdown completion unconditionally.
if matches!(event.msg, EventMsg::ShutdownComplete)
&& self.suppress_shutdown_complete_thread_id == self.current_displayed_thread_id()
&& self.suppress_shutdown_complete_thread_id.is_some()
{
self.suppress_shutdown_complete_thread_id = None;
return;
@@ -5717,6 +5723,72 @@ mod tests {
);
}
#[tokio::test]
async fn suppressed_shutdown_complete_after_replacing_chat_widget_does_not_exit() {
let (mut app, mut app_event_rx, _op_rx) = make_test_app_with_channels().await;
let old_thread_id = ThreadId::new();
let new_thread_id = ThreadId::new();
app.chat_widget.handle_codex_event(Event {
id: "old-session".to_string(),
msg: EventMsg::SessionConfigured(SessionConfiguredEvent {
session_id: old_thread_id,
forked_from_id: None,
thread_name: None,
model: "gpt-test".to_string(),
model_provider_id: "test-provider".to_string(),
service_tier: None,
approval_policy: AskForApproval::Never,
approvals_reviewer: ApprovalsReviewer::User,
sandbox_policy: SandboxPolicy::new_read_only_policy(),
cwd: PathBuf::from("/tmp/project"),
reasoning_effort: None,
history_log_id: 0,
history_entry_count: 0,
initial_messages: None,
network_proxy: None,
rollout_path: Some(PathBuf::new()),
}),
});
while app_event_rx.try_recv().is_ok() {}
app.suppress_shutdown_complete_thread_id = Some(old_thread_id);
let (mut replacement, _app_event_tx, _rx, _new_op_rx) =
make_chatwidget_manual_with_sender().await;
replacement.handle_codex_event(Event {
id: "new-session".to_string(),
msg: EventMsg::SessionConfigured(SessionConfiguredEvent {
session_id: new_thread_id,
forked_from_id: None,
thread_name: None,
model: "gpt-test".to_string(),
model_provider_id: "test-provider".to_string(),
service_tier: None,
approval_policy: AskForApproval::Never,
approvals_reviewer: ApprovalsReviewer::User,
sandbox_policy: SandboxPolicy::new_read_only_policy(),
cwd: PathBuf::from("/tmp/project"),
reasoning_effort: None,
history_log_id: 0,
history_entry_count: 0,
initial_messages: None,
network_proxy: None,
rollout_path: Some(PathBuf::new()),
}),
});
app.replace_chat_widget(replacement);
while app_event_rx.try_recv().is_ok() {}
app.handle_codex_event_now(Event {
id: "shutdown".to_string(),
msg: EventMsg::ShutdownComplete,
});
assert_eq!(app.suppress_shutdown_complete_thread_id, None);
assert!(app_event_rx.try_recv().is_err());
}
#[tokio::test]
async fn replay_thread_snapshot_restores_pending_pastes_for_submit() {
let (mut app, _app_event_rx, _op_rx) = make_test_app_with_channels().await;