Hey, first of all, thank you so much for publishing this. It's a huge help.
I have a TCP server running that will accept QEMU connections locally. I'm passing it to a dedicated Tokio thread and will shuttle messages to and from a TUI to help me debug a kernel I'm working on.
Details
use tokio::{net::TcpListener, sync::mpsc::Sender};
pub enum Event {
Online,
Offline,
}
pub fn start_monitor_daemon(config: crate::Config, sender: Sender<Event>) {
std::thread::spawn(move || {
tokio::runtime::Runtime::new()
.unwrap()
.block_on(run_monitor(config, sender));
panic!("monitor thread exited!");
});
}
async fn run_monitor(config: crate::Config, sender: Sender<Event>) {
let listener = TcpListener::bind(("0.0.0.0", config.port)).await.unwrap();
loop {
sender.send(Event::Offline).await.unwrap();
let Ok((socket, _)) = listener.accept().await else {
continue;
};
sender.send(Event::Online).await.unwrap();
println!("got connection");
let qmp_negotiation = qapi::futures::QmpStreamTokio::open(socket).await.unwrap();
println!("negotiation...");
let mut qmp = qmp_negotiation.negotiate().await.unwrap();
println!("negotiated");
// loop and sleep for 5 seconds each time
loop {
let status = match qmp.execute(qapi::qmp::query_status {}).await {
Ok(status) => status,
Err(err) => {
println!("error: {:#?}", err);
break;
}
};
println!("status: {status:#?}");
tokio::time::sleep(tokio::time::Duration::from_secs(5)).await;
}
}
}
This so far works. However, when the remote QEMU instance is killed, the connection should close but for some reason it doesn't ever register here (the println!("error: {:#?}", err); line is never hit). For some reason, the library isn't handling closed connections appropriately.
Is there a reason for this? Or is this a bug? I'll try to investigate a bit to see if I can't figure it out myself, but figured I'd post here either way.
Thanks again for all the work on this.
Hey, first of all, thank you so much for publishing this. It's a huge help.
I have a TCP server running that will accept QEMU connections locally. I'm passing it to a dedicated Tokio thread and will shuttle messages to and from a TUI to help me debug a kernel I'm working on.
Details
This so far works. However, when the remote QEMU instance is killed, the connection should close but for some reason it doesn't ever register here (the
println!("error: {:#?}", err);line is never hit). For some reason, the library isn't handling closed connections appropriately.Is there a reason for this? Or is this a bug? I'll try to investigate a bit to see if I can't figure it out myself, but figured I'd post here either way.
Thanks again for all the work on this.