1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
//! LISTEN/NOTIFY support for PostgreSQL connections.
//!
//! PostgreSQL sends `NotificationResponse` messages asynchronously when
//! a channel the connection is LISTENing on receives a NOTIFY.
//!
//! This module provides:
//! - `Notification` struct — channel name + payload
//! - `listen()` / `unlisten()` — subscribe/unsubscribe to channels
//! - `poll_notifications()` — drain buffered notifications (non-blocking)
//! - `recv_notification()` — block-wait for the next notification
use super::{
PgConnection, PgError, PgResult, io::MAX_MESSAGE_SIZE, is_ignorable_session_message,
unexpected_backend_message,
};
use crate::protocol::PgEncoder;
/// A notification received from PostgreSQL LISTEN/NOTIFY.
#[derive(Debug, Clone)]
pub struct Notification {
/// The PID of the notifying backend process
pub process_id: i32,
/// The channel name
pub channel: String,
/// The payload (may be empty)
pub payload: String,
}
#[inline]
fn return_with_desync<T>(conn: &mut PgConnection, err: PgError) -> PgResult<T> {
if matches!(
err,
PgError::Protocol(_) | PgError::Connection(_) | PgError::Timeout(_)
) {
conn.mark_io_desynced();
}
Err(err)
}
impl PgConnection {
/// Subscribe to a notification channel.
///
/// ```ignore
/// conn.listen("price_calendar_changed").await?;
/// ```
pub async fn listen(&mut self, channel: &str) -> PgResult<()> {
// Channel names are identifiers, quote them to prevent injection
let sql = format!("LISTEN \"{}\"", channel.replace('"', "\"\""));
self.execute_simple(&sql).await
}
/// Unsubscribe from a notification channel.
pub async fn unlisten(&mut self, channel: &str) -> PgResult<()> {
let sql = format!("UNLISTEN \"{}\"", channel.replace('"', "\"\""));
self.execute_simple(&sql).await
}
/// Unsubscribe from all notification channels.
pub async fn unlisten_all(&mut self) -> PgResult<()> {
self.execute_simple("UNLISTEN *").await
}
/// Drain all buffered notifications without blocking.
///
/// Notifications arrive asynchronously from PostgreSQL and are buffered
/// whenever `recv()` encounters a `NotificationResponse`. This method
/// returns all currently buffered notifications.
pub fn poll_notifications(&mut self) -> Vec<Notification> {
self.notifications.drain(..).collect()
}
/// Wait for the next notification, blocking until one arrives.
///
/// Unlike `recv()`, this does NOT use the 30-second Slowloris timeout
/// guard. LISTEN connections idle for long periods — that's normal,
/// not a DoS attack.
///
/// Useful for a dedicated LISTEN connection in a background task.
pub async fn recv_notification(&mut self) -> PgResult<Notification> {
use crate::protocol::BackendMessage;
// Return buffered notification immediately if available
if let Some(n) = self.notifications.pop_front() {
return Ok(n);
}
// Send empty query to flush any pending notifications from server
let bytes = PgEncoder::try_encode_query_string("")?;
self.write_all_with_timeout(&bytes, "stream write").await?;
// Read messages — use recv() for the initial empty query response
// (which completes quickly), then switch to no-timeout reads
let mut got_ready = false;
loop {
// Try to decode from the existing buffer first
if self.buffer.len() >= 5 {
let msg_len = u32::from_be_bytes([
self.buffer[1],
self.buffer[2],
self.buffer[3],
self.buffer[4],
]) as usize;
if msg_len < 4 {
return return_with_desync(
self,
PgError::Protocol(format!(
"Invalid message length: {} (minimum 4)",
msg_len
)),
);
}
if msg_len > MAX_MESSAGE_SIZE {
return return_with_desync(
self,
PgError::Protocol(format!(
"Message too large: {} bytes (max {})",
msg_len, MAX_MESSAGE_SIZE
)),
);
}
if self.buffer.len() > msg_len {
let msg_bytes = self.buffer.split_to(msg_len + 1);
let (msg, _) = match BackendMessage::decode(&msg_bytes) {
Ok(decoded) => decoded,
Err(err) => return return_with_desync(self, PgError::Protocol(err)),
};
match msg {
BackendMessage::NotificationResponse {
process_id,
channel,
payload,
} => {
let notification = Notification {
process_id,
channel,
payload,
};
if got_ready {
return Ok(notification);
}
self.notifications.push_back(notification);
continue;
}
BackendMessage::EmptyQueryResponse => continue,
BackendMessage::NoticeResponse(_) => continue,
BackendMessage::ParameterStatus { .. } => continue,
BackendMessage::CommandComplete(_) => continue,
BackendMessage::ReadyForQuery(_) => {
got_ready = true;
// Check buffer for notifications that arrived with this batch
if let Some(n) = self.notifications.pop_front() {
return Ok(n);
}
continue;
}
BackendMessage::ErrorResponse(err) => {
return Err(PgError::QueryServer(err.into()));
}
msg if is_ignorable_session_message(&msg) => continue,
other => {
return return_with_desync(
self,
unexpected_backend_message("listen/notify wait", &other),
);
}
}
}
}
// Read from socket — use tokio read (no timeout!) if we've
// already gotten ReadyForQuery (now we're just waiting for NOTIFY)
if self.buffer.capacity() - self.buffer.len() < 65536 {
self.buffer.reserve(131072);
}
if got_ready {
// LISTEN connections can stay idle for hours (empty buffer),
// but a partially buffered backend frame should still timeout
// to fail-closed on slowloris-style partial writes.
let n = if self.buffer.is_empty() {
self.read_without_timeout().await?
} else {
self.read_with_timeout().await?
};
if n == 0 {
return return_with_desync(
self,
PgError::Connection("Connection closed".to_string()),
);
}
} else {
// Initial flush — use the normal timeout to avoid hanging
// if the server is unresponsive during the empty query
let n = self.read_with_timeout().await?;
if n == 0 {
return return_with_desync(
self,
PgError::Connection("Connection closed".to_string()),
);
}
}
}
}
}
#[cfg(test)]
mod tests {
use super::return_with_desync;
use crate::driver::{PgConnection, PgError};
#[cfg(unix)]
fn test_conn_with_peer() -> (PgConnection, tokio::net::UnixStream) {
use crate::driver::connection::StatementCache;
use crate::driver::stream::PgStream;
use bytes::BytesMut;
use std::collections::{HashMap, VecDeque};
use std::num::NonZeroUsize;
use tokio::net::UnixStream;
let (unix_stream, peer) = UnixStream::pair().expect("unix stream pair");
(
PgConnection {
stream: PgStream::Unix(unix_stream),
buffer: BytesMut::with_capacity(1024),
write_buf: BytesMut::with_capacity(1024),
sql_buf: BytesMut::with_capacity(256),
params_buf: Vec::new(),
prepared_statements: HashMap::new(),
stmt_cache: StatementCache::new(NonZeroUsize::new(2).expect("non-zero")),
column_info_cache: HashMap::new(),
process_id: 0,
cancel_key_bytes: Vec::new(),
requested_protocol_minor: PgConnection::default_protocol_minor(),
negotiated_protocol_minor: PgConnection::default_protocol_minor(),
notifications: VecDeque::new(),
replication_stream_active: false,
replication_mode_enabled: false,
last_replication_wal_end: None,
io_desynced: false,
pending_statement_closes: Vec::new(),
draining_statement_closes: false,
},
peer,
)
}
#[cfg(unix)]
fn test_conn() -> PgConnection {
test_conn_with_peer().0
}
#[cfg(unix)]
fn push_backend_frame(conn: &mut PgConnection, msg_type: u8, payload: &[u8]) {
conn.buffer.extend_from_slice(&[msg_type]);
conn.buffer
.extend_from_slice(&((payload.len() + 4) as u32).to_be_bytes());
conn.buffer.extend_from_slice(payload);
}
#[cfg(unix)]
fn notification_payload(process_id: i32, channel: &str, payload: &str) -> Vec<u8> {
let mut bytes = Vec::new();
bytes.extend_from_slice(&process_id.to_be_bytes());
bytes.extend_from_slice(channel.as_bytes());
bytes.push(0);
bytes.extend_from_slice(payload.as_bytes());
bytes.push(0);
bytes
}
#[cfg(unix)]
#[tokio::test]
async fn notification_return_with_desync_marks_protocol_error() {
let mut conn = test_conn();
let err =
return_with_desync::<()>(&mut conn, PgError::Protocol("bad notify frame".to_string()))
.expect_err("protocol error must be returned");
assert!(err.to_string().contains("bad notify frame"));
assert!(conn.is_io_desynced());
}
#[cfg(unix)]
#[tokio::test]
async fn recv_notification_drains_empty_query_before_returning_pre_ready_notify() {
let (mut conn, _peer) = test_conn_with_peer();
let payload = notification_payload(42, "jobs", "ready");
push_backend_frame(&mut conn, b'A', &payload);
push_backend_frame(&mut conn, b'I', &[]);
push_backend_frame(&mut conn, b'Z', b"I");
let notification = conn
.recv_notification()
.await
.expect("pre-ready notification should be returned after flush drain");
assert_eq!(notification.process_id, 42);
assert_eq!(notification.channel, "jobs");
assert_eq!(notification.payload, "ready");
assert!(
conn.buffer.is_empty(),
"empty-query flush frames must not remain buffered"
);
assert!(!conn.is_io_desynced());
}
}