Skip to contentSkip to conversation

rxrpc: Fix recvmsg() unconditional requeue

PATCH New Jay Wang 1 participant 1 message Apr 22

Subsystems (3)
lore
Share & tools
lore

Jay Wang

author
#0 lore raw

PATCH 1/1 rxrpc: Fix recvmsg() unconditional requeue

From: David Howells <[email protected]>

[ Upstream commit 2c28769a51deb6022d7fbd499987e237a01dd63a ]

If rxrpc_recvmsg() fails because MSG_DONTWAIT was specified but the call at the front of the recvmsg queue already has its mutex locked, it requeues the call - whether or not the call is already queued. The call may be on the queue because MSG_PEEK was also passed and so the call was not dequeued or because the I/O thread requeued it.

The unconditional requeue may then corrupt the recvmsg queue, leading to things like UAFs or refcount underruns.

Fix this by only requeuing the call if it isn't already on the queue - and moving it to the front if it is already queued. If we don't queue it, we have to put the ref we obtained by dequeuing it.

Also, MSG_PEEK doesn't dequeue the call so shouldn't call rxrpc_notify_socket() for the call if we didn't use up all the data on the queue, so fix that also.

Fixes 540b1c48c37a ("rxrpc: Fix deadlock between call creation and sendmsg/recvmsg")

[Adapted to 5.15: use write_lock_bh/write_unlock_bh, trace_rxrpc_call
directly for see-call tracing, 5.15 trace enum naming convention, and added entries to both plain enum and EM() macro list.]

---
 include/trace/events/rxrpc.h | 8 ++++++++
 net/rxrpc/recvmsg.c          | 20 ++++++++++++++++----
 2 files changed, 24 insertions(+), 4 deletions(-)
include/trace/events/rxrpc.h +8 -0
@@ -93,9 +93,13 @@ enum rxrpc_call_trace {
 	rxrpc_call_put_notimer,
 	rxrpc_call_put_timer,
 	rxrpc_call_put_userid,
+	rxrpc_call_put_recvmsg_peek_nowait,
 	rxrpc_call_queued,
 	rxrpc_call_queued_ref,
 	rxrpc_call_release,
+	rxrpc_call_see_recvmsg_requeue,
+	rxrpc_call_see_recvmsg_requeue_first,
+	rxrpc_call_see_recvmsg_requeue_move,
 	rxrpc_call_seen,
 };
 
@@ -291,9 +295,13 @@ enum rxrpc_tx_point {
 	EM(rxrpc_call_put_notimer,		"PnT") \
 	EM(rxrpc_call_put_timer,		"PTM") \
 	EM(rxrpc_call_put_userid,		"Pus") \
+	EM(rxrpc_call_put_recvmsg_peek_nowait,	"PpN") \
 	EM(rxrpc_call_queued,			"QUE") \
 	EM(rxrpc_call_queued_ref,		"QUR") \
 	EM(rxrpc_call_release,			"RLS") \
+	EM(rxrpc_call_see_recvmsg_requeue,	"SrQ") \
+	EM(rxrpc_call_see_recvmsg_requeue_first,"SrF") \
+	EM(rxrpc_call_see_recvmsg_requeue_move,	"SrM") \
 	E_(rxrpc_call_seen,			"SEE")
 
 #define rxrpc_transmit_traces \
net/rxrpc/recvmsg.c +18 -4
@@ -607,7 +607,8 @@ int rxrpc_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 
 		if (after(call->rx_top, call->rx_hard_ack) &&
 		    call->rxtx_buffer[(call->rx_hard_ack + 1) & RXRPC_RXTX_BUFF_MASK])
-			rxrpc_notify_socket(call);
+			if (!(flags & MSG_PEEK))
+				rxrpc_notify_socket(call);
 		break;
 	default:
 		ret = 0;
@@ -642,11 +643,24 @@ int rxrpc_recvmsg(struct socket *sock, struct msghdr *msg, size_t len,
 error_requeue_call:
 	if (!(flags & MSG_PEEK)) {
 		write_lock_bh(&rx->recvmsg_lock);
-		list_add(&call->recvmsg_link, &rx->recvmsg_q);
-		write_unlock_bh(&rx->recvmsg_lock);
+		if (list_empty(&call->recvmsg_link)) {
+			list_add(&call->recvmsg_link, &rx->recvmsg_q);
+			trace_rxrpc_call(call->debug_id,
+					 rxrpc_call_see_recvmsg_requeue,
+					 refcount_read(&call->ref),
+					 __builtin_return_address(0), NULL);
+			write_unlock_bh(&rx->recvmsg_lock);
+		} else if (list_is_first(&call->recvmsg_link, &rx->recvmsg_q)) {
+			write_unlock_bh(&rx->recvmsg_lock);
+			rxrpc_put_call(call, rxrpc_call_see_recvmsg_requeue_first);
+		} else {
+			list_move(&call->recvmsg_link, &rx->recvmsg_q);
+			write_unlock_bh(&rx->recvmsg_lock);
+			rxrpc_put_call(call, rxrpc_call_see_recvmsg_requeue_move);
+		}
 		trace_rxrpc_recvmsg(call, rxrpc_recvmsg_requeue, 0, 0, 0, 0);
 	} else {
-		rxrpc_put_call(call, rxrpc_call_put);
+		rxrpc_put_call(call, rxrpc_call_put_recvmsg_peek_nowait);
 	}
 error_no_call:
 	release_sock(&rx->sk);
Signature · 1 line
2.43.0