[PATCH 6.6 032/111] af_packet: fix soft lockup issue caused by tpacket_snd()

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

6.6-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Yun Lu <luyun@kylinos.cn>

commit 55f0bfc0370539213202f4ce1a07615327ac4713 upstream.

When MSG_DONTWAIT is not set, the tpacket_snd operation will wait for
pending_refcnt to decrement to zero before returning. The pending_refcnt
is decremented by 1 when the skb->destructor function is called,
indicating that the skb has been successfully sent and needs to be
destroyed.

If an error occurs during this process, the tpacket_snd() function will
exit and return error, but pending_refcnt may not yet have decremented to
zero. Assuming the next send operation is executed immediately, but there
are no available frames to be sent in tx_ring (i.e., packet_current_frame
returns NULL), and skb is also NULL, the function will not execute
wait_for_completion_interruptible_timeout() to yield the CPU. Instead, it
will enter a do-while loop, waiting for pending_refcnt to be zero. Even
if the previous skb has completed transmission, the skb->destructor
function can only be invoked in the ksoftirqd thread (assuming NAPI
threading is enabled). When both the ksoftirqd thread and the tpacket_snd
operation happen to run on the same CPU, and the CPU trapped in the
do-while loop without yielding, the ksoftirqd thread will not get
scheduled to run. As a result, pending_refcnt will never be reduced to
zero, and the do-while loop cannot exit, eventually leading to a CPU soft
lockup issue.

In fact, skb is true for all but the first iterations of that loop, and
as long as pending_refcnt is not zero, even if incremented by a previous
call, wait_for_completion_interruptible_timeout() should be executed to
yield the CPU, allowing the ksoftirqd thread to be scheduled. Therefore,
the execution condition of this function should be modified to check if
pending_refcnt is not zero, instead of check skb.

-	if (need_wait && skb) {
+	if (need_wait && packet_read_pending(&po->tx_ring)) {

As a result, the judgment conditions are duplicated with the end code of
the while loop, and packet_read_pending() is a very expensive function.
Actually, this loop can only exit when ph is NULL, so the loop condition
can be changed to while (1), and in the "ph = NULL" branch, if the
subsequent condition of if is not met,  the loop can break directly. Now,
the loop logic remains the same as origin but is clearer and more obvious.

Fixes: 89ed5b519004 ("af_packet: Block execution of tasks waiting for transmit to complete in AF_PACKET")
Cc: stable@kernel.org
Suggested-by: LongJun Tang <tanglongjun@kylinos.cn>
Signed-off-by: Yun Lu <luyun@kylinos.cn>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/packet/af_packet.c |   23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)

--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -2852,15 +2852,21 @@ static int tpacket_snd(struct packet_soc
 		ph = packet_current_frame(po, &po->tx_ring,
 					  TP_STATUS_SEND_REQUEST);
 		if (unlikely(ph == NULL)) {
-			if (need_wait && skb) {
+			/* Note: packet_read_pending() might be slow if we
+			 * have to call it as it's per_cpu variable, but in
+			 * fast-path we don't have to call it, only when ph
+			 * is NULL, we need to check the pending_refcnt.
+			 */
+			if (need_wait && packet_read_pending(&po->tx_ring)) {
 				timeo = wait_for_completion_interruptible_timeout(&po->skb_completion, timeo);
 				if (timeo <= 0) {
 					err = !timeo ? -ETIMEDOUT : -ERESTARTSYS;
 					goto out_put;
 				}
-			}
-			/* check for additional frames */
-			continue;
+				/* check for additional frames */
+				continue;
+			} else
+				break;
 		}
 
 		skb = NULL;
@@ -2949,14 +2955,7 @@ tpacket_error:
 		}
 		packet_increment_head(&po->tx_ring);
 		len_sum += tp_len;
-	} while (likely((ph != NULL) ||
-		/* Note: packet_read_pending() might be slow if we have
-		 * to call it as it's per_cpu variable, but in fast-path
-		 * we already short-circuit the loop with the first
-		 * condition, and luckily don't have to go that path
-		 * anyway.
-		 */
-		 (need_wait && packet_read_pending(&po->tx_ring))));
+	} while (1);
 
 	err = len_sum;
 	goto out_put;