[PATCH 3/3] nvme-tcp: do not queue io_work on a full socket

[Hannes Reinecke <hare@kernel.org>]

There really is no point in scheduling io_work from ->queue_rq()
if the socket is full; we'll be notified by the ->write_space()
callback once space on the socket becomes available.
Consequently we also need to remove the check for 'sk_stream_is_writeable()'
in the ->write_space() callback as we need to schedule io_work
to receive packets even if the sending side is blocked.

Signed-off-by: Hannes Reinecke <hare@kernel.org>
---
 drivers/nvme/host/tcp.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index 0e178115dc04..e4dd1620dc28 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -411,6 +411,9 @@ static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req,
 	empty = llist_add(&req->lentry, &queue->req_list) &&
 		list_empty(&queue->send_list) && !queue->request;
 
+	if (!sk_stream_is_writeable(queue->sock->sk))
+		empty = false;
+
 	/*
 	 * if we're the first on the send_list and we can try to send
 	 * directly, otherwise queue io_work. Also, only do that if we
@@ -422,7 +425,8 @@ static inline void nvme_tcp_queue_request(struct nvme_tcp_request *req,
 		mutex_unlock(&queue->send_mutex);
 	}
 
-	if (last && nvme_tcp_queue_has_pending(queue))
+	if (last && nvme_tcp_queue_has_pending(queue) &&
+	    sk_stream_is_writeable(queue->sock->sk))
 		queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
 }
 
@@ -1074,7 +1078,7 @@ static void nvme_tcp_write_space(struct sock *sk)
 
 	read_lock_bh(&sk->sk_callback_lock);
 	queue = sk->sk_user_data;
-	if (likely(queue && sk_stream_is_writeable(sk))) {
+	if (likely(queue)) {
 		clear_bit(SOCK_NOSPACE, &sk->sk_socket->flags);
 		queue_work_on(queue->io_cpu, nvme_tcp_wq, &queue->io_work);
 	}
-- 
2.35.3