RDS: don't trust the LL_SEND_FULL bit

We are seeing connections stuck with the LL_SEND_FULL bit getting
set and never cleared.  This changes RDS to stop trusting the
LL_SEND_FULL bit and kick krdsd after any time we
see -ENOMEM from the ring allocation code.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Bang Nguyen <bang.nguyen@oracle.com>

diff --git a/net/rds/send.c b/net/rds/send.c
index 16e26151f3e07a2025f813ed8e013d9477e69c90..0cbddbc942f516bfb338e7d8f45ee43a99e78edc 100644 (file)
--- a/net/rds/send.c
+++ b/net/rds/send.c
@@ -1096,8 +1096,10 @@ int rds_sendmsg(struct kiocb *iocb, struct socket *sock, struct msghdr *msg,
         */
        rds_stats_inc(s_send_queued);
 
-       if (!test_bit(RDS_LL_SEND_FULL, &conn->c_flags))
-               rds_send_xmit(conn);
+       ret = rds_send_xmit(conn);
+       if (ret == -ENOMEM || ret == -EAGAIN)
+               queue_delayed_work(rds_wq, &conn->c_send_w, 1);
+
 
        rds_message_put(rm);
        return payload_len;
@@ -1153,8 +1155,9 @@ rds_send_pong(struct rds_connection *conn, __be16 dport)
        rds_stats_inc(s_send_queued);
        rds_stats_inc(s_send_pong);
 
-       if (!test_bit(RDS_LL_SEND_FULL, &conn->c_flags))
-               rds_send_xmit(conn);
+       ret = rds_send_xmit(conn);
+       if (ret == -ENOMEM || ret == -EAGAIN)
+               queue_delayed_work(rds_wq, &conn->c_send_w, 1);
 
        rds_message_put(rm);
        return 0;

diff --git a/net/rds/threads.c b/net/rds/threads.c
index 21aeecca483167b56aa618c9082243052ae2fcd0..84a06c6c1964dfcec19e2ae0916c672a4cce63c6 100644 (file)
--- a/net/rds/threads.c
+++ b/net/rds/threads.c
@@ -169,7 +169,9 @@ void rds_send_worker(struct work_struct *work)
        int ret;
 
        if (rds_conn_state(conn) == RDS_CONN_UP) {
+               clear_bit(RDS_LL_SEND_FULL, &conn->c_flags);
                ret = rds_send_xmit(conn);
+               cond_resched();
                rdsdebug("conn %p ret %d\n", conn, ret);
                switch (ret) {
                case -EAGAIN: