Whamcloud - gitweb
LU-10682 lnd: pending transmits dropped silently 20/31520/2
authorAmir Shehata <amir.shehata@intel.com>
Thu, 22 Feb 2018 00:21:02 +0000 (16:21 -0800)
committerJohn L. Hammond <john.hammond@intel.com>
Thu, 5 Apr 2018 20:07:54 +0000 (20:07 +0000)
list_add was being used erroneously. The logic should be to move
the txs on ibp_tx_queue on a local list which is then processed.
The code, however, did the reverse, which would result in the
pending txs not processed and thus dropped silently. This in turn
would lead to peers reference counts at the LNet layer not
decremented since lnet_finalize() might not be called for a message.

Initialize local list and use list_splice_init() to move
transmits on the ibp_tx_queue to the local list.

Lustre-change: https://review.whamcloud.com/31374
Lustre-commit: f5c6228f0351787aefb76cb8f807eede3371ea2e

Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Change-Id: I6b36f709db2c89e53e0b3354883a8a1b1052a1dd
Reviewed-by: Doug Oucharek <dougso@me.com>
Reviewed-by: Sonia Sharma <sonia.sharma@intel.com>
Signed-off-by: Minh Diep <minh.diep@intel.com>
Reviewed-on: https://review.whamcloud.com/31520
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
lnet/klnds/o2iblnd/o2iblnd_cb.c

index 8db6c2d..9e01f97 100644 (file)
@@ -2158,8 +2158,8 @@ kiblnd_connreq_done(kib_conn_t *conn, int status)
         }
 
        /* grab pending txs while I have the lock */
-       list_add(&txs, &peer_ni->ibp_tx_queue);
-       list_del_init(&peer_ni->ibp_tx_queue);
+       INIT_LIST_HEAD(&txs);
+       list_splice_init(&peer_ni->ibp_tx_queue, &txs);
 
         if (!kiblnd_peer_active(peer_ni) ||        /* peer_ni has been deleted */
             conn->ibc_comms_error != 0) {       /* error has happened already */