[lustre-devel] [PATCH 15/20] lnet: o2iblnd: fix handling of RDMA_CM_EVENT_UNREACHABLE

[James Simmons]

From: Serguei Smirnov <ssmirnov at whamcloud.com>

RDMA_CM_EVENT_UNREACHABLE may be received not only when connection
is being connected, but also when it is being closed. Fix handing
of this event accordingly.

WC-bug-id: https://jira.whamcloud.com/browse/LU-15885
Lustre-commit: 3925b1669d519e6c0 ("LU-15885 o2iblnd: fix handling of RDMA_CM_EVENT_UNREACHABLE")
Signed-off-by: Serguei Smirnov <ssmirnov at whamcloud.com>
Reviewed-on: https://review.whamcloud.com/c/fs/lustre-release/+/48492
Reviewed-by: Frank Sehr <fsehr at whamcloud.com>
Reviewed-by: Chris Horn <chris.horn at hpe.com>
Reviewed-by: Oleg Drokin <green at whamcloud.com>
Signed-off-by: James Simmons <jsimmons at infradead.org>
---
 net/lnet/klnds/o2iblnd/o2iblnd_cb.c | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/net/lnet/klnds/o2iblnd/o2iblnd_cb.c b/net/lnet/klnds/o2iblnd/o2iblnd_cb.c
index 30e77c0382f1..919b83d5c6e2 100644
--- a/net/lnet/klnds/o2iblnd/o2iblnd_cb.c
+++ b/net/lnet/klnds/o2iblnd/o2iblnd_cb.c
@@ -3171,13 +3171,17 @@ kiblnd_cm_callback(struct rdma_cm_id *cmid, struct rdma_cm_event *event)
 		return event->status;	/* rc destroys cmid */
 
 	case RDMA_CM_EVENT_UNREACHABLE:
-		conn = (struct kib_conn *)cmid->context;
-		LASSERT(conn->ibc_state == IBLND_CONN_ACTIVE_CONNECT ||
-			conn->ibc_state == IBLND_CONN_PASSIVE_WAIT);
-		CNETERR("%s: UNREACHABLE %d\n",
-			libcfs_nid2str(conn->ibc_peer->ibp_nid), event->status);
-		kiblnd_connreq_done(conn, -ENETDOWN);
-		kiblnd_conn_decref(conn);
+		CNETERR("%s: UNREACHABLE %d, ibc_state: %d\n",
+			libcfs_nid2str(conn->ibc_peer->ibp_nid),
+			event->status,
+			conn->ibc_state);
+		LASSERT(conn->ibc_state != IBLND_CONN_ESTABLISHED &&
+			conn->ibc_state != IBLND_CONN_INIT);
+		if (conn->ibc_state == IBLND_CONN_ACTIVE_CONNECT ||
+		    conn->ibc_state == IBLND_CONN_PASSIVE_WAIT) {
+			kiblnd_connreq_done(conn, -ENETDOWN);
+			kiblnd_conn_decref(conn);
+		}
 		return 0;
 
 	case RDMA_CM_EVENT_CONNECT_ERROR:
-- 
2.27.0