[PATCH 3.19.y-ckt 048/251] IB/srp: Fix reconnection failure handling

[Kamal Mostafa]

3.19.8-ckt4 -stable review patch.  If anyone has any objections, please let me know.

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

From: Bart Van Assche <bart.vanassche at sandisk.com>

commit a44074f14ba1ea0747ea737026eb929b81993dc3 upstream.

Although it is possible to let SRP I/O continue if a reconnect
results in a reduction of the number of channels, the current
code does not handle this scenario correctly. Instead of making
the reconnect code more complex, consider this as a reconnection
failure.

Signed-off-by: Bart Van Assche <bart.vanassche at sandisk.com>
Cc: Sagi Grimberg <sagig at mellanox.com>
Cc: Sebastian Parschauer <sebastian.riemer at profitbricks.com>
Signed-off-by: Doug Ledford <dledford at redhat.com>
Signed-off-by: Kamal Mostafa <kamal at canonical.com>
---
 drivers/infiniband/ulp/srp/ib_srp.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 5297447..313dfad 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1242,11 +1242,8 @@ static int srp_rport_reconnect(struct srp_rport *rport)
 
 	for (i = 0; i < target->ch_count; i++) {
 		ch = &target->ch[i];
-		if (ret || !ch->target) {
-			if (i > 1)
-				ret = 0;
+		if (ret || !ch->target)
 			break;
-		}
 		ret = srp_connect_ch(ch, multich);
 		multich = true;
 	}
-- 
1.9.1