From: Andreas Dilger <andreas.dilger@intel.com>
Date: Sat, 25 Jan 2014 01:16:43 +0000 (-0700)
Subject: LU-4413 ptlrpc: don't try to recover no_recov connection
X-Git-Tag: 2.5.56~15
X-Git-Url: https://git.whamcloud.com/?p=fs%2Flustre-release.git;a=commitdiff_plain;h=c2fdaa739a0ff8e04531a3f153f132d631d1a16f

LU-4413 ptlrpc: don't try to recover no_recov connection

If a connection has been stopped with ptlrpc_pinger_del_import() and
marked obd_no_recov, don't reconnect in ptlrpc_disconnect_import() if
the import is already disconnected.  Otherwise, without the pinger it
will just wait there indefinitely for the reconnection that will never
happen.

Put the obd_no_recov check inside ptlrpc_import_in_recovery() so that
any threads waiting on the connection to recover would also be broken
out of their sleep if obd_no_recov is set.

Signed-off-by: Andreas Dilger <andreas.dilger@intel.com>
Change-Id: Icd8041be0ce344add8d67b026353df1b1e0cab07
Reviewed-on: http://review.whamcloud.com/8996
Tested-by: Jenkins
Tested-by: Maloo <hpdd-maloo@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
---

diff --git a/lustre/osp/lwp_dev.c b/lustre/osp/lwp_dev.c
index 9560504..dc4f978 100644
--- a/lustre/osp/lwp_dev.c
+++ b/lustre/osp/lwp_dev.c
@@ -146,11 +146,9 @@ static int lwp_disconnect(struct lwp_device *d)
 	 * never added.) */
 	ptlrpc_pinger_del_import(imp);
 	rc = ptlrpc_disconnect_import(imp, 0);
-	if (rc != 0 && rc != -ETIMEDOUT && rc != -ENOTCONN && rc != -ESHUTDOWN)
+	if (rc != 0)
 		CWARN("%s: can't disconnect: rc = %d\n",
 		      d->lpd_obd->obd_name, rc);
-	else
-		rc = 0;
 
 	ptlrpc_invalidate_import(imp);
 
diff --git a/lustre/osp/osp_dev.c b/lustre/osp/osp_dev.c
index e3aad21..41272f0 100644
--- a/lustre/osp/osp_dev.c
+++ b/lustre/osp/osp_dev.c
@@ -329,9 +329,7 @@ static int osp_disconnect(struct osp_device *d)
 	(void)ptlrpc_pinger_del_import(imp);
 
 	rc = ptlrpc_disconnect_import(imp, 0);
-	if (rc == -ETIMEDOUT || rc == -ENOTCONN || rc == -ESHUTDOWN)
-		rc = 0;
-	if (rc)
+	if (rc != 0)
 		CERROR("%s: can't disconnect: rc = %d\n",
 		       d->opd_obd->obd_name, rc);
 
diff --git a/lustre/ptlrpc/import.c b/lustre/ptlrpc/import.c
index 3ba64d7..05d14e0 100644
--- a/lustre/ptlrpc/import.c
+++ b/lustre/ptlrpc/import.c
@@ -1439,29 +1439,36 @@ out:
 
 int ptlrpc_disconnect_import(struct obd_import *imp, int noclose)
 {
-        struct ptlrpc_request *req;
-        int rq_opc, rc = 0;
-        int nowait = imp->imp_obd->obd_force;
-        ENTRY;
+	struct ptlrpc_request *req;
+	int rq_opc, rc = 0;
+	ENTRY;
 
-        if (nowait)
-                GOTO(set_state, rc);
-
-        switch (imp->imp_connect_op) {
-        case OST_CONNECT: rq_opc = OST_DISCONNECT; break;
-        case MDS_CONNECT: rq_opc = MDS_DISCONNECT; break;
-        case MGS_CONNECT: rq_opc = MGS_DISCONNECT; break;
-        default:
-                CERROR("don't know how to disconnect from %s (connect_op %d)\n",
-                       obd2cli_tgt(imp->imp_obd), imp->imp_connect_op);
-                RETURN(-EINVAL);
-        }
+	if (imp->imp_obd->obd_force)
+		GOTO(set_state, rc);
+
+	switch (imp->imp_connect_op) {
+	case OST_CONNECT:
+		rq_opc = OST_DISCONNECT;
+		break;
+	case MDS_CONNECT:
+		rq_opc = MDS_DISCONNECT;
+		break;
+	case MGS_CONNECT:
+		rq_opc = MGS_DISCONNECT;
+		break;
+	default:
+		rc = -EINVAL;
+		CERROR("%s: don't know how to disconnect from %s "
+		       "(connect_op %d): rc = %d\n",
+		       imp->imp_obd->obd_name, obd2cli_tgt(imp->imp_obd),
+		       imp->imp_connect_op, rc);
+		RETURN(rc);
+	}
 
         if (ptlrpc_import_in_recovery(imp)) {
                 struct l_wait_info lwi;
                 cfs_duration_t timeout;
 
-
                 if (AT_OFF) {
                         if (imp->imp_server_timeout)
                                 timeout = cfs_time_seconds(obd_timeout / 2);
@@ -1484,7 +1491,6 @@ int ptlrpc_disconnect_import(struct obd_import *imp, int noclose)
 	spin_lock(&imp->imp_lock);
 	if (imp->imp_state != LUSTRE_IMP_FULL)
 		GOTO(out, rc);
-
 	spin_unlock(&imp->imp_lock);
 
         req = ptlrpc_request_alloc_pack(imp, &RQF_MDS_DISCONNECT,
@@ -1517,6 +1523,8 @@ out:
 	memset(&imp->imp_remote_handle, 0, sizeof(imp->imp_remote_handle));
 	spin_unlock(&imp->imp_lock);
 
+	if (rc == -ETIMEDOUT || rc == -ENOTCONN || rc == -ESHUTDOWN)
+		rc = 0;
 	RETURN(rc);
 }
 EXPORT_SYMBOL(ptlrpc_disconnect_import);
diff --git a/lustre/ptlrpc/recover.c b/lustre/ptlrpc/recover.c
index 266e826..a5d6ae7 100644
--- a/lustre/ptlrpc/recover.c
+++ b/lustre/ptlrpc/recover.c
@@ -386,11 +386,14 @@ EXPORT_SYMBOL(ptlrpc_recover_import);
 int ptlrpc_import_in_recovery(struct obd_import *imp)
 {
 	int in_recovery = 1;
+
 	spin_lock(&imp->imp_lock);
 	if (imp->imp_state == LUSTRE_IMP_FULL ||
 	    imp->imp_state == LUSTRE_IMP_CLOSED ||
-	    imp->imp_state == LUSTRE_IMP_DISCON)
+	    imp->imp_state == LUSTRE_IMP_DISCON ||
+	    imp->imp_obd->obd_no_recov)
 		in_recovery = 0;
 	spin_unlock(&imp->imp_lock);
+
 	return in_recovery;
 }