prevent spurious unlocks from afr selfheal

afr selfheal now remembers all the nodes on which locks were successfully
held and sends unlocks only to those nodes

Signed-off-by: Anand V. Avati <avati@dev.gluster.com>

BUG: 112 (parallel deletion of files mounted by different clients on the same back-end hangs and/or does not completely delete)
URL: http://bugs.gluster.com/cgi-bin/bugzilla3/show_bug.cgi?id=112

1 files changed, 22 insertions, 6 deletions

diff --git a/xlators/cluster/afr/src/afr-self-heal-data.c b/xlators/cluster/afr/src/afr-self-heal-data.c
index 620ad1a2c87..d2224ec9263 100644
--- a/xlators/cluster/afr/src/afr-self-heal-data.c
+++ b/xlators/cluster/afr/src/afr-self-heal-data.c
@@ -56,6 +56,7 @@ afr_sh_data_done (call_frame_t *frame, xlator_t *this)
 	afr_local_t     *local = NULL;
 	afr_self_heal_t *sh = NULL;
 	afr_private_t   *priv = NULL;
+        int              i = 0;
 
 	local = frame->local;
 	sh = &local->self_heal;
@@ -65,6 +66,14 @@ afr_sh_data_done (call_frame_t *frame, xlator_t *this)
 	   TODO: cleanup sh->* 
 	 */
 
+        if (sh->healing_fd) {
+		fd_unref (sh->healing_fd);
+		sh->healing_fd = NULL;
+        }
+
+        for (i = 0; i < priv->child_count; i++)
+                sh->locked_nodes[i] = 0;
+
 	gf_log (this->name, GF_LOG_TRACE,
 		"self heal of %s completed",
 		local->loc.path);
@@ -96,8 +105,6 @@ afr_sh_data_flush_cbk (call_frame_t *frame, void *cookie, xlator_t *this,
 	call_count = afr_frame_return (frame);
 
 	if (call_count == 0) {
-		fd_unref (sh->healing_fd);
-		sh->healing_fd = NULL;
 		afr_sh_data_done (frame, this);
 	}
 
@@ -261,7 +268,15 @@ afr_sh_data_unlock (call_frame_t *frame, xlator_t *this)
 	sh = &local->self_heal;
 	priv = this->private;
 
-	call_count = local->child_count;
+        for (i = 0; i < priv->child_count; i++) {
+                if (sh->locked_nodes[i])
+                        call_count++;
+        }
+
+        if (call_count == 0) {
+                afr_sh_data_close (frame, this);
+                return 0;
+        }
 
 	local->call_count = call_count;		
 
@@ -270,7 +285,7 @@ afr_sh_data_unlock (call_frame_t *frame, xlator_t *this)
 	flock.l_type  = F_UNLCK;
 
 	for (i = 0; i < priv->child_count; i++) {
-		if (local->child_up[i]) {
+		if (sh->locked_nodes[i]) {
 			gf_log (this->name, GF_LOG_TRACE,
 				"unlocking %s on subvolume %s",
 				local->loc.path, priv->children[i]->name);
@@ -861,12 +876,13 @@ afr_sh_data_lock_cbk (call_frame_t *frame, void *cookie, xlator_t *this,
 		if (op_ret == -1) {
 			sh->op_failed = 1;
 
-			gf_log (this->name,
-                                GF_LOG_DEBUG,
+                        sh->locked_nodes[child_index] = 0;
+			gf_log (this->name, GF_LOG_DEBUG,
 				"locking of %s on child %d failed: %s",
 				local->loc.path, child_index,
 				strerror (op_errno));
 		} else {
+                        sh->locked_nodes[child_index] = 1;
 			gf_log (this->name, GF_LOG_TRACE,
 				"inode of %s on child %d locked",
 				local->loc.path, child_index);