Skip to contentSkip to conversation

dlm: delayed reply message warning

PATCH New David Teigland 1 participant 1 message May 23, 2011 Distributed lock manager (DLM)

lore
Share & tools
lore

David Teigland

author maintainer
#0 lore raw

PATCH 2/4 dlm: delayed reply message warning

Add an option (disabled by default) to print a warning message when a lock has been waiting a configurable amount of time for a reply message from another node. This is mainly for debugging.

---
 fs/dlm/config.c       |    9 ++++-
 fs/dlm/config.h       |    1 +
 fs/dlm/dlm_internal.h |    2 +
 fs/dlm/lock.c         |  100 +++++++++++++++++++++++++++++++++++++++++++++---
 fs/dlm/lock.h         |    1 +
 fs/dlm/lockspace.c    |    6 +-
 6 files changed, 108 insertions(+), 11 deletions(-)
fs/dlm/config.c +8 -1
index 0d329ff..9b026ea 100644
@@ -100,6 +100,7 @@ struct dlm_cluster {
 	unsigned int cl_log_debug;
 	unsigned int cl_protocol;
 	unsigned int cl_timewarn_cs;
+	unsigned int cl_waitwarn_us;
 };
 
 enum {
@@ -114,6 +115,7 @@ enum {
 	CLUSTER_ATTR_LOG_DEBUG,
 	CLUSTER_ATTR_PROTOCOL,
 	CLUSTER_ATTR_TIMEWARN_CS,
+	CLUSTER_ATTR_WAITWARN_US,
 };
 
 struct cluster_attribute {
@@ -166,6 +168,7 @@ CLUSTER_ATTR(scan_secs, 1);
 CLUSTER_ATTR(log_debug, 0);
 CLUSTER_ATTR(protocol, 0);
 CLUSTER_ATTR(timewarn_cs, 1);
+CLUSTER_ATTR(waitwarn_us, 0);
 
 static struct configfs_attribute *cluster_attrs[] = {
 	[CLUSTER_ATTR_TCP_PORT] = &cluster_attr_tcp_port.attr,
@@ -179,6 +182,7 @@ static struct configfs_attribute *cluster_attrs[] = {
 	[CLUSTER_ATTR_LOG_DEBUG] = &cluster_attr_log_debug.attr,
 	[CLUSTER_ATTR_PROTOCOL] = &cluster_attr_protocol.attr,
 	[CLUSTER_ATTR_TIMEWARN_CS] = &cluster_attr_timewarn_cs.attr,
+	[CLUSTER_ATTR_WAITWARN_US] = &cluster_attr_waitwarn_us.attr,
 	NULL,
 };
 
@@ -439,6 +443,7 @@ static struct config_group *make_cluster(struct config_group *g,
 	cl->cl_log_debug = dlm_config.ci_log_debug;
 	cl->cl_protocol = dlm_config.ci_protocol;
 	cl->cl_timewarn_cs = dlm_config.ci_timewarn_cs;
+	cl->cl_waitwarn_us = dlm_config.ci_waitwarn_us;
 
 	space_list = &sps->ss_group;
 	comm_list = &cms->cs_group;
@@ -986,6 +991,7 @@ int dlm_our_addr(struct sockaddr_storage *addr, int num)
 #define DEFAULT_LOG_DEBUG          0
 #define DEFAULT_PROTOCOL           0
 #define DEFAULT_TIMEWARN_CS      500 /* 5 sec = 500 centiseconds */
+#define DEFAULT_WAITWARN_US	   0
 
 struct dlm_config_info dlm_config = {
 	.ci_tcp_port = DEFAULT_TCP_PORT,
@@ -998,6 +1004,7 @@ struct dlm_config_info dlm_config = {
 	.ci_scan_secs = DEFAULT_SCAN_SECS,
 	.ci_log_debug = DEFAULT_LOG_DEBUG,
 	.ci_protocol = DEFAULT_PROTOCOL,
-	.ci_timewarn_cs = DEFAULT_TIMEWARN_CS
+	.ci_timewarn_cs = DEFAULT_TIMEWARN_CS,
+	.ci_waitwarn_us = DEFAULT_WAITWARN_US
 };
 
fs/dlm/config.h +1 -0
index 4f1d6fc..dd0ce24 100644
@@ -28,6 +28,7 @@ struct dlm_config_info {
 	int ci_log_debug;
 	int ci_protocol;
 	int ci_timewarn_cs;
+	int ci_waitwarn_us;
 };
 
 extern struct dlm_config_info dlm_config;
fs/dlm/dlm_internal.h +2 -0
index b942049..6a92478 100644
@@ -245,6 +245,7 @@ struct dlm_lkb {
 
 	int8_t			lkb_wait_type;	/* type of reply waiting for */
 	int8_t			lkb_wait_count;
+	int			lkb_wait_nodeid; /* for debugging */
 
 	struct list_head	lkb_idtbl_list;	/* lockspace lkbtbl */
 	struct list_head	lkb_statequeue;	/* rsb g/c/w list */
@@ -254,6 +255,7 @@ struct dlm_lkb {
 	struct list_head	lkb_ownqueue;	/* list of locks for a process */
 	struct list_head	lkb_time_list;
 	ktime_t			lkb_timestamp;
+	ktime_t			lkb_wait_time;
 	unsigned long		lkb_timeout_cs;
 
 	struct dlm_callback	lkb_callbacks[DLM_CALLBACKS_SIZE];
fs/dlm/lock.c +93 -7

139 lines not on this page · raw message

fs/dlm/lock.h +1 -0
index 88e93c8..265017a 100644
@@ -24,6 +24,7 @@ int dlm_put_lkb(struct dlm_lkb *lkb);
 void dlm_scan_rsbs(struct dlm_ls *ls);
 int dlm_lock_recovery_try(struct dlm_ls *ls);
 void dlm_unlock_recovery(struct dlm_ls *ls);
+void dlm_scan_waiters(struct dlm_ls *ls);
 void dlm_scan_timeout(struct dlm_ls *ls);
 void dlm_adjust_timeouts(struct dlm_ls *ls);
 
fs/dlm/lockspace.c +3 -3
index f994a7d..14cbf40 100644
@@ -243,7 +243,6 @@ static struct dlm_ls *find_ls_to_scan(void)
 static int dlm_scand(void *data)
 {
 	struct dlm_ls *ls;
-	int timeout_jiffies = dlm_config.ci_scan_secs * HZ;
 
 	while (!kthread_should_stop()) {
 		ls = find_ls_to_scan();
@@ -252,13 +251,14 @@ static int dlm_scand(void *data)
 				ls->ls_scan_time = jiffies;
 				dlm_scan_rsbs(ls);
 				dlm_scan_timeout(ls);
+				dlm_scan_waiters(ls);
 				dlm_unlock_recovery(ls);
 			} else {
 				ls->ls_scan_time += HZ;
 			}
-		} else {
-			schedule_timeout_interruptible(timeout_jiffies);
+			continue;
 		}
+		schedule_timeout_interruptible(dlm_config.ci_scan_secs * HZ);
 	}
 	return 0;
 }
Signature · 1 line
1.7.1.1