summaryrefslogtreecommitdiff
path: root/fs/ocfs2/dlm
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ocfs2/dlm')
-rw-r--r--fs/ocfs2/dlm/dlmdomain.c27
-rw-r--r--fs/ocfs2/dlm/dlmmaster.c8
-rw-r--r--fs/ocfs2/dlm/dlmrecovery.c29
3 files changed, 39 insertions, 25 deletions
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index 33660a4a52fa..c973690dc0bc 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -1123,7 +1123,6 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
struct dlm_ctxt *dlm = NULL;
char *local = NULL;
int status = 0;
- int locked = 0;
qr = (struct dlm_query_region *) msg->buf;
@@ -1132,10 +1131,8 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
/* buffer used in dlm_mast_regions() */
local = kmalloc(sizeof(qr->qr_regions), GFP_KERNEL);
- if (!local) {
- status = -ENOMEM;
- goto bail;
- }
+ if (!local)
+ return -ENOMEM;
status = -EINVAL;
@@ -1144,16 +1141,15 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
if (!dlm) {
mlog(ML_ERROR, "Node %d queried hb regions on domain %s "
"before join domain\n", qr->qr_node, qr->qr_domain);
- goto bail;
+ goto out_domain_lock;
}
spin_lock(&dlm->spinlock);
- locked = 1;
if (dlm->joining_node != qr->qr_node) {
mlog(ML_ERROR, "Node %d queried hb regions on domain %s "
"but joining node is %d\n", qr->qr_node, qr->qr_domain,
dlm->joining_node);
- goto bail;
+ goto out_dlm_lock;
}
/* Support for global heartbeat was added in 1.1 */
@@ -1163,14 +1159,15 @@ static int dlm_query_region_handler(struct o2net_msg *msg, u32 len,
"but active dlm protocol is %d.%d\n", qr->qr_node,
qr->qr_domain, dlm->dlm_locking_proto.pv_major,
dlm->dlm_locking_proto.pv_minor);
- goto bail;
+ goto out_dlm_lock;
}
status = dlm_match_regions(dlm, qr, local, sizeof(qr->qr_regions));
-bail:
- if (locked)
- spin_unlock(&dlm->spinlock);
+out_dlm_lock:
+ spin_unlock(&dlm->spinlock);
+
+out_domain_lock:
spin_unlock(&dlm_domain_lock);
kfree(local);
@@ -1877,19 +1874,19 @@ static int dlm_join_domain(struct dlm_ctxt *dlm)
goto bail;
}
- status = dlm_debug_init(dlm);
+ status = dlm_launch_thread(dlm);
if (status < 0) {
mlog_errno(status);
goto bail;
}
- status = dlm_launch_thread(dlm);
+ status = dlm_launch_recovery_thread(dlm);
if (status < 0) {
mlog_errno(status);
goto bail;
}
- status = dlm_launch_recovery_thread(dlm);
+ status = dlm_debug_init(dlm);
if (status < 0) {
mlog_errno(status);
goto bail;
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index af3f7aa73e13..ee1f88419cb0 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -472,11 +472,15 @@ bail:
void dlm_destroy_master_caches(void)
{
- if (dlm_lockname_cache)
+ if (dlm_lockname_cache) {
kmem_cache_destroy(dlm_lockname_cache);
+ dlm_lockname_cache = NULL;
+ }
- if (dlm_lockres_cache)
+ if (dlm_lockres_cache) {
kmem_cache_destroy(dlm_lockres_cache);
+ dlm_lockres_cache = NULL;
+ }
}
static void dlm_lockres_release(struct kref *kref)
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 7035af09cc03..fe29f7978f81 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -537,7 +537,10 @@ master_here:
/* success! see if any other nodes need recovery */
mlog(0, "DONE mastering recovery of %s:%u here(this=%u)!\n",
dlm->name, dlm->reco.dead_node, dlm->node_num);
- dlm_reset_recovery(dlm);
+ spin_lock(&dlm->spinlock);
+ __dlm_reset_recovery(dlm);
+ dlm->reco.state &= ~DLM_RECO_STATE_FINALIZE;
+ spin_unlock(&dlm->spinlock);
}
dlm_end_recovery(dlm);
@@ -695,6 +698,14 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
if (all_nodes_done) {
int ret;
+ /* Set this flag on recovery master to avoid
+ * a new recovery for another dead node start
+ * before the recovery is not done. That may
+ * cause recovery hung.*/
+ spin_lock(&dlm->spinlock);
+ dlm->reco.state |= DLM_RECO_STATE_FINALIZE;
+ spin_unlock(&dlm->spinlock);
+
/* all nodes are now in DLM_RECO_NODE_DATA_DONE state
* just send a finalize message to everyone and
* clean up */
@@ -1750,13 +1761,13 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm,
struct dlm_migratable_lockres *mres)
{
struct dlm_migratable_lock *ml;
- struct list_head *queue;
+ struct list_head *queue, *iter;
struct list_head *tmpq = NULL;
struct dlm_lock *newlock = NULL;
struct dlm_lockstatus *lksb = NULL;
int ret = 0;
int i, j, bad;
- struct dlm_lock *lock = NULL;
+ struct dlm_lock *lock;
u8 from = O2NM_MAX_NODES;
unsigned int added = 0;
__be64 c;
@@ -1791,14 +1802,16 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm,
/* MIGRATION ONLY! */
BUG_ON(!(mres->flags & DLM_MRES_MIGRATION));
+ lock = NULL;
spin_lock(&res->spinlock);
for (j = DLM_GRANTED_LIST; j <= DLM_BLOCKED_LIST; j++) {
tmpq = dlm_list_idx_to_ptr(res, j);
- list_for_each_entry(lock, tmpq, list) {
- if (lock->ml.cookie != ml->cookie)
- lock = NULL;
- else
+ list_for_each(iter, tmpq) {
+ lock = list_entry(iter,
+ struct dlm_lock, list);
+ if (lock->ml.cookie == ml->cookie)
break;
+ lock = NULL;
}
if (lock)
break;
@@ -2882,8 +2895,8 @@ int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data,
BUG();
}
dlm->reco.state &= ~DLM_RECO_STATE_FINALIZE;
+ __dlm_reset_recovery(dlm);
spin_unlock(&dlm->spinlock);
- dlm_reset_recovery(dlm);
dlm_kick_recovery_thread(dlm);
break;
default:
generated by cgit v1.2.3 (git 2.0.4) at 2025-06-11 04:38:49 +0000