aboutsummaryrefslogtreecommitdiff
path: root/fs/ocfs2
diff options
context:
space:
mode:
Diffstat (limited to 'fs/ocfs2')
-rw-r--r--fs/ocfs2/buffer_head_io.c10
-rw-r--r--fs/ocfs2/cluster/heartbeat.c5
-rw-r--r--fs/ocfs2/cluster/masklog.c1
-rw-r--r--fs/ocfs2/cluster/masklog.h2
-rw-r--r--fs/ocfs2/cluster/nodemanager.c4
-rw-r--r--fs/ocfs2/cluster/tcp.c30
-rw-r--r--fs/ocfs2/cluster/tcp.h5
-rw-r--r--fs/ocfs2/dlm/dlmcommon.h13
-rw-r--r--fs/ocfs2/dlm/dlmconvert.c12
-rw-r--r--fs/ocfs2/dlm/dlmdebug.c12
-rw-r--r--fs/ocfs2/dlm/dlmdomain.c57
-rw-r--r--fs/ocfs2/dlm/dlmlock.c25
-rw-r--r--fs/ocfs2/dlm/dlmmaster.c35
-rw-r--r--fs/ocfs2/dlm/dlmrecovery.c315
-rw-r--r--fs/ocfs2/dlm/dlmunlock.c13
-rw-r--r--fs/ocfs2/dlm/userdlm.c2
-rw-r--r--fs/ocfs2/extent_map.c50
-rw-r--r--fs/ocfs2/file.c61
-rw-r--r--fs/ocfs2/heartbeat.c1
-rw-r--r--fs/ocfs2/inode.c52
-rw-r--r--fs/ocfs2/inode.h4
-rw-r--r--fs/ocfs2/journal.c163
-rw-r--r--fs/ocfs2/journal.h2
-rw-r--r--fs/ocfs2/ocfs2.h10
-rw-r--r--fs/ocfs2/ocfs2_fs.h1
-rw-r--r--fs/ocfs2/super.c22
-rw-r--r--fs/ocfs2/sysfile.c6
-rw-r--r--fs/ocfs2/uptodate.c12
-rw-r--r--fs/ocfs2/uptodate.h2
29 files changed, 660 insertions, 267 deletions
diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c
index d424041b38e..bae3d7548be 100644
--- a/fs/ocfs2/buffer_head_io.c
+++ b/fs/ocfs2/buffer_head_io.c
@@ -58,7 +58,7 @@ int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
goto out;
}
- down(&OCFS2_I(inode)->ip_io_sem);
+ mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
lock_buffer(bh);
set_buffer_uptodate(bh);
@@ -82,7 +82,7 @@ int ocfs2_write_block(struct ocfs2_super *osb, struct buffer_head *bh,
brelse(bh);
}
- up(&OCFS2_I(inode)->ip_io_sem);
+ mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
out:
mlog_exit(ret);
return ret;
@@ -125,13 +125,13 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
flags &= ~OCFS2_BH_CACHED;
if (inode)
- down(&OCFS2_I(inode)->ip_io_sem);
+ mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
for (i = 0 ; i < nr ; i++) {
if (bhs[i] == NULL) {
bhs[i] = sb_getblk(sb, block++);
if (bhs[i] == NULL) {
if (inode)
- up(&OCFS2_I(inode)->ip_io_sem);
+ mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
status = -EIO;
mlog_errno(status);
goto bail;
@@ -220,7 +220,7 @@ int ocfs2_read_blocks(struct ocfs2_super *osb, u64 block, int nr,
ocfs2_set_buffer_uptodate(inode, bh);
}
if (inode)
- up(&OCFS2_I(inode)->ip_io_sem);
+ mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
mlog(ML_BH_IO, "block=(%"MLFu64"), nr=(%d), cached=%s\n", block, nr,
(!(flags & OCFS2_BH_CACHED) || ignore_cache) ? "no" : "yes");
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 7307ba52891..d08971d29b6 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -917,8 +917,9 @@ static int o2hb_thread(void *data)
elapsed_msec = o2hb_elapsed_msecs(&before_hb, &after_hb);
mlog(0, "start = %lu.%lu, end = %lu.%lu, msec = %u\n",
- before_hb.tv_sec, before_hb.tv_usec,
- after_hb.tv_sec, after_hb.tv_usec, elapsed_msec);
+ before_hb.tv_sec, (unsigned long) before_hb.tv_usec,
+ after_hb.tv_sec, (unsigned long) after_hb.tv_usec,
+ elapsed_msec);
if (elapsed_msec < reg->hr_timeout_ms) {
/* the kthread api has blocked signals for us so no
diff --git a/fs/ocfs2/cluster/masklog.c b/fs/ocfs2/cluster/masklog.c
index fd741cea570..636593bf4d1 100644
--- a/fs/ocfs2/cluster/masklog.c
+++ b/fs/ocfs2/cluster/masklog.c
@@ -74,6 +74,7 @@ struct mlog_attribute {
#define define_mask(_name) { \
.attr = { \
.name = #_name, \
+ .owner = THIS_MODULE, \
.mode = S_IRUGO | S_IWUSR, \
}, \
.mask = ML_##_name, \
diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h
index e8c56a3d9c6..2cadc3009c8 100644
--- a/fs/ocfs2/cluster/masklog.h
+++ b/fs/ocfs2/cluster/masklog.h
@@ -256,7 +256,7 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits;
} \
} while (0)
-#if (BITS_PER_LONG == 32) || defined(CONFIG_X86_64)
+#if (BITS_PER_LONG == 32) || defined(CONFIG_X86_64) || (defined(CONFIG_UML_X86) && defined(CONFIG_64BIT))
#define MLFi64 "lld"
#define MLFu64 "llu"
#define MLFx64 "llx"
diff --git a/fs/ocfs2/cluster/nodemanager.c b/fs/ocfs2/cluster/nodemanager.c
index cf7828f2336..e1fceb8aa32 100644
--- a/fs/ocfs2/cluster/nodemanager.c
+++ b/fs/ocfs2/cluster/nodemanager.c
@@ -756,7 +756,7 @@ static int __init init_o2nm(void)
if (!ocfs2_table_header) {
printk(KERN_ERR "nodemanager: unable to register sysctl\n");
ret = -ENOMEM; /* or something. */
- goto out;
+ goto out_o2net;
}
ret = o2net_register_hb_callbacks();
@@ -780,6 +780,8 @@ out_callbacks:
o2net_unregister_hb_callbacks();
out_sysctl:
unregister_sysctl_table(ocfs2_table_header);
+out_o2net:
+ o2net_exit();
out:
return ret;
}
diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index 35d92c01a97..0f60cc0d398 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -1285,14 +1285,16 @@ static void o2net_idle_timer(unsigned long data)
mlog(ML_NOTICE, "here are some times that might help debug the "
"situation: (tmr %ld.%ld now %ld.%ld dr %ld.%ld adv "
"%ld.%ld:%ld.%ld func (%08x:%u) %ld.%ld:%ld.%ld)\n",
- sc->sc_tv_timer.tv_sec, sc->sc_tv_timer.tv_usec,
- now.tv_sec, now.tv_usec,
- sc->sc_tv_data_ready.tv_sec, sc->sc_tv_data_ready.tv_usec,
- sc->sc_tv_advance_start.tv_sec, sc->sc_tv_advance_start.tv_usec,
- sc->sc_tv_advance_stop.tv_sec, sc->sc_tv_advance_stop.tv_usec,
+ sc->sc_tv_timer.tv_sec, (long) sc->sc_tv_timer.tv_usec,
+ now.tv_sec, (long) now.tv_usec,
+ sc->sc_tv_data_ready.tv_sec, (long) sc->sc_tv_data_ready.tv_usec,
+ sc->sc_tv_advance_start.tv_sec,
+ (long) sc->sc_tv_advance_start.tv_usec,
+ sc->sc_tv_advance_stop.tv_sec,
+ (long) sc->sc_tv_advance_stop.tv_usec,
sc->sc_msg_key, sc->sc_msg_type,
- sc->sc_tv_func_start.tv_sec, sc->sc_tv_func_start.tv_usec,
- sc->sc_tv_func_stop.tv_sec, sc->sc_tv_func_stop.tv_usec);
+ sc->sc_tv_func_start.tv_sec, (long) sc->sc_tv_func_start.tv_usec,
+ sc->sc_tv_func_stop.tv_sec, (long) sc->sc_tv_func_stop.tv_usec);
o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
}
@@ -1316,7 +1318,7 @@ static void o2net_start_connect(void *arg)
{
struct o2net_node *nn = arg;
struct o2net_sock_container *sc = NULL;
- struct o2nm_node *node = NULL;
+ struct o2nm_node *node = NULL, *mynode = NULL;
struct socket *sock = NULL;
struct sockaddr_in myaddr = {0, }, remoteaddr = {0, };
int ret = 0;
@@ -1332,6 +1334,12 @@ static void o2net_start_connect(void *arg)
goto out;
}
+ mynode = o2nm_get_node_by_num(o2nm_this_node());
+ if (mynode == NULL) {
+ ret = 0;
+ goto out;
+ }
+
spin_lock(&nn->nn_lock);
/* see if we already have one pending or have given up */
if (nn->nn_sc || nn->nn_persistent_error)
@@ -1359,12 +1367,14 @@ static void o2net_start_connect(void *arg)
sock->sk->sk_allocation = GFP_ATOMIC;
myaddr.sin_family = AF_INET;
+ myaddr.sin_addr.s_addr = (__force u32)mynode->nd_ipv4_address;
myaddr.sin_port = (__force u16)htons(0); /* any port */
ret = sock->ops->bind(sock, (struct sockaddr *)&myaddr,
sizeof(myaddr));
if (ret) {
- mlog(0, "bind failed: %d\n", ret);
+ mlog(ML_ERROR, "bind failed with %d at address %u.%u.%u.%u\n",
+ ret, NIPQUAD(mynode->nd_ipv4_address));
goto out;
}
@@ -1405,6 +1415,8 @@ out:
sc_put(sc);
if (node)
o2nm_node_put(node);
+ if (mynode)
+ o2nm_node_put(mynode);
return;
}
diff --git a/fs/ocfs2/cluster/tcp.h b/fs/ocfs2/cluster/tcp.h
index a6f4585501c..616ff2b8434 100644
--- a/fs/ocfs2/cluster/tcp.h
+++ b/fs/ocfs2/cluster/tcp.h
@@ -85,13 +85,10 @@ enum {
O2NET_DRIVER_READY,
};
-int o2net_init_tcp_sock(struct inode *inode);
int o2net_send_message(u32 msg_type, u32 key, void *data, u32 len,
u8 target_node, int *status);
int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *vec,
size_t veclen, u8 target_node, int *status);
-int o2net_broadcast_message(u32 msg_type, u32 key, void *data, u32 len,
- struct inode *group);
int o2net_register_handler(u32 msg_type, u32 key, u32 max_len,
o2net_msg_handler_func *func, void *data,
@@ -107,7 +104,5 @@ void o2net_disconnect_node(struct o2nm_node *node);
int o2net_init(void);
void o2net_exit(void);
-int o2net_proc_init(struct proc_dir_entry *parent);
-void o2net_proc_exit(struct proc_dir_entry *parent);
#endif /* O2CLUSTER_TCP_H */
diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h
index 3fecba0a602..9c772583744 100644
--- a/fs/ocfs2/dlm/dlmcommon.h
+++ b/fs/ocfs2/dlm/dlmcommon.h
@@ -37,9 +37,7 @@
#define DLM_THREAD_SHUFFLE_INTERVAL 5 // flush everything every 5 passes
#define DLM_THREAD_MS 200 // flush at least every 200 ms
-#define DLM_HASH_BITS 7
-#define DLM_HASH_SIZE (1 << DLM_HASH_BITS)
-#define DLM_HASH_MASK (DLM_HASH_SIZE - 1)
+#define DLM_HASH_BUCKETS (PAGE_SIZE / sizeof(struct hlist_head))
enum dlm_ast_type {
DLM_AST = 0,
@@ -87,7 +85,7 @@ enum dlm_ctxt_state {
struct dlm_ctxt
{
struct list_head list;
- struct list_head *resources;
+ struct hlist_head *lockres_hash;
struct list_head dirty_list;
struct list_head purge_list;
struct list_head pending_asts;
@@ -208,13 +206,16 @@ static inline void __dlm_set_joining_node(struct dlm_ctxt *dlm,
#define DLM_LOCK_RES_IN_PROGRESS 0x00000010
#define DLM_LOCK_RES_MIGRATING 0x00000020
+/* max milliseconds to wait to sync up a network failure with a node death */
+#define DLM_NODE_DEATH_WAIT_MAX (5 * 1000)
+
#define DLM_PURGE_INTERVAL_MS (8 * 1000)
struct dlm_lock_resource
{
/* WARNING: Please see the comment in dlm_init_lockres before
* adding fields here. */
- struct list_head list;
+ struct hlist_node hash_node;
struct kref refs;
/* please keep these next 3 in this order
@@ -657,6 +658,8 @@ void dlm_complete_thread(struct dlm_ctxt *dlm);
int dlm_launch_recovery_thread(struct dlm_ctxt *dlm);
void dlm_complete_recovery_thread(struct dlm_ctxt *dlm);
void dlm_wait_for_recovery(struct dlm_ctxt *dlm);
+int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node);
+int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
void dlm_put(struct dlm_ctxt *dlm);
struct dlm_ctxt *dlm_grab(struct dlm_ctxt *dlm);
diff --git a/fs/ocfs2/dlm/dlmconvert.c b/fs/ocfs2/dlm/dlmconvert.c
index 6001b22a997..f66e2d818cc 100644
--- a/fs/ocfs2/dlm/dlmconvert.c
+++ b/fs/ocfs2/dlm/dlmconvert.c
@@ -392,6 +392,11 @@ static enum dlm_status dlm_send_remote_convert_request(struct dlm_ctxt *dlm,
} else {
mlog_errno(tmpret);
if (dlm_is_host_down(tmpret)) {
+ /* instead of logging the same network error over
+ * and over, sleep here and wait for the heartbeat
+ * to notice the node is dead. times out after 5s. */
+ dlm_wait_for_node_death(dlm, res->owner,
+ DLM_NODE_DEATH_WAIT_MAX);
ret = DLM_RECOVERING;
mlog(0, "node %u died so returning DLM_RECOVERING "
"from convert message!\n", res->owner);
@@ -421,7 +426,7 @@ int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data)
struct dlm_lockstatus *lksb;
enum dlm_status status = DLM_NORMAL;
u32 flags;
- int call_ast = 0, kick_thread = 0;
+ int call_ast = 0, kick_thread = 0, ast_reserved = 0;
if (!dlm_grab(dlm)) {
dlm_error(DLM_REJECTED);
@@ -490,6 +495,7 @@ int dlm_convert_lock_handler(struct o2net_msg *msg, u32 len, void *data)
status = __dlm_lockres_state_to_status(res);
if (status == DLM_NORMAL) {
__dlm_lockres_reserve_ast(res);
+ ast_reserved = 1;
res->state |= DLM_LOCK_RES_IN_PROGRESS;
status = __dlmconvert_master(dlm, res, lock, flags,
cnv->requested_type,
@@ -512,10 +518,10 @@ leave:
else
dlm_lock_put(lock);
- /* either queue the ast or release it */
+ /* either queue the ast or release it, if reserved */
if (call_ast)
dlm_queue_ast(dlm, lock);
- else
+ else if (ast_reserved)
dlm_lockres_release_ast(dlm, res);
if (kick_thread)
diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c
index f339fe27975..54f61b76ab5 100644
--- a/fs/ocfs2/dlm/dlmdebug.c
+++ b/fs/ocfs2/dlm/dlmdebug.c
@@ -117,8 +117,8 @@ EXPORT_SYMBOL_GPL(dlm_print_one_lock);
void dlm_dump_lock_resources(struct dlm_ctxt *dlm)
{
struct dlm_lock_resource *res;
- struct list_head *iter;
- struct list_head *bucket;
+ struct hlist_node *iter;
+ struct hlist_head *bucket;
int i;
mlog(ML_NOTICE, "struct dlm_ctxt: %s, node=%u, key=%u\n",
@@ -129,12 +129,10 @@ void dlm_dump_lock_resources(struct dlm_ctxt *dlm)
}
spin_lock(&dlm->spinlock);
- for (i=0; i<DLM_HASH_SIZE; i++) {
- bucket = &(dlm->resources[i]);
- list_for_each(iter, bucket) {
- res = list_entry(iter, struct dlm_lock_resource, list);
+ for (i=0; i<DLM_HASH_BUCKETS; i++) {
+ bucket = &(dlm->lockres_hash[i]);
+ hlist_for_each_entry(res, iter, bucket, hash_node)
dlm_print_one_lock_resource(res);
- }
}
spin_unlock(&dlm->spinlock);
}
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index da3c22045f8..8f3a9e3106f 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -77,26 +77,26 @@ static void dlm_unregister_domain_handlers(struct dlm_ctxt *dlm);
void __dlm_unhash_lockres(struct dlm_lock_resource *lockres)
{
- list_del_init(&lockres->list);
+ hlist_del_init(&lockres->hash_node);
dlm_lockres_put(lockres);
}
void __dlm_insert_lockres(struct dlm_ctxt *dlm,
struct dlm_lock_resource *res)
{
- struct list_head *bucket;
+ struct hlist_head *bucket;
struct qstr *q;
assert_spin_locked(&dlm->spinlock);
q = &res->lockname;
q->hash = full_name_hash(q->name, q->len);
- bucket = &(dlm->resources[q->hash & DLM_HASH_MASK]);
+ bucket = &(dlm->lockres_hash[q->hash % DLM_HASH_BUCKETS]);
/* get a reference for our hashtable */
dlm_lockres_get(res);
- list_add_tail(&res->list, bucket);
+ hlist_add_head(&res->hash_node, bucket);
}
struct dlm_lock_resource * __dlm_lookup_lockres(struct dlm_ctxt *dlm,
@@ -104,9 +104,9 @@ struct dlm_lock_resource * __dlm_lookup_lockres(struct dlm_ctxt *dlm,
unsigned int len)
{
unsigned int hash;
- struct list_head *iter;
+ struct hlist_node *iter;
struct dlm_lock_resource *tmpres=NULL;
- struct list_head *bucket;
+ struct hlist_head *bucket;
mlog_entry("%.*s\n", len, name);
@@ -114,11 +114,11 @@ struct dlm_lock_resource * __dlm_lookup_lockres(struct dlm_ctxt *dlm,
hash = full_name_hash(name, len);
- bucket = &(dlm->resources[hash & DLM_HASH_MASK]);
+ bucket = &(dlm->lockres_hash[hash % DLM_HASH_BUCKETS]);
/* check for pre-existing lock */
- list_for_each(iter, bucket) {
- tmpres = list_entry(iter, struct dlm_lock_resource, list);
+ hlist_for_each(iter, bucket) {
+ tmpres = hlist_entry(iter, struct dlm_lock_resource, hash_node);
if (tmpres->lockname.len == len &&
memcmp(tmpres->lockname.name, name, len) == 0) {
dlm_lockres_get(tmpres);
@@ -193,8 +193,8 @@ static int dlm_wait_on_domain_helper(const char *domain)
static void dlm_free_ctxt_mem(struct dlm_ctxt *dlm)
{
- if (dlm->resources)
- free_page((unsigned long) dlm->resources);
+ if (dlm->lockres_hash)
+ free_page((unsigned long) dlm->lockres_hash);
if (dlm->name)
kfree(dlm->name);
@@ -303,10 +303,10 @@ static void dlm_migrate_all_locks(struct dlm_ctxt *dlm)
mlog(0, "Migrating locks from domain %s\n", dlm->name);
restart:
spin_lock(&dlm->spinlock);
- for (i=0; i<DLM_HASH_SIZE; i++) {
- while (!list_empty(&dlm->resources[i])) {
- res = list_entry(dlm->resources[i].next,
- struct dlm_lock_resource, list);
+ for (i = 0; i < DLM_HASH_BUCKETS; i++) {
+ while (!hlist_empty(&dlm->lockres_hash[i])) {
+ res = hlist_entry(dlm->lockres_hash[i].first,
+ struct dlm_lock_resource, hash_node);
/* need reference when manually grabbing lockres */
dlm_lockres_get(res);
/* this should unhash the lockres
@@ -573,8 +573,11 @@ static int dlm_query_join_handler(struct o2net_msg *msg, u32 len, void *data)
spin_lock(&dlm_domain_lock);
dlm = __dlm_lookup_domain_full(query->domain, query->name_len);
/* Once the dlm ctxt is marked as leaving then we don't want
- * to be put in someone's domain map. */
+ * to be put in someone's domain map.
+ * Also, explicitly disallow joining at certain troublesome
+ * times (ie. during recovery). */
if (dlm && dlm->dlm_state != DLM_CTXT_LEAVING) {
+ int bit = query->node_idx;
spin_lock(&dlm->spinlock);
if (dlm->dlm_state == DLM_CTXT_NEW &&
@@ -586,6 +589,19 @@ static int dlm_query_join_handler(struct o2net_msg *msg, u32 len, void *data)
} else if (dlm->joining_node != DLM_LOCK_RES_OWNER_UNKNOWN) {
/* Disallow parallel joins. */
response = JOIN_DISALLOW;
+ } else if (dlm->reco.state & DLM_RECO_STATE_ACTIVE) {
+ mlog(ML_NOTICE, "node %u trying to join, but recovery "
+ "is ongoing.\n", bit);
+ response = JOIN_DISALLOW;
+ } else if (test_bit(bit, dlm->recovery_map)) {
+ mlog(ML_NOTICE, "node %u trying to join, but it "
+ "still needs recovery.\n", bit);
+ response = JOIN_DISALLOW;
+ } else if (test_bit(bit, dlm->domain_map)) {
+ mlog(ML_NOTICE, "node %u trying to join, but it "
+ "is still in the domain! needs recovery?\n",
+ bit);
+ response = JOIN_DISALLOW;
} else {
/* Alright we're fully a part of this domain
* so we keep some state as to who's joining
@@ -1175,18 +1191,17 @@ static struct dlm_ctxt *dlm_alloc_ctxt(const char *domain,
goto leave;
}
- dlm->resources = (struct list_head *) __get_free_page(GFP_KERNEL);
- if (!dlm->resources) {
+ dlm->lockres_hash = (struct hlist_head *) __get_free_page(GFP_KERNEL);
+ if (!dlm->lockres_hash) {
mlog_errno(-ENOMEM);
kfree(dlm->name);
kfree(dlm);
dlm = NULL;
goto leave;
}
- memset(dlm->resources, 0, PAGE_SIZE);
- for (i=0; i<DLM_HASH_SIZE; i++)
- INIT_LIST_HEAD(&dlm->resources[i]);
+ for (i=0; i<DLM_HASH_BUCKETS; i++)
+ INIT_HLIST_HEAD(&dlm->lockres_hash[i]);
strcpy(dlm->name, domain);
dlm->key = key;
diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c
index d1a0038557a..671d4ff222c 100644
--- a/fs/ocfs2/dlm/dlmlock.c
+++ b/fs/ocfs2/dlm/dlmlock.c
@@ -220,6 +220,17 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm,
dlm_error(status);
dlm_revert_pending_lock(res, lock);
dlm_lock_put(lock);
+ } else if (dlm_is_recovery_lock(res->lockname.name,
+ res->lockname.len)) {
+ /* special case for the $RECOVERY lock.
+ * there will never be an AST delivered to put
+ * this lock on the proper secondary queue
+ * (granted), so do it manually. */
+ mlog(0, "%s: $RECOVERY lock for this node (%u) is "
+ "mastered by %u; got lock, manually granting (no ast)\n",
+ dlm->name, dlm->node_num, res->owner);
+ list_del_init(&lock->list);
+ list_add_tail(&lock->list, &res->granted);
}
spin_unlock(&res->spinlock);
@@ -646,7 +657,19 @@ retry_lock:
mlog(0, "retrying lock with migration/"
"recovery/in progress\n");
msleep(100);
- dlm_wait_for_recovery(dlm);
+ /* no waiting for dlm_reco_thread */
+ if (recovery) {
+ if (status == DLM_RECOVERING) {
+ mlog(0, "%s: got RECOVERING "
+ "for $REOCVERY lock, master "
+ "was %u\n", dlm->name,
+ res->owner);
+ dlm_wait_for_node_death(dlm, res->owner,
+ DLM_NODE_DEATH_WAIT_MAX);
+ }
+ } else {
+ dlm_wait_for_recovery(dlm);
+ }
goto retry_lock;
}
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 27e984f7e4c..847dd3cc4cf 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -564,7 +564,7 @@ static void dlm_lockres_release(struct kref *kref)
/* By the time we're ready to blow this guy away, we shouldn't
* be on any lists. */
- BUG_ON(!list_empty(&res->list));
+ BUG_ON(!hlist_unhashed(&res->hash_node));
BUG_ON(!list_empty(&res->granted));
BUG_ON(!list_empty(&res->converting));
BUG_ON(!list_empty(&res->blocked));
@@ -605,7 +605,7 @@ static void dlm_init_lockres(struct dlm_ctxt *dlm,
init_waitqueue_head(&res->wq);
spin_lock_init(&res->spinlock);
- INIT_LIST_HEAD(&res->list);
+ INIT_HLIST_NODE(&res->hash_node);
INIT_LIST_HEAD(&res->granted);
INIT_LIST_HEAD(&res->converting);
INIT_LIST_HEAD(&res->blocked);
@@ -1050,17 +1050,10 @@ static int dlm_restart_lock_mastery(struct dlm_ctxt *dlm,
node = dlm_bitmap_diff_iter_next(&bdi, &sc);
while (node >= 0) {
if (sc == NODE_UP) {
- /* a node came up. easy. might not even need
- * to talk to it if its node number is higher
- * or if we are already blocked. */
- mlog(0, "node up! %d\n", node);
- if (blocked)
- goto next;
-
- if (node > dlm->node_num) {
- mlog(0, "node > this node. skipping.\n");
- goto next;
- }
+ /* a node came up. clear any old vote from
+ * the response map and set it in the vote map
+ * then restart the mastery. */
+ mlog(ML_NOTICE, "node %d up while restarting\n", node);
/* redo the master request, but only for the new node */
mlog(0, "sending request to new node\n");
@@ -2005,6 +1998,15 @@ fail:
break;
mlog(0, "timed out during migration\n");
+ /* avoid hang during shutdown when migrating lockres
+ * to a node which also goes down */
+ if (dlm_is_node_dead(dlm, target)) {
+ mlog(0, "%s:%.*s: expected migration target %u "
+ "is no longer up. restarting.\n",
+ dlm->name, res->lockname.len,
+ res->lockname.name, target);
+ ret = -ERESTARTSYS;
+ }
}
if (ret == -ERESTARTSYS) {
/* migration failed, detach and clean up mle */
@@ -2480,7 +2482,9 @@ top:
atomic_set(&mle->woken, 1);
spin_unlock(&mle->spinlock);
wake_up(&mle->wq);
- /* final put will take care of list removal */
+ /* do not need events any longer, so detach
+ * from heartbeat */
+ __dlm_mle_detach_hb_events(dlm, mle);
__dlm_put_mle(mle);
}
continue;
@@ -2535,6 +2539,9 @@ top:
spin_unlock(&res->spinlock);
dlm_lockres_put(res);
+ /* about to get rid of mle, detach from heartbeat */
+ __dlm_mle_detach_hb_events(dlm, mle);
+
/* dump the mle */
spin_lock(&dlm->master_lock);
__dlm_put_mle(mle);
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 0c8eb1093f0..1e232000f3f 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -39,6 +39,7 @@
#include <linux/inet.h>
#include <linux/timer.h>
#include <linux/kthread.h>
+#include <linux/delay.h>
#include "cluster/heartbeat.h"
@@ -256,6 +257,45 @@ static int dlm_recovery_thread(void *data)
return 0;
}
+/* returns true when the recovery master has contacted us */
+static int dlm_reco_master_ready(struct dlm_ctxt *dlm)
+{
+ int ready;
+ spin_lock(&dlm->spinlock);
+ ready = (dlm->reco.new_master != O2NM_INVALID_NODE_NUM);
+ spin_unlock(&dlm->spinlock);
+ return ready;
+}
+
+/* returns true if node is no longer in the domain
+ * could be dead or just not joined */
+int dlm_is_node_dead(struct dlm_ctxt *dlm, u8 node)
+{
+ int dead;
+ spin_lock(&dlm->spinlock);
+ dead = test_bit(node, dlm->domain_map);
+ spin_unlock(&dlm->spinlock);
+ return dead;
+}
+
+int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
+{
+ if (timeout) {
+ mlog(ML_NOTICE, "%s: waiting %dms for notification of "
+ "death of node %u\n", dlm->name, timeout, node);
+ wait_event_timeout(dlm->dlm_reco_thread_wq,
+ dlm_is_node_dead(dlm, node),
+ msecs_to_jiffies(timeout));
+ } else {
+ mlog(ML_NOTICE, "%s: waiting indefinitely for notification "
+ "of death of node %u\n", dlm->name, node);
+ wait_event(dlm->dlm_reco_thread_wq,
+ dlm_is_node_dead(dlm, node));
+ }
+ /* for now, return 0 */
+ return 0;
+}
+
/* callers of the top-level api calls (dlmlock/dlmunlock) should
* block on the dlm->reco.event when recovery is in progress.
* the dlm recovery thread will set this state when it begins
@@ -297,6 +337,7 @@ static void dlm_end_recovery(struct dlm_ctxt *dlm)
static int dlm_do_recovery(struct dlm_ctxt *dlm)
{
int status = 0;
+ int ret;
spin_lock(&dlm->spinlock);
@@ -343,10 +384,13 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
goto master_here;
if (dlm->reco.new_master == O2NM_INVALID_NODE_NUM) {
- /* choose a new master */
- if (!dlm_pick_recovery_master(dlm)) {
+ /* choose a new master, returns 0 if this node
+ * is the master, -EEXIST if it's another node.
+ * this does not return until a new master is chosen
+ * or recovery completes entirely. */
+ ret = dlm_pick_recovery_master(dlm);
+ if (!ret) {
/* already notified everyone. go. */
- dlm->reco.new_master = dlm->node_num;
goto master_here;
}
mlog(0, "another node will master this recovery session.\n");
@@ -371,8 +415,13 @@ master_here:
if (status < 0) {
mlog(ML_ERROR, "error %d remastering locks for node %u, "
"retrying.\n", status, dlm->reco.dead_node);
+ /* yield a bit to allow any final network messages
+ * to get handled on remaining nodes */
+ msleep(100);
} else {
/* 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);
}
dlm_end_recovery(dlm);
@@ -477,7 +526,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
BUG();
break;
case DLM_RECO_NODE_DATA_DEAD:
- mlog(0, "node %u died after "
+ mlog(ML_NOTICE, "node %u died after "
"requesting recovery info for "
"node %u\n", ndata->node_num,
dead_node);
@@ -485,6 +534,19 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
// start all over
destroy = 1;
status = -EAGAIN;
+ /* instead of spinning like crazy here,
+ * wait for the domain map to catch up
+ * with the network state. otherwise this
+ * can be hit hundreds of times before
+ * the node is really seen as dead. */
+ wait_event_timeout(dlm->dlm_reco_thread_wq,
+ dlm_is_node_dead(dlm,
+ ndata->node_num),
+ msecs_to_jiffies(1000));
+ mlog(0, "waited 1 sec for %u, "
+ "dead? %s\n", ndata->node_num,
+ dlm_is_node_dead(dlm, ndata->node_num) ?
+ "yes" : "no");
goto leave;
case DLM_RECO_NODE_DATA_RECEIVING:
case DLM_RECO_NODE_DATA_REQUESTED:
@@ -678,11 +740,27 @@ static void dlm_request_all_locks_worker(struct dlm_work_item *item, void *data)
dlm = item->dlm;
dead_node = item->u.ral.dead_node;
reco_master = item->u.ral.reco_master;
+ mres = (struct dlm_migratable_lockres *)data;
+
+ if (dead_node != dlm->reco.dead_node ||
+ reco_master != dlm->reco.new_master) {
+ /* show extra debug info if the recovery state is messed */
+ mlog(ML_ERROR, "%s: bad reco state: reco(dead=%u, master=%u), "
+ "request(dead=%u, master=%u)\n",
+ dlm->name, dlm->reco.dead_node, dlm->reco.new_master,
+ dead_node, reco_master);
+ mlog(ML_ERROR, "%s: name=%.*s master=%u locks=%u/%u flags=%u "
+ "entry[0]={c=%"MLFu64",l=%u,f=%u,t=%d,ct=%d,hb=%d,n=%u}\n",
+ dlm->name, mres->lockname_len, mres->lockname, mres->master,
+ mres->num_locks, mres->total_locks, mres->flags,
+ mres->ml[0].cookie, mres->ml[0].list, mres->ml[0].flags,
+ mres->ml[0].type, mres->ml[0].convert_type,
+ mres->ml[0].highest_blocked, mres->ml[0].node);
+ BUG();
+ }
BUG_ON(dead_node != dlm->reco.dead_node);
BUG_ON(reco_master != dlm->reco.new_master);
- mres = (struct dlm_migratable_lockres *)data;
-
/* lock resources should have already been moved to the
* dlm->reco.resources list. now move items from that list
* to a temp list if the dead owner matches. note that the
@@ -757,15 +835,18 @@ int dlm_reco_data_done_handler(struct o2net_msg *msg, u32 len, void *data)
continue;
switch (ndata->state) {
+ /* should have moved beyond INIT but not to FINALIZE yet */
case DLM_RECO_NODE_DATA_INIT:
case DLM_RECO_NODE_DATA_DEAD:
- case DLM_RECO_NODE_DATA_DONE:
case DLM_RECO_NODE_DATA_FINALIZE_SENT:
mlog(ML_ERROR, "bad ndata state for node %u:"
" state=%d\n", ndata->node_num,
ndata->state);
BUG();
break;
+ /* these states are possible at this point, anywhere along
+ * the line of recovery */
+ case DLM_RECO_NODE_DATA_DONE:
case DLM_RECO_NODE_DATA_RECEIVING:
case DLM_RECO_NODE_DATA_REQUESTED:
case DLM_RECO_NODE_DATA_REQUESTING:
@@ -799,13 +880,31 @@ static void dlm_move_reco_locks_to_list(struct dlm_ctxt *dlm,
{
struct dlm_lock_resource *res;
struct list_head *iter, *iter2;
+ struct dlm_lock *lock;
spin_lock(&dlm->spinlock);
list_for_each_safe(iter, iter2, &dlm->reco.resources) {
res = list_entry (iter, struct dlm_lock_resource, recovering);
+ /* always prune any $RECOVERY entries for dead nodes,
+ * otherwise hangs can occur during later recovery */
if (dlm_is_recovery_lock(res->lockname.name,
- res->lockname.len))
+ res->lockname.len)) {
+ spin_lock(&res->spinlock);
+ list_for_each_entry(lock, &res->granted, list) {
+ if (lock->ml.node == dead_node) {
+ mlog(0, "AHA! there was "
+ "a $RECOVERY lock for dead "
+ "node %u (%s)!\n",
+ dead_node, dlm->name);
+ list_del_init(&lock->list);
+ dlm_lock_put(lock);
+ break;
+ }
+ }
+ spin_unlock(&res->spinlock);
continue;
+ }
+
if (res->owner == dead_node) {
mlog(0, "found lockres owned by dead node while "
"doing recovery for node %u. sending it.\n",
@@ -1179,7 +1278,7 @@ static void dlm_mig_lockres_worker(struct dlm_work_item *item, void *data)
again:
ret = dlm_lockres_master_requery(dlm, res, &real_master);
if (ret < 0) {
- mlog(0, "dlm_lockres_master_requery failure: %d\n",
+ mlog(0, "dlm_lockres_master_requery ret=%d\n",
ret);
goto again;
}
@@ -1594,7 +1693,10 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
u8 dead_node, u8 new_master)
{
int i;
- struct list_head *iter, *iter2, *bucket;
+ struct list_head *iter, *iter2;
+ struct hlist_node *hash_iter;
+ struct hlist_head *bucket;
+
struct dlm_lock_resource *res;
mlog_entry_void();
@@ -1618,10 +1720,9 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
* for now we need to run the whole hash, clear
* the RECOVERING state and set the owner
* if necessary */
- for (i=0; i<DLM_HASH_SIZE; i++) {
- bucket = &(dlm->resources[i]);
- list_for_each(iter, bucket) {
- res = list_entry (iter, struct dlm_lock_resource, list);
+ for (i = 0; i < DLM_HASH_BUCKETS; i++) {
+ bucket = &(dlm->lockres_hash[i]);
+ hlist_for_each_entry(res, hash_iter, bucket, hash_node) {
if (res->state & DLM_LOCK_RES_RECOVERING) {
if (res->owner == dead_node) {
mlog(0, "(this=%u) res %.*s owner=%u "
@@ -1753,10 +1854,11 @@ static void dlm_free_dead_locks(struct dlm_ctxt *dlm,
static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node)
{
- struct list_head *iter;
+ struct hlist_node *iter;
struct dlm_lock_resource *res;
int i;
- struct list_head *bucket;
+ struct hlist_head *bucket;
+ struct dlm_lock *lock;
/* purge any stale mles */
@@ -1776,14 +1878,28 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node)
* can be kicked again to see if any ASTs or BASTs
* need to be fired as a result.
*/
- for (i=0; i<DLM_HASH_SIZE; i++) {
- bucket = &(dlm->resources[i]);
- list_for_each(iter, bucket) {
- res = list_entry (iter, struct dlm_lock_resource, list);
+ for (i = 0; i < DLM_HASH_BUCKETS; i++) {
+ bucket = &(dlm->lockres_hash[i]);
+ hlist_for_each_entry(res, iter, bucket, hash_node) {
+ /* always prune any $RECOVERY entries for dead nodes,
+ * otherwise hangs can occur during later recovery */
if (dlm_is_recovery_lock(res->lockname.name,
- res->lockname.len))
+ res->lockname.len)) {
+ spin_lock(&res->spinlock);
+ list_for_each_entry(lock, &res->granted, list) {
+ if (lock->ml.node == dead_node) {
+ mlog(0, "AHA! there was "
+ "a $RECOVERY lock for dead "
+ "node %u (%s)!\n",
+ dead_node, dlm->name);
+ list_del_init(&lock->list);
+ dlm_lock_put(lock);
+ break;
+ }
+ }
+ spin_unlock(&res->spinlock);
continue;
-
+ }
spin_lock(&res->spinlock);
/* zero the lvb if necessary */
dlm_revalidate_lvb(dlm, res, dead_node);
@@ -1869,12 +1985,9 @@ void dlm_hb_node_up_cb(struct o2nm_node *node, int idx, void *data)
return;
spin_lock(&dlm->spinlock);
-
set_bit(idx, dlm->live_nodes_map);
-
- /* notify any mles attached to the heartbeat events */
- dlm_hb_event_notify_attached(dlm, idx, 1);
-
+ /* do NOT notify mle attached to the heartbeat events.
+ * new nodes are not interesting in mastery until joined. */
spin_unlock(&dlm->spinlock);
dlm_put(dlm);
@@ -1897,7 +2010,18 @@ static void dlm_reco_unlock_ast(void *astdata, enum dlm_status st)
mlog(0, "unlockast for recovery lock fired!\n");
}
-
+/*
+ * dlm_pick_recovery_master will continually attempt to use
+ * dlmlock() on the special "$RECOVERY" lockres with the
+ * LKM_NOQUEUE flag to get an EX. every thread that enters
+ * this function on each node racing to become the recovery
+ * master will not stop attempting this until either:
+ * a) this node gets the EX (and becomes the recovery master),
+ * or b) dlm->reco.new_master gets set to some nodenum
+ * != O2NM_INVALID_NODE_NUM (another node will do the reco).
+ * so each time a recovery master is needed, the entire cluster
+ * will sync at this point. if the new master dies, that will
+ * be detected in dlm_do_recovery */
static int dlm_pick_recovery_master(struct dlm_ctxt *dlm)
{
enum dlm_status ret;
@@ -1906,23 +2030,69 @@ static int dlm_pick_recovery_master(struct dlm_ctxt *dlm)
mlog(0, "starting recovery of %s at %lu, dead=%u, this=%u\n",
dlm->name, jiffies, dlm->reco.dead_node, dlm->node_num);
-retry:
+again:
memset(&lksb, 0, sizeof(lksb));
ret = dlmlock(dlm, LKM_EXMODE, &lksb, LKM_NOQUEUE|LKM_RECOVERY,
DLM_RECOVERY_LOCK_NAME, dlm_reco_ast, dlm, dlm_reco_bast);
+ mlog(0, "%s: dlmlock($RECOVERY) returned %d, lksb=%d\n",
+ dlm->name, ret, lksb.status);
+
if (ret == DLM_NORMAL) {
mlog(0, "dlm=%s dlmlock says I got it (this=%u)\n",
dlm->name, dlm->node_num);
- /* I am master, send message to all nodes saying
- * that I am beginning a recovery session */
- status = dlm_send_begin_reco_message(dlm,
- dlm->reco.dead_node);
+
+ /* got the EX lock. check to see if another node
+ * just became the reco master */
+ if (dlm_reco_master_ready(dlm)) {
+ mlog(0, "%s: got reco EX lock, but %u will "
+ "do the recovery\n", dlm->name,
+ dlm->reco.new_master);
+ status = -EEXIST;
+ } else {
+ status = 0;
+
+ /* see if recovery was already finished elsewhere */
+ spin_lock(&dlm->spinlock);
+ if (dlm->reco.dead_node == O2NM_INVALID_NODE_NUM) {
+ status = -EINVAL;
+ mlog(0, "%s: got reco EX lock, but "
+ "node got recovered already\n", dlm->name);
+ if (dlm->reco.new_master != O2NM_INVALID_NODE_NUM) {
+ mlog(ML_ERROR, "%s: new master is %u "
+ "but no dead node!\n",
+ dlm->name, dlm->reco.new_master);
+ BUG();
+ }
+ }
+ spin_unlock(&dlm->spinlock);
+ }
+
+ /* if this node has actually become the recovery master,
+ * set the master and send the messages to begin recovery */
+ if (!status) {
+ mlog(0, "%s: dead=%u, this=%u, sending "
+ "begin_reco now\n", dlm->name,
+ dlm->reco.dead_node, dlm->node_num);
+ status = dlm_send_begin_reco_message(dlm,
+ dlm->reco.dead_node);
+ /* this always succeeds */
+ BUG_ON(status);
+
+ /* set the new_master to this node */
+ spin_lock(&dlm->spinlock);
+ dlm->reco.new_master = dlm->node_num;
+ spin_unlock(&dlm->spinlock);
+ }
/* recovery lock is a special case. ast will not get fired,
* so just go ahead and unlock it. */
ret = dlmunlock(dlm, &lksb, 0, dlm_reco_unlock_ast, dlm);
+ if (ret == DLM_DENIED) {
+ mlog(0, "got DLM_DENIED, trying LKM_CANCEL\n");
+ ret = dlmunlock(dlm, &lksb, LKM_CANCEL, dlm_reco_unlock_ast, dlm);
+ }
if (ret != DLM_NORMAL) {
/* this would really suck. this could only happen
* if there was a network error during the unlock
@@ -1930,20 +2100,42 @@ retry:
* is actually "done" and the lock structure is
* even freed. we can continue, but only
* because this specific lock name is special. */
- mlog(0, "dlmunlock returned %d\n", ret);
- }
-
- if (status < 0) {
- mlog(0, "failed to send recovery message. "
- "must retry with new node map.\n");
- goto retry;
+ mlog(ML_ERROR, "dlmunlock returned %d\n", ret);
}
} else if (ret == DLM_NOTQUEUED) {
mlog(0, "dlm=%s dlmlock says another node got it (this=%u)\n",
dlm->name, dlm->node_num);
/* another node is master. wait on
- * reco.new_master != O2NM_INVALID_NODE_NUM */
+ * reco.new_master != O2NM_INVALID_NODE_NUM
+ * for at most one second */
+ wait_event_timeout(dlm->dlm_reco_thread_wq,
+ dlm_reco_master_ready(dlm),
+ msecs_to_jiffies(1000));
+ if (!dlm_reco_master_ready(dlm)) {
+ mlog(0, "%s: reco master taking awhile\n",
+ dlm->name);
+ goto again;
+ }
+ /* another node has informed this one that it is reco master */
+ mlog(0, "%s: reco master %u is ready to recover %u\n",
+ dlm->name, dlm->reco.new_master, dlm->reco.dead_node);
status = -EEXIST;
+ } else {
+ struct dlm_lock_resource *res;
+
+ /* dlmlock returned something other than NOTQUEUED or NORMAL */
+ mlog(ML_ERROR, "%s: got %s from dlmlock($RECOVERY), "
+ "lksb.status=%s\n", dlm->name, dlm_errname(ret),
+ dlm_errname(lksb.status));
+ res = dlm_lookup_lockres(dlm, DLM_RECOVERY_LOCK_NAME,
+ DLM_RECOVERY_LOCK_NAME_LEN);
+ if (res) {
+ dlm_print_one_lock_resource(res);
+ dlm_lockres_put(res);
+ } else {
+ mlog(ML_ERROR, "recovery lock not found\n");
+ }
+ BUG();
}
return status;
@@ -1982,7 +2174,7 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node)
mlog(0, "not sending begin reco to self\n");
continue;
}
-
+retry:
ret = -EINVAL;
mlog(0, "attempting to send begin reco msg to %d\n",
nodenum);
@@ -1991,8 +2183,17 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node)
/* negative status is handled ok by caller here */
if (ret >= 0)
ret = status;
+ if (dlm_is_host_down(ret)) {
+ /* node is down. not involved in recovery
+ * so just keep going */
+ mlog(0, "%s: node %u was down when sending "
+ "begin reco msg (%d)\n", dlm->name, nodenum, ret);
+ ret = 0;
+ }
if (ret < 0) {
struct dlm_lock_resource *res;
+ /* this is now a serious problem, possibly ENOMEM
+ * in the network stack. must retry */
mlog_errno(ret);
mlog(ML_ERROR, "begin reco of dlm %s to node %u "
" returned %d\n", dlm->name, nodenum, ret);
@@ -2004,7 +2205,10 @@ static int dlm_send_begin_reco_message(struct dlm_ctxt *dlm, u8 dead_node)
} else {
mlog(ML_ERROR, "recovery lock not found\n");
}
- break;
+ /* sleep for a bit in hopes that we can avoid
+ * another ENOMEM */
+ msleep(100);
+ goto retry;
}
}
@@ -2027,19 +2231,34 @@ int dlm_begin_reco_handler(struct o2net_msg *msg, u32 len, void *data)
spin_lock(&dlm->spinlock);
if (dlm->reco.new_master != O2NM_INVALID_NODE_NUM) {
- mlog(0, "new_master already set to %u!\n",
- dlm->reco.new_master);
+ if (test_bit(dlm->reco.new_master, dlm->recovery_map)) {
+ mlog(0, "%s: new_master %u died, changing "
+ "to %u\n", dlm->name, dlm->reco.new_master,
+ br->node_idx);
+ } else {
+ mlog(0, "%s: new_master %u NOT DEAD, changing "
+ "to %u\n", dlm->name, dlm->reco.new_master,
+ br->node_idx);
+ /* may not have seen the new master as dead yet */
+ }
}
if (dlm->reco.dead_node != O2NM_INVALID_NODE_NUM) {
- mlog(0, "dead_node already set to %u!\n",
- dlm->reco.dead_node);
+ mlog(ML_NOTICE, "%s: dead_node previously set to %u, "
+ "node %u changing it to %u\n", dlm->name,
+ dlm->reco.dead_node, br->node_idx, br->dead_node);
}
dlm->reco.new_master = br->node_idx;
dlm->reco.dead_node = br->dead_node;
if (!test_bit(br->dead_node, dlm->recovery_map)) {
- mlog(ML_ERROR, "recovery master %u sees %u as dead, but this "
+ mlog(0, "recovery master %u sees %u as dead, but this "
"node has not yet. marking %u as dead\n",
br->node_idx, br->dead_node, br->dead_node);
+ if (!test_bit(br->dead_node, dlm->domain_map) ||
+ !test_bit(br->dead_node, dlm->live_nodes_map))
+ mlog(0, "%u not in domain/live_nodes map "
+ "so setting it in reco map manually\n",
+ br->dead_node);
+ set_bit(br->dead_node, dlm->recovery_map);
__dlm_hb_node_down(dlm, br->dead_node);
}
spin_unlock(&dlm->spinlock);
diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c
index cec2ce1cd31..c95f08d2e92 100644
--- a/fs/ocfs2/dlm/dlmunlock.c
+++ b/fs/ocfs2/dlm/dlmunlock.c
@@ -188,6 +188,19 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm,
actions &= ~(DLM_UNLOCK_REMOVE_LOCK|
DLM_UNLOCK_REGRANT_LOCK|
DLM_UNLOCK_CLEAR_CONVERT_TYPE);
+ } else if (status == DLM_RECOVERING ||
+ status == DLM_MIGRATING ||
+ status == DLM_FORWARD) {
+ /* must clear the actions because this unlock
+ * is about to be retried. cannot free or do
+ * any list manipulation. */
+ mlog(0, "%s:%.*s: clearing actions, %s\n",
+ dlm->name, res->lockname.len,
+ res->lockname.name,
+ status==DLM_RECOVERING?"recovering":
+ (status==DLM_MIGRATING?"migrating":
+ "forward"));
+ actions = 0;
}
if (flags & LKM_CANCEL)
lock->cancel_pending = 0;
diff --git a/fs/ocfs2/dlm/userdlm.c b/fs/ocfs2/dlm/userdlm.c
index e1fdd288796..c3764f4744e 100644
--- a/fs/ocfs2/dlm/userdlm.c
+++ b/fs/ocfs2/dlm/userdlm.c
@@ -27,7 +27,7 @@
* Boston, MA 021110-1307, USA.
*/
-#include <asm/signal.h>
+#include <linux/signal.h>
#include <linux/module.h>
#include <linux/fs.h>
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index f2fb40cd296..e6f207eebab 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -181,6 +181,12 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode,
ret = -EBADR;
if (rec_end > OCFS2_I(inode)->ip_clusters) {
mlog_errno(ret);
+ ocfs2_error(inode->i_sb,
+ "Extent %d at e_blkno %"MLFu64" of inode %"MLFu64" goes past ip_clusters of %u\n",
+ i,
+ le64_to_cpu(rec->e_blkno),
+ OCFS2_I(inode)->ip_blkno,
+ OCFS2_I(inode)->ip_clusters);
goto out_free;
}
@@ -226,6 +232,12 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode,
ret = -EBADR;
if (blkno) {
mlog_errno(ret);
+ ocfs2_error(inode->i_sb,
+ "Multiple extents for (cpos = %u, clusters = %u) on inode %"MLFu64"; e_blkno %"MLFu64" and rec %d at e_blkno %"MLFu64"\n",
+ cpos, clusters,
+ OCFS2_I(inode)->ip_blkno,
+ blkno, i,
+ le64_to_cpu(rec->e_blkno));
goto out_free;
}
@@ -238,6 +250,10 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode,
*/
ret = -EBADR;
if (!blkno) {
+ ocfs2_error(inode->i_sb,
+ "No record found for (cpos = %u, clusters = %u) on inode %"MLFu64"\n",
+ cpos, clusters,
+ OCFS2_I(inode)->ip_blkno);
mlog_errno(ret);
goto out_free;
}
@@ -262,11 +278,24 @@ static int ocfs2_extent_map_find_leaf(struct inode *inode,
el = &eb->h_list;
}
- if (el->l_tree_depth)
- BUG();
+ BUG_ON(el->l_tree_depth);
for (i = 0; i < le16_to_cpu(el->l_next_free_rec); i++) {
rec = &el->l_recs[i];
+
+ if ((le32_to_cpu(rec->e_cpos) + le32_to_cpu(rec->e_clusters)) >
+ OCFS2_I(inode)->ip_clusters) {
+ ret = -EBADR;
+ mlog_errno(ret);
+ ocfs2_error(inode->i_sb,
+ "Extent %d at e_blkno %"MLFu64" of inode %"MLFu64" goes past ip_clusters of %u\n",
+ i,
+ le64_to_cpu(rec->e_blkno),
+ OCFS2_I(inode)->ip_blkno,
+ OCFS2_I(inode)->ip_clusters);
+ return ret;
+ }
+
ret = ocfs2_extent_map_insert(inode, rec,
le16_to_cpu(el->l_tree_depth));
if (ret) {
@@ -364,8 +393,8 @@ static int ocfs2_extent_map_lookup_read(struct inode *inode,
return ret;
}
- if (ent->e_tree_depth)
- BUG(); /* FIXME: Make sure this isn't a corruption */
+ /* FIXME: Make sure this isn't a corruption */
+ BUG_ON(ent->e_tree_depth);
*ret_ent = ent;
@@ -423,8 +452,7 @@ static int ocfs2_extent_map_try_insert(struct inode *inode,
le32_to_cpu(rec->e_clusters), NULL,
NULL);
- if (!old_ent)
- BUG();
+ BUG_ON(!old_ent);
ret = -EEXIST;
if (old_ent->e_tree_depth < tree_depth)
@@ -528,6 +556,10 @@ static int ocfs2_extent_map_insert(struct inode *inode,
OCFS2_I(inode)->ip_map.em_clusters) {
ret = -EBADR;
mlog_errno(ret);
+ ocfs2_error(inode->i_sb,
+ "Zero e_clusters on non-tail extent record at e_blkno %"MLFu64" on inode %"MLFu64"\n",
+ le64_to_cpu(rec->e_blkno),
+ OCFS2_I(inode)->ip_blkno);
return ret;
}
@@ -590,12 +622,12 @@ static int ocfs2_extent_map_insert(struct inode *inode,
* Existing record in the extent map:
*
* cpos = 10, len = 10
- * |---------|
+ * |---------|
*
* New Record:
*
* cpos = 10, len = 20
- * |------------------|
+ * |------------------|
*
* The passed record is the new on-disk record. The new_clusters value
* is how many clusters were added to the file. If the append is a
@@ -988,7 +1020,7 @@ int __init init_ocfs2_extent_maps(void)
return 0;
}
-void __exit exit_ocfs2_extent_maps(void)
+void exit_ocfs2_extent_maps(void)
{
kmem_cache_destroy(ocfs2_em_ent_cachep);
}
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index eaf33caa0a1..8a4048b55fd 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -933,9 +933,6 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
struct file *filp = iocb->ki_filp;
struct inode *inode = filp->f_dentry->d_inode;
loff_t newsize, saved_pos;
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-#endif
mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", filp, buf,
(unsigned int)count,
@@ -951,14 +948,6 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
return -EIO;
}
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- /* ugh, work around some applications which open everything O_DIRECT +
- * O_APPEND and really don't mean to use O_DIRECT. */
- if (osb->s_mount_opt & OCFS2_MOUNT_COMPAT_OCFS &&
- (filp->f_flags & O_APPEND) && (filp->f_flags & O_DIRECT))
- filp->f_flags &= ~O_DIRECT;
-#endif
-
mutex_lock(&inode->i_mutex);
/* to match setattr's i_mutex -> i_alloc_sem -> rw_lock ordering */
if (filp->f_flags & O_DIRECT) {
@@ -1022,8 +1011,9 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
}
newsize = count + saved_pos;
- mlog(0, "pos=%lld newsize=%"MLFu64" cursize=%lld\n",
- saved_pos, newsize, i_size_read(inode));
+ mlog(0, "pos=%lld newsize=%lld cursize=%lld\n",
+ (long long) saved_pos, (long long) newsize,
+ (long long) i_size_read(inode));
/* No need for a higher level metadata lock if we're
* never going past i_size. */
@@ -1042,8 +1032,9 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
spin_unlock(&OCFS2_I(inode)->ip_lock);
mlog(0, "Writing at EOF, may need more allocation: "
- "i_size = %lld, newsize = %"MLFu64", need %u clusters\n",
- i_size_read(inode), newsize, clusters);
+ "i_size = %lld, newsize = %lld, need %u clusters\n",
+ (long long) i_size_read(inode), (long long) newsize,
+ clusters);
/* We only want to continue the rest of this loop if
* our extend will actually require more
@@ -1077,27 +1068,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
/* communicate with ocfs2_dio_end_io */
ocfs2_iocb_set_rw_locked(iocb);
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- if (osb->s_mount_opt & OCFS2_MOUNT_COMPAT_OCFS &&
- filp->f_flags & O_DIRECT) {
- unsigned int saved_flags = filp->f_flags;
- int sector_size = 1 << osb->s_sectsize_bits;
-
- if ((saved_pos & (sector_size - 1)) ||
- (count & (sector_size - 1)) ||
- ((unsigned long)buf & (sector_size - 1))) {
- filp->f_flags |= O_SYNC;
- filp->f_flags &= ~O_DIRECT;
- }
-
- ret = generic_file_aio_write_nolock(iocb, &local_iov, 1,
- &iocb->ki_pos);
-
- filp->f_flags = saved_flags;
- } else
-#endif
- ret = generic_file_aio_write_nolock(iocb, &local_iov, 1,
- &iocb->ki_pos);
+ ret = generic_file_aio_write_nolock(iocb, &local_iov, 1, &iocb->ki_pos);
/* buffered aio wouldn't have proper lock coverage today */
BUG_ON(ret == -EIOCBQUEUED && !(filp->f_flags & O_DIRECT));
@@ -1138,9 +1109,6 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
int ret = 0, rw_level = -1, have_alloc_sem = 0;
struct file *filp = iocb->ki_filp;
struct inode *inode = filp->f_dentry->d_inode;
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
-#endif
mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", filp, buf,
(unsigned int)count,
@@ -1153,21 +1121,6 @@ static ssize_t ocfs2_file_aio_read(struct kiocb *iocb,
goto bail;
}
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- if (osb->s_mount_opt & OCFS2_MOUNT_COMPAT_OCFS) {
- if (filp->f_flags & O_DIRECT) {
- int sector_size = 1 << osb->s_sectsize_bits;
-
- if ((pos & (sector_size - 1)) ||
- (count & (sector_size - 1)) ||
- ((unsigned long)buf & (sector_size - 1)) ||
- (i_size_read(inode) & (sector_size -1))) {
- filp->f_flags &= ~O_DIRECT;
- }
- }
- }
-#endif
-
/*
* buffered reads protect themselves in ->readpage(). O_DIRECT reads
* need locks to protect pending reads from racing with truncate.
diff --git a/fs/ocfs2/heartbeat.c b/fs/ocfs2/heartbeat.c
index 0bbd22f46c8..cbfd45a97a6 100644
--- a/fs/ocfs2/heartbeat.c
+++ b/fs/ocfs2/heartbeat.c
@@ -67,6 +67,7 @@ void ocfs2_init_node_maps(struct ocfs2_super *osb)
ocfs2_node_map_init(&osb->mounted_map);
ocfs2_node_map_init(&osb->recovery_map);
ocfs2_node_map_init(&osb->umount_map);
+ ocfs2_node_map_init(&osb->osb_recovering_orphan_dirs);
}
static void ocfs2_do_node_down(int node_num,
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index d4ecc062771..315472a5c19 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -41,6 +41,7 @@
#include "dlmglue.h"
#include "extent_map.h"
#include "file.h"
+#include "heartbeat.h"
#include "inode.h"
#include "journal.h"
#include "namei.h"
@@ -544,6 +545,42 @@ bail:
return status;
}
+/*
+ * Serialize with orphan dir recovery. If the process doing
+ * recovery on this orphan dir does an iget() with the dir
+ * i_mutex held, we'll deadlock here. Instead we detect this
+ * and exit early - recovery will wipe this inode for us.
+ */
+static int ocfs2_check_orphan_recovery_state(struct ocfs2_super *osb,
+ int slot)
+{
+ int ret = 0;
+
+ spin_lock(&osb->osb_lock);
+ if (ocfs2_node_map_test_bit(osb, &osb->osb_recovering_orphan_dirs, slot)) {
+ mlog(0, "Recovery is happening on orphan dir %d, will skip "
+ "this inode\n", slot);
+ ret = -EDEADLK;
+ goto out;
+ }
+ /* This signals to the orphan recovery process that it should
+ * wait for us to handle the wipe. */
+ osb->osb_orphan_wipes[slot]++;
+out:
+ spin_unlock(&osb->osb_lock);
+ return ret;
+}
+
+static void ocfs2_signal_wipe_completion(struct ocfs2_super *osb,
+ int slot)
+{
+ spin_lock(&osb->osb_lock);
+ osb->osb_orphan_wipes[slot]--;
+ spin_unlock(&osb->osb_lock);
+
+ wake_up(&osb->osb_wipe_event);
+}
+
static int ocfs2_wipe_inode(struct inode *inode,
struct buffer_head *di_bh)
{
@@ -555,6 +592,11 @@ static int ocfs2_wipe_inode(struct inode *inode,
/* We've already voted on this so it should be readonly - no
* spinlock needed. */
orphaned_slot = OCFS2_I(inode)->ip_orphaned_slot;
+
+ status = ocfs2_check_orphan_recovery_state(osb, orphaned_slot);
+ if (status)
+ return status;
+
orphan_dir_inode = ocfs2_get_system_file_inode(osb,
ORPHAN_DIR_SYSTEM_INODE,
orphaned_slot);
@@ -597,6 +639,7 @@ bail_unlock_dir:
brelse(orphan_dir_bh);
bail:
iput(orphan_dir_inode);
+ ocfs2_signal_wipe_completion(osb, orphaned_slot);
return status;
}
@@ -822,7 +865,8 @@ void ocfs2_delete_inode(struct inode *inode)
status = ocfs2_wipe_inode(inode, di_bh);
if (status < 0) {
- mlog_errno(status);
+ if (status != -EDEADLK)
+ mlog_errno(status);
goto bail_unlock_inode;
}
@@ -903,10 +947,10 @@ void ocfs2_clear_inode(struct inode *inode)
"Clear inode of %"MLFu64", inode is locked\n",
oi->ip_blkno);
- mlog_bug_on_msg(down_trylock(&oi->ip_io_sem),
- "Clear inode of %"MLFu64", io_sem is locked\n",
+ mlog_bug_on_msg(!mutex_trylock(&oi->ip_io_mutex),
+ "Clear inode of %"MLFu64", io_mutex is locked\n",
oi->ip_blkno);
- up(&oi->ip_io_sem);
+ mutex_unlock(&oi->ip_io_mutex);
/*
* down_trylock() returns 0, down_write_trylock() returns 1
diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 9b017743365..84c50796128 100644
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -46,10 +46,10 @@ struct ocfs2_inode_info
struct list_head ip_io_markers;
int ip_orphaned_slot;
- struct semaphore ip_io_sem;
+ struct mutex ip_io_mutex;
/* Used by the journalling code to attach an inode to a
- * handle. These are protected by ip_io_sem in order to lock
+ * handle. These are protected by ip_io_mutex in order to lock
* out other I/O to the inode until we either commit or
* abort. */
struct list_head ip_handle_list;
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 303c8d96457..4be801f4559 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -147,8 +147,7 @@ struct ocfs2_journal_handle *ocfs2_start_trans(struct ocfs2_super *osb,
mlog_entry("(max_buffs = %d)\n", max_buffs);
- if (!osb || !osb->journal->j_journal)
- BUG();
+ BUG_ON(!osb || !osb->journal->j_journal);
if (ocfs2_is_hard_readonly(osb)) {
ret = -EROFS;
@@ -401,7 +400,7 @@ int ocfs2_journal_access(struct ocfs2_journal_handle *handle,
* j_trans_barrier for us. */
ocfs2_set_inode_lock_trans(OCFS2_SB(inode->i_sb)->journal, inode);
- down(&OCFS2_I(inode)->ip_io_sem);
+ mutex_lock(&OCFS2_I(inode)->ip_io_mutex);
switch (type) {
case OCFS2_JOURNAL_ACCESS_CREATE:
case OCFS2_JOURNAL_ACCESS_WRITE:
@@ -416,7 +415,7 @@ int ocfs2_journal_access(struct ocfs2_journal_handle *handle,
status = -EINVAL;
mlog(ML_ERROR, "Uknown access type!\n");
}
- up(&OCFS2_I(inode)->ip_io_sem);
+ mutex_unlock(&OCFS2_I(inode)->ip_io_mutex);
if (status < 0)
mlog(ML_ERROR, "Error %d getting %d access to buffer!\n",
@@ -561,7 +560,11 @@ int ocfs2_journal_init(struct ocfs2_journal *journal, int *dirty)
SET_INODE_JOURNAL(inode);
OCFS2_I(inode)->ip_open_count++;
- status = ocfs2_meta_lock(inode, NULL, &bh, 1);
+ /* Skip recovery waits here - journal inode metadata never
+ * changes in a live cluster so it can be considered an
+ * exception to the rule. */
+ status = ocfs2_meta_lock_full(inode, NULL, &bh, 1,
+ OCFS2_META_LOCK_RECOVERY);
if (status < 0) {
if (status != -ERESTARTSYS)
mlog(ML_ERROR, "Could not get lock on journal!\n");
@@ -672,8 +675,7 @@ void ocfs2_journal_shutdown(struct ocfs2_super *osb)
mlog_entry_void();
- if (!osb)
- BUG();
+ BUG_ON(!osb);
journal = osb->journal;
if (!journal)
@@ -805,8 +807,7 @@ int ocfs2_journal_wipe(struct ocfs2_journal *journal, int full)
mlog_entry_void();
- if (!journal)
- BUG();
+ BUG_ON(!journal);
status = journal_wipe(journal->j_journal, full);
if (status < 0) {
@@ -1072,10 +1073,10 @@ restart:
NULL);
bail:
- down(&osb->recovery_lock);
+ mutex_lock(&osb->recovery_lock);
if (!status &&
!ocfs2_node_map_is_empty(osb, &osb->recovery_map)) {
- up(&osb->recovery_lock);
+ mutex_unlock(&osb->recovery_lock);
goto restart;
}
@@ -1083,7 +1084,7 @@ bail:
mb(); /* sync with ocfs2_recovery_thread_running */
wake_up(&osb->recovery_event);
- up(&osb->recovery_lock);
+ mutex_unlock(&osb->recovery_lock);
mlog_exit(status);
/* no one is callint kthread_stop() for us so the kthread() api
@@ -1098,7 +1099,7 @@ void ocfs2_recovery_thread(struct ocfs2_super *osb, int node_num)
mlog_entry("(node_num=%d, osb->node_num = %d)\n",
node_num, osb->node_num);
- down(&osb->recovery_lock);
+ mutex_lock(&osb->recovery_lock);
if (osb->disable_recovery)
goto out;
@@ -1120,7 +1121,7 @@ void ocfs2_recovery_thread(struct ocfs2_super *osb, int node_num)
}
out:
- up(&osb->recovery_lock);
+ mutex_unlock(&osb->recovery_lock);
wake_up(&osb->recovery_event);
mlog_exit_void();
@@ -1271,8 +1272,7 @@ static int ocfs2_recover_node(struct ocfs2_super *osb,
/* Should not ever be called to recover ourselves -- in that
* case we should've called ocfs2_journal_load instead. */
- if (osb->node_num == node_num)
- BUG();
+ BUG_ON(osb->node_num == node_num);
slot_num = ocfs2_node_num_to_slot(si, node_num);
if (slot_num == OCFS2_INVALID_SLOT) {
@@ -1408,21 +1408,17 @@ bail:
return status;
}
-static int ocfs2_recover_orphans(struct ocfs2_super *osb,
- int slot)
+static int ocfs2_queue_orphans(struct ocfs2_super *osb,
+ int slot,
+ struct inode **head)
{
- int status = 0;
- int have_disk_lock = 0;
- struct inode *inode = NULL;
- struct inode *iter;
+ int status;
struct inode *orphan_dir_inode = NULL;
+ struct inode *iter;
unsigned long offset, blk, local;
struct buffer_head *bh = NULL;
struct ocfs2_dir_entry *de;
struct super_block *sb = osb->sb;
- struct ocfs2_inode_info *oi;
-
- mlog(0, "Recover inodes from orphan dir in slot %d\n", slot);
orphan_dir_inode = ocfs2_get_system_file_inode(osb,
ORPHAN_DIR_SYSTEM_INODE,
@@ -1430,17 +1426,15 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
if (!orphan_dir_inode) {
status = -ENOENT;
mlog_errno(status);
- goto out;
- }
+ return status;
+ }
mutex_lock(&orphan_dir_inode->i_mutex);
status = ocfs2_meta_lock(orphan_dir_inode, NULL, NULL, 0);
if (status < 0) {
- mutex_unlock(&orphan_dir_inode->i_mutex);
mlog_errno(status);
goto out;
}
- have_disk_lock = 1;
offset = 0;
iter = NULL;
@@ -1451,11 +1445,10 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
if (!bh)
status = -EINVAL;
if (status < 0) {
- mutex_unlock(&orphan_dir_inode->i_mutex);
if (bh)
brelse(bh);
mlog_errno(status);
- goto out;
+ goto out_unlock;
}
local = 0;
@@ -1465,11 +1458,10 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
if (!ocfs2_check_dir_entry(orphan_dir_inode,
de, bh, local)) {
- mutex_unlock(&orphan_dir_inode->i_mutex);
status = -EINVAL;
mlog_errno(status);
brelse(bh);
- goto out;
+ goto out_unlock;
}
local += le16_to_cpu(de->rec_len);
@@ -1504,18 +1496,95 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
mlog(0, "queue orphan %"MLFu64"\n",
OCFS2_I(iter)->ip_blkno);
- OCFS2_I(iter)->ip_next_orphan = inode;
- inode = iter;
+ /* No locking is required for the next_orphan
+ * queue as there is only ever a single
+ * process doing orphan recovery. */
+ OCFS2_I(iter)->ip_next_orphan = *head;
+ *head = iter;
}
brelse(bh);
}
- mutex_unlock(&orphan_dir_inode->i_mutex);
+out_unlock:
ocfs2_meta_unlock(orphan_dir_inode, 0);
- have_disk_lock = 0;
-
+out:
+ mutex_unlock(&orphan_dir_inode->i_mutex);
iput(orphan_dir_inode);
- orphan_dir_inode = NULL;
+ return status;
+}
+
+static int ocfs2_orphan_recovery_can_continue(struct ocfs2_super *osb,
+ int slot)
+{
+ int ret;
+
+ spin_lock(&osb->osb_lock);
+ ret = !osb->osb_orphan_wipes[slot];
+ spin_unlock(&osb->osb_lock);
+ return ret;
+}
+
+static void ocfs2_mark_recovering_orphan_dir(struct ocfs2_super *osb,
+ int slot)
+{
+ spin_lock(&osb->osb_lock);
+ /* Mark ourselves such that new processes in delete_inode()
+ * know to quit early. */
+ ocfs2_node_map_set_bit(osb, &osb->osb_recovering_orphan_dirs, slot);
+ while (osb->osb_orphan_wipes[slot]) {
+ /* If any processes are already in the middle of an
+ * orphan wipe on this dir, then we need to wait for
+ * them. */
+ spin_unlock(&osb->osb_lock);
+ wait_event_interruptible(osb->osb_wipe_event,
+ ocfs2_orphan_recovery_can_continue(osb, slot));
+ spin_lock(&osb->osb_lock);
+ }
+ spin_unlock(&osb->osb_lock);
+}
+
+static void ocfs2_clear_recovering_orphan_dir(struct ocfs2_super *osb,
+ int slot)
+{
+ ocfs2_node_map_clear_bit(osb, &osb->osb_recovering_orphan_dirs, slot);
+}
+
+/*
+ * Orphan recovery. Each mounted node has it's own orphan dir which we
+ * must run during recovery. Our strategy here is to build a list of
+ * the inodes in the orphan dir and iget/iput them. The VFS does
+ * (most) of the rest of the work.
+ *
+ * Orphan recovery can happen at any time, not just mount so we have a
+ * couple of extra considerations.
+ *
+ * - We grab as many inodes as we can under the orphan dir lock -
+ * doing iget() outside the orphan dir risks getting a reference on
+ * an invalid inode.
+ * - We must be sure not to deadlock with other processes on the
+ * system wanting to run delete_inode(). This can happen when they go
+ * to lock the orphan dir and the orphan recovery process attempts to
+ * iget() inside the orphan dir lock. This can be avoided by
+ * advertising our state to ocfs2_delete_inode().
+ */
+static int ocfs2_recover_orphans(struct ocfs2_super *osb,
+ int slot)
+{
+ int ret = 0;
+ struct inode *inode = NULL;
+ struct inode *iter;
+ struct ocfs2_inode_info *oi;
+
+ mlog(0, "Recover inodes from orphan dir in slot %d\n", slot);
+
+ ocfs2_mark_recovering_orphan_dir(osb, slot);
+ ret = ocfs2_queue_orphans(osb, slot, &inode);
+ ocfs2_clear_recovering_orphan_dir(osb, slot);
+
+ /* Error here should be noted, but we want to continue with as
+ * many queued inodes as we've got. */
+ if (ret)
+ mlog_errno(ret);
while (inode) {
oi = OCFS2_I(inode);
@@ -1541,14 +1610,7 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb,
inode = iter;
}
-out:
- if (have_disk_lock)
- ocfs2_meta_unlock(orphan_dir_inode, 0);
-
- if (orphan_dir_inode)
- iput(orphan_dir_inode);
-
- return status;
+ return ret;
}
static int ocfs2_wait_on_mount(struct ocfs2_super *osb)
@@ -1584,10 +1646,9 @@ static int ocfs2_commit_thread(void *arg)
while (!(kthread_should_stop() &&
atomic_read(&journal->j_num_trans) == 0)) {
- wait_event_interruptible_timeout(osb->checkpoint_event,
- atomic_read(&journal->j_num_trans)
- || kthread_should_stop(),
- OCFS2_CHECKPOINT_INTERVAL);
+ wait_event_interruptible(osb->checkpoint_event,
+ atomic_read(&journal->j_num_trans)
+ || kthread_should_stop());
status = ocfs2_commit_cache(osb);
if (status < 0)
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 7d0a816184f..2f3a6acdac4 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -29,8 +29,6 @@
#include <linux/fs.h>
#include <linux/jbd.h>
-#define OCFS2_CHECKPOINT_INTERVAL (8 * HZ)
-
enum ocfs2_journal_state {
OCFS2_JOURNAL_FREE = 0,
OCFS2_JOURNAL_LOADED,
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index f468c600cf9..e89de9b6e49 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -33,6 +33,7 @@
#include <linux/rbtree.h>
#include <linux/workqueue.h>
#include <linux/kref.h>
+#include <linux/mutex.h>
#include "cluster/nodemanager.h"
#include "cluster/heartbeat.h"
@@ -173,9 +174,6 @@ enum ocfs2_mount_options
OCFS2_MOUNT_NOINTR = 1 << 2, /* Don't catch signals */
OCFS2_MOUNT_ERRORS_PANIC = 1 << 3, /* Panic on errors */
OCFS2_MOUNT_DATA_WRITEBACK = 1 << 4, /* No data ordering */
-#ifdef OCFS2_ORACORE_WORKAROUNDS
- OCFS2_MOUNT_COMPAT_OCFS = 1 << 30, /* ocfs1 compatibility mode */
-#endif
};
#define OCFS2_OSB_SOFT_RO 0x0001
@@ -233,7 +231,7 @@ struct ocfs2_super
struct proc_dir_entry *proc_sub_dir; /* points to /proc/fs/ocfs2/<maj_min> */
atomic_t vol_state;
- struct semaphore recovery_lock;
+ struct mutex recovery_lock;
struct task_struct *recovery_thread_task;
int disable_recovery;
wait_queue_head_t checkpoint_event;
@@ -289,6 +287,10 @@ struct ocfs2_super
struct inode *osb_tl_inode;
struct buffer_head *osb_tl_bh;
struct work_struct osb_truncate_log_wq;
+
+ struct ocfs2_node_map osb_recovering_orphan_dirs;
+ unsigned int *osb_orphan_wipes;
+ wait_queue_head_t osb_wipe_event;
};
#define OCFS2_SB(sb) ((struct ocfs2_super *)(sb)->s_fs_info)
diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h
index dfb8a5bedfc..c5b1ac547c1 100644
--- a/fs/ocfs2/ocfs2_fs.h
+++ b/fs/ocfs2/ocfs2_fs.h
@@ -138,7 +138,6 @@
/* Journal limits (in bytes) */
#define OCFS2_MIN_JOURNAL_SIZE (4 * 1024 * 1024)
-#define OCFS2_MAX_JOURNAL_SIZE (500 * 1024 * 1024)
struct ocfs2_system_inode_info {
char *si_name;
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 364d64bd5f1..8dd3aafec49 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -932,7 +932,7 @@ static void ocfs2_inode_init_once(void *data,
oi->ip_dir_start_lookup = 0;
init_rwsem(&oi->ip_alloc_sem);
- init_MUTEX(&(oi->ip_io_sem));
+ mutex_init(&oi->ip_io_mutex);
oi->ip_blkno = 0ULL;
oi->ip_clusters = 0;
@@ -1137,9 +1137,9 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
/* disable any new recovery threads and wait for any currently
* running ones to exit. Do this before setting the vol_state. */
- down(&osb->recovery_lock);
+ mutex_lock(&osb->recovery_lock);
osb->disable_recovery = 1;
- up(&osb->recovery_lock);
+ mutex_unlock(&osb->recovery_lock);
wait_event(osb->recovery_event, !ocfs2_recovery_thread_running(osb));
/* At this point, we know that no more recovery threads can be
@@ -1254,8 +1254,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
osb->sb = sb;
/* Save off for ocfs2_rw_direct */
osb->s_sectsize_bits = blksize_bits(sector_size);
- if (!osb->s_sectsize_bits)
- BUG();
+ BUG_ON(!osb->s_sectsize_bits);
osb->net_response_ids = 0;
spin_lock_init(&osb->net_response_lock);
@@ -1283,7 +1282,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
snprintf(osb->dev_str, sizeof(osb->dev_str), "%u,%u",
MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
- init_MUTEX(&osb->recovery_lock);
+ mutex_init(&osb->recovery_lock);
osb->disable_recovery = 0;
osb->recovery_thread_task = NULL;
@@ -1326,6 +1325,16 @@ static int ocfs2_initialize_super(struct super_block *sb,
}
mlog(ML_NOTICE, "max_slots for this device: %u\n", osb->max_slots);
+ init_waitqueue_head(&osb->osb_wipe_event);
+ osb->osb_orphan_wipes = kcalloc(osb->max_slots,
+ sizeof(*osb->osb_orphan_wipes),
+ GFP_KERNEL);
+ if (!osb->osb_orphan_wipes) {
+ status = -ENOMEM;
+ mlog_errno(status);
+ goto bail;
+ }
+
osb->s_feature_compat =
le32_to_cpu(OCFS2_RAW_SB(di)->s_feature_compat);
osb->s_feature_ro_compat =
@@ -1639,6 +1648,7 @@ static void ocfs2_delete_osb(struct ocfs2_super *osb)
if (osb->slot_info)
ocfs2_free_slot_info(osb->slot_info);
+ kfree(osb->osb_orphan_wipes);
/* FIXME
* This belongs in journal shutdown, but because we have to
* allocate osb->journal at the start of ocfs2_initalize_osb(),
diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c
index 600a8bc5b54..fc29cb7a437 100644
--- a/fs/ocfs2/sysfile.c
+++ b/fs/ocfs2/sysfile.c
@@ -77,8 +77,7 @@ struct inode *ocfs2_get_system_file_inode(struct ocfs2_super *osb,
if (arr && ((inode = *arr) != NULL)) {
/* get a ref in addition to the array ref */
inode = igrab(inode);
- if (!inode)
- BUG();
+ BUG_ON(!inode);
return inode;
}
@@ -89,8 +88,7 @@ struct inode *ocfs2_get_system_file_inode(struct ocfs2_super *osb,
/* add one more if putting into array for first time */
if (arr && inode) {
*arr = igrab(inode);
- if (!*arr)
- BUG();
+ BUG_ON(!*arr);
}
return inode;
}
diff --git a/fs/ocfs2/uptodate.c b/fs/ocfs2/uptodate.c
index 3a0458fd3e1..300b5bedfb2 100644
--- a/fs/ocfs2/uptodate.c
+++ b/fs/ocfs2/uptodate.c
@@ -388,7 +388,7 @@ out_free:
}
}
-/* Item insertion is guarded by ip_io_sem, so the insertion path takes
+/* Item insertion is guarded by ip_io_mutex, so the insertion path takes
* advantage of this by not rechecking for a duplicate insert during
* the slow case. Additionally, if the cache needs to be bumped up to
* a tree, the code will not recheck after acquiring the lock --
@@ -418,7 +418,7 @@ void ocfs2_set_buffer_uptodate(struct inode *inode,
(unsigned long long) bh->b_blocknr);
/* No need to recheck under spinlock - insertion is guarded by
- * ip_io_sem */
+ * ip_io_mutex */
spin_lock(&oi->ip_lock);
if (ocfs2_insert_can_use_array(oi, ci)) {
/* Fast case - it's an array and there's a free
@@ -440,7 +440,7 @@ void ocfs2_set_buffer_uptodate(struct inode *inode,
/* Called against a newly allocated buffer. Most likely nobody should
* be able to read this sort of metadata while it's still being
- * allocated, but this is careful to take ip_io_sem anyway. */
+ * allocated, but this is careful to take ip_io_mutex anyway. */
void ocfs2_set_new_buffer_uptodate(struct inode *inode,
struct buffer_head *bh)
{
@@ -451,9 +451,9 @@ void ocfs2_set_new_buffer_uptodate(struct inode *inode,
set_buffer_uptodate(bh);
- down(&oi->ip_io_sem);
+ mutex_lock(&oi->ip_io_mutex);
ocfs2_set_buffer_uptodate(inode, bh);
- up(&oi->ip_io_sem);
+ mutex_unlock(&oi->ip_io_mutex);
}
/* Requires ip_lock. */
@@ -537,7 +537,7 @@ int __init init_ocfs2_uptodate_cache(void)
return 0;
}
-void __exit exit_ocfs2_uptodate_cache(void)
+void exit_ocfs2_uptodate_cache(void)
{
if (ocfs2_uptodate_cachep)
kmem_cache_destroy(ocfs2_uptodate_cachep);
diff --git a/fs/ocfs2/uptodate.h b/fs/ocfs2/uptodate.h
index e5aacdf4eab..01cd32d26b0 100644
--- a/fs/ocfs2/uptodate.h
+++ b/fs/ocfs2/uptodate.h
@@ -27,7 +27,7 @@
#define OCFS2_UPTODATE_H
int __init init_ocfs2_uptodate_cache(void);
-void __exit exit_ocfs2_uptodate_cache(void);
+void exit_ocfs2_uptodate_cache(void);
void ocfs2_metadata_cache_init(struct inode *inode);
void ocfs2_metadata_cache_purge(struct inode *inode);
generated by cgit v1.2.3 (git 2.25.1) at 2024-07-25 09:33:05 +0000