summaryrefslogtreecommitdiffstats
path: root/fs
diff options
context:
space:
mode:
Diffstat (limited to 'fs')
-rw-r--r--fs/btrfs/ctree.h3
-rw-r--r--fs/btrfs/extent-tree.c126
-rw-r--r--fs/btrfs/extent_io.c51
-rw-r--r--fs/btrfs/free-space-cache.c2
-rw-r--r--fs/btrfs/inode.c2
-rw-r--r--fs/btrfs/ioctl.c2
-rw-r--r--fs/btrfs/scrub.c5
-rw-r--r--fs/btrfs/super.c6
-rw-r--r--fs/btrfs/volumes.c2
-rw-r--r--fs/cifs/connect.c2
-rw-r--r--fs/cifs/file.c26
-rw-r--r--fs/cifs/readdir.c10
-rw-r--r--fs/cifs/smbencrypt.c6
-rw-r--r--fs/dcache.c71
-rw-r--r--fs/ext4/inode.c2
-rw-r--r--fs/namespace.c20
-rw-r--r--fs/ocfs2/alloc.c2
-rw-r--r--fs/ocfs2/aops.c69
-rw-r--r--fs/ocfs2/aops.h14
-rw-r--r--fs/ocfs2/cluster/heartbeat.c194
-rw-r--r--fs/ocfs2/cluster/netdebug.c102
-rw-r--r--fs/ocfs2/cluster/tcp.c138
-rw-r--r--fs/ocfs2/cluster/tcp.h2
-rw-r--r--fs/ocfs2/dir.c3
-rw-r--r--fs/ocfs2/dlm/dlmcommon.h56
-rw-r--r--fs/ocfs2/dlm/dlmdomain.c44
-rw-r--r--fs/ocfs2/dlm/dlmlock.c54
-rw-r--r--fs/ocfs2/dlm/dlmmaster.c175
-rw-r--r--fs/ocfs2/dlm/dlmrecovery.c164
-rw-r--r--fs/ocfs2/dlm/dlmthread.c16
-rw-r--r--fs/ocfs2/dlmglue.c21
-rw-r--r--fs/ocfs2/extent_map.c96
-rw-r--r--fs/ocfs2/extent_map.h2
-rw-r--r--fs/ocfs2/file.c96
-rw-r--r--fs/ocfs2/inode.c2
-rw-r--r--fs/ocfs2/inode.h3
-rw-r--r--fs/ocfs2/ioctl.c11
-rw-r--r--fs/ocfs2/journal.c23
-rw-r--r--fs/ocfs2/journal.h5
-rw-r--r--fs/ocfs2/mmap.c53
-rw-r--r--fs/ocfs2/move_extents.c2
-rw-r--r--fs/ocfs2/ocfs2.h51
-rw-r--r--fs/ocfs2/quota_local.c23
-rw-r--r--fs/ocfs2/slot_map.c4
-rw-r--r--fs/ocfs2/stack_o2cb.c71
-rw-r--r--fs/ocfs2/super.c25
-rw-r--r--fs/ocfs2/xattr.c10
-rw-r--r--fs/proc/meminfo.c7
-rw-r--r--fs/proc/stat.c4
-rw-r--r--fs/pstore/platform.c13
-rw-r--r--fs/seq_file.c6
-rw-r--r--fs/xfs/xfs_acl.c2
-rw-r--r--fs/xfs/xfs_attr_leaf.c64
-rw-r--r--fs/xfs/xfs_bmap.c20
-rw-r--r--fs/xfs/xfs_export.c8
-rw-r--r--fs/xfs/xfs_inode.c21
-rw-r--r--fs/xfs/xfs_inode.h1
-rw-r--r--fs/xfs/xfs_log.c348
-rw-r--r--fs/xfs/xfs_sync.c11
-rw-r--r--fs/xfs/xfs_trace.h12
60 files changed, 1501 insertions, 883 deletions
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 04a5dfcee5a1..50634abef9b4 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2369,6 +2369,9 @@ int btrfs_block_rsv_check(struct btrfs_root *root,
int btrfs_block_rsv_refill(struct btrfs_root *root,
struct btrfs_block_rsv *block_rsv,
u64 min_reserved);
+int btrfs_block_rsv_refill_noflush(struct btrfs_root *root,
+ struct btrfs_block_rsv *block_rsv,
+ u64 min_reserved);
int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv,
struct btrfs_block_rsv *dst_rsv,
u64 num_bytes);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 930ae8949737..2ad813674d77 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -3888,9 +3888,9 @@ int btrfs_block_rsv_check(struct btrfs_root *root,
return ret;
}
-int btrfs_block_rsv_refill(struct btrfs_root *root,
- struct btrfs_block_rsv *block_rsv,
- u64 min_reserved)
+static inline int __btrfs_block_rsv_refill(struct btrfs_root *root,
+ struct btrfs_block_rsv *block_rsv,
+ u64 min_reserved, int flush)
{
u64 num_bytes = 0;
int ret = -ENOSPC;
@@ -3909,7 +3909,7 @@ int btrfs_block_rsv_refill(struct btrfs_root *root,
if (!ret)
return 0;
- ret = reserve_metadata_bytes(root, block_rsv, num_bytes, 1);
+ ret = reserve_metadata_bytes(root, block_rsv, num_bytes, flush);
if (!ret) {
block_rsv_add_bytes(block_rsv, num_bytes, 0);
return 0;
@@ -3918,6 +3918,20 @@ int btrfs_block_rsv_refill(struct btrfs_root *root,
return ret;
}
+int btrfs_block_rsv_refill(struct btrfs_root *root,
+ struct btrfs_block_rsv *block_rsv,
+ u64 min_reserved)
+{
+ return __btrfs_block_rsv_refill(root, block_rsv, min_reserved, 1);
+}
+
+int btrfs_block_rsv_refill_noflush(struct btrfs_root *root,
+ struct btrfs_block_rsv *block_rsv,
+ u64 min_reserved)
+{
+ return __btrfs_block_rsv_refill(root, block_rsv, min_reserved, 0);
+}
+
int btrfs_block_rsv_migrate(struct btrfs_block_rsv *src_rsv,
struct btrfs_block_rsv *dst_rsv,
u64 num_bytes)
@@ -5093,11 +5107,11 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
struct btrfs_root *root = orig_root->fs_info->extent_root;
struct btrfs_free_cluster *last_ptr = NULL;
struct btrfs_block_group_cache *block_group = NULL;
+ struct btrfs_block_group_cache *used_block_group;
int empty_cluster = 2 * 1024 * 1024;
int allowed_chunk_alloc = 0;
int done_chunk_alloc = 0;
struct btrfs_space_info *space_info;
- int last_ptr_loop = 0;
int loop = 0;
int index = 0;
int alloc_type = (data & BTRFS_BLOCK_GROUP_DATA) ?
@@ -5159,6 +5173,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
ideal_cache:
block_group = btrfs_lookup_block_group(root->fs_info,
search_start);
+ used_block_group = block_group;
/*
* we don't want to use the block group if it doesn't match our
* allocation bits, or if its not cached.
@@ -5196,6 +5211,7 @@ search:
u64 offset;
int cached;
+ used_block_group = block_group;
btrfs_get_block_group(block_group);
search_start = block_group->key.objectid;
@@ -5265,84 +5281,73 @@ alloc:
spin_lock(&block_group->free_space_ctl->tree_lock);
if (cached &&
block_group->free_space_ctl->free_space <
- num_bytes + empty_size) {
+ num_bytes + empty_cluster + empty_size) {
spin_unlock(&block_group->free_space_ctl->tree_lock);
goto loop;
}
spin_unlock(&block_group->free_space_ctl->tree_lock);
/*
- * Ok we want to try and use the cluster allocator, so lets look
- * there, unless we are on LOOP_NO_EMPTY_SIZE, since we will
- * have tried the cluster allocator plenty of times at this
- * point and not have found anything, so we are likely way too
- * fragmented for the clustering stuff to find anything, so lets
- * just skip it and let the allocator find whatever block it can
- * find
+ * Ok we want to try and use the cluster allocator, so
+ * lets look there
*/
- if (last_ptr && loop < LOOP_NO_EMPTY_SIZE) {
+ if (last_ptr) {
/*
* the refill lock keeps out other
* people trying to start a new cluster
*/
spin_lock(&last_ptr->refill_lock);
- if (last_ptr->block_group &&
- (last_ptr->block_group->ro ||
- !block_group_bits(last_ptr->block_group, data))) {
- offset = 0;
+ used_block_group = last_ptr->block_group;
+ if (used_block_group != block_group &&
+ (!used_block_group ||
+ used_block_group->ro ||
+ !block_group_bits(used_block_group, data))) {
+ used_block_group = block_group;
goto refill_cluster;
}
- offset = btrfs_alloc_from_cluster(block_group, last_ptr,
- num_bytes, search_start);
+ if (used_block_group != block_group)
+ btrfs_get_block_group(used_block_group);
+
+ offset = btrfs_alloc_from_cluster(used_block_group,
+ last_ptr, num_bytes, used_block_group->key.objectid);
if (offset) {
/* we have a block, we're done */
spin_unlock(&last_ptr->refill_lock);
goto checks;
}
- spin_lock(&last_ptr->lock);
- /*
- * whoops, this cluster doesn't actually point to
- * this block group. Get a ref on the block
- * group is does point to and try again
- */
- if (!last_ptr_loop && last_ptr->block_group &&
- last_ptr->block_group != block_group &&
- index <=
- get_block_group_index(last_ptr->block_group)) {
-
- btrfs_put_block_group(block_group);
- block_group = last_ptr->block_group;
- btrfs_get_block_group(block_group);
- spin_unlock(&last_ptr->lock);
- spin_unlock(&last_ptr->refill_lock);
-
- last_ptr_loop = 1;
- search_start = block_group->key.objectid;
- /*
- * we know this block group is properly
- * in the list because
- * btrfs_remove_block_group, drops the
- * cluster before it removes the block
- * group from the list
- */
- goto have_block_group;
+ WARN_ON(last_ptr->block_group != used_block_group);
+ if (used_block_group != block_group) {
+ btrfs_put_block_group(used_block_group);
+ used_block_group = block_group;
}
- spin_unlock(&last_ptr->lock);
refill_cluster:
+ BUG_ON(used_block_group != block_group);
+ /* If we are on LOOP_NO_EMPTY_SIZE, we can't
+ * set up a new clusters, so lets just skip it
+ * and let the allocator find whatever block
+ * it can find. If we reach this point, we
+ * will have tried the cluster allocator
+ * plenty of times and not have found
+ * anything, so we are likely way too
+ * fragmented for the clustering stuff to find
+ * anything. */
+ if (loop >= LOOP_NO_EMPTY_SIZE) {
+ spin_unlock(&last_ptr->refill_lock);
+ goto unclustered_alloc;
+ }
+
/*
* this cluster didn't work out, free it and
* start over
*/
btrfs_return_cluster_to_free_space(NULL, last_ptr);
- last_ptr_loop = 0;
-
/* allocate a cluster in this block group */
ret = btrfs_find_space_cluster(trans, root,
block_group, last_ptr,
- offset, num_bytes,
+ search_start, num_bytes,
empty_cluster + empty_size);
if (ret == 0) {
/*
@@ -5378,6 +5383,7 @@ refill_cluster:
goto loop;
}
+unclustered_alloc:
offset = btrfs_find_space_for_alloc(block_group, search_start,
num_bytes, empty_size);
/*
@@ -5404,14 +5410,14 @@ checks:
search_start = stripe_align(root, offset);
/* move on to the next group */
if (search_start + num_bytes >= search_end) {
- btrfs_add_free_space(block_group, offset, num_bytes);
+ btrfs_add_free_space(used_block_group, offset, num_bytes);
goto loop;
}
/* move on to the next group */
if (search_start + num_bytes >
- block_group->key.objectid + block_group->key.offset) {
- btrfs_add_free_space(block_group, offset, num_bytes);
+ used_block_group->key.objectid + used_block_group->key.offset) {
+ btrfs_add_free_space(used_block_group, offset, num_bytes);
goto loop;
}
@@ -5419,14 +5425,14 @@ checks:
ins->offset = num_bytes;
if (offset < search_start)
- btrfs_add_free_space(block_group, offset,
+ btrfs_add_free_space(used_block_group, offset,
search_start - offset);
BUG_ON(offset > search_start);
- ret = btrfs_update_reserved_bytes(block_group, num_bytes,
+ ret = btrfs_update_reserved_bytes(used_block_group, num_bytes,
alloc_type);
if (ret == -EAGAIN) {
- btrfs_add_free_space(block_group, offset, num_bytes);
+ btrfs_add_free_space(used_block_group, offset, num_bytes);
goto loop;
}
@@ -5435,15 +5441,19 @@ checks:
ins->offset = num_bytes;
if (offset < search_start)
- btrfs_add_free_space(block_group, offset,
+ btrfs_add_free_space(used_block_group, offset,
search_start - offset);
BUG_ON(offset > search_start);
+ if (used_block_group != block_group)
+ btrfs_put_block_group(used_block_group);
btrfs_put_block_group(block_group);
break;
loop:
failed_cluster_refill = false;
failed_alloc = false;
BUG_ON(index != get_block_group_index(block_group));
+ if (used_block_group != block_group)
+ btrfs_put_block_group(used_block_group);
btrfs_put_block_group(block_group);
}
up_read(&space_info->groups_sem);
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 9472d3de5e52..49f3c9dc09f4 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -935,8 +935,10 @@ again:
node = tree_search(tree, start);
if (!node) {
prealloc = alloc_extent_state_atomic(prealloc);
- if (!prealloc)
- return -ENOMEM;
+ if (!prealloc) {
+ err = -ENOMEM;
+ goto out;
+ }
err = insert_state(tree, prealloc, start, end, &bits);
prealloc = NULL;
BUG_ON(err == -EEXIST);
@@ -992,8 +994,10 @@ hit_next:
*/
if (state->start < start) {
prealloc = alloc_extent_state_atomic(prealloc);
- if (!prealloc)
- return -ENOMEM;
+ if (!prealloc) {
+ err = -ENOMEM;
+ goto out;
+ }
err = split_state(tree, state, prealloc, start);
BUG_ON(err == -EEXIST);
prealloc = NULL;
@@ -1024,8 +1028,10 @@ hit_next:
this_end = last_start - 1;
prealloc = alloc_extent_state_atomic(prealloc);
- if (!prealloc)
- return -ENOMEM;
+ if (!prealloc) {
+ err = -ENOMEM;
+ goto out;
+ }
/*
* Avoid to free 'prealloc' if it can be merged with
@@ -1051,8 +1057,10 @@ hit_next:
*/
if (state->start <= end && state->end > end) {
prealloc = alloc_extent_state_atomic(prealloc);
- if (!prealloc)
- return -ENOMEM;
+ if (!prealloc) {
+ err = -ENOMEM;
+ goto out;
+ }
err = split_state(tree, state, prealloc, end + 1);
BUG_ON(err == -EEXIST);
@@ -2287,14 +2295,20 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
if (!uptodate) {
int failed_mirror;
failed_mirror = (int)(unsigned long)bio->bi_bdev;
- if (tree->ops && tree->ops->readpage_io_failed_hook)
- ret = tree->ops->readpage_io_failed_hook(
- bio, page, start, end,
- failed_mirror, state);
- else
- ret = bio_readpage_error(bio, page, start, end,
- failed_mirror, NULL);
+ /*
+ * The generic bio_readpage_error handles errors the
+ * following way: If possible, new read requests are
+ * created and submitted and will end up in
+ * end_bio_extent_readpage as well (if we're lucky, not
+ * in the !uptodate case). In that case it returns 0 and
+ * we just go on with the next page in our bio. If it
+ * can't handle the error it will return -EIO and we
+ * remain responsible for that page.
+ */
+ ret = bio_readpage_error(bio, page, start, end,
+ failed_mirror, NULL);
if (ret == 0) {
+error_handled:
uptodate =
test_bit(BIO_UPTODATE, &bio->bi_flags);
if (err)
@@ -2302,6 +2316,13 @@ static void end_bio_extent_readpage(struct bio *bio, int err)
uncache_state(&cached);
continue;
}
+ if (tree->ops && tree->ops->readpage_io_failed_hook) {
+ ret = tree->ops->readpage_io_failed_hook(
+ bio, page, start, end,
+ failed_mirror, state);
+ if (ret == 0)
+ goto error_handled;
+ }
}
if (uptodate) {
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 6e5b7e463698..ec23d43d0c35 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -1470,6 +1470,7 @@ static void add_new_bitmap(struct btrfs_free_space_ctl *ctl,
{
info->offset = offset_to_bitmap(ctl, offset);
info->bytes = 0;
+ INIT_LIST_HEAD(&info->list);
link_free_space(ctl, info);
ctl->total_bitmaps++;
@@ -2319,6 +2320,7 @@ again:
if (!found) {
start = i;
+ cluster->max_size = 0;
found = true;
}
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 526dd51a1966..2c984f7d4c2a 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3490,7 +3490,7 @@ void btrfs_evict_inode(struct inode *inode)
* doing the truncate.
*/
while (1) {
- ret = btrfs_block_rsv_refill(root, rsv, min_size);
+ ret = btrfs_block_rsv_refill_noflush(root, rsv, min_size);
/*
* Try and steal from the global reserve since we will
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index a90e749ed6d2..72d461656f60 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1278,7 +1278,7 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root,
}
ret = btrfs_grow_device(trans, device, new_size);
btrfs_commit_transaction(trans, root);
- } else {
+ } else if (new_size < old_size) {
ret = btrfs_shrink_device(device, new_size);
}
diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index fab420db5121..c27bcb67f330 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -256,6 +256,11 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, void *ctx)
btrfs_release_path(swarn->path);
ipath = init_ipath(4096, local_root, swarn->path);
+ if (IS_ERR(ipath)) {
+ ret = PTR_ERR(ipath);
+ ipath = NULL;
+ goto err;
+ }
ret = paths_from_inode(inum, ipath);
if (ret < 0)
diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c
index 17ee7fc5e64e..e28ad4baf483 100644
--- a/fs/btrfs/super.c
+++ b/fs/btrfs/super.c
@@ -1057,7 +1057,7 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
int i = 0, nr_devices;
int ret;
- nr_devices = fs_info->fs_devices->rw_devices;
+ nr_devices = fs_info->fs_devices->open_devices;
BUG_ON(!nr_devices);
devices_info = kmalloc(sizeof(*devices_info) * nr_devices,
@@ -1079,8 +1079,8 @@ static int btrfs_calc_avail_data_space(struct btrfs_root *root, u64 *free_bytes)
else
min_stripe_size = BTRFS_STRIPE_LEN;
- list_for_each_entry(device, &fs_devices->alloc_list, dev_alloc_list) {
- if (!device->in_fs_metadata)
+ list_for_each_entry(device, &fs_devices->devices, dev_list) {
+ if (!device->in_fs_metadata || !device->bdev)
continue;
avail_space = device->total_bytes - device->bytes_used;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index c37433d3cd82..0a8c8f8304b1 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1611,7 +1611,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
if ((sb->s_flags & MS_RDONLY) && !root->fs_info->fs_devices->seeding)
return -EINVAL;
- bdev = blkdev_get_by_path(device_path, FMODE_EXCL,
+ bdev = blkdev_get_by_path(device_path, FMODE_WRITE | FMODE_EXCL,
root->fs_info->bdev_holder);
if (IS_ERR(bdev))
return PTR_ERR(bdev);
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c
index d6a972df0338..8cd4b52d4217 100644
--- a/fs/cifs/connect.c
+++ b/fs/cifs/connect.c
@@ -441,6 +441,8 @@ cifs_readv_from_socket(struct TCP_Server_Info *server, struct kvec *iov_orig,
smb_msg.msg_controllen = 0;
for (total_read = 0; to_read; total_read += length, to_read -= length) {
+ try_to_freeze();
+
if (server_unresponsive(server)) {
total_read = -EAGAIN;
break;
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index cf0b1539b321..4dd9283885e7 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -702,6 +702,13 @@ cifs_find_lock_conflict(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock,
lock->type, lock->netfid, conf_lock);
}
+/*
+ * Check if there is another lock that prevents us to set the lock (mandatory
+ * style). If such a lock exists, update the flock structure with its
+ * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks
+ * or leave it the same if we can't. Returns 0 if we don't need to request to
+ * the server or 1 otherwise.
+ */
static int
cifs_lock_test(struct cifsInodeInfo *cinode, __u64 offset, __u64 length,
__u8 type, __u16 netfid, struct file_lock *flock)
@@ -739,6 +746,12 @@ cifs_lock_add(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock)
mutex_unlock(&cinode->lock_mutex);
}
+/*
+ * Set the byte-range lock (mandatory style). Returns:
+ * 1) 0, if we set the lock and don't need to request to the server;
+ * 2) 1, if no locks prevent us but we need to request to the server;
+ * 3) -EACCESS, if there is a lock that prevents us and wait is false.
+ */
static int
cifs_lock_add_if(struct cifsInodeInfo *cinode, struct cifsLockInfo *lock,
bool wait)
@@ -778,6 +791,13 @@ try_again:
return rc;
}
+/*
+ * Check if there is another lock that prevents us to set the lock (posix
+ * style). If such a lock exists, update the flock structure with its
+ * properties. Otherwise, set the flock type to F_UNLCK if we can cache brlocks
+ * or leave it the same if we can't. Returns 0 if we don't need to request to
+ * the server or 1 otherwise.
+ */
static int
cifs_posix_lock_test(struct file *file, struct file_lock *flock)
{
@@ -800,6 +820,12 @@ cifs_posix_lock_test(struct file *file, struct file_lock *flock)
return rc;
}
+/*
+ * Set the byte-range lock (posix style). Returns:
+ * 1) 0, if we set the lock and don't need to request to the server;
+ * 2) 1, if we need to request to the server;
+ * 3) <0, if the error occurs while setting the lock.
+ */
static int
cifs_posix_lock_set(struct file *file, struct file_lock *flock)
{
diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c
index 5de03ec20144..a090bbe6ee29 100644
--- a/fs/cifs/readdir.c
+++ b/fs/cifs/readdir.c
@@ -554,7 +554,10 @@ static int find_cifs_entry(const int xid, struct cifs_tcon *pTcon,
rc);
return rc;
}
- cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile);
+ /* FindFirst/Next set last_entry to NULL on malformed reply */
+ if (cifsFile->srch_inf.last_entry)
+ cifs_save_resume_key(cifsFile->srch_inf.last_entry,
+ cifsFile);
}
while ((index_to_find >= cifsFile->srch_inf.index_of_last_entry) &&
@@ -562,7 +565,10 @@ static int find_cifs_entry(const int xid, struct cifs_tcon *pTcon,
cFYI(1, "calling findnext2");
rc = CIFSFindNext(xid, pTcon, cifsFile->netfid,
&cifsFile->srch_inf);
- cifs_save_resume_key(cifsFile->srch_inf.last_entry, cifsFile);
+ /* FindFirst/Next set last_entry to NULL on malformed reply */
+ if (cifsFile->srch_inf.last_entry)
+ cifs_save_resume_key(cifsFile->srch_inf.last_entry,
+ cifsFile);
if (rc)
return -ENOENT;
}
diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c
index 7cacba12b8f1..80d850881938 100644
--- a/fs/cifs/smbencrypt.c
+++ b/fs/cifs/smbencrypt.c
@@ -209,7 +209,7 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16,
{
int rc;
int len;
- __u16 wpwd[129];
+ __le16 wpwd[129];
/* Password cannot be longer than 128 characters */
if (passwd) /* Password must be converted to NT unicode */
@@ -219,8 +219,8 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16,
*wpwd = 0; /* Ensure string is null terminated */
}
- rc = mdfour(p16, (unsigned char *) wpwd, len * sizeof(__u16));
- memset(wpwd, 0, 129 * sizeof(__u16));
+ rc = mdfour(p16, (unsigned char *) wpwd, len * sizeof(__le16));
+ memset(wpwd, 0, 129 * sizeof(__le16));
return rc;
}
diff --git a/fs/dcache.c b/fs/dcache.c
index 10ba92def3f6..89509b5a090e 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -2439,16 +2439,14 @@ static int prepend_name(char **buffer, int *buflen, struct qstr *name)
/**
* prepend_path - Prepend path string to a buffer
* @path: the dentry/vfsmount to report
- * @root: root vfsmnt/dentry (may be modified by this function)
+ * @root: root vfsmnt/dentry
* @buffer: pointer to the end of the buffer
* @buflen: pointer to buffer length
*
* Caller holds the rename_lock.
- *
- * If path is not reachable from the supplied root, then the value of
- * root is changed (without modifying refcounts).
*/
-static int prepend_path(const struct path *path, struct path *root,
+static int prepend_path(const struct path *path,
+ const struct path *root,
char **buffer, int *buflen)
{
struct dentry *dentry = path->dentry;
@@ -2483,10 +2481,10 @@ static int prepend_path(const struct path *path, struct path *root,
dentry = parent;
}
-out:
if (!error && !slash)
error = prepend(buffer, buflen, "/", 1);
+out:
br_read_unlock(vfsmount_lock);
return error;
@@ -2500,15 +2498,17 @@ global_root:
WARN(1, "Root dentry has weird name <%.*s>\n",
(int) dentry->d_name.len, dentry->d_name.name);
}
- root->mnt = vfsmnt;
- root->dentry = dentry;
+ if (!slash)
+ error = prepend(buffer, buflen, "/", 1);
+ if (!error)
+ error = vfsmnt->mnt_ns ? 1 : 2;
goto out;
}
/**
* __d_path - return the path of a dentry
* @path: the dentry/vfsmount to report
- * @root: root vfsmnt/dentry (may be modified by this function)
+ * @root: root vfsmnt/dentry
* @buf: buffer to return value in
* @buflen: buffer length
*
@@ -2519,10 +2519,10 @@ global_root:
*
* "buflen" should be positive.
*
- * If path is not reachable from the supplied root, then the value of
- * root is changed (without modifying refcounts).
+ * If the path is not reachable from the supplied root, return %NULL.
*/
-char *__d_path(const struct path *path, struct path *root,
+char *__d_path(const struct path *path,
+ const struct path *root,
char *buf, int buflen)
{
char *res = buf + buflen;
@@ -2533,7 +2533,28 @@ char *__d_path(const struct path *path, struct path *root,
error = prepend_path(path, root, &res, &buflen);
write_sequnlock(&rename_lock);
- if (error)
+ if (error < 0)
+ return ERR_PTR(error);
+ if (error > 0)
+ return NULL;
+ return res;
+}
+
+char *d_absolute_path(const struct path *path,
+ char *buf, int buflen)
+{
+ struct path root = {};
+ char *res = buf + buflen;
+ int error;
+
+ prepend(&res, &buflen, "\0", 1);
+ write_seqlock(&rename_lock);
+ error = prepend_path(path, &root, &res, &buflen);
+ write_sequnlock(&rename_lock);
+
+ if (error > 1)
+ error = -EINVAL;
+ if (error < 0)
return ERR_PTR(error);
return res;
}
@@ -2541,8 +2562,9 @@ char *__d_path(const struct path *path, struct path *root,
/*
* same as __d_path but appends "(deleted)" for unlinked files.
*/
-static int path_with_deleted(const struct path *path, struct path *root,
- char **buf, int *buflen)
+static int path_with_deleted(const struct path *path,
+ const struct path *root,
+ char **buf, int *buflen)
{
prepend(buf, buflen, "\0", 1);
if (d_unlinked(path->dentry)) {
@@ -2579,7 +2601,6 @@ char *d_path(const struct path *path, char *buf, int buflen)
{
char *res = buf + buflen;
struct path root;
- struct path tmp;
int error;
/*
@@ -2594,9 +2615,8 @@ char *d_path(const struct path *path, char *buf, int buflen)
get_fs_root(current->fs, &root);
write_seqlock(&rename_lock);
- tmp = root;
- error = path_with_deleted(path, &tmp, &res, &buflen);
- if (error)
+ error = path_with_deleted(path, &root, &res, &buflen);
+ if (error < 0)
res = ERR_PTR(error);
write_sequnlock(&rename_lock);
path_put(&root);
@@ -2617,7 +2637,6 @@ char *d_path_with_unreachable(const struct path *path, char *buf, int buflen)
{
char *res = buf + buflen;
struct path root;
- struct path tmp;
int error;
if (path->dentry->d_op && path->dentry->d_op->d_dname)
@@ -2625,9 +2644,8 @@ char *d_path_with_unreachable(const struct path *path, char *buf, int buflen)
get_fs_root(current->fs, &root);
write_seqlock(&rename_lock);
- tmp = root;
- error = path_with_deleted(path, &tmp, &res, &buflen);
- if (!error && !path_equal(&tmp, &root))
+ error = path_with_deleted(path, &root, &res, &buflen);
+ if (error > 0)
error = prepend_unreachable(&res, &buflen);
write_sequnlock(&rename_lock);
path_put(&root);
@@ -2758,19 +2776,18 @@ SYSCALL_DEFINE2(getcwd, char __user *, buf, unsigned long, size)
write_seqlock(&rename_lock);
if (!d_unlinked(pwd.dentry)) {
unsigned long len;
- struct path tmp = root;
char *cwd = page + PAGE_SIZE;
int buflen = PAGE_SIZE;
prepend(&cwd, &buflen, "\0", 1);
- error = prepend_path(&pwd, &tmp, &cwd, &buflen);
+ error = prepend_path(&pwd, &root, &cwd, &buflen);
write_sequnlock(&rename_lock);
- if (error)
+ if (error < 0)
goto out;
/* Unreachable from current root */
- if (!path_equal(&tmp, &root)) {
+ if (error > 0) {
error = prepend_unreachable(&cwd, &buflen);
if (error)
goto out;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index fffec40d5996..848f436df29f 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2807,8 +2807,8 @@ out:
spin_unlock_irqrestore(&ei->i_completed_io_lock, flags);
/* queue the work to convert unwritten extents to written */
- queue_work(wq, &io_end->work);
iocb->private = NULL;
+ queue_work(wq, &io_end->work);
/* XXX: probably should move into the real I/O completion handler */
inode_dio_done(inode);
diff --git a/fs/namespace.c b/fs/namespace.c
index 6d3a1963879b..cfc6d4448aa5 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1048,15 +1048,12 @@ static int show_mountinfo(struct seq_file *m, void *v)
if (err)
goto out;
seq_putc(m, ' ');
- seq_path_root(m, &mnt_path, &root, " \t\n\\");
- if (root.mnt != p->root.mnt || root.dentry != p->root.dentry) {
- /*
- * Mountpoint is outside root, discard that one. Ugly,
- * but less so than trying to do that in iterator in a
- * race-free way (due to renames).
- */
- return SEQ_SKIP;
- }
+
+ /* mountpoints outside of chroot jail will give SEQ_SKIP on this */
+ err = seq_path_root(m, &mnt_path, &root, " \t\n\\");
+ if (err)
+ goto out;
+
seq_puts(m, mnt->mnt_flags & MNT_READONLY ? " ro" : " rw");
show_mnt_opts(m, mnt);
@@ -2776,3 +2773,8 @@ void kern_unmount(struct vfsmount *mnt)
}
}
EXPORT_SYMBOL(kern_unmount);
+
+bool our_mnt(struct vfsmount *mnt)
+{
+ return check_mnt(mnt);
+}
diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index ed553c60de82..3165aebb43c8 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -5699,7 +5699,7 @@ int ocfs2_remove_btree_range(struct inode *inode,
OCFS2_JOURNAL_ACCESS_WRITE);
if (ret) {
mlog_errno(ret);
- goto out;
+ goto out_commit;
}
dquot_free_space_nodirty(inode,
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index c1efe939c774..78b68af3b0e3 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -290,7 +290,15 @@ static int ocfs2_readpage(struct file *file, struct page *page)
}
if (down_read_trylock(&oi->ip_alloc_sem) == 0) {
+ /*
+ * Unlock the page and cycle ip_alloc_sem so that we don't
+ * busyloop waiting for ip_alloc_sem to unlock
+ */
ret = AOP_TRUNCATED_PAGE;
+ unlock_page(page);
+ unlock = 0;
+ down_read(&oi->ip_alloc_sem);
+ up_read(&oi->ip_alloc_sem);
goto out_inode_unlock;
}
@@ -563,6 +571,7 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
{
struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode;
int level;
+ wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
/* this io's submitter should not have unlocked this before we could */
BUG_ON(!ocfs2_iocb_is_rw_locked(iocb));
@@ -570,6 +579,15 @@ static void ocfs2_dio_end_io(struct kiocb *iocb,
if (ocfs2_iocb_is_sem_locked(iocb))
ocfs2_iocb_clear_sem_locked(iocb);
+ if (ocfs2_iocb_is_unaligned_aio(iocb)) {
+ ocfs2_iocb_clear_unaligned_aio(iocb);
+
+ if (atomic_dec_and_test(&OCFS2_I(inode)->ip_unaligned_aio) &&
+ waitqueue_active(wq)) {
+ wake_up_all(wq);
+ }
+ }
+
ocfs2_iocb_clear_rw_locked(iocb);
level = ocfs2_iocb_rw_locked_level(iocb);
@@ -863,6 +881,12 @@ struct ocfs2_write_ctxt {
struct page *w_target_page;
/*
+ * w_target_locked is used for page_mkwrite path indicating no unlocking
+ * against w_target_page in ocfs2_write_end_nolock.
+ */
+ unsigned int w_target_locked:1;
+
+ /*
* ocfs2_write_end() uses this to know what the real range to
* write in the target should be.
*/
@@ -895,6 +919,24 @@ void ocfs2_unlock_and_free_pages(struct page **pages, int num_pages)
static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc)
{
+ int i;
+
+ /*
+ * w_target_locked is only set to true in the page_mkwrite() case.
+ * The intent is to allow us to lock the target page from write_begin()
+ * to write_end(). The caller must hold a ref on w_target_page.
+ */
+ if (wc->w_target_locked) {
+ BUG_ON(!wc->w_target_page);
+ for (i = 0; i < wc->w_num_pages; i++) {
+ if (wc->w_target_page == wc->w_pages[i]) {
+ wc->w_pages[i] = NULL;
+ break;
+ }
+ }
+ mark_page_accessed(wc->w_target_page);
+ page_cache_release(wc->w_target_page);
+ }
ocfs2_unlock_and_free_pages(wc->w_pages, wc->w_num_pages);
brelse(wc->w_di_bh);
@@ -1132,20 +1174,17 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
*/
lock_page(mmap_page);
+ /* Exit and let the caller retry */
if (mmap_page->mapping != mapping) {
+ WARN_ON(mmap_page->mapping);
unlock_page(mmap_page);
- /*
- * Sanity check - the locking in
- * ocfs2_pagemkwrite() should ensure
- * that this code doesn't trigger.
- */
- ret = -EINVAL;
- mlog_errno(ret);
+ ret = -EAGAIN;
goto out;
}
page_cache_get(mmap_page);
wc->w_pages[i] = mmap_page;
+ wc->w_target_locked = true;
} else {
wc->w_pages[i] = find_or_create_page(mapping, index,
GFP_NOFS);
@@ -1160,6 +1199,8 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping,
wc->w_target_page = wc->w_pages[i];
}
out:
+ if (ret)
+ wc->w_target_locked = false;
return ret;
}
@@ -1817,11 +1858,23 @@ try_again:
*/
ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos, len,
cluster_of_pages, mmap_page);
- if (ret) {
+ if (ret && ret != -EAGAIN) {
mlog_errno(ret);
goto out_quota;
}
+ /*
+ * ocfs2_grab_pages_for_write() returns -EAGAIN if it could not lock
+ * the target page. In this case, we exit with no error and no target
+ * page. This will trigger the caller, page_mkwrite(), to re-try
+ * the operation.
+ */
+ if (ret == -EAGAIN) {
+ BUG_ON(wc->w_target_page);
+ ret = 0;
+ goto out_quota;
+ }
+
ret = ocfs2_write_cluster_by_desc(mapping, data_ac, meta_ac, wc, pos,
len);
if (ret) {
diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h
index 75cf3ad987a6..ffb2da370a99 100644
--- a/fs/ocfs2/aops.h
+++ b/fs/ocfs2/aops.h
@@ -78,6 +78,7 @@ enum ocfs2_iocb_lock_bits {
OCFS2_IOCB_RW_LOCK = 0,
OCFS2_IOCB_RW_LOCK_LEVEL,
OCFS2_IOCB_SEM,
+ OCFS2_IOCB_UNALIGNED_IO,
OCFS2_IOCB_NUM_LOCKS
};
@@ -91,4 +92,17 @@ enum ocfs2_iocb_lock_bits {
clear_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
#define ocfs2_iocb_is_sem_locked(iocb) \
test_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private)
+
+#define ocfs2_iocb_set_unaligned_aio(iocb) \
+ set_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_clear_unaligned_aio(iocb) \
+ clear_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+#define ocfs2_iocb_is_unaligned_aio(iocb) \
+ test_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private)
+
+#define OCFS2_IOEND_WQ_HASH_SZ 37
+#define ocfs2_ioend_wq(v) (&ocfs2__ioend_wq[((unsigned long)(v)) %\
+ OCFS2_IOEND_WQ_HASH_SZ])
+extern wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
#endif /* OCFS2_FILE_H */
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 9a3e6bbff27b..a4e855e3690e 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -216,6 +216,7 @@ struct o2hb_region {
struct list_head hr_all_item;
unsigned hr_unclean_stop:1,
+ hr_aborted_start:1,
hr_item_pinned:1,
hr_item_dropped:1;
@@ -254,6 +255,10 @@ struct o2hb_region {
* a more complete api that doesn't lead to this sort of fragility. */
atomic_t hr_steady_iterations;
+ /* terminate o2hb thread if it does not reach steady state
+ * (hr_steady_iterations == 0) within hr_unsteady_iterations */
+ atomic_t hr_unsteady_iterations;
+
char hr_dev_name[BDEVNAME_SIZE];
unsigned int hr_timeout_ms;
@@ -324,6 +329,10 @@ static void o2hb_write_timeout(struct work_struct *work)
static void o2hb_arm_write_timeout(struct o2hb_region *reg)
{
+ /* Arm writeout only after thread reaches steady state */
+ if (atomic_read(&reg->hr_steady_iterations) != 0)
+ return;
+
mlog(ML_HEARTBEAT, "Queue write timeout for %u ms\n",
O2HB_MAX_WRITE_TIMEOUT_MS);
@@ -537,9 +546,14 @@ static int o2hb_verify_crc(struct o2hb_region *reg,
return read == computed;
}
-/* We want to make sure that nobody is heartbeating on top of us --
- * this will help detect an invalid configuration. */
-static void o2hb_check_last_timestamp(struct o2hb_region *reg)
+/*
+ * Compare the slot data with what we wrote in the last iteration.
+ * If the match fails, print an appropriate error message. This is to
+ * detect errors like... another node hearting on the same slot,
+ * flaky device that is losing writes, etc.
+ * Returns 1 if check succeeds, 0 otherwise.
+ */
+static int o2hb_check_own_slot(struct o2hb_region *reg)
{
struct o2hb_disk_slot *slot;
struct o2hb_disk_heartbeat_block *hb_block;
@@ -548,13 +562,13 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
slot = &reg->hr_slots[o2nm_this_node()];
/* Don't check on our 1st timestamp */
if (!slot->ds_last_time)
- return;
+ return 0;
hb_block = slot->ds_raw_block;
if (le64_to_cpu(hb_block->hb_seq) == slot->ds_last_time &&
le64_to_cpu(hb_block->hb_generation) == slot->ds_last_generation &&
hb_block->hb_node == slot->ds_node_num)
- return;
+ return 1;
#define ERRSTR1 "Another node is heartbeating on device"
#define ERRSTR2 "Heartbeat generation mismatch on device"
@@ -574,6 +588,8 @@ static void o2hb_check_last_timestamp(struct o2hb_region *reg)
(unsigned long long)slot->ds_last_time, hb_block->hb_node,
(unsigned long long)le64_to_cpu(hb_block->hb_generation),
(unsigned long long)le64_to_cpu(hb_block->hb_seq));
+
+ return 0;
}
static inline void o2hb_prepare_block(struct o2hb_region *reg,
@@ -719,17 +735,24 @@ static void o2hb_shutdown_slot(struct o2hb_disk_slot *slot)
o2nm_node_put(node);
}
-static void o2hb_set_quorum_device(struct o2hb_region *reg,
- struct o2hb_disk_slot *slot)
+static void o2hb_set_quorum_device(struct o2hb_region *reg)
{
- assert_spin_locked(&o2hb_live_lock);
-
if (!o2hb_global_heartbeat_active())
return;
- if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+ /* Prevent race with o2hb_heartbeat_group_drop_item() */
+ if (kthread_should_stop())
+ return;
+
+ /* Tag region as quorum only after thread reaches steady state */
+ if (atomic_read(&reg->hr_steady_iterations) != 0)
return;
+ spin_lock(&o2hb_live_lock);
+
+ if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+ goto unlock;
+
/*
* A region can be added to the quorum only when it sees all
* live nodes heartbeat on it. In other words, the region has been
@@ -737,13 +760,10 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
*/
if (memcmp(reg->hr_live_node_bitmap, o2hb_live_node_bitmap,
sizeof(o2hb_live_node_bitmap)))
- return;
-
- if (slot->ds_changed_samples < O2HB_LIVE_THRESHOLD)
- return;
+ goto unlock;
- printk(KERN_NOTICE "o2hb: Region %s is now a quorum device\n",
- config_item_name(&reg->hr_item));
+ printk(KERN_NOTICE "o2hb: Region %s (%s) is now a quorum device\n",
+ config_item_name(&reg->hr_item), reg->hr_dev_name);
set_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
@@ -754,6 +774,8 @@ static void o2hb_set_quorum_device(struct o2hb_region *reg,
if (o2hb_pop_count(&o2hb_quorum_region_bitmap,
O2NM_MAX_REGIONS) > O2HB_PIN_CUT_OFF)
o2hb_region_unpin(NULL);
+unlock:
+ spin_unlock(&o2hb_live_lock);
}
static int o2hb_check_slot(struct o2hb_region *reg,
@@ -925,8 +947,6 @@ fire_callbacks:
slot->ds_equal_samples = 0;
}
out:
- o2hb_set_quorum_device(reg, slot);
-
spin_unlock(&o2hb_live_lock);
o2hb_run_event_list(&event);
@@ -957,7 +977,8 @@ static int o2hb_highest_node(unsigned long *nodes,
static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
{
- int i, ret, highest_node, change = 0;
+ int i, ret, highest_node;
+ int membership_change = 0, own_slot_ok = 0;
unsigned long configured_nodes[BITS_TO_LONGS(O2NM_MAX_NODES)];
unsigned long live_node_bitmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
struct o2hb_bio_wait_ctxt write_wc;
@@ -966,7 +987,7 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
sizeof(configured_nodes));
if (ret) {
mlog_errno(ret);
- return ret;
+ goto bail;
}
/*
@@ -982,8 +1003,9 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
highest_node = o2hb_highest_node(configured_nodes, O2NM_MAX_NODES);
if (highest_node >= O2NM_MAX_NODES) {
- mlog(ML_NOTICE, "ocfs2_heartbeat: no configured nodes found!\n");
- return -EINVAL;
+ mlog(ML_NOTICE, "o2hb: No configured nodes found!\n");
+ ret = -EINVAL;
+ goto bail;
}
/* No sense in reading the slots of nodes that don't exist
@@ -993,29 +1015,27 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
ret = o2hb_read_slots(reg, highest_node + 1);
if (ret < 0) {
mlog_errno(ret);
- return ret;
+ goto bail;
}
/* With an up to date view of the slots, we can check that no
* other node has been improperly configured to heartbeat in
* our slot. */
- o2hb_check_last_timestamp(reg);
+ own_slot_ok = o2hb_check_own_slot(reg);
/* fill in the proper info for our next heartbeat */
o2hb_prepare_block(reg, reg->hr_generation);
- /* And fire off the write. Note that we don't wait on this I/O
- * until later. */
ret = o2hb_issue_node_write(reg, &write_wc);
if (ret < 0) {
mlog_errno(ret);
- return ret;
+ goto bail;
}
i = -1;
while((i = find_next_bit(configured_nodes,
O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) {
- change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
+ membership_change |= o2hb_check_slot(reg, &reg->hr_slots[i]);
}
/*
@@ -1030,18 +1050,39 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg)
* disk */
mlog(ML_ERROR, "Write error %d on device \"%s\"\n",
write_wc.wc_error, reg->hr_dev_name);
- return write_wc.wc_error;
+ ret = write_wc.wc_error;
+ goto bail;
}
- o2hb_arm_write_timeout(reg);
+ /* Skip disarming the timeout if own slot has stale/bad data */
+ if (own_slot_ok) {
+ o2hb_set_quorum_device(reg);
+ o2hb_arm_write_timeout(reg);
+ }
+bail:
/* let the person who launched us know when things are steady */
- if (!change && (atomic_read(&reg->hr_steady_iterations) != 0)) {
- if (atomic_dec_and_test(&reg->hr_steady_iterations))
+ if (atomic_read(&reg->hr_steady_iterations) != 0) {
+ if (!ret && own_slot_ok && !membership_change) {
+ if (atomic_dec_and_test(&reg->hr_steady_iterations))
+ wake_up(&o2hb_steady_queue);
+ }
+ }
+
+ if (atomic_read(&reg->hr_steady_iterations) != 0) {
+ if (atomic_dec_and_test(&reg->hr_unsteady_iterations)) {
+ printk(KERN_NOTICE "o2hb: Unable to stabilize "
+ "heartbeart on region %s (%s)\n",
+ config_item_name(&reg->hr_item),
+ reg->hr_dev_name);
+ atomic_set(&reg->hr_steady_iterations, 0);
+ reg->hr_aborted_start = 1;
wake_up(&o2hb_steady_queue);
+ ret = -EIO;
+ }
}
- return 0;
+ return ret;
}
/* Subtract b from a, storing the result in a. a *must* have a larger
@@ -1095,7 +1136,8 @@ static int o2hb_thread(void *data)
/* Pin node */
o2nm_depend_this_node();
- while (!kthread_should_stop() && !reg->hr_unclean_stop) {
+ while (!kthread_should_stop() &&
+ !reg->hr_unclean_stop && !reg->hr_aborted_start) {
/* We track the time spent inside
* o2hb_do_disk_heartbeat so that we avoid more than
* hr_timeout_ms between disk writes. On busy systems
@@ -1103,10 +1145,7 @@ static int o2hb_thread(void *data)
* likely to time itself out. */
do_gettimeofday(&before_hb);
- i = 0;
- do {
- ret = o2hb_do_disk_heartbeat(reg);
- } while (ret && ++i < 2);
+ ret = o2hb_do_disk_heartbeat(reg);
do_gettimeofday(&after_hb);
elapsed_msec = o2hb_elapsed_msecs(&before_hb, &after_hb);
@@ -1117,7 +1156,8 @@ static int o2hb_thread(void *data)
after_hb.tv_sec, (unsigned long) after_hb.tv_usec,
elapsed_msec);
- if (elapsed_msec < reg->hr_timeout_ms) {
+ if (!kthread_should_stop() &&
+ elapsed_msec < reg->hr_timeout_ms) {
/* the kthread api has blocked signals for us so no
* need to record the return value. */
msleep_interruptible(reg->hr_timeout_ms - elapsed_msec);
@@ -1134,20 +1174,20 @@ static int o2hb_thread(void *data)
* to timeout on this region when we could just as easily
* write a clear generation - thus indicating to them that
* this node has left this region.
- *
- * XXX: Should we skip this on unclean_stop? */
- o2hb_prepare_block(reg, 0);
- ret = o2hb_issue_node_write(reg, &write_wc);
- if (ret == 0) {
- o2hb_wait_on_io(reg, &write_wc);
- } else {
- mlog_errno(ret);
+ */
+ if (!reg->hr_unclean_stop && !reg->hr_aborted_start) {
+ o2hb_prepare_block(reg, 0);
+ ret = o2hb_issue_node_write(reg, &write_wc);
+ if (ret == 0)
+ o2hb_wait_on_io(reg, &write_wc);
+ else
+ mlog_errno(ret);
}
/* Unpin node */
o2nm_undepend_this_node();
- mlog(ML_HEARTBEAT|ML_KTHREAD, "hb thread exiting\n");
+ mlog(ML_HEARTBEAT|ML_KTHREAD, "o2hb thread exiting\n");
return 0;
}
@@ -1158,6 +1198,7 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
struct o2hb_debug_buf *db = inode->i_private;
struct o2hb_region *reg;
unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+ unsigned long lts;
char *buf = NULL;
int i = -1;
int out = 0;
@@ -1194,9 +1235,11 @@ static int o2hb_debug_open(struct inode *inode, struct file *file)
case O2HB_DB_TYPE_REGION_ELAPSED_TIME:
reg = (struct o2hb_region *)db->db_data;
- out += snprintf(buf + out, PAGE_SIZE - out, "%u\n",
- jiffies_to_msecs(jiffies -
- reg->hr_last_timeout_start));
+ lts = reg->hr_last_timeout_start;
+ /* If 0, it has never been set before */
+ if (lts)
+ lts = jiffies_to_msecs(jiffies - lts);
+ out += snprintf(buf + out, PAGE_SIZE - out, "%lu\n", lts);
goto done;
case O2HB_DB_TYPE_REGION_PINNED:
@@ -1426,6 +1469,8 @@ static void o2hb_region_release(struct config_item *item)
struct page *page;
struct o2hb_region *reg = to_o2hb_region(item);
+ mlog(ML_HEARTBEAT, "hb region release (%s)\n", reg->hr_dev_name);
+
if (reg->hr_tmp_block)
kfree(reg->hr_tmp_block);
@@ -1792,7 +1837,10 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
live_threshold <<= 1;
spin_unlock(&o2hb_live_lock);
}
- atomic_set(&reg->hr_steady_iterations, live_threshold + 1);
+ ++live_threshold;
+ atomic_set(&reg->hr_steady_iterations, live_threshold);
+ /* unsteady_iterations is double the steady_iterations */
+ atomic_set(&reg->hr_unsteady_iterations, (live_threshold << 1));
hb_task = kthread_run(o2hb_thread, reg, "o2hb-%s",
reg->hr_item.ci_name);
@@ -1809,14 +1857,12 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
ret = wait_event_interruptible(o2hb_steady_queue,
atomic_read(&reg->hr_steady_iterations) == 0);
if (ret) {
- /* We got interrupted (hello ptrace!). Clean up */
- spin_lock(&o2hb_live_lock);
- hb_task = reg->hr_task;
- reg->hr_task = NULL;
- spin_unlock(&o2hb_live_lock);
+ atomic_set(&reg->hr_steady_iterations, 0);
+ reg->hr_aborted_start = 1;
+ }
- if (hb_task)
- kthread_stop(hb_task);
+ if (reg->hr_aborted_start) {
+ ret = -EIO;
goto out;
}
@@ -1833,8 +1879,8 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg,
ret = -EIO;
if (hb_task && o2hb_global_heartbeat_active())
- printk(KERN_NOTICE "o2hb: Heartbeat started on region %s\n",
- config_item_name(&reg->hr_item));
+ printk(KERN_NOTICE "o2hb: Heartbeat started on region %s (%s)\n",
+ config_item_name(&reg->hr_item), reg->hr_dev_name);
out:
if (filp)
@@ -2092,13 +2138,6 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
/* stop the thread when the user removes the region dir */
spin_lock(&o2hb_live_lock);
- if (o2hb_global_heartbeat_active()) {
- clear_bit(reg->hr_region_num, o2hb_region_bitmap);
- clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
- if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
- quorum_region = 1;
- clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
- }
hb_task = reg->hr_task;
reg->hr_task = NULL;
reg->hr_item_dropped = 1;
@@ -2107,19 +2146,30 @@ static void o2hb_heartbeat_group_drop_item(struct config_group *group,
if (hb_task)
kthread_stop(hb_task);
+ if (o2hb_global_heartbeat_active()) {
+ spin_lock(&o2hb_live_lock);
+ clear_bit(reg->hr_region_num, o2hb_region_bitmap);
+ clear_bit(reg->hr_region_num, o2hb_live_region_bitmap);
+ if (test_bit(reg->hr_region_num, o2hb_quorum_region_bitmap))
+ quorum_region = 1;
+ clear_bit(reg->hr_region_num, o2hb_quorum_region_bitmap);
+ spin_unlock(&o2hb_live_lock);
+ printk(KERN_NOTICE "o2hb: Heartbeat %s on region %s (%s)\n",
+ ((atomic_read(&reg->hr_steady_iterations) == 0) ?
+ "stopped" : "start aborted"), config_item_name(item),
+ reg->hr_dev_name);
+ }
+
/*
* If we're racing a dev_write(), we need to wake them. They will
* check reg->hr_task
*/
if (atomic_read(&reg->hr_steady_iterations) != 0) {
+ reg->hr_aborted_start = 1;
atomic_set(&reg->hr_steady_iterations, 0);
wake_up(&o2hb_steady_queue);
}
- if (o2hb_global_heartbeat_active())
- printk(KERN_NOTICE "o2hb: Heartbeat stopped on region %s\n",
- config_item_name(&reg->hr_item));
-
config_item_put(item);
if (!o2hb_global_heartbeat_active() || !quorum_region)
diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c
index 3a5835904b3d..dc45deb19e68 100644
--- a/fs/ocfs2/cluster/netdebug.c
+++ b/fs/ocfs2/cluster/netdebug.c
@@ -47,6 +47,7 @@
#define SC_DEBUG_NAME "sock_containers"
#define NST_DEBUG_NAME "send_tracking"
#define STATS_DEBUG_NAME "stats"
+#define NODES_DEBUG_NAME "connected_nodes"
#define SHOW_SOCK_CONTAINERS 0
#define SHOW_SOCK_STATS 1
@@ -55,6 +56,7 @@ static struct dentry *o2net_dentry;
static struct dentry *sc_dentry;
static struct dentry *nst_dentry;
static struct dentry *stats_dentry;
+static struct dentry *nodes_dentry;
static DEFINE_SPINLOCK(o2net_debug_lock);
@@ -491,53 +493,87 @@ static const struct file_operations sc_seq_fops = {
.release = sc_fop_release,
};
-int o2net_debugfs_init(void)
+static int o2net_fill_bitmap(char *buf, int len)
{
- o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
- if (!o2net_dentry) {
- mlog_errno(-ENOMEM);
- goto bail;
- }
+ unsigned long map[BITS_TO_LONGS(O2NM_MAX_NODES)];
+ int i = -1, out = 0;
- nst_dentry = debugfs_create_file(NST_DEBUG_NAME, S_IFREG|S_IRUSR,
- o2net_dentry, NULL,
- &nst_seq_fops);
- if (!nst_dentry) {
- mlog_errno(-ENOMEM);
- goto bail;
- }
+ o2net_fill_node_map(map, sizeof(map));
- sc_dentry = debugfs_create_file(SC_DEBUG_NAME, S_IFREG|S_IRUSR,
- o2net_dentry, NULL,
- &sc_seq_fops);
- if (!sc_dentry) {
- mlog_errno(-ENOMEM);
- goto bail;
- }
+ while ((i = find_next_bit(map, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES)
+ out += snprintf(buf + out, PAGE_SIZE - out, "%d ", i);
+ out += snprintf(buf + out, PAGE_SIZE - out, "\n");
- stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, S_IFREG|S_IRUSR,
- o2net_dentry, NULL,
- &stats_seq_fops);
- if (!stats_dentry) {
- mlog_errno(-ENOMEM);
- goto bail;
- }
+ return out;
+}
+
+static int nodes_fop_open(struct inode *inode, struct file *file)
+{
+ char *buf;
+
+ buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+ if (!buf)
+ return -ENOMEM;
+
+ i_size_write(inode, o2net_fill_bitmap(buf, PAGE_SIZE));
+
+ file->private_data = buf;
return 0;
-bail:
- debugfs_remove(stats_dentry);
- debugfs_remove(sc_dentry);
- debugfs_remove(nst_dentry);
- debugfs_remove(o2net_dentry);
- return -ENOMEM;
}
+static int o2net_debug_release(struct inode *inode, struct file *file)
+{
+ kfree(file->private_data);
+ return 0;
+}
+
+static ssize_t o2net_debug_read(struct file *file, char __user *buf,
+ size_t nbytes, loff_t *ppos)
+{
+ return simple_read_from_buffer(buf, nbytes, ppos, file->private_data,
+ i_size_read(file->f_mapping->host));
+}
+
+static const struct file_operations nodes_fops = {
+ .open = nodes_fop_open,
+ .release = o2net_debug_release,
+ .read = o2net_debug_read,
+ .llseek = generic_file_llseek,
+};
+
void o2net_debugfs_exit(void)
{
+ debugfs_remove(nodes_dentry);
debugfs_remove(stats_dentry);
debugfs_remove(sc_dentry);
debugfs_remove(nst_dentry);
debugfs_remove(o2net_dentry);
}
+int o2net_debugfs_init(void)
+{
+ mode_t mode = S_IFREG|S_IRUSR;
+
+ o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL);
+ if (o2net_dentry)
+ nst_dentry = debugfs_create_file(NST_DEBUG_NAME, mode,
+ o2net_dentry, NULL, &nst_seq_fops);
+ if (nst_dentry)
+ sc_dentry = debugfs_create_file(SC_DEBUG_NAME, mode,
+ o2net_dentry, NULL, &sc_seq_fops);
+ if (sc_dentry)
+ stats_dentry = debugfs_create_file(STATS_DEBUG_NAME, mode,
+ o2net_dentry, NULL, &stats_seq_fops);
+ if (stats_dentry)
+ nodes_dentry = debugfs_create_file(NODES_DEBUG_NAME, mode,
+ o2net_dentry, NULL, &nodes_fops);
+ if (nodes_dentry)
+ return 0;
+
+ o2net_debugfs_exit();
+ mlog_errno(-ENOMEM);
+ return -ENOMEM;
+}
+
#endif /* CONFIG_DEBUG_FS */
diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index ad7d0c155de4..044e7b58d31c 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -546,7 +546,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
}
if (was_valid && !valid) {
- printk(KERN_NOTICE "o2net: no longer connected to "
+ printk(KERN_NOTICE "o2net: No longer connected to "
SC_NODEF_FMT "\n", SC_NODEF_ARGS(old_sc));
o2net_complete_nodes_nsw(nn);
}
@@ -556,7 +556,7 @@ static void o2net_set_nn_state(struct o2net_node *nn,
cancel_delayed_work(&nn->nn_connect_expired);
printk(KERN_NOTICE "o2net: %s " SC_NODEF_FMT "\n",
o2nm_this_node() > sc->sc_node->nd_num ?
- "connected to" : "accepted connection from",
+ "Connected to" : "Accepted connection from",
SC_NODEF_ARGS(sc));
}
@@ -644,7 +644,7 @@ static void o2net_state_change(struct sock *sk)
o2net_sc_queue_work(sc, &sc->sc_connect_work);
break;
default:
- printk(KERN_INFO "o2net: connection to " SC_NODEF_FMT
+ printk(KERN_INFO "o2net: Connection to " SC_NODEF_FMT
" shutdown, state %d\n",
SC_NODEF_ARGS(sc), sk->sk_state);
o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
@@ -1035,6 +1035,25 @@ static int o2net_tx_can_proceed(struct o2net_node *nn,
return ret;
}
+/* Get a map of all nodes to which this node is currently connected to */
+void o2net_fill_node_map(unsigned long *map, unsigned bytes)
+{
+ struct o2net_sock_container *sc;
+ int node, ret;
+
+ BUG_ON(bytes < (BITS_TO_LONGS(O2NM_MAX_NODES) * sizeof(unsigned long)));
+
+ memset(map, 0, bytes);
+ for (node = 0; node < O2NM_MAX_NODES; ++node) {
+ o2net_tx_can_proceed(o2net_nn_from_num(node), &sc, &ret);
+ if (!ret) {
+ set_bit(node, map);
+ sc_put(sc);
+ }
+ }
+}
+EXPORT_SYMBOL_GPL(o2net_fill_node_map);
+
int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec,
size_t caller_veclen, u8 target_node, int *status)
{
@@ -1285,11 +1304,11 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
if (hand->protocol_version != cpu_to_be64(O2NET_PROTOCOL_VERSION)) {
- mlog(ML_NOTICE, SC_NODEF_FMT " advertised net protocol "
- "version %llu but %llu is required, disconnecting\n",
- SC_NODEF_ARGS(sc),
- (unsigned long long)be64_to_cpu(hand->protocol_version),
- O2NET_PROTOCOL_VERSION);
+ printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " Advertised net "
+ "protocol version %llu but %llu is required. "
+ "Disconnecting.\n", SC_NODEF_ARGS(sc),
+ (unsigned long long)be64_to_cpu(hand->protocol_version),
+ O2NET_PROTOCOL_VERSION);
/* don't bother reconnecting if its the wrong version. */
o2net_ensure_shutdown(nn, sc, -ENOTCONN);
@@ -1303,33 +1322,33 @@ static int o2net_check_handshake(struct o2net_sock_container *sc)
*/
if (be32_to_cpu(hand->o2net_idle_timeout_ms) !=
o2net_idle_timeout()) {
- mlog(ML_NOTICE, SC_NODEF_FMT " uses a network idle timeout of "
- "%u ms, but we use %u ms locally. disconnecting\n",
- SC_NODEF_ARGS(sc),
- be32_to_cpu(hand->o2net_idle_timeout_ms),
- o2net_idle_timeout());
+ printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a network "
+ "idle timeout of %u ms, but we use %u ms locally. "
+ "Disconnecting.\n", SC_NODEF_ARGS(sc),
+ be32_to_cpu(hand->o2net_idle_timeout_ms),
+ o2net_idle_timeout());
o2net_ensure_shutdown(nn, sc, -ENOTCONN);
return -1;
}
if (be32_to_cpu(hand->o2net_keepalive_delay_ms) !=
o2net_keepalive_delay()) {
- mlog(ML_NOTICE, SC_NODEF_FMT " uses a keepalive delay of "
- "%u ms, but we use %u ms locally. disconnecting\n",
- SC_NODEF_ARGS(sc),
- be32_to_cpu(hand->o2net_keepalive_delay_ms),
- o2net_keepalive_delay());
+ printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a keepalive "
+ "delay of %u ms, but we use %u ms locally. "
+ "Disconnecting.\n", SC_NODEF_ARGS(sc),
+ be32_to_cpu(hand->o2net_keepalive_delay_ms),
+ o2net_keepalive_delay());
o2net_ensure_shutdown(nn, sc, -ENOTCONN);
return -1;
}
if (be32_to_cpu(hand->o2hb_heartbeat_timeout_ms) !=
O2HB_MAX_WRITE_TIMEOUT_MS) {
- mlog(ML_NOTICE, SC_NODEF_FMT " uses a heartbeat timeout of "
- "%u ms, but we use %u ms locally. disconnecting\n",
- SC_NODEF_ARGS(sc),
- be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
- O2HB_MAX_WRITE_TIMEOUT_MS);
+ printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a heartbeat "
+ "timeout of %u ms, but we use %u ms locally. "
+ "Disconnecting.\n", SC_NODEF_ARGS(sc),
+ be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
+ O2HB_MAX_WRITE_TIMEOUT_MS);
o2net_ensure_shutdown(nn, sc, -ENOTCONN);
return -1;
}
@@ -1540,28 +1559,16 @@ static void o2net_idle_timer(unsigned long data)
{
struct o2net_sock_container *sc = (struct o2net_sock_container *)data;
struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
-
#ifdef CONFIG_DEBUG_FS
- ktime_t now = ktime_get();
+ unsigned long msecs = ktime_to_ms(ktime_get()) -
+ ktime_to_ms(sc->sc_tv_timer);
+#else
+ unsigned long msecs = o2net_idle_timeout();
#endif
- printk(KERN_NOTICE "o2net: connection to " SC_NODEF_FMT " has been idle for %u.%u "
- "seconds, shutting it down.\n", SC_NODEF_ARGS(sc),
- o2net_idle_timeout() / 1000,
- o2net_idle_timeout() % 1000);
-
-#ifdef CONFIG_DEBUG_FS
- mlog(ML_NOTICE, "Here are some times that might help debug the "
- "situation: (Timer: %lld, Now %lld, DataReady %lld, Advance %lld-%lld, "
- "Key 0x%08x, Func %u, FuncTime %lld-%lld)\n",
- (long long)ktime_to_us(sc->sc_tv_timer), (long long)ktime_to_us(now),
- (long long)ktime_to_us(sc->sc_tv_data_ready),
- (long long)ktime_to_us(sc->sc_tv_advance_start),
- (long long)ktime_to_us(sc->sc_tv_advance_stop),
- sc->sc_msg_key, sc->sc_msg_type,
- (long long)ktime_to_us(sc->sc_tv_func_start),
- (long long)ktime_to_us(sc->sc_tv_func_stop));
-#endif
+ printk(KERN_NOTICE "o2net: Connection to " SC_NODEF_FMT " has been "
+ "idle for %lu.%lu secs, shutting it down.\n", SC_NODEF_ARGS(sc),
+ msecs / 1000, msecs % 1000);
/*
* Initialize the nn_timeout so that the next connection attempt
@@ -1694,8 +1701,8 @@ static void o2net_start_connect(struct work_struct *work)
out:
if (ret) {
- mlog(ML_NOTICE, "connect attempt to " SC_NODEF_FMT " failed "
- "with errno %d\n", SC_NODEF_ARGS(sc), ret);
+ printk(KERN_NOTICE "o2net: Connect attempt to " SC_NODEF_FMT
+ " failed with errno %d\n", SC_NODEF_ARGS(sc), ret);
/* 0 err so that another will be queued and attempted
* from set_nn_state */
if (sc)
@@ -1718,8 +1725,8 @@ static void o2net_connect_expired(struct work_struct *work)
spin_lock(&nn->nn_lock);
if (!nn->nn_sc_valid) {
- mlog(ML_ERROR, "no connection established with node %u after "
- "%u.%u seconds, giving up and returning errors.\n",
+ printk(KERN_NOTICE "o2net: No connection established with "
+ "node %u after %u.%u seconds, giving up.\n",
o2net_num_from_nn(nn),
o2net_idle_timeout() / 1000,
o2net_idle_timeout() % 1000);
@@ -1862,21 +1869,21 @@ static int o2net_accept_one(struct socket *sock)
node = o2nm_get_node_by_ip(sin.sin_addr.s_addr);
if (node == NULL) {
- mlog(ML_NOTICE, "attempt to connect from unknown node at %pI4:%d\n",
- &sin.sin_addr.s_addr, ntohs(sin.sin_port));
+ printk(KERN_NOTICE "o2net: Attempt to connect from unknown "
+ "node at %pI4:%d\n", &sin.sin_addr.s_addr,
+ ntohs(sin.sin_port));
ret = -EINVAL;
goto out;
}
if (o2nm_this_node() >= node->nd_num) {
local_node = o2nm_get_node_by_num(o2nm_this_node());
- mlog(ML_NOTICE, "unexpected connect attempt seen at node '%s' ("
- "%u, %pI4:%d) from node '%s' (%u, %pI4:%d)\n",
- local_node->nd_name, local_node->nd_num,
- &(local_node->nd_ipv4_address),
- ntohs(local_node->nd_ipv4_port),
- node->nd_name, node->nd_num, &sin.sin_addr.s_addr,
- ntohs(sin.sin_port));
+ printk(KERN_NOTICE "o2net: Unexpected connect attempt seen "
+ "at node '%s' (%u, %pI4:%d) from node '%s' (%u, "
+ "%pI4:%d)\n", local_node->nd_name, local_node->nd_num,
+ &(local_node->nd_ipv4_address),
+ ntohs(local_node->nd_ipv4_port), node->nd_name,
+ node->nd_num, &sin.sin_addr.s_addr, ntohs(sin.sin_port));
ret = -EINVAL;
goto out;
}
@@ -1901,10 +1908,10 @@ static int o2net_accept_one(struct socket *sock)
ret = 0;
spin_unlock(&nn->nn_lock);
if (ret) {
- mlog(ML_NOTICE, "attempt to connect from node '%s' at "
- "%pI4:%d but it already has an open connection\n",
- node->nd_name, &sin.sin_addr.s_addr,
- ntohs(sin.sin_port));
+ printk(KERN_NOTICE "o2net: Attempt to connect from node '%s' "
+ "at %pI4:%d but it already has an open connection\n",
+ node->nd_name, &sin.sin_addr.s_addr,
+ ntohs(sin.sin_port));
goto out;
}
@@ -1984,7 +1991,7 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
if (ret < 0) {
- mlog(ML_ERROR, "unable to create socket, ret=%d\n", ret);
+ printk(KERN_ERR "o2net: Error %d while creating socket\n", ret);
goto out;
}
@@ -2001,16 +2008,15 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
sock->sk->sk_reuse = 1;
ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin));
if (ret < 0) {
- mlog(ML_ERROR, "unable to bind socket at %pI4:%u, "
- "ret=%d\n", &addr, ntohs(port), ret);
+ printk(KERN_ERR "o2net: Error %d while binding socket at "
+ "%pI4:%u\n", ret, &addr, ntohs(port));
goto out;
}
ret = sock->ops->listen(sock, 64);
- if (ret < 0) {
- mlog(ML_ERROR, "unable to listen on %pI4:%u, ret=%d\n",
- &addr, ntohs(port), ret);
- }
+ if (ret < 0)
+ printk(KERN_ERR "o2net: Error %d while listening on %pI4:%u\n",
+ ret, &addr, ntohs(port));
out:
if (ret) {
diff --git a/fs/ocfs2/cluster/tcp.h b/fs/ocfs2/cluster/tcp.h
index fd6179eb26d4..5bada2a69b50 100644
--- a/fs/ocfs2/cluster/tcp.h
+++ b/fs/ocfs2/cluster/tcp.h
@@ -106,6 +106,8 @@ int o2net_register_handler(u32 msg_type, u32 key, u32 max_len,
struct list_head *unreg_list);
void o2net_unregister_handler_list(struct list_head *list);
+void o2net_fill_node_map(unsigned long *map, unsigned bytes);
+
struct o2nm_node;
int o2net_register_hb_callbacks(void);
void o2net_unregister_hb_callbacks(void);
diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c
index e2878b5895fb..8fe4e2892ab9 100644
--- a/fs/ocfs2/dir.c
+++ b/fs/ocfs2/dir.c
@@ -1184,8 +1184,7 @@ static int __ocfs2_delete_entry(handle_t *handle, struct inode *dir,
if (pde)
le16_add_cpu(&pde->rec_len,
le16_to_cpu(de->rec_len));
- else
- de->inode = 0;
+ de->inode = 0;
dir->i_version++;
ocfs2_journal_dirty(handle, bh);
goto bail;
diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h
index d602abb51b61..a5952ceecba5 100644
--- a/fs/ocfs2/dlm/dlmcommon.h
+++ b/fs/ocfs2/dlm/dlmcommon.h
@@ -859,8 +859,8 @@ void dlm_complete_recovery_thread(struct dlm_ctxt *dlm);
void dlm_wait_for_recovery(struct dlm_ctxt *dlm);
void dlm_kick_recovery_thread(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);
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout);
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout);
void dlm_put(struct dlm_ctxt *dlm);
struct dlm_ctxt *dlm_grab(struct dlm_ctxt *dlm);
@@ -877,9 +877,8 @@ static inline void dlm_lockres_get(struct dlm_lock_resource *res)
kref_get(&res->refs);
}
void dlm_lockres_put(struct dlm_lock_resource *res);
-void __dlm_unhash_lockres(struct dlm_lock_resource *res);
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res);
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res);
struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
const char *name,
unsigned int len,
@@ -902,46 +901,15 @@ struct dlm_lock_resource *dlm_new_lockres(struct dlm_ctxt *dlm,
const char *name,
unsigned int namelen);
-#define dlm_lockres_set_refmap_bit(bit,res) \
- __dlm_lockres_set_refmap_bit(bit,res,__FILE__,__LINE__)
-#define dlm_lockres_clear_refmap_bit(bit,res) \
- __dlm_lockres_clear_refmap_bit(bit,res,__FILE__,__LINE__)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res, int bit);
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res, int bit);
-static inline void __dlm_lockres_set_refmap_bit(int bit,
- struct dlm_lock_resource *res,
- const char *file,
- int line)
-{
- //printk("%s:%d:%.*s: setting bit %d\n", file, line,
- // res->lockname.len, res->lockname.name, bit);
- set_bit(bit, res->refmap);
-}
-
-static inline void __dlm_lockres_clear_refmap_bit(int bit,
- struct dlm_lock_resource *res,
- const char *file,
- int line)
-{
- //printk("%s:%d:%.*s: clearing bit %d\n", file, line,
- // res->lockname.len, res->lockname.name, bit);
- clear_bit(bit, res->refmap);
-}
-
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res,
- const char *file,
- int line);
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res,
- int new_lockres,
- const char *file,
- int line);
-#define dlm_lockres_drop_inflight_ref(d,r) \
- __dlm_lockres_drop_inflight_ref(d,r,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref(d,r) \
- __dlm_lockres_grab_inflight_ref(d,r,0,__FILE__,__LINE__)
-#define dlm_lockres_grab_inflight_ref_new(d,r) \
- __dlm_lockres_grab_inflight_ref(d,r,1,__FILE__,__LINE__)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res);
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res);
void dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
void dlm_queue_bast(struct dlm_ctxt *dlm, struct dlm_lock *lock);
diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c
index 6ed6b95dcf93..92f2ead0fab6 100644
--- a/fs/ocfs2/dlm/dlmdomain.c
+++ b/fs/ocfs2/dlm/dlmdomain.c
@@ -157,16 +157,18 @@ static int dlm_protocol_compare(struct dlm_protocol_version *existing,
static void dlm_unregister_domain_handlers(struct dlm_ctxt *dlm);
-void __dlm_unhash_lockres(struct dlm_lock_resource *lockres)
+void __dlm_unhash_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
{
- if (!hlist_unhashed(&lockres->hash_node)) {
- hlist_del_init(&lockres->hash_node);
- dlm_lockres_put(lockres);
- }
+ if (hlist_unhashed(&res->hash_node))
+ return;
+
+ mlog(0, "%s: Unhash res %.*s\n", dlm->name, res->lockname.len,
+ res->lockname.name);
+ hlist_del_init(&res->hash_node);
+ dlm_lockres_put(res);
}
-void __dlm_insert_lockres(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res)
+void __dlm_insert_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
{
struct hlist_head *bucket;
struct qstr *q;
@@ -180,6 +182,9 @@ void __dlm_insert_lockres(struct dlm_ctxt *dlm,
dlm_lockres_get(res);
hlist_add_head(&res->hash_node, bucket);
+
+ mlog(0, "%s: Hash res %.*s\n", dlm->name, res->lockname.len,
+ res->lockname.name);
}
struct dlm_lock_resource * __dlm_lookup_lockres_full(struct dlm_ctxt *dlm,
@@ -539,17 +544,17 @@ again:
static void __dlm_print_nodes(struct dlm_ctxt *dlm)
{
- int node = -1;
+ int node = -1, num = 0;
assert_spin_locked(&dlm->spinlock);
- printk(KERN_NOTICE "o2dlm: Nodes in domain %s: ", dlm->name);
-
+ printk("( ");
while ((node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES,
node + 1)) < O2NM_MAX_NODES) {
printk("%d ", node);
+ ++num;
}
- printk("\n");
+ printk(") %u nodes\n", num);
}
static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
@@ -566,11 +571,10 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data,
node = exit_msg->node_idx;
- printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s\n", node, dlm->name);
-
spin_lock(&dlm->spinlock);
clear_bit(node, dlm->domain_map);
clear_bit(node, dlm->exit_domain_map);
+ printk(KERN_NOTICE "o2dlm: Node %u leaves domain %s ", node, dlm->name);
__dlm_print_nodes(dlm);
/* notify anything attached to the heartbeat events */
@@ -755,6 +759,7 @@ void dlm_unregister_domain(struct dlm_ctxt *dlm)
dlm_mark_domain_leaving(dlm);
dlm_leave_domain(dlm);
+ printk(KERN_NOTICE "o2dlm: Leaving domain %s\n", dlm->name);
dlm_force_free_mles(dlm);
dlm_complete_dlm_shutdown(dlm);
}
@@ -970,7 +975,7 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data,
clear_bit(assert->node_idx, dlm->exit_domain_map);
__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
- printk(KERN_NOTICE "o2dlm: Node %u joins domain %s\n",
+ printk(KERN_NOTICE "o2dlm: Node %u joins domain %s ",
assert->node_idx, dlm->name);
__dlm_print_nodes(dlm);
@@ -1701,8 +1706,10 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm)
bail:
spin_lock(&dlm->spinlock);
__dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN);
- if (!status)
+ if (!status) {
+ printk(KERN_NOTICE "o2dlm: Joining domain %s ", dlm->name);
__dlm_print_nodes(dlm);
+ }
spin_unlock(&dlm->spinlock);
if (ctxt) {
@@ -2131,13 +2138,6 @@ struct dlm_ctxt * dlm_register_domain(const char *domain,
goto leave;
}
- if (!o2hb_check_local_node_heartbeating()) {
- mlog(ML_ERROR, "the local node has not been configured, or is "
- "not heartbeating\n");
- ret = -EPROTO;
- goto leave;
- }
-
mlog(0, "register called for domain \"%s\"\n", domain);
retry:
diff --git a/fs/ocfs2/dlm/dlmlock.c b/fs/ocfs2/dlm/dlmlock.c
index 8d39e0fd66f7..975810b98492 100644
--- a/fs/ocfs2/dlm/dlmlock.c
+++ b/fs/ocfs2/dlm/dlmlock.c
@@ -183,10 +183,6 @@ static enum dlm_status dlmlock_master(struct dlm_ctxt *dlm,
kick_thread = 1;
}
}
- /* reduce the inflight count, this may result in the lockres
- * being purged below during calc_usage */
- if (lock->ml.node == dlm->node_num)
- dlm_lockres_drop_inflight_ref(dlm, res);
spin_unlock(&res->spinlock);
wake_up(&res->wq);
@@ -231,10 +227,16 @@ static enum dlm_status dlmlock_remote(struct dlm_ctxt *dlm,
lock->ml.type, res->lockname.len,
res->lockname.name, flags);
+ /*
+ * Wait if resource is getting recovered, remastered, etc.
+ * If the resource was remastered and new owner is self, then exit.
+ */
spin_lock(&res->spinlock);
-
- /* will exit this call with spinlock held */
__dlm_wait_on_lockres(res);
+ if (res->owner == dlm->node_num) {
+ spin_unlock(&res->spinlock);
+ return DLM_RECOVERING;
+ }
res->state |= DLM_LOCK_RES_IN_PROGRESS;
/* add lock to local (secondary) queue */
@@ -319,27 +321,23 @@ static enum dlm_status dlm_send_remote_lock_request(struct dlm_ctxt *dlm,
tmpret = o2net_send_message(DLM_CREATE_LOCK_MSG, dlm->key, &create,
sizeof(create), res->owner, &status);
if (tmpret >= 0) {
- // successfully sent and received
- ret = status; // this is already a dlm_status
+ ret = status;
if (ret == DLM_REJECTED) {
- mlog(ML_ERROR, "%s:%.*s: BUG. this is a stale lockres "
- "no longer owned by %u. that node is coming back "
- "up currently.\n", dlm->name, create.namelen,
+ mlog(ML_ERROR, "%s: res %.*s, Stale lockres no longer "
+ "owned by node %u. That node is coming back up "
+ "currently.\n", dlm->name, create.namelen,
create.name, res->owner);
dlm_print_one_lock_resource(res);
BUG();
}
} else {
- mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
- "node %u\n", tmpret, DLM_CREATE_LOCK_MSG, dlm->key,
- res->owner);
- if (dlm_is_host_down(tmpret)) {
+ mlog(ML_ERROR, "%s: res %.*s, Error %d send CREATE LOCK to "
+ "node %u\n", dlm->name, create.namelen, create.name,
+ tmpret, res->owner);
+ if (dlm_is_host_down(tmpret))
ret = DLM_RECOVERING;
- mlog(0, "node %u died so returning DLM_RECOVERING "
- "from lock message!\n", res->owner);
- } else {
+ else
ret = dlm_err_to_dlm_status(tmpret);
- }
}
return ret;
@@ -440,7 +438,7 @@ struct dlm_lock * dlm_new_lock(int type, u8 node, u64 cookie,
/* zero memory only if kernel-allocated */
lksb = kzalloc(sizeof(*lksb), GFP_NOFS);
if (!lksb) {
- kfree(lock);
+ kmem_cache_free(dlm_lock_cache, lock);
return NULL;
}
kernel_allocated = 1;
@@ -718,18 +716,10 @@ retry_lock:
if (status == DLM_RECOVERING || status == DLM_MIGRATING ||
status == DLM_FORWARD) {
- mlog(0, "retrying lock with migration/"
- "recovery/in progress\n");
msleep(100);
- /* no waiting for dlm_reco_thread */
if (recovery) {
if (status != DLM_RECOVERING)
goto retry_lock;
-
- mlog(0, "%s: got RECOVERING "
- "for $RECOVERY lock, master "
- "was %u\n", dlm->name,
- res->owner);
/* wait to see the node go down, then
* drop down and allow the lockres to
* get cleaned up. need to remaster. */
@@ -741,6 +731,14 @@ retry_lock:
}
}
+ /* Inflight taken in dlm_get_lock_resource() is dropped here */
+ spin_lock(&res->spinlock);
+ dlm_lockres_drop_inflight_ref(dlm, res);
+ spin_unlock(&res->spinlock);
+
+ dlm_lockres_calc_usage(dlm, res);
+ dlm_kick_thread(dlm, res);
+
if (status != DLM_NORMAL) {
lock->lksb->flags &= ~DLM_LKSB_GET_LVB;
if (status != DLM_NOTQUEUED)
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 11eefb8c12e9..005261c333b0 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -631,39 +631,54 @@ error:
return NULL;
}
-void __dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res,
- int new_lockres,
- const char *file,
- int line)
+void dlm_lockres_set_refmap_bit(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res, int bit)
{
- if (!new_lockres)
- assert_spin_locked(&res->spinlock);
+ assert_spin_locked(&res->spinlock);
+
+ mlog(0, "res %.*s, set node %u, %ps()\n", res->lockname.len,
+ res->lockname.name, bit, __builtin_return_address(0));
+
+ set_bit(bit, res->refmap);
+}
+
+void dlm_lockres_clear_refmap_bit(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res, int bit)
+{
+ assert_spin_locked(&res->spinlock);
+
+ mlog(0, "res %.*s, clr node %u, %ps()\n", res->lockname.len,
+ res->lockname.name, bit, __builtin_return_address(0));
+
+ clear_bit(bit, res->refmap);
+}
+
+
+void dlm_lockres_grab_inflight_ref(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
+{
+ assert_spin_locked(&res->spinlock);
- if (!test_bit(dlm->node_num, res->refmap)) {
- BUG_ON(res->inflight_locks != 0);
- dlm_lockres_set_refmap_bit(dlm->node_num, res);
- }
res->inflight_locks++;
- mlog(0, "%s:%.*s: inflight++: now %u\n",
- dlm->name, res->lockname.len, res->lockname.name,
- res->inflight_locks);
+
+ mlog(0, "%s: res %.*s, inflight++: now %u, %ps()\n", dlm->name,
+ res->lockname.len, res->lockname.name, res->inflight_locks,
+ __builtin_return_address(0));
}
-void __dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
- struct dlm_lock_resource *res,
- const char *file,
- int line)
+void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
{
assert_spin_locked(&res->spinlock);
BUG_ON(res->inflight_locks == 0);
+
res->inflight_locks--;
- mlog(0, "%s:%.*s: inflight--: now %u\n",
- dlm->name, res->lockname.len, res->lockname.name,
- res->inflight_locks);
- if (res->inflight_locks == 0)
- dlm_lockres_clear_refmap_bit(dlm->node_num, res);
+
+ mlog(0, "%s: res %.*s, inflight--: now %u, %ps()\n", dlm->name,
+ res->lockname.len, res->lockname.name, res->inflight_locks,
+ __builtin_return_address(0));
+
wake_up(&res->wq);
}
@@ -697,7 +712,6 @@ struct dlm_lock_resource * dlm_get_lock_resource(struct dlm_ctxt *dlm,
unsigned int hash;
int tries = 0;
int bit, wait_on_recovery = 0;
- int drop_inflight_if_nonlocal = 0;
BUG_ON(!lockid);
@@ -709,36 +723,33 @@ lookup:
spin_lock(&dlm->spinlock);
tmpres = __dlm_lookup_lockres_full(dlm, lockid, namelen, hash);
if (tmpres) {
- int dropping_ref = 0;
-
spin_unlock(&dlm->spinlock);
-
spin_lock(&tmpres->spinlock);
- /* We wait for the other thread that is mastering the resource */
+ /* Wait on the thread that is mastering the resource */
if (tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN) {
__dlm_wait_on_lockres(tmpres);
BUG_ON(tmpres->owner == DLM_LOCK_RES_OWNER_UNKNOWN);
+ spin_unlock(&tmpres->spinlock);
+ dlm_lockres_put(tmpres);
+ tmpres = NULL;
+ goto lookup;
}
- if (tmpres->owner == dlm->node_num) {
- BUG_ON(tmpres->state & DLM_LOCK_RES_DROPPING_REF);
- dlm_lockres_grab_inflight_ref(dlm, tmpres);
- } else if (tmpres->state & DLM_LOCK_RES_DROPPING_REF)
- dropping_ref = 1;
- spin_unlock(&tmpres->spinlock);
-
- /* wait until done messaging the master, drop our ref to allow
- * the lockres to be purged, start over. */
- if (dropping_ref) {
- spin_lock(&tmpres->spinlock);
- __dlm_wait_on_lockres_flags(tmpres, DLM_LOCK_RES_DROPPING_REF);
+ /* Wait on the resource purge to complete before continuing */
+ if (tmpres->state & DLM_LOCK_RES_DROPPING_REF) {
+ BUG_ON(tmpres->owner == dlm->node_num);
+ __dlm_wait_on_lockres_flags(tmpres,
+ DLM_LOCK_RES_DROPPING_REF);
spin_unlock(&tmpres->spinlock);
dlm_lockres_put(tmpres);
tmpres = NULL;
goto lookup;
}
- mlog(0, "found in hash!\n");
+ /* Grab inflight ref to pin the resource */
+ dlm_lockres_grab_inflight_ref(dlm, tmpres);
+
+ spin_unlock(&tmpres->spinlock);
if (res)
dlm_lockres_put(res);
res = tmpres;
@@ -829,8 +840,8 @@ lookup:
* but they might own this lockres. wait on them. */
bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
if (bit < O2NM_MAX_NODES) {
- mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
- "recover before lock mastery can begin\n",
+ mlog(0, "%s: res %.*s, At least one node (%d) "
+ "to recover before lock mastery can begin\n",
dlm->name, namelen, (char *)lockid, bit);
wait_on_recovery = 1;
}
@@ -843,12 +854,11 @@ lookup:
/* finally add the lockres to its hash bucket */
__dlm_insert_lockres(dlm, res);
- /* since this lockres is new it doesn't not require the spinlock */
- dlm_lockres_grab_inflight_ref_new(dlm, res);
- /* if this node does not become the master make sure to drop
- * this inflight reference below */
- drop_inflight_if_nonlocal = 1;
+ /* Grab inflight ref to pin the resource */
+ spin_lock(&res->spinlock);
+ dlm_lockres_grab_inflight_ref(dlm, res);
+ spin_unlock(&res->spinlock);
/* get an extra ref on the mle in case this is a BLOCK
* if so, the creator of the BLOCK may try to put the last
@@ -864,8 +874,8 @@ redo_request:
* dlm spinlock would be detectable be a change on the mle,
* so we only need to clear out the recovery map once. */
if (dlm_is_recovery_lock(lockid, namelen)) {
- mlog(ML_NOTICE, "%s: recovery map is not empty, but "
- "must master $RECOVERY lock now\n", dlm->name);
+ mlog(0, "%s: Recovery map is not empty, but must "
+ "master $RECOVERY lock now\n", dlm->name);
if (!dlm_pre_master_reco_lockres(dlm, res))
wait_on_recovery = 0;
else {
@@ -883,8 +893,8 @@ redo_request:
spin_lock(&dlm->spinlock);
bit = find_next_bit(dlm->recovery_map, O2NM_MAX_NODES, 0);
if (bit < O2NM_MAX_NODES) {
- mlog(ML_NOTICE, "%s:%.*s: at least one node (%d) to "
- "recover before lock mastery can begin\n",
+ mlog(0, "%s: res %.*s, At least one node (%d) "
+ "to recover before lock mastery can begin\n",
dlm->name, namelen, (char *)lockid, bit);
wait_on_recovery = 1;
} else
@@ -913,8 +923,8 @@ redo_request:
* yet, keep going until it does. this is how the
* master will know that asserts are needed back to
* the lower nodes. */
- mlog(0, "%s:%.*s: requests only up to %u but master "
- "is %u, keep going\n", dlm->name, namelen,
+ mlog(0, "%s: res %.*s, Requests only up to %u but "
+ "master is %u, keep going\n", dlm->name, namelen,
lockid, nodenum, mle->master);
}
}
@@ -924,13 +934,12 @@ wait:
ret = dlm_wait_for_lock_mastery(dlm, res, mle, &blocked);
if (ret < 0) {
wait_on_recovery = 1;
- mlog(0, "%s:%.*s: node map changed, redo the "
- "master request now, blocked=%d\n",
- dlm->name, res->lockname.len,
+ mlog(0, "%s: res %.*s, Node map changed, redo the master "
+ "request now, blocked=%d\n", dlm->name, res->lockname.len,
res->lockname.name, blocked);
if (++tries > 20) {
- mlog(ML_ERROR, "%s:%.*s: spinning on "
- "dlm_wait_for_lock_mastery, blocked=%d\n",
+ mlog(ML_ERROR, "%s: res %.*s, Spinning on "
+ "dlm_wait_for_lock_mastery, blocked = %d\n",
dlm->name, res->lockname.len,
res->lockname.name, blocked);
dlm_print_one_lock_resource(res);
@@ -940,7 +949,8 @@ wait:
goto redo_request;
}
- mlog(0, "lockres mastered by %u\n", res->owner);
+ mlog(0, "%s: res %.*s, Mastered by %u\n", dlm->name, res->lockname.len,
+ res->lockname.name, res->owner);
/* make sure we never continue without this */
BUG_ON(res->owner == O2NM_MAX_NODES);
@@ -952,8 +962,6 @@ wait:
wake_waiters:
spin_lock(&res->spinlock);
- if (res->owner != dlm->node_num && drop_inflight_if_nonlocal)
- dlm_lockres_drop_inflight_ref(dlm, res);
res->state &= ~DLM_LOCK_RES_IN_PROGRESS;
spin_unlock(&res->spinlock);
wake_up(&res->wq);
@@ -1426,9 +1434,7 @@ way_up_top:
}
if (res->owner == dlm->node_num) {
- mlog(0, "%s:%.*s: setting bit %u in refmap\n",
- dlm->name, namelen, name, request->node_idx);
- dlm_lockres_set_refmap_bit(request->node_idx, res);
+ dlm_lockres_set_refmap_bit(dlm, res, request->node_idx);
spin_unlock(&res->spinlock);
response = DLM_MASTER_RESP_YES;
if (mle)
@@ -1493,10 +1499,8 @@ way_up_top:
* go back and clean the mles on any
* other nodes */
dispatch_assert = 1;
- dlm_lockres_set_refmap_bit(request->node_idx, res);
- mlog(0, "%s:%.*s: setting bit %u in refmap\n",
- dlm->name, namelen, name,
- request->node_idx);
+ dlm_lockres_set_refmap_bit(dlm, res,
+ request->node_idx);
} else
response = DLM_MASTER_RESP_NO;
} else {
@@ -1702,7 +1706,7 @@ again:
"lockres, set the bit in the refmap\n",
namelen, lockname, to);
spin_lock(&res->spinlock);
- dlm_lockres_set_refmap_bit(to, res);
+ dlm_lockres_set_refmap_bit(dlm, res, to);
spin_unlock(&res->spinlock);
}
}
@@ -2187,8 +2191,6 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
namelen = res->lockname.len;
BUG_ON(namelen > O2NM_MAX_NAME_LEN);
- mlog(0, "%s:%.*s: sending deref to %d\n",
- dlm->name, namelen, lockname, res->owner);
memset(&deref, 0, sizeof(deref));
deref.node_idx = dlm->node_num;
deref.namelen = namelen;
@@ -2197,14 +2199,12 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res)
ret = o2net_send_message(DLM_DEREF_LOCKRES_MSG, dlm->key,
&deref, sizeof(deref), res->owner, &r);
if (ret < 0)
- mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to "
- "node %u\n", ret, DLM_DEREF_LOCKRES_MSG, dlm->key,
- res->owner);
+ mlog(ML_ERROR, "%s: res %.*s, error %d send DEREF to node %u\n",
+ dlm->name, namelen, lockname, ret, res->owner);
else if (r < 0) {
/* BAD. other node says I did not have a ref. */
- mlog(ML_ERROR,"while dropping ref on %s:%.*s "
- "(master=%u) got %d.\n", dlm->name, namelen,
- lockname, res->owner, r);
+ mlog(ML_ERROR, "%s: res %.*s, DEREF to node %u got %d\n",
+ dlm->name, namelen, lockname, res->owner, r);
dlm_print_one_lock_resource(res);
BUG();
}
@@ -2260,7 +2260,7 @@ int dlm_deref_lockres_handler(struct o2net_msg *msg, u32 len, void *data,
else {
BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
if (test_bit(node, res->refmap)) {
- dlm_lockres_clear_refmap_bit(node, res);
+ dlm_lockres_clear_refmap_bit(dlm, res, node);
cleared = 1;
}
}
@@ -2320,7 +2320,7 @@ static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data)
BUG_ON(res->state & DLM_LOCK_RES_DROPPING_REF);
if (test_bit(node, res->refmap)) {
__dlm_wait_on_lockres_flags(res, DLM_LOCK_RES_SETREF_INPROG);
- dlm_lockres_clear_refmap_bit(node, res);
+ dlm_lockres_clear_refmap_bit(dlm, res, node);
cleared = 1;
}
spin_unlock(&res->spinlock);
@@ -2802,7 +2802,8 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
BUG_ON(!list_empty(&lock->bast_list));
BUG_ON(lock->ast_pending);
BUG_ON(lock->bast_pending);
- dlm_lockres_clear_refmap_bit(lock->ml.node, res);
+ dlm_lockres_clear_refmap_bit(dlm, res,
+ lock->ml.node);
list_del_init(&lock->list);
dlm_lock_put(lock);
/* In a normal unlock, we would have added a
@@ -2823,7 +2824,7 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm,
mlog(0, "%s:%.*s: node %u had a ref to this "
"migrating lockres, clearing\n", dlm->name,
res->lockname.len, res->lockname.name, bit);
- dlm_lockres_clear_refmap_bit(bit, res);
+ dlm_lockres_clear_refmap_bit(dlm, res, bit);
}
bit++;
}
@@ -2916,9 +2917,9 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
&migrate, sizeof(migrate), nodenum,
&status);
if (ret < 0) {
- mlog(ML_ERROR, "Error %d when sending message %u (key "
- "0x%x) to node %u\n", ret, DLM_MIGRATE_REQUEST_MSG,
- dlm->key, nodenum);
+ mlog(ML_ERROR, "%s: res %.*s, Error %d send "
+ "MIGRATE_REQUEST to node %u\n", dlm->name,
+ migrate.namelen, migrate.name, ret, nodenum);
if (!dlm_is_host_down(ret)) {
mlog(ML_ERROR, "unhandled error=%d!\n", ret);
BUG();
@@ -2937,7 +2938,7 @@ static int dlm_do_migrate_request(struct dlm_ctxt *dlm,
dlm->name, res->lockname.len, res->lockname.name,
nodenum);
spin_lock(&res->spinlock);
- dlm_lockres_set_refmap_bit(nodenum, res);
+ dlm_lockres_set_refmap_bit(dlm, res, nodenum);
spin_unlock(&res->spinlock);
}
}
@@ -3271,7 +3272,7 @@ int dlm_finish_migration(struct dlm_ctxt *dlm, struct dlm_lock_resource *res,
* mastery reference here since old_master will briefly have
* a reference after the migration completes */
spin_lock(&res->spinlock);
- dlm_lockres_set_refmap_bit(old_master, res);
+ dlm_lockres_set_refmap_bit(dlm, res, old_master);
spin_unlock(&res->spinlock);
mlog(0, "now time to do a migrate request to other nodes\n");
diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c
index 7efab6d28a21..01ebfd0bdad7 100644
--- a/fs/ocfs2/dlm/dlmrecovery.c
+++ b/fs/ocfs2/dlm/dlmrecovery.c
@@ -362,40 +362,38 @@ static int dlm_is_node_recovered(struct dlm_ctxt *dlm, u8 node)
}
-int dlm_wait_for_node_death(struct dlm_ctxt *dlm, u8 node, int timeout)
+void 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);
+ if (dlm_is_node_dead(dlm, node))
+ return;
+
+ printk(KERN_NOTICE "o2dlm: Waiting on the death of node %u in "
+ "domain %s\n", node, dlm->name);
+
+ if (timeout)
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);
+ dlm_is_node_dead(dlm, node),
+ msecs_to_jiffies(timeout));
+ else
wait_event(dlm->dlm_reco_thread_wq,
dlm_is_node_dead(dlm, node));
- }
- /* for now, return 0 */
- return 0;
}
-int dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
+void dlm_wait_for_node_recovery(struct dlm_ctxt *dlm, u8 node, int timeout)
{
- if (timeout) {
- mlog(0, "%s: waiting %dms for notification of "
- "recovery of node %u\n", dlm->name, timeout, node);
+ if (dlm_is_node_recovered(dlm, node))
+ return;
+
+ printk(KERN_NOTICE "o2dlm: Waiting on the recovery of node %u in "
+ "domain %s\n", node, dlm->name);
+
+ if (timeout)
wait_event_timeout(dlm->dlm_reco_thread_wq,
- dlm_is_node_recovered(dlm, node),
- msecs_to_jiffies(timeout));
- } else {
- mlog(0, "%s: waiting indefinitely for notification "
- "of recovery of node %u\n", dlm->name, node);
+ dlm_is_node_recovered(dlm, node),
+ msecs_to_jiffies(timeout));
+ else
wait_event(dlm->dlm_reco_thread_wq,
dlm_is_node_recovered(dlm, node));
- }
- /* for now, return 0 */
- return 0;
}
/* callers of the top-level api calls (dlmlock/dlmunlock) should
@@ -430,6 +428,8 @@ static void dlm_begin_recovery(struct dlm_ctxt *dlm)
{
spin_lock(&dlm->spinlock);
BUG_ON(dlm->reco.state & DLM_RECO_STATE_ACTIVE);
+ printk(KERN_NOTICE "o2dlm: Begin recovery on domain %s for node %u\n",
+ dlm->name, dlm->reco.dead_node);
dlm->reco.state |= DLM_RECO_STATE_ACTIVE;
spin_unlock(&dlm->spinlock);
}
@@ -440,9 +440,18 @@ static void dlm_end_recovery(struct dlm_ctxt *dlm)
BUG_ON(!(dlm->reco.state & DLM_RECO_STATE_ACTIVE));
dlm->reco.state &= ~DLM_RECO_STATE_ACTIVE;
spin_unlock(&dlm->spinlock);
+ printk(KERN_NOTICE "o2dlm: End recovery on domain %s\n", dlm->name);
wake_up(&dlm->reco.event);
}
+static void dlm_print_recovery_master(struct dlm_ctxt *dlm)
+{
+ printk(KERN_NOTICE "o2dlm: Node %u (%s) is the Recovery Master for the "
+ "dead node %u in domain %s\n", dlm->reco.new_master,
+ (dlm->node_num == dlm->reco.new_master ? "me" : "he"),
+ dlm->reco.dead_node, dlm->name);
+}
+
static int dlm_do_recovery(struct dlm_ctxt *dlm)
{
int status = 0;
@@ -505,9 +514,8 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
}
mlog(0, "another node will master this recovery session.\n");
}
- mlog(0, "dlm=%s (%d), new_master=%u, this node=%u, dead_node=%u\n",
- dlm->name, task_pid_nr(dlm->dlm_reco_thread_task), dlm->reco.new_master,
- dlm->node_num, dlm->reco.dead_node);
+
+ dlm_print_recovery_master(dlm);
/* it is safe to start everything back up here
* because all of the dead node's lock resources
@@ -518,15 +526,13 @@ static int dlm_do_recovery(struct dlm_ctxt *dlm)
return 0;
master_here:
- mlog(ML_NOTICE, "(%d) Node %u is the Recovery Master for the Dead Node "
- "%u for Domain %s\n", task_pid_nr(dlm->dlm_reco_thread_task),
- dlm->node_num, dlm->reco.dead_node, dlm->name);
+ dlm_print_recovery_master(dlm);
status = dlm_remaster_locks(dlm, dlm->reco.dead_node);
if (status < 0) {
/* we should never hit this anymore */
- mlog(ML_ERROR, "error %d remastering locks for node %u, "
- "retrying.\n", status, dlm->reco.dead_node);
+ mlog(ML_ERROR, "%s: Error %d remastering locks for node %u, "
+ "retrying.\n", dlm->name, status, dlm->reco.dead_node);
/* yield a bit to allow any final network messages
* to get handled on remaining nodes */
msleep(100);
@@ -567,7 +573,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
BUG_ON(ndata->state != DLM_RECO_NODE_DATA_INIT);
ndata->state = DLM_RECO_NODE_DATA_REQUESTING;
- mlog(0, "requesting lock info from node %u\n",
+ mlog(0, "%s: Requesting lock info from node %u\n", dlm->name,
ndata->node_num);
if (ndata->node_num == dlm->node_num) {
@@ -640,7 +646,7 @@ static int dlm_remaster_locks(struct dlm_ctxt *dlm, u8 dead_node)
spin_unlock(&dlm_reco_state_lock);
}
- mlog(0, "done requesting all lock info\n");
+ mlog(0, "%s: Done requesting all lock info\n", dlm->name);
/* nodes should be sending reco data now
* just need to wait */
@@ -802,10 +808,9 @@ static int dlm_request_all_locks(struct dlm_ctxt *dlm, u8 request_from,
/* negative status is handled by caller */
if (ret < 0)
- mlog(ML_ERROR, "Error %d when sending message %u (key "
- "0x%x) to node %u\n", ret, DLM_LOCK_REQUEST_MSG,
- dlm->key, request_from);
-
+ mlog(ML_ERROR, "%s: Error %d send LOCK_REQUEST to node %u "
+ "to recover dead node %u\n", dlm->name, ret,
+ request_from, dead_node);
// return from here, then
// sleep until all received or error
return ret;
@@ -956,9 +961,9 @@ static int dlm_send_all_done_msg(struct dlm_ctxt *dlm, u8 dead_node, u8 send_to)
ret = o2net_send_message(DLM_RECO_DATA_DONE_MSG, dlm->key, &done_msg,
sizeof(done_msg), send_to, &tmpret);
if (ret < 0) {
- mlog(ML_ERROR, "Error %d when sending message %u (key "
- "0x%x) to node %u\n", ret, DLM_RECO_DATA_DONE_MSG,
- dlm->key, send_to);
+ mlog(ML_ERROR, "%s: Error %d send RECO_DATA_DONE to node %u "
+ "to recover dead node %u\n", dlm->name, ret, send_to,
+ dead_node);
if (!dlm_is_host_down(ret)) {
BUG();
}
@@ -1127,9 +1132,11 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm,
if (ret < 0) {
/* XXX: negative status is not handled.
* this will end up killing this node. */
- mlog(ML_ERROR, "Error %d when sending message %u (key "
- "0x%x) to node %u\n", ret, DLM_MIG_LOCKRES_MSG,
- dlm->key, send_to);
+ mlog(ML_ERROR, "%s: res %.*s, Error %d send MIG_LOCKRES to "
+ "node %u (%s)\n", dlm->name, mres->lockname_len,
+ mres->lockname, ret, send_to,
+ (orig_flags & DLM_MRES_MIGRATION ?
+ "migration" : "recovery"));
} else {
/* might get an -ENOMEM back here */
ret = status;
@@ -1767,7 +1774,7 @@ static int dlm_process_recovery_data(struct dlm_ctxt *dlm,
dlm->name, mres->lockname_len, mres->lockname,
from);
spin_lock(&res->spinlock);
- dlm_lockres_set_refmap_bit(from, res);
+ dlm_lockres_set_refmap_bit(dlm, res, from);
spin_unlock(&res->spinlock);
added++;
break;
@@ -1965,7 +1972,7 @@ skip_lvb:
mlog(0, "%s:%.*s: added lock for node %u, "
"setting refmap bit\n", dlm->name,
res->lockname.len, res->lockname.name, ml->node);
- dlm_lockres_set_refmap_bit(ml->node, res);
+ dlm_lockres_set_refmap_bit(dlm, res, ml->node);
added++;
}
spin_unlock(&res->spinlock);
@@ -2084,6 +2091,9 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
list_for_each_entry_safe(res, next, &dlm->reco.resources, recovering) {
if (res->owner == dead_node) {
+ mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+ dlm->name, res->lockname.len, res->lockname.name,
+ res->owner, new_master);
list_del_init(&res->recovering);
spin_lock(&res->spinlock);
/* new_master has our reference from
@@ -2105,40 +2115,30 @@ static void dlm_finish_local_lockres_recovery(struct dlm_ctxt *dlm,
for (i = 0; i < DLM_HASH_BUCKETS; i++) {
bucket = dlm_lockres_hash(dlm, 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 "
- "was not on recovering list, but "
- "clearing state anyway\n",
- dlm->node_num, res->lockname.len,
- res->lockname.name, new_master);
- } else if (res->owner == dlm->node_num) {
- mlog(0, "(this=%u) res %.*s owner=%u "
- "was not on recovering list, "
- "owner is THIS node, clearing\n",
- dlm->node_num, res->lockname.len,
- res->lockname.name, new_master);
- } else
- continue;
+ if (!(res->state & DLM_LOCK_RES_RECOVERING))
+ continue;
- if (!list_empty(&res->recovering)) {
- mlog(0, "%s:%.*s: lockres was "
- "marked RECOVERING, owner=%u\n",
- dlm->name, res->lockname.len,
- res->lockname.name, res->owner);
- list_del_init(&res->recovering);
- dlm_lockres_put(res);
- }
- spin_lock(&res->spinlock);
- /* new_master has our reference from
- * the lock state sent during recovery */
- dlm_change_lockres_owner(dlm, res, new_master);
- res->state &= ~DLM_LOCK_RES_RECOVERING;
- if (__dlm_lockres_has_locks(res))
- __dlm_dirty_lockres(dlm, res);
- spin_unlock(&res->spinlock);
- wake_up(&res->wq);
+ if (res->owner != dead_node &&
+ res->owner != dlm->node_num)
+ continue;
+
+ if (!list_empty(&res->recovering)) {
+ list_del_init(&res->recovering);
+ dlm_lockres_put(res);
}
+
+ /* new_master has our reference from
+ * the lock state sent during recovery */
+ mlog(0, "%s: res %.*s, Changing owner from %u to %u\n",
+ dlm->name, res->lockname.len, res->lockname.name,
+ res->owner, new_master);
+ spin_lock(&res->spinlock);
+ dlm_change_lockres_owner(dlm, res, new_master);
+ res->state &= ~DLM_LOCK_RES_RECOVERING;
+ if (__dlm_lockres_has_locks(res))
+ __dlm_dirty_lockres(dlm, res);
+ spin_unlock(&res->spinlock);
+ wake_up(&res->wq);
}
}
}
@@ -2252,12 +2252,12 @@ static void dlm_free_dead_locks(struct dlm_ctxt *dlm,
res->lockname.len, res->lockname.name, freed, dead_node);
__dlm_print_one_lock_resource(res);
}
- dlm_lockres_clear_refmap_bit(dead_node, res);
+ dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
} else if (test_bit(dead_node, res->refmap)) {
mlog(0, "%s:%.*s: dead node %u had a ref, but had "
"no locks and had not purged before dying\n", dlm->name,
res->lockname.len, res->lockname.name, dead_node);
- dlm_lockres_clear_refmap_bit(dead_node, res);
+ dlm_lockres_clear_refmap_bit(dlm, res, dead_node);
}
/* do not kick thread yet */
@@ -2324,9 +2324,9 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node)
dlm_revalidate_lvb(dlm, res, dead_node);
if (res->owner == dead_node) {
if (res->state & DLM_LOCK_RES_DROPPING_REF) {
- mlog(ML_NOTICE, "Ignore %.*s for "
+ mlog(ML_NOTICE, "%s: res %.*s, Skip "
"recovery as it is being freed\n",
- res->lockname.len,
+ dlm->name, res->lockname.len,
res->lockname.name);
} else
dlm_move_lockres_to_recovery_list(dlm,
diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c
index 1d6d1d22c471..e73c833fc2a1 100644
--- a/fs/ocfs2/dlm/dlmthread.c
+++ b/fs/ocfs2/dlm/dlmthread.c
@@ -94,24 +94,26 @@ int __dlm_lockres_unused(struct dlm_lock_resource *res)
{
int bit;
+ assert_spin_locked(&res->spinlock);
+
if (__dlm_lockres_has_locks(res))
return 0;
+ /* Locks are in the process of being created */
+ if (res->inflight_locks)
+ return 0;
+
if (!list_empty(&res->dirty) || res->state & DLM_LOCK_RES_DIRTY)
return 0;
if (res->state & DLM_LOCK_RES_RECOVERING)
return 0;
+ /* Another node has this resource with this node as the master */
bit = find_next_bit(res->refmap, O2NM_MAX_NODES, 0);
if (bit < O2NM_MAX_NODES)
return 0;
- /*
- * since the bit for dlm->node_num is not set, inflight_locks better
- * be zero
- */
- BUG_ON(res->inflight_locks != 0);
return 1;
}
@@ -185,8 +187,6 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
/* clear our bit from the master's refmap, ignore errors */
ret = dlm_drop_lockres_ref(dlm, res);
if (ret < 0) {
- mlog(ML_ERROR, "%s: deref %.*s failed %d\n", dlm->name,
- res->lockname.len, res->lockname.name, ret);
if (!dlm_is_host_down(ret))
BUG();
}
@@ -209,7 +209,7 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm,
BUG();
}
- __dlm_unhash_lockres(res);
+ __dlm_unhash_lockres(dlm, res);
/* lockres is not in the hash now. drop the flag and wake up
* any processes waiting in dlm_get_lock_resource. */
diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index e1ed5e502ff2..81a4cd22f80b 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -1692,7 +1692,7 @@ int ocfs2_open_lock(struct inode *inode)
mlog(0, "inode %llu take PRMODE open lock\n",
(unsigned long long)OCFS2_I(inode)->ip_blkno);
- if (ocfs2_mount_local(osb))
+ if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb))
goto out;
lockres = &OCFS2_I(inode)->ip_open_lockres;
@@ -1718,6 +1718,12 @@ int ocfs2_try_open_lock(struct inode *inode, int write)
(unsigned long long)OCFS2_I(inode)->ip_blkno,
write ? "EXMODE" : "PRMODE");
+ if (ocfs2_is_hard_readonly(osb)) {
+ if (write)
+ status = -EROFS;
+ goto out;
+ }
+
if (ocfs2_mount_local(osb))
goto out;
@@ -2298,7 +2304,7 @@ int ocfs2_inode_lock_full_nested(struct inode *inode,
if (ocfs2_is_hard_readonly(osb)) {
if (ex)
status = -EROFS;
- goto bail;
+ goto getbh;
}
if (ocfs2_mount_local(osb))
@@ -2356,7 +2362,7 @@ local:
mlog_errno(status);
goto bail;
}
-
+getbh:
if (ret_bh) {
status = ocfs2_assign_bh(inode, ret_bh, local_bh);
if (status < 0) {
@@ -2628,8 +2634,11 @@ int ocfs2_dentry_lock(struct dentry *dentry, int ex)
BUG_ON(!dl);
- if (ocfs2_is_hard_readonly(osb))
- return -EROFS;
+ if (ocfs2_is_hard_readonly(osb)) {
+ if (ex)
+ return -EROFS;
+ return 0;
+ }
if (ocfs2_mount_local(osb))
return 0;
@@ -2647,7 +2656,7 @@ void ocfs2_dentry_unlock(struct dentry *dentry, int ex)
struct ocfs2_dentry_lock *dl = dentry->d_fsdata;
struct ocfs2_super *osb = OCFS2_SB(dentry->d_sb);
- if (!ocfs2_mount_local(osb))
+ if (!ocfs2_is_hard_readonly(osb) && !ocfs2_mount_local(osb))
ocfs2_cluster_unlock(osb, &dl->dl_lockres, level);
}
diff --git a/fs/ocfs2/extent_map.c b/fs/ocfs2/extent_map.c
index 23457b491e8c..2f5b92ef0e53 100644
--- a/fs/ocfs2/extent_map.c
+++ b/fs/ocfs2/extent_map.c
@@ -832,6 +832,102 @@ out:
return ret;
}
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin)
+{
+ struct inode *inode = file->f_mapping->host;
+ int ret;
+ unsigned int is_last = 0, is_data = 0;
+ u16 cs_bits = OCFS2_SB(inode->i_sb)->s_clustersize_bits;
+ u32 cpos, cend, clen, hole_size;
+ u64 extoff, extlen;
+ struct buffer_head *di_bh = NULL;
+ struct ocfs2_extent_rec rec;
+
+ BUG_ON(origin != SEEK_DATA && origin != SEEK_HOLE);
+
+ ret = ocfs2_inode_lock(inode, &di_bh, 0);
+ if (ret) {
+ mlog_errno(ret);
+ goto out;
+ }
+
+ down_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+ if (*offset >= inode->i_size) {
+ ret = -ENXIO;
+ goto out_unlock;
+ }
+
+ if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
+ if (origin == SEEK_HOLE)
+ *offset = inode->i_size;
+ goto out_unlock;
+ }
+
+ clen = 0;
+ cpos = *offset >> cs_bits;
+ cend = ocfs2_clusters_for_bytes(inode->i_sb, inode->i_size);
+
+ while (cpos < cend && !is_last) {
+ ret = ocfs2_get_clusters_nocache(inode, di_bh, cpos, &hole_size,
+ &rec, &is_last);
+ if (ret) {
+ mlog_errno(ret);
+ goto out_unlock;
+ }
+
+ extoff = cpos;
+ extoff <<= cs_bits;
+
+ if (rec.e_blkno == 0ULL) {
+ clen = hole_size;
+ is_data = 0;
+ } else {
+ clen = le16_to_cpu(rec.e_leaf_clusters) -
+ (cpos - le32_to_cpu(rec.e_cpos));
+ is_data = (rec.e_flags & OCFS2_EXT_UNWRITTEN) ? 0 : 1;
+ }
+
+ if ((!is_data && origin == SEEK_HOLE) ||
+ (is_data && origin == SEEK_DATA)) {
+ if (extoff > *offset)
+ *offset = extoff;
+ goto out_unlock;
+ }
+
+ if (!is_last)
+ cpos += clen;
+ }
+
+ if (origin == SEEK_HOLE) {
+ extoff = cpos;
+ extoff <<= cs_bits;
+ extlen = clen;
+ extlen <<= cs_bits;
+
+ if ((extoff + extlen) > inode->i_size)
+ extlen = inode->i_size - extoff;
+ extoff += extlen;
+ if (extoff > *offset)
+ *offset = extoff;
+ goto out_unlock;
+ }
+
+ ret = -ENXIO;
+
+out_unlock:
+
+ brelse(di_bh);
+
+ up_read(&OCFS2_I(inode)->ip_alloc_sem);
+
+ ocfs2_inode_unlock(inode, 0);
+out:
+ if (ret && ret != -ENXIO)
+ ret = -ENXIO;
+ return ret;
+}
+
int ocfs2_read_virt_blocks(struct inode *inode, u64 v_block, int nr,
struct buffer_head *bhs[], int flags,
int (*validate)(struct super_block *sb,
diff --git a/fs/ocfs2/extent_map.h b/fs/ocfs2/extent_map.h
index e79d41c2c909..67ea57d2fd59 100644
--- a/fs/ocfs2/extent_map.h
+++ b/fs/ocfs2/extent_map.h
@@ -53,6 +53,8 @@ int ocfs2_extent_map_get_blocks(struct inode *inode, u64 v_blkno, u64 *p_blkno,
int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
u64 map_start, u64 map_len);
+int ocfs2_seek_data_hole_offset(struct file *file, loff_t *offset, int origin);
+
int ocfs2_xattr_get_clusters(struct inode *inode, u32 v_cluster,
u32 *p_cluster, u32 *num_clusters,
struct ocfs2_extent_list *el,
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index de4ea1af041b..6e396683c3d4 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1950,6 +1950,9 @@ static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
if (ret < 0)
mlog_errno(ret);
+ if (file->f_flags & O_SYNC)
+ handle->h_sync = 1;
+
ocfs2_commit_trans(osb, handle);
out_inode_unlock:
@@ -2052,6 +2055,23 @@ out:
return ret;
}
+static void ocfs2_aiodio_wait(struct inode *inode)
+{
+ wait_queue_head_t *wq = ocfs2_ioend_wq(inode);
+
+ wait_event(*wq, (atomic_read(&OCFS2_I(inode)->ip_unaligned_aio) == 0));
+}
+
+static int ocfs2_is_io_unaligned(struct inode *inode, size_t count, loff_t pos)
+{
+ int blockmask = inode->i_sb->s_blocksize - 1;
+ loff_t final_size = pos + count;
+
+ if ((pos & blockmask) || (final_size & blockmask))
+ return 1;
+ return 0;
+}
+
static int ocfs2_prepare_inode_for_refcount(struct inode *inode,
struct file *file,
loff_t pos, size_t count,
@@ -2230,6 +2250,7 @@ static ssize_t ocfs2_file_aio_write(struct kiocb *iocb,
struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
int full_coherency = !(osb->s_mount_opt &
OCFS2_MOUNT_COHERENCY_BUFFERED);
+ int unaligned_dio = 0;
trace_ocfs2_file_aio_write(inode, file, file->f_path.dentry,
(unsigned long long)OCFS2_I(inode)->ip_blkno,
@@ -2297,6 +2318,10 @@ relock:
goto out;
}
+ if (direct_io && !is_sync_kiocb(iocb))
+ unaligned_dio = ocfs2_is_io_unaligned(inode, iocb->ki_left,
+ *ppos);
+
/*
* We can't complete the direct I/O as requested, fall back to
* buffered I/O.
@@ -2311,6 +2336,18 @@ relock:
goto relock;
}
+ if (unaligned_dio) {
+ /*
+ * Wait on previous unaligned aio to complete before
+ * proceeding.
+ */
+ ocfs2_aiodio_wait(inode);
+
+ /* Mark the iocb as needing a decrement in ocfs2_dio_end_io */
+ atomic_inc(&OCFS2_I(inode)->ip_unaligned_aio);
+ ocfs2_iocb_set_unaligned_aio(iocb);
+ }
+
/*
* To later detect whether a journal commit for sync writes is
* necessary, we sample i_size, and cluster count here.
@@ -2382,8 +2419,12 @@ out_dio:
if ((ret == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) {
rw_level = -1;
have_alloc_sem = 0;
+ unaligned_dio = 0;
}
+ if (unaligned_dio)
+ atomic_dec(&OCFS2_I(inode)->ip_unaligned_aio);
+
out:
if (rw_level != -1)
ocfs2_rw_unlock(inode, rw_level);
@@ -2591,6 +2632,57 @@ bail:
return ret;
}
+/* Refer generic_file_llseek_unlocked() */
+static loff_t ocfs2_file_llseek(struct file *file, loff_t offset, int origin)
+{
+ struct inode *inode = file->f_mapping->host;
+ int ret = 0;
+
+ mutex_lock(&inode->i_mutex);
+
+ switch (origin) {
+ case SEEK_SET:
+ break;
+ case SEEK_END:
+ offset += inode->i_size;
+ break;
+ case SEEK_CUR:
+ if (offset == 0) {
+ offset = file->f_pos;
+ goto out;
+ }
+ offset += file->f_pos;
+ break;
+ case SEEK_DATA:
+ case SEEK_HOLE:
+ ret = ocfs2_seek_data_hole_offset(file, &offset, origin);
+ if (ret)
+ goto out;
+ break;
+ default:
+ ret = -EINVAL;
+ goto out;
+ }
+
+ if (offset < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET))
+ ret = -EINVAL;
+ if (!ret && offset > inode->i_sb->s_maxbytes)
+ ret = -EINVAL;
+ if (ret)
+ goto out;
+
+ if (offset != file->f_pos) {
+ file->f_pos = offset;
+ file->f_version = 0;
+ }
+
+out:
+ mutex_unlock(&inode->i_mutex);
+ if (ret)
+ return ret;
+ return offset;
+}
+
const struct inode_operations ocfs2_file_iops = {
.setattr = ocfs2_setattr,
.getattr = ocfs2_getattr,
@@ -2615,7 +2707,7 @@ const struct inode_operations ocfs2_special_file_iops = {
* ocfs2_fops_no_plocks and ocfs2_dops_no_plocks!
*/
const struct file_operations ocfs2_fops = {
- .llseek = generic_file_llseek,
+ .llseek = ocfs2_file_llseek,
.read = do_sync_read,
.write = do_sync_write,
.mmap = ocfs2_mmap,
@@ -2663,7 +2755,7 @@ const struct file_operations ocfs2_dops = {
* the cluster.
*/
const struct file_operations ocfs2_fops_no_plocks = {
- .llseek = generic_file_llseek,
+ .llseek = ocfs2_file_llseek,
.read = do_sync_read,
.write = do_sync_write,
.mmap = ocfs2_mmap,
diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c
index a22d2c098890..17454a904d7b 100644
--- a/fs/ocfs2/inode.c
+++ b/fs/ocfs2/inode.c
@@ -951,7 +951,7 @@ static void ocfs2_cleanup_delete_inode(struct inode *inode,
trace_ocfs2_cleanup_delete_inode(
(unsigned long long)OCFS2_I(inode)->ip_blkno, sync_data);
if (sync_data)
- write_inode_now(inode, 1);
+ filemap_write_and_wait(inode->i_mapping);
truncate_inode_pages(&inode->i_data, 0);
}
diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h
index 1c508b149b3a..88924a3133fa 100644
--- a/fs/ocfs2/inode.h
+++ b/fs/ocfs2/inode.h
@@ -43,6 +43,9 @@ struct ocfs2_inode_info
/* protects extended attribute changes on this inode */
struct rw_semaphore ip_xattr_sem;
+ /* Number of outstanding AIO's which are not page aligned */
+ atomic_t ip_unaligned_aio;
+
/* These fields are protected by ip_lock */
spinlock_t ip_lock;
u32 ip_open_count;
diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c
index bc91072b7219..726ff265b296 100644
--- a/fs/ocfs2/ioctl.c
+++ b/fs/ocfs2/ioctl.c
@@ -122,7 +122,7 @@ static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
if ((oldflags & OCFS2_IMMUTABLE_FL) || ((flags ^ oldflags) &
(OCFS2_APPEND_FL | OCFS2_IMMUTABLE_FL))) {
if (!capable(CAP_LINUX_IMMUTABLE))
- goto bail_unlock;
+ goto bail_commit;
}
ocfs2_inode->ip_attr = flags;
@@ -132,6 +132,7 @@ static int ocfs2_set_inode_attr(struct inode *inode, unsigned flags,
if (status < 0)
mlog_errno(status);
+bail_commit:
ocfs2_commit_trans(osb, handle);
bail_unlock:
ocfs2_inode_unlock(inode, 1);
@@ -381,7 +382,7 @@ int ocfs2_info_handle_freeinode(struct inode *inode,
if (!oifi) {
status = -ENOMEM;
mlog_errno(status);
- goto bail;
+ goto out_err;
}
if (o2info_from_user(*oifi, req))
@@ -431,7 +432,7 @@ bail:
o2info_set_request_error(&oifi->ifi_req, req);
kfree(oifi);
-
+out_err:
return status;
}
@@ -666,7 +667,7 @@ int ocfs2_info_handle_freefrag(struct inode *inode,
if (!oiff) {
status = -ENOMEM;
mlog_errno(status);
- goto bail;
+ goto out_err;
}
if (o2info_from_user(*oiff, req))
@@ -716,7 +717,7 @@ bail:
o2info_set_request_error(&oiff->iff_req, req);
kfree(oiff);
-
+out_err:
return status;
}
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 295d56454e8b..0a42ae96dca7 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -1544,9 +1544,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
/* we need to run complete recovery for offline orphan slots */
ocfs2_replay_map_set_state(osb, REPLAY_NEEDED);
- mlog(ML_NOTICE, "Recovering node %d from slot %d on device (%u,%u)\n",
- node_num, slot_num,
- MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev));
+ printk(KERN_NOTICE "ocfs2: Begin replay journal (node %d, slot %d) on "\
+ "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+ MINOR(osb->sb->s_dev));
OCFS2_I(inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
@@ -1601,6 +1601,9 @@ static int ocfs2_replay_journal(struct ocfs2_super *osb,
jbd2_journal_destroy(journal);
+ printk(KERN_NOTICE "ocfs2: End replay journal (node %d, slot %d) on "\
+ "device (%u,%u)\n", node_num, slot_num, MAJOR(osb->sb->s_dev),
+ MINOR(osb->sb->s_dev));
done:
/* drop the lock on this nodes journal */
if (got_lock)
@@ -1808,6 +1811,20 @@ static inline unsigned long ocfs2_orphan_scan_timeout(void)
* every slot, queuing a recovery of the slot on the ocfs2_wq thread. This
* is done to catch any orphans that are left over in orphan directories.
*
+ * It scans all slots, even ones that are in use. It does so to handle the
+ * case described below:
+ *
+ * Node 1 has an inode it was using. The dentry went away due to memory
+ * pressure. Node 1 closes the inode, but it's on the free list. The node
+ * has the open lock.
+ * Node 2 unlinks the inode. It grabs the dentry lock to notify others,
+ * but node 1 has no dentry and doesn't get the message. It trylocks the
+ * open lock, sees that another node has a PR, and does nothing.
+ * Later node 2 runs its orphan dir. It igets the inode, trylocks the
+ * open lock, sees the PR still, and does nothing.
+ * Basically, we have to trigger an orphan iput on node 1. The only way
+ * for this to happen is if node 1 runs node 2's orphan dir.
+ *
* ocfs2_queue_orphan_scan gets called every ORPHAN_SCAN_SCHEDULE_TIMEOUT
* seconds. It gets an EX lock on os_lockres and checks sequence number
* stored in LVB. If the sequence number has changed, it means some other
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 68cf2f6d3c6a..a3385b63ff5e 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -441,10 +441,11 @@ static inline int ocfs2_mknod_credits(struct super_block *sb, int is_dir,
#define OCFS2_SIMPLE_DIR_EXTEND_CREDITS (2)
/* file update (nlink, etc) + directory mtime/ctime + dir entry block + quota
- * update on dir + index leaf + dx root update for free list */
+ * update on dir + index leaf + dx root update for free list +
+ * previous dirblock update in the free list */
static inline int ocfs2_link_credits(struct super_block *sb)
{
- return 2*OCFS2_INODE_UPDATE_CREDITS + 3 +
+ return 2*OCFS2_INODE_UPDATE_CREDITS + 4 +
ocfs2_quota_trans_credits(sb);
}
diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c
index 3e9393ca39eb..9cd41083e991 100644
--- a/fs/ocfs2/mmap.c
+++ b/fs/ocfs2/mmap.c
@@ -61,7 +61,7 @@ static int ocfs2_fault(struct vm_area_struct *area, struct vm_fault *vmf)
static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
struct page *page)
{
- int ret;
+ int ret = VM_FAULT_NOPAGE;
struct inode *inode = file->f_path.dentry->d_inode;
struct address_space *mapping = inode->i_mapping;
loff_t pos = page_offset(page);
@@ -71,32 +71,25 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
void *fsdata;
loff_t size = i_size_read(inode);
- /*
- * Another node might have truncated while we were waiting on
- * cluster locks.
- * We don't check size == 0 before the shift. This is borrowed
- * from do_generic_file_read.
- */
last_index = (size - 1) >> PAGE_CACHE_SHIFT;
- if (unlikely(!size || page->index > last_index)) {
- ret = -EINVAL;
- goto out;
- }
/*
- * The i_size check above doesn't catch the case where nodes
- * truncated and then re-extended the file. We'll re-check the
- * page mapping after taking the page lock inside of
- * ocfs2_write_begin_nolock().
+ * There are cases that lead to the page no longer bebongs to the
+ * mapping.
+ * 1) pagecache truncates locally due to memory pressure.
+ * 2) pagecache truncates when another is taking EX lock against
+ * inode lock. see ocfs2_data_convert_worker.
+ *
+ * The i_size check doesn't catch the case where nodes truncated and
+ * then re-extended the file. We'll re-check the page mapping after
+ * taking the page lock inside of ocfs2_write_begin_nolock().
+ *
+ * Let VM retry with these cases.
*/
- if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
- /*
- * the page has been umapped in ocfs2_data_downconvert_worker.
- * So return 0 here and let VFS retry.
- */
- ret = 0;
+ if ((page->mapping != inode->i_mapping) ||
+ (!PageUptodate(page)) ||
+ (page_offset(page) >= size))
goto out;
- }
/*
* Call ocfs2_write_begin() and ocfs2_write_end() to take
@@ -116,17 +109,21 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh,
if (ret) {
if (ret != -ENOSPC)
mlog_errno(ret);
+ if (ret == -ENOMEM)
+ ret = VM_FAULT_OOM;
+ else
+ ret = VM_FAULT_SIGBUS;
goto out;
}
- ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
- fsdata);
- if (ret < 0) {
- mlog_errno(ret);
+ if (!locked_page) {
+ ret = VM_FAULT_NOPAGE;
goto out;
}
+ ret = ocfs2_write_end_nolock(mapping, pos, len, len, locked_page,
+ fsdata);
BUG_ON(ret != len);
- ret = 0;
+ ret = VM_FAULT_LOCKED;
out:
return ret;
}
@@ -168,8 +165,6 @@ static int ocfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
out:
ocfs2_unblock_signals(&oldset);
- if (ret)
- ret = VM_FAULT_SIGBUS;
return ret;
}
diff --git a/fs/ocfs2/move_extents.c b/fs/ocfs2/move_extents.c
index d53cb706f14c..184c76b8c293 100644
--- a/fs/ocfs2/move_extents.c
+++ b/fs/ocfs2/move_extents.c
@@ -745,7 +745,7 @@ static int ocfs2_move_extent(struct ocfs2_move_extents_context *context,
*/
ocfs2_probe_alloc_group(inode, gd_bh, &goal_bit, len, move_max_hop,
new_phys_cpos);
- if (!new_phys_cpos) {
+ if (!*new_phys_cpos) {
ret = -ENOSPC;
goto out_commit;
}
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index 409285854f64..d355e6e36b36 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -836,18 +836,65 @@ static inline unsigned int ocfs2_clusters_to_megabytes(struct super_block *sb,
static inline void _ocfs2_set_bit(unsigned int bit, unsigned long *bitmap)
{
- __test_and_set_bit_le(bit, bitmap);
+ __set_bit_le(bit, bitmap);
}
#define ocfs2_set_bit(bit, addr) _ocfs2_set_bit((bit), (unsigned long *)(addr))
static inline void _ocfs2_clear_bit(unsigned int bit, unsigned long *bitmap)
{
- __test_and_clear_bit_le(bit, bitmap);
+ __clear_bit_le(bit, bitmap);
}
#define ocfs2_clear_bit(bit, addr) _ocfs2_clear_bit((bit), (unsigned long *)(addr))
#define ocfs2_test_bit test_bit_le
#define ocfs2_find_next_zero_bit find_next_zero_bit_le
#define ocfs2_find_next_bit find_next_bit_le
+
+static inline void *correct_addr_and_bit_unaligned(int *bit, void *addr)
+{
+#if BITS_PER_LONG == 64
+ *bit += ((unsigned long) addr & 7UL) << 3;
+ addr = (void *) ((unsigned long) addr & ~7UL);
+#elif BITS_PER_LONG == 32
+ *bit += ((unsigned long) addr & 3UL) << 3;
+ addr = (void *) ((unsigned long) addr & ~3UL);
+#else
+#error "how many bits you are?!"
+#endif
+ return addr;
+}
+
+static inline void ocfs2_set_bit_unaligned(int bit, void *bitmap)
+{
+ bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+ ocfs2_set_bit(bit, bitmap);
+}
+
+static inline void ocfs2_clear_bit_unaligned(int bit, void *bitmap)
+{
+ bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+ ocfs2_clear_bit(bit, bitmap);
+}
+
+static inline int ocfs2_test_bit_unaligned(int bit, void *bitmap)
+{
+ bitmap = correct_addr_and_bit_unaligned(&bit, bitmap);
+ return ocfs2_test_bit(bit, bitmap);
+}
+
+static inline int ocfs2_find_next_zero_bit_unaligned(void *bitmap, int max,
+ int start)
+{
+ int fix = 0, ret, tmpmax;
+ bitmap = correct_addr_and_bit_unaligned(&fix, bitmap);
+ tmpmax = max + fix;
+ start += fix;
+
+ ret = ocfs2_find_next_zero_bit(bitmap, tmpmax, start) - fix;
+ if (ret > max)
+ return max;
+ return ret;
+}
+
#endif /* OCFS2_H */
diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c
index dc8007fc9247..f100bf70a906 100644
--- a/fs/ocfs2/quota_local.c
+++ b/fs/ocfs2/quota_local.c
@@ -404,7 +404,9 @@ struct ocfs2_quota_recovery *ocfs2_begin_quota_recovery(
int status = 0;
struct ocfs2_quota_recovery *rec;
- mlog(ML_NOTICE, "Beginning quota recovery in slot %u\n", slot_num);
+ printk(KERN_NOTICE "ocfs2: Beginning quota recovery on device (%s) for "
+ "slot %u\n", osb->dev_str, slot_num);
+
rec = ocfs2_alloc_quota_recovery();
if (!rec)
return ERR_PTR(-ENOMEM);
@@ -549,8 +551,8 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode,
goto out_commit;
}
lock_buffer(qbh);
- WARN_ON(!ocfs2_test_bit(bit, dchunk->dqc_bitmap));
- ocfs2_clear_bit(bit, dchunk->dqc_bitmap);
+ WARN_ON(!ocfs2_test_bit_unaligned(bit, dchunk->dqc_bitmap));
+ ocfs2_clear_bit_unaligned(bit, dchunk->dqc_bitmap);
le32_add_cpu(&dchunk->dqc_free, 1);
unlock_buffer(qbh);
ocfs2_journal_dirty(handle, qbh);
@@ -596,7 +598,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
struct inode *lqinode;
unsigned int flags;
- mlog(ML_NOTICE, "Finishing quota recovery in slot %u\n", slot_num);
+ printk(KERN_NOTICE "ocfs2: Finishing quota recovery on device (%s) for "
+ "slot %u\n", osb->dev_str, slot_num);
+
mutex_lock(&sb_dqopt(sb)->dqonoff_mutex);
for (type = 0; type < MAXQUOTAS; type++) {
if (list_empty(&(rec->r_list[type])))
@@ -612,8 +616,9 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb,
/* Someone else is holding the lock? Then he must be
* doing the recovery. Just skip the file... */
if (status == -EAGAIN) {
- mlog(ML_NOTICE, "skipping quota recovery for slot %d "
- "because quota file is locked.\n", slot_num);
+ printk(KERN_NOTICE "ocfs2: Skipping quota recovery on "
+ "device (%s) for slot %d because quota file is "
+ "locked.\n", osb->dev_str, slot_num);
status = 0;
goto out_put;
} else if (status < 0) {
@@ -944,7 +949,7 @@ static struct ocfs2_quota_chunk *ocfs2_find_free_entry(struct super_block *sb,
* ol_quota_entries_per_block(sb);
}
- found = ocfs2_find_next_zero_bit(dchunk->dqc_bitmap, len, 0);
+ found = ocfs2_find_next_zero_bit_unaligned(dchunk->dqc_bitmap, len, 0);
/* We failed? */
if (found == len) {
mlog(ML_ERROR, "Did not find empty entry in chunk %d with %u"
@@ -1208,7 +1213,7 @@ static void olq_alloc_dquot(struct buffer_head *bh, void *private)
struct ocfs2_local_disk_chunk *dchunk;
dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data;
- ocfs2_set_bit(*offset, dchunk->dqc_bitmap);
+ ocfs2_set_bit_unaligned(*offset, dchunk->dqc_bitmap);
le32_add_cpu(&dchunk->dqc_free, -1);
}
@@ -1289,7 +1294,7 @@ int ocfs2_local_release_dquot(handle_t *handle, struct dquot *dquot)
(od->dq_chunk->qc_headerbh->b_data);
/* Mark structure as freed */
lock_buffer(od->dq_chunk->qc_headerbh);
- ocfs2_clear_bit(offset, dchunk->dqc_bitmap);
+ ocfs2_clear_bit_unaligned(offset, dchunk->dqc_bitmap);
le32_add_cpu(&dchunk->dqc_free, 1);
unlock_buffer(od->dq_chunk->qc_headerbh);
ocfs2_journal_dirty(handle, od->dq_chunk->qc_headerbh);
diff --git a/fs/ocfs2/slot_map.c b/fs/ocfs2/slot_map.c
index 26fc0014d509..1424c151cccc 100644
--- a/fs/ocfs2/slot_map.c
+++ b/fs/ocfs2/slot_map.c
@@ -493,8 +493,8 @@ int ocfs2_find_slot(struct ocfs2_super *osb)
goto bail;
}
} else
- mlog(ML_NOTICE, "slot %d is already allocated to this node!\n",
- slot);
+ printk(KERN_INFO "ocfs2: Slot %d on device (%s) was already "
+ "allocated to this node!\n", slot, osb->dev_str);
ocfs2_set_slot(si, slot, osb->node_num);
osb->slot_num = slot;
diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c
index 19965b00c43c..94368017edb3 100644
--- a/fs/ocfs2/stack_o2cb.c
+++ b/fs/ocfs2/stack_o2cb.c
@@ -28,6 +28,7 @@
#include "cluster/masklog.h"
#include "cluster/nodemanager.h"
#include "cluster/heartbeat.h"
+#include "cluster/tcp.h"
#include "stackglue.h"
@@ -256,6 +257,61 @@ static void o2cb_dump_lksb(struct ocfs2_dlm_lksb *lksb)
}
/*
+ * Check if this node is heartbeating and is connected to all other
+ * heartbeating nodes.
+ */
+static int o2cb_cluster_check(void)
+{
+ u8 node_num;
+ int i;
+ unsigned long hbmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+ unsigned long netmap[BITS_TO_LONGS(O2NM_MAX_NODES)];
+
+ node_num = o2nm_this_node();
+ if (node_num == O2NM_MAX_NODES) {
+ printk(KERN_ERR "o2cb: This node has not been configured.\n");
+ return -EINVAL;
+ }
+
+ /*
+ * o2dlm expects o2net sockets to be created. If not, then
+ * dlm_join_domain() fails with a stack of errors which are both cryptic
+ * and incomplete. The idea here is to detect upfront whether we have
+ * managed to connect to all nodes or not. If not, then list the nodes
+ * to allow the user to check the configuration (incorrect IP, firewall,
+ * etc.) Yes, this is racy. But its not the end of the world.
+ */
+#define O2CB_MAP_STABILIZE_COUNT 60
+ for (i = 0; i < O2CB_MAP_STABILIZE_COUNT; ++i) {
+ o2hb_fill_node_map(hbmap, sizeof(hbmap));
+ if (!test_bit(node_num, hbmap)) {
+ printk(KERN_ERR "o2cb: %s heartbeat has not been "
+ "started.\n", (o2hb_global_heartbeat_active() ?
+ "Global" : "Local"));
+ return -EINVAL;
+ }
+ o2net_fill_node_map(netmap, sizeof(netmap));
+ /* Force set the current node to allow easy compare */
+ set_bit(node_num, netmap);
+ if (!memcmp(hbmap, netmap, sizeof(hbmap)))
+ return 0;
+ if (i < O2CB_MAP_STABILIZE_COUNT)
+ msleep(1000);
+ }
+
+ printk(KERN_ERR "o2cb: This node could not connect to nodes:");
+ i = -1;
+ while ((i = find_next_bit(hbmap, O2NM_MAX_NODES,
+ i + 1)) < O2NM_MAX_NODES) {
+ if (!test_bit(i, netmap))
+ printk(" %u", i);
+ }
+ printk(".\n");
+
+ return -ENOTCONN;
+}
+
+/*
* Called from the dlm when it's about to evict a node. This is how the
* classic stack signals node death.
*/
@@ -263,8 +319,8 @@ static void o2dlm_eviction_cb(int node_num, void *data)
{
struct ocfs2_cluster_connection *conn = data;
- mlog(ML_NOTICE, "o2dlm has evicted node %d from group %.*s\n",
- node_num, conn->cc_namelen, conn->cc_name);
+ printk(KERN_NOTICE "o2cb: o2dlm has evicted node %d from domain %.*s\n",
+ node_num, conn->cc_namelen, conn->cc_name);
conn->cc_recovery_handler(node_num, conn->cc_recovery_data);
}
@@ -280,12 +336,11 @@ static int o2cb_cluster_connect(struct ocfs2_cluster_connection *conn)
BUG_ON(conn == NULL);
BUG_ON(conn->cc_proto == NULL);
- /* for now we only have one cluster/node, make sure we see it
- * in the heartbeat universe */
- if (!o2hb_check_local_node_heartbeating()) {
- if (o2hb_global_heartbeat_active())
- mlog(ML_ERROR, "Global heartbeat not started\n");
- rc = -EINVAL;
+ /* Ensure cluster stack is up and all nodes are connected */
+ rc = o2cb_cluster_check();
+ if (rc) {
+ printk(KERN_ERR "o2cb: Cluster check failed. Fix errors "
+ "before retrying.\n");
goto out;
}
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 56f61027236b..4994f8b0e604 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -54,6 +54,7 @@
#include "ocfs1_fs_compat.h"
#include "alloc.h"
+#include "aops.h"
#include "blockcheck.h"
#include "dlmglue.h"
#include "export.h"
@@ -1107,9 +1108,9 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent)
ocfs2_set_ro_flag(osb, 1);
- printk(KERN_NOTICE "Readonly device detected. No cluster "
- "services will be utilized for this mount. Recovery "
- "will be skipped.\n");
+ printk(KERN_NOTICE "ocfs2: Readonly device (%s) detected. "
+ "Cluster services will not be used for this mount. "
+ "Recovery will be skipped.\n", osb->dev_str);
}
if (!ocfs2_is_hard_readonly(osb)) {
@@ -1616,12 +1617,17 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt)
return 0;
}
+wait_queue_head_t ocfs2__ioend_wq[OCFS2_IOEND_WQ_HASH_SZ];
+
static int __init ocfs2_init(void)
{
- int status;
+ int status, i;
ocfs2_print_version();
+ for (i = 0; i < OCFS2_IOEND_WQ_HASH_SZ; i++)
+ init_waitqueue_head(&ocfs2__ioend_wq[i]);
+
status = init_ocfs2_uptodate_cache();
if (status < 0) {
mlog_errno(status);
@@ -1760,7 +1766,7 @@ static void ocfs2_inode_init_once(void *data)
ocfs2_extent_map_init(&oi->vfs_inode);
INIT_LIST_HEAD(&oi->ip_io_markers);
oi->ip_dir_start_lookup = 0;
-
+ atomic_set(&oi->ip_unaligned_aio, 0);
init_rwsem(&oi->ip_alloc_sem);
init_rwsem(&oi->ip_xattr_sem);
mutex_init(&oi->ip_io_mutex);
@@ -1974,7 +1980,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err)
* If we failed before we got a uuid_str yet, we can't stop
* heartbeat. Otherwise, do it.
*/
- if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str)
+ if (!mnt_err && !ocfs2_mount_local(osb) && osb->uuid_str &&
+ !ocfs2_is_hard_readonly(osb))
hangup_needed = 1;
if (osb->cconn)
@@ -2353,7 +2360,7 @@ static int ocfs2_initialize_super(struct super_block *sb,
mlog_errno(status);
goto bail;
}
- cleancache_init_shared_fs((char *)&uuid_net_key, sb);
+ cleancache_init_shared_fs((char *)&di->id2.i_super.s_uuid, sb);
bail:
return status;
@@ -2462,8 +2469,8 @@ static int ocfs2_check_volume(struct ocfs2_super *osb)
goto finally;
}
} else {
- mlog(ML_NOTICE, "File system was not unmounted cleanly, "
- "recovering volume.\n");
+ printk(KERN_NOTICE "ocfs2: File system on device (%s) was not "
+ "unmounted cleanly, recovering it.\n", osb->dev_str);
}
local = ocfs2_mount_local(osb);
diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c
index 194fb22ef79d..aa9e8777b09a 100644
--- a/fs/ocfs2/xattr.c
+++ b/fs/ocfs2/xattr.c
@@ -2376,16 +2376,18 @@ static int ocfs2_remove_value_outside(struct inode*inode,
}
ret = ocfs2_xattr_value_truncate(inode, vb, 0, &ctxt);
- if (ret < 0) {
- mlog_errno(ret);
- break;
- }
ocfs2_commit_trans(osb, ctxt.handle);
if (ctxt.meta_ac) {
ocfs2_free_alloc_context(ctxt.meta_ac);
ctxt.meta_ac = NULL;
}
+
+ if (ret < 0) {
+ mlog_errno(ret);
+ break;
+ }
+
}
if (ctxt.meta_ac)
diff --git a/fs/proc/meminfo.c b/fs/proc/meminfo.c
index 586174168e2a..80e4645f7990 100644
--- a/fs/proc/meminfo.c
+++ b/fs/proc/meminfo.c
@@ -131,12 +131,13 @@ static int meminfo_proc_show(struct seq_file *m, void *v)
K(i.freeswap),
K(global_page_state(NR_FILE_DIRTY)),
K(global_page_state(NR_WRITEBACK)),
- K(global_page_state(NR_ANON_PAGES)
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+ K(global_page_state(NR_ANON_PAGES)
+ global_page_state(NR_ANON_TRANSPARENT_HUGEPAGES) *
- HPAGE_PMD_NR
+ HPAGE_PMD_NR),
+#else
+ K(global_page_state(NR_ANON_PAGES)),
#endif
- ),
K(global_page_state(NR_FILE_MAPPED)),
K(global_page_state(NR_SHMEM)),
K(global_page_state(NR_SLAB_RECLAIMABLE) +
diff --git a/fs/proc/stat.c b/fs/proc/stat.c
index 42b274da92c3..2a30d67dd6b8 100644
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@ -32,7 +32,7 @@ static cputime64_t get_idle_time(int cpu)
idle = kstat_cpu(cpu).cpustat.idle;
idle = cputime64_add(idle, arch_idle_time(cpu));
} else
- idle = usecs_to_cputime(idle_time);
+ idle = nsecs_to_jiffies64(1000 * idle_time);
return idle;
}
@@ -46,7 +46,7 @@ static cputime64_t get_iowait_time(int cpu)
/* !NO_HZ so we can rely on cpustat.iowait */
iowait = kstat_cpu(cpu).cpustat.iowait;
else
- iowait = usecs_to_cputime(iowait_time);
+ iowait = nsecs_to_jiffies64(1000 * iowait_time);
return iowait;
}
diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c
index 2bd620f0d796..57bbf9078ac8 100644
--- a/fs/pstore/platform.c
+++ b/fs/pstore/platform.c
@@ -167,6 +167,7 @@ int pstore_register(struct pstore_info *psi)
}
psinfo = psi;
+ mutex_init(&psinfo->read_mutex);
spin_unlock(&pstore_lock);
if (owner && !try_module_get(owner)) {
@@ -195,30 +196,32 @@ EXPORT_SYMBOL_GPL(pstore_register);
void pstore_get_records(int quiet)
{
struct pstore_info *psi = psinfo;
+ char *buf = NULL;
ssize_t size;
u64 id;
enum pstore_type_id type;
struct timespec time;
int failed = 0, rc;
- unsigned long flags;
if (!psi)
return;
- spin_lock_irqsave(&psinfo->buf_lock, flags);
+ mutex_lock(&psi->read_mutex);
rc = psi->open(psi);
if (rc)
goto out;
- while ((size = psi->read(&id, &type, &time, psi)) > 0) {
- rc = pstore_mkfile(type, psi->name, id, psi->buf, (size_t)size,
+ while ((size = psi->read(&id, &type, &time, &buf, psi)) > 0) {
+ rc = pstore_mkfile(type, psi->name, id, buf, (size_t)size,
time, psi);
+ kfree(buf);
+ buf = NULL;
if (rc && (rc != -EEXIST || !quiet))
failed++;
}
psi->close(psi);
out:
- spin_unlock_irqrestore(&psinfo->buf_lock, flags);
+ mutex_unlock(&psi->read_mutex);
if (failed)
printk(KERN_WARNING "pstore: failed to load %d record(s) from '%s'\n",
diff --git a/fs/seq_file.c b/fs/seq_file.c
index 05d6b0e78c95..dba43c3ea3af 100644
--- a/fs/seq_file.c
+++ b/fs/seq_file.c
@@ -449,8 +449,6 @@ EXPORT_SYMBOL(seq_path);
/*
* Same as seq_path, but relative to supplied root.
- *
- * root may be changed, see __d_path().
*/
int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
char *esc)
@@ -463,6 +461,8 @@ int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
char *p;
p = __d_path(path, root, buf, size);
+ if (!p)
+ return SEQ_SKIP;
res = PTR_ERR(p);
if (!IS_ERR(p)) {
char *end = mangle_path(buf, p, esc);
@@ -474,7 +474,7 @@ int seq_path_root(struct seq_file *m, struct path *path, struct path *root,
}
seq_commit(m, res);
- return res < 0 ? res : 0;
+ return res < 0 && res != -ENAMETOOLONG ? res : 0;
}
/*
diff --git a/fs/xfs/xfs_acl.c b/fs/xfs/xfs_acl.c
index b6c4b3795c4a..76e4266d2e7e 100644
--- a/fs/xfs/xfs_acl.c
+++ b/fs/xfs/xfs_acl.c
@@ -42,6 +42,8 @@ xfs_acl_from_disk(struct xfs_acl *aclp)
int count, i;
count = be32_to_cpu(aclp->acl_cnt);
+ if (count > XFS_ACL_MAX_ENTRIES)
+ return ERR_PTR(-EFSCORRUPTED);
acl = posix_acl_alloc(count, GFP_KERNEL);
if (!acl)
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index d4906e7c9787..c1b55e596551 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -110,6 +110,7 @@ xfs_attr_namesp_match(int arg_flags, int ondisk_flags)
/*
* Query whether the requested number of additional bytes of extended
* attribute space will be able to fit inline.
+ *
* Returns zero if not, else the di_forkoff fork offset to be used in the
* literal area for attribute data once the new bytes have been added.
*
@@ -122,7 +123,7 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
int offset;
int minforkoff; /* lower limit on valid forkoff locations */
int maxforkoff; /* upper limit on valid forkoff locations */
- int dsize;
+ int dsize;
xfs_mount_t *mp = dp->i_mount;
offset = (XFS_LITINO(mp) - bytes) >> 3; /* rounded down */
@@ -136,47 +137,60 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
return (offset >= minforkoff) ? minforkoff : 0;
}
- if (!(mp->m_flags & XFS_MOUNT_ATTR2)) {
- if (bytes <= XFS_IFORK_ASIZE(dp))
- return dp->i_d.di_forkoff;
+ /*
+ * If the requested numbers of bytes is smaller or equal to the
+ * current attribute fork size we can always proceed.
+ *
+ * Note that if_bytes in the data fork might actually be larger than
+ * the current data fork size is due to delalloc extents. In that
+ * case either the extent count will go down when they are converted
+ * to real extents, or the delalloc conversion will take care of the
+ * literal area rebalancing.
+ */
+ if (bytes <= XFS_IFORK_ASIZE(dp))
+ return dp->i_d.di_forkoff;
+
+ /*
+ * For attr2 we can try to move the forkoff if there is space in the
+ * literal area, but for the old format we are done if there is no
+ * space in the fixed attribute fork.
+ */
+ if (!(mp->m_flags & XFS_MOUNT_ATTR2))
return 0;
- }
dsize = dp->i_df.if_bytes;
-
+
switch (dp->i_d.di_format) {
case XFS_DINODE_FMT_EXTENTS:
- /*
+ /*
* If there is no attr fork and the data fork is extents,
- * determine if creating the default attr fork will result
- * in the extents form migrating to btree. If so, the
- * minimum offset only needs to be the space required for
+ * determine if creating the default attr fork will result
+ * in the extents form migrating to btree. If so, the
+ * minimum offset only needs to be the space required for
* the btree root.
- */
+ */
if (!dp->i_d.di_forkoff && dp->i_df.if_bytes >
xfs_default_attroffset(dp))
dsize = XFS_BMDR_SPACE_CALC(MINDBTPTRS);
break;
-
case XFS_DINODE_FMT_BTREE:
/*
- * If have data btree then keep forkoff if we have one,
- * otherwise we are adding a new attr, so then we set
- * minforkoff to where the btree root can finish so we have
+ * If we have a data btree then keep forkoff if we have one,
+ * otherwise we are adding a new attr, so then we set
+ * minforkoff to where the btree root can finish so we have
* plenty of room for attrs
*/
if (dp->i_d.di_forkoff) {
- if (offset < dp->i_d.di_forkoff)
+ if (offset < dp->i_d.di_forkoff)
return 0;
- else
- return dp->i_d.di_forkoff;
- } else
- dsize = XFS_BMAP_BROOT_SPACE(dp->i_df.if_broot);
+ return dp->i_d.di_forkoff;
+ }
+ dsize = XFS_BMAP_BROOT_SPACE(dp->i_df.if_broot);
break;
}
-
- /*
- * A data fork btree root must have space for at least
+
+ /*
+ * A data fork btree root must have space for at least
* MINDBTPTRS key/ptr pairs if the data fork is small or empty.
*/
minforkoff = MAX(dsize, XFS_BMDR_SPACE_CALC(MINDBTPTRS));
@@ -186,10 +200,10 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
maxforkoff = XFS_LITINO(mp) - XFS_BMDR_SPACE_CALC(MINABTPTRS);
maxforkoff = maxforkoff >> 3; /* rounded down */
- if (offset >= minforkoff && offset < maxforkoff)
- return offset;
if (offset >= maxforkoff)
return maxforkoff;
+ if (offset >= minforkoff)
+ return offset;
return 0;
}
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index c68baeb0974a..d0ab78837057 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -2383,6 +2383,8 @@ xfs_bmap_btalloc(
int tryagain;
int error;
+ ASSERT(ap->length);
+
mp = ap->ip->i_mount;
align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
if (unlikely(align)) {
@@ -4629,6 +4631,8 @@ xfs_bmapi_allocate(
int error;
int rt;
+ ASSERT(bma->length > 0);
+
rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(bma->ip);
/*
@@ -4849,6 +4853,7 @@ xfs_bmapi_write(
ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
ASSERT(!(flags & XFS_BMAPI_IGSTATE));
ASSERT(tp != NULL);
+ ASSERT(len > 0);
whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
XFS_ATTR_FORK : XFS_DATA_FORK;
@@ -4918,9 +4923,22 @@ xfs_bmapi_write(
bma.eof = eof;
bma.conv = !!(flags & XFS_BMAPI_CONVERT);
bma.wasdel = wasdelay;
- bma.length = len;
bma.offset = bno;
+ /*
+ * There's a 32/64 bit type mismatch between the
+ * allocation length request (which can be 64 bits in
+ * length) and the bma length request, which is
+ * xfs_extlen_t and therefore 32 bits. Hence we have to
+ * check for 32-bit overflows and handle them here.
+ */
+ if (len > (xfs_filblks_t)MAXEXTLEN)
+ bma.length = MAXEXTLEN;
+ else
+ bma.length = len;
+
+ ASSERT(len > 0);
+ ASSERT(bma.length > 0);
error = xfs_bmapi_allocate(&bma, flags);
if (error)
goto error0;
diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c
index da108977b21f..558910f5e3c0 100644
--- a/fs/xfs/xfs_export.c
+++ b/fs/xfs/xfs_export.c
@@ -98,22 +98,22 @@ xfs_fs_encode_fh(
switch (fileid_type) {
case FILEID_INO32_GEN_PARENT:
spin_lock(&dentry->d_lock);
- fid->i32.parent_ino = dentry->d_parent->d_inode->i_ino;
+ fid->i32.parent_ino = XFS_I(dentry->d_parent->d_inode)->i_ino;
fid->i32.parent_gen = dentry->d_parent->d_inode->i_generation;
spin_unlock(&dentry->d_lock);
/*FALLTHRU*/
case FILEID_INO32_GEN:
- fid->i32.ino = inode->i_ino;
+ fid->i32.ino = XFS_I(inode)->i_ino;
fid->i32.gen = inode->i_generation;
break;
case FILEID_INO32_GEN_PARENT | XFS_FILEID_TYPE_64FLAG:
spin_lock(&dentry->d_lock);
- fid64->parent_ino = dentry->d_parent->d_inode->i_ino;
+ fid64->parent_ino = XFS_I(dentry->d_parent->d_inode)->i_ino;
fid64->parent_gen = dentry->d_parent->d_inode->i_generation;
spin_unlock(&dentry->d_lock);
/*FALLTHRU*/
case FILEID_INO32_GEN | XFS_FILEID_TYPE_64FLAG:
- fid64->ino = inode->i_ino;
+ fid64->ino = XFS_I(inode)->i_ino;
fid64->gen = inode->i_generation;
break;
}
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index c0237c602f11..755ee8164880 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -2835,6 +2835,27 @@ corrupt_out:
return XFS_ERROR(EFSCORRUPTED);
}
+void
+xfs_promote_inode(
+ struct xfs_inode *ip)
+{
+ struct xfs_buf *bp;
+
+ ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
+
+ bp = xfs_incore(ip->i_mount->m_ddev_targp, ip->i_imap.im_blkno,
+ ip->i_imap.im_len, XBF_TRYLOCK);
+ if (!bp)
+ return;
+
+ if (XFS_BUF_ISDELAYWRITE(bp)) {
+ xfs_buf_delwri_promote(bp);
+ wake_up_process(ip->i_mount->m_ddev_targp->bt_task);
+ }
+
+ xfs_buf_relse(bp);
+}
+
/*
* Return a pointer to the extent record at file index idx.
*/
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 760140d1dd66..b4cd4739f98e 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -498,6 +498,7 @@ int xfs_iunlink(struct xfs_trans *, xfs_inode_t *);
void xfs_iext_realloc(xfs_inode_t *, int, int);
void xfs_iunpin_wait(xfs_inode_t *);
int xfs_iflush(xfs_inode_t *, uint);
+void xfs_promote_inode(struct xfs_inode *);
void xfs_lock_inodes(xfs_inode_t **, int, uint);
void xfs_lock_two_inodes(xfs_inode_t *, xfs_inode_t *, uint);
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index a14cd89fe465..34817adf4b9e 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -150,6 +150,117 @@ xlog_grant_add_space(
} while (head_val != old);
}
+STATIC bool
+xlog_reserveq_wake(
+ struct log *log,
+ int *free_bytes)
+{
+ struct xlog_ticket *tic;
+ int need_bytes;
+
+ list_for_each_entry(tic, &log->l_reserveq, t_queue) {
+ if (tic->t_flags & XLOG_TIC_PERM_RESERV)
+ need_bytes = tic->t_unit_res * tic->t_cnt;
+ else
+ need_bytes = tic->t_unit_res;
+
+ if (*free_bytes < need_bytes)
+ return false;
+ *free_bytes -= need_bytes;
+
+ trace_xfs_log_grant_wake_up(log, tic);
+ wake_up(&tic->t_wait);
+ }
+
+ return true;
+}
+
+STATIC bool
+xlog_writeq_wake(
+ struct log *log,
+ int *free_bytes)
+{
+ struct xlog_ticket *tic;
+ int need_bytes;
+
+ list_for_each_entry(tic, &log->l_writeq, t_queue) {
+ ASSERT(tic->t_flags & XLOG_TIC_PERM_RESERV);
+
+ need_bytes = tic->t_unit_res;
+
+ if (*free_bytes < need_bytes)
+ return false;
+ *free_bytes -= need_bytes;
+
+ trace_xfs_log_regrant_write_wake_up(log, tic);
+ wake_up(&tic->t_wait);
+ }
+
+ return true;
+}
+
+STATIC int
+xlog_reserveq_wait(
+ struct log *log,
+ struct xlog_ticket *tic,
+ int need_bytes)
+{
+ list_add_tail(&tic->t_queue, &log->l_reserveq);
+
+ do {
+ if (XLOG_FORCED_SHUTDOWN(log))
+ goto shutdown;
+ xlog_grant_push_ail(log, need_bytes);
+
+ XFS_STATS_INC(xs_sleep_logspace);
+ trace_xfs_log_grant_sleep(log, tic);
+
+ xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
+ trace_xfs_log_grant_wake(log, tic);
+
+ spin_lock(&log->l_grant_reserve_lock);
+ if (XLOG_FORCED_SHUTDOWN(log))
+ goto shutdown;
+ } while (xlog_space_left(log, &log->l_grant_reserve_head) < need_bytes);
+
+ list_del_init(&tic->t_queue);
+ return 0;
+shutdown:
+ list_del_init(&tic->t_queue);
+ return XFS_ERROR(EIO);
+}
+
+STATIC int
+xlog_writeq_wait(
+ struct log *log,
+ struct xlog_ticket *tic,
+ int need_bytes)
+{
+ list_add_tail(&tic->t_queue, &log->l_writeq);
+
+ do {
+ if (XLOG_FORCED_SHUTDOWN(log))
+ goto shutdown;
+ xlog_grant_push_ail(log, need_bytes);
+
+ XFS_STATS_INC(xs_sleep_logspace);
+ trace_xfs_log_regrant_write_sleep(log, tic);
+
+ xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
+ trace_xfs_log_regrant_write_wake(log, tic);
+
+ spin_lock(&log->l_grant_write_lock);
+ if (XLOG_FORCED_SHUTDOWN(log))
+ goto shutdown;
+ } while (xlog_space_left(log, &log->l_grant_write_head) < need_bytes);
+
+ list_del_init(&tic->t_queue);
+ return 0;
+shutdown:
+ list_del_init(&tic->t_queue);
+ return XFS_ERROR(EIO);
+}
+
static void
xlog_tic_reset_res(xlog_ticket_t *tic)
{
@@ -350,8 +461,19 @@ xfs_log_reserve(
retval = xlog_grant_log_space(log, internal_ticket);
}
+ if (unlikely(retval)) {
+ /*
+ * If we are failing, make sure the ticket doesn't have any
+ * current reservations. We don't want to add this back
+ * when the ticket/ transaction gets cancelled.
+ */
+ internal_ticket->t_curr_res = 0;
+ /* ungrant will give back unit_res * t_cnt. */
+ internal_ticket->t_cnt = 0;
+ }
+
return retval;
-} /* xfs_log_reserve */
+}
/*
@@ -2481,8 +2603,8 @@ restart:
/*
* Atomically get the log space required for a log ticket.
*
- * Once a ticket gets put onto the reserveq, it will only return after
- * the needed reservation is satisfied.
+ * Once a ticket gets put onto the reserveq, it will only return after the
+ * needed reservation is satisfied.
*
* This function is structured so that it has a lock free fast path. This is
* necessary because every new transaction reservation will come through this
@@ -2490,113 +2612,53 @@ restart:
* every pass.
*
* As tickets are only ever moved on and off the reserveq under the
- * l_grant_reserve_lock, we only need to take that lock if we are going
- * to add the ticket to the queue and sleep. We can avoid taking the lock if the
- * ticket was never added to the reserveq because the t_queue list head will be
- * empty and we hold the only reference to it so it can safely be checked
- * unlocked.
+ * l_grant_reserve_lock, we only need to take that lock if we are going to add
+ * the ticket to the queue and sleep. We can avoid taking the lock if the ticket
+ * was never added to the reserveq because the t_queue list head will be empty
+ * and we hold the only reference to it so it can safely be checked unlocked.
*/
STATIC int
-xlog_grant_log_space(xlog_t *log,
- xlog_ticket_t *tic)
+xlog_grant_log_space(
+ struct log *log,
+ struct xlog_ticket *tic)
{
- int free_bytes;
- int need_bytes;
+ int free_bytes, need_bytes;
+ int error = 0;
-#ifdef DEBUG
- if (log->l_flags & XLOG_ACTIVE_RECOVERY)
- panic("grant Recovery problem");
-#endif
+ ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
trace_xfs_log_grant_enter(log, tic);
+ /*
+ * If there are other waiters on the queue then give them a chance at
+ * logspace before us. Wake up the first waiters, if we do not wake
+ * up all the waiters then go to sleep waiting for more free space,
+ * otherwise try to get some space for this transaction.
+ */
need_bytes = tic->t_unit_res;
if (tic->t_flags & XFS_LOG_PERM_RESERV)
need_bytes *= tic->t_ocnt;
-
- /* something is already sleeping; insert new transaction at end */
- if (!list_empty_careful(&log->l_reserveq)) {
- spin_lock(&log->l_grant_reserve_lock);
- /* recheck the queue now we are locked */
- if (list_empty(&log->l_reserveq)) {
- spin_unlock(&log->l_grant_reserve_lock);
- goto redo;
- }
- list_add_tail(&tic->t_queue, &log->l_reserveq);
-
- trace_xfs_log_grant_sleep1(log, tic);
-
- /*
- * Gotta check this before going to sleep, while we're
- * holding the grant lock.
- */
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return;
-
- XFS_STATS_INC(xs_sleep_logspace);
- xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
-
- /*
- * If we got an error, and the filesystem is shutting down,
- * we'll catch it down below. So just continue...
- */
- trace_xfs_log_grant_wake1(log, tic);
- }
-
-redo:
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return_unlocked;
-
free_bytes = xlog_space_left(log, &log->l_grant_reserve_head);
- if (free_bytes < need_bytes) {
+ if (!list_empty_careful(&log->l_reserveq)) {
spin_lock(&log->l_grant_reserve_lock);
- if (list_empty(&tic->t_queue))
- list_add_tail(&tic->t_queue, &log->l_reserveq);
-
- trace_xfs_log_grant_sleep2(log, tic);
-
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return;
-
- xlog_grant_push_ail(log, need_bytes);
-
- XFS_STATS_INC(xs_sleep_logspace);
- xlog_wait(&tic->t_wait, &log->l_grant_reserve_lock);
-
- trace_xfs_log_grant_wake2(log, tic);
- goto redo;
- }
-
- if (!list_empty(&tic->t_queue)) {
+ if (!xlog_reserveq_wake(log, &free_bytes) ||
+ free_bytes < need_bytes)
+ error = xlog_reserveq_wait(log, tic, need_bytes);
+ spin_unlock(&log->l_grant_reserve_lock);
+ } else if (free_bytes < need_bytes) {
spin_lock(&log->l_grant_reserve_lock);
- list_del_init(&tic->t_queue);
+ error = xlog_reserveq_wait(log, tic, need_bytes);
spin_unlock(&log->l_grant_reserve_lock);
}
+ if (error)
+ return error;
- /* we've got enough space */
xlog_grant_add_space(log, &log->l_grant_reserve_head, need_bytes);
xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
trace_xfs_log_grant_exit(log, tic);
xlog_verify_grant_tail(log);
return 0;
-
-error_return_unlocked:
- spin_lock(&log->l_grant_reserve_lock);
-error_return:
- list_del_init(&tic->t_queue);
- spin_unlock(&log->l_grant_reserve_lock);
- trace_xfs_log_grant_error(log, tic);
-
- /*
- * If we are failing, make sure the ticket doesn't have any
- * current reservations. We don't want to add this back when
- * the ticket/transaction gets cancelled.
- */
- tic->t_curr_res = 0;
- tic->t_cnt = 0; /* ungrant will give back unit_res * t_cnt. */
- return XFS_ERROR(EIO);
-} /* xlog_grant_log_space */
-
+}
/*
* Replenish the byte reservation required by moving the grant write head.
@@ -2605,10 +2667,12 @@ error_return:
* free fast path.
*/
STATIC int
-xlog_regrant_write_log_space(xlog_t *log,
- xlog_ticket_t *tic)
+xlog_regrant_write_log_space(
+ struct log *log,
+ struct xlog_ticket *tic)
{
- int free_bytes, need_bytes;
+ int free_bytes, need_bytes;
+ int error = 0;
tic->t_curr_res = tic->t_unit_res;
xlog_tic_reset_res(tic);
@@ -2616,104 +2680,38 @@ xlog_regrant_write_log_space(xlog_t *log,
if (tic->t_cnt > 0)
return 0;
-#ifdef DEBUG
- if (log->l_flags & XLOG_ACTIVE_RECOVERY)
- panic("regrant Recovery problem");
-#endif
+ ASSERT(!(log->l_flags & XLOG_ACTIVE_RECOVERY));
trace_xfs_log_regrant_write_enter(log, tic);
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return_unlocked;
- /* If there are other waiters on the queue then give them a
- * chance at logspace before us. Wake up the first waiters,
- * if we do not wake up all the waiters then go to sleep waiting
- * for more free space, otherwise try to get some space for
- * this transaction.
+ /*
+ * If there are other waiters on the queue then give them a chance at
+ * logspace before us. Wake up the first waiters, if we do not wake
+ * up all the waiters then go to sleep waiting for more free space,
+ * otherwise try to get some space for this transaction.
*/
need_bytes = tic->t_unit_res;
- if (!list_empty_careful(&log->l_writeq)) {
- struct xlog_ticket *ntic;
-
- spin_lock(&log->l_grant_write_lock);
- free_bytes = xlog_space_left(log, &log->l_grant_write_head);
- list_for_each_entry(ntic, &log->l_writeq, t_queue) {
- ASSERT(ntic->t_flags & XLOG_TIC_PERM_RESERV);
-
- if (free_bytes < ntic->t_unit_res)
- break;
- free_bytes -= ntic->t_unit_res;
- wake_up(&ntic->t_wait);
- }
-
- if (ntic != list_first_entry(&log->l_writeq,
- struct xlog_ticket, t_queue)) {
- if (list_empty(&tic->t_queue))
- list_add_tail(&tic->t_queue, &log->l_writeq);
- trace_xfs_log_regrant_write_sleep1(log, tic);
-
- xlog_grant_push_ail(log, need_bytes);
-
- XFS_STATS_INC(xs_sleep_logspace);
- xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
- trace_xfs_log_regrant_write_wake1(log, tic);
- } else
- spin_unlock(&log->l_grant_write_lock);
- }
-
-redo:
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return_unlocked;
-
free_bytes = xlog_space_left(log, &log->l_grant_write_head);
- if (free_bytes < need_bytes) {
+ if (!list_empty_careful(&log->l_writeq)) {
spin_lock(&log->l_grant_write_lock);
- if (list_empty(&tic->t_queue))
- list_add_tail(&tic->t_queue, &log->l_writeq);
-
- if (XLOG_FORCED_SHUTDOWN(log))
- goto error_return;
-
- xlog_grant_push_ail(log, need_bytes);
-
- XFS_STATS_INC(xs_sleep_logspace);
- trace_xfs_log_regrant_write_sleep2(log, tic);
- xlog_wait(&tic->t_wait, &log->l_grant_write_lock);
-
- trace_xfs_log_regrant_write_wake2(log, tic);
- goto redo;
- }
-
- if (!list_empty(&tic->t_queue)) {
+ if (!xlog_writeq_wake(log, &free_bytes) ||
+ free_bytes < need_bytes)
+ error = xlog_writeq_wait(log, tic, need_bytes);
+ spin_unlock(&log->l_grant_write_lock);
+ } else if (free_bytes < need_bytes) {
spin_lock(&log->l_grant_write_lock);
- list_del_init(&tic->t_queue);
+ error = xlog_writeq_wait(log, tic, need_bytes);
spin_unlock(&log->l_grant_write_lock);
}
- /* we've got enough space */
+ if (error)
+ return error;
+
xlog_grant_add_space(log, &log->l_grant_write_head, need_bytes);
trace_xfs_log_regrant_write_exit(log, tic);
xlog_verify_grant_tail(log);
return 0;
-
-
- error_return_unlocked:
- spin_lock(&log->l_grant_write_lock);
- error_return:
- list_del_init(&tic->t_queue);
- spin_unlock(&log->l_grant_write_lock);
- trace_xfs_log_regrant_write_error(log, tic);
-
- /*
- * If we are failing, make sure the ticket doesn't have any
- * current reservations. We don't want to add this back when
- * the ticket/transaction gets cancelled.
- */
- tic->t_curr_res = 0;
- tic->t_cnt = 0; /* ungrant will give back unit_res * t_cnt. */
- return XFS_ERROR(EIO);
-} /* xlog_regrant_write_log_space */
-
+}
/* The first cnt-1 times through here we don't need to
* move the grant write head because the permanent
diff --git a/fs/xfs/xfs_sync.c b/fs/xfs/xfs_sync.c
index aa3dc1a4d53d..be5c51d8f757 100644
--- a/fs/xfs/xfs_sync.c
+++ b/fs/xfs/xfs_sync.c
@@ -770,6 +770,17 @@ restart:
if (!xfs_iflock_nowait(ip)) {
if (!(sync_mode & SYNC_WAIT))
goto out;
+
+ /*
+ * If we only have a single dirty inode in a cluster there is
+ * a fair chance that the AIL push may have pushed it into
+ * the buffer, but xfsbufd won't touch it until 30 seconds
+ * from now, and thus we will lock up here.
+ *
+ * Promote the inode buffer to the front of the delwri list
+ * and wake up xfsbufd now.
+ */
+ xfs_promote_inode(ip);
xfs_iflock(ip);
}
diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index f1d2802b2f07..494035798873 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -834,18 +834,14 @@ DEFINE_LOGGRANT_EVENT(xfs_log_umount_write);
DEFINE_LOGGRANT_EVENT(xfs_log_grant_enter);
DEFINE_LOGGRANT_EVENT(xfs_log_grant_exit);
DEFINE_LOGGRANT_EVENT(xfs_log_grant_error);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep1);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake1);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep2);
-DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake2);
+DEFINE_LOGGRANT_EVENT(xfs_log_grant_sleep);
+DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake);
DEFINE_LOGGRANT_EVENT(xfs_log_grant_wake_up);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_enter);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_exit);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_error);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep1);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake1);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep2);
-DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake2);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_sleep);
+DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_write_wake_up);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_enter);
DEFINE_LOGGRANT_EVENT(xfs_log_regrant_reserve_exit);