#include "locking.h"
#include "free-space-cache.h"
#include "math.h"
+#include "sysfs.h"
#undef SCRAMBLE_DELAYED_REFS
again:
mutex_lock(&caching_ctl->mutex);
/* need to make sure the commit_root doesn't disappear */
- down_read(&fs_info->extent_commit_sem);
+ down_read(&fs_info->commit_root_sem);
next:
ret = btrfs_search_slot(NULL, extent_root, &key, path, 0, 0);
if (ret)
break;
- if (need_resched()) {
+ if (need_resched() ||
+ rwsem_is_contended(&fs_info->commit_root_sem)) {
caching_ctl->progress = last;
btrfs_release_path(path);
- up_read(&fs_info->extent_commit_sem);
+ up_read(&fs_info->commit_root_sem);
mutex_unlock(&caching_ctl->mutex);
cond_resched();
goto again;
err:
btrfs_free_path(path);
- up_read(&fs_info->extent_commit_sem);
+ up_read(&fs_info->commit_root_sem);
free_excluded_extents(extent_root, block_group);
caching_ctl->block_group = cache;
caching_ctl->progress = cache->key.objectid;
atomic_set(&caching_ctl->count, 1);
- caching_ctl->work.func = caching_thread;
+ btrfs_init_work(&caching_ctl->work, caching_thread, NULL, NULL);
spin_lock(&cache->lock);
/*
return 0;
}
- down_write(&fs_info->extent_commit_sem);
+ down_write(&fs_info->commit_root_sem);
atomic_inc(&caching_ctl->count);
list_add_tail(&caching_ctl->list, &fs_info->caching_block_groups);
- up_write(&fs_info->extent_commit_sem);
+ up_write(&fs_info->commit_root_sem);
btrfs_get_block_group(cache);
- btrfs_queue_worker(&fs_info->caching_workers, &caching_ctl->work);
+ btrfs_queue_work(fs_info->caching_workers, &caching_ctl->work);
return ret;
}
btrfs_put_delayed_ref(&head->node);
goto search_again;
}
+ spin_lock(&head->lock);
if (head->extent_op && head->extent_op->update_flags)
extent_flags |= head->extent_op->flags_to_set;
else
BUG_ON(num_refs == 0);
num_refs += head->node.ref_mod;
+ spin_unlock(&head->lock);
mutex_unlock(&head->mutex);
}
spin_unlock(&delayed_refs->lock);
__le64 lenum;
lenum = cpu_to_le64(root_objectid);
- high_crc = crc32c(high_crc, &lenum, sizeof(lenum));
+ high_crc = btrfs_crc32c(high_crc, &lenum, sizeof(lenum));
lenum = cpu_to_le64(owner);
- low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
+ low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
lenum = cpu_to_le64(offset);
- low_crc = crc32c(low_crc, &lenum, sizeof(lenum));
+ low_crc = btrfs_crc32c(low_crc, &lenum, sizeof(lenum));
return ((u64)high_crc << 31) ^ (u64)low_crc;
}
select_delayed_ref(struct btrfs_delayed_ref_head *head)
{
struct rb_node *node;
- struct btrfs_delayed_ref_node *ref;
- int action = BTRFS_ADD_DELAYED_REF;
-again:
+ struct btrfs_delayed_ref_node *ref, *last = NULL;;
+
/*
* select delayed ref of type BTRFS_ADD_DELAYED_REF first.
* this prevents ref count from going down to zero when
* there still are pending delayed ref.
*/
- node = rb_prev(&head->node.rb_node);
- while (1) {
- if (!node)
- break;
+ node = rb_first(&head->ref_root);
+ while (node) {
ref = rb_entry(node, struct btrfs_delayed_ref_node,
rb_node);
- if (ref->bytenr != head->node.bytenr)
- break;
- if (ref->action == action)
+ if (ref->action == BTRFS_ADD_DELAYED_REF)
return ref;
- node = rb_prev(node);
- }
- if (action == BTRFS_ADD_DELAYED_REF) {
- action = BTRFS_DROP_DELAYED_REF;
- goto again;
+ else if (last == NULL)
+ last = ref;
+ node = rb_next(node);
}
- return NULL;
+ return last;
}
/*
* Returns 0 on success or if called with an already aborted transaction.
* Returns -ENOMEM or -EIO on failure and will abort the transaction.
*/
-static noinline int run_clustered_refs(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
- struct list_head *cluster)
+static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root,
+ unsigned long nr)
{
struct btrfs_delayed_ref_root *delayed_refs;
struct btrfs_delayed_ref_node *ref;
struct btrfs_delayed_ref_head *locked_ref = NULL;
struct btrfs_delayed_extent_op *extent_op;
struct btrfs_fs_info *fs_info = root->fs_info;
+ ktime_t start = ktime_get();
int ret;
- int count = 0;
+ unsigned long count = 0;
+ unsigned long actual_count = 0;
int must_insert_reserved = 0;
delayed_refs = &trans->transaction->delayed_refs;
while (1) {
if (!locked_ref) {
- /* pick a new head ref from the cluster list */
- if (list_empty(cluster))
+ if (count >= nr)
break;
- locked_ref = list_entry(cluster->next,
- struct btrfs_delayed_ref_head, cluster);
+ spin_lock(&delayed_refs->lock);
+ locked_ref = btrfs_select_ref_head(trans);
+ if (!locked_ref) {
+ spin_unlock(&delayed_refs->lock);
+ break;
+ }
/* grab the lock that says we are going to process
* all the refs for this head */
ret = btrfs_delayed_ref_lock(trans, locked_ref);
-
+ spin_unlock(&delayed_refs->lock);
/*
* we may have dropped the spin lock to get the head
* mutex lock, and that might have given someone else
* finish. If we merged anything we need to re-loop so we can
* get a good ref.
*/
+ spin_lock(&locked_ref->lock);
btrfs_merge_delayed_refs(trans, fs_info, delayed_refs,
locked_ref);
if (ref && ref->seq &&
btrfs_check_delayed_seq(fs_info, delayed_refs, ref->seq)) {
- /*
- * there are still refs with lower seq numbers in the
- * process of being added. Don't run this ref yet.
- */
- list_del_init(&locked_ref->cluster);
+ spin_unlock(&locked_ref->lock);
btrfs_delayed_ref_unlock(locked_ref);
- locked_ref = NULL;
+ spin_lock(&delayed_refs->lock);
+ locked_ref->processing = 0;
delayed_refs->num_heads_ready++;
spin_unlock(&delayed_refs->lock);
+ locked_ref = NULL;
cond_resched();
- spin_lock(&delayed_refs->lock);
+ count++;
continue;
}
locked_ref->extent_op = NULL;
if (!ref) {
+
+
/* All delayed refs have been processed, Go ahead
* and send the head node to run_one_delayed_ref,
* so that any accounting fixes can happen
}
if (extent_op) {
- spin_unlock(&delayed_refs->lock);
-
+ spin_unlock(&locked_ref->lock);
ret = run_delayed_extent_op(trans, root,
ref, extent_op);
btrfs_free_delayed_extent_op(extent_op);
*/
if (must_insert_reserved)
locked_ref->must_insert_reserved = 1;
+ locked_ref->processing = 0;
btrfs_debug(fs_info, "run_delayed_extent_op returned %d", ret);
- spin_lock(&delayed_refs->lock);
btrfs_delayed_ref_unlock(locked_ref);
return ret;
}
+ continue;
+ }
- goto next;
+ /*
+ * Need to drop our head ref lock and re-aqcuire the
+ * delayed ref lock and then re-check to make sure
+ * nobody got added.
+ */
+ spin_unlock(&locked_ref->lock);
+ spin_lock(&delayed_refs->lock);
+ spin_lock(&locked_ref->lock);
+ if (rb_first(&locked_ref->ref_root) ||
+ locked_ref->extent_op) {
+ spin_unlock(&locked_ref->lock);
+ spin_unlock(&delayed_refs->lock);
+ continue;
}
+ ref->in_tree = 0;
+ delayed_refs->num_heads--;
+ rb_erase(&locked_ref->href_node,
+ &delayed_refs->href_root);
+ spin_unlock(&delayed_refs->lock);
+ } else {
+ actual_count++;
+ ref->in_tree = 0;
+ rb_erase(&ref->rb_node, &locked_ref->ref_root);
}
+ atomic_dec(&delayed_refs->num_entries);
- ref->in_tree = 0;
- rb_erase(&ref->rb_node, &delayed_refs->root);
- delayed_refs->num_entries--;
if (!btrfs_delayed_ref_is_head(ref)) {
/*
* when we play the delayed ref, also correct the
default:
WARN_ON(1);
}
- } else {
- list_del_init(&locked_ref->cluster);
}
- spin_unlock(&delayed_refs->lock);
+ spin_unlock(&locked_ref->lock);
ret = run_one_delayed_ref(trans, root, ref, extent_op,
must_insert_reserved);
btrfs_free_delayed_extent_op(extent_op);
if (ret) {
+ locked_ref->processing = 0;
btrfs_delayed_ref_unlock(locked_ref);
btrfs_put_delayed_ref(ref);
btrfs_debug(fs_info, "run_one_delayed_ref returned %d", ret);
- spin_lock(&delayed_refs->lock);
return ret;
}
}
btrfs_put_delayed_ref(ref);
count++;
-next:
cond_resched();
+ }
+
+ /*
+ * We don't want to include ref heads since we can have empty ref heads
+ * and those will drastically skew our runtime down since we just do
+ * accounting, no actual extent tree updates.
+ */
+ if (actual_count > 0) {
+ u64 runtime = ktime_to_ns(ktime_sub(ktime_get(), start));
+ u64 avg;
+
+ /*
+ * We weigh the current average higher than our current runtime
+ * to avoid large swings in the average.
+ */
spin_lock(&delayed_refs->lock);
+ avg = fs_info->avg_delayed_ref_runtime * 3 + runtime;
+ avg = div64_u64(avg, 4);
+ fs_info->avg_delayed_ref_runtime = avg;
+ spin_unlock(&delayed_refs->lock);
}
- return count;
+ return 0;
}
#ifdef SCRAMBLE_DELAYED_REFS
return ret;
}
-static int refs_newer(struct btrfs_delayed_ref_root *delayed_refs, int seq,
- int count)
-{
- int val = atomic_read(&delayed_refs->ref_seq);
-
- if (val < seq || val >= seq + count)
- return 1;
- return 0;
-}
-
static inline u64 heads_to_leaves(struct btrfs_root *root, u64 heads)
{
u64 num_bytes;
return div64_u64(num_bytes, BTRFS_LEAF_DATA_SIZE(root));
}
-int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
+int btrfs_check_space_for_delayed_refs(struct btrfs_trans_handle *trans,
struct btrfs_root *root)
{
struct btrfs_block_rsv *global_rsv;
return ret;
}
+int btrfs_should_throttle_delayed_refs(struct btrfs_trans_handle *trans,
+ struct btrfs_root *root)
+{
+ struct btrfs_fs_info *fs_info = root->fs_info;
+ u64 num_entries =
+ atomic_read(&trans->transaction->delayed_refs.num_entries);
+ u64 avg_runtime;
+
+ smp_mb();
+ avg_runtime = fs_info->avg_delayed_ref_runtime;
+ if (num_entries * avg_runtime >= NSEC_PER_SEC)
+ return 1;
+
+ return btrfs_check_space_for_delayed_refs(trans, root);
+}
+
/*
* this starts processing the delayed reference count updates and
* extent insertions we have queued up so far. count can be
{
struct rb_node *node;
struct btrfs_delayed_ref_root *delayed_refs;
- struct btrfs_delayed_ref_node *ref;
- struct list_head cluster;
+ struct btrfs_delayed_ref_head *head;
int ret;
- u64 delayed_start;
int run_all = count == (unsigned long)-1;
int run_most = 0;
- int loops;
/* We'll clean this up in btrfs_cleanup_transaction */
if (trans->aborted)
btrfs_delayed_refs_qgroup_accounting(trans, root->fs_info);
delayed_refs = &trans->transaction->delayed_refs;
- INIT_LIST_HEAD(&cluster);
if (count == 0) {
- count = delayed_refs->num_entries * 2;
+ count = atomic_read(&delayed_refs->num_entries) * 2;
run_most = 1;
}
- if (!run_all && !run_most) {
- int old;
- int seq = atomic_read(&delayed_refs->ref_seq);
-
-progress:
- old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1);
- if (old) {
- DEFINE_WAIT(__wait);
- if (delayed_refs->flushing ||
- !btrfs_should_throttle_delayed_refs(trans, root))
- return 0;
-
- prepare_to_wait(&delayed_refs->wait, &__wait,
- TASK_UNINTERRUPTIBLE);
-
- old = atomic_cmpxchg(&delayed_refs->procs_running_refs, 0, 1);
- if (old) {
- schedule();
- finish_wait(&delayed_refs->wait, &__wait);
-
- if (!refs_newer(delayed_refs, seq, 256))
- goto progress;
- else
- return 0;
- } else {
- finish_wait(&delayed_refs->wait, &__wait);
- goto again;
- }
- }
-
- } else {
- atomic_inc(&delayed_refs->procs_running_refs);
- }
-
again:
- loops = 0;
- spin_lock(&delayed_refs->lock);
-
#ifdef SCRAMBLE_DELAYED_REFS
delayed_refs->run_delayed_start = find_middle(&delayed_refs->root);
#endif
-
- while (1) {
- if (!(run_all || run_most) &&
- !btrfs_should_throttle_delayed_refs(trans, root))
- break;
-
- /*
- * go find something we can process in the rbtree. We start at
- * the beginning of the tree, and then build a cluster
- * of refs to process starting at the first one we are able to
- * lock
- */
- delayed_start = delayed_refs->run_delayed_start;
- ret = btrfs_find_ref_cluster(trans, &cluster,
- delayed_refs->run_delayed_start);
- if (ret)
- break;
-
- ret = run_clustered_refs(trans, root, &cluster);
- if (ret < 0) {
- btrfs_release_ref_cluster(&cluster);
- spin_unlock(&delayed_refs->lock);
- btrfs_abort_transaction(trans, root, ret);
- atomic_dec(&delayed_refs->procs_running_refs);
- wake_up(&delayed_refs->wait);
- return ret;
- }
-
- atomic_add(ret, &delayed_refs->ref_seq);
-
- count -= min_t(unsigned long, ret, count);
-
- if (count == 0)
- break;
-
- if (delayed_start >= delayed_refs->run_delayed_start) {
- if (loops == 0) {
- /*
- * btrfs_find_ref_cluster looped. let's do one
- * more cycle. if we don't run any delayed ref
- * during that cycle (because we can't because
- * all of them are blocked), bail out.
- */
- loops = 1;
- } else {
- /*
- * no runnable refs left, stop trying
- */
- BUG_ON(run_all);
- break;
- }
- }
- if (ret) {
- /* refs were run, let's reset staleness detection */
- loops = 0;
- }
+ ret = __btrfs_run_delayed_refs(trans, root, count);
+ if (ret < 0) {
+ btrfs_abort_transaction(trans, root, ret);
+ return ret;
}
if (run_all) {
- if (!list_empty(&trans->new_bgs)) {
- spin_unlock(&delayed_refs->lock);
+ if (!list_empty(&trans->new_bgs))
btrfs_create_pending_block_groups(trans, root);
- spin_lock(&delayed_refs->lock);
- }
- node = rb_first(&delayed_refs->root);
- if (!node)
+ spin_lock(&delayed_refs->lock);
+ node = rb_first(&delayed_refs->href_root);
+ if (!node) {
+ spin_unlock(&delayed_refs->lock);
goto out;
+ }
count = (unsigned long)-1;
while (node) {
- ref = rb_entry(node, struct btrfs_delayed_ref_node,
- rb_node);
- if (btrfs_delayed_ref_is_head(ref)) {
- struct btrfs_delayed_ref_head *head;
+ head = rb_entry(node, struct btrfs_delayed_ref_head,
+ href_node);
+ if (btrfs_delayed_ref_is_head(&head->node)) {
+ struct btrfs_delayed_ref_node *ref;
- head = btrfs_delayed_node_to_head(ref);
+ ref = &head->node;
atomic_inc(&ref->refs);
spin_unlock(&delayed_refs->lock);
btrfs_put_delayed_ref(ref);
cond_resched();
goto again;
+ } else {
+ WARN_ON(1);
}
node = rb_next(node);
}
spin_unlock(&delayed_refs->lock);
- schedule_timeout(1);
+ cond_resched();
goto again;
}
out:
- atomic_dec(&delayed_refs->procs_running_refs);
- smp_mb();
- if (waitqueue_active(&delayed_refs->wait))
- wake_up(&delayed_refs->wait);
-
- spin_unlock(&delayed_refs->lock);
assert_qgroups_uptodate(trans);
return 0;
}
struct rb_node *node;
int ret = 0;
- ret = -ENOENT;
delayed_refs = &trans->transaction->delayed_refs;
spin_lock(&delayed_refs->lock);
head = btrfs_find_delayed_ref_head(trans, bytenr);
- if (!head)
- goto out;
+ if (!head) {
+ spin_unlock(&delayed_refs->lock);
+ return 0;
+ }
if (!mutex_trylock(&head->mutex)) {
atomic_inc(&head->node.refs);
btrfs_put_delayed_ref(&head->node);
return -EAGAIN;
}
+ spin_unlock(&delayed_refs->lock);
- node = rb_prev(&head->node.rb_node);
- if (!node)
- goto out_unlock;
-
- ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-
- if (ref->bytenr != bytenr)
- goto out_unlock;
-
- ret = 1;
- if (ref->type != BTRFS_EXTENT_DATA_REF_KEY)
- goto out_unlock;
+ spin_lock(&head->lock);
+ node = rb_first(&head->ref_root);
+ while (node) {
+ ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
+ node = rb_next(node);
- data_ref = btrfs_delayed_node_to_data_ref(ref);
+ /* If it's a shared ref we know a cross reference exists */
+ if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) {
+ ret = 1;
+ break;
+ }
- node = rb_prev(node);
- if (node) {
- int seq = ref->seq;
+ data_ref = btrfs_delayed_node_to_data_ref(ref);
- ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
- if (ref->bytenr == bytenr && ref->seq == seq)
- goto out_unlock;
+ /*
+ * If our ref doesn't match the one we're currently looking at
+ * then we have a cross reference.
+ */
+ if (data_ref->root != root->root_key.objectid ||
+ data_ref->objectid != objectid ||
+ data_ref->offset != offset) {
+ ret = 1;
+ break;
+ }
}
-
- if (data_ref->root != root->root_key.objectid ||
- data_ref->objectid != objectid || data_ref->offset != offset)
- goto out_unlock;
-
- ret = 0;
-out_unlock:
+ spin_unlock(&head->lock);
mutex_unlock(&head->mutex);
-out:
- spin_unlock(&delayed_refs->lock);
return ret;
}
return readonly;
}
+static const char *alloc_name(u64 flags)
+{
+ switch (flags) {
+ case BTRFS_BLOCK_GROUP_METADATA|BTRFS_BLOCK_GROUP_DATA:
+ return "mixed";
+ case BTRFS_BLOCK_GROUP_METADATA:
+ return "metadata";
+ case BTRFS_BLOCK_GROUP_DATA:
+ return "data";
+ case BTRFS_BLOCK_GROUP_SYSTEM:
+ return "system";
+ default:
+ WARN_ON(1);
+ return "invalid-combination";
+ };
+}
+
static int update_space_info(struct btrfs_fs_info *info, u64 flags,
u64 total_bytes, u64 bytes_used,
struct btrfs_space_info **space_info)
return ret;
}
- for (i = 0; i < BTRFS_NR_RAID_TYPES; i++)
+ for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
INIT_LIST_HEAD(&found->block_groups[i]);
+ kobject_init(&found->block_group_kobjs[i], &btrfs_raid_ktype);
+ }
init_rwsem(&found->groups_sem);
spin_lock_init(&found->lock);
found->flags = flags & BTRFS_BLOCK_GROUP_TYPE_MASK;
found->chunk_alloc = 0;
found->flush = 0;
init_waitqueue_head(&found->wait);
+
+ ret = kobject_init_and_add(&found->kobj, &space_info_ktype,
+ info->space_info_kobj, "%s",
+ alloc_name(found->flags));
+ if (ret) {
+ kfree(found);
+ return ret;
+ }
+
*space_info = found;
list_add_rcu(&found->list, &info->space_info);
if (flags & BTRFS_BLOCK_GROUP_DATA)
info->data_sinfo = found;
- return 0;
+
+ return ret;
}
static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
}
static void btrfs_writeback_inodes_sb_nr(struct btrfs_root *root,
- unsigned long nr_pages)
+ unsigned long nr_pages, int nr_items)
{
struct super_block *sb = root->fs_info->sb;
* the filesystem is readonly(all dirty pages are written to
* the disk).
*/
- btrfs_start_delalloc_roots(root->fs_info, 0);
+ btrfs_start_delalloc_roots(root->fs_info, 0, nr_items);
if (!current->journal_info)
- btrfs_wait_ordered_roots(root->fs_info, -1);
+ btrfs_wait_ordered_roots(root->fs_info, nr_items);
}
}
while (delalloc_bytes && loops < 3) {
max_reclaim = min(delalloc_bytes, to_reclaim);
nr_pages = max_reclaim >> PAGE_CACHE_SHIFT;
- btrfs_writeback_inodes_sb_nr(root, nr_pages);
+ btrfs_writeback_inodes_sb_nr(root, nr_pages, items);
/*
* We need to wait for the async pages to actually start before
* we do anything.
goto commit;
/* See if there is enough pinned space to make this reservation */
- spin_lock(&space_info->lock);
if (percpu_counter_compare(&space_info->total_bytes_pinned,
- bytes) >= 0) {
- spin_unlock(&space_info->lock);
+ bytes) >= 0)
goto commit;
- }
- spin_unlock(&space_info->lock);
/*
* See if there is some space in the delayed insertion reservation for
if (space_info != delayed_rsv->space_info)
return -ENOSPC;
- spin_lock(&space_info->lock);
spin_lock(&delayed_rsv->lock);
if (percpu_counter_compare(&space_info->total_bytes_pinned,
bytes - delayed_rsv->size) >= 0) {
spin_unlock(&delayed_rsv->lock);
- spin_unlock(&space_info->lock);
return -ENOSPC;
}
spin_unlock(&delayed_rsv->lock);
- spin_unlock(&space_info->lock);
commit:
trans = btrfs_join_transaction(root);
break;
case FLUSH_DELALLOC:
case FLUSH_DELALLOC_WAIT:
- shrink_delalloc(root, num_bytes, orig_bytes,
+ shrink_delalloc(root, num_bytes * 2, orig_bytes,
state == FLUSH_DELALLOC_WAIT);
break;
case ALLOC_CHUNK:
u64 num_bytes)
{
struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
- if (global_rsv->full || global_rsv == block_rsv ||
+ if (global_rsv == block_rsv ||
block_rsv->space_info != global_rsv->space_info)
global_rsv = NULL;
block_rsv_release_bytes(root->fs_info, block_rsv, global_rsv,
struct btrfs_block_group_cache *cache;
struct btrfs_space_info *space_info;
- down_write(&fs_info->extent_commit_sem);
+ down_write(&fs_info->commit_root_sem);
list_for_each_entry_safe(caching_ctl, next,
&fs_info->caching_block_groups, list) {
else
fs_info->pinned_extents = &fs_info->freed_extents[0];
- up_write(&fs_info->extent_commit_sem);
+ up_write(&fs_info->commit_root_sem);
list_for_each_entry_rcu(space_info, &fs_info->space_info, list)
percpu_counter_set(&space_info->total_bytes_pinned, 0);
{
struct btrfs_delayed_ref_head *head;
struct btrfs_delayed_ref_root *delayed_refs;
- struct btrfs_delayed_ref_node *ref;
- struct rb_node *node;
int ret = 0;
delayed_refs = &trans->transaction->delayed_refs;
spin_lock(&delayed_refs->lock);
head = btrfs_find_delayed_ref_head(trans, bytenr);
if (!head)
- goto out;
+ goto out_delayed_unlock;
- node = rb_prev(&head->node.rb_node);
- if (!node)
- goto out;
-
- ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node);
-
- /* there are still entries for this ref, we can't drop it */
- if (ref->bytenr == bytenr)
+ spin_lock(&head->lock);
+ if (rb_first(&head->ref_root))
goto out;
if (head->extent_op) {
* ahead and process it.
*/
head->node.in_tree = 0;
- rb_erase(&head->node.rb_node, &delayed_refs->root);
+ rb_erase(&head->href_node, &delayed_refs->href_root);
- delayed_refs->num_entries--;
+ atomic_dec(&delayed_refs->num_entries);
/*
* we don't take a ref on the node because we're removing it from the
* tree, so we just steal the ref the tree was holding.
*/
delayed_refs->num_heads--;
- if (list_empty(&head->cluster))
+ if (head->processing == 0)
delayed_refs->num_heads_ready--;
-
- list_del_init(&head->cluster);
+ head->processing = 0;
+ spin_unlock(&head->lock);
spin_unlock(&delayed_refs->lock);
BUG_ON(head->extent_op);
btrfs_put_delayed_ref(&head->node);
return ret;
out:
+ spin_unlock(&head->lock);
+
+out_delayed_unlock:
spin_unlock(&delayed_refs->lock);
return 0;
}
return BTRFS_RAID_SINGLE; /* BTRFS_BLOCK_GROUP_SINGLE */
}
-static int get_block_group_index(struct btrfs_block_group_cache *cache)
+int get_block_group_index(struct btrfs_block_group_cache *cache)
{
return __get_raid_index(cache->flags);
}
+static const char *btrfs_raid_type_names[BTRFS_NR_RAID_TYPES] = {
+ [BTRFS_RAID_RAID10] = "raid10",
+ [BTRFS_RAID_RAID1] = "raid1",
+ [BTRFS_RAID_DUP] = "dup",
+ [BTRFS_RAID_RAID0] = "raid0",
+ [BTRFS_RAID_SINGLE] = "single",
+ [BTRFS_RAID_RAID5] = "raid5",
+ [BTRFS_RAID_RAID6] = "raid6",
+};
+
+static const char *get_raid_name(enum btrfs_raid_types type)
+{
+ if (type >= BTRFS_NR_RAID_TYPES)
+ return NULL;
+
+ return btrfs_raid_type_names[type];
+}
+
enum btrfs_loop_type {
LOOP_CACHING_NOWAIT = 0,
LOOP_CACHING_WAIT = 1,
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;
u64 search_start = 0;
u64 max_extent_size = 0;
int empty_cluster = 2 * 1024 * 1024;
int index = __get_raid_index(flags);
int alloc_type = (flags & BTRFS_BLOCK_GROUP_DATA) ?
RESERVE_ALLOC_NO_ACCOUNT : RESERVE_ALLOC;
- bool found_uncached_bg = false;
bool failed_cluster_refill = false;
bool failed_alloc = false;
bool use_cluster = true;
if (search_start == hint_byte) {
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.
u64 offset;
int cached;
- used_block_group = block_group;
btrfs_get_block_group(block_group);
search_start = block_group->key.objectid;
have_block_group:
cached = block_group_cache_done(block_group);
if (unlikely(!cached)) {
- found_uncached_bg = true;
ret = cache_block_group(block_group, 0);
BUG_ON(ret < 0);
ret = 0;
* lets look there
*/
if (last_ptr) {
+ struct btrfs_block_group_cache *used_block_group;
unsigned long aligned_cluster;
/*
* the refill lock keeps out other
if (used_block_group != block_group &&
(!used_block_group ||
used_block_group->ro ||
- !block_group_bits(used_block_group, flags))) {
- used_block_group = block_group;
+ !block_group_bits(used_block_group, flags)))
goto refill_cluster;
- }
if (used_block_group != block_group)
btrfs_get_block_group(used_block_group);
/* we have a block, we're done */
spin_unlock(&last_ptr->refill_lock);
trace_btrfs_reserve_extent_cluster(root,
- block_group, search_start, num_bytes);
+ used_block_group,
+ search_start, num_bytes);
+ if (used_block_group != block_group) {
+ btrfs_put_block_group(block_group);
+ block_group = used_block_group;
+ }
goto checks;
}
WARN_ON(last_ptr->block_group != used_block_group);
- if (used_block_group != block_group) {
+ if (used_block_group != block_group)
btrfs_put_block_group(used_block_group);
- used_block_group = block_group;
- }
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
goto loop;
}
checks:
- search_start = stripe_align(root, used_block_group,
+ search_start = stripe_align(root, block_group,
offset, num_bytes);
/* move on to the next group */
if (search_start + num_bytes >
- used_block_group->key.objectid + used_block_group->key.offset) {
- btrfs_add_free_space(used_block_group, offset, num_bytes);
+ block_group->key.objectid + block_group->key.offset) {
+ btrfs_add_free_space(block_group, offset, num_bytes);
goto loop;
}
if (offset < search_start)
- btrfs_add_free_space(used_block_group, offset,
+ btrfs_add_free_space(block_group, offset,
search_start - offset);
BUG_ON(offset > search_start);
- ret = btrfs_update_reserved_bytes(used_block_group, num_bytes,
+ ret = btrfs_update_reserved_bytes(block_group, num_bytes,
alloc_type);
if (ret == -EAGAIN) {
- btrfs_add_free_space(used_block_group, offset, num_bytes);
+ btrfs_add_free_space(block_group, offset, num_bytes);
goto loop;
}
trace_btrfs_reserve_extent(orig_root, block_group,
search_start, num_bytes);
- 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);
int index = 0;
spin_lock(&info->lock);
- printk(KERN_INFO "space_info %llu has %llu free, is %sfull\n",
+ printk(KERN_INFO "BTRFS: space_info %llu has %llu free, is %sfull\n",
info->flags,
info->total_bytes - info->bytes_used - info->bytes_pinned -
info->bytes_reserved - info->bytes_readonly,
(info->full) ? "" : "not ");
- printk(KERN_INFO "space_info total=%llu, used=%llu, pinned=%llu, "
+ printk(KERN_INFO "BTRFS: space_info total=%llu, used=%llu, pinned=%llu, "
"reserved=%llu, may_use=%llu, readonly=%llu\n",
info->total_bytes, info->bytes_used, info->bytes_pinned,
info->bytes_reserved, info->bytes_may_use,
again:
list_for_each_entry(cache, &info->block_groups[index], list) {
spin_lock(&cache->lock);
- printk(KERN_INFO "block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s\n",
+ printk(KERN_INFO "BTRFS: "
+ "block group %llu has %llu bytes, "
+ "%llu used %llu pinned %llu reserved %s\n",
cache->key.objectid, cache->key.offset,
btrfs_block_group_used(&cache->item), cache->pinned,
cache->reserved, cache->ro ? "[readonly]" : "");
/*DEFAULT_RATELIMIT_BURST*/ 1);
if (__ratelimit(&_rs))
WARN(1, KERN_DEBUG
- "btrfs: block rsv returned %d\n", ret);
+ "BTRFS: block rsv returned %d\n", ret);
}
try_reserve:
ret = reserve_metadata_bytes(root, block_rsv, blocksize,
btrfs_end_transaction_throttle(trans, tree_root);
if (!for_reloc && btrfs_need_cleaner_sleep(root)) {
- pr_debug("btrfs: drop snapshot early exit\n");
+ pr_debug("BTRFS: drop snapshot early exit\n");
err = -EAGAIN;
goto out_free;
}
*/
if (!for_reloc && root_dropped == false)
btrfs_add_dead_root(root);
- if (err)
+ if (err && err != -EAGAIN)
btrfs_std_error(root->fs_info, err);
return err;
}
struct btrfs_caching_control *caching_ctl;
struct rb_node *n;
- down_write(&info->extent_commit_sem);
+ down_write(&info->commit_root_sem);
while (!list_empty(&info->caching_block_groups)) {
caching_ctl = list_entry(info->caching_block_groups.next,
struct btrfs_caching_control, list);
list_del(&caching_ctl->list);
put_caching_control(caching_ctl);
}
- up_write(&info->extent_commit_sem);
+ up_write(&info->commit_root_sem);
spin_lock(&info->block_group_cache_lock);
while ((n = rb_last(&info->block_group_cache_tree)) != NULL) {
release_global_block_rsv(info);
while (!list_empty(&info->space_info)) {
+ int i;
+
space_info = list_entry(info->space_info.next,
struct btrfs_space_info,
list);
dump_space_info(space_info, 0, 0);
}
}
- percpu_counter_destroy(&space_info->total_bytes_pinned);
list_del(&space_info->list);
- kfree(space_info);
+ for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
+ struct kobject *kobj;
+ kobj = &space_info->block_group_kobjs[i];
+ if (kobj->parent) {
+ kobject_del(kobj);
+ kobject_put(kobj);
+ }
+ }
+ kobject_del(&space_info->kobj);
+ kobject_put(&space_info->kobj);
}
return 0;
}
int index = get_block_group_index(cache);
down_write(&space_info->groups_sem);
+ if (list_empty(&space_info->block_groups[index])) {
+ struct kobject *kobj = &space_info->block_group_kobjs[index];
+ int ret;
+
+ kobject_get(&space_info->kobj); /* put in release */
+ ret = kobject_add(kobj, &space_info->kobj, "%s",
+ get_raid_name(index));
+ if (ret) {
+ pr_warn("BTRFS: failed to add kobject for block cache. ignoring.\n");
+ kobject_put(&space_info->kobj);
+ }
+ }
list_add_tail(&cache->list, &space_info->block_groups[index]);
up_write(&space_info->groups_sem);
}
+static struct btrfs_block_group_cache *
+btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size)
+{
+ struct btrfs_block_group_cache *cache;
+
+ cache = kzalloc(sizeof(*cache), GFP_NOFS);
+ if (!cache)
+ return NULL;
+
+ cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
+ GFP_NOFS);
+ if (!cache->free_space_ctl) {
+ kfree(cache);
+ return NULL;
+ }
+
+ cache->key.objectid = start;
+ cache->key.offset = size;
+ cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
+
+ cache->sectorsize = root->sectorsize;
+ cache->fs_info = root->fs_info;
+ cache->full_stripe_len = btrfs_full_stripe_len(root,
+ &root->fs_info->mapping_tree,
+ start);
+ atomic_set(&cache->count, 1);
+ spin_lock_init(&cache->lock);
+ INIT_LIST_HEAD(&cache->list);
+ INIT_LIST_HEAD(&cache->cluster_list);
+ INIT_LIST_HEAD(&cache->new_bg_list);
+ btrfs_init_free_space_ctl(cache);
+
+ return cache;
+}
+
int btrfs_read_block_groups(struct btrfs_root *root)
{
struct btrfs_path *path;
break;
if (ret != 0)
goto error;
+
leaf = path->nodes[0];
btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
- cache = kzalloc(sizeof(*cache), GFP_NOFS);
+
+ cache = btrfs_create_block_group_cache(root, found_key.objectid,
+ found_key.offset);
if (!cache) {
ret = -ENOMEM;
goto error;
}
- cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
- GFP_NOFS);
- if (!cache->free_space_ctl) {
- kfree(cache);
- ret = -ENOMEM;
- goto error;
- }
-
- atomic_set(&cache->count, 1);
- spin_lock_init(&cache->lock);
- cache->fs_info = info;
- INIT_LIST_HEAD(&cache->list);
- INIT_LIST_HEAD(&cache->cluster_list);
if (need_clear) {
/*
read_extent_buffer(leaf, &cache->item,
btrfs_item_ptr_offset(leaf, path->slots[0]),
sizeof(cache->item));
- memcpy(&cache->key, &found_key, sizeof(found_key));
+ cache->flags = btrfs_block_group_flags(&cache->item);
key.objectid = found_key.objectid + found_key.offset;
btrfs_release_path(path);
- cache->flags = btrfs_block_group_flags(&cache->item);
- cache->sectorsize = root->sectorsize;
- cache->full_stripe_len = btrfs_full_stripe_len(root,
- &root->fs_info->mapping_tree,
- found_key.objectid);
- btrfs_init_free_space_ctl(cache);
/*
* We need to exclude the super stripes now so that the space
* case.
*/
free_excluded_extents(root, cache);
- kfree(cache->free_space_ctl);
- kfree(cache);
+ btrfs_put_block_group(cache);
goto error;
}
root->fs_info->last_trans_log_full_commit = trans->transid;
- cache = kzalloc(sizeof(*cache), GFP_NOFS);
+ cache = btrfs_create_block_group_cache(root, chunk_offset, size);
if (!cache)
return -ENOMEM;
- cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl),
- GFP_NOFS);
- if (!cache->free_space_ctl) {
- kfree(cache);
- return -ENOMEM;
- }
-
- cache->key.objectid = chunk_offset;
- cache->key.offset = size;
- cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY;
- cache->sectorsize = root->sectorsize;
- cache->fs_info = root->fs_info;
- cache->full_stripe_len = btrfs_full_stripe_len(root,
- &root->fs_info->mapping_tree,
- chunk_offset);
-
- atomic_set(&cache->count, 1);
- spin_lock_init(&cache->lock);
- INIT_LIST_HEAD(&cache->list);
- INIT_LIST_HEAD(&cache->cluster_list);
- INIT_LIST_HEAD(&cache->new_bg_list);
-
- btrfs_init_free_space_ctl(cache);
btrfs_set_block_group_used(&cache->item, bytes_used);
btrfs_set_block_group_chunk_objectid(&cache->item, chunk_objectid);
- cache->flags = type;
btrfs_set_block_group_flags(&cache->item, type);
+ cache->flags = type;
cache->last_byte_to_unpin = (u64)-1;
cache->cached = BTRFS_CACHE_FINISHED;
ret = exclude_super_stripes(root, cache);
* case.
*/
free_excluded_extents(root, cache);
- kfree(cache->free_space_ctl);
- kfree(cache);
+ btrfs_put_block_group(cache);
return ret;
}
* are still on the list after taking the semaphore
*/
list_del_init(&block_group->list);
- if (list_empty(&block_group->space_info->block_groups[index]))
+ if (list_empty(&block_group->space_info->block_groups[index])) {
+ kobject_del(&block_group->space_info->block_group_kobjs[index]);
+ kobject_put(&block_group->space_info->block_group_kobjs[index]);
clear_avail_alloc_bits(root->fs_info, block_group->flags);
+ }
up_write(&block_group->space_info->groups_sem);
if (block_group->cached == BTRFS_CACHE_STARTED)
range->len = trimmed;
return ret;
}
+
+/*
+ * btrfs_{start,end}_write() is similar to mnt_{want, drop}_write(),
+ * they are used to prevent the some tasks writing data into the page cache
+ * by nocow before the subvolume is snapshoted, but flush the data into
+ * the disk after the snapshot creation.
+ */
+void btrfs_end_nocow_write(struct btrfs_root *root)
+{
+ percpu_counter_dec(&root->subv_writers->counter);
+ /*
+ * Make sure counter is updated before we wake up
+ * waiters.
+ */
+ smp_mb();
+ if (waitqueue_active(&root->subv_writers->wait))
+ wake_up(&root->subv_writers->wait);
+}
+
+int btrfs_start_nocow_write(struct btrfs_root *root)
+{
+ if (unlikely(atomic_read(&root->will_be_snapshoted)))
+ return 0;
+
+ percpu_counter_inc(&root->subv_writers->counter);
+ /*
+ * Make sure counter is updated before we check for snapshot creation.
+ */
+ smp_mb();
+ if (unlikely(atomic_read(&root->will_be_snapshoted))) {
+ btrfs_end_nocow_write(root);
+ return 0;
+ }
+ return 1;
+}
projects / cascardo / linux.git / blobdiff