mirror of
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git
synced 2024-08-26 10:49:33 +00:00
![Chris Mason](/assets/img/avatar_default.png)
Every transaction in btrfs creates a new snapshot, and then schedules the snapshot from the last transaction for deletion. Snapshot deletion works by walking down the btree and dropping the reference counts on each btree block during the walk. If if a given leaf or node has a reference count greater than one, the reference count is decremented and the subtree pointed to by that node is ignored. If the reference count is one, walking continues down into that node or leaf, and the references of everything it points to are decremented. The old code would try to work in small pieces, walking down the tree until it found the lowest leaf or node to free and then returning. This was very friendly to the rest of the FS because it didn't have a huge impact on other operations. But it wouldn't always keep up with the rate that new commits added new snapshots for deletion, and it wasn't very optimal for the extent allocation tree because it wasn't finding leaves that were close together on disk and processing them at the same time. This changes things to walk down to a level 1 node and then process it in bulk. All the leaf pointers are sorted and the leaves are dropped in order based on their extent number. The extent allocation tree and commit code are now fast enough for this kind of bulk processing to work without slowing the rest of the FS down. Overall it does less IO and is better able to keep up with snapshot deletions under high load. Signed-off-by: Chris Mason <chris.mason@oracle.com>
231 lines
5.5 KiB
C
231 lines
5.5 KiB
C
/*
|
|
* Copyright (C) 2008 Oracle. All rights reserved.
|
|
*
|
|
* This program is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU General Public
|
|
* License v2 as published by the Free Software Foundation.
|
|
*
|
|
* This program is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
* General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU General Public
|
|
* License along with this program; if not, write to the
|
|
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
* Boston, MA 021110-1307, USA.
|
|
*/
|
|
|
|
#include <linux/sched.h>
|
|
#include <linux/sort.h>
|
|
#include "ctree.h"
|
|
#include "ref-cache.h"
|
|
#include "transaction.h"
|
|
|
|
/*
|
|
* leaf refs are used to cache the information about which extents
|
|
* a given leaf has references on. This allows us to process that leaf
|
|
* in btrfs_drop_snapshot without needing to read it back from disk.
|
|
*/
|
|
|
|
/*
|
|
* kmalloc a leaf reference struct and update the counters for the
|
|
* total ref cache size
|
|
*/
|
|
struct btrfs_leaf_ref *btrfs_alloc_leaf_ref(struct btrfs_root *root,
|
|
int nr_extents)
|
|
{
|
|
struct btrfs_leaf_ref *ref;
|
|
size_t size = btrfs_leaf_ref_size(nr_extents);
|
|
|
|
ref = kmalloc(size, GFP_NOFS);
|
|
if (ref) {
|
|
spin_lock(&root->fs_info->ref_cache_lock);
|
|
root->fs_info->total_ref_cache_size += size;
|
|
spin_unlock(&root->fs_info->ref_cache_lock);
|
|
|
|
memset(ref, 0, sizeof(*ref));
|
|
atomic_set(&ref->usage, 1);
|
|
INIT_LIST_HEAD(&ref->list);
|
|
}
|
|
return ref;
|
|
}
|
|
|
|
/*
|
|
* free a leaf reference struct and update the counters for the
|
|
* total ref cache size
|
|
*/
|
|
void btrfs_free_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref)
|
|
{
|
|
if (!ref)
|
|
return;
|
|
WARN_ON(atomic_read(&ref->usage) == 0);
|
|
if (atomic_dec_and_test(&ref->usage)) {
|
|
size_t size = btrfs_leaf_ref_size(ref->nritems);
|
|
|
|
BUG_ON(ref->in_tree);
|
|
kfree(ref);
|
|
|
|
spin_lock(&root->fs_info->ref_cache_lock);
|
|
root->fs_info->total_ref_cache_size -= size;
|
|
spin_unlock(&root->fs_info->ref_cache_lock);
|
|
}
|
|
}
|
|
|
|
static struct rb_node *tree_insert(struct rb_root *root, u64 bytenr,
|
|
struct rb_node *node)
|
|
{
|
|
struct rb_node **p = &root->rb_node;
|
|
struct rb_node *parent = NULL;
|
|
struct btrfs_leaf_ref *entry;
|
|
|
|
while (*p) {
|
|
parent = *p;
|
|
entry = rb_entry(parent, struct btrfs_leaf_ref, rb_node);
|
|
|
|
if (bytenr < entry->bytenr)
|
|
p = &(*p)->rb_left;
|
|
else if (bytenr > entry->bytenr)
|
|
p = &(*p)->rb_right;
|
|
else
|
|
return parent;
|
|
}
|
|
|
|
entry = rb_entry(node, struct btrfs_leaf_ref, rb_node);
|
|
rb_link_node(node, parent, p);
|
|
rb_insert_color(node, root);
|
|
return NULL;
|
|
}
|
|
|
|
static struct rb_node *tree_search(struct rb_root *root, u64 bytenr)
|
|
{
|
|
struct rb_node *n = root->rb_node;
|
|
struct btrfs_leaf_ref *entry;
|
|
|
|
while (n) {
|
|
entry = rb_entry(n, struct btrfs_leaf_ref, rb_node);
|
|
WARN_ON(!entry->in_tree);
|
|
|
|
if (bytenr < entry->bytenr)
|
|
n = n->rb_left;
|
|
else if (bytenr > entry->bytenr)
|
|
n = n->rb_right;
|
|
else
|
|
return n;
|
|
}
|
|
return NULL;
|
|
}
|
|
|
|
int btrfs_remove_leaf_refs(struct btrfs_root *root, u64 max_root_gen,
|
|
int shared)
|
|
{
|
|
struct btrfs_leaf_ref *ref = NULL;
|
|
struct btrfs_leaf_ref_tree *tree = root->ref_tree;
|
|
|
|
if (shared)
|
|
tree = &root->fs_info->shared_ref_tree;
|
|
if (!tree)
|
|
return 0;
|
|
|
|
spin_lock(&tree->lock);
|
|
while (!list_empty(&tree->list)) {
|
|
ref = list_entry(tree->list.next, struct btrfs_leaf_ref, list);
|
|
BUG_ON(ref->tree != tree);
|
|
if (ref->root_gen > max_root_gen)
|
|
break;
|
|
if (!xchg(&ref->in_tree, 0)) {
|
|
cond_resched_lock(&tree->lock);
|
|
continue;
|
|
}
|
|
|
|
rb_erase(&ref->rb_node, &tree->root);
|
|
list_del_init(&ref->list);
|
|
|
|
spin_unlock(&tree->lock);
|
|
btrfs_free_leaf_ref(root, ref);
|
|
cond_resched();
|
|
spin_lock(&tree->lock);
|
|
}
|
|
spin_unlock(&tree->lock);
|
|
return 0;
|
|
}
|
|
|
|
/*
|
|
* find the leaf ref for a given extent. This returns the ref struct with
|
|
* a usage reference incremented
|
|
*/
|
|
struct btrfs_leaf_ref *btrfs_lookup_leaf_ref(struct btrfs_root *root,
|
|
u64 bytenr)
|
|
{
|
|
struct rb_node *rb;
|
|
struct btrfs_leaf_ref *ref = NULL;
|
|
struct btrfs_leaf_ref_tree *tree = root->ref_tree;
|
|
again:
|
|
if (tree) {
|
|
spin_lock(&tree->lock);
|
|
rb = tree_search(&tree->root, bytenr);
|
|
if (rb)
|
|
ref = rb_entry(rb, struct btrfs_leaf_ref, rb_node);
|
|
if (ref)
|
|
atomic_inc(&ref->usage);
|
|
spin_unlock(&tree->lock);
|
|
if (ref)
|
|
return ref;
|
|
}
|
|
if (tree != &root->fs_info->shared_ref_tree) {
|
|
tree = &root->fs_info->shared_ref_tree;
|
|
goto again;
|
|
}
|
|
return NULL;
|
|
}
|
|
|
|
/*
|
|
* add a fully filled in leaf ref struct
|
|
* remove all the refs older than a given root generation
|
|
*/
|
|
int btrfs_add_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref,
|
|
int shared)
|
|
{
|
|
int ret = 0;
|
|
struct rb_node *rb;
|
|
struct btrfs_leaf_ref_tree *tree = root->ref_tree;
|
|
|
|
if (shared)
|
|
tree = &root->fs_info->shared_ref_tree;
|
|
|
|
spin_lock(&tree->lock);
|
|
rb = tree_insert(&tree->root, ref->bytenr, &ref->rb_node);
|
|
if (rb) {
|
|
ret = -EEXIST;
|
|
} else {
|
|
atomic_inc(&ref->usage);
|
|
ref->tree = tree;
|
|
ref->in_tree = 1;
|
|
list_add_tail(&ref->list, &tree->list);
|
|
}
|
|
spin_unlock(&tree->lock);
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
* remove a single leaf ref from the tree. This drops the ref held by the tree
|
|
* only
|
|
*/
|
|
int btrfs_remove_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref)
|
|
{
|
|
struct btrfs_leaf_ref_tree *tree;
|
|
|
|
if (!xchg(&ref->in_tree, 0))
|
|
return 0;
|
|
|
|
tree = ref->tree;
|
|
spin_lock(&tree->lock);
|
|
|
|
rb_erase(&ref->rb_node, &tree->root);
|
|
list_del_init(&ref->list);
|
|
|
|
spin_unlock(&tree->lock);
|
|
|
|
btrfs_free_leaf_ref(root, ref);
|
|
return 0;
|
|
}
|