[PATCH v5 12/19] btrfs: dedup: Add support for on-disk hash search
From: Qu Wenruo <hidden>
Date: 2016-02-02 03:08:04
Subsystem:
btrfs file system, filesystems (vfs and infrastructure), the rest · Maintainers:
Chris Mason, David Sterba, Alexander Viro, Christian Brauner, Linus Torvalds
Now on-disk backend should be able to search hash now. Signed-off-by: Wang Xiaoguang <redacted> Signed-off-by: Qu Wenruo <redacted> --- fs/btrfs/dedup.c | 133 ++++++++++++++++++++++++++++++++++++++++++++++++------- fs/btrfs/dedup.h | 1 + 2 files changed, 118 insertions(+), 16 deletions(-)
diff --git a/fs/btrfs/dedup.c b/fs/btrfs/dedup.c
index 4dd07b7..c2b6a03 100644
--- a/fs/btrfs/dedup.c
+++ b/fs/btrfs/dedup.c@@ -576,6 +576,79 @@ int btrfs_dedup_disable(struct btrfs_fs_info *fs_info) return ret; } + /* + * Return 0 for not found + * Return >0 for found and set bytenr_ret + * Return <0 for error + */ +static int ondisk_search_hash(struct btrfs_dedup_info *dedup_info, u8 *hash, + u64 *bytenr_ret, u32 *num_bytes_ret) +{ + struct btrfs_path *path; + struct btrfs_key key; + struct btrfs_root *dedup_root = dedup_info->dedup_root; + u8 *buf = NULL; + u64 hash_key; + int hash_len = btrfs_dedup_sizes[dedup_info->hash_type]; + int ret; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + buf = kmalloc(hash_len, GFP_NOFS); + if (!buf) { + ret = -ENOMEM; + goto out; + } + + memcpy(&hash_key, hash + hash_len - 8, 8); + key.objectid = hash_key; + key.type = BTRFS_DEDUP_HASH_ITEM_KEY; + key.offset = (u64)-1; + + ret = btrfs_search_slot(NULL, dedup_root, &key, path, 0, 0); + if (ret < 0) + goto out; + WARN_ON(ret == 0); + while (1) { + struct extent_buffer *node; + struct btrfs_dedup_hash_item *hash_item; + int slot; + + ret = btrfs_previous_item(dedup_root, path, hash_key, + BTRFS_DEDUP_HASH_ITEM_KEY); + if (ret < 0) + goto out; + if (ret > 0) { + ret = 0; + goto out; + } + + node = path->nodes[0]; + slot = path->slots[0]; + btrfs_item_key_to_cpu(node, &key, slot); + + if (key.type != BTRFS_DEDUP_HASH_ITEM_KEY || + memcmp(&key.objectid, hash + hash_len - 8, 8)) + break; + hash_item = btrfs_item_ptr(node, slot, + struct btrfs_dedup_hash_item); + read_extent_buffer(node, buf, (unsigned long)(hash_item + 1), + hash_len); + if (!memcmp(buf, hash, hash_len)) { + ret = 1; + *bytenr_ret = key.offset; + *num_bytes_ret = btrfs_dedup_hash_len(node, hash_item); + break; + } + } +out: + kfree(buf); + btrfs_free_path(path); + return ret; +} + /* * Caller must ensure the corresponding ref head is not being run. */
@@ -606,9 +679,36 @@ inmem_search_hash(struct btrfs_dedup_info *dedup_info, u8 *hash) return NULL; } -static int inmem_search(struct btrfs_dedup_info *dedup_info, - struct inode *inode, u64 file_pos, - struct btrfs_dedup_hash *hash) +/* Wapper for different backends, caller needs to hold dedup_info->lock */ +static inline int generic_search_hash(struct btrfs_dedup_info *dedup_info, + u8 *hash, u64 *bytenr_ret, + u32 *num_bytes_ret) +{ + if (dedup_info->backend == BTRFS_DEDUP_BACKEND_INMEMORY) { + struct inmem_hash *found_hash; + int ret; + + found_hash = inmem_search_hash(dedup_info, hash); + if (found_hash) { + ret = 1; + *bytenr_ret = found_hash->bytenr; + *num_bytes_ret = found_hash->num_bytes; + } else { + ret = 0; + *bytenr_ret = 0; + *num_bytes_ret = 0; + } + return ret; + } else if (dedup_info->backend == BTRFS_DEDUP_BACKEND_ONDISK) { + return ondisk_search_hash(dedup_info, hash, bytenr_ret, + num_bytes_ret); + } + return -EINVAL; +} + +static int generic_search(struct btrfs_dedup_info *dedup_info, + struct inode *inode, u64 file_pos, + struct btrfs_dedup_hash *hash) { int ret; struct btrfs_root *root = BTRFS_I(inode)->root;
@@ -618,9 +718,9 @@ static int inmem_search(struct btrfs_dedup_info *dedup_info, struct btrfs_delayed_ref_head *insert_head; struct btrfs_delayed_data_ref *insert_dref; struct btrfs_qgroup_extent_record *insert_qrecord = NULL; - struct inmem_hash *found_hash; int free_insert = 1; u64 bytenr; + u64 tmp_bytenr; u32 num_bytes; insert_head = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS);
@@ -652,14 +752,9 @@ static int inmem_search(struct btrfs_dedup_info *dedup_info, again: mutex_lock(&dedup_info->lock); - found_hash = inmem_search_hash(dedup_info, hash->hash); - /* If we don't find a duplicated extent, just return. */ - if (!found_hash) { - ret = 0; + ret = generic_search_hash(dedup_info, hash->hash, &bytenr, &num_bytes); + if (ret <= 0) goto out; - } - bytenr = found_hash->bytenr; - num_bytes = found_hash->num_bytes; delayed_refs = &trans->transaction->delayed_refs;
@@ -698,12 +793,17 @@ again: mutex_lock(&dedup_info->lock); /* Search again to ensure the hash is still here */ - found_hash = inmem_search_hash(dedup_info, hash->hash); - if (!found_hash) { - ret = 0; + ret = generic_search_hash(dedup_info, hash->hash, &tmp_bytenr, + &num_bytes); + if (ret <= 0) { mutex_unlock(&head->mutex); goto out; } + if (tmp_bytenr != bytenr) { + mutex_unlock(&head->mutex); + mutex_unlock(&dedup_info->lock); + goto again; + } hash->bytenr = bytenr; hash->num_bytes = num_bytes;
@@ -737,8 +837,9 @@ int btrfs_dedup_search(struct btrfs_dedup_info *dedup_info, if (WARN_ON(!dedup_info || !hash)) return 0; - if (dedup_info->backend == BTRFS_DEDUP_BACKEND_INMEMORY) - ret = inmem_search(dedup_info, inode, file_pos, hash); + if (dedup_info->backend == BTRFS_DEDUP_BACKEND_INMEMORY || + dedup_info->backend == BTRFS_DEDUP_BACKEND_ONDISK) + ret = generic_search(dedup_info, inode, file_pos, hash); /* It's possible hash->bytenr/num_bytenr already changed */ if (ret == 0) {
diff --git a/fs/btrfs/dedup.h b/fs/btrfs/dedup.h
index cf8eae6..1661fe6 100644
--- a/fs/btrfs/dedup.h
+++ b/fs/btrfs/dedup.h@@ -172,6 +172,7 @@ int btrfs_dedup_calc_hash(struct btrfs_dedup_info *dedup_info, * *INCREASED*, and hash->bytenr/num_bytes will record the existing * extent data. * Return 0 for a hash miss. Nothing is done + * Return < 0 for error */ int btrfs_dedup_search(struct btrfs_dedup_info *dedup_info, struct inode *inode, u64 file_pos,
--
2.7.0