Commit 0f96f517 authored by Robbie Ko's avatar Robbie Ko Committed by David Sterba

btrfs: incremental send, improve rmdir performance for large directory

Currently when checking if a directory can be deleted, we always check
if all its children have been processed.

Example: A directory with 2,000,000 files was deleted

original: 1994m57.071s
patch:       1m38.554s

[FIX]
Instead of checking all children on all calls to can_rmdir(), we keep
track of the directory index offset of the child last checked in the
last call to can_rmdir(), and then use it as the starting point for
future calls to can_rmdir().
Signed-off-by: default avatarRobbie Ko <robbieko@synology.com>
Reviewed-by: default avatarFilipe Manana <fdmanana@suse.com>
[ update changelog ]
Signed-off-by: default avatarDavid Sterba <dsterba@suse.com>
parent 35c8eda1
...@@ -235,6 +235,7 @@ struct orphan_dir_info { ...@@ -235,6 +235,7 @@ struct orphan_dir_info {
struct rb_node node; struct rb_node node;
u64 ino; u64 ino;
u64 gen; u64 gen;
u64 last_dir_index_offset;
}; };
struct name_cache_entry { struct name_cache_entry {
...@@ -2861,6 +2862,7 @@ add_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino) ...@@ -2861,6 +2862,7 @@ add_orphan_dir_info(struct send_ctx *sctx, u64 dir_ino)
return ERR_PTR(-ENOMEM); return ERR_PTR(-ENOMEM);
odi->ino = dir_ino; odi->ino = dir_ino;
odi->gen = 0; odi->gen = 0;
odi->last_dir_index_offset = 0;
rb_link_node(&odi->node, parent, p); rb_link_node(&odi->node, parent, p);
rb_insert_color(&odi->node, &sctx->orphan_dirs); rb_insert_color(&odi->node, &sctx->orphan_dirs);
...@@ -2916,6 +2918,7 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen, ...@@ -2916,6 +2918,7 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen,
struct btrfs_key found_key; struct btrfs_key found_key;
struct btrfs_key loc; struct btrfs_key loc;
struct btrfs_dir_item *di; struct btrfs_dir_item *di;
struct orphan_dir_info *odi = NULL;
/* /*
* Don't try to rmdir the top/root subvolume dir. * Don't try to rmdir the top/root subvolume dir.
...@@ -2930,6 +2933,11 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen, ...@@ -2930,6 +2933,11 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen,
key.objectid = dir; key.objectid = dir;
key.type = BTRFS_DIR_INDEX_KEY; key.type = BTRFS_DIR_INDEX_KEY;
key.offset = 0; key.offset = 0;
odi = get_orphan_dir_info(sctx, dir);
if (odi)
key.offset = odi->last_dir_index_offset;
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
if (ret < 0) if (ret < 0)
goto out; goto out;
...@@ -2957,30 +2965,33 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen, ...@@ -2957,30 +2965,33 @@ static int can_rmdir(struct send_ctx *sctx, u64 dir, u64 dir_gen,
dm = get_waiting_dir_move(sctx, loc.objectid); dm = get_waiting_dir_move(sctx, loc.objectid);
if (dm) { if (dm) {
struct orphan_dir_info *odi;
odi = add_orphan_dir_info(sctx, dir); odi = add_orphan_dir_info(sctx, dir);
if (IS_ERR(odi)) { if (IS_ERR(odi)) {
ret = PTR_ERR(odi); ret = PTR_ERR(odi);
goto out; goto out;
} }
odi->gen = dir_gen; odi->gen = dir_gen;
odi->last_dir_index_offset = found_key.offset;
dm->rmdir_ino = dir; dm->rmdir_ino = dir;
ret = 0; ret = 0;
goto out; goto out;
} }
if (loc.objectid > send_progress) { if (loc.objectid > send_progress) {
struct orphan_dir_info *odi; odi = add_orphan_dir_info(sctx, dir);
if (IS_ERR(odi)) {
odi = get_orphan_dir_info(sctx, dir); ret = PTR_ERR(odi);
free_orphan_dir_info(sctx, odi); goto out;
}
odi->gen = dir_gen;
odi->last_dir_index_offset = found_key.offset;
ret = 0; ret = 0;
goto out; goto out;
} }
path->slots[0]++; path->slots[0]++;
} }
free_orphan_dir_info(sctx, odi);
ret = 1; ret = 1;
...@@ -3258,13 +3269,16 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm) ...@@ -3258,13 +3269,16 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm)
if (rmdir_ino) { if (rmdir_ino) {
struct orphan_dir_info *odi; struct orphan_dir_info *odi;
u64 gen;
odi = get_orphan_dir_info(sctx, rmdir_ino); odi = get_orphan_dir_info(sctx, rmdir_ino);
if (!odi) { if (!odi) {
/* already deleted */ /* already deleted */
goto finish; goto finish;
} }
ret = can_rmdir(sctx, rmdir_ino, odi->gen, sctx->cur_ino); gen = odi->gen;
ret = can_rmdir(sctx, rmdir_ino, gen, sctx->cur_ino);
if (ret < 0) if (ret < 0)
goto out; goto out;
if (!ret) if (!ret)
...@@ -3275,13 +3289,12 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm) ...@@ -3275,13 +3289,12 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm)
ret = -ENOMEM; ret = -ENOMEM;
goto out; goto out;
} }
ret = get_cur_path(sctx, rmdir_ino, odi->gen, name); ret = get_cur_path(sctx, rmdir_ino, gen, name);
if (ret < 0) if (ret < 0)
goto out; goto out;
ret = send_rmdir(sctx, name); ret = send_rmdir(sctx, name);
if (ret < 0) if (ret < 0)
goto out; goto out;
free_orphan_dir_info(sctx, odi);
} }
finish: finish:
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment