From: Naohiro Aota <naohiro.aota@xxxxxxx>
[ Upstream commit 5f0addf7b89085f8e0a2593faa419d6111612b9b ]
Currently, we use btrfs_inode_{lock,unlock}() to grant an exclusive
writeback of the relocation data inode in
btrfs_zoned_data_reloc_{lock,unlock}(). However, that can cause a deadlock
in the following path.
Thread A takes btrfs_inode_lock() and waits for metadata reservation by
e.g, waiting for writeback:
prealloc_file_extent_cluster()
- btrfs_inode_lock(&inode->vfs_inode, 0);
- btrfs_prealloc_file_range()
...
- btrfs_replace_file_extents()
- btrfs_start_transaction
...
- btrfs_reserve_metadata_bytes()
Thread B (e.g, doing a writeback work) needs to wait for the inode lock to
continue writeback process:
do_writepages
- btrfs_writepages
- extent_writpages
- btrfs_zoned_data_reloc_lock(BTRFS_I(inode));
- btrfs_inode_lock()
The deadlock is caused by relying on the vfs_inode's lock. By using it, we
introduced unnecessary exclusion of writeback and
btrfs_prealloc_file_range(). Also, the lock at this point is useless as we
don't have any dirty pages in the inode yet.
Introduce fs_info->zoned_data_reloc_io_lock and use it for the exclusive
writeback.
Fixes: 35156d852762 ("btrfs: zoned: only allow one process to add pages to a relocation inode")
CC: stable@xxxxxxxxxxxxxxx # 5.16.x: 869f4cdc73f9: btrfs: zoned: encapsulate inode locking for zoned relocation
CC: stable@xxxxxxxxxxxxxxx # 5.16.x
CC: stable@xxxxxxxxxxxxxxx # 5.17
Cc: Johannes Thumshirn <johannes.thumshirn@xxxxxxx>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@xxxxxxx>
Signed-off-by: Naohiro Aota <naohiro.aota@xxxxxxx>
Signed-off-by: David Sterba <dsterba@xxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
fs/btrfs/ctree.h | 1 +
fs/btrfs/disk-io.c | 1 +
fs/btrfs/zoned.h | 4 ++--
3 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index cc72d8981c47..d1838de0b39c 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -1027,6 +1027,7 @@ struct btrfs_fs_info {
*/
spinlock_t relocation_bg_lock;
u64 data_reloc_bg;
+ struct mutex zoned_data_reloc_io_lock;
#ifdef CONFIG_BTRFS_FS_REF_VERIFY
spinlock_t ref_verify_lock;
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 233d894f6feb..909d19656316 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -2914,6 +2914,7 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
mutex_init(&fs_info->reloc_mutex);
mutex_init(&fs_info->delalloc_root_mutex);
mutex_init(&fs_info->zoned_meta_io_lock);
+ mutex_init(&fs_info->zoned_data_reloc_io_lock);
seqlock_init(&fs_info->profiles_lock);
INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
diff --git a/fs/btrfs/zoned.h b/fs/btrfs/zoned.h
index d680c3ee918a..3a826f7c2040 100644
--- a/fs/btrfs/zoned.h
+++ b/fs/btrfs/zoned.h
@@ -330,7 +330,7 @@ static inline void btrfs_zoned_data_reloc_lock(struct btrfs_inode *inode)
struct btrfs_root *root = inode->root;
if (btrfs_is_data_reloc_root(root) && btrfs_is_zoned(root->fs_info))
- btrfs_inode_lock(&inode->vfs_inode, 0);
+ mutex_lock(&root->fs_info->zoned_data_reloc_io_lock);
}
static inline void btrfs_zoned_data_reloc_unlock(struct btrfs_inode *inode)
@@ -338,7 +338,7 @@ static inline void btrfs_zoned_data_reloc_unlock(struct btrfs_inode *inode)
struct btrfs_root *root = inode->root;
if (btrfs_is_data_reloc_root(root) && btrfs_is_zoned(root->fs_info))
- btrfs_inode_unlock(&inode->vfs_inode, 0);
+ mutex_unlock(&root->fs_info->zoned_data_reloc_io_lock);
}
#endif
--
2.35.1
