When a snapshot is being created, the atomic counter snapshot_force_cow
is incremented to force incoming writes to fallback to COW. This is a
critical mechanism to protect the consistency of the snapshot being taken.

Currently, can_nocow_file_extent() checks this counter only after
performing several checks, most notably the expensive cross-reference
check via btrfs_cross_ref_exist(). btrfs_cross_ref_exist() releases the
path and performs a search in the extent tree or backref cache, which
involves btree traversals and locking overhead.

This patch moves the snapshot_force_cow check to the very beginning of
can_nocow_file_extent().

This reordering is safe and beneficial because:
1. args->writeback_path is invariant for the duration of the call (set
   by caller run_delalloc_nocow).
2. is_freespace_inode is a static property of the inode.
3. The state of snapshot_force_cow is driven by the btrfs_mksnapshot
   process. Checking it earlier does not change the outcome of the NOCOW
   decision, but effectively prunes the expensive code path when a
   fallback to COW is inevitable.

By failing fast when a snapshot is pending, we avoid the unnecessary
overhead of btrfs_cross_ref_exist() and other extent item checks in the
scenario where NOCOW is already known to be impossible.

Signed-off-by: jkchen <jkchen@ugreen.com>
---
 fs/btrfs/inode.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 845164485..3549031f3 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1921,6 +1921,11 @@ static int can_nocow_file_extent(struct btrfs_path *path,
 	int ret = 0;
 	bool nowait = path->nowait;
 
+	/* If there are pending snapshots for this root, we must COW. */
+	if (args->writeback_path && !is_freespace_inode &&
+	    atomic_read(&root->snapshot_force_cow))
+		goto out;
+
 	fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item);
 	extent_type = btrfs_file_extent_type(leaf, fi);
 
@@ -1982,11 +1987,6 @@ static int can_nocow_file_extent(struct btrfs_path *path,
 		path = NULL;
 	}
 
-	/* If there are pending snapshots for this root, we must COW. */
-	if (args->writeback_path && !is_freespace_inode &&
-	    atomic_read(&root->snapshot_force_cow))
-		goto out;
-
 	args->file_extent.num_bytes = min(args->end + 1, extent_end) - args->start;
 	args->file_extent.offset += args->start - key->offset;
 	io_start = args->file_extent.disk_bytenr + args->file_extent.offset;
-- 
2.43.0

On Mon, Feb 16, 2026 at 10:17 PM jkchen <jk.chen1095@gmail.com> wrote:
>
> When a snapshot is being created, the atomic counter snapshot_force_cow
> is incremented to force incoming writes to fallback to COW. This is a
> critical mechanism to protect the consistency of the snapshot being taken.
>
> Currently, can_nocow_file_extent() checks this counter only after
> performing several checks, most notably the expensive cross-reference
> check via btrfs_cross_ref_exist(). btrfs_cross_ref_exist() releases the
> path and performs a search in the extent tree or backref cache, which
> involves btree traversals and locking overhead.
>
> This patch moves the snapshot_force_cow check to the very beginning of
> can_nocow_file_extent().
>
> This reordering is safe and beneficial because:
> 1. args->writeback_path is invariant for the duration of the call (set
>    by caller run_delalloc_nocow).
> 2. is_freespace_inode is a static property of the inode.
> 3. The state of snapshot_force_cow is driven by the btrfs_mksnapshot
>    process. Checking it earlier does not change the outcome of the NOCOW
>    decision, but effectively prunes the expensive code path when a
>    fallback to COW is inevitable.
>
> By failing fast when a snapshot is pending, we avoid the unnecessary
> overhead of btrfs_cross_ref_exist() and other extent item checks in the
> scenario where NOCOW is already known to be impossible.
>
> Signed-off-by: jkchen <jkchen@ugreen.com>

Reviewed-by: Filipe Manana <fdmanana@suse.com>

Looks good, I'll add it to the github for-next branch, thanks.

> ---
>  fs/btrfs/inode.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
> index 845164485..3549031f3 100644
> --- a/fs/btrfs/inode.c
> +++ b/fs/btrfs/inode.c
> @@ -1921,6 +1921,11 @@ static int can_nocow_file_extent(struct btrfs_path *path,
>         int ret = 0;
>         bool nowait = path->nowait;
>
> +       /* If there are pending snapshots for this root, we must COW. */
> +       if (args->writeback_path && !is_freespace_inode &&
> +           atomic_read(&root->snapshot_force_cow))
> +               goto out;
> +
>         fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item);
>         extent_type = btrfs_file_extent_type(leaf, fi);
>
> @@ -1982,11 +1987,6 @@ static int can_nocow_file_extent(struct btrfs_path *path,
>                 path = NULL;
>         }
>
> -       /* If there are pending snapshots for this root, we must COW. */
> -       if (args->writeback_path && !is_freespace_inode &&
> -           atomic_read(&root->snapshot_force_cow))
> -               goto out;
> -
>         args->file_extent.num_bytes = min(args->end + 1, extent_end) - args->start;
>         args->file_extent.offset += args->start - key->offset;
>         io_start = args->file_extent.disk_bytenr + args->file_extent.offset;
> --
> 2.43.0
>
>