[PATCH v2] btrfs: Refactor count handling in btrfs_unpin_free_ino()

[PATCH v2] btrfs: Refactor count handling in btrfs_unpin_free_ino()

[Geert Uytterhoeven]

With gcc 4.1.2:

    fs/btrfs/inode-map.c: In function ‘btrfs_unpin_free_ino’:
    fs/btrfs/inode-map.c:241: warning: ‘count’ may be used uninitialized in this function

While this warning is a false-positive, it can easily be killed by
refactoring the code.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
---
Please review. Compile-tested only.

v2:
  - Rebased.
---
 fs/btrfs/inode-map.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c
index 12fcd8897c3368df..a56f88ea9c735d18 100644
--- a/fs/btrfs/inode-map.c
+++ b/fs/btrfs/inode-map.c
@@ -244,8 +244,6 @@ void btrfs_unpin_free_ino(struct btrfs_root *root)
 		return;
 
 	while (1) {
-		bool add_to_ctl = true;
-
 		spin_lock(rbroot_lock);
 		n = rb_first(rbroot);
 		if (!n) {
@@ -257,15 +255,14 @@ void btrfs_unpin_free_ino(struct btrfs_root *root)
 		BUG_ON(info->bitmap); /* Logic error */
 
 		if (info->offset > root->ino_cache_progress)
-			add_to_ctl = false;
-		else if (info->offset + info->bytes > root->ino_cache_progress)
-			count = root->ino_cache_progress - info->offset + 1;
+			count = 0;
 		else
-			count = info->bytes;
+			count = min(root->ino_cache_progress - info->offset + 1,
+				    info->bytes);
 
 		rb_erase(&info->offset_index, rbroot);
 		spin_unlock(rbroot_lock);
-		if (add_to_ctl)
+		if (count)
 			__btrfs_add_free_space(root->fs_info, ctl,
 					       info->offset, count);
 		kmem_cache_free(btrfs_free_space_cachep, info);
-- 
2.17.1

Re: [PATCH v2] btrfs: Refactor count handling in btrfs_unpin_free_ino()

[David Sterba]

On Fri, Jun 22, 2018 at 09:18:29AM +0200, Geert Uytterhoeven wrote:
> With gcc 4.1.2:
> 
>     fs/btrfs/inode-map.c: In function ‘btrfs_unpin_free_ino’:
>     fs/btrfs/inode-map.c:241: warning: ‘count’ may be used uninitialized in this function
> 
> While this warning is a false-positive, it can easily be killed by
> refactoring the code.
> 
> Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
> ---
> Please review. Compile-tested only.

Reviewed-by: David Sterba <dsterba@suse.com>