Btrfs devel - Jul 2013 - [patch] btrfs/raid56: fix and cleanup some error paths

The alloc_rbio() frees "raid_map" and "bbio" on error, so
there is a
potential double free bug in raid56_parity_write().  The
raid56_parity_write() and raid56_parity_recover() functions should still
free "raid_map" and "bbio" on error if other errors occur
though, so I
have added some more calls to kfree().

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>

diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c
index 0525e13..0db856c 100644
--- a/fs/btrfs/raid56.c
+++ b/fs/btrfs/raid56.c
@@ -1687,11 +1687,8 @@ int raid56_parity_write(struct btrfs_root *root, struct
bio *bio,
 	struct blk_plug_cb *cb;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
-		kfree(raid_map);
-		kfree(bbio);
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 	bio_list_add(&rbio->bio_list, bio);
 	rbio->bio_list_bytes = bio->bi_size;
 
@@ -2041,9 +2038,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct
bio *bio,
 	int ret;
 
 	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
-	if (IS_ERR(rbio)) {
+	if (IS_ERR(rbio))
 		return PTR_ERR(rbio);
-	}
 
 	rbio->read_rebuild = 1;
 	bio_list_add(&rbio->bio_list, bio);
@@ -2052,6 +2048,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct
bio *bio,
 	rbio->faila = find_logical_bio_stripe(rbio, bio);
 	if (rbio->faila == -1) {
 		BUG();
+		kfree(raid_map);
+		kfree(bbio);
 		kfree(rbio);
 		return -EIO;
 	}
--
To unsubscribe from this list: send the line "unsubscribe
kernel-janitors" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On mon, 22 Jul 2013 09:55:15 +0300, Dan Carpenter wrote:
> The alloc_rbio() frees "raid_map" and "bbio" on error,
so there is a
> potential double free bug in raid56_parity_write().  The
> raid56_parity_write() and raid56_parity_recover() functions should still
> free "raid_map" and "bbio" on error if other errors
occur though, so I
> have added some more calls to kfree().
> 
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
From the viewpoint of the readability, it is better to free raid_map and bbio
in the caller, I think. But it is up to you.

Reviewed-by: Miao Xie <miaox@cn.fujitsu.com>
> 
> diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c
> index 0525e13..0db856c 100644
> --- a/fs/btrfs/raid56.c
> +++ b/fs/btrfs/raid56.c
> @@ -1687,11 +1687,8 @@ int raid56_parity_write(struct btrfs_root *root,
struct bio *bio,
>  	struct blk_plug_cb *cb;
>  
>  	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
> -	if (IS_ERR(rbio)) {
> -		kfree(raid_map);
> -		kfree(bbio);
> +	if (IS_ERR(rbio))
>  		return PTR_ERR(rbio);
> -	}
>  	bio_list_add(&rbio->bio_list, bio);
>  	rbio->bio_list_bytes = bio->bi_size;
>  
> @@ -2041,9 +2038,8 @@ int raid56_parity_recover(struct btrfs_root *root,
struct bio *bio,
>  	int ret;
>  
>  	rbio = alloc_rbio(root, bbio, raid_map, stripe_len);
> -	if (IS_ERR(rbio)) {
> +	if (IS_ERR(rbio))
>  		return PTR_ERR(rbio);
> -	}
>  
>  	rbio->read_rebuild = 1;
>  	bio_list_add(&rbio->bio_list, bio);
> @@ -2052,6 +2048,8 @@ int raid56_parity_recover(struct btrfs_root *root,
struct bio *bio,
>  	rbio->faila = find_logical_bio_stripe(rbio, bio);
>  	if (rbio->faila == -1) {
>  		BUG();
> +		kfree(raid_map);
> +		kfree(bbio);
>  		kfree(rbio);
>  		return -EIO;
>  	}
> --
> To unsubscribe from this list: send the line "unsubscribe
linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
--
To unsubscribe from this list: send the line "unsubscribe
kernel-janitors" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Mon, Jul 22, 2013 at 04:47:00PM +0800, Miao Xie
wrote:
> On mon, 22 Jul 2013 09:55:15 +0300, Dan Carpenter wrote:
> > The alloc_rbio() frees "raid_map" and "bbio" on
error, so there is a
> > potential double free bug in raid56_parity_write().  The
> > raid56_parity_write() and raid56_parity_recover() functions should
still
> > free "raid_map" and "bbio" on error if other
errors occur though, so I
> > have added some more calls to kfree().
> > 
> > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> 
> >From the viewpoint of the readability, it is better to free raid_map
and bbio
> in the caller, I think. But it is up to you.
Normally, yes, you are right, but in this case the code makes sense
as is.  We would have to free it in btrfs_map_bio() but we actually
allocate it in __btrfs_map_block().  It would be just as weird as
the current code.

The current code is not ugly, it''s just that complicated things
require complicated code.

regards,
dan carpenter

--
To unsubscribe from this list: send the line "unsubscribe
kernel-janitors" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html