On Fri 11-01-13 03:56:45, shilong wang wrote:
> From: Wang Shilong <wangsl-fnst@xxxxxxxxxxxxxx>
>
> Add a necessary check when a io error happens.
> If io error happens,free the memory has been allocated and forget buffer head.
>
> Signed-off-by: Wang Shilong <wangsl-fnst@xxxxxxxxxxxxxx>
> ---
> fs/ext2/inode.c | 12 ++++++++++++
> 1 files changed, 12 insertions(+), 0 deletions(-)
>
> diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
> index 6363ac6..661c816 100644
> --- a/fs/ext2/inode.c
> +++ b/fs/ext2/inode.c
> @@ -495,6 +495,10 @@ static int ext2_alloc_branch(struct inode *inode,
> * parent to disk.
> */
> bh = sb_getblk(inode->i_sb, new_blocks[n-1]);
> + if (!bh) {
> + err = -EIO;
> + goto failed;
> + }
I think it should rather be ENOMEM because that's the only real reason
why sb_getblk() can fail...
> branch[n].bh = bh;
> lock_buffer(bh);
> memset(bh->b_data, 0, blocksize);
> @@ -523,6 +527,14 @@ static int ext2_alloc_branch(struct inode *inode,
> }
> *blks = num;
> return err;
> +
> +failed:
> + for (i = 1; i < n; i++)
> + bforget(branch[i].bh);
I don't think we should bforget() the buffer here. They will be cleaned
up later in ext2_get_blocks().
> + for (i = 0; i < indirect_blks; i++)
> + ext2_free_blocks(inode, new_blocks[i], 1);
> + ext2_free_blocks(inode, new_blocks[i], num);
> + return err;
> }
Honza
--
Jan Kara <jack@xxxxxxx>
SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
