Re: [PATCH 1/2 linux-next] fs/affs/inode.c: mark new inode before hashtable insertion

From: Jan Kara
Date: Wed May 06 2015 - 03:04:19 EST


On Tue 05-05-15 20:06:09, Fabian Frederick wrote:
> AFFS didn't mark new inode to I_NEW state like other FS
> using insert_inode_locked() during creation.
> This patch also unlocks inode when setup is complete
> or operation failed.
>
> Problem appeared when trying to add tmpfile support.
>
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: Alexander Viro <viro@xxxxxxxxxxxxxxxxxx>
> Cc: Jan Kara <jack@xxxxxxx>
> Suggested-by: Jan Kara <jack@xxxxxxx>
> Signed-off-by: Fabian Frederick <fabf@xxxxxxxxx>
> ---
> fs/affs/inode.c | 2 ++
> fs/affs/namei.c | 6 ++++++
> 2 files changed, 8 insertions(+)
>
> diff --git a/fs/affs/inode.c b/fs/affs/inode.c
> index 1734950..859a9cb 100644
> --- a/fs/affs/inode.c
> +++ b/fs/affs/inode.c
> @@ -323,6 +323,8 @@ affs_new_inode(struct inode *dir)
> AFFS_I(inode)->i_pa_cnt = 0;
> AFFS_I(inode)->i_extcnt = 1;
> AFFS_I(inode)->i_ext_last = ~1;
> + if (insert_inode_locked(inode) < 0)
> + goto err_inode;
>
> insert_inode_hash(inode);
But you need to also remove insert_inode_hash() to avoid double insert.
Otherwise the patch looks good.

Honza

>
> diff --git a/fs/affs/namei.c b/fs/affs/namei.c
> index 181e05b..b183540 100644
> --- a/fs/affs/namei.c
> +++ b/fs/affs/namei.c
> @@ -281,9 +281,11 @@ affs_create(struct inode *dir, struct dentry *dentry, umode_t mode, bool excl)
> error = affs_add_entry(dir, inode, dentry, ST_FILE);
> if (error) {
> clear_nlink(inode);
> + unlock_new_inode(inode);
> iput(inode);
> return error;
> }
> + unlock_new_inode(inode);
> return 0;
> }
>
> @@ -310,9 +312,11 @@ affs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
> if (error) {
> clear_nlink(inode);
> mark_inode_dirty(inode);
> + unlock_new_inode(inode);
> iput(inode);
> return error;
> }
> + unlock_new_inode(inode);
> return 0;
> }
>
> @@ -391,11 +395,13 @@ affs_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
> if (error)
> goto err;
>
> + unlock_new_inode(inode);
> return 0;
>
> err:
> clear_nlink(inode);
> mark_inode_dirty(inode);
> + unlock_new_inode(inode);
> iput(inode);
> return error;
> }
> --
> 1.9.1
>
--
Jan Kara <jack@xxxxxxx>
SUSE Labs, CR
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/