Re: [PATCH] tmpfs: security xattr setting on inode creation

["Ware, Ryan R" <ryan.r.ware@xxxxxxxxx>]

On Tue, Feb 28, 2012 at 7:46 AM, Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:

On Fri, 24 Feb 2012 19:19:22 -0800 (PST)
Hugh Dickins <hughd@xxxxxxxxxx> wrote:

> +/*
> + * Callback for security_inode_init_security() for acquiring xattrs.
> + */
> +static int shmem_initxattrs(struct inode *inode,
> +                         const struct xattr *xattr_array,
> +                         void *fs_info)
> +{
> +     struct shmem_inode_info *info = SHMEM_I(inode);
> +     const struct xattr *xattr;
> +     struct shmem_xattr *new_xattr;
> +     size_t len;
> +
> +     for (xattr = xattr_array; xattr->name != NULL; xattr++) {
> +             new_xattr = shmem_xattr_alloc(xattr->value, xattr->value_len);
> +             if (!new_xattr)
> +                     return -ENOMEM;
> +
> +             len = strlen(xattr->name) + 1;
> +             new_xattr->name = kmalloc(XATTR_SECURITY_PREFIX_LEN + len,
> +                                       GFP_KERNEL);
> +             if (!new_xattr->name) {
> +                     kfree(new_xattr);
> +                     return -ENOMEM;
> +             }
> +
> +             memcpy(new_xattr->name, XATTR_SECURITY_PREFIX,
> +                    XATTR_SECURITY_PREFIX_LEN);
> +             memcpy(new_xattr->name + XATTR_SECURITY_PREFIX_LEN,
> +                    xattr->name, len);
> +
> +             spin_lock(&info->lock);
> +             list_add(&new_xattr->list, &info->xattr_list);
> +             spin_unlock(&info->lock);
> +     }
> +
> +     return 0;
> +}

So if there's a kmalloc failure partway through the array, we leave a
partially xattrified inode in place.

Are we sure this is OK?

I'm guessing Jarkko can clean that up a bit.  It wouldn't be a good idea to leave inaccurate data structures laying around during failure cases.

Ryan