Date: Thu, 25 Oct 2012 10:33:26 -0700
From: Tejun Heo <tj@kernel.org>
To: Aristeu Rozanski <aris@redhat.com>
Cc: linux-kernel@vger.kernel.org, Li Zefan <lizefan@huawei.com>,
        Al Viro <viro@ZenIV.linux.org.uk>, linux-fsdevel@vger.kernel.org,
        Hugh Dickins <hughd@google.com>
Subject: Re: [PATCH] fs: xattr: rewrite simple_xattr_set()
Message-ID: <20121025173326.GH11442@htj.dyndns.org>
References: <20121025152613.GF14085@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20121025152613.GF14085@redhat.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 4843
Lines: 189

(cc'ing Hugh and keeping the whole body)

Hello,

On Thu, Oct 25, 2012 at 11:26:14AM -0400, Aristeu Rozanski wrote:
> The way this function was written is confusing and already caused problems.
> Rewriting it to be easier to understand and maintain.
> 
> Cc: Tejun Heo <tj@kernel.org>
> Cc: Li Zefan <lizefan@huawei.com>
> Cc: Al Viro <viro@ZenIV.linux.org.uk>
> Signed-off-by: Aristeu Rozanski <aris@redhat.com>

Generally looks okay to me but I think the return value from removal
path is wrong.  More below.

> ---
>  fs/xattr.c |  124 +++++++++++++++++++++++++++++++++++++------------------------
>  1 file changed, 76 insertions(+), 48 deletions(-)
> 
> Index: github/fs/xattr.c
> ===================================================================
> --- github.orig/fs/xattr.c	2012-10-23 16:02:41.155857391 -0400
> +++ github/fs/xattr.c	2012-10-25 11:17:15.118197552 -0400
> @@ -842,55 +842,46 @@
>  	return ret;
>  }
>  
> -static int __simple_xattr_set(struct simple_xattrs *xattrs, const char *name,
> -			      const void *value, size_t size, int flags)
> +static struct simple_xattr *__find_xattr(struct simple_xattrs *xattrs,
> +					 const char *name)
>  {
>  	struct simple_xattr *xattr;
> -	struct simple_xattr *new_xattr = NULL;
> -	int err = 0;
> -
> -	/* value == NULL means remove */
> -	if (value) {
> -		new_xattr = simple_xattr_alloc(value, size);
> -		if (!new_xattr)
> -			return -ENOMEM;
> -
> -		new_xattr->name = kstrdup(name, GFP_KERNEL);
> -		if (!new_xattr->name) {
> -			kfree(new_xattr);
> -			return -ENOMEM;
> -		}
> -	}
>  
> -	spin_lock(&xattrs->lock);
>  	list_for_each_entry(xattr, &xattrs->head, list) {
> -		if (!strcmp(name, xattr->name)) {
> -			if (flags & XATTR_CREATE) {
> -				xattr = new_xattr;
> -				err = -EEXIST;
> -			} else if (new_xattr) {
> -				list_replace(&xattr->list, &new_xattr->list);
> -			} else {
> -				list_del(&xattr->list);
> -			}
> -			goto out;
> -		}
> +		if (!strcmp(name, xattr->name))
> +			return xattr;
>  	}
> -	if (flags & XATTR_REPLACE) {
> -		xattr = new_xattr;
> -		err = -ENODATA;
> -	} else {
> -		list_add(&new_xattr->list, &xattrs->head);
> -		xattr = NULL;
> -	}
> -out:
> -	spin_unlock(&xattrs->lock);
> +	return NULL;
> +}
> +
> +static int __simple_xattr_remove(struct simple_xattrs *xattrs,
> +				 const char *name)
> +{
> +	struct simple_xattr *xattr;
> +
> +	xattr = __find_xattr(xattrs, name);
>  	if (xattr) {
> +		list_del(&xattr->list);
>  		kfree(xattr->name);
>  		kfree(xattr);
> +		return 0;
>  	}
> -	return err;
>  
> +	return 1;
> +}

So, it returns 0 on success and 1 on failure, which in itself isn't a
particularly good idea.

> +
> +/*
> + * xattr REMOVE operation for in-memory/pseudo filesystems
> + */
> +int simple_xattr_remove(struct simple_xattrs *xattrs, const char *name)
> +{
> +	int rc;
> +
> +	spin_lock(&xattrs->lock);
> +	rc = __simple_xattr_remove(xattrs, name);
> +	spin_unlock(&xattrs->lock);
> +
> +	return rc;
>  }
>  
>  /**
> @@ -910,17 +901,54 @@
>  int simple_xattr_set(struct simple_xattrs *xattrs, const char *name,
>  		     const void *value, size_t size, int flags)
>  {
> +	struct simple_xattr *found, *new_xattr;
> +	int err = 0;
> +
>  	if (size == 0)
> -		value = ""; /* empty EA, do not remove */
> -	return __simple_xattr_set(xattrs, name, value, size, flags);
> -}
> +		value = ""; /* empty EA */
>  
> -/*
> - * xattr REMOVE operation for in-memory/pseudo filesystems
> - */
> -int simple_xattr_remove(struct simple_xattrs *xattrs, const char *name)
> -{
> -	return __simple_xattr_set(xattrs, name, NULL, 0, XATTR_REPLACE);
> +	/* if value == NULL is specified, remove the item */
> +	if (value == NULL)
> +		return simple_xattr_remove(xattrs, name);

And gets relayed to the caller.

> +
> +	new_xattr = simple_xattr_alloc(value, size);
> +	if (!new_xattr)
> +		return -ENOMEM;
> +
> +	new_xattr->name = kstrdup(name, GFP_KERNEL);
> +	if (!new_xattr->name) {
> +		kfree(new_xattr);
> +		return -ENOMEM;
> +	}
> +
> +	spin_lock(&xattrs->lock);
> +
> +	found = __find_xattr(xattrs, name);
> +	if (found) {
> +		if (flags & XATTR_CREATE) {
> +			err = -EEXIST;
> +			goto free_new;
> +		}
> +
> +		list_replace(&found->list, &new_xattr->list);
> +		kfree(found->name);
> +		kfree(found);
> +	} else {
> +		if (flags & XATTR_REPLACE) {
> +			err = -ENODATA;
> +			goto free_new;
> +		}
> +
> +		list_add_tail(&new_xattr->list, &xattrs->head);
> +	}
> +
> +out:
> +	spin_unlock(&xattrs->lock);
> +	return err;
> +free_new:
> +	kfree(new_xattr->name);
> +	kfree(new_xattr);
> +	goto out;
>  }
>  
>  static bool xattr_is_trusted(const char *name)

-- 
tejun
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/