Skip to contentSkip to conversation

Introduce vfs_listxattr

PATCH In review Josef Sipek 4 participants 6 messages Oct 9, 2006 – Oct 10, 2006 VFS

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Introduce vfs_listxattr

From: Bill Nottingham <[email protected]>

This patch moves code out of fs/xattr.c:listxattr into a new function - vfs_listxattr. The code for vfs_listxattr was originally submitted by Bill Nottingham <[email protected]> to Unionfs.

fs/xattr.c +21 -12
diff -r 0231458fbb78 fs/xattr.c
@@ -135,6 +135,26 @@ vfs_getxattr(struct dentry *dentry, char
 }
 EXPORT_SYMBOL_GPL(vfs_getxattr);
 
+ssize_t
+vfs_listxattr(struct dentry *d, char *list, size_t size)
+{
+	ssize_t error;
+
+	error = security_inode_listxattr(d);
+	if (error)
+		return error;
+	error = -EOPNOTSUPP;
+	if (d->d_inode->i_op && d->d_inode->i_op->listxattr) {
+		error = d->d_inode->i_op->listxattr(d, list, size);
+	} else {
+		error = security_inode_listsecurity(d->d_inode, list, size);
+		if (size && error > size)
+			error = -ERANGE;
+	}
+	return error;
+}
+EXPORT_SYMBOL_GPL(vfs_listxattr);
+
 int
 vfs_removexattr(struct dentry *dentry, char *name)
 {
@@ -346,17 +366,7 @@ listxattr(struct dentry *d, char __user 
 			return -ENOMEM;
 	}
 
-	error = security_inode_listxattr(d);
-	if (error)
-		goto out;
-	error = -EOPNOTSUPP;
-	if (d->d_inode->i_op && d->d_inode->i_op->listxattr) {
-		error = d->d_inode->i_op->listxattr(d, klist, size);
-	} else {
-		error = security_inode_listsecurity(d->d_inode, klist, size);
-		if (size && error > size)
-			error = -ERANGE;
-	}
+	error = vfs_listxattr(d, klist, size);
 	if (error > 0) {
 		if (size && copy_to_user(list, klist, error))
 			error = -EFAULT;
@@ -365,7 +375,6 @@ listxattr(struct dentry *d, char __user 
 		   than XATTR_LIST_MAX bytes. Not possible. */
 		error = -E2BIG;
 	}
-out:
 	kfree(klist);
 	return error;
 }
include/linux/xattr.h +1 -0
diff -r 0231458fbb78 include/linux/xattr.h
@@ -41,6 +41,7 @@ struct xattr_handler {
 };
 
 ssize_t vfs_getxattr(struct dentry *, char *, void *, size_t);
+ssize_t vfs_listxattr(struct dentry *d, char *list, size_t size);
 int vfs_setxattr(struct dentry *, char *, void *, size_t, int);
 int vfs_removexattr(struct dentry *, char *);
 
Signature · 1 line
UNIX is user-friendly ... it's just selective about who it's friends are
#1 lore raw
Quoted 3 lines · On Mon, 9 Oct 2006 16:10:48 -0400 Josef Sipek <[email protected]> wrote:

This patch moves code out of fs/xattr.c:listxattr into a new function - vfs_listxattr. The code for vfs_listxattr was originally submitted by Bill Nottingham <[email protected]> to Unionfs.

That tells us what the patch does. In general, please be sure to also tell us *why* you prepared a patch.

Does this patch allow unionfs to be loaded into an otherwise unpatched kernel.org kernel? If so, that seems to be a good reason for including this patch into the mainline kernel.

#2 lore raw
Quoted 7 lines · Andrew Morton
Quoted 3 lines · On Mon, 9 Oct 2006 16:10:48 -0400 Josef Sipek <[email protected]> wrote:

This patch moves code out of fs/xattr.c:listxattr into a new function - vfs_listxattr. The code for vfs_listxattr was originally submitted by Bill Nottingham <[email protected]> to Unionfs.

Quoted 6 lines

That tells us what the patch does. In general, please be sure to also tell us *why* you prepared a patch.

Does this patch allow unionfs to be loaded into an otherwise unpatched kernel.org kernel? If so, that seems to be a good reason for including this patch into the mainline kernel.

Generally I'd say that it makes sense. Anything that wants to access the method in question would either have to play with set_fs() or open-code it; neither is good.

It makes sense to localize calls of a method when we have pretty much mandatory framing for it (security_... stuff).

So the only question is whether it makes sense for anything other than syscall itself to access the method in question. AFAICS, the answer's yes...

#3 lore raw
In reply to Andrew Morton
Quoted 7 lines · Andrew Morton
Quoted 3 lines · On Mon, 9 Oct 2006 16:10:48 -0400 Josef Sipek <[email protected]> wrote:

This patch moves code out of fs/xattr.c:listxattr into a new function - vfs_listxattr. The code for vfs_listxattr was originally submitted by Bill Nottingham <[email protected]> to Unionfs.

Quoted 6 lines

That tells us what the patch does. In general, please be sure to also tell us *why* you prepared a patch.

Does this patch allow unionfs to be loaded into an otherwise unpatched kernel.org kernel? If so, that seems to be a good reason for including this patch into the mainline kernel.

Sorry about that. The reason for this submission is to make the listxattr code in fs/xattr.c a little cleaner (as well as to clean up some code in Unionfs.)

Currently, Unionfs has vfs_listxattr defined in its code. I think that's very ugly, and I'd like to see it (re)moved. The logical place to put it, is along side of all the other vfs_*xattr functions. Overall, I think this patch is benefitial for both kernel.org kernel and Unionfs.

Josef "Jeff" Sipek.

Signature · 1 line
Humans were created by water to transport it upward.
#5 lore raw
In reply to Josef Sipek
Quoted 5 lines · Josef Sipek

From: Bill Nottingham <[email protected]>

This patch moves code out of fs/xattr.c:listxattr into a new function - vfs_listxattr. The code for vfs_listxattr was originally submitted by Bill Nottingham <[email protected]> to Unionfs.

Looks fine, thanks. I didn't do it back then because nfsd wasn't using it, but it's an obvious addition to the vfs_* APIs.