On 07/17/2017 02:58 PM, Vivek Goyal wrote:

On Tue, Jul 11, 2017 at 11:05:11AM -0400, Stefan Berger wrote:

[..]

quoted

+/*
+ * xattr_list_userns_rewrite - Rewrite list of xattr names for user namespaces
+ *                             or determine needed size for attribute list
+ *                             in case size == 0
+ *
+ * In a user namespace we do not present all extended attributes to the
+ * user. We filter out those that are in the list of userns supported xattr.
+ * Besides that we filter out those with @uid=<uid> when there is no mapping
+ * for that uid in the current user namespace.
+ *
+ * @list:        list of 0-byte separated xattr names
+ * @size:        the size of the list; may be 0 to determine needed list size
+ * @list_maxlen: allocated buffer size of list
+ */
+static ssize_t
+xattr_list_userns_rewrite(char *list, ssize_t size, size_t list_maxlen)
+{
+	char *nlist = NULL;
+	size_t s_off, len, nlen;
+	ssize_t d_off;
+	char *name, *newname;
+
+	if (!list || size < 0 || current_user_ns() == &init_user_ns)

size will never be less than 0 here. Only caller calls this function only
if size is >0. So can we remove this?

Correct.

What about case of "!list". So if user space called listxattr(foo, NULL,
0), we want to return the size of buffer as if all the xattrs will be
returned to user space. But in practice we probably will filter out some
xattrs so actually returned string will be smaller than size reported
previously.

This case of size=0 is a problem in userns. Depending on the mapping of 
the userid's the list can expand. A security.foo at uid=100 can become 
security.foo at uid=100000, if the mapping is set up so that uid 100 on the 
host becomes uid 100000 inside the container. So for now we only have 
security.capability and the way I solved this is by allocating a 65k 
buffer when calling from a userns. In this buffer where we gather the 
xattr names and then walk them to determine the size that's needed for 
the buffer by simulating the rewriting. It's not nice but I don't know 
of any other solution.

Looks like that's the intent of "!list" condition here. Just wanted to
make sure, hence asking.

Thanks for asking. I thought I had this case covered, but obviously I 
did not.

BTW, I am testing this with overlayfs and trying to figure out if
switching of creds will create issues. Simple operations like listxattr
and getxattr and setxattr so far worked for me. And reason seems to be
that name transformation we are doing in top layer based on creds of
caller (and not based on creds of mounter).

Vivek

    Stefan

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html