linux-kernel - RE: [PATCH] fs: don't call posix_acl_listxattr in generic

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Mon, 4 Sep 2023 20:36:54 +0000
From:   Ondrej Valousek <ondrej.valousek.xm@...esas.com>
To:     Jeff Layton <jlayton@...nel.org>,
        Alexander Viro <viro@...iv.linux.org.uk>,
        Christian Brauner <brauner@...nel.org>
CC:     "trondmy@...merspace.com" <trondmy@...merspace.com>,
        "eggert@...ucla.edu" <eggert@...ucla.edu>,
        "bruno@...sp.org" <bruno@...sp.org>,
        "linux-fsdevel@...r.kernel.org" <linux-fsdevel@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH] fs: don't call posix_acl_listxattr in generic_listxattr

Hi Jeff,

I can confirm that with rawhide kernel 6.5 the error is gone, i.e.

Listxattr() shows only "system.nfs4_acl" attribute on NFSv4 filesystem,

Problem is, that (on the same kernel) getxattr(name,XATTR_NAME_POSIX_ACL_ACCESS, 0,0)
Sets errno to ENODATA where "name" is file on NFSv4.

This is different behavior to the previous versions, i.e. on RHEL8 getxattr() sets errno to ENOTSUP in the same scenario - which is what I'd expect more.

Is the change of the getxattr() behavior expected or not?

Thanks,
Ondrej

-----Original Message-----
From: Jeff Layton <jlayton@...nel.org> 
Sent: Dienstag, 16. Mai 2023 14:47
To: Alexander Viro <viro@...iv.linux.org.uk>; Christian Brauner <brauner@...nel.org>
Cc: trondmy@...merspace.com; eggert@...ucla.edu; bruno@...sp.org; Ondrej Valousek <ondrej.valousek.xm@...esas.com>; linux-fsdevel@...r.kernel.org; linux-kernel@...r.kernel.org
Subject: [PATCH] fs: don't call posix_acl_listxattr in generic_listxattr

Commit f2620f166e2a caused the kernel to start emitting POSIX ACL xattrs for NFSv4 inodes, which it doesn't support. The only other user of generic_listxattr is HFS (classic) and it doesn't support POSIX ACLs either.

Fixes: f2620f166e2a xattr: simplify listxattr helpers
Reported-by: Ondrej Valousek <ondrej.valousek.xm@...esas.com>
Signed-off-by: Jeff Layton <jlayton@...nel.org>
---
 fs/xattr.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)

diff --git a/fs/xattr.c b/fs/xattr.c
index fcf67d80d7f9..e7bbb7f57557 100644
--- a/fs/xattr.c
+++ b/fs/xattr.c
@@ -985,9 +985,16 @@ int xattr_list_one(char **buffer, ssize_t *remaining_size, const char *name)
 	return 0;
 }
 
-/*
+/**
+ * generic_listxattr - run through a dentry's xattr list() operations
+ * @dentry: dentry to list the xattrs
+ * @buffer: result buffer
+ * @buffer_size: size of @buffer
+ *
  * Combine the results of the list() operation from every xattr_handler in the
- * list.
+ * xattr_handler stack.
+ *
+ * Note that this will not include the entries for POSIX ACLs.
  */
 ssize_t
 generic_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size) @@ -996,10 +1003,6 @@ generic_listxattr(struct dentry *dentry, char *buffer, size_t buffer_size)
 	ssize_t remaining_size = buffer_size;
 	int err = 0;
 
-	err = posix_acl_listxattr(d_inode(dentry), &buffer, &remaining_size);
-	if (err)
-		return err;
-
 	for_each_xattr_handler(handlers, handler) {
 		if (!handler->name || (handler->list && !handler->list(dentry)))
 			continue;
--
2.40.1