posix_acl: Inode acl caching fixes

When get_acl() is called for an inode whose ACL is not cached yet, the
get_acl inode operation is called to fetch the ACL from the filesystem.
The inode operation is responsible for updating the cached acl with
set_cached_acl().  This is done without locking at the VFS level, so
another task can call set_cached_acl() or forget_cached_acl() before the
get_acl inode operation gets to calling set_cached_acl(), and then
get_acl's call to set_cached_acl() results in caching an outdate ACL.

Prevent this from happening by setting the cached ACL pointer to a
task-specific sentinel value before calling the get_acl inode operation.
Move the responsibility for updating the cached ACL from the get_acl
inode operations to get_acl().  There, only set the cached ACL if the
sentinel value hasn't changed.

The sentinel values are chosen to have odd values.  Likewise, the value
of ACL_NOT_CACHED is odd.  In contrast, ACL object pointers always have
an even value (ACLs are aligned in memory).  This allows to distinguish
uncached ACLs values from ACL objects.

In addition, switch from guarding inode->i_acl and inode->i_default_acl
upates by the inode->i_lock spinlock to using xchg() and cmpxchg().

Filesystems that do not want ACLs returned from their get_acl inode
operations to be cached must call forget_cached_acl() to prevent the VFS
from doing so.

(Patch written by Al Viro and Andreas Gruenbacher.)

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

1 files changed, 41 insertions, 2 deletions

diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c
index 17c0fa1eccfa..720d92f5abfb 100644
--- a/fs/nfs/nfs3acl.c
+++ b/fs/nfs/nfs3acl.c
@@ -11,6 +11,38 @@
 
 #define NFSDBG_FACILITY	NFSDBG_PROC
 
+/*
+ * nfs3_prepare_get_acl, nfs3_complete_get_acl, nfs3_abort_get_acl: Helpers for
+ * caching get_acl results in a race-free way.  See fs/posix_acl.c:get_acl()
+ * for explanations.
+ */
+static void nfs3_prepare_get_acl(struct posix_acl **p)
+{
+	struct posix_acl *sentinel = uncached_acl_sentinel(current);
+
+	if (cmpxchg(p, ACL_NOT_CACHED, sentinel) != ACL_NOT_CACHED) {
+		/* Not the first reader or sentinel already in place. */
+	}
+}
+
+static void nfs3_complete_get_acl(struct posix_acl **p, struct posix_acl *acl)
+{
+	struct posix_acl *sentinel = uncached_acl_sentinel(current);
+
+	/* Only cache the ACL if our sentinel is still in place. */
+	posix_acl_dup(acl);
+	if (cmpxchg(p, sentinel, acl) != sentinel)
+		posix_acl_release(acl);
+}
+
+static void nfs3_abort_get_acl(struct posix_acl **p)
+{
+	struct posix_acl *sentinel = uncached_acl_sentinel(current);
+
+	/* Remove our sentinel upon failure. */
+	cmpxchg(p, sentinel, ACL_NOT_CACHED);
+}
+
 struct posix_acl *nfs3_get_acl(struct inode *inode, int type)
 {
 	struct nfs_server *server = NFS_SERVER(inode);
@@ -55,6 +87,11 @@ struct posix_acl *nfs3_get_acl(struct inode *inode, int type)
 	if (res.fattr == NULL)
 		return ERR_PTR(-ENOMEM);
 
+	if (args.mask & NFS_ACL)
+		nfs3_prepare_get_acl(&inode->i_acl);
+	if (args.mask & NFS_DFACL)
+		nfs3_prepare_get_acl(&inode->i_default_acl);
+
 	status = rpc_call_sync(server->client_acl, &msg, 0);
 	dprintk("NFS reply getacl: %d\n", status);
 
@@ -89,12 +126,12 @@ struct posix_acl *nfs3_get_acl(struct inode *inode, int type)
 	}
 
 	if (res.mask & NFS_ACL)
-		set_cached_acl(inode, ACL_TYPE_ACCESS, res.acl_access);
+		nfs3_complete_get_acl(&inode->i_acl, res.acl_access);
 	else
 		forget_cached_acl(inode, ACL_TYPE_ACCESS);
 
 	if (res.mask & NFS_DFACL)
-		set_cached_acl(inode, ACL_TYPE_DEFAULT, res.acl_default);
+		nfs3_complete_get_acl(&inode->i_default_acl, res.acl_default);
 	else
 		forget_cached_acl(inode, ACL_TYPE_DEFAULT);
 
@@ -108,6 +145,8 @@ struct posix_acl *nfs3_get_acl(struct inode *inode, int type)
 	}
 
 getout:
+	nfs3_abort_get_acl(&inode->i_acl);
+	nfs3_abort_get_acl(&inode->i_default_acl);
 	posix_acl_release(res.acl_access);
 	posix_acl_release(res.acl_default);
 	nfs_free_fattr(res.fattr);