[PATCH v2 08/28] sys: __sys_setfsgid(): handle fsid mappings

[Christian Brauner]

Switch setfsgid() to lookup fsids in the fsid mappings. If no fsid mappings are
setup the behavior is unchanged, i.e. fsids are looked up in the id mappings.

A caller can only setfs{g,u}id() to a given id if the id maps to a valid kid in
both the id and fsid maps of the caller's user namespace. This is always the
case when no id mappings and fsid mappings have been written. It is also always
the case when an id mapping has been written which includes the target id and
but no fsid mappings have been written. All non-fsid mapping aware workloads
will thus work just as before.
Requiring a valid mapping for the target id in both the id and fsid mappings of
the container simplifies permission checking for userns visible filesystems
such as proc.

Signed-off-by: Christian Brauner <christian.brauner@xxxxxxxxxx>
---
/* v2 */
- Christian Brauner <christian.brauner@xxxxxxxxxx>:
  - Set unmapped fsid as well.
---
 kernel/sys.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/kernel/sys.c b/kernel/sys.c
index 13f790dbda71..864fa78f25a7 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -849,15 +849,19 @@ long __sys_setfsgid(gid_t gid)
 	const struct cred *old;
 	struct cred *new;
 	gid_t old_fsgid;
-	kgid_t kgid;
+	kgid_t kgid, kfsgid;
 
 	old = current_cred();
-	old_fsgid = from_kgid_munged(old->user_ns, old->fsgid);
+	old_fsgid = from_kfsgid_munged(old->user_ns, old->fsgid);
 
-	kgid = make_kgid(old->user_ns, gid);
+	kgid = make_kfsgid(old->user_ns, gid);
 	if (!gid_valid(kgid))
 		return old_fsgid;
 
+	kfsgid = make_kgid(old->user_ns, gid);
+	if (!gid_valid(kfsgid))
+		return old_fsgid;
+
 	new = prepare_creds();
 	if (!new)
 		return old_fsgid;
@@ -867,6 +871,7 @@ long __sys_setfsgid(gid_t gid)
 	    ns_capable(old->user_ns, CAP_SETGID)) {
 		if (!gid_eq(kgid, old->fsgid)) {
 			new->fsgid = kgid;
+			new->kfsgid = kfsgid;
 			goto change_okay;
 		}
 	}
-- 
2.25.0