Unbreak mount_capable()

In "consolidate the capability checks in sget_{fc,userns}())" the
wrong argument had been passed to mount_capable() by sget_fc().
That mistake had been further obscured later, when switching
mount_capable() to fs_context has moved the calculation of
bogus argument from sget_fc() to mount_capable() itself.  It
should've been fc->user_ns all along.

Screwed-up-by: Al Viro <viro@zeniv.linux.org.uk>
Reported-by: Christian Brauner <christian@brauner.io>
Tested-by: Christian Brauner <christian@brauner.io>
Reviewed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

fs/super.c

@@ -478,13 +478,10 @@ EXPORT_SYMBOL(generic_shutdown_super);
 
 bool mount_capable(struct fs_context *fc)
 {
-	struct user_namespace *user_ns = fc->global ? &init_user_ns
-						    : fc->user_ns;
-
 	if (!(fc->fs_type->fs_flags & FS_USERNS_MOUNT))
 		return capable(CAP_SYS_ADMIN);
 	else
-		return ns_capable(user_ns, CAP_SYS_ADMIN);
+		return ns_capable(fc->user_ns, CAP_SYS_ADMIN);
 }
 
 /**