nsfs: support exhaustive file handles

Pidfd file handles are exhaustive meaning they don't require a handle on
another pidfd to pass to open_by_handle_at() so it can derive the
filesystem to decode in. Instead it can be derived from the file
handle itself. The same is possible for namespace file handles.

Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>

fs/fhandle.c

@@ -11,6 +11,7 @@
 #include <linux/personality.h>
 #include <linux/uaccess.h>
 #include <linux/compat.h>
+#include <linux/nsfs.h>
 #include "internal.h"
 #include "mount.h"
 
@@ -189,6 +190,11 @@ static int get_path_anchor(int fd, struct path *root)
 		return 0;
 	}
 
+	if (fd == FD_NSFS_ROOT) {
+		nsfs_get_root(root);
+		return 0;
+	}
+
 	return -EBADF;
 }
 

fs/internal.h

@@ -355,3 +355,4 @@ int anon_inode_getattr(struct mnt_idmap *idmap, const struct path *path,
 int anon_inode_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 		       struct iattr *attr);
 void pidfs_get_root(struct path *path);
+void nsfs_get_root(struct path *path);

fs/nsfs.c

@@ -25,6 +25,14 @@
 
 static struct vfsmount *nsfs_mnt;
 
+static struct path nsfs_root_path = {};
+
+void nsfs_get_root(struct path *path)
+{
+	*path = nsfs_root_path;
+	path_get(path);
+}
+
 static long ns_ioctl(struct file *filp, unsigned int ioctl,
 			unsigned long arg);
 static const struct file_operations ns_file_operations = {
@@ -598,4 +606,6 @@ void __init nsfs_init(void)
 	if (IS_ERR(nsfs_mnt))
 		panic("can't set nsfs up\n");
 	nsfs_mnt->mnt_sb->s_flags &= ~SB_NOUSER;
+	nsfs_root_path.mnt = nsfs_mnt;
+	nsfs_root_path.dentry = nsfs_mnt->mnt_root;
 }

include/uapi/linux/fcntl.h

@@ -111,6 +111,7 @@
 #define PIDFD_SELF_THREAD_GROUP		-10001 /* Current thread group leader. */
 
 #define FD_PIDFS_ROOT			-10002 /* Root of the pidfs filesystem */
+#define FD_NSFS_ROOT			-10003 /* Root of the nsfs filesystem */
 #define FD_INVALID			-10009 /* Invalid file descriptor: -10000 - EBADF = -10009 */
 
 /* Generic flags for the *at(2) family of syscalls. */