snapdir: add 'disabled' value to make .zfs inaccessible

In some environments, just making the .zfs control dir hidden from sight
might not be enough. In particular, the following scenarios might
warrant not allowing access at all:
- old snapshots with wrong permissions/ownership
- old snapshots with exploitable setuid/setgid binaries
- old snapshots with sensitive contents

Introducing a new 'disabled' value that not only hides the control dir,
but prevents access to its contents by returning ENOENT solves all of
the above.

The new property value takes advantage of 'iuv' semantics ("ignore
unknown value") to automatically fall back to the old default value when
a pool is accessed by an older version of ZFS that doesn't yet know
about 'disabled' semantics.

I think that technically the zfs_dirlook change is enough to prevent
access, but preventing lookups and dir entries in an already opened .zfs
handle might also be a good idea to prevent races when modifying the
property at runtime.

Add zfs_snapshot_no_setuid parameter to control whether automatically
mounted snapshots have the setuid mount option set or not.

this could be considered a partial fix for one of the scenarios
mentioned in desired.

Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Tino Reichardt <milky-zfs@mcmilk.de>
Signed-off-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Co-authored-by: Fabian Grünbichler <f.gruenbichler@proxmox.com>
Closes #3963
Closes #16587

module/os/linux/zfs/zfs_ctldir.c

@@ -111,6 +111,7 @@ static krwlock_t zfs_snapshot_lock;
  */
 int zfs_expire_snapshot = ZFSCTL_EXPIRE_SNAPSHOT;
 static int zfs_admin_snapshot = 0;
+static int zfs_snapshot_no_setuid = 0;
 
 typedef struct {
 	char		*se_name;	/* full snapshot name */
@@ -807,7 +808,9 @@ zfsctl_root_lookup(struct inode *dip, const char *name, struct inode **ipp,
 	if ((error = zfs_enter(zfsvfs, FTAG)) != 0)
 		return (error);
 
-	if (strcmp(name, "..") == 0) {
+	if (zfsvfs->z_show_ctldir == ZFS_SNAPDIR_DISABLED) {
+		*ipp = NULL;
+	} else if (strcmp(name, "..") == 0) {
 		*ipp = dip->i_sb->s_root->d_inode;
 	} else if (strcmp(name, ZFS_SNAPDIR_NAME) == 0) {
 		*ipp = zfsctl_inode_lookup(zfsvfs, ZFSCTL_INO_SNAPDIR,
@@ -1097,9 +1100,9 @@ zfsctl_snapshot_mount(struct path *path, int flags)
 	zfsvfs_t *zfsvfs;
 	zfsvfs_t *snap_zfsvfs;
 	zfs_snapentry_t *se;
-	char *full_name, *full_path;
+	char *full_name, *full_path, *options;
 	char *argv[] = { "/usr/bin/env", "mount", "-i", "-t", "zfs", "-n",
-	    NULL, NULL, NULL };
+	    "-o", NULL, NULL, NULL, NULL };
 	char *envp[] = { NULL };
 	int error;
 	struct path spath;
@@ -1113,6 +1116,7 @@ zfsctl_snapshot_mount(struct path *path, int flags)
 
 	full_name = kmem_zalloc(ZFS_MAX_DATASET_NAME_LEN, KM_SLEEP);
 	full_path = kmem_zalloc(MAXPATHLEN, KM_SLEEP);
+	options = kmem_zalloc(7, KM_SLEEP);
 
 	error = zfsctl_snapshot_name(zfsvfs, dname(dentry),
 	    ZFS_MAX_DATASET_NAME_LEN, full_name);
@@ -1128,6 +1132,9 @@ zfsctl_snapshot_mount(struct path *path, int flags)
 	    zfsvfs->z_vfs->vfs_mntpoint ? zfsvfs->z_vfs->vfs_mntpoint : "",
 	    dname(dentry));
 
+	snprintf(options, 7, "%s",
+	    zfs_snapshot_no_setuid ? "nosuid" : "suid");
+
 	/*
 	 * Multiple concurrent automounts of a snapshot are never allowed.
 	 * The snapshot may be manually mounted as many times as desired.
@@ -1150,8 +1157,9 @@ zfsctl_snapshot_mount(struct path *path, int flags)
 	 * value from call_usermodehelper() will be (exitcode << 8 + signal).
 	 */
 	dprintf("mount; name=%s path=%s\n", full_name, full_path);
-	argv[6] = full_name;
-	argv[7] = full_path;
+	argv[7] = options;
+	argv[8] = full_name;
+	argv[9] = full_path;
 	error = call_usermodehelper(argv[0], argv, envp, UMH_WAIT_PROC);
 	if (error) {
 		if (!(error & MOUNT_BUSY << 8)) {
@@ -1312,3 +1320,7 @@ MODULE_PARM_DESC(zfs_admin_snapshot, "Enable mkdir/rmdir/mv in .zfs/snapshot");
 
 module_param(zfs_expire_snapshot, int, 0644);
 MODULE_PARM_DESC(zfs_expire_snapshot, "Seconds to expire .zfs/snapshot");
+
+module_param(zfs_snapshot_no_setuid, int, 0644);
+MODULE_PARM_DESC(zfs_snapshot_no_setuid,
+	"Disable setuid/setgid for automounts in .zfs/snapshot");

module/os/linux/zfs/zfs_dir.c

@@ -415,6 +415,9 @@ zfs_dirlook(znode_t *dzp, char *name, znode_t **zpp, int flags,
 			*zpp = zp;
 		rw_exit(&dzp->z_parent_lock);
 	} else if (zfs_has_ctldir(dzp) && strcmp(name, ZFS_CTLDIR_NAME) == 0) {
+		if (ZTOZSB(dzp)->z_show_ctldir == ZFS_SNAPDIR_DISABLED) {
+			return (SET_ERROR(ENOENT));
+		}
 		ip = zfsctl_root(dzp);
 		*zpp = ITOZ(ip);
 	} else {

module/os/linux/zfs/zfs_vfsops.c

@@ -1702,6 +1702,11 @@ zfs_vget(struct super_block *sb, struct inode **ipp, fid_t *fidp)
 	    (object == ZFSCTL_INO_ROOT || object == ZFSCTL_INO_SNAPDIR)) {
 		*ipp = zfsvfs->z_ctldir;
 		ASSERT(*ipp != NULL);
+
+		if (zfsvfs->z_show_ctldir == ZFS_SNAPDIR_DISABLED) {
+			return (SET_ERROR(ENOENT));
+		}
+
 		if (object == ZFSCTL_INO_SNAPDIR) {
 			VERIFY(zfsctl_root_lookup(*ipp, "snapshot", ipp,
 			    0, kcred, NULL, NULL) == 0);

module/os/linux/zfs/zpl_ctldir.c

@@ -57,6 +57,10 @@ zpl_root_iterate(struct file *filp, struct dir_context *ctx)
 	zfsvfs_t *zfsvfs = ITOZSB(file_inode(filp));
 	int error = 0;
 
+	if (zfsvfs->z_show_ctldir == ZFS_SNAPDIR_DISABLED) {
+		return (SET_ERROR(ENOENT));
+	}
+
 	if ((error = zpl_enter(zfsvfs, FTAG)) != 0)
 		return (error);