main.c

/* fuse-overlayfs: Overlay Filesystem in Userspace

   Copyright (C) 2018 Giuseppe Scrivano <giuseppe@scrivano.org>
   Copyright (C) 2018-2020 Red Hat Inc.
   Copyright (C) 2001-2007  Miklos Szeredi <miklos@szeredi.hu>

   This program is free software: you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2 of the License, or
   (at your option) any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#define _GNU_SOURCE
#define FUSE_USE_VERSION 32
#define _FILE_OFFSET_BITS 64

#include <config.h>

#include "fuse-overlayfs.h"

#include <fuse_lowlevel.h>
#include <unistd.h>
#include <stdlib.h>
#include <stdio.h>
#include <stddef.h>
#include <stdbool.h>
#include <stdalign.h>
#include <string.h>
#include <dirent.h>
#include <assert.h>
#include <errno.h>
#include <linux/magic.h>
#include <err.h>
#include <sys/vfs.h>
#include <sys/ioctl.h>
#ifdef HAVE_SYS_SENDFILE_H
#  include <sys/sendfile.h>
#endif

#include <fuse_overlayfs_error.h>

#include <inttypes.h>
#include <fcntl.h>
#include <hash.h>
#include <sys/statvfs.h>
#include <sys/file.h>
#include <signal.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <sys/syscall.h>
#include <sys/sysmacros.h>
#include <sys/xattr.h>
#include <linux/fs.h>
#include <linux/xattr.h>
#include <sys/time.h>
#include <sys/resource.h>
#include <pthread.h>

#include <utils.h>
#include <plugin.h>

#ifndef TEMP_FAILURE_RETRY
#  define TEMP_FAILURE_RETRY(expression) \
    (__extension__({ long int __result;                 \
       do __result = (long int) (expression);           \
       while (__result == -1L && errno == EINTR);       \
       __result; }))
#endif

static bool disable_locking;
static pthread_mutex_t lock;

static int
enter_big_lock ()
{
  if (disable_locking)
    return 0;

  pthread_mutex_lock (&lock);
  return 1;
}

static int
release_big_lock ()
{
  if (disable_locking)
    return 0;

  pthread_mutex_unlock (&lock);
  return 0;
}

static inline void
cleanup_lockp (int *l)
{
  if (*l == 0)
    return;

  pthread_mutex_unlock (&lock);
  *l = 0;
}

#define cleanup_lock __attribute__ ((cleanup (cleanup_lockp)))

#ifndef HAVE_OPEN_BY_HANDLE_AT
struct file_handle
{
  unsigned int handle_bytes; /* Size of f_handle [in, out] */
  int handle_type;           /* Handle type [out] */
  unsigned char f_handle[0]; /* File identifier (sized by
                                caller) [out] */
};

int
open_by_handle_at (int mount_fd, struct file_handle *handle, int flags)
{
  return syscall (SYS_open_by_handle_at, mount_fd, handle, flags);
}
#endif

#ifndef RENAME_NOREPLACE
#  define RENAME_NOREPLACE (1 << 0)
#endif

#ifndef RENAME_EXCHANGE
#  define RENAME_EXCHANGE (1 << 1)
#endif

#ifndef RENAME_WHITEOUT
#  define RENAME_WHITEOUT (1 << 2)
#endif

#define XATTR_PREFIX "user.fuseoverlayfs."
#define ORIGIN_XATTR "user.fuseoverlayfs.origin"
#define OPAQUE_XATTR "user.fuseoverlayfs.opaque"
#define XATTR_CONTAINERS_PREFIX "user.containers."
#define XATTR_CONTAINERS_OVERRIDE_PREFIX "user.containers.override_"
#define UNPRIVILEGED_XATTR_PREFIX "user.overlay."
#define UNPRIVILEGED_OPAQUE_XATTR "user.overlay.opaque"
#define PRIVILEGED_XATTR_PREFIX "trusted.overlay."
#define PRIVILEGED_OPAQUE_XATTR "trusted.overlay.opaque"
#define PRIVILEGED_ORIGIN_XATTR "trusted.overlay.origin"
#define OPAQUE_WHITEOUT ".wh..wh..opq"
#define WHITEOUT_MAX_LEN (sizeof (".wh.") - 1)

#if ! defined FICLONE && defined __linux__
#  define FICLONE _IOW (0x94, 9, int)
#endif

#if defined(__GNUC__) && (__GNUC__ > 4 || __GNUC__ == 4 && __GNUC_MINOR__ >= 6) && ! defined __cplusplus
_Static_assert(sizeof (fuse_ino_t) >= sizeof (uintptr_t),
               "fuse_ino_t too small to hold uintptr_t values!");
#else
struct _uintptr_to_must_hold_fuse_ino_t_dummy_struct
{
  unsigned _uintptr_to_must_hold_fuse_ino_t : ((sizeof (fuse_ino_t) >= sizeof (uintptr_t)) ? 1 : -1);
};
#endif

static uid_t overflow_uid;
static gid_t overflow_gid;

static struct ovl_ino dummy_ino;

struct stats_s
{
  size_t nodes;
  size_t inodes;
};

static volatile struct stats_s stats;

static void
print_stats (int sig)
{
  char fmt[128];
  int l = snprintf (fmt, sizeof (fmt) - 1, "# INODES: %zu\n# NODES: %zu\n", stats.inodes, stats.nodes);
  fmt[l] = '\0';
  (void) write (STDERR_FILENO, fmt, l + 1);
}

static double
get_timeout (struct ovl_data *lo)
{
  return lo->timeout;
}

static const struct fuse_opt ovl_opts[] = {
  { "redirect_dir=%s",
    offsetof (struct ovl_data, redirect_dir), 0 },
  { "context=%s",
    offsetof (struct ovl_data, context), 0 },
  { "lowerdir=%s",
    offsetof (struct ovl_data, lowerdir), 0 },
  { "upperdir=%s",
    offsetof (struct ovl_data, upperdir), 0 },
  { "workdir=%s",
    offsetof (struct ovl_data, workdir), 0 },
  { "uidmapping=%s",
    offsetof (struct ovl_data, uid_str), 0 },
  { "gidmapping=%s",
    offsetof (struct ovl_data, gid_str), 0 },
  { "timeout=%s",
    offsetof (struct ovl_data, timeout_str), 0 },
  { "threaded=%d",
    offsetof (struct ovl_data, threaded), 0 },
  { "fsync=%d",
    offsetof (struct ovl_data, fsync), 1 },
  { "fast_ino=%d",
    offsetof (struct ovl_data, fast_ino_check), 0 },
  { "writeback=%d",
    offsetof (struct ovl_data, writeback), 1 },
  { "noxattrs=%d",
    offsetof (struct ovl_data, disable_xattrs), 1 },
  { "plugins=%s",
    offsetof (struct ovl_data, plugins), 0 },
  { "xattr_permissions=%d",
    offsetof (struct ovl_data, xattr_permissions), 0 },
  { "squash_to_root",
    offsetof (struct ovl_data, squash_to_root), 1 },
  { "squash_to_uid=%d",
    offsetof (struct ovl_data, squash_to_uid), 1 },
  { "squash_to_gid=%d",
    offsetof (struct ovl_data, squash_to_gid), 1 },
  { "ino32_t",
    offsetof (struct ovl_data, ino_t_32), 1 },
  { "static_nlink",
    offsetof (struct ovl_data, static_nlink), 1 },
  { "volatile", /* native overlay supports "volatile" to mean fsync=0.  */
    offsetof (struct ovl_data, volatile_mode), 1 },
  { "noacl",
    offsetof (struct ovl_data, noacl), 1 },
  FUSE_OPT_KEY ("xino=off", 1),
  FUSE_OPT_KEY ("xino=auto", 2),
  FUSE_OPT_KEY ("xino=on", 3),
  FUSE_OPT_END
};

/* The current process has enough privileges to use mknod.  */
static bool can_mknod = true;

/* Kernel definitions.  */

typedef unsigned char u8;
typedef unsigned char uuid_t[16];

/* The type returned by overlay exportfs ops when encoding an ovl_fh handle */
#define OVL_FILEID 0xfb

/* On-disk and in-memory format for redirect by file handle */
struct ovl_fh
{
  u8 version;  /* 0 */
  u8 magic;    /* 0xfb */
  u8 len;      /* size of this header + size of fid */
  u8 flags;    /* OVL_FH_FLAG_* */
  u8 type;     /* fid_type of fid */
  uuid_t uuid; /* uuid of filesystem */
  u8 fid[0];   /* file identifier */
} __packed;

static struct ovl_data *
ovl_data (fuse_req_t req)
{
  return (struct ovl_data *) fuse_req_userdata (req);
}

static unsigned long
get_next_wd_counter ()
{
  static unsigned long counter = 1;
  return counter++;
}

static ino_t
node_to_inode (const struct ovl_node *n)
{
  return (ino_t) n->ino;
}

static struct ovl_ino *
lookup_inode (struct ovl_data *lo, ino_t n)
{
  return (struct ovl_ino *) n;
}

static struct ovl_node *
inode_to_node (struct ovl_data *lo, ino_t n)
{
  return lookup_inode (lo, n)->node;
}

static void
check_writeable_proc ()
{
  struct statfs svfs;
  int ret;

  ret = statfs ("/proc", &svfs);
  if (ret < 0)
    {
      fprintf (stderr, "error stating /proc: %m\n");
      return;
    }

  if (svfs.f_type != PROC_SUPER_MAGIC)
    {
      fprintf (stderr, "invalid file system type found on /proc: %ld, expected %d\n", (long) svfs.f_type, PROC_SUPER_MAGIC);
      return;
    }

  if (svfs.f_flags & ST_RDONLY)
    {
      fprintf (stderr, "/proc seems to be mounted as readonly, it can lead to unexpected failures");
      return;
    }
}

static void
check_can_mknod (struct ovl_data *lo)
{
  int ret;
  char path[PATH_MAX];

  if (getenv ("FUSE_OVERLAYFS_DISABLE_OVL_WHITEOUT"))
    {
      can_mknod = false;
      return;
    }

  sprintf (path, "%lu", get_next_wd_counter ());

  ret = mknodat (lo->workdir_fd, path, S_IFCHR | 0700, makedev (0, 0));
  if (ret == 0)
    unlinkat (lo->workdir_fd, path, 0);
  if (ret < 0 && errno == EPERM)
    can_mknod = false;
}

static struct ovl_mapping *
read_mappings (const char *str)
{
  char *buf = NULL, *saveptr = NULL, *it, *endptr;
  struct ovl_mapping *tmp, *ret = NULL;
  unsigned int a, b, c;
  int state = 0;

  buf = alloca (strlen (str) + 1);
  strcpy (buf, str);

  for (it = strtok_r (buf, ":", &saveptr); it; it = strtok_r (NULL, ":", &saveptr))
    {
      switch (state)
        {
        case 0:
          a = strtol (it, &endptr, 10);
          if (*endptr != 0)
            error (EXIT_FAILURE, 0, "invalid mapping specified: %s", str);
          state++;
          break;

        case 1:
          b = strtol (it, &endptr, 10);
          if (*endptr != 0)
            error (EXIT_FAILURE, 0, "invalid mapping specified: %s", str);
          state++;
          break;

        case 2:
          c = strtol (it, &endptr, 10);
          if (*endptr != 0)
            error (EXIT_FAILURE, 0, "invalid mapping specified: %s", str);
          state = 0;

          tmp = malloc (sizeof (*tmp));
          if (tmp == NULL)
            return NULL;
          tmp->next = ret;
          tmp->host = a;
          tmp->to = b;
          tmp->len = c;
          ret = tmp;
          break;
        }
    }

  if (state != 0)
    error (EXIT_FAILURE, 0, "invalid mapping specified: %s", str);

  return ret;
}

static void
free_mapping (struct ovl_mapping *it)
{
  struct ovl_mapping *next = NULL;
  for (; it; it = next)
    {
      next = it->next;
      free (it);
    }
}

/* Useful in a gdb session.  */
void
dump_directory (struct ovl_node *node)
{
  struct ovl_node *it;

  if (node->children == NULL)
    return;

  for (it = hash_get_first (node->children); it; it = hash_get_next (node->children, it))
    printf ("ENTRY: %s (%s)\n", it->name, it->path);
}

static long int
read_file_as_int (const char *file)
{
  cleanup_close int fd = -1;
  long int ret;
  char buffer[256];
  int r;

  fd = open (file, O_RDONLY);
  if (fd < 0)
    error (EXIT_FAILURE, errno, "can't open %s", file);

  r = read (fd, buffer, sizeof (buffer) - 1);
  if (r < 0)
    error (EXIT_FAILURE, errno, "can't read from %s", file);
  buffer[r] = '\0';

  ret = strtol (buffer, NULL, 10);
  if (ret == 0)
    error (EXIT_FAILURE, errno, "can't parse %s", file);

  return ret;
}

static void
read_overflowids (void)
{
  overflow_uid = read_file_as_int ("/proc/sys/kernel/overflowuid");
  overflow_gid = read_file_as_int ("/proc/sys/kernel/overflowgid");
}

static bool
ovl_debug (fuse_req_t req)
{
  return ovl_data (req)->debug != 0;
}

static void
ovl_init (void *userdata, struct fuse_conn_info *conn)
{
  struct ovl_data *lo = (struct ovl_data *) userdata;

  if ((conn->capable & FUSE_CAP_WRITEBACK_CACHE) == 0)
    lo->writeback = 0;

  if ((lo->noacl == 0) && (conn->capable & FUSE_CAP_POSIX_ACL))
    conn->want |= FUSE_CAP_POSIX_ACL;

  conn->want |= FUSE_CAP_DONT_MASK | FUSE_CAP_SPLICE_READ | FUSE_CAP_SPLICE_WRITE | FUSE_CAP_SPLICE_MOVE;
  if (lo->writeback)
    conn->want |= FUSE_CAP_WRITEBACK_CACHE;
#ifdef FUSE_CAP_NO_EXPORT_SUPPORT
  if (! lo->ino_passthrough && ! lo->nfs_filehandles && (conn->capable & FUSE_CAP_NO_EXPORT_SUPPORT))
    conn->want |= FUSE_CAP_NO_EXPORT_SUPPORT;
#endif
}

static struct ovl_layer *
get_first_layer (struct ovl_data *lo)
{
  return lo->layers;
}

static struct ovl_layer *
get_upper_layer (struct ovl_data *lo)
{
  if (lo->upperdir == NULL)
    return NULL;

  return lo->layers;
}

static struct ovl_layer *
get_lower_layers (struct ovl_data *lo)
{
  if (lo->upperdir == NULL)
    return lo->layers;

  return lo->layers->next;
}

static inline bool
node_dirp (struct ovl_node *n)
{
  return n->children != NULL;
}

static int
node_dirfd (struct ovl_node *n)
{
  if (n->hidden)
    return n->hidden_dirfd;
  return n->layer->fd;
}

static bool
can_access_xattr (const struct ovl_layer *l, const char *name)
{
  return ! (has_prefix (name, XATTR_PREFIX)
            || has_prefix (name, PRIVILEGED_XATTR_PREFIX)
            || has_prefix (name, UNPRIVILEGED_XATTR_PREFIX)
            || (l->stat_override_mode == STAT_OVERRIDE_CONTAINERS && has_prefix (name, XATTR_SECURITY_PREFIX)));
}

static bool
encoded_xattr_name (const struct ovl_layer *l, const char *name)
{
  return has_prefix (name, XATTR_CONTAINERS_OVERRIDE_PREFIX) && ! can_access_xattr (l, name + sizeof (XATTR_CONTAINERS_OVERRIDE_PREFIX) - 1);
}

static const char *
decode_xattr_name (const struct ovl_layer *l, const char *name)
{
  if (encoded_xattr_name (l, name))
    return name + sizeof (XATTR_CONTAINERS_OVERRIDE_PREFIX) - 1;

  if (can_access_xattr (l, name))
    return name;

  return NULL;
}

static const char *
encode_xattr_name (const struct ovl_layer *l, char *buf,
                   const char *name)
{
  if (can_access_xattr (l, name))
    return name;

  if (l->stat_override_mode != STAT_OVERRIDE_CONTAINERS || strlen (name) > XATTR_NAME_MAX + 1 - sizeof (XATTR_CONTAINERS_OVERRIDE_PREFIX))
    return NULL;

  strcpy (buf, XATTR_CONTAINERS_OVERRIDE_PREFIX);
  strcpy (buf + sizeof (XATTR_CONTAINERS_OVERRIDE_PREFIX) - 1, name);

  return buf;
}

static ssize_t
write_permission_xattr (struct ovl_data *lo, int fd, const char *path, uid_t uid, gid_t gid, mode_t mode)
{
  char buf[64];
  size_t len;
  int ret;
  const char *name = NULL;

  switch (get_upper_layer (lo)->stat_override_mode)
    {
    case STAT_OVERRIDE_NONE:
      return 0;

    case STAT_OVERRIDE_USER:
      name = XATTR_OVERRIDE_STAT;
      break;

    case STAT_OVERRIDE_PRIVILEGED:
      name = XATTR_PRIVILEGED_OVERRIDE_STAT;
      break;

    case STAT_OVERRIDE_CONTAINERS:
      name = XATTR_OVERRIDE_CONTAINERS_STAT;
      break;

    default:
      errno = EINVAL;
      return -1;
    }

  if (path == NULL && fd < 0)
    {
      errno = EINVAL;
      return -1;
    }

  len = sprintf (buf, "%d:%d:%o", uid, gid, mode);
  if (fd >= 0)
    return fsetxattr (fd, name, buf, len, 0);

  ret = lsetxattr (path, name, buf, len, 0);
  /* Ignore EPERM in unprivileged mode.  */
  if (ret < 0 && lo->xattr_permissions == 2 && errno == EPERM)
    return 0;
  return ret;
}

static int
do_fchown (struct ovl_data *lo, int fd, uid_t uid, gid_t gid, mode_t mode)
{
  int ret;
  if (lo->xattr_permissions)
    ret = write_permission_xattr (lo, fd, NULL, uid, gid, mode);
  else
    ret = fchown (fd, uid, gid);
  return (lo->squash_to_root || lo->squash_to_uid != -1 || lo->squash_to_gid != -1) ? 0 : ret;
}
/* Make sure it is not used anymore.  */
#define fchown ERROR

static int
do_chown (struct ovl_data *lo, const char *path, uid_t uid, gid_t gid, mode_t mode)
{
  int ret;
  if (lo->xattr_permissions)
    ret = write_permission_xattr (lo, -1, path, uid, gid, mode);
  else
    ret = chown (path, uid, gid);
  return (lo->squash_to_root || lo->squash_to_uid != -1 || lo->squash_to_gid != -1) ? 0 : ret;
}
/* Make sure it is not used anymore.  */
#define chown ERROR

static int
do_fchownat (struct ovl_data *lo, int dfd, const char *path, uid_t uid, gid_t gid, mode_t mode, int flags)
{
  int ret;
  if (lo->xattr_permissions)
    {
      char proc_path[32];
      cleanup_close int fd = openat (dfd, path, O_NOFOLLOW | O_PATH);
      if (fd < 0)
        return fd;

      sprintf (proc_path, "/proc/self/fd/%d", fd);
      ret = write_permission_xattr (lo, -1, proc_path, uid, gid, mode);
    }
  else
    ret = fchownat (dfd, path, uid, gid, flags);
  return (lo->squash_to_root || lo->squash_to_uid != -1 || lo->squash_to_gid != -1) ? 0 : ret;
}
/* Make sure it is not used anymore.  */
#define fchownat ERROR

static int
do_stat (struct ovl_node *node, int fd, const char *path, struct stat *st)
{
  struct ovl_layer *l = node->layer;

  if (fd >= 0)
    return l->ds->fstat (l, fd, path, STATX_BASIC_STATS, st);

  if (path != NULL)
    return stat (path, st);

  if (node->hidden)
    return fstatat (node_dirfd (node), node->path, st, AT_SYMLINK_NOFOLLOW);

  return l->ds->statat (l, node->path, st, AT_SYMLINK_NOFOLLOW, STATX_BASIC_STATS);
}

static int
do_fchmod (struct ovl_data *lo, struct ovl_node *node, int fd, mode_t mode)
{
  if (lo->xattr_permissions)
    {
      struct stat st;

      st.st_uid = 0;
      st.st_gid = 0;
      if (do_stat (node, fd, NULL, &st) < 0)
        return -1;

      return write_permission_xattr (lo, fd, NULL, st.st_uid, st.st_gid, mode);
    }
  return fchmod (fd, mode);
}
/* Make sure it is not used anymore.  */
#define fchmod ERROR

static int
do_chmod (struct ovl_data *lo, struct ovl_node *node, const char *path, mode_t mode)
{
  if (lo->xattr_permissions)
    {
      struct stat st;

      st.st_uid = 0;
      st.st_gid = 0;
      if (do_stat (node, -1, path, &st) < 0)
        return -1;

      return write_permission_xattr (lo, -1, path, st.st_uid, st.st_gid, mode);
    }
  return chmod (path, mode);
}
/* Make sure it is not used anymore.  */
#define chmod ERROR

static int
set_fd_origin (int fd, const char *origin)
{
  cleanup_close int opq_whiteout_fd = -1;
  size_t len = strlen (origin) + 1;
  int ret;

  ret = fsetxattr (fd, ORIGIN_XATTR, origin, len, 0);
  if (ret < 0)
    {
      if (errno == ENOTSUP)
        return 0;
    }
  return ret;
}

static int
set_fd_opaque (int fd)
{
  cleanup_close int opq_whiteout_fd = -1;
  int ret;

  ret = fsetxattr (fd, PRIVILEGED_OPAQUE_XATTR, "y", 1, 0);
  if (ret < 0)
    {
      if (errno == ENOTSUP)
        goto create_opq_whiteout;
      if (errno != EPERM || (fsetxattr (fd, OPAQUE_XATTR, "y", 1, 0) < 0 && errno != ENOTSUP))
        return -1;
    }
create_opq_whiteout:
  opq_whiteout_fd = TEMP_FAILURE_RETRY (safe_openat (fd, OPAQUE_WHITEOUT, O_CREAT | O_WRONLY | O_NONBLOCK, 0700));
  return (opq_whiteout_fd >= 0 || ret == 0) ? 0 : -1;
}

static int
is_directory_opaque (struct ovl_layer *l, const char *path)
{
  char b[16];
  ssize_t s;

  s = l->ds->getxattr (l, path, PRIVILEGED_OPAQUE_XATTR, b, sizeof (b));
  if (s < 0 && errno == ENODATA)
    s = l->ds->getxattr (l, path, UNPRIVILEGED_OPAQUE_XATTR, b, sizeof (b));
  if (s < 0 && errno == ENODATA)
    s = l->ds->getxattr (l, path, OPAQUE_XATTR, b, sizeof (b));

  if (s < 0)
    {
      if (errno == ENOTSUP || errno == ENODATA)
        {
          char whiteout_opq_path[PATH_MAX];

          strconcat3 (whiteout_opq_path, PATH_MAX, path, "/" OPAQUE_WHITEOUT, NULL);

          if (l->ds->file_exists (l, whiteout_opq_path) == 0)
            return 1;

          return (errno == ENOENT) ? 0 : -1;
        }
      return -1;
    }

  return b[0] == 'y' ? 1 : 0;
}

static int
create_whiteout (struct ovl_data *lo, struct ovl_node *parent, const char *name, bool skip_mknod, bool force_create)
{
  char whiteout_wh_path[PATH_MAX];
  cleanup_close int fd = -1;
  int ret;

  if (! force_create)
    {
      char path[PATH_MAX];
      struct ovl_layer *l;
      bool found = false;

      strconcat3 (path, PATH_MAX, parent->path, "/", name);

      for (l = get_lower_layers (lo); l; l = l->next)
        {
          ret = l->ds->file_exists (l, path);
          if (ret < 0 && errno == ENOENT)
            continue;

          found = true;
          break;
        }
      /* Not present in the lower layers, do not do anything.  */
      if (! found)
        return 0;
    }

  if (! skip_mknod && can_mknod)
    {
      char whiteout_path[PATH_MAX];

      strconcat3 (whiteout_path, PATH_MAX, parent->path, "/", name);

      ret = mknodat (get_upper_layer (lo)->fd, whiteout_path, S_IFCHR | 0700, makedev (0, 0));
      if (ret == 0)
        return 0;

      if (errno == EEXIST)
        {
          int saved_errno = errno;
          struct stat st;

          /* Check whether it is already a whiteout.  */
          if (TEMP_FAILURE_RETRY (fstatat (get_upper_layer (lo)->fd, whiteout_path, &st, AT_SYMLINK_NOFOLLOW)) == 0
              && (st.st_mode & S_IFMT) == S_IFCHR
              && major (st.st_rdev) == 0
              && minor (st.st_rdev) == 0)
            return 0;

          errno = saved_errno;
        }

      if (errno != EPERM && errno != ENOTSUP)
        return -1;

      /* if it fails with EPERM then do not attempt mknod again.  */
      can_mknod = false;
    }

  strconcat3 (whiteout_wh_path, PATH_MAX, parent->path, "/.wh.", name);

  fd = get_upper_layer (lo)->ds->openat (get_upper_layer (lo), whiteout_wh_path, O_CREAT | O_WRONLY | O_NONBLOCK, 0700);
  if (fd < 0 && errno != EEXIST)
    return -1;

  return 0;
}

static int
delete_whiteout (struct ovl_data *lo, int dirfd, struct ovl_node *parent, const char *name)
{
  struct stat st;

  if (can_mknod)
    {
      if (dirfd >= 0)
        {
          if (TEMP_FAILURE_RETRY (fstatat (dirfd, name, &st, AT_SYMLINK_NOFOLLOW)) == 0
              && (st.st_mode & S_IFMT) == S_IFCHR
              && major (st.st_rdev) == 0
              && minor (st.st_rdev) == 0)
            {
              if (unlinkat (dirfd, name, 0) < 0)
                return -1;
            }
        }
      else
        {
          char whiteout_path[PATH_MAX];

          strconcat3 (whiteout_path, PATH_MAX, parent->path, "/", name);

          if (get_upper_layer (lo)->ds->statat (get_upper_layer (lo), whiteout_path, &st, AT_SYMLINK_NOFOLLOW, STATX_MODE | STATX_TYPE) == 0
              && (st.st_mode & S_IFMT) == S_IFCHR
              && major (st.st_rdev) == 0
              && minor (st.st_rdev) == 0)
            {
              if (unlinkat (get_upper_layer (lo)->fd, whiteout_path, 0) < 0)
                return -1;
            }
        }
    }

  /* Look for the .wh. alternative as well.  */

  if (dirfd >= 0)
    {
      char whiteout_path[PATH_MAX];

      strconcat3 (whiteout_path, PATH_MAX, ".wh.", name, NULL);

      if (unlinkat (dirfd, whiteout_path, 0) < 0 && errno != ENOENT)
        return -1;
    }
  else
    {
      char whiteout_path[PATH_MAX];

      strconcat3 (whiteout_path, PATH_MAX, parent->path, "/.wh.", name);

      if (unlinkat (get_upper_layer (lo)->fd, whiteout_path, 0) < 0 && errno != ENOENT)
        return -1;
    }

  return 0;
}

static unsigned int
find_mapping (unsigned int id, const struct ovl_data *data,
              bool direct, bool uid)
{
  const struct ovl_mapping *mapping = (uid ? data->uid_mappings
                                           : data->gid_mappings);

  if (direct && uid && data->squash_to_uid != -1)
    return data->squash_to_uid;
  if (direct && ! uid && data->squash_to_gid != -1)
    return data->squash_to_gid;
  if (direct && data->squash_to_root)
    return 0;

  if (mapping == NULL)
    return id;
  for (; mapping; mapping = mapping->next)
    {
      if (direct)
        {
          if (id >= mapping->host && id < mapping->host + mapping->len)
            return mapping->to + (id - mapping->host);
        }
      else
        {
          if (id >= mapping->to && id < mapping->to + mapping->len)
            return mapping->host + (id - mapping->to);
        }
    }
  return uid ? overflow_uid : overflow_gid;
}

static uid_t
get_uid (struct ovl_data *data, uid_t id)
{
  return find_mapping (id, data, false, true);
}

static uid_t
get_gid (struct ovl_data *data, gid_t id)
{
  return find_mapping (id, data, false, false);
}

static ino_t
get_st_ino (const struct ovl_node *node, const struct ovl_data *data)
{
  const struct ovl_layer *l = node->layer;

  uint64_t ino;

  if (data->ino_passthrough)
    ino = node->tmp_ino;
  else if (data->nfs_filehandles)
    ino = l->ds->get_nfs_filehandle (l, node->path);
  else
    ino = node_to_inode (node) >> __builtin_ctz (alignof (max_align_t));

  if (data->ino_t_32)
    return (uint32_t) ino;

  return ino;
}

static int
rpl_stat (fuse_req_t req, struct ovl_node *node, int fd, const char *path, struct stat *st_in, struct stat *st)
{
  int ret = 0;
  struct ovl_layer *l = node->layer;
  struct ovl_data *data = ovl_data (req);

  if (st_in)
    memcpy (st, st_in, sizeof (*st));
  else
    ret = do_stat (node, fd, path, st);

  if (ret < 0)
    return ret;

  st->st_uid = find_mapping (st->st_uid, data, true, true);
  st->st_gid = find_mapping (st->st_gid, data, true, false);
  st->st_ino = get_st_ino (node, data);

  if (node->loaded && node->n_links > 0)
    st->st_nlink = node->n_links;
  else if (node_dirp (node))
    {
      if (data->static_nlink)
        st->st_nlink = 1;
      else
        {
          struct ovl_node *it;

          st->st_nlink = 2;

          for (it = hash_get_first (node->children); it; it = hash_get_next (node->children, it))
            {
              if (node_dirp (it))
                st->st_nlink++;
            }
        }

      node->n_links = st->st_nlink;
    }
  else
    {
      struct ovl_node *n;

      st->st_nlink = 0;
      for (n = node->ino->node; n; n = n->next_link)
        st->st_nlink++;
    }

  return ret;
}

static void
node_mark_all_free (void *p)
{
  struct ovl_node *it, *n = (struct ovl_node *) p;

  for (it = n->next_link; it; it = it->next_link)
    it->ino->lookups = 0;

  n->ino->lookups = 0;

  if (n->children)
    {
      for (it = hash_get_first (n->children); it; it = hash_get_next (n->children, it))
        node_mark_all_free (it);
    }
}

static void
node_free (void *p)
{
  struct ovl_node *n = (struct ovl_node *) p;

  if (n->parent)
    {
      if (n->parent->children && hash_lookup (n->parent->children, n) == n)
        hash_remove (n->parent->children, n);
      n->parent->loaded = 0;
      n->parent = NULL;
    }

  if ((n->ino && n->ino != &dummy_ino) || n->node_lookups > 0)
    return;

  if (n->children)
    {
      struct ovl_node *it;

      for (it = hash_get_first (n->children); it; it = hash_get_next (n->children, it))
        it->parent = NULL;

      hash_free (n->children);
      n->children = NULL;
    }

  if (n->do_unlink)
    unlinkat (n->hidden_dirfd, n->path, 0);
  if (n->do_rmdir)
    unlinkat (n->hidden_dirfd, n->path, AT_REMOVEDIR);

  stats.nodes--;
  free (n->name);
  free (n->path);
  free (n);
}

static void
inode_free (void *p)
{
  struct ovl_node *n, *tmp;
  struct ovl_ino *i = (struct ovl_ino *) p;

  n = i->node;
  while (n)
    {
      tmp = n;
      n = n->next_link;

      tmp->ino = NULL;
      node_free (tmp);
    }

  stats.inodes--;
  free (i);
}

static void
drop_node_from_ino (Hash_table *inodes, struct ovl_node *node)
{
  struct ovl_ino *ino;
  struct ovl_node *it, *prev = NULL;

  ino = node->ino;

  if (ino->lookups == 0)
    {
      hash_remove (inodes, ino);
      inode_free (ino);
      return;
    }

  /* If it is the only node referenced by the inode, do not destroy it.  */
  if (ino->node == node && node->next_link == NULL)
    return;

  node->ino = NULL;

  for (it = ino->node; it; it = it->next_link)
    {
      if (it == node)
        {
          if (prev)
            prev->next_link = it->next_link;
          else
            ino->node = it->next_link;
          break;
        }
      prev = it;
    }
}

static int
direct_renameat2 (int olddirfd, const char *oldpath,
                  int newdirfd, const char *newpath, unsigned int flags)
{
  return syscall (SYS_renameat2, olddirfd, oldpath, newdirfd, newpath, flags);
}

static int
hide_node (struct ovl_data *lo, struct ovl_node *node, bool unlink_src)
{
  cleanup_free char *newpath = NULL;
  int ret;

  ret = asprintf (&newpath, "%lu", get_next_wd_counter ());
  if (ret < 0)
    return ret;

  assert (node->layer == get_upper_layer (lo));

  if (unlink_src)
    {
      bool moved = false;
      bool whiteout_created = false;
      bool needs_whiteout;

      needs_whiteout = (node->last_layer != get_upper_layer (lo)) && (node->parent && node->parent->last_layer != get_upper_layer (lo));
      if (! needs_whiteout && node_dirp (node))
        {
          ret = is_directory_opaque (get_upper_layer (lo), node->path);
          if (ret < 0)
            return ret;
          if (ret)
            needs_whiteout = true;
        }

      // if the parent directory is opaque, there's no need to put a whiteout in it.
      if (node->parent != NULL)
        needs_whiteout = needs_whiteout && (is_directory_opaque (get_upper_layer (lo), node->parent->path) < 1);

      if (needs_whiteout)
        {
          /* If the atomic rename+mknod failed, then fallback into doing it in two steps.  */
          if (can_mknod && syscall (SYS_renameat2, node_dirfd (node), node->path, lo->workdir_fd, newpath, RENAME_WHITEOUT) == 0)
            {
              whiteout_created = true;
              moved = true;
            }

          if (! whiteout_created)
            {
              if (node->parent)
                {
                  /* If we are here, it means we have no permissions to use mknod.  Also
                     since the file is not yet moved, creating a whiteout would fail on
                     the mknodat call.  */
                  if (create_whiteout (lo, node->parent, node->name, true, false) < 0)
                    return -1;
                }
            }
        }

      if (! moved)
        {
          if (renameat (node_dirfd (node), node->path, lo->workdir_fd, newpath) < 0)
            return -1;
        }
    }
  else
    {
      if (node_dirp (node))
        {
          if (mkdirat (lo->workdir_fd, newpath, 0700) < 0)
            return -1;
        }
      else
        {
          if (linkat (node_dirfd (node), node->path, lo->workdir_fd, newpath, 0) < 0)
            return -1;
        }
    }
  drop_node_from_ino (lo->inodes, node);

  node->hidden_dirfd = lo->workdir_fd;
  free (node->path);
  node->path = newpath;
  newpath = NULL; /* Do not auto cleanup.  */

  node->hidden = 1;
  if (node->parent)
    node->parent->loaded = 0;
  node->parent = NULL;

  if (node_dirp (node))
    node->do_rmdir = 1;
  else
    node->do_unlink = 1;
  return 0;
}

static size_t
node_inode_hasher (const void *p, size_t s)
{
  struct ovl_ino *n = (struct ovl_ino *) p;

  return (n->ino ^ n->dev) % s;
}

static bool
node_inode_compare (const void *n1, const void *n2)
{
  struct ovl_ino *i1 = (struct ovl_ino *) n1;
  struct ovl_ino *i2 = (struct ovl_ino *) n2;

  return i1->ino == i2->ino && i1->dev == i2->dev;
}

static size_t
node_hasher (const void *p, size_t s)
{
  struct ovl_node *n = (struct ovl_node *) p;
  return n->name_hash % s;
}

static bool
node_compare (const void *n1, const void *n2)
{
  struct ovl_node *node1 = (struct ovl_node *) n1;
  struct ovl_node *node2 = (struct ovl_node *) n2;

  if (node1->name_hash != node2->name_hash)
    return false;

  return strcmp (node1->name, node2->name) == 0 ? true : false;
}

static struct ovl_node *
register_inode (struct ovl_data *lo, struct ovl_node *n, mode_t mode)
{
  struct ovl_ino key;
  struct ovl_ino *ino = NULL;

  key.ino = n->tmp_ino;
  key.dev = n->tmp_dev;

  /* Already registered.  */
  if (n->ino)
    return n;

  ino = hash_lookup (lo->inodes, &key);
  if (ino)
    {
      struct ovl_node *it;

      for (it = ino->node; it; it = it->next_link)
        {
          if (node_dirp (it) || strcmp (n->path, it->path) == 0)
            {
              node_free (n);
              return it;
            }
        }

      n->next_link = ino->node;
      ino->node = n;
      ino->mode = mode;
      n->ino = ino;
      return n;
    }

  ino = calloc (1, sizeof (*ino));
  if (ino == NULL)
    return NULL;

  ino->ino = n->tmp_ino;
  ino->dev = n->tmp_dev;
  ino->node = n;
  n->ino = ino;
  ino->mode = mode;

  if (hash_insert (lo->inodes, ino) == NULL)
    {
      free (ino);
      node_free (n);
      return NULL;
    }

  stats.inodes++;
  return ino->node;
}

static bool
do_forget (struct ovl_data *lo, fuse_ino_t ino, uint64_t nlookup)
{
  struct ovl_ino *i;

  if (ino == FUSE_ROOT_ID || ino == 0)
    return false;

  i = lookup_inode (lo, ino);
  if (i == NULL || i == &dummy_ino)
    return false;

  i->lookups -= nlookup;
  if (i->lookups <= 0)
    {
      hash_remove (lo->inodes, i);
      inode_free (i);
    }
  return true;
}

/* cleanup any inode that has 0 lookups.  */
static void
cleanup_inodes (struct ovl_data *lo)
{
  cleanup_free struct ovl_ino **to_cleanup = NULL;
  size_t no_lookups = 0;
  struct ovl_ino *it;
  size_t i;

  /* Also attempt to cleanup any inode that has 0 lookups.  */
  for (it = hash_get_first (lo->inodes); it; it = hash_get_next (lo->inodes, it))
    {
      if (it->lookups == 0)
        no_lookups++;
    }
  if (no_lookups > 0)
    {
      to_cleanup = malloc (sizeof (*to_cleanup) * no_lookups);
      if (! to_cleanup)
        return;

      for (i = 0, it = hash_get_first (lo->inodes); it; it = hash_get_next (lo->inodes, it))
        {
          if (it->lookups == 0)
            to_cleanup[i++] = it;
        }
      for (i = 0; i < no_lookups; i++)
        do_forget (lo, (fuse_ino_t) to_cleanup[i], 0);
    }
}

static void
ovl_forget (fuse_req_t req, fuse_ino_t ino, uint64_t nlookup)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_forget(ino=%" PRIu64 ", nlookup=%lu)\n",
             ino, nlookup);
  do_forget (lo, ino, nlookup);
  fuse_reply_none (req);
}

static void
ovl_forget_multi (fuse_req_t req, size_t count, struct fuse_forget_data *forgets)
{
  size_t i;
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_forget_multi(count=%zu, forgets=%p)\n",
             count, forgets);

  for (i = 0; i < count; i++)
    do_forget (lo, forgets[i].ino, forgets[i].nlookup);

  cleanup_inodes (lo);

  fuse_reply_none (req);
}

static inline void
cleanup_node_initp (struct ovl_node **p)
{
  struct ovl_node *n = *p;
  if (n == NULL)
    return;
  if (n->children)
    hash_free (n->children);
  free (n->name);
  free (n->path);
  free (n);
}

#define cleanup_node_init __attribute__ ((cleanup (cleanup_node_initp)))

static void
node_set_name (struct ovl_node *node, char *name)
{
  node->name = name;
  if (name == NULL)
    node->name_hash = 0;
  else
    node->name_hash = hash_string (name, SIZE_MAX);
}

static struct ovl_node *
make_whiteout_node (const char *path, const char *name)
{
  cleanup_node_init struct ovl_node *ret = NULL;
  struct ovl_node *ret_xchg;
  char *new_name;

  ret = calloc (1, sizeof (*ret));
  if (ret == NULL)
    return NULL;

  new_name = strdup (name);
  if (new_name == NULL)
    return NULL;

  node_set_name (ret, new_name);

  ret->path = strdup (path);
  if (ret->path == NULL)
    return NULL;

  ret->whiteout = 1;
  ret->ino = &dummy_ino;

  ret_xchg = ret;
  ret = NULL;

  stats.nodes++;
  return ret_xchg;
}

static ssize_t
safe_read_xattr (char **ret, int sfd, const char *name, size_t initial_size)
{
  cleanup_free char *buffer = NULL;
  size_t current_size;
  ssize_t s;

  current_size = initial_size;
  buffer = malloc (current_size + 1);
  if (buffer == NULL)
    return -1;

  while (1)
    {
      char *tmp;

      s = fgetxattr (sfd, name, buffer, current_size);
      if (s >= 0 && s < current_size)
        break;

      if (s < 0 && errno != ERANGE)
        break;

      current_size *= 2;
      tmp = realloc (buffer, current_size + 1);
      if (tmp == NULL)
        return -1;

      buffer = tmp;
    }

  if (s <= 0)
    return s;

  buffer[s] = '\0';

  /* Change owner.  */
  *ret = buffer;
  buffer = NULL;

  return s;
}

static struct ovl_node *
make_ovl_node (struct ovl_data *lo, const char *path, struct ovl_layer *layer, const char *name, ino_t ino, dev_t dev, bool dir_p, struct ovl_node *parent, bool fast_ino_check)
{
  mode_t mode = 0;
  char *new_name;
  struct ovl_node *ret_xchg;
  bool has_origin = true;
  cleanup_node_init struct ovl_node *ret = NULL;

  ret = calloc (1, sizeof (*ret));
  if (ret == NULL)
    return NULL;

  ret->parent = parent;
  ret->layer = layer;
  ret->tmp_ino = ino;
  ret->tmp_dev = dev;
  ret->hidden_dirfd = -1;
  ret->inodes = lo->inodes;
  ret->next_link = NULL;
  ret->ino = NULL;
  ret->node_lookups = 0;

  new_name = strdup (name);
  if (new_name == NULL)
    return NULL;
  node_set_name (ret, new_name);

  if (has_prefix (path, "./") && path[2])
    path += 2;

  ret->path = strdup (path);
  if (ret->path == NULL)
    return NULL;

  if (! dir_p)
    ret->children = NULL;
  else
    {
      ret->children = hash_initialize (128, NULL, node_hasher, node_compare, node_free);
      if (ret->children == NULL)
        return NULL;
    }

  if (ret->tmp_ino == 0)
    {
      struct stat st;
      struct ovl_layer *it;
      cleanup_free char *npath = NULL;
      char whiteout_path[PATH_MAX];
      bool stop_lookup = false;

      npath = strdup (ret->path);
      if (npath == NULL)
        return NULL;

      if (parent)
        strconcat3 (whiteout_path, PATH_MAX, parent->path, "/.wh.", name);
      else
        strconcat3 (whiteout_path, PATH_MAX, "/.wh.", name, NULL);

      for (it = layer; it && ! stop_lookup; it = it->next)
        {
          ssize_t s;
          cleanup_free char *val = NULL;
          cleanup_free char *origin = NULL;
          cleanup_close int fd = -1;

          if (parent && parent->last_layer == it)
            stop_lookup = true;

          if (dir_p)
            {
              int r;

              r = it->ds->file_exists (it, whiteout_path);
              if (r < 0 && errno == EACCES)
                break;

              if (r < 0 && errno != ENOENT && errno != ENOTDIR && errno != ENAMETOOLONG)
                return NULL;

              if (r == 0)
                break;
            }

          if (! fast_ino_check)
            fd = it->ds->openat (it, npath, O_RDONLY | O_NONBLOCK | O_NOFOLLOW, 0755);

          if (fd < 0)
            {
              if (it->ds->statat (it, npath, &st, AT_SYMLINK_NOFOLLOW, STATX_TYPE | STATX_MODE | STATX_INO) == 0)
                {
                  if (has_origin)
                    {
                      ret->tmp_ino = st.st_ino;
                      ret->tmp_dev = st.st_dev;
                      if (mode == 0)
                        mode = st.st_mode;
                    }
                  ret->last_layer = it;
                }
              has_origin = false;
              goto no_fd;
            }

          /* It is an open FD, stat the file and read the origin xattrs.  */
          if (it->ds->fstat (it, fd, npath, STATX_TYPE | STATX_MODE | STATX_INO, &st) == 0)
            {
              if (has_origin)
                {
                  ret->tmp_ino = st.st_ino;
                  ret->tmp_dev = st.st_dev;
                  if (mode == 0)
                    mode = st.st_mode;
                }
              ret->last_layer = it;
            }

          s = safe_read_xattr (&val, fd, PRIVILEGED_ORIGIN_XATTR, PATH_MAX);
          if (s > 0)
            {
              char buf[512];
              struct ovl_fh *ofh = (struct ovl_fh *) val;
              size_t s = ofh->len - sizeof (*ofh);
              struct file_handle *fh = (struct file_handle *) buf;

              if (s < sizeof (buf) - sizeof (int) * 2)
                {
                  cleanup_close int originfd = -1;

                  /*
                    overlay in the kernel stores a file handle in the .origin xattr.
                    Honor it when present, but don't fail on errors as an unprivileged
                    user cannot open a file handle.
                  */
                  fh->handle_bytes = s;
                  fh->handle_type = ofh->type;
                  memcpy (fh->f_handle, ofh->fid, s);

                  originfd = open_by_handle_at (AT_FDCWD, fh, O_RDONLY);
                  if (originfd >= 0)
                    {
                      if (it->ds->fstat (it, originfd, npath, STATX_TYPE | STATX_MODE | STATX_INO, &st) == 0)
                        {
                          ret->tmp_ino = st.st_ino;
                          ret->tmp_dev = st.st_dev;
                          mode = st.st_mode;
                          break;
                        }
                    }
                }
            }

          /* If an origin is specified, use it for the next layer lookup.  */
          s = safe_read_xattr (&origin, fd, ORIGIN_XATTR, PATH_MAX);
          if (s <= 0)
            has_origin = false;
          else
            {
              free (npath);
              npath = origin;
              origin = NULL;
            }

        no_fd:
          if (parent && parent->last_layer == it)
            break;
        }
    }

  ret_xchg = ret;
  ret = NULL;

  stats.nodes++;
  return register_inode (lo, ret_xchg, mode);
}

static struct ovl_node *
insert_node (struct ovl_node *parent, struct ovl_node *item, bool replace)
{
  struct ovl_node *old = NULL, *prev_parent = item->parent;
  int ret;

  if (prev_parent)
    {
      if (hash_lookup (prev_parent->children, item) == item)
        hash_remove (prev_parent->children, item);
    }

  if (replace)
    {
      old = hash_remove (parent->children, item);
      if (old)
        node_free (old);
    }

  ret = hash_insert_if_absent (parent->children, item, (const void **) &old);
  if (ret < 0)
    {
      node_free (item);
      errno = ENOMEM;
      return NULL;
    }
  if (ret == 0)
    {
      node_free (item);
      return old;
    }

  item->parent = parent;

  return item;
}

static const char *
get_whiteout_name (const char *name, struct stat *st)
{
  if (has_prefix (name, ".wh."))
    return name + 4;
  if (st
      && (st->st_mode & S_IFMT) == S_IFCHR
      && major (st->st_rdev) == 0
      && minor (st->st_rdev) == 0)
    return name;
  return NULL;
}

static struct ovl_node *
load_dir (struct ovl_data *lo, struct ovl_node *n, struct ovl_layer *layer, char *path, char *name)
{
  struct dirent *dent;
  bool stop_lookup = false;
  struct ovl_layer *it, *upper_layer = get_upper_layer (lo);
  char parent_whiteout_path[PATH_MAX];

  if (! n)
    {
      n = make_ovl_node (lo, path, layer, name, 0, 0, true, NULL, lo->fast_ino_check);
      if (n == NULL)
        {
          errno = ENOMEM;
          return NULL;
        }
    }

  if (n->parent)
    strconcat3 (parent_whiteout_path, PATH_MAX, n->parent->path, "/.wh.", name);
  else
    strconcat3 (parent_whiteout_path, PATH_MAX, ".wh.", name, NULL);

  for (it = lo->layers; it && ! stop_lookup; it = it->next)
    {
      struct stat st;
      DIR *dp = NULL;
      int ret;

      if (n->last_layer == it)
        stop_lookup = true;

      ret = it->ds->file_exists (it, parent_whiteout_path);
      if (ret < 0 && errno != ENOENT && errno != ENOTDIR && errno != ENAMETOOLONG)
        return NULL;

      if (ret == 0)
        break;

      ret = it->ds->statat (it, path, &st, AT_SYMLINK_NOFOLLOW, STATX_TYPE);
      if (ret < 0)
        {
          if (errno == ENOENT || errno == ENOTDIR || errno == ENAMETOOLONG)
            continue;
          return NULL;
        }
      /* not a directory, stop lookup in lower layers.  */
      if ((st.st_mode & S_IFMT) != S_IFDIR)
        break;

      dp = it->ds->opendir (it, path);
      if (dp == NULL)
        continue;

      for (;;)
        {
          struct ovl_node key;
          struct ovl_node *child = NULL;
          char node_path[PATH_MAX];
          char whiteout_path[PATH_MAX];

          errno = 0;
          dent = it->ds->readdir (dp);
          if (dent == NULL)
            {
              if (errno)
                {
                  it->ds->closedir (dp);
                  return NULL;
                }

              break;
            }

          node_set_name (&key, dent->d_name);

          if ((strcmp (dent->d_name, ".") == 0) || strcmp (dent->d_name, "..") == 0)
            continue;

          child = hash_lookup (n->children, &key);
          if (child)
            {
              child->last_layer = it;
              if (! child->whiteout || it != upper_layer)
                continue;
              else
                {
                  hash_remove (n->children, child);
                  node_free (child);
                  child = NULL;
                }
            }

          strconcat3 (whiteout_path, PATH_MAX, path, "/.wh.", dent->d_name);

          strconcat3 (node_path, PATH_MAX, n->path, "/", dent->d_name);

          ret = it->ds->file_exists (it, whiteout_path);
          if (ret < 0 && errno == EACCES)
            continue;
          if (ret < 0 && errno != ENOENT && errno != ENOTDIR && errno != ENAMETOOLONG)
            {
              it->ds->closedir (dp);
              return NULL;
            }

          if (ret == 0)
            {
              child = make_whiteout_node (node_path, dent->d_name);
              if (child == NULL)
                {
                  errno = ENOMEM;
                  it->ds->closedir (dp);
                  return NULL;
                }
            }
          else
            {
              const char *wh = NULL;
              bool dirp = dent->d_type == DT_DIR;

              if ((dent->d_type != DT_CHR) && (dent->d_type != DT_UNKNOWN))
                wh = get_whiteout_name (dent->d_name, NULL);
              else
                {
                  /* A stat is required either if the type is not known, or if it is a character device as it could be
                     a whiteout file.  */
                  struct stat st;

                  ret = it->ds->statat (it, node_path, &st, AT_SYMLINK_NOFOLLOW, STATX_TYPE);
                  if (ret < 0)
                    {
                      it->ds->closedir (dp);
                      return NULL;
                    }

                  dirp = st.st_mode & S_IFDIR;
                  wh = get_whiteout_name (dent->d_name, &st);
                }

              if (wh)
                {
                  child = make_whiteout_node (node_path, wh);
                  if (child == NULL)
                    {
                      errno = ENOMEM;
                      it->ds->closedir (dp);
                      return NULL;
                    }
                }
              else
                {
                  ino_t ino = 0;

                  if (lo->fast_ino_check)
                    ino = dent->d_ino;

                  child = make_ovl_node (lo, node_path, it, dent->d_name, ino, 0, dirp, n, lo->fast_ino_check);
                  if (child == NULL)
                    {
                      errno = ENOMEM;
                      it->ds->closedir (dp);
                      return NULL;
                    }
                  child->last_layer = it;
                }
            }

          if (insert_node (n, child, false) == NULL)
            {
              errno = ENOMEM;
              it->ds->closedir (dp);
              return NULL;
            }
        }

      ret = is_directory_opaque (it, path);
      if (ret < 0)
        {
          it->ds->closedir (dp);
          return NULL;
        }
      if (ret > 0)
        {
          n->last_layer = it;
          stop_lookup = true;
        }
      it->ds->closedir (dp);
    }

  if (get_timeout (lo) > 0)
    n->loaded = 1;
  return n;
}

static struct ovl_node *
reload_dir (struct ovl_data *lo, struct ovl_node *node)
{
  if (! node->loaded)
    node = load_dir (lo, node, node->layer, node->path, node->name);
  return node;
}

static void
free_layers (struct ovl_layer *layers)
{
  if (layers == NULL)
    return;
  free_layers (layers->next);
  free (layers->path);
  if (layers->fd >= 0)
    close (layers->fd);
  free (layers);
}

static void
cleanup_layerp (struct ovl_layer **p)
{
  struct ovl_layer *l = *p;
  free_layers (l);
}

#define cleanup_layer __attribute__ ((cleanup (cleanup_layerp)))

static void
unescape (char *input)
{
  char *dest = input;

  if (input == NULL)
    return;

  for (; *input; input++)
    {
      if (*input == '\\')
        continue;

      *dest++ = *input;
    }
  *dest = '\0';
}

static char *
get_next_path (char *it, char **saveptr)
{
  char *ret;

  if (*saveptr == NULL)
    *saveptr = it;

  ret = *saveptr;

  if (*ret == '\0')
    return NULL;

  while (1)
    {
      if (**saveptr == '\0')
        break;

      if (**saveptr == ':')
        {
          **saveptr = '\0';
          (*saveptr)++;
          break;
        }
      else if (**saveptr == '\\')
        {
          memmove (*saveptr, *saveptr + 1, strlen (*saveptr) + 1);
        }

      (*saveptr)++;
    }
  return ret;
}

static struct ovl_layer *
read_dirs (struct ovl_data *lo, char *path, bool low, struct ovl_layer *layers)
{
  char *saveptr = NULL, *it;
  struct ovl_layer *last;
  cleanup_free char *buf = NULL;

  if (path == NULL)
    return NULL;

  buf = strdup (path);
  if (buf == NULL)
    return NULL;

  last = layers;
  while (last && last->next)
    last = last->next;

  for (it = get_next_path (buf, &saveptr); it; it = get_next_path (NULL, &saveptr))
    {
      char *name, *data;
      char *it_path = it;
      int i, n_layers;
      cleanup_layer struct ovl_layer *l = NULL;
      struct data_source *ds;

      if (it[0] != '/' || it[1] != '/')
        {
          /* By default use the direct access data store.  */
          ds = &direct_access_ds;

          data = NULL;
          path = it_path;
        }
      else
        {
          struct ovl_plugin *p;
          char *plugin_data_sep, *plugin_sep;

          if (! low)
            {
              fprintf (stderr, "plugins are supported only with lower layers\n");
              return NULL;
            }

          plugin_sep = strchr (it + 2, '/');
          if (! plugin_sep)
            {
              fprintf (stderr, "invalid separator for plugin\n");
              return NULL;
            }

          *plugin_sep = '\0';

          name = it + 2;
          data = plugin_sep + 1;

          plugin_data_sep = strchr (data, '/');
          if (! plugin_data_sep)
            {
              fprintf (stderr, "invalid separator for plugin\n");
              return NULL;
            }

          *plugin_data_sep = '\0';
          path = plugin_data_sep + 1;

          p = plugin_find (lo->plugins_ctx, name);
          if (! p)
            {
              fprintf (stderr, "cannot find plugin %s\n", name);
              return NULL;
            }

          ds = p->load (data, path);
          if (ds == NULL)
            {
              fprintf (stderr, "cannot load plugin %s\n", name);
              return NULL;
            }
        }

      n_layers = ds->num_of_layers (data, path);
      if (n_layers < 0)
        {
          fprintf (stderr, "cannot retrieve number of layers for %s\n", path);
          return NULL;
        }

      for (i = 0; i < n_layers; i++)
        {
          l = calloc (1, sizeof (*l));
          if (l == NULL)
            return NULL;

          l->ds = ds;

          l->ovl_data = lo;

          l->path = NULL;
          l->fd = -1;

          /* If the plugin doesn't override it, we set st_dev to a hash of the
           * text configuring the plugin.  This means that different layers
           * will be deemed to be on different devices.
           */
          l->st_dev = hash_string (it, -1);

          if (l->ds->load_data_source (l, data, path, i) < 0)
            {
              fprintf (stderr, "cannot load store %s at %s\n", data, path);
              return NULL;
            }

          l->low = low;
          if (low)
            {
              l->next = NULL;
              if (last == NULL)
                last = layers = l;
              else
                {
                  last->next = l;
                  last = l;
                }
            }
          else
            {
              l->next = layers;
              layers = l;
            }
          l = NULL;
        }
    }
  return layers;
}

static struct ovl_node *
do_lookup_file (struct ovl_data *lo, fuse_ino_t parent, const char *name)
{
  struct ovl_node key;
  struct ovl_node *node, *pnode;

  if (parent == FUSE_ROOT_ID)
    pnode = lo->root;
  else
    pnode = inode_to_node (lo, parent);

  if (name == NULL)
    {
      /* Prefer a version that is loaded.  */
      for (node = pnode; node; node = node->next_link)
        if (node->parent)
          return node;
      return pnode;
    }

  if (has_prefix (name, ".wh."))
    {
      errno = EINVAL;
      return NULL;
    }

  node_set_name (&key, (char *) name);
  node = hash_lookup (pnode->children, &key);
  if (node == NULL && ! pnode->loaded)
    {
      int ret;
      struct ovl_layer *it;
      struct stat st;
      bool stop_lookup = false;

      for (it = lo->layers; it && ! stop_lookup; it = it->next)
        {
          char path[PATH_MAX];
          char whpath[PATH_MAX];
          const char *wh_name;

          if (pnode->last_layer == it)
            stop_lookup = true;

          strconcat3 (path, PATH_MAX, pnode->path, "/", name);

          ret = it->ds->statat (it, path, &st, AT_SYMLINK_NOFOLLOW, STATX_TYPE | STATX_MODE | STATX_INO);
          if (ret < 0)
            {
              int saved_errno = errno;

              if (errno == ENOENT || errno == ENOTDIR || errno == EACCES)
                {
                  if (node)
                    continue;

                  strconcat3 (whpath, PATH_MAX, pnode->path, "/.wh.", name);

                  ret = it->ds->file_exists (it, whpath);
                  if (ret < 0 && errno != ENOENT && errno != ENOTDIR && errno != ENAMETOOLONG)
                    return NULL;
                  if (ret == 0)
                    {
                      node = make_whiteout_node (path, name);
                      if (node == NULL)
                        {
                          errno = ENOMEM;
                          return NULL;
                        }
                      goto insert_node;
                    }
                  continue;
                }
              errno = saved_errno;
              return NULL;
            }

          /* If we already know the node, simply update the ino.  */
          if (node)
            {
              node->tmp_ino = st.st_ino;
              node->tmp_dev = st.st_dev;
              node->last_layer = it;
              continue;
            }

          strconcat3 (whpath, PATH_MAX, pnode->path, "/.wh.", name);
          ret = it->ds->file_exists (it, whpath);
          if (ret < 0 && errno != ENOENT && errno != ENOTDIR && errno != ENAMETOOLONG)
            return NULL;
          if (ret == 0)
            node = make_whiteout_node (path, name);
          else
            {
              wh_name = get_whiteout_name (name, &st);
              if (wh_name)
                node = make_whiteout_node (path, wh_name);
              else
                node = make_ovl_node (lo, path, it, name, 0, 0, st.st_mode & S_IFDIR, pnode, lo->fast_ino_check);
            }
          if (node == NULL)
            {
              errno = ENOMEM;
              return NULL;
            }

          if (st.st_mode & S_IFDIR)
            {
              ret = is_directory_opaque (it, path);
              if (ret < 0)
                {
                  node_free (node);
                  return NULL;
                }
              if (ret > 0)
                {
                  node->last_layer = it;
                  stop_lookup = true;
                }
            }
        insert_node:
          if (insert_node (pnode, node, false) == NULL)
            {
              node_free (node);
              errno = ENOMEM;
              return NULL;
            }
        }
    }

  return node;
}

static void
ovl_lookup (fuse_req_t req, fuse_ino_t parent, const char *name)
{
  cleanup_lock int l = enter_big_lock ();
  struct fuse_entry_param e;
  int err = 0;
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_lookup(parent=%" PRIu64 ", name=%s)\n",
             parent, name);

  memset (&e, 0, sizeof (e));

  node = do_lookup_file (lo, parent, name);
  if (node == NULL || node->whiteout)
    {
      e.ino = 0;
      e.attr_timeout = get_timeout (lo);
      e.entry_timeout = get_timeout (lo);
      fuse_reply_entry (req, &e);
      return;
    }

  if (! lo->static_nlink && node_dirp (node))
    {
      node = reload_dir (lo, node);
      if (node == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  err = rpl_stat (req, node, -1, NULL, NULL, &e.attr);
  if (err)
    {
      fuse_reply_err (req, errno);
      return;
    }

  e.ino = node_to_inode (node);
  node->ino->lookups++;
  e.attr_timeout = get_timeout (lo);
  e.entry_timeout = get_timeout (lo);
  fuse_reply_entry (req, &e);
}

struct ovl_dirp
{
  struct ovl_data *lo;
  struct ovl_node *parent;
  struct ovl_node **tbl;
  size_t tbl_size;
  size_t offset;
};

static struct ovl_dirp *
ovl_dirp (struct fuse_file_info *fi)
{
  return (struct ovl_dirp *) (uintptr_t) fi->fh;
}

/* Comparison function for qsort to ensure stable directory entry order.
   Entries are sorted by name to guarantee consistent offsets across
   multiple opendir sessions. */
static int
compare_nodes_by_name (const void *a, const void *b)
{
  struct ovl_node *node_a = *(struct ovl_node **) a;
  struct ovl_node *node_b = *(struct ovl_node **) b;
  return strcmp (node_a->name, node_b->name);
}

static int
reload_tbl (struct ovl_data *lo, struct ovl_dirp *d, struct ovl_node *node)
{
  size_t counter = 0;
  struct ovl_node *it;

  node = reload_dir (lo, node);
  if (node == NULL)
    return -1;

  if (d->tbl)
    free (d->tbl);

  d->offset = 0;
  d->parent = node;
  d->tbl_size = hash_get_n_entries (node->children) + 2;
  d->tbl = calloc (sizeof (struct ovl_node *), d->tbl_size);
  if (d->tbl == NULL)
    {
      errno = ENOMEM;
      return -1;
    }

  d->tbl[counter++] = node;
  d->tbl[counter++] = node->parent;

  for (it = hash_get_first (node->children); it; it = hash_get_next (node->children, it))
    {
      it->ino->lookups++;
      it->node_lookups++;
      d->tbl[counter++] = it;
    }

  /* Sort the directory entries by name to ensure stable offsets.

    The kernel FUSE layer caches directory listings using offsets as
    identifiers. If entries change positions between opendir sessions
    (which can happen when nodes are freed and recreated between loads),
    the kernel may incorrectly skip or duplicate entries.

    Sorting ensures the same filename always appears at the same offset. */
  qsort (&d->tbl[2], counter - 2, sizeof (struct ovl_node *), compare_nodes_by_name);

  return 0;
}

static void
ovl_opendir (fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
{
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  struct ovl_dirp *d = calloc (1, sizeof (struct ovl_dirp));
  cleanup_lock int l = enter_big_lock ();

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_opendir(ino=%" PRIu64 ")\n", ino);

  if (d == NULL)
    {
      errno = ENOMEM;
      goto out_errno;
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL)
    {
      errno = ENOENT;
      goto out_errno;
    }

  if (! node_dirp (node))
    {
      errno = ENOTDIR;
      goto out_errno;
    }

  d->parent = node;

  fi->fh = (uintptr_t) d;
  if (get_timeout (lo) > 0)
    {
      fi->keep_cache = 1;
#if HAVE_FUSE_CACHE_READDIR
      fi->cache_readdir = 1;
#endif
    }
  node->in_readdir++;
  fuse_reply_open (req, fi);
  return;

out_errno:
  if (d)
    {
      if (d->tbl)
        free (d->tbl);
      free (d);
    }
  fuse_reply_err (req, errno);
}

static int
create_missing_whiteouts (struct ovl_data *lo, struct ovl_node *node, const char *from)
{
  struct ovl_layer *l;

  if (! node_dirp (node))
    return 0;

  node = reload_dir (lo, node);
  if (node == NULL)
    return -1;

  for (l = get_lower_layers (lo); l; l = l->next)
    {
      cleanup_dir DIR *dp = NULL;

      dp = l->ds->opendir (l, from);
      if (dp == NULL)
        {
          if (errno == ENOTDIR)
            break;
          if (errno == ENOENT)
            continue;
          return -1;
        }
      else
        {
          struct dirent *dent;

          for (;;)
            {
              struct ovl_node key;
              struct ovl_node *n;

              errno = 0;
              dent = readdir (dp);
              if (dent == NULL)
                {
                  if (errno)
                    return -1;

                  break;
                }

              if (strcmp (dent->d_name, ".") == 0)
                continue;
              if (strcmp (dent->d_name, "..") == 0)
                continue;
              if (has_prefix (dent->d_name, ".wh."))
                continue;

              node_set_name (&key, (char *) dent->d_name);

              n = hash_lookup (node->children, &key);
              if (n)
                {
                  if (node_dirp (n))
                    {
                      char c[PATH_MAX];

                      n = reload_dir (lo, n);
                      if (n == NULL)
                        return -1;

                      strconcat3 (c, PATH_MAX, from, "/", n->name);

                      if (create_missing_whiteouts (lo, n, c) < 0)
                        return -1;
                    }
                  continue;
                }

              if (create_whiteout (lo, node, dent->d_name, false, true) < 0)
                return -1;
            }
        }
    }
  return 0;
}

static void
ovl_do_readdir (fuse_req_t req, fuse_ino_t ino, size_t size,
                off_t offset, struct fuse_file_info *fi, int plus)
{
  struct ovl_data *lo = ovl_data (req);
  struct ovl_dirp *d = ovl_dirp (fi);
  size_t remaining = size;
  char *p;
  cleanup_free char *buffer = NULL;

  buffer = malloc (size);
  if (buffer == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (offset == 0 || d->tbl == NULL)
    {
      if (reload_tbl (lo, d, d->parent) < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  p = buffer;
  for (; remaining > 0 && offset < d->tbl_size; offset++)
    {
      int ret;
      size_t entsize;
      const char *name;
      struct ovl_node *node = d->tbl[offset];
      struct fuse_entry_param e;
      struct stat *st = &e.attr;

      if (node == NULL || node->whiteout || node->hidden)
        continue;

      if (offset == 0)
        name = ".";
      else if (offset == 1)
        name = "..";
      else
        {
          if (node->parent != d->parent)
            continue;
          name = node->name;
        }

      if (! plus)
        {
          /* From the 'stbuf' argument the st_ino field and bits 12-15 of the
           * st_mode field are used.  The other fields are ignored.
           */
          st->st_ino = get_st_ino (node, lo);
          st->st_mode = node->ino->mode;

          entsize = fuse_add_direntry (req, p, remaining, name, st, offset + 1);
        }
      else
        {
          if (! lo->static_nlink && node_dirp (node))
            {
              node = reload_dir (lo, node);
              if (node == NULL)
                {
                  fuse_reply_err (req, errno);
                  return;
                }
            }
          memset (&e, 0, sizeof (e));
          ret = rpl_stat (req, node, -1, NULL, NULL, st);
          if (ret < 0)
            {
              fuse_reply_err (req, errno);
              return;
            }

          e.attr_timeout = get_timeout (lo);
          e.entry_timeout = get_timeout (lo);
          e.ino = node_to_inode (node);
          entsize = fuse_add_direntry_plus (req, p, remaining, name, &e, offset + 1);
          if (entsize <= remaining)
            {
              /* First two entries are . and .. */
              if (offset >= 2)
                node->ino->lookups++;
            }
        }

      if (entsize > remaining)
        break;

      p += entsize;
      remaining -= entsize;
    }
  fuse_reply_buf (req, buffer, size - remaining);
}

static void
ovl_readdir (fuse_req_t req, fuse_ino_t ino, size_t size,
             off_t offset, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_readdir(ino=%" PRIu64 ", size=%zu, offset=%lo)\n", ino, size, offset);
  ovl_do_readdir (req, ino, size, offset, fi, 0);
}

static void
ovl_readdirplus (fuse_req_t req, fuse_ino_t ino, size_t size,
                 off_t offset, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_readdirplus(ino=%" PRIu64 ", size=%zu, offset=%lo)\n", ino, size, offset);
  ovl_do_readdir (req, ino, size, offset, fi, 1);
}

static void
ovl_releasedir (fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  size_t s;
  struct ovl_dirp *d = ovl_dirp (fi);
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node = NULL;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_releasedir(ino=%" PRIu64 ")\n", ino);

  for (s = 2; s < d->tbl_size; s++)
    {
      d->tbl[s]->node_lookups--;
      if (! do_forget (lo, (fuse_ino_t) d->tbl[s]->ino, 1))
        {
          if (d->tbl[s]->node_lookups == 0)
            node_free (d->tbl[s]);
        }
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node)
    node->in_readdir--;

  free (d->tbl);
  free (d);
  fuse_reply_err (req, 0);
}

static int
inherit_acl (struct ovl_data *lo, struct ovl_node *parent, int targetfd, const char *path)
{
  cleanup_free char *v = NULL;
  cleanup_close int dfd = -1;
  int s;

  if (parent == NULL || lo->noacl)
    return 0;

  dfd = safe_openat (node_dirfd (parent), parent->path, O_DIRECTORY, 0);
  if (dfd < 0)
    return -1;

  s = safe_read_xattr (&v, dfd, ACL_XATTR, 4096);
  if (s < 0)
    {
      if (errno == ENODATA || errno == ENOTSUP)
        return 0;
      return -1;
    }
  if (targetfd >= 0)
    return fsetxattr (targetfd, ACL_XATTR, v, s, 0);

  return setxattr (path, ACL_XATTR, v, s, 0);
}

/* in-place filter xattrs that cannot be accessed.  */
static ssize_t
filter_xattrs_list (struct ovl_layer *l, char *buf, ssize_t len)
{
  ssize_t ret = 0;
  char *it;

  if (buf == NULL)
    return len;

  it = buf;

  while (it < buf + len)
    {
      size_t it_len;

      it_len = strlen (it) + 1;

      if (can_access_xattr (l, it))
        {
          it += it_len;
          ret += it_len;
        }
      else
        {
          char *next = it;

          next += encoded_xattr_name (l, it) ? sizeof (XATTR_CONTAINERS_OVERRIDE_PREFIX) - 1 : it_len;

          memmove (it, next, buf + len - next);
          len -= it_len;
        }
    }

  return ret;
}

static void
ovl_listxattr (fuse_req_t req, fuse_ino_t ino, size_t size)
{
  cleanup_lock int l = enter_big_lock ();
  ssize_t len;
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  cleanup_free char *buf = NULL;
  int ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_listxattr(ino=%" PRIu64 ", size=%zu)\n", ino, size);

  if (lo->disable_xattrs)
    {
      fuse_reply_err (req, ENOSYS);
      return;
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (size > 0)
    {
      buf = malloc (size);
      if (buf == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (! node->hidden)
    ret = node->layer->ds->listxattr (node->layer, node->path, buf, size);
  else
    {
      char path[PATH_MAX];
      strconcat3 (path, PATH_MAX, lo->workdir, "/", node->path);
      ret = listxattr (path, buf, size);
    }
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  len = filter_xattrs_list (node->layer, buf, ret);

  if (size == 0)
    fuse_reply_xattr (req, len);
  else if (len <= size)
    fuse_reply_buf (req, buf, len);
}

static void
ovl_getxattr (fuse_req_t req, fuse_ino_t ino, const char *name, size_t size)
{
  cleanup_lock int l = enter_big_lock ();
  ssize_t len;
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  cleanup_free char *value_buf = NULL;
  char name_buf[XATTR_NAME_MAX + 1];
  int ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_getxattr(ino=%" PRIu64 ", name=%s, size=%zu)\n", ino, name, size);

  if (lo->disable_xattrs)
    {
      fuse_reply_err (req, ENOSYS);
      return;
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  name = encode_xattr_name (node->layer, name_buf, name);
  if (! name)
    {
      fuse_reply_err (req, ENODATA);
      return;
    }

  if (size > 0)
    {
      value_buf = malloc (size);
      if (value_buf == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (! node->hidden)
    ret = node->layer->ds->getxattr (node->layer, node->path, name, value_buf, size);
  else
    {
      char path[PATH_MAX];
      strconcat3 (path, PATH_MAX, lo->workdir, "/", node->path);
      ret = getxattr (path, name, value_buf, size);
    }

  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  len = ret;

  if (size == 0)
    fuse_reply_xattr (req, len);
  else
    fuse_reply_buf (req, value_buf, len);
}

static void
ovl_access (fuse_req_t req, fuse_ino_t ino, int mask)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *n = do_lookup_file (lo, ino, NULL);

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_access(ino=%" PRIu64 ", mask=%d)\n",
             ino, mask);

  if ((mask & n->ino->mode) == mask)
    fuse_reply_err (req, 0);
  else
    fuse_reply_err (req, EPERM);
}

static int
copy_xattr (const struct ovl_layer *sl, int sfd,
            const struct ovl_layer *dl, int dfd, char *buf, size_t buf_size)
{
  ssize_t xattr_len;

  xattr_len = flistxattr (sfd, buf, buf_size);
  if (xattr_len > 0)
    {
      char *it;
      for (it = buf; it - buf < xattr_len; it += strlen (it) + 1)
        {
          cleanup_free char *v = NULL;
          const char *decoded_name = decode_xattr_name (sl, it);
          const char *encoded_name;
          char buf[XATTR_NAME_MAX + 1];
          ssize_t s;

          if (! decoded_name)
            continue;

          encoded_name = encode_xattr_name (dl, buf, decoded_name);
          if (! encoded_name)
            continue;

          s = safe_read_xattr (&v, sfd, it, 256);
          if (s < 0)
            {
              if (errno == EOVERFLOW)
                continue;
              return -1;
            }

          if (fsetxattr (dfd, encoded_name, v, s, 0) < 0)
            {
              if (errno == EINVAL || errno == EOPNOTSUPP)
                continue;
              return -1;
            }
        }
    }
  return 0;
}

static int
empty_dirfd (int fd)
{
  cleanup_dir DIR *dp = NULL;
  struct dirent *dent;

  dp = fdopendir (fd);
  if (dp == NULL)
    {
      close (fd);
      return -1;
    }

  for (;;)
    {
      int ret;

      errno = 0;
      dent = readdir (dp);
      if (dent == NULL)
        {
          if (errno)
            return -1;

          break;
        }
      if (strcmp (dent->d_name, ".") == 0)
        continue;
      if (strcmp (dent->d_name, "..") == 0)
        continue;

      ret = unlinkat (dirfd (dp), dent->d_name, 0);
      if (ret < 0 && errno == EISDIR)
        {
          ret = unlinkat (dirfd (dp), dent->d_name, AT_REMOVEDIR);
          if (ret < 0 && errno == ENOTEMPTY)
            {
              int dfd;

              dfd = safe_openat (dirfd (dp), dent->d_name, O_DIRECTORY, 0);
              if (dfd < 0)
                return -1;

              ret = empty_dirfd (dfd);
              if (ret < 0)
                return -1;

              ret = unlinkat (dirfd (dp), dent->d_name, AT_REMOVEDIR);
              if (ret < 0)
                return -1;

              continue;
            }
        }
      if (ret < 0)
        return ret;
    }

  return 0;
}

static int create_node_directory (struct ovl_data *lo, struct ovl_node *src);

static int
create_directory (struct ovl_data *lo, int dirfd, const char *name, const struct timespec *times,
                  struct ovl_node *parent, struct ovl_layer *sl, int xattr_sfd,
                  uid_t uid, gid_t gid, mode_t mode, bool set_opaque, struct stat *st_out)
{
  int ret;
  int saved_errno;
  cleanup_close int dfd = -1;
  cleanup_free char *buf = NULL;
  char wd_tmp_file_name[32];
  bool need_rename;
  mode_t backing_file_mode = mode | (lo->xattr_permissions ? 0755 : 0);

  need_rename = set_opaque || times || xattr_sfd >= 0 || uid != lo->uid || gid != lo->gid || get_upper_layer (lo)->stat_override_mode != STAT_OVERRIDE_NONE;
  if (! need_rename)
    {
      /* mkdir can be used directly without a temporary directory in the working directory.  */
      ret = mkdirat (dirfd, name, mode);
      if (ret < 0)
        {
          if (errno == EEXIST)
            {
              unlinkat (dirfd, name, 0);
              ret = mkdirat (dirfd, name, mode);
            }
          if (ret < 0)
            return ret;
        }
      if (st_out)
        return fstatat (dirfd, name, st_out, AT_SYMLINK_NOFOLLOW);
      return 0;
    }

  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

  ret = mkdirat (lo->workdir_fd, wd_tmp_file_name, backing_file_mode);
  if (ret < 0)
    goto out;

  ret = dfd = TEMP_FAILURE_RETRY (safe_openat (lo->workdir_fd, wd_tmp_file_name, O_RDONLY, 0));
  if (ret < 0)
    goto out;

  if (uid != lo->uid || gid != lo->gid || get_upper_layer (lo)->stat_override_mode != STAT_OVERRIDE_NONE)
    {
      ret = do_fchown (lo, dfd, uid, gid, mode);
      if (ret < 0)
        goto out;
    }

  if (times)
    {
      ret = futimens (dfd, times);
      if (ret < 0)
        goto out;
    }

  if (ret == 0 && xattr_sfd >= 0)
    {
      const size_t buf_size = 1 << 20;
      buf = malloc (buf_size);
      if (buf == NULL)
        {
          ret = -1;
          goto out;
        }

      ret = copy_xattr (sl, xattr_sfd, get_upper_layer (lo), dfd, buf, buf_size);
      if (ret < 0)
        goto out;
    }

  if (set_opaque)
    {
      ret = set_fd_opaque (dfd);
      if (ret < 0)
        goto out;
    }

  if (st_out)
    {
      ret = fstat (dfd, st_out);
      if (ret < 0)
        goto out;
      st_out->st_mode = (st_out->st_mode & S_IFMT) | (mode & ~S_IFMT);
    }

  ret = inherit_acl (lo, parent, dfd, NULL);
  if (ret < 0)
    goto out;

  ret = renameat (lo->workdir_fd, wd_tmp_file_name, dirfd, name);
  if (ret < 0)
    {
      if (errno == EEXIST)
        {
          int dfd = -1;

          ret = direct_renameat2 (lo->workdir_fd, wd_tmp_file_name, dirfd, name, RENAME_EXCHANGE);
          if (ret < 0)
            goto out;

          dfd = TEMP_FAILURE_RETRY (safe_openat (lo->workdir_fd, wd_tmp_file_name, O_DIRECTORY, 0));
          if (dfd < 0)
            return -1;

          ret = empty_dirfd (dfd);
          if (ret < 0)
            goto out;

          return unlinkat (lo->workdir_fd, wd_tmp_file_name, AT_REMOVEDIR);
        }
      if (errno == ENOTDIR)
        unlinkat (dirfd, name, 0);
      if (errno == ENOENT && parent)
        {
          ret = create_node_directory (lo, parent);
          if (ret != 0)
            goto out;
        }

      ret = renameat (lo->workdir_fd, wd_tmp_file_name, dirfd, name);
    }
out:
  saved_errno = errno;
  if (ret < 0)
    unlinkat (lo->workdir_fd, wd_tmp_file_name, AT_REMOVEDIR);
  errno = saved_errno;

  return ret;
}

static int
create_node_directory (struct ovl_data *lo, struct ovl_node *src)
{
  int ret;
  struct stat st;
  cleanup_close int sfd = -1;
  struct timespec times[2];

  if (src == NULL)
    return 0;

  if (src->layer == get_upper_layer (lo))
    return 0;

  ret = sfd = src->layer->ds->openat (src->layer, src->path, O_RDONLY | O_NONBLOCK, 0755);
  if (ret < 0)
    return ret;

  ret = TEMP_FAILURE_RETRY (fstat (sfd, &st));
  if (ret < 0)
    return ret;

  times[0] = st.st_atim;
  times[1] = st.st_mtim;

  if (override_mode (src->layer, sfd, NULL, NULL, &st) < 0 && errno != ENODATA && errno != EOPNOTSUPP)
    return -1;

  ret = create_directory (lo, get_upper_layer (lo)->fd, src->path, times, src->parent, src->layer, sfd, st.st_uid, st.st_gid, st.st_mode, false, NULL);
  if (ret == 0)
    {
      src->layer = get_upper_layer (lo);

      if (src->parent)
        delete_whiteout (lo, -1, src->parent, src->name);
    }

  return ret;
}

static int
copy_fd_to_fd (int sfd, int dfd, char *buf, size_t buf_size)
{
  int ret;

  for (;;)
    {
      int written;
      int nread;

      nread = TEMP_FAILURE_RETRY (read (sfd, buf, buf_size));
      if (nread < 0)
        return nread;

      if (nread == 0)
        break;

      written = 0;
      do
        {
          ret = TEMP_FAILURE_RETRY (write (dfd, buf + written, nread));
          if (ret < 0)
            return ret;
          nread -= ret;
          written += ret;
      } while (nread);
    }
  return 0;
}

static int
copyup (struct ovl_data *lo, struct ovl_node *node)
{
  int saved_errno;
  int ret = -1;
  cleanup_close int dfd = -1;
  cleanup_close int sfd = -1;
  struct stat st;
  const size_t buf_size = 1 << 20;
  cleanup_free char *buf = NULL;
  struct timespec times[2];
  char wd_tmp_file_name[32];
  static bool support_reflinks = true;
  bool data_copied = false;
  mode_t mode;

  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

  ret = node->layer->ds->statat (node->layer, node->path, &st, AT_SYMLINK_NOFOLLOW, STATX_BASIC_STATS);
  if (ret < 0)
    return ret;

  if (node->parent)
    {
      ret = create_node_directory (lo, node->parent);
      if (ret < 0)
        return ret;
    }

  mode = st.st_mode;
  if (lo->xattr_permissions)
    mode |= 0755;
  if (lo->euid > 0)
    mode |= 0200;

  if ((mode & S_IFMT) == S_IFDIR)
    {
      ret = create_node_directory (lo, node);
      if (ret < 0)
        goto exit;
      goto success;
    }

  if ((mode & S_IFMT) == S_IFLNK)
    {
      size_t current_size = PATH_MAX + 1;
      cleanup_free char *p = malloc (current_size);

      while (1)
        {
          char *new;

          ret = node->layer->ds->readlinkat (node->layer, node->path, p, current_size - 1);
          if (ret < 0)
            goto exit;
          if (ret < current_size - 1)
            break;

          current_size = current_size * 2;
          new = realloc (p, current_size);
          if (new == NULL)
            goto exit;
          p = new;
        }
      p[ret] = '\0';
      ret = symlinkat (p, get_upper_layer (lo)->fd, node->path);
      if (ret < 0)
        goto exit;
      goto success;
    }

  ret = sfd = node->layer->ds->openat (node->layer, node->path, O_RDONLY | O_NONBLOCK, 0);
  if (sfd < 0)
    goto exit;

  ret = dfd = TEMP_FAILURE_RETRY (safe_openat (lo->workdir_fd, wd_tmp_file_name, O_CREAT | O_WRONLY, mode));
  if (dfd < 0)
    goto exit;

  if (st.st_uid != lo->uid || st.st_gid != lo->gid || get_upper_layer (lo)->stat_override_mode != STAT_OVERRIDE_NONE)
    {
      ret = do_fchown (lo, dfd, st.st_uid, st.st_gid, mode);
      if (ret < 0)
        goto exit;
    }

  buf = malloc (buf_size);
  if (buf == NULL)
    goto exit;

  if (support_reflinks)
    {
      if (ioctl (dfd, FICLONE, sfd) >= 0)
        data_copied = true;
      else if (errno == ENOTSUP || errno == EINVAL)
        {
          /* Fallback to data copy and don't attempt again FICLONE.  */
          support_reflinks = false;
        }
    }

#ifdef HAVE_SYS_SENDFILE_H
  if (! data_copied)
    {
      off_t copied = 0;

      while (copied < st.st_size)
        {
          off_t tocopy = st.st_size - copied;
          ssize_t n = TEMP_FAILURE_RETRY (sendfile (dfd, sfd, NULL, tocopy > SIZE_MAX ? SIZE_MAX : (size_t) tocopy));
          if (n < 0)
            {
              /* On failure, fallback to the read/write loop.  */
              ret = copy_fd_to_fd (sfd, dfd, buf, buf_size);
              if (ret < 0)
                goto exit;
              break;
            }
          copied += n;
        }
      data_copied = true;
    }
#endif

  if (! data_copied)
    {
      ret = copy_fd_to_fd (sfd, dfd, buf, buf_size);
      if (ret < 0)
        goto exit;
    }

  times[0] = st.st_atim;
  times[1] = st.st_mtim;
  ret = futimens (dfd, times);
  if (ret < 0)
    goto exit;

  ret = copy_xattr (node->layer, sfd, get_upper_layer (lo), dfd, buf, buf_size);
  if (ret < 0)
    goto exit;

  ret = set_fd_origin (dfd, node->path);
  if (ret < 0)
    goto exit;

  /* Finally, move the file to its destination.  */
  ret = renameat (lo->workdir_fd, wd_tmp_file_name, get_upper_layer (lo)->fd, node->path);
  if (ret < 0)
    goto exit;

  if (node->parent)
    {
      char whpath[PATH_MAX];

      strconcat3 (whpath, PATH_MAX, node->parent->path, "/.wh.", node->name);

      if (unlinkat (get_upper_layer (lo)->fd, whpath, 0) < 0 && errno != ENOENT)
        goto exit;
    }

success:
  ret = 0;

  node->layer = get_upper_layer (lo);

exit:
  saved_errno = errno;
  if (ret < 0)
    unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
  errno = saved_errno;

  return ret;
}

static struct ovl_node *
get_node_up (struct ovl_data *lo, struct ovl_node *node)
{
  int ret;

  if (lo->upperdir == NULL)
    {
      errno = EROFS;
      return NULL;
    }

  if (node->layer == get_upper_layer (lo))
    return node;

  ret = copyup (lo, node);
  if (ret < 0)
    return NULL;

  assert (node->layer == get_upper_layer (lo));

  return node;
}

static size_t
count_dir_entries (struct ovl_node *node, size_t *whiteouts)
{
  size_t c = 0;
  struct ovl_node *it;

  if (whiteouts)
    *whiteouts = 0;

  for (it = hash_get_first (node->children); it; it = hash_get_next (node->children, it))
    {
      if (it->whiteout)
        {
          if (whiteouts)
            (*whiteouts)++;
          continue;
        }
      if (strcmp (it->name, ".") == 0)
        continue;
      if (strcmp (it->name, "..") == 0)
        continue;
      c++;
    }
  return c;
}

static int
update_paths (struct ovl_node *node)
{
  struct ovl_node *it;

  if (node == NULL)
    return 0;

  if (node->parent)
    {
      free (node->path);
      if (asprintf (&node->path, "%s/%s", node->parent->path, node->name) < 0)
        {
          node->path = NULL;
          return -1;
        }
    }

  if (node->children)
    {
      for (it = hash_get_first (node->children); it; it = hash_get_next (node->children, it))
        {
          if (update_paths (it) < 0)
            return -1;
        }
    }

  return 0;
}

static int
empty_dir (struct ovl_layer *l, const char *path)
{
  cleanup_close int cleanup_fd = -1;
  int ret;

  cleanup_fd = TEMP_FAILURE_RETRY (safe_openat (l->fd, path, O_DIRECTORY, 0));
  if (cleanup_fd < 0)
    return -1;

  if (set_fd_opaque (cleanup_fd) < 0)
    return -1;

  ret = empty_dirfd (cleanup_fd);

  cleanup_fd = -1;

  return ret;
}

static void
do_rm (fuse_req_t req, fuse_ino_t parent, const char *name, bool dirp)
{
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *pnode;
  int ret = 0;
  size_t whiteouts = 0;
  struct ovl_node key, *rm;

  node = do_lookup_file (lo, parent, name);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (dirp)
    {
      size_t c;

      /* Re-load the directory.  */
      node = reload_dir (lo, node);
      if (node == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }

      c = count_dir_entries (node, &whiteouts);
      if (c)
        {
          fuse_reply_err (req, ENOTEMPTY);
          return;
        }
    }

  if (node->layer == get_upper_layer (lo))
    {
      if (! dirp)
        node->do_unlink = 1;
      else
        {
          if (whiteouts > 0)
            {
              if (empty_dir (get_upper_layer (lo), node->path) < 0)
                {
                  fuse_reply_err (req, errno);
                  return;
                }
            }

          node->do_rmdir = 1;
        }
    }

  pnode = do_lookup_file (lo, parent, NULL);
  if (pnode == NULL || pnode->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  /* If the node is still accessible then be sure we
     can write to it.  Fix it to be done when a write is
     really done, not now.  */
  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  node_set_name (&key, (char *) name);

  rm = hash_remove (pnode->children, &key);
  fuse_lowlevel_notify_inval_inode (lo->se, node_to_inode (node), -1, 0);
  if (rm)
    {
      ret = hide_node (lo, rm, true);
      if (ret < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }

      node_free (rm);
    }

  fuse_reply_err (req, ret);
}

static void
ovl_unlink (fuse_req_t req, fuse_ino_t parent, const char *name)
{
  cleanup_lock int l = enter_big_lock ();
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_unlink(parent=%" PRIu64 ", name=%s)\n",
             parent, name);
  do_rm (req, parent, name, false);
}

static void
ovl_rmdir (fuse_req_t req, fuse_ino_t parent, const char *name)
{
  cleanup_lock int l = enter_big_lock ();
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_rmdir(parent=%" PRIu64 ", name=%s)\n",
             parent, name);
  do_rm (req, parent, name, true);
}

static int
direct_setxattr (struct ovl_layer *l, const char *path, const char *name, const char *buf, size_t size, int flags)
{
  cleanup_close int fd = -1;
  char full_path[PATH_MAX];
  int ret;

  full_path[0] = '\0';
  ret = open_fd_or_get_path (l, path, full_path, &fd, O_WRONLY);
  if (ret < 0)
    return ret;

  if (fd >= 0)
    return fsetxattr (fd, name, buf, size, flags);

  return setxattr (full_path, name, buf, size, flags);
}

static void
ovl_setxattr (fuse_req_t req, fuse_ino_t ino, const char *name,
              const char *value, size_t size, int flags)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node;
  char name_buf[XATTR_NAME_MAX + 1];
  int ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_setxattr(ino=%" PRIu64 ", name=%s, value=%s, size=%zu, flags=%d)\n", ino, name,
             value, size, flags);

  if (lo->disable_xattrs)
    {
      fuse_reply_err (req, ENOSYS);
      return;
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  name = encode_xattr_name (node->layer, name_buf, name);
  if (! name)
    {
      fuse_reply_err (req, EPERM);
      return;
    }

  if (! node->hidden)
    ret = direct_setxattr (node->layer, node->path, name, value, size, flags);
  else
    {
      char path[PATH_MAX];
      strconcat3 (path, PATH_MAX, lo->workdir, "/", node->path);
      ret = setxattr (path, name, value, size, flags);
    }
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fuse_reply_err (req, 0);
}

static int
direct_removexattr (struct ovl_layer *l, const char *path, const char *name)
{
  cleanup_close int fd = -1;
  char full_path[PATH_MAX];
  int ret;

  full_path[0] = '\0';
  ret = open_fd_or_get_path (l, path, full_path, &fd, O_WRONLY);
  if (ret < 0)
    return ret;

  if (fd >= 0)
    return fremovexattr (fd, name);

  return lremovexattr (full_path, name);
}

static void
ovl_removexattr (fuse_req_t req, fuse_ino_t ino, const char *name)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  char name_buf[XATTR_NAME_MAX + 1];
  int ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_removexattr(ino=%" PRIu64 ", name=%s)\n", ino, name);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  name = encode_xattr_name (node->layer, name_buf, name);
  if (! name)
    {
      fuse_reply_err (req, EPERM);
      return;
    }

  if (! node->hidden)
    ret = direct_removexattr (node->layer, node->path, name);
  else
    {
      char path[PATH_MAX];
      strconcat3 (path, PATH_MAX, lo->workdir, "/", node->path);
      ret = removexattr (path, name);
    }

  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fuse_reply_err (req, 0);
}

static int
direct_create_file (struct ovl_layer *l, int dirfd, const char *path, uid_t uid, gid_t gid, int flags, mode_t mode)
{
  struct ovl_data *lo = l->ovl_data;
  mode_t backing_file_mode = mode | (lo->xattr_permissions ? 0755 : 0);
  cleanup_close int fd = -1;
  char wd_tmp_file_name[32];
  int ret;

  /* try to create directly the file if it doesn't need to be chowned.  */
  if (uid == lo->uid && gid == lo->gid && l->stat_override_mode == STAT_OVERRIDE_NONE)
    {
      ret = TEMP_FAILURE_RETRY (safe_openat (get_upper_layer (lo)->fd, path, flags, backing_file_mode));
      if (ret >= 0)
        return ret;
      /* if it fails (e.g. there is a whiteout) then fallback to create it in
         the working dir + rename.  */
    }

  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

  fd = TEMP_FAILURE_RETRY (safe_openat (lo->workdir_fd, wd_tmp_file_name, flags, backing_file_mode));
  if (fd < 0)
    return -1;
  if (uid != lo->uid || gid != lo->gid || l->stat_override_mode != STAT_OVERRIDE_NONE)
    {
      if (do_fchown (lo, fd, uid, gid, mode) < 0)
        {
          unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
          return -1;
        }
    }

  if (renameat (lo->workdir_fd, wd_tmp_file_name, get_upper_layer (lo)->fd, path) < 0)
    {
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return -1;
    }

  ret = fd;
  fd = -1;
  return ret;
}

static int
ovl_do_open (fuse_req_t req, fuse_ino_t parent, const char *name, int flags, mode_t mode, struct ovl_node **retnode, struct stat *st)
{
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *n;
  bool readonly = (flags & (O_APPEND | O_RDWR | O_WRONLY | O_CREAT | O_TRUNC)) == 0;
  cleanup_free char *path = NULL;
  cleanup_close int fd = -1;
  uid_t uid;
  gid_t gid;
  bool need_delete_whiteout = true;
  bool is_whiteout = false;

  flags |= O_NOFOLLOW;

  flags &= ~O_DIRECT;

  if (lo->writeback)
    {
      if ((flags & O_ACCMODE) == O_WRONLY)
        {
          flags &= ~O_ACCMODE;
          flags |= O_RDWR;
        }
      if (flags & O_APPEND)
        flags &= ~O_APPEND;
    }

  if (name && has_prefix (name, ".wh."))
    {
      errno = EINVAL;
      return -1;
    }

  n = do_lookup_file (lo, parent, name);
  if (n && n->hidden)
    {
      if (retnode)
        *retnode = n;

      return openat (n->hidden_dirfd, n->path, flags, mode);
    }
  if (n && ! n->whiteout && (flags & O_CREAT))
    {
      errno = EEXIST;
      return -1;
    }
  if (n && n->whiteout)
    {
      n = NULL;
      is_whiteout = true;
    }

  if (! n)
    {
      int ret;
      struct ovl_node *p;
      const struct fuse_ctx *ctx = fuse_req_ctx (req);
      char wd_tmp_file_name[32];
      struct stat st_tmp;

      if ((flags & O_CREAT) == 0)
        {
          errno = ENOENT;
          return -1;
        }

      p = do_lookup_file (lo, parent, NULL);
      if (p == NULL)
        {
          errno = ENOENT;
          return -1;
        }

      p = get_node_up (lo, p);
      if (p == NULL)
        return -1;

      if (p->loaded && ! is_whiteout)
        need_delete_whiteout = false;

      sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

      ret = asprintf (&path, "%s/%s", p->path, name);
      if (ret < 0)
        return ret;

      uid = get_uid (lo, ctx->uid);
      gid = get_gid (lo, ctx->gid);

      fd = direct_create_file (get_upper_layer (lo), get_upper_layer (lo)->fd, path, uid, gid, flags, mode & ~ctx->umask);
      if (fd < 0)
        return fd;

      ret = inherit_acl (lo, n, fd, NULL);
      if (ret < 0)
        return ret;

      if (need_delete_whiteout && delete_whiteout (lo, -1, p, name) < 0)
        return -1;

      if (st == NULL)
        st = &st_tmp;

      if (get_upper_layer (lo)->ds->fstat (get_upper_layer (lo), fd, path, STATX_BASIC_STATS, st) < 0)
        return -1;

      n = make_ovl_node (lo, path, get_upper_layer (lo), name, st->st_ino, st->st_dev, false, p, lo->fast_ino_check);
      if (n == NULL)
        {
          errno = ENOMEM;
          return -1;
        }
      if (! is_whiteout)
        n->last_layer = get_upper_layer (lo);

      n = insert_node (p, n, true);
      if (n == NULL)
        {
          errno = ENOMEM;
          return -1;
        }
      ret = fd;
      fd = -1; /*  We use a temporary variable so we don't close it at cleanup.  */
      if (retnode)
        *retnode = n;
      return ret;
    }

  /* readonly, we can use both lowerdir and upperdir.  */
  if (readonly)
    {
      struct ovl_layer *l = n->layer;
      if (retnode)
        *retnode = n;
      return l->ds->openat (l, n->path, flags, mode);
    }
  else
    {
      struct ovl_layer *l;

      n = get_node_up (lo, n);
      if (n == NULL)
        return -1;

      if (retnode)
        *retnode = n;

      l = n->layer;

      return l->ds->openat (l, n->path, flags, mode);
    }
}

static void
ovl_read (fuse_req_t req, fuse_ino_t ino, size_t size,
          off_t offset, struct fuse_file_info *fi)
{
  struct fuse_bufvec buf = FUSE_BUFVEC_INIT (size);
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_read(ino=%" PRIu64 ", size=%zd, "
                     "off=%lu)\n",
             ino, size, (unsigned long) offset);
  buf.buf[0].flags = FUSE_BUF_IS_FD | FUSE_BUF_FD_SEEK | FUSE_BUF_FD_RETRY;
  buf.buf[0].fd = fi->fh;
  buf.buf[0].pos = offset;
  fuse_reply_data (req, &buf, 0);
}

static void
ovl_write_buf (fuse_req_t req, fuse_ino_t ino,
               struct fuse_bufvec *in_buf, off_t off,
               struct fuse_file_info *fi)
{
  struct ovl_data *lo = ovl_data (req);
  ssize_t res;
  struct ovl_ino *inode;
  int saved_errno;
  struct fuse_bufvec out_buf = FUSE_BUFVEC_INIT (fuse_buf_size (in_buf));

  out_buf.buf[0].flags = FUSE_BUF_IS_FD | FUSE_BUF_FD_SEEK | FUSE_BUF_FD_RETRY;
  out_buf.buf[0].fd = fi->fh;
  out_buf.buf[0].pos = off;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_write_buf(ino=%" PRIu64 ", size=%zd, off=%lu, fd=%d)\n",
             ino, out_buf.buf[0].size, (unsigned long) off, (int) fi->fh);

  inode = lookup_inode (lo, ino);

  errno = 0;
  res = fuse_buf_copy (&out_buf, in_buf, 0);
  saved_errno = errno;

  /* if it is a writepage request, make sure to restore the setuid bit.  */
  if (fi->writepage && (inode->mode & (S_ISUID | S_ISGID)))
    {
      if (do_fchmod (lo, inode->node, fi->fh, inode->mode) < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (res < 0)
    fuse_reply_err (req, saved_errno);
  else
    fuse_reply_write (req, (size_t) res);
}

static void
ovl_release (fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
{
  int ret;
  (void) ino;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_release(ino=%" PRIu64 ")\n", ino);

  ret = close (fi->fh);
  fuse_reply_err (req, ret == 0 ? 0 : errno);
}

static int
do_getattr (fuse_req_t req, struct fuse_entry_param *e, struct ovl_node *node, int fd, const char *path)
{
  struct ovl_data *lo = ovl_data (req);
  int err = 0;

  memset (e, 0, sizeof (*e));

  err = rpl_stat (req, node, fd, path, NULL, &e->attr);
  if (err < 0)
    return err;

  e->ino = node_to_inode (node);
  e->attr_timeout = get_timeout (lo);
  e->entry_timeout = get_timeout (lo);

  return 0;
}

static int
do_statfs (struct ovl_data *lo, struct statvfs *sfs)
{
  int ret, fd;

  fd = get_first_layer (lo)->fd;

  if (fd >= 0)
    ret = fstatvfs (fd, sfs);
  else
    ret = statvfs (lo->mountpoint, sfs);
  if (ret < 0)
    return ret;

  sfs->f_namemax -= WHITEOUT_MAX_LEN;
  return 0;
}

static short
get_fs_namemax (struct ovl_data *lo)
{
  static short namemax = 0;
  if (namemax == 0)
    {
      struct statvfs sfs;
      int ret;

      ret = do_statfs (lo, &sfs);
      /* On errors use a sane default.  */
      if (ret < 0)
        namemax = 255 - WHITEOUT_MAX_LEN;
      else
        namemax = sfs.f_namemax;
    }
  return namemax;
}

static void
ovl_create (fuse_req_t req, fuse_ino_t parent, const char *name,
            mode_t mode, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  cleanup_close int fd = -1;
  struct fuse_entry_param e;
  struct ovl_node *p, *node = NULL;
  struct ovl_data *lo = ovl_data (req);
  struct stat st;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_create(parent=%" PRIu64 ", name=%s)\n",
             parent, name);

  if (strlen (name) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  fi->flags = fi->flags | O_CREAT;

  fd = ovl_do_open (req, parent, name, fi->flags, mode, &node, &st);
  if (fd < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  p = do_lookup_file (lo, parent, NULL);
  /* Make sure the cache is invalidated, if the parent is in the middle of a readdir. */
  if (p && p->in_readdir)
    fuse_lowlevel_notify_inval_inode (lo->se, parent, 0, 0);

  if (node == NULL || do_getattr (req, &e, node, fd, NULL) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fi->fh = fd;
  fd = -1; /* Do not clean it up.  */

  node->ino->lookups++;
  fuse_reply_create (req, &e, fi);
}

static void
ovl_open (fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
{
  struct ovl_data *lo = ovl_data (req);
  cleanup_lock int l = enter_big_lock ();
  cleanup_close int fd = -1;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_open(ino=%" PRIu64 ")\n", ino);

  fd = ovl_do_open (req, ino, NULL, fi->flags, 0700, NULL, NULL);
  if (fd < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }
  fi->fh = fd;
  if (get_timeout (lo) > 0)
    fi->keep_cache = 1;
  fd = -1; /* Do not clean it up.  */
  fuse_reply_open (req, fi);
}

static void
ovl_getattr (fuse_req_t req, fuse_ino_t ino, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node;
  struct fuse_entry_param e;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_getattr(ino=%" PRIu64 ")\n", ino);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (do_getattr (req, &e, node, -1, NULL) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fuse_reply_attr (req, &e.attr, get_timeout (lo));
}

static void
ovl_setattr (fuse_req_t req, fuse_ino_t ino, struct stat *attr, int to_set, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  cleanup_close int cleaned_up_fd = -1;
  struct ovl_node *node;
  struct fuse_entry_param e;
  struct timespec times[2];
  uid_t uid;
  gid_t gid;
  int ret;
  int fd = -1;
  char path[PATH_MAX];

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_setattr(ino=%" PRIu64 ", to_set=%d)\n", ino, to_set);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (to_set & FUSE_SET_ATTR_CTIME)
    {
      /* Ignore request.  */
    }

  uid = -1;
  gid = -1;
  if (to_set & FUSE_SET_ATTR_UID)
    uid = get_uid (lo, attr->st_uid);
  if (to_set & FUSE_SET_ATTR_GID)
    gid = get_gid (lo, attr->st_gid);

  if (fi != NULL)
    fd = fi->fh; // use existing fd if fuse_file_info is available
  else
    {
      mode_t mode = node->ino->mode;
      int dirfd = node_dirfd (node);

      if (mode == 0)
        {
          struct stat st;

          ret = fstatat (dirfd, node->path, &st, AT_SYMLINK_NOFOLLOW);
          if (ret < 0)
            {
              fuse_reply_err (req, errno);
              return;
            }
          node->ino->mode = mode = st.st_mode;
        }

      switch (mode & S_IFMT)
        {
        case S_IFREG:
          cleaned_up_fd = fd = TEMP_FAILURE_RETRY (safe_openat (dirfd, node->path, O_NOFOLLOW | O_NONBLOCK | (to_set & FUSE_SET_ATTR_SIZE ? O_WRONLY : 0), 0));
          if (fd < 0)
            strconcat3 (path, PATH_MAX, get_upper_layer (lo)->path, "/", node->path);
          break;

        case S_IFDIR:
          cleaned_up_fd = fd = TEMP_FAILURE_RETRY (safe_openat (dirfd, node->path, O_NOFOLLOW | O_NONBLOCK, 0));
          if (fd < 0)
            {
              if (errno != ELOOP)
                strconcat3 (path, PATH_MAX, get_upper_layer (lo)->path, "/", node->path);
            }
          break;

        case S_IFLNK:
          cleaned_up_fd = TEMP_FAILURE_RETRY (safe_openat (dirfd, node->path, O_PATH | O_NOFOLLOW | O_NONBLOCK, 0));
          if (cleaned_up_fd < 0)
            {
              fuse_reply_err (req, errno);
              return;
            }
          sprintf (path, "/proc/self/fd/%d", cleaned_up_fd);
          break;

        default:
          strconcat3 (path, PATH_MAX, get_upper_layer (lo)->path, "/", node->path);
          break;
        }
    }

  l = release_big_lock ();

  memset (times, 0, sizeof (times));
  times[0].tv_nsec = UTIME_OMIT;
  times[1].tv_nsec = UTIME_OMIT;
  if (to_set & FUSE_SET_ATTR_ATIME)
    {
      times[0] = attr->st_atim;
      if (to_set & FUSE_SET_ATTR_ATIME_NOW)
        times[0].tv_nsec = UTIME_NOW;
    }

  if (to_set & FUSE_SET_ATTR_MTIME)
    {
      times[1] = attr->st_mtim;
      if (to_set & FUSE_SET_ATTR_MTIME_NOW)
        times[1].tv_nsec = UTIME_NOW;
    }

  if (times[0].tv_nsec != UTIME_OMIT || times[1].tv_nsec != UTIME_OMIT)
    {
      if (fd >= 0)
        ret = futimens (fd, times);
      else
        ret = utimensat (AT_FDCWD, path, times, 0);
      if (ret < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (to_set & FUSE_SET_ATTR_MODE)
    {
      if (fd >= 0)
        ret = do_fchmod (lo, node, fd, attr->st_mode);
      else
        ret = do_chmod (lo, node, path, attr->st_mode);
      if (ret < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
      node->ino->mode = attr->st_mode;
    }

  if (to_set & FUSE_SET_ATTR_SIZE)
    {
      if (fd >= 0)
        ret = ftruncate (fd, attr->st_size);
      else
        ret = truncate (path, attr->st_size);
      if (ret < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (uid != -1 || gid != -1)
    {
      struct stat st;

      if (do_stat (node, fd, NULL, &st) < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }

      if (uid == -1)
        {
          uid = st.st_uid;
        }

      if (gid == -1)
        {
          gid = st.st_gid;
        }

      if (fd >= 0)
        ret = do_fchown (lo, fd, uid, gid, node->ino->mode);
      else
        ret = do_chown (lo, path, uid, gid, node->ino->mode);
      if (ret < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  if (do_getattr (req, &e, node, fd, path) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fuse_reply_attr (req, &e.attr, get_timeout (lo));
}

static int
direct_linkat (struct ovl_layer *l, const char *oldpath, const char *newpath, int flags)
{
  return linkat (l->fd, oldpath, l->fd, newpath, 0);
}

static void
ovl_link (fuse_req_t req, fuse_ino_t ino, fuse_ino_t newparent, const char *newname)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node, *newparentnode, *destnode;
  cleanup_free char *path = NULL;
  int ret;
  struct fuse_entry_param e;
  char wd_tmp_file_name[32];

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_link(ino=%" PRIu64 ", newparent=%" PRIu64 ", newname=%s)\n", ino, newparent, newname);

  if (strlen (newname) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  newparentnode = do_lookup_file (lo, newparent, NULL);
  if (newparentnode == NULL || newparentnode->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  destnode = do_lookup_file (lo, newparent, newname);
  if (destnode && ! destnode->whiteout)
    {
      fuse_reply_err (req, EEXIST);
      return;
    }

  newparentnode = get_node_up (lo, newparentnode);
  if (newparentnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (delete_whiteout (lo, -1, newparentnode, newname) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

  ret = asprintf (&path, "%s/%s", newparentnode->path, newname);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  ret = direct_linkat (get_upper_layer (lo), node->path, path, 0);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  node = make_ovl_node (lo, path, get_upper_layer (lo), newname, node->tmp_ino, node->tmp_dev, false, newparentnode, lo->fast_ino_check);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }
  if (destnode && ! destnode->whiteout)
    node->last_layer = get_upper_layer (lo);

  node = insert_node (newparentnode, node, true);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  memset (&e, 0, sizeof (e));

  ret = rpl_stat (req, node, -1, NULL, NULL, &e.attr);
  if (ret)
    {
      fuse_reply_err (req, errno);
      return;
    }

  e.ino = node_to_inode (node);
  node->ino->lookups++;
  e.attr_timeout = get_timeout (lo);
  e.entry_timeout = get_timeout (lo);
  fuse_reply_entry (req, &e);
}

static int
direct_symlinkat (struct ovl_layer *l, const char *target, const char *linkpath, uid_t uid, gid_t gid)
{
  struct ovl_data *lo = l->ovl_data;
  char wd_tmp_file_name[32];
  int ret;

  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());

  unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
  ret = symlinkat (linkpath, lo->workdir_fd, wd_tmp_file_name);
  if (ret < 0)
    return ret;

  if (uid != lo->uid || gid != lo->gid || l->stat_override_mode != STAT_OVERRIDE_NONE)
    {
      ret = do_fchownat (lo, lo->workdir_fd, wd_tmp_file_name, uid, gid, 0755, AT_SYMLINK_NOFOLLOW);
      if (ret < 0)
        {
          unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
          return ret;
        }
    }

  ret = renameat (lo->workdir_fd, wd_tmp_file_name, get_upper_layer (lo)->fd, target);
  if (ret < 0)
    {
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return ret;
    }

  return 0;
}

static void
ovl_symlink (fuse_req_t req, const char *link, fuse_ino_t parent, const char *name)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *pnode, *node;
  int ret;
  struct fuse_entry_param e;
  const struct fuse_ctx *ctx = fuse_req_ctx (req);
  char wd_tmp_file_name[32];
  bool need_delete_whiteout = true;
  cleanup_free char *path = NULL;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_symlink(link=%s, ino=%" PRIu64 ", name=%s)\n", link, parent, name);

  if (strlen (name) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  pnode = do_lookup_file (lo, parent, NULL);
  if (pnode == NULL || pnode->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  node = do_lookup_file (lo, parent, name);
  if (node != NULL && ! node->whiteout)
    {
      fuse_reply_err (req, EEXIST);
      return;
    }

  if (pnode->loaded && node == NULL)
    need_delete_whiteout = false;

  ret = asprintf (&path, "%s/%s", pnode->path, name);
  if (ret < 0)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  ret = direct_symlinkat (get_upper_layer (lo), path, link, get_uid (lo, ctx->uid), get_gid (lo, ctx->gid));
  if (ret < 0)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  if (need_delete_whiteout && delete_whiteout (lo, -1, pnode, name) < 0)
    {
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      fuse_reply_err (req, errno);
      return;
    }

  node = make_ovl_node (lo, path, get_upper_layer (lo), name, 0, 0, false, pnode, lo->fast_ino_check);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  node = insert_node (pnode, node, true);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  memset (&e, 0, sizeof (e));

  ret = rpl_stat (req, node, -1, NULL, NULL, &e.attr);
  if (ret)
    {
      fuse_reply_err (req, errno);
      return;
    }

  e.ino = node_to_inode (node);
  node->ino->lookups++;
  e.attr_timeout = get_timeout (lo);
  e.entry_timeout = get_timeout (lo);
  fuse_reply_entry (req, &e);
}

static void
ovl_rename_exchange (fuse_req_t req, fuse_ino_t parent, const char *name,
                     fuse_ino_t newparent, const char *newname,
                     unsigned int flags)
{
  struct ovl_node *pnode, *node, *destnode, *destpnode;
  struct ovl_data *lo = ovl_data (req);
  int ret;
  cleanup_close int srcfd = -1;
  cleanup_close int destfd = -1;
  struct ovl_node *rm1, *rm2;
  char *tmp;

  node = do_lookup_file (lo, parent, name);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (node_dirp (node))
    {
      node = reload_dir (lo, node);
      if (node == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }

      if (node->layer != get_upper_layer (lo) || node->last_layer != get_upper_layer (lo))
        {
          fuse_reply_err (req, EXDEV);
          return;
        }
    }
  pnode = node->parent;

  destpnode = do_lookup_file (lo, newparent, NULL);
  destnode = NULL;

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    goto error;

  ret = TEMP_FAILURE_RETRY (safe_openat (node_dirfd (pnode), pnode->path, O_DIRECTORY, 0));
  if (ret < 0)
    goto error;
  srcfd = ret;

  destpnode = get_node_up (lo, destpnode);
  if (destpnode == NULL)
    goto error;

  ret = TEMP_FAILURE_RETRY (safe_openat (node_dirfd (destpnode), destpnode->path, O_DIRECTORY, 0));
  if (ret < 0)
    goto error;
  destfd = ret;

  destnode = do_lookup_file (lo, newparent, newname);

  node = get_node_up (lo, node);
  if (node == NULL)
    goto error;

  if (destnode == NULL)
    {
      errno = ENOENT;
      goto error;
    }
  if (node_dirp (node) && destnode->last_layer != get_upper_layer (lo))
    {
      fuse_reply_err (req, EXDEV);
      return;
    }
  destnode = get_node_up (lo, destnode);
  if (destnode == NULL)
    goto error;

  ret = direct_renameat2 (srcfd, name, destfd, newname, flags);
  if (ret < 0)
    goto error;

  rm1 = hash_remove (destpnode->children, destnode);
  rm2 = hash_remove (pnode->children, node);

  tmp = node->path;
  node->path = destnode->path;
  destnode->path = tmp;

  tmp = node->name;
  node_set_name (node, destnode->name);
  node_set_name (destnode, tmp);

  node = insert_node (destpnode, node, true);
  if (node == NULL)
    {
      node_free (rm1);
      node_free (rm2);
      goto error;
    }
  destnode = insert_node (pnode, destnode, true);
  if (destnode == NULL)
    {
      node_free (rm1);
      node_free (rm2);
      goto error;
    }
  if ((update_paths (node) < 0) || (update_paths (destnode) < 0))
    goto error;

  if (delete_whiteout (lo, destfd, NULL, newname) < 0)
    goto error;

  ret = 0;
  goto cleanup;

error:
  ret = -1;

cleanup:
  fuse_reply_err (req, ret == 0 ? 0 : errno);
}

static void
ovl_rename_direct (fuse_req_t req, fuse_ino_t parent, const char *name,
                   fuse_ino_t newparent, const char *newname,
                   unsigned int flags)
{
  struct ovl_node *pnode, *node, *destnode, *destpnode;
  struct ovl_data *lo = ovl_data (req);
  int ret;
  cleanup_close int srcfd = -1;
  cleanup_close int destfd = -1;
  struct ovl_node key;
  bool destnode_is_whiteout = false;

  pnode = do_lookup_file (lo, parent, NULL);
  if (pnode == NULL || pnode->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = do_lookup_file (lo, parent, name);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (node_dirp (node))
    {
      node = reload_dir (lo, node);
      if (node == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }

      if (node->layer != get_upper_layer (lo) || node->last_layer != get_upper_layer (lo))
        {
          fuse_reply_err (req, EXDEV);
          return;
        }
    }

  destpnode = do_lookup_file (lo, newparent, NULL);
  destnode = NULL;

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    goto error;

  ret = TEMP_FAILURE_RETRY (safe_openat (node_dirfd (pnode), pnode->path, O_DIRECTORY, 0));
  if (ret < 0)
    goto error;
  srcfd = ret;

  destpnode = get_node_up (lo, destpnode);
  if (destpnode == NULL)
    goto error;

  ret = TEMP_FAILURE_RETRY (safe_openat (node_dirfd (destpnode), destpnode->path, O_DIRECTORY, 0));
  if (ret < 0)
    goto error;
  destfd = ret;

  node_set_name (&key, (char *) newname);
  destnode = hash_lookup (destpnode->children, &key);

  node = get_node_up (lo, node);
  if (node == NULL)
    goto error;

  /* If NOREPLACE flag is given, check if we should throw an error now.
     If not, just remove the flag as it might cause problems with replacing whiteouts later.  */
  if (flags & RENAME_NOREPLACE && destnode && ! destnode->whiteout)
    {
      errno = EEXIST;
      goto error;
    }
  else
    flags = flags & ~RENAME_NOREPLACE;

  if (destnode)
    {
      size_t destnode_whiteouts = 0;

      errno = EINVAL;
      if (! destnode->whiteout && destnode->tmp_ino == node->tmp_ino && destnode->tmp_dev == node->tmp_dev)
        goto error;

      destnode_is_whiteout = destnode->whiteout;

      if (! destnode->whiteout && node_dirp (destnode))
        {
          destnode = reload_dir (lo, destnode);
          if (destnode == NULL)
            goto error;

          if (count_dir_entries (destnode, &destnode_whiteouts) > 0)
            {
              errno = ENOTEMPTY;
              goto error;
            }
          if (destnode_whiteouts && empty_dir (get_upper_layer (lo), destnode->path) < 0)
            goto error;
        }

      if (node_dirp (node) && create_missing_whiteouts (lo, node, destnode->path) < 0)
        goto error;

      if (destnode->ino->lookups > 0)
        node_free (destnode);
      else
        {
          node_free (destnode);
          destnode = NULL;
        }

      if (destnode && ! destnode_is_whiteout)
        {
          /* If the node is still accessible then be sure we
             can write to it.  Fix it to be done when a write is
             really done, not now.  */
          destnode = get_node_up (lo, destnode);
          if (destnode == NULL)
            {
              fuse_reply_err (req, errno);
              return;
            }

          if (hide_node (lo, destnode, true) < 0)
            goto error;
        }
    }

  /* If the node is a directory we must ensure there is no whiteout at the
     destination, otherwise the renameat2 will fail.  Create a .wh.$NAME style
     whiteout file until the renameat2 is completed.  */
  if (node_dirp (node))
    {
      ret = create_whiteout (lo, destpnode, newname, true, true);
      if (ret < 0)
        goto error;
      unlinkat (destfd, newname, 0);
    }

  bool src_needs_whiteout = (node->last_layer != get_upper_layer (lo));

  if (src_needs_whiteout)
    {
      /* Trying to atomically both rename and create the whiteout.
         If destination is a whiteout, we can EXCHANGE source and destination and reuse the old whiteout.
         If not, we can try to atomically create one with the WHITEOUT flag.  */
      if (destnode_is_whiteout)
        ret = direct_renameat2 (srcfd, name, destfd, newname, flags | RENAME_EXCHANGE);
      else
        {
          if (! can_mknod)
            {
              ret = -1;
              errno = EPERM;
            }
          else
            ret = direct_renameat2 (srcfd, name, destfd, newname, flags | RENAME_WHITEOUT);
        }

      /* If atomic whiteout creation failed, fall back to separate rename and whiteout creation.  */
      if (ret < 0)
        {
          ret = direct_renameat2 (srcfd, name, destfd, newname, flags);
          if (ret < 0)
            goto error;

          ret = create_whiteout (lo, pnode, name, false, true);
          if (ret < 0)
            goto error;
        }
    }
  else
    {
      ret = direct_renameat2 (srcfd, name, destfd, newname, flags);
      if (ret < 0)
        goto error;
    }

  pnode->loaded = 0;

  /* If the destination was .wh. style whiteout, it was not replaced automatically, so delete it.  */
  if (delete_whiteout (lo, destfd, NULL, newname) < 0)
    goto error;

  hash_remove (pnode->children, node);

  free (node->name);
  node_set_name (node, strdup (newname));
  if (node->name == NULL)
    goto error;

  node = insert_node (destpnode, node, true);
  if (node == NULL)
    goto error;
  if (update_paths (node) < 0)
    goto error;

  node->loaded = 0;

  ret = 0;
  fuse_reply_err (req, 0);
  return;

error:
  ret = -1;
  fuse_reply_err (req, errno);
}

static void
ovl_rename (fuse_req_t req, fuse_ino_t parent, const char *name,
            fuse_ino_t newparent, const char *newname,
            unsigned int flags)
{
  struct ovl_node *p;
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_rename(ino=%" PRIu64 ", name=%s , ino=%" PRIu64 ", name=%s)\n", parent, name, newparent, newname);

  if (strlen (newname) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  if (flags & RENAME_EXCHANGE)
    ovl_rename_exchange (req, parent, name, newparent, newname, flags);
  else
    ovl_rename_direct (req, parent, name, newparent, newname, flags);

  /* Make sure the cache is invalidated, if the parent is in the middle of a readdir. */
  p = do_lookup_file (lo, parent, NULL);
  if (p && p->in_readdir)
    fuse_lowlevel_notify_inval_inode (lo->se, parent, 0, 0);
  p = do_lookup_file (lo, newparent, NULL);
  if (p && p->in_readdir)
    fuse_lowlevel_notify_inval_inode (lo->se, newparent, 0, 0);
}

static void
ovl_statfs (fuse_req_t req, fuse_ino_t ino)
{
  int ret;
  struct statvfs sfs;
  struct ovl_data *lo = ovl_data (req);

  ret = do_statfs (lo, &sfs);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fuse_reply_statfs (req, &sfs);
}

static void
ovl_readlink (fuse_req_t req, fuse_ino_t ino)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  cleanup_free char *buf = NULL;
  struct ovl_node *node;
  size_t current_size;
  int ret = 0;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_readlink(ino=%" PRIu64 ")\n", ino);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  current_size = PATH_MAX + 1;
  buf = malloc (current_size);
  if (buf == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  while (1)
    {
      char *tmp;

      ret = node->layer->ds->readlinkat (node->layer, node->path, buf, current_size - 1);
      if (ret == -1)
        {
          fuse_reply_err (req, errno);
          return;
        }
      if (ret < current_size - 1)
        break;

      current_size *= 2;
      tmp = realloc (buf, current_size);
      if (tmp == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }
      buf = tmp;
    }

  buf[ret] = '\0';
  fuse_reply_readlink (req, buf);
}

static void
ovl_mknod (fuse_req_t req, fuse_ino_t parent, const char *name, mode_t mode, dev_t rdev)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *pnode;
  int ret = 0;
  cleanup_free char *path = NULL;
  struct fuse_entry_param e;
  const struct fuse_ctx *ctx = fuse_req_ctx (req);
  char wd_tmp_file_name[32];
  mode_t backing_file_mode = mode | (lo->xattr_permissions ? 0755 : 0);

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_mknod(ino=%" PRIu64 ", name=%s, mode=%d, rdev=%lu)\n",
             parent, name, mode, rdev);

  if (strlen (name) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  mode = mode & ~ctx->umask;

  node = do_lookup_file (lo, parent, name);
  if (node != NULL && ! node->whiteout)
    {
      fuse_reply_err (req, EEXIST);
      return;
    }

  pnode = do_lookup_file (lo, parent, NULL);
  if (pnode == NULL)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }
  sprintf (wd_tmp_file_name, "%lu", get_next_wd_counter ());
  ret = mknodat (lo->workdir_fd, wd_tmp_file_name, backing_file_mode, rdev);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (do_fchownat (lo, lo->workdir_fd, wd_tmp_file_name, get_uid (lo, ctx->uid), get_gid (lo, ctx->gid), mode, 0) < 0)
    {
      fuse_reply_err (req, errno);
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return;
    }

  ret = asprintf (&path, "%s/%s", pnode->path, name);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return;
    }

  ret = inherit_acl (lo, pnode, -1, path);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return;
    }

  ret = renameat (lo->workdir_fd, wd_tmp_file_name, get_upper_layer (lo)->fd, path);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      unlinkat (lo->workdir_fd, wd_tmp_file_name, 0);
      return;
    }

  node = make_ovl_node (lo, path, get_upper_layer (lo), name, 0, 0, false, pnode, lo->fast_ino_check);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  node = insert_node (pnode, node, true);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  if (delete_whiteout (lo, -1, pnode, name) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  memset (&e, 0, sizeof (e));

  ret = rpl_stat (req, node, -1, NULL, NULL, &e.attr);
  if (ret)
    {
      fuse_reply_err (req, errno);
      return;
    }

  /* Make sure the cache is invalidated, if the parent is in the middle of a readdir. */
  if (pnode->in_readdir)
    fuse_lowlevel_notify_inval_inode (lo->se, parent, 0, 0);

  e.ino = node_to_inode (node);
  e.attr_timeout = get_timeout (lo);
  e.entry_timeout = get_timeout (lo);
  node->ino->lookups++;
  fuse_reply_entry (req, &e);
}

static void
ovl_mkdir (fuse_req_t req, fuse_ino_t parent, const char *name, mode_t mode)
{
  const struct fuse_ctx *ctx = fuse_req_ctx (req);
  struct ovl_data *lo = ovl_data (req);
  struct fuse_entry_param e;
  bool parent_upperdir_only;
  struct ovl_node *pnode;
  struct ovl_node *node;
  struct stat st;
  ino_t ino = 0;
  dev_t dev = 0;
  int ret = 0;
  cleanup_free char *path = NULL;
  bool need_delete_whiteout = true;
  cleanup_lock int l = enter_big_lock ();

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_mkdir(ino=%" PRIu64 ", name=%s, mode=%d)\n",
             parent, name, mode);

  if (strlen (name) > get_fs_namemax (lo))
    {
      fuse_reply_err (req, ENAMETOOLONG);
      return;
    }

  node = do_lookup_file (lo, parent, name);
  if (node != NULL && ! node->whiteout)
    {
      fuse_reply_err (req, EEXIST);
      return;
    }

  pnode = do_lookup_file (lo, parent, NULL);
  if (pnode == NULL)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  pnode = get_node_up (lo, pnode);
  if (pnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (pnode->loaded && node == NULL)
    need_delete_whiteout = false;

  parent_upperdir_only = pnode->last_layer == get_upper_layer (lo);

  ret = asprintf (&path, "%s/%s", pnode->path, name);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  ret = create_directory (lo, get_upper_layer (lo)->fd, path, NULL, pnode, NULL, -1,
                          get_uid (lo, ctx->uid), get_gid (lo, ctx->gid), mode & ~ctx->umask,
                          true, &st);
  if (ret < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (need_delete_whiteout && delete_whiteout (lo, -1, pnode, name) < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  /* if the parent is on the upper layer, it doesn't need to lookup the ino in the lower layers.  */
  if (parent_upperdir_only)
    {
      ino = st.st_ino;
      dev = st.st_dev;
    }

  node = make_ovl_node (lo, path, get_upper_layer (lo), name, ino, dev, true, pnode, lo->fast_ino_check);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  node = insert_node (pnode, node, true);
  if (node == NULL)
    {
      fuse_reply_err (req, ENOMEM);
      return;
    }

  node->last_layer = get_upper_layer (lo);
  if (get_timeout (lo) > 0)
    node->loaded = 1;

  memset (&e, 0, sizeof (e));

  ret = rpl_stat (req, node, -1, NULL, parent_upperdir_only ? &st : NULL, &e.attr);
  if (ret)
    {
      fuse_reply_err (req, errno);
      return;
    }

  e.ino = node_to_inode (node);
  e.attr_timeout = get_timeout (lo);
  e.entry_timeout = get_timeout (lo);
  node->ino->lookups++;
  pnode->n_links++;
  fuse_reply_entry (req, &e);
}

static int
direct_fsync (struct ovl_layer *l, int fd, const char *path, int datasync)
{
  cleanup_close int cfd = -1;

  if (fd < 0)
    {
      cfd = safe_openat (l->fd, path, O_NOFOLLOW | O_DIRECTORY, 0);
      if (cfd < 0)
        return cfd;
      fd = cfd;
    }

  return datasync ? fdatasync (fd) : fsync (fd);
}

static void
do_fsync (fuse_req_t req, fuse_ino_t ino, int datasync, int fd)
{
  int ret = 0;
  bool do_fsync;
  struct ovl_node *node;
  struct ovl_data *lo = ovl_data (req);
  cleanup_lock int l = 0;

  if (! lo->fsync)
    {
      fuse_reply_err (req, ENOSYS);
      return;
    }

  l = enter_big_lock ();

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  /* Skip fsync for lower layers.  */
  do_fsync = node && node->layer == get_upper_layer (lo);

  if (node->layer == NULL)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  if (! do_fsync)
    {
      fuse_reply_err (req, 0);
      return;
    }

  if (do_fsync)
    ret = direct_fsync (node->layer, fd, node->path, datasync);

  fuse_reply_err (req, ret == 0 ? 0 : errno);
}

static void
ovl_fsync (fuse_req_t req, fuse_ino_t ino, int datasync, struct fuse_file_info *fi)
{
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_fsync(ino=%" PRIu64 ", datasync=%d, fi=%p)\n",
             ino, datasync, fi);

  return do_fsync (req, ino, datasync, fi->fh);
}

static void
ovl_fsyncdir (fuse_req_t req, fuse_ino_t ino, int datasync, struct fuse_file_info *fi)
{
  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_fsyncdir(ino=%" PRIu64 ", datasync=%d, fi=%p)\n",
             ino, datasync, fi);

  return do_fsync (req, ino, datasync, -1);
}

static void
ovl_ioctl (fuse_req_t req, fuse_ino_t ino, int cmd, void *arg,
           struct fuse_file_info *fi, unsigned int flags,
           const void *in_buf, size_t in_bufsz, size_t out_bufsz)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  cleanup_close int cleaned_fd = -1;
  struct ovl_node *node;
  int fd = -1;
  int r = 0;

  if (flags & FUSE_IOCTL_COMPAT)
    {
      fuse_reply_err (req, ENOSYS);
      return;
    }

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_ioctl(ino=%" PRIu64 ", cmd=%d, arg=%p, fi=%p, flags=%d, buf=%p, in_bufsz=%zu, out_bufsz=%zu)\n",
             ino, cmd, arg, fi, flags, in_buf, in_bufsz, out_bufsz);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  switch (cmd)
    {
    case FS_IOC_GETVERSION:
    case FS_IOC_GETFLAGS:
      if (! node_dirp (node))
        fd = fi->fh;
      break;

    case FS_IOC_SETVERSION:
    case FS_IOC_SETFLAGS:
      node = get_node_up (lo, node);
      if (node == NULL)
        {
          fuse_reply_err (req, errno);
          return;
        }
      if (in_bufsz >= sizeof (r))
        r = *(unsigned long *) in_buf;
      break;

    default:
      fuse_reply_err (req, ENOSYS);
      return;
    }

  if (fd < 0)
    {
      fd = cleaned_fd = node->layer->ds->openat (node->layer, node->path, O_RDONLY | O_NONBLOCK, 0755);
      if (fd < 0)
        {
          fuse_reply_err (req, errno);
          return;
        }
    }

  l = release_big_lock ();

  if (ioctl (fd, cmd, &r, sizeof (r)) < 0)
    fuse_reply_err (req, errno);
  else
    fuse_reply_ioctl (req, 0, &r, out_bufsz ? sizeof (r) : 0);
}

static int
direct_fallocate (struct ovl_layer *l, int fd, int mode, off_t offset, off_t len)
{
  return fallocate (fd, mode, offset, len);
}

static void
ovl_fallocate (fuse_req_t req, fuse_ino_t ino, int mode, off_t offset, off_t length, struct fuse_file_info *fi)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  cleanup_close int fd = -1;
  struct ovl_node *node;
  int dirfd;
  int ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_fallocate(ino=%" PRIu64 ", mode=%d, offset=%lo, length=%lu, fi=%p)\n",
             ino, mode, offset, length, fi);

  node = do_lookup_file (lo, ino, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  node = get_node_up (lo, node);
  if (node == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  dirfd = node_dirfd (node);
  fd = safe_openat (dirfd, node->path, O_NONBLOCK | O_NOFOLLOW | O_WRONLY, 0);
  if (fd < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  l = release_big_lock ();

  ret = direct_fallocate (node->layer, fd, mode, offset, length);
  fuse_reply_err (req, ret < 0 ? errno : 0);
}

#ifdef HAVE_COPY_FILE_RANGE

static ssize_t
direct_copy_file_range (struct ovl_layer *l, int fd_in, off_t *off_in,
                        int fd_out, off_t *off_out,
                        size_t len, unsigned int flags)
{
  return copy_file_range (fd_in, off_in, fd_out, off_out, len, flags);
}

static void
ovl_copy_file_range (fuse_req_t req, fuse_ino_t ino_in, off_t off_in, struct fuse_file_info *fi_in, fuse_ino_t ino_out, off_t off_out, struct fuse_file_info *fi_out, size_t len, int flags)
{
  cleanup_lock int l = enter_big_lock ();
  struct ovl_data *lo = ovl_data (req);
  struct ovl_node *node, *dnode;
  cleanup_close int fd_dest = -1;
  cleanup_close int fd = -1;
  ssize_t ret;

  if (UNLIKELY (ovl_debug (req)))
    fprintf (stderr, "ovl_copy_file_range(ino_in=%" PRIu64 ", off_in=%lo, fi_in=%p, ino_out=%" PRIu64 ", off_out=%lo, fi_out=%p, size=%zu, flags=%d)\n",
             ino_in, off_in, fi_in, ino_out, off_out, fi_out, len, flags);

  node = do_lookup_file (lo, ino_in, NULL);
  if (node == NULL || node->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  dnode = do_lookup_file (lo, ino_out, NULL);
  if (dnode == NULL || dnode->whiteout)
    {
      fuse_reply_err (req, ENOENT);
      return;
    }

  dnode = get_node_up (lo, dnode);
  if (dnode == NULL)
    {
      fuse_reply_err (req, errno);
      return;
    }

  if (node->hidden)
    fd = openat (node->hidden_dirfd, node->path, O_NONBLOCK | O_NOFOLLOW | O_RDONLY, 0755);
  else
    fd = node->layer->ds->openat (node->layer, node->path, O_NONBLOCK | O_NOFOLLOW | O_RDONLY, 0755);
  if (fd < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  fd_dest = TEMP_FAILURE_RETRY (safe_openat (node_dirfd (dnode), dnode->path, O_NONBLOCK | O_NOFOLLOW | O_WRONLY, 0));
  if (fd_dest < 0)
    {
      fuse_reply_err (req, errno);
      return;
    }

  l = release_big_lock ();

  ret = direct_copy_file_range (node->layer, fd, &off_in, fd_dest, &off_out, len, flags);
  if (ret < 0)
    fuse_reply_err (req, errno);
  else
    fuse_reply_write (req, ret);
}
#endif

static struct fuse_lowlevel_ops ovl_oper = {
  .statfs = ovl_statfs,
  .access = ovl_access,
  .getxattr = ovl_getxattr,
  .removexattr = ovl_removexattr,
  .setxattr = ovl_setxattr,
  .listxattr = ovl_listxattr,
  .init = ovl_init,
  .lookup = ovl_lookup,
  .forget = ovl_forget,
  .forget_multi = ovl_forget_multi,
  .getattr = ovl_getattr,
  .readlink = ovl_readlink,
  .opendir = ovl_opendir,
  .readdir = ovl_readdir,
  .readdirplus = ovl_readdirplus,
  .releasedir = ovl_releasedir,
  .create = ovl_create,
  .open = ovl_open,
  .release = ovl_release,
  .read = ovl_read,
  .write_buf = ovl_write_buf,
  .unlink = ovl_unlink,
  .rmdir = ovl_rmdir,
  .setattr = ovl_setattr,
  .symlink = ovl_symlink,
  .rename = ovl_rename,
  .mkdir = ovl_mkdir,
  .mknod = ovl_mknod,
  .link = ovl_link,
  .fsync = ovl_fsync,
  .fsyncdir = ovl_fsyncdir,
  .ioctl = ovl_ioctl,
  .fallocate = ovl_fallocate,
#if HAVE_COPY_FILE_RANGE && HAVE_FUSE_COPY_FILE_RANGE
  .copy_file_range = ovl_copy_file_range,
#endif
};

static int
fuse_opt_proc (void *data, const char *arg, int key, struct fuse_args *outargs)
{
  struct ovl_data *ovl_data = data;

  if (strcmp (arg, "-f") == 0)
    return 1;
  if (strcmp (arg, "--help") == 0)
    return 1;
  if (strcmp (arg, "-h") == 0)
    return 1;
  if (strcmp (arg, "--version") == 0)
    return 1;
  if (strcmp (arg, "-V") == 0)
    return 1;
  if ((strcmp (arg, "--debug") == 0) || (strcmp (arg, "-d") == 0) || (strcmp (arg, "debug") == 0))
    {
      ovl_data->debug = 1;
      return 1;
    }

  if (strcmp (arg, "allow_root") == 0)
    return 1;
  if (strcmp (arg, "default_permissions") == 0)
    return 1;
  if (strcmp (arg, "allow_other") == 0)
    return 1;
  if (strcmp (arg, "suid") == 0)
    return 1;
  if (strcmp (arg, "dev") == 0)
    return 1;
  if (strcmp (arg, "nosuid") == 0)
    return 1;
  if (strcmp (arg, "nodev") == 0)
    return 1;
  if (strcmp (arg, "exec") == 0)
    return 1;
  if (strcmp (arg, "noexec") == 0)
    return 1;
  if (strcmp (arg, "atime") == 0)
    return 1;
  if (strcmp (arg, "noatime") == 0)
    return 1;
  if (strcmp (arg, "diratime") == 0)
    return 1;
  if (strcmp (arg, "nodiratime") == 0)
    return 1;
  if (strcmp (arg, "splice_write") == 0)
    return 1;
  if (strcmp (arg, "splice_read") == 0)
    return 1;
  if (strcmp (arg, "splice_move") == 0)
    return 1;
  if (strcmp (arg, "kernel_cache") == 0)
    return 1;
  if (strcmp (arg, "max_write") == 0)
    return 1;
  if (strcmp (arg, "ro") == 0)
    return 1;

  if (key >= 1 && key <= 3)
    {
      ovl_data->nfs_filehandles = key - 1;
      return 0;
    }

  if (key == FUSE_OPT_KEY_NONOPT)
    {
      if (ovl_data->mountpoint)
        free (ovl_data->mountpoint);

      ovl_data->mountpoint = strdup (arg);
      return 0;
    }
  /* Ignore unknown arguments.  */
  if (key == -1)
    {
      fprintf (stderr, "unknown argument ignored: %s\n", arg);
      return 0;
    }

  return 1;
}

char **
get_new_args (int *argc, char **argv)
{
  int i;
  char **newargv;

  newargv = malloc (sizeof (char *) * (*argc + 2));
  if (newargv == NULL)
    error (EXIT_FAILURE, 0, "error allocating memory");

  newargv[0] = argv[0];
  if (geteuid () == 0)
    newargv[1] = "-odefault_permissions,allow_other,suid,noatime";
  else
    newargv[1] = "-odefault_permissions,noatime";
  for (i = 1; i < *argc; i++)
    newargv[i + 1] = argv[i];
  (*argc)++;
  return newargv;
}

static void
set_limits ()
{
  struct rlimit l;

  if (getrlimit (RLIMIT_NOFILE, &l) < 0)
    error (EXIT_FAILURE, errno, "cannot read nofile rlimit");

  /* Set the soft limit to the hard limit.  */
  l.rlim_cur = l.rlim_max;

  if (setrlimit (RLIMIT_NOFILE, &l) < 0)
    error (EXIT_FAILURE, errno, "cannot set nofile rlimit");
}

static char *
load_default_plugins ()
{
  DIR *dp = NULL;
  char *plugins = strdup ("");

  dp = opendir (PKGLIBEXECDIR);
  if (dp == NULL)
    return plugins;

  for (;;)
    {
      struct dirent *dent;

      dent = readdir (dp);
      if (dent == NULL)
        break;
      if (dent->d_type != DT_DIR)
        {
          char *new_plugins = NULL;
          if (asprintf (&new_plugins, "%s/%s:%s", PKGLIBEXECDIR, dent->d_name, plugins) < 0)
            {
              free (plugins);
              closedir (dp);
              return NULL;
            }
          free (plugins);
          plugins = new_plugins;
        }
    }
  closedir (dp);

  return plugins;
}

int
main (int argc, char *argv[])
{
  struct fuse_session *se;
  struct fuse_cmdline_opts opts;
  char **newargv = get_new_args (&argc, argv);
  struct ovl_data lo = {
    .debug = 0,
    .uid_mappings = NULL,
    .gid_mappings = NULL,
    .uid_str = NULL,
    .gid_str = NULL,
    .root = NULL,
    .lowerdir = NULL,
    .redirect_dir = NULL,
    .mountpoint = NULL,
    .fsync = 1,
    .noacl = 0,
    .squash_to_uid = -1,
    .squash_to_gid = -1,
    .static_nlink = 0,
    .ino_passthrough = 1,
    .nfs_filehandles = 0,
    .ino_t_32 = 0,
    .xattr_permissions = 0,
    .euid = geteuid (),
    .timeout = 1000000000.0,
    .timeout_str = NULL,
    .writeback = 1,
    .volatile_mode = 0,
  };
  struct fuse_loop_config fuse_conf = {
    .clone_fd = 1,
    .max_idle_threads = 10,
  };
  int ret = -1;
  cleanup_layer struct ovl_layer *layers = NULL;
  struct ovl_layer *tmp_layer = NULL;
  struct fuse_args args = FUSE_ARGS_INIT (argc, newargv);

  memset (&opts, 0, sizeof (opts));
  if (fuse_opt_parse (&args, &lo, ovl_opts, fuse_opt_proc) == -1)
    error (EXIT_FAILURE, 0, "error parsing options");
  if (fuse_parse_cmdline (&args, &opts) != 0)
    error (EXIT_FAILURE, 0, "error parsing cmdline");

  if (opts.mountpoint)
    free (opts.mountpoint);

  read_overflowids ();

  pthread_mutex_init (&lock, NULL);

  if (opts.show_help)
    {
      printf ("usage: %s [options] <mountpoint>\n\n", argv[0]);
      fuse_cmdline_help ();
      fuse_lowlevel_help ();
      exit (EXIT_SUCCESS);
    }
  else if (opts.show_version)
    {
      printf ("fuse-overlayfs: version %s\n", PACKAGE_VERSION);
      printf ("FUSE library version %s\n", fuse_pkgversion ());
      fuse_lowlevel_version ();
      exit (EXIT_SUCCESS);
    }

  lo.uid = geteuid ();
  lo.gid = getegid ();

  if (lo.redirect_dir && strcmp (lo.redirect_dir, "off"))
    error (EXIT_FAILURE, 0, "fuse-overlayfs only supports redirect_dir=off");

  if (lo.mountpoint == NULL)
    error (EXIT_FAILURE, 0, "no mountpoint specified");

  unescape (lo.workdir);

  set_limits ();
  check_can_mknod (&lo);
  check_writeable_proc ();

  if (lo.volatile_mode)
    lo.fsync = 0;

  if (lo.debug)
    {
      fprintf (stderr, "uid=%s\n", lo.uid_str ?: "unchanged");
      fprintf (stderr, "gid=%s\n", lo.gid_str ?: "unchanged");
      fprintf (stderr, "upperdir=%s\n", lo.upperdir ? lo.upperdir : "NOT USED");
      fprintf (stderr, "workdir=%s\n", lo.workdir ? lo.workdir : "NOT USED");
      fprintf (stderr, "lowerdir=%s\n", lo.lowerdir);
      fprintf (stderr, "mountpoint=%s\n", lo.mountpoint);
      fprintf (stderr, "plugins=%s\n", lo.plugins ? lo.plugins : "<none>");
      fprintf (stderr, "fsync=%s\n", lo.fsync ? "enabled" : "disabled");
    }

  lo.uid_mappings = lo.uid_str ? read_mappings (lo.uid_str) : NULL;
  lo.gid_mappings = lo.gid_str ? read_mappings (lo.gid_str) : NULL;

  errno = 0;
  if (lo.timeout_str)
    {
      lo.timeout = strtod (lo.timeout_str, NULL);
      if (errno == ERANGE)
        error (EXIT_FAILURE, errno, "cannot convert %s", lo.timeout_str);
    }

  if (lo.plugins == NULL)
    lo.plugins = load_default_plugins ();

  lo.plugins_ctx = load_plugins (lo.plugins);

  layers = read_dirs (&lo, lo.lowerdir, true, NULL);
  if (layers == NULL)
    {
      error (EXIT_FAILURE, errno, "cannot read lower dirs");
    }

  if (lo.upperdir != NULL)
    {
      tmp_layer = read_dirs (&lo, lo.upperdir, false, layers);
      if (tmp_layer == NULL)
        error (EXIT_FAILURE, errno, "cannot read upper dir");
      layers = tmp_layer;
    }

  lo.layers = layers;

  for (tmp_layer = layers; tmp_layer; tmp_layer = tmp_layer->next)
    {
      if (tmp_layer->st_dev != layers->st_dev)
        lo.ino_passthrough = 0;
      if (tmp_layer->nfs_filehandles == -1)
        {
          if (lo.nfs_filehandles == 2)
            error (EXIT_FAILURE, 0, "Unexpected error on %s while checking for nfs filehandle support", tmp_layer->path);
          else
            warn ("Unexpected error on %s while checking for nfs filehandle support", tmp_layer->path);
          lo.nfs_filehandles = 0;
        }
      else if (tmp_layer->nfs_filehandles == 0)
        {
          if (lo.nfs_filehandles == 2)
            error (EXIT_FAILURE, 0, "xino=on but %s does not support nfs filehandles", tmp_layer->path);
          lo.nfs_filehandles = 0;
        }
      if (! tmp_layer->ds->support_acls (tmp_layer))
        lo.noacl = 1;
    }

  if (lo.upperdir)
    {
      if (lo.xattr_permissions)
        {
          const char *name = NULL;
          char data[64];
          ssize_t s;
          if (lo.xattr_permissions == 1)
            {
              get_upper_layer (&lo)->stat_override_mode = STAT_OVERRIDE_PRIVILEGED;
              name = XATTR_PRIVILEGED_OVERRIDE_STAT;
            }
          else if (lo.xattr_permissions == 2)
            {
              get_upper_layer (&lo)->stat_override_mode = STAT_OVERRIDE_CONTAINERS;
              name = XATTR_OVERRIDE_CONTAINERS_STAT;
            }
          else
            error (EXIT_FAILURE, 0, "invalid value for xattr_permissions");

          s = fgetxattr (get_upper_layer (&lo)->fd, name, data, sizeof (data));
          if (s < 0 && errno == ENODATA && lo.xattr_permissions == 2)
            {
              s = fgetxattr (get_upper_layer (&lo)->fd, XATTR_OVERRIDE_STAT, data, sizeof (data));
              if (s >= 0)
                {
                  get_upper_layer (&lo)->stat_override_mode = STAT_OVERRIDE_USER;
                  name = XATTR_OVERRIDE_STAT;
                }
            }
          if (s < 0)
            {
              bool found = false;
              struct ovl_layer *l;

              if (errno != ENODATA)
                error (EXIT_FAILURE, errno, "read xattr `%s` from upperdir", name);

              for (l = get_lower_layers (&lo); l; l = l->next)
                {
                  switch (lo.xattr_permissions)
                    {
                    case 1:
                      s = fgetxattr (l->fd, name, data, sizeof (data));
                      break;

                    case 2:
                      s = fgetxattr (l->fd, XATTR_OVERRIDE_CONTAINERS_STAT, data, sizeof (data));
                      if (s < 0 && errno == ENODATA)
                        s = fgetxattr (l->fd, XATTR_OVERRIDE_STAT, data, sizeof (data));
                      break;
                    }

                  if (s > 0)
                    {
                      ret = fsetxattr (get_upper_layer (&lo)->fd, name, data, s, 0);
                      if (ret < 0)
                        error (EXIT_FAILURE, errno, "fsetxattr mode to upper layer");
                      found = true;
                      break;
                    }
                }
              if (! found)
                {
                  /* If the mode is missing, set a standard value.  */
                  ret = write_permission_xattr (&lo, get_upper_layer (&lo)->fd, get_upper_layer (&lo)->path, 0, 0, 0555);
                  if (ret < 0)
                    error (EXIT_FAILURE, errno, "write xattr `%s` to upperdir", name);
                }
            }
        }
    }

  lo.inodes = hash_initialize (2048, NULL, node_inode_hasher, node_inode_compare, inode_free);

  lo.root = load_dir (&lo, NULL, lo.layers, ".", "");
  if (lo.root == NULL)
    error (EXIT_FAILURE, errno, "cannot read upper dir");
  lo.root->ino->lookups = 2;

  if (lo.workdir == NULL && lo.upperdir != NULL)
    error (EXIT_FAILURE, 0, "workdir not specified");

  if (lo.workdir)
    {
      int dfd;
      cleanup_free char *path = NULL;

      path = realpath (lo.workdir, NULL);
      if (path == NULL)
        goto err_out1;
      mkdir (path, 0700);
      path = realloc (path, strlen (path) + strlen ("/work") + 1);
      if (! path)
        error (EXIT_FAILURE, errno, "allocating workdir path");
      strcat (path, "/work");
      mkdir (path, 0700);
      free (lo.workdir);
      lo.workdir = strdup (path);
      if (lo.workdir == NULL)
        error (EXIT_FAILURE, errno, "allocating workdir path");

      lo.workdir_fd = open (lo.workdir, O_DIRECTORY);
      if (lo.workdir_fd < 0)
        error (EXIT_FAILURE, errno, "cannot open workdir");

      dfd = dup (lo.workdir_fd);
      if (dfd < 0)
        error (EXIT_FAILURE, errno, "dup workdir file descriptor");
      empty_dirfd (dfd);
    }

  umask (0);
  disable_locking = ! lo.threaded;

  se = fuse_session_new (&args, &ovl_oper, sizeof (ovl_oper), &lo);
  lo.se = se;
  if (se == NULL)
    {
      error (0, errno, "cannot create FUSE session");
      goto err_out1;
    }
  if (fuse_set_signal_handlers (se) != 0)
    {
      error (0, errno, "cannot set signal handler");
      goto err_out2;
    }

  signal (SIGUSR1, print_stats);

  if (fuse_session_mount (se, lo.mountpoint) != 0)
    {
      error (0, errno, "cannot mount");
      goto err_out3;
    }
  fuse_daemonize (opts.foreground);

  if (lo.threaded)
    ret = fuse_session_loop_mt (se, &fuse_conf);
  else
    ret = fuse_session_loop (se);

  fuse_session_unmount (se);
err_out3:
  fuse_remove_signal_handlers (se);
err_out2:
  fuse_session_destroy (se);
err_out1:

  for (tmp_layer = lo.layers; tmp_layer; tmp_layer = tmp_layer->next)
    tmp_layer->ds->cleanup (tmp_layer);

  node_mark_all_free (lo.root);

  hash_free (lo.inodes);

  plugin_free_all (lo.plugins_ctx);

  free_mapping (lo.uid_mappings);
  free_mapping (lo.gid_mappings);

  close (lo.workdir_fd);

  fuse_opt_free_args (&args);

  exit (ret ? EXIT_FAILURE : EXIT_SUCCESS);
  return 1;
}