passenger 3.0.21 → 3.9.1.beta

data/ext/libeio/eio.c

@@ -0,0 +1,2816 @@
+/*
+ * libeio implementation
+ *
+ * Copyright (c) 2007,2008,2009,2010,2011,2012 Marc Alexander Lehmann <libeio@schmorp.de>
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modifica-
+ * tion, are permitted provided that the following conditions are met:
+ * 
+ *   1.  Redistributions of source code must retain the above copyright notice,
+ *       this list of conditions and the following disclaimer.
+ * 
+ *   2.  Redistributions in binary form must reproduce the above copyright
+ *       notice, this list of conditions and the following disclaimer in the
+ *       documentation and/or other materials provided with the distribution.
+ * 
+ * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
+ * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER-
+ * CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.  IN NO
+ * EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE-
+ * CIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+ * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
+ * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
+ * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTH-
+ * ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ * OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * Alternatively, the contents of this file may be used under the terms of
+ * the GNU General Public License ("GPL") version 2 or any later version,
+ * in which case the provisions of the GPL are applicable instead of
+ * the above. If you wish to allow the use of your version of this file
+ * only under the terms of the GPL and not to allow others to use your
+ * version of this file under the BSD license, indicate your decision
+ * by deleting the provisions above and replace them with the notice
+ * and other provisions required by the GPL. If you do not delete the
+ * provisions above, a recipient may use your version of this file under
+ * either the BSD or the GPL.
+ */
+
+#ifndef _WIN32
+# include "config.h"
+#endif
+
+#include "eio.h"
+#include "ecb.h"
+
+#ifdef EIO_STACKSIZE
+# define X_STACKSIZE EIO_STACKSIZE
+#endif
+#include "xthread.h"
+
+#include <errno.h>
+#include <stddef.h>
+#include <stdlib.h>
+#include <string.h>
+#include <errno.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <limits.h>
+#include <fcntl.h>
+#include <assert.h>
+
+/* intptr_t comes from unistd.h, says POSIX/UNIX/tradition */
+/* intptr_t only comes from stdint.h, says idiot openbsd coder */
+#if HAVE_STDINT_H
+# include <stdint.h>
+#endif
+
+#ifndef ECANCELED
+# define ECANCELED EDOM
+#endif
+#ifndef ELOOP
+# define ELOOP EDOM
+#endif
+
+#if !defined(ENOTSOCK) && defined(WSAENOTSOCK)
+# define ENOTSOCK WSAENOTSOCK
+#endif
+
+static void eio_destroy (eio_req *req);
+
+#ifndef EIO_FINISH
+# define EIO_FINISH(req)  ((req)->finish) && !EIO_CANCELLED (req) ? (req)->finish (req) : 0
+#endif
+
+#ifndef EIO_DESTROY
+# define EIO_DESTROY(req) do { if ((req)->destroy) (req)->destroy (req); } while (0)
+#endif
+
+#ifndef EIO_FEED
+# define EIO_FEED(req)    do { if ((req)->feed   ) (req)->feed    (req); } while (0)
+#endif
+
+#ifndef EIO_FD_TO_WIN32_HANDLE
+# define EIO_FD_TO_WIN32_HANDLE(fd) _get_osfhandle (fd)
+#endif
+#ifndef EIO_WIN32_HANDLE_TO_FD
+# define EIO_WIN32_HANDLE_TO_FD(handle) _open_osfhandle (handle, 0)
+#endif
+
+#define EIO_ERRNO(errval,retval) ((errno = errval), retval)
+
+#define EIO_ENOSYS() EIO_ERRNO (ENOSYS, -1)
+
+#ifdef _WIN32
+
+  #undef PAGESIZE
+  #define PAGESIZE 4096 /* GetSystemInfo? */
+
+  /* TODO: look at how perl does stat (non-sloppy), unlink (ro-files), utime, link */
+
+  #ifdef EIO_STRUCT_STATI64
+    /* look at perl's non-sloppy stat */
+    #define stat(path,buf)       _stati64 (path,buf)
+    #define fstat(fd,buf)        _fstati64 (fd,buf)
+  #endif
+  #define lstat(path,buf)      stat (path,buf)
+  #define fsync(fd)            (FlushFileBuffers ((HANDLE)EIO_FD_TO_WIN32_HANDLE (fd)) ? 0 : EIO_ERRNO (EBADF, -1))
+  #define mkdir(path,mode)     _mkdir (path)
+  #define link(old,neu)        (CreateHardLink (neu, old, 0) ? 0 : EIO_ERRNO (ENOENT, -1))
+
+  #define chmod(path,mode)     _chmod (path, mode)
+  #define dup(fd)              _dup (fd)
+  #define dup2(fd1,fd2)        _dup2 (fd1, fd2)
+
+  #define fchmod(fd,mode)      EIO_ENOSYS ()
+  #define chown(path,uid,gid)  EIO_ENOSYS ()
+  #define fchown(fd,uid,gid)   EIO_ENOSYS ()
+  #define truncate(path,offs)  EIO_ENOSYS () /* far-miss: SetEndOfFile */
+  #define ftruncate(fd,offs)   EIO_ENOSYS () /* near-miss: SetEndOfFile */
+  #define mknod(path,mode,dev) EIO_ENOSYS ()
+  #define sync()               EIO_ENOSYS ()
+  #define readlink(path,buf,s) EIO_ENOSYS ()
+  #define statvfs(path,buf)    EIO_ENOSYS ()
+  #define fstatvfs(fd,buf)     EIO_ENOSYS ()
+
+  /* rename() uses MoveFile, which fails to overwrite */
+  #define rename(old,neu)      eio__rename (old, neu)
+
+  static int
+  eio__rename (const char *old, const char *neu)
+  {
+    if (MoveFileEx (old, neu, MOVEFILE_REPLACE_EXISTING))
+      return 0;
+
+    /* should steal _dosmaperr */
+    switch (GetLastError ())
+      {
+        case ERROR_FILE_NOT_FOUND:
+        case ERROR_PATH_NOT_FOUND:
+        case ERROR_INVALID_DRIVE:
+        case ERROR_NO_MORE_FILES:
+        case ERROR_BAD_NETPATH:
+        case ERROR_BAD_NET_NAME:
+        case ERROR_BAD_PATHNAME:
+        case ERROR_FILENAME_EXCED_RANGE:
+          errno = ENOENT;
+          break;
+
+        default:
+          errno = EACCES;
+          break;
+      }
+
+    return -1;
+  }
+
+  /* we could even stat and see if it exists */
+  static int
+  symlink (const char *old, const char *neu)
+  {
+    #if WINVER >= 0x0600
+      if (CreateSymbolicLink (neu, old, 1))
+        return 0;
+
+      if (CreateSymbolicLink (neu, old, 0))
+        return 0;
+    #endif
+
+    return EIO_ERRNO (ENOENT, -1);
+  }
+
+  /* POSIX API only */
+  #define CreateHardLink(neu,old,flags) 0
+  #define CreateSymbolicLink(neu,old,flags) 0
+
+  struct statvfs
+  {
+    int dummy;
+  };
+
+  #define DT_DIR EIO_DT_DIR
+  #define DT_REG EIO_DT_REG
+  #define D_NAME(entp) entp.cFileName
+  #define D_TYPE(entp) (entp.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY ? DT_DIR : DT_REG)
+
+#else
+
+  #include <sys/time.h>
+  #include <sys/select.h>
+  #include <sys/statvfs.h>
+  #include <unistd.h>
+  #include <signal.h>
+  #include <dirent.h>
+
+  #if _POSIX_MEMLOCK || _POSIX_MEMLOCK_RANGE || _POSIX_MAPPED_FILES
+    #include <sys/mman.h>
+  #endif
+
+  #define D_NAME(entp) entp->d_name
+
+  /* POSIX_SOURCE is useless on bsd's, and XOPEN_SOURCE is unreliable there, too */
+  #if __FreeBSD__ || __NetBSD__ || __OpenBSD__
+    #define _DIRENT_HAVE_D_TYPE /* sigh */
+    #define D_INO(de) (de)->d_fileno
+    #define D_NAMLEN(de) (de)->d_namlen
+  #elif __linux || defined d_ino || _XOPEN_SOURCE >= 600
+    #define D_INO(de) (de)->d_ino
+  #endif
+
+  #ifdef _D_EXACT_NAMLEN
+    #undef D_NAMLEN
+    #define D_NAMLEN(de) _D_EXACT_NAMLEN (de)
+  #endif
+
+  #ifdef _DIRENT_HAVE_D_TYPE
+    #define D_TYPE(de) (de)->d_type
+  #endif
+
+  #ifndef EIO_STRUCT_DIRENT
+    #define EIO_STRUCT_DIRENT struct dirent
+  #endif
+
+#endif
+
+#if HAVE_UTIMES
+# include <utime.h>
+#endif
+
+#if HAVE_SYS_SYSCALL_H
+# include <sys/syscall.h>
+#endif
+
+#if HAVE_SYS_PRCTL_H
+# include <sys/prctl.h>
+#endif
+
+#if HAVE_SENDFILE
+# if __linux
+#  include <sys/sendfile.h>
+# elif __FreeBSD__ || defined __APPLE__
+#  include <sys/socket.h>
+#  include <sys/uio.h>
+# elif __hpux
+#  include <sys/socket.h>
+# elif __solaris
+#  include <sys/sendfile.h>
+# else
+#  error sendfile support requested but not available
+# endif
+#endif
+
+#ifndef D_TYPE
+# define D_TYPE(de) 0
+#endif
+#ifndef D_INO
+# define D_INO(de) 0
+#endif
+#ifndef D_NAMLEN
+# define D_NAMLEN(entp) strlen (D_NAME (entp))
+#endif
+
+/* used for struct dirent, AIX doesn't provide it */
+#ifndef NAME_MAX
+# define NAME_MAX 4096
+#endif
+
+/* used for readlink etc. */
+#ifndef PATH_MAX
+# define PATH_MAX 4096
+#endif
+
+/* buffer size for various temporary buffers */
+#define EIO_BUFSIZE 65536
+
+#define dBUF	 				\
+  char *eio_buf = malloc (EIO_BUFSIZE);		\
+  errno = ENOMEM;				\
+  if (!eio_buf)					\
+    return -1
+
+#define FUBd					\
+  free (eio_buf)
+
+#define EIO_TICKS ((1000000 + 1023) >> 10)
+
+/*****************************************************************************/
+
+struct tmpbuf
+{
+  void *ptr;
+  int len;
+};
+
+static void *
+tmpbuf_get (struct tmpbuf *buf, int len)
+{
+  if (buf->len < len)
+    {
+      free (buf->ptr);
+      buf->ptr = malloc (buf->len = len);
+    }
+
+  return buf->ptr;
+}
+
+struct tmpbuf;
+
+#if _POSIX_VERSION >= 200809L
+  #define HAVE_AT 1
+  #define WD2FD(wd) ((wd) ? (wd)->fd : AT_FDCWD)
+  #ifndef O_SEARCH
+    #define O_SEARCH O_RDONLY
+  #endif
+#else
+  #define HAVE_AT 0
+  static const char *wd_expand (struct tmpbuf *tmpbuf, eio_wd wd, const char *path);
+#endif
+
+struct eio_pwd
+{
+#if HAVE_AT
+  int fd;
+#endif
+  int len;
+  char str[1]; /* actually, a 0-terminated canonical path */
+};
+
+/*****************************************************************************/
+
+#define ETP_PRI_MIN EIO_PRI_MIN
+#define ETP_PRI_MAX EIO_PRI_MAX
+
+struct etp_worker;
+
+#define ETP_REQ eio_req
+#define ETP_DESTROY(req) eio_destroy (req)
+static int eio_finish (eio_req *req);
+#define ETP_FINISH(req)  eio_finish (req)
+static void eio_execute (struct etp_worker *self, eio_req *req);
+#define ETP_EXECUTE(wrk,req) eio_execute (wrk,req)
+
+/*****************************************************************************/
+
+#define ETP_NUM_PRI (ETP_PRI_MAX - ETP_PRI_MIN + 1)
+
+/* calculate time difference in ~1/EIO_TICKS of a second */
+ecb_inline int
+tvdiff (struct timeval *tv1, struct timeval *tv2)
+{
+  return  (tv2->tv_sec  - tv1->tv_sec ) * EIO_TICKS
+       + ((tv2->tv_usec - tv1->tv_usec) >> 10);
+}
+
+static unsigned int started, idle, wanted = 4;
+
+static void (*want_poll_cb) (void);
+static void (*done_poll_cb) (void);
+ 
+static unsigned int max_poll_time;     /* reslock */
+static unsigned int max_poll_reqs;     /* reslock */
+
+static unsigned int nreqs;    /* reqlock */
+static unsigned int nready;   /* reqlock */
+static unsigned int npending; /* reqlock */
+static unsigned int max_idle = 4;      /* maximum number of threads that can idle indefinitely */
+static unsigned int idle_timeout = 10; /* number of seconds after which an idle threads exit */
+
+static xmutex_t wrklock;
+static xmutex_t reslock;
+static xmutex_t reqlock;
+static xcond_t  reqwait;
+
+#if !HAVE_PREADWRITE
+/*
+ * make our pread/pwrite emulation safe against themselves, but not against
+ * normal read/write by using a mutex. slows down execution a lot,
+ * but that's your problem, not mine.
+ */
+static xmutex_t preadwritelock;
+#endif
+
+typedef struct etp_worker
+{
+  struct tmpbuf tmpbuf;
+
+  /* locked by wrklock */
+  struct etp_worker *prev, *next;
+
+  xthread_t tid;
+
+#ifdef ETP_WORKER_COMMON
+  ETP_WORKER_COMMON
+#endif
+} etp_worker;
+
+static etp_worker wrk_first; /* NOT etp */
+
+#define ETP_WORKER_LOCK(wrk)   X_LOCK   (wrklock)
+#define ETP_WORKER_UNLOCK(wrk) X_UNLOCK (wrklock)
+
+/* worker threads management */
+
+static void
+etp_worker_clear (etp_worker *wrk)
+{
+}
+
+static void ecb_cold
+etp_worker_free (etp_worker *wrk)
+{
+  free (wrk->tmpbuf.ptr);
+
+  wrk->next->prev = wrk->prev;
+  wrk->prev->next = wrk->next;
+
+  free (wrk);
+}
+
+static unsigned int
+etp_nreqs (void)
+{
+  int retval;
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  retval = nreqs;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+  return retval;
+}
+
+static unsigned int
+etp_nready (void)
+{
+  unsigned int retval;
+
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  retval = nready;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+
+  return retval;
+}
+
+static unsigned int
+etp_npending (void)
+{
+  unsigned int retval;
+
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  retval = npending;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+
+  return retval;
+}
+
+static unsigned int
+etp_nthreads (void)
+{
+  unsigned int retval;
+
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  retval = started;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+
+  return retval;
+}
+
+/*
+ * a somewhat faster data structure might be nice, but
+ * with 8 priorities this actually needs <20 insns
+ * per shift, the most expensive operation.
+ */
+typedef struct {
+  ETP_REQ *qs[ETP_NUM_PRI], *qe[ETP_NUM_PRI]; /* qstart, qend */
+  int size;
+} etp_reqq;
+
+static etp_reqq req_queue;
+static etp_reqq res_queue;
+
+static void ecb_noinline ecb_cold
+reqq_init (etp_reqq *q)
+{
+  int pri;
+
+  for (pri = 0; pri < ETP_NUM_PRI; ++pri)
+    q->qs[pri] = q->qe[pri] = 0;
+
+  q->size = 0;
+}
+
+static int ecb_noinline
+reqq_push (etp_reqq *q, ETP_REQ *req)
+{
+  int pri = req->pri;
+  req->next = 0;
+
+  if (q->qe[pri])
+    {
+      q->qe[pri]->next = req;
+      q->qe[pri] = req;
+    }
+  else
+    q->qe[pri] = q->qs[pri] = req;
+
+  return q->size++;
+}
+
+static ETP_REQ * ecb_noinline
+reqq_shift (etp_reqq *q)
+{
+  int pri;
+
+  if (!q->size)
+    return 0;
+
+  --q->size;
+
+  for (pri = ETP_NUM_PRI; pri--; )
+    {
+      eio_req *req = q->qs[pri];
+
+      if (req)
+        {
+          if (!(q->qs[pri] = (eio_req *)req->next))
+            q->qe[pri] = 0;
+
+          return req;
+        }
+    }
+
+  abort ();
+}
+
+static int ecb_cold
+etp_init (void (*want_poll)(void), void (*done_poll)(void))
+{
+  X_MUTEX_CREATE (wrklock);
+  X_MUTEX_CREATE (reslock);
+  X_MUTEX_CREATE (reqlock);
+  X_COND_CREATE  (reqwait);
+
+  reqq_init (&req_queue);
+  reqq_init (&res_queue);
+
+  wrk_first.next =
+  wrk_first.prev = &wrk_first;
+
+  started  = 0;
+  idle     = 0;
+  nreqs    = 0;
+  nready   = 0;
+  npending = 0;
+
+  want_poll_cb = want_poll;
+  done_poll_cb = done_poll;
+
+  return 0;
+}
+
+X_THREAD_PROC (etp_proc);
+
+static void ecb_cold
+etp_start_thread (void)
+{
+  etp_worker *wrk = calloc (1, sizeof (etp_worker));
+
+  /*TODO*/
+  assert (("unable to allocate worker thread data", wrk));
+
+  X_LOCK (wrklock);
+
+  if (thread_create (&wrk->tid, etp_proc, (void *)wrk))
+    {
+      wrk->prev = &wrk_first;
+      wrk->next = wrk_first.next;
+      wrk_first.next->prev = wrk;
+      wrk_first.next = wrk;
+      ++started;
+    }
+  else
+    free (wrk);
+
+  X_UNLOCK (wrklock);
+}
+
+static void
+etp_maybe_start_thread (void)
+{
+  if (ecb_expect_true (etp_nthreads () >= wanted))
+    return;
+  
+  /* todo: maybe use idle here, but might be less exact */
+  if (ecb_expect_true (0 <= (int)etp_nthreads () + (int)etp_npending () - (int)etp_nreqs ()))
+    return;
+
+  etp_start_thread ();
+}
+
+static void ecb_cold
+etp_end_thread (void)
+{
+  eio_req *req = calloc (1, sizeof (eio_req)); /* will be freed by worker */
+
+  req->type = -1;
+  req->pri  = ETP_PRI_MAX - ETP_PRI_MIN;
+
+  X_LOCK (reqlock);
+  reqq_push (&req_queue, req);
+  X_COND_SIGNAL (reqwait);
+  X_UNLOCK (reqlock);
+
+  X_LOCK (wrklock);
+  --started;
+  X_UNLOCK (wrklock);
+}
+
+static int
+etp_poll (void)
+{
+  unsigned int maxreqs;
+  unsigned int maxtime;
+  struct timeval tv_start, tv_now;
+
+  X_LOCK (reslock);
+  maxreqs = max_poll_reqs;
+  maxtime = max_poll_time;
+  X_UNLOCK (reslock);
+
+  if (maxtime)
+    gettimeofday (&tv_start, 0);
+
+  for (;;)
+    {
+      ETP_REQ *req;
+
+      etp_maybe_start_thread ();
+
+      X_LOCK (reslock);
+      req = reqq_shift (&res_queue);
+
+      if (req)
+        {
+          --npending;
+
+          if (!res_queue.size && done_poll_cb)
+            done_poll_cb ();
+        }
+
+      X_UNLOCK (reslock);
+
+      if (!req)
+        return 0;
+
+      X_LOCK (reqlock);
+      --nreqs;
+      X_UNLOCK (reqlock);
+
+      if (ecb_expect_false (req->type == EIO_GROUP && req->size))
+        {
+          req->int1 = 1; /* mark request as delayed */
+          continue;
+        }
+      else
+        {
+          int res = ETP_FINISH (req);
+          if (ecb_expect_false (res))
+            return res;
+        }
+
+      if (ecb_expect_false (maxreqs && !--maxreqs))
+        break;
+
+      if (maxtime)
+        {
+          gettimeofday (&tv_now, 0);
+
+          if (tvdiff (&tv_start, &tv_now) >= maxtime)
+            break;
+        }
+    }
+
+  errno = EAGAIN;
+  return -1;
+}
+
+static void
+etp_cancel (ETP_REQ *req)
+{
+  req->cancelled = 1;
+
+  eio_grp_cancel (req);
+}
+
+static void
+etp_submit (ETP_REQ *req)
+{
+  req->pri -= ETP_PRI_MIN;
+
+  if (ecb_expect_false (req->pri < ETP_PRI_MIN - ETP_PRI_MIN)) req->pri = ETP_PRI_MIN - ETP_PRI_MIN;
+  if (ecb_expect_false (req->pri > ETP_PRI_MAX - ETP_PRI_MIN)) req->pri = ETP_PRI_MAX - ETP_PRI_MIN;
+
+  if (ecb_expect_false (req->type == EIO_GROUP))
+    {
+      /* I hope this is worth it :/ */
+      X_LOCK (reqlock);
+      ++nreqs;
+      X_UNLOCK (reqlock);
+
+      X_LOCK (reslock);
+
+      ++npending;
+
+      if (!reqq_push (&res_queue, req) && want_poll_cb)
+        want_poll_cb ();
+
+      X_UNLOCK (reslock);
+    }
+  else
+    {
+      X_LOCK (reqlock);
+      ++nreqs;
+      ++nready;
+      reqq_push (&req_queue, req);
+      X_COND_SIGNAL (reqwait);
+      X_UNLOCK (reqlock);
+
+      etp_maybe_start_thread ();
+    }
+}
+
+static void ecb_cold
+etp_set_max_poll_time (double nseconds)
+{
+  if (WORDACCESS_UNSAFE) X_LOCK   (reslock);
+  max_poll_time = nseconds * EIO_TICKS;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
+}
+
+static void ecb_cold
+etp_set_max_poll_reqs (unsigned int maxreqs)
+{
+  if (WORDACCESS_UNSAFE) X_LOCK   (reslock);
+  max_poll_reqs = maxreqs;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reslock);
+}
+
+static void ecb_cold
+etp_set_max_idle (unsigned int nthreads)
+{
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  max_idle = nthreads;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+}
+
+static void ecb_cold
+etp_set_idle_timeout (unsigned int seconds)
+{
+  if (WORDACCESS_UNSAFE) X_LOCK   (reqlock);
+  idle_timeout = seconds;
+  if (WORDACCESS_UNSAFE) X_UNLOCK (reqlock);
+}
+
+static void ecb_cold
+etp_set_min_parallel (unsigned int nthreads)
+{
+  if (wanted < nthreads)
+    wanted = nthreads;
+}
+
+static void ecb_cold
+etp_set_max_parallel (unsigned int nthreads)
+{
+  if (wanted > nthreads)
+    wanted = nthreads;
+
+  while (started > wanted)
+    etp_end_thread ();
+}
+
+/*****************************************************************************/
+
+static void
+grp_try_feed (eio_req *grp)
+{
+  while (grp->size < grp->int2 && !EIO_CANCELLED (grp))
+    {
+      grp->flags &= ~EIO_FLAG_GROUPADD;
+
+      EIO_FEED (grp);
+
+      /* stop if no progress has been made */
+      if (!(grp->flags & EIO_FLAG_GROUPADD))
+        {
+          grp->feed = 0;
+          break;
+        }
+    }
+}
+
+static int
+grp_dec (eio_req *grp)
+{
+  --grp->size;
+
+  /* call feeder, if applicable */
+  grp_try_feed (grp);
+
+  /* finish, if done */
+  if (!grp->size && grp->int1)
+    return eio_finish (grp);
+  else
+    return 0;
+}
+
+static void
+eio_destroy (eio_req *req)
+{
+  if ((req)->flags & EIO_FLAG_PTR1_FREE) free (req->ptr1);
+  if ((req)->flags & EIO_FLAG_PTR2_FREE) free (req->ptr2);
+
+  EIO_DESTROY (req);
+}
+
+static int
+eio_finish (eio_req *req)
+{
+  int res = EIO_FINISH (req);
+
+  if (req->grp)
+    {
+      int res2;
+      eio_req *grp = req->grp;
+
+      /* unlink request */
+      if (req->grp_next) req->grp_next->grp_prev = req->grp_prev;
+      if (req->grp_prev) req->grp_prev->grp_next = req->grp_next;
+
+      if (grp->grp_first == req)
+        grp->grp_first = req->grp_next;
+
+      res2 = grp_dec (grp);
+
+      if (!res)
+        res = res2;
+    }
+
+  eio_destroy (req);
+
+  return res;
+}
+
+void
+eio_grp_cancel (eio_req *grp)
+{
+  for (grp = grp->grp_first; grp; grp = grp->grp_next)
+    eio_cancel (grp);
+}
+
+void
+eio_cancel (eio_req *req)
+{
+  etp_cancel (req);
+}
+
+void
+eio_submit (eio_req *req)
+{
+  etp_submit (req);
+}
+
+unsigned int
+eio_nreqs (void)
+{
+  return etp_nreqs ();
+}
+
+unsigned int
+eio_nready (void)
+{
+  return etp_nready ();
+}
+
+unsigned int
+eio_npending (void)
+{
+  return etp_npending ();
+}
+
+unsigned int ecb_cold
+eio_nthreads (void)
+{
+  return etp_nthreads ();
+}
+
+void ecb_cold
+eio_set_max_poll_time (double nseconds)
+{
+  etp_set_max_poll_time (nseconds);
+}
+
+void ecb_cold
+eio_set_max_poll_reqs (unsigned int maxreqs)
+{
+  etp_set_max_poll_reqs (maxreqs);
+}
+
+void ecb_cold
+eio_set_max_idle (unsigned int nthreads)
+{
+  etp_set_max_idle (nthreads);
+}
+
+void ecb_cold
+eio_set_idle_timeout (unsigned int seconds)
+{
+  etp_set_idle_timeout (seconds);
+}
+
+void ecb_cold
+eio_set_min_parallel (unsigned int nthreads)
+{
+  etp_set_min_parallel (nthreads);
+}
+
+void ecb_cold
+eio_set_max_parallel (unsigned int nthreads)
+{
+  etp_set_max_parallel (nthreads);
+}
+
+int eio_poll (void)
+{
+  return etp_poll ();
+}
+
+/*****************************************************************************/
+/* work around various missing functions */
+
+#if !HAVE_PREADWRITE
+# undef pread
+# undef pwrite
+# define pread  eio__pread
+# define pwrite eio__pwrite
+
+static eio_ssize_t
+eio__pread (int fd, void *buf, size_t count, off_t offset)
+{
+  eio_ssize_t res;
+  off_t ooffset;
+
+  X_LOCK (preadwritelock);
+  ooffset = lseek (fd, 0, SEEK_CUR);
+  lseek (fd, offset, SEEK_SET);
+  res = read (fd, buf, count);
+  lseek (fd, ooffset, SEEK_SET);
+  X_UNLOCK (preadwritelock);
+
+  return res;
+}
+
+static eio_ssize_t
+eio__pwrite (int fd, void *buf, size_t count, off_t offset)
+{
+  eio_ssize_t res;
+  off_t ooffset;
+
+  X_LOCK (preadwritelock);
+  ooffset = lseek (fd, 0, SEEK_CUR);
+  lseek (fd, offset, SEEK_SET);
+  res = write (fd, buf, count);
+  lseek (fd, ooffset, SEEK_SET);
+  X_UNLOCK (preadwritelock);
+
+  return res;
+}
+#endif
+
+#ifndef HAVE_UTIMES
+
+# undef utimes
+# define utimes(path,times)  eio__utimes (path, times)
+
+static int
+eio__utimes (const char *filename, const struct timeval times[2])
+{
+  if (times)
+    {
+      struct utimbuf buf;
+
+      buf.actime  = times[0].tv_sec;
+      buf.modtime = times[1].tv_sec;
+
+      return utime (filename, &buf);
+    }
+  else
+    return utime (filename, 0);
+}
+
+#endif
+
+#ifndef HAVE_FUTIMES
+
+# undef futimes
+# define futimes(fd,times) eio__futimes (fd, times)
+
+static int
+eio__futimes (int fd, const struct timeval tv[2])
+{
+  errno = ENOSYS;
+  return -1;
+}
+
+#endif
+
+#if !HAVE_FDATASYNC
+# undef fdatasync
+# define fdatasync(fd) fsync (fd)
+#endif
+
+static int
+eio__syncfs (int fd)
+{
+  int res;
+
+#if HAVE_SYS_SYNCFS
+  res = (int)syscall (__NR_syncfs, (int)(fd));
+#else
+  res = EIO_ENOSYS ();
+#endif
+
+  if (res < 0 && errno == ENOSYS && fd >= 0)
+    sync ();
+
+  return res;
+}
+
+/* sync_file_range always needs emulation */
+static int
+eio__sync_file_range (int fd, off_t offset, size_t nbytes, unsigned int flags)
+{
+#if HAVE_SYNC_FILE_RANGE
+  int res;
+
+  if (EIO_SYNC_FILE_RANGE_WAIT_BEFORE   != SYNC_FILE_RANGE_WAIT_BEFORE
+      || EIO_SYNC_FILE_RANGE_WRITE      != SYNC_FILE_RANGE_WRITE
+      || EIO_SYNC_FILE_RANGE_WAIT_AFTER != SYNC_FILE_RANGE_WAIT_AFTER)
+    {
+      flags = 0
+         | (flags & EIO_SYNC_FILE_RANGE_WAIT_BEFORE ? SYNC_FILE_RANGE_WAIT_BEFORE : 0)
+         | (flags & EIO_SYNC_FILE_RANGE_WRITE       ? SYNC_FILE_RANGE_WRITE       : 0)
+         | (flags & EIO_SYNC_FILE_RANGE_WAIT_AFTER  ? SYNC_FILE_RANGE_WAIT_AFTER  : 0);
+    }
+
+  res = sync_file_range (fd, offset, nbytes, flags);
+
+  if (!res || errno != ENOSYS)
+    return res;
+#endif
+
+  /* even though we could play tricks with the flags, it's better to always
+   * call fdatasync, as that matches the expectation of its users best */
+  return fdatasync (fd);
+}
+
+static int
+eio__fallocate (int fd, int mode, off_t offset, size_t len)
+{
+#if HAVE_LINUX_FALLOCATE
+  return fallocate (fd, mode, offset, len);
+#else
+  return EIO_ENOSYS ();
+#endif
+}
+
+#if !HAVE_READAHEAD
+# undef readahead
+# define readahead(fd,offset,count) eio__readahead (fd, offset, count, self)
+
+static eio_ssize_t
+eio__readahead (int fd, off_t offset, size_t count, etp_worker *self)
+{
+  size_t todo = count;
+  dBUF;
+
+  while (todo > 0)
+    {
+      size_t len = todo < EIO_BUFSIZE ? todo : EIO_BUFSIZE;
+
+      pread (fd, eio_buf, len, offset);
+      offset += len;
+      todo   -= len;
+    }
+
+  FUBd;
+
+  /* linux's readahead basically only fails for EBADF or EINVAL (not mmappable) */
+  /* but not for e.g. EIO or eof, so we also never fail */
+  return 0;
+}
+
+#endif
+
+/* sendfile always needs emulation */
+static eio_ssize_t
+eio__sendfile (int ofd, int ifd, off_t offset, size_t count)
+{
+  eio_ssize_t written = 0;
+  eio_ssize_t res;
+
+  if (!count)
+    return 0;
+
+  for (;;)
+    {
+#ifdef __APPLE__
+# undef HAVE_SENDFILE /* broken, as everything on os x */
+#endif
+#if HAVE_SENDFILE
+# if __linux
+      off_t soffset = offset;
+      res = sendfile (ofd, ifd, &soffset, count);
+
+# elif __FreeBSD__
+      /*
+       * Of course, the freebsd sendfile is a dire hack with no thoughts
+       * wasted on making it similar to other I/O functions.
+       */
+      off_t sbytes;
+      res = sendfile (ifd, ofd, offset, count, 0, &sbytes, 0);
+
+      #if 0 /* according to the manpage, this is correct, but broken behaviour */
+      /* freebsd' sendfile will return 0 on success */
+      /* freebsd 8 documents it as only setting *sbytes on EINTR and EAGAIN, but */
+      /* not on e.g. EIO or EPIPE - sounds broken */
+      if ((res < 0 && (errno == EAGAIN || errno == EINTR) && sbytes) || res == 0)
+        res = sbytes;
+      #endif
+
+      /* according to source inspection, this is correct, and useful behaviour */
+      if (sbytes)
+        res = sbytes;
+
+# elif defined __APPLE__
+      off_t sbytes = count;
+      res = sendfile (ifd, ofd, offset, &sbytes, 0, 0);
+
+      /* according to the manpage, sbytes is always valid */
+      if (sbytes)
+        res = sbytes;
+
+# elif __hpux
+      res = sendfile (ofd, ifd, offset, count, 0, 0);
+
+# elif __solaris
+      struct sendfilevec vec;
+      size_t sbytes;
+
+      vec.sfv_fd   = ifd;
+      vec.sfv_flag = 0;
+      vec.sfv_off  = offset;
+      vec.sfv_len  = count;
+
+      res = sendfilev (ofd, &vec, 1, &sbytes);
+
+      if (res < 0 && sbytes)
+        res = sbytes;
+
+# endif
+
+#elif defined (_WIN32) && 0
+      /* does not work, just for documentation of what would need to be done */
+      /* actually, cannot be done like this, as TransmitFile changes the file offset, */
+      /* libeio guarantees that the file offset does not change, and windows */
+      /* has no way to get an independent handle to the same file description */
+      HANDLE h = TO_SOCKET (ifd);
+      SetFilePointer (h, offset, 0, FILE_BEGIN);
+      res = TransmitFile (TO_SOCKET (ofd), h, count, 0, 0, 0, 0);
+
+#else
+      res = EIO_ENOSYS ();
+#endif
+
+      /* we assume sendfile can copy at least 128mb in one go */
+      if (res <= 128 * 1024 * 1024)
+        {
+          if (res > 0)
+            written += res;
+
+          if (written)
+            return written;
+
+          break;
+        }
+      else
+        {
+          /* if we requested more, then probably the kernel was lazy */
+          written += res;
+          offset  += res;
+          count   -= res;
+
+          if (!count)
+            return written;
+        }
+    }
+
+  if (res < 0
+      && (errno == ENOSYS || errno == EINVAL || errno == ENOTSOCK
+          /* BSDs */
+#ifdef ENOTSUP /* sigh, if the steenking pile called openbsd would only try to at least compile posix code... */
+          || errno == ENOTSUP
+#endif
+#ifdef EOPNOTSUPP /* windows */
+          || errno == EOPNOTSUPP /* BSDs */
+#endif
+#if __solaris
+          || errno == EAFNOSUPPORT || errno == EPROTOTYPE
+#endif
+         )
+      )
+    {
+      /* emulate sendfile. this is a major pain in the ass */
+      dBUF;
+
+      res = 0;
+
+      while (count)
+        {
+          eio_ssize_t cnt;
+          
+          cnt = pread (ifd, eio_buf, count > EIO_BUFSIZE ? EIO_BUFSIZE : count, offset);
+
+          if (cnt <= 0)
+            {
+              if (cnt && !res) res = -1;
+              break;
+            }
+
+          cnt = write (ofd, eio_buf, cnt);
+
+          if (cnt <= 0)
+            {
+              if (cnt && !res) res = -1;
+              break;
+            }
+
+          offset += cnt;
+          res    += cnt;
+          count  -= cnt;
+        }
+
+      FUBd;
+    }
+
+  return res;
+}
+
+#ifdef PAGESIZE
+# define eio_pagesize() PAGESIZE
+#else
+static intptr_t
+eio_pagesize (void)
+{
+  static intptr_t page;
+
+  if (!page)
+    page = sysconf (_SC_PAGESIZE);
+
+  return page;
+}
+#endif
+
+static void
+eio_page_align (void **addr, size_t *length)
+{
+  intptr_t mask = eio_pagesize () - 1;
+
+  /* round down addr */
+  intptr_t adj = mask & (intptr_t)*addr;
+
+  *addr   = (void *)((intptr_t)*addr - adj);
+  *length += adj;
+
+  /* round up length */
+  *length = (*length + mask) & ~mask;
+}
+
+#if !_POSIX_MEMLOCK
+# define eio__mlockall(a) EIO_ENOSYS ()
+#else
+
+static int
+eio__mlockall (int flags)
+{
+  #if __GLIBC__ == 2 && __GLIBC_MINOR__ <= 7
+    extern int mallopt (int, int);
+    mallopt (-6, 238); /* http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=473812 */
+  #endif
+
+  if (EIO_MCL_CURRENT   != MCL_CURRENT
+      || EIO_MCL_FUTURE != MCL_FUTURE)
+    {
+      flags = 0
+         | (flags & EIO_MCL_CURRENT ? MCL_CURRENT : 0)
+         | (flags & EIO_MCL_FUTURE  ? MCL_FUTURE : 0);
+    }
+
+  return mlockall (flags);
+}
+#endif
+
+#if !_POSIX_MEMLOCK_RANGE
+# define eio__mlock(a,b) EIO_ENOSYS ()
+#else
+
+static int
+eio__mlock (void *addr, size_t length)
+{
+  eio_page_align (&addr, &length);
+
+  return mlock (addr, length);
+}
+
+#endif
+
+#if !(_POSIX_MAPPED_FILES && _POSIX_SYNCHRONIZED_IO)
+# define eio__msync(a,b,c) EIO_ENOSYS ()
+#else
+
+static int
+eio__msync (void *mem, size_t len, int flags)
+{
+  eio_page_align (&mem, &len);
+
+  if (EIO_MS_ASYNC         != MS_SYNC
+      || EIO_MS_INVALIDATE != MS_INVALIDATE
+      || EIO_MS_SYNC       != MS_SYNC)
+    {
+      flags = 0
+         | (flags & EIO_MS_ASYNC      ? MS_ASYNC : 0)
+         | (flags & EIO_MS_INVALIDATE ? MS_INVALIDATE : 0)
+         | (flags & EIO_MS_SYNC       ? MS_SYNC : 0);
+    }
+
+  return msync (mem, len, flags);
+}
+
+#endif
+
+static int
+eio__mtouch (eio_req *req)
+{
+  void *mem  = req->ptr2;
+  size_t len = req->size;
+  int flags  = req->int1;
+
+  eio_page_align (&mem, &len);
+
+  {
+    intptr_t addr = (intptr_t)mem;
+    intptr_t end = addr + len;
+    intptr_t page = eio_pagesize ();
+
+    if (addr < end)
+      if (flags & EIO_MT_MODIFY) /* modify */
+        do { *((volatile sig_atomic_t *)addr) |= 0; } while ((addr += page) < len && !EIO_CANCELLED (req));
+      else
+        do { *((volatile sig_atomic_t *)addr)     ; } while ((addr += page) < len && !EIO_CANCELLED (req));
+  }
+
+  return 0;
+}
+
+/*****************************************************************************/
+/* requests implemented outside eio_execute, because they are so large */
+
+static void
+eio__lseek (eio_req *req)
+{
+  /* this usually gets optimised away completely, or your compiler sucks, */
+  /* or the whence constants really are not 0, 1, 2 */
+  int whence = req->int2 == EIO_SEEK_SET ? SEEK_SET
+             : req->int2 == EIO_SEEK_CUR ? SEEK_CUR
+             : req->int2 == EIO_SEEK_END ? SEEK_END
+             : req->int2;
+
+  req->offs   = lseek (req->int1, req->offs, whence);
+  req->result = req->offs == (off_t)-1 ? -1 : 0;
+}
+
+/* result will always end up in tmpbuf, there is always space for adding a 0-byte */
+static int
+eio__realpath (struct tmpbuf *tmpbuf, eio_wd wd, const char *path)
+{
+  const char *rel = path;
+  char *res;
+  char *tmp1, *tmp2;
+#if SYMLOOP_MAX > 32
+  int symlinks = SYMLOOP_MAX;
+#else
+  int symlinks = 32;
+#endif
+
+  errno = EINVAL;
+  if (!rel)
+    return -1;
+
+  errno = ENOENT;
+  if (!*rel)
+    return -1;
+
+  res  = tmpbuf_get (tmpbuf, PATH_MAX * 3);
+  tmp1 = res  + PATH_MAX;
+  tmp2 = tmp1 + PATH_MAX;
+
+#if 0 /* disabled, the musl way to do things is just too racy */
+#if __linux && defined(O_NONBLOCK) && defined(O_NOATIME)
+  /* on linux we may be able to ask the kernel */
+  {
+    int fd = open (rel, O_RDONLY | O_NONBLOCK | O_NOCTTY | O_NOATIME);
+
+    if (fd >= 0)
+      {
+        sprintf (tmp1, "/proc/self/fd/%d", fd);
+        req->result = readlink (tmp1, res, PATH_MAX);
+        close (fd);
+
+        /* here we should probably stat the open file and the disk file, to make sure they still match */
+
+        if (req->result > 0)
+          goto done;
+      }
+    else if (errno == ELOOP || errno == ENAMETOOLONG || errno == ENOENT || errno == ENOTDIR || errno == EIO)
+      return;
+  }
+#endif
+#endif
+
+  if (*rel != '/')
+    {
+      int len;
+
+      errno = ENOENT;
+      if (wd == EIO_INVALID_WD)
+        return -1;
+      
+      if (wd == EIO_CWD)
+        {
+          if (!getcwd (res, PATH_MAX))
+            return -1;
+
+          len = strlen (res);
+        }
+      else
+        memcpy (res, wd->str, len = wd->len);
+
+      if (res [1]) /* only use if not / */
+        res += len;
+    }
+
+  while (*rel)
+    {
+      eio_ssize_t len, linklen;
+      const char *beg = rel;
+
+      while (*rel && *rel != '/')
+        ++rel;
+
+      len = rel - beg;
+
+      if (!len) /* skip slashes */
+        {
+          ++rel;
+          continue;
+        }
+
+      if (beg [0] == '.')
+        {
+          if (len == 1)
+            continue; /* . - nop */
+
+          if (beg [1] == '.' && len == 2)
+            {
+              /* .. - back up one component, if possible */
+
+              while (res != tmpbuf->ptr)
+                if (*--res == '/')
+                  break;
+
+              continue;
+            }
+        }
+
+        errno = ENAMETOOLONG;
+        if (res + 1 + len + 1 >= tmp1)
+          return -1;
+
+        /* copy one component */
+        *res = '/';
+        memcpy (res + 1, beg, len);
+
+        /* zero-terminate, for readlink */
+        res [len + 1] = 0;
+
+        /* now check if it's a symlink */
+        linklen = readlink (tmpbuf->ptr, tmp1, PATH_MAX);
+
+        if (linklen < 0)
+          {
+            if (errno != EINVAL)
+              return -1;
+
+            /* it's a normal directory. hopefully */
+            res += len + 1;
+          }
+        else
+          {
+            /* yay, it was a symlink - build new path in tmp2 */
+            int rellen = strlen (rel);
+
+            errno = ENAMETOOLONG;
+            if (linklen + 1 + rellen >= PATH_MAX)
+              return -1;
+
+            errno = ELOOP;
+            if (!--symlinks)
+              return -1;
+
+            if (*tmp1 == '/')
+              res = tmpbuf->ptr; /* symlink resolves to an absolute path */
+
+            /* we need to be careful, as rel might point into tmp2 already */
+            memmove (tmp2 + linklen + 1, rel, rellen + 1);
+            tmp2 [linklen] = '/';
+            memcpy (tmp2, tmp1, linklen);
+
+            rel = tmp2;
+          }
+    }
+
+  /* special case for the lone root path */
+  if (res == tmpbuf->ptr)
+    *res++ = '/';
+
+  return res - (char *)tmpbuf->ptr;
+}
+
+static signed char
+eio_dent_cmp (const eio_dirent *a, const eio_dirent *b)
+{
+  return a->score - b->score ? a->score - b->score /* works because our signed char is always 0..100 */
+       : a->inode < b->inode ? -1
+       : a->inode > b->inode ?  1
+       :                        0;
+}
+
+#define EIO_DENT_CMP(i,op,j) eio_dent_cmp (&i, &j) op 0
+
+#define EIO_SORT_CUTOFF 30 /* quite high, but performs well on many filesystems */
+#define EIO_SORT_FAST   60 /* when to only use insertion sort */
+
+static void
+eio_dent_radix_sort (eio_dirent *dents, int size, signed char score_bits, eio_ino_t inode_bits)
+{
+  unsigned char bits [9 + sizeof (eio_ino_t) * 8];
+  unsigned char *bit = bits;
+
+  assert (CHAR_BIT == 8);
+  assert (sizeof (eio_dirent) * 8 < 256);
+  assert (offsetof (eio_dirent, inode)); /* we use bit #0 as sentinel */
+  assert (offsetof (eio_dirent, score)); /* we use bit #0 as sentinel */
+
+  if (size <= EIO_SORT_FAST)
+    return;
+
+  /* first prepare an array of bits to test in our radix sort */
+  /* try to take endianness into account, as well as differences in eio_ino_t sizes */
+  /* inode_bits must contain all inodes ORed together */
+  /* which is used to skip bits that are 0 everywhere, which is very common */
+  {
+    eio_ino_t endianness;
+    int i, j;
+
+    /* we store the byte offset of byte n into byte n of "endianness" */
+    for (i = 0; i < sizeof (eio_ino_t); ++i)
+      ((unsigned char *)&endianness)[i] = i;
+
+    *bit++ = 0;
+
+    for (i = 0; i < sizeof (eio_ino_t); ++i)
+      {
+        /* shifting off the byte offsets out of "endianness" */
+        int offs = (offsetof (eio_dirent, inode) + (endianness & 0xff)) * 8;
+        endianness >>= 8;
+
+        for (j = 0; j < 8; ++j)
+          if (inode_bits & (((eio_ino_t)1) << (i * 8 + j)))
+            *bit++ = offs + j;
+      }
+
+    for (j = 0; j < 8; ++j)
+      if (score_bits & (1 << j))
+        *bit++ = offsetof (eio_dirent, score) * 8 + j;
+  }
+
+  /* now actually do the sorting (a variant of MSD radix sort) */
+  {
+    eio_dirent    *base_stk [9 + sizeof (eio_ino_t) * 8], *base;
+    eio_dirent    *end_stk  [9 + sizeof (eio_ino_t) * 8], *end;
+    unsigned char *bit_stk  [9 + sizeof (eio_ino_t) * 8];
+    int stk_idx = 0;
+
+    base_stk [stk_idx] = dents;
+    end_stk  [stk_idx] = dents + size;
+    bit_stk  [stk_idx] = bit - 1;
+
+    do
+      {
+        base = base_stk [stk_idx];
+        end  = end_stk  [stk_idx];
+        bit  = bit_stk  [stk_idx];
+
+        for (;;)
+          {
+            unsigned char O = *bit >> 3;
+            unsigned char M = 1 << (*bit & 7);
+
+            eio_dirent *a = base;
+            eio_dirent *b = end;
+
+            if (b - a < EIO_SORT_CUTOFF)
+              break;
+
+            /* now bit-partition the array on the bit */
+            /* this ugly asymmetric loop seems to perform much better than typical */
+            /* partition algos found in the literature */
+            do
+              if (!(((unsigned char *)a)[O] & M))
+                ++a;
+              else if (!(((unsigned char *)--b)[O] & M))
+                {
+                  eio_dirent tmp = *a; *a = *b; *b = tmp;
+                  ++a;
+                }
+            while (b > a);
+
+            /* next bit, or stop, if no bits left in this path */
+            if (!*--bit)
+              break;
+
+            base_stk [stk_idx] = a;
+            end_stk  [stk_idx] = end;
+            bit_stk  [stk_idx] = bit;
+            ++stk_idx;
+
+            end = a;
+          }
+      }
+    while (stk_idx--);
+  }
+}
+
+static void
+eio_dent_insertion_sort (eio_dirent *dents, int size)
+{
+  /* first move the smallest element to the front, to act as a sentinel */
+  {
+    int i;
+    eio_dirent *min = dents;
+    
+    /* the radix pre-pass ensures that the minimum element is in the first EIO_SORT_CUTOFF + 1 elements */
+    for (i = size > EIO_SORT_FAST ? EIO_SORT_CUTOFF + 1 : size; --i; )
+      if (EIO_DENT_CMP (dents [i], <, *min))
+        min = &dents [i];
+
+    /* swap elements 0 and j (minimum) */
+    {
+      eio_dirent tmp = *dents; *dents = *min; *min = tmp;
+    }
+  }
+
+  /* then do standard insertion sort, assuming that all elements are >= dents [0] */
+  {
+    eio_dirent *i, *j;
+
+    for (i = dents + 1; i < dents + size; ++i)
+      {
+        eio_dirent value = *i;
+
+        for (j = i - 1; EIO_DENT_CMP (*j, >, value); --j)
+          j [1] = j [0];
+
+        j [1] = value;
+      }
+  }
+}
+
+static void
+eio_dent_sort (eio_dirent *dents, int size, signed char score_bits, eio_ino_t inode_bits)
+{
+  if (size <= 1)
+    return; /* our insertion sort relies on size > 0 */
+
+  /* first we use a radix sort, but only for dirs >= EIO_SORT_FAST */
+  /* and stop sorting when the partitions are <= EIO_SORT_CUTOFF */
+  eio_dent_radix_sort (dents, size, score_bits, inode_bits);
+
+  /* use an insertion sort at the end, or for small arrays, */
+  /* as insertion sort is more efficient for small partitions */
+  eio_dent_insertion_sort (dents, size);
+}
+
+/* read a full directory */
+static void
+eio__scandir (eio_req *req, etp_worker *self)
+{
+  char *name, *names;
+  int namesalloc = 4096 - sizeof (void *) * 4;
+  int namesoffs = 0;
+  int flags = req->int1;
+  eio_dirent *dents = 0;
+  int dentalloc = 128;
+  int dentoffs = 0;
+  eio_ino_t inode_bits = 0;
+#ifdef _WIN32
+  HANDLE dirp;
+  WIN32_FIND_DATA entp;
+#else
+  DIR *dirp;
+  EIO_STRUCT_DIRENT *entp;
+#endif
+
+  req->result = -1;
+
+  if (!(flags & EIO_READDIR_DENTS))
+    flags &= ~(EIO_READDIR_DIRS_FIRST | EIO_READDIR_STAT_ORDER);
+
+#ifdef _WIN32
+  {
+    int len = strlen ((const char *)req->ptr1);
+    char *path = malloc (MAX_PATH);
+    const char *fmt;
+    const char *reqpath = wd_expand (&self->tmpbuf, req->wd, req->ptr1);
+
+    if (!len)
+      fmt = "./*";
+    else if (reqpath[len - 1] == '/' || reqpath[len - 1] == '\\')
+      fmt = "%s*";
+    else
+      fmt = "%s/*";
+
+    _snprintf (path, MAX_PATH, fmt, reqpath);
+    dirp = FindFirstFile (path, &entp);
+    free (path);
+
+    if (dirp == INVALID_HANDLE_VALUE)
+     {
+       /* should steal _dosmaperr */
+       switch (GetLastError ())
+         {
+           case ERROR_FILE_NOT_FOUND:
+             req->result = 0;
+             break;
+
+           case ERROR_INVALID_NAME:
+           case ERROR_PATH_NOT_FOUND:
+           case ERROR_NO_MORE_FILES:
+             errno = ENOENT;
+             break;
+
+           case ERROR_NOT_ENOUGH_MEMORY:
+             errno = ENOMEM;
+             break;
+
+           default:
+             errno = EINVAL;
+             break;
+         }
+
+       return;
+     }
+  }
+#else
+  #if HAVE_AT
+    if (req->wd)
+      {
+        int fd = openat (WD2FD (req->wd), req->ptr1, O_CLOEXEC | O_SEARCH | O_DIRECTORY);
+
+        if (fd < 0)
+          return;
+
+        dirp = fdopendir (fd);
+
+        if (!dirp)
+          close (fd);
+      }
+    else
+      dirp = opendir (req->ptr1);
+  #else
+    dirp = opendir (wd_expand (&self->tmpbuf, req->wd, req->ptr1));
+  #endif
+
+  if (!dirp)
+    return;
+#endif
+
+  if (req->flags & EIO_FLAG_PTR1_FREE)
+    free (req->ptr1);
+
+  req->flags |= EIO_FLAG_PTR1_FREE | EIO_FLAG_PTR2_FREE;
+  req->ptr1 = dents = flags ? malloc (dentalloc * sizeof (eio_dirent)) : 0;
+  req->ptr2 = names = malloc (namesalloc);
+
+  if (!names || (flags && !dents))
+    return;
+
+  for (;;)
+    {
+      int done;
+
+#ifdef _WIN32
+      done = !dirp;
+#else
+      errno = 0;
+      entp = readdir (dirp);
+      done = !entp;
+#endif
+
+      if (done)
+        {
+#ifndef _WIN32
+          int old_errno = errno;
+          closedir (dirp);
+          errno = old_errno;
+
+          if (errno)
+            break;
+#endif
+
+          /* sort etc. */
+          req->int1   = flags;
+          req->result = dentoffs;
+
+          if (flags & EIO_READDIR_STAT_ORDER)
+            eio_dent_sort (dents, dentoffs, flags & EIO_READDIR_DIRS_FIRST ? 7 : 0, inode_bits);
+          else if (flags & EIO_READDIR_DIRS_FIRST)
+            if (flags & EIO_READDIR_FOUND_UNKNOWN)
+              eio_dent_sort (dents, dentoffs, 7, inode_bits); /* sort by score and inode */
+            else
+              {
+                /* in this case, all is known, and we just put dirs first and sort them */
+                eio_dirent *oth = dents + dentoffs;
+                eio_dirent *dir = dents;
+
+                /* now partition dirs to the front, and non-dirs to the back */
+                /* by walking from both sides and swapping if necessary */
+                while (oth > dir)
+                  {
+                    if (dir->type == EIO_DT_DIR)
+                      ++dir;
+                    else if ((--oth)->type == EIO_DT_DIR)
+                      {
+                        eio_dirent tmp = *dir; *dir = *oth; *oth = tmp;
+
+                        ++dir;
+                      }
+                  }
+
+                /* now sort the dirs only (dirs all have the same score) */
+                eio_dent_sort (dents, dir - dents, 0, inode_bits);
+              }
+
+          break;
+        }
+
+      /* now add the entry to our list(s) */
+      name = D_NAME (entp);
+
+      /* skip . and .. entries */
+      if (name [0] != '.' || (name [1] && (name [1] != '.' || name [2])))
+        {
+          int len = D_NAMLEN (entp) + 1;
+
+          while (ecb_expect_false (namesoffs + len > namesalloc))
+            {
+              namesalloc *= 2;
+              req->ptr2 = names = realloc (names, namesalloc);
+
+              if (!names)
+                break;
+            }
+
+          memcpy (names + namesoffs, name, len);
+
+          if (dents)
+            {
+              struct eio_dirent *ent;
+
+              if (ecb_expect_false (dentoffs == dentalloc))
+                {
+                  dentalloc *= 2;
+                  req->ptr1 = dents = realloc (dents, dentalloc * sizeof (eio_dirent));
+
+                  if (!dents)
+                    break;
+                }
+
+              ent = dents + dentoffs;
+
+              ent->nameofs = namesoffs; /* rather dirtily we store the offset in the pointer */
+              ent->namelen = len - 1;
+              ent->inode   = D_INO (entp);
+
+              inode_bits |= ent->inode;
+
+              switch (D_TYPE (entp))
+                {
+                  default:
+                    ent->type = EIO_DT_UNKNOWN;
+                    flags |= EIO_READDIR_FOUND_UNKNOWN;
+                    break;
+
+                  #ifdef DT_FIFO
+                    case DT_FIFO: ent->type = EIO_DT_FIFO; break;
+                  #endif
+                  #ifdef DT_CHR
+                    case DT_CHR:  ent->type = EIO_DT_CHR;  break;
+                  #endif          
+                  #ifdef DT_MPC
+                    case DT_MPC:  ent->type = EIO_DT_MPC;  break;
+                  #endif          
+                  #ifdef DT_DIR
+                    case DT_DIR:  ent->type = EIO_DT_DIR;  break;
+                  #endif          
+                  #ifdef DT_NAM
+                    case DT_NAM:  ent->type = EIO_DT_NAM;  break;
+                  #endif          
+                  #ifdef DT_BLK
+                    case DT_BLK:  ent->type = EIO_DT_BLK;  break;
+                  #endif          
+                  #ifdef DT_MPB
+                    case DT_MPB:  ent->type = EIO_DT_MPB;  break;
+                  #endif          
+                  #ifdef DT_REG
+                    case DT_REG:  ent->type = EIO_DT_REG;  break;
+                  #endif          
+                  #ifdef DT_NWK
+                    case DT_NWK:  ent->type = EIO_DT_NWK;  break;
+                  #endif          
+                  #ifdef DT_CMP
+                    case DT_CMP:  ent->type = EIO_DT_CMP;  break;
+                  #endif          
+                  #ifdef DT_LNK
+                    case DT_LNK:  ent->type = EIO_DT_LNK;  break;
+                  #endif
+                  #ifdef DT_SOCK
+                    case DT_SOCK: ent->type = EIO_DT_SOCK; break;
+                  #endif
+                  #ifdef DT_DOOR
+                    case DT_DOOR: ent->type = EIO_DT_DOOR; break;
+                  #endif
+                  #ifdef DT_WHT
+                    case DT_WHT:  ent->type = EIO_DT_WHT;  break;
+                  #endif
+                }
+
+              ent->score = 7;
+
+              if (flags & EIO_READDIR_DIRS_FIRST)
+                {
+                  if (ent->type == EIO_DT_UNKNOWN)
+                    {
+                      if (*name == '.') /* leading dots are likely directories, and, in any case, rare */
+                        ent->score = 1;
+                      else if (!strchr (name, '.')) /* absence of dots indicate likely dirs */
+                        ent->score = len <= 2 ? 4 - len : len <= 4 ? 4 : len <= 7 ? 5 : 6; /* shorter == more likely dir, but avoid too many classes */
+                    }
+                  else if (ent->type == EIO_DT_DIR)
+                    ent->score = 0;
+                }
+            }
+
+          namesoffs += len;
+          ++dentoffs;
+        }
+
+      if (EIO_CANCELLED (req))
+        {
+          errno = ECANCELED;
+          break;
+        }
+
+#ifdef _WIN32
+      if (!FindNextFile (dirp, &entp))
+        {
+          FindClose (dirp);
+          dirp = 0;
+        }
+#endif
+    }
+}
+
+/*****************************************************************************/
+/* working directory stuff */
+/* various deficiencies in the posix 2008 api force us to */
+/* keep the absolute path in string form at all times */
+/* fuck yeah. */
+
+#if !HAVE_AT
+
+/* a bit like realpath, but usually faster because it doesn'T have to return */
+/* an absolute or canonical path */
+static const char *
+wd_expand (struct tmpbuf *tmpbuf, eio_wd wd, const char *path)
+{
+  if (!wd || *path == '/')
+    return path;
+
+  if (path [0] == '.' && !path [1])
+    return wd->str;
+
+  {
+    int l1 = wd->len;
+    int l2 = strlen (path);
+
+    char *res = tmpbuf_get (tmpbuf, l1 + l2 + 2);
+
+    memcpy (res, wd->str, l1);
+    res [l1] = '/';
+    memcpy (res + l1 + 1, path, l2 + 1);
+
+    return res;
+  }
+}
+
+#endif
+
+static eio_wd
+eio__wd_open_sync (struct tmpbuf *tmpbuf, eio_wd wd, const char *path)
+{
+  int fd;
+  eio_wd res;
+  int len = eio__realpath (tmpbuf, wd, path);
+
+  if (len < 0)
+    return EIO_INVALID_WD;
+
+#if HAVE_AT
+  fd = openat (WD2FD (wd), path, O_CLOEXEC | O_SEARCH | O_DIRECTORY);
+
+  if (fd < 0)
+    return EIO_INVALID_WD;
+#endif
+
+  res = malloc (sizeof (*res) + len); /* one extra 0-byte */
+
+#if HAVE_AT
+  res->fd = fd;
+#endif
+
+  res->len = len;
+  memcpy (res->str, tmpbuf->ptr, len);
+  res->str [len] = 0;
+
+  return res;
+}
+
+eio_wd
+eio_wd_open_sync (eio_wd wd, const char *path)
+{
+  struct tmpbuf tmpbuf = { 0 };
+  wd = eio__wd_open_sync (&tmpbuf, wd, path);
+  free (tmpbuf.ptr);
+
+  return wd;
+}
+
+void
+eio_wd_close_sync (eio_wd wd)
+{
+  if (wd != EIO_INVALID_WD && wd != EIO_CWD)
+    {
+      #if HAVE_AT
+      close (wd->fd);
+      #endif
+      free (wd);
+    }
+}
+
+#if HAVE_AT
+
+/* they forgot these */
+
+static int
+eio__truncateat (int dirfd, const char *path, off_t length)
+{
+  int fd = openat (dirfd, path, O_WRONLY | O_CLOEXEC);
+  int res;
+
+  if (fd < 0)
+    return fd;
+
+  res = ftruncate (fd, length);
+  close (fd);
+  return res;
+}
+
+static int
+eio__statvfsat (int dirfd, const char *path, struct statvfs *buf)
+{
+  int fd = openat (dirfd, path, O_SEARCH | O_CLOEXEC);
+  int res;
+
+  if (fd < 0)
+    return fd;
+
+  res = fstatvfs (fd, buf);
+  close (fd);
+  return res;
+
+}
+
+#endif
+
+/*****************************************************************************/
+
+#define ALLOC(len)				\
+  if (!req->ptr2)				\
+    {						\
+      X_LOCK (wrklock);				\
+      req->flags |= EIO_FLAG_PTR2_FREE;		\
+      X_UNLOCK (wrklock);			\
+      req->ptr2 = malloc (len);			\
+      if (!req->ptr2)				\
+        {					\
+          errno       = ENOMEM;			\
+          req->result = -1;			\
+          break;				\
+        }					\
+    }
+
+static void ecb_noinline ecb_cold
+etp_proc_init (void)
+{
+#if HAVE_PRCTL_SET_NAME
+  /* provide a more sensible "thread name" */
+  char name[16 + 1];
+  const int namelen = sizeof (name) - 1;
+  int len;
+
+  prctl (PR_GET_NAME, (unsigned long)name, 0, 0, 0);
+  name [namelen] = 0;
+  len = strlen (name);
+  strcpy (name + (len <= namelen - 4 ? len : namelen - 4), "/eio");
+  prctl (PR_SET_NAME, (unsigned long)name, 0, 0, 0);
+#endif
+}
+
+X_THREAD_PROC (etp_proc)
+{
+  ETP_REQ *req;
+  struct timespec ts;
+  etp_worker *self = (etp_worker *)thr_arg;
+
+  etp_proc_init ();
+
+  /* try to distribute timeouts somewhat evenly */
+  ts.tv_nsec = ((unsigned long)self & 1023UL) * (1000000000UL / 1024UL);
+
+  for (;;)
+    {
+      ts.tv_sec = 0;
+
+      X_LOCK (reqlock);
+
+      for (;;)
+        {
+          req = reqq_shift (&req_queue);
+
+          if (req)
+            break;
+
+          if (ts.tv_sec == 1) /* no request, but timeout detected, let's quit */
+            {
+              X_UNLOCK (reqlock);
+              X_LOCK (wrklock);
+              --started;
+              X_UNLOCK (wrklock);
+              goto quit;
+            }
+
+          ++idle;
+
+          if (idle <= max_idle)
+            /* we are allowed to idle, so do so without any timeout */
+            X_COND_WAIT (reqwait, reqlock);
+          else
+            {
+              /* initialise timeout once */
+              if (!ts.tv_sec)
+                ts.tv_sec = time (0) + idle_timeout;
+
+              if (X_COND_TIMEDWAIT (reqwait, reqlock, ts) == ETIMEDOUT)
+                ts.tv_sec = 1; /* assuming this is not a value computed above.,.. */
+            }
+
+          --idle;
+        }
+
+      --nready;
+
+      X_UNLOCK (reqlock);
+     
+      if (req->type < 0)
+        goto quit;
+
+      ETP_EXECUTE (self, req);
+
+      X_LOCK (reslock);
+
+      ++npending;
+
+      if (!reqq_push (&res_queue, req) && want_poll_cb)
+        want_poll_cb ();
+
+      etp_worker_clear (self);
+
+      X_UNLOCK (reslock);
+    }
+
+quit:
+  free (req);
+
+  X_LOCK (wrklock);
+  etp_worker_free (self);
+  X_UNLOCK (wrklock);
+}
+
+/*****************************************************************************/
+
+int ecb_cold
+eio_init (void (*want_poll)(void), void (*done_poll)(void))
+{
+#if !HAVE_PREADWRITE
+  X_MUTEX_CREATE (preadwritelock);
+#endif
+
+  return etp_init (want_poll, done_poll);
+}
+
+ecb_inline void
+eio_api_destroy (eio_req *req)
+{
+  free (req);
+}
+
+#define REQ(rtype)                                            	\
+  eio_req *req;                                                 \
+                                                                \
+  req = (eio_req *)calloc (1, sizeof *req);                     \
+  if (!req)                                                     \
+    return 0;                                                   \
+                                                                \
+  req->type    = rtype;                                         \
+  req->pri     = pri;						\
+  req->finish  = cb;						\
+  req->data    = data;						\
+  req->destroy = eio_api_destroy;
+
+#define SEND eio_submit (req); return req
+
+#define PATH							\
+  req->flags |= EIO_FLAG_PTR1_FREE;				\
+  req->ptr1 = strdup (path);					\
+  if (!req->ptr1)						\
+    {								\
+      eio_api_destroy (req);					\
+      return 0;							\
+    }
+
+static void
+eio_execute (etp_worker *self, eio_req *req)
+{
+#if HAVE_AT
+  int dirfd;
+#else
+  const char *path;
+#endif
+
+  if (ecb_expect_false (EIO_CANCELLED (req)))
+    {
+      req->result  = -1;
+      req->errorno = ECANCELED;
+      return;
+    }
+
+  if (ecb_expect_false (req->wd == EIO_INVALID_WD))
+    {
+      req->result  = -1;
+      req->errorno = ENOENT;
+      return;
+    }
+
+  if (req->type >= EIO_OPEN)
+    {
+      #if HAVE_AT
+        dirfd = WD2FD (req->wd);
+      #else
+        path = wd_expand (&self->tmpbuf, req->wd, req->ptr1);
+      #endif
+    }
+
+  switch (req->type)
+    {
+      case EIO_WD_OPEN:   req->wd = eio__wd_open_sync (&self->tmpbuf, req->wd, req->ptr1);
+                          req->result = req->wd == EIO_INVALID_WD ? -1 : 0;
+                          break;
+      case EIO_WD_CLOSE:  req->result = 0;
+                          eio_wd_close_sync (req->wd); break;
+
+      case EIO_SEEK:      eio__lseek (req); break;
+      case EIO_READ:      ALLOC (req->size);
+                          req->result = req->offs >= 0
+                                      ? pread     (req->int1, req->ptr2, req->size, req->offs)
+                                      : read      (req->int1, req->ptr2, req->size); break;
+      case EIO_WRITE:     req->result = req->offs >= 0
+                                      ? pwrite    (req->int1, req->ptr2, req->size, req->offs)
+                                      : write     (req->int1, req->ptr2, req->size); break;
+
+      case EIO_READAHEAD: req->result = readahead     (req->int1, req->offs, req->size); break;
+      case EIO_SENDFILE:  req->result = eio__sendfile (req->int1, req->int2, req->offs, req->size); break;
+
+#if HAVE_AT
+
+      case EIO_STAT:      ALLOC (sizeof (EIO_STRUCT_STAT));
+                          req->result = fstatat   (dirfd, req->ptr1, (EIO_STRUCT_STAT *)req->ptr2, 0); break;
+      case EIO_LSTAT:     ALLOC (sizeof (EIO_STRUCT_STAT));
+                          req->result = fstatat   (dirfd, req->ptr1, (EIO_STRUCT_STAT *)req->ptr2, AT_SYMLINK_NOFOLLOW); break;
+      case EIO_CHOWN:     req->result = fchownat  (dirfd, req->ptr1, req->int2, req->int3, 0); break;
+      case EIO_CHMOD:     req->result = fchmodat  (dirfd, req->ptr1, (mode_t)req->int2, 0); break;
+      case EIO_TRUNCATE:  req->result = eio__truncateat (dirfd, req->ptr1, req->offs); break;
+      case EIO_OPEN:      req->result = openat    (dirfd, req->ptr1, req->int1, (mode_t)req->int2); break;
+
+      case EIO_UNLINK:    req->result = unlinkat  (dirfd, req->ptr1, 0); break;
+      case EIO_RMDIR:     req->result = unlinkat  (dirfd, req->ptr1, AT_REMOVEDIR); break;
+      case EIO_MKDIR:     req->result = mkdirat   (dirfd, req->ptr1, (mode_t)req->int2); break;
+      case EIO_RENAME:    req->result = renameat  (dirfd, req->ptr1, WD2FD ((eio_wd)req->int3), req->ptr2); break;
+      case EIO_LINK:      req->result = linkat    (dirfd, req->ptr1, WD2FD ((eio_wd)req->int3), req->ptr2, 0); break;
+      case EIO_SYMLINK:   req->result = symlinkat (req->ptr1, dirfd, req->ptr2); break;
+      case EIO_MKNOD:     req->result = mknodat   (dirfd, req->ptr1, (mode_t)req->int2, (dev_t)req->offs); break;
+      case EIO_READLINK:  ALLOC (PATH_MAX);
+                          req->result = readlinkat (dirfd, req->ptr1, req->ptr2, PATH_MAX); break;
+      case EIO_STATVFS:   ALLOC (sizeof (EIO_STRUCT_STATVFS));
+                          req->result = eio__statvfsat (dirfd, req->ptr1, (EIO_STRUCT_STATVFS *)req->ptr2); break;
+      case EIO_UTIME:
+      case EIO_FUTIME:
+        {
+          struct timespec ts[2];
+          struct timespec *times;
+
+          if (req->nv1 != -1. || req->nv2 != -1.)
+            {
+              ts[0].tv_sec  = req->nv1;
+              ts[0].tv_nsec = (req->nv1 - ts[0].tv_sec) * 1e9;
+              ts[1].tv_sec  = req->nv2;
+              ts[1].tv_nsec = (req->nv2 - ts[1].tv_sec) * 1e9;
+
+              times = ts;
+            }
+          else
+            times = 0;
+
+          req->result = req->type == EIO_FUTIME
+                        ? futimens  (req->int1, times)
+                        : utimensat (dirfd, req->ptr1, times, 0);
+        }
+        break;
+
+#else
+
+      case EIO_STAT:      ALLOC (sizeof (EIO_STRUCT_STAT));
+                          req->result = stat      (path     , (EIO_STRUCT_STAT *)req->ptr2); break;
+      case EIO_LSTAT:     ALLOC (sizeof (EIO_STRUCT_STAT));
+                          req->result = lstat     (path     , (EIO_STRUCT_STAT *)req->ptr2); break;
+      case EIO_CHOWN:     req->result = chown     (path     , req->int2, req->int3); break;
+      case EIO_CHMOD:     req->result = chmod     (path     , (mode_t)req->int2); break;
+      case EIO_TRUNCATE:  req->result = truncate  (path     , req->offs); break;
+      case EIO_OPEN:      req->result = open      (path     , req->int1, (mode_t)req->int2); break;
+
+      case EIO_UNLINK:    req->result = unlink    (path     ); break;
+      case EIO_RMDIR:     req->result = rmdir     (path     ); break;
+      case EIO_MKDIR:     req->result = mkdir     (path     , (mode_t)req->int2); break;
+      case EIO_RENAME:    req->result = rename    (path     , req->ptr2); break;
+      case EIO_LINK:      req->result = link      (path     , req->ptr2); break;
+      case EIO_SYMLINK:   req->result = symlink   (path     , req->ptr2); break;
+      case EIO_MKNOD:     req->result = mknod     (path     , (mode_t)req->int2, (dev_t)req->offs); break;
+      case EIO_READLINK:  ALLOC (PATH_MAX);
+                          req->result = readlink  (path, req->ptr2, PATH_MAX); break;
+      case EIO_STATVFS:   ALLOC (sizeof (EIO_STRUCT_STATVFS));
+                          req->result = statvfs   (path     , (EIO_STRUCT_STATVFS *)req->ptr2); break;
+
+      case EIO_UTIME:
+      case EIO_FUTIME:
+        {
+          struct timeval tv[2];
+          struct timeval *times;
+
+          if (req->nv1 != -1. || req->nv2 != -1.)
+            {
+              tv[0].tv_sec  = req->nv1;
+              tv[0].tv_usec = (req->nv1 - tv[0].tv_sec) * 1e6;
+              tv[1].tv_sec  = req->nv2;
+              tv[1].tv_usec = (req->nv2 - tv[1].tv_sec) * 1e6;
+
+              times = tv;
+            }
+          else
+            times = 0;
+
+          req->result = req->type == EIO_FUTIME
+                        ? futimes (req->int1, times)
+                        : utimes  (req->ptr1, times);
+        }
+        break;
+
+#endif
+
+      case EIO_REALPATH:  if (0 <= (req->result = eio__realpath (&self->tmpbuf, req->wd, req->ptr1)))
+                            {
+                              ALLOC (req->result);
+                              memcpy (req->ptr2, self->tmpbuf.ptr, req->result);
+                            }
+                           break;
+
+      case EIO_FSTAT:     ALLOC (sizeof (EIO_STRUCT_STAT));
+                          req->result = fstat     (req->int1, (EIO_STRUCT_STAT *)req->ptr2); break;
+
+      case EIO_FSTATVFS:  ALLOC (sizeof (EIO_STRUCT_STATVFS));
+                          req->result = fstatvfs  (req->int1, (EIO_STRUCT_STATVFS *)req->ptr2); break;
+
+      case EIO_FCHOWN:    req->result = fchown    (req->int1, req->int2, req->int3); break;
+      case EIO_FCHMOD:    req->result = fchmod    (req->int1, (mode_t)req->int2); break;
+      case EIO_FTRUNCATE: req->result = ftruncate (req->int1, req->offs); break;
+
+      case EIO_CLOSE:     req->result = close     (req->int1); break;
+      case EIO_DUP2:      req->result = dup2      (req->int1, req->int2); break;
+      case EIO_SYNC:      req->result = 0; sync (); break;
+      case EIO_FSYNC:     req->result = fsync     (req->int1); break;
+      case EIO_FDATASYNC: req->result = fdatasync (req->int1); break;
+      case EIO_SYNCFS:    req->result = eio__syncfs (req->int1); break;
+      case EIO_SYNC_FILE_RANGE: req->result = eio__sync_file_range (req->int1, req->offs, req->size, req->int2); break;
+      case EIO_MSYNC:     req->result = eio__msync (req->ptr2, req->size, req->int1); break;
+      case EIO_MTOUCH:    req->result = eio__mtouch (req); break;
+      case EIO_MLOCK:     req->result = eio__mlock (req->ptr2, req->size); break;
+      case EIO_MLOCKALL:  req->result = eio__mlockall (req->int1); break;
+      case EIO_FALLOCATE: req->result = eio__fallocate (req->int1, req->int2, req->offs, req->size); break;
+
+      case EIO_READDIR:   eio__scandir (req, self); break;
+
+      case EIO_BUSY:
+#ifdef _WIN32
+	Sleep (req->nv1 * 1e3);
+#else
+        {
+          struct timeval tv;
+
+          tv.tv_sec  = req->nv1;
+          tv.tv_usec = (req->nv1 - tv.tv_sec) * 1e6;
+
+          req->result = select (0, 0, 0, 0, &tv);
+        }
+#endif
+        break;
+
+      case EIO_GROUP:
+        abort (); /* handled in eio_request */
+
+      case EIO_NOP:
+        req->result = 0;
+        break;
+
+      case EIO_CUSTOM:
+        req->feed (req);
+        break;
+
+      default:
+        req->result = EIO_ENOSYS ();
+        break;
+    }
+
+  req->errorno = errno;
+}
+
+#ifndef EIO_NO_WRAPPERS
+
+eio_req *eio_wd_open (const char *path, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_WD_OPEN); PATH; SEND;
+}
+
+eio_req *eio_wd_close (eio_wd wd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_WD_CLOSE); req->wd = wd; SEND;
+}
+
+eio_req *eio_nop (int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_NOP); SEND;
+}
+
+eio_req *eio_busy (double delay, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_BUSY); req->nv1 = delay; SEND;
+}
+
+eio_req *eio_sync (int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_SYNC); SEND;
+}
+
+eio_req *eio_fsync (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FSYNC); req->int1 = fd; SEND;
+}
+
+eio_req *eio_msync (void *addr, size_t length, int flags, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MSYNC); req->ptr2 = addr; req->size = length; req->int1 = flags; SEND;
+}
+
+eio_req *eio_fdatasync (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FDATASYNC); req->int1 = fd; SEND;
+}
+
+eio_req *eio_syncfs (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_SYNCFS); req->int1 = fd; SEND;
+}
+
+eio_req *eio_sync_file_range (int fd, off_t offset, size_t nbytes, unsigned int flags, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_SYNC_FILE_RANGE); req->int1 = fd; req->offs = offset; req->size = nbytes; req->int2 = flags; SEND;
+}
+
+eio_req *eio_mtouch (void *addr, size_t length, int flags, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MTOUCH); req->ptr2 = addr; req->size = length; req->int1 = flags; SEND;
+}
+
+eio_req *eio_mlock (void *addr, size_t length, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MLOCK); req->ptr2 = addr; req->size = length; SEND;
+}
+
+eio_req *eio_mlockall (int flags, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MLOCKALL); req->int1 = flags; SEND;
+}
+
+eio_req *eio_fallocate (int fd, int mode, off_t offset, size_t len, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FALLOCATE); req->int1 = fd; req->int2 = mode; req->offs = offset; req->size = len; SEND;
+}
+
+eio_req *eio_close (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_CLOSE); req->int1 = fd; SEND;
+}
+
+eio_req *eio_readahead (int fd, off_t offset, size_t length, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_READAHEAD); req->int1 = fd; req->offs = offset; req->size = length; SEND;
+}
+
+eio_req *eio_seek (int fd, off_t offset, int whence, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_SEEK); req->int1 = fd; req->offs = offset; req->int2 = whence; SEND;
+}
+
+eio_req *eio_read (int fd, void *buf, size_t length, off_t offset, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_READ); req->int1 = fd; req->offs = offset; req->size = length; req->ptr2 = buf; SEND;
+}
+
+eio_req *eio_write (int fd, void *buf, size_t length, off_t offset, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_WRITE); req->int1 = fd; req->offs = offset; req->size = length; req->ptr2 = buf; SEND;
+}
+
+eio_req *eio_fstat (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FSTAT); req->int1 = fd; SEND;
+}
+
+eio_req *eio_fstatvfs (int fd, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FSTATVFS); req->int1 = fd; SEND;
+}
+
+eio_req *eio_futime (int fd, double atime, double mtime, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FUTIME); req->int1 = fd; req->nv1 = atime; req->nv2 = mtime; SEND;
+}
+
+eio_req *eio_ftruncate (int fd, off_t offset, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FTRUNCATE); req->int1 = fd; req->offs = offset; SEND;
+}
+
+eio_req *eio_fchmod (int fd, mode_t mode, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FCHMOD); req->int1 = fd; req->int2 = (long)mode; SEND;
+}
+
+eio_req *eio_fchown (int fd, eio_uid_t uid, eio_gid_t gid, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_FCHOWN); req->int1 = fd; req->int2 = (long)uid; req->int3 = (long)gid; SEND;
+}
+
+eio_req *eio_dup2 (int fd, int fd2, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_DUP2); req->int1 = fd; req->int2 = fd2; SEND;
+}
+
+eio_req *eio_sendfile (int out_fd, int in_fd, off_t in_offset, size_t length, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_SENDFILE); req->int1 = out_fd; req->int2 = in_fd; req->offs = in_offset; req->size = length; SEND;
+}
+
+eio_req *eio_open (const char *path, int flags, mode_t mode, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_OPEN); PATH; req->int1 = flags; req->int2 = (long)mode; SEND;
+}
+
+eio_req *eio_utime (const char *path, double atime, double mtime, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_UTIME); PATH; req->nv1 = atime; req->nv2 = mtime; SEND;
+}
+
+eio_req *eio_truncate (const char *path, off_t offset, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_TRUNCATE); PATH; req->offs = offset; SEND;
+}
+
+eio_req *eio_chown (const char *path, eio_uid_t uid, eio_gid_t gid, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_CHOWN); PATH; req->int2 = (long)uid; req->int3 = (long)gid; SEND;
+}
+
+eio_req *eio_chmod (const char *path, mode_t mode, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_CHMOD); PATH; req->int2 = (long)mode; SEND;
+}
+
+eio_req *eio_mkdir (const char *path, mode_t mode, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MKDIR); PATH; req->int2 = (long)mode; SEND;
+}
+
+static eio_req *
+eio__1path (int type, const char *path, int pri, eio_cb cb, void *data)
+{
+  REQ (type); PATH; SEND;
+}
+
+eio_req *eio_readlink (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_READLINK, path, pri, cb, data);
+}
+
+eio_req *eio_realpath (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_REALPATH, path, pri, cb, data);
+}
+
+eio_req *eio_stat (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_STAT, path, pri, cb, data);
+}
+
+eio_req *eio_lstat (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_LSTAT, path, pri, cb, data);
+}
+
+eio_req *eio_statvfs (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_STATVFS, path, pri, cb, data);
+}
+
+eio_req *eio_unlink (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_UNLINK, path, pri, cb, data);
+}
+
+eio_req *eio_rmdir (const char *path, int pri, eio_cb cb, void *data)
+{
+  return eio__1path (EIO_RMDIR, path, pri, cb, data);
+}
+
+eio_req *eio_readdir (const char *path, int flags, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_READDIR); PATH; req->int1 = flags; SEND;
+}
+
+eio_req *eio_mknod (const char *path, mode_t mode, dev_t dev, int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_MKNOD); PATH; req->int2 = (long)mode; req->offs = (off_t)dev; SEND;
+}
+
+static eio_req *
+eio__2path (int type, const char *path, const char *new_path, int pri, eio_cb cb, void *data)
+{
+  REQ (type); PATH;
+
+  req->flags |= EIO_FLAG_PTR2_FREE;
+  req->ptr2 = strdup (new_path);
+  if (!req->ptr2)
+    {
+      eio_api_destroy (req);
+      return 0;
+    }
+
+  SEND;
+}
+
+eio_req *eio_link (const char *path, const char *new_path, int pri, eio_cb cb, void *data)
+{
+  return eio__2path (EIO_LINK, path, new_path, pri, cb, data);
+}
+
+eio_req *eio_symlink (const char *path, const char *new_path, int pri, eio_cb cb, void *data)
+{
+  return eio__2path (EIO_SYMLINK, path, new_path, pri, cb, data);
+}
+
+eio_req *eio_rename (const char *path, const char *new_path, int pri, eio_cb cb, void *data)
+{
+  return eio__2path (EIO_RENAME, path, new_path, pri, cb, data);
+}
+
+eio_req *eio_custom (void (*execute)(eio_req *), int pri, eio_cb cb, void *data)
+{
+  REQ (EIO_CUSTOM); req->feed = execute; SEND;
+}
+
+#endif
+
+eio_req *eio_grp (eio_cb cb, void *data)
+{
+  const int pri = EIO_PRI_MAX;
+
+  REQ (EIO_GROUP); SEND;
+}
+
+#undef REQ
+#undef PATH
+#undef SEND
+
+/*****************************************************************************/
+/* grp functions */
+
+void
+eio_grp_feed (eio_req *grp, void (*feed)(eio_req *req), int limit)
+{
+  grp->int2 = limit;
+  grp->feed = feed;
+
+  grp_try_feed (grp);
+}
+
+void
+eio_grp_limit (eio_req *grp, int limit)
+{
+  grp->int2 = limit;
+
+  grp_try_feed (grp);
+}
+
+void
+eio_grp_add (eio_req *grp, eio_req *req)
+{
+  assert (("cannot add requests to IO::AIO::GRP after the group finished", grp->int1 != 2));
+
+  grp->flags |= EIO_FLAG_GROUPADD;
+
+  ++grp->size;
+  req->grp = grp;
+
+  req->grp_prev = 0;
+  req->grp_next = grp->grp_first;
+
+  if (grp->grp_first)
+    grp->grp_first->grp_prev = req;
+
+  grp->grp_first = req;
+}
+
+/*****************************************************************************/
+/* misc garbage */
+
+eio_ssize_t
+eio_sendfile_sync (int ofd, int ifd, off_t offset, size_t count)
+{
+  return eio__sendfile (ofd, ifd, offset, count);
+}
+