libev/ev_iouring.c

/*
 * libev linux io_uring fd activity backend
 *
 * Copyright (c) 2019 Marc Alexander Lehmann <libev@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.
 */

/*
 * general notes about linux io_uring:
 *
 * a) it's the best interface I have seen so far. on linux.
 * b) best is not necessarily very good.
 * c) it's better than the aio mess, doesn't suffer from the fork problems
 *    of linux aio or epoll and so on and so on. and you could do event stuff
 *    without any syscalls. what's not to like?
 * d) ok, it's vastly more complex, but that's ok, really.
 * e) why 3 mmaps instead of one? one would be more space-efficient,
 *    and I can't see what benefit three would have (other than being
 *    somehow resizable/relocatable, but that's apparently not possible).
 *    (FIXME: newer kernels can use 2 mmaps only, need to look into this).
 * f) hmm, it's practiclaly undebuggable (gdb can't access the memory, and
 *    the bizarre way structure offsets are communicated makes it hard to
 *    just print the ring buffer heads, even *iff* the memory were visible
 *    in gdb. but then, that's also ok, really.
 * g) well, you cannot specify a timeout when waiting for events. no,
 *    seriously, the interface doesn't support a timeout. never seen _that_
 *    before. sure, you can use a timerfd, but that's another syscall
 *    you could have avoided. overall, this bizarre omission smells
 *    like a µ-optimisation by the io_uring author for his personal
 *    applications, to the detriment of everybody else who just wants
 *    an event loop. but, umm, ok, if that's all, it could be worse.
 *    (FIXME: jens mentioned timeout commands, need to investigate)
 * h) there is a hardcoded limit of 4096 outstanding events. okay,
 *    at least there is no arbitrary low system-wide limit...
 *    (FIXME: apparently, this was increased to 32768 in later kernels(
 * i) unlike linux aio, you *can* register more then the limit
 *    of fd events, and the kernel will "gracefully" signal an
 *    overflow, after which you could destroy and recreate the kernel
 *    state, a bit bigger, or fall back to e.g. poll. thats not
 *    totally insane, but kind of questions the point a high
 *    performance I/O framework when it doesn't really work
 *    under stress.
 *    (FIXME: iouring should no longer drop events, need to investigate)
 * j) but, oh my! is has exactly the same bugs as the linux aio backend,
 *    where some undocumented poll combinations just fail.
 *    so we need epoll AGAIN as a fallback. AGAIN! epoll!! and of course,
 *    this is completely undocumented, have I mantioned this already?
 * k) overall, the *API* itself is, I dare to say, not a total trainwreck.
 *    the big isuess with it are the bugs requiring epoll, which might
 *    or might not get fixed (do I hold my breath?).
 */

#include <sys/timerfd.h>
#include <sys/mman.h>
#include <poll.h>

#define IOURING_INIT_ENTRIES 32

/*****************************************************************************/
/* syscall wrapdadoop - this section has the raw api/abi definitions */

#include <linux/fs.h>
#include <linux/types.h>

/* mostly directly taken from the kernel or documentation */

struct io_uring_sqe
{
  __u8 opcode;
  __u8 flags;
  __u16 ioprio;
  __s32 fd;
  __u64 off;
  __u64 addr;
  __u32 len;
  union {
    __kernel_rwf_t rw_flags;
    __u32 fsync_flags;
    __u16 poll_events;
    __u32 sync_range_flags;
    __u32 msg_flags;
  };
  __u64 user_data;
  union {
    __u16 buf_index;
    __u64 __pad2[3];
  };
};

struct io_uring_cqe
{
  __u64 user_data;
  __s32 res;
  __u32 flags;
};

struct io_sqring_offsets
{
  __u32 head;
  __u32 tail;
  __u32 ring_mask;
  __u32 ring_entries;
  __u32 flags;
  __u32 dropped;
  __u32 array;
  __u32 resv1;
  __u64 resv2;
};

struct io_cqring_offsets
{
  __u32 head;
  __u32 tail;
  __u32 ring_mask;
  __u32 ring_entries;
  __u32 overflow;
  __u32 cqes;
  __u64 resv[2];
};

struct io_uring_params
{
  __u32 sq_entries;
  __u32 cq_entries;
  __u32 flags;
  __u32 sq_thread_cpu;
  __u32 sq_thread_idle;
  __u32 resv[5];
  struct io_sqring_offsets sq_off;
  struct io_cqring_offsets cq_off;
};

#define IORING_OP_POLL_ADD    6
#define IORING_OP_POLL_REMOVE 7

#define IORING_ENTER_GETEVENTS 0x01

#define IORING_OFF_SQ_RING 0x00000000ULL
#define IORING_OFF_CQ_RING 0x08000000ULL
#define IORING_OFF_SQES    0x10000000ULL

inline_size
int
evsys_io_uring_setup (unsigned entries, struct io_uring_params *params)
{
  return ev_syscall2 (SYS_io_uring_setup, entries, params);
}

inline_size
int
evsys_io_uring_enter (int fd, unsigned to_submit, unsigned min_complete, unsigned flags, const sigset_t *sig, size_t sigsz)
{
  return ev_syscall6 (SYS_io_uring_enter, fd, to_submit, min_complete, flags, sig, sigsz);
}

/*****************************************************************************/
/* actual backed implementation */

/* we hope that volatile will make the compiler access this variables only once */
#define EV_SQ_VAR(name) *(volatile unsigned *)((char *)iouring_sq_ring + iouring_sq_ ## name)
#define EV_CQ_VAR(name) *(volatile unsigned *)((char *)iouring_cq_ring + iouring_cq_ ## name)

/* the index array */
#define EV_SQ_ARRAY     ((unsigned *)((char *)iouring_sq_ring + iouring_sq_array))

/* the submit/completion queue entries */
#define EV_SQES         ((struct io_uring_sqe *)         iouring_sqes)
#define EV_CQES         ((struct io_uring_cqe *)((char *)iouring_cq_ring + iouring_cq_cqes))

static
struct io_uring_sqe *
iouring_sqe_get (EV_P)
{
  unsigned tail = EV_SQ_VAR (tail);

  if (tail + 1 - EV_SQ_VAR (head) > EV_SQ_VAR (ring_entries))
    {
      /* queue full, flush */
      evsys_io_uring_enter (iouring_fd, iouring_to_submit, 0, 0, 0, 0);
      iouring_to_submit = 0;
    }

  assert (("libev: io_uring queue full after flush", tail + 1 - EV_SQ_VAR (head) <= EV_SQ_VAR (ring_entries)));

  return EV_SQES + (tail & EV_SQ_VAR (ring_mask));
}

inline_size
struct io_uring_sqe *
iouring_sqe_submit (EV_P_ struct io_uring_sqe *sqe)
{
  unsigned idx = sqe - EV_SQES;

  EV_SQ_ARRAY [idx] = idx;
  ECB_MEMORY_FENCE_RELEASE;
  ++EV_SQ_VAR (tail);
  /*ECB_MEMORY_FENCE_RELEASE; /* for the time being we assume this is not needed */
  ++iouring_to_submit;
}

/*****************************************************************************/

/* when the timerfd expires we simply note the fact,
 * as the purpose of the timerfd is to wake us up, nothing else.
 * the next iteration should re-set it.
 */
static void
iouring_tfd_cb (EV_P_ struct ev_io *w, int revents)
{
  iouring_tfd_to = EV_TSTAMP_HUGE;
}

static void
iouring_epoll_cb (EV_P_ struct ev_io *w, int revents)
{
  epoll_poll (EV_A_ 0);
}

/* called for full and partial cleanup */
ecb_cold
static int
iouring_internal_destroy (EV_P)
{
  close (iouring_tfd);
  close (iouring_fd);

  if (iouring_sq_ring != MAP_FAILED) munmap (iouring_sq_ring, iouring_sq_ring_size);
  if (iouring_cq_ring != MAP_FAILED) munmap (iouring_cq_ring, iouring_cq_ring_size);
  if (iouring_sqes    != MAP_FAILED) munmap (iouring_sqes   , iouring_sqes_size   );

  if (ev_is_active (&iouring_epoll_w)) ev_ref (EV_A); ev_io_stop (EV_A_ &iouring_epoll_w);
  if (ev_is_active (&iouring_tfd_w  )) ev_ref (EV_A); ev_io_stop (EV_A_ &iouring_tfd_w  );
}

ecb_cold
static int
iouring_internal_init (EV_P)
{
  struct io_uring_params params = { 0 };

  iouring_to_submit = 0;

  iouring_tfd     = -1;
  iouring_sq_ring = MAP_FAILED;
  iouring_cq_ring = MAP_FAILED;
  iouring_sqes    = MAP_FAILED;

  for (;;)
    {
      iouring_fd = evsys_io_uring_setup (iouring_entries, &params);

      if (iouring_fd >= 0)
        break; /* yippie */

      if (errno != EINVAL)
        return -1; /* we failed */

      /* EINVAL: lots of possible reasons, but maybe
       * it is because we hit the unqueryable hardcoded size limit
       */

      /* we hit the limit already, give up */
      if (iouring_max_entries)
        return -1;

      /* first time we hit EINVAL? assume we hit the limit, so go back and retry */
      iouring_entries >>= 1;
      iouring_max_entries = iouring_entries;
    }

  iouring_sq_ring_size = params.sq_off.array + params.sq_entries * sizeof (unsigned);
  iouring_cq_ring_size = params.cq_off.cqes  + params.cq_entries * sizeof (struct io_uring_cqe);
  iouring_sqes_size    =                       params.sq_entries * sizeof (struct io_uring_sqe);

  iouring_sq_ring = mmap (0, iouring_sq_ring_size, PROT_READ | PROT_WRITE,
                          MAP_SHARED | MAP_POPULATE, iouring_fd, IORING_OFF_SQ_RING);
  iouring_cq_ring = mmap (0, iouring_cq_ring_size, PROT_READ | PROT_WRITE,
                          MAP_SHARED | MAP_POPULATE, iouring_fd, IORING_OFF_CQ_RING);
  iouring_sqes    = mmap (0, iouring_sqes_size, PROT_READ | PROT_WRITE,
                          MAP_SHARED | MAP_POPULATE, iouring_fd, IORING_OFF_SQES);

  if (iouring_sq_ring == MAP_FAILED || iouring_cq_ring == MAP_FAILED || iouring_sqes == MAP_FAILED)
    return -1;

  iouring_sq_head         = params.sq_off.head;
  iouring_sq_tail         = params.sq_off.tail;
  iouring_sq_ring_mask    = params.sq_off.ring_mask;
  iouring_sq_ring_entries = params.sq_off.ring_entries;
  iouring_sq_flags        = params.sq_off.flags;
  iouring_sq_dropped      = params.sq_off.dropped;
  iouring_sq_array        = params.sq_off.array;

  iouring_cq_head         = params.cq_off.head;
  iouring_cq_tail         = params.cq_off.tail;
  iouring_cq_ring_mask    = params.cq_off.ring_mask;
  iouring_cq_ring_entries = params.cq_off.ring_entries;
  iouring_cq_overflow     = params.cq_off.overflow;
  iouring_cq_cqes         = params.cq_off.cqes;

  iouring_tfd = timerfd_create (CLOCK_MONOTONIC, TFD_CLOEXEC);

  if (iouring_tfd < 0)
    return iouring_tfd;

  iouring_tfd_to = EV_TSTAMP_HUGE;

  return 0;
}

ecb_cold
static void
iouring_fork (EV_P)
{
  iouring_internal_destroy (EV_A);

  while (iouring_internal_init (EV_A) < 0)
    ev_syserr ("(libev) io_uring_setup");

  /* forking epoll should also effectively unregister all fds from the backend */
  epoll_fork (EV_A);
  /* epoll_fork already did this. hopefully */
  /*fd_rearm_all (EV_A);*/

  ev_io_stop  (EV_A_ &iouring_epoll_w);
  ev_io_set   (EV_A_ &iouring_epoll_w, backend_fd, EV_READ);
  ev_io_start (EV_A_ &iouring_epoll_w);

  ev_io_stop  (EV_A_ &iouring_tfd_w);
  ev_io_set   (EV_A_ &iouring_tfd_w, iouring_tfd, EV_READ);
  ev_io_start (EV_A_ &iouring_tfd_w);
}

/*****************************************************************************/

static void
iouring_modify (EV_P_ int fd, int oev, int nev)
{
  if (ecb_expect_false (anfds [fd].eflags))
    {
      /* we handed this fd over to epoll, so undo this first */
      /* we do it manually because the optimisations on epoll_modify won't do us any good */
      epoll_ctl (iouring_fd, EPOLL_CTL_DEL, fd, 0);
      anfds [fd].eflags = 0;
      oev = 0;
    }

  if (oev)
    {
      /* we assume the sqe's are all "properly" initialised */
      struct io_uring_sqe *sqe = iouring_sqe_get (EV_A);
      sqe->opcode    = IORING_OP_POLL_REMOVE;
      sqe->fd        = fd;
      sqe->user_data = -1;
      iouring_sqe_submit (EV_A_ sqe);

      /* increment generation counter to avoid handling old events */
      ++anfds [fd].egen;
    }

  if (nev)
    {
      struct io_uring_sqe *sqe = iouring_sqe_get (EV_A);
      sqe->opcode      = IORING_OP_POLL_ADD;
      sqe->fd          = fd;
      sqe->user_data   = (uint32_t)fd | ((__u64)(uint32_t)anfds [fd].egen << 32);
      sqe->poll_events =
        (nev & EV_READ ? POLLIN : 0)
        | (nev & EV_WRITE ? POLLOUT : 0);
      iouring_sqe_submit (EV_A_ sqe);
    }
}

inline_size
void
iouring_tfd_update (EV_P_ ev_tstamp timeout)
{
  ev_tstamp tfd_to = mn_now + timeout;

  /* we assume there will be many iterations per timer change, so
   * we only re-set the timerfd when we have to because its expiry
   * is too late.
   */
  if (ecb_expect_false (tfd_to < iouring_tfd_to))
    {
       struct itimerspec its;

       iouring_tfd_to = tfd_to;
       EV_TS_SET (its.it_interval, 0.);
       EV_TS_SET (its.it_value, tfd_to);

       if (timerfd_settime (iouring_tfd, TFD_TIMER_ABSTIME, &its, 0) < 0)
         assert (("libev: iouring timerfd_settime failed", 0));
    }
}

inline_size
void
iouring_process_cqe (EV_P_ struct io_uring_cqe *cqe)
{
  int      fd  = cqe->user_data & 0xffffffffU;
  uint32_t gen = cqe->user_data >> 32;
  int      res = cqe->res;

  /* ignore fd removal events, if there are any. TODO: verify */
  if (cqe->user_data == (__u64)-1)
    abort ();//D

  assert (("libev: io_uring fd must be in-bounds", fd >= 0 && fd < anfdmax));

  /* documentation lies, of course. the result value is NOT like
   * normal syscalls, but like linux raw syscalls, i.e. negative
   * error numbers. fortunate, as otherwise there would be no way
   * to get error codes at all. still, why not document this?
   */

  /* ignore event if generation doesn't match */
  /* this should actually be very rare */
  if (ecb_expect_false (gen != (uint32_t)anfds [fd].egen))
    return;

  if (ecb_expect_false (res < 0))
    {
      if (res == -EINVAL)
        {
          /* we assume this error code means the fd/poll combination is buggy
           * and fall back to epoll.
           * this error code might also indicate a bug, but the kernel doesn't
           * distinguish between those two conditions, so... sigh...
           */

          epoll_modify (EV_A_ fd, 0, anfds [fd].events);
        }
      else if (res == -EBADF)
        {
          assert (("libev: event loop rejected bad fd", res != -EBADF));
          fd_kill (EV_A_ fd);
        }
      else
        {
          errno = -res;
          ev_syserr ("(libev) IORING_OP_POLL_ADD");
        }

      return;
    }

  /* feed events, we do not expect or handle POLLNVAL */
  fd_event (
    EV_A_
    fd,
    (res & (POLLOUT | POLLERR | POLLHUP) ? EV_WRITE : 0)
    | (res & (POLLIN | POLLERR | POLLHUP) ? EV_READ : 0)
  );

  /* io_uring is oneshot, so we need to re-arm the fd next iteration */
  /* this also means we usually have to do at least one syscall per iteration */
  anfds [fd].events = 0;
  fd_change (EV_A_ fd, EV_ANFD_REIFY);
}

/* called when the event queue overflows */
ecb_cold
static void
iouring_overflow (EV_P)
{
  /* we have two options, resize the queue (by tearing down
   * everything and recreating it, or living with it
   * and polling.
   * we implement this by resizing tghe queue, and, if that fails,
   * we just recreate the state on every failure, which
   * kind of is a very inefficient poll.
   * one danger is, due to the bios toward lower fds,
   * we will only really get events for those, so
   * maybe we need a poll() fallback, after all.
   */
  /*EV_CQ_VAR (overflow) = 0;*/ /* need to do this if we keep the state and poll manually */

  fd_rearm_all (EV_A);

  /* we double the size until we hit the hard-to-probe maximum */
  if (!iouring_max_entries)
    {
      iouring_entries <<= 1;
      iouring_fork (EV_A);
    }
  else
    {
      /* we hit the kernel limit, we should fall back to something else.
       * we can either poll() a few times and hope for the best,
       * poll always, or switch to epoll.
       * since we use epoll anyways, go epoll.
       */

      iouring_internal_destroy (EV_A);

      /* this should make it so that on return, we don'T call any uring functions */
      iouring_to_submit = 0;

      for (;;)
        {
          backend = epoll_init (EV_A_ 0);

          if (backend)
            break;

          ev_syserr ("(libev) iouring switch to epoll");
        }
    }
}

/* handle any events in the completion queue, return true if there were any */
static int
iouring_handle_cq (EV_P)
{
  unsigned head, tail, mask;
  
  head = EV_CQ_VAR (head);
  ECB_MEMORY_FENCE_ACQUIRE;
  tail = EV_CQ_VAR (tail);

  if (head == tail)
    return 0;

  /* it can only overflow if we have events, yes, yes? */
  if (ecb_expect_false (EV_CQ_VAR (overflow)))
    {
      iouring_overflow (EV_A);
      return 1;
    }

  mask = EV_CQ_VAR (ring_mask);

  do
    iouring_process_cqe (EV_A_ &EV_CQES [head++ & mask]);
  while (head != tail);

  EV_CQ_VAR (head) = head;
  ECB_MEMORY_FENCE_RELEASE;

  return 1;
}

static void
iouring_poll (EV_P_ ev_tstamp timeout)
{
  /* if we have events, no need for extra syscalls, but we might have to queue events */
  if (iouring_handle_cq (EV_A))
    timeout = EV_TS_CONST (0.);
  else
    /* no events, so maybe wait for some */
    iouring_tfd_update (EV_A_ timeout);

  /* only enter the kernel if we have something to submit, or we need to wait */
  if (timeout || iouring_to_submit)
    {
      int res;

      EV_RELEASE_CB;

      res = evsys_io_uring_enter (iouring_fd, iouring_to_submit, 1,
                                  timeout > EV_TS_CONST (0.) ? IORING_ENTER_GETEVENTS : 0, 0, 0);
      iouring_to_submit = 0;

      EV_ACQUIRE_CB;

      if (ecb_expect_false (res < 0))
        if (errno == EINTR)
          /* ignore */;
        else
          ev_syserr ("(libev) iouring setup");
      else
        iouring_handle_cq (EV_A);
    }
}

inline_size
int
iouring_init (EV_P_ int flags)
{
  if (!epoll_init (EV_A_ 0))
    return 0;

  iouring_entries     = IOURING_INIT_ENTRIES;
  iouring_max_entries = 0;

  if (iouring_internal_init (EV_A) < 0)
    {
      iouring_internal_destroy (EV_A);
      return 0;
    }

  ev_io_init  (&iouring_epoll_w, iouring_epoll_cb, backend_fd, EV_READ);
  ev_set_priority (&iouring_epoll_w, EV_MAXPRI);

  ev_io_init  (&iouring_tfd_w, iouring_tfd_cb, iouring_tfd, EV_READ);
  ev_set_priority (&iouring_tfd_w, EV_MAXPRI);

  ev_io_start (EV_A_ &iouring_epoll_w);
  ev_unref (EV_A); /* watcher should not keep loop alive */

  ev_io_start (EV_A_ &iouring_tfd_w);
  ev_unref (EV_A); /* watcher should not keep loop alive */

  backend_modify = iouring_modify;
  backend_poll   = iouring_poll;

  return EVBACKEND_IOURING;
}

inline_size
void
iouring_destroy (EV_P)
{
  iouring_internal_destroy (EV_A);
  epoll_destroy (EV_A);
}

Revision:	1.7
Committed:	Sun Dec 22 15:05:32 2019 UTC (4 years, 4 months ago) by root
Content type:	text/plain
Branch:	MAIN
Changes since 1.6:	+5 -1 lines
Log Message:	* empty log message *
#	Content
1	/*
2	* libev linux io_uring fd activity backend
3	*
4	* Copyright (c) 2019 Marc Alexander Lehmann <libev@schmorp.de>
5	* All rights reserved.
6	*
7	* Redistribution and use in source and binary forms, with or without modifica-
8	* tion, are permitted provided that the following conditions are met:
9	*
10	* 1. Redistributions of source code must retain the above copyright notice,
11	* this list of conditions and the following disclaimer.
12	*
13	* 2. Redistributions in binary form must reproduce the above copyright
14	* notice, this list of conditions and the following disclaimer in the
15	* documentation and/or other materials provided with the distribution.
16	*
17	* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED
18	* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MER-
19	* CHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO
20	* EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPE-
21	* CIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
22	* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
23	* OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
24	* WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTH-
25	* ERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
26	* OF THE POSSIBILITY OF SUCH DAMAGE.
27	*
28	* Alternatively, the contents of this file may be used under the terms of
29	* the GNU General Public License ("GPL") version 2 or any later version,
30	* in which case the provisions of the GPL are applicable instead of
31	* the above. If you wish to allow the use of your version of this file
32	* only under the terms of the GPL and not to allow others to use your
33	* version of this file under the BSD license, indicate your decision
34	* by deleting the provisions above and replace them with the notice
35	* and other provisions required by the GPL. If you do not delete the
36	* provisions above, a recipient may use your version of this file under
37	* either the BSD or the GPL.
38	*/
39
40	/*
41	* general notes about linux io_uring:
42	*
43	* a) it's the best interface I have seen so far. on linux.
44	* b) best is not necessarily very good.
45	* c) it's better than the aio mess, doesn't suffer from the fork problems
46	* of linux aio or epoll and so on and so on. and you could do event stuff
47	* without any syscalls. what's not to like?
48	* d) ok, it's vastly more complex, but that's ok, really.
49	* e) why 3 mmaps instead of one? one would be more space-efficient,
50	* and I can't see what benefit three would have (other than being
51	* somehow resizable/relocatable, but that's apparently not possible).
52	* (FIXME: newer kernels can use 2 mmaps only, need to look into this).
53	* f) hmm, it's practiclaly undebuggable (gdb can't access the memory, and
54	* the bizarre way structure offsets are communicated makes it hard to
55	* just print the ring buffer heads, even iff the memory were visible
56	* in gdb. but then, that's also ok, really.
57	* g) well, you cannot specify a timeout when waiting for events. no,
58	* seriously, the interface doesn't support a timeout. never seen _that_
59	* before. sure, you can use a timerfd, but that's another syscall
60	* you could have avoided. overall, this bizarre omission smells
61	* like a µ-optimisation by the io_uring author for his personal
62	* applications, to the detriment of everybody else who just wants
63	* an event loop. but, umm, ok, if that's all, it could be worse.
64	* (FIXME: jens mentioned timeout commands, need to investigate)
65	* h) there is a hardcoded limit of 4096 outstanding events. okay,
66	* at least there is no arbitrary low system-wide limit...
67	* (FIXME: apparently, this was increased to 32768 in later kernels(
68	* i) unlike linux aio, you can register more then the limit
69	* of fd events, and the kernel will "gracefully" signal an
70	* overflow, after which you could destroy and recreate the kernel
71	* state, a bit bigger, or fall back to e.g. poll. thats not
72	* totally insane, but kind of questions the point a high
73	* performance I/O framework when it doesn't really work
74	* under stress.
75	* (FIXME: iouring should no longer drop events, need to investigate)
76	* j) but, oh my! is has exactly the same bugs as the linux aio backend,
77	* where some undocumented poll combinations just fail.
78	* so we need epoll AGAIN as a fallback. AGAIN! epoll!! and of course,
79	* this is completely undocumented, have I mantioned this already?
80	* k) overall, the API itself is, I dare to say, not a total trainwreck.
81	* the big isuess with it are the bugs requiring epoll, which might
82	* or might not get fixed (do I hold my breath?).
83	*/
84
85	#include <sys/timerfd.h>
86	#include <sys/mman.h>
87	#include <poll.h>
88
89	#define IOURING_INIT_ENTRIES 32
90
91	/*****************************************************************************/
92	/* syscall wrapdadoop - this section has the raw api/abi definitions */
93
94	#include <linux/fs.h>
95	#include <linux/types.h>
96
97	/* mostly directly taken from the kernel or documentation */
98
99	struct io_uring_sqe
100	{
101	__u8 opcode;
102	__u8 flags;
103	__u16 ioprio;
104	__s32 fd;
105	__u64 off;
106	__u64 addr;
107	__u32 len;
108	union {
109	__kernel_rwf_t rw_flags;
110	__u32 fsync_flags;
111	__u16 poll_events;
112	__u32 sync_range_flags;
113	__u32 msg_flags;
114	};
115	__u64 user_data;
116	union {
117	__u16 buf_index;
118	__u64 __pad2[3];
119	};
120	};
121
122	struct io_uring_cqe
123	{
124	__u64 user_data;
125	__s32 res;
126	__u32 flags;
127	};
128
129	struct io_sqring_offsets
130	{
131	__u32 head;
132	__u32 tail;
133	__u32 ring_mask;
134	__u32 ring_entries;
135	__u32 flags;
136	__u32 dropped;
137	__u32 array;
138	__u32 resv1;
139	__u64 resv2;
140	};
141
142	struct io_cqring_offsets
143	{
144	__u32 head;
145	__u32 tail;
146	__u32 ring_mask;
147	__u32 ring_entries;
148	__u32 overflow;
149	__u32 cqes;
150	__u64 resv[2];
151	};
152
153	struct io_uring_params
154	{
155	__u32 sq_entries;
156	__u32 cq_entries;
157	__u32 flags;
158	__u32 sq_thread_cpu;
159	__u32 sq_thread_idle;
160	__u32 resv[5];
161	struct io_sqring_offsets sq_off;
162	struct io_cqring_offsets cq_off;
163	};
164
165	#define IORING_OP_POLL_ADD 6
166	#define IORING_OP_POLL_REMOVE 7
167
168	#define IORING_ENTER_GETEVENTS 0x01
169
170	#define IORING_OFF_SQ_RING 0x00000000ULL
171	#define IORING_OFF_CQ_RING 0x08000000ULL
172	#define IORING_OFF_SQES 0x10000000ULL
173
174	inline_size
175	int
176	evsys_io_uring_setup (unsigned entries, struct io_uring_params *params)
177	{
178	return ev_syscall2 (SYS_io_uring_setup, entries, params);
179	}
180
181	inline_size
182	int
183	evsys_io_uring_enter (int fd, unsigned to_submit, unsigned min_complete, unsigned flags, const sigset_t *sig, size_t sigsz)
184	{
185	return ev_syscall6 (SYS_io_uring_enter, fd, to_submit, min_complete, flags, sig, sigsz);
186	}
187
188	/*****************************************************************************/
189	/* actual backed implementation */
190
191	/* we hope that volatile will make the compiler access this variables only once */
192	#define EV_SQ_VAR(name) (volatile unsigned )((char *)iouring_sq_ring + iouring_sq_ ## name)
193	#define EV_CQ_VAR(name) (volatile unsigned )((char *)iouring_cq_ring + iouring_cq_ ## name)
194
195	/* the index array */
196	#define EV_SQ_ARRAY ((unsigned )((char )iouring_sq_ring + iouring_sq_array))
197
198	/* the submit/completion queue entries */
199	#define EV_SQES ((struct io_uring_sqe *) iouring_sqes)
200	#define EV_CQES ((struct io_uring_cqe )((char )iouring_cq_ring + iouring_cq_cqes))
201
202	static
203	struct io_uring_sqe *
204	iouring_sqe_get (EV_P)
205	{
206	unsigned tail = EV_SQ_VAR (tail);
207
208	if (tail + 1 - EV_SQ_VAR (head) > EV_SQ_VAR (ring_entries))
209	{
210	/* queue full, flush */
211	evsys_io_uring_enter (iouring_fd, iouring_to_submit, 0, 0, 0, 0);
212	iouring_to_submit = 0;
213	}
214
215	assert (("libev: io_uring queue full after flush", tail + 1 - EV_SQ_VAR (head) <= EV_SQ_VAR (ring_entries)));
216
217	return EV_SQES + (tail & EV_SQ_VAR (ring_mask));
218	}
219
220	inline_size
221	struct io_uring_sqe *
222	iouring_sqe_submit (EV_P_ struct io_uring_sqe *sqe)
223	{
224	unsigned idx = sqe - EV_SQES;
225
226	EV_SQ_ARRAY [idx] = idx;
227	ECB_MEMORY_FENCE_RELEASE;
228	++EV_SQ_VAR (tail);
229	/ECB_MEMORY_FENCE_RELEASE; / for the time being we assume this is not needed */
230	++iouring_to_submit;
231	}
232
233	/*****************************************************************************/
234
235	/* when the timerfd expires we simply note the fact,
236	* as the purpose of the timerfd is to wake us up, nothing else.
237	* the next iteration should re-set it.
238	*/
239	static void
240	iouring_tfd_cb (EV_P_ struct ev_io *w, int revents)
241	{
242	iouring_tfd_to = EV_TSTAMP_HUGE;
243	}
244
245	static void
246	iouring_epoll_cb (EV_P_ struct ev_io *w, int revents)
247	{
248	epoll_poll (EV_A_ 0);
249	}
250
251	/* called for full and partial cleanup */
252	ecb_cold
253	static int
254	iouring_internal_destroy (EV_P)
255	{
256	close (iouring_tfd);
257	close (iouring_fd);
258
259	if (iouring_sq_ring != MAP_FAILED) munmap (iouring_sq_ring, iouring_sq_ring_size);
260	if (iouring_cq_ring != MAP_FAILED) munmap (iouring_cq_ring, iouring_cq_ring_size);
261	if (iouring_sqes != MAP_FAILED) munmap (iouring_sqes , iouring_sqes_size );
262
263	if (ev_is_active (&iouring_epoll_w)) ev_ref (EV_A); ev_io_stop (EV_A_ &iouring_epoll_w);
264	if (ev_is_active (&iouring_tfd_w )) ev_ref (EV_A); ev_io_stop (EV_A_ &iouring_tfd_w );
265	}
266
267	ecb_cold
268	static int
269	iouring_internal_init (EV_P)
270	{
271	struct io_uring_params params = { 0 };
272
273	iouring_to_submit = 0;
274
275	iouring_tfd = -1;
276	iouring_sq_ring = MAP_FAILED;
277	iouring_cq_ring = MAP_FAILED;
278	iouring_sqes = MAP_FAILED;
279
280	for (;;)
281	{
282	iouring_fd = evsys_io_uring_setup (iouring_entries, &params);
283
284	if (iouring_fd >= 0)
285	break; /* yippie */
286
287	if (errno != EINVAL)
288	return -1; /* we failed */
289
290	/* EINVAL: lots of possible reasons, but maybe
291	* it is because we hit the unqueryable hardcoded size limit
292	*/
293
294	/* we hit the limit already, give up */
295	if (iouring_max_entries)
296	return -1;
297
298	/* first time we hit EINVAL? assume we hit the limit, so go back and retry */
299	iouring_entries >>= 1;
300	iouring_max_entries = iouring_entries;
301	}
302
303	iouring_sq_ring_size = params.sq_off.array + params.sq_entries * sizeof (unsigned);
304	iouring_cq_ring_size = params.cq_off.cqes + params.cq_entries * sizeof (struct io_uring_cqe);
305	iouring_sqes_size = params.sq_entries * sizeof (struct io_uring_sqe);
306
307	iouring_sq_ring = mmap (0, iouring_sq_ring_size, PROT_READ \| PROT_WRITE,
308	MAP_SHARED \| MAP_POPULATE, iouring_fd, IORING_OFF_SQ_RING);
309	iouring_cq_ring = mmap (0, iouring_cq_ring_size, PROT_READ \| PROT_WRITE,
310	MAP_SHARED \| MAP_POPULATE, iouring_fd, IORING_OFF_CQ_RING);
311	iouring_sqes = mmap (0, iouring_sqes_size, PROT_READ \| PROT_WRITE,
312	MAP_SHARED \| MAP_POPULATE, iouring_fd, IORING_OFF_SQES);
313
314	if (iouring_sq_ring == MAP_FAILED \|\| iouring_cq_ring == MAP_FAILED \|\| iouring_sqes == MAP_FAILED)
315	return -1;
316
317	iouring_sq_head = params.sq_off.head;
318	iouring_sq_tail = params.sq_off.tail;
319	iouring_sq_ring_mask = params.sq_off.ring_mask;
320	iouring_sq_ring_entries = params.sq_off.ring_entries;
321	iouring_sq_flags = params.sq_off.flags;
322	iouring_sq_dropped = params.sq_off.dropped;
323	iouring_sq_array = params.sq_off.array;
324
325	iouring_cq_head = params.cq_off.head;
326	iouring_cq_tail = params.cq_off.tail;
327	iouring_cq_ring_mask = params.cq_off.ring_mask;
328	iouring_cq_ring_entries = params.cq_off.ring_entries;
329	iouring_cq_overflow = params.cq_off.overflow;
330	iouring_cq_cqes = params.cq_off.cqes;
331
332	iouring_tfd = timerfd_create (CLOCK_MONOTONIC, TFD_CLOEXEC);
333
334	if (iouring_tfd < 0)
335	return iouring_tfd;
336
337	iouring_tfd_to = EV_TSTAMP_HUGE;
338
339	return 0;
340	}
341
342	ecb_cold
343	static void
344	iouring_fork (EV_P)
345	{
346	iouring_internal_destroy (EV_A);
347
348	while (iouring_internal_init (EV_A) < 0)
349	ev_syserr ("(libev) io_uring_setup");
350
351	/* forking epoll should also effectively unregister all fds from the backend */
352	epoll_fork (EV_A);
353	/* epoll_fork already did this. hopefully */
354	/fd_rearm_all (EV_A);/
355
356	ev_io_stop (EV_A_ &iouring_epoll_w);
357	ev_io_set (EV_A_ &iouring_epoll_w, backend_fd, EV_READ);
358	ev_io_start (EV_A_ &iouring_epoll_w);
359
360	ev_io_stop (EV_A_ &iouring_tfd_w);
361	ev_io_set (EV_A_ &iouring_tfd_w, iouring_tfd, EV_READ);
362	ev_io_start (EV_A_ &iouring_tfd_w);
363	}
364
365	/*****************************************************************************/
366
367	static void
368	iouring_modify (EV_P_ int fd, int oev, int nev)
369	{
370	if (ecb_expect_false (anfds [fd].eflags))
371	{
372	/* we handed this fd over to epoll, so undo this first */
373	/* we do it manually because the optimisations on epoll_modify won't do us any good */
374	epoll_ctl (iouring_fd, EPOLL_CTL_DEL, fd, 0);
375	anfds [fd].eflags = 0;
376	oev = 0;
377	}
378
379	if (oev)
380	{
381	/* we assume the sqe's are all "properly" initialised */
382	struct io_uring_sqe *sqe = iouring_sqe_get (EV_A);
383	sqe->opcode = IORING_OP_POLL_REMOVE;
384	sqe->fd = fd;
385	sqe->user_data = -1;
386	iouring_sqe_submit (EV_A_ sqe);
387
388	/* increment generation counter to avoid handling old events */
389	++anfds [fd].egen;
390	}
391
392	if (nev)
393	{
394	struct io_uring_sqe *sqe = iouring_sqe_get (EV_A);
395	sqe->opcode = IORING_OP_POLL_ADD;
396	sqe->fd = fd;
397	sqe->user_data = (uint32_t)fd \| ((__u64)(uint32_t)anfds [fd].egen << 32);
398	sqe->poll_events =
399	(nev & EV_READ ? POLLIN : 0)
400	\| (nev & EV_WRITE ? POLLOUT : 0);
401	iouring_sqe_submit (EV_A_ sqe);
402	}
403	}
404
405	inline_size
406	void
407	iouring_tfd_update (EV_P_ ev_tstamp timeout)
408	{
409	ev_tstamp tfd_to = mn_now + timeout;
410
411	/* we assume there will be many iterations per timer change, so
412	* we only re-set the timerfd when we have to because its expiry
413	* is too late.
414	*/
415	if (ecb_expect_false (tfd_to < iouring_tfd_to))
416	{
417	struct itimerspec its;
418
419	iouring_tfd_to = tfd_to;
420	EV_TS_SET (its.it_interval, 0.);
421	EV_TS_SET (its.it_value, tfd_to);
422
423	if (timerfd_settime (iouring_tfd, TFD_TIMER_ABSTIME, &its, 0) < 0)
424	assert (("libev: iouring timerfd_settime failed", 0));
425	}
426	}
427
428	inline_size
429	void
430	iouring_process_cqe (EV_P_ struct io_uring_cqe *cqe)
431	{
432	int fd = cqe->user_data & 0xffffffffU;
433	uint32_t gen = cqe->user_data >> 32;
434	int res = cqe->res;
435
436	/* ignore fd removal events, if there are any. TODO: verify */
437	if (cqe->user_data == (__u64)-1)
438	abort ();//D
439
440	assert (("libev: io_uring fd must be in-bounds", fd >= 0 && fd < anfdmax));
441
442	/* documentation lies, of course. the result value is NOT like
443	* normal syscalls, but like linux raw syscalls, i.e. negative
444	* error numbers. fortunate, as otherwise there would be no way
445	* to get error codes at all. still, why not document this?
446	*/
447
448	/* ignore event if generation doesn't match */
449	/* this should actually be very rare */
450	if (ecb_expect_false (gen != (uint32_t)anfds [fd].egen))
451	return;
452
453	if (ecb_expect_false (res < 0))
454	{
455	if (res == -EINVAL)
456	{
457	/* we assume this error code means the fd/poll combination is buggy
458	* and fall back to epoll.
459	* this error code might also indicate a bug, but the kernel doesn't
460	* distinguish between those two conditions, so... sigh...
461	*/
462
463	epoll_modify (EV_A_ fd, 0, anfds [fd].events);
464	}
465	else if (res == -EBADF)
466	{
467	assert (("libev: event loop rejected bad fd", res != -EBADF));
468	fd_kill (EV_A_ fd);
469	}
470	else
471	{
472	errno = -res;
473	ev_syserr ("(libev) IORING_OP_POLL_ADD");
474	}
475
476	return;
477	}
478
479	/* feed events, we do not expect or handle POLLNVAL */
480	fd_event (
481	EV_A_
482	fd,
483	(res & (POLLOUT \| POLLERR \| POLLHUP) ? EV_WRITE : 0)
484	\| (res & (POLLIN \| POLLERR \| POLLHUP) ? EV_READ : 0)
485	);
486
487	/* io_uring is oneshot, so we need to re-arm the fd next iteration */
488	/* this also means we usually have to do at least one syscall per iteration */
489	anfds [fd].events = 0;
490	fd_change (EV_A_ fd, EV_ANFD_REIFY);
491	}
492
493	/* called when the event queue overflows */
494	ecb_cold
495	static void
496	iouring_overflow (EV_P)
497	{
498	/* we have two options, resize the queue (by tearing down
499	* everything and recreating it, or living with it
500	* and polling.
501	* we implement this by resizing tghe queue, and, if that fails,
502	* we just recreate the state on every failure, which
503	* kind of is a very inefficient poll.
504	* one danger is, due to the bios toward lower fds,
505	* we will only really get events for those, so
506	* maybe we need a poll() fallback, after all.
507	*/
508	/EV_CQ_VAR (overflow) = 0;/ /* need to do this if we keep the state and poll manually */
509
510	fd_rearm_all (EV_A);
511
512	/* we double the size until we hit the hard-to-probe maximum */
513	if (!iouring_max_entries)
514	{
515	iouring_entries <<= 1;
516	iouring_fork (EV_A);
517	}
518	else
519	{
520	/* we hit the kernel limit, we should fall back to something else.
521	* we can either poll() a few times and hope for the best,
522	* poll always, or switch to epoll.
523	* since we use epoll anyways, go epoll.
524	*/
525
526	iouring_internal_destroy (EV_A);
527
528	/* this should make it so that on return, we don'T call any uring functions */
529	iouring_to_submit = 0;
530
531	for (;;)
532	{
533	backend = epoll_init (EV_A_ 0);
534
535	if (backend)
536	break;
537
538	ev_syserr ("(libev) iouring switch to epoll");
539	}
540	}
541	}
542
543	/* handle any events in the completion queue, return true if there were any */
544	static int
545	iouring_handle_cq (EV_P)
546	{
547	unsigned head, tail, mask;
548
549	head = EV_CQ_VAR (head);
550	ECB_MEMORY_FENCE_ACQUIRE;
551	tail = EV_CQ_VAR (tail);
552
553	if (head == tail)
554	return 0;
555
556	/* it can only overflow if we have events, yes, yes? */
557	if (ecb_expect_false (EV_CQ_VAR (overflow)))
558	{
559	iouring_overflow (EV_A);
560	return 1;
561	}
562
563	mask = EV_CQ_VAR (ring_mask);
564
565	do
566	iouring_process_cqe (EV_A_ &EV_CQES [head++ & mask]);
567	while (head != tail);
568
569	EV_CQ_VAR (head) = head;
570	ECB_MEMORY_FENCE_RELEASE;
571
572	return 1;
573	}
574
575	static void
576	iouring_poll (EV_P_ ev_tstamp timeout)
577	{
578	/* if we have events, no need for extra syscalls, but we might have to queue events */
579	if (iouring_handle_cq (EV_A))
580	timeout = EV_TS_CONST (0.);
581	else
582	/* no events, so maybe wait for some */
583	iouring_tfd_update (EV_A_ timeout);
584
585	/* only enter the kernel if we have something to submit, or we need to wait */
586	if (timeout \|\| iouring_to_submit)
587	{
588	int res;
589
590	EV_RELEASE_CB;
591
592	res = evsys_io_uring_enter (iouring_fd, iouring_to_submit, 1,
593	timeout > EV_TS_CONST (0.) ? IORING_ENTER_GETEVENTS : 0, 0, 0);
594	iouring_to_submit = 0;
595
596	EV_ACQUIRE_CB;
597
598	if (ecb_expect_false (res < 0))
599	if (errno == EINTR)
600	/* ignore */;
601	else
602	ev_syserr ("(libev) iouring setup");
603	else
604	iouring_handle_cq (EV_A);
605	}
606	}
607
608	inline_size
609	int
610	iouring_init (EV_P_ int flags)
611	{
612	if (!epoll_init (EV_A_ 0))
613	return 0;
614
615	iouring_entries = IOURING_INIT_ENTRIES;
616	iouring_max_entries = 0;
617
618	if (iouring_internal_init (EV_A) < 0)
619	{
620	iouring_internal_destroy (EV_A);
621	return 0;
622	}
623
624	ev_io_init (&iouring_epoll_w, iouring_epoll_cb, backend_fd, EV_READ);
625	ev_set_priority (&iouring_epoll_w, EV_MAXPRI);
626
627	ev_io_init (&iouring_tfd_w, iouring_tfd_cb, iouring_tfd, EV_READ);
628	ev_set_priority (&iouring_tfd_w, EV_MAXPRI);
629
630	ev_io_start (EV_A_ &iouring_epoll_w);
631	ev_unref (EV_A); /* watcher should not keep loop alive */
632
633	ev_io_start (EV_A_ &iouring_tfd_w);
634	ev_unref (EV_A); /* watcher should not keep loop alive */
635
636	backend_modify = iouring_modify;
637	backend_poll = iouring_poll;
638
639	return EVBACKEND_IOURING;
640	}
641
642	inline_size
643	void
644	iouring_destroy (EV_P)
645	{
646	iouring_internal_destroy (EV_A);
647	epoll_destroy (EV_A);
648	}
649