[ViewVC] Diff of: cvs/libev/ev

Comparing libev/ev_iouring.c (file contents):
Revision 1.8 by root, Fri Dec 27 16:08:24 2019 UTC vs.
Revision 1.13 by root, Sat Dec 28 03:29:50 2019 UTC

  * 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?).
  */
+/* TODO: use internal TIMEOUT */
+/* TODO: take advantage of single mmap, NODROP etc. */
+/* TODO: resize cq/sq size independently */
 #include <sys/timerfd.h>
 #include <sys/mman.h>
 #include <poll.h>
+#include <stdint.h>
 #define IOURING_INIT_ENTRIES 32
 /*****************************************************************************/
 /* syscall wrapdadoop - this section has the raw api/abi definitions */
   __u32 resv[4];
   struct io_sqring_offsets sq_off;
   struct io_cqring_offsets cq_off;
 };
+#define IORING_SETUP_CQSIZE 0x00000008
-#define IORING_OP_POLL_ADD    6
+#define IORING_OP_POLL_ADD        6
-#define IORING_OP_POLL_REMOVE 7
+#define IORING_OP_POLL_REMOVE     7
+#define IORING_OP_TIMEOUT        11
+#define IORING_OP_TIMEOUT_REMOVE 12
+/* relative or absolute, reference clock is CLOCK_MONOTONIC */
+struct iouring_kernel_timespec
+{
+  int64_t tv_sec;
+  long long tv_nsec;
+};
+#define IORING_TIMEOUT_ABS 0x00000001
 #define IORING_ENTER_GETEVENTS 0x01
 #define IORING_OFF_SQ_RING 0x00000000ULL
 #define IORING_OFF_CQ_RING 0x08000000ULL
 #define IORING_OFF_SQES	   0x10000000ULL
-#define IORING_FEAT_SINGLE_MMAP   0x1
+#define IORING_FEAT_SINGLE_MMAP   0x00000001
-#define IORING_FEAT_NODROP        0x2
+#define IORING_FEAT_NODROP        0x00000002
-#define IORING_FEAT_SUBMIT_STABLE 0x4
+#define IORING_FEAT_SUBMIT_STABLE 0x00000004
 inline_size
 int
 evsys_io_uring_setup (unsigned entries, struct io_uring_params *params)
 {
   iouring_tfd     = -1;
   iouring_sq_ring = MAP_FAILED;
   iouring_cq_ring = MAP_FAILED;
   iouring_sqes    = MAP_FAILED;
+  if (!have_monotonic) /* cannot really happen, but what if11 */
+    return -1;
   for (;;)
     {
       iouring_fd = evsys_io_uring_setup (iouring_entries, &params);
       if (iouring_fd >= 0)
         break; /* yippie */
       if (errno != EINVAL)
         return -1; /* we failed */
+#if TODO
+      if ((~params.features) & (IORING_FEAT_NODROP | IORING_FEATURE_SINGLE_MMAP))
+        return -1; /* we require the above features */
+#endif
       /* EINVAL: lots of possible reasons, but maybe
        * it is because we hit the unqueryable hardcoded size limit
        */
     {
       /* 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;
+      /* Jens Axboe notified me that user_data is not what is documented, but is
+       * some kind of unique ID that has to match, otherwise the request cannot
+       * be removed. Since we don't *really* have that, we pass in the old
+       * generation counter - if that fails, too bad, it will hopefully be removed
+       * at close time and then be ignored. */
+      sqe->user_data = (uint32_t)fd | ((__u64)(uint32_t)anfds [fd].egen << 32);
       iouring_sqe_submit (EV_A_ sqe);
       /* increment generation counter to avoid handling old events */
       ++anfds [fd].egen;
     }
 {
   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 */
+  /* other than skipping removal events, */
   /* this should actually be very rare */
   if (ecb_expect_false (gen != (uint32_t)anfds [fd].egen))
     return;
   if (ecb_expect_false (res < 0))
     {
-      //TODO: EINVAL handling (was something failed with this fd)
+      /*TODO: EINVAL handling (was something failed with this fd)*/
-      //TODO: EBUSY happens when?
+      /*TODO: EBUSY happens when?*/
       if (res == -EBADF)
         {
           assert (("libev: event loop rejected bad fd", res != -EBADF));
           fd_kill (EV_A_ fd);
 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 implement this by resizing the 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.
   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.
+       * TODO: is this necessary with newer kernels?
        */
       iouring_internal_destroy (EV_A);
-      /* this should make it so that on return, we don'T call any uring functions */
+      /* 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);

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing libev/ev_iouring.c (file contents): Revision 1.8 by root, Fri Dec 27 16:08:24 2019 UTC vs. Revision 1.13 by root, Sat Dec 28 03:29:50 2019 UTC

Diff Legend

Comparing libev/ev_iouring.c (file contents):
Revision 1.8 by root, Fri Dec 27 16:08:24 2019 UTC vs.
Revision 1.13 by root, Sat Dec 28 03:29:50 2019 UTC