aio-posix: split poll check from ready handler

Adaptive polling measures the execution time of the polling check plus
handlers called when a polled event becomes ready. Handlers can take a
significant amount of time, making it look like polling was running for
a long time when in fact the event handler was running for a long time.

For example, on Linux the io_submit(2) syscall invoked when a virtio-blk
device's virtqueue becomes ready can take 10s of microseconds. This
can exceed the default polling interval (32 microseconds) and cause
adaptive polling to stop polling.

By excluding the handler's execution time from the polling check we make
the adaptive polling calculation more accurate. As a result, the event
loop now stays in polling mode where previously it would have fallen
back to file descriptor monitoring.

The following data was collected with virtio-blk num-queues=2
event_idx=off using an IOThread. Before:

168k IOPS, IOThread syscalls:

  9837.115 ( 0.020 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 16, iocbpp: 0x7fcb9f937db0)    = 16
  9837.158 ( 0.002 ms): IO iothread1/620155 write(fd: 103, buf: 0x556a2ef71b88, count: 8)                         = 8
  9837.161 ( 0.001 ms): IO iothread1/620155 write(fd: 104, buf: 0x556a2ef71b88, count: 8)                         = 8
  9837.163 ( 0.001 ms): IO iothread1/620155 ppoll(ufds: 0x7fcb90002800, nfds: 4, tsp: 0x7fcb9f1342d0, sigsetsize: 8) = 3
  9837.164 ( 0.001 ms): IO iothread1/620155 read(fd: 107, buf: 0x7fcb9f939cc0, count: 512)                        = 8
  9837.174 ( 0.001 ms): IO iothread1/620155 read(fd: 105, buf: 0x7fcb9f939cc0, count: 512)                        = 8
  9837.176 ( 0.001 ms): IO iothread1/620155 read(fd: 106, buf: 0x7fcb9f939cc0, count: 512)                        = 8
  9837.209 ( 0.035 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 32, iocbpp: 0x7fca7d0cebe0)    = 32

174k IOPS (+3.6%), IOThread syscalls:

  9809.566 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0cdd62be0)    = 32
  9809.625 ( 0.001 ms): IO iothread1/623061 write(fd: 103, buf: 0x5647cfba5f58, count: 8)                         = 8
  9809.627 ( 0.002 ms): IO iothread1/623061 write(fd: 104, buf: 0x5647cfba5f58, count: 8)                         = 8
  9809.663 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0d0388b50)    = 32

Notice that ppoll(2) and eventfd read(2) syscalls are eliminated because
the IOThread stays in polling mode instead of falling back to file
descriptor monitoring.

As usual, polling is not implemented on Windows so this patch ignores
the new io_poll_read() callback in aio-win32.c.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20211207132336.36627-2-stefanha@redhat.com

[Fixed up aio_set_event_notifier() calls in
tests/unit/test-fdmon-epoll.c added after this series was queued.
--Stefan]

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

1 files changed, 67 insertions, 22 deletions

diff --git a/util/aio-posix.c b/util/aio-posix.c
index 2b86777e91..7b9f629218 100644
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -23,6 +23,15 @@
 #include "trace.h"
 #include "aio-posix.h"
 
+/*
+ * G_IO_IN and G_IO_OUT are not appropriate revents values for polling, since
+ * the handler may not need to access the file descriptor. For example, the
+ * handler doesn't need to read from an EventNotifier if it polled a memory
+ * location and a read syscall would be slow. Define our own unique revents
+ * value to indicate that polling determined this AioHandler is ready.
+ */
+#define REVENTS_POLL_READY 0
+
 /* Stop userspace polling on a handler if it isn't active for some time */
 #define POLL_IDLE_INTERVAL_NS (7 * NANOSECONDS_PER_SECOND)
 
@@ -93,6 +102,7 @@ void aio_set_fd_handler(AioContext *ctx,
                         IOHandler *io_read,
                         IOHandler *io_write,
                         AioPollFn *io_poll,
+                        IOHandler *io_poll_ready,
                         void *opaque)
 {
     AioHandler *node;
@@ -101,6 +111,10 @@ void aio_set_fd_handler(AioContext *ctx,
     bool deleted = false;
     int poll_disable_change;
 
+    if (io_poll && !io_poll_ready) {
+        io_poll = NULL; /* polling only makes sense if there is a handler */
+    }
+
     qemu_lockcnt_lock(&ctx->list_lock);
 
     node = find_aio_handler(ctx, fd);
@@ -127,6 +141,7 @@ void aio_set_fd_handler(AioContext *ctx,
         new_node->io_read = io_read;
         new_node->io_write = io_write;
         new_node->io_poll = io_poll;
+        new_node->io_poll_ready = io_poll_ready;
         new_node->opaque = opaque;
         new_node->is_external = is_external;
 
@@ -182,10 +197,12 @@ void aio_set_event_notifier(AioContext *ctx,
                             EventNotifier *notifier,
                             bool is_external,
                             EventNotifierHandler *io_read,
-                            AioPollFn *io_poll)
+                            AioPollFn *io_poll,
+                            EventNotifierHandler *io_poll_ready)
 {
     aio_set_fd_handler(ctx, event_notifier_get_fd(notifier), is_external,
-                       (IOHandler *)io_read, NULL, io_poll, notifier);
+                       (IOHandler *)io_read, NULL, io_poll,
+                       (IOHandler *)io_poll_ready, notifier);
 }
 
 void aio_set_event_notifier_poll(AioContext *ctx,
@@ -198,7 +215,8 @@ void aio_set_event_notifier_poll(AioContext *ctx,
                     (IOHandler *)io_poll_end);
 }
 
-static bool poll_set_started(AioContext *ctx, bool started)
+static bool poll_set_started(AioContext *ctx, AioHandlerList *ready_list,
+                             bool started)
 {
     AioHandler *node;
     bool progress = false;
@@ -228,8 +246,9 @@ static bool poll_set_started(AioContext *ctx, bool started)
         }
 
         /* Poll one last time in case ->io_poll_end() raced with the event */
-        if (!started) {
-            progress = node->io_poll(node->opaque) || progress;
+        if (!started && node->io_poll(node->opaque)) {
+            aio_add_ready_handler(ready_list, node, REVENTS_POLL_READY);
+            progress = true;
         }
     }
     qemu_lockcnt_dec(&ctx->list_lock);
@@ -240,8 +259,11 @@ static bool poll_set_started(AioContext *ctx, bool started)
 
 bool aio_prepare(AioContext *ctx)
 {
+    AioHandlerList ready_list = QLIST_HEAD_INITIALIZER(ready_list);
+
     /* Poll mode cannot be used with glib's event loop, disable it. */
-    poll_set_started(ctx, false);
+    poll_set_started(ctx, &ready_list, false);
+    /* TODO what to do with this list? */
 
     return false;
 }
@@ -321,6 +343,18 @@ static bool aio_dispatch_handler(AioContext *ctx, AioHandler *node)
         }
         QLIST_INSERT_HEAD(&ctx->poll_aio_handlers, node, node_poll);
     }
+    if (!QLIST_IS_INSERTED(node, node_deleted) &&
+        revents == 0 &&
+        aio_node_check(ctx, node->is_external) &&
+        node->io_poll_ready) {
+        node->io_poll_ready(node->opaque);
+
+        /*
+         * Return early since revents was zero. aio_notify() does not count as
+         * progress.
+         */
+        return node->opaque != &ctx->notifier;
+    }
 
     if (!QLIST_IS_INSERTED(node, node_deleted) &&
         (revents & (G_IO_IN | G_IO_HUP | G_IO_ERR)) &&
@@ -387,6 +421,7 @@ void aio_dispatch(AioContext *ctx)
 }
 
 static bool run_poll_handlers_once(AioContext *ctx,
+                                   AioHandlerList *ready_list,
                                    int64_t now,
                                    int64_t *timeout)
 {
@@ -397,6 +432,8 @@ static bool run_poll_handlers_once(AioContext *ctx,
     QLIST_FOREACH_SAFE(node, &ctx->poll_aio_handlers, node_poll, tmp) {
         if (aio_node_check(ctx, node->is_external) &&
             node->io_poll(node->opaque)) {
+            aio_add_ready_handler(ready_list, node, REVENTS_POLL_READY);
+
             node->poll_idle_timeout = now + POLL_IDLE_INTERVAL_NS;
 
             /*
@@ -420,7 +457,9 @@ static bool fdmon_supports_polling(AioContext *ctx)
     return ctx->fdmon_ops->need_wait != aio_poll_disabled;
 }
 
-static bool remove_idle_poll_handlers(AioContext *ctx, int64_t now)
+static bool remove_idle_poll_handlers(AioContext *ctx,
+                                      AioHandlerList *ready_list,
+                                      int64_t now)
 {
     AioHandler *node;
     AioHandler *tmp;
@@ -451,7 +490,11 @@ static bool remove_idle_poll_handlers(AioContext *ctx, int64_t now)
                  * Nevermind about re-adding the handler in the rare case where
                  * this causes progress.
                  */
-                progress = node->io_poll(node->opaque) || progress;
+                if (node->io_poll(node->opaque)) {
+                    aio_add_ready_handler(ready_list, node,
+                                          REVENTS_POLL_READY);
+                    progress = true;
+                }
             }
         }
     }
@@ -461,6 +504,7 @@ static bool remove_idle_poll_handlers(AioContext *ctx, int64_t now)
 
 /* run_poll_handlers:
  * @ctx: the AioContext
+ * @ready_list: the list to place ready handlers on
  * @max_ns: maximum time to poll for, in nanoseconds
  *
  * Polls for a given time.
@@ -469,7 +513,8 @@ static bool remove_idle_poll_handlers(AioContext *ctx, int64_t now)
  *
  * Returns: true if progress was made, false otherwise
  */
-static bool run_poll_handlers(AioContext *ctx, int64_t max_ns, int64_t *timeout)
+static bool run_poll_handlers(AioContext *ctx, AioHandlerList *ready_list,
+                              int64_t max_ns, int64_t *timeout)
 {
     bool progress;
     int64_t start_time, elapsed_time;
@@ -490,13 +535,15 @@ static bool run_poll_handlers(AioContext *ctx, int64_t max_ns, int64_t *timeout)
 
     start_time = qemu_clock_get_ns(QEMU_CLOCK_REALTIME);
     do {
-        progress = run_poll_handlers_once(ctx, start_time, timeout);
+        progress = run_poll_handlers_once(ctx, ready_list,
+                                          start_time, timeout);
         elapsed_time = qemu_clock_get_ns(QEMU_CLOCK_REALTIME) - start_time;
         max_ns = qemu_soonest_timeout(*timeout, max_ns);
         assert(!(max_ns && progress));
     } while (elapsed_time < max_ns && !ctx->fdmon_ops->need_wait(ctx));
 
-    if (remove_idle_poll_handlers(ctx, start_time + elapsed_time)) {
+    if (remove_idle_poll_handlers(ctx, ready_list,
+                                  start_time + elapsed_time)) {
         *timeout = 0;
         progress = true;
     }
@@ -514,6 +561,7 @@ static bool run_poll_handlers(AioContext *ctx, int64_t max_ns, int64_t *timeout)
 
 /* try_poll_mode:
  * @ctx: the AioContext
+ * @ready_list: list to add handlers that need to be run
  * @timeout: timeout for blocking wait, computed by the caller and updated if
  *    polling succeeds.
  *
@@ -521,7 +569,8 @@ static bool run_poll_handlers(AioContext *ctx, int64_t max_ns, int64_t *timeout)
  *
  * Returns: true if progress was made, false otherwise
  */
-static bool try_poll_mode(AioContext *ctx, int64_t *timeout)
+static bool try_poll_mode(AioContext *ctx, AioHandlerList *ready_list,
+                          int64_t *timeout)
 {
     int64_t max_ns;
 
@@ -531,14 +580,14 @@ static bool try_poll_mode(AioContext *ctx, int64_t *timeout)
 
     max_ns = qemu_soonest_timeout(*timeout, ctx->poll_ns);
     if (max_ns && !ctx->fdmon_ops->need_wait(ctx)) {
-        poll_set_started(ctx, true);
+        poll_set_started(ctx, ready_list, true);
 
-        if (run_poll_handlers(ctx, max_ns, timeout)) {
+        if (run_poll_handlers(ctx, ready_list, max_ns, timeout)) {
             return true;
         }
     }
 
-    if (poll_set_started(ctx, false)) {
+    if (poll_set_started(ctx, ready_list, false)) {
         *timeout = 0;
         return true;
     }
@@ -549,7 +598,6 @@ static bool try_poll_mode(AioContext *ctx, int64_t *timeout)
 bool aio_poll(AioContext *ctx, bool blocking)
 {
     AioHandlerList ready_list = QLIST_HEAD_INITIALIZER(ready_list);
-    int ret = 0;
     bool progress;
     bool use_notify_me;
     int64_t timeout;
@@ -574,7 +622,7 @@ bool aio_poll(AioContext *ctx, bool blocking)
     }
 
     timeout = blocking ? aio_compute_timeout(ctx) : 0;
-    progress = try_poll_mode(ctx, &timeout);
+    progress = try_poll_mode(ctx, &ready_list, &timeout);
     assert(!(timeout && progress));
 
     /*
@@ -604,7 +652,7 @@ bool aio_poll(AioContext *ctx, bool blocking)
      * system call---a single round of run_poll_handlers_once suffices.
      */
     if (timeout || ctx->fdmon_ops->need_wait(ctx)) {
-        ret = ctx->fdmon_ops->wait(ctx, &ready_list, timeout);
+        ctx->fdmon_ops->wait(ctx, &ready_list, timeout);
     }
 
     if (use_notify_me) {
@@ -657,10 +705,7 @@ bool aio_poll(AioContext *ctx, bool blocking)
     }
 
     progress |= aio_bh_poll(ctx);
-
-    if (ret > 0) {
-        progress |= aio_dispatch_ready_handlers(ctx, &ready_list);
-    }
+    progress |= aio_dispatch_ready_handlers(ctx, &ready_list);
 
     aio_free_deleted_handlers(ctx);