aio-posix: Separate AioPolledEvent per AioHandler

Adaptive polling has a big problem: It doesn't consider that an event
loop can wait for many different events that may have very different
typical latencies.

For example, think of a guest that tends to send a new I/O request soon
after the previous I/O request completes, but the storage on the host is
rather slow. In this case, getting the new request from guest quickly
means that polling is enabled, but the next thing is performing the I/O
request on the backend, which is slow and disables polling again for the
next guest request. This means that in such a scenario, polling could
help for every other event, but is only ever enabled when it can't
succeed.

In order to fix this, keep a separate AioPolledEvent for each
AioHandler. We will then know that the backend file descriptor always
has a high latency and isn't worth polling for, but we also know that
the guest is always fast and we should poll for it. This solves at least
half of the problem, we can now keep polling for those cases where it
makes sense and get the improved performance from it.

Since the event loop doesn't know which event will be next, we still do
some unnecessary polling while we're waiting for the slow disk. I made
some attempts to be more clever than just randomly growing and shrinking
the polling time, and even to let callers be explicit about when they
expect a new event, but so far this hasn't resulted in improved
performance or even caused performance regressions. For now, let's just
fix the part that is easy enough to fix, we can revisit the rest later.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250307221634.71951-6-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

diff --git a/include/block/aio.h b/include/block/aio.h
index c9fcfe7ccfd327854d6c5c455a3ecfff4275396e..99ff48420b1f67fa41701d4fa8fe2eeaf97bb994 100644 (file)
--- a/include/block/aio.h
+++ b/include/block/aio.h
@@ -233,7 +233,6 @@ struct AioContext {
     int poll_disable_cnt;
 
     /* Polling mode parameters */
-    AioPolledEvent poll;
     int64_t poll_max_ns;    /* maximum polling time in nanoseconds */
     int64_t poll_grow;      /* polling time growth factor */
     int64_t poll_shrink;    /* polling time shrink factor */

diff --git a/util/aio-posix.c b/util/aio-posix.c
index 259827c7addc22f0233d4253a00599d55d3bb022..80785c29d2c09c41620aaf7ef3c6622795872edb 100644 (file)
--- a/util/aio-posix.c
+++ b/util/aio-posix.c
@@ -579,13 +579,19 @@ static bool run_poll_handlers(AioContext *ctx, AioHandlerList *ready_list,
 static bool try_poll_mode(AioContext *ctx, AioHandlerList *ready_list,
                           int64_t *timeout)
 {
+    AioHandler *node;
     int64_t max_ns;
 
     if (QLIST_EMPTY_RCU(&ctx->poll_aio_handlers)) {
         return false;
     }
 
-    max_ns = qemu_soonest_timeout(*timeout, ctx->poll.ns);
+    max_ns = 0;
+    QLIST_FOREACH(node, &ctx->poll_aio_handlers, node_poll) {
+        max_ns = MAX(max_ns, node->poll.ns);
+    }
+    max_ns = qemu_soonest_timeout(*timeout, max_ns);
+
     if (max_ns && !ctx->fdmon_ops->need_wait(ctx)) {
         /*
          * Enable poll mode. It pairs with the poll_set_started() in
@@ -721,8 +727,14 @@ bool aio_poll(AioContext *ctx, bool blocking)
 
     /* Adjust polling time */
     if (ctx->poll_max_ns) {
+        AioHandler *node;
         int64_t block_ns = qemu_clock_get_ns(QEMU_CLOCK_REALTIME) - start;
-        adjust_polling_time(ctx, &ctx->poll, block_ns);
+
+        QLIST_FOREACH(node, &ctx->poll_aio_handlers, node_poll) {
+            if (QLIST_IS_INSERTED(node, node_ready)) {
+                adjust_polling_time(ctx, &node->poll, block_ns);
+            }
+        }
     }
 
     progress |= aio_bh_poll(ctx);
@@ -772,11 +784,17 @@ void aio_context_use_g_source(AioContext *ctx)
 void aio_context_set_poll_params(AioContext *ctx, int64_t max_ns,
                                  int64_t grow, int64_t shrink, Error **errp)
 {
+    AioHandler *node;
+
+    qemu_lockcnt_inc(&ctx->list_lock);
+    QLIST_FOREACH(node, &ctx->aio_handlers, node) {
+        node->poll.ns = 0;
+    }
+    qemu_lockcnt_dec(&ctx->list_lock);
+
     /* No thread synchronization here, it doesn't matter if an incorrect value
      * is used once.
      */
-    ctx->poll.ns = 0;
-
     ctx->poll_max_ns = max_ns;
     ctx->poll_grow = grow;
     ctx->poll_shrink = shrink;

diff --git a/util/aio-posix.h b/util/aio-posix.h
index 4264c518be012e78d3615174f6848c76b4cd1302..82a0201ea48b1bc5a928f401595e36d750be2069 100644 (file)
--- a/util/aio-posix.h
+++ b/util/aio-posix.h
@@ -38,6 +38,7 @@ struct AioHandler {
 #endif
     int64_t poll_idle_timeout; /* when to stop userspace polling */
     bool poll_ready; /* has polling detected an event? */
+    AioPolledEvent poll;
 };
 
 /* Add a handler to a ready list */

diff --git a/util/async.c b/util/async.c
index fc8a78aa79d50a7a8b141698d149de77b10db842..863416dee929fc2415e1f80f24cf39566fa896f2 100644 (file)
--- a/util/async.c
+++ b/util/async.c
@@ -609,8 +609,6 @@ AioContext *aio_context_new(Error **errp)
     qemu_rec_mutex_init(&ctx->lock);
     timerlistgroup_init(&ctx->tlg, aio_timerlist_notify, ctx);
 
-    ctx->poll.ns = 0;
-
     ctx->poll_max_ns = 0;
     ctx->poll_grow = 0;
     ctx->poll_shrink = 0;