qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: qemu-devel@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>,
	Peter Maydell <peter.maydell@linaro.org>,
	qemu-block@nongnu.org, Max Reitz <mreitz@redhat.com>,
	Stefan Hajnoczi <stefanha@redhat.com>,
	Paolo Bonzini <pbonzini@redhat.com>, Fam Zheng <fam@euphon.net>
Subject: [PULL 10/24] virtio: don't enable notifications during polling
Date: Thu, 19 Dec 2019 08:27:43 -0500	[thread overview]
Message-ID: <20191219132621.16595-11-mst@redhat.com> (raw)
In-Reply-To: <20191219132621.16595-1-mst@redhat.com>

From: Stefan Hajnoczi <stefanha@redhat.com>

Virtqueue notifications are not necessary during polling, so we disable
them.  This allows the guest driver to avoid MMIO vmexits.
Unfortunately the virtio-blk and virtio-scsi handler functions re-enable
notifications, defeating this optimization.

Fix virtio-blk and virtio-scsi emulation so they leave notifications
disabled.  The key thing to remember for correctness is that polling
always checks one last time after ending its loop, therefore it's safe
to lose the race when re-enabling notifications at the end of polling.

There is a measurable performance improvement of 5-10% with the null-co
block driver.  Real-life storage configurations will see a smaller
improvement because the MMIO vmexit overhead contributes less to
latency.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20191209210957.65087-1-stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
 include/hw/virtio/virtio.h |  1 +
 hw/block/virtio-blk.c      |  9 +++++++--
 hw/scsi/virtio-scsi.c      |  9 +++++++--
 hw/virtio/virtio.c         | 12 ++++++------
 4 files changed, 21 insertions(+), 10 deletions(-)

diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index 777772475c..b69d517496 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -228,6 +228,7 @@ int virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id);
 
 void virtio_notify_config(VirtIODevice *vdev);
 
+bool virtio_queue_get_notification(VirtQueue *vq);
 void virtio_queue_set_notification(VirtQueue *vq, int enable);
 
 int virtio_queue_ready(VirtQueue *vq);
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index 4c357d2928..c4e55fb3de 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -764,13 +764,16 @@ bool virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
 {
     VirtIOBlockReq *req;
     MultiReqBuffer mrb = {};
+    bool suppress_notifications = virtio_queue_get_notification(vq);
     bool progress = false;
 
     aio_context_acquire(blk_get_aio_context(s->blk));
     blk_io_plug(s->blk);
 
     do {
-        virtio_queue_set_notification(vq, 0);
+        if (suppress_notifications) {
+            virtio_queue_set_notification(vq, 0);
+        }
 
         while ((req = virtio_blk_get_request(s, vq))) {
             progress = true;
@@ -781,7 +784,9 @@ bool virtio_blk_handle_vq(VirtIOBlock *s, VirtQueue *vq)
             }
         }
 
-        virtio_queue_set_notification(vq, 1);
+        if (suppress_notifications) {
+            virtio_queue_set_notification(vq, 1);
+        }
     } while (!virtio_queue_empty(vq));
 
     if (mrb.num_reqs) {
diff --git a/hw/scsi/virtio-scsi.c b/hw/scsi/virtio-scsi.c
index e8b2b64d09..f080545f48 100644
--- a/hw/scsi/virtio-scsi.c
+++ b/hw/scsi/virtio-scsi.c
@@ -597,12 +597,15 @@ bool virtio_scsi_handle_cmd_vq(VirtIOSCSI *s, VirtQueue *vq)
 {
     VirtIOSCSIReq *req, *next;
     int ret = 0;
+    bool suppress_notifications = virtio_queue_get_notification(vq);
     bool progress = false;
 
     QTAILQ_HEAD(, VirtIOSCSIReq) reqs = QTAILQ_HEAD_INITIALIZER(reqs);
 
     do {
-        virtio_queue_set_notification(vq, 0);
+        if (suppress_notifications) {
+            virtio_queue_set_notification(vq, 0);
+        }
 
         while ((req = virtio_scsi_pop_req(s, vq))) {
             progress = true;
@@ -622,7 +625,9 @@ bool virtio_scsi_handle_cmd_vq(VirtIOSCSI *s, VirtQueue *vq)
             }
         }
 
-        virtio_queue_set_notification(vq, 1);
+        if (suppress_notifications) {
+            virtio_queue_set_notification(vq, 1);
+        }
     } while (ret != -EINVAL && !virtio_queue_empty(vq));
 
     QTAILQ_FOREACH_SAFE(req, &reqs, next, next) {
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 7bc6a9455e..95d8ff8508 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -432,6 +432,11 @@ static void virtio_queue_packed_set_notification(VirtQueue *vq, int enable)
     }
 }
 
+bool virtio_queue_get_notification(VirtQueue *vq)
+{
+    return vq->notification;
+}
+
 void virtio_queue_set_notification(VirtQueue *vq, int enable)
 {
     vq->notification = enable;
@@ -3410,17 +3415,12 @@ static bool virtio_queue_host_notifier_aio_poll(void *opaque)
 {
     EventNotifier *n = opaque;
     VirtQueue *vq = container_of(n, VirtQueue, host_notifier);
-    bool progress;
 
     if (!vq->vring.desc || virtio_queue_empty(vq)) {
         return false;
     }
 
-    progress = virtio_queue_notify_aio_vq(vq);
-
-    /* In case the handler function re-enabled notifications */
-    virtio_queue_set_notification(vq, 0);
-    return progress;
+    return virtio_queue_notify_aio_vq(vq);
 }
 
 static void virtio_queue_host_notifier_aio_poll_end(EventNotifier *n)
-- 
MST



  parent reply	other threads:[~2019-12-19 13:35 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-12-19 13:26 [PULL 00/24] virtio, pci, pc: fixes, features Michael S. Tsirkin
2019-12-19 13:26 ` [PULL 01/24] virtio: add ability to delete vq through a pointer Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 02/24] virtio: make virtio_delete_queue idempotent Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 03/24] virtio-balloon: fix memory leak while attach virtio-balloon device Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 04/24] virtio-serial-bus: fix memory leak while attach virtio-serial-bus Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 05/24] virtio-input: convert to new virtio_delete_queue Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 06/24] intel_iommu: fix bug to read DMAR_RTADDR_REG Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 07/24] virtio: update queue size on guest write Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 08/24] virtio-pci: disable vring processing when bus-mastering is disabled Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 09/24] Implement backend program convention command for vhost-user-blk Michael S. Tsirkin
2019-12-19 13:27 ` Michael S. Tsirkin [this message]
2019-12-19 13:27 ` [PULL 11/24] numa: Extend CLI to provide initiator information for numa nodes Michael S. Tsirkin
2019-12-19 13:27 ` [PULL 12/24] numa: Extend CLI to provide memory latency and bandwidth information Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 13/24] numa: Extend CLI to provide memory side cache information Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 14/24] hmat acpi: Build Memory Proximity Domain Attributes Structure(s) Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 15/24] hmat acpi: Build System Locality Latency and Bandwidth Information Structure(s) Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 16/24] hmat acpi: Build Memory Side Cache " Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 17/24] tests/numa: Add case for QMP build HMAT Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 18/24] tests/bios-tables-test: add test cases for ACPI HMAT Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 19/24] ACPI: add expected files for HMAT tests (acpihmat) Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 20/24] virtio-mmio: Clear v2 transport state on soft reset Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 21/24] hw/pci/pci_host: Remove redundant PCI_DPRINTF() Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 22/24] hw/pci/pci_host: Let pci_data_[read/write] use unsigned 'size' argument Michael S. Tsirkin
2019-12-19 13:28 ` [PULL 23/24] vhost-user: add VHOST_USER_RESET_DEVICE to reset devices Michael S. Tsirkin
2019-12-19 13:29 ` [PULL 24/24] vhost-user-scsi: reset the device if supported Michael S. Tsirkin
2019-12-20 18:24 ` [PULL 00/24] virtio, pci, pc: fixes, features Peter Maydell
2019-12-22 12:59   ` Michael S. Tsirkin
2019-12-22 13:02   ` Michael S. Tsirkin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20191219132621.16595-11-mst@redhat.com \
    --to=mst@redhat.com \
    --cc=fam@euphon.net \
    --cc=kwolf@redhat.com \
    --cc=mreitz@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=stefanha@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).