[PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

This patches improve the guest receive and transmit performance.
On the handle_tx side, we poll the sock receive queue at the same time.
handle_rx do that in the same way.

For more performance report, see patch 4.

v4 -> v5:
fix some issues

v3 -> v4:
fix some issues

v2 -> v3:
This patches are splited from previous big patch:
http://patchwork.ozlabs.org/patch/934673/

Tonghao Zhang (4):
  vhost: lock the vqs one by one
  net: vhost: replace magic number of lock annotation
  net: vhost: factor out busy polling logic to vhost_net_busy_poll()
  net: vhost: add rx busy polling in tx path

 drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
 drivers/vhost/vhost.c |  24 ++++-------
 2 files changed, 67 insertions(+), 65 deletions(-)

-- 
1.8.3.1

[PATCH net-next v5 1/4] net: vhost: lock the vqs one by one

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

This patch changes the way that lock all vqs
at the same, to lock them one by one. It will
be used for next patch to avoid the deadlock.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/vhost/vhost.c | 24 +++++++-----------------
 1 file changed, 7 insertions(+), 17 deletions(-)

diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 895eaa2..4ca9383 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -294,8 +294,11 @@ static void vhost_vq_meta_reset(struct vhost_dev *d)
 {
 	int i;
 
-	for (i = 0; i < d->nvqs; ++i)
+	for (i = 0; i < d->nvqs; ++i) {
+		mutex_lock(&d->vqs[i]->mutex);
 		__vhost_vq_meta_reset(d->vqs[i]);
+		mutex_unlock(&d->vqs[i]->mutex);
+	}
 }
 
 static void vhost_vq_reset(struct vhost_dev *dev,
@@ -887,20 +890,6 @@ static inline void __user *__vhost_get_user(struct vhost_virtqueue *vq,
 #define vhost_get_used(vq, x, ptr) \
 	vhost_get_user(vq, x, ptr, VHOST_ADDR_USED)
 
-static void vhost_dev_lock_vqs(struct vhost_dev *d)
-{
-	int i = 0;
-	for (i = 0; i < d->nvqs; ++i)
-		mutex_lock_nested(&d->vqs[i]->mutex, i);
-}
-
-static void vhost_dev_unlock_vqs(struct vhost_dev *d)
-{
-	int i = 0;
-	for (i = 0; i < d->nvqs; ++i)
-		mutex_unlock(&d->vqs[i]->mutex);
-}
-
 static int vhost_new_umem_range(struct vhost_umem *umem,
 				u64 start, u64 size, u64 end,
 				u64 userspace_addr, int perm)
@@ -950,7 +939,10 @@ static void vhost_iotlb_notify_vq(struct vhost_dev *d,
 		if (msg->iova <= vq_msg->iova &&
 		    msg->iova + msg->size - 1 > vq_msg->iova &&
 		    vq_msg->type == VHOST_IOTLB_MISS) {
+			mutex_lock(&node->vq->mutex);
 			vhost_poll_queue(&node->vq->poll);
+			mutex_unlock(&node->vq->mutex);
+
 			list_del(&node->node);
 			kfree(node);
 		}
@@ -982,7 +974,6 @@ static int vhost_process_iotlb_msg(struct vhost_dev *dev,
 	int ret = 0;
 
 	mutex_lock(&dev->mutex);
-	vhost_dev_lock_vqs(dev);
 	switch (msg->type) {
 	case VHOST_IOTLB_UPDATE:
 		if (!dev->iotlb) {
@@ -1016,7 +1007,6 @@ static int vhost_process_iotlb_msg(struct vhost_dev *dev,
 		break;
 	}
 
-	vhost_dev_unlock_vqs(dev);
 	mutex_unlock(&dev->mutex);
 
 	return ret;
-- 
1.8.3.1

[PATCH net-next v5 1/4] net: vhost: lock the vqs one by one

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

This patch changes the way that lock all vqs
at the same, to lock them one by one. It will
be used for next patch to avoid the deadlock.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Jason Wang <jasowang@redhat.com>
---
 drivers/vhost/vhost.c | 24 +++++++-----------------
 1 file changed, 7 insertions(+), 17 deletions(-)

diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 895eaa2..4ca9383 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -294,8 +294,11 @@ static void vhost_vq_meta_reset(struct vhost_dev *d)
 {
 	int i;
 
-	for (i = 0; i < d->nvqs; ++i)
+	for (i = 0; i < d->nvqs; ++i) {
+		mutex_lock(&d->vqs[i]->mutex);
 		__vhost_vq_meta_reset(d->vqs[i]);
+		mutex_unlock(&d->vqs[i]->mutex);
+	}
 }
 
 static void vhost_vq_reset(struct vhost_dev *dev,
@@ -887,20 +890,6 @@ static inline void __user *__vhost_get_user(struct vhost_virtqueue *vq,
 #define vhost_get_used(vq, x, ptr) \
 	vhost_get_user(vq, x, ptr, VHOST_ADDR_USED)
 
-static void vhost_dev_lock_vqs(struct vhost_dev *d)
-{
-	int i = 0;
-	for (i = 0; i < d->nvqs; ++i)
-		mutex_lock_nested(&d->vqs[i]->mutex, i);
-}
-
-static void vhost_dev_unlock_vqs(struct vhost_dev *d)
-{
-	int i = 0;
-	for (i = 0; i < d->nvqs; ++i)
-		mutex_unlock(&d->vqs[i]->mutex);
-}
-
 static int vhost_new_umem_range(struct vhost_umem *umem,
 				u64 start, u64 size, u64 end,
 				u64 userspace_addr, int perm)
@@ -950,7 +939,10 @@ static void vhost_iotlb_notify_vq(struct vhost_dev *d,
 		if (msg->iova <= vq_msg->iova &&
 		    msg->iova + msg->size - 1 > vq_msg->iova &&
 		    vq_msg->type == VHOST_IOTLB_MISS) {
+			mutex_lock(&node->vq->mutex);
 			vhost_poll_queue(&node->vq->poll);
+			mutex_unlock(&node->vq->mutex);
+
 			list_del(&node->node);
 			kfree(node);
 		}
@@ -982,7 +974,6 @@ static int vhost_process_iotlb_msg(struct vhost_dev *dev,
 	int ret = 0;
 
 	mutex_lock(&dev->mutex);
-	vhost_dev_lock_vqs(dev);
 	switch (msg->type) {
 	case VHOST_IOTLB_UPDATE:
 		if (!dev->iotlb) {
@@ -1016,7 +1007,6 @@ static int vhost_process_iotlb_msg(struct vhost_dev *dev,
 		break;
 	}
 
-	vhost_dev_unlock_vqs(dev);
 	mutex_unlock(&dev->mutex);
 
 	return ret;
-- 
1.8.3.1

[PATCH net-next v5 2/4] net: vhost: replace magic number of lock annotation

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

Use the VHOST_NET_VQ_XXX as a subclass for mutex_lock_nested.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
Acked-by: Jason Wang <jasowang@redhat.com>
---
 drivers/vhost/net.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index e7cf7d2..62bb8e8 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -484,7 +484,7 @@ static void handle_tx(struct vhost_net *net)
 	bool zcopy, zcopy_used;
 	int sent_pkts = 0;
 
-	mutex_lock(&vq->mutex);
+	mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
 	sock = vq->private_data;
 	if (!sock)
 		goto out;
@@ -655,7 +655,7 @@ static int vhost_net_rx_peek_head_len(struct vhost_net *net, struct sock *sk)
 		/* Flush batched heads first */
 		vhost_rx_signal_used(rvq);
 		/* Both tx vq and rx socket were polled here */
-		mutex_lock_nested(&vq->mutex, 1);
+		mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
 		vhost_disable_notify(&net->dev, vq);
 
 		preempt_disable();
@@ -789,7 +789,7 @@ static void handle_rx(struct vhost_net *net)
 	__virtio16 num_buffers;
 	int recv_pkts = 0;
 
-	mutex_lock_nested(&vq->mutex, 0);
+	mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_RX);
 	sock = vq->private_data;
 	if (!sock)
 		goto out;
-- 
1.8.3.1

[PATCH net-next v5 2/4] net: vhost: replace magic number of lock annotation

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

Use the VHOST_NET_VQ_XXX as a subclass for mutex_lock_nested.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
Acked-by: Jason Wang <jasowang@redhat.com>
---
 drivers/vhost/net.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index e7cf7d2..62bb8e8 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -484,7 +484,7 @@ static void handle_tx(struct vhost_net *net)
 	bool zcopy, zcopy_used;
 	int sent_pkts = 0;
 
-	mutex_lock(&vq->mutex);
+	mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
 	sock = vq->private_data;
 	if (!sock)
 		goto out;
@@ -655,7 +655,7 @@ static int vhost_net_rx_peek_head_len(struct vhost_net *net, struct sock *sk)
 		/* Flush batched heads first */
 		vhost_rx_signal_used(rvq);
 		/* Both tx vq and rx socket were polled here */
-		mutex_lock_nested(&vq->mutex, 1);
+		mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
 		vhost_disable_notify(&net->dev, vq);
 
 		preempt_disable();
@@ -789,7 +789,7 @@ static void handle_rx(struct vhost_net *net)
 	__virtio16 num_buffers;
 	int recv_pkts = 0;
 
-	mutex_lock_nested(&vq->mutex, 0);
+	mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_RX);
 	sock = vq->private_data;
 	if (!sock)
 		goto out;
-- 
1.8.3.1

[PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

Factor out generic busy polling logic and will be
used for in tx path in the next patch. And with the patch,
qemu can set differently the busyloop_timeout for rx queue.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
---
 drivers/vhost/net.c | 94 +++++++++++++++++++++++++++++++----------------------
 1 file changed, 55 insertions(+), 39 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 62bb8e8..2790959 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -429,6 +429,52 @@ static int vhost_net_enable_vq(struct vhost_net *n,
 	return vhost_poll_start(poll, sock->file);
 }
 
+static int sk_has_rx_data(struct sock *sk)
+{
+	struct socket *sock = sk->sk_socket;
+
+	if (sock->ops->peek_len)
+		return sock->ops->peek_len(sock);
+
+	return skb_queue_empty(&sk->sk_receive_queue);
+}
+
+static void vhost_net_busy_poll(struct vhost_net *net,
+				struct vhost_virtqueue *rvq,
+				struct vhost_virtqueue *tvq,
+				bool rx)
+{
+	unsigned long uninitialized_var(endtime);
+	unsigned long busyloop_timeout;
+	struct socket *sock;
+	struct vhost_virtqueue *vq = rx ? tvq : rvq;
+
+	mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
+
+	vhost_disable_notify(&net->dev, vq);
+	sock = rvq->private_data;
+	busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
+
+	preempt_disable();
+	endtime = busy_clock() + busyloop_timeout;
+	while (vhost_can_busy_poll(tvq->dev, endtime) &&
+	       !(sock && sk_has_rx_data(sock->sk)) &&
+	       vhost_vq_avail_empty(tvq->dev, tvq))
+		cpu_relax();
+	preempt_enable();
+
+	if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
+	    (!rx && (sock && sk_has_rx_data(sock->sk)))) {
+		vhost_poll_queue(&vq->poll);
+	} else if (vhost_enable_notify(&net->dev, vq) && rx) {
+		vhost_disable_notify(&net->dev, vq);
+		vhost_poll_queue(&vq->poll);
+	}
+
+	mutex_unlock(&vq->mutex);
+}
+
+
 static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
 				    struct vhost_virtqueue *vq,
 				    struct iovec iov[], unsigned int iov_size,
@@ -621,16 +667,6 @@ static int peek_head_len(struct vhost_net_virtqueue *rvq, struct sock *sk)
 	return len;
 }
 
-static int sk_has_rx_data(struct sock *sk)
-{
-	struct socket *sock = sk->sk_socket;
-
-	if (sock->ops->peek_len)
-		return sock->ops->peek_len(sock);
-
-	return skb_queue_empty(&sk->sk_receive_queue);
-}
-
 static void vhost_rx_signal_used(struct vhost_net_virtqueue *nvq)
 {
 	struct vhost_virtqueue *vq = &nvq->vq;
@@ -645,39 +681,19 @@ static void vhost_rx_signal_used(struct vhost_net_virtqueue *nvq)
 
 static int vhost_net_rx_peek_head_len(struct vhost_net *net, struct sock *sk)
 {
-	struct vhost_net_virtqueue *rvq = &net->vqs[VHOST_NET_VQ_RX];
-	struct vhost_net_virtqueue *nvq = &net->vqs[VHOST_NET_VQ_TX];
-	struct vhost_virtqueue *vq = &nvq->vq;
-	unsigned long uninitialized_var(endtime);
-	int len = peek_head_len(rvq, sk);
+	struct vhost_net_virtqueue *rnvq = &net->vqs[VHOST_NET_VQ_RX];
+	struct vhost_net_virtqueue *tnvq = &net->vqs[VHOST_NET_VQ_TX];
 
-	if (!len && vq->busyloop_timeout) {
-		/* Flush batched heads first */
-		vhost_rx_signal_used(rvq);
-		/* Both tx vq and rx socket were polled here */
-		mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
-		vhost_disable_notify(&net->dev, vq);
+	int len = peek_head_len(rnvq, sk);
 
-		preempt_disable();
-		endtime = busy_clock() + vq->busyloop_timeout;
-
-		while (vhost_can_busy_poll(&net->dev, endtime) &&
-		       !sk_has_rx_data(sk) &&
-		       vhost_vq_avail_empty(&net->dev, vq))
-			cpu_relax();
-
-		preempt_enable();
-
-		if (!vhost_vq_avail_empty(&net->dev, vq))
-			vhost_poll_queue(&vq->poll);
-		else if (unlikely(vhost_enable_notify(&net->dev, vq))) {
-			vhost_disable_notify(&net->dev, vq);
-			vhost_poll_queue(&vq->poll);
-		}
+	if (!len && rnvq->vq.busyloop_timeout) {
+		/* Flush batched heads first */
+		vhost_rx_signal_used(rnvq);
 
-		mutex_unlock(&vq->mutex);
+		/* Both tx vq and rx socket were polled here */
+		vhost_net_busy_poll(net, &rnvq->vq, &tnvq->vq, true);
 
-		len = peek_head_len(rvq, sk);
+		len = peek_head_len(rnvq, sk);
 	}
 
 	return len;
-- 
1.8.3.1

[PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

Factor out generic busy polling logic and will be
used for in tx path in the next patch. And with the patch,
qemu can set differently the busyloop_timeout for rx queue.

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
---
 drivers/vhost/net.c | 94 +++++++++++++++++++++++++++++++----------------------
 1 file changed, 55 insertions(+), 39 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 62bb8e8..2790959 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -429,6 +429,52 @@ static int vhost_net_enable_vq(struct vhost_net *n,
 	return vhost_poll_start(poll, sock->file);
 }
 
+static int sk_has_rx_data(struct sock *sk)
+{
+	struct socket *sock = sk->sk_socket;
+
+	if (sock->ops->peek_len)
+		return sock->ops->peek_len(sock);
+
+	return skb_queue_empty(&sk->sk_receive_queue);
+}
+
+static void vhost_net_busy_poll(struct vhost_net *net,
+				struct vhost_virtqueue *rvq,
+				struct vhost_virtqueue *tvq,
+				bool rx)
+{
+	unsigned long uninitialized_var(endtime);
+	unsigned long busyloop_timeout;
+	struct socket *sock;
+	struct vhost_virtqueue *vq = rx ? tvq : rvq;
+
+	mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
+
+	vhost_disable_notify(&net->dev, vq);
+	sock = rvq->private_data;
+	busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
+
+	preempt_disable();
+	endtime = busy_clock() + busyloop_timeout;
+	while (vhost_can_busy_poll(tvq->dev, endtime) &&
+	       !(sock && sk_has_rx_data(sock->sk)) &&
+	       vhost_vq_avail_empty(tvq->dev, tvq))
+		cpu_relax();
+	preempt_enable();
+
+	if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
+	    (!rx && (sock && sk_has_rx_data(sock->sk)))) {
+		vhost_poll_queue(&vq->poll);
+	} else if (vhost_enable_notify(&net->dev, vq) && rx) {
+		vhost_disable_notify(&net->dev, vq);
+		vhost_poll_queue(&vq->poll);
+	}
+
+	mutex_unlock(&vq->mutex);
+}
+
+
 static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
 				    struct vhost_virtqueue *vq,
 				    struct iovec iov[], unsigned int iov_size,
@@ -621,16 +667,6 @@ static int peek_head_len(struct vhost_net_virtqueue *rvq, struct sock *sk)
 	return len;
 }
 
-static int sk_has_rx_data(struct sock *sk)
-{
-	struct socket *sock = sk->sk_socket;
-
-	if (sock->ops->peek_len)
-		return sock->ops->peek_len(sock);
-
-	return skb_queue_empty(&sk->sk_receive_queue);
-}
-
 static void vhost_rx_signal_used(struct vhost_net_virtqueue *nvq)
 {
 	struct vhost_virtqueue *vq = &nvq->vq;
@@ -645,39 +681,19 @@ static void vhost_rx_signal_used(struct vhost_net_virtqueue *nvq)
 
 static int vhost_net_rx_peek_head_len(struct vhost_net *net, struct sock *sk)
 {
-	struct vhost_net_virtqueue *rvq = &net->vqs[VHOST_NET_VQ_RX];
-	struct vhost_net_virtqueue *nvq = &net->vqs[VHOST_NET_VQ_TX];
-	struct vhost_virtqueue *vq = &nvq->vq;
-	unsigned long uninitialized_var(endtime);
-	int len = peek_head_len(rvq, sk);
+	struct vhost_net_virtqueue *rnvq = &net->vqs[VHOST_NET_VQ_RX];
+	struct vhost_net_virtqueue *tnvq = &net->vqs[VHOST_NET_VQ_TX];
 
-	if (!len && vq->busyloop_timeout) {
-		/* Flush batched heads first */
-		vhost_rx_signal_used(rvq);
-		/* Both tx vq and rx socket were polled here */
-		mutex_lock_nested(&vq->mutex, VHOST_NET_VQ_TX);
-		vhost_disable_notify(&net->dev, vq);
+	int len = peek_head_len(rnvq, sk);
 
-		preempt_disable();
-		endtime = busy_clock() + vq->busyloop_timeout;
-
-		while (vhost_can_busy_poll(&net->dev, endtime) &&
-		       !sk_has_rx_data(sk) &&
-		       vhost_vq_avail_empty(&net->dev, vq))
-			cpu_relax();
-
-		preempt_enable();
-
-		if (!vhost_vq_avail_empty(&net->dev, vq))
-			vhost_poll_queue(&vq->poll);
-		else if (unlikely(vhost_enable_notify(&net->dev, vq))) {
-			vhost_disable_notify(&net->dev, vq);
-			vhost_poll_queue(&vq->poll);
-		}
+	if (!len && rnvq->vq.busyloop_timeout) {
+		/* Flush batched heads first */
+		vhost_rx_signal_used(rnvq);
 
-		mutex_unlock(&vq->mutex);
+		/* Both tx vq and rx socket were polled here */
+		vhost_net_busy_poll(net, &rnvq->vq, &tnvq->vq, true);
 
-		len = peek_head_len(rvq, sk);
+		len = peek_head_len(rnvq, sk);
 	}
 
 	return len;
-- 
1.8.3.1

[PATCH net-next v5 4/4] net: vhost: add rx busy polling in tx path

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

This patch improves the guest receive and transmit performance.
On the handle_tx side, we poll the sock receive queue at the
same time. handle_rx do that in the same way.

We set the poll-us=100us and use the iperf3 to test
its bandwidth, use the netperf to test throughput and mean
latency. When running the tests, the vhost-net kthread of
that VM, is alway 100% CPU. The commands are shown as below.

iperf3  -s -D
iperf3  -c IP -i 1 -P 1 -t 20 -M 1400

or
netserver
netperf -H IP -t TCP_RR -l 20 -- -O "THROUGHPUT,MEAN_LATENCY"

host -> guest:
iperf3:
* With the patch:     27.0 Gbits/sec
* Without the patch:  14.4 Gbits/sec

netperf (TCP_RR):
* With the patch:     48039.56 trans/s, 20.64us mean latency
* Without the patch:  46027.07 trans/s, 21.58us mean latency

This patch also improves the guest transmit performance.

guest -> host:
iperf3:
* With the patch:     27.2 Gbits/sec
* Without the patch:  24.4 Gbits/sec

netperf (TCP_RR):
* With the patch:     47963.25 trans/s, 20.71us mean latency
* Without the patch:  45796.70 trans/s, 21.68us mean latency

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
---
 drivers/vhost/net.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 2790959..3f26547 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -480,17 +480,13 @@ static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
 				    struct iovec iov[], unsigned int iov_size,
 				    unsigned int *out_num, unsigned int *in_num)
 {
-	unsigned long uninitialized_var(endtime);
+	struct vhost_net_virtqueue *rnvq = &net->vqs[VHOST_NET_VQ_RX];
 	int r = vhost_get_vq_desc(vq, vq->iov, ARRAY_SIZE(vq->iov),
 				  out_num, in_num, NULL, NULL);
 
 	if (r == vq->num && vq->busyloop_timeout) {
-		preempt_disable();
-		endtime = busy_clock() + vq->busyloop_timeout;
-		while (vhost_can_busy_poll(vq->dev, endtime) &&
-		       vhost_vq_avail_empty(vq->dev, vq))
-			cpu_relax();
-		preempt_enable();
+		vhost_net_busy_poll(net, &rnvq->vq, vq, false);
+
 		r = vhost_get_vq_desc(vq, vq->iov, ARRAY_SIZE(vq->iov),
 				      out_num, in_num, NULL, NULL);
 	}
-- 
1.8.3.1

[PATCH net-next v5 4/4] net: vhost: add rx busy polling in tx path

[xiangxia.m.yue]

From: Tonghao Zhang <xiangxia.m.yue@gmail.com>

This patch improves the guest receive and transmit performance.
On the handle_tx side, we poll the sock receive queue at the
same time. handle_rx do that in the same way.

We set the poll-us=100us and use the iperf3 to test
its bandwidth, use the netperf to test throughput and mean
latency. When running the tests, the vhost-net kthread of
that VM, is alway 100% CPU. The commands are shown as below.

iperf3  -s -D
iperf3  -c IP -i 1 -P 1 -t 20 -M 1400

or
netserver
netperf -H IP -t TCP_RR -l 20 -- -O "THROUGHPUT,MEAN_LATENCY"

host -> guest:
iperf3:
* With the patch:     27.0 Gbits/sec
* Without the patch:  14.4 Gbits/sec

netperf (TCP_RR):
* With the patch:     48039.56 trans/s, 20.64us mean latency
* Without the patch:  46027.07 trans/s, 21.58us mean latency

This patch also improves the guest transmit performance.

guest -> host:
iperf3:
* With the patch:     27.2 Gbits/sec
* Without the patch:  24.4 Gbits/sec

netperf (TCP_RR):
* With the patch:     47963.25 trans/s, 20.71us mean latency
* Without the patch:  45796.70 trans/s, 21.68us mean latency

Signed-off-by: Tonghao Zhang <zhangtonghao@didichuxing.com>
---
 drivers/vhost/net.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 2790959..3f26547 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -480,17 +480,13 @@ static int vhost_net_tx_get_vq_desc(struct vhost_net *net,
 				    struct iovec iov[], unsigned int iov_size,
 				    unsigned int *out_num, unsigned int *in_num)
 {
-	unsigned long uninitialized_var(endtime);
+	struct vhost_net_virtqueue *rnvq = &net->vqs[VHOST_NET_VQ_RX];
 	int r = vhost_get_vq_desc(vq, vq->iov, ARRAY_SIZE(vq->iov),
 				  out_num, in_num, NULL, NULL);
 
 	if (r == vq->num && vq->busyloop_timeout) {
-		preempt_disable();
-		endtime = busy_clock() + vq->busyloop_timeout;
-		while (vhost_can_busy_poll(vq->dev, endtime) &&
-		       vhost_vq_avail_empty(vq->dev, vq))
-			cpu_relax();
-		preempt_enable();
+		vhost_net_busy_poll(net, &rnvq->vq, vq, false);
+
 		r = vhost_get_vq_desc(vq, vq->iov, ARRAY_SIZE(vq->iov),
 				      out_num, in_num, NULL, NULL);
 	}
-- 
1.8.3.1

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>
> This patches improve the guest receive and transmit performance.
> On the handle_tx side, we poll the sock receive queue at the same time.
> handle_rx do that in the same way.
>
> For more performance report, see patch 4.
>
> v4 -> v5:
> fix some issues
>
> v3 -> v4:
> fix some issues
>
> v2 -> v3:
> This patches are splited from previous big patch:
> http://patchwork.ozlabs.org/patch/934673/
>
> Tonghao Zhang (4):
>    vhost: lock the vqs one by one
>    net: vhost: replace magic number of lock annotation
>    net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>    net: vhost: add rx busy polling in tx path
>
>   drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>   drivers/vhost/vhost.c |  24 ++++-------
>   2 files changed, 67 insertions(+), 65 deletions(-)
>

Acked-by: Jason Wang <jasowang@redhat.com>

Note: you probably need a rebase on top of Makita's recent patches. It 
depends on the order of being merged, let's see.

Thanks

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[Toshiaki Makita]

On 2018/07/04 13:31, xiangxia.m.yue@gmail.com wrote:
...
> +static void vhost_net_busy_poll(struct vhost_net *net,
> +				struct vhost_virtqueue *rvq,
> +				struct vhost_virtqueue *tvq,
> +				bool rx)
> +{
> +	unsigned long uninitialized_var(endtime);
> +	unsigned long busyloop_timeout;
> +	struct socket *sock;
> +	struct vhost_virtqueue *vq = rx ? tvq : rvq;
> +
> +	mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
> +
> +	vhost_disable_notify(&net->dev, vq);
> +	sock = rvq->private_data;
> +	busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
> +
> +	preempt_disable();
> +	endtime = busy_clock() + busyloop_timeout;
> +	while (vhost_can_busy_poll(tvq->dev, endtime) &&
> +	       !(sock && sk_has_rx_data(sock->sk)) &&
> +	       vhost_vq_avail_empty(tvq->dev, tvq))
> +		cpu_relax();
> +	preempt_enable();
> +
> +	if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
> +	    (!rx && (sock && sk_has_rx_data(sock->sk)))) {
> +		vhost_poll_queue(&vq->poll);
> +	} else if (vhost_enable_notify(&net->dev, vq) && rx) {

Hmm... on tx here sock has no rx data, so you are waiting for sock
wakeup for rx and vhost_enable_notify() seems not needed. Do you want
this actually?

} else if (rx && vhost_enable_notify(&net->dev, vq)) {

> +		vhost_disable_notify(&net->dev, vq);
> +		vhost_poll_queue(&vq->poll);
> +	}

-- 
Toshiaki Makita

Re: [PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[Jason Wang]

On 2018年07月04日 15:59, Toshiaki Makita wrote:
> On 2018/07/04 13:31, xiangxia.m.yue@gmail.com wrote:
> ...
>> +static void vhost_net_busy_poll(struct vhost_net *net,
>> +				struct vhost_virtqueue *rvq,
>> +				struct vhost_virtqueue *tvq,
>> +				bool rx)
>> +{
>> +	unsigned long uninitialized_var(endtime);
>> +	unsigned long busyloop_timeout;
>> +	struct socket *sock;
>> +	struct vhost_virtqueue *vq = rx ? tvq : rvq;
>> +
>> +	mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
>> +
>> +	vhost_disable_notify(&net->dev, vq);
>> +	sock = rvq->private_data;
>> +	busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
>> +
>> +	preempt_disable();
>> +	endtime = busy_clock() + busyloop_timeout;
>> +	while (vhost_can_busy_poll(tvq->dev, endtime) &&
>> +	       !(sock && sk_has_rx_data(sock->sk)) &&
>> +	       vhost_vq_avail_empty(tvq->dev, tvq))
>> +		cpu_relax();
>> +	preempt_enable();
>> +
>> +	if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
>> +	    (!rx && (sock && sk_has_rx_data(sock->sk)))) {
>> +		vhost_poll_queue(&vq->poll);
>> +	} else if (vhost_enable_notify(&net->dev, vq) && rx) {
> Hmm... on tx here sock has no rx data, so you are waiting for sock
> wakeup for rx and vhost_enable_notify() seems not needed. Do you want
> this actually?
>
> } else if (rx && vhost_enable_notify(&net->dev, vq)) {

Right, rx need to be checked first here.

Thanks

>> +		vhost_disable_notify(&net->dev, vq);
>> +		vhost_poll_queue(&vq->poll);
>> +	}

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[Tonghao Zhang]

On Wed, Jul 4, 2018 at 5:18 PM Jason Wang <jasowang@redhat.com> wrote:
>
>
>
> On 2018年07月04日 15:59, Toshiaki Makita wrote:
> > On 2018/07/04 13:31, xiangxia.m.yue@gmail.com wrote:
> > ...
> >> +static void vhost_net_busy_poll(struct vhost_net *net,
> >> +                            struct vhost_virtqueue *rvq,
> >> +                            struct vhost_virtqueue *tvq,
> >> +                            bool rx)
> >> +{
> >> +    unsigned long uninitialized_var(endtime);
> >> +    unsigned long busyloop_timeout;
> >> +    struct socket *sock;
> >> +    struct vhost_virtqueue *vq = rx ? tvq : rvq;
> >> +
> >> +    mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
> >> +
> >> +    vhost_disable_notify(&net->dev, vq);
> >> +    sock = rvq->private_data;
> >> +    busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
> >> +
> >> +    preempt_disable();
> >> +    endtime = busy_clock() + busyloop_timeout;
> >> +    while (vhost_can_busy_poll(tvq->dev, endtime) &&
> >> +           !(sock && sk_has_rx_data(sock->sk)) &&
> >> +           vhost_vq_avail_empty(tvq->dev, tvq))
> >> +            cpu_relax();
> >> +    preempt_enable();
> >> +
> >> +    if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
> >> +        (!rx && (sock && sk_has_rx_data(sock->sk)))) {
> >> +            vhost_poll_queue(&vq->poll);
> >> +    } else if (vhost_enable_notify(&net->dev, vq) && rx) {
> > Hmm... on tx here sock has no rx data, so you are waiting for sock
> > wakeup for rx and vhost_enable_notify() seems not needed. Do you want
> > this actually?
> >
> > } else if (rx && vhost_enable_notify(&net->dev, vq)) {
>
> Right, rx need to be checked first here.
thanks， if we dont call the vhost_enable_notify for tx. so we dont
need to call vhost_disable_notify for tx?

@@ -451,7 +451,9 @@ static void vhost_net_busy_poll(struct vhost_net *net,
                                              tvq->busyloop_timeout;

        mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
-       vhost_disable_notify(&net->dev, vq);
+
+       if (rx)
+               vhost_disable_notify(&net->dev, vq);

        preempt_disable();
        endtime = busy_clock() + busyloop_timeout;

> Thanks
>
> >> +            vhost_disable_notify(&net->dev, vq);
> >> +            vhost_poll_queue(&vq->poll);
> >> +    }
>
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[Jason Wang]

On 2018年07月04日 17:46, Tonghao Zhang wrote:
> On Wed, Jul 4, 2018 at 5:18 PM Jason Wang <jasowang@redhat.com> wrote:
>>
>>
>> On 2018年07月04日 15:59, Toshiaki Makita wrote:
>>> On 2018/07/04 13:31, xiangxia.m.yue@gmail.com wrote:
>>> ...
>>>> +static void vhost_net_busy_poll(struct vhost_net *net,
>>>> +                            struct vhost_virtqueue *rvq,
>>>> +                            struct vhost_virtqueue *tvq,
>>>> +                            bool rx)
>>>> +{
>>>> +    unsigned long uninitialized_var(endtime);
>>>> +    unsigned long busyloop_timeout;
>>>> +    struct socket *sock;
>>>> +    struct vhost_virtqueue *vq = rx ? tvq : rvq;
>>>> +
>>>> +    mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
>>>> +
>>>> +    vhost_disable_notify(&net->dev, vq);
>>>> +    sock = rvq->private_data;
>>>> +    busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
>>>> +
>>>> +    preempt_disable();
>>>> +    endtime = busy_clock() + busyloop_timeout;
>>>> +    while (vhost_can_busy_poll(tvq->dev, endtime) &&
>>>> +           !(sock && sk_has_rx_data(sock->sk)) &&
>>>> +           vhost_vq_avail_empty(tvq->dev, tvq))
>>>> +            cpu_relax();
>>>> +    preempt_enable();
>>>> +
>>>> +    if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
>>>> +        (!rx && (sock && sk_has_rx_data(sock->sk)))) {
>>>> +            vhost_poll_queue(&vq->poll);
>>>> +    } else if (vhost_enable_notify(&net->dev, vq) && rx) {
>>> Hmm... on tx here sock has no rx data, so you are waiting for sock
>>> wakeup for rx and vhost_enable_notify() seems not needed. Do you want
>>> this actually?
>>>
>>> } else if (rx && vhost_enable_notify(&net->dev, vq)) {
>> Right, rx need to be checked first here.
> thanks， if we dont call the vhost_enable_notify for tx. so we dont
> need to call vhost_disable_notify for tx?
>
> @@ -451,7 +451,9 @@ static void vhost_net_busy_poll(struct vhost_net *net,
>                                                tvq->busyloop_timeout;
>
>          mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
> -       vhost_disable_notify(&net->dev, vq);
> +
> +       if (rx)
> +               vhost_disable_notify(&net->dev, vq);
>
>          preempt_disable();
>          endtime = busy_clock() + busyloop_timeout;

Sorry for being unclear. We need enable tx notification at end end of 
the loop.

I meant we need replace

+    } else if (vhost_enable_notify(&net->dev, vq) && rx) {

with

+    } else if (rx && vhost_enable_notify(&net->dev, vq)) {

We only need rx notification when there's no avail buffers. This means 
we need only enable notification for tx.

And maybe we can simplify the logic as

if (rx) {
......
} else {
......
}

here. (not a must).

Thanks


>
>> Thanks
>>
>>>> +            vhost_disable_notify(&net->dev, vq);
>>>> +            vhost_poll_queue(&vq->poll);
>>>> +    }

Re: [PATCH net-next v5 3/4] net: vhost: factor out busy polling logic to vhost_net_busy_poll()

[Jason Wang]

On 2018年07月04日 17:46, Tonghao Zhang wrote:
> On Wed, Jul 4, 2018 at 5:18 PM Jason Wang <jasowang@redhat.com> wrote:
>>
>>
>> On 2018年07月04日 15:59, Toshiaki Makita wrote:
>>> On 2018/07/04 13:31, xiangxia.m.yue@gmail.com wrote:
>>> ...
>>>> +static void vhost_net_busy_poll(struct vhost_net *net,
>>>> +                            struct vhost_virtqueue *rvq,
>>>> +                            struct vhost_virtqueue *tvq,
>>>> +                            bool rx)
>>>> +{
>>>> +    unsigned long uninitialized_var(endtime);
>>>> +    unsigned long busyloop_timeout;
>>>> +    struct socket *sock;
>>>> +    struct vhost_virtqueue *vq = rx ? tvq : rvq;
>>>> +
>>>> +    mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
>>>> +
>>>> +    vhost_disable_notify(&net->dev, vq);
>>>> +    sock = rvq->private_data;
>>>> +    busyloop_timeout = rx ? rvq->busyloop_timeout : tvq->busyloop_timeout;
>>>> +
>>>> +    preempt_disable();
>>>> +    endtime = busy_clock() + busyloop_timeout;
>>>> +    while (vhost_can_busy_poll(tvq->dev, endtime) &&
>>>> +           !(sock && sk_has_rx_data(sock->sk)) &&
>>>> +           vhost_vq_avail_empty(tvq->dev, tvq))
>>>> +            cpu_relax();
>>>> +    preempt_enable();
>>>> +
>>>> +    if ((rx && !vhost_vq_avail_empty(&net->dev, vq)) ||
>>>> +        (!rx && (sock && sk_has_rx_data(sock->sk)))) {
>>>> +            vhost_poll_queue(&vq->poll);
>>>> +    } else if (vhost_enable_notify(&net->dev, vq) && rx) {
>>> Hmm... on tx here sock has no rx data, so you are waiting for sock
>>> wakeup for rx and vhost_enable_notify() seems not needed. Do you want
>>> this actually?
>>>
>>> } else if (rx && vhost_enable_notify(&net->dev, vq)) {
>> Right, rx need to be checked first here.
> thanks， if we dont call the vhost_enable_notify for tx. so we dont
> need to call vhost_disable_notify for tx?
>
> @@ -451,7 +451,9 @@ static void vhost_net_busy_poll(struct vhost_net *net,
>                                                tvq->busyloop_timeout;
>
>          mutex_lock_nested(&vq->mutex, rx ? VHOST_NET_VQ_TX: VHOST_NET_VQ_RX);
> -       vhost_disable_notify(&net->dev, vq);
> +
> +       if (rx)
> +               vhost_disable_notify(&net->dev, vq);
>
>          preempt_disable();
>          endtime = busy_clock() + busyloop_timeout;

Sorry for being unclear. We need enable tx notification at end end of 
the loop.

I meant we need replace

+    } else if (vhost_enable_notify(&net->dev, vq) && rx) {

with

+    } else if (rx && vhost_enable_notify(&net->dev, vq)) {

We only need rx notification when there's no avail buffers. This means 
we need only enable notification for tx.

And maybe we can simplify the logic as

if (rx) {
......
} else {
......
}

here. (not a must).

Thanks


>
>> Thanks
>>
>>>> +            vhost_disable_notify(&net->dev, vq);
>>>> +            vhost_poll_queue(&vq->poll);
>>>> +    }

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>
> This patches improve the guest receive and transmit performance.
> On the handle_tx side, we poll the sock receive queue at the same time.
> handle_rx do that in the same way.
>
> For more performance report, see patch 4.
>
> v4 -> v5:
> fix some issues
>
> v3 -> v4:
> fix some issues
>
> v2 -> v3:
> This patches are splited from previous big patch:
> http://patchwork.ozlabs.org/patch/934673/
>
> Tonghao Zhang (4):
>    vhost: lock the vqs one by one
>    net: vhost: replace magic number of lock annotation
>    net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>    net: vhost: add rx busy polling in tx path
>
>   drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>   drivers/vhost/vhost.c |  24 ++++-------
>   2 files changed, 67 insertions(+), 65 deletions(-)
>

Hi, any progress on the new version?

I plan to send a new series of packed virtqueue support of vhost. If you 
plan to send it soon, I can wait. Otherwise, I will send my series.

Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>
> This patches improve the guest receive and transmit performance.
> On the handle_tx side, we poll the sock receive queue at the same time.
> handle_rx do that in the same way.
>
> For more performance report, see patch 4.
>
> v4 -> v5:
> fix some issues
>
> v3 -> v4:
> fix some issues
>
> v2 -> v3:
> This patches are splited from previous big patch:
> http://patchwork.ozlabs.org/patch/934673/
>
> Tonghao Zhang (4):
>    vhost: lock the vqs one by one
>    net: vhost: replace magic number of lock annotation
>    net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>    net: vhost: add rx busy polling in tx path
>
>   drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>   drivers/vhost/vhost.c |  24 ++++-------
>   2 files changed, 67 insertions(+), 65 deletions(-)
>

Hi, any progress on the new version?

I plan to send a new series of packed virtqueue support of vhost. If you 
plan to send it soon, I can wait. Otherwise, I will send my series.

Thanks
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Tonghao Zhang]

On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>
>
>
> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> >
> > This patches improve the guest receive and transmit performance.
> > On the handle_tx side, we poll the sock receive queue at the same time.
> > handle_rx do that in the same way.
> >
> > For more performance report, see patch 4.
> >
> > v4 -> v5:
> > fix some issues
> >
> > v3 -> v4:
> > fix some issues
> >
> > v2 -> v3:
> > This patches are splited from previous big patch:
> > http://patchwork.ozlabs.org/patch/934673/
> >
> > Tonghao Zhang (4):
> >    vhost: lock the vqs one by one
> >    net: vhost: replace magic number of lock annotation
> >    net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> >    net: vhost: add rx busy polling in tx path
> >
> >   drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> >   drivers/vhost/vhost.c |  24 ++++-------
> >   2 files changed, 67 insertions(+), 65 deletions(-)
> >
>
> Hi, any progress on the new version?
>
> I plan to send a new series of packed virtqueue support of vhost. If you
> plan to send it soon, I can wait. Otherwise, I will send my series.
I rebase the codes. and find there is no improvement anymore, the
patches of  makita  may solve the problem. jason you may send your
patches, and I will do some research on busypoll.

> Thanks
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月11日 11:49, Tonghao Zhang wrote:
> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>
>>
>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>
>>> This patches improve the guest receive and transmit performance.
>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>> handle_rx do that in the same way.
>>>
>>> For more performance report, see patch 4.
>>>
>>> v4 -> v5:
>>> fix some issues
>>>
>>> v3 -> v4:
>>> fix some issues
>>>
>>> v2 -> v3:
>>> This patches are splited from previous big patch:
>>> http://patchwork.ozlabs.org/patch/934673/
>>>
>>> Tonghao Zhang (4):
>>>     vhost: lock the vqs one by one
>>>     net: vhost: replace magic number of lock annotation
>>>     net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>     net: vhost: add rx busy polling in tx path
>>>
>>>    drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>    drivers/vhost/vhost.c |  24 ++++-------
>>>    2 files changed, 67 insertions(+), 65 deletions(-)
>>>
>> Hi, any progress on the new version?
>>
>> I plan to send a new series of packed virtqueue support of vhost. If you
>> plan to send it soon, I can wait. Otherwise, I will send my series.
> I rebase the codes. and find there is no improvement anymore, the
> patches of  makita  may solve the problem. jason you may send your
> patches, and I will do some research on busypoll.

I see. Maybe you can try some bi-directional traffic.

Btw, lots of optimizations could be done for busy polling. E.g 
integrating with host NAPI busy polling or a 100% busy polling 
vhost_net. You're welcome to work or propose new ideas.

Thanks

>
>> Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月11日 11:49, Tonghao Zhang wrote:
> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>
>>
>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>
>>> This patches improve the guest receive and transmit performance.
>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>> handle_rx do that in the same way.
>>>
>>> For more performance report, see patch 4.
>>>
>>> v4 -> v5:
>>> fix some issues
>>>
>>> v3 -> v4:
>>> fix some issues
>>>
>>> v2 -> v3:
>>> This patches are splited from previous big patch:
>>> http://patchwork.ozlabs.org/patch/934673/
>>>
>>> Tonghao Zhang (4):
>>>     vhost: lock the vqs one by one
>>>     net: vhost: replace magic number of lock annotation
>>>     net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>     net: vhost: add rx busy polling in tx path
>>>
>>>    drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>    drivers/vhost/vhost.c |  24 ++++-------
>>>    2 files changed, 67 insertions(+), 65 deletions(-)
>>>
>> Hi, any progress on the new version?
>>
>> I plan to send a new series of packed virtqueue support of vhost. If you
>> plan to send it soon, I can wait. Otherwise, I will send my series.
> I rebase the codes. and find there is no improvement anymore, the
> patches of  makita  may solve the problem. jason you may send your
> patches, and I will do some research on busypoll.

I see. Maybe you can try some bi-directional traffic.

Btw, lots of optimizations could be done for busy polling. E.g 
integrating with host NAPI busy polling or a 100% busy polling 
vhost_net. You're welcome to work or propose new ideas.

Thanks

>
>> Thanks

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Michael S. Tsirkin]

On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
> 
> 
> On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > > 
> > > 
> > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > > 
> > > > This patches improve the guest receive and transmit performance.
> > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > handle_rx do that in the same way.
> > > > 
> > > > For more performance report, see patch 4.
> > > > 
> > > > v4 -> v5:
> > > > fix some issues
> > > > 
> > > > v3 -> v4:
> > > > fix some issues
> > > > 
> > > > v2 -> v3:
> > > > This patches are splited from previous big patch:
> > > > http://patchwork.ozlabs.org/patch/934673/
> > > > 
> > > > Tonghao Zhang (4):
> > > >     vhost: lock the vqs one by one
> > > >     net: vhost: replace magic number of lock annotation
> > > >     net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > >     net: vhost: add rx busy polling in tx path
> > > > 
> > > >    drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> > > >    drivers/vhost/vhost.c |  24 ++++-------
> > > >    2 files changed, 67 insertions(+), 65 deletions(-)
> > > > 
> > > Hi, any progress on the new version?
> > > 
> > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > I rebase the codes. and find there is no improvement anymore, the
> > patches of  makita  may solve the problem. jason you may send your
> > patches, and I will do some research on busypoll.
> 
> I see. Maybe you can try some bi-directional traffic.
> 
> Btw, lots of optimizations could be done for busy polling. E.g integrating
> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> to work or propose new ideas.
> 
> Thanks

It seems clear we do need adaptive polling.  The difficulty with NAPI
polling is it can't access guest memory easily. But maybe
get_user_pages on the polled memory+NAPI polling can work.

> > 
> > > Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Michael S. Tsirkin]

On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
> 
> 
> On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > > 
> > > 
> > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > > 
> > > > This patches improve the guest receive and transmit performance.
> > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > handle_rx do that in the same way.
> > > > 
> > > > For more performance report, see patch 4.
> > > > 
> > > > v4 -> v5:
> > > > fix some issues
> > > > 
> > > > v3 -> v4:
> > > > fix some issues
> > > > 
> > > > v2 -> v3:
> > > > This patches are splited from previous big patch:
> > > > http://patchwork.ozlabs.org/patch/934673/
> > > > 
> > > > Tonghao Zhang (4):
> > > >     vhost: lock the vqs one by one
> > > >     net: vhost: replace magic number of lock annotation
> > > >     net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > >     net: vhost: add rx busy polling in tx path
> > > > 
> > > >    drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> > > >    drivers/vhost/vhost.c |  24 ++++-------
> > > >    2 files changed, 67 insertions(+), 65 deletions(-)
> > > > 
> > > Hi, any progress on the new version?
> > > 
> > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > I rebase the codes. and find there is no improvement anymore, the
> > patches of  makita  may solve the problem. jason you may send your
> > patches, and I will do some research on busypoll.
> 
> I see. Maybe you can try some bi-directional traffic.
> 
> Btw, lots of optimizations could be done for busy polling. E.g integrating
> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> to work or propose new ideas.
> 
> Thanks

It seems clear we do need adaptive polling.  The difficulty with NAPI
polling is it can't access guest memory easily. But maybe
get_user_pages on the polled memory+NAPI polling can work.

> > 
> > > Thanks
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
> On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>>
>> On 2018年07月11日 11:49, Tonghao Zhang wrote:
>>> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>>>
>>>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>>>
>>>>> This patches improve the guest receive and transmit performance.
>>>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>>>> handle_rx do that in the same way.
>>>>>
>>>>> For more performance report, see patch 4.
>>>>>
>>>>> v4 -> v5:
>>>>> fix some issues
>>>>>
>>>>> v3 -> v4:
>>>>> fix some issues
>>>>>
>>>>> v2 -> v3:
>>>>> This patches are splited from previous big patch:
>>>>> http://patchwork.ozlabs.org/patch/934673/
>>>>>
>>>>> Tonghao Zhang (4):
>>>>>      vhost: lock the vqs one by one
>>>>>      net: vhost: replace magic number of lock annotation
>>>>>      net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>>>      net: vhost: add rx busy polling in tx path
>>>>>
>>>>>     drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>>>     drivers/vhost/vhost.c |  24 ++++-------
>>>>>     2 files changed, 67 insertions(+), 65 deletions(-)
>>>>>
>>>> Hi, any progress on the new version?
>>>>
>>>> I plan to send a new series of packed virtqueue support of vhost. If you
>>>> plan to send it soon, I can wait. Otherwise, I will send my series.
>>> I rebase the codes. and find there is no improvement anymore, the
>>> patches of  makita  may solve the problem. jason you may send your
>>> patches, and I will do some research on busypoll.
>> I see. Maybe you can try some bi-directional traffic.
>>
>> Btw, lots of optimizations could be done for busy polling. E.g integrating
>> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
>> to work or propose new ideas.
>>
>> Thanks
> It seems clear we do need adaptive polling.

Yes.

>   The difficulty with NAPI
> polling is it can't access guest memory easily. But maybe
> get_user_pages on the polled memory+NAPI polling can work.

You mean something like zerocopy? Looks like we can do busy polling 
without it. I mean something like 
https://patchwork.kernel.org/patch/8707511/.

Thanks

>
>>>> Thanks

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Michael S. Tsirkin]

On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
> 
> 
> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
> > On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
> > > 
> > > On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > > > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > > > > 
> > > > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > > > > 
> > > > > > This patches improve the guest receive and transmit performance.
> > > > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > > > handle_rx do that in the same way.
> > > > > > 
> > > > > > For more performance report, see patch 4.
> > > > > > 
> > > > > > v4 -> v5:
> > > > > > fix some issues
> > > > > > 
> > > > > > v3 -> v4:
> > > > > > fix some issues
> > > > > > 
> > > > > > v2 -> v3:
> > > > > > This patches are splited from previous big patch:
> > > > > > http://patchwork.ozlabs.org/patch/934673/
> > > > > > 
> > > > > > Tonghao Zhang (4):
> > > > > >      vhost: lock the vqs one by one
> > > > > >      net: vhost: replace magic number of lock annotation
> > > > > >      net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > > > >      net: vhost: add rx busy polling in tx path
> > > > > > 
> > > > > >     drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> > > > > >     drivers/vhost/vhost.c |  24 ++++-------
> > > > > >     2 files changed, 67 insertions(+), 65 deletions(-)
> > > > > > 
> > > > > Hi, any progress on the new version?
> > > > > 
> > > > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > > > I rebase the codes. and find there is no improvement anymore, the
> > > > patches of  makita  may solve the problem. jason you may send your
> > > > patches, and I will do some research on busypoll.
> > > I see. Maybe you can try some bi-directional traffic.
> > > 
> > > Btw, lots of optimizations could be done for busy polling. E.g integrating
> > > with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> > > to work or propose new ideas.
> > > 
> > > Thanks
> > It seems clear we do need adaptive polling.
> 
> Yes.
> 
> >   The difficulty with NAPI
> > polling is it can't access guest memory easily. But maybe
> > get_user_pages on the polled memory+NAPI polling can work.
> 
> You mean something like zerocopy? Looks like we can do busy polling without
> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
> 
> Thanks

How does this patch work? vhost_vq_avail_empty can sleep,
you are calling it within an rcu read side critical section.

That's not the only problem btw, another one is that the
CPU time spent polling isn't accounted with the VM.

> > 
> > > > > Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Michael S. Tsirkin]

On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
> 
> 
> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
> > On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
> > > 
> > > On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > > > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > > > > 
> > > > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > > > > 
> > > > > > This patches improve the guest receive and transmit performance.
> > > > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > > > handle_rx do that in the same way.
> > > > > > 
> > > > > > For more performance report, see patch 4.
> > > > > > 
> > > > > > v4 -> v5:
> > > > > > fix some issues
> > > > > > 
> > > > > > v3 -> v4:
> > > > > > fix some issues
> > > > > > 
> > > > > > v2 -> v3:
> > > > > > This patches are splited from previous big patch:
> > > > > > http://patchwork.ozlabs.org/patch/934673/
> > > > > > 
> > > > > > Tonghao Zhang (4):
> > > > > >      vhost: lock the vqs one by one
> > > > > >      net: vhost: replace magic number of lock annotation
> > > > > >      net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > > > >      net: vhost: add rx busy polling in tx path
> > > > > > 
> > > > > >     drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> > > > > >     drivers/vhost/vhost.c |  24 ++++-------
> > > > > >     2 files changed, 67 insertions(+), 65 deletions(-)
> > > > > > 
> > > > > Hi, any progress on the new version?
> > > > > 
> > > > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > > > I rebase the codes. and find there is no improvement anymore, the
> > > > patches of  makita  may solve the problem. jason you may send your
> > > > patches, and I will do some research on busypoll.
> > > I see. Maybe you can try some bi-directional traffic.
> > > 
> > > Btw, lots of optimizations could be done for busy polling. E.g integrating
> > > with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> > > to work or propose new ideas.
> > > 
> > > Thanks
> > It seems clear we do need adaptive polling.
> 
> Yes.
> 
> >   The difficulty with NAPI
> > polling is it can't access guest memory easily. But maybe
> > get_user_pages on the polled memory+NAPI polling can work.
> 
> You mean something like zerocopy? Looks like we can do busy polling without
> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
> 
> Thanks

How does this patch work? vhost_vq_avail_empty can sleep,
you are calling it within an rcu read side critical section.

That's not the only problem btw, another one is that the
CPU time spent polling isn't accounted with the VM.

> > 
> > > > > Thanks
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月12日 11:34, Michael S. Tsirkin wrote:
> On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
>>
>> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
>>> On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>>>> On 2018年07月11日 11:49, Tonghao Zhang wrote:
>>>>> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>>>>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>>>>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>>>>>
>>>>>>> This patches improve the guest receive and transmit performance.
>>>>>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>>>>>> handle_rx do that in the same way.
>>>>>>>
>>>>>>> For more performance report, see patch 4.
>>>>>>>
>>>>>>> v4 -> v5:
>>>>>>> fix some issues
>>>>>>>
>>>>>>> v3 -> v4:
>>>>>>> fix some issues
>>>>>>>
>>>>>>> v2 -> v3:
>>>>>>> This patches are splited from previous big patch:
>>>>>>> http://patchwork.ozlabs.org/patch/934673/
>>>>>>>
>>>>>>> Tonghao Zhang (4):
>>>>>>>       vhost: lock the vqs one by one
>>>>>>>       net: vhost: replace magic number of lock annotation
>>>>>>>       net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>>>>>       net: vhost: add rx busy polling in tx path
>>>>>>>
>>>>>>>      drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>>>>>      drivers/vhost/vhost.c |  24 ++++-------
>>>>>>>      2 files changed, 67 insertions(+), 65 deletions(-)
>>>>>>>
>>>>>> Hi, any progress on the new version?
>>>>>>
>>>>>> I plan to send a new series of packed virtqueue support of vhost. If you
>>>>>> plan to send it soon, I can wait. Otherwise, I will send my series.
>>>>> I rebase the codes. and find there is no improvement anymore, the
>>>>> patches of  makita  may solve the problem. jason you may send your
>>>>> patches, and I will do some research on busypoll.
>>>> I see. Maybe you can try some bi-directional traffic.
>>>>
>>>> Btw, lots of optimizations could be done for busy polling. E.g integrating
>>>> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
>>>> to work or propose new ideas.
>>>>
>>>> Thanks
>>> It seems clear we do need adaptive polling.
>> Yes.
>>
>>>    The difficulty with NAPI
>>> polling is it can't access guest memory easily. But maybe
>>> get_user_pages on the polled memory+NAPI polling can work.
>> You mean something like zerocopy? Looks like we can do busy polling without
>> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
>>
>> Thanks
> How does this patch work? vhost_vq_avail_empty can sleep,
> you are calling it within an rcu read side critical section.

Ok, I get your meaning. I have patches to access vring through 
get_user_pages + vmap() which should help here. (And it increase PPS 
about 10%-20%).

>
> That's not the only problem btw, another one is that the
> CPU time spent polling isn't accounted with the VM.


Yes, but it's not the 'issue' of this patch. And I believe cgroup can help?

Thanks

>
>>>>>> Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月12日 11:34, Michael S. Tsirkin wrote:
> On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
>>
>> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
>>> On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>>>> On 2018年07月11日 11:49, Tonghao Zhang wrote:
>>>>> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>>>>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>>>>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>>>>>
>>>>>>> This patches improve the guest receive and transmit performance.
>>>>>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>>>>>> handle_rx do that in the same way.
>>>>>>>
>>>>>>> For more performance report, see patch 4.
>>>>>>>
>>>>>>> v4 -> v5:
>>>>>>> fix some issues
>>>>>>>
>>>>>>> v3 -> v4:
>>>>>>> fix some issues
>>>>>>>
>>>>>>> v2 -> v3:
>>>>>>> This patches are splited from previous big patch:
>>>>>>> http://patchwork.ozlabs.org/patch/934673/
>>>>>>>
>>>>>>> Tonghao Zhang (4):
>>>>>>>       vhost: lock the vqs one by one
>>>>>>>       net: vhost: replace magic number of lock annotation
>>>>>>>       net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>>>>>       net: vhost: add rx busy polling in tx path
>>>>>>>
>>>>>>>      drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>>>>>      drivers/vhost/vhost.c |  24 ++++-------
>>>>>>>      2 files changed, 67 insertions(+), 65 deletions(-)
>>>>>>>
>>>>>> Hi, any progress on the new version?
>>>>>>
>>>>>> I plan to send a new series of packed virtqueue support of vhost. If you
>>>>>> plan to send it soon, I can wait. Otherwise, I will send my series.
>>>>> I rebase the codes. and find there is no improvement anymore, the
>>>>> patches of  makita  may solve the problem. jason you may send your
>>>>> patches, and I will do some research on busypoll.
>>>> I see. Maybe you can try some bi-directional traffic.
>>>>
>>>> Btw, lots of optimizations could be done for busy polling. E.g integrating
>>>> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
>>>> to work or propose new ideas.
>>>>
>>>> Thanks
>>> It seems clear we do need adaptive polling.
>> Yes.
>>
>>>    The difficulty with NAPI
>>> polling is it can't access guest memory easily. But maybe
>>> get_user_pages on the polled memory+NAPI polling can work.
>> You mean something like zerocopy? Looks like we can do busy polling without
>> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
>>
>> Thanks
> How does this patch work? vhost_vq_avail_empty can sleep,
> you are calling it within an rcu read side critical section.

Ok, I get your meaning. I have patches to access vring through 
get_user_pages + vmap() which should help here. (And it increase PPS 
about 10%-20%).

>
> That's not the only problem btw, another one is that the
> CPU time spent polling isn't accounted with the VM.


Yes, but it's not the 'issue' of this patch. And I believe cgroup can help?

Thanks

>
>>>>>> Thanks

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Michael S. Tsirkin]

On Thu, Jul 12, 2018 at 01:21:03PM +0800, Jason Wang wrote:
> 
> 
> On 2018年07月12日 11:34, Michael S. Tsirkin wrote:
> > On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
> > > 
> > > On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
> > > > On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
> > > > > On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > > > > > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > > > > > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > > > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > > > > > > 
> > > > > > > > This patches improve the guest receive and transmit performance.
> > > > > > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > > > > > handle_rx do that in the same way.
> > > > > > > > 
> > > > > > > > For more performance report, see patch 4.
> > > > > > > > 
> > > > > > > > v4 -> v5:
> > > > > > > > fix some issues
> > > > > > > > 
> > > > > > > > v3 -> v4:
> > > > > > > > fix some issues
> > > > > > > > 
> > > > > > > > v2 -> v3:
> > > > > > > > This patches are splited from previous big patch:
> > > > > > > > http://patchwork.ozlabs.org/patch/934673/
> > > > > > > > 
> > > > > > > > Tonghao Zhang (4):
> > > > > > > >       vhost: lock the vqs one by one
> > > > > > > >       net: vhost: replace magic number of lock annotation
> > > > > > > >       net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > > > > > >       net: vhost: add rx busy polling in tx path
> > > > > > > > 
> > > > > > > >      drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
> > > > > > > >      drivers/vhost/vhost.c |  24 ++++-------
> > > > > > > >      2 files changed, 67 insertions(+), 65 deletions(-)
> > > > > > > > 
> > > > > > > Hi, any progress on the new version?
> > > > > > > 
> > > > > > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > > > > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > > > > > I rebase the codes. and find there is no improvement anymore, the
> > > > > > patches of  makita  may solve the problem. jason you may send your
> > > > > > patches, and I will do some research on busypoll.
> > > > > I see. Maybe you can try some bi-directional traffic.
> > > > > 
> > > > > Btw, lots of optimizations could be done for busy polling. E.g integrating
> > > > > with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> > > > > to work or propose new ideas.
> > > > > 
> > > > > Thanks
> > > > It seems clear we do need adaptive polling.
> > > Yes.
> > > 
> > > >    The difficulty with NAPI
> > > > polling is it can't access guest memory easily. But maybe
> > > > get_user_pages on the polled memory+NAPI polling can work.
> > > You mean something like zerocopy? Looks like we can do busy polling without
> > > it. I mean something like https://patchwork.kernel.org/patch/8707511/.
> > > 
> > > Thanks
> > How does this patch work? vhost_vq_avail_empty can sleep,
> > you are calling it within an rcu read side critical section.
> 
> Ok, I get your meaning. I have patches to access vring through
> get_user_pages + vmap() which should help here. (And it increase PPS about
> 10%-20%).

Remember you must mark it as dirty on unpin too ...


> > 
> > That's not the only problem btw, another one is that the
> > CPU time spent polling isn't accounted with the VM.
> 
> 
> Yes, but it's not the 'issue' of this patch.

Yes it is. polling within thread context accounts CPU correctly.

> And I believe cgroup can help?
> 
> Thanks


cgroups are what's broken by polling in irq context.

> > 
> > > > > > > Thanks
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月12日 13:24, Michael S. Tsirkin wrote:
> On Thu, Jul 12, 2018 at 01:21:03PM +0800, Jason Wang wrote:
>>
>> On 2018年07月12日 11:34, Michael S. Tsirkin wrote:
>>> On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
>>>> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
>>>>> On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>>>>>> On 2018年07月11日 11:49, Tonghao Zhang wrote:
>>>>>>> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>>>>>>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>>>>>>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>>>>>>>
>>>>>>>>> This patches improve the guest receive and transmit performance.
>>>>>>>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>>>>>>>> handle_rx do that in the same way.
>>>>>>>>>
>>>>>>>>> For more performance report, see patch 4.
>>>>>>>>>
>>>>>>>>> v4 -> v5:
>>>>>>>>> fix some issues
>>>>>>>>>
>>>>>>>>> v3 -> v4:
>>>>>>>>> fix some issues
>>>>>>>>>
>>>>>>>>> v2 -> v3:
>>>>>>>>> This patches are splited from previous big patch:
>>>>>>>>> http://patchwork.ozlabs.org/patch/934673/
>>>>>>>>>
>>>>>>>>> Tonghao Zhang (4):
>>>>>>>>>        vhost: lock the vqs one by one
>>>>>>>>>        net: vhost: replace magic number of lock annotation
>>>>>>>>>        net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>>>>>>>        net: vhost: add rx busy polling in tx path
>>>>>>>>>
>>>>>>>>>       drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>>>>>>>       drivers/vhost/vhost.c |  24 ++++-------
>>>>>>>>>       2 files changed, 67 insertions(+), 65 deletions(-)
>>>>>>>>>
>>>>>>>> Hi, any progress on the new version?
>>>>>>>>
>>>>>>>> I plan to send a new series of packed virtqueue support of vhost. If you
>>>>>>>> plan to send it soon, I can wait. Otherwise, I will send my series.
>>>>>>> I rebase the codes. and find there is no improvement anymore, the
>>>>>>> patches of  makita  may solve the problem. jason you may send your
>>>>>>> patches, and I will do some research on busypoll.
>>>>>> I see. Maybe you can try some bi-directional traffic.
>>>>>>
>>>>>> Btw, lots of optimizations could be done for busy polling. E.g integrating
>>>>>> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
>>>>>> to work or propose new ideas.
>>>>>>
>>>>>> Thanks
>>>>> It seems clear we do need adaptive polling.
>>>> Yes.
>>>>
>>>>>     The difficulty with NAPI
>>>>> polling is it can't access guest memory easily. But maybe
>>>>> get_user_pages on the polled memory+NAPI polling can work.
>>>> You mean something like zerocopy? Looks like we can do busy polling without
>>>> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
>>>>
>>>> Thanks
>>> How does this patch work? vhost_vq_avail_empty can sleep,
>>> you are calling it within an rcu read side critical section.
>> Ok, I get your meaning. I have patches to access vring through
>> get_user_pages + vmap() which should help here. (And it increase PPS about
>> 10%-20%).
> Remember you must mark it as dirty on unpin too ...

Ok.

>
>
>>> That's not the only problem btw, another one is that the
>>> CPU time spent polling isn't accounted with the VM.
>>
>> Yes, but it's not the 'issue' of this patch.
> Yes it is. polling within thread context accounts CPU correctly.
>
>> And I believe cgroup can help?
>>
>> Thanks
>
> cgroups are what's broken by polling in irq context.

But I think the NAPI busy polling is still done in process context.

Thanks

>
>>>>>>>> Thanks

Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop

[Jason Wang]

On 2018年07月12日 13:24, Michael S. Tsirkin wrote:
> On Thu, Jul 12, 2018 at 01:21:03PM +0800, Jason Wang wrote:
>>
>> On 2018年07月12日 11:34, Michael S. Tsirkin wrote:
>>> On Thu, Jul 12, 2018 at 11:26:12AM +0800, Jason Wang wrote:
>>>> On 2018年07月11日 19:59, Michael S. Tsirkin wrote:
>>>>> On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>>>>>> On 2018年07月11日 11:49, Tonghao Zhang wrote:
>>>>>>> On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
>>>>>>>> On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
>>>>>>>>> From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
>>>>>>>>>
>>>>>>>>> This patches improve the guest receive and transmit performance.
>>>>>>>>> On the handle_tx side, we poll the sock receive queue at the same time.
>>>>>>>>> handle_rx do that in the same way.
>>>>>>>>>
>>>>>>>>> For more performance report, see patch 4.
>>>>>>>>>
>>>>>>>>> v4 -> v5:
>>>>>>>>> fix some issues
>>>>>>>>>
>>>>>>>>> v3 -> v4:
>>>>>>>>> fix some issues
>>>>>>>>>
>>>>>>>>> v2 -> v3:
>>>>>>>>> This patches are splited from previous big patch:
>>>>>>>>> http://patchwork.ozlabs.org/patch/934673/
>>>>>>>>>
>>>>>>>>> Tonghao Zhang (4):
>>>>>>>>>        vhost: lock the vqs one by one
>>>>>>>>>        net: vhost: replace magic number of lock annotation
>>>>>>>>>        net: vhost: factor out busy polling logic to vhost_net_busy_poll()
>>>>>>>>>        net: vhost: add rx busy polling in tx path
>>>>>>>>>
>>>>>>>>>       drivers/vhost/net.c   | 108 ++++++++++++++++++++++++++++----------------------
>>>>>>>>>       drivers/vhost/vhost.c |  24 ++++-------
>>>>>>>>>       2 files changed, 67 insertions(+), 65 deletions(-)
>>>>>>>>>
>>>>>>>> Hi, any progress on the new version?
>>>>>>>>
>>>>>>>> I plan to send a new series of packed virtqueue support of vhost. If you
>>>>>>>> plan to send it soon, I can wait. Otherwise, I will send my series.
>>>>>>> I rebase the codes. and find there is no improvement anymore, the
>>>>>>> patches of  makita  may solve the problem. jason you may send your
>>>>>>> patches, and I will do some research on busypoll.
>>>>>> I see. Maybe you can try some bi-directional traffic.
>>>>>>
>>>>>> Btw, lots of optimizations could be done for busy polling. E.g integrating
>>>>>> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
>>>>>> to work or propose new ideas.
>>>>>>
>>>>>> Thanks
>>>>> It seems clear we do need adaptive polling.
>>>> Yes.
>>>>
>>>>>     The difficulty with NAPI
>>>>> polling is it can't access guest memory easily. But maybe
>>>>> get_user_pages on the polled memory+NAPI polling can work.
>>>> You mean something like zerocopy? Looks like we can do busy polling without
>>>> it. I mean something like https://patchwork.kernel.org/patch/8707511/.
>>>>
>>>> Thanks
>>> How does this patch work? vhost_vq_avail_empty can sleep,
>>> you are calling it within an rcu read side critical section.
>> Ok, I get your meaning. I have patches to access vring through
>> get_user_pages + vmap() which should help here. (And it increase PPS about
>> 10%-20%).
> Remember you must mark it as dirty on unpin too ...

Ok.

>
>
>>> That's not the only problem btw, another one is that the
>>> CPU time spent polling isn't accounted with the VM.
>>
>> Yes, but it's not the 'issue' of this patch.
> Yes it is. polling within thread context accounts CPU correctly.
>
>> And I believe cgroup can help?
>>
>> Thanks
>
> cgroups are what's broken by polling in irq context.

But I think the NAPI busy polling is still done in process context.

Thanks

>
>>>>>>>> Thanks

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization