[dpdk-dev,v2] net/virtio-user: fix multiple queues fail in server mode
Checks
Commit Message
This patch fixes multiple queues failure when virtio-user works in
server mode.
This patch adds feature negotiation in the processing of virtio-user
reccnnection.
Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
---
Changes in V2:
1. fix a comment typo.
2. add feature negotiation in the processing of reconnection.
drivers/net/virtio/virtio_user/vhost_user.c | 3 +++
drivers/net/virtio/virtio_user/virtio_user_dev.c | 14 ++++++++++----
drivers/net/virtio/virtio_user_ethdev.c | 20 ++++++++++++++++++++
3 files changed, 33 insertions(+), 4 deletions(-)
Comments
On Thu, May 10, 2018 at 05:36:23PM +0800, zhiyong.yang@intel.com wrote:
> This patch fixes multiple queues failure when virtio-user works in
> server mode.
>
> This patch adds feature negotiation in the processing of virtio-user
> reccnnection.
typo: reccnnection
>
> Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
>
> Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
> ---
>
> Changes in V2:
> 1. fix a comment typo.
> 2. add feature negotiation in the processing of reconnection.
>
> drivers/net/virtio/virtio_user/vhost_user.c | 3 +++
> drivers/net/virtio/virtio_user/virtio_user_dev.c | 14 ++++++++++----
> drivers/net/virtio/virtio_user_ethdev.c | 20 ++++++++++++++++++++
> 3 files changed, 33 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/net/virtio/virtio_user/vhost_user.c b/drivers/net/virtio/virtio_user/vhost_user.c
> index a6df97a..93e4d92 100644
> --- a/drivers/net/virtio/virtio_user/vhost_user.c
> +++ b/drivers/net/virtio/virtio_user/vhost_user.c
> @@ -263,6 +263,9 @@ struct hugepage_file_info {
>
> PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
>
> + if (dev->is_server && vhostfd < 0)
> + return -1;
> +
> msg.request = req;
> msg.flags = VHOST_USER_VERSION;
> msg.size = 0;
> diff --git a/drivers/net/virtio/virtio_user/virtio_user_dev.c b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> index 38b8bc9..b7e1915 100644
> --- a/drivers/net/virtio/virtio_user/virtio_user_dev.c
> +++ b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> @@ -447,10 +447,16 @@ int virtio_user_stop_device(struct virtio_user_dev *dev)
> return -1;
> }
>
> - for (i = 0; i < q_pairs; ++i)
> - ret |= dev->ops->enable_qp(dev, i, 1);
> - for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> - ret |= dev->ops->enable_qp(dev, i, 0);
> + /* Server mode can't enable queue pairs if vhostfd is invalid,
> + * always return 0 in this case.
> + */
> + if (dev->vhostfd >= 0) {
> + for (i = 0; i < q_pairs; ++i)
> + ret |= dev->ops->enable_qp(dev, i, 1);
> + for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> + ret |= dev->ops->enable_qp(dev, i, 0);
> + } else if (!dev->is_server)
> + ret = ~0;
You need to find a chance to enable these queue pairs
when the connection is established.
>
> dev->queue_pairs = q_pairs;
>
> diff --git a/drivers/net/virtio/virtio_user_ethdev.c b/drivers/net/virtio/virtio_user_ethdev.c
> index 4e7b3c3..91a0c44 100644
> --- a/drivers/net/virtio/virtio_user_ethdev.c
> +++ b/drivers/net/virtio/virtio_user_ethdev.c
> @@ -30,6 +30,7 @@
> int ret;
> int flag;
> int connectfd;
> + uint64_t features = dev->device_features, status;
> struct rte_eth_dev *eth_dev = &rte_eth_devices[dev->port_id];
>
> connectfd = accept(dev->listenfd, NULL, NULL);
> @@ -37,6 +38,25 @@
> return -1;
>
> dev->vhostfd = connectfd;
> + if (dev->ops->send_request(dev, VHOST_USER_GET_FEATURES,
> + &dev->device_features) < 0) {
> + PMD_INIT_LOG(ERR, "get_features failed: %s",
> + strerror(errno));
> + return -1;
> + }
> +
> + status = features & ~dev->device_features;
There is no need to introduce `status`.
> + /* For following bits, vhost-user doesn't really need to know */
> + status &= ~(1ull << VIRTIO_NET_F_MAC);
> + status &= ~(1ull << VIRTIO_NET_F_CTRL_VLAN);
> + status &= ~(1ull << VIRTIO_NET_F_CTRL_MAC_ADDR);
> + status &= ~(1ull << VIRTIO_NET_F_STATUS);
> + if (status)
> + PMD_INIT_LOG(ERR, "WARNING: Some features (0x%lx) are not supported by vhost-user!",
You will want to use PRIx64 here.
Thanks
> + status);
> +
> + dev->features &= dev->device_features;
> +
> flag = fcntl(connectfd, F_GETFD);
> fcntl(connectfd, F_SETFL, flag | O_NONBLOCK);
>
> --
> 1.8.3.1
>
Hi tiwei,
Thanks for your review firstly. Reply inline.
> -----Original Message-----
> From: Bie, Tiwei
> Sent: Thursday, May 10, 2018 6:23 PM
> To: Yang, Zhiyong <zhiyong.yang@intel.com>
> Cc: dev@dpdk.org; stable@dpdk.org; maxime.coquelin@redhat.com; Yigit,
> Ferruh <ferruh.yigit@intel.com>
> Subject: Re: [PATCH v2] net/virtio-user: fix multiple queues fail in server
> mode
>
> On Thu, May 10, 2018 at 05:36:23PM +0800, zhiyong.yang@intel.com wrote:
> > This patch fixes multiple queues failure when virtio-user works in
> > server mode.
> >
> > This patch adds feature negotiation in the processing of virtio-user
> > reccnnection.
>
> typo: reccnnection
>
Fix it.
> >
> > Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
> >
> > Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
> > ---
> >
> > Changes in V2:
> > 1. fix a comment typo.
> > 2. add feature negotiation in the processing of reconnection.
> >
> > drivers/net/virtio/virtio_user/vhost_user.c | 3 +++
> > drivers/net/virtio/virtio_user/virtio_user_dev.c | 14 ++++++++++----
> > drivers/net/virtio/virtio_user_ethdev.c | 20
> ++++++++++++++++++++
> > 3 files changed, 33 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/net/virtio/virtio_user/vhost_user.c
> > b/drivers/net/virtio/virtio_user/vhost_user.c
> > index a6df97a..93e4d92 100644
> > --- a/drivers/net/virtio/virtio_user/vhost_user.c
> > +++ b/drivers/net/virtio/virtio_user/vhost_user.c
> > @@ -263,6 +263,9 @@ struct hugepage_file_info {
> >
> > PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
> >
> > + if (dev->is_server && vhostfd < 0)
> > + return -1;
> > +
> > msg.request = req;
> > msg.flags = VHOST_USER_VERSION;
> > msg.size = 0;
> > diff --git a/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > index 38b8bc9..b7e1915 100644
> > --- a/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > +++ b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > @@ -447,10 +447,16 @@ int virtio_user_stop_device(struct
> virtio_user_dev *dev)
> > return -1;
> > }
> >
> > - for (i = 0; i < q_pairs; ++i)
> > - ret |= dev->ops->enable_qp(dev, i, 1);
> > - for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> > - ret |= dev->ops->enable_qp(dev, i, 0);
> > + /* Server mode can't enable queue pairs if vhostfd is invalid,
> > + * always return 0 in this case.
> > + */
> > + if (dev->vhostfd >= 0) {
> > + for (i = 0; i < q_pairs; ++i)
> > + ret |= dev->ops->enable_qp(dev, i, 1);
> > + for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> > + ret |= dev->ops->enable_qp(dev, i, 0);
> > + } else if (!dev->is_server)
> > + ret = ~0;
>
> You need to find a chance to enable these queue pairs when the connection
> is established.
>
Enable only one queue pairs in virtio_user_start_device.
I want to call virtio_user_handle_mq after virtio_user_start_device in virtio_user_server_reconnect. How about ?
> >
> > dev->queue_pairs = q_pairs;
> >
> > diff --git a/drivers/net/virtio/virtio_user_ethdev.c
> > b/drivers/net/virtio/virtio_user_ethdev.c
> > index 4e7b3c3..91a0c44 100644
> > --- a/drivers/net/virtio/virtio_user_ethdev.c
> > +++ b/drivers/net/virtio/virtio_user_ethdev.c
> > @@ -30,6 +30,7 @@
> > int ret;
> > int flag;
> > int connectfd;
> > + uint64_t features = dev->device_features, status;
> > struct rte_eth_dev *eth_dev = &rte_eth_devices[dev->port_id];
> >
> > connectfd = accept(dev->listenfd, NULL, NULL); @@ -37,6 +38,25
> @@
> > return -1;
> >
> > dev->vhostfd = connectfd;
> > + if (dev->ops->send_request(dev, VHOST_USER_GET_FEATURES,
> > + &dev->device_features) < 0) {
> > + PMD_INIT_LOG(ERR, "get_features failed: %s",
> > + strerror(errno));
> > + return -1;
> > + }
> > +
> > + status = features & ~dev->device_features;
>
> There is no need to introduce `status`.
>
Ok, remove it.
> > + /* For following bits, vhost-user doesn't really need to know */
> > + status &= ~(1ull << VIRTIO_NET_F_MAC);
> > + status &= ~(1ull << VIRTIO_NET_F_CTRL_VLAN);
> > + status &= ~(1ull << VIRTIO_NET_F_CTRL_MAC_ADDR);
> > + status &= ~(1ull << VIRTIO_NET_F_STATUS);
> > + if (status)
> > + PMD_INIT_LOG(ERR, "WARNING: Some features (0x%lx) are
> not supported
> > +by vhost-user!",
>
> You will want to use PRIx64 here.
>
Ok.
Thanks
Zhiyong
On Thu, May 10, 2018 at 10:01:55PM +0800, Yang, Zhiyong wrote:
> Hi tiwei,
>
> Thanks for your review firstly. Reply inline.
>
> > -----Original Message-----
> > From: Bie, Tiwei
> > Sent: Thursday, May 10, 2018 6:23 PM
> > To: Yang, Zhiyong <zhiyong.yang@intel.com>
> > Cc: dev@dpdk.org; stable@dpdk.org; maxime.coquelin@redhat.com; Yigit,
> > Ferruh <ferruh.yigit@intel.com>
> > Subject: Re: [PATCH v2] net/virtio-user: fix multiple queues fail in server
> > mode
> >
> > On Thu, May 10, 2018 at 05:36:23PM +0800, zhiyong.yang@intel.com wrote:
> > > This patch fixes multiple queues failure when virtio-user works in
> > > server mode.
> > >
> > > This patch adds feature negotiation in the processing of virtio-user
> > > reccnnection.
> >
> > typo: reccnnection
> >
>
> Fix it.
>
> > >
> > > Fixes: bd8f50a45d0f ("net/virtio-user: support server mode")
> > >
> > > Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
> > > ---
> > >
> > > Changes in V2:
> > > 1. fix a comment typo.
> > > 2. add feature negotiation in the processing of reconnection.
> > >
> > > drivers/net/virtio/virtio_user/vhost_user.c | 3 +++
> > > drivers/net/virtio/virtio_user/virtio_user_dev.c | 14 ++++++++++----
> > > drivers/net/virtio/virtio_user_ethdev.c | 20
> > ++++++++++++++++++++
> > > 3 files changed, 33 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/drivers/net/virtio/virtio_user/vhost_user.c
> > > b/drivers/net/virtio/virtio_user/vhost_user.c
> > > index a6df97a..93e4d92 100644
> > > --- a/drivers/net/virtio/virtio_user/vhost_user.c
> > > +++ b/drivers/net/virtio/virtio_user/vhost_user.c
> > > @@ -263,6 +263,9 @@ struct hugepage_file_info {
> > >
> > > PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
> > >
> > > + if (dev->is_server && vhostfd < 0)
> > > + return -1;
> > > +
> > > msg.request = req;
> > > msg.flags = VHOST_USER_VERSION;
> > > msg.size = 0;
> > > diff --git a/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > > b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > > index 38b8bc9..b7e1915 100644
> > > --- a/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > > +++ b/drivers/net/virtio/virtio_user/virtio_user_dev.c
> > > @@ -447,10 +447,16 @@ int virtio_user_stop_device(struct
> > virtio_user_dev *dev)
> > > return -1;
> > > }
> > >
> > > - for (i = 0; i < q_pairs; ++i)
> > > - ret |= dev->ops->enable_qp(dev, i, 1);
> > > - for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> > > - ret |= dev->ops->enable_qp(dev, i, 0);
> > > + /* Server mode can't enable queue pairs if vhostfd is invalid,
> > > + * always return 0 in this case.
> > > + */
> > > + if (dev->vhostfd >= 0) {
> > > + for (i = 0; i < q_pairs; ++i)
> > > + ret |= dev->ops->enable_qp(dev, i, 1);
> > > + for (i = q_pairs; i < dev->max_queue_pairs; ++i)
> > > + ret |= dev->ops->enable_qp(dev, i, 0);
> > > + } else if (!dev->is_server)
> > > + ret = ~0;
> >
> > You need to find a chance to enable these queue pairs when the connection
> > is established.
> >
> Enable only one queue pairs in virtio_user_start_device.
> I want to call virtio_user_handle_mq after virtio_user_start_device in virtio_user_server_reconnect. How about ?
Sound good.
Thanks
@@ -263,6 +263,9 @@ struct hugepage_file_info {
PMD_DRV_LOG(INFO, "%s", vhost_msg_strings[req]);
+ if (dev->is_server && vhostfd < 0)
+ return -1;
+
msg.request = req;
msg.flags = VHOST_USER_VERSION;
msg.size = 0;
@@ -447,10 +447,16 @@ int virtio_user_stop_device(struct virtio_user_dev *dev)
return -1;
}
- for (i = 0; i < q_pairs; ++i)
- ret |= dev->ops->enable_qp(dev, i, 1);
- for (i = q_pairs; i < dev->max_queue_pairs; ++i)
- ret |= dev->ops->enable_qp(dev, i, 0);
+ /* Server mode can't enable queue pairs if vhostfd is invalid,
+ * always return 0 in this case.
+ */
+ if (dev->vhostfd >= 0) {
+ for (i = 0; i < q_pairs; ++i)
+ ret |= dev->ops->enable_qp(dev, i, 1);
+ for (i = q_pairs; i < dev->max_queue_pairs; ++i)
+ ret |= dev->ops->enable_qp(dev, i, 0);
+ } else if (!dev->is_server)
+ ret = ~0;
dev->queue_pairs = q_pairs;
@@ -30,6 +30,7 @@
int ret;
int flag;
int connectfd;
+ uint64_t features = dev->device_features, status;
struct rte_eth_dev *eth_dev = &rte_eth_devices[dev->port_id];
connectfd = accept(dev->listenfd, NULL, NULL);
@@ -37,6 +38,25 @@
return -1;
dev->vhostfd = connectfd;
+ if (dev->ops->send_request(dev, VHOST_USER_GET_FEATURES,
+ &dev->device_features) < 0) {
+ PMD_INIT_LOG(ERR, "get_features failed: %s",
+ strerror(errno));
+ return -1;
+ }
+
+ status = features & ~dev->device_features;
+ /* For following bits, vhost-user doesn't really need to know */
+ status &= ~(1ull << VIRTIO_NET_F_MAC);
+ status &= ~(1ull << VIRTIO_NET_F_CTRL_VLAN);
+ status &= ~(1ull << VIRTIO_NET_F_CTRL_MAC_ADDR);
+ status &= ~(1ull << VIRTIO_NET_F_STATUS);
+ if (status)
+ PMD_INIT_LOG(ERR, "WARNING: Some features (0x%lx) are not supported by vhost-user!",
+ status);
+
+ dev->features &= dev->device_features;
+
flag = fcntl(connectfd, F_GETFD);
fcntl(connectfd, F_SETFL, flag | O_NONBLOCK);