Re: [PATCH v6 5/9] vhost:add support for configure interrupt

[Jason Wang]

在 2021/4/27 上午11:39, Cindy Lu 写道:
> Add configure notifier support in vhost and related driver
> When backend support VIRTIO_NET_F_STATUS,setup the configure
> interrupt function in vhost_dev_start and release the related
> resource when vhost_dev_stop
>
> Signed-off-by: Cindy Lu <lulu@redhat.com>
> ---
>   hw/net/vhost_net.c         |  9 +++++
>   hw/net/virtio-net.c        |  6 ++++
>   hw/virtio/vhost.c          | 70 ++++++++++++++++++++++++++++++++++++--
>   hw/virtio/virtio.c         | 22 ++++++++++++
>   include/hw/virtio/vhost.h  |  3 ++
>   include/hw/virtio/virtio.h |  4 +++
>   include/net/vhost_net.h    |  3 ++
>   7 files changed, 115 insertions(+), 2 deletions(-)
>
> diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c
> index 24d555e764..12e30dc25e 100644
> --- a/hw/net/vhost_net.c
> +++ b/hw/net/vhost_net.c
> @@ -426,6 +426,15 @@ void vhost_net_virtqueue_mask(VHostNetState *net, 
> VirtIODevice *dev,
>       vhost_virtqueue_mask(&net->dev, dev, idx, mask);
>   }
>   
> +bool vhost_net_config_pending(VHostNetState *net, int idx)
> +{
> +    return vhost_config_pending(&net->dev, idx);
> +}
> +void vhost_net_config_mask(VHostNetState *net, VirtIODevice *dev,
> +                              bool mask)
> +{
> +    vhost_config_mask(&net->dev, dev,  mask);
> +}
>   VHostNetState *get_vhost_net(NetClientState *nc)
>   {
>       VHostNetState *vhost_net = 0;
> diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
> index 78ccaa228c..43b912453a 100644
> --- a/hw/net/virtio-net.c
> +++ b/hw/net/virtio-net.c
> @@ -3063,6 +3063,9 @@ static bool 
> virtio_net_guest_notifier_pending(VirtIODevice *vdev, int idx)
>       if (idx != -1) {
>           return vhost_net_virtqueue_pending(get_vhost_net(nc->peer), idx);
>       }
> +    if (idx == -1) {
> +        return vhost_net_config_pending(get_vhost_net(nc->peer), idx);
> +   }


This looks wrong. Have you tested the case of multiqueue?

In the case of multiqeueu, there could be N 1:1 mappings between nc and 
vhost_dev. And what's more important, nc is not related to config 
interrupt but network queue pair.


>       return false;
>   }
>   
> @@ -3075,6 +3078,9 @@ static void virtio_net_guest_notifier_mask(VirtIODevice *vdev, int idx,
>       if (idx != -1) {
>           vhost_net_virtqueue_mask(get_vhost_net(nc->peer), vdev, idx, mask);
>        }
> +    if (idx == -1) {
> +        vhost_net_config_mask(get_vhost_net(nc->peer), vdev, mask);
> +     }
>   }
>   
>   static void virtio_net_set_config_size(VirtIONet *n, uint64_t host_features)
> diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
> index 614ccc2bcb..162a5dd90c 100644
> --- a/hw/virtio/vhost.c
> +++ b/hw/virtio/vhost.c
> @@ -21,6 +21,7 @@
>   #include "qemu/error-report.h"
>   #include "qemu/memfd.h"
>   #include "standard-headers/linux/vhost_types.h"
> +#include "standard-headers/linux/virtio_net.h"
>   #include "exec/address-spaces.h"
>   #include "hw/virtio/virtio-bus.h"
>   #include "hw/virtio/virtio-access.h"
> @@ -1313,6 +1314,10 @@ int vhost_dev_init(struct vhost_dev *hdev, void *opaque,
>               goto fail;
>           }
>       }
> +    r = event_notifier_init(&hdev->masked_config_notifier, 0);
> +    if (r < 0) {
> +        return r;
> +    }


Similarly, we don't need per hdev masked_config_notifier.


>   
>       if (busyloop_timeout) {
>           for (i = 0; i < hdev->nvqs; ++i) {
> @@ -1405,6 +1410,7 @@ void vhost_dev_cleanup(struct vhost_dev *hdev)
>       for (i = 0; i < hdev->nvqs; ++i) {
>           vhost_virtqueue_cleanup(hdev->vqs + i);
>       }
> +    event_notifier_cleanup(&hdev->masked_config_notifier);
>       if (hdev->mem) {
>           /* those are only safe after successful init */
>           memory_listener_unregister(&hdev->memory_listener);
> @@ -1498,6 +1504,16 @@ bool vhost_virtqueue_pending(struct vhost_dev *hdev, int 
> n)
>       return event_notifier_test_and_clear(&vq->masked_notifier);
>   }
>   
> +bool vhost_config_pending(struct vhost_dev *hdev, int n)
> +{
> +    assert(hdev->vhost_ops);
> +
> +    if ((hdev->started == false) ||
> +        (hdev->vhost_ops->vhost_set_config_call == NULL)) {
> +        return false;
> +    }
> +    return event_notifier_test_and_clear(&hdev->masked_config_notifier);
> +}
>   /* Mask/unmask events from this vq. */
>   void vhost_virtqueue_mask(struct vhost_dev *hdev, VirtIODevice *vdev, int n,
>                            bool mask)
> @@ -1522,6 +1538,30 @@ void vhost_virtqueue_mask(struct vhost_dev *hdev, 
> VirtIODevice *vdev, int n,
>           VHOST_OPS_DEBUG("vhost_set_vring_call failed");
>       }
>   }
> +void vhost_config_mask(struct vhost_dev *hdev, VirtIODevice *vdev,
> +                         bool mask)
> +{
> +   int fd;
> +   int r;
> +   EventNotifier *masked_config_notifier = &hdev->masked_config_notifier;
> +   EventNotifier *config_notifier = &vdev->config_notifier;
> +   assert(hdev->vhost_ops);
> +
> +   if ((hdev->started == false) ||
> +        (hdev->vhost_ops->vhost_set_config_call == NULL)) {
> +        return ;
> +    }
> +    if (mask) {
> +        assert(vdev->use_guest_notifier_mask);
> +        fd = event_notifier_get_fd(masked_config_notifier);
> +    } else {
> +        fd = event_notifier_get_fd(config_notifier);
> +    }
> +   r = hdev->vhost_ops->vhost_set_config_call(hdev, &fd);
> +   if (r < 0) {
> +        error_report("vhost_set_config_call failed");
> +    }
> +}
>   
>   uint64_t vhost_get_features(struct vhost_dev *hdev, const int *feature_bits,
>                               uint64_t features)
> @@ -1701,6 +1741,7 @@ int vhost_dev_get_inflight(struct vhost_dev *dev, 
> uint16_t queue_size,
>   int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice *vdev)
>   {
>       int i, r;
> +    int fd = 0;
>   
>       /* should only be called after backend is connected */
>       assert(hdev->vhost_ops);
> @@ -1732,7 +1773,10 @@ int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice 
> *vdev)
>               goto fail_vq;
>           }
>       }
> -
> +    event_notifier_test_and_clear(&hdev->masked_config_notifier);
> +    if (!vdev->use_guest_notifier_mask) {
> +        vhost_config_mask(hdev, vdev,  true);
> +    }
>       if (hdev->log_enabled) {
>           uint64_t log_base;
>   
> @@ -1749,6 +1793,7 @@ int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice *vdev)
>               goto fail_log;
>           }
>       }
> +
>       if (hdev->vhost_ops->vhost_dev_start) {
>           r = hdev->vhost_ops->vhost_dev_start(hdev, true);
>           if (r) {
> @@ -1766,6 +1811,19 @@ int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice 
> *vdev)
>               vhost_device_iotlb_miss(hdev, vq->used_phys, true);
>           }
>       }
> +   if (!(hdev->features & (0x1ULL << VIRTIO_NET_F_STATUS))) {
> +        return 0;
> +    }
> +    if (hdev->vhost_ops->vhost_set_config_call) {
> +        fd = event_notifier_get_fd(&vdev->config_notifier);
> +        r = hdev->vhost_ops->vhost_set_config_call(hdev, &fd);
> +        if (!r) {
> +            event_notifier_set(&vdev->config_notifier);
> +        }
> +        if (r) {
> +            goto fail_log;
> +         }
> +    }
>       return 0;
>   fail_log:
>       vhost_log_put(hdev, false);
> @@ -1788,10 +1846,18 @@ fail_features:
>   void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev)
>   {
>       int i;
> +    int fd;
>   
>       /* should only be called after backend is connected */
>       assert(hdev->vhost_ops);
> -
> +    event_notifier_test_and_clear(&hdev->masked_config_notifier);
> +    event_notifier_test_and_clear(&vdev->config_notifier);
> +    if ((hdev->features & (0x1ULL << VIRTIO_NET_F_STATUS))) {


Any reason for such check. Let's try not check per device feature in the 
generic vhost core.

Note that the config interrupt is a basic facility which could be used 
by various other devices (e.g block).


> +        if (hdev->vhost_ops->vhost_set_config_call) {
> +            fd = -1;
> +            hdev->vhost_ops->vhost_set_config_call(hdev, &fd);
> +        }
> +    }
>       if (hdev->vhost_ops->vhost_dev_start) {
>           hdev->vhost_ops->vhost_dev_start(hdev, false);
>       }
> diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
> index ceb58fda6c..5dff29c981 100644
> --- a/hw/virtio/virtio.c
> +++ b/hw/virtio/virtio.c
> @@ -3502,6 +3502,14 @@ static void 
> virtio_queue_guest_notifier_read(EventNotifier *n)
>       }
>   }
>   
> +static void virtio_config_read(EventNotifier *n)
> +{
> +    VirtIODevice *vdev = container_of(n, VirtIODevice, config_notifier);
> +
> +    if (event_notifier_test_and_clear(n)) {
> +        virtio_notify_config(vdev);
> +    }
> +}
>   void virtio_queue_set_guest_notifier_fd_handler(VirtQueue *vq, bool assign,
>                                                   bool with_irqfd)
>   {
> @@ -3517,6 +3525,16 @@ void 
> virtio_queue_set_guest_notifier_fd_handler(VirtQueue *vq, bool assign,
>           virtio_queue_guest_notifier_read(&vq->guest_notifier);
>       }
>   }
> +void virtio_set_config_notifier_fd_handler(VirtIODevice *vdev, bool assign,
> +                                                bool with_irqfd)
> +{
> +    if (assign && !with_irqfd) {
> +        event_notifier_set_handler(&vdev->config_notifier,
> +                                   virtio_config_read);
> +    } else {
> +       event_notifier_set_handler(&vdev->config_notifier, NULL);
> +    }
> +}
>   
>   EventNotifier *virtio_queue_get_guest_notifier(VirtQueue *vq)
>   {
> @@ -3591,6 +3609,10 @@ EventNotifier *virtio_queue_get_host_notifier(VirtQueue 
> *vq)
>       return &vq->host_notifier;
>   }
>   
> +EventNotifier *virtio_get_config_notifier(VirtIODevice *vdev)
> +{
> +    return &vdev->config_notifier;
> +}
>   void virtio_queue_set_host_notifier_enabled(VirtQueue *vq, bool enabled)
>   {
>       vq->host_notifier_enabled = enabled;
> diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
> index 4a8bc75415..22efa7008e 100644
> --- a/include/hw/virtio/vhost.h
> +++ b/include/hw/virtio/vhost.h
> @@ -91,6 +91,7 @@ struct vhost_dev {
>       QLIST_HEAD(, vhost_iommu) iommu_list;
>       IOMMUNotifier n;
>       const VhostDevConfigOps *config_ops;
> +    EventNotifier masked_config_notifier;


So I think it's wrong to store the masked_config_notifier in vhost_dev. 
See my above reply for the case of multiqueue. The correct way is to 
store them somewhere else, probably VirtIODevice.

Thanks


>   };
>   
>   struct vhost_net {
> @@ -108,6 +109,8 @@ int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice 
> *vdev);
>   void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev);
>   int vhost_dev_enable_notifiers(struct vhost_dev *hdev, VirtIODevice *vdev);
>   void vhost_dev_disable_notifiers(struct vhost_dev *hdev, VirtIODevice *vdev);
> +bool vhost_config_pending(struct vhost_dev *hdev, int n);
> +void vhost_config_mask(struct vhost_dev *hdev, VirtIODevice *vdev,  bool mask);
>   
>   /* Test and clear masked event pending status.
>    * Should be called after unmask to avoid losing events.
> diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
> index b7ece7a6a8..b0b714f6d4 100644
> --- a/include/hw/virtio/virtio.h
> +++ b/include/hw/virtio/virtio.h
> @@ -108,6 +108,7 @@ struct VirtIODevice
>       bool use_guest_notifier_mask;
>       AddressSpace *dma_as;
>       QLIST_HEAD(, VirtQueue) *vector_queues;
> +    EventNotifier config_notifier;
>   };
>   
>   struct VirtioDeviceClass {
> @@ -310,11 +311,14 @@ uint16_t virtio_get_queue_index(VirtQueue *vq);
>   EventNotifier *virtio_queue_get_guest_notifier(VirtQueue *vq);
>   void virtio_queue_set_guest_notifier_fd_handler(VirtQueue *vq, bool assign,
>                                                   bool with_irqfd);
> +void virtio_set_config_notifier_fd_handler(VirtIODevice *vdev, bool assign,
> +                                                bool with_irqfd);
>   int virtio_device_start_ioeventfd(VirtIODevice *vdev);
>   int virtio_device_grab_ioeventfd(VirtIODevice *vdev);
>   void virtio_device_release_ioeventfd(VirtIODevice *vdev);
>   bool virtio_device_ioeventfd_enabled(VirtIODevice *vdev);
>   EventNotifier *virtio_queue_get_host_notifier(VirtQueue *vq);
> +EventNotifier *virtio_get_config_notifier(VirtIODevice *vdev);
>   void virtio_queue_set_host_notifier_enabled(VirtQueue *vq, bool enabled);
>   void virtio_queue_host_notifier_read(EventNotifier *n);
>   void virtio_queue_aio_set_host_notifier_handler(VirtQueue *vq, AioContext 
> *ctx,
> diff --git a/include/net/vhost_net.h b/include/net/vhost_net.h
> index 172b0051d8..0d38c97c94 100644
> --- a/include/net/vhost_net.h
> +++ b/include/net/vhost_net.h
> @@ -36,6 +36,9 @@ int vhost_net_set_config(struct vhost_net *net, const uint8_t 
> *data,
>   bool vhost_net_virtqueue_pending(VHostNetState *net, int n);
>   void vhost_net_virtqueue_mask(VHostNetState *net, VirtIODevice *dev,
>                                 int idx, bool mask);
> +bool vhost_net_config_pending(VHostNetState *net, int n);
> +void vhost_net_config_mask(VHostNetState *net, VirtIODevice *dev,
> +                              bool mask);
>   int vhost_net_notify_migration_done(VHostNetState *net, char* mac_addr);
>   VHostNetState *get_vhost_net(NetClientState *nc);
>