[PATCH 0/2] virtio: add VIRTIO_F_NOTIFICATION_DATA feature support

[PATCH 0/2] virtio: add VIRTIO_F_NOTIFICATION_DATA feature support

[Alvaro Karsz]

Add VIRTIO_F_NOTIFICATION_DATA feature support for the MMIO, channel
I/O, modern PCI and vDPA transports.

This patchset binds 2 patches that were sent separately to the mailing
lists.

The first one [1] adds support for the MMIO, channel I/O and modern PCI transports.
The second one [2] adds support to the vDPA transport.

[1] https://lore.kernel.org/lkml/20230324195029.2410503-1-viktor@daynix.com/
[2] https://lore.kernel.org/virtualization/20230409070706.3288876-1-alvaro.karsz@solid-run.com/

Alvaro Karsz (1):
  virtio-vdpa: add VIRTIO_F_NOTIFICATION_DATA feature support

Viktor Prutyanov (1):
  virtio: add VIRTIO_F_NOTIFICATION_DATA feature support

 drivers/s390/virtio/virtio_ccw.c   | 22 +++++++++++++++++++---
 drivers/virtio/virtio_mmio.c       | 18 +++++++++++++++++-
 drivers/virtio/virtio_pci_modern.c | 17 ++++++++++++++++-
 drivers/virtio/virtio_ring.c       | 19 +++++++++++++++++++
 drivers/virtio/virtio_vdpa.c       | 23 +++++++++++++++++++++--
 include/linux/vdpa.h               |  9 +++++++++
 include/linux/virtio_ring.h        |  2 ++
 include/uapi/linux/virtio_config.h |  6 ++++++
 8 files changed, 109 insertions(+), 7 deletions(-)

-- 
2.34.1

[PATCH 1/2] virtio: add VIRTIO_F_NOTIFICATION_DATA feature support

[Alvaro Karsz]

From: Viktor Prutyanov <viktor@daynix.com>

According to VirtIO spec v1.2, VIRTIO_F_NOTIFICATION_DATA feature
indicates that the driver passes extra data along with the queue
notifications.

In a split queue case, the extra data is 16-bit available index. In a
packed queue case, the extra data is 1-bit wrap counter and 15-bit
available index.

Add support for this feature for MMIO, channel I/O and modern PCI
transports.

Signed-off-by: Viktor Prutyanov <viktor@daynix.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
---
 drivers/s390/virtio/virtio_ccw.c   | 22 +++++++++++++++++++---
 drivers/virtio/virtio_mmio.c       | 18 +++++++++++++++++-
 drivers/virtio/virtio_pci_modern.c | 17 ++++++++++++++++-
 drivers/virtio/virtio_ring.c       | 19 +++++++++++++++++++
 include/linux/virtio_ring.h        |  2 ++
 include/uapi/linux/virtio_config.h |  6 ++++++
 6 files changed, 79 insertions(+), 5 deletions(-)

diff --git a/drivers/s390/virtio/virtio_ccw.c b/drivers/s390/virtio/virtio_ccw.c
index 954fc31b4bc..02922768b12 100644
--- a/drivers/s390/virtio/virtio_ccw.c
+++ b/drivers/s390/virtio/virtio_ccw.c
@@ -391,7 +391,7 @@ static void virtio_ccw_drop_indicator(struct virtio_ccw_device *vcdev,
 	ccw_device_dma_free(vcdev->cdev, thinint_area, sizeof(*thinint_area));
 }
 
-static bool virtio_ccw_kvm_notify(struct virtqueue *vq)
+static inline bool virtio_ccw_do_kvm_notify(struct virtqueue *vq, u32 data)
 {
 	struct virtio_ccw_vq_info *info = vq->priv;
 	struct virtio_ccw_device *vcdev;
@@ -402,12 +402,22 @@ static bool virtio_ccw_kvm_notify(struct virtqueue *vq)
 	BUILD_BUG_ON(sizeof(struct subchannel_id) != sizeof(unsigned int));
 	info->cookie = kvm_hypercall3(KVM_S390_VIRTIO_CCW_NOTIFY,
 				      *((unsigned int *)&schid),
-				      vq->index, info->cookie);
+				      data, info->cookie);
 	if (info->cookie < 0)
 		return false;
 	return true;
 }
 
+static bool virtio_ccw_kvm_notify(struct virtqueue *vq)
+{
+	return virtio_ccw_do_kvm_notify(vq, vq->index);
+}
+
+static bool virtio_ccw_kvm_notify_with_data(struct virtqueue *vq)
+{
+	return virtio_ccw_do_kvm_notify(vq, vring_notification_data(vq));
+}
+
 static int virtio_ccw_read_vq_conf(struct virtio_ccw_device *vcdev,
 				   struct ccw1 *ccw, int index)
 {
@@ -495,6 +505,7 @@ static struct virtqueue *virtio_ccw_setup_vq(struct virtio_device *vdev,
 					     struct ccw1 *ccw)
 {
 	struct virtio_ccw_device *vcdev = to_vc_device(vdev);
+	bool (*notify)(struct virtqueue *vq);
 	int err;
 	struct virtqueue *vq = NULL;
 	struct virtio_ccw_vq_info *info;
@@ -502,6 +513,11 @@ static struct virtqueue *virtio_ccw_setup_vq(struct virtio_device *vdev,
 	unsigned long flags;
 	bool may_reduce;
 
+	if (__virtio_test_bit(vdev, VIRTIO_F_NOTIFICATION_DATA))
+		notify = virtio_ccw_kvm_notify_with_data;
+	else
+		notify = virtio_ccw_kvm_notify;
+
 	/* Allocate queue. */
 	info = kzalloc(sizeof(struct virtio_ccw_vq_info), GFP_KERNEL);
 	if (!info) {
@@ -524,7 +540,7 @@ static struct virtqueue *virtio_ccw_setup_vq(struct virtio_device *vdev,
 	may_reduce = vcdev->revision > 0;
 	vq = vring_create_virtqueue(i, info->num, KVM_VIRTIO_CCW_RING_ALIGN,
 				    vdev, true, may_reduce, ctx,
-				    virtio_ccw_kvm_notify, callback, name);
+				    notify, callback, name);
 
 	if (!vq) {
 		/* For now, we fail if we can't get the requested size. */
diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c
index 3ff746e3f24..dd4424c1423 100644
--- a/drivers/virtio/virtio_mmio.c
+++ b/drivers/virtio/virtio_mmio.c
@@ -285,6 +285,16 @@ static bool vm_notify(struct virtqueue *vq)
 	return true;
 }
 
+static bool vm_notify_with_data(struct virtqueue *vq)
+{
+	struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vq->vdev);
+	u32 data = vring_notification_data(vq);
+
+	writel(data, vm_dev->base + VIRTIO_MMIO_QUEUE_NOTIFY);
+
+	return true;
+}
+
 /* Notify all virtqueues on an interrupt. */
 static irqreturn_t vm_interrupt(int irq, void *opaque)
 {
@@ -363,12 +373,18 @@ static struct virtqueue *vm_setup_vq(struct virtio_device *vdev, unsigned int in
 				  const char *name, bool ctx)
 {
 	struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vdev);
+	bool (*notify)(struct virtqueue *vq);
 	struct virtio_mmio_vq_info *info;
 	struct virtqueue *vq;
 	unsigned long flags;
 	unsigned int num;
 	int err;
 
+	if (__virtio_test_bit(vdev, VIRTIO_F_NOTIFICATION_DATA))
+		notify = vm_notify_with_data;
+	else
+		notify = vm_notify;
+
 	if (!name)
 		return NULL;
 
@@ -397,7 +413,7 @@ static struct virtqueue *vm_setup_vq(struct virtio_device *vdev, unsigned int in
 
 	/* Create the vring */
 	vq = vring_create_virtqueue(index, num, VIRTIO_MMIO_VRING_ALIGN, vdev,
-				 true, true, ctx, vm_notify, callback, name);
+				 true, true, ctx, notify, callback, name);
 	if (!vq) {
 		err = -ENOMEM;
 		goto error_new_virtqueue;
diff --git a/drivers/virtio/virtio_pci_modern.c b/drivers/virtio/virtio_pci_modern.c
index 9e496e288cf..05deba5153b 100644
--- a/drivers/virtio/virtio_pci_modern.c
+++ b/drivers/virtio/virtio_pci_modern.c
@@ -288,6 +288,15 @@ static u16 vp_config_vector(struct virtio_pci_device *vp_dev, u16 vector)
 	return vp_modern_config_vector(&vp_dev->mdev, vector);
 }
 
+static bool vp_notify_with_data(struct virtqueue *vq)
+{
+	u32 data = vring_notification_data(vq);
+
+	iowrite32(data, (void __iomem *)vq->priv);
+
+	return true;
+}
+
 static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 				  struct virtio_pci_vq_info *info,
 				  unsigned int index,
@@ -298,10 +307,16 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 {
 
 	struct virtio_pci_modern_device *mdev = &vp_dev->mdev;
+	bool (*notify)(struct virtqueue *vq);
 	struct virtqueue *vq;
 	u16 num;
 	int err;
 
+	if (__virtio_test_bit(&vp_dev->vdev, VIRTIO_F_NOTIFICATION_DATA))
+		notify = vp_notify_with_data;
+	else
+		notify = vp_notify;
+
 	if (index >= vp_modern_get_num_queues(mdev))
 		return ERR_PTR(-EINVAL);
 
@@ -321,7 +336,7 @@ static struct virtqueue *setup_vq(struct virtio_pci_device *vp_dev,
 	vq = vring_create_virtqueue(index, num,
 				    SMP_CACHE_BYTES, &vp_dev->vdev,
 				    true, true, ctx,
-				    vp_notify, callback, name);
+				    notify, callback, name);
 	if (!vq)
 		return ERR_PTR(-ENOMEM);
 
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index 41144b5246a..7749bd5f99f 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c
@@ -2752,6 +2752,23 @@ void vring_del_virtqueue(struct virtqueue *_vq)
 }
 EXPORT_SYMBOL_GPL(vring_del_virtqueue);
 
+u32 vring_notification_data(struct virtqueue *_vq)
+{
+	struct vring_virtqueue *vq = to_vvq(_vq);
+	u16 next;
+
+	if (vq->packed_ring)
+		next = (vq->packed.next_avail_idx &
+				~(-(1 << VRING_PACKED_EVENT_F_WRAP_CTR))) |
+			vq->packed.avail_wrap_counter <<
+				VRING_PACKED_EVENT_F_WRAP_CTR;
+	else
+		next = vq->split.avail_idx_shadow;
+
+	return next << 16 | _vq->index;
+}
+EXPORT_SYMBOL_GPL(vring_notification_data);
+
 /* Manipulates transport-specific feature bits. */
 void vring_transport_features(struct virtio_device *vdev)
 {
@@ -2771,6 +2788,8 @@ void vring_transport_features(struct virtio_device *vdev)
 			break;
 		case VIRTIO_F_ORDER_PLATFORM:
 			break;
+		case VIRTIO_F_NOTIFICATION_DATA:
+			break;
 		default:
 			/* We don't understand this bit. */
 			__virtio_clear_bit(vdev, i);
diff --git a/include/linux/virtio_ring.h b/include/linux/virtio_ring.h
index 8b95b69ef69..2550c9170f4 100644
--- a/include/linux/virtio_ring.h
+++ b/include/linux/virtio_ring.h
@@ -117,4 +117,6 @@ void vring_del_virtqueue(struct virtqueue *vq);
 void vring_transport_features(struct virtio_device *vdev);
 
 irqreturn_t vring_interrupt(int irq, void *_vq);
+
+u32 vring_notification_data(struct virtqueue *_vq);
 #endif /* _LINUX_VIRTIO_RING_H */
diff --git a/include/uapi/linux/virtio_config.h b/include/uapi/linux/virtio_config.h
index 3c05162bc98..2c712c65416 100644
--- a/include/uapi/linux/virtio_config.h
+++ b/include/uapi/linux/virtio_config.h
@@ -99,6 +99,12 @@
  */
 #define VIRTIO_F_SR_IOV			37
 
+/*
+ * This feature indicates that the driver passes extra data (besides
+ * identifying the virtqueue) in its device notifications.
+ */
+#define VIRTIO_F_NOTIFICATION_DATA	38
+
 /*
  * This feature indicates that the driver can reset a queue individually.
  */
-- 
2.34.1

[PATCH 2/2] virtio-vdpa: add VIRTIO_F_NOTIFICATION_DATA feature support

[Alvaro Karsz]

Add VIRTIO_F_NOTIFICATION_DATA support for vDPA transport.
If this feature is negotiated, the driver passes extra data when kicking
a virtqueue.

A device that offers this feature needs to implement the
kick_vq_with_data callback.

kick_vq_with_data receives the vDPA device and data.
data includes:
16 bits vqn and 16 bits next available index for split virtqueues.
16 bits vqs, 15 least significant bits of next available index
and 1 bit next_wrap for packed virtqueues.

This patch follows a patch [1] by Viktor Prutyanov which adds support
for the MMIO, channel I/O and modern PCI transports.

Signed-off-by: Alvaro Karsz <alvaro.karsz@solid-run.com>
---
 drivers/virtio/virtio_vdpa.c | 23 +++++++++++++++++++++--
 include/linux/vdpa.h         |  9 +++++++++
 2 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/drivers/virtio/virtio_vdpa.c b/drivers/virtio/virtio_vdpa.c
index d7f5af62dda..737c1f36d32 100644
--- a/drivers/virtio/virtio_vdpa.c
+++ b/drivers/virtio/virtio_vdpa.c
@@ -112,6 +112,17 @@ static bool virtio_vdpa_notify(struct virtqueue *vq)
 	return true;
 }
 
+static bool virtio_vdpa_notify_with_data(struct virtqueue *vq)
+{
+	struct vdpa_device *vdpa = vd_get_vdpa(vq->vdev);
+	const struct vdpa_config_ops *ops = vdpa->config;
+	u32 data = vring_notification_data(vq);
+
+	ops->kick_vq_with_data(vdpa, data);
+
+	return true;
+}
+
 static irqreturn_t virtio_vdpa_config_cb(void *private)
 {
 	struct virtio_vdpa_device *vd_dev = private;
@@ -138,6 +149,7 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
 	struct device *dma_dev;
 	const struct vdpa_config_ops *ops = vdpa->config;
 	struct virtio_vdpa_vq_info *info;
+	bool (*notify)(struct virtqueue *vq) = virtio_vdpa_notify;
 	struct vdpa_callback cb;
 	struct virtqueue *vq;
 	u64 desc_addr, driver_addr, device_addr;
@@ -154,6 +166,14 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
 	if (index >= vdpa->nvqs)
 		return ERR_PTR(-ENOENT);
 
+	/* We cannot accept VIRTIO_F_NOTIFICATION_DATA without kick_vq_with_data */
+	if (__virtio_test_bit(vdev, VIRTIO_F_NOTIFICATION_DATA)) {
+		if (ops->kick_vq_with_data)
+			notify = virtio_vdpa_notify_with_data;
+		else
+			__virtio_clear_bit(vdev, VIRTIO_F_NOTIFICATION_DATA);
+	}
+
 	/* Queue shouldn't already be set up. */
 	if (ops->get_vq_ready(vdpa, index))
 		return ERR_PTR(-ENOENT);
@@ -183,8 +203,7 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
 		dma_dev = vdpa_get_dma_dev(vdpa);
 	vq = vring_create_virtqueue_dma(index, max_num, align, vdev,
 					true, may_reduce_num, ctx,
-					virtio_vdpa_notify, callback,
-					name, dma_dev);
+					notify, callback, name, dma_dev);
 	if (!vq) {
 		err = -ENOMEM;
 		goto error_new_virtqueue;
diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h
index 43f59ef10cc..04cdaad77dd 100644
--- a/include/linux/vdpa.h
+++ b/include/linux/vdpa.h
@@ -143,6 +143,14 @@ struct vdpa_map_file {
  * @kick_vq:			Kick the virtqueue
  *				@vdev: vdpa device
  *				@idx: virtqueue index
+ * @kick_vq_with_data:		Kick the virtqueue and supply extra data
+ *				(only if VIRTIO_F_NOTIFICATION_DATA is negotiated)
+ *				@vdev: vdpa device
+ *				@data for split virtqueue:
+ *				16 bits vqn and 16 bits next available index.
+ *				@data for packed virtqueue:
+ *				16 bits vqn, 15 least significant bits of
+ *				next available index and 1 bit next_wrap.
  * @set_vq_cb:			Set the interrupt callback function for
  *				a virtqueue
  *				@vdev: vdpa device
@@ -300,6 +308,7 @@ struct vdpa_config_ops {
 			      u64 device_area);
 	void (*set_vq_num)(struct vdpa_device *vdev, u16 idx, u32 num);
 	void (*kick_vq)(struct vdpa_device *vdev, u16 idx);
+	void (*kick_vq_with_data)(struct vdpa_device *vdev, u32 data);
 	void (*set_vq_cb)(struct vdpa_device *vdev, u16 idx,
 			  struct vdpa_callback *cb);
 	void (*set_vq_ready)(struct vdpa_device *vdev, u16 idx, bool ready);
-- 
2.34.1

Re: [PATCH 2/2] virtio-vdpa: add VIRTIO_F_NOTIFICATION_DATA feature support

[Jason Wang]

On Thu, Apr 13, 2023 at 4:19 PM Alvaro Karsz <alvaro.karsz@solid-run.com> wrote:
>
> Add VIRTIO_F_NOTIFICATION_DATA support for vDPA transport.
> If this feature is negotiated, the driver passes extra data when kicking
> a virtqueue.
>
> A device that offers this feature needs to implement the
> kick_vq_with_data callback.
>
> kick_vq_with_data receives the vDPA device and data.
> data includes:
> 16 bits vqn and 16 bits next available index for split virtqueues.
> 16 bits vqs, 15 least significant bits of next available index
> and 1 bit next_wrap for packed virtqueues.
>
> This patch follows a patch [1] by Viktor Prutyanov which adds support
> for the MMIO, channel I/O and modern PCI transports.
>
> Signed-off-by: Alvaro Karsz <alvaro.karsz@solid-run.com>

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

Thanks

> ---
>  drivers/virtio/virtio_vdpa.c | 23 +++++++++++++++++++++--
>  include/linux/vdpa.h         |  9 +++++++++
>  2 files changed, 30 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/virtio/virtio_vdpa.c b/drivers/virtio/virtio_vdpa.c
> index d7f5af62dda..737c1f36d32 100644
> --- a/drivers/virtio/virtio_vdpa.c
> +++ b/drivers/virtio/virtio_vdpa.c
> @@ -112,6 +112,17 @@ static bool virtio_vdpa_notify(struct virtqueue *vq)
>         return true;
>  }
>
> +static bool virtio_vdpa_notify_with_data(struct virtqueue *vq)
> +{
> +       struct vdpa_device *vdpa = vd_get_vdpa(vq->vdev);
> +       const struct vdpa_config_ops *ops = vdpa->config;
> +       u32 data = vring_notification_data(vq);
> +
> +       ops->kick_vq_with_data(vdpa, data);
> +
> +       return true;
> +}
> +
>  static irqreturn_t virtio_vdpa_config_cb(void *private)
>  {
>         struct virtio_vdpa_device *vd_dev = private;
> @@ -138,6 +149,7 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
>         struct device *dma_dev;
>         const struct vdpa_config_ops *ops = vdpa->config;
>         struct virtio_vdpa_vq_info *info;
> +       bool (*notify)(struct virtqueue *vq) = virtio_vdpa_notify;
>         struct vdpa_callback cb;
>         struct virtqueue *vq;
>         u64 desc_addr, driver_addr, device_addr;
> @@ -154,6 +166,14 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
>         if (index >= vdpa->nvqs)
>                 return ERR_PTR(-ENOENT);
>
> +       /* We cannot accept VIRTIO_F_NOTIFICATION_DATA without kick_vq_with_data */
> +       if (__virtio_test_bit(vdev, VIRTIO_F_NOTIFICATION_DATA)) {
> +               if (ops->kick_vq_with_data)
> +                       notify = virtio_vdpa_notify_with_data;
> +               else
> +                       __virtio_clear_bit(vdev, VIRTIO_F_NOTIFICATION_DATA);
> +       }
> +
>         /* Queue shouldn't already be set up. */
>         if (ops->get_vq_ready(vdpa, index))
>                 return ERR_PTR(-ENOENT);
> @@ -183,8 +203,7 @@ virtio_vdpa_setup_vq(struct virtio_device *vdev, unsigned int index,
>                 dma_dev = vdpa_get_dma_dev(vdpa);
>         vq = vring_create_virtqueue_dma(index, max_num, align, vdev,
>                                         true, may_reduce_num, ctx,
> -                                       virtio_vdpa_notify, callback,
> -                                       name, dma_dev);
> +                                       notify, callback, name, dma_dev);
>         if (!vq) {
>                 err = -ENOMEM;
>                 goto error_new_virtqueue;
> diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h
> index 43f59ef10cc..04cdaad77dd 100644
> --- a/include/linux/vdpa.h
> +++ b/include/linux/vdpa.h
> @@ -143,6 +143,14 @@ struct vdpa_map_file {
>   * @kick_vq:                   Kick the virtqueue
>   *                             @vdev: vdpa device
>   *                             @idx: virtqueue index
> + * @kick_vq_with_data:         Kick the virtqueue and supply extra data
> + *                             (only if VIRTIO_F_NOTIFICATION_DATA is negotiated)
> + *                             @vdev: vdpa device
> + *                             @data for split virtqueue:
> + *                             16 bits vqn and 16 bits next available index.
> + *                             @data for packed virtqueue:
> + *                             16 bits vqn, 15 least significant bits of
> + *                             next available index and 1 bit next_wrap.
>   * @set_vq_cb:                 Set the interrupt callback function for
>   *                             a virtqueue
>   *                             @vdev: vdpa device
> @@ -300,6 +308,7 @@ struct vdpa_config_ops {
>                               u64 device_area);
>         void (*set_vq_num)(struct vdpa_device *vdev, u16 idx, u32 num);
>         void (*kick_vq)(struct vdpa_device *vdev, u16 idx);
> +       void (*kick_vq_with_data)(struct vdpa_device *vdev, u32 data);
>         void (*set_vq_cb)(struct vdpa_device *vdev, u16 idx,
>                           struct vdpa_callback *cb);
>         void (*set_vq_ready)(struct vdpa_device *vdev, u16 idx, bool ready);
> --
> 2.34.1
>