[PATCH net-next] virtio-net: support xsk wake up

[PATCH net-next] virtio-net: support xsk wake up

[Menglong Dong]

For now, XDP_RING_NEED_WAKEUP is not supported properly by the virtio-net.
Take the tx path for example, we set xsk_set_tx_need_wakeup() in
virtnet_xsk_xmit(), but we didn't call xsk_clear_tx_need_wakeup()
anywhere, which means the user will call send() for every packet.

For the tx path, we will call xsk_set_tx_need_wakeup() after
virtnet_xsk_xmit_batch() if sq->vq is empty, as we can't be wakeup by the
skb_xmit_done() in this case. Otherwise, we will clear the wakeup flag.

For the rx path, we will call xsk_set_rx_need_wakeup() if we have free
buffers in rq->vq.

Race condition is considered for both rx and tx path.

Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
---
 drivers/net/virtio_net.c | 52 ++++++++++++++++++++++++++++++++++------
 1 file changed, 45 insertions(+), 7 deletions(-)

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 9b3da9f9786c..25e895b849a6 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -1323,16 +1323,27 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
 				   struct xsk_buff_pool *pool, gfp_t gfp)
 {
 	struct xdp_buff **xsk_buffs;
+	bool need_wakeup;
 	dma_addr_t addr;
 	int err = 0;
 	u32 len, i;
 	int num;
 
+	need_wakeup = xsk_uses_need_wakeup(pool);
 	xsk_buffs = rq->xsk_buffs;
 
+	/* If both rq->vq and fill ring are empty, and then the user submit
+	 * all the chunks to the fill ring and check the wake up flag
+	 * after xsk_buff_alloc_batch() and before xsk_set_rx_need_wakeup(),
+	 * we will lose the chance to wake up the rx napi, so we have to
+	 * set the need_wakeup flag here.
+	 */
+	if (need_wakeup && virtqueue_get_vring_size(rq->vq) == rq->vq->num_free)
+		xsk_set_rx_need_wakeup(pool);
+
 	num = xsk_buff_alloc_batch(pool, xsk_buffs, rq->vq->num_free);
 	if (!num) {
-		if (xsk_uses_need_wakeup(pool)) {
+		if (need_wakeup) {
 			xsk_set_rx_need_wakeup(pool);
 			/* Return 0 instead of -ENOMEM so that NAPI is
 			 * descheduled.
@@ -1341,8 +1352,6 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
 		}
 
 		return -ENOMEM;
-	} else {
-		xsk_clear_rx_need_wakeup(pool);
 	}
 
 	len = xsk_pool_get_rx_frame_size(pool) + vi->hdr_len;
@@ -1363,6 +1372,16 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
 			goto err;
 	}
 
+	if (need_wakeup) {
+		if (rq->vq->num_free)
+			/* We have free buffers, so we'd better wake up the
+			 * rx napi as soon as possible.
+			 */
+			xsk_set_rx_need_wakeup(pool);
+		else
+			xsk_clear_rx_need_wakeup(pool);
+	}
+
 	return num;
 
 err:
@@ -1440,8 +1459,9 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
 	struct virtnet_info *vi = sq->vq->vdev->priv;
 	struct virtnet_sq_free_stats stats = {};
 	struct net_device *dev = vi->dev;
+	int sent, vring_size;
+	bool need_wakeup;
 	u64 kicks = 0;
-	int sent;
 
 	/* Avoid to wakeup napi meanless, so call __free_old_xmit instead of
 	 * free_old_xmit().
@@ -1451,8 +1471,29 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
 	if (stats.xsk)
 		xsk_tx_completed(sq->xsk_pool, stats.xsk);
 
+	vring_size = virtqueue_get_vring_size(sq->vq);
+	need_wakeup = xsk_uses_need_wakeup(pool);
+	/* If the sq->vq is empty, and the tx ring is empty, and the user
+	 * submit an entry to the tx ring after virtnet_xsk_xmit_batch() and
+	 * before xsk_set_tx_need_wakeup(), we will lose the chance to wake
+	 * up the tx napi, so we have to set the need_wakeup flag here.
+	 */
+	if (need_wakeup && vring_size == sq->vq->num_free)
+		xsk_set_tx_need_wakeup(pool);
+
 	sent = virtnet_xsk_xmit_batch(sq, pool, budget, &kicks);
 
+	if (need_wakeup) {
+		if (vring_size == sq->vq->num_free)
+			/* we can't wake up by ourself, and it should be done
+			 * by the user.
+			 */
+			xsk_set_tx_need_wakeup(pool);
+		else
+			/* we can wake up from skb_xmit_done() */
+			xsk_clear_tx_need_wakeup(pool);
+	}
+
 	if (!is_xdp_raw_buffer_queue(vi, sq - vi->sq))
 		check_sq_full_and_disable(vi, vi->dev, sq);
 
@@ -1470,9 +1511,6 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
 	u64_stats_add(&sq->stats.xdp_tx,  sent);
 	u64_stats_update_end(&sq->stats.syncp);
 
-	if (xsk_uses_need_wakeup(pool))
-		xsk_set_tx_need_wakeup(pool);
-
 	return sent;
 }
 
-- 
2.54.0

Re: [PATCH net-next] virtio-net: support xsk wake up

[Eugenio Perez Martin]

On Wed, Jun 10, 2026 at 10:17 AM Menglong Dong <menglong8.dong@gmail.com> wrote:
>
> For now, XDP_RING_NEED_WAKEUP is not supported properly by the virtio-net.
> Take the tx path for example, we set xsk_set_tx_need_wakeup() in
> virtnet_xsk_xmit(), but we didn't call xsk_clear_tx_need_wakeup()
> anywhere, which means the user will call send() for every packet.
>
> For the tx path, we will call xsk_set_tx_need_wakeup() after
> virtnet_xsk_xmit_batch() if sq->vq is empty, as we can't be wakeup by the
> skb_xmit_done() in this case. Otherwise, we will clear the wakeup flag.
>
> For the rx path, we will call xsk_set_rx_need_wakeup() if we have free
> buffers in rq->vq.
>
> Race condition is considered for both rx and tx path.
>

I didn't review the code itself, but the patch message lacks a Fixes:
tag. And the From and Signed-off-by emails don't match, which I'm not
sure is valid.

> Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
> ---
>  drivers/net/virtio_net.c | 52 ++++++++++++++++++++++++++++++++++------
>  1 file changed, 45 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> index 9b3da9f9786c..25e895b849a6 100644
> --- a/drivers/net/virtio_net.c
> +++ b/drivers/net/virtio_net.c
> @@ -1323,16 +1323,27 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
>                                    struct xsk_buff_pool *pool, gfp_t gfp)
>  {
>         struct xdp_buff **xsk_buffs;
> +       bool need_wakeup;
>         dma_addr_t addr;
>         int err = 0;
>         u32 len, i;
>         int num;
>
> +       need_wakeup = xsk_uses_need_wakeup(pool);
>         xsk_buffs = rq->xsk_buffs;
>
> +       /* If both rq->vq and fill ring are empty, and then the user submit
> +        * all the chunks to the fill ring and check the wake up flag
> +        * after xsk_buff_alloc_batch() and before xsk_set_rx_need_wakeup(),
> +        * we will lose the chance to wake up the rx napi, so we have to
> +        * set the need_wakeup flag here.
> +        */
> +       if (need_wakeup && virtqueue_get_vring_size(rq->vq) == rq->vq->num_free)
> +               xsk_set_rx_need_wakeup(pool);
> +
>         num = xsk_buff_alloc_batch(pool, xsk_buffs, rq->vq->num_free);
>         if (!num) {
> -               if (xsk_uses_need_wakeup(pool)) {
> +               if (need_wakeup) {
>                         xsk_set_rx_need_wakeup(pool);
>                         /* Return 0 instead of -ENOMEM so that NAPI is
>                          * descheduled.
> @@ -1341,8 +1352,6 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
>                 }
>
>                 return -ENOMEM;
> -       } else {
> -               xsk_clear_rx_need_wakeup(pool);
>         }
>
>         len = xsk_pool_get_rx_frame_size(pool) + vi->hdr_len;
> @@ -1363,6 +1372,16 @@ static int virtnet_add_recvbuf_xsk(struct virtnet_info *vi, struct receive_queue
>                         goto err;
>         }
>
> +       if (need_wakeup) {
> +               if (rq->vq->num_free)
> +                       /* We have free buffers, so we'd better wake up the
> +                        * rx napi as soon as possible.
> +                        */
> +                       xsk_set_rx_need_wakeup(pool);
> +               else
> +                       xsk_clear_rx_need_wakeup(pool);
> +       }
> +
>         return num;
>
>  err:
> @@ -1440,8 +1459,9 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
>         struct virtnet_info *vi = sq->vq->vdev->priv;
>         struct virtnet_sq_free_stats stats = {};
>         struct net_device *dev = vi->dev;
> +       int sent, vring_size;
> +       bool need_wakeup;
>         u64 kicks = 0;
> -       int sent;
>
>         /* Avoid to wakeup napi meanless, so call __free_old_xmit instead of
>          * free_old_xmit().
> @@ -1451,8 +1471,29 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
>         if (stats.xsk)
>                 xsk_tx_completed(sq->xsk_pool, stats.xsk);
>
> +       vring_size = virtqueue_get_vring_size(sq->vq);
> +       need_wakeup = xsk_uses_need_wakeup(pool);
> +       /* If the sq->vq is empty, and the tx ring is empty, and the user
> +        * submit an entry to the tx ring after virtnet_xsk_xmit_batch() and
> +        * before xsk_set_tx_need_wakeup(), we will lose the chance to wake
> +        * up the tx napi, so we have to set the need_wakeup flag here.
> +        */
> +       if (need_wakeup && vring_size == sq->vq->num_free)
> +               xsk_set_tx_need_wakeup(pool);
> +
>         sent = virtnet_xsk_xmit_batch(sq, pool, budget, &kicks);
>
> +       if (need_wakeup) {
> +               if (vring_size == sq->vq->num_free)
> +                       /* we can't wake up by ourself, and it should be done
> +                        * by the user.
> +                        */
> +                       xsk_set_tx_need_wakeup(pool);
> +               else
> +                       /* we can wake up from skb_xmit_done() */
> +                       xsk_clear_tx_need_wakeup(pool);
> +       }
> +
>         if (!is_xdp_raw_buffer_queue(vi, sq - vi->sq))
>                 check_sq_full_and_disable(vi, vi->dev, sq);
>
> @@ -1470,9 +1511,6 @@ static bool virtnet_xsk_xmit(struct send_queue *sq, struct xsk_buff_pool *pool,
>         u64_stats_add(&sq->stats.xdp_tx,  sent);
>         u64_stats_update_end(&sq->stats.syncp);
>
> -       if (xsk_uses_need_wakeup(pool))
> -               xsk_set_tx_need_wakeup(pool);
> -
>         return sent;
>  }
>
> --
> 2.54.0
>

Re: [PATCH net-next] virtio-net: support xsk wake up

[Menglong Dong]

On Wed, Jun 10, 2026 at 4:28 PM Eugenio Perez Martin
<eperezma@redhat.com> wrote:
>
> On Wed, Jun 10, 2026 at 10:17 AM Menglong Dong <menglong8.dong@gmail.com> wrote:
> >
> > For now, XDP_RING_NEED_WAKEUP is not supported properly by the virtio-net.
> > Take the tx path for example, we set xsk_set_tx_need_wakeup() in
> > virtnet_xsk_xmit(), but we didn't call xsk_clear_tx_need_wakeup()
> > anywhere, which means the user will call send() for every packet.
> >
> > For the tx path, we will call xsk_set_tx_need_wakeup() after
> > virtnet_xsk_xmit_batch() if sq->vq is empty, as we can't be wakeup by the
> > skb_xmit_done() in this case. Otherwise, we will clear the wakeup flag.
> >
> > For the rx path, we will call xsk_set_rx_need_wakeup() if we have free
> > buffers in rq->vq.
> >
> > Race condition is considered for both rx and tx path.
> >
>
> I didn't review the code itself, but the patch message lacks a Fixes:

Yeah, you are right, it's better to use the Fix tag here, and I'll
add it in the next version.

> tag. And the From and Signed-off-by emails don't match, which I'm not
> sure is valid.

Sorry about that. I use my company email in the SOB. I used
my company email for the From too. However, my company
email system can trigger "Delivery failure" messages sometimes,
and Alexei said it is annoying. So I didn't use it for the From
and removed it from the CC since then :/

Thanks!
Menglong Dong

>
> > Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
> > ---
> >  drivers/net/virtio_net.c | 52 ++++++++++++++++++++++++++++++++++------
> >  1 file changed, 45 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
> > index 9b3da9f9786c..25e895b849a6 100644
> > --- a/drivers/net/virtio_net.c
> > +++ b/drivers/net/virtio_net.c
[...]
> >
> > --
> > 2.54.0
> >
>

Re: [PATCH net-next] virtio-net: support xsk wake up

[Jakub Kicinski]

On Wed, 10 Jun 2026 16:16:48 +0800 Menglong Dong wrote:
> +	/* If both rq->vq and fill ring are empty, and then the user submit
> +	 * all the chunks to the fill ring and check the wake up flag
> +	 * after xsk_buff_alloc_batch() and before xsk_set_rx_need_wakeup(),
> +	 * we will lose the chance to wake up the rx napi, so we have to
> +	 * set the need_wakeup flag here.
> +	 */

TBH all the comments you're adding are harder to understand than the
code itself ;( Please try to phrase them better or just remove them.

Re: [PATCH net-next] virtio-net: support xsk wake up

[Jakub Kicinski]

On Wed, 10 Jun 2026 10:27:28 +0200 Eugenio Perez Martin wrote:
> And the From and Signed-off-by emails don't match, which I'm not sure is valid.

It's clearly the same person. Please focus on the code, not trivial
process issues.

Quoting documentation:

  Reviewer guidance
  -----------------

  [...]

  Reviewers are highly encouraged to do more in-depth review of submissions
  and not focus exclusively on process issues, trivial or subjective
  matters like code formatting, tags etc.
  
See: https://www.kernel.org/doc/html/next/process/maintainer-netdev.html#reviewer-guidance