Re: [PATCH v4 3/5] udp_vu: Move virtqueue management from udp_vu_sock_recv() to its caller

[David Gibson <david@gibson.dropbear.id.au>]

[-- Attachment #1: Type: text/plain, Size: 6440 bytes --]

On Mon, Mar 23, 2026 at 03:31:49PM +0100, Laurent Vivier wrote:
> udp_vu_sock_recv() currently mixes two concerns: receiving data from the
> socket and managing virtqueue buffers (collecting, rewinding, releasing).
> This makes the function harder to reason about and couples socket I/O
> with virtqueue state.
> 
> Move all virtqueue operations, vu_collect(), vu_init_elem(),
> vu_queue_rewind(), vu_set_vnethdr(), and the queue-readiness check, into
> udp_vu_sock_to_tap(), which is the only caller.  This turns
> udp_vu_sock_recv() into a pure socket receive function that simply reads
> into the provided iov array and adjusts its length.
> 
> Signed-off-by: Laurent Vivier <lvivier@redhat.com>

Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

> ---
>  udp_vu.c | 110 +++++++++++++++++++++++++------------------------------
>  1 file changed, 49 insertions(+), 61 deletions(-)
> 
> diff --git a/udp_vu.c b/udp_vu.c
> index 6638cb0ddc5f..dc949e30a793 100644
> --- a/udp_vu.c
> +++ b/udp_vu.c
> @@ -33,9 +33,6 @@
>  #include "udp_vu.h"
>  #include "vu_common.h"
>  
> -static struct iovec     iov_vu		[VIRTQUEUE_MAX_SIZE];
> -static struct vu_virtq_element	elem		[VIRTQUEUE_MAX_SIZE];
> -
>  /**
>   * udp_vu_hdrlen() - Sum size of all headers, from UDP to virtio-net
>   * @v6:		Set for IPv6 packet
> @@ -58,78 +55,35 @@ static size_t udp_vu_hdrlen(bool v6)
>  
>  /**
>   * udp_vu_sock_recv() - Receive datagrams from socket into vhost-user buffers
> - * @c:		Execution context
>   * @iov:	IO vector for the frame (in/out)
>   * @cnt:	Number of IO vector entries (in/out)
> - * @vq:		virtqueue to use to receive data
>   * @s:		Socket to receive from
>   * @v6:		Set for IPv6 connections
>   *
> - * Return: size of received data, 0 if the datagram
> - *         was discarded because the virtqueue is not ready, -1 on error
> + * Return: size of received data, -1 on error
>   */
> -static ssize_t udp_vu_sock_recv(const struct ctx *c, struct iovec *iov,
> -				size_t *cnt, unsigned *elem_used,
> -				struct vu_virtq *vq, int s, bool v6)
> +static ssize_t udp_vu_sock_recv(struct iovec *iov, size_t *cnt, int s, bool v6)
>  {
> -	const struct vu_dev *vdev = c->vdev;
> -	struct msghdr msg  = { 0 };
> +	struct iovec msg_iov[*cnt];
> +	struct msghdr msg = { 0 };
>  	struct iov_tail payload;
> -	size_t hdrlen, iov_used;
> -	unsigned elem_cnt;
> -	unsigned i, j;
> +	size_t hdrlen;
>  	ssize_t dlen;
>  
> -	assert(!c->no_udp);
> -
> -	if (!vu_queue_enabled(vq) || !vu_queue_started(vq)) {
> -		debug("Got UDP packet, but RX virtqueue not usable yet");
> -
> -		if (recvmsg(s, &msg, MSG_DONTWAIT) < 0)
> -			debug_perror("Failed to discard datagram");
> -
> -		*cnt = 0;
> -		return 0;
> -	}
> -
>  	/* compute L2 header length */
>  	hdrlen = udp_vu_hdrlen(v6);
>  
> -	elem_cnt = vu_collect(vdev, vq, elem, ARRAY_SIZE(elem),
> -			      iov, *cnt, &iov_used,
> -			      IP_MAX_MTU + ETH_HLEN + VNET_HLEN, NULL);
> -	if (elem_cnt == 0)
> -		return -1;
> -
> -	assert((size_t)elem_cnt == iov_used);	/* one iovec per element */
> -
> -	payload = IOV_TAIL(iov, iov_used, hdrlen);
> +	payload = IOV_TAIL(iov, *cnt, hdrlen);
>  
> -	struct iovec msg_iov[payload.cnt];
>  	msg.msg_iov = msg_iov;
>  	msg.msg_iovlen = iov_tail_clone(msg.msg_iov, payload.cnt, &payload);
>  
>  	/* read data from the socket */
>  	dlen = recvmsg(s, &msg, 0);
> -	if (dlen < 0) {
> -		vu_queue_rewind(vq, elem_cnt);
> +	if (dlen < 0)
>  		return -1;
> -	}
> -
> -	*cnt = vu_pad(iov, iov_used, 0, dlen + hdrlen);
> -
> -	*elem_used = 0;
> -	for (i = 0, j = 0; j < *cnt && i < elem_cnt; i++) {
> -		if (j + elem[i].in_num > *cnt)
> -			elem[i].in_num = *cnt - j;
> -		j += elem[i].in_num;
> -		(*elem_used)++;
> -	}
>  
> -	vu_set_vnethdr(iov[0].iov_base, *elem_used);
> -
> -	/* release unused buffers */
> -	vu_queue_rewind(vq, elem_cnt - *elem_used);
> +	*cnt = vu_pad(iov, *cnt, 0, dlen + hdrlen);
>  
>  	return dlen;
>  }
> @@ -217,26 +171,60 @@ static void udp_vu_csum(const struct flowside *toside,
>   */
>  void udp_vu_sock_to_tap(const struct ctx *c, int s, int n, flow_sidx_t tosidx)
>  {
> +	static struct iovec     iov_vu		[VIRTQUEUE_MAX_SIZE];
> +	static struct vu_virtq_element	elem	[VIRTQUEUE_MAX_SIZE];
>  	const struct flowside *toside = flowside_at_sidx(tosidx);
>  	bool v6 = !(inany_v4(&toside->eaddr) && inany_v4(&toside->oaddr));
>  	struct vu_dev *vdev = c->vdev;
>  	struct vu_virtq *vq = &vdev->vq[VHOST_USER_RX_QUEUE];
> -	struct iov_tail data;
>  	int i;
>  
> +	assert(!c->no_udp);
> +
> +	if (!vu_queue_enabled(vq) || !vu_queue_started(vq)) {
> +		struct msghdr msg = { 0 };
> +
> +		debug("Got UDP packet, but RX virtqueue not usable yet");
> +
> +		for (i = 0; i < n; i++) {
> +			if (recvmsg(s, &msg, MSG_DONTWAIT) < 0)
> +				debug_perror("Failed to discard datagram");
> +		}
> +
> +		return;
> +	}
> +
>  	for (i = 0; i < n; i++) {
> -		unsigned elem_used;
> +		unsigned elem_used, elem_cnt, j, k;
>  		size_t iov_cnt;
>  		ssize_t dlen;
>  
> -		iov_cnt = ARRAY_SIZE(iov_vu);
> -		dlen = udp_vu_sock_recv(c, iov_vu, &iov_cnt, &elem_used, vq,
> -					s, v6);
> -		if (dlen < 0)
> +		elem_cnt = vu_collect(vdev, vq, elem, ARRAY_SIZE(elem),
> +				      iov_vu, ARRAY_SIZE(iov_vu), &iov_cnt,
> +				      IP_MAX_MTU + ETH_HLEN + VNET_HLEN, NULL);
> +		if (elem_cnt == 0)
> +			break;
> +
> +		dlen = udp_vu_sock_recv(iov_vu, &iov_cnt, s, v6);
> +		if (dlen < 0) {
> +			vu_queue_rewind(vq, elem_cnt);
>  			break;
> +		}
> +
> +		elem_used = 0;
> +		for (j = 0, k = 0; k < iov_cnt && j < elem_cnt; j++) {
> +			if (k + elem[j].in_num > iov_cnt)
> +				elem[j].in_num = iov_cnt - k;
> +			k += elem[j].in_num;
> +			elem_used++;
> +		}
> +
> +		/* release unused buffers */
> +		vu_queue_rewind(vq, elem_cnt - elem_used);
>  
>  		if (iov_cnt > 0) {
> -			data = IOV_TAIL(iov_vu, iov_cnt, 0);
> +			struct iov_tail data = IOV_TAIL(iov_vu, iov_cnt, 0);
> +			vu_set_vnethdr(iov_vu[0].iov_base, elem_used);
>  			udp_vu_prepare(c, &data, toside, dlen);
>  			if (*c->pcap) {
>  				udp_vu_csum(toside, &data);
> -- 
> 2.53.0
> 

-- 
David Gibson (he or they)	| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you, not the other way
				| around.
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]