public inbox for passt-dev@passt.top
 help / color / mirror / code / Atom feed
From: David Gibson <david@gibson.dropbear.id.au>
To: Laurent Vivier <lvivier@redhat.com>
Cc: passt-dev@passt.top
Subject: Re: [PATCH v2 3/6] vhost-user: Add queue pair parameter throughout the network stack
Date: Wed, 26 Nov 2025 13:35:29 +1100	[thread overview]
Message-ID: <aSZncX2NUHFBqqGu@zatzit> (raw)
In-Reply-To: <20251121165902.1014964-4-lvivier@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 29778 bytes --]

On Fri, Nov 21, 2025 at 05:58:59PM +0100, Laurent Vivier wrote:
> Add a queue pair parameter to vu_send_single() and propagate this parameter
> through the entire network stack call chain. The queue pair parameter specifies
> which queue pair to use for sending packets in vhost-user mode.

"sending" in this case meaning passt -> guest, right?  Which is the Rx
queue in vu terminology.

> Functions modified to accept and propagate the queue pair parameter:
> - Core sending: tap_send_single(), vu_send_single()
> - UDP/ICMP helpers: tap_udp4_send(), tap_udp6_send(), tap_icmp4_send(),
>   tap_icmp6_send()
> - Protocol handlers: arp(), dhcp(), dhcpv6(), ndp(), tcp_rst_no_conn()
> - ARP helpers: arp_announce()
> - NDP helpers: ndp_send(), ndp_na(), ndp_ra(), ndp_unsolicited_na()
> - UDP error handling: udp_send_tap_icmp4(), udp_send_tap_icmp6()
> 
> All callers currently pass queue pair #0 to preserve existing
> behavior. This is a preparatory step for enabling multi-queue and
> per-queue worker threads in vhost-user mode.
> 
> No functional change.
> 
> Signed-off-by: Laurent Vivier <lvivier@redhat.com>

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

Couple of trivial queries below.

> ---
>  arp.c       | 12 +++++++-----
>  arp.h       |  4 ++--
>  dhcp.c      |  5 +++--
>  dhcp.h      |  2 +-
>  dhcpv6.c    | 12 +++++++-----
>  dhcpv6.h    |  2 +-
>  fwd.c       |  4 ++--
>  icmp.c      |  6 ++++--
>  ndp.c       | 32 +++++++++++++++++++-------------
>  ndp.h       |  5 +++--
>  tap.c       | 37 ++++++++++++++++++++++---------------
>  tap.h       | 13 +++++++------
>  tcp.c       |  8 +++++---
>  udp.c       | 14 ++++++++------
>  vu_common.c |  9 ++++++---
>  vu_common.h |  3 ++-
>  16 files changed, 99 insertions(+), 69 deletions(-)
> 
> diff --git a/arp.c b/arp.c
> index 33b03cf6c316..c43d33ced2c3 100644
> --- a/arp.c
> +++ b/arp.c
> @@ -63,11 +63,12 @@ static bool ignore_arp(const struct ctx *c,
>  /**
>   * arp() - Check if this is a supported ARP message, reply as needed
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the reply
>   * @data:	Single packet with Ethernet buffer
>   *
>   * Return: 1 if handled, -1 on failure
>   */
> -int arp(const struct ctx *c, struct iov_tail *data)
> +int arp(const struct ctx *c, int qpair, struct iov_tail *data)
>  {
>  	union inany_addr tgt;
>  	struct {
> @@ -112,7 +113,7 @@ int arp(const struct ctx *c, struct iov_tail *data)
>  	memcpy(resp.am.tha,		am->sha,	sizeof(resp.am.tha));
>  	memcpy(resp.am.tip,		am->sip,	sizeof(resp.am.tip));
>  
> -	tap_send_single(c, &resp, sizeof(resp));
> +	tap_send_single(c, qpair, &resp, sizeof(resp));
>  
>  	return 1;
>  }
> @@ -148,16 +149,17 @@ void arp_send_init_req(const struct ctx *c)
>  	memcpy(req.am.tip,	&c->ip4.addr,		sizeof(req.am.tip));
>  
>  	debug("Sending initial ARP request for guest MAC address");
> -	tap_send_single(c, &req, sizeof(req));
> +	tap_send_single(c, 0, &req, sizeof(req));

Should this one also go into the caller?  I don't think it really
matters, it's only a question of what seems logically more sensible.

>  }
>  
>  /**
>   * arp_announce() - Send an ARP announcement for an IPv4 host
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the announcement
>   * @ip:	IPv4 address we announce as owned by @mac
>   * @mac:	MAC address to advertise for @ip
>   */
> -void arp_announce(const struct ctx *c, struct in_addr *ip,
> +void arp_announce(const struct ctx *c, int qpair, struct in_addr *ip,
>  		  const unsigned char *mac)
>  {
>  	char ip_str[INET_ADDRSTRLEN];
> @@ -199,5 +201,5 @@ void arp_announce(const struct ctx *c, struct in_addr *ip,
>  	eth_ntop(mac, mac_str, sizeof(mac_str));
>  	debug("ARP announcement for %s / %s", ip_str, mac_str);
>  
> -	tap_send_single(c, &msg, sizeof(msg));
> +	tap_send_single(c, qpair, &msg, sizeof(msg));
>  }
> diff --git a/arp.h b/arp.h
> index 4862e90a14ee..7dd872809340 100644
> --- a/arp.h
> +++ b/arp.h
> @@ -20,9 +20,9 @@ struct arpmsg {
>  	unsigned char tip[4];
>  } __attribute__((__packed__));
>  
> -int arp(const struct ctx *c, struct iov_tail *data);
> +int arp(const struct ctx *c, int qpair, struct iov_tail *data);
>  void arp_send_init_req(const struct ctx *c);
> -void arp_announce(const struct ctx *c, struct in_addr *ip,
> +void arp_announce(const struct ctx *c, int qpair, struct in_addr *ip,
>  		  const unsigned char *mac);
>  
>  #endif /* ARP_H */
> diff --git a/dhcp.c b/dhcp.c
> index 6b9c2e3b9e5a..dd3c67f52724 100644
> --- a/dhcp.c
> +++ b/dhcp.c
> @@ -296,11 +296,12 @@ static void opt_set_dns_search(const struct ctx *c, size_t max_len)
>  /**
>   * dhcp() - Check if this is a DHCP message, reply as needed
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the reply
>   * @data:	Single packet with Ethernet buffer
>   *
>   * Return: 0 if it's not a DHCP message, 1 if handled, -1 on failure
>   */
> -int dhcp(const struct ctx *c, struct iov_tail *data)
> +int dhcp(const struct ctx *c, int qpair, struct iov_tail *data)
>  {
>  	char macstr[ETH_ADDRSTRLEN];
>  	size_t mlen, dlen, opt_len;
> @@ -471,7 +472,7 @@ int dhcp(const struct ctx *c, struct iov_tail *data)
>  	else
>  		dst = c->ip4.addr;
>  
> -	tap_udp4_send(c, c->ip4.our_tap_addr, 67, dst, 68, &reply, dlen);
> +	tap_udp4_send(c, qpair, c->ip4.our_tap_addr, 67, dst, 68, &reply, dlen);
>  
>  	return 1;
>  }
> diff --git a/dhcp.h b/dhcp.h
> index cd50c99b8856..a96991633e58 100644
> --- a/dhcp.h
> +++ b/dhcp.h
> @@ -6,7 +6,7 @@
>  #ifndef DHCP_H
>  #define DHCP_H
>  
> -int dhcp(const struct ctx *c, struct iov_tail *data);
> +int dhcp(const struct ctx *c, int qpair, struct iov_tail *data);
>  void dhcp_init(void);
>  
>  #endif /* DHCP_H */
> diff --git a/dhcpv6.c b/dhcpv6.c
> index e4df0db562e6..b1b792612615 100644
> --- a/dhcpv6.c
> +++ b/dhcpv6.c
> @@ -369,12 +369,13 @@ notonlink:
>  /**
>   * dhcpv6_send_ia_notonlink() - Send NotOnLink status
>   * @c:			Execution context
> + * @qpair:		Queue pair on which to send the reply
>   * @ia_base:		Non-appropriate IA_NA or IA_TA base
>   * @client_id_base:	Client ID message option base
>   * @len:		Client ID length
>   * @xid:		Transaction ID for message exchange
>   */
> -static void dhcpv6_send_ia_notonlink(struct ctx *c,
> +static void dhcpv6_send_ia_notonlink(struct ctx *c, int qpair,
>  				     const struct iov_tail *ia_base,
>  				     const struct iov_tail *client_id_base,
>  				     int len, uint32_t xid)
> @@ -404,7 +405,7 @@ static void dhcpv6_send_ia_notonlink(struct ctx *c,
>  
>  	resp_not_on_link.hdr.xid = xid;
>  
> -	tap_udp6_send(c, src, 547, tap_ip6_daddr(c, src), 546,
> +	tap_udp6_send(c, qpair, src, 547, tap_ip6_daddr(c, src), 546,
>  		      xid, &resp_not_on_link, n);
>  }
>  
> @@ -539,13 +540,14 @@ static size_t dhcpv6_client_fqdn_fill(const struct iov_tail *data,
>  /**
>   * dhcpv6() - Check if this is a DHCPv6 message, reply as needed
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the reply
>   * @data:	Single packet starting from UDP header
>   * @saddr:	Source IPv6 address of original message
>   * @daddr:	Destination IPv6 address of original message
>   *
>   * Return: 0 if it's not a DHCPv6 message, 1 if handled, -1 on failure
>   */
> -int dhcpv6(struct ctx *c, struct iov_tail *data,
> +int dhcpv6(struct ctx *c, int qpair, struct iov_tail *data,
>  	   const struct in6_addr *saddr, const struct in6_addr *daddr)
>  {
>  	const struct opt_server_id *server_id = NULL;
> @@ -627,7 +629,7 @@ int dhcpv6(struct ctx *c, struct iov_tail *data,
>  
>  		if (dhcpv6_ia_notonlink(data, &c->ip6.addr)) {
>  
> -			dhcpv6_send_ia_notonlink(c, data, &client_id_base,
> +			dhcpv6_send_ia_notonlink(c, qpair, data, &client_id_base,
>  						 ntohs(client_id->l), mh->xid);
>  
>  			return 1;
> @@ -677,7 +679,7 @@ int dhcpv6(struct ctx *c, struct iov_tail *data,
>  
>  	resp.hdr.xid = mh->xid;
>  
> -	tap_udp6_send(c, src, 547, tap_ip6_daddr(c, src), 546,
> +	tap_udp6_send(c, qpair, src, 547, tap_ip6_daddr(c, src), 546,
>  		      mh->xid, &resp, n);
>  	c->ip6.addr_seen = c->ip6.addr;
>  
> diff --git a/dhcpv6.h b/dhcpv6.h
> index c706dfdbb2ac..420caf8b7169 100644
> --- a/dhcpv6.h
> +++ b/dhcpv6.h
> @@ -6,7 +6,7 @@
>  #ifndef DHCPV6_H
>  #define DHCPV6_H
>  
> -int dhcpv6(struct ctx *c, struct iov_tail *data,
> +int dhcpv6(struct ctx *c, int qpair, struct iov_tail *data,
>  	   struct in6_addr *saddr, struct in6_addr *daddr);
>  void dhcpv6_init(const struct ctx *c);
>  
> diff --git a/fwd.c b/fwd.c
> index 68bb11663c46..60c6ec3b6af9 100644
> --- a/fwd.c
> +++ b/fwd.c
> @@ -147,9 +147,9 @@ void fwd_neigh_table_update(const struct ctx *c, const union inany_addr *addr,
>  		return;
>  
>  	if (inany_v4(addr))
> -		arp_announce(c, inany_v4(addr), e->mac);
> +		arp_announce(c, 0, inany_v4(addr), e->mac);
>  	else
> -		ndp_unsolicited_na(c, &addr->a6);
> +		ndp_unsolicited_na(c, 0, &addr->a6);
>  }
>  
>  /**
> diff --git a/icmp.c b/icmp.c
> index 35faefb91870..a9f0518c2f61 100644
> --- a/icmp.c
> +++ b/icmp.c
> @@ -132,12 +132,14 @@ void icmp_sock_handler(const struct ctx *c, union epoll_ref ref)
>  		const struct in_addr *daddr = inany_v4(&ini->eaddr);
>  
>  		ASSERT(saddr && daddr); /* Must have IPv4 addresses */
> -		tap_icmp4_send(c, *saddr, *daddr, buf, pingf->f.tap_omac, n);
> +		tap_icmp4_send(c, 0, *saddr, *daddr, buf,
> +			       pingf->f.tap_omac, n);
>  	} else if (pingf->f.type == FLOW_PING6) {
>  		const struct in6_addr *saddr = &ini->oaddr.a6;
>  		const struct in6_addr *daddr = &ini->eaddr.a6;
>  
> -		tap_icmp6_send(c, saddr, daddr, buf, pingf->f.tap_omac, n);
> +		tap_icmp6_send(c, 0, saddr, daddr, buf,
> +			       pingf->f.tap_omac, n);
>  	}
>  	return;
>  
> diff --git a/ndp.c b/ndp.c
> index a33239d4aa81..0963a6392655 100644
> --- a/ndp.c
> +++ b/ndp.c
> @@ -175,25 +175,27 @@ struct ndp_ns {
>  /**
>   * ndp_send() - Send an NDP message
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the message
>   * @dst:	IPv6 address to send the message to
>   * @buf:	ICMPv6 header + message payload
>   * @l4len:	Length of message, including ICMPv6 header
>   */
> -static void ndp_send(const struct ctx *c, const struct in6_addr *dst,
> +static void ndp_send(const struct ctx *c, int qpair, const struct in6_addr *dst,
>  		     const void *buf, size_t l4len)
>  {
>  	const struct in6_addr *src = &c->ip6.our_tap_ll;
>  
> -	tap_icmp6_send(c, src, dst, buf, c->our_tap_mac, l4len);
> +	tap_icmp6_send(c, qpair, src, dst, buf, c->our_tap_mac, l4len);
>  }
>  
>  /**
>   * ndp_na() - Send an NDP Neighbour Advertisement (NA) message
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the NA
>   * @dst:	IPv6 address to send the NA to
>   * @addr:	IPv6 address to advertise
>   */
> -static void ndp_na(const struct ctx *c, const struct in6_addr *dst,
> +static void ndp_na(const struct ctx *c, int qpair, const struct in6_addr *dst,
>  	    const struct in6_addr *addr)
>  {
>  	union inany_addr tgt;
> @@ -217,25 +219,28 @@ static void ndp_na(const struct ctx *c, const struct in6_addr *dst,
>  	inany_from_af(&tgt, AF_INET6, addr);
>  	fwd_neigh_mac_get(c, &tgt, na.target_l2_addr.mac);
>  
> -	ndp_send(c, dst, &na, sizeof(na));
> +	ndp_send(c, qpair, dst, &na, sizeof(na));
>  }
>  
>  /**
>   * ndp_unsolicited_na() - Send unsolicited NA
>   * @c:         Execution context
> + * @qpair:     Queue pair on which to send the RA
>   * @addr:      IPv6 address to advertise
>   */
> -void ndp_unsolicited_na(const struct ctx *c, const struct in6_addr *addr)
> +void ndp_unsolicited_na(const struct ctx *c, int qpair,
> +			const struct in6_addr *addr)
>  {
> -	ndp_na(c, &in6addr_ll_all_nodes, addr);
> +	ndp_na(c, qpair, &in6addr_ll_all_nodes, addr);
>  }
>  
>  /**
>   * ndp_ra() - Send an NDP Router Advertisement (RA) message
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the RA
>   * @dst:	IPv6 address to send the RA to
>   */
> -static void ndp_ra(const struct ctx *c, const struct in6_addr *dst)
> +static void ndp_ra(const struct ctx *c, int qpair, const struct in6_addr *dst)
>  {
>  	struct ndp_ra ra = {
>  		.ih = {
> @@ -341,18 +346,19 @@ static void ndp_ra(const struct ctx *c, const struct in6_addr *dst)
>  	memcpy(&ra.source_ll.mac, c->our_tap_mac, ETH_ALEN);
>  
>  	/* NOLINTNEXTLINE(clang-analyzer-security.PointerSub) */
> -	ndp_send(c, dst, &ra, ptr - (unsigned char *)&ra);
> +	ndp_send(c, qpair, dst, &ra, ptr - (unsigned char *)&ra);
>  }
>  
>  /**
>   * ndp() - Check for NDP solicitations, reply as needed
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send replies
>   * @saddr:	Source IPv6 address
>   * @data:	Single packet with ICMPv6 header
>   *
>   * Return: 0 if not handled here, 1 if handled, -1 on failure
>   */
> -int ndp(const struct ctx *c, const struct in6_addr *saddr,
> +int ndp(const struct ctx *c, int qpair, const struct in6_addr *saddr,
>  	struct iov_tail *data)
>  {
>  	struct icmp6hdr ih_storage;
> @@ -381,13 +387,13 @@ int ndp(const struct ctx *c, const struct in6_addr *saddr,
>  
>  		info("NDP: received NS, sending NA");
>  
> -		ndp_na(c, saddr, &ns->target_addr);
> +		ndp_na(c, qpair, saddr, &ns->target_addr);
>  	} else if (ih->icmp6_type == RS) {
>  		if (c->no_ra)
>  			return 1;
>  
>  		info("NDP: received RS, sending RA");
> -		ndp_ra(c, saddr);
> +		ndp_ra(c, qpair, saddr);
>  	}
>  
>  	return 1;
> @@ -445,7 +451,7 @@ void ndp_timer(const struct ctx *c, const struct timespec *now)
>  
>  	info("NDP: sending unsolicited RA, next in %llds", (long long)interval);
>  
> -	ndp_ra(c, &in6addr_ll_all_nodes);
> +	ndp_ra(c, 0, &in6addr_ll_all_nodes);
>  
>  first:
>  	next_ra = now->tv_sec + interval;
> @@ -468,5 +474,5 @@ void ndp_send_init_req(const struct ctx *c)
>  		.target_addr = c->ip6.addr
>  	};
>  	debug("Sending initial NDP NS request for guest MAC address");
> -	ndp_send(c, &c->ip6.addr, &ns, sizeof(ns));
> +	ndp_send(c, 0, &c->ip6.addr, &ns, sizeof(ns));

As for arp_send_init_req(), would the 0 here make more sense in the
caller?

>  }
> diff --git a/ndp.h b/ndp.h
> index 56b756d8400b..927e69eb4649 100644
> --- a/ndp.h
> +++ b/ndp.h
> @@ -8,10 +8,11 @@
>  
>  struct icmp6hdr;
>  
> -int ndp(const struct ctx *c, const struct in6_addr *saddr,
> +int ndp(const struct ctx *c, int qpair, const struct in6_addr *saddr,
>  	struct iov_tail *data);
>  void ndp_timer(const struct ctx *c, const struct timespec *now);
>  void ndp_send_init_req(const struct ctx *c);
> -void ndp_unsolicited_na(const struct ctx *c, const struct in6_addr *addr);
> +void ndp_unsolicited_na(const struct ctx *c, int qpair,
> +			const struct in6_addr *addr);
>  
>  #endif /* NDP_H */
> diff --git a/tap.c b/tap.c
> index d098061ed559..a842104687b7 100644
> --- a/tap.c
> +++ b/tap.c
> @@ -125,10 +125,12 @@ unsigned long tap_l2_max_len(const struct ctx *c)
>  /**
>   * tap_send_single() - Send a single frame
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the frame
>   * @data:	Packet buffer
>   * @l2len:	Total L2 packet length
>   */
> -void tap_send_single(const struct ctx *c, const void *data, size_t l2len)
> +void tap_send_single(const struct ctx *c, int qpair, const void *data,
> +		     size_t l2len)
>  {
>  	uint32_t vnet_len = htonl(l2len);
>  	struct iovec iov[2];
> @@ -147,7 +149,7 @@ void tap_send_single(const struct ctx *c, const void *data, size_t l2len)
>  		tap_send_frames(c, iov, iovcnt, 1);
>  		break;
>  	case MODE_VU:
> -		vu_send_single(c, data, l2len);
> +		vu_send_single(c, qpair, data, l2len);
>  		break;
>  	}
>  }
> @@ -250,6 +252,7 @@ void *tap_push_uh4(struct udphdr *uh, struct in_addr src, in_port_t sport,
>  /**
>   * tap_udp4_send() - Send UDP over IPv4 packet
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send packet
>   * @src:	IPv4 source address
>   * @sport:	UDP source port
>   * @dst:	IPv4 destination address
> @@ -257,7 +260,7 @@ void *tap_push_uh4(struct udphdr *uh, struct in_addr src, in_port_t sport,
>   * @in:	UDP payload contents (not including UDP header)
>   * @dlen:	UDP payload length (not including UDP header)
>   */
> -void tap_udp4_send(const struct ctx *c, struct in_addr src, in_port_t sport,
> +void tap_udp4_send(const struct ctx *c, int qpair, struct in_addr src, in_port_t sport,
>  		   struct in_addr dst, in_port_t dport,
>  		   const void *in, size_t dlen)
>  {
> @@ -268,20 +271,22 @@ void tap_udp4_send(const struct ctx *c, struct in_addr src, in_port_t sport,
>  	char *data = tap_push_uh4(uh, src, sport, dst, dport, in, dlen);
>  
>  	memcpy(data, in, dlen);
> -	tap_send_single(c, buf, dlen + (data - buf));
> +	tap_send_single(c, qpair, buf, dlen + (data - buf));
>  }
>  
>  /**
>   * tap_icmp4_send() - Send ICMPv4 packet
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send packet
>   * @src:	IPv4 source address
>   * @dst:	IPv4 destination address
>   * @in:		ICMP packet, including ICMP header
>   * @src_mac:	MAC address to be used as source for message
>   * @l4len:	ICMP packet length, including ICMP header
>   */
> -void tap_icmp4_send(const struct ctx *c, struct in_addr src, struct in_addr dst,
> -		    const void *in, const void *src_mac, size_t l4len)
> +void tap_icmp4_send(const struct ctx *c, int qpair, struct in_addr src,
> +		    struct in_addr dst, const void *in, const void *src_mac,
> +		    size_t l4len)
>  {
>  	char buf[USHRT_MAX];
>  	struct iphdr *ip4h = tap_push_l2h(c, buf, src_mac, ETH_P_IP);
> @@ -291,7 +296,7 @@ void tap_icmp4_send(const struct ctx *c, struct in_addr src, struct in_addr dst,
>  	memcpy(icmp4h, in, l4len);
>  	csum_icmp4(icmp4h, icmp4h + 1, l4len - sizeof(*icmp4h));
>  
> -	tap_send_single(c, buf, l4len + ((char *)icmp4h - buf));
> +	tap_send_single(c, qpair, buf, l4len + ((char *)icmp4h - buf));
>  }
>  
>  /**
> @@ -355,6 +360,7 @@ void *tap_push_uh6(struct udphdr *uh,
>  /**
>   * tap_udp6_send() - Send UDP over IPv6 packet
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send packet
>   * @src:	IPv6 source address
>   * @sport:	UDP source port
>   * @dst:	IPv6 destination address
> @@ -363,7 +369,7 @@ void *tap_push_uh6(struct udphdr *uh,
>   * @in:	UDP payload contents (not including UDP header)
>   * @dlen:	UDP payload length (not including UDP header)
>   */
> -void tap_udp6_send(const struct ctx *c,
> +void tap_udp6_send(const struct ctx *c, int qpair,
>  		   const struct in6_addr *src, in_port_t sport,
>  		   const struct in6_addr *dst, in_port_t dport,
>  		   uint32_t flow, void *in, size_t dlen)
> @@ -376,19 +382,20 @@ void tap_udp6_send(const struct ctx *c,
>  	char *data = tap_push_uh6(uh, src, sport, dst, dport, in, dlen);
>  
>  	memcpy(data, in, dlen);
> -	tap_send_single(c, buf, dlen + (data - buf));
> +	tap_send_single(c, qpair, buf, dlen + (data - buf));
>  }
>  
>  /**
>   * tap_icmp6_send() - Send ICMPv6 packet
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send packet
>   * @src:	IPv6 source address
>   * @dst:	IPv6 destination address
>   * @in:		ICMP packet, including ICMP header
>   * @src_mac:	MAC address to be used as source for message
>   * @l4len:	ICMP packet length, including ICMP header
>   */
> -void tap_icmp6_send(const struct ctx *c,
> +void tap_icmp6_send(const struct ctx *c, int qpair,
>  		    const struct in6_addr *src, const struct in6_addr *dst,
>  		    const void *in, const void *src_mac, size_t l4len)
>  {
> @@ -400,7 +407,7 @@ void tap_icmp6_send(const struct ctx *c,
>  	memcpy(icmp6h, in, l4len);
>  	csum_icmp6(icmp6h, src, dst, icmp6h + 1, l4len - sizeof(*icmp6h));
>  
> -	tap_send_single(c, buf, l4len + ((char *)icmp6h - buf));
> +	tap_send_single(c, qpair, buf, l4len + ((char *)icmp6h - buf));
>  }
>  
>  /**
> @@ -727,7 +734,7 @@ resume:
>  		if (!eh)
>  			continue;
>  		if (ntohs(eh->h_proto) == ETH_P_ARP) {
> -			arp(c, &data);
> +			arp(c, 0, &data);
>  			continue;
>  		}
>  
> @@ -788,7 +795,7 @@ resume:
>  			struct iov_tail eh_data;
>  
>  			packet_get(pool_tap4, i, &eh_data);
> -			if (dhcp(c, &eh_data))
> +			if (dhcp(c, 0, &eh_data))
>  				continue;
>  		}
>  
> @@ -954,7 +961,7 @@ resume:
>  				continue;
>  
>  			ndp_data = data;
> -			if (ndp(c, saddr, &ndp_data))
> +			if (ndp(c, 0, saddr, &ndp_data))
>  				continue;
>  
>  			tap_packet_debug(NULL, ip6h, NULL, proto, NULL, 1);
> @@ -973,7 +980,7 @@ resume:
>  		if (proto == IPPROTO_UDP) {
>  			struct iov_tail uh_data = data;
>  
> -			if (dhcpv6(c, &uh_data, saddr, daddr))
> +			if (dhcpv6(c, 0, &uh_data, saddr, daddr))
>  				continue;
>  		}
>  
> diff --git a/tap.h b/tap.h
> index 1864173cc9b0..92d3e5446991 100644
> --- a/tap.h
> +++ b/tap.h
> @@ -87,24 +87,25 @@ void *tap_push_ip6h(struct ipv6hdr *ip6h,
>  		    const struct in6_addr *src,
>  		    const struct in6_addr *dst,
>  		    size_t l4len, uint8_t proto, uint32_t flow);
> -void tap_udp4_send(const struct ctx *c, struct in_addr src, in_port_t sport,
> +void tap_udp4_send(const struct ctx *c, int qpair, struct in_addr src, in_port_t sport,
>  		   struct in_addr dst, in_port_t dport,
>  		   const void *in, size_t dlen);
> -void tap_icmp4_send(const struct ctx *c, struct in_addr src, struct in_addr dst,
> -		    const void *in, const void *src_mac, size_t l4len);
> +void tap_icmp4_send(const struct ctx *c, int qpair, struct in_addr src,
> +		    struct in_addr dst, const void *in, const void *src_mac,
> +		    size_t l4len);
>  const struct in6_addr *tap_ip6_daddr(const struct ctx *c,
>  				     const struct in6_addr *src);
>  void *tap_push_ip6h(struct ipv6hdr *ip6h,
>  		    const struct in6_addr *src, const struct in6_addr *dst,
>  		    size_t l4len, uint8_t proto, uint32_t flow);
> -void tap_udp6_send(const struct ctx *c,
> +void tap_udp6_send(const struct ctx *c, int qpair,
>  		   const struct in6_addr *src, in_port_t sport,
>  		   const struct in6_addr *dst, in_port_t dport,
>  		   uint32_t flow, void *in, size_t dlen);
> -void tap_icmp6_send(const struct ctx *c,
> +void tap_icmp6_send(const struct ctx *c, int qpair,
>  		    const struct in6_addr *src, const struct in6_addr *dst,
>  		    const void *in, const void *src_mac, size_t l4len);
> -void tap_send_single(const struct ctx *c, const void *data, size_t l2len);
> +void tap_send_single(const struct ctx *c, int qpair, const void *data, size_t l2len);
>  size_t tap_send_frames(const struct ctx *c, const struct iovec *iov,
>  		       size_t bufs_per_frame, size_t nframes);
>  void eth_update_mac(struct ethhdr *eh,
> diff --git a/tcp.c b/tcp.c
> index 3202d3385a63..76f3273bb93f 100644
> --- a/tcp.c
> +++ b/tcp.c
> @@ -1985,6 +1985,7 @@ static void tcp_conn_from_sock_finish(const struct ctx *c,
>  /**
>   * tcp_rst_no_conn() - Send RST in response to a packet with no connection
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the reply
>   * @af:		Address family, AF_INET or AF_INET6
>   * @saddr:	Source address of the packet we're responding to
>   * @daddr:	Destination address of the packet we're responding to
> @@ -1992,7 +1993,7 @@ static void tcp_conn_from_sock_finish(const struct ctx *c,
>   * @th:		TCP header of the packet we're responding to
>   * @l4len:	Packet length, including TCP header
>   */
> -static void tcp_rst_no_conn(const struct ctx *c, int af,
> +static void tcp_rst_no_conn(const struct ctx *c, int qpair, int af,
>  			    const void *saddr, const void *daddr,
>  			    uint32_t flow_lbl,
>  			    const struct tcphdr *th, size_t l4len)
> @@ -2050,7 +2051,7 @@ static void tcp_rst_no_conn(const struct ctx *c, int af,
>  
>  	tcp_update_csum(psum, rsth, &payload);
>  	rst_l2len = ((char *)rsth - buf) + sizeof(*rsth);
> -	tap_send_single(c, buf, rst_l2len);
> +	tap_send_single(c, qpair, buf, rst_l2len);
>  }
>  
>  /**
> @@ -2109,7 +2110,8 @@ int tcp_tap_handler(const struct ctx *c, uint8_t pif, sa_family_t af,
>  			tcp_conn_from_tap(c, af, saddr, daddr, th,
>  					  opts, optlen, now);
>  		else
> -			tcp_rst_no_conn(c, af, saddr, daddr, flow_lbl, th, l4len);
> +			tcp_rst_no_conn(c, 0, af, saddr, daddr, flow_lbl, th,
> +					l4len);
>  		return 1;
>  	}
>  
> diff --git a/udp.c b/udp.c
> index 9c00950250a0..2c74b42a3d95 100644
> --- a/udp.c
> +++ b/udp.c
> @@ -384,13 +384,14 @@ static void udp_tap_prepare(const struct mmsghdr *mmh,
>  /**
>   * udp_send_tap_icmp4() - Construct and send ICMPv4 to local peer
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the ICMPv4 packet
>   * @ee:	Extended error descriptor
>   * @toside:	Destination side of flow
>   * @saddr:	Address of ICMP generating node
>   * @in:	First bytes (max 8) of original UDP message body
>   * @dlen:	Length of the read part of original UDP message body
>   */
> -static void udp_send_tap_icmp4(const struct ctx *c,
> +static void udp_send_tap_icmp4(const struct ctx *c, int qpair,
>  			       const struct sock_extended_err *ee,
>  			       const struct flowside *toside,
>  			       struct in_addr saddr,
> @@ -426,13 +427,14 @@ static void udp_send_tap_icmp4(const struct ctx *c,
>  	/* Try to obtain the MAC address of the generating node */
>  	saddr_any = inany_from_v4(saddr);
>  	fwd_neigh_mac_get(c, &saddr_any, tap_omac);
> -	tap_icmp4_send(c, saddr, eaddr, &msg, tap_omac, msglen);
> +	tap_icmp4_send(c, qpair, saddr, eaddr, &msg, tap_omac, msglen);
>  }
>  
>  
>  /**
>   * udp_send_tap_icmp6() - Construct and send ICMPv6 to local peer
>   * @c:		Execution context
> + * @qpair:	Queue pair on which to send the ICMPv6 packet
>   * @ee:	Extended error descriptor
>   * @toside:	Destination side of flow
>   * @saddr:	Address of ICMP generating node
> @@ -440,7 +442,7 @@ static void udp_send_tap_icmp4(const struct ctx *c,
>   * @dlen:	Length of the read part of original UDP message body
>   * @flow:	IPv6 flow identifier
>   */
> -static void udp_send_tap_icmp6(const struct ctx *c,
> +static void udp_send_tap_icmp6(const struct ctx *c, int qpair,
>  			       const struct sock_extended_err *ee,
>  			       const struct flowside *toside,
>  			       const struct in6_addr *saddr,
> @@ -474,7 +476,7 @@ static void udp_send_tap_icmp6(const struct ctx *c,
>  
>  	/* Try to obtain the MAC address of the generating node */
>  	fwd_neigh_mac_get(c, (union inany_addr *) saddr, tap_omac);
> -	tap_icmp6_send(c, saddr, eaddr, &msg, tap_omac, msglen);
> +	tap_icmp6_send(c, qpair, saddr, eaddr, &msg, tap_omac, msglen);
>  }
>  
>  /**
> @@ -634,12 +636,12 @@ static int udp_sock_recverr(const struct ctx *c, int s, flow_sidx_t sidx,
>  	if (hdr->cmsg_level == IPPROTO_IP &&
>  	    (o4 = inany_v4(&otap)) && inany_v4(&toside->eaddr)) {
>  		dlen = MIN(dlen, ICMP4_MAX_DLEN);
> -		udp_send_tap_icmp4(c, ee, toside, *o4, data, dlen);
> +		udp_send_tap_icmp4(c, 0, ee, toside, *o4, data, dlen);
>  		return 1;
>  	}
>  
>  	if (hdr->cmsg_level == IPPROTO_IPV6 && !inany_v4(&toside->eaddr)) {
> -		udp_send_tap_icmp6(c, ee, toside, &otap.a6, data, dlen,
> +		udp_send_tap_icmp6(c, 0, ee, toside, &otap.a6, data, dlen,
>  				   FLOW_IDX(uflow));
>  		return 1;
>  	}
> diff --git a/vu_common.c b/vu_common.c
> index b13b7c308fd8..040ad067ffbf 100644
> --- a/vu_common.c
> +++ b/vu_common.c
> @@ -235,23 +235,26 @@ void vu_kick_cb(struct vu_dev *vdev, union epoll_ref ref,
>  }
>  
>  /**
> - * vu_send_single() - Send a buffer to the front-end using the RX virtqueue
> + * vu_send_single() - Send a buffer to the front-end using a specified virtqueue
>   * @c:		execution context
> + * @qpair:	Queue pair on which to send the buffer
>   * @buf:	address of the buffer
>   * @size:	size of the buffer
>   *
>   * Return: number of bytes sent, -1 if there is an error
>   */
> -int vu_send_single(const struct ctx *c, const void *buf, size_t size)
> +int vu_send_single(const struct ctx *c, int qpair, const void *buf, size_t size)
>  {
>  	struct vu_dev *vdev = c->vdev;
> -	struct vu_virtq *vq = &vdev->vq[VHOST_USER_RX_QUEUE];
>  	struct vu_virtq_element elem[VIRTQUEUE_MAX_SIZE];
>  	struct iovec in_sg[VIRTQUEUE_MAX_SIZE];
> +	struct vu_virtq *vq;
>  	size_t total;
>  	int elem_cnt;
>  	int i;
>  
> +	vq = &vdev->vq[qpair << 1];
> +
>  	trace("vu_send_single size %zu", size);
>  
>  	if (!vu_queue_enabled(vq) || !vu_queue_started(vq)) {
> diff --git a/vu_common.h b/vu_common.h
> index f538f237790b..25b824c51d1d 100644
> --- a/vu_common.h
> +++ b/vu_common.h
> @@ -56,6 +56,7 @@ void vu_flush(const struct vu_dev *vdev, struct vu_virtq *vq,
>  	      struct vu_virtq_element *elem, int elem_cnt);
>  void vu_kick_cb(struct vu_dev *vdev, union epoll_ref ref,
>  		const struct timespec *now);
> -int vu_send_single(const struct ctx *c, const void *buf, size_t size);
> +int vu_send_single(const struct ctx *c, int qpair, const void *buf,
> +		   size_t size);
>  
>  #endif /* VU_COMMON_H */
> -- 
> 2.51.0
> 

-- 
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 --]

  reply	other threads:[~2025-11-26  3:17 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-11-21 16:58 [PATCH v2 0/6] vhost-user: Add multiqueue support Laurent Vivier
2025-11-21 16:58 ` [PATCH v2 1/6] tap: Remove pool parameter from tap4_handler() and tap6_handler() Laurent Vivier
2025-11-21 16:58 ` [PATCH v2 2/6] vhost-user: Enable multiqueue Laurent Vivier
2025-11-26  2:20   ` David Gibson
2025-11-26  8:04     ` Laurent Vivier
2025-11-21 16:58 ` [PATCH v2 3/6] vhost-user: Add queue pair parameter throughout the network stack Laurent Vivier
2025-11-26  2:35   ` David Gibson [this message]
2025-11-21 16:59 ` [PATCH v2 4/6] tap: Add queue pair parameter throughout the packet processing path Laurent Vivier
2025-11-26  3:16   ` David Gibson
2025-11-21 16:59 ` [PATCH v2 5/6] flow: Add queue pair tracking to flow management Laurent Vivier
2025-11-26  3:41   ` David Gibson
2025-11-27  9:17     ` Laurent Vivier
2025-11-28  0:53       ` David Gibson
2025-11-21 16:59 ` [PATCH v2 6/6] test: Add multiqueue support to vhost-user test infrastructure Laurent Vivier
2025-11-26  3:45   ` David Gibson
2025-11-27  9:20     ` Laurent Vivier
2025-11-28  0:54       ` David Gibson

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=aSZncX2NUHFBqqGu@zatzit \
    --to=david@gibson.dropbear.id.au \
    --cc=lvivier@redhat.com \
    --cc=passt-dev@passt.top \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://passt.top/passt

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for IMAP folder(s).