From: David Gibson <david@gibson.dropbear.id.au>
To: Jon Maloy <jmaloy@redhat.com>
Cc: sbrivio@redhat.com, dgibson@redhat.com, passt-dev@passt.top
Subject: Re: [RFC 12/12] netlink: Rename tap interface when late binding discovers template name
Date: Thu, 18 Dec 2025 16:02:27 +1100 [thread overview]
Message-ID: <aUOK41mlF_7jn-jf@zatzit> (raw)
In-Reply-To: <20251215015441.887736-13-jmaloy@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 4960 bytes --]
On Sun, Dec 14, 2025 at 08:54:41PM -0500, Jon Maloy wrote:
> When pasta starts without a template interface (-I), it creates the
> tap
Template interface is -i not -I.
-I is the guest interface name - that should take precedence if
specified, regardless of late binding or not.
> device with the default name (tap0). Later, when late binding discovers
> the actual template interface, we now rename the device to match it.
>
> This ensures the namespace interface has the expected name, matching
> the discovered host interface.
>
> Key changes:
> - Add nl_link_rename() function to rename network interfaces
> - When late binding triggers with the default TAP name, bring it
> down, rename it to match the discovered interface, then bring it up
>
> Signed-off-by: Jon Maloy <jmaloy@redhat.com>
> ---
> netlink.c | 37 +++++++++++++++++++++++++++++++++++++
> tap.c | 27 +++++++++++++++++++++++++++
> tap.h | 1 +
> 3 files changed, 65 insertions(+)
>
> diff --git a/netlink.c b/netlink.c
> index de04fb7..71089ab 100644
> --- a/netlink.c
> +++ b/netlink.c
> @@ -44,6 +44,8 @@
> /* Default namespace interface name from conf.c */
> extern const char *pasta_default_ifn;
>
> +static int nl_link_rename(int s, unsigned int ifi, const char *name);
> +
> /* Same as RTA_NEXT() but for nexthops: RTNH_NEXT() doesn't take 'attrlen' */
> #define RTNH_NEXT_AND_DEC(rtnh, attrlen) \
> ((attrlen) -= RTNH_ALIGN((rtnh)->rtnh_len), RTNH_NEXT(rtnh))
> @@ -315,6 +317,14 @@ static void nl_linkaddr_host_msg_read(struct ctx *c, const struct nlmsghdr *nh)
> }
> late_binding = true;
>
> + /* Rename interface if it is still using default name */
> + if (is_default && strcmp(ifname, pasta_default_ifn)) {
> + nl_link_set_flags(nl_sock_ns, c->pasta_ifi,
> + 0, IFF_UP);
> + nl_link_rename(nl_sock_ns, c->pasta_ifi, ifname);
> + debug("Renamed tap: %s -> %s",
> + pasta_default_ifn, ifname);
> + }
> if (is_default)
> snprintf(c->pasta_ifn, sizeof(c->pasta_ifn),
> "%s", ifname);
> @@ -1943,6 +1953,33 @@ int nl_link_set_flags(int s, unsigned int ifi,
> return nl_do(s, &req, RTM_NEWLINK, 0, sizeof(req));
> }
>
> +/**
> + * nl_link_rename() - Rename a network interface
> + * @s: Netlink socket
> + * @ifi: Interface index
> + * @name: New interface name
> + *
> + * Return: 0 on success, negative error code on failure
> + */
> +static int nl_link_rename(int s, unsigned int ifi, const char *name)
> +{
> + struct req_t {
> + struct nlmsghdr nlh;
> + struct ifinfomsg ifm;
> + struct rtattr rta;
> + char name[IFNAMSIZ];
> + } req = {
> + .ifm.ifi_family = AF_UNSPEC,
> + .ifm.ifi_index = ifi,
> + .rta.rta_type = IFLA_IFNAME,
> + .rta.rta_len = RTA_LENGTH(IFNAMSIZ),
> + };
> +
> + snprintf(req.name, IFNAMSIZ, "%s", name);
> +
> + return nl_do(s, &req, RTM_NEWLINK, 0, sizeof(req));
> +}
> +
> /**
> * nl_neigh_msg_read() - Interpret a neighbour state message from netlink
> * @c: Execution context
> diff --git a/tap.c b/tap.c
> index a2a4459..cd59160 100644
> --- a/tap.c
> +++ b/tap.c
> @@ -1502,6 +1502,33 @@ static void tap_sock_tun_init(struct ctx *c)
> tap_start_connection(c);
> }
>
> +/**
> + * tap_backend_init_late() - Create tap device for late binding
> + * @c: Execution context
> + *
> + * Called when late binding discovers the template interface name.
> + * Creates the TAP device with the discovered name.
> + */
> +void tap_backend_init_late(struct ctx *c)
Uh... why would you need a second tap device? Also this function
doesn't seem to be called.
> +{
> + if (c->mode != MODE_PASTA || c->fd_tap != -1)
> + return;
> +
> + if (!*c->pasta_ifn) {
> + warn("%s called with empty pasta_ifn", __func__);
> + return;
> + }
> +
> + NS_CALL(tap_ns_tun, c);
> + if (c->fd_tap == -1) {
> + err("Failed to set up tap device in namespace");
> + return;
> + }
> +
> + tap_start_connection(c);
> + info("Created tap device %s for late binding", c->pasta_ifn);
> +}
> +
> /**
> * tap_sock_update_pool() - Set the buffer base and size for the pool of packets
> * @base: Buffer base
> diff --git a/tap.h b/tap.h
> index ee22a9d..b41eae9 100644
> --- a/tap.h
> +++ b/tap.h
> @@ -118,6 +118,7 @@ void tap_handler_passt(struct ctx *c, uint32_t events,
> int tap_sock_unix_open(char *sock_path);
> void tap_sock_reset(struct ctx *c);
> void tap_backend_init(struct ctx *c);
> +void tap_backend_init_late(struct ctx *c);
> void tap_flush_pools(void);
> void tap_handler(struct ctx *c, const struct timespec *now);
> void tap_add_packet(struct ctx *c, struct iov_tail *data,
> --
> 2.51.1
>
--
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 --]
prev parent reply other threads:[~2025-12-18 5:02 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-15 1:54 [RFC 00/12] Support for multiple address and late binding Jon Maloy
2025-12-15 1:54 ` [RFC 01/12] ip: Introduce multi-address data structures for IPv4 and IPv6 Jon Maloy
2025-12-15 9:40 ` David Gibson
2025-12-15 22:05 ` Jon Maloy
2025-12-16 1:58 ` Jon Maloy
2025-12-16 3:14 ` David Gibson
2025-12-15 9:46 ` David Gibson
2025-12-15 1:54 ` [RFC 02/12] ip: Add ip4_default_prefix_len() helper function for class-based prefix Jon Maloy
2025-12-15 9:41 ` David Gibson
2025-12-15 1:54 ` [RFC 03/12] conf: Allow multiple -a/--address options per address family Jon Maloy
2025-12-15 9:53 ` David Gibson
2025-12-15 1:54 ` [RFC 04/12] conf: Apply -n/--netmask to most recently added address Jon Maloy
2025-12-15 9:54 ` David Gibson
2025-12-15 22:43 ` Jon Maloy
2025-12-15 1:54 ` [RFC 05/12] fwd: Check all configured addresses in guest accessibility functions Jon Maloy
2025-12-15 10:06 ` David Gibson
2025-12-15 1:54 ` [RFC 06/12] arp: Check all configured addresses in ARP filtering Jon Maloy
2025-12-15 10:07 ` David Gibson
2025-12-15 1:54 ` [RFC 07/12] netlink: Subscribe to link/address changes in namespace Jon Maloy
2025-12-15 10:32 ` David Gibson
2025-12-15 23:25 ` Jon Maloy
2025-12-16 3:21 ` David Gibson
2025-12-15 1:54 ` [RFC 08/12] netlink: Subscribe to route " Jon Maloy
2025-12-15 10:38 ` David Gibson
2025-12-15 1:54 ` [RFC 09/12] netlink: Add host-side monitoring for late template interface binding Jon Maloy
2025-12-18 4:44 ` David Gibson
2025-12-15 1:54 ` [RFC 10/12] netlink: Add host-side route monitoring and propagation Jon Maloy
2025-12-18 4:53 ` David Gibson
2025-12-15 1:54 ` [RFC 11/12] netlink: Prevent host route events from overwriting guest-configured gateway Jon Maloy
2025-12-18 4:59 ` David Gibson
2025-12-15 1:54 ` [RFC 12/12] netlink: Rename tap interface when late binding discovers template name Jon Maloy
2025-12-18 5:02 ` David Gibson [this message]
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=aUOK41mlF_7jn-jf@zatzit \
--to=david@gibson.dropbear.id.au \
--cc=dgibson@redhat.com \
--cc=jmaloy@redhat.com \
--cc=passt-dev@passt.top \
--cc=sbrivio@redhat.com \
/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).