public inbox for passt-dev@passt.top
 help / color / mirror / code / Atom feed
From: David Gibson <david@gibson.dropbear.id.au>
To: passt-dev@passt.top, Stefano Brivio <sbrivio@redhat.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Subject: [PATCH 3/3] tcp, udp: Bind outbound listening sockets by interface instead of address
Date: Fri, 17 Oct 2025 11:34:47 +1100	[thread overview]
Message-ID: <20251017003447.414103-4-david@gibson.dropbear.id.au> (raw)
In-Reply-To: <20251017003447.414103-1-david@gibson.dropbear.id.au>

Currently, outbound forwards (-T, -U) are handled by sockets bound to the
loopback address.  Typically we create two sockets, one for 127.0.0.1 and
one for ::1.

This has some disadvantages:
 * The guest can't connect to these services using its global IP address,
   it must explicitly use 127.0.0.1 or ::1 (bug 100)
 * The guest can't even connect via 127.0.0.0/8 addresses other than
   127.0.0.1
 * We can't use dual-stack sockets, we have to have separate sockets for
   IPv4 and IPv6.

The restriction exist for a reason though.  If the guest has any interfaces
other than pasta (e.g. a VPN tunnel) external hosts could reach the host
via the forwards.  Especially combined with -T auto / -U auto this would
make it very easy to make a mistake with nasty security implications.

We can achieve both goals, however, if we don't bind the outbound listening
sockets to a particular address, but _do_ use SO_BINDTODEVICE to restrict
them to the "lo" interface.

Link: https://bugs.passt.top/show_bug.cgi?id=100

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 pif.c |  6 ------
 tcp.c | 18 ++----------------
 udp.c | 27 ++++++++++-----------------
 3 files changed, 12 insertions(+), 39 deletions(-)

diff --git a/pif.c b/pif.c
index 592fafaa..84e3ceae 100644
--- a/pif.c
+++ b/pif.c
@@ -87,12 +87,6 @@ int pif_sock_l4(const struct ctx *c, enum epoll_type type, uint8_t pif,
 
 	ASSERT(pif_is_socket(pif));
 
-	if (pif == PIF_SPLICE) {
-		/* Sanity checks */
-		ASSERT(!ifname);
-		ASSERT(addr && inany_is_loopback(addr));
-	}
-
 	if (!addr)
 		return sock_l4_sa(c, type, &sa, sizeof(sa.sa6),
 				  ifname, false, data);
diff --git a/tcp.c b/tcp.c
index 15c012d7..982c9190 100644
--- a/tcp.c
+++ b/tcp.c
@@ -2592,20 +2592,6 @@ int tcp_sock_init(const struct ctx *c, uint8_t pif,
 
 	return r4 < 0 ? r4 : r6;
 }
-/**
- * tcp_ns_sock_init() - Init socket to listen for spliced outbound connections
- * @c:		Execution context
- * @port:	Port, host order
- */
-static void tcp_ns_sock_init(const struct ctx *c, in_port_t port)
-{
-	ASSERT(!c->no_tcp);
-
-	if (c->ifi4)
-		tcp_sock_init_one(c, PIF_SPLICE, &inany_loopback4, NULL, port);
-	if (c->ifi6)
-		tcp_sock_init_one(c, PIF_SPLICE, &inany_loopback6, NULL, port);
-}
 
 /**
  * tcp_ns_socks_init() - Bind sockets in namespace for outbound connections
@@ -2625,7 +2611,7 @@ static int tcp_ns_socks_init(void *arg)
 		if (!bitmap_isset(c->tcp.fwd_out.map, port))
 			continue;
 
-		tcp_ns_sock_init(c, port);
+		tcp_sock_init(c, PIF_SPLICE, NULL, "lo", port);
 	}
 
 	return 0;
@@ -2805,7 +2791,7 @@ static void tcp_port_rebind(struct ctx *c, bool outbound)
 		if ((c->ifi4 && socks[port][V4] == -1) ||
 		    (c->ifi6 && socks[port][V6] == -1)) {
 			if (outbound)
-				tcp_ns_sock_init(c, port);
+				tcp_sock_init(c, PIF_SPLICE, NULL, "lo", port);
 			else
 				tcp_sock_init(c, PIF_HOST, NULL, NULL, port);
 		}
diff --git a/udp.c b/udp.c
index 49dd0144..e38114eb 100644
--- a/udp.c
+++ b/udp.c
@@ -1127,26 +1127,16 @@ int udp_sock_init(const struct ctx *c, uint8_t pif,
 	}
 
 	if ((!addr || inany_v4(addr)) && c->ifi4) {
-		const union inany_addr *a = addr ?
-			addr : &inany_any4;
-
-		if (pif == PIF_SPLICE)
-			a = &inany_loopback4;
-
-		r4 = pif_sock_l4(c, EPOLL_TYPE_UDP_LISTEN, pif, a, ifname,
+		r4 = pif_sock_l4(c, EPOLL_TYPE_UDP_LISTEN, pif,
+				 addr ? addr : &inany_any4, ifname,
 				 port, uref.u32);
 
 		socks[V4][port] = r4 < 0 ? -1 : r4;
 	}
 
 	if ((!addr || !inany_v4(addr)) && c->ifi6) {
-		const union inany_addr *a = addr ?
-			addr : &inany_any6;
-
-		if (pif == PIF_SPLICE)
-			a = &inany_loopback6;
-
-		r6 = pif_sock_l4(c, EPOLL_TYPE_UDP_LISTEN, pif, a, ifname,
+		r6 = pif_sock_l4(c, EPOLL_TYPE_UDP_LISTEN, pif,
+				 addr ? addr : &inany_any6, ifname,
 				 port, uref.u32);
 
 		socks[V6][port] = r6 < 0 ? -1 : r6;
@@ -1214,9 +1204,12 @@ static void udp_port_rebind(struct ctx *c, bool outbound)
 			continue;
 
 		if ((c->ifi4 && socks[V4][port] == -1) ||
-		    (c->ifi6 && socks[V6][port] == -1))
-			udp_sock_init(c, outbound ? PIF_SPLICE : PIF_HOST,
-				      NULL, NULL, port);
+		    (c->ifi6 && socks[V6][port] == -1)) {
+			if (outbound)
+				udp_sock_init(c, PIF_SPLICE, NULL, "lo", port);
+			else
+				udp_sock_init(c, PIF_HOST, NULL, NULL, port);
+		}
 	}
 }
 
-- 
2.51.0


  parent reply	other threads:[~2025-10-17  0:34 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-17  0:34 [PATCH 0/3] RFC: Reduce differences between inbound and outbound socket binding David Gibson
2025-10-17  0:34 ` [PATCH 1/3] tcp: Merge tcp_ns_sock_init[46]() into tcp_sock_init_one() David Gibson
2025-10-20  6:08   ` Stefano Brivio
2025-10-20  9:24     ` David Gibson
2025-10-20  6:09   ` Stefano Brivio
2025-10-20  9:25     ` David Gibson
2025-10-17  0:34 ` [PATCH 2/3] udp: Unify some more inbound/outbound parts of udp_sock_init() David Gibson
2025-10-21 21:51   ` Stefano Brivio
2025-10-22  0:08     ` David Gibson
2025-10-17  0:34 ` David Gibson [this message]
2025-10-21 21:51   ` [PATCH 3/3] tcp, udp: Bind outbound listening sockets by interface instead of address Stefano Brivio
2025-10-22  0:34     ` David Gibson
2025-10-22  8:59       ` Stefano Brivio
2025-10-23  1:18         ` 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=20251017003447.414103-4-david@gibson.dropbear.id.au \
    --to=david@gibson.dropbear.id.au \
    --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).