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 v3 13/20] tcp_splice: Make tcp_splice_connect() create its own sockets
Date: Wed, 28 Feb 2024 22:25:13 +1100	[thread overview]
Message-ID: <20240228112520.2078220-14-david@gibson.dropbear.id.au> (raw)
In-Reply-To: <20240228112520.2078220-1-david@gibson.dropbear.id.au>

Currently creating the connected socket for a splice is split between
tcp_splice_conn_from_sock(), which opens the socket, and
tcp_splice_connect() which connects it.  Alter tcp_splice_connect() to
open its own socket based on an address family and pif we pass it.

This does require a second conditional on pif, but makes for a more
logical split of functionality: tcp_splice_conn_from_sock() picks the
target, tcp_splice_connect() creates the connection.  While we're
there improve reporting of errors

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 tcp_splice.c | 31 +++++++++++++++++--------------
 1 file changed, 17 insertions(+), 14 deletions(-)

diff --git a/tcp_splice.c b/tcp_splice.c
index 269265e8..0e5b6524 100644
--- a/tcp_splice.c
+++ b/tcp_splice.c
@@ -91,6 +91,7 @@ static const char *tcp_splice_flag_str[] __attribute((__unused__)) = {
 
 /* Forward declaration */
 static int tcp_sock_refill_ns(void *arg);
+static int tcp_conn_sock_ns(const struct ctx *c, sa_family_t af);
 
 /**
  * tcp_splice_conn_epoll_events() - epoll events masks for given state
@@ -319,13 +320,14 @@ static int tcp_splice_connect_finish(const struct ctx *c,
  * tcp_splice_connect() - Create and connect socket for new spliced connection
  * @c:		Execution context
  * @conn:	Connection pointer
- * @s:		Accepted socket
+ * @af:		Address family
+ * @pif:	pif on which to create socket
  * @port:	Destination port, host order
  *
  * Return: 0 for connect() succeeded or in progress, negative value on error
  */
 static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn,
-			      int sock_conn, in_port_t port)
+			      sa_family_t af, uint8_t pif, in_port_t port)
 {
 	struct sockaddr_in6 addr6 = {
 		.sin6_family = AF_INET6,
@@ -340,7 +342,15 @@ static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn,
 	const struct sockaddr *sa;
 	socklen_t sl;
 
-	conn->s[1] = sock_conn;
+	if (pif == PIF_HOST)
+		conn->s[1] = tcp_conn_sock(c, af);
+	else if (pif == PIF_SPLICE)
+		conn->s[1] = tcp_conn_sock_ns(c, af);
+	else
+		ASSERT(0);
+
+	if (conn->s[1] < 0)
+		return -1;
 
 	if (setsockopt(conn->s[1], SOL_TCP, TCP_QUICKACK,
 		       &((int){ 1 }), sizeof(int))) {
@@ -416,7 +426,7 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
 	struct tcp_splice_conn *conn;
 	union inany_addr src;
 	sa_family_t af;
-	int s1;
+	uint8_t pif1;
 
 	ASSERT(c->mode == MODE_PASTA);
 
@@ -438,23 +448,16 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
 		flow_trace(conn, "failed to set TCP_QUICKACK on %i", s0);
 
 	if (ref.pif == PIF_SPLICE) {
+		pif1 = PIF_HOST;
 		dstport += c->tcp.fwd_out.delta[dstport];
-
-		s1 = tcp_conn_sock(c, af);
 	} else {
 		ASSERT(ref.pif == PIF_HOST);
 
+		pif1 = PIF_SPLICE;
 		dstport += c->tcp.fwd_in.delta[dstport];
-
-		s1 = tcp_conn_sock_ns(c, af);
-	}
-
-	if (s1 < 0) {
-		conn_flag(c, conn, CLOSING);
-		return true;
 	}
 
-	if (tcp_splice_connect(c, conn, s1, dstport))
+	if (tcp_splice_connect(c, conn, af, pif1, dstport))
 		conn_flag(c, conn, CLOSING);
 
 	return true;
-- 
@@ -91,6 +91,7 @@ static const char *tcp_splice_flag_str[] __attribute((__unused__)) = {
 
 /* Forward declaration */
 static int tcp_sock_refill_ns(void *arg);
+static int tcp_conn_sock_ns(const struct ctx *c, sa_family_t af);
 
 /**
  * tcp_splice_conn_epoll_events() - epoll events masks for given state
@@ -319,13 +320,14 @@ static int tcp_splice_connect_finish(const struct ctx *c,
  * tcp_splice_connect() - Create and connect socket for new spliced connection
  * @c:		Execution context
  * @conn:	Connection pointer
- * @s:		Accepted socket
+ * @af:		Address family
+ * @pif:	pif on which to create socket
  * @port:	Destination port, host order
  *
  * Return: 0 for connect() succeeded or in progress, negative value on error
  */
 static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn,
-			      int sock_conn, in_port_t port)
+			      sa_family_t af, uint8_t pif, in_port_t port)
 {
 	struct sockaddr_in6 addr6 = {
 		.sin6_family = AF_INET6,
@@ -340,7 +342,15 @@ static int tcp_splice_connect(const struct ctx *c, struct tcp_splice_conn *conn,
 	const struct sockaddr *sa;
 	socklen_t sl;
 
-	conn->s[1] = sock_conn;
+	if (pif == PIF_HOST)
+		conn->s[1] = tcp_conn_sock(c, af);
+	else if (pif == PIF_SPLICE)
+		conn->s[1] = tcp_conn_sock_ns(c, af);
+	else
+		ASSERT(0);
+
+	if (conn->s[1] < 0)
+		return -1;
 
 	if (setsockopt(conn->s[1], SOL_TCP, TCP_QUICKACK,
 		       &((int){ 1 }), sizeof(int))) {
@@ -416,7 +426,7 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
 	struct tcp_splice_conn *conn;
 	union inany_addr src;
 	sa_family_t af;
-	int s1;
+	uint8_t pif1;
 
 	ASSERT(c->mode == MODE_PASTA);
 
@@ -438,23 +448,16 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
 		flow_trace(conn, "failed to set TCP_QUICKACK on %i", s0);
 
 	if (ref.pif == PIF_SPLICE) {
+		pif1 = PIF_HOST;
 		dstport += c->tcp.fwd_out.delta[dstport];
-
-		s1 = tcp_conn_sock(c, af);
 	} else {
 		ASSERT(ref.pif == PIF_HOST);
 
+		pif1 = PIF_SPLICE;
 		dstport += c->tcp.fwd_in.delta[dstport];
-
-		s1 = tcp_conn_sock_ns(c, af);
-	}
-
-	if (s1 < 0) {
-		conn_flag(c, conn, CLOSING);
-		return true;
 	}
 
-	if (tcp_splice_connect(c, conn, s1, dstport))
+	if (tcp_splice_connect(c, conn, af, pif1, dstport))
 		conn_flag(c, conn, CLOSING);
 
 	return true;
-- 
2.43.2


  parent reply	other threads:[~2024-02-28 11:25 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-28 11:25 [PATCH v3 00/20] More flow table preliminaries: address handling improvements David Gibson
2024-02-28 11:25 ` [PATCH v3 01/20] inany: Helper to test for various address types David Gibson
2024-02-28 11:25 ` [PATCH v3 02/20] inany: Add inany_ntop() helper David Gibson
2024-02-28 11:25 ` [PATCH v3 03/20] inany: Provide more conveniently typed constants for special addresses David Gibson
2024-02-28 11:25 ` [PATCH v3 04/20] inany: Introduce union sockaddr_inany David Gibson
2024-02-28 11:25 ` [PATCH v3 05/20] util: Allow IN4_IS_* macros to operate on untyped addresses David Gibson
2024-02-28 11:25 ` [PATCH v3 06/20] tcp, udp: Don't precompute port remappings in epoll references David Gibson
2024-02-28 11:25 ` [PATCH v3 07/20] flow: Add helper to determine a flow's protocol David Gibson
2024-02-28 11:25 ` [PATCH v3 08/20] tcp_splice: Simplify clean up logic David Gibson
2024-02-28 11:25 ` [PATCH v3 09/20] tcp_splice: Don't use flow_trace() before setting flow type David Gibson
2024-02-28 11:25 ` [PATCH v3 10/20] flow: Clarify flow entry life cycle, introduce uniform logging David Gibson
2024-02-28 11:25 ` [PATCH v3 11/20] tcp_splice: More specific variable names in new splice path David Gibson
2024-02-28 11:25 ` [PATCH v3 12/20] tcp_splice: Merge tcp_splice_new() into its caller David Gibson
2024-02-28 11:25 ` David Gibson [this message]
2024-02-28 11:25 ` [PATCH v3 14/20] tcp_splice: Improve error reporting on connect path David Gibson
2024-02-28 11:25 ` [PATCH v3 15/20] tcp_splice: Improve logic deciding when to splice David Gibson
2024-02-28 11:25 ` [PATCH v3 16/20] tcp, tcp_splice: Parse listening socket epoll ref in tcp_listen_handler() David Gibson
2024-02-28 11:25 ` [PATCH v3 17/20] tcp: Validate TCP endpoint addresses David Gibson
2024-02-28 11:25 ` [PATCH v3 18/20] tap: Disallow loopback addresses on tap interface David Gibson
2024-02-28 11:25 ` [PATCH v3 19/20] port_fwd: Fix copypasta error in port_fwd_scan_udp() comments David Gibson
2024-02-28 11:25 ` [PATCH v3 20/20] fwd: Rename port_fwd.[ch] and their contents David Gibson
2024-02-29 10:53 ` [PATCH v3 00/20] More flow table preliminaries: address handling improvements Stefano Brivio

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=20240228112520.2078220-14-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).