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 02/16] flow, tcp: Generalise connection types
Date: Thu, 30 Nov 2023 13:02:08 +1100 [thread overview]
Message-ID: <20231130020222.4056647-3-david@gibson.dropbear.id.au> (raw)
In-Reply-To: <20231130020222.4056647-1-david@gibson.dropbear.id.au>
Currently TCP connections use a 1-bit selector, 'spliced', to determine the
rest of the contents of the structure. We want to generalise the TCP
connection table to other types of flows in other protocols. Make a start
on this by replacing the tcp_conn_common structure with a new flow_common
structure with an enum rather than a simple boolean indicating the type of
flow.
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
Makefile | 8 +++----
flow.c | 18 +++++++++++++++
flow.h | 36 ++++++++++++++++++++++++++++++
tcp.c | 63 +++++++++++++++++++++++++++++++++++++---------------
tcp_conn.h | 24 ++++++--------------
tcp_splice.c | 3 ++-
6 files changed, 112 insertions(+), 40 deletions(-)
create mode 100644 flow.c
create mode 100644 flow.h
diff --git a/Makefile b/Makefile
index 6c53695..e2970e3 100644
--- a/Makefile
+++ b/Makefile
@@ -44,15 +44,15 @@ FLAGS += -DARCH=\"$(TARGET_ARCH)\"
FLAGS += -DVERSION=\"$(VERSION)\"
FLAGS += -DDUAL_STACK_SOCKETS=$(DUAL_STACK_SOCKETS)
-PASST_SRCS = arch.c arp.c checksum.c conf.c dhcp.c dhcpv6.c icmp.c igmp.c \
- isolation.c lineread.c log.c mld.c ndp.c netlink.c packet.c passt.c \
- pasta.c pcap.c port_fwd.c tap.c tcp.c tcp_splice.c udp.c util.c
+PASST_SRCS = arch.c arp.c checksum.c conf.c dhcp.c dhcpv6.c flow.c icmp.c \
+ igmp.c isolation.c lineread.c log.c mld.c ndp.c netlink.c packet.c \
+ passt.c pasta.c pcap.c port_fwd.c tap.c tcp.c tcp_splice.c udp.c util.c
QRAP_SRCS = qrap.c
SRCS = $(PASST_SRCS) $(QRAP_SRCS)
MANPAGES = passt.1 pasta.1 qrap.1
-PASST_HEADERS = arch.h arp.h checksum.h conf.h dhcp.h dhcpv6.h icmp.h \
+PASST_HEADERS = arch.h arp.h checksum.h conf.h dhcp.h dhcpv6.h flow.h icmp.h \
inany.h isolation.h lineread.h log.h ndp.h netlink.h packet.h passt.h \
pasta.h pcap.h pif.h port_fwd.h siphash.h tap.h tcp.h tcp_conn.h \
tcp_splice.h udp.h util.h
diff --git a/flow.c b/flow.c
new file mode 100644
index 0000000..b71358a
--- /dev/null
+++ b/flow.c
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later
+ * Copyright Red Hat
+ * Author: David Gibson <david@gibson.dropbear.id.au>
+ *
+ * Tracking for logical "flows" of packets.
+ */
+
+#include <stdint.h>
+
+#include "flow.h"
+
+const char *flow_type_str[] = {
+ [FLOW_TYPE_NONE] = "<none>",
+ [FLOW_TCP] = "TCP connection",
+ [FLOW_TCP_SPLICE] = "TCP connection (spliced)",
+};
+static_assert(ARRAY_SIZE(flow_type_str) == FLOW_NUM_TYPES,
+ "flow_type_str[] doesn't match enum flow_type");
diff --git a/flow.h b/flow.h
new file mode 100644
index 0000000..351837d
--- /dev/null
+++ b/flow.h
@@ -0,0 +1,36 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later
+ * Copyright Red Hat
+ * Author: David Gibson <david@gibson.dropbear.id.au>
+ *
+ * Tracking for logical "flows" of packets.
+ */
+#ifndef FLOW_H
+#define FLOW_H
+
+/**
+ * enum flow_type - Different types of packet flows we track
+ */
+enum flow_type {
+ /* Represents an invalid or unused flow */
+ FLOW_TYPE_NONE = 0,
+ /* A TCP connection between a socket and tap interface */
+ FLOW_TCP,
+ /* A TCP connection between a host socket and ns socket */
+ FLOW_TCP_SPLICE,
+
+ FLOW_NUM_TYPES,
+};
+
+extern const char *flow_type_str[];
+#define FLOW_TYPE(f) \
+ ((f)->type < FLOW_NUM_TYPES ? flow_type_str[(f)->type] : "?")
+
+/**
+ * struct flow_common - Common fields for packet flows
+ * @type: Type of packet flow
+ */
+struct flow_common {
+ uint8_t type;
+};
+
+#endif /* FLOW_H */
diff --git a/tcp.c b/tcp.c
index 40e3dec..1c25032 100644
--- a/tcp.c
+++ b/tcp.c
@@ -299,6 +299,7 @@
#include "tcp_splice.h"
#include "log.h"
#include "inany.h"
+#include "flow.h"
#include "tcp_conn.h"
@@ -584,7 +585,7 @@ static inline struct tcp_tap_conn *conn_at_idx(int idx)
{
if ((idx < 0) || (idx >= TCP_MAX_CONNS))
return NULL;
- ASSERT(!(CONN(idx)->c.spliced));
+ ASSERT(CONN(idx)->f.type == FLOW_TCP);
return CONN(idx);
}
@@ -1319,14 +1320,21 @@ void tcp_table_compact(struct ctx *c, union tcp_conn *hole)
from = tc + c->tcp.conn_count;
memcpy(hole, from, sizeof(*hole));
- if (from->c.spliced)
- tcp_splice_conn_update(c, &hole->splice);
- else
+ switch (from->f.type) {
+ case FLOW_TCP:
tcp_tap_conn_update(c, &from->tap, &hole->tap);
+ break;
+ case FLOW_TCP_SPLICE:
+ tcp_splice_conn_update(c, &hole->splice);
+ break;
+ default:
+ die("Unexpected %s in tcp_table_compact()",
+ FLOW_TYPE(&from->f));
+ }
- debug("TCP: table compaction (spliced=%d): old index %li, new index %li, "
+ debug("TCP: table compaction (%s): old index %li, new index %li, "
"from: %p, to: %p",
- from->c.spliced, CONN_IDX(from), CONN_IDX(hole),
+ FLOW_TYPE(&from->f), CONN_IDX(from), CONN_IDX(hole),
(void *)from, (void *)hole);
memset(from, 0, sizeof(*from));
@@ -1402,12 +1410,18 @@ void tcp_defer_handler(struct ctx *c)
tcp_l2_data_buf_flush(c);
for (conn = tc + c->tcp.conn_count - 1; conn >= tc; conn--) {
- if (conn->c.spliced) {
- if (conn->splice.flags & CLOSING)
- tcp_splice_destroy(c, conn);
- } else {
+ switch (conn->f.type) {
+ case FLOW_TCP:
if (conn->tap.events == CLOSED)
tcp_conn_destroy(c, conn);
+ break;
+ case FLOW_TCP_SPLICE:
+ if (conn->splice.flags & CLOSING)
+ tcp_splice_destroy(c, conn);
+ break;
+ default:
+ die("Unexpected %s in tcp_defer_handler()",
+ FLOW_TYPE(&conn->f));
}
}
}
@@ -2017,7 +2031,7 @@ static void tcp_conn_from_tap(struct ctx *c,
}
conn = CONN(c->tcp.conn_count++);
- conn->c.spliced = false;
+ conn->f.type = FLOW_TCP;
conn->sock = s;
conn->timer = -1;
conn_event(c, conn, TAP_SYN_RCVD);
@@ -2726,7 +2740,7 @@ static void tcp_tap_conn_from_sock(struct ctx *c,
const struct sockaddr *sa,
const struct timespec *now)
{
- conn->c.spliced = false;
+ conn->f.type = FLOW_TCP;
conn->sock = s;
conn->timer = -1;
conn->ws_to_tap = conn->ws_from_tap = 0;
@@ -2909,10 +2923,17 @@ void tcp_sock_handler(struct ctx *c, union epoll_ref ref, uint32_t events)
{
union tcp_conn *conn = tc + ref.tcp.index;
- if (conn->c.spliced)
- tcp_splice_sock_handler(c, &conn->splice, ref.fd, events);
- else
+ switch (conn->f.type) {
+ case FLOW_TCP:
tcp_tap_sock_handler(c, &conn->tap, events);
+ break;
+ case FLOW_TCP_SPLICE:
+ tcp_splice_sock_handler(c, &conn->splice, ref.fd, events);
+ break;
+ default:
+ die("Unexpected %s in tcp_sock_handler_compact()",
+ FLOW_TYPE(&conn->f));
+ }
}
/**
@@ -3244,11 +3265,17 @@ void tcp_timer(struct ctx *c, const struct timespec *ts)
}
for (conn = tc + c->tcp.conn_count - 1; conn >= tc; conn--) {
- if (conn->c.spliced) {
- tcp_splice_timer(c, conn);
- } else {
+ switch (conn->f.type) {
+ case FLOW_TCP:
if (conn->tap.events == CLOSED)
tcp_conn_destroy(c, conn);
+ break;
+ case FLOW_TCP_SPLICE:
+ tcp_splice_timer(c, conn);
+ break;
+ default:
+ die("Unexpected %s in tcp_timer()",
+ FLOW_TYPE(&conn->f));
}
}
diff --git a/tcp_conn.h b/tcp_conn.h
index 0c6a35b..136f8da 100644
--- a/tcp_conn.h
+++ b/tcp_conn.h
@@ -9,19 +9,9 @@
#ifndef TCP_CONN_H
#define TCP_CONN_H
-/**
- * struct tcp_conn_common - Common fields for spliced and non-spliced
- * @spliced: Is this a spliced connection?
- */
-struct tcp_conn_common {
- bool spliced :1;
-};
-
-extern const char *tcp_common_flag_str[];
-
/**
* struct tcp_tap_conn - Descriptor for a TCP connection (not spliced)
- * @c: Fields common with tcp_splice_conn
+ * @f: Generic flow information
* @in_epoll: Is the connection in the epoll set?
* @next_index: Connection index of next item in hash chain, -1 for none
* @tap_mss: MSS advertised by tap/guest, rounded to 2 ^ TCP_MSS_BITS
@@ -46,8 +36,8 @@ extern const char *tcp_common_flag_str[];
* @seq_init_from_tap: Initial sequence number from tap
*/
struct tcp_tap_conn {
- /* Must be first element to match tcp_splice_conn */
- struct tcp_conn_common c;
+ /* Must be first element */
+ struct flow_common f;
bool in_epoll :1;
int next_index :TCP_CONN_INDEX_BITS + 2;
@@ -121,7 +111,7 @@ struct tcp_tap_conn {
#define SIDES 2
/**
* struct tcp_splice_conn - Descriptor for a spliced TCP connection
- * @c: Fields common with tcp_tap_conn
+ * @f: Generic flow information
* @in_epoll: Is the connection in the epoll set?
* @s: File descriptor for sockets
* @pipe: File descriptors for pipes
@@ -131,8 +121,8 @@ struct tcp_tap_conn {
* @written: Bytes written (not fully written from one other side read)
*/
struct tcp_splice_conn {
- /* Must be first element to match tcp_tap_conn */
- struct tcp_conn_common c;
+ /* Must be first element */
+ struct flow_common f;
bool in_epoll :1;
int s[SIDES];
@@ -168,7 +158,7 @@ struct tcp_splice_conn {
* @splice: Fields specific to spliced connections
*/
union tcp_conn {
- struct tcp_conn_common c;
+ struct flow_common f;
struct tcp_tap_conn tap;
struct tcp_splice_conn splice;
};
diff --git a/tcp_splice.c b/tcp_splice.c
index a5c1332..bfd2bd1 100644
--- a/tcp_splice.c
+++ b/tcp_splice.c
@@ -54,6 +54,7 @@
#include "tcp_splice.h"
#include "siphash.h"
#include "inany.h"
+#include "flow.h"
#include "tcp_conn.h"
@@ -476,7 +477,7 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
if (setsockopt(s, SOL_TCP, TCP_QUICKACK, &((int){ 1 }), sizeof(int)))
trace("TCP (spliced): failed to set TCP_QUICKACK on %i", s);
- conn->c.spliced = true;
+ conn->f.type = FLOW_TCP_SPLICE;
conn->s[0] = s;
if (tcp_splice_new(c, conn, ref.port, ref.pif))
--
@@ -54,6 +54,7 @@
#include "tcp_splice.h"
#include "siphash.h"
#include "inany.h"
+#include "flow.h"
#include "tcp_conn.h"
@@ -476,7 +477,7 @@ bool tcp_splice_conn_from_sock(const struct ctx *c,
if (setsockopt(s, SOL_TCP, TCP_QUICKACK, &((int){ 1 }), sizeof(int)))
trace("TCP (spliced): failed to set TCP_QUICKACK on %i", s);
- conn->c.spliced = true;
+ conn->f.type = FLOW_TCP_SPLICE;
conn->s[0] = s;
if (tcp_splice_new(c, conn, ref.port, ref.pif))
--
2.43.0
next prev parent reply other threads:[~2023-11-30 2:02 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-30 2:02 [PATCH v3 00/16] Introduce unified flow table, first steps David Gibson
2023-11-30 2:02 ` [PATCH v3 01/16] treewide: Add messages to static_assert() calls David Gibson
2023-11-30 2:02 ` David Gibson [this message]
2023-11-30 2:02 ` [PATCH v3 03/16] flow, tcp: Move TCP connection table to unified flow table David Gibson
2023-11-30 2:02 ` [PATCH v3 04/16] flow, tcp: Consolidate flow pointer<->index helpers David Gibson
2023-11-30 2:02 ` [PATCH v3 05/16] util: MAX_FROM_BITS() should be unsigned David Gibson
2023-11-30 2:02 ` [PATCH v3 06/16] flow: Make unified version of flow table compaction David Gibson
2023-11-30 2:02 ` [PATCH v3 07/16] flow, tcp: Add logging helpers for connection related messages David Gibson
2023-11-30 2:02 ` [PATCH v3 08/16] flow: Introduce 'sidx' type to represent one side of one flow David Gibson
2023-12-02 4:35 ` Stefano Brivio
2023-11-30 2:02 ` [PATCH v3 09/16] tcp: Remove unneccessary bounds check in tcp_timer_handler() David Gibson
2023-11-30 2:02 ` [PATCH v3 10/16] flow,tcp: Generalise TCP epoll_ref to generic flows David Gibson
2023-11-30 2:02 ` [PATCH v3 11/16] tcp_splice: Use unsigned to represent side David Gibson
2023-11-30 2:02 ` [PATCH v3 12/16] flow,tcp: Use epoll_ref type including flow and side David Gibson
2023-11-30 2:02 ` [PATCH v3 13/16] test: Avoid hitting guestfish command length limits David Gibson
2023-11-30 2:02 ` [PATCH v3 14/16] pif: Add helpers to get the name of a pif David Gibson
2023-11-30 12:45 ` Stefano Brivio
2023-12-01 0:08 ` David Gibson
2023-11-30 2:02 ` [PATCH v3 15/16] tcp: "TCP" hash secret doesn't need to be TCP specific David Gibson
2023-11-30 2:02 ` [PATCH v3 16/16] tcp: Don't defer hash table removal David Gibson
2023-11-30 12:45 ` Stefano Brivio
2023-12-01 0:07 ` David Gibson
2023-12-02 4:34 ` Stefano Brivio
2023-12-04 0:43 ` David Gibson
2023-12-04 9:54 ` [PATCH v3 00/16] Introduce unified flow table, first steps 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=20231130020222.4056647-3-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).