[PATCH bpf-next v9 00/11] use network helpers, part 8

[PATCH bpf-next v9 00/11] use network helpers, part 8

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

v9:
 - new patches 5-7, new struct member expect_errno for network_helper_opts.
 - patches 1-4, 8-9 unchanged.
 - update patches 10-11 to make sure all tests pass.

v8:
 - only patch 8 updated, to fix errors reported by CI.

v7:
 - address Martin's comments in v6. (thanks)
 - use MAX(opts->backlog, 0) instead of opts->backlog.
 - use connect_to_fd_opts instead connect_to_fd.
 - more ASSERT_* to check errors.

v6:
 - update patch 6 as Daniel suggested. (thanks)

v5:
 - keep make_server and make_client as Eduard suggested.

v4:
 - a new patch to use make_sockaddr in sockmap_ktls.
 - a new patch to close fd in error path in drop_on_reuseport.
 - drop make_server() in patch 7.
 - drop make_client() too in patch 9.

v3:
 - a new patch to add backlog for network_helper_opts.
 - use start_server_str in sockmap_ktls now, not start_server.

v2:
 - address Eduard's comments in v1. (thanks)
 - fix errors reported by CI.

This patch set uses network helpers in sockmap_ktls and sk_lookup, and
drop three local helpers tcp_server(), inetaddr_len() and make_socket()
in them.

Geliang Tang (11):
  selftests/bpf: Add backlog for network_helper_opts
  selftests/bpf: Use start_server_str in sockmap_ktls
  selftests/bpf: Use connect_to_fd_opts in sockmap_ktls
  selftests/bpf: Use make_sockaddr in sockmap_ktls
  selftests/bpf: Add network_helper_opts for connect_fd_to_fd
  selftests/bpf: Add expect_errno for network_helper_opts
  selftests/bpf: Set expect_errno for cgroup_skb_sk_lookup
  selftests/bpf: Close fd in error path in drop_on_reuseport
  selftests/bpf: Use start_server_str in sk_lookup
  selftests/bpf: Use connect_to_addr in sk_lookup
  selftests/bpf: Drop make_socket in sk_lookup

 tools/testing/selftests/bpf/network_helpers.c |  23 ++-
 tools/testing/selftests/bpf/network_helpers.h |   8 +-
 .../testing/selftests/bpf/prog_tests/bpf_nf.c |   5 +-
 .../bpf/prog_tests/cgroup_skb_sk_lookup.c     |   8 +-
 .../selftests/bpf/prog_tests/cgroup_tcp_skb.c |   4 +-
 .../selftests/bpf/prog_tests/cgroup_v1v2.c    |   1 +
 .../selftests/bpf/prog_tests/sk_lookup.c      | 152 +++++++-----------
 .../selftests/bpf/prog_tests/sockmap_ktls.c   |  53 ++----
 8 files changed, 106 insertions(+), 148 deletions(-)

-- 
2.43.0

[PATCH bpf-next v9 01/11] selftests/bpf: Add backlog for network_helper_opts

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Some callers expect __start_server() helper to pass their own "backlog"
value to listen() instead of the default of 1. So this patch adds struct
member "backlog" for network_helper_opts to allow callers to set "backlog"
value via start_server_str() helper.

listen(fd, 0 /* backlog */) can be used to enforce syncookie. Meaning
backlog 0 is a legit value.

Using 0 as a default and changing it to 1 here is fine. It makes the test
program easier to write for the common case. Enforcing syncookie mode by
using backlog 0 is a niche use case but it should at least have a way for
the caller to do that.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/bpf/network_helpers.c | 2 +-
 tools/testing/selftests/bpf/network_helpers.h | 4 ++++
 2 files changed, 5 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/bpf/network_helpers.c b/tools/testing/selftests/bpf/network_helpers.c
index 44c2c8fa542a..e0cba4178e41 100644
--- a/tools/testing/selftests/bpf/network_helpers.c
+++ b/tools/testing/selftests/bpf/network_helpers.c
@@ -106,7 +106,7 @@ static int __start_server(int type, const struct sockaddr *addr, socklen_t addrl
 	}
 
 	if (type == SOCK_STREAM) {
-		if (listen(fd, 1) < 0) {
+		if (listen(fd, opts->backlog ? MAX(opts->backlog, 0) : 1) < 0) {
 			log_err("Failed to listed on socket");
 			goto error_close;
 		}
diff --git a/tools/testing/selftests/bpf/network_helpers.h b/tools/testing/selftests/bpf/network_helpers.h
index 9ea36524b9db..4f26bfc2dbf5 100644
--- a/tools/testing/selftests/bpf/network_helpers.h
+++ b/tools/testing/selftests/bpf/network_helpers.h
@@ -25,6 +25,10 @@ struct network_helper_opts {
 	int timeout_ms;
 	bool must_fail;
 	int proto;
+	/* The backlog argument for listen(), defines the maximum length to which
+	 * the queue of pending connections for sockfd may grow.
+	 */
+	int backlog;
 	int (*post_socket_cb)(int fd, void *opts);
 	void *cb_opts;
 };
-- 
2.43.0

[PATCH bpf-next v9 02/11] selftests/bpf: Use start_server_str in sockmap_ktls

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Include network_helpers.h in prog_tests/sockmap_ktls.c, use public network
helper start_server_str() instead of local defined function tcp_server().
This can avoid duplicate code.

Technically, this is not a one-for-one replacement, as start_server_str()
also does bind(). But the difference does not seem to matter.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sockmap_ktls.c   | 23 +++++--------------
 1 file changed, 6 insertions(+), 17 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
index 2d0796314862..32be112967a5 100644
--- a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
+++ b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
@@ -6,25 +6,11 @@
 
 #include <netinet/tcp.h>
 #include "test_progs.h"
+#include "network_helpers.h"
 
 #define MAX_TEST_NAME 80
 #define TCP_ULP 31
 
-static int tcp_server(int family)
-{
-	int err, s;
-
-	s = socket(family, SOCK_STREAM, 0);
-	if (!ASSERT_GE(s, 0, "socket"))
-		return -1;
-
-	err = listen(s, SOMAXCONN);
-	if (!ASSERT_OK(err, "listen"))
-		return -1;
-
-	return s;
-}
-
 static int disconnect(int fd)
 {
 	struct sockaddr unspec = { AF_UNSPEC };
@@ -35,12 +21,15 @@ static int disconnect(int fd)
 /* Disconnect (unhash) a kTLS socket after removing it from sockmap. */
 static void test_sockmap_ktls_disconnect_after_delete(int family, int map)
 {
+	struct network_helper_opts opts = {
+		.backlog = SOMAXCONN,
+	};
 	struct sockaddr_storage addr = {0};
 	socklen_t len = sizeof(addr);
 	int err, cli, srv, zero = 0;
 
-	srv = tcp_server(family);
-	if (srv == -1)
+	srv = start_server_str(family, SOCK_STREAM, NULL, 0, &opts);
+	if (!ASSERT_GE(srv, 0, "start_server_str"))
 		return;
 
 	err = getsockname(srv, (struct sockaddr *)&addr, &len);
-- 
2.43.0

[PATCH bpf-next v9 03/11] selftests/bpf: Use connect_to_fd_opts in sockmap_ktls

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Use public network helper connect_to_fd_opts() instead of open-coding it
in prog_tests/sockmap_ktls.c. This can avoid duplicate code.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sockmap_ktls.c        | 14 ++------------
 1 file changed, 2 insertions(+), 12 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
index 32be112967a5..a794aa688ab5 100644
--- a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
+++ b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
@@ -24,26 +24,16 @@ static void test_sockmap_ktls_disconnect_after_delete(int family, int map)
 	struct network_helper_opts opts = {
 		.backlog = SOMAXCONN,
 	};
-	struct sockaddr_storage addr = {0};
-	socklen_t len = sizeof(addr);
 	int err, cli, srv, zero = 0;
 
 	srv = start_server_str(family, SOCK_STREAM, NULL, 0, &opts);
 	if (!ASSERT_GE(srv, 0, "start_server_str"))
 		return;
 
-	err = getsockname(srv, (struct sockaddr *)&addr, &len);
-	if (!ASSERT_OK(err, "getsockopt"))
-		goto close_srv;
-
-	cli = socket(family, SOCK_STREAM, 0);
-	if (!ASSERT_GE(cli, 0, "socket"))
+	cli = connect_to_fd_opts(srv, SOCK_STREAM, NULL);
+	if (!ASSERT_GE(cli, 0, "connect_to_fd_opts"))
 		goto close_srv;
 
-	err = connect(cli, (struct sockaddr *)&addr, len);
-	if (!ASSERT_OK(err, "connect"))
-		goto close_cli;
-
 	err = bpf_map_update_elem(map, &zero, &cli, 0);
 	if (!ASSERT_OK(err, "bpf_map_update_elem"))
 		goto close_cli;
-- 
2.43.0

[PATCH bpf-next v9 04/11] selftests/bpf: Use make_sockaddr in sockmap_ktls

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

This patch uses public helper make_sockaddr() exported in network_helpers.h
instead of open-coding in sockmap_ktls.c. This can avoid duplicate code.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sockmap_ktls.c      | 16 ++--------------
 1 file changed, 2 insertions(+), 14 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
index a794aa688ab5..3c223601e8b6 100644
--- a/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
+++ b/tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c
@@ -59,23 +59,11 @@ static void test_sockmap_ktls_update_fails_when_sock_has_ulp(int family, int map
 {
 	struct sockaddr_storage addr = {};
 	socklen_t len = sizeof(addr);
-	struct sockaddr_in6 *v6;
-	struct sockaddr_in *v4;
 	int err, s, zero = 0;
 
-	switch (family) {
-	case AF_INET:
-		v4 = (struct sockaddr_in *)&addr;
-		v4->sin_family = AF_INET;
-		break;
-	case AF_INET6:
-		v6 = (struct sockaddr_in6 *)&addr;
-		v6->sin6_family = AF_INET6;
-		break;
-	default:
-		PRINT_FAIL("unsupported socket family %d", family);
+	err = make_sockaddr(family, NULL, 0, &addr, &len);
+	if (!ASSERT_OK(err, "make_sockaddr"))
 		return;
-	}
 
 	s = socket(family, SOCK_STREAM, 0);
 	if (!ASSERT_GE(s, 0, "socket"))
-- 
2.43.0

[PATCH bpf-next v9 05/11] selftests/bpf: Add network_helper_opts for connect_fd_to_fd

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

To make connect_fd_to_fd() more flexible, use a struct network_helper_opts
parameter "opts" instead of timeout_ms. With this change, other opts like
"must_fail" can be used in connect_fd_to_fd() too.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/bpf/network_helpers.c          | 10 +++++++---
 tools/testing/selftests/bpf/network_helpers.h          |  3 ++-
 tools/testing/selftests/bpf/prog_tests/bpf_nf.c        |  5 ++++-
 .../selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c    |  5 ++++-
 .../testing/selftests/bpf/prog_tests/cgroup_tcp_skb.c  |  4 ++--
 5 files changed, 19 insertions(+), 8 deletions(-)

diff --git a/tools/testing/selftests/bpf/network_helpers.c b/tools/testing/selftests/bpf/network_helpers.c
index e0cba4178e41..475a5a04e61e 100644
--- a/tools/testing/selftests/bpf/network_helpers.c
+++ b/tools/testing/selftests/bpf/network_helpers.c
@@ -369,12 +369,16 @@ int connect_to_fd(int server_fd, int timeout_ms)
 	return connect_to_fd_opts(server_fd, type, &opts);
 }
 
-int connect_fd_to_fd(int client_fd, int server_fd, int timeout_ms)
+int connect_fd_to_fd(int client_fd, int server_fd,
+		     const struct network_helper_opts *opts)
 {
 	struct sockaddr_storage addr;
 	socklen_t len = sizeof(addr);
 
-	if (settimeo(client_fd, timeout_ms))
+	if (!opts)
+		opts = &default_opts;
+
+	if (settimeo(client_fd, opts->timeout_ms))
 		return -1;
 
 	if (getsockname(server_fd, (struct sockaddr *)&addr, &len)) {
@@ -382,7 +386,7 @@ int connect_fd_to_fd(int client_fd, int server_fd, int timeout_ms)
 		return -1;
 	}
 
-	if (connect_fd_to_addr(client_fd, &addr, len, false))
+	if (connect_fd_to_addr(client_fd, &addr, len, opts->must_fail))
 		return -1;
 
 	return 0;
diff --git a/tools/testing/selftests/bpf/network_helpers.h b/tools/testing/selftests/bpf/network_helpers.h
index 4f26bfc2dbf5..fcda6b2333ad 100644
--- a/tools/testing/selftests/bpf/network_helpers.h
+++ b/tools/testing/selftests/bpf/network_helpers.h
@@ -66,7 +66,8 @@ int connect_to_addr(int type, const struct sockaddr_storage *addr, socklen_t len
 		    const struct network_helper_opts *opts);
 int connect_to_fd(int server_fd, int timeout_ms);
 int connect_to_fd_opts(int server_fd, int type, const struct network_helper_opts *opts);
-int connect_fd_to_fd(int client_fd, int server_fd, int timeout_ms);
+int connect_fd_to_fd(int client_fd, int server_fd,
+		     const struct network_helper_opts *opts);
 int fastopen_connect(int server_fd, const char *data, unsigned int data_len,
 		     int timeout_ms);
 int make_sockaddr(int family, const char *addr_str, __u16 port,
diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_nf.c b/tools/testing/selftests/bpf/prog_tests/bpf_nf.c
index a4a1f93878d4..cd4dce5bd7f1 100644
--- a/tools/testing/selftests/bpf/prog_tests/bpf_nf.c
+++ b/tools/testing/selftests/bpf/prog_tests/bpf_nf.c
@@ -33,13 +33,16 @@ enum {
 
 static int connect_to_server(int srv_fd)
 {
+	struct network_helper_opts opts = {
+		.timeout_ms = TIMEOUT_MS,
+	};
 	int fd = -1;
 
 	fd = socket(AF_INET, SOCK_STREAM, 0);
 	if (!ASSERT_GE(fd, 0, "socket"))
 		goto out;
 
-	if (!ASSERT_EQ(connect_fd_to_fd(fd, srv_fd, TIMEOUT_MS), 0, "connect_fd_to_fd")) {
+	if (!ASSERT_EQ(connect_fd_to_fd(fd, srv_fd, &opts), 0, "connect_fd_to_fd")) {
 		close(fd);
 		fd = -1;
 	}
diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
index b9dc4ec655b5..bcf142779cc6 100644
--- a/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
@@ -8,6 +8,9 @@
 
 static void run_lookup_test(__u16 *g_serv_port, int out_sk)
 {
+	struct network_helper_opts opts = {
+		.timeout_ms	= 1000,
+	};
 	int serv_sk = -1, in_sk = -1, serv_in_sk = -1, err;
 	struct sockaddr_in6 addr = {};
 	socklen_t addr_len = sizeof(addr);
@@ -24,7 +27,7 @@ static void run_lookup_test(__u16 *g_serv_port, int out_sk)
 	*g_serv_port = addr.sin6_port;
 
 	/* Client outside of test cgroup should fail to connect by timeout. */
-	err = connect_fd_to_fd(out_sk, serv_sk, 1000);
+	err = connect_fd_to_fd(out_sk, serv_sk, &opts);
 	if (CHECK(!err || errno != EINPROGRESS, "connect_fd_to_fd",
 		  "unexpected result err %d errno %d\n", err, errno))
 		goto cleanup;
diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_tcp_skb.c b/tools/testing/selftests/bpf/prog_tests/cgroup_tcp_skb.c
index a1542faf7873..f35a9897db24 100644
--- a/tools/testing/selftests/bpf/prog_tests/cgroup_tcp_skb.c
+++ b/tools/testing/selftests/bpf/prog_tests/cgroup_tcp_skb.c
@@ -86,7 +86,7 @@ static int talk_to_cgroup(int *client_fd, int *listen_fd, int *service_fd,
 	skel->bss->g_sock_port = ntohs(port);
 
 	/* Connect client to server */
-	err = connect_fd_to_fd(*client_fd, *listen_fd, 0);
+	err = connect_fd_to_fd(*client_fd, *listen_fd, NULL);
 	if (!ASSERT_OK(err, "connect_fd_to_fd"))
 		return -1;
 	*service_fd = accept(*listen_fd, NULL, NULL);
@@ -136,7 +136,7 @@ static int talk_to_outside(int *client_fd, int *listen_fd, int *service_fd,
 	skel->bss->g_sock_port = ntohs(port);
 
 	/* Connect client to server */
-	err = connect_fd_to_fd(*client_fd, *listen_fd, 0);
+	err = connect_fd_to_fd(*client_fd, *listen_fd, NULL);
 	if (!ASSERT_OK(err, "connect_fd_to_fd"))
 		return -1;
 	*service_fd = accept(*listen_fd, NULL, NULL);
-- 
2.43.0

[PATCH bpf-next v9 06/11] selftests/bpf: Add expect_errno for network_helper_opts

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

The errno EPERM is skipped in connect_fd_to_addr() by cgroup_v1v2 tests.
More generally, it makes sense to add a "expect_errno" struct member for
network_helper_opts to identify the expect errno to be skipped.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/bpf/network_helpers.c        | 11 +++++++----
 tools/testing/selftests/bpf/network_helpers.h        |  1 +
 tools/testing/selftests/bpf/prog_tests/cgroup_v1v2.c |  1 +
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/tools/testing/selftests/bpf/network_helpers.c b/tools/testing/selftests/bpf/network_helpers.c
index 475a5a04e61e..062170d6be1c 100644
--- a/tools/testing/selftests/bpf/network_helpers.c
+++ b/tools/testing/selftests/bpf/network_helpers.c
@@ -279,7 +279,8 @@ int client_socket(int family, int type,
 
 static int connect_fd_to_addr(int fd,
 			      const struct sockaddr_storage *addr,
-			      socklen_t addrlen, const bool must_fail)
+			      socklen_t addrlen, const bool must_fail,
+			      const int expect_errno)
 {
 	int ret;
 
@@ -290,7 +291,7 @@ static int connect_fd_to_addr(int fd,
 			log_err("Unexpected success to connect to server");
 			return -1;
 		}
-		if (errno != EPERM) {
+		if (errno != expect_errno) {
 			log_err("Unexpected error from connect to server");
 			return -1;
 		}
@@ -318,7 +319,8 @@ int connect_to_addr(int type, const struct sockaddr_storage *addr, socklen_t add
 		return -1;
 	}
 
-	if (connect_fd_to_addr(fd, addr, addrlen, opts->must_fail))
+	if (connect_fd_to_addr(fd, addr, addrlen, opts->must_fail,
+			       opts->expect_errno))
 		goto error_close;
 
 	return fd;
@@ -386,7 +388,8 @@ int connect_fd_to_fd(int client_fd, int server_fd,
 		return -1;
 	}
 
-	if (connect_fd_to_addr(client_fd, &addr, len, opts->must_fail))
+	if (connect_fd_to_addr(client_fd, &addr, len, opts->must_fail,
+			       opts->expect_errno))
 		return -1;
 
 	return 0;
diff --git a/tools/testing/selftests/bpf/network_helpers.h b/tools/testing/selftests/bpf/network_helpers.h
index fcda6b2333ad..14d161d35248 100644
--- a/tools/testing/selftests/bpf/network_helpers.h
+++ b/tools/testing/selftests/bpf/network_helpers.h
@@ -24,6 +24,7 @@ typedef __u16 __sum16;
 struct network_helper_opts {
 	int timeout_ms;
 	bool must_fail;
+	int expect_errno;
 	int proto;
 	/* The backlog argument for listen(), defines the maximum length to which
 	 * the queue of pending connections for sockfd may grow.
diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_v1v2.c b/tools/testing/selftests/bpf/prog_tests/cgroup_v1v2.c
index 9709c8db7275..ff477163f0ea 100644
--- a/tools/testing/selftests/bpf/prog_tests/cgroup_v1v2.c
+++ b/tools/testing/selftests/bpf/prog_tests/cgroup_v1v2.c
@@ -11,6 +11,7 @@ static int run_test(int cgroup_fd, int server_fd, bool classid)
 {
 	struct network_helper_opts opts = {
 		.must_fail = true,
+		.expect_errno = EPERM,
 	};
 	struct connect4_dropper *skel;
 	int fd, err = 0;
-- 
2.43.0

[PATCH bpf-next v9 07/11] selftests/bpf: Set expect_errno for cgroup_skb_sk_lookup

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

EINPROGRESS is skipped in run_lookup_test() in cgroup_skb_sk_lookup
tests, but it is still showed in the log:

 ./test_progs -t cgroup_skb_sk_lookup -v

 run_cgroup_bpf_test:PASS:skel_open_load 0 nsec
 run_cgroup_bpf_test:PASS:cgroup_join 0 nsec
 run_cgroup_bpf_test:PASS:cgroup_attach 0 nsec
 run_lookup_test:PASS:start_server 0 nsec
 run_lookup_test:PASS:getsockname 0 nsec
 (network_helpers.c:300: errno: Operation now in progress) Failed to \
                                                connect to server
 run_lookup_test:PASS:connect_fd_to_fd 0 nsec
 run_lookup_test:PASS:connect_to_fd 0 nsec
 run_lookup_test:PASS:accept 0 nsec
 #51      cgroup_skb_sk_lookup:OK

To fix this, set EINPROGRESS as the expect_errno of network_helper_opts and
pass it to connect_fd_to_fd(). Skip this expect_errno when must_fail is
false too in connect_fd_to_addr().

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/bpf/network_helpers.c                 | 2 +-
 tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/bpf/network_helpers.c b/tools/testing/selftests/bpf/network_helpers.c
index 062170d6be1c..1403487c5921 100644
--- a/tools/testing/selftests/bpf/network_helpers.c
+++ b/tools/testing/selftests/bpf/network_helpers.c
@@ -296,7 +296,7 @@ static int connect_fd_to_addr(int fd,
 			return -1;
 		}
 	} else {
-		if (ret) {
+		if (ret && errno != expect_errno) {
 			log_err("Failed to connect to server");
 			return -1;
 		}
diff --git a/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
index bcf142779cc6..0ef2e9a6f47c 100644
--- a/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/cgroup_skb_sk_lookup.c
@@ -10,6 +10,7 @@ static void run_lookup_test(__u16 *g_serv_port, int out_sk)
 {
 	struct network_helper_opts opts = {
 		.timeout_ms	= 1000,
+		.expect_errno	= EINPROGRESS,
 	};
 	int serv_sk = -1, in_sk = -1, serv_in_sk = -1, err;
 	struct sockaddr_in6 addr = {};
@@ -28,7 +29,7 @@ static void run_lookup_test(__u16 *g_serv_port, int out_sk)
 
 	/* Client outside of test cgroup should fail to connect by timeout. */
 	err = connect_fd_to_fd(out_sk, serv_sk, &opts);
-	if (CHECK(!err || errno != EINPROGRESS, "connect_fd_to_fd",
+	if (CHECK(err || errno != EINPROGRESS, "connect_fd_to_fd",
 		  "unexpected result err %d errno %d\n", err, errno))
 		goto cleanup;
 
-- 
2.43.0

[PATCH bpf-next v9 08/11] selftests/bpf: Close fd in error path in drop_on_reuseport

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Server 1 fd should be closed in the error path when update_lookup_map()
fails. This patch fixes it by goto "close_srv1" instead of "detach"
lable in that case.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 tools/testing/selftests/bpf/prog_tests/sk_lookup.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
index 597d0467a926..de2466547efe 100644
--- a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
@@ -994,7 +994,7 @@ static void drop_on_reuseport(const struct test *t)
 
 	err = update_lookup_map(t->sock_map, SERVER_A, server1);
 	if (err)
-		goto detach;
+		goto close_srv1;
 
 	/* second server on destination address we should never reach */
 	server2 = make_server(t->sotype, t->connect_to.ip, t->connect_to.port,
-- 
2.43.0

[PATCH bpf-next v9 09/11] selftests/bpf: Use start_server_str in sk_lookup

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

This patch uses public helper start_server_addr() instead of local
defined function make_server() in prog_tests/sk_lookup.c to avoid
duplicate code.

Add a helper setsockopts() to set SOL_CUSTOM sockopt looply, set
it to setsockopt pointer of struct network_helper_opts, and pass it
to start_server_addr().

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sk_lookup.c      | 58 +++++++++++--------
 1 file changed, 34 insertions(+), 24 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
index de2466547efe..d03ea3e64a2e 100644
--- a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
@@ -77,6 +77,12 @@ struct test {
 	bool reuseport_has_conns; /* Add a connected socket to reuseport group */
 };
 
+struct cb_opts {
+	int family;
+	int sotype;
+	bool reuseport;
+};
+
 static __u32 duration;		/* for CHECK macro */
 
 static bool is_ipv6(const char *ip)
@@ -142,19 +148,14 @@ static int make_socket(int sotype, const char *ip, int port,
 	return fd;
 }
 
-static int make_server(int sotype, const char *ip, int port,
-		       struct bpf_program *reuseport_prog)
+static int setsockopts(int fd, void *opts)
 {
-	struct sockaddr_storage addr = {0};
+	struct cb_opts *co = (struct cb_opts *)opts;
 	const int one = 1;
-	int err, fd = -1;
-
-	fd = make_socket(sotype, ip, port, &addr);
-	if (fd < 0)
-		return -1;
+	int err = 0;
 
 	/* Enabled for UDPv6 sockets for IPv4-mapped IPv6 to work. */
-	if (sotype == SOCK_DGRAM) {
+	if (co->sotype == SOCK_DGRAM) {
 		err = setsockopt(fd, SOL_IP, IP_RECVORIGDSTADDR, &one,
 				 sizeof(one));
 		if (CHECK(err, "setsockopt(IP_RECVORIGDSTADDR)", "failed\n")) {
@@ -163,7 +164,7 @@ static int make_server(int sotype, const char *ip, int port,
 		}
 	}
 
-	if (sotype == SOCK_DGRAM && addr.ss_family == AF_INET6) {
+	if (co->sotype == SOCK_DGRAM && co->family == AF_INET6) {
 		err = setsockopt(fd, SOL_IPV6, IPV6_RECVORIGDSTADDR, &one,
 				 sizeof(one));
 		if (CHECK(err, "setsockopt(IPV6_RECVORIGDSTADDR)", "failed\n")) {
@@ -172,7 +173,7 @@ static int make_server(int sotype, const char *ip, int port,
 		}
 	}
 
-	if (sotype == SOCK_STREAM) {
+	if (co->sotype == SOCK_STREAM) {
 		err = setsockopt(fd, SOL_SOCKET, SO_REUSEADDR, &one,
 				 sizeof(one));
 		if (CHECK(err, "setsockopt(SO_REUSEADDR)", "failed\n")) {
@@ -181,7 +182,7 @@ static int make_server(int sotype, const char *ip, int port,
 		}
 	}
 
-	if (reuseport_prog) {
+	if (co->reuseport) {
 		err = setsockopt(fd, SOL_SOCKET, SO_REUSEPORT, &one,
 				 sizeof(one));
 		if (CHECK(err, "setsockopt(SO_REUSEPORT)", "failed\n")) {
@@ -190,19 +191,28 @@ static int make_server(int sotype, const char *ip, int port,
 		}
 	}
 
-	err = bind(fd, (void *)&addr, inetaddr_len(&addr));
-	if (CHECK(err, "bind", "failed\n")) {
-		log_err("failed to bind listen socket");
-		goto fail;
-	}
+fail:
+	return err;
+}
 
-	if (sotype == SOCK_STREAM) {
-		err = listen(fd, SOMAXCONN);
-		if (CHECK(err, "make_server", "listen")) {
-			log_err("failed to listen on port %d", port);
-			goto fail;
-		}
-	}
+static int make_server(int sotype, const char *ip, int port,
+		       struct bpf_program *reuseport_prog)
+{
+	struct cb_opts cb_opts = {
+		.family = is_ipv6(ip) ? AF_INET6 : AF_INET,
+		.sotype = sotype,
+		.reuseport = reuseport_prog,
+	};
+	struct network_helper_opts opts = {
+		.backlog	= SOMAXCONN,
+		.post_socket_cb = setsockopts,
+		.cb_opts	= &cb_opts,
+	};
+	int err, fd;
+
+	fd = start_server_str(cb_opts.family, sotype, ip, port, &opts);
+	if (!ASSERT_GE(fd, 0, "start_server_str"))
+		return -1;
 
 	/* Late attach reuseport prog so we can have one init path */
 	if (reuseport_prog) {
-- 
2.43.0

[PATCH bpf-next v9 10/11] selftests/bpf: Use connect_to_addr in sk_lookup

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Add a new parameter expect_errno for make_client(), and use public network
helpers make_sockaddr() + connect_to_addr() in it instead of using local
defined function make_socket() + connect(). This local function can be
dropped latter.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sk_lookup.c      | 46 +++++++++----------
 1 file changed, 23 insertions(+), 23 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
index d03ea3e64a2e..f9b8415c16a6 100644
--- a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
@@ -229,25 +229,27 @@ static int make_server(int sotype, const char *ip, int port,
 	return -1;
 }
 
-static int make_client(int sotype, const char *ip, int port)
+static int make_client(int sotype, const char *ip, int port,
+		       const int expect_errno)
 {
-	struct sockaddr_storage addr = {0};
+	int family = is_ipv6(ip) ? AF_INET6 : AF_INET;
+	struct network_helper_opts opts = {
+		.timeout_ms	= IO_TIMEOUT_SEC,
+		.expect_errno	= expect_errno,
+	};
+	struct sockaddr_storage addr;
+	socklen_t len;
 	int err, fd;
 
-	fd = make_socket(sotype, ip, port, &addr);
-	if (fd < 0)
+	err = make_sockaddr(family, ip, port, &addr, &len);
+	if (!ASSERT_OK(err, "make_sockaddr"))
 		return -1;
 
-	err = connect(fd, (void *)&addr, inetaddr_len(&addr));
-	if (CHECK(err, "make_client", "connect")) {
-		log_err("failed to connect client socket");
-		goto fail;
-	}
+	fd = connect_to_addr(sotype, &addr, len, &opts);
+	if (!ASSERT_GE(fd, 0, "connect_to_addr"))
+		return -1;
 
 	return fd;
-fail:
-	close(fd);
-	return -1;
 }
 
 static __u64 socket_cookie(int fd)
@@ -658,7 +660,7 @@ static void run_lookup_prog(const struct test *t)
 			goto close;
 	}
 
-	client_fd = make_client(t->sotype, t->connect_to.ip, t->connect_to.port);
+	client_fd = make_client(t->sotype, t->connect_to.ip, t->connect_to.port, 0);
 	if (client_fd < 0)
 		goto close;
 
@@ -874,7 +876,6 @@ static void test_redirect_lookup(struct test_sk_lookup *skel)
 
 static void drop_on_lookup(const struct test *t)
 {
-	struct sockaddr_storage dst = {};
 	int client_fd, server_fd, err;
 	struct bpf_link *lookup_link;
 	ssize_t n;
@@ -888,12 +889,12 @@ static void drop_on_lookup(const struct test *t)
 	if (server_fd < 0)
 		goto detach;
 
-	client_fd = make_socket(t->sotype, t->connect_to.ip,
-				t->connect_to.port, &dst);
+	client_fd = make_client(t->sotype, t->connect_to.ip,
+				t->connect_to.port, ECONNREFUSED);
 	if (client_fd < 0)
 		goto close_srv;
 
-	err = connect(client_fd, (void *)&dst, inetaddr_len(&dst));
+	err = -1;
 	if (t->sotype == SOCK_DGRAM) {
 		err = send_byte(client_fd);
 		if (err)
@@ -988,7 +989,6 @@ static void test_drop_on_lookup(struct test_sk_lookup *skel)
 
 static void drop_on_reuseport(const struct test *t)
 {
-	struct sockaddr_storage dst = { 0 };
 	int client, server1, server2, err;
 	struct bpf_link *lookup_link;
 	ssize_t n;
@@ -1012,12 +1012,12 @@ static void drop_on_reuseport(const struct test *t)
 	if (server2 < 0)
 		goto close_srv1;
 
-	client = make_socket(t->sotype, t->connect_to.ip,
-			     t->connect_to.port, &dst);
+	client = make_client(t->sotype, t->connect_to.ip,
+			     t->connect_to.port, ECONNREFUSED);
 	if (client < 0)
 		goto close_srv2;
 
-	err = connect(client, (void *)&dst, inetaddr_len(&dst));
+	err = -1;
 	if (t->sotype == SOCK_DGRAM) {
 		err = send_byte(client);
 		if (err)
@@ -1164,7 +1164,7 @@ static void run_sk_assign_connected(struct test_sk_lookup *skel,
 	if (server_fd < 0)
 		return;
 
-	connected_fd = make_client(sotype, EXT_IP4, EXT_PORT);
+	connected_fd = make_client(sotype, EXT_IP4, EXT_PORT, 0);
 	if (connected_fd < 0)
 		goto out_close_server;
 
@@ -1178,7 +1178,7 @@ static void run_sk_assign_connected(struct test_sk_lookup *skel,
 		goto out_close_connected;
 
 	/* Try to redirect TCP SYN / UDP packet to a connected socket */
-	client_fd = make_client(sotype, EXT_IP4, EXT_PORT);
+	client_fd = make_client(sotype, EXT_IP4, EXT_PORT, 0);
 	if (client_fd < 0)
 		goto out_unlink_prog;
 	if (sotype == SOCK_DGRAM) {
-- 
2.43.0

[PATCH bpf-next v9 11/11] selftests/bpf: Drop make_socket in sk_lookup

[Geliang Tang]

From: Geliang Tang <tanggeliang@kylinos.cn>

Use local helper make_client() in run_multi_prog_lookup() instead of using
make_socket() + connect(). Then make_socket() and inetaddr_len() can be
dropped now.

Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
---
 .../selftests/bpf/prog_tests/sk_lookup.c      | 46 ++-----------------
 1 file changed, 3 insertions(+), 43 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
index f9b8415c16a6..dc84a3f90701 100644
--- a/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
+++ b/tools/testing/selftests/bpf/prog_tests/sk_lookup.c
@@ -108,46 +108,6 @@ static int attach_reuseport(int sock_fd, struct bpf_program *reuseport_prog)
 	return 0;
 }
 
-static socklen_t inetaddr_len(const struct sockaddr_storage *addr)
-{
-	return (addr->ss_family == AF_INET ? sizeof(struct sockaddr_in) :
-		addr->ss_family == AF_INET6 ? sizeof(struct sockaddr_in6) : 0);
-}
-
-static int make_socket(int sotype, const char *ip, int port,
-		       struct sockaddr_storage *addr)
-{
-	struct timeval timeo = { .tv_sec = IO_TIMEOUT_SEC };
-	int err, family, fd;
-
-	family = is_ipv6(ip) ? AF_INET6 : AF_INET;
-	err = make_sockaddr(family, ip, port, addr, NULL);
-	if (CHECK(err, "make_address", "failed\n"))
-		return -1;
-
-	fd = socket(addr->ss_family, sotype, 0);
-	if (CHECK(fd < 0, "socket", "failed\n")) {
-		log_err("failed to make socket");
-		return -1;
-	}
-
-	err = setsockopt(fd, SOL_SOCKET, SO_SNDTIMEO, &timeo, sizeof(timeo));
-	if (CHECK(err, "setsockopt(SO_SNDTIMEO)", "failed\n")) {
-		log_err("failed to set SNDTIMEO");
-		close(fd);
-		return -1;
-	}
-
-	err = setsockopt(fd, SOL_SOCKET, SO_RCVTIMEO, &timeo, sizeof(timeo));
-	if (CHECK(err, "setsockopt(SO_RCVTIMEO)", "failed\n")) {
-		log_err("failed to set RCVTIMEO");
-		close(fd);
-		return -1;
-	}
-
-	return fd;
-}
-
 static int setsockopts(int fd, void *opts)
 {
 	struct cb_opts *co = (struct cb_opts *)opts;
@@ -1227,7 +1187,6 @@ struct test_multi_prog {
 
 static void run_multi_prog_lookup(const struct test_multi_prog *t)
 {
-	struct sockaddr_storage dst = {};
 	int map_fd, server_fd, client_fd;
 	struct bpf_link *link1, *link2;
 	int prog_idx, done, err;
@@ -1260,11 +1219,12 @@ static void run_multi_prog_lookup(const struct test_multi_prog *t)
 	if (err)
 		goto out_close_server;
 
-	client_fd = make_socket(SOCK_STREAM, EXT_IP4, EXT_PORT, &dst);
+	client_fd = make_client(SOCK_STREAM, EXT_IP4, EXT_PORT,
+				t->expect_errno);
 	if (client_fd < 0)
 		goto out_close_server;
 
-	err = connect(client_fd, (void *)&dst, inetaddr_len(&dst));
+	err = t->expect_errno ? -1 : 0;
 	if (CHECK(err && !t->expect_errno, "connect",
 		  "unexpected error %d\n", errno))
 		goto out_close_client;
-- 
2.43.0

Re: [PATCH bpf-next v9 00/11] use network helpers, part 8

[Geliang Tang]

On Sat, 2024-07-06 at 22:45 +0800, Geliang Tang wrote:
> From: Geliang Tang <tanggeliang@kylinos.cn>
> 
> v9:
>  - new patches 5-7, new struct member expect_errno for
> network_helper_opts.
>  - patches 1-4, 8-9 unchanged.
>  - update patches 10-11 to make sure all tests pass.

Superseded, v10 is sent.

Thanks,
-Geliang

> 
> v8:
>  - only patch 8 updated, to fix errors reported by CI.
> 
> v7:
>  - address Martin's comments in v6. (thanks)
>  - use MAX(opts->backlog, 0) instead of opts->backlog.
>  - use connect_to_fd_opts instead connect_to_fd.
>  - more ASSERT_* to check errors.
> 
> v6:
>  - update patch 6 as Daniel suggested. (thanks)
> 
> v5:
>  - keep make_server and make_client as Eduard suggested.
> 
> v4:
>  - a new patch to use make_sockaddr in sockmap_ktls.
>  - a new patch to close fd in error path in drop_on_reuseport.
>  - drop make_server() in patch 7.
>  - drop make_client() too in patch 9.
> 
> v3:
>  - a new patch to add backlog for network_helper_opts.
>  - use start_server_str in sockmap_ktls now, not start_server.
> 
> v2:
>  - address Eduard's comments in v1. (thanks)
>  - fix errors reported by CI.
> 
> This patch set uses network helpers in sockmap_ktls and sk_lookup,
> and
> drop three local helpers tcp_server(), inetaddr_len() and
> make_socket()
> in them.
> 
> Geliang Tang (11):
>   selftests/bpf: Add backlog for network_helper_opts
>   selftests/bpf: Use start_server_str in sockmap_ktls
>   selftests/bpf: Use connect_to_fd_opts in sockmap_ktls
>   selftests/bpf: Use make_sockaddr in sockmap_ktls
>   selftests/bpf: Add network_helper_opts for connect_fd_to_fd
>   selftests/bpf: Add expect_errno for network_helper_opts
>   selftests/bpf: Set expect_errno for cgroup_skb_sk_lookup
>   selftests/bpf: Close fd in error path in drop_on_reuseport
>   selftests/bpf: Use start_server_str in sk_lookup
>   selftests/bpf: Use connect_to_addr in sk_lookup
>   selftests/bpf: Drop make_socket in sk_lookup
> 
>  tools/testing/selftests/bpf/network_helpers.c |  23 ++-
>  tools/testing/selftests/bpf/network_helpers.h |   8 +-
>  .../testing/selftests/bpf/prog_tests/bpf_nf.c |   5 +-
>  .../bpf/prog_tests/cgroup_skb_sk_lookup.c     |   8 +-
>  .../selftests/bpf/prog_tests/cgroup_tcp_skb.c |   4 +-
>  .../selftests/bpf/prog_tests/cgroup_v1v2.c    |   1 +
>  .../selftests/bpf/prog_tests/sk_lookup.c      | 152 +++++++---------
> --
>  .../selftests/bpf/prog_tests/sockmap_ktls.c   |  53 ++----
>  8 files changed, 106 insertions(+), 148 deletions(-)
>