netdev - Re: [PATCHv2 bpf-next] selftests/bpf: run mptcp in a dedicated netns

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 22 Feb 2023 15:44:17 -0800
From:   Martin KaFai Lau <martin.lau@...ux.dev>
To:     Hangbin Liu <liuhangbin@...il.com>
Cc:     Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andrii@...nel.org>, bpf@...r.kernel.org,
        Jakub Kicinski <kuba@...nel.org>,
        "David S . Miller" <davem@...emloft.net>,
        Song Liu <song@...nel.org>, Yonghong Song <yhs@...com>,
        John Fastabend <john.fastabend@...il.com>,
        KP Singh <kpsingh@...nel.org>,
        Stanislav Fomichev <sdf@...gle.com>,
        Hao Luo <haoluo@...gle.com>, Jiri Olsa <jolsa@...nel.org>,
        Mykola Lysenko <mykolal@...com>,
        Felix Maurer <fmaurer@...hat.com>,
        Matthieu Baerts <matthieu.baerts@...sares.net>,
        mptcp@...ts.linux.dev, netdev@...r.kernel.org
Subject: Re: [PATCHv2 bpf-next] selftests/bpf: run mptcp in a dedicated netns

On 2/18/23 11:01 PM, Hangbin Liu wrote:
> The current mptcp test is run in init netns. If the user or default
> system config disabled mptcp, the test will fail. Let's run the mptcp
> test in a dedicated netns to avoid none kernel default mptcp setting.
> 
> Suggested-by: Martin KaFai Lau <martin.lau@...ux.dev>
> Signed-off-by: Hangbin Liu <liuhangbin@...il.com>
> ---
> v2: remove unneed close_cgroup_fd goto label.
> ---
>   .../testing/selftests/bpf/prog_tests/mptcp.c  | 27 +++++++++++++++++--
>   1 file changed, 25 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c
> index 59f08d6d1d53..dbe2bcfd3b38 100644
> --- a/tools/testing/selftests/bpf/prog_tests/mptcp.c
> +++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c
> @@ -7,6 +7,16 @@
>   #include "network_helpers.h"
>   #include "mptcp_sock.skel.h"
>   
> +#define SYS(fmt, ...)						\
> +	({							\
> +		char cmd[1024];					\
> +		snprintf(cmd, sizeof(cmd), fmt, ##__VA_ARGS__);	\
> +		if (!ASSERT_OK(system(cmd), cmd))		\
> +			goto fail;				\
> +	})
> +
> +#define NS_TEST "mptcp_ns"
> +
>   #ifndef TCP_CA_NAME_MAX
>   #define TCP_CA_NAME_MAX	16
>   #endif
> @@ -138,12 +148,20 @@ static int run_test(int cgroup_fd, int server_fd, bool is_mptcp)
>   
>   static void test_base(void)
>   {
> +	struct nstoken *nstoken = NULL;
>   	int server_fd, cgroup_fd;
>   
>   	cgroup_fd = test__join_cgroup("/mptcp");
>   	if (!ASSERT_GE(cgroup_fd, 0, "test__join_cgroup"))
>   		return;
>   
> +	SYS("ip netns add %s", NS_TEST);
> +	SYS("ip -net %s link set dev lo up", NS_TEST);
> +
> +	nstoken = open_netns(NS_TEST);
> +	if (!ASSERT_OK_PTR(nstoken, "open_netns"))
> +		goto fail;
> +
>   	/* without MPTCP */
>   	server_fd = start_server(AF_INET, SOCK_STREAM, NULL, 0, 0);
>   	if (!ASSERT_GE(server_fd, 0, "start_server"))
> @@ -157,13 +175,18 @@ static void test_base(void)
>   	/* with MPTCP */
>   	server_fd = start_mptcp_server(AF_INET, NULL, 0, 0);
>   	if (!ASSERT_GE(server_fd, 0, "start_mptcp_server"))
> -		goto close_cgroup_fd;
> +		goto fail;
>   
>   	ASSERT_OK(run_test(cgroup_fd, server_fd, true), "run_test mptcp");
>   
>   	close(server_fd);
>   
> -close_cgroup_fd:
> +fail:
> +	if (nstoken)
> +		close_netns(nstoken);
> +
> +	system("ip netns del " NS_TEST " >& /dev/null");

It needs to be "&>", like the fix in commit 98e13848cf43 ("selftests/bpf: Fix 
decap_sanity_ns cleanup").

Since it needs to respin, could you help and take this chance to put the above 
SYS() macro into the test_progs.h. Other selftests are doing similar thing also. 
If possible, it may be easier to have a configurable "goto_label" as the first arg.