Re: [PATCH bpf-next v9 4/4] selftests/bpf: Add mptcpify test

[Yonghong Song <yonghong.song@linux.dev>]

On 8/3/23 7:24 PM, Geliang Tang wrote:
> Hi Yonghong,
> 
> On Thu, Aug 03, 2023 at 06:23:57PM -0700, Yonghong Song wrote:
>>
>>
>> On 8/3/23 6:41 AM, Geliang Tang wrote:
>>> Implement a new test program mptcpify: if the family is AF_INET or
>>> AF_INET6, the type is SOCK_STREAM, and the protocol ID is 0 or
>>> IPPROTO_TCP, set it to IPPROTO_MPTCP. It will be hooked in
>>> update_socket_protocol().
>>>
>>> Extend the MPTCP test base, add a selftest test_mptcpify() for the
>>> mptcpify case. Open and load the mptcpify test prog to mptcpify the
>>> TCP sockets dynamically, then use start_server() and connect_to_fd()
>>> to create a TCP socket, but actually what's created is an MPTCP
>>> socket, which can be verified through the outputs of 'ss' and 'nstat'
>>> commands.
>>>
>>> Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
>>> Signed-off-by: Geliang Tang <geliang.tang@suse.com>
>>> ---
>>>    .../testing/selftests/bpf/prog_tests/mptcp.c  | 94 +++++++++++++++++++
>>>    tools/testing/selftests/bpf/progs/mptcpify.c  | 25 +++++
>>>    2 files changed, 119 insertions(+)
>>>    create mode 100644 tools/testing/selftests/bpf/progs/mptcpify.c
>>>
>>> diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c
>>> index 4407bd5c9e9a..caab3aa6a162 100644
>>> --- a/tools/testing/selftests/bpf/prog_tests/mptcp.c
>>> +++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c
>>> @@ -6,6 +6,7 @@
>>>    #include "cgroup_helpers.h"
>>>    #include "network_helpers.h"
>>>    #include "mptcp_sock.skel.h"
>>> +#include "mptcpify.skel.h"
>>>    char NS_TEST[32];
>>> @@ -195,8 +196,101 @@ static void test_base(void)
>>>    	close(cgroup_fd);
>>>    }
>>> +static void send_byte(int fd)
>>> +{
>>> +	char b = 0x55;
>>> +
>>> +	ASSERT_EQ(write(fd, &b, sizeof(b)), 1, "send single byte");
>>> +}
>>> +
>>> +static int verify_mptcpify(void)
>>> +{
>>> +	char cmd[256];
>>> +	int err = 0;
>>> +
>>> +	snprintf(cmd, sizeof(cmd),
>>> +		 "ip netns exec %s ss -tOni | grep -q '%s'",
>>> +		 NS_TEST, "tcp-ulp-mptcp");
>>
>> Could you show what is the expected output from the above command line
>>    ip netns exec %s ss -tOni
>> ?
>> This way, users can easily reason about the ss states based on tests.
> 
> There're too many items in the output of command 'ip netns exec %s ss -tOni':
> 
> '''
> State Recv-Q Send-Q Local Address:Port  Peer Address:Port Process
> ESTAB 0      0          127.0.0.1:42225    127.0.0.1:44180 cubic wscale:7,7 rto:201 rtt:0.034/0.017 ato:40 mss:16640 pmtu:65535 rcvmss:536 advmss:65483 cwnd:10 bytes_received:1 segs_out:1 segs_in:3 data_segs_in:1 send 39152941176bps lastsnd:7 lastrcv:7 lastack:7 pacing_rate 78305882352bps delivered:1 app_limited rcv_space:33280 rcv_ssthresh:33280 minrtt:0.034 snd_wnd:33280 tcp-ulp-mptcp flags:Mec token:0000(id:0)/3a1e0d3c(id:0) seq:c2802f11c5228db6 sfseq:1 ssnoff:49d3c135 maplen:1
> ESTAB 0      0          127.0.0.1:44180    127.0.0.1:42225 cubic wscale:7,7 rto:201 rtt:0.036/0.02 mss:16640 pmtu:65535 rcvmss:536 advmss:65483 cwnd:10 bytes_sent:1 bytes_acked:2 segs_out:3 segs_in:2 data_segs_out:1 send 36977777778bps lastsnd:7 lastrcv:7 lastack:7 pacing_rate 72200677960bps delivery_rate 8874666664bps delivered:2 rcv_space:33280 rcv_ssthresh:33280 minrtt:0.015 snd_wnd:33280 tcp-ulp-mptcp flags:Mmec token:0000(id:0)/39429ce(id:0) seq:e3ed00de37c805c sfseq:1 ssnoff:d4e4d561 maplen:0
> '''
> 
> We only care about this 'tcp-ulp-mptcp' item.
> 
> Show all output will confuse users. So we just pick and test the only
> item we care.

Thanks. Originally I thought at least we should put one line in
the comment which has 'tcp-ulp-mptcp' like

ESTAB 0      0          127.0.0.1:44180    127.0.0.1:42225 cubic 
wscale:7,7 rto:201 rtt:0.036/0.02 mss:16640 pmtu:65535 rcvmss:536 
advmss:65483 cwnd:10 bytes_sent:1 bytes_acked:2 segs_out:3 segs_in:2 
data_segs_out:1 send 36977777778bps lastsnd:7 lastrcv:7 lastack:7 
pacing_rate 72200677960bps delivery_rate 8874666664bps delivered:2 
rcv_space:33280 rcv_ssthresh:33280 minrtt:0.015 snd_wnd:33280 
tcp-ulp-mptcp flags:Mmec token:0000(id:0)/39429ce(id:0) 
seq:e3ed00de37c805c sfseq:1 ssnoff:d4e4d561 maplen:0

or simplified version

ESTAB 0      0          127.0.0.1:44180    127.0.0.1:42225 cubic
... tcp-ulp-mptcp flags:Mmec ...

But people familiar with 'ss' should be able to dump it and get
the above (maybe without tcp-ulp-mptcp) easily. So I am okay
with no additional comments.

> 
>>
>>> +	if (!ASSERT_OK(system(cmd), "No tcp-ulp-mptcp found!"))
>>> +		err++;
>>> +
>>> +	snprintf(cmd, sizeof(cmd),
>>> +		 "ip netns exec %s nstat -asz %s | awk '%s' | grep -q '%s'",
>>> +		 NS_TEST, "MPTcpExtMPCapableSYNACKRX",
>>> +		 "NR==1 {next} {print $2}", "1");
>>
>> The same thing here. Could you show the expected output with
>>     ip netns exec %s nstat -asz %s
>> ?
> 
> The output of 'ip netns exec %s nstat -asz %s' is:
> 
> '''
> #kernel
> MPTcpExtMPCapableSYNACKRX       1                  0.0
> '''
> 
> The same, we only check if it contains an MPTcpExtMPCapableSYNACKRX, not
> show the output.
> 
> -Geliang
> 
>>
>>> +	if (!ASSERT_OK(system(cmd), "No MPTcpExtMPCapableSYNACKRX found!"))
>>> +		err++;
>>> +
>>> +	return err;
>>> +}
>>> +
>> [...]