Re: [RFC PATCH v1 2/3] test/vsock: add big message test

From: Arseniy Krasnov
Date: Mon Nov 21 2022 - 16:41:01 EST


On 21.11.2022 19:50, Arseniy Krasnov wrote:
> On 21.11.2022 17:52, Stefano Garzarella wrote:
>> On Tue, Nov 15, 2022 at 08:52:35PM +0000, Arseniy Krasnov wrote:
>>> This adds test for sending message, bigger than peer's buffer size.
>>> For SOCK_SEQPACKET socket it must fail, as this type of socket has
>>> message size limit.
>>>
>>> Signed-off-by: Arseniy Krasnov <AVKrasnov@xxxxxxxxxxxxxx>
>>> ---
>>> tools/testing/vsock/vsock_test.c | 62 ++++++++++++++++++++++++++++++++
>>> 1 file changed, 62 insertions(+)
>>>
>>> diff --git a/tools/testing/vsock/vsock_test.c b/tools/testing/vsock/vsock_test.c
>>> index 107c11165887..bb4e8657f1d6 100644
>>> --- a/tools/testing/vsock/vsock_test.c
>>> +++ b/tools/testing/vsock/vsock_test.c
>>> @@ -560,6 +560,63 @@ static void test_seqpacket_timeout_server(const struct test_opts *opts)
>>>     close(fd);
>>> }
>>>
>>> +static void test_seqpacket_bigmsg_client(const struct test_opts *opts)
>>> +{
>>> +    unsigned long sock_buf_size;
>>> +    ssize_t send_size;
>>> +    socklen_t len;
>>> +    void *data;
>>> +    int fd;
>>> +
>>> +    len = sizeof(sock_buf_size);
>>> +
>>> +    fd = vsock_seqpacket_connect(opts->peer_cid, 1234);
>>
>> Not for this patch, but someday we should add a macro for this port and maybe even make it configurable :-)
>>
>>> +    if (fd < 0) {
>>> +        perror("connect");
>>> +        exit(EXIT_FAILURE);
>>> +    }
>>> +
>>> +    if (getsockopt(fd, AF_VSOCK, SO_VM_SOCKETS_BUFFER_SIZE,
>>> +               &sock_buf_size, &len)) {
>>> +        perror("getsockopt");
>>> +        exit(EXIT_FAILURE);
>>> +    }
>>> +
>>> +    sock_buf_size++;
>>> +
>>> +    data = malloc(sock_buf_size);
>>> +    if (!data) {
>>> +        perror("malloc");
>>> +        exit(EXIT_FAILURE);
>>> +    }
>>> +
>>> +    send_size = send(fd, data, sock_buf_size, 0);
>>> +    if (send_size != -1) {
>>
>> Can we check also `errno`?
>> IIUC it should contains EMSGSIZE.
Hm, seems current implementation is a little bit broken and returns ENOMEM, because any negative value, returned by
transport callback is always replaced to ENOMEM. I think i need this patch from Bobby:
https://lore.kernel.org/lkml/d81818b868216c774613dd03641fcfe63cc55a45.1660362668.git.bobby.eshleman@xxxxxxxxxxxxx/
May be i can include it to this patchset also fixing review comments(of course keeping Bobby as author). Or more
simple way is to check ENOMEM instead of EMSGSIZE in this test(simple, but a little bit dumb i think).
>>
>>> +        fprintf(stderr, "expected 'send(2)' failure, got %zi\n",
>>> +            send_size);
>>> +    }
>>> +
>>> +    control_writeln("CLISENT");
>>> +
>>> +    free(data);
>>> +    close(fd);
>>> +}
>>> +
>>> +static void test_seqpacket_bigmsg_server(const struct test_opts *opts)
>>> +{
>>> +    int fd;
>>> +
>>> +    fd = vsock_seqpacket_accept(VMADDR_CID_ANY, 1234, NULL);
>>> +    if (fd < 0) {
>>> +        perror("accept");
>>> +        exit(EXIT_FAILURE);
>>> +    }
>>> +
>>> +    control_expectln("CLISENT");
>>> +
>>> +    close(fd);
>>> +}
>>> +
>>> #define BUF_PATTERN_1 'a'
>>> #define BUF_PATTERN_2 'b'
>>>
>>> @@ -832,6 +889,11 @@ static struct test_case test_cases[] = {
>>>         .run_client = test_seqpacket_timeout_client,
>>>         .run_server = test_seqpacket_timeout_server,
>>>     },
>>> +    {
>>> +        .name = "SOCK_SEQPACKET big message",
>>> +        .run_client = test_seqpacket_bigmsg_client,
>>> +        .run_server = test_seqpacket_bigmsg_server,
>>> +    },
>>
>> I would add new tests always at the end, so if some CI uses --skip, we don't have to update the scripts to skip some tests.
> Ack this and all above
>>
>>>     {
>>>         .name = "SOCK_SEQPACKET invalid receive buffer",
>>>         .run_client = test_seqpacket_invalid_rec_buffer_client,
>>> -- 
>>> 2.25.1
>>
>