mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Krasnov Arseniy Vladimirovich <AVKrasnov@sberdevices.ru>
To: Stefano Garzarella <sgarzare@redhat.com>
Cc: Krasnov Arseniy <oxffffaa@gmail.com>,
	Rokosov Dmitry Dmitrievich <DDRokosov@sberdevices.ru>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"virtualization@lists.linux-foundation.org" 
	<virtualization@lists.linux-foundation.org>,
	"netdev@vger.kernel.org" <netdev@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [RFC PATCH v1 3/3] af_vsock: SOCK_SEQPACKET broken buffer test
Date: Tue, 15 Mar 2022 09:34:35 +0000	[thread overview]
Message-ID: <b452aeac-9628-5e37-e0e6-d33f8bb47b22@sberdevices.ru> (raw)
In-Reply-To: <20220315083617.n33naazzf3se4ozo@sgarzare-redhat>

On 15.03.2022 11:36, Stefano Garzarella wrote:
> On Fri, Mar 11, 2022 at 10:58:32AM +0000, Krasnov Arseniy Vladimirovich wrote:
>> Add test where sender sends two message, each with own
>> data pattern. Reader tries to read first to broken buffer:
>> it has three pages size, but middle page is unmapped. Then,
>> reader tries to read second message to valid buffer. Test
>> checks, that uncopied part of first message was dropped
>> and thus not copied as part of second message.
>>
>> Signed-off-by: Arseniy Krasnov <AVKrasnov@sberdevices.ru>
>> ---
>> tools/testing/vsock/vsock_test.c | 121 +++++++++++++++++++++++++++++++
>> 1 file changed, 121 insertions(+)
>>
>> diff --git a/tools/testing/vsock/vsock_test.c b/tools/testing/vsock/vsock_test.c
>> index aa2de27d0f77..686af712b4ad 100644
>> --- a/tools/testing/vsock/vsock_test.c
>> +++ b/tools/testing/vsock/vsock_test.c
>> @@ -16,6 +16,7 @@
>> #include <linux/kernel.h>
>> #include <sys/types.h>
>> #include <sys/socket.h>
>> +#include <sys/mman.h>
>>
>> #include "timeout.h"
>> #include "control.h"
>> @@ -435,6 +436,121 @@ static void test_seqpacket_timeout_server(const struct test_opts *opts)
>>     close(fd);
>> }
>>
>> +#define BUF_PATTERN_1 'a'
>> +#define BUF_PATTERN_2 'b'
>> +
>> +static void test_seqpacket_invalid_rec_buffer_client(const struct test_opts *opts)
>> +{
>> +    int fd;
>> +    unsigned char *buf1;
>> +    unsigned char *buf2;
>> +    int buf_size = getpagesize() * 3;
>> +
>> +    fd = vsock_seqpacket_connect(opts->peer_cid, 1234);
>> +    if (fd < 0) {
>> +        perror("connect");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    buf1 = malloc(buf_size);
>> +    if (buf1 == NULL) {
>> +        perror("'malloc()' for 'buf1'");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    buf2 = malloc(buf_size);
>> +    if (buf2 == NULL) {
>> +        perror("'malloc()' for 'buf2'");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    memset(buf1, BUF_PATTERN_1, buf_size);
>> +    memset(buf2, BUF_PATTERN_2, buf_size);
>> +
>> +    if (send(fd, buf1, buf_size, 0) != buf_size) {
>> +        perror("send failed");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    if (send(fd, buf2, buf_size, 0) != buf_size) {
>> +        perror("send failed");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    close(fd);
>> +}
>> +
>> +static void test_seqpacket_invalid_rec_buffer_server(const struct test_opts *opts)
>> +{
>> +    int fd;
>> +    unsigned char *broken_buf;
>> +    unsigned char *valid_buf;
>> +    int page_size = getpagesize();
>> +    int buf_size = page_size * 3;
>> +    ssize_t res;
>> +    int prot = PROT_READ | PROT_WRITE;
>> +    int flags = MAP_PRIVATE | MAP_ANONYMOUS;
>> +    int i;
>> +
>> +    fd = vsock_seqpacket_accept(VMADDR_CID_ANY, 1234, NULL);
>> +    if (fd < 0) {
>> +        perror("accept");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    /* Setup first buffer. */
>> +    broken_buf = mmap(NULL, buf_size, prot, flags, -1, 0);
>> +    if (broken_buf == MAP_FAILED) {
>> +        perror("mmap for 'broken_buf'");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    /* Unmap "hole" in buffer. */
>> +    if (munmap(broken_buf + page_size, page_size)) {
>> +        perror("'broken_buf' setup");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    valid_buf = mmap(NULL, buf_size, prot, flags, -1, 0);
>> +    if (valid_buf == MAP_FAILED) {
>> +        perror("mmap for 'valid_buf'");
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    /* Try to fill buffer with unmapped middle. */
>> +    res = read(fd, broken_buf, buf_size);
>> +    if (res != -1) {
>> +        perror("invalid read result of 'broken_buf'");
> 
> if `res` is valid, errno is not set, better to use fprintf(stderr, ...) printing the expected and received result.
> Take a look at test_stream_connection_reset()

Ack, fix it in v2

> 
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    if (errno != ENOMEM) {
>> +        perror("invalid errno of 'broken_buf'");
> 
> Instead of "invalid", I would say "unexpected".

Ack

> 
>> +        exit(EXIT_FAILURE);
>> +    }
> 
> 
>> +
>> +    /* Try to fill valid buffer. */
>> +    res = read(fd, valid_buf, buf_size);
>> +    if (res != buf_size) {
>> +        perror("invalid read result of 'valid_buf'");
> 
> I would split in 2 checks:
> - (res < 0) then use perror()
> - (res != buf_size) then use fprintf(stderr, ...) printing the expected   and received result.

Ack

> 
>> +        exit(EXIT_FAILURE);
>> +    }
>> +
>> +    for (i = 0; i < buf_size; i++) {
>> +        if (valid_buf[i] != BUF_PATTERN_2) {
>> +            perror("invalid pattern for valid buf");
> 
> errno is not set here, better to use fprintf(stderr, ...)

Ack

> 
>> +            exit(EXIT_FAILURE);
>> +        }
>> +    }
> 
> What about replace this for with a memcmp()?

Ack

> 
>> +
>> +
>> +    /* Unmap buffers. */
>> +    munmap(broken_buf, page_size);
>> +    munmap(broken_buf + page_size * 2, page_size);
>> +    munmap(valid_buf, buf_size);
>> +    close(fd);
>> +}
>> +
>> static struct test_case test_cases[] = {
>>     {
>>         .name = "SOCK_STREAM connection reset",
>> @@ -480,6 +596,11 @@ static struct test_case test_cases[] = {
>>         .run_client = test_seqpacket_timeout_client,
>>         .run_server = test_seqpacket_timeout_server,
>>     },
>> +    {
>> +        .name = "SOCK_SEQPACKET invalid receive buffer",
>> +        .run_client = test_seqpacket_invalid_rec_buffer_client,
>> +        .run_server = test_seqpacket_invalid_rec_buffer_server,
>> +    },
> 
> 
> Is this the right behavior? If read() fails because the buffer is invalid, do we throw out the whole packet?
> 
> I was expecting the packet not to be consumed, have you tried AF_UNIX, does it have the same behavior?

I've just checked AF_UNIX implementation of SEQPACKET receive in net/unix/af_unix.c. So, if 'skb_copy_datagram_msg()'
fails, it calls 'skb_free_datagram()'. I think this means that whole sk buff will be dropped, but anyway, i'll check
this behaviour in practice. See '__unix_dgram_recvmsg()' in net/unix/af_unix.c.


> 
> Thanks,
> Stefano
> 


  reply	other threads:[~2022-03-15  9:35 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-11 10:52 [RFC PATCH v1 1/3] af_vsock: add two new tests for SOCK_SEQPACKET Krasnov Arseniy Vladimirovich
2022-03-11 10:55 ` [RFC PATCH v1 2/3] af_vsock: SOCK_SEQPACKET receive timeout test Krasnov Arseniy Vladimirovich
2022-03-15  8:15   ` Stefano Garzarella
2022-03-15  8:35     ` Krasnov Arseniy Vladimirovich
2022-03-11 10:58 ` [RFC PATCH v1 3/3] af_vsock: SOCK_SEQPACKET broken buffer test Krasnov Arseniy Vladimirovich
2022-03-15  8:36   ` Stefano Garzarella
2022-03-15  9:34     ` Krasnov Arseniy Vladimirovich [this message]
2022-03-15 11:49       ` Stefano Garzarella
2022-03-15 12:43       ` Krasnov Arseniy Vladimirovich
2022-03-15 12:55         ` Stefano Garzarella
2022-03-15  8:42 ` [RFC PATCH v1 1/3] af_vsock: add two new tests for SOCK_SEQPACKET Stefano Garzarella
2022-03-15 11:00   ` Krasnov Arseniy Vladimirovich

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=b452aeac-9628-5e37-e0e6-d33f8bb47b22@sberdevices.ru \
    --to=avkrasnov@sberdevices.ru \
    --cc=DDRokosov@sberdevices.ru \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=oxffffaa@gmail.com \
    --cc=sgarzare@redhat.com \
    --cc=virtualization@lists.linux-foundation.org \
    /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.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®