netdev - Function to get bytes/s received on a socket

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>] [day] [month] [year] [list]

Message-ID: <CABzZ0PHPshVp6+8skcawTaW64=kwxwFiWpaDCWpKhnYvm2ugKg@mail.gmail.com>
Date:	Wed, 29 Feb 2012 18:41:26 +0100
From:	Riccardo Trebbi <riccardo.trebbi89@...il.com>
To:	netdev@...r.kernel.org
Subject: Function to get bytes/s received on a socket

Hi all,

I am working on a project for the university and I needed a function
that could calculate the throughput of a socket in terms of bytes/s
received.

I followed the line of  tcp_rcv_space_adjust() in net/ipv4/tcp_input.c
 and I did something like this.

First I modified the struct tcp_sock to add this:

+++++++++++++++++++++++++++++++++++++++++++++++++
/* Throughput estimation of the receiver */
    struct {
        u32 time;   // time we got the seq
        u32 seq;    // seq number we expected at time
        u32 thrpt;   // last value of throughput
    } thrpt_est;
+++++++++++++++++++++++++++++++++++++++++++++++++

Then I implemented this function in net/ipv4/tcp_input.c:

+++++++++++++++++++++++++++++++++++++++++++++++++

/* This function is used to calculate the throughput of a
 * tcp_sock that is receiveing data */
void my_function(struct sock *sk)
{
    struct tcp_sock *tp = tcp_sk(sk);
u32 time;
u32 rcvd_bytes;

if (tp->thrpt_est.time == 0 || tp->thrpt_est.seq == 0)
goto new_measure;

time = tcp_time_stamp - tp->thrpt_est.time;

if(time < 100)
        return;

    if(tp->rcv_nxt < tp->thrpt_est.seq)
        printk("tp->rcv_nxt < tp->thrpt_est.seq\n");

    rcvd_bytes = tp->rcv_nxt - tp->thrpt_est.seq;

    if(time == 0)
        goto new_measure;

    tp->thrpt_est.thrpt = rcvd_bytes / time;

new_measure:
tp->thrpt_est.seq = tp->rcv_nxt;
tp->thrpt_est.time = tcp_time_stamp;
}

+++++++++++++++++++++++++++++++++++++++++++++++++

The my_function() is called when tcp_rcv_space_adjust() is called.

The value in tp->thrpt_est.thrpt is bytes/ms but I can get the bytes/s
with a simple conversion.

I am a total newbie in kernel hacking and there's so many things I don't know,
but I thought that this was quite simple and should have worked.

The problem is the value in tp->thrpt_est.thrpt is way higher than the
real throughput.
I get values 4-5 times higher than the reality.

Would you please help me understand?

Thanks,

Riccardo
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html