Bug 135 – Ipv4 tracing events do not have enough parameters

Bug 135 - Ipv4 tracing events do not have enough parameters

Summary:

Ipv4 tracing events do not have enough parameters

Status:	RESOLVED FIXED

Product:	ns-3
Component:	internet-stack
Version:	pre-release
Platform:	All All

Importance:	P3 enhancement
Assigned To:	ns-bugs

URL:
Keywords:

Depends on:
Blocks:
	Show dependency tree / graph

Reported:	2008-02-07 13:20 EDT by Gustavo J. A. M. Carneiro
Modified:	2008-02-11 10:03 EDT (History)

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note

You need to log in before you can comment on or make changes to this bug.

Description From Gustavo J. A. M. Carneiro 2008-02-07 13:20:13 EDT

Regarding the trace event "/nodes/*/ipv4/drop",

void
Ipv4L3Protocol::SendRealOut (bool found,
                             Ipv4Route const &route,
                             Ptr<Packet> packet,
                             Ipv4Header const &ipHeader)
{
  NS_LOG_FUNCTION;
  NS_LOG_PARAMS (this << found << &route << packet << &ipHeader);

  if (!found)
    {
      NS_LOG_WARN ("No route to host.  Drop.");
      m_dropTrace (packet);
      return;
    }
[...]

In the user callback, only the IPv4 payload is received, but there is not
information regarding the payload type, so I do not know how to decode the
packet (is it UDP, TCP, or...?).

I propose to change the trace callback signature to include also the Ipv4
header:

      m_dropTrace (packet, ipHeader);

------- Comment #1 From Gustavo J. A. M. Carneiro 2008-02-07 13:25:11 EDT -------

Oh, same thing applies for tx and rx events.  Currently it is not possible to
find out what kind of IPv4 packet I am receiving or sending.  The Ipv4Header
would allow that.

PS: the code in question is in src/internet-node/ipv4-l3-protocol.{cc,h}.

------- Comment #2 From Tom Henderson 2008-02-11 00:21:26 EDT -------

is it really the same for tx/rx?  It looks like the header is pushed on before
those trace events are called.

I don't mind expanding the callback but I am curious whether moving 
packet->AddHeader (ipHeader);
 to in front of the dropTrace() call would solve this also?

------- Comment #3 From Gustavo J. A. M. Carneiro 2008-02-11 06:14:27 EDT -------

(In reply to comment #2)
> is it really the same for tx/rx?  It looks like the header is pushed on before
> those trace events are called.

You're right.

> 
> I don't mind expanding the callback but I am curious whether moving 
> packet->AddHeader (ipHeader);
>  to in front of the dropTrace() call would solve this also?
> 

It could.

Thanks for the suggestions.  I think you're right: rx/tx don't need to change;
drop could be simply modified to add the IPv4 header before dropping.  Shall I
go ahead and make this change?

------- Comment #4 From Tom Henderson 2008-02-11 09:40:26 EDT -------

I support the change.

------- Comment #5 From Gustavo J. A. M. Carneiro 2008-02-11 10:03:41 EDT -------

-> http://code.nsnam.org/ns-3-dev/rev/084b07dceafb