[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [ssm] Igmpv3 - Experiment-1



Hi,

> - I want to know also whether this full state igmpv3  support is ethernet
> card
> vendor specific,means some vendor cards have this support and
> some may not.
> like 3c59x driver may have and 8139too realtek driver .does not have.
>

In modern up-to-date Operating Systems usually the Ethernet-Frame
is built in Software, means by a loadable driver. Therefore (Ethernet)
Hardware provides only a MAC-Interface (Medium Access Control).

Example:
You want to transmit a frame, let us say an RTP/UDP/IP frame.
Usually you built an application where it is your task to build
an RTP frame including RTP header. Then you ask the socket
to establish an UDP (Datagram) Socket and put this self-built frame
to the socket (so far nothing new, I assume).
Then usually the TCP/IP Kernel Driver catches this frame and adds
in this case the UDP header. After that the IP header is built in front.
Now nearly all Operating Systems have built-in a tool with the meaning
"Protocol Manager".
This Protocol Manager manages now that your IP frame is
diverted to the correct netcard. The netcard provides access by the MAC
interface.
(In MS Windows OS this is a part of the NDIS (Network Describing
Interface Specification) interface, in Novell this interface is called ODI,
for Linux
I need a hint.)
Usually the netcard driver adds the protocol specific headers and the
netcard hardware adds the necessary checksums and introductions (PLLs...).

That was a long story. The result: In the last 15 years the netcards
provided more or less a raw access. It is easier to exchange a piece of
software
than exchanging a monolithic netcard.
So, IGMP is a topic of the IP stack and is not influenced by your netcard.
(Of course there could be bugs in the software that do not initialize the
netcard correct or set filters for the several frames wrong or ....)

If I am wrong, please correct me.

Michael


_______________________________________________
ssm mailing list
ssm@ietf.org
https://www1.ietf.org/mailman/listinfo/ssm