[net2-wg] tos msg vs ADU send for collection

Rodrigo Fonseca rfonseca at cs.berkeley.edu
Fri Jan 13 15:54:08 PST 2006


I think the more important point is zero-copying policy that exposing
TOS_Msg allows. As Phil pointed out, after you depart from the TOS_Msg
for sending, you need to start copying data at least once.

It also has an implication on storage: if the application has the
TOS_Msgs allocated and passes uses that for sending, then the stack
doesn't need to allocate extra message buffers. On the other hand, if
the application has the data in a buffer somewhere, this data needs to
be copied to a message by the stack somewhere. Who has these buffers,
and how many are they?

I would say it is slightly less convenient for application writers to
have the TOS_Msgs instead of just worrying about data buffers, but
there are some efficiency benefits.

I would say that some applications might be happy with just the
TOS_Msg, and if needed be we could add a layer doing the packaging of
data into packets as a higher level primitive. This packaging can even
be shared with other forms of network protocol sending, and not only
collection.

Thoughts?

Rodrigo


On 1/13/06, Omprakash Gnawali <gnawali at usc.edu> wrote:
>
> This is a follow up on a specific thread from our meeting
> yesterday. It was mentioned that providing tos msg send will enable
> implementing ADU send on top of it but not necessarily the other way
> around.
>
> It is not clear why this is the case. One can call the ADU send with a
> tos msg as an argument; the ADU send which will package it inside its
> own tos msg. The argument why this does not work is because there are
> headers in a tos msg. As far as I can tell, collection applications do
> not care about the header in a tos msg. Users of collection should
> only care about data.
>
> Are there instances when users of collection care about the header ?
> If there are instances of this then collection must provide a tos msg
> send interface.
>
> - om_p
> _______________________________________________
> net2-wg mailing list
> net2-wg at millennium.berkeley.edu
> https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/net2-wg
>




More information about the net2-wg mailing list