[P4-dev] Reg: Suggestion on P4 implementation

Huynhtu Dang huynh.tu.dang at usi.ch
Wed Jun 22 08:24:06 EDT 2016

Hi Mohan,

I think you could break GTP-U header into two headers: one with the fields that always exist, and the other is the optional field. Then you could “select” based on the bit value that determines the existence of the optional field. If it is set, then parse the header with optional field(s).

Tu Dang

On Jun 22, 2016, at 2:09 PM, Mohan Krishna <mohankrishna290787 at gmail.com<mailto:mohankrishna290787 at gmail.com>> wrote:


I apologize, I should have been more specific.

GTP-U has a bit and if this bit is set, a optional field is present.  It does not have type and length field present.
I am trying to implement GTP-U using P4. If this has to be done, can someone suggest, how it could be done.

This is for academic purpose, hence any idea on how it could be done now/future could be helpful.


On Tue, Jun 21, 2016 at 8:31 AM, Mohan Krishna <mohankrishna290787 at gmail.com<mailto:mohankrishna290787 at gmail.com>> wrote:

I am trying to find the feasibility of implementing GTP-U tunneling Protocol using P4. The header has a message length, followed by extension headers which are optional. The extension headers are identified using the next extension header field(Not TLV type).

I did not understand how this could be implemented using P4. Please let me know your suggestions regarding the implementation of the same.

+       Bits 0-7        8-23    24-31
0       Message length  Contents
...     ...
...     Contents        Next extension header


P4-dev mailing list
P4-dev at lists.p4.org<mailto:P4-dev at lists.p4.org>

More information about the P4-dev mailing list