This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

How to use PA and SA in order to fully offload CPU?

Hi Ti Folks,

In the given forum entry it is said that fully-offloaded PDU processing of the IPSEC-encrypted GTPU data plane is possible. I have doubt how can one configure PA&SA to do RoHC. Please confirm fully-offloaded scheme is possible in case non compression case.

http://e2e.ti.com/support/dsp/c6000_multi-core_dsps/f/639/p/264607/926245.aspx#926245

Burhan.

  • Yes, we only support fully offloaded GTP-U processing with no ROHC or ROHC transparent mode.

    Eric

  • Thank you for reply.

    I also wonder if we can take advantage of PA on tx path(from air to IP) for GTP-U/UDP/IP header bulding. If appending outer IP/UDP and GTP-U header on hardware is not possible software should handle it.

    In addition to this, will you give advice for fast path. Should we use separate QMSS queues for every GTP-U tunnel or one queue for user and a parser on the ingress side? Of course this a bit design specific question but it would be nice to have your opinion.

    Burhan.