Attention is currently required from: daniel, lynxis lazus.

View Change

6 comments:

• File gtp/gtp.c:

  • Patch Set #11, Line 899: struct

    Im surprised that the destination peer is not 'const'. Likeweise I'd normally expect any othe rinput data like local_ref or ie to be 'const' as the called function shouldn't modify any of these. Probably you're doing it this way as a lot / all of the existing code does it this way?

  • Patch Set #11, Line 1134: int gtp_decode_pdp_ctx(const uint8_t *buf, unsigned int size, struct pdp_t *pdp, uint16_t *sapi)

    I think the function could deserve some documentation as a comment [doxygen style or not]

  • Patch Set #11, Line 1273: int gtp_encode_pdp_ctx(uint8_t *buf, unsigned int size, const struct pdp_t *pdp, uint16_t sapi)

    I think the function could deserve some documentation as a comment [doxygen style or not]

  • Patch Set #11, Line 1403: LOGP(DLGTP, LOGL_NOTICE,

    shouldn't we print some kind of context? Like which GSN peer (address) has sent us the message with the wrong version?

  • Patch Set #11, Line 1410: "SGSN Context Response expected Sequence flag set. Flags %d\n", packet->flags);

    shouldn't we print some kind of context? Like which GSN peer (address) has sent us the message with the wrong flags?

  • Patch Set #11, Line 1458: "SGSN Context Request expected only on GTPCv1: %u\n", version);

    same here

To view, visit change 38938. To unsubscribe, or for help writing mail filters, visit settings.