bgpd: Ignore repeated SRv6 SID Information Sub-TLVs

[cscarpitta]

RFC 9252 says that when multiple SID Information Sub-TLVs are present, the ingress PE SHOULD use the SID from the first instance and ignore subsequent ones.

The parser currently treats a subsequent SID Information Sub-TLV as a malformed Prefix-SID attribute, contrary to RFC 9252.

Fix this by ignoring subsequent SID Information Sub-TLVs after the first SID has been processed, aligning the parser with RFC 9252.

[greptile-apps]

Confidence Score: 5/5

[Medium impact] The PR appears safe to merge, with the previous non-blocking test coverage gap still partly open.

Findings

1.  Test misses the changed check ▶

Fix with agent prompt

### Issue 1
tests/bgpd/test_mp_attr.c:954-958
The new fixture repeats outer SRv6 L3 Service TLVs, not SID Information Sub-TLVs within one service TLV. `bgp_attr_prefix_sid` already skips the second outer TLV, so this test never reaches the changed duplicate check and would pass without this fix.

Put both SID Information Sub-TLVs inside one service TLV and check the retained SID and endpoint behavior. `SHOULD_PARSE` alone does not prove that the first SID wins.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Summary

Updates bgp_attr_srv6_service to skip a repeated SID Information Sub-TLV instead of reporting a malformed attribute.

• BGP keeps the first SID when a service repeats SID information.

_{Reviews (2) · Last reviewed commit: "tests: Cover repeated SRv6 SID Informati..." · Reviewed by Greptile}

[cscarpitta]

@greptile review