The following table show the current status of the Implementation. Once the implementation is completed, the implementation specific columns will be removed.
Name | Implemented | Finalized | OS specific | Result FreeBSD | Result Linux |
---|---|---|---|---|---|
sctp-bdc-i-7-1 | Yes | Yes | No | Passed (Note 1) | Passed |
sctp-bdc-i-7-2 | Yes (Note 2) | Yes | Yes | Passed | Passed |
sctp-bdc-i-7-3 | Yes (Note 3) | Yes | Yes | Passed | Passed |
sctp-bdc-v-7-4 | Yes (Note 4) | Yes | No | Passed | Passed |
sctp-bdc-v-7-5 | Yes | Yes | No | Passed | Passed |
sctp-bdc-v-7-6 | Yes | Yes | No | Passed | Passed |
sctp-bdc-v-7-7 | Yes | Yes | No | Passed | Passed |
sctp-bdc-v-7-8 | Yes (Note 5) | Yes | No | Passed | Failed (Note 6) |
- Fix for sctp-bdc-i-7-1: r287535.
- FreeBSD just ignores the packet containing the INIT-ACK chunk bundled with a DATA chunk whereas Linux responds with an ABORT chunk indicating a protocol violation.
- FreeBSD just ignores the packet containing the SHUTDOWN-COMPLETE chunk bundled with a DATA chunk whereas Linux responds with an ABORT chunk indicating a protocol violation.
- It is assumed that the SUT bundles the COOKIE-ACK chunk with the SACK chunk.
- It is assumed that the SUT ignores the DATA chunk, when it is bundled with the SHUTDOWN-ACK chunk, and sends a SHUTDOWN-COMPLETE chunk.
- The message flow is correct. The problem is
read()
not returning 0 after the teardown handshake is complete.