22April2021 [FreeBSD WIKI]

This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.
Attendees:  jtl@, rscheff@, tuexen@, markj@

====== Discussion ======
  * We continued our discussion on the use of sockaddr_in on IPv6 sockets and mapped IPv4 addresses. As far back as stable/9, it looks like using AF_INET sockaddr's on AF_INET6 TCP sockets would not have worked correctly. tuexen@ explained the subtleties of the way SCTP handles sockaddrs for AF_INET6 sockets. markj@ will ensure his patch is correct with respect to SCTP. Otherwise, it seems like a safe change to make.
  * [[https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=252913|PR 252913]]: tuexen@ got a tracefile and some information from Netflix. This impacts stable/12. Now that 13.0 has been released, tuexen@ will look at this in the near future.
  * [[https://reviews.freebsd.org/D23230|D23230]]: ECN++, rscheff@ updated the sysctl interface so sysctls are not checked in fast path. Needs review.
  * [[https://reviews.freebsd.org/D28702|D28702]]: Planned for main, mid March. Still in progress, waiting for 13 release. rrs@ will review.
  * [[https://reviews.freebsd.org/D29835|D29835]] Fix for rescue retransmission panics when rfc_6675_pipe/sack.revised is enabled.
  * Lost retransmission detection: [[https://reviews.freebsd.org/D28931|D28931]] is an idea for doing SACK loss recovery in the base stack. (Documented in an [[https://datatracker.ietf.org/doc/draft-scheffenegger-tcpm-lrd/|I-D]].) We discussed the idea further. rscheff@ will modify code to include a sysctl to enable/disable the new behavior.
  * [[https://reviews.freebsd.org/D29515|D29515]]: rscheff@ committed change to use jenkins hash with random salt and allow an administrator to re-salt with a sysctl. rscheff@ will MFC to stable/13.
  * tuexen@ committed support for TCP over UDP to main. He plans to MFC to stable/13. tuexen@ will do this.
    * RACK will be updated after TCP over UDP support has been committed. rrs@ will do this.
  * [[https://reviews.freebsd.org/D29411|D29411]]: PRR for ECN and congestion window reduction. Committed to head; waiting for MFC.
    * [[https://reviews.freebsd.org/D29440|D29440]]: Do preparatory work to enable PRR for more cases. Committed to head; waiting for MFC.
    * [[https://reviews.freebsd.org/D29441|D29441]]: Implement PRR for !SACK. Committed to head; waiting for MFC.
  * [[https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=254725|254725]]: LRO panic. Peer ACKs up to last data byte. Peer SACKs FIN - see reconstructed trace from in-core packet headers {{transportprotocols/22april2021/pr254725.trc?linkonly|PR254725.trc}}. snd_una is correct, but send buffer is empty. Still no progress on reproduction. Randall suggested changing the code to wait to send FIN until there was nothing outstanding (see [[https://reviews.freebsd.org/D29565|D29565]]). This change will delay the remote side recognizing session shutdown, which might impact applications which take action when a session closes. General consensus is that it would be better to find and fix the bugs.
  * Follow up on the RST issue reported by NFS developer: We send a challenge ACK. Remote side sends a RST. We don't process the RST. rscheff@ proposed [[https://reviews.freebsd.org/D29690|D29690]]. tuexen@ and jtl@ have a small concern that we should do more work on finding the root cause before we commit a fix. rscheff@ will debug further on a machine seeing this problem.

Next meeting: 06 May @ 1400 UTC