Attendees
- glebius@
- rrs@
- rscheff@
- tuexen@
- Cheng C
Ongoing Short Term Projects
It was discussed whether the TCP_LOG socket option should be inherited from the listener. jtl@ said that this should be the case, but interaction with the sysctl-variable net.inet.tcp.bb.log_auto_ratio should be taken into account. The patch is under review: D38436.
netstat does not support reading connection information from a core file anymore. tuexen@ will document that in the man-page.
- glebius@ asks whether it makes sense to improve the TCP timer precision. tuexen@ states that this is very useful for testing purposes. glebius@ will provide a patch to tuexen@ for testing and tuexen@ will provide feedback, if such a patch us useful for FreeBSD in general. This applies to the default stack only, not the the RACK or BBR stack, which uses HPTS for timers.
At least for testing purposes (packetdrill) it would be good to have LRO support for tun interfaces. tuexen@ suggested to generalize LRO to not depend on an Ethernet header being present. gallatin@ prefers to not make the LRO code more complex, since it most likely will impact the performance. tjh@ has D37437 under review, which allows the tun interface to add an Ethernet header. tuexen@ will test this to see if it can also be used to enable LRO for the tun interface. This seems netmap specific. tuexen@ will consider to use tap interfaces in packetdrill.
D28822 improves PRR. rscheff@ will do some testing, rrs@ and tuexen@ will review.
While working on allowing BBLog being used as a replacement of trpt, tuexen@ figured out that trpt is broken in several ways in stable/13. After a short discussion is was decided that it is better to document this brokenness than to fix it by direct commits in the stable/13 branch. tuexen@ will add a note to the trpt man-page.
- Discussion on use cases of having multiple TCP connections in parallel.
- tuexen@ will try to improve the error handling in PRU functions to allow the reporting of more error cases and put it up for review. glebius@ will review it and make sure it works with the recent changes in the inp/tcb area.
Ongoing Longer Term Projects
rscheff@ is improving AccECN code: D36303.
D23230 is implementing ECN++.
tuexen@ wants to write a tool which dumps the BBLog information of a TCP endpoint from a life system or a core. Using kvm_read() has drawbacks as glebius@ pointed out, using a python kgdb script may be an alternative. An alternative is to write a python kgdb script which pumps data into tcplog_dumper. This would allow to minimize code duplication. @glebius will try to work on this soon.
- glebius@ and markj@ are improving the inp code and as part of this work will make the transport protocols inp code agnostic.
- thj@ describes a problem related to SYN-ACK retransmissions occurring in a setup using jails and epairs. There seem to be a 1 second pause.
Next Meeting
09 March 2023 @ 1500 UTC