* [MPTCP] [Weekly meetings] MoM - 26th of March 2020
@ 2020-03-26 17:04 Matthieu Baerts
0 siblings, 0 replies; only message in thread
From: Matthieu Baerts @ 2020-03-26 17:04 UTC (permalink / raw)
To: mptcp
[-- Attachment #1: Type: text/plain, Size: 8387 bytes --]
Hello,
We just had our 92nd meeting with Mat, Peter and Ossama (Intel OTC),
Christoph (Apple), Paolo, Davide and Florian (RedHat) and myself (Tessares).
Thanks again for this new good meeting!
Here are the minutes of the meeting:
Accepted patches:
- The list of accepted patches can be seen on PatchWork:
https://patchwork.ozlabs.org/project/mptcp/list/?state=3
netdev (if mptcp ML is in cc) (by Davide):
1258480 [net-next] net: mptcp: don't hang in mptcp_sendmsg() after TCP
fallba...
our repo: (Mat Martineau, Matthieu Baerts, Paolo Abeni, Peter Krystad)
1261955 [net-next,v3] Squash-to: "mptcp: add netlink-based PM"
1261703 squash-to: selftests: add PM netlink functional tests
1260815 [3/3] mptcp: fix "fn parameter not described" warning
1260813 [2/3] mptcp: fix "fn parameter not described" warnings
1260812 [1/3] mptcp: fix "fn parameter not described" warning
1260233 mptcp:netlink: fix compilation warning
1259160 [v1,5/5] Squash-to: "mptcp: update per unacked sequence on pkt
1259159 [v1,4/5] Squash-to: "mptcp: Add handling of outgoing MP_JOIN re
1259162 [v1,3/5] Squash-to: "mptcp: Add handling of incoming MP_JOIN re
1259156 [v1,2/5] Squash-to: "mptcp: Add path manager interface"
1259157 [v1,1/5] Squash-to: "mptcp: Add ADD_ADDR handling"
1259137 [2/2] mptcp: Include ADD_ADDR port value for path manager
1259136 [1/2] mptcp: Parse optional port field of ADD_ADDR
1258564 [v1,4/4] Squash-to: "subflow: place further subflows on new 'jo
1258562 [v1,3/4] Squash-to: "mptcp: Add handling of incoming MP_JOIN re
1258566 [v1,2/4] Squash-to: "mptcp: Add path manager interface"
1258565 [v1,1/4] Squash-to: "mptcp: Add ADD_ADDR handling"
1258550 squash-to: mptcp: Add path manager interface
1258547 topmsg: mptcp: add netlink based PM
1258546 topmsg: mptcp: implement and use MPTCP-level retransmission
1258548 topmsg: mptcp: update per unacked sequence on pkt reception
1258549 topmsg: mptcp: Implement path manager interface commands
1258545 topmsg: mptcp: Add path manager interface
1258046 [2/2] mptcp: fix compilation warning without IPV6
1258047 [1/2] mptcp: fix compilation error without IPV6
1255551 [v1,3/3] Squash-to: "selftests: add test-cases for MPTCP MP_JOI
1255552 [v1,2/3] Squash-to: "mptcp: add netlink based PM"
1255550 [v1,1/3] Squash-to: "mptcp: Implement path manager interface co
Pending patches:
- The list of pending patches can be seen on PatchWork:
https://patchwork.ozlabs.org/project/mptcp/list/?state=*
netdev: (Florian)
1258245 [mptcp-next,v2] tcp: mptcp: use mptcp receive buffer space to s
our repo: (Florian, Paolo)
1253668: Changes Requested: mptcp: propagate sendbuf setting to subflow
1253672: Changes Requested: mptcp: copy mptcp receive buffer setting to:
- wait to know what we do here (setsockopt, etc.)
- so no new version planned here
1262061: New: [RFC,1/2] uapi: update linux/mptcp.h
1262063: New: [RFC,2/2] add support for mptcp netlink interface.:
- iproute2 modifications
- Paolo is asking for help for the man page
- any review welcome
- especially with the syntax, see patch 2 → usage()
https://patchwork.ozlabs.org/patch/1262063/
- related changes:
- Davide has some patches to extract data from subflow (ss)
- can be queued with netlink patches
FYI: Current Roadmap:
- Part 3 (this merge window):
- PM: [Done]
- locking [Done]
- basic → netlink support [Done]
- MP_JOIN [Done]
- ADD_ADDR for MPTCP v1 [DONE]
- clean the current patches [DONE]
- squash the current patches [DONE]:
- "subflow: place further subflows on new 'join_list'"
- PM basic [Done]
- fix sender ID in MP_JOIN [Done]
- Part 4 (to fully support MPTCP):
- Shared recv window (full support)
- IPv6 - IPv4 mapped support
- not dropping MPTCP options (ADD_ADDR, etc.)
- FAST_CLOSE
- full MPTCP v1 support (reliable add_addr, etc.)
- after a few attempts of failed MPTCP, we fallback to TCP
(like TFO is doing)
- PM server (more advanced)
- Full DATA_FIN support [WIP]:
- could be nice to have it: if ready
- Active backup support [WIP]
- ADD_ADDR for MPTCPv1: echo bit
- Opti in TCP option structures (unions) [to be rebased]
- Part 5 (extra needed for prod):
- opti/perfs
- TFO
- PM netlink
- PM bpf
- Scheduler bpf
- syncookies
- [gs]etsockopt per subflow
- notify the userspace when a subflow is added/removed → cmsg
Part 3 remaining stuff:
- Review commit messages:
- DONE
- there are also a couple of patches at the end of the series we
will have to rebase/squash:
- DONE
- ADD_ADDR for MPTCPv1:
- DONE (without the echo)
- v1:
- sent on Monday
- v2 in preparation:
- rebase on latest net-next (with resolved conflicts)
- small fixes:
- one comment in token.c
- netlink:
- add spinlock
- relax admin constraint on 'getters'
- let user space set a single limit at once
- fix reported error message
- nothing else
- tag → netdev-for-5.7-part3-v2
- 2a6d194b94e7...c8cb6ce85550 netdev-for-5.7-part3-v2 ->
netdev-for-5.7-part3-v2 (forced update)
Shared received windows:
- is "tcp: mptcp: use mptcp receive buffer space to select rcv
window" patch enough?
- can be sent to netdev when part 3 is applied
- can be queued at the end of the export branch
- Matth: TODO
Extra tests:
- news about Syzkaller? (Christoph):
- 2 new issues:
- one is a known issue, already on Github → like
"mptcp_connect.sh -m mmap"
- a new one but a nasty one:
- Paolo looked at it and he is trying to collect data
- Looks quite bad. Paolo will send an email about that.
- when ran without KASAN, DEBUG, more issues are found!
- new about Intel's kbuild? (Mat):
- They said they found an issue in BPF selftests, we are
waiting for more news about that
- packetdrill (Davide):
- Can we merge:
https://github.com/multipath-tcp/packetdrill/pull/3 :
- yes we can
- now validating: DACK4 + DSN4. But also MP_JOIN
- CI (Matth):
- Now testing W=1 for files in net/mptcp/
- and testing commits at the end of "selftests: add test-cases
for MPTCP MP_JOIN" as well
Filed some small buglet as github issues:
- https://github.com/multipath-tcp/mptcp_net-next/issues
- feel free to add new ones
- feel free to pick any of them
2: reduce hooking in TCP code
3: fix 'mmap' related race
4: keep a single work struct in mptcp socket
5: Allow ss/netstat etc. to show program name for client and
listener MPTCP subflows
6: loss and delay without reordering causes very slow transfer 2:
- https://github.com/multipath-tcp/mptcp_net-next/issues/6
-
7: cleanup sendmsg_frag allocation
8: fix possible race in subflow_finish_connect()
9: fix "IPv4: Attempt to release TCP socket in state 1 "... on shutdown
10: reduce mptcp options space usage
11: fix fallback to TCP... @dcaratti
TODO for next meeting:
- Paolo: Use MPTCP with multiple paths not to be disconnected if
one fails.
Next meeting:
- We propose to have the next meeting on Thursday, the 2nd of April.
- /!\ *BACK* to the usual time: 16:00 UTC (9am PDT, 8pm CEST) /!\
- Still open to everyone!
- https://annuel2.framapad.org/p/mptcp_upstreaming_20200402
Feel free to comment on these points and propose new ones for the next
meeting!
Talk to you next week,
Matt
--
Matthieu Baerts | R&D Engineer
matthieu.baerts(a)tessares.net
Tessares SA | Hybrid Access Solutions
www.tessares.net
1 Avenue Jean Monnet, 1348 Louvain-la-Neuve, Belgium
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2020-03-26 17:04 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-03-26 17:04 [MPTCP] [Weekly meetings] MoM - 26th of March 2020 Matthieu Baerts
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.