Page MenuHomeFreeBSD
Differential D5734

VNET: cleanup TCP on vent teardown
ClosedPublic
Actions

Authored by bz on Mar 24 2016, 9:39 PM.
Tags
None
Referenced Files
Unknown Object (File)
Oct 14 2024, 10:10 AM
Unknown Object (File)
Sep 16 2024, 10:35 PM
Unknown Object (File)
Sep 8 2024, 6:02 PM
Unknown Object (File)
Sep 8 2024, 12:01 PM
Unknown Object (File)
Sep 7 2024, 10:36 AM
Unknown Object (File)
Sep 5 2024, 4:57 AM
Unknown Object (File)
Sep 5 2024, 4:57 AM
Unknown Object (File)
Sep 5 2024, 4:57 AM
View All 49 Files
Subscribers
emaste
imp
network
transport

Details

Reviewers
emaste
gnn
Group Reviewers
transport
Commits
rS297742: Mfp: r296310,r296343
Summary

Mfp: r296310,r296343

It looks like as with the safety belt of DELAY(hz) fastened (*) we can
completely tear down and free all memory for TCP (after r281599).

(*) in theory a few ticks should be good enough to make sure the timers
are all really gone. I wonder if we could use a better matric here and
check a tcbcb count as an optimization.

Sponsored by: The FreeBSD Foundation

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Not Applicable
Unit
Tests Not Applicable

Event Timeline

bz updated this revision to Diff 14588.Mar 24 2016, 9:39 PM
bz retitled this revision from to VNET: cleanup TCP on vent teardown.
bz updated this object.
bz edited the test plan for this revision. (Show Details)
bz added a reviewer: gnn.
bz set the repository for this revision to rS FreeBSD src repository - subversion.
bz added subscribers: transport, network.
Herald added a reviewer: transport. · View Herald TranscriptMar 24 2016, 9:39 PM
Herald added a subscriber: imp. · View Herald Transcript
gnn accepted this revision.Mar 25 2016, 12:45 AM
gnn edited edge metadata.

That delay is kind of hackish. But it's a start.

This revision is now accepted and ready to land.Mar 25 2016, 12:45 AM
emaste added a subscriber: emaste.Apr 5 2016, 1:48 PM
emaste added inline comments.
sys/netinet/tcp_subr.c
744 ↗(On Diff #14588)

I wonder if your comment from the review check a tcbcb count as an optimization should go in this comment.

bz added inline comments.Apr 5 2016, 1:58 PM
sys/netinet/tcp_subr.c
744 ↗(On Diff #14588)

I'll do. Also one which should be a we instead in that comment anyway.

emaste accepted this revision.Apr 6 2016, 1:14 PM
emaste added a reviewer: emaste.

For future uploads please upload with full (or at least more, e.g. -U999) context.

I agree with @gnn on the DELAY but the additional comment will help clarify and suggest a path forward for a future change.

Closed by commit rS297742: Mfp: r296310,r296343 (authored by bz). · Explain WhyApr 9 2016, 12:05 PM
This revision was automatically updated to reflect the committed changes.

Revision Contents
Changeset List

PathSize
head/
sys/
netinet/
15 lines

Diff 15037

View Options

head/sys/netinet/tcp_subr.c

Loading...