Don't start grouptaskqueue threads for unused grouptaskqueues
Needs RevisionPublic
Actions

Authored by kmacy on Aug 28 2017, 2:19 AM.

Details

Reviewers

shurd
scottl
sbruno
jtl
glebius
kbowling
erj

Group Reviewers

Intel Networking
iflib

Summary

Defer starting grouptaskqueue threads until the first time a grouptask is associated with the group.

This keeps iflib from cluttering ddb ps when there are no iflib drivers in the system.

Diff Detail

Repository

rS FreeBSD src repository - subversion

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 11273

Event Timeline

kmacy created this revision.Aug 28 2017, 2:19 AM

Herald added subscribers: ae, imp. · View Herald TranscriptAug 28 2017, 2:19 AM

kmacy added a subscriber: kbowling.Aug 28 2017, 2:19 AM

kbowling added a reviewer: glebius.Aug 28 2017, 6:57 AM

Only launch as many threads as we actually need
@sbruno dedicated branch vs iflib/ithread_dispatch called iflib/pollution reduction

Looks good to me.

This revision is now accepted and ready to land.Sep 5 2017, 12:29 AM

shurd mentioned this in D12235: iflib rollup patch..Sep 5 2017, 9:15 PM

shurd mentioned this in D12446: Support attaching tx queues to cpus.Sep 23 2017, 1:03 AM

krzysztof.galazka_intel.com added a reviewer: Intel Networking.Nov 30 2017, 6:20 PM

krzysztof.galazka_intel.com added a subscriber: krzysztof.galazka_intel.com.

@shurd does this still work or should it be abandoned?

May as well abandon it until/unless there's interest. I don't think there's much gain for the added complexity.

erj resigned from this revision.Oct 3 2018, 5:21 PM

@krzysztof.galazka_intel.com can you rebase this on main?

This revision now requires changes to proceed.Sep 25 2024, 9:52 PM

Herald added a reviewer: iflib. · View Herald TranscriptSep 25 2024, 9:52 PM

Herald added a subscriber: melifaro. · View Herald Transcript

In D12142#1066726, @kbowling wrote:

@krzysztof.galazka_intel.com can you rebase this on main?

Did autocompletion tricked you, or am I really the target of your message? I'm not sure if @kmacy will be happy with me messing with his patch.

In D12142#1068530, @krzysztof.galazka_intel.com wrote:

In D12142#1066726, @kbowling wrote:

@krzysztof.galazka_intel.com can you rebase this on main?

Did autocompletion tricked you, or am I really the target of your message? I'm not sure if @kmacy will be happy with me messing with his patch.

Sorry I was looking at another review and got the wires crossed, for some reason I thought this was an Intel patch but I was looking at something else.

In D12142#1068558, @kbowling wrote:

In D12142#1068530, @krzysztof.galazka_intel.com wrote:

In D12142#1066726, @kbowling wrote:

@krzysztof.galazka_intel.com can you rebase this on main?

Did autocompletion tricked you, or am I really the target of your message? I'm not sure if @kmacy will be happy with me messing with his patch.

Sorry I was looking at another review and got the wires crossed, for some reason I thought this was an Intel patch but I was looking at something else.

No problem. I was thinking about adding some way of dynamic control of number of threads in taskqgroup while working on https://reviews.freebsd.org/D46062 but it seemed to make things overcomplicated for tasks not critical for traffic performance.