All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jakub Kicinski <kuba@kernel.org>
To: Eric Dumazet <edumazet@google.com>
Cc: Wei Wang <weiwan@google.com>,
	"David S . Miller" <davem@davemloft.net>,
	netdev <netdev@vger.kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Hannes Frederic Sowa <hannes@stressinduktion.org>,
	Felix Fietkau <nbd@nbd.name>
Subject: Re: [RFC PATCH net-next 0/6] implement kthread based napi poll
Date: Tue, 29 Sep 2020 12:19:02 -0700	[thread overview]
Message-ID: <20200929121902.7ee1c700@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com> (raw)
In-Reply-To: <CANn89iJDM97U15Znrx4k4bOFKunQp7dwJ9mtPwvMmB4S+rSSbA@mail.gmail.com>

On Mon, 28 Sep 2020 19:43:36 +0200 Eric Dumazet wrote:
> Wei, this is a very nice work.
> 
> Please re-send it without the RFC tag, so that we can hopefully merge it ASAP.

The problem is for the application I'm testing with this implementation
is significantly slower (in terms of RPS) than Felix's code:

              |        L  A  T  E  N  C  Y       |  App   |     C P U     |
       |  RPS |   AVG  |  P50  |   P99  |   P999 | Overld |  busy |  PSI  |
thread | 1.1% | -15.6% | -0.3% | -42.5% |  -8.1% | -83.4% | -2.3% | 60.6% |
work q | 4.3% | -13.1% |  0.1% | -44.4% |  -1.1% |   2.3% | -1.2% | 90.1% |
TAPI   | 4.4% | -17.1% | -1.4% | -43.8% | -11.0% | -60.2% | -2.3% | 46.7% |

thread is this code, "work q" is Felix's code, TAPI is my hacks.

The numbers are comparing performance to normal NAPI.

In all cases (but not the baseline) I configured timer-based polling
(defer_hard_irqs), with around 100us timeout. Without deferring hard
IRQs threaded NAPI is actually slower for this app. Also I'm not
modifying niceness, this again causes application performance
regression here.

1 NUMA node. 18 NAPI instances each is around 25% of a single CPU.

I was initially hoping that TAPI would fit nicely as an extension 
of this code, but I don't think that will be the case.

Are there any assumptions you're making about the configuration that 
I should try to replicate?

  parent reply	other threads:[~2020-09-29 19:19 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-14 17:24 [RFC PATCH net-next 0/6] implement kthread based napi poll Wei Wang
2020-09-14 17:24 ` [RFC PATCH net-next 1/6] net: implement threaded-able napi poll loop support Wei Wang
2020-09-25 19:45   ` Hannes Frederic Sowa
2020-09-25 23:50     ` Wei Wang
2020-09-26 14:22       ` Hannes Frederic Sowa
2020-09-28  8:45         ` Paolo Abeni
2020-09-28 18:13           ` Wei Wang
2020-09-14 17:24 ` [RFC PATCH net-next 2/6] net: add sysfs attribute to control napi threaded mode Wei Wang
2020-09-15  2:50   ` kernel test robot
2020-09-15  3:47   ` kernel test robot
2020-09-14 17:24 ` [RFC PATCH net-next 3/6] net: extract napi poll functionality to __napi_poll() Wei Wang
2020-09-14 17:24 ` [RFC PATCH net-next 4/6] net: modify kthread handler to use __napi_poll() Wei Wang
2020-09-14 17:24 ` [RFC PATCH net-next 5/6] net: process RPS/RFS work in kthread context Wei Wang
2020-09-18 22:44   ` Wei Wang
2020-09-21  8:11     ` Eric Dumazet
2020-09-14 17:24 ` [RFC PATCH net-next 6/6] net: improve napi threaded config Wei Wang
2020-09-25 13:48 ` [RFC PATCH net-next 0/6] implement kthread based napi poll Magnus Karlsson
2020-09-25 17:15   ` Wei Wang
2020-09-25 17:30     ` Eric Dumazet
2020-09-25 18:16     ` Stephen Hemminger
2020-09-25 18:23       ` Eric Dumazet
2020-09-25 19:00         ` Stephen Hemminger
2020-09-25 19:06   ` Jakub Kicinski
2020-09-28 14:07     ` Magnus Karlsson
2020-09-28 17:43 ` Eric Dumazet
2020-09-28 18:15   ` Wei Wang
2020-09-29 19:19   ` Jakub Kicinski [this message]
2020-09-29 20:16     ` Wei Wang
2020-09-29 21:48       ` Jakub Kicinski
2020-09-30  8:23         ` David Laight
2020-09-30  8:58         ` Paolo Abeni
2020-09-30 15:58           ` Jakub Kicinski

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200929121902.7ee1c700@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com \
    --to=kuba@kernel.org \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=hannes@stressinduktion.org \
    --cc=nbd@nbd.name \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=weiwan@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
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.