netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Stephen Hemminger <stephen@networkplumber.org>
To: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Cc: netdev@vger.kernel.org, roopa@cumulusnetworks.com, davem@davemloft.net
Subject: Re: [PATCH RFC net-next 3/4] bridge: move write-heavy fdb members in their own cache line
Date: Tue, 31 Jan 2017 08:37:57 -0800	[thread overview]
Message-ID: <20170131083757.17b9a2aa@xeon-e3> (raw)
In-Reply-To: <1485876718-18091-4-git-send-email-nikolay@cumulusnetworks.com>

On Tue, 31 Jan 2017 16:31:57 +0100
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> wrote:

> Fdb's used and updated fields are written to on every packet forward and
> packet receive respectively. Thus if we are receiving packets from a
> particular fdb, they'll cause false-sharing with everyone who has looked
> it up (even if it didn't match, since mac/vid share cache line!). The
> "used" field is even worse since it is updated on every packet forward
> to that fdb, thus the standard config where X ports use a single gateway
> results in 100% fdb false-sharing. Note that this patch does not prevent
> the last scenario, but it makes it better for other bridge participants
> which are not using that fdb (and are only doing lookups over it).
> The point is with this move we make sure that only communicating parties
> get the false-sharing, in a later patch we'll show how to avoid that too.
> 
> Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
> ---

What about making updated a per-cpu value?

  reply	other threads:[~2017-01-31 16:38 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-31 15:31 [PATCH RFC net-next 0/4] bridge: improve cache utilization Nikolay Aleksandrov
2017-01-31 15:31 ` [PATCH RFC net-next 1/4] bridge: modify bridge and port to have often accessed fields in one cache line Nikolay Aleksandrov
2017-01-31 15:31 ` [PATCH RFC net-next 2/4] bridge: move to workqueue gc Nikolay Aleksandrov
2017-01-31 15:31 ` [PATCH RFC net-next 3/4] bridge: move write-heavy fdb members in their own cache line Nikolay Aleksandrov
2017-01-31 16:37   ` Stephen Hemminger [this message]
2017-01-31 16:39     ` Nikolay Aleksandrov
2017-01-31 15:31 ` [PATCH RFC net-next 4/4] bridge: add ability to turn off fdb used updates Nikolay Aleksandrov
2017-02-03  2:47   ` David Miller
2017-02-03  8:30     ` Nikolay Aleksandrov
2017-02-03 18:28       ` Stephen Hemminger
2017-02-03 18:34         ` Nikolay Aleksandrov
2017-02-03 22:24           ` Stephen Hemminger
2017-02-03 22:27             ` Nikolay Aleksandrov
2017-02-04 16:45     ` Nikolay Aleksandrov
2017-01-31 16:39 ` [PATCH RFC net-next 0/4] bridge: improve cache utilization Stephen Hemminger
2017-01-31 16:41   ` Nikolay Aleksandrov
2017-01-31 18:09     ` Nikolay Aleksandrov
2017-01-31 18:21       ` Stephen Hemminger
2017-01-31 18:45         ` Nikolay Aleksandrov
2017-01-31 18:51           ` Nikolay Aleksandrov

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=20170131083757.17b9a2aa@xeon-e3 \
    --to=stephen@networkplumber.org \
    --cc=davem@davemloft.net \
    --cc=netdev@vger.kernel.org \
    --cc=nikolay@cumulusnetworks.com \
    --cc=roopa@cumulusnetworks.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).