From: Jesper Dangaard Brouer <brouer@redhat.com>
To: John Fastabend <john.fastabend@gmail.com>
Cc: David Miller <davem@davemloft.net>,
netdev@vger.kernel.org, andy@greyhouse.net, daniel@iogearbox.net,
ast@fb.com, alexander.duyck@gmail.com, bjorn.topel@intel.com,
jakub.kicinski@netronome.com, ecree@solarflare.com,
sgoutham@cavium.com, Yuval.Mintz@cavium.com, saeedm@mellanox.com,
Andi Kleen <andi@firstfloor.org>,
jesse.brandeburg@intel.com, brouer@redhat.com
Subject: Re: [RFC PATCH 00/12] Implement XDP bpf_redirect vairants
Date: Sun, 16 Jul 2017 10:23:08 +0200 [thread overview]
Message-ID: <20170716102308.163b9f91@redhat.com> (raw)
In-Reply-To: <596528E6.2080707@gmail.com>
On Tue, 11 Jul 2017 12:37:10 -0700 John Fastabend <john.fastabend@gmail.com> wrote:
> > I have a really strange observation... if I change the CPU powersave
> > settings, then the xdp_redirect_map performance drops in half! Above
> > was with "tuned-adm profile powersave" (because, this is a really noisy
> > server, and I'm sitting next to it). I can see that the CPU under-load
> > goes into "turbomode", rest going into low-power, including the
> > Hyper-thread siblings.
> >
> > If I change the profile to: # tuned-adm profile network-latency
> >
> > ifindex 6: 12964879 pkt/s
> > ifindex 6: 12964683 pkt/s
> > ifindex 6: 12961497 pkt/s
> > ifindex 6: 11779966 pkt/s <-- change to tuned-adm profile network-latency
> > ifindex 6: 6853959 pkt/s
> > ifindex 6: 6851120 pkt/s
> > ifindex 6: 6856934 pkt/s
> > ifindex 6: 6857344 pkt/s
> > ifindex 6: 6857161 pkt/s
> >
> > The CPU efficiency goes from 2.35 to 1.24 insn per cycle.
> >
> > John do you know some Intel people that could help me understand what
> > is going on?!? This is very strange...
> >
> > I tried Andi's toplev tool, which AFAIK indicate that this is a
> > Frontend problem, e.g. in decoding the instructions?!?
> >
>
> hmm maybe Jesse or Alex have some clues. Adding them to the CC list.
This seems related to Hyper-Threading. I tried to disable
hyperthreading in the BIOS, and the problem goes away. That is, the
benchmarks are no-longer affected by the CPU tuned-adm profile.
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Principal Kernel Engineer at Red Hat
LinkedIn: http://www.linkedin.com/in/brouer
next prev parent reply other threads:[~2017-07-16 8:23 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-07-07 17:34 [RFC PATCH 00/12] Implement XDP bpf_redirect vairants John Fastabend
2017-07-07 17:34 ` [RFC PATCH 01/12] ixgbe: NULL xdp_tx rings on resource cleanup John Fastabend
2017-07-07 17:35 ` [RFC PATCH 02/12] net: xdp: support xdp generic on virtual devices John Fastabend
2017-07-07 17:35 ` [RFC PATCH 03/12] xdp: add bpf_redirect helper function John Fastabend
2017-07-09 13:37 ` Saeed Mahameed
2017-07-10 17:23 ` John Fastabend
2017-07-11 14:09 ` Andy Gospodarek
2017-07-11 18:38 ` John Fastabend
2017-07-11 19:38 ` Jesper Dangaard Brouer
2017-07-12 11:00 ` Saeed Mahameed
2017-07-07 17:35 ` [RFC PATCH 04/12] xdp: sample program for new bpf_redirect helper John Fastabend
2017-07-07 17:36 ` [RFC PATCH 05/12] net: implement XDP_REDIRECT for xdp generic John Fastabend
2017-07-07 17:36 ` [RFC PATCH 06/12] ixgbe: add initial support for xdp redirect John Fastabend
2017-07-07 17:36 ` [RFC PATCH 07/12] xdp: add trace event " John Fastabend
2017-07-07 17:37 ` [RFC PATCH 08/12] bpf: add devmap, a map for storing net device references John Fastabend
2017-07-08 18:57 ` Jesper Dangaard Brouer
2017-07-07 17:37 ` [RFC PATCH 09/12] bpf: add bpf_redirect_map helper routine John Fastabend
2017-07-07 17:37 ` [RFC PATCH 10/12] xdp: Add batching support to redirect map John Fastabend
2017-07-10 17:53 ` Jesper Dangaard Brouer
2017-07-10 17:56 ` John Fastabend
2017-07-07 17:38 ` [RFC PATCH 11/12] net: add notifier hooks for devmap bpf map John Fastabend
2017-07-07 17:38 ` [RFC PATCH 12/12] xdp: bpf redirect with map sample program John Fastabend
2017-07-07 17:48 ` [RFC PATCH 00/12] Implement XDP bpf_redirect vairants John Fastabend
2017-07-08 9:46 ` David Miller
2017-07-08 19:06 ` Jesper Dangaard Brouer
2017-07-10 18:30 ` Jesper Dangaard Brouer
2017-07-11 0:59 ` John Fastabend
2017-07-11 14:23 ` Jesper Dangaard Brouer
2017-07-11 18:26 ` John Fastabend
2017-07-13 11:14 ` Jesper Dangaard Brouer
2017-07-13 16:16 ` Jesper Dangaard Brouer
2017-07-13 17:00 ` John Fastabend
2017-07-13 18:21 ` David Miller
2017-07-11 15:36 ` Jesper Dangaard Brouer
2017-07-11 17:48 ` John Fastabend
2017-07-11 18:01 ` Jesper Dangaard Brouer
2017-07-11 18:29 ` John Fastabend
2017-07-11 18:44 ` Jesper Dangaard Brouer
2017-07-11 18:56 ` John Fastabend
2017-07-11 19:19 ` Jesper Dangaard Brouer
2017-07-11 19:37 ` John Fastabend
2017-07-16 8:23 ` Jesper Dangaard Brouer [this message]
2017-07-17 17:04 ` Jesse Brandeburg
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=20170716102308.163b9f91@redhat.com \
--to=brouer@redhat.com \
--cc=Yuval.Mintz@cavium.com \
--cc=alexander.duyck@gmail.com \
--cc=andi@firstfloor.org \
--cc=andy@greyhouse.net \
--cc=ast@fb.com \
--cc=bjorn.topel@intel.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=ecree@solarflare.com \
--cc=jakub.kicinski@netronome.com \
--cc=jesse.brandeburg@intel.com \
--cc=john.fastabend@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=saeedm@mellanox.com \
--cc=sgoutham@cavium.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).