From: Yuanhan Liu <yuanhan.liu@linux.intel.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: "Wang, Zhihong" <zhihong.wang@intel.com>,
Maxime Coquelin <maxime.coquelin@redhat.com>,
Stephen Hemminger <stephen@networkplumber.org>,
"dev@dpdk.org" <dev@dpdk.org>,
"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Subject: Re: [dpdk-dev] [Qemu-devel] [PATCH 1/2] vhost: enable any layout feature
Date: Tue, 11 Oct 2016 14:57:49 +0800 [thread overview]
Message-ID: <20161011065749.GO1597@yliu-dev.sh.intel.com> (raw)
In-Reply-To: <20161010073721-mutt-send-email-mst@kernel.org>
On Mon, Oct 10, 2016 at 07:39:59AM +0300, Michael S. Tsirkin wrote:
> > > > > > 1. why is this memset expensive?
> > > > >
> > > > > I don't have the exact answer, but just some rough thoughts:
> > > > >
> > > > > It's an external clib function: there is a call stack and the
> > > > > IP register will bounch back and forth.
> > > >
> > > > for memset 0? gcc 5.3.1 on fedora happily inlines it.
> > >
> > > Good to know!
> > >
> > > > > overkill to use that for resetting 14 bytes structure.
> > > > >
> > > > > Some trick like
> > > > > *(struct virtio_net_hdr *)hdr = {0, };
> > > > >
> > > > > Or even
> > > > > hdr->xxx = 0;
> > > > > hdr->yyy = 0;
> > > > >
> > > > > should behaviour better.
> > > > >
> > > > > There was an example: the vhost enqueue optmization patchset from
> > > > > Zhihong [0] uses memset, and it introduces more than 15% drop (IIRC)
> > > > > on my Ivybridge server: it has no such issue on his server though.
> > > > >
> > > > > [0]: http://dpdk.org/ml/archives/dev/2016-August/045272.html
> > > > >
> > > > > --yliu
> > > >
> > > > I'd say that's weird. what's your config? any chance you
> > > > are using an old compiler?
> > >
> > > Not really, it's gcc 5.3.1. Maybe Zhihong could explain more. IIRC,
> > > he said the memset is not well optimized for Ivybridge server.
> >
> > The dst is remote in that case. It's fine on Haswell but has complication
> > in Ivy Bridge which (wasn't supposed to but) causes serious frontend issue.
> >
> > I don't think gcc inlined it there. I'm using fc24 gcc 6.1.1.
>
>
> So try something like this then:
Yes, I saw memset is inlined when this diff is applied.
So, mind to send a formal patch? You might want to try build at least:
it doesn't build.
> Generally pointer chasing in vq->hw->vtnet_hdr_size can't be good
> for performance. Move fields used on data path into vq
> and use from there to avoid indirections?
Good suggestion!
--yliu
next prev parent reply other threads:[~2016-10-11 6:57 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-26 6:40 [dpdk-dev] [PATCH 0/2] enables vhost/virtio " Yuanhan Liu
2016-09-26 6:40 ` [dpdk-dev] [PATCH 1/2] vhost: enable " Yuanhan Liu
2016-09-26 18:01 ` Stephen Hemminger
2016-09-26 19:24 ` Michael S. Tsirkin
2016-09-27 3:11 ` Yuanhan Liu
2016-09-27 19:48 ` Stephen Hemminger
2016-09-27 19:56 ` Michael S. Tsirkin
2016-09-28 2:28 ` Yuanhan Liu
2016-09-29 15:30 ` [dpdk-dev] [Qemu-devel] " Maxime Coquelin
2016-09-29 17:57 ` Michael S. Tsirkin
2016-09-29 20:05 ` Maxime Coquelin
2016-09-29 20:21 ` Michael S. Tsirkin
2016-09-29 21:23 ` Maxime Coquelin
2016-09-30 12:05 ` Maxime Coquelin
2016-09-30 19:16 ` Michael S. Tsirkin
2016-10-10 4:05 ` Yuanhan Liu
2016-10-10 4:17 ` Michael S. Tsirkin
2016-10-10 4:22 ` Yuanhan Liu
2016-10-10 4:25 ` Michael S. Tsirkin
2016-10-10 12:40 ` Maxime Coquelin
2016-10-10 14:42 ` Yuanhan Liu
2016-10-10 14:54 ` Maxime Coquelin
2016-10-11 6:04 ` Yuanhan Liu
2016-10-11 6:39 ` Maxime Coquelin
2016-10-11 6:49 ` Yuanhan Liu
2016-10-03 14:20 ` Maxime Coquelin
2016-10-10 3:37 ` Yuanhan Liu
2016-10-10 3:46 ` Michael S. Tsirkin
2016-10-10 3:59 ` Yuanhan Liu
2016-10-10 4:16 ` Wang, Zhihong
2016-10-10 4:24 ` Michael S. Tsirkin
2016-10-10 4:39 ` Michael S. Tsirkin
2016-10-11 6:57 ` Yuanhan Liu [this message]
2016-10-12 3:21 ` Yuanhan Liu
[not found] ` <F5DF4F0E3AFEF648ADC1C3C33AD4DBF16C2409EB@SHSMSX101.ccr.corp.intel.com>
2016-10-13 2:52 ` Yang, Zhiyong
2016-10-10 3:50 ` Yuanhan Liu
2016-10-09 23:20 ` [dpdk-dev] " Michael S. Tsirkin
2016-10-10 3:03 ` Yuanhan Liu
2016-10-10 3:04 ` Michael S. Tsirkin
2016-10-10 3:10 ` Yuanhan Liu
2016-09-26 6:40 ` [dpdk-dev] [PATCH 2/2] net/virtio: " Yuanhan Liu
2016-09-26 18:04 ` Stephen Hemminger
2016-09-29 18:00 ` Michael S. Tsirkin
2016-09-29 18:01 ` Michael S. Tsirkin
2016-11-10 15:18 ` [dpdk-dev] [PATCH 0/2] enables vhost/virtio " Michael S. Tsirkin
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=20161011065749.GO1597@yliu-dev.sh.intel.com \
--to=yuanhan.liu@linux.intel.com \
--cc=dev@dpdk.org \
--cc=maxime.coquelin@redhat.com \
--cc=mst@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stephen@networkplumber.org \
--cc=zhihong.wang@intel.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).