From: Stephen Hemminger <stephen@networkplumber.org>
To: Bruce Richardson <bruce.richardson@intel.com>
Cc: Marcin Zapolski <marcinx.a.zapolski@intel.com>, dev@dpdk.org
Subject: Re: [dpdk-dev] [RFC 19.11 1/2] ethdev: make DPDK core functions non-inline
Date: Tue, 30 Jul 2019 08:54:13 -0700 [thread overview]
Message-ID: <20190730085413.550c979b@hermes.lan> (raw)
In-Reply-To: <20190730153355.GB1689@bricha3-MOBL.ger.corp.intel.com>
On Tue, 30 Jul 2019 16:33:55 +0100
Bruce Richardson <bruce.richardson@intel.com> wrote:
> On Tue, Jul 30, 2019 at 08:25:34AM -0700, Stephen Hemminger wrote:
> > On Tue, 30 Jul 2019 14:49:49 +0200
> > Marcin Zapolski <marcinx.a.zapolski@intel.com> wrote:
> >
> > > Make rte_eth_rx_burst, rte_eth_tx_burst and other static inline ethdev
> > > functions not inline. They are referencing DPDK internal structures and
> > > inlining forces those structures to be exposed to user applications.
> > >
> > > In internal testing with i40e NICs a performance drop of about 2% was
> > > observed with testpmd.
> > >
> > > Signed-off-by: Marcin Zapolski <marcinx.a.zapolski@intel.com>
> >
> > Sorry 2% matters.
>
> Note that this is with testpmd. Are there many apps out there where a 2%
> drop in IO cost would be noticable?
Why not find a way to get the 2% back elsewhere? Maybe analyzing the code/cache
in more detail. Perhaps some prefetching could help, or getting rid of
indirect calls elsewhere in the code. At the extreme, maybe implementing
something like the kernel static branches (self-modifying code) would
get a lot back.
next prev parent reply other threads:[~2019-07-30 15:54 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-30 12:49 [dpdk-dev] [RFC 19.11 0/2] Hide DPDK internal struct from public API Marcin Zapolski
2019-07-30 12:49 ` [dpdk-dev] [RFC 19.11 1/2] ethdev: make DPDK core functions non-inline Marcin Zapolski
2019-07-30 15:01 ` Jerin Jacob Kollanukkaran
2019-07-30 15:32 ` Bruce Richardson
2019-07-30 15:25 ` Stephen Hemminger
2019-07-30 15:33 ` Bruce Richardson
2019-07-30 15:54 ` Stephen Hemminger [this message]
2019-07-30 16:04 ` Wiles, Keith
2019-07-30 16:11 ` Bruce Richardson
2019-07-30 16:23 ` Stephen Hemminger
2019-07-30 12:49 ` [dpdk-dev] [RFC 19.11 2/2] ethdev: hide DPDK internal struct from public API Marcin Zapolski
2019-07-30 14:53 ` Ferruh Yigit
2019-09-06 13:18 ` [dpdk-dev] [RFC 19.11 v2 0/3] Hide " Marcin Zapolski
2019-09-06 14:00 ` Bruce Richardson
2019-09-06 13:18 ` [dpdk-dev] [RFC 19.11 v2 1/3] ethdev: hide key ethdev structures " Marcin Zapolski
2019-09-06 14:37 ` Ferruh Yigit
2019-09-09 8:07 ` Zapolski, MarcinX A
2019-09-09 9:59 ` Ferruh Yigit
2019-09-09 10:02 ` Zapolski, MarcinX A
2019-09-09 10:24 ` Ferruh Yigit
2019-09-09 11:41 ` Zapolski, MarcinX A
2019-09-14 10:34 ` Jerin Jacob
2019-09-06 17:24 ` Stephen Hemminger
2019-09-09 9:01 ` Zapolski, MarcinX A
2019-09-10 9:59 ` Zapolski, MarcinX A
2019-09-10 10:06 ` Bruce Richardson
2019-09-10 10:13 ` Zapolski, MarcinX A
2019-09-10 12:19 ` Bruce Richardson
2019-09-10 12:22 ` Zapolski, MarcinX A
2019-09-06 13:18 ` [dpdk-dev] [RFC 19.11 v2 2/3] i40e: make driver compatible with changes in ethdev Marcin Zapolski
2019-09-06 13:18 ` [dpdk-dev] [RFC 19.11 v2 3/3] ixgbe: " Marcin Zapolski
2019-07-30 15:45 [dpdk-dev] [RFC 19.11 1/2] ethdev: make DPDK core functions non-inline Jerin Jacob Kollanukkaran
2019-07-30 16:05 ` Bruce Richardson
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=20190730085413.550c979b@hermes.lan \
--to=stephen@networkplumber.org \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=marcinx.a.zapolski@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).