From: Neil Horman <nhorman@tuxdriver.com>
To: Bruce Richardson <bruce.richardson@intel.com>
Cc: Ferruh Yigit <ferruh.yigit@intel.com>,
John McNamara <john.mcnamara@intel.com>,
Marko Kovacevic <marko.kovacevic@intel.com>,
Thomas Monjalon <thomas@monjalon.net>,
Reshma Pattan <reshma.pattan@intel.com>,
dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v5] ethdev: return named opaque type instead of void pointer
Date: Fri, 23 Mar 2018 22:08:32 -0400 [thread overview]
Message-ID: <20180324020832.GA28338@neilslaptop.think-freely.org> (raw)
In-Reply-To: <20180323170034.GA8888@bricha3-MOBL.ger.corp.intel.com>
On Fri, Mar 23, 2018 at 05:00:34PM +0000, Bruce Richardson wrote:
> On Wed, Mar 21, 2018 at 09:04:01AM -0400, Neil Horman wrote:
> > On Tue, Mar 20, 2018 at 04:34:04PM +0000, Ferruh Yigit wrote:
> > > "struct rte_eth_rxtx_callback" is defined as internal data structure and
> > > used as named opaque type.
> > >
> > > So the functions that are adding callbacks can return objects in this
> > > type instead of void pointer.
> > >
> > > Also const qualifier added to "struct rte_eth_rxtx_callback *" to
> > > protect it better from application modification.
> > >
> > > Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
> > > ---
> > > v2:
> > > * keep using struct * in parameters, instead add callback functions
> > > return struct rte_eth_rxtx_callback pointer.
> > >
> > > v4:
> > > * Remove deprecation notice. LIBABIVER already increased in this release
> > >
> > > v5:
> > > * add const qualifier to rte_eth_rxtx_callback
> > I still wish we could find a way to remove the inline functions and truly
> > protect that struct, but a const is definately better than nothing
> >
> > Acked-by: Neil Horman <nhorman@tuxdriver.com>
> >
> Actually, I think we should do exactly that - convert the rx and tx burst
> calls into actual function calls (and consider any other inlined functions
> too). The cost would be the overhead of making an additional function call
> per-burst, which is likely to be pretty minimal for most common burst sizes
> e.g. 32.
>
> We did some quick tests here with the i40e driver, and for a burst size of
> 32 saw less than 1% perf drop, and for even a small burst of 8 saw less
> than 5% drop. Note that this is testing with testpmd, which has nothing but
> I/O in the datapath. A real-world app is likely to do far more with the
> packets and therefore see proportionally far less perf hit.
>
> Thoughts?
>
> /Bruce
>
> PS: This un-inlining could probably be applied to other device types too,
> e.g. cryptodev (though probably not eventdev as it tends to have smaller
> bursts in some use-cases).
>
I would be 1000% on board with this conversion.
Best
Neil
next prev parent reply other threads:[~2018-03-24 2:09 UTC|newest]
Thread overview: 76+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-12-01 2:29 [dpdk-dev] [PATCH 1/7] ethdev: remove unused flag from header Ferruh Yigit
2017-12-01 2:29 ` [dpdk-dev] [PATCH 2/7] ethdev: fix port id storage Ferruh Yigit
2017-12-01 9:01 ` Hemant Agrawal
2017-12-01 23:52 ` Ferruh Yigit
2017-12-01 2:29 ` [dpdk-dev] [PATCH 3/7] ethdev: separate driver APIs Ferruh Yigit
2017-12-01 6:24 ` Shreyansh Jain
2017-12-01 8:59 ` Hemant Agrawal
2017-12-01 23:51 ` Ferruh Yigit
2017-12-01 2:29 ` [dpdk-dev] [PATCH 4/7] ethdev: separate internal structures into own header Ferruh Yigit
2017-12-01 5:44 ` Hemant Agrawal
2017-12-01 2:29 ` [dpdk-dev] [PATCH 5/7] ethdev: reorder inline functions Ferruh Yigit
2017-12-01 2:29 ` [dpdk-dev] [PATCH 6/7] ethdev: rename function parameter for consistency Ferruh Yigit
2017-12-01 2:29 ` [dpdk-dev] [PATCH 7/7] ethdev: use opaque user callback object Ferruh Yigit
2017-12-01 10:33 ` Bruce Richardson
2017-12-01 11:22 ` Ananyev, Konstantin
2017-12-01 13:17 ` Bruce Richardson
2017-12-01 23:51 ` Ferruh Yigit
2017-12-04 10:31 ` Bruce Richardson
2017-12-04 17:49 ` Ferruh Yigit
[not found] ` <49eb13f1-e12f-071a-b58b-1cd4852b7c8b@scylladb.com>
2017-12-01 22:30 ` [dpdk-dev] [PATCH 1/7] ethdev: remove unused flag from header Vlad Zolotarov
2017-12-01 23:51 ` Ferruh Yigit
2017-12-02 2:22 ` Vlad Zolotarov
2017-12-04 17:54 ` Ferruh Yigit
2017-12-04 19:37 ` Vlad Zolotarov
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 1/6] ethdev: fix port id storage Ferruh Yigit
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 2/6] ethdev: separate driver APIs Ferruh Yigit
2018-01-09 19:19 ` Andrew Rybchenko
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 3/6] ethdev: separate internal structures into own header Ferruh Yigit
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 4/6] ethdev: reorder inline functions Ferruh Yigit
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 5/6] ethdev: rename function parameter for consistency Ferruh Yigit
2018-01-09 16:23 ` [dpdk-dev] [PATCH v2 6/6] ethdev: return named opaque type instead of void pointer Ferruh Yigit
2018-01-09 18:55 ` Stephen Hemminger
2018-01-09 18:58 ` Stephen Hemminger
2018-01-17 21:57 ` [dpdk-dev] [PATCH v3 1/6] ethdev: fix port id storage Ferruh Yigit
2018-01-17 21:57 ` [dpdk-dev] [PATCH v3 2/6] ethdev: return named opaque type instead of void pointer Ferruh Yigit
2018-01-17 22:11 ` [dpdk-dev] [dpdk-stable] " Thomas Monjalon
2018-01-18 10:09 ` Ferruh Yigit
2018-03-09 11:25 ` [dpdk-dev] [PATCH v4] " Ferruh Yigit
[not found] ` <20180309123651.GB19004@hmswarspite.think-freely.org>
2018-03-09 13:00 ` Ferruh Yigit
2018-03-09 15:16 ` Neil Horman
2018-03-09 15:45 ` Ferruh Yigit
2018-03-09 19:06 ` Neil Horman
2018-03-20 15:51 ` Ferruh Yigit
2018-03-20 16:34 ` [dpdk-dev] [PATCH v5] " Ferruh Yigit
2018-03-21 13:04 ` Neil Horman
2018-03-23 17:00 ` Bruce Richardson
2018-03-24 2:08 ` Neil Horman [this message]
2018-03-26 8:47 ` Ananyev, Konstantin
2018-03-27 19:10 ` Ferruh Yigit
2018-01-17 21:57 ` [dpdk-dev] [PATCH v3 3/6] ethdev: separate driver APIs Ferruh Yigit
2018-01-17 21:58 ` [dpdk-dev] [PATCH v3 4/6] ethdev: separate internal structures into own header Ferruh Yigit
2018-01-17 22:24 ` [dpdk-dev] [dpdk-stable] " Thomas Monjalon
2018-01-18 10:09 ` Ferruh Yigit
2018-01-17 21:58 ` [dpdk-dev] [PATCH v3 5/6] ethdev: reorder inline functions Ferruh Yigit
2018-01-17 21:58 ` [dpdk-dev] [PATCH v3 6/6] ethdev: rename function parameter for consistency Ferruh Yigit
2018-01-17 22:09 ` [dpdk-dev] [PATCH v3 1/6] ethdev: fix port id storage Thomas Monjalon
2018-01-17 22:19 ` Thomas Monjalon
2018-01-18 10:15 ` Ferruh Yigit
2018-01-18 11:29 ` Thomas Monjalon
2018-01-20 16:57 ` [dpdk-dev] [PATCH v4 1/4] ethdev: separate driver APIs Ferruh Yigit
2018-01-20 16:57 ` [dpdk-dev] [PATCH v4 2/4] ethdev: separate internal structures into own header Ferruh Yigit
2018-01-20 16:57 ` [dpdk-dev] [PATCH v4 3/4] ethdev: reorder inline functions Ferruh Yigit
2018-01-20 16:57 ` [dpdk-dev] [PATCH v4 4/4] ethdev: rename function parameter for consistency Ferruh Yigit
2018-01-21 23:12 ` [dpdk-dev] [PATCH v4 1/4] ethdev: separate driver APIs Thomas Monjalon
2018-01-21 23:35 ` [dpdk-dev] [PATCH v5 " Ferruh Yigit
2018-01-21 23:35 ` [dpdk-dev] [PATCH v5 2/4] ethdev: separate internal structures into own header Ferruh Yigit
2018-01-21 23:35 ` [dpdk-dev] [PATCH v5 3/4] ethdev: reorder inline functions Ferruh Yigit
2018-01-21 23:35 ` [dpdk-dev] [PATCH v5 4/4] ethdev: rename function parameter for consistency Ferruh Yigit
2018-01-22 0:16 ` [dpdk-dev] [PATCH v6 1/4] ethdev: separate driver APIs Ferruh Yigit
2018-01-22 0:16 ` [dpdk-dev] [PATCH v6 2/4] ethdev: separate internal structures into own header Ferruh Yigit
2018-01-22 0:16 ` [dpdk-dev] [PATCH v6 3/4] ethdev: reorder inline functions Ferruh Yigit
2018-01-22 0:16 ` [dpdk-dev] [PATCH v6 4/4] ethdev: rename function parameter for consistency Ferruh Yigit
2018-01-22 0:51 ` [dpdk-dev] [PATCH v6 1/4] ethdev: separate driver APIs Thomas Monjalon
2018-03-09 11:27 ` [dpdk-dev] [PATCH v4] ethdev: fix port id storage Ferruh Yigit
2018-03-09 12:58 ` Thomas Monjalon
2018-03-26 20:25 ` Ferruh Yigit
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=20180324020832.GA28338@neilslaptop.think-freely.org \
--to=nhorman@tuxdriver.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@intel.com \
--cc=john.mcnamara@intel.com \
--cc=marko.kovacevic@intel.com \
--cc=reshma.pattan@intel.com \
--cc=thomas@monjalon.net \
/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).