DPDK patches and discussions
 help / color / mirror / Atom feed
From: kefu chai <tchaikov@gmail.com>
To: Bruce Richardson <bruce.richardson@intel.com>,
	Avi Kivity <avi@scylladb.com>
Cc: dev@dpdk.org, anatoly.burakov@intel.com
Subject: Re: [dpdk-dev] [PATCH v4] build: add meson options of max_memseg_lists and atomic_mbuf_ref_counts
Date: Mon, 27 Sep 2021 23:03:12 +0800
Message-ID: <CAJE9aOOFq=z6wSSFF=MhSkCCz-2vfx28Fwc5w9VSDYChOtPU2A@mail.gmail.com> (raw)
In-Reply-To: <YUhBdHm8hHAxZHL+@bricha3-MOBL.ger.corp.intel.com>

On Mon, Sep 20, 2021 at 4:08 PM Bruce Richardson
<bruce.richardson@intel.com> wrote:
>
> On Mon, Sep 20, 2021 at 03:51:06PM +0800, kefu chai wrote:
> > hello Bruce,
> >
> > do you have any further concerns? is there anything i can do to move
> > this forward?
> >
> > cheers,
> >
>
> +Anatoly, for his input for the memory segments change.
>
> I still would prefer not to have these as config options, but perhaps one
> or both needs to be. The atomic refcount seems more reasonable to add of
> the two. For the max memseg lists, what is the impact if we were to
> increase this value globally?

hi Bruce, thank you for your insights.

yeah, as i explained in the previous email.the atomic refcount is more
critical for my work on integration of DPDK+SPDK+Seastar. since
Seastar enforces share-nothing in its design, there is no need to use
atomic refcount under almost all circumstances. regarding to the max
memseg list, what i am trying is to port the change of
https://github.com/scylladb/seastar/commit/716c7c04db693c266f52de6b0cced0252d70b3bf
to the DPDK used by the latest release of SPDK.

i am copying Avi also for his input. he is the author of the change above.

cheers,

>
> /Bruce
>
> > On Thu, Sep 9, 2021 at 12:51 AM Kefu Chai <tchaikov@gmail.com> wrote:
> > >
> > > RTE_MAX_MEMSEG_LISTS = 128 is not enough for high-memory machines, in our
> > > case, we need to increase it to 8192. so add an option so user can
> > > override it. RTE_MBUF_REFCNT_ATOMIC = 0 is not necessary for applications
> > > like Seastar, where it's safe to assume that the mbuf refcnt is only
> > > updated by a single core only.
> > >
> > > ---
> > >
> > > v4:
> > >
> > > fix the coding style issue by reduce the line length to under 75.
> > > this change should silence the warning like:
> > >
> > > WARNING:COMMIT_LOG_LONG_LINE: Possible unwrapped commit description (prefer a maximum 75 chars per line)
> > > #81:
> > > RTE_MAX_MEMSEG_LISTS = 128 is not enough for high-memory machines, in our case,
> > >
> > > total: 0 errors, 1 warnings, 35 lines checked
> > >
> > > Signed-off-by: Kefu Chai <tchaikov@gmail.com>
> > > ---
> > >  config/meson.build  | 5 ++++-
> > >  config/rte_config.h | 2 --
> > >  meson_options.txt   | 4 ++++
> > >  3 files changed, 8 insertions(+), 3 deletions(-)
> > >
>
> <snip>



-- 
Regards
Kefu Chai

  reply	other threads:[~2021-09-27 15:03 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-02 14:48 [dpdk-dev] [PATCH] build: add meson option of "max_memseg_lists" and "mbuf_refcnt_atomic" Kefu Chai
2021-09-02 15:02 ` Bruce Richardson
2021-09-02 15:05 ` [dpdk-dev] [PATCH v2] " Kefu Chai
2021-09-02 15:11 ` [dpdk-dev] [PATCH v3] build: add meson options of max_memseg_lists and atomic_mbuf_ref_counts Kefu Chai
2021-09-08 16:51   ` [dpdk-dev] [PATCH v4] " Kefu Chai
2021-09-20  7:51     ` kefu chai
2021-09-20  8:08       ` Bruce Richardson
2021-09-27 15:03         ` kefu chai [this message]
2021-10-13 15:38           ` Thomas Monjalon
2021-10-13 20:54 ` [dpdk-dev] [PATCH v5 0/2] build: add meson option of "max_memseg_lists" and "mbuf_refcnt_atomic" Kefu Chai
2021-10-13 20:54 ` [dpdk-dev] [PATCH v5 1/2] build: add meson options of atomic_mbuf_ref_counts Kefu Chai
2021-10-14  8:20   ` Bruce Richardson
2021-10-25 15:55     ` Thomas Monjalon
2021-10-13 20:54 ` [dpdk-dev] [PATCH v5 2/2] build: add meson options of max_memseg_lists Kefu Chai
2021-10-14  8:25   ` Bruce Richardson
2021-10-14  8:29     ` Thomas Monjalon

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='CAJE9aOOFq=z6wSSFF=MhSkCCz-2vfx28Fwc5w9VSDYChOtPU2A@mail.gmail.com' \
    --to=tchaikov@gmail.com \
    --cc=anatoly.burakov@intel.com \
    --cc=avi@scylladb.com \
    --cc=bruce.richardson@intel.com \
    --cc=dev@dpdk.org \
    /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

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror http://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ http://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git