From: Gilles Roudiere <gilles.roudiere@gmail.com>
To: users@dpdk.org
Subject: Re: [dpdk-users] How to fix the number of RX desc to the highest value possible ?
Date: Sun, 5 Feb 2017 11:38:27 +0100 [thread overview]
Message-ID: <CAAaJdm9yVP8cz87vcapuMADMjFROqC9JF1esnfztB_8QGDBbxg@mail.gmail.com> (raw)
In-Reply-To: <CAAaJdm8LL1guR9hcKtY4TgYWsHpXGyH_fcjTDYy=c_D_n0tx+g@mail.gmail.com>
Hello,
In my DPDK app, I would like to fix the number of RX descriptors to the
maximum value available. As this is a trafic capture app (dpdkcap
<https://github.com/dpdkcap/dpdkcap>), I do not want to loose any packet,
so I want the buffers as big as possible.
I have been trying to use the *dev_info.rx_desc_lim.nb_max* field, but it
seems to produce memory allocation failures in some situation. (This value
is obtained using* rte_eth_dev_info_get
<http://dpdk.org/doc/api/rte__ethdev_8h.html#a469a98f6245c85d14f58af8f73168d8b>(portid,
&dev_info) *, and used as the number of requested descriptors when calling*
rte_eth_rx_queue_setup
<http://dpdk.org/doc/api/rte__ethdev_8h.html#a8db38c1bb895250a99b217bb24126d8e>(...)*).
One user is claiming having a message saying that the requested number of
rx desc is not aligned to a power of 2, and that the requested value have
been increased. In this situation I do understand why it does not work (the
value goes over the max availabe), but I do not get why this value is not
given aligned.
Therefore, I wondered what the *dev_info.rx_desc_lim.nb_max *really mean:
is that a maximum per queue ? Or the total number of allowed descriptors ?
And what is the nb_align for ? The documentation
<http://dpdk.org/doc/api/structrte__eth__desc__lim.html> is kind a
incomplete about that. :/
So if any of you has more information about those values, I would really
appreciate it.
Thank you for your time,
Regards,
Gilles Roudiere
parent reply other threads:[~2017-02-05 10:38 UTC|newest]
Thread overview: expand[flat|nested] mbox.gz Atom feed
[parent not found: <CAAaJdm8LL1guR9hcKtY4TgYWsHpXGyH_fcjTDYy=c_D_n0tx+g@mail.gmail.com>]
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=CAAaJdm9yVP8cz87vcapuMADMjFROqC9JF1esnfztB_8QGDBbxg@mail.gmail.com \
--to=gilles.roudiere@gmail.com \
--cc=users@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
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).