DPDK usage discussions
 help / color / mirror / Atom feed
From: Massimo Girondi <massimo@girondi.net>
To: Filip Janiszewski <contact@filipjaniszewski.com>,
	"users@dpdk.org" <users@dpdk.org>
Subject: Re: rte flow rule not clear with DPDK 21.11 and Intel E810
Date: Thu, 2 Dec 2021 12:24:10 +0100	[thread overview]
Message-ID: <c456649c-0a85-f03d-f0de-bf85d8030fd8@girondi.net> (raw)
In-Reply-To: <81372222-08c1-ae43-0962-4670e72fc3aa@filipjaniszewski.com>

Hi Filip,
I saw similar issues on a E810 too when I tested them a while ago.
Looks like the supported patterns and actions are more limited than 
other cards.
I'd suggest to look at the Intel DDP Tech Guide for some hints.
You may also want to play with priorities and pipeline modes, since they 
change what is accepted in terms of patterns.

See below

On 01/12/2021 11.17, Filip Janiszewski wrote:
 > Just updating on other similar issue I'm finding, it appears this rte
 > flow rule is completely broken for Intel E810 and most probably useless,
 > for example, this rule works:
 >
 > .
 > testpmd> flow create 0 ingress pattern eth / ipv4 / tcp dst is 2044 /
 > end actions queue index 1 / end
 > Flow rule #4 created
 > .
 >
 > But the UDP version does not work:
 >
 > .
 > testpmd> flow create 0 ingress pattern eth / ipv4 / udp dst is 51005 /
 > end actions queue index 1 / end
 > Flow rule #1 created
This I think it worked when I tried. Try to match a specific IP.
 > .
 >
 > Also vlan id based filtering do not work, ip ranges (using spec+last) do
 > not work et cetera.
Be careful with the VLAN, only some specific patterns are supported.
Also, check the "protocol extraction per queue" on thje DPDK documentation.
 >
 > Am I doing something wrong with testpmd or anybody can confirm that this
 > is just not working until fixed?
 >
 > Thanks
 >
 > Il 12/1/21 8:48 AM, Filip Janiszewski ha scritto:
 >> Hi,
 >>
 >> Is there any sensible reason for which this flow rule works:
 >>
 >> .
 >> testpmd> flow create 0 ingress pattern eth / ipv4 dst spec 199.168.152.2
 >> dst mask 255.255.0.255 / end actions queue index 1 / end
 >> Flow rule #0 created
 >> .
 >>
 >> But this one not?:
 >>
 >> .
 >> testpmd> flow create 0 ingress pattern eth / ipv4 dst spec 199.168.152.2
 >> dst mask 255.255.255.0 / end actions queue index 1 / end
 >> ice_flow_create(): Failed to create flow
 >> port_flow_complain(): Caught PMD error type 10 (item specification):
 >> cause: 0x7ffe040edaa8, Invalid input set: Invalid argument
This may be due to the 0 not at the end, so that you can't do LPM. I saw 
similar problems on other drivers too with this.
 >> .
 >>
 >> I've just changed the mask from 255.255.0.255 to 255.255.255.0, it's
 >> really confusing why the latter one is not accepted..
 >>
 >> Thanks
 >>
 >
Hope that helps
Massimo


      reply	other threads:[~2021-12-02 11:24 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-01  7:48 Filip Janiszewski
2021-12-01 10:17 ` Filip Janiszewski
2021-12-02 11:24   ` Massimo Girondi [this message]

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=c456649c-0a85-f03d-f0de-bf85d8030fd8@girondi.net \
    --to=massimo@girondi.net \
    --cc=contact@filipjaniszewski.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).