Re: [dpdk-users] users Digest, Vol 176, Issue 6

Re: [dpdk-users] users Digest, Vol 176, Issue 6

[puneet_kumar kumar]

Hi Everyone, 
I am trying to install dpdk on freebsd and got stuck at one point. DPDK is in /usr/ports/net/dpdk. It has been compiled there. I also compiled via downloading dpdk from git.
Now I get these following errors


# kenv hw.contigmem.buffer_size=10737418

hw.contigmem.buffer_size="10737418"

# sysctl -a hw.contigmem

sysctl: unknown oid 'hw.contigmem'

# kldload ./kmod/contigmem.ko

kldload: can't load ./kmod/contigmem.ko: No such file or directory

# setenv contigmem_load "YES"

# kldload ./kmod/contigmem.ko

kldload: can't load ./kmod/contigmem.ko: No such file or directory

# kldload ./kmod/nic_uio.ko
kldload: can't load ./kmod/nic_uio.ko: No such file or directory

Can someone please help me? It is really urgent :(
Puneet


    On Saturday, 9 March, 2019, 3:00:05 am GMT-8, <users-request@dpdk.org> wrote:  
 
 Send users mailing list submissions to
    users@dpdk.org

To subscribe or unsubscribe via the World Wide Web, visit
    https://mails.dpdk.org/listinfo/users
or, via email, send a message with subject or body 'help' to
    users-request@dpdk.org

You can reach the person managing the list at
    users-owner@dpdk.org

When replying, please edit your Subject line so it is more specific
than "Re: Contents of users digest..."


Today's Topics:

  1. Cannot enable VFs for QAT PF (Changchun Zhang)
  2. Re: QAT cryto PMD queue pair and Ring Libary (Changchun Zhang)


----------------------------------------------------------------------

Message: 1
Date: Fri, 8 Mar 2019 13:13:08 -0800 (PST)
From: Changchun Zhang <changchun.zhang@oracle.com>
To: users@dpdk.org
Subject: [dpdk-users] Cannot enable VFs for QAT PF
Message-ID: <6ec28615-daea-473e-9f4f-cee9b6237207@default>
Content-Type: text/plain; charset=us-ascii

Hi,

 

I cannot enable VFs for the QAT device, my steps are as follows:

1.    Enable VT-d in BIOS

2.    Add the "intel_iommu=on iommu=pt" in the kernel command line

3.    QAT device driver has been inserted:

~ # lsmod | grep qa

qat_c62x              16384  0 

intel_qat            114688  1 qat_c62x

 

4.    Find the BDF of the PF:

~ # lspci -d:37c8

b5:00.0 Class 0b40: Device 8086:37c8 (rev 04)

b7:00.0 Class 0b40: Device 8086:37c8 (rev 04)

 

5.    Trying to enable VFs, but I did not find sriov_numvfs in directories:

/sys/bus/pci/drivers/c6xx

/sys/bus/pci/drivers/c6xx # ls -ltr 

--w-------    1      4096 Jan 22 21:41 uevent

--w-------    1      4096 Jan 22 21:43 unbind

--w-------    1      4096 Jan 22 21:43 remove_id

--w-------    1      4096 Jan 22 21:43 new_id

lrwxrwxrwx    1        0 Jan 22 21:43 module -> ../../../../module/qat_c62x

--w-------    1      4096 Jan 22 21:43 bind

 

But I can find sriov_numvfs in other directories:

/sys # find ./ -name sriov_numvfs

./devices/pci0000:b2/0000:b2:00.0/0000:b3:00.0/0000:b4:00.0/0000:b5:00.0/sriov_numvfs

./devices/pci0000:b2/0000:b2:00.0/0000:b3:00.0/0000:b4:02.0/0000:b7:00.0/sriov_numvfs

 

Unfortunately, when trying to enable the VFs with above sriov_numvfs file, cannot write to this file, like:

/sys # echo 16 > ./devices/pci0000:b2/0000:b2:00.0/0000:b3:00.0/0000:b4:02.0/0000:b7:00.0/sriov_numvfs 

sh: write error: No such file or directory

 

Can any guys know some clue about it?

Thanks,

 


------------------------------

Message: 2
Date: Fri, 8 Mar 2019 13:14:56 -0800 (PST)
From: Changchun Zhang <changchun.zhang@oracle.com>
To: "Trahe, Fiona" <fiona.trahe@intel.com>, users@dpdk.org
Subject: Re: [dpdk-users] QAT cryto PMD queue pair and Ring Libary
Message-ID: <2e0a0ee6-e626-415f-82ec-2eeffc0f7048@default>
Content-Type: text/plain; charset=us-ascii

Thanks.
The related question, for Ethernet device, if it has been bound to igb_uio driver, who is going to actually implement the rx_ring and tx_ring for the Ethernet PMD? The igb_uio driver?

Changchun (Alex)


-----Original Message-----
From: Trahe, Fiona [mailto:fiona.trahe@intel.com] 
Sent: Friday, March 08, 2019 5:56 AM
To: Changchun Zhang <changchun.zhang@oracle.com>; users@dpdk.org
Cc: Trahe, Fiona <fiona.trahe@intel.com>
Subject: RE: [dpdk-users] QAT cryto PMD queue pair and Ring Libary


> -----Original Message-----
> From: users [mailto:users-bounces@dpdk.org] On Behalf Of Changchun 
> Zhang
> Sent: Thursday, March 7, 2019 9:29 PM
> To: users@dpdk.org
> Subject: [dpdk-users] QAT cryto PMD queue pair and Ring Libary
> 
> Hi,
> 
> 
> 
> QAT PMD driver uses the queue pair to communication with QAT device. 
> My question is, whether or this QAP queue pair is implemented over DPDK ring library or just self-defined queue implementation?
[Fiona] This doesn't use the DPDK ring library. They are QAT-specific, hw-supported tx and rx rings.


End of users Digest, Vol 176, Issue 6
*************************************
  
From tomas.jansky@flowmon.com  Mon Mar 11 10:09:19 2019
Return-Path: <tomas.jansky@flowmon.com>
Received: from mail-it1-f181.google.com (mail-it1-f181.google.com
 [209.85.166.181]) by dpdk.org (Postfix) with ESMTP id 0BA144C9F
 for <users@dpdk.org>; Mon, 11 Mar 2019 10:09:18 +0100 (CET)
Received: by mail-it1-f181.google.com with SMTP id z131so5989541itf.5
 for <users@dpdk.org>; Mon, 11 Mar 2019 02:09:18 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=flowmon.com; s=google;
 h=mime-version:from:date:message-id:subject:to;
 bh=6zW36vd09d0J+Zod2M3TOFxk7xWZ5hOEujIq8w2PVhw=;
 b=zO6TeQrWhTRnPvaqHx8wpSAMn9h1bZ3K4zPqOTIHL6NDmSMQQ/ZBmF84hyBpkCR8dd
 nw/MSs33H83Flhml2xlrMf453C3LAHfSwHd8x8ZNlcZ4CM/FhfVS0I854JJbqJleX64h
 kHZxXnyOysZpTw8i3Q3c9uLoaRQYawGh27Tgn4/oU0s25rwq5ShzeN/tiJ3naCM2YLHd
 bUWFHyAZso4MRbTYQ5a4Xxum/LKFaRmk8VNVdO81tllVcDw2WhpKGF0tsCJi3sUTAyka
 HzrsfWftY4b32chVdAw+bf0KoZOpF85tVfg7Q5PlBP8kBAOdQVxMttL+4Wyhoc7JeoXK
 OQxg=X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d\x1e100.net; s 161025;
 h=x-gm-message-state:mime-version:from:date:message-id:subject:to;
 bh=6zW36vd09d0J+Zod2M3TOFxk7xWZ5hOEujIq8w2PVhw=;
 b=i9oRYaaeskkZL0HVsF8vGeXtwHmPKum3EE4vTv5wtFnSx792YQHZ49JA9li79bU0Uj
 56BxBJ/0U9EfPnI6fGDYosyktO+uHYT+CBcbmZItFihMnww93Yu7kNRDH1MhP+QA7VhX
 RsP9yvF2LrEQAG2FH/3UB3X5/aQm0Vpf022hfyP0sbqvMHRI1mE1VqN2sSyJ3KpJojEx
 DuDmxPzLVmyVmFPN5b8P2vjKs23sxA9J3dJZXbLEjEMw9ZjAvDsVfu9GaM7JVY/aQTtu
 y7ZUGXVlTRYyCS9KJHNjeSsRq9mYrXq3bYGPR5hroGh6n0/+cAArs33Pj6TMpsZ4yZyB
 9zlQ=X-Gm-Message-State: APjAAAXlNCtz+dJbpWCLTrxoytYJ7BYK4ZhkRyjeQ9gEFoGlY8lXxS1Q
 bCm2fsxV2EXFd/bHPWi8mijQjfXJoAQuXOezAx7L9mVx7REX-Google-Smtp-Source: APXvYqziGNx6G3/PFK2IkfkfN0vEGFM/HHoepc7EuIUTD9oPiN1zW+yfA4VIQcxGv47rg4/Wxiy4g8BZjwmmLAi1RHcX-Received: by 2002:a02:228f:: with SMTP id o137mr17121007jao.76.1552295358267;
 Mon, 11 Mar 2019 02:09:18 -0700 (PDT)
MIME-Version: 1.0
From: =?UTF-8?B?VG9tw6HFoSBKw6Fuc2vDvQ==?= <tomas.jansky@flowmon.com>
Date: Mon, 11 Mar 2019 10:09:07 +0100
Message-ID: <CAPP7y6whcx3NCS3CBeLimYrx6s8bJ5O8ZOPsmATWqSi5k6UAVw@mail.gmail.com>
To: users@dpdk.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Content-Filtered-By: Mailman/MimeDel 2.1.15
Subject: [dpdk-users] Outer VLAN stripping X710 NIC
X-BeenThere: users@dpdk.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: DPDK usage discussions <users.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/users>,
 <mailto:users-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/users/>
List-Post: <mailto:users@dpdk.org>
List-Help: <mailto:users-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/users>,
 <mailto:users-request@dpdk.org?subject=subscribe>
X-List-Received-Date: Mon, 11 Mar 2019 09:09:19 -0000

Hello DPDK Team,

I have already asked this once, however nobody responded, so I am giving it
one last try.

I am having an issue with the behavior of Intel X710 NIC and would like
some clarification.
When the card receives packets with QinQ (double 802.1Q) the RX_RSS based
on L3 and L4 information is not working without specifying
DEV_RX_OFFLOAD_VLAN_EXTEND.
However, by specifying the mentioned flag, the card automatically strips
outer VLAN from the packet (even without using other offload flags like
DEV_RX_OFFLOAD_VLAN_STRIP) while leaving only the inner VLAN in the packet.
I know that the TCI from the stripped VLAN can be found in rte_mbuf
structure, but is it possible to force the card to keep the outer VLAN in
the packet (and still maintain working RSS)?
I do not know, whether this is an issue with DPDK design or the X710 NIC
itself. I was unable to find any useful information in the card datasheet.

Any clarification on this topic would be greatly appreciated.
Thanks in advance.

All the best
Tom

Tomáš Jánský | Developer | Flowmon Networks

M: +420 736 203 153 | W: www.flowmon.com

[image: Flowmon_logo_mail.png] <https://www.flowmon.com>