From: "Avi Cohen (A)" <avi.cohen@huawei.com>
To: "Gowda, Sandesh" <sandesh.gowda@intel.com>,
"users@dpdk.org" <users@dpdk.org>
Subject: Re: [dpdk-users] IPSEC-SECGW sample application
Date: Tue, 9 Jan 2018 09:38:41 +0000 [thread overview]
Message-ID: <B84047ECBD981D4B93EAE5A6245AA36101594C58@FRAEML521-MBX.china.huawei.com> (raw)
In-Reply-To: <EDE1359882508442A045AB0CF959E30B70AEC7B8@PGSMSX102.gar.corp.intel.com>
Thank you Sandesh
I'm trying to run the l2fwd-crypto sample app. but got this error msg :
"No crypto devices available
EAL: Error - exiting with code: 1
Cause: Failed to initialize crypto devices"
My setup include 2 physical ports (intel x540) bound to dpdk - these Nics are ipsec offload capable.
The cmd-line I use is:
./l2fwd-crypto -l 0-1 -n 4 -- -p 0x3 --cdev_type HW --chain CIPHER_HASH --cipher_op ENCRYPT --cipher_algo aes-cbc --cipher_key 00:01:02:03:04:05:06:07:08:09:0a:0b:0c:0d:0e:0f --auth_op GENERATE --auth_algo aes-xcbc-mac --auth_key 10:11:12:13:14:15:16:17:18:19:1a:1b:1c:1d:1e:1f
Best Regards
Avi
> -----Original Message-----
> From: Gowda, Sandesh [mailto:sandesh.gowda@intel.com]
> Sent: Monday, 08 January, 2018 7:23 PM
> To: Avi Cohen (A); users@dpdk.org
> Subject: RE: IPSEC-SECGW sample application
>
> Hi Avi,
>
> My response inline.
>
> > 1. I see in the documentation that this app. Supports only **complete
> > offload**.
> > But Intel NICS x540 and 82599 which supports ipsec offload requires
> > that the SW will add/remove the ESP headers How can I run this app with
> x540 nic ?
>
> The SA rule "type" field lets you choose the kind of offload.
> Following is the description from the ipsecgw app guide:
>
> <type>
>
> Action type to specify the security action. This option specify the SA to be
> performed with look aside protocol offload to HW accelerator or protocol
> offload on ethernet device or inline crypto processing on the ethernet device
> during transmission.
> Optional: Yes, default type no-offload
> Available options:
> lookaside-protocol-offload: look aside protocol offload to HW accelerator
> inline-protocol-offload: inline protocol offload on ethernet device
> inline-crypto-offload: inline crypto processing on ethernet device
> no-offload: no offloading to hardware
>
> Correct your SA rules to have the desired "type" field.
>
> The ipsecgw application must work fine for QAT PCIe as well as Ethernet NIC
> with IPSec feature provided the VFs as correctly bound to DPDK.
>
>
> > 2. I added support for ESP header and trailer insertion for
> > inline-protocol- offload for intel x540 Can you tell me the exact
> > command line to run the application for this mode ?
> > is vdev required ?
>
> The ipsecgw application must work fine for QAT PCIe as well as Ethernet NIC
> with IPSec feature provided the VFs as correctly bound to DPDK.
> Please try running a more basic L2Fwd Crypto application on your NIC to make
> sure the Crypto feature works.
>
> Regards,
> Sandesh
>
>
>
> > -----Original Message-----
> > From: Avi Cohen (A) [mailto:avi.cohen@huawei.com]
> > Sent: Monday, January 08, 2018 10:05 PM
> > To: Gowda, Sandesh <sandesh.gowda@intel.com>; users@dpdk.org
> > Subject: RE: IPSEC-SECGW sample application
> >
> >
> > Hi Sandesh [I added one more question] Thank you - I already
> > understood that.
> > 1. I see in the documentation that this app. Supports only **complete
> > offload**.
> > But Intel NICS x540 and 82599 which supports ipsec offload requires
> > that the SW will add/remove the ESP headers How can I run this app with
> x540 nic ?
> >
> > 2. I added support for ESP header and trailer insertion for
> > inline-protocol- offload for intel x540 Can you tell me the exact
> > command line to run the application for this mode ?
> > is vdev required ?
> > Best Regards
> > Avi
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: Gowda, Sandesh [mailto:sandesh.gowda@intel.com]
> > > > Sent: Monday, 08 January, 2018 10:47 AM
> > > > To: Avi Cohen (A); users@dpdk.org
> > > > Subject: RE: IPSEC-SECGW sample application
> > > >
> > > >
> > > > Hi Avi,
> > > >
> > > > The application classifies the ports as Protected and Unprotected.
> > > > Thus,
> > > traffic
> > > > received on an Unprotected or Protected port is consider Inbound
> > > > or
> > > Outbound
> > > > respectively.
> > > > ( Refer :
> > > > http://dpdk.org/doc/guides/sample_app_ug/ipsec_secgw.html
> > > > )
> > > >
> > > > The Packets sent on a Unprotected network requires Encryption
> > > > whereas packets on Protected Network can be plain text.
> > > > This is the expected behavior.
> > > >
> > > > Regards,
> > > > Sandesh
> > > >
> > > >
> > > >
> > > >
> > > > -----Original Message-----
> > > > From: users [mailto:users-bounces@dpdk.org] On Behalf Of Avi Cohen
> > > > (A)
> > > > Sent: Sunday, January 07, 2018 9:12 PM
> > > > To: users@dpdk.org
> > > > Subject: [dpdk-users] IPSEC-SECGW sample application
> > > >
> > > >
> > > > Hello
> > > > I'm using the DPDK17.11 and running the sample app. Ipsec_secgw.
> > > > I have 2 ports port 0 is protected and port 1 is unprotected
> > > > Traffic is received
> > > in
> > > > the unprotected and should be sent to the protected port for
> > > > encryption But the traffic processing for the traffic received in
> > > > the unprotected port is going through the **process_pkts_inbound ** .
> > > > I expect that the traffic should be directed to the
> > > **process_pkts_outbound**
> > > > [where ESP headers are added etc.] Can someone help ?
> > > >
> > > >
> > > > This is the config file:
> > > >
> > > > #SP rules
> > > > sp ipv4 in esp protect 5 src 1.1.1.2/32 dst 1.1.2.10/32 #SA rules
> > > > sa in 5 cipher_algo aes-128-cbc cipher_key
> > > > 0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0 \ auth_algo sha1-hmac auth_key
> > > > 0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0:0 \ mode ipv4-tunnel src
> > > > 172.16.1.5 dst 172.16.2.5 \ type inline-protocol-offload port_id 0
> > > > #Routing rules rt ipv4 dst 172.16.2.5/32 port 0 rt ipv4 dst
> > > > 1.1.2.0/24 port 0 rt ipv4 dst
> > > > 1.1.1.0/24 port 0
> > > >
> > > >
> > > > and this is the command line to run the applic:
> > > >
> > > > ./ipsec-secgw -l 1 -n 2 -- -p 0x3 -P -u 0x2
> > > > --config="(0,0,1),(1,0,1)" -f ../ep1.cfg
> > > >
> > > >
> > > > Best Regards
> > > > Avi
next prev parent reply other threads:[~2018-01-09 9:38 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-01-07 15:41 Avi Cohen (A)
2018-01-08 8:47 ` Gowda, Sandesh
2018-01-08 11:47 ` Avi Cohen (A)
2018-01-08 16:34 ` Avi Cohen (A)
2018-01-08 17:23 ` Gowda, Sandesh
2018-01-09 9:38 ` Avi Cohen (A) [this message]
2018-01-09 10:00 ` De Lara Guarch, Pablo
2018-01-09 15:15 ` Avi Cohen (A)
2018-01-14 13:53 ` Avi Cohen (A)
2018-01-15 13:54 ` Avi Cohen (A)
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=B84047ECBD981D4B93EAE5A6245AA36101594C58@FRAEML521-MBX.china.huawei.com \
--to=avi.cohen@huawei.com \
--cc=sandesh.gowda@intel.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).