DPDK usage discussions
 help / color / mirror / Atom feed
From: David Griswold <DGriswold@advaoptical.com>
To: "Trahe, Fiona" <fiona.trahe@intel.com>,
	"users@dpdk.org" <users@dpdk.org>
Subject: Re: [dpdk-users] Crypto Framework with QAT Device
Date: Tue, 9 Aug 2016 14:17:53 +0000	[thread overview]
Message-ID: <c8b7134108c342a4a90e693a0a91041f@advaoptical.com> (raw)
In-Reply-To: <348A99DA5F5B7549AA880327E580B4358907B371@IRSMSX101.ger.corp.intel.com>


Thanks Fiona,

Unfortunately, I'm wanting to run on C2xxx (Atom) devices.  I don't believe any of these SKUs have the newer QAT device. Is that correct?
I did get it working with the QAT sample application as you point out.
Just for context, could you explain why the cryptodev framework did not use the ADF driver's API, like the QAT sample uses?
I suspect it had to do with performance but perhaps the API is limiting in some way?

Thanks for your help,

David


From: Trahe, Fiona [mailto:fiona.trahe@intel.com]
Sent: Tuesday, August 09, 2016 10:08 AM
To: users@dpdk.org; David Griswold <DGriswold@advaoptical.com>
Cc: Trahe, Fiona <fiona.trahe@intel.com>
Subject: re: Crypto Framework with QAT Device

Hi David,

The cryptodev framework can't be used with the earlier QuickAssist devices.(8900-8920)
Even if you got past the differing CSRs the firmware interface is significantly different, so a
complete rewrite of the code which builds the message descriptors would be necessary.
Sorry, that's probably not what you wanted to hear.

A dpdk sample app wrapping the released Intel QuickAssist driver was developed prior to the cryptodev framework:
http://dpdk.org/doc/guides/sample_app_ug/intel_quickassist.html?highlight=quickassist

A better alternative would be to move to one of the newer QuickAssist devices, is that an option?

Regards,
Fiona


Hello,

I'd like to use the crypto framework with an older QAT device from the 8900-8920 (QAT1.5) series.
The current support is for the 8925-8955 series (QAT1.6).
I've installed the ADF software and gotten the crypto framework to recognize the PCIe device,
but the BAR register layout seems to be different from what the framework is expecting when it
goes to create the queue pairs.

Could the experts comment on how difficult this would be and if possible what my next steps might be?

Thanks in advance for your advice,

David

  reply	other threads:[~2016-08-09 14:17 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-08-09 14:07 Trahe, Fiona
2016-08-09 14:17 ` David Griswold [this message]
2016-08-09 15:15   ` Trahe, Fiona
  -- strict thread matches above, loose matches on Subject: below --
2016-08-04 17:48 David Griswold

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=c8b7134108c342a4a90e693a0a91041f@advaoptical.com \
    --to=dgriswold@advaoptical.com \
    --cc=fiona.trahe@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).