From: "De Lara Guarch, Pablo" <pablo.de.lara.guarch@intel.com>
To: Ravi Kumar <Ravi1.kumar@amd.com>, "dev@dpdk.org" <dev@dpdk.org>
Cc: "thomas@monjalon.net" <thomas@monjalon.net>,
"hemant.agrawal@nxp.com" <hemant.agrawal@nxp.com>
Subject: Re: [dpdk-dev] [PATCH 2/2] crypto/ccp: change to command line option from compile time
Date: Wed, 9 May 2018 08:34:52 +0000 [thread overview]
Message-ID: <E115CCD9D858EF4F90C690B0DCB4D8976CD04BE2@IRSMSX108.ger.corp.intel.com> (raw)
In-Reply-To: <1525327041-128382-2-git-send-email-Ravi1.kumar@amd.com>
> -----Original Message-----
> From: Ravi Kumar [mailto:Ravi1.kumar@amd.com]
> Sent: Thursday, May 3, 2018 6:57 AM
> To: dev@dpdk.org
> Cc: De Lara Guarch, Pablo <pablo.de.lara.guarch@intel.com>;
> thomas@monjalon.net; hemant.agrawal@nxp.com
> Subject: [PATCH 2/2] crypto/ccp: change to command line option from compile
> time
>
> CCP PMD supports authentication offload to either of CCP or CPU. The earlier
> version of patch provided this option as compile time. This patch changes this
> option from compile time to run time. User can pass "ccp_auth_opt=1" as an
> additional argument to vdev parameter to enable authentication operations on
> CPU.
>
> Signed-off-by: Ravi Kumar <Ravi1.kumar@amd.com>
...
> +++ b/drivers/crypto/ccp/ccp_pmd_ops.c
> @@ -13,7 +13,6 @@
> #include "ccp_crypto.h"
>
> static const struct rte_cryptodev_capabilities ccp_pmd_capabilities[] = { -#ifdef
> RTE_LIBRTE_PMD_CCP_CPU_AUTH
> { /* MD5 HMAC */
> .op = RTE_CRYPTO_OP_TYPE_SYMMETRIC,
> {.sym = {
> @@ -35,7 +34,6 @@ static const struct rte_cryptodev_capabilities
> ccp_pmd_capabilities[] = {
> }, }
> }, }
> },
> -#endif
MD5 won't be supported unless the auth_opt is enabled.
I would suggest doing something similar to what QAT does.
You can have an array with the base capabilities and another one
for the extended ones (with MD5).
When "auth_opt" is enabled, you can have the combination of both.
When it is not, then you can just have the base capabilities.
> { /* SHA1 */
> .op = RTE_CRYPTO_OP_TYPE_SYMMETRIC,
> {.sym = {
...
> RTE_LOG(INFO, PMD, "Initialising %s on NUMA node %d\n", name,
> - init_params.socket_id);
> + init_params.def_p.socket_id);
> RTE_LOG(INFO, PMD, "Max number of queue pairs = %d\n",
> - init_params.max_nb_queue_pairs);
> + init_params.def_p.max_nb_queue_pairs);
> RTE_LOG(INFO, PMD, "Max number of sessions = %d\n",
> - init_params.max_nb_sessions);
> + init_params.def_p.max_nb_sessions);
> + RTE_LOG(INFO, PMD, "Authentication offload to %s\n",
> + ((init_params.auth_opt == 0) ? "CCP" : "CPU"));
> + snprintf(die_num, 20, "die_num-%d", init_params.die_num);
> + RTE_LOG(INFO, PMD, "Combining %s CCP devices\n",
> + ((init_params.die_num == -1) ? "all" : die_num));
>
> rc = cryptodev_ccp_create(name, vdev, &init_params);
> if (rc)
> --
> 2.7.4
You are adding new parameters, so you should extend the parameters registration at the bottom of this file:
RTE_PMD_REGISTER_PARAM_STRING(CRYPTODEV_NAME_CCP_PMD,
"max_nb_queue_pairs=<int> max_nb_sessions=<int> socket_id=<int>");
next prev parent reply other threads:[~2018-05-09 8:34 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-05-03 5:57 [dpdk-dev] [PATCH 1/2] crypto/ccp: enable CCP crypto poll mode driver by default Ravi Kumar
2018-05-03 5:57 ` [dpdk-dev] [PATCH 2/2] crypto/ccp: change to command line option from compile time Ravi Kumar
2018-05-09 8:34 ` De Lara Guarch, Pablo [this message]
2018-05-03 7:23 ` [dpdk-dev] [PATCH 1/2] crypto/ccp: enable CCP crypto poll mode driver by default De Lara Guarch, Pablo
2018-05-09 5:53 ` Kumar, Ravi1
2018-05-09 7:58 ` Thomas Monjalon
2018-05-09 8:11 ` Kumar, Ravi1
2018-05-09 8:19 ` Thomas Monjalon
2018-05-09 8:36 ` De Lara Guarch, Pablo
2018-05-09 8:58 ` Kumar, Ravi1
2018-05-09 12:46 ` [dpdk-dev] [PATCH v2] crypto/ccp: support command line option to enable CPU based auth Ravi Kumar
2018-05-09 15:27 ` De Lara Guarch, Pablo
2018-05-10 8:43 ` Kumar, Ravi1
2018-05-10 8:46 ` De Lara Guarch, Pablo
2018-05-10 9:20 ` [dpdk-dev] [PATCH v3] " Ravi Kumar
2018-05-10 12:22 ` De Lara Guarch, Pablo
2018-05-11 5:37 ` Kumar, Ravi1
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=E115CCD9D858EF4F90C690B0DCB4D8976CD04BE2@IRSMSX108.ger.corp.intel.com \
--to=pablo.de.lara.guarch@intel.com \
--cc=Ravi1.kumar@amd.com \
--cc=dev@dpdk.org \
--cc=hemant.agrawal@nxp.com \
--cc=thomas@monjalon.net \
/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).