From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <olivier.matz@6wind.com>
Received: from mail-wm0-f47.google.com (mail-wm0-f47.google.com [74.125.82.47])
 by dpdk.org (Postfix) with ESMTP id D929858D1
 for <dev@dpdk.org>; Fri, 30 Jun 2017 16:13:06 +0200 (CEST)
Received: by mail-wm0-f47.google.com with SMTP id 62so111540454wmw.1
 for <dev@dpdk.org>; Fri, 30 Jun 2017 07:13:06 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d=6wind-com.20150623.gappssmtp.com; s=20150623;
 h=date:from:to:cc:subject:message-id:in-reply-to:references
 :mime-version:content-transfer-encoding;
 bh=whvenEmJi6p/A4R+kNRo5udFdU65GIS3/cFnTf4Ru5o=;
 b=PvxARVlYRMnSGayqJPUIPbGKQtUg7XQQg5a4xxKdYTuNKE49elWl3EYNFuwviaU1HP
 CyS2w2Qao2u4CqC0hB7lWBoKfLXRG+T2zyajpeBNgjHSOGyoDEKl6L53uzV5yp4PVc0F
 OIXGcUXZPtI0FeNZ6UNqwlG6yr2RffLnIcORauNLVtl5T/2AafIpvJ4uYp4FpmdUu2wA
 27mXT+SOBhb3O3ijyoafyqMB+20RSHgze6WGbMMeFQmbCZhHyc8IFkaWDEJK6Lkxc8YT
 DydYM3vKyQNA766rNmBaPtKyO6tXbmTmZXPnB4WteZsQHbsX8Ifgs7GAt/yaXAFMc62J
 wJxQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d=1e100.net; s=20161025;
 h=x-gm-message-state:date:from:to:cc:subject:message-id:in-reply-to
 :references:mime-version:content-transfer-encoding;
 bh=whvenEmJi6p/A4R+kNRo5udFdU65GIS3/cFnTf4Ru5o=;
 b=UT+s3k/5XW4SOH5rK3YfHd+1XM4TidNZ7Y4sJgMYoYcW9cBfUVe6Ksn8FVUxMhnbxE
 nRFSBaGzUVvkZlhTfpAcNQRlhstibcHlrAjYTEDTyPXZ+IyaZsGMtRZ3tw86ETBVOVL2
 +Y+NmtXCAlw/NGaLH5NGnGF6GKgrw7pasBT2rVRNeqjcP818wWZFuE80uG7ufPdSePPc
 /gudwqugBbB6xymTvtZVT5DEyaXj48kGWn/+BerSoTHN+caEYgUlMm7vTmhp2s0GdUIE
 nEbA8aeuXwY2YB34V6JmfW9GSD97NcHGihr+9XH1wSttOR+edDOQk0IEFebrTA9v/CeH
 OFuA==
X-Gm-Message-State: AKS2vOweFIO3USL6kj35110xGXPvHSFgVIQSAfPp+BzZLou0Ex9/mrCf
 7iUa/7WMSu+lRMcg
X-Received: by 10.28.35.201 with SMTP id j192mr15128868wmj.26.1498831985236;
 Fri, 30 Jun 2017 07:13:05 -0700 (PDT)
Received: from platinum (2a01cb0c03c651000226b0fffeed02fc.ipv6.abo.wanadoo.fr.
 [2a01:cb0c:3c6:5100:226:b0ff:feed:2fc])
 by smtp.gmail.com with ESMTPSA id c55sm7730230wrc.7.2017.06.30.07.13.04
 (version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256);
 Fri, 30 Jun 2017 07:13:05 -0700 (PDT)
Date: Fri, 30 Jun 2017 16:13:02 +0200
From: Olivier Matz <olivier.matz@6wind.com>
To: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Cc: dev@dpdk.org, hemant.agrawal@nxp.com, jerin.jacob@caviumnetworks.com
Message-ID: <20170630161302.1c11ca46@platinum>
In-Reply-To: <20170601080559.10684-3-santosh.shukla@caviumnetworks.com>
References: <20170601080559.10684-1-santosh.shukla@caviumnetworks.com>
 <20170601080559.10684-3-santosh.shukla@caviumnetworks.com>
X-Mailer: Claws Mail 3.14.1 (GTK+ 2.24.31; x86_64-pc-linux-gnu)
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Subject: Re: [dpdk-dev] [PATCH 2/2] ether/ethdev: Allow pmd to advertise
 preferred pool capability
X-BeenThere: dev@dpdk.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <http://dpdk.org/ml/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <http://dpdk.org/ml/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
X-List-Received-Date: Fri, 30 Jun 2017 14:13:07 -0000

On Thu,  1 Jun 2017 13:35:59 +0530, Santosh Shukla <santosh.shukla@caviumnetworks.com> wrote:
> Platform with two different NICs like external PCI NIC and
> Integrated NIC, May want to use their preferred pool handle.
> Right now there is no way that two different NICs on same board,
> Could use their choice of a pool.
> Both NICs forced to use same pool, Which is statically configured
> by setting CONFIG_RTE_MEMPOOL_DEFAULT_OPS=<pool-name>.
> 
> So Introducing get_preferred_pool() API. Which allows PMD driver
> to advertise their pool capability to Application.
> Based on that hint, Application creates separate pool for
> That driver.
> 
> Signed-off-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
> ---
>  lib/librte_ether/rte_ethdev.c          | 16 ++++++++++++++++
>  lib/librte_ether/rte_ethdev.h          | 21 +++++++++++++++++++++
>  lib/librte_ether/rte_ether_version.map |  7 +++++++
>  3 files changed, 44 insertions(+)
> 
> diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
> index 83898a8f7..4068a05b1 100644
> --- a/lib/librte_ether/rte_ethdev.c
> +++ b/lib/librte_ether/rte_ethdev.c
> @@ -3472,3 +3472,19 @@ rte_eth_dev_l2_tunnel_offload_set(uint8_t port_id,
>  				-ENOTSUP);
>  	return (*dev->dev_ops->l2_tunnel_offload_set)(dev, l2_tunnel, mask, en);
>  }
> +
> +int
> +rte_eth_dev_get_preferred_pool(uint8_t port_id, const char *pool)
> +{
> +	struct rte_eth_dev *dev;
> +
> +	RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
> +
> +	dev = &rte_eth_devices[port_id];
> +
> +	if (*dev->dev_ops->get_preferred_pool == NULL) {
> +		pool = RTE_MBUF_DEFAULT_MEMPOOL_OPS;
> +		return 0;
> +	}
> +	return (*dev->dev_ops->get_preferred_pool)(dev, pool);
> +}

Instead of this, what about:

/*
 * Return values:
 *   - -ENOTSUP: error, pool type is not supported
 *   - on success, return the priority of the mempool (0 = highest)
 */
int
rte_eth_dev_pool_ops_supported(uint8_t port_id, const char *pool)

By default, always return 0 (i.e. all pools are supported).

With this API, we can announce several supported pools (not only
one preferred), and order them by preference.

I also wonder if we should use a ops_index instead of a pool name
for the second argument.



> diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h
> index 0f38b45f8..8e5b06af7 100644
> --- a/lib/librte_ether/rte_ethdev.h
> +++ b/lib/librte_ether/rte_ethdev.h
> @@ -1381,6 +1381,10 @@ typedef int (*eth_l2_tunnel_offload_set_t)
>  	 uint8_t en);
>  /**< @internal enable/disable the l2 tunnel offload functions */
>  
> +typedef int (*eth_get_preferred_pool_t)(struct rte_eth_dev *dev,
> +						const char *pool);
> +/**< @internal Get preferred pool handler for a device */
> +
>  #ifdef RTE_NIC_BYPASS
>  
>  enum {
> @@ -1573,6 +1577,8 @@ struct eth_dev_ops {
>  	/**< Get extended device statistic values by ID. */
>  	eth_xstats_get_names_by_id_t xstats_get_names_by_id;
>  	/**< Get name of extended device statistics by ID. */
> +	eth_get_preferred_pool_t get_preferred_pool;
> +	/**< Get preferred pool handler for a device */
>  };
>  
>  /**
> @@ -4607,6 +4613,21 @@ rte_eth_dev_get_port_by_name(const char *name, uint8_t *port_id);
>  int
>  rte_eth_dev_get_name_by_port(uint8_t port_id, char *name);
>  
> +/**
> + * Get preferred pool handle for a device
> + *
> + * @param port_id
> + *   port identifier of the device
> + * @param [out] pool
> + *   Preferred pool handle for this device.
> + *   Pool len shouldn't more than 256B. Allocated by pmd driver.

[out] ??
I don't get why it is allocated by the driver



> + * @return
> + *   - (0) if successful.
> + *   - (-EINVAL) on failure.
> + */
> +int
> +rte_eth_dev_get_preferred_pool(uint8_t port_id, const char *pool);
> +
>  #ifdef __cplusplus
>  }
>  #endif
> diff --git a/lib/librte_ether/rte_ether_version.map b/lib/librte_ether/rte_ether_version.map
> index d6726bb1b..819fe800e 100644
> --- a/lib/librte_ether/rte_ether_version.map
> +++ b/lib/librte_ether/rte_ether_version.map
> @@ -156,3 +156,10 @@ DPDK_17.05 {
>  	rte_eth_xstats_get_names_by_id;
>  
>  } DPDK_17.02;
> +
> +DPDK_17.08 {
> +	global:
> +
> +	rte_eth_dev_get_preferred_pool;
> +
> +} DPDK_17.05;