From mboxrd@z Thu Jan  1 00:00:00 1970
From: Shahaf Shuler <shahafs@mellanox.com>
Subject: Re: [PATCH v3 2/2] net/mlx5: fix instruction hotspot
	on	replenishing Rx buffer
Date: Thu, 10 Jan 2019 19:10:04 +0000
Message-ID: <DB7PR05MB4426EBCE0B4B424A504CA78CC3840@DB7PR05MB4426.eurprd05.prod.outlook.com>
References: <20190109085426.39965-1-yskoh@mellanox.com>
 <20190110183528.42503-1-yskoh@mellanox.com>
 <20190110183528.42503-2-yskoh@mellanox.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: quoted-printable
Cc: "dev@dpdk.org" <dev@dpdk.org>, "roszenrami@gmail.com"
 <roszenrami@gmail.com>, "david.marchand@redhat.com"
 <david.marchand@redhat.com>, "stable@dpdk.org" <stable@dpdk.org>
To: Yongseok Koh <yskoh@mellanox.com>, "olivier.matz@6wind.com"
 <olivier.matz@6wind.com>
Return-path: <dev-bounces@dpdk.org>
In-Reply-To: <20190110183528.42503-2-yskoh@mellanox.com>
Content-Language: en-US
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://mails.dpdk.org/options/dev>,
 <mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://mails.dpdk.org/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://mails.dpdk.org/listinfo/dev>,
 <mailto:dev-request@dpdk.org?subject=subscribe>
Errors-To: dev-bounces@dpdk.org
Sender: "dev" <dev-bounces@dpdk.org>

Thursday, January 10, 2019 8:35 PM, Yongseok Koh:
> Subject: [dpdk-dev] [PATCH v3 2/2] net/mlx5: fix instruction hotspot on
> replenishing Rx buffer
>=20
> On replenishing Rx buffers for vectorized Rx, mbuf->buf_addr isn't needed=
 to
> be accessed as it is static and easily calculated from the mbuf address.
> Accessing the mbuf content causes unnecessary load stall and it is worsen=
ed on
> ARM.
>=20
> Fixes: 545b884b1da3 ("net/mlx5: fix buffer address posting in SSE Rx")
> Cc: stable@dpdk.org
>=20
> Signed-off-by: Yongseok Koh <yskoh@mellanox.com>

Acked-by: Shahaf Shuler <shahafs@mellanox.com>

> ---
>=20
> v3:
> * rte_mbuf_buf_addr_default() -> rte_mbuf_buf_addr()
>=20
> v2:
> * use the newly introduced API - rte_mbuf_buf_addr_default()
> * fix error in assert
>=20
>  drivers/net/mlx5/mlx5_rxtx_vec.h | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
>=20
> diff --git a/drivers/net/mlx5/mlx5_rxtx_vec.h
> b/drivers/net/mlx5/mlx5_rxtx_vec.h
> index fda7004e2d..989a1fdce5 100644
> --- a/drivers/net/mlx5/mlx5_rxtx_vec.h
> +++ b/drivers/net/mlx5/mlx5_rxtx_vec.h
> @@ -102,8 +102,10 @@ mlx5_rx_replenish_bulk_mbuf(struct mlx5_rxq_data
> *rxq, uint16_t n)
>  		return;
>  	}
>  	for (i =3D 0; i < n; ++i) {
> -		wq[i].addr =3D rte_cpu_to_be_64((uintptr_t)elts[i]->buf_addr +
> -					      RTE_PKTMBUF_HEADROOM);
> +		void *buf_addr =3D rte_mbuf_buf_addr(elts[i], rxq->mp);
> +
> +		assert(buf_addr =3D=3D elts[i]->buf_addr);
> +		wq[i].addr =3D rte_cpu_to_be_64((uintptr_t)buf_addr +
> +RTE_PKTMBUF_HEADROOM);
>  		/* If there's only one MR, no need to replace LKey in WQE. */
>  		if (unlikely(mlx5_mr_btree_len(&rxq->mr_ctrl.cache_bh) > 1))
>  			wq[i].lkey =3D mlx5_rx_mb2mr(rxq, elts[i]);
> --
> 2.11.0