DPDK patches and discussions
 help / color / mirror / Atom feed
From: Raslan Darawsheh <rasland@mellanox.com>
To: Alexander Kozyrev <akozyrev@mellanox.com>, "dev@dpdk.org" <dev@dpdk.org>
Cc: Slava Ovsiienko <viacheslavo@mellanox.com>,
	"stable@dpdk.org" <stable@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH] net/mlx5: set dynamic flow metadata in Rx queues
Date: Tue, 21 Apr 2020 17:55:15 +0000	[thread overview]
Message-ID: <AM0PR05MB67076B45006640451F52EDBBC2D50@AM0PR05MB6707.eurprd05.prod.outlook.com> (raw)
In-Reply-To: <1587143693-22724-1-git-send-email-akozyrev@mellanox.com>

Hi,

> -----Original Message-----
> From: Alexander Kozyrev <akozyrev@mellanox.com>
> Sent: Friday, April 17, 2020 8:15 PM
> To: dev@dpdk.org
> Cc: Raslan Darawsheh <rasland@mellanox.com>; Slava Ovsiienko
> <viacheslavo@mellanox.com>; stable@dpdk.org
> Subject: [PATCH] net/mlx5: set dynamic flow metadata in Rx queues
> 
> Using a global mbuf dynamic field for metadata incurs some
> performance penalty on a datapath. Store this information in
> the Rx queue descriptor for a better cache locality.
> 
> Fixes: a18ac6113331 ("net/mlx5: add metadata support to Rx datapath")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Alexander Kozyrev <akozyrev@mellanox.com>
> Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
> ---
>  drivers/net/mlx5/mlx5.h                  |  1 +
>  drivers/net/mlx5/mlx5_flow.c             | 29 +++++++++++++++++++
>  drivers/net/mlx5/mlx5_rxtx.c             |  7 +++--
>  drivers/net/mlx5/mlx5_rxtx.h             |  4 ++-
>  drivers/net/mlx5/mlx5_rxtx_vec_altivec.h | 27 +++++++++++-------
>  drivers/net/mlx5/mlx5_rxtx_vec_neon.h    | 49 ++++++++++++++++++------
> --------
>  drivers/net/mlx5/mlx5_rxtx_vec_sse.h     | 49 ++++++++++++++++++--------
> ------
>  drivers/net/mlx5/mlx5_trigger.c          |  2 ++
>  lib/librte_ethdev/rte_flow.c             |  2 +-
>  lib/librte_ethdev/rte_flow.h             |  2 +-
>  10 files changed, 113 insertions(+), 59 deletions(-)
> 
> diff --git a/drivers/net/mlx5/mlx5.h b/drivers/net/mlx5/mlx5.h
> index 318618e..16c60c8 100644
> --- a/drivers/net/mlx5/mlx5.h
> +++ b/drivers/net/mlx5/mlx5.h
> @@ -753,6 +753,7 @@ int mlx5_counter_query(struct rte_eth_dev *dev,
> uint32_t cnt,
>  		       bool clear, uint64_t *pkts, uint64_t *bytes);
>  int mlx5_flow_dev_dump(struct rte_eth_dev *dev, FILE *file,
>  		       struct rte_flow_error *error);
> +void mlx5_flow_rxq_dynf_metadata_set(struct rte_eth_dev *dev);
> 
>  /* mlx5_mp.c */
>  int mlx5_mp_primary_handle(const struct rte_mp_msg *mp_msg, const
> void *peer);
> diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
> index c529aa3..70e42a6 100644
> --- a/drivers/net/mlx5/mlx5_flow.c
> +++ b/drivers/net/mlx5/mlx5_flow.c
> @@ -894,6 +894,35 @@ uint32_t mlx5_flow_adjust_priority(struct
> rte_eth_dev *dev, int32_t priority,
>  	}
>  }
> 
> +/**
> + * Set the Rx queue dynamic metadata (mask and offset) for a flow
> + *
> + * @param[in] dev
> + *   Pointer to the Ethernet device structure.
> + */
> +void
> +mlx5_flow_rxq_dynf_metadata_set(struct rte_eth_dev *dev)
> +{
> +	struct mlx5_priv *priv = dev->data->dev_private;
> +	struct mlx5_rxq_data *data;
> +	unsigned int i;
> +
> +	for (i = 0; i != priv->rxqs_n; ++i) {
> +		if (!(*priv->rxqs)[i])
> +			continue;
> +		data = (*priv->rxqs)[i];
> +		if (!rte_flow_dynf_metadata_avail()) {
> +			data->dynf_meta = 0;
> +			data->flow_meta_mask = 0;
> +			data->flow_meta_offset = -1;
> +		} else {
> +			data->dynf_meta = 1;
> +			data->flow_meta_mask =
> rte_flow_dynf_metadata_mask;
> +			data->flow_meta_offset =
> rte_flow_dynf_metadata_offs;
> +		}
> +	}
> +}
> +
>  /*
>   * return a pointer to the desired action in the list of actions.
>   *
> diff --git a/drivers/net/mlx5/mlx5_rxtx.c b/drivers/net/mlx5/mlx5_rxtx.c
> index 3e583d4..0010b42 100644
> --- a/drivers/net/mlx5/mlx5_rxtx.c
> +++ b/drivers/net/mlx5/mlx5_rxtx.c
> @@ -1339,9 +1339,10 @@ enum mlx5_txcmp_code {
>  			pkt->hash.fdir.hi = mlx5_flow_mark_get(mark);
>  		}
>  	}
> -	if (rte_flow_dynf_metadata_avail() && cqe->flow_table_metadata)
> {
> -		pkt->ol_flags |= PKT_RX_DYNF_METADATA;
> -		*RTE_FLOW_DYNF_METADATA(pkt) = cqe-
> >flow_table_metadata;
> +	if (rxq->dynf_meta && cqe->flow_table_metadata) {
> +		pkt->ol_flags |= rxq->flow_meta_mask;
> +		*RTE_MBUF_DYNFIELD(pkt, rxq->flow_meta_offset,
> uint32_t *) =
> +			cqe->flow_table_metadata;
>  	}
>  	if (rxq->csum)
>  		pkt->ol_flags |= rxq_cq_to_ol_flags(cqe);
> diff --git a/drivers/net/mlx5/mlx5_rxtx.h b/drivers/net/mlx5/mlx5_rxtx.h
> index 8695218..48f2b79 100644
> --- a/drivers/net/mlx5/mlx5_rxtx.h
> +++ b/drivers/net/mlx5/mlx5_rxtx.h
> @@ -121,7 +121,7 @@ struct mlx5_rxq_data {
>  	unsigned int err_state:2; /* enum mlx5_rxq_err_state. */
>  	unsigned int strd_scatter_en:1; /* Scattered packets from a stride. */
>  	unsigned int lro:1; /* Enable LRO. */
> -	unsigned int :1; /* Remaining bits. */
> +	unsigned int dynf_meta:1; /* Dynamic metadata is configured. */
>  	volatile uint32_t *rq_db;
>  	volatile uint32_t *cq_db;
>  	uint16_t port_id;
> @@ -159,6 +159,8 @@ struct mlx5_rxq_data {
>  	/* CQ (UAR) access lock required for 32bit implementations */
>  #endif
>  	uint32_t tunnel; /* Tunnel information. */
> +	uint64_t flow_meta_mask;
> +	int32_t flow_meta_offset;
>  } __rte_cache_aligned;
> 
>  enum mlx5_rxq_obj_type {
> diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_altivec.h
> b/drivers/net/mlx5/mlx5_rxtx_vec_altivec.h
> index 0bb8b7a..03c0be2 100644
> --- a/drivers/net/mlx5/mlx5_rxtx_vec_altivec.h
> +++ b/drivers/net/mlx5/mlx5_rxtx_vec_altivec.h
> @@ -264,17 +264,22 @@
>  			elts[pos + 2]->hash.fdir.hi = flow_tag;
>  			elts[pos + 3]->hash.fdir.hi = flow_tag;
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> -			const uint32_t meta =
> *RTE_FLOW_DYNF_METADATA(t_pkt);
> +		if (!!rxq->flow_meta_mask) {
> +			int32_t offs = rxq->flow_meta_offset;
> +			const uint32_t meta =
> +				*RTE_MBUF_DYNFIELD(t_pkt, offs, uint32_t
> *);
> 
>  			/* Check if title packet has valid metadata. */
>  			if (meta) {
> -				MLX5_ASSERT(t_pkt->ol_flags &
> -					    PKT_RX_DYNF_METADATA);
> -				*RTE_FLOW_DYNF_METADATA(elts[pos]) =
> meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 1])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 2])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 3])
> = meta;
> +				MLX5_ASSERT(t_pkt->ol_flags & offs);
> +				*RTE_MBUF_DYNFIELD(elts[pos], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 1], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 2], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 3], offs,
> +							uint32_t *) = meta;
>  			}
>  		}
> 
> @@ -1023,9 +1028,9 @@
>  			pkts[pos + 3]->timestamp =
>  				rte_be_to_cpu_64(cq[pos + p3].timestamp);
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> -			uint64_t flag = rte_flow_dynf_metadata_mask;
> -			int offs = rte_flow_dynf_metadata_offs;
> +		if (!!rxq->flow_meta_mask) {
> +			uint64_t flag = rxq->flow_meta_mask;
> +			int32_t offs = rxq->flow_meta_offset;
>  			uint32_t metadata;
> 
>  			/* This code is subject for futher optimization. */
> diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
> b/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
> index 7b6c5db..d39e726 100644
> --- a/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
> +++ b/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
> @@ -205,17 +205,22 @@
>  			elts[pos + 2]->hash.fdir.hi = flow_tag;
>  			elts[pos + 3]->hash.fdir.hi = flow_tag;
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> -			const uint32_t meta =
> *RTE_FLOW_DYNF_METADATA(t_pkt);
> +		if (!!rxq->flow_meta_mask) {
> +			int32_t offs = rxq->flow_meta_offset;
> +			const uint32_t meta =
> +				*RTE_MBUF_DYNFIELD(t_pkt, offs, uint32_t
> *);
> 
>  			/* Check if title packet has valid metadata. */
>  			if (meta) {
> -				MLX5_ASSERT(t_pkt->ol_flags &
> -					    PKT_RX_DYNF_METADATA);
> -				*RTE_FLOW_DYNF_METADATA(elts[pos]) =
> meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 1])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 2])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 3])
> = meta;
> +				MLX5_ASSERT(t_pkt->ol_flags & offs);
> +				*RTE_MBUF_DYNFIELD(elts[pos], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 1], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 2], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 3], offs,
> +							uint32_t *) = meta;
>  			}
>  		}
>  		pos += MLX5_VPMD_DESCS_PER_LOOP;
> @@ -701,28 +706,30 @@
>  					container_of(p3, struct mlx5_cqe,
>  						     pkt_info)->timestamp);
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> +		if (!!rxq->flow_meta_mask) {
>  			/* This code is subject for futher optimization. */
> -			*RTE_FLOW_DYNF_METADATA(elts[pos]) =
> +			int32_t offs = rxq->flow_meta_offset;
> +
> +			*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t *) =
>  				container_of(p0, struct mlx5_cqe,
>  					     pkt_info)->flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(elts[pos + 1]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t *) =
>  				container_of(p1, struct mlx5_cqe,
>  					     pkt_info)->flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(elts[pos + 2]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t *) =
>  				container_of(p2, struct mlx5_cqe,
>  					     pkt_info)->flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(elts[pos + 3]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t *) =
>  				container_of(p3, struct mlx5_cqe,
>  					     pkt_info)->flow_table_metadata;
> -			if (*RTE_FLOW_DYNF_METADATA(elts[pos]))
> -				elts[pos]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(elts[pos + 1]))
> -				elts[pos + 1]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(elts[pos + 2]))
> -				elts[pos + 2]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(elts[pos + 3]))
> -				elts[pos + 3]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t
> *))
> +				elts[pos]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 1], offs,
> uint32_t *))
> +				elts[pos + 1]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 2], offs,
> uint32_t *))
> +				elts[pos + 2]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 3], offs,
> uint32_t *))
> +				elts[pos + 3]->ol_flags |= rxq-
> >flow_meta_mask;
>  		}
>  #ifdef MLX5_PMD_SOFT_COUNTERS
>  		/* Add up received bytes count. */
> diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
> b/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
> index 4b711f0..f110f73 100644
> --- a/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
> +++ b/drivers/net/mlx5/mlx5_rxtx_vec_sse.h
> @@ -192,17 +192,22 @@
>  			elts[pos + 2]->hash.fdir.hi = flow_tag;
>  			elts[pos + 3]->hash.fdir.hi = flow_tag;
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> -			const uint32_t meta =
> *RTE_FLOW_DYNF_METADATA(t_pkt);
> +		if (rxq->dynf_meta) {
> +			int32_t offs = rxq->flow_meta_offset;
> +			const uint32_t meta =
> +				*RTE_MBUF_DYNFIELD(t_pkt, offs, uint32_t
> *);
> 
>  			/* Check if title packet has valid metadata. */
>  			if (meta) {
> -				MLX5_ASSERT(t_pkt->ol_flags &
> -					    PKT_RX_DYNF_METADATA);
> -				*RTE_FLOW_DYNF_METADATA(elts[pos]) =
> meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 1])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 2])
> = meta;
> -				*RTE_FLOW_DYNF_METADATA(elts[pos + 3])
> = meta;
> +				MLX5_ASSERT(t_pkt->ol_flags & offs);
> +				*RTE_MBUF_DYNFIELD(elts[pos], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 1], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 2], offs,
> +							uint32_t *) = meta;
> +				*RTE_MBUF_DYNFIELD(elts[pos + 3], offs,
> +							uint32_t *) = meta;
>  			}
>  		}
>  		pos += MLX5_VPMD_DESCS_PER_LOOP;
> @@ -654,24 +659,26 @@
>  			pkts[pos + 3]->timestamp =
>  				rte_be_to_cpu_64(cq[pos + p3].timestamp);
>  		}
> -		if (rte_flow_dynf_metadata_avail()) {
> +		if (rxq->dynf_meta) {
>  			/* This code is subject for futher optimization. */
> -			*RTE_FLOW_DYNF_METADATA(pkts[pos]) =
> +			int32_t offs = rxq->flow_meta_offset;
> +
> +			*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t *) =
>  				cq[pos].flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(pkts[pos + 1]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos + 1], offs, uint32_t
> *) =
>  				cq[pos + p1].flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(pkts[pos + 2]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos + 2], offs, uint32_t
> *) =
>  				cq[pos + p2].flow_table_metadata;
> -			*RTE_FLOW_DYNF_METADATA(pkts[pos + 3]) =
> +			*RTE_MBUF_DYNFIELD(pkts[pos + 3], offs, uint32_t
> *) =
>  				cq[pos + p3].flow_table_metadata;
> -			if (*RTE_FLOW_DYNF_METADATA(pkts[pos]))
> -				pkts[pos]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(pkts[pos + 1]))
> -				pkts[pos + 1]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(pkts[pos + 2]))
> -				pkts[pos + 2]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> -			if (*RTE_FLOW_DYNF_METADATA(pkts[pos + 3]))
> -				pkts[pos + 3]->ol_flags |=
> PKT_RX_DYNF_METADATA;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos], offs, uint32_t
> *))
> +				pkts[pos]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 1], offs,
> uint32_t *))
> +				pkts[pos + 1]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 2], offs,
> uint32_t *))
> +				pkts[pos + 2]->ol_flags |= rxq-
> >flow_meta_mask;
> +			if (*RTE_MBUF_DYNFIELD(pkts[pos + 3], offs,
> uint32_t *))
> +				pkts[pos + 3]->ol_flags |= rxq-
> >flow_meta_mask;
>  		}
>  #ifdef MLX5_PMD_SOFT_COUNTERS
>  		/* Add up received bytes count. */
> diff --git a/drivers/net/mlx5/mlx5_trigger.c
> b/drivers/net/mlx5/mlx5_trigger.c
> index 7596704..feb9154 100644
> --- a/drivers/net/mlx5/mlx5_trigger.c
> +++ b/drivers/net/mlx5/mlx5_trigger.c
> @@ -323,6 +323,8 @@
>  			dev->data->port_id);
>  		goto error;
>  	}
> +	/* Set a mask and offset of dynamic metadata flows into Rx
> queues*/
> +	mlx5_flow_rxq_dynf_metadata_set(dev);
>  	/*
>  	 * In non-cached mode, it only needs to start the default mreg copy
>  	 * action and no flow created by application exists anymore.
> diff --git a/lib/librte_ethdev/rte_flow.c b/lib/librte_ethdev/rte_flow.c
> index 3699edc..1685be5 100644
> --- a/lib/librte_ethdev/rte_flow.c
> +++ b/lib/librte_ethdev/rte_flow.c
> @@ -19,7 +19,7 @@
>  #include "rte_flow.h"
> 
>  /* Mbuf dynamic field name for metadata. */
> -int rte_flow_dynf_metadata_offs = -1;
> +int32_t rte_flow_dynf_metadata_offs = -1;
> 
>  /* Mbuf dynamic field flag bit number for metadata. */
>  uint64_t rte_flow_dynf_metadata_mask;
> diff --git a/lib/librte_ethdev/rte_flow.h b/lib/librte_ethdev/rte_flow.h
> index fab44f6..132b44e 100644
> --- a/lib/librte_ethdev/rte_flow.h
> +++ b/lib/librte_ethdev/rte_flow.h
> @@ -2653,7 +2653,7 @@ struct rte_flow_action_set_dscp {
>  };
> 
>  /* Mbuf dynamic field offset for metadata. */
> -extern int rte_flow_dynf_metadata_offs;
> +extern int32_t rte_flow_dynf_metadata_offs;
> 
>  /* Mbuf dynamic field flag mask for metadata. */
>  extern uint64_t rte_flow_dynf_metadata_mask;
> --
> 1.8.3.1


Patch applied to next-net-mlx,

Kindest regards,
Raslan Darawsheh

      reply	other threads:[~2020-04-21 17:55 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-17 17:14 Alexander Kozyrev
2020-04-21 17:55 ` Raslan Darawsheh [this message]

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=AM0PR05MB67076B45006640451F52EDBBC2D50@AM0PR05MB6707.eurprd05.prod.outlook.com \
    --to=rasland@mellanox.com \
    --cc=akozyrev@mellanox.com \
    --cc=dev@dpdk.org \
    --cc=stable@dpdk.org \
    --cc=viacheslavo@mellanox.com \
    /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).