[dpdk-dev] [PATCH v3 01/14] net/mlx5: support 16 hardware priorities

[Xueming(Steven) Li]

> -----Original Message-----
> From: Nélio Laranjeiro <nelio.laranjeiro at 6wind.com>
> Sent: Friday, April 13, 2018 7:59 PM
> To: Xueming(Steven) Li <xuemingl at mellanox.com>
> Cc: Shahaf Shuler <shahafs at mellanox.com>; dev at dpdk.org
> Subject: Re: [PATCH v3 01/14] net/mlx5: support 16 hardware priorities
> 
> Hi Xueming,
> 
> Small nips and documentation issues,
> 
> On Fri, Apr 13, 2018 at 07:20:10PM +0800, Xueming Li wrote:
> > This patch supports new 16 Verbs flow priorities by trying to create a
> > simple flow of priority 15. If 16 priorities not available, fallback
> > to traditional 8 priorities.
> >
> > Verb priority mapping:
> > 			8 priorities	>=16 priorities
> > Control flow:		4-7		8-15
> > User normal flow:	1-3		4-7
> > User tunnel flow:	0-2		0-3
> 
> There is an overlap between tunnel and normal flows it is expected?

For 8 priorities, (4-7), (1-3) and (0-2) are the behavior of today, 
1 Verbs shift to make tunnel flow higher priority, please refer to 
commit #74936571

> 
> >
> > Signed-off-by: Xueming Li <xuemingl at mellanox.com>
> > ---
> >  drivers/net/mlx5/mlx5.c         |  18 +++++++
> >  drivers/net/mlx5/mlx5.h         |   5 ++
> >  drivers/net/mlx5/mlx5_flow.c    | 112
> +++++++++++++++++++++++++++++++++-------
> >  drivers/net/mlx5/mlx5_trigger.c |   8 ---
> >  4 files changed, 115 insertions(+), 28 deletions(-)
> >
> > diff --git a/drivers/net/mlx5/mlx5.c b/drivers/net/mlx5/mlx5.c index
> > cfab55897..38118e524 100644
> > --- a/drivers/net/mlx5/mlx5.c
> > +++ b/drivers/net/mlx5/mlx5.c
> > @@ -197,6 +197,7 @@ mlx5_dev_close(struct rte_eth_dev *dev)
> >  		priv->txqs_n = 0;
> >  		priv->txqs = NULL;
> >  	}
> > +	mlx5_flow_delete_drop_queue(dev);
> >  	if (priv->pd != NULL) {
> >  		assert(priv->ctx != NULL);
> >  		claim_zero(mlx5_glue->dealloc_pd(priv->pd));
> > @@ -612,6 +613,7 @@ mlx5_pci_probe(struct rte_pci_driver *pci_drv
> __rte_unused,
> >  	unsigned int mps;
> >  	unsigned int cqe_comp;
> >  	unsigned int tunnel_en = 0;
> > +	unsigned int verb_priorities = 0;
> >  	int idx;
> >  	int i;
> >  	struct mlx5dv_context attrs_out = {0}; @@ -993,6 +995,22 @@
> > mlx5_pci_probe(struct rte_pci_driver *pci_drv __rte_unused,
> >  		mlx5_set_link_up(eth_dev);
> >  		/* Store device configuration on private structure. */
> >  		priv->config = config;
> > +		/* Create drop queue. */
> > +		err = mlx5_flow_create_drop_queue(eth_dev);
> > +		if (err) {
> > +			DRV_LOG(ERR, "port %u drop queue allocation failed: %s",
> > +				eth_dev->data->port_id, strerror(rte_errno));
> > +			goto port_error;
> > +		}
> > +		/* Supported Verbs flow priority number detection. */
> > +		if (verb_priorities == 0)
> > +			verb_priorities = priv_get_max_verbs_prio(eth_dev);
> 
> No more priv*() rename it to mlx5_get_max_verbs_prio()
> 
> > +		if (verb_priorities < MLX5_VERBS_FLOW_PRIO_8) {
> > +			DRV_LOG(ERR, "port %u wrong Verbs flow priorities: %u",
> > +				eth_dev->data->port_id, verb_priorities);
> > +			goto port_error;
> > +		}
> > +		priv->config.max_verb_prio = verb_priorities;
> 
> s/verb/verbs/
> 
> >  		continue;
> >  port_error:
> >  		if (priv)
> > diff --git a/drivers/net/mlx5/mlx5.h b/drivers/net/mlx5/mlx5.h index
> > 63b24e6bb..6e4613fe0 100644
> > --- a/drivers/net/mlx5/mlx5.h
> > +++ b/drivers/net/mlx5/mlx5.h
> > @@ -89,6 +89,7 @@ struct mlx5_dev_config {
> >  	unsigned int rx_vec_en:1; /* Rx vector is enabled. */
> >  	unsigned int mpw_hdr_dseg:1; /* Enable DSEGs in the title WQEBB. */
> >  	unsigned int vf_nl_en:1; /* Enable Netlink requests in VF mode. */
> > +	unsigned int max_verb_prio; /* Number of Verb flow priorities. */
> >  	unsigned int tso_max_payload_sz; /* Maximum TCP payload for TSO. */
> >  	unsigned int ind_table_max_size; /* Maximum indirection table size.
> */
> >  	int txq_inline; /* Maximum packet size for inlining. */ @@ -105,6
> > +106,9 @@ enum mlx5_verbs_alloc_type {
> >  	MLX5_VERBS_ALLOC_TYPE_RX_QUEUE,
> >  };
> >
> > +/* 8 Verbs priorities. */
> > +#define MLX5_VERBS_FLOW_PRIO_8 8
> > +
> >  /**
> >   * Verbs allocator needs a context to know in the callback which kind
> of
> >   * resources it is allocating.
> > @@ -253,6 +257,7 @@ int mlx5_traffic_restart(struct rte_eth_dev *dev);
> >
> >  /* mlx5_flow.c */
> >
> > +unsigned int priv_get_max_verbs_prio(struct rte_eth_dev *dev);
> >  int mlx5_flow_validate(struct rte_eth_dev *dev,
> >  		       const struct rte_flow_attr *attr,
> >  		       const struct rte_flow_item items[], diff --git
> > a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c index
> > 288610620..5c4f0b586 100644
> > --- a/drivers/net/mlx5/mlx5_flow.c
> > +++ b/drivers/net/mlx5/mlx5_flow.c
> > @@ -32,8 +32,8 @@
> >  #include "mlx5_prm.h"
> >  #include "mlx5_glue.h"
> >
> > -/* Define minimal priority for control plane flows. */ -#define
> > MLX5_CTRL_FLOW_PRIORITY 4
> > +/* Flow priority for control plane flows. */ #define
> > +MLX5_CTRL_FLOW_PRIORITY 1
> >
> >  /* Internet Protocol versions. */
> >  #define MLX5_IPV4 4
> > @@ -129,7 +129,7 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_SRC_PORT_TCP |
> >  				IBV_RX_HASH_DST_PORT_TCP),
> >  		.dpdk_rss_hf = ETH_RSS_NONFRAG_IPV4_TCP,
> > -		.flow_priority = 1,
> > +		.flow_priority = 0,
> >  		.ip_version = MLX5_IPV4,
> >  	},
> >  	[HASH_RXQ_UDPV4] = {
> > @@ -138,7 +138,7 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_SRC_PORT_UDP |
> >  				IBV_RX_HASH_DST_PORT_UDP),
> >  		.dpdk_rss_hf = ETH_RSS_NONFRAG_IPV4_UDP,
> > -		.flow_priority = 1,
> > +		.flow_priority = 0,
> >  		.ip_version = MLX5_IPV4,
> >  	},
> >  	[HASH_RXQ_IPV4] = {
> > @@ -146,7 +146,7 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_DST_IPV4),
> >  		.dpdk_rss_hf = (ETH_RSS_IPV4 |
> >  				ETH_RSS_FRAG_IPV4),
> > -		.flow_priority = 2,
> > +		.flow_priority = 1,
> >  		.ip_version = MLX5_IPV4,
> >  	},
> >  	[HASH_RXQ_TCPV6] = {
> > @@ -155,7 +155,7 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_SRC_PORT_TCP |
> >  				IBV_RX_HASH_DST_PORT_TCP),
> >  		.dpdk_rss_hf = ETH_RSS_NONFRAG_IPV6_TCP,
> > -		.flow_priority = 1,
> > +		.flow_priority = 0,
> >  		.ip_version = MLX5_IPV6,
> >  	},
> >  	[HASH_RXQ_UDPV6] = {
> > @@ -164,7 +164,7 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_SRC_PORT_UDP |
> >  				IBV_RX_HASH_DST_PORT_UDP),
> >  		.dpdk_rss_hf = ETH_RSS_NONFRAG_IPV6_UDP,
> > -		.flow_priority = 1,
> > +		.flow_priority = 0,
> >  		.ip_version = MLX5_IPV6,
> >  	},
> >  	[HASH_RXQ_IPV6] = {
> > @@ -172,13 +172,13 @@ const struct hash_rxq_init hash_rxq_init[] = {
> >  				IBV_RX_HASH_DST_IPV6),
> >  		.dpdk_rss_hf = (ETH_RSS_IPV6 |
> >  				ETH_RSS_FRAG_IPV6),
> > -		.flow_priority = 2,
> > +		.flow_priority = 1,
> >  		.ip_version = MLX5_IPV6,
> >  	},
> >  	[HASH_RXQ_ETH] = {
> >  		.hash_fields = 0,
> >  		.dpdk_rss_hf = 0,
> > -		.flow_priority = 3,
> > +		.flow_priority = 2,
> >  	},
> >  };
> >
> > @@ -900,30 +900,50 @@ mlx5_flow_convert_allocate(unsigned int size,
> struct rte_flow_error *error)
> >   * Make inner packet matching with an higher priority from the non
> Inner
> >   * matching.
> >   *
> > + * @param dev
> > + *   Pointer to Ethernet device.
> >   * @param[in, out] parser
> >   *   Internal parser structure.
> >   * @param attr
> >   *   User flow attribute.
> >   */
> >  static void
> > -mlx5_flow_update_priority(struct mlx5_flow_parse *parser,
> > +mlx5_flow_update_priority(struct rte_eth_dev *dev,
> > +			  struct mlx5_flow_parse *parser,
> >  			  const struct rte_flow_attr *attr)  {
> > +	struct priv *priv = dev->data->dev_private;
> >  	unsigned int i;
> > +	uint16_t priority;
> >
> > +	/*			8 priorities	>= 16 priorities
> > +	 * Control flow:	4-7		8-15
> > +	 * User normal flow:	1-3		4-7
> > +	 * User tunnel flow:	0-2		0-3
> > +	 */
> 
> Same comment here, the tunnel flow overlap when there are only 8
> priorities.
> 
> > +	priority = attr->priority * MLX5_VERBS_FLOW_PRIO_8;
> > +	if (priv->config.max_verb_prio == MLX5_VERBS_FLOW_PRIO_8)
> > +		priority /= 2;
> > +	/*
> > +	 * Lower non-tunnel flow Verbs priority 1 if only support 8 Verbs
> > +	 * priorities, lower 4 otherwise.
> > +	 */
> > +	if (!parser->inner) {
> > +		if (priv->config.max_verb_prio == MLX5_VERBS_FLOW_PRIO_8)
> > +			priority += 1;
> > +		else
> > +			priority += MLX5_VERBS_FLOW_PRIO_8 / 2;
> > +	}
> >  	if (parser->drop) {
> > -		parser->queue[HASH_RXQ_ETH].ibv_attr->priority =
> > -			attr->priority +
> > -			hash_rxq_init[HASH_RXQ_ETH].flow_priority;
> > +		parser->queue[HASH_RXQ_ETH].ibv_attr->priority = priority +
> > +				hash_rxq_init[HASH_RXQ_ETH].flow_priority;
> >  		return;
> >  	}
> >  	for (i = 0; i != hash_rxq_init_n; ++i) {
> > -		if (parser->queue[i].ibv_attr) {
> > -			parser->queue[i].ibv_attr->priority =
> > -				attr->priority +
> > -				hash_rxq_init[i].flow_priority -
> > -				(parser->inner ? 1 : 0);
> > -		}
> > +		if (!parser->queue[i].ibv_attr)
> > +			continue;
> > +		parser->queue[i].ibv_attr->priority = priority +
> > +				hash_rxq_init[i].flow_priority;
> >  	}
> >  }
> >
> > @@ -1158,7 +1178,7 @@ mlx5_flow_convert(struct rte_eth_dev *dev,
> >  	 */
> >  	if (!parser->drop)
> >  		mlx5_flow_convert_finalise(parser);
> > -	mlx5_flow_update_priority(parser, attr);
> > +	mlx5_flow_update_priority(dev, parser, attr);
> >  exit_free:
> >  	/* Only verification is expected, all resources should be released.
> */
> >  	if (!parser->create) {
> > @@ -3161,3 +3181,55 @@ mlx5_dev_filter_ctrl(struct rte_eth_dev *dev,
> >  	}
> >  	return 0;
> >  }
> > +
> > +/**
> > + * Detect number of Verbs flow priorities supported.
> > + *
> > + * @param dev
> > + *   Pointer to Ethernet device.
> > + *
> > + * @return
> > + *   number of supported Verbs flow priority.
> > + */
> > +unsigned int
> > +priv_get_max_verbs_prio(struct rte_eth_dev *dev) {
> > +	struct priv *priv = dev->data->dev_private;
> > +	unsigned int verb_priorities = MLX5_VERBS_FLOW_PRIO_8;
> > +	struct {
> > +		struct ibv_flow_attr attr;
> > +		struct ibv_flow_spec_eth eth;
> > +		struct ibv_flow_spec_action_drop drop;
> > +	} flow_attr = {
> > +		.attr = {
> > +			.num_of_specs = 2,
> > +		},
> > +		.eth = {
> > +			.type = IBV_FLOW_SPEC_ETH,
> > +			.size = sizeof(struct ibv_flow_spec_eth),
> > +		},
> > +		.drop = {
> > +			.size = sizeof(struct ibv_flow_spec_action_drop),
> > +			.type = IBV_FLOW_SPEC_ACTION_DROP,
> > +		},
> > +	};
> > +	struct ibv_flow *flow;
> > +
> > +	do {
> > +		flow_attr.attr.priority = verb_priorities - 1;
> > +		flow = mlx5_glue->create_flow(priv->flow_drop_queue->qp,
> > +					      &flow_attr.attr);
> > +		if (flow) {
> > +			claim_zero(mlx5_glue->destroy_flow(flow));
> > +			/* Try more priorities. */
> > +			verb_priorities *= 2;
> > +		} else {
> > +			/* Failed, restore last right number. */
> > +			verb_priorities /= 2;
> > +			break;
> > +		}
> > +	} while (1);
> > +	DRV_LOG(INFO, "port %u Verbs flow priorities: %d",
> > +		dev->data->port_id, verb_priorities);
> 
> Please remove this developer log, it will confuse the user who will
> believe he have N priorities which is absolutely not the case.

How about change it to DEBUG level, this should be very useful in real deployment
trouble shooting.

I could append something like "user flow priorities: %d" to avoid confusion..

> 
> > +	return verb_priorities;
> > +}
> > diff --git a/drivers/net/mlx5/mlx5_trigger.c
> > b/drivers/net/mlx5/mlx5_trigger.c index 6bb4ffb14..d80a2e688 100644
> > --- a/drivers/net/mlx5/mlx5_trigger.c
> > +++ b/drivers/net/mlx5/mlx5_trigger.c
> > @@ -148,12 +148,6 @@ mlx5_dev_start(struct rte_eth_dev *dev)
> >  	int ret;
> >
> >  	dev->data->dev_started = 1;
> > -	ret = mlx5_flow_create_drop_queue(dev);
> > -	if (ret) {
> > -		DRV_LOG(ERR, "port %u drop queue allocation failed: %s",
> > -			dev->data->port_id, strerror(rte_errno));
> > -		goto error;
> > -	}
> >  	DRV_LOG(DEBUG, "port %u allocating and configuring hash Rx queues",
> >  		dev->data->port_id);
> >  	rte_mempool_walk(mlx5_mp2mr_iter, priv); @@ -202,7 +196,6 @@
> > mlx5_dev_start(struct rte_eth_dev *dev)
> >  	mlx5_traffic_disable(dev);
> >  	mlx5_txq_stop(dev);
> >  	mlx5_rxq_stop(dev);
> > -	mlx5_flow_delete_drop_queue(dev);
> >  	rte_errno = ret; /* Restore rte_errno. */
> >  	return -rte_errno;
> >  }
> > @@ -237,7 +230,6 @@ mlx5_dev_stop(struct rte_eth_dev *dev)
> >  	mlx5_rxq_stop(dev);
> >  	for (mr = LIST_FIRST(&priv->mr); mr; mr = LIST_FIRST(&priv->mr))
> >  		mlx5_mr_release(mr);
> > -	mlx5_flow_delete_drop_queue(dev);
> >  }
> >
> >  /**
> > --
> > 2.13.3
> 
> Thanks,
> 
> --
> Nélio Laranjeiro
> 6WIND