[dpdk-dev] [PATCH v2 1/3] net/ice: enable switch filter

Xing, Beilei beilei.xing at intel.com
Tue Jun 18 03:50:29 CEST 2019


Hi Wei,

> -----Original Message-----
> From: Zhao1, Wei
> Sent: Monday, June 17, 2019 4:52 PM
> To: Xing, Beilei <beilei.xing at intel.com>; Yang, Qiming
> <qiming.yang at intel.com>; dev at dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v2 1/3] net/ice: enable switch filter
> 
> Hi , Beilei
> 
> 
> > -----Original Message-----
> > From: Xing, Beilei
> > Sent: Monday, June 17, 2019 1:27 PM
> > To: Yang, Qiming <qiming.yang at intel.com>; dev at dpdk.org
> > Cc: Zhao1, Wei <wei.zhao1 at intel.com>
> > Subject: RE: [dpdk-dev] [PATCH v2 1/3] net/ice: enable switch filter
> >
> >
> >
> > ...
> >
> > > +
> > > +/* By now ice switch filter action code implement only
> > > +* supports QUEUE or DROP.
> > > +*/
> > > +static int
> > > +ice_parse_switch_action(struct ice_pf *pf,
> > > +				 const struct rte_flow_action *actions,
> > > +				 struct rte_flow_error *error,
> > > +				 struct ice_adv_rule_info *rule_info) {
> > > +	struct ice_hw *hw = ICE_PF_TO_HW(pf);
> > > +	struct ice_vsi *vsi = pf->main_vsi;
> > > +	const struct rte_flow_action *act;
> > > +	const struct rte_flow_action_queue *act_q;
> > > +	uint16_t base_queue, index = 0;
> > > +	uint32_t reg;
> > > +
> > > +	/* Check if the first non-void action is QUEUE or DROP. */
> > > +	NEXT_ITEM_OF_ACTION(act, actions, index);
> > > +	if (act->type != RTE_FLOW_ACTION_TYPE_QUEUE &&
> > > +	    act->type != RTE_FLOW_ACTION_TYPE_DROP) {
> > > +		rte_flow_error_set(error, EINVAL,
> > > RTE_FLOW_ERROR_TYPE_ACTION,
> > > +				   act, "Not supported action.");
> > > +		return -rte_errno;
> > > +	}
> > > +	reg = ICE_READ_REG(hw, PFLAN_RX_QALLOC);
> > > +	if (reg & PFLAN_RX_QALLOC_VALID_M) {
> > > +		base_queue = reg & PFLAN_RX_QALLOC_FIRSTQ_M;
> > > +	} else {
> > > +		rte_flow_error_set(error, EINVAL,
> > > +			RTE_FLOW_ERROR_TYPE_ACTION,
> > > +			act, "Invalid queue register");
> > > +		return -rte_errno;
> > > +	}
> > > +	if (act->type == RTE_FLOW_ACTION_TYPE_QUEUE) {
> > > +		act_q = act->conf;
> > > +		rule_info->sw_act.fltr_act = ICE_FWD_TO_Q;
> > > +		rule_info->sw_act.fwd_id.q_id = base_queue + act_q->index;
> > > +		if (act_q->index >= pf->dev_data->nb_rx_queues) {
> > > +			rte_flow_error_set(error, EINVAL,
> > > +				RTE_FLOW_ERROR_TYPE_ACTION,
> > > +				act, "Invalid queue ID for"
> > > +				" switch filter.");
> > > +			return -rte_errno;
> > > +		}
> > > +	} else {
> > > +		rule_info->sw_act.fltr_act = ICE_DROP_PACKET;
> > > +	}
> > > +
> > > +	rule_info->sw_act.vsi_handle = vsi->idx;
> > > +	rule_info->rx = 1;
> > > +	rule_info->sw_act.src = vsi->idx;
> > > +
> > > +	/* Check if the next non-void item is END */
> > > +	index++;
> > > +	NEXT_ITEM_OF_ACTION(act, actions, index);
> > > +	if (act->type != RTE_FLOW_ACTION_TYPE_END) {
> > > +		rte_flow_error_set(error, EINVAL,
> > > RTE_FLOW_ERROR_TYPE_ACTION,
> > > +				   act, "Not supported action.");
> > > +		return -rte_errno;
> > > +	}
> > > +
> > > +	return 0;
> > > +}
> >
> >
> > How about use supported array to replace NEXT_ITEM_OF_ACTION? Just
> > like pattern.
> 
> This seems no need to change, i40e also implement in this way.

Code in I40e is not perfect, we can try to improve our PMD in new driver. 
I think supported array is more clear and friendly, what do you think?



More information about the dev mailing list