Re: [dpdk-dev] [PATCH] net/i40e: Fail rte_flow MARK requests if RX func was vectorized

["Ergin, Mesut A" <mesut.a.ergin@intel.com>]

Hi Qi,

> -----Original Message-----
> From: Zhang, Qi Z
> Sent: Thursday, May 23, 2019 7:49 PM
> To: Ergin, Mesut A <mesut.a.ergin@intel.com>; Xing, Beilei
> <beilei.xing@intel.com>
> Cc: dev@dpdk.org
> Subject: RE: [PATCH] net/i40e: Fail rte_flow MARK requests if RX func was
> vectorized
> 
> Hi Mesut:
> 
> > -----Original Message-----
> > From: Ergin, Mesut A
> > Sent: Friday, May 24, 2019 1:57 AM
> > To: Xing, Beilei <beilei.xing@intel.com>; Zhang, Qi Z <qi.z.zhang@intel.com>
> > Cc: dev@dpdk.org; Ergin, Mesut A <mesut.a.ergin@intel.com>
> > Subject: [PATCH] net/i40e: Fail rte_flow MARK requests if RX func was
> > vectorized
> >
> > Runtime requests to install an rte_flow with MARK action should fail if a vector
> > RX function was already chosen for the device during configuration time.
> >
> > Currently, i40e rte_flow driver would successfully install the flow with MARK
> > action, even when vector RX functions are in use. However, those vector RX
> > functions will fail to retrieve the MARK data from the device descriptor into
> > the mbuf. The original app  installing the flow would never know what went
> > wrong. The change introduced in this patch must be reverted if/when vector
> > RX functions start supporting correct FDIR processing for MARK actions.
> >
> > Signed-off-by: Mesut Ali Ergin <mesut.a.ergin@intel.com>
> > ---
> >  drivers/net/i40e/i40e_flow.c | 21 +++++++++++++++++++++
> >  1 file changed, 21 insertions(+)
> >
> > diff --git a/drivers/net/i40e/i40e_flow.c b/drivers/net/i40e/i40e_flow.c index
> > 3afd779..9a14a1c 100644
> > --- a/drivers/net/i40e/i40e_flow.c
> > +++ b/drivers/net/i40e/i40e_flow.c
> > @@ -22,6 +22,7 @@
> >  #include "base/i40e_type.h"
> >  #include "base/i40e_prototype.h"
> >  #include "i40e_ethdev.h"
> > +#include "i40e_rxtx.h"
> >
> >  #define I40E_IPV6_TC_MASK	(0xFF << I40E_FDIR_IPv6_TC_OFFSET)
> >  #define I40E_IPV6_FRAG_HEADER	44
> > @@ -3079,6 +3080,16 @@ i40e_flow_parse_fdir_action(struct rte_eth_dev
> > *dev,
> >  		filter->action.behavior = I40E_FDIR_PASSTHRU;
> >  		break;
> >  	case RTE_FLOW_ACTION_TYPE_MARK:
> > +		if (dev->rx_pkt_burst == i40e_recv_scattered_pkts_vec ||
> > +		    dev->rx_pkt_burst == i40e_recv_pkts_vec ||
> > +		    dev->rx_pkt_burst == i40e_recv_scattered_pkts_vec_avx2 ||
> > +		    dev->rx_pkt_burst == i40e_recv_pkts_vec_avx2) {
> > +			/* Vector RX Functions do not support MARK*/
> > +			rte_flow_error_set(error, EINVAL,
> > +					   RTE_FLOW_ERROR_TYPE_ACTION,
> act,
> > +					   "Action not supported when vector
> RX is in use.");
> > +			return -rte_errno;
> > +		}
> 
> We may not need to prevent this if the device is already stopped (check dev-
> >data->dev_started),
> since with the fdir flow be created here, vector path will not be selected in
> following dev_start.

Makes sense, sending a v2 in a bit.

> 
> >  		filter->action.behavior = I40E_FDIR_PASSTHRU;
> >  		mark_spec = act->conf;
> >  		filter->action.report_status = I40E_FDIR_REPORT_ID; @@ -
> 3103,6
> > +3114,16 @@ i40e_flow_parse_fdir_action(struct rte_eth_dev *dev,
> >  			   "Invalid action.");
> >  			return -rte_errno;
> >  		}
> > +		if (dev->rx_pkt_burst == i40e_recv_scattered_pkts_vec ||
> > +		    dev->rx_pkt_burst == i40e_recv_pkts_vec ||
> > +		    dev->rx_pkt_burst == i40e_recv_scattered_pkts_vec_avx2 ||
> > +		    dev->rx_pkt_burst == i40e_recv_pkts_vec_avx2) {
> > +			/* Vector RX Functions do not support MARK*/
> > +			rte_flow_error_set(error, EINVAL,
> > +					   RTE_FLOW_ERROR_TYPE_ACTION,
> act,
> > +					   "Action not supported when vector
> RX is in use.");
> > +			return -rte_errno;
> > +		}
> >  		mark_spec = act->conf;
> >  		filter->action.report_status = I40E_FDIR_REPORT_ID;
> >  		filter->soft_id = mark_spec->id;
> > --
> > 2.7.4