[dpdk-dev] [PATCH] net/ixgbe: fix ntuple filter support for sctp

Zhao1, Wei wei.zhao1 at intel.com
Thu Apr 27 07:39:22 CEST 2017


Hi,  Ferruh

> -----Original Message-----
> From: Yigit, Ferruh
> Sent: Thursday, April 27, 2017 12:53 PM
> To: Zhao1, Wei <wei.zhao1 at intel.com>; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] net/ixgbe: fix ntuple filter support for sctp
> 
> On 4/27/2017 4:14 AM, Wei Zhao wrote:
> > Add the support of RTE_FLOW_ITEM_TYPE_SCTP type packet for ixgbe
> > ntuple filter.
> 
> The function comment also should be updated for this pattern type.
> 
> And is this a fix or adding new type support? If this is not fixing something
> existing, lets postpone this to next release.
> 

No, it is a fix patch for our missing support for sctp packet type filter.
There are also other 2 had better been merge to this release also, they are high risk bug need to be fix.  
http://dpdk.org/dev/patchwork/patch/23761/
http://dpdk.org/dev/patchwork/patch/23760/

> >
> > Fixes: 672be56d76a ("net/ixgbe: parse n-tuple filter")
> >
> > Signed-off-by: Wei Zhao <wei.zhao1 at intel.com>
> > ---
> >  drivers/net/ixgbe/ixgbe_flow.c | 27 ++++++++++++++++++++++++++-
> >  1 file changed, 26 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/ixgbe/ixgbe_flow.c
> > b/drivers/net/ixgbe/ixgbe_flow.c index e2ba9c2..2c18bcd 100644
> > --- a/drivers/net/ixgbe/ixgbe_flow.c
> > +++ b/drivers/net/ixgbe/ixgbe_flow.c
> > @@ -142,6 +142,8 @@ cons_parse_ntuple_filter(const struct rte_flow_attr
> *attr,
> >  	const struct rte_flow_item_tcp *tcp_mask;
> >  	const struct rte_flow_item_udp *udp_spec;
> >  	const struct rte_flow_item_udp *udp_mask;
> > +	const struct rte_flow_item_sctp *sctp_spec;
> > +	const struct rte_flow_item_sctp *sctp_mask;
> >  	uint32_t index;
> >
> >  	if (!pattern) {
> > @@ -319,7 +321,7 @@ cons_parse_ntuple_filter(const struct rte_flow_attr
> *attr,
> >  		filter->dst_port  = tcp_spec->hdr.dst_port;
> >  		filter->src_port  = tcp_spec->hdr.src_port;
> >  		filter->tcp_flags = tcp_spec->hdr.tcp_flags;
> > -	} else {
> > +	} else if (item->type == RTE_FLOW_ITEM_TYPE_UDP) {
> >  		udp_mask = (const struct rte_flow_item_udp *)item->mask;
> >
> >  		/**
> > @@ -342,6 +344,29 @@ cons_parse_ntuple_filter(const struct
> rte_flow_attr *attr,
> >  		udp_spec = (const struct rte_flow_item_udp *)item->spec;
> >  		filter->dst_port = udp_spec->hdr.dst_port;
> >  		filter->src_port = udp_spec->hdr.src_port;
> > +	} else {
> 
> There is a check for this item if it is TCP or UDP, else it returns an error (in line
> 255), so it should not be hitting this else at all, am I missing something?

Yes,  you are right, I miss SOME in line 255, a v2 later to be commit.

> 
> > +		sctp_mask = (const struct rte_flow_item_sctp *)item->mask;
> > +
> > +		/**
> > +		 * Only support src & dst ports,
> > +		 * others should be masked.
> > +		 */
> > +		if (sctp_mask->hdr.tag ||
> > +		    sctp_mask->hdr.cksum) {
> > +			memset(filter, 0,
> > +				sizeof(struct rte_eth_ntuple_filter));
> > +			rte_flow_error_set(error, EINVAL,
> > +				RTE_FLOW_ERROR_TYPE_ITEM,
> > +				item, "Not supported by ntuple filter");
> > +			return -rte_errno;
> > +		}
> > +
> > +		filter->dst_port_mask = sctp_mask->hdr.dst_port;
> > +		filter->src_port_mask = sctp_mask->hdr.src_port;
> > +
> > +		sctp_spec = (const struct rte_flow_item_sctp *)item->spec;
> > +		filter->dst_port = sctp_spec->hdr.dst_port;
> > +		filter->src_port = sctp_spec->hdr.src_port;
> >  	}
> >
> >  	/* check if the next not void item is END */
> >



More information about the dev mailing list