From: "Guo, Jia" <jia.guo@intel.com>
To: "Wang, Haiyue" <haiyue.wang@intel.com>,
"Wu, Jingjing" <jingjing.wu@intel.com>,
"Zhang, Qi Z" <qi.z.zhang@intel.com>,
"Xing, Beilei" <beilei.xing@intel.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v1] net/iavf: support flex desc metadata extraction
Date: Fri, 18 Sep 2020 02:41:27 +0000 [thread overview]
Message-ID: <ae77f15b55cb405f892d6d0e244ead51@intel.com> (raw)
In-Reply-To: <BN8PR11MB379554D9FDC9350AF082E93FF73E0@BN8PR11MB3795.namprd11.prod.outlook.com>
hi, haiyue
> -----Original Message-----
> From: Wang, Haiyue <haiyue.wang@intel.com>
> Sent: Thursday, September 17, 2020 11:00 AM
> To: Guo, Jia <jia.guo@intel.com>; Wu, Jingjing <jingjing.wu@intel.com>;
> Zhang, Qi Z <qi.z.zhang@intel.com>; Xing, Beilei <beilei.xing@intel.com>
> Cc: dev@dpdk.org; Guo, Jia <jia.guo@intel.com>
> Subject: RE: [dpdk-dev] [PATCH v1] net/iavf: support flex desc metadata
> extraction
>
> Hi Jeff,
>
> > -----Original Message-----
> > From: dev <dev-bounces@dpdk.org> On Behalf Of Jeff Guo
> > Sent: Wednesday, September 9, 2020 10:54
> > To: Wu, Jingjing <jingjing.wu@intel.com>; Zhang, Qi Z
> > <qi.z.zhang@intel.com>; Xing, Beilei <beilei.xing@intel.com>
> > Cc: dev@dpdk.org; Guo, Jia <jia.guo@intel.com>
> > Subject: [dpdk-dev] [PATCH v1] net/iavf: support flex desc metadata
> > extraction
> >
> > Enable metadata extraction for flexible descriptors in AVF, that would
> > allow network function directly get metadata without additional
> > parsing which would reduce the CPU cost for VFs. The enabling metadata
> > extractions involve the metadata of VLAN/IPv4/IPv6/IPv6-FLOW/TCP/OVS/
> > MPLS flexible descriptors, and the VF could negotiate the capability
> > of the flexible descriptor with PF and correspondingly configure the
> > specific offload at receiving queues.
> >
> > Signed-off-by: Jeff Guo <jia.guo@intel.com>
> > ---
> > doc/guides/rel_notes/release_20_11.rst | 6 +
> > drivers/net/iavf/Makefile | 1 +
> > drivers/net/iavf/iavf.h | 25 +-
> > drivers/net/iavf/iavf_ethdev.c | 398 +++++++++++++++++++++++++
> > drivers/net/iavf/iavf_rxtx.c | 230 +++++++++++++-
> > drivers/net/iavf/iavf_rxtx.h | 17 ++
> > drivers/net/iavf/iavf_vchnl.c | 22 +-
> > drivers/net/iavf/meson.build | 2 +
> > drivers/net/iavf/rte_pmd_iavf.h | 258 ++++++++++++++++
> > 9 files changed, 937 insertions(+), 22 deletions(-) create mode
> > 100644 drivers/net/iavf/rte_pmd_iavf.h
> >
>
>
> > -------------
> > diff --git a/drivers/net/iavf/Makefile b/drivers/net/iavf/Makefile
> > index 792cbb7f7..05fcbdc47 100644
> > --- a/drivers/net/iavf/Makefile
> > +++ b/drivers/net/iavf/Makefile
>
> meson build only now, remove the Makefile
>
Oh, that is exactly.
>
> > diff --git a/drivers/net/iavf/iavf_rxtx.c
> > b/drivers/net/iavf/iavf_rxtx.c index 05a7dd898..fa71b4a80 100644
> > --- a/drivers/net/iavf/iavf_rxtx.c
> > +++ b/drivers/net/iavf/iavf_rxtx.c
> > @@ -26,6 +26,74 @@
>
>
> > +
> > +/* Translate the rx flex descriptor status to pkt flags */ static
> > +inline void iavf_rxd_to_pkt_fields(struct rte_mbuf *mb,
> > + volatile union iavf_rx_flex_desc *rxdp, uint8_t rxdid) { if
> > +(rxdid == IAVF_RXDID_COMMS_GENERIC ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_VLAN ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_IPV4 ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_IPV6 ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_IPV6_FLOW ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_TCP ||
> > + rxdid == IAVF_RXDID_COMMS_AUX_IP_OFFSET)
> > +iavf_rxd_to_pkt_fields_aux(mb, rxdp); else if (rxdid ==
> > +IAVF_RXDID_COMMS_OVS_1) iavf_rxd_to_pkt_fields_ovs(mb, rxdp); }
>
> We can optimize this by calling function handle:
>
> struct iavf_rx_queue *rxq->rxd_to_pkt_fields(mb, rxdp)
>
> and when setup the queue, assign the right handle according to the rxdid.
> if (rxdid == IAVF_RXDID_COMMS_GENERIC ...)
> rxq->rxd_to_pkt_fields = iavf_rxd_to_pkt_fields_aux;
> else if (OVS_1)
> rxq->rxd_to_pkt_fields = iavf_rxd_to_pkt_fields_ovs;
>
>
Sounds not bad if it could let it more clear for handling these diversity protocols, let's see what we could bring some modification in coming version.
> > --- a/drivers/net/iavf/meson.build
> > +++ b/drivers/net/iavf/meson.build
> > @@ -35,3 +35,5 @@ if arch_subdir == 'x86'
> > objs += iavf_avx2_lib.extract_objects('iavf_rxtx_vec_avx2.c')
> > endif
> > endif
> > +
> > +install_headers('rte_pmd_iavf.h')
> > diff --git a/drivers/net/iavf/rte_pmd_iavf.h
> > b/drivers/net/iavf/rte_pmd_iavf.h new file mode 100644 index
> > 000000000..858201bd7
> > --- /dev/null
> > +++ b/drivers/net/iavf/rte_pmd_iavf.h
> > @@ -0,0 +1,258 @@
> > +/* SPDX-Liavfnse-Identifier: BSD-3-Clause
> > + * Copyright(c) 2019 Intel Corporation */
> > +
> > +#ifndef _RTE_PMD_IAVF_H_
> > +#define _RTE_PMD_IAVF_H_
> > +
> > +/**
> > + * @file rte_pmd_iavf.h
> > + *
> > + * iavf PMD specific functions.
> > + *
> > + * @b EXPERIMENTAL: this API may change, or be removed, without prior
> > +notiavf
> > + *
> > + */
> > +
> > +#include <stdio.h>
> > +#include <rte_mbuf.h>
> > +#include <rte_mbuf_dyn.h>
> > +
> > +#ifdef __cplusplus
> > +extern "C" {
> > +#endif
> > +
> > +/**
> > + * The supported network flexible descriptor's extraction metadata
> format.
> > + */
> > +union rte_net_iavf_flex_desc_metadata { uint32_t metadata;
> > +
> > +struct {
> > +uint16_t data0;
> > +uint16_t data1;
> > +} raw;
> > +
> > +struct {
> > +uint16_t stag_vid:12,
> > + stag_dei:1,
> > + stag_pcp:3;
> > +uint16_t ctag_vid:12,
> > + ctag_dei:1,
> > + ctag_pcp:3;
> > +} vlan;
> > +
> > +struct {
> > +uint16_t protocol:8,
> > + ttl:8;
> > +uint16_t tos:8,
> > + ihl:4,
> > + version:4;
> > +} ipv4;
> > +
> > +struct {
> > +uint16_t hoplimit:8,
> > + nexthdr:8;
> > +uint16_t flowhi4:4,
> > + tc:8,
> > + version:4;
> > +} ipv6;
> > +
> > +struct {
> > +uint16_t flowlo16;
> > +uint16_t flowhi4:4,
> > + tc:8,
> > + version:4;
> > +} ipv6_flow;
> > +
> > +struct {
> > +uint16_t fin:1,
> > + syn:1,
> > + rst:1,
> > + psh:1,
> > + ack:1,
> > + urg:1,
> > + ece:1,
> > + cwr:1,
> > + res1:4,
> > + doff:4;
> > +uint16_t rsvd;
> > +} tcp;
> > +
> > +uint32_t ip_ofs;
> > +};
> > +
> > +/* Offset of mbuf dynamic field for flexible descriptor's extraction
> > +data */ extern int rte_net_iavf_dynfield_flex_desc_metadata_offs;
> > +
> > +/* Mask of mbuf dynamic flags for flexible descriptor's extraction
> > +type */ extern uint64_t rte_net_iavf_dynflag_flex_desc_vlan_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_ipv4_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_ipv6_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_ipv6_flow_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_tcp_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_ovs_mask;
> > +extern uint64_t rte_net_iavf_dynflag_flex_desc_ip_offset_mask;
> > +
> > +/**
> > + * The mbuf dynamic field pointer for flexible descriptor's extraction
> metadata.
> > + */
> > +#define RTE_NET_IAVF_DYNF_FLEX_DESC_METADATA(m) \
> > +RTE_MBUF_DYNFIELD((m), \
> > + rte_net_iavf_dynfield_flex_desc_metadata_offs, \
> > + uint32_t *)
> > +
> > +/**
> > + * The mbuf dynamic flag for VLAN protocol extraction metadata, it is
> > +valid
> > + * when dev_args 'flex_desc' has 'vlan' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_VLAN \
> > +(rte_net_iavf_dynflag_flex_desc_vlan_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for IPv4 protocol extraction metadata, it is
> > +valid
> > + * when dev_args 'flex_desc' has 'ipv4' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV4 \
> > +(rte_net_iavf_dynflag_flex_desc_ipv4_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for IPv6 protocol extraction metadata, it is
> > +valid
> > + * when dev_args 'flex_desc' has 'ipv6' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV6 \
> > +(rte_net_iavf_dynflag_flex_desc_ipv6_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for IPv6 with flow protocol extraction
> > +metadata, it is
> > + * valid when dev_args 'flex_desc' has 'ipv6_flow' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV6_FLOW \
> > +(rte_net_iavf_dynflag_flex_desc_ipv6_flow_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for TCP protocol extraction metadata, it is
> > +valid
> > + * when dev_args 'flex_desc' has 'tcp' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_TCP \
> > +(rte_net_iavf_dynflag_flex_desc_tcp_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for the extraction metadata of OVS flexible
> > + * descriptor, it is valid when dev_args 'flex_desc' has 'ovs' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_OVS \
> > +(rte_net_iavf_dynflag_flex_desc_ovs_mask)
> > +
> > +/**
> > + * The mbuf dynamic flag for IP_OFFSET extraction metadata, it is
> > +valid
> > + * when dev_args 'flex_desc' has 'ip_offset' specified.
> > + */
> > +#define RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IP_OFFSET \
> > +(rte_net_iavf_dynflag_flex_desc_ip_offset_mask)
> > +
> > +/**
> > + * Check if mbuf dynamic field for flexible descriptor's extraction
> > +metadata
> > + * is registered.
> > + *
> > + * @return
> > + * True if registered, false otherwise.
> > + */
> > +__rte_experimental
> > +static __rte_always_inline int
> > +rte_net_iavf_dynf_flex_desc_metadata_avail(void)
> > +{
> > +return rte_net_iavf_dynfield_flex_desc_metadata_offs != -1; }
> > +
> > +/**
> > + * Get the mbuf dynamic field for flexible descriptor's extraction metadata.
> > + *
> > + * @param m
> > + * The pointer to the mbuf.
> > + * @return
> > + * The saved protocol extraction metadata.
> > + */
> > +__rte_experimental
> > +static __rte_always_inline uint32_t
> > +rte_net_iavf_dynf_flex_desc_metadata_get(struct rte_mbuf *m)
> { return
> > +*RTE_NET_IAVF_DYNF_FLEX_DESC_METADATA(m);
> > +}
> > +
> > +/**
> > + * Dump the mbuf dynamic field for flexible descriptor's extraction
> metadata.
> > + *
> > + * @param m
> > + * The pointer to the mbuf.
> > + */
> > +__rte_experimental
> > +static inline void
> > +rte_net_iavf_dump_flex_desc_metadata(struct rte_mbuf *m) { union
> > +rte_net_iavf_flex_desc_metadata data;
> > +
> > +if (!rte_net_iavf_dynf_flex_desc_metadata_avail())
> > +return;
> > +
> > +data.metadata = rte_net_iavf_dynf_flex_desc_metadata_get(m);
> > +
> > +if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_VLAN)
> > +printf(" - Flexible descriptor's Metadata: [0x%04x:0x%04x],"
> > + "vlan,stag=%u:%u:%u,ctag=%u:%u:%u",
> > + data.raw.data0, data.raw.data1,
> > + data.vlan.stag_pcp,
> > + data.vlan.stag_dei,
> > + data.vlan.stag_vid,
> > + data.vlan.ctag_pcp,
> > + data.vlan.ctag_dei,
> > + data.vlan.ctag_vid);
> > +else if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV4)
> > +printf(" - Flexible descriptor's Metadata: [0x%04x:0x%04x],"
> > + "ipv4,ver=%u,hdrlen=%u,tos=%u,ttl=%u,proto=%u",
> > + data.raw.data0, data.raw.data1,
> > + data.ipv4.version,
> > + data.ipv4.ihl,
> > + data.ipv4.tos,
> > + data.ipv4.ttl,
> > + data.ipv4.protocol);
> > +else if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV6)
> > +printf(" - Flexible descriptor's Metadata: [0x%04x:0x%04x],"
> > + "ipv6,ver=%u,tc=%u,flow_hi4=0x%x,nexthdr=%u,hoplimit=%u",
> > + data.raw.data0, data.raw.data1,
> > + data.ipv6.version,
> > + data.ipv6.tc,
> > + data.ipv6.flowhi4,
> > + data.ipv6.nexthdr,
> > + data.ipv6.hoplimit);
> > +else if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IPV6_FLOW)
> > +printf(" - Flexible descriptor's Metadata: [0x%04x:0x%04x],"
> > + "ipv6_flow,ver=%u,tc=%u,flow=0x%x%04x",
> > + data.raw.data0, data.raw.data1,
> > + data.ipv6_flow.version,
> > + data.ipv6_flow.tc,
> > + data.ipv6_flow.flowhi4,
> > + data.ipv6_flow.flowlo16);
> > +else if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_TCP)
> > +printf(" - Flexible descriptor's Metadata: [0x%04x:0x%04x],"
> > + "tcp,doff=%u,flags=%s%s%s%s%s%s%s%s",
> > + data.raw.data0, data.raw.data1,
> > + data.tcp.doff,
> > + data.tcp.cwr ? "C" : "",
> > + data.tcp.ece ? "E" : "",
> > + data.tcp.urg ? "U" : "",
> > + data.tcp.ack ? "A" : "",
> > + data.tcp.psh ? "P" : "",
> > + data.tcp.rst ? "R" : "",
> > + data.tcp.syn ? "S" : "",
> > + data.tcp.fin ? "F" : "");
> > +else if (m->ol_flags & RTE_IAVF_PKT_RX_DYNF_FLEX_DESC_IP_OFFSET)
> > +printf(" - Flexible descriptor's Extraction: ip_offset=%u",
> > + data.ip_ofs);
> > +}
> > +
> > +#ifdef __cplusplus
> > +}
> > +#endif
> > +
> > +#endif /* _RTE_PMD_IAVF_H_ */
>
> You need to export these global symbols into rte_pmd_iavf_version.map like:
>
Ok.
> EXPERIMENTAL {
> global:
>
> rte_net_iavf_dynfield_proto_xtr_metadata_offs;
> ...
> };
>
> > --
> > 2.20.1
>
next prev parent reply other threads:[~2020-09-18 2:41 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-09 2:54 Jeff Guo
2020-09-17 3:00 ` Wang, Haiyue
2020-09-18 2:41 ` Guo, Jia [this message]
2020-09-23 7:45 ` [dpdk-dev] [PATCH v2] " Jeff Guo
2020-09-23 7:52 ` [dpdk-dev] [PATCH v3] " Jeff Guo
2020-09-23 8:10 ` Wang, Haiyue
2020-09-23 8:22 ` Guo, Jia
2020-09-23 15:36 ` [dpdk-dev] [PATCH v4] " Jeff Guo
2020-09-25 6:23 ` [dpdk-dev] [PATCH v5] " Jeff Guo
2020-09-25 6:33 ` Wang, Haiyue
2020-09-27 2:08 ` [dpdk-dev] [PATCH v6] " Jeff Guo
2020-09-27 3:00 ` Zhang, Qi Z
2020-09-28 15:59 ` Ferruh Yigit
2020-09-28 16:17 ` Wang, Haiyue
2020-09-28 16:21 ` Bruce Richardson
2020-09-28 16:29 ` Wang, Haiyue
2020-09-29 2:27 ` Guo, Jia
2020-09-29 6:10 ` [dpdk-dev] [PATCH v7] " Jeff Guo
2020-09-29 6:12 ` Jeff Guo
2020-10-13 8:17 ` [dpdk-dev] [PATCH v8] " Jeff Guo
2020-10-13 10:10 ` Zhang, Qi Z
2020-10-14 12:31 ` Ferruh Yigit
2020-10-14 14:03 ` Bruce Richardson
2020-10-15 3:40 ` Guo, Jia
2020-10-15 5:26 ` Guo, Jia
2020-10-15 8:33 ` Ferruh Yigit
2020-10-26 9:37 ` Olivier Matz
2020-10-26 11:41 ` Wang, Haiyue
2020-10-15 3:41 ` [dpdk-dev] [PATCH v9] " Jeff Guo
2020-10-27 5:04 ` [dpdk-dev] [PATCH v10] " Jeff Guo
2020-10-27 5:21 ` Wang, Haiyue
2020-10-27 8:27 ` Guo, Jia
2020-10-27 11:55 ` Zhang, Qi Z
2020-10-30 2:54 ` [dpdk-dev] [PATCH v11] " Jeff Guo
2020-10-30 8:34 ` [dpdk-dev] [PATCH v12] " Jeff Guo
2020-10-30 8:40 ` Jeff Guo
2020-10-30 9:35 ` Zhang, Qi Z
2020-10-30 10:51 ` Ferruh Yigit
2020-10-30 11:14 ` Zhang, Qi Z
2020-10-30 16:03 ` Ferruh Yigit
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=ae77f15b55cb405f892d6d0e244ead51@intel.com \
--to=jia.guo@intel.com \
--cc=beilei.xing@intel.com \
--cc=dev@dpdk.org \
--cc=haiyue.wang@intel.com \
--cc=jingjing.wu@intel.com \
--cc=qi.z.zhang@intel.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).