RE: [PATCH v3 2/4] ethdev: introduce protocol hdr based buffer split

["Wang, YuanX" <yuanx.wang@intel.com>]

Hi Andrew,

> -----Original Message-----
> From: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
> Sent: Tuesday, September 20, 2022 1:35 PM
> To: Wang, YuanX <yuanx.wang@intel.com>; dev@dpdk.org; Thomas
> Monjalon <thomas@monjalon.net>; Ferruh Yigit <ferruh.yigit@xilinx.com>
> Cc: mdr@ashroe.eu; Li, Xiaoyun <xiaoyun.li@intel.com>; Singh, Aman Deep
> <aman.deep.singh@intel.com>; Zhang, Yuying <yuying.zhang@intel.com>;
> Zhang, Qi Z <qi.z.zhang@intel.com>; Yang, Qiming <qiming.yang@intel.com>;
> jerinjacobk@gmail.com; viacheslavo@nvidia.com;
> stephen@networkplumber.org; Ding, Xuan <xuan.ding@intel.com>;
> hpothula@marvell.com; Tang, Yaqi <yaqi.tang@intel.com>; Wenxuan Wu
> <wenxuanx.wu@intel.com>
> Subject: Re: [PATCH v3 2/4] ethdev: introduce protocol hdr based buffer split
> 
> On 9/16/22 11:38, Wang, YuanX wrote:
> > Hi Andrew,
> >
> >> -----Original Message-----
> >> From: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
> >> Sent: Monday, September 12, 2022 7:47 PM
> >> To: Wang, YuanX <yuanx.wang@intel.com>; dev@dpdk.org; Thomas
> Monjalon
> >> <thomas@monjalon.net>; Ferruh Yigit <ferruh.yigit@xilinx.com>
> >> Cc: mdr@ashroe.eu; Li, Xiaoyun <xiaoyun.li@intel.com>; Singh, Aman
> >> Deep <aman.deep.singh@intel.com>; Zhang, Yuying
> >> <yuying.zhang@intel.com>; Zhang, Qi Z <qi.z.zhang@intel.com>; Yang,
> >> Qiming <qiming.yang@intel.com>; jerinjacobk@gmail.com;
> >> viacheslavo@nvidia.com; stephen@networkplumber.org; Ding, Xuan
> >> <xuan.ding@intel.com>; hpothula@marvell.com; Tang, Yaqi
> >> <yaqi.tang@intel.com>; Wenxuan Wu <wenxuanx.wu@intel.com>
> >> Subject: Re: [PATCH v3 2/4] ethdev: introduce protocol hdr based
> >> buffer split
> >>
> >> On 9/2/22 22:10, Yuan Wang wrote:
> >>> Currently, Rx buffer split supports length based split. With Rx
> >>> queue offload RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT enabled and Rx
> packet
> >> segment
> >>> configured, PMD will be able to split the received packets into
> >>> multiple segments.
> >>>
> >>> However, length based buffer split is not suitable for NICs that do
> >>> split based on protocol headers. Given an arbitrarily variable
> >>> length in Rx packet segment, it is almost impossible to pass a fixed
> >>> protocol header to driver. Besides, the existence of tunneling
> >>> results in the composition of a packet is various, which makes the
> situation even worse.
> >>>
> >>> This patch extends current buffer split to support protocol header
> >>> based buffer split. A new proto_hdr field is introduced in the
> >>> reserved field of rte_eth_rxseg_split structure to specify protocol
> >>> header. The proto_hdr field defines the split position of packet,
> >>> splitting will always happens after the protocol header defined in
> >>> the Rx packet segment. When Rx queue offload
> >>> RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT is enabled and corresponding
> >> protocol
> >>> header is configured, driver will split the ingress packets into
> >>> multiple
> >> segments.
> >>>
> >>> struct rte_eth_rxseg_split {
> >>>           struct rte_mempool *mp; /* memory pools to allocate
> >>> segment from
> >> */
> >>>           uint16_t length; /* segment maximal data length,
> >>>                               configures split point */
> >>>           uint16_t offset; /* data offset from beginning
> >>>                               of mbuf data buffer */
> >>>           uint32_t proto_hdr; /* supported ptype of a specific pmd,
> >>>                                  configures split point.
> >>> 			       It should be defined by RTE_PTYPE_*
> >>
> >> If I understand correctly, the statement is a bit misleading since it
> >> should be a bit mask of RTE_PTYPE_* defines. Not exactly one
> RTE_PTYPE_*.
> >
> > Do you mean that a segment should support multiple protocol headers,
> such as splitting both tcp and udp headers?
> 
> No-no. Look. In order to split after some protocol, for example UDP, NIC
> should recognice all previous protocols. Moreover, in the case of a tunenl
> UDP could be inner and outer. At which point would you like to split if NIC
> supports both?
> Another example, if NIC support stplit after Eth-IPv4-UDP and after Eth-IPv6-
> UDP, how to request to split just after Eth-IPv4-UDP, but not Eth-IPv6-UDP?

Thank you for your patience. 

We have a proposal to solve this problem. 
We define the proto_hdr as a bit mask. Each mask should contain the composition of the packet, but inner and outer are written separately (to avoid unnecessary trouble).
We use the highest RTE_PTYPE* in the mask to define the split position.

For the first example, since ptype has distinguished between outer and inner, external UDP can be simply written as RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_UDP.
The inner UDP can be simply written as RTE_PTYPE_TUNNEL_GRENAT | RTE_PTYPE_INNER_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_INNER_L4_UDP.

 For the second example, Eth-IPv4-UDP can be written as RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV4_EXT_UNKNOWN | RTE_PTYPE_L4_UDP.
 Eth-IPv6-UDP can be written as RTE_PTYPE_L2_ETHER | RTE_PTYPE_L3_IPV6_EXT_UNKNOWN | RTE_PTYPE_L4_UDP.

What do you think?

Thanks,
Yuan

> 
> >
> >>
> >>> 			     */
> >>> };
> >>>
> >>> If protocol header split can be supported by a PMD. The
> >>> rte_eth_buffer_split_get_supported_hdr_ptypes function can be use
> to
> >>> obtain a list of these protocol headers.
> >>>
> >>> For example, let's suppose we configured the Rx queue with the
> >>> following segments:
> >>>           seg0 - pool0, proto_hdr0=RTE_PTYPE_L3_IPV4, off0=2B
> >>>           seg1 - pool1, proto_hdr1=RTE_PTYPE_L4_UDP, off1=128B
> >>>           seg2 - pool2, off1=0B
> >>>
> >>> The packet consists of MAC_IPV4_UDP_PAYLOAD will be split like
> >>
> >> What is MAC_IPV4_UDP_PAYLOAD? Do you mean
> ETH_IPV4_UDP_PAYLOAD?
> >
> > Thanks for your correction, it should be ETH_IPV4_UDP_PAYLOAD.
> >
> >>
> >>> following:
> >>>           seg0 - ipv4 header @ RTE_PKTMBUF_HEADROOM + 2 in mbuf from
> >> pool0
> >>>           seg1 - udp header @ 128 in mbuf from pool1
> >>>           seg2 - payload @ 0 in mbuf from pool2
> >>>
> >>> Note: NIC will only do split when the packets exactly match all the
> >>> protocol headers in the segments. For example, if ARP packets
> >>> received with above config, the NIC won't do split for ARP packets
> >>> since it does not contains ipv4 header and udp header.
> >>
> >> You must define which mempool is used in the case.
> >
> > IMHO I don't think we can define which mempool to use, it depends on NIC
> behavior.
> > For our NIC, packets that are unable to split will be put into the last valid
> pool, with zero offset.
> > So here we would like to define to put these packets into the last valid
> mempool, with zero offset.
> 
> 
> Anyway API should not be silent about the case since it is the first question at
> least in my head. IMHO the last segment is the only sensible option since it is
> typically will be big enough. Other mempool for protocol headers are likely to
> be small. So, I suggest to define this way.
> 
> >>
> >>>
> >>> Now buffer split can be configured in two modes. For length based
> >>> buffer split, the mp, length, offset field in Rx packet segment
> >>> should be configured, while the proto_hdr field will be ignored.
> >>> For protocol header based buffer split, the mp, offset, proto_hdr
> >>> field in Rx packet segment should be configured, while the length
> >>> field will be ignored.
> >>>
> >>> The split limitations imposed by underlying driver is reported in
> >>> the rte_eth_dev_info->rx_seg_capa field. The memory attributes for
> >>> the split parts may differ either, dpdk memory and external memory,
> >> respectively.
> >>>
> >>> Signed-off-by: Yuan Wang <yuanx.wang@intel.com>
> >>> Signed-off-by: Xuan Ding <xuan.ding@intel.com>
> >>> Signed-off-by: Wenxuan Wu <wenxuanx.wu@intel.com>
> >>> ---
> >>>    doc/guides/rel_notes/release_22_11.rst |  5 +++
> >>>    lib/ethdev/rte_ethdev.c                | 55 ++++++++++++++++++++------
> >>>    lib/ethdev/rte_ethdev.h                | 17 +++++++-
> >>>    3 files changed, 65 insertions(+), 12 deletions(-)
> >>>
> >>> diff --git a/doc/guides/rel_notes/release_22_11.rst
> >>> b/doc/guides/rel_notes/release_22_11.rst
> >>> index 4d90514a9a..f3b58c7895 100644
> >>> --- a/doc/guides/rel_notes/release_22_11.rst
> >>> +++ b/doc/guides/rel_notes/release_22_11.rst
> >>> @@ -60,6 +60,11 @@ New Features
> >>>      Added ``rte_eth_buffer_split_get_supported_hdr_ptypes()``, to
> >>> get
> >> supported
> >>>      header protocols of a PMD to split.
> >>>
> >>> +* **Added protocol header based buffer split.**
> >>> +  Ethdev: The ``reserved`` field in the  ``rte_eth_rxseg_split``
> >>> +structure is
> >>> +  replaced with ``proto_hdr`` to support protocol header based buffer
> split.
> >>> +  User can choose length or protocol header to configure buffer
> >>> +split
> >>> +  according to NIC's capability.
> >>
> >> Add one more empty line to have two before the next sectoin.
> >
> > Thanks for your catch.
> >
> >>
> >>>
> >>>    Removed Items
> >>>    -------------
> >>> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c index
> >>> 093c577add..dfceb723ee 100644
> >>> --- a/lib/ethdev/rte_ethdev.c
> >>> +++ b/lib/ethdev/rte_ethdev.c
> >>> @@ -1635,9 +1635,10 @@ rte_eth_dev_is_removed(uint16_t port_id)
> >>>    }
> >>>
> >>>    static int
> >>> -rte_eth_rx_queue_check_split(const struct rte_eth_rxseg_split
> *rx_seg,
> >>> -			     uint16_t n_seg, uint32_t *mbp_buf_size,
> >>> -			     const struct rte_eth_dev_info *dev_info)
> >>> +rte_eth_rx_queue_check_split(uint16_t port_id,
> >>> +			const struct rte_eth_rxseg_split *rx_seg,
> >>> +			uint16_t n_seg, uint32_t *mbp_buf_size,
> >>> +			const struct rte_eth_dev_info *dev_info)
> >>>    {
> >>>    	const struct rte_eth_rxseg_capa *seg_capa = &dev_info-
> >>> rx_seg_capa;
> >>>    	struct rte_mempool *mp_first;
> >>> @@ -1660,6 +1661,7 @@ rte_eth_rx_queue_check_split(const struct
> >> rte_eth_rxseg_split *rx_seg,
> >>>    		struct rte_mempool *mpl = rx_seg[seg_idx].mp;
> >>>    		uint32_t length = rx_seg[seg_idx].length;
> >>>    		uint32_t offset = rx_seg[seg_idx].offset;
> >>> +		uint32_t proto_hdr = rx_seg[seg_idx].proto_hdr;
> >>>
> >>>    		if (mpl == NULL) {
> >>>    			RTE_ETHDEV_LOG(ERR, "null mempool pointer\n");
> >> @@ -1693,13
> >>> +1695,44 @@ rte_eth_rx_queue_check_split(const struct
> >> rte_eth_rxseg_split *rx_seg,
> >>>    		}
> >>>    		offset += seg_idx != 0 ? 0 : RTE_PKTMBUF_HEADROOM;
> >>>    		*mbp_buf_size = rte_pktmbuf_data_room_size(mpl);
> >>> -		length = length != 0 ? length : *mbp_buf_size;
> >>> -		if (*mbp_buf_size < length + offset) {
> >>> -			RTE_ETHDEV_LOG(ERR,
> >>> -				       "%s mbuf_data_room_size %u < %u
> >> (segment length=%u + segment offset=%u)\n",
> >>> -				       mpl->name, *mbp_buf_size,
> >>> -				       length + offset, length, offset);
> >>> -			return -EINVAL;
> >>> +
> >>> +		int ret =
> >> rte_eth_buffer_split_get_supported_hdr_ptypes(port_id,
> >>> +NULL, 0);
> >>
> >> Do not mix variable declaration and code.
> >> It is better to give the variable some sensible name.
> >> Otherwise else branch code is hard to read.
> >
> > Thanks for the suggestion, will  take care of naming.
> >
> >>
> >>> +		if (ret <= 0) {
> >>
> >> May be I'm missing something, but nothing prevetns a driver/HW to
> >> support both protocol-based and fixed-length split.
> >> So, ability to support protocol based split should be treated as a
> >> request to do it. It must be based on rx_seg->proto_hdr content (for all
> segments).
> >>
> >> Also nothing should prevent to mix protocol and fixed-length split.
> >> I.e. split just after UDP in the first segment,
> >> 40 bytes in the second segment, everything else in the third.
> >
> > Mix mode is an interesting idea. Currently testpmd and driver do not
> support mixed mode, but it does not affect the library to support this mode.
> 
> That's OK.
> 
> >
> >>
> >>> +			/* Split at fixed length. */
> >>> +			length = length != 0 ? length : *mbp_buf_size;
> >>> +			if (*mbp_buf_size < length + offset) {
> >>> +				RTE_ETHDEV_LOG(ERR,
> >>> +					"%s mbuf_data_room_size %u < %u
> >> (segment length=%u + segment offset=%u)\n",
> >>> +					mpl->name, *mbp_buf_size,
> >>> +					length + offset, length, offset);
> >>> +				return -EINVAL;
> >>> +			}
> >>> +		} else {
> >>> +			/* Split after specified protocol header. */
> >>> +			uint32_t ptypes[ret];
> >>> +			int i;
> >>> +
> >>> +			ret =
> >> rte_eth_buffer_split_get_supported_hdr_ptypes(port_id,
> >>> +ptypes, ret);
> >>
> >> In theory, the funciton could fail since input arguments differ. So,
> >> it should be handled.
> >
> > Thanks for your catch, will fix in the next version.
> >
> >>
> >>> +			for (i = 0; i < ret; i++)
> >>> +				if (ptypes[i] & proto_hdr)
> >>
> >> IMHO it should be ==, not &. I think that
> >> rte_eth_buffer_split_get_supported_hdr_ptypes() should define points
> >> at which split could happen and we should match the point exactly.
> >
> > Sure, == is better. Thanks for the suggestion.
> >
> >>
> >>> +					break;
> >>> +
> >>> +			if (i == ret) {
> >>> +#define PTYPE_NAMESIZE	256
> >>
> >> Why? It is looks really strange that it is defined here.
> >
> > I intend to display the protocol name in the log, but if the proto_hdr is a bit
> mask, can I just show the number?
> > Please see v4 for this modification.
> >
> >>
> >>> +				char ptype_name[PTYPE_NAMESIZE];
> >>> +				rte_get_ptype_name(proto_hdr,
> >> ptype_name, sizeof(ptype_name));
> >>> +				RTE_ETHDEV_LOG(ERR,
> >>> +					"Protocol header %s is not
> >> supported.\n",
> >>> +					ptype_name);
> >>> +				return -EINVAL;
> >>> +			}
> >>> +			if (*mbp_buf_size < offset) {
> >>> +				RTE_ETHDEV_LOG(ERR,
> >>> +						"%s
> >> mbuf_data_room_size %u < %u segment offset)\n",
> >>> +						mpl->name, *mbp_buf_size,
> >>> +						offset);
> >>> +				return -EINVAL;
> >>> +			}
> >>>    		}
> >>>    	}
> >>>    	return 0;
> >>> @@ -1778,7 +1811,7 @@ rte_eth_rx_queue_setup(uint16_t port_id,
> >> uint16_t rx_queue_id,
> >>>    		n_seg = rx_conf->rx_nseg;
> >>>
> >>>    		if (rx_conf->offloads &
> >> RTE_ETH_RX_OFFLOAD_BUFFER_SPLIT) {
> >>> -			ret = rte_eth_rx_queue_check_split(rx_seg, n_seg,
> >>> +			ret = rte_eth_rx_queue_check_split(port_id, rx_seg,
> >> n_seg,
> >>>    							   &mbp_buf_size,
> >>>    							   &dev_info);
> >>>    			if (ret != 0)
> >>> diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h index
> >>> c58c908c3a..410fba5eab 100644
> >>> --- a/lib/ethdev/rte_ethdev.h
> >>> +++ b/lib/ethdev/rte_ethdev.h
> >>> @@ -1175,6 +1175,9 @@ struct rte_eth_txmode {
> >>>     *   specified in the first array element, the second buffer, from the
> >>>     *   pool in the second element, and so on.
> >>>     *
> >>> + * - The proto_hdrs in the elements define the split position of
> >>> + *   received packets.
> >>> + *
> >>>     * - The offsets from the segment description elements specify
> >>>     *   the data offset from the buffer beginning except the first mbuf.
> >>>     *   The first segment offset is added with RTE_PKTMBUF_HEADROOM.
> >>> @@ -1196,12 +1199,24 @@ struct rte_eth_txmode {
> >>>     *     - pool from the last valid element
> >>>     *     - the buffer size from this pool
> >>>     *     - zero offset
> >>> + *
> >>> + * - Length based buffer split:
> >>> + *     - mp, length, offset should be configured.
> >>> + *     - The proto_hdr field will be ignored.
> >>> + *
> >>> + * - Protocol header based buffer split:
> >>> + *     - mp, offset, proto_hdr should be configured.
> >>> + *     - The length field will be ignored.
> >>>     */
> >>>    struct rte_eth_rxseg_split {
> >>>    	struct rte_mempool *mp; /**< Memory pool to allocate segment
> >> from. */
> >>>    	uint16_t length; /**< Segment data length, configures split point. */
> >>>    	uint16_t offset; /**< Data offset from beginning of mbuf data buffer.
> >> */
> >>> -	uint32_t reserved; /**< Reserved field. */
> >>> +	/**
> >>> +	 * Supported ptype of a specific pmd, configures split point.
> >>> +	 * It should be defined by RTE_PTYPE_*.
> >>> +	 */
> >>> +	uint32_t proto_hdr;
> >>>    };
> >>>
> >>>    /**
> >