[dpdk-dev] [PATCH v10 1/3] lib: add Generic Receive Offload API framework
Hu, Jiayu
jiayu.hu at intel.com
Mon Jul 3 07:56:20 CEST 2017
Hi Jianfeng,
> -----Original Message-----
> From: Tan, Jianfeng
> Sent: Sunday, July 2, 2017 6:20 PM
> To: Hu, Jiayu <jiayu.hu at intel.com>; dev at dpdk.org
> Cc: Ananyev, Konstantin <konstantin.ananyev at intel.com>;
> stephen at networkplumber.org; yliu at fridaylinux.org; Wu, Jingjing
> <jingjing.wu at intel.com>; Yao, Lei A <lei.a.yao at intel.com>; Bie, Tiwei
> <tiwei.bie at intel.com>
> Subject: Re: [PATCH v10 1/3] lib: add Generic Receive Offload API framework
>
>
>
> On 7/1/2017 7:08 PM, Jiayu Hu wrote:
> > Generic Receive Offload (GRO) is a widely used SW-based offloading
> > technique to reduce per-packet processing overhead. It gains
> > performance by reassembling small packets into large ones. This
> > patchset is to support GRO in DPDK. To support GRO, this patch
> > implements a GRO API framework.
> >
> > To enable more flexibility to applications, DPDK GRO is implemented as
> > a user library. Applications explicitly use the GRO library to merge
> > small packets into large ones. DPDK GRO provides two reassembly modes.
> > One is called lightweight mode, the other is called heavyweight mode.
> > If applications want to merge packets in a simple way and the number
> > of packets is relatively small, they can use the lightweight mode.
> > If applications need more fine-grained controls, they can choose the
> > heavyweight mode.
> >
> > rte_gro_reassemble_burst is the main reassembly API which is used in
> > lightweight mode and processes N packets at a time. For applications,
> > performing GRO in lightweight mode is simple. They just need to invoke
> > rte_gro_reassemble_burst. Applications can get GROed packets as soon as
> > rte_gro_reassemble_burst returns.
> >
> > rte_gro_reassemble is the main reassembly API which is used in
> > heavyweight mode and tries to merge N inputted packets with the packets
> > in a givn GRO table. For applications, performing GRO in heavyweight
> > mode is relatively complicated. Before performing GRO, applications need
> > to create a GRO table by rte_gro_tbl_create. Then they can use
> > rte_gro_reassemble to merge packets. The GROed packets are in the GRO
> > table. If applications want to get them, applications need to manually
> > flush them by flush API.
>
> > Signed-off-by: Jiayu Hu <jiayu.hu at intel.com>
> > ---
> > config/common_base | 5 ++
> > lib/Makefile | 2 +
> > lib/librte_gro/Makefile | 50 +++++++++++
> > lib/librte_gro/rte_gro.c | 176
> +++++++++++++++++++++++++++++++++++++
> > lib/librte_gro/rte_gro.h | 176
> +++++++++++++++++++++++++++++++++++++
> > lib/librte_gro/rte_gro_version.map | 12 +++
> > mk/rte.app.mk | 1 +
> > 7 files changed, 422 insertions(+)
> > create mode 100644 lib/librte_gro/Makefile
> > create mode 100644 lib/librte_gro/rte_gro.c
> > create mode 100644 lib/librte_gro/rte_gro.h
> > create mode 100644 lib/librte_gro/rte_gro_version.map
> >
> > diff --git a/config/common_base b/config/common_base
> > index f6aafd1..167f5ef 100644
> > --- a/config/common_base
> > +++ b/config/common_base
> > @@ -712,6 +712,11 @@ CONFIG_RTE_LIBRTE_VHOST_DEBUG=n
> > CONFIG_RTE_LIBRTE_PMD_VHOST=n
> >
> > #
> > +# Compile GRO library
> > +#
> > +CONFIG_RTE_LIBRTE_GRO=y
> > +
> > +#
> > #Compile Xen domain0 support
> > #
> > CONFIG_RTE_LIBRTE_XEN_DOM0=n
> > diff --git a/lib/Makefile b/lib/Makefile
> > index 07e1fd0..ac1c2f6 100644
> > --- a/lib/Makefile
> > +++ b/lib/Makefile
> > @@ -106,6 +106,8 @@ DIRS-$(CONFIG_RTE_LIBRTE_REORDER) +=
> librte_reorder
> > DEPDIRS-librte_reorder := librte_eal librte_mempool librte_mbuf
> > DIRS-$(CONFIG_RTE_LIBRTE_PDUMP) += librte_pdump
> > DEPDIRS-librte_pdump := librte_eal librte_mempool librte_mbuf
> librte_ether
> > +DIRS-$(CONFIG_RTE_LIBRTE_GRO) += librte_gro
> > +DEPDIRS-librte_gro := librte_eal librte_mbuf librte_ether librte_net
> >
> > ifeq ($(CONFIG_RTE_EXEC_ENV_LINUXAPP),y)
> > DIRS-$(CONFIG_RTE_LIBRTE_KNI) += librte_kni
> > diff --git a/lib/librte_gro/Makefile b/lib/librte_gro/Makefile
> > new file mode 100644
> > index 0000000..7e0f128
> > --- /dev/null
> > +++ b/lib/librte_gro/Makefile
> > @@ -0,0 +1,50 @@
> > +# BSD LICENSE
> > +#
> > +# Copyright(c) 2017 Intel Corporation. All rights reserved.
> > +# All rights reserved.
> > +#
> > +# Redistribution and use in source and binary forms, with or without
> > +# modification, are permitted provided that the following conditions
> > +# are met:
> > +#
> > +# * Redistributions of source code must retain the above copyright
> > +# notice, this list of conditions and the following disclaimer.
> > +# * Redistributions in binary form must reproduce the above copyright
> > +# notice, this list of conditions and the following disclaimer in
> > +# the documentation and/or other materials provided with the
> > +# distribution.
> > +# * Neither the name of Intel Corporation nor the names of its
> > +# contributors may be used to endorse or promote products derived
> > +# from this software without specific prior written permission.
> > +#
> > +# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
> CONTRIBUTORS
> > +# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT
> NOT
> > +# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
> FITNESS FOR
> > +# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> COPYRIGHT
> > +# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> INCIDENTAL,
> > +# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
> BUT NOT
> > +# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> LOSS OF USE,
> > +# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
> AND ON ANY
> > +# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
> TORT
> > +# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
> OF THE USE
> > +# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
> DAMAGE.
> > +
> > +include $(RTE_SDK)/mk/rte.vars.mk
> > +
> > +# library name
> > +LIB = librte_gro.a
> > +
> > +CFLAGS += -O3
> > +CFLAGS += $(WERROR_FLAGS) -I$(SRCDIR)
> > +
> > +EXPORT_MAP := rte_gro_version.map
> > +
> > +LIBABIVER := 1
> > +
> > +# source files
> > +SRCS-$(CONFIG_RTE_LIBRTE_GRO) += rte_gro.c
> > +
> > +# install this header file
> > +SYMLINK-$(CONFIG_RTE_LIBRTE_GRO)-include += rte_gro.h
> > +
> > +include $(RTE_SDK)/mk/rte.lib.mk
> > diff --git a/lib/librte_gro/rte_gro.c b/lib/librte_gro/rte_gro.c
> > new file mode 100644
> > index 0000000..648835b
> > --- /dev/null
> > +++ b/lib/librte_gro/rte_gro.c
> > @@ -0,0 +1,176 @@
> > +/*-
> > + * BSD LICENSE
> > + *
> > + * Copyright(c) 2017 Intel Corporation. All rights reserved.
> > + *
> > + * Redistribution and use in source and binary forms, with or without
> > + * modification, are permitted provided that the following conditions
> > + * are met:
> > + *
> > + * * Redistributions of source code must retain the above copyright
> > + * notice, this list of conditions and the following disclaimer.
> > + * * Redistributions in binary form must reproduce the above copyright
> > + * notice, this list of conditions and the following disclaimer in
> > + * the documentation and/or other materials provided with the
> > + * distribution.
> > + * * Neither the name of Intel Corporation nor the names of its
> > + * contributors may be used to endorse or promote products derived
> > + * from this software without specific prior written permission.
> > + *
> > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
> CONTRIBUTORS
> > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT
> NOT
> > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
> FITNESS FOR
> > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> COPYRIGHT
> > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> INCIDENTAL,
> > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
> BUT NOT
> > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> LOSS OF USE,
> > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
> AND ON ANY
> > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
> TORT
> > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
> OF THE USE
> > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
> DAMAGE.
> > + */
> > +
> > +#include <rte_malloc.h>
> > +#include <rte_mbuf.h>
> > +
> > +#include "rte_gro.h"
> > +
> > +typedef void *(*gro_tbl_create_fn)(uint16_t socket_id,
> > + uint16_t max_flow_num,
> > + uint16_t max_item_per_flow);
> > +typedef void (*gro_tbl_destroy_fn)(void *tbl);
> > +typedef uint32_t (*gro_tbl_item_num_fn)(void *tbl);
> > +
> > +static gro_tbl_create_fn
> tbl_create_functions[RTE_GRO_TYPE_MAX_NUM];
> > +static gro_tbl_destroy_fn
> tbl_destroy_functions[RTE_GRO_TYPE_MAX_NUM];
> > +static gro_tbl_item_num_fn
> tbl_item_num_functions[RTE_GRO_TYPE_MAX_NUM];
> > +
> > +/**
> > + * GRO table, which is used to merge packets. It keeps many reassembly
> > + * tables of desired GRO types. Applications need to create GRO tables
> > + * before using rte_gro_reassemble to perform GRO.
> > + */
> > +struct gro_tbl {
> > + uint64_t desired_gro_types; /**< GRO types to perform */
> > + /* max TTL measured in nanosecond */
> > + uint64_t max_timeout_cycles;
> > + /* max length of merged packet measured in byte */
> > + uint32_t max_packet_size;
> > + /* reassebly tables of desired GRO types */
> > + void *tbls[RTE_GRO_TYPE_MAX_NUM];
> > +};
> > +
> > +void *rte_gro_tbl_create(const
> > + const struct rte_gro_param *param)
>
> The name of this API and the definition of struct gro_tbl involve some
> confusion. A gro table contains gro tables? I suppose a better name is
> needed, for example, struct gro_ctl.
Actually, a GRO table includes N reassembly tables. But gro_tbl is not a good
name. I will change the name. Thanks.
>
> > +{
> > + gro_tbl_create_fn create_tbl_fn;
> > + gro_tbl_destroy_fn destroy_tbl_fn;
> > + struct gro_tbl *gro_tbl;
> > + uint64_t gro_type_flag = 0;
> > + uint8_t i, j;
> > +
> > + gro_tbl = rte_zmalloc_socket(__func__,
> > + sizeof(struct gro_tbl),
> > + RTE_CACHE_LINE_SIZE,
> > + param->socket_id);
> > + if (gro_tbl == NULL)
> > + return NULL;
> > + gro_tbl->max_packet_size = param->max_packet_size;
> > + gro_tbl->max_timeout_cycles = param->max_timeout_cycles;
> > + gro_tbl->desired_gro_types = param->desired_gro_types;
> > +
> > + for (i = 0; i < RTE_GRO_TYPE_MAX_NUM; i++) {
> > + gro_type_flag = 1 << i;
> > +
> > + if ((param->desired_gro_types & gro_type_flag) == 0)
> > + continue;
> > + create_tbl_fn = tbl_create_functions[i];
> > + if (create_tbl_fn == NULL)
> > + continue;
> > +
> > + gro_tbl->tbls[i] = create_tbl_fn(
> > + param->socket_id,
> > + param->max_flow_num,
> > + param->max_item_per_flow);
>
> Here and somewhere else: the alignment seems not correct.
> gro_tbl->tbls[i] = create_tbl_fn(param->socket_id,
> /* keep
> all parameters aligned like this */
> param->max_flow_num,
> param->max_item_per_flow);
Thanks, I will modify it.
> > + if (gro_tbl->tbls[i] == NULL) {
> > + /* destroy all allocated tables */
> > + for (j = 0; j < i; j++) {
> > + gro_type_flag = 1 << j;
> > + if ((param->desired_gro_types &
> gro_type_flag) == 0)
> > + continue;
> > + destroy_tbl_fn = tbl_destroy_functions[j];
> > + if (destroy_tbl_fn)
> > + destroy_tbl_fn(gro_tbl->tbls[j]);
> > + }
> > + rte_free(gro_tbl);
> > + return NULL;
> > + }
> > + }
> > + return gro_tbl;
> > +}
> > +
> > +void rte_gro_tbl_destroy(void *tbl)
> > +{
> > + gro_tbl_destroy_fn destroy_tbl_fn;
> > + struct gro_tbl *gro_tbl = (struct gro_tbl *)tbl;
> > + uint64_t gro_type_flag;
> > + uint8_t i;
> > +
> > + if (gro_tbl == NULL)
> > + return;
> > + for (i = 0; i < RTE_GRO_TYPE_MAX_NUM; i++) {
> > + gro_type_flag = 1 << i;
> > + if ((gro_tbl->desired_gro_types & gro_type_flag) == 0)
> > + continue;
> > + destroy_tbl_fn = tbl_destroy_functions[i];
> > + if (destroy_tbl_fn)
> > + destroy_tbl_fn(gro_tbl->tbls[i]);
> > + }
> > + rte_free(gro_tbl);
> > +}
> > +
> > +uint16_t
> > +rte_gro_reassemble_burst(struct rte_mbuf **pkts __rte_unused,
> > + uint16_t nb_pkts,
> > + const struct rte_gro_param *param __rte_unused)
> > +{
> > + return nb_pkts;
> > +}
> > +
> > +uint16_t
> > +rte_gro_reassemble(struct rte_mbuf **pkts __rte_unused,
> > + uint16_t nb_pkts,
> > + void *tbl __rte_unused)
> > +{
> > + return nb_pkts;
> > +}
> > +
> > +uint16_t
> > +rte_gro_timeout_flush(void *tbl __rte_unused,
> > + uint64_t desired_gro_types __rte_unused,
> > + struct rte_mbuf **out __rte_unused,
> > + uint16_t max_nb_out __rte_unused)
> > +{
> > + return 0;
> > +}
> > +
> > +uint64_t rte_gro_tbl_item_num(void *tbl)
>
> Does rte_gro_get_count() sound better?
OK, I will change the name.
>
> > +{
> > + struct gro_tbl *gro_tbl = (struct gro_tbl *)tbl;
> > + gro_tbl_item_num_fn item_num_fn;
> > + uint64_t item_num = 0;
> > + uint64_t gro_type_flag;
> > + uint8_t i;
> > +
> > + for (i = 0; i < RTE_GRO_TYPE_MAX_NUM; i++) {
> > + gro_type_flag = 1 << i;
> > + if ((gro_tbl->desired_gro_types & gro_type_flag) == 0)
> > + continue;
> > +
> > + item_num_fn = tbl_item_num_functions[i];
> > + if (item_num_fn == NULL)
> > + continue;
> > + item_num += item_num_fn(gro_tbl->tbls[i]);
> > + }
> > + return item_num;
> > +}
> > diff --git a/lib/librte_gro/rte_gro.h b/lib/librte_gro/rte_gro.h
> > new file mode 100644
> > index 0000000..02c9113
> > --- /dev/null
> > +++ b/lib/librte_gro/rte_gro.h
> > @@ -0,0 +1,176 @@
> > +/*-
> > + * BSD LICENSE
> > + *
> > + * Copyright(c) 2017 Intel Corporation. All rights reserved.
> > + *
> > + * Redistribution and use in source and binary forms, with or without
> > + * modification, are permitted provided that the following conditions
> > + * are met:
> > + *
> > + * * Redistributions of source code must retain the above copyright
> > + * notice, this list of conditions and the following disclaimer.
> > + * * Redistributions in binary form must reproduce the above copyright
> > + * notice, this list of conditions and the following disclaimer in
> > + * the documentation and/or other materials provided with the
> > + * distribution.
> > + * * Neither the name of Intel Corporation nor the names of its
> > + * contributors may be used to endorse or promote products derived
> > + * from this software without specific prior written permission.
> > + *
> > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
> CONTRIBUTORS
> > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT
> NOT
> > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
> FITNESS FOR
> > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> COPYRIGHT
> > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> INCIDENTAL,
> > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
> BUT NOT
> > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> LOSS OF USE,
> > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
> AND ON ANY
> > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
> TORT
> > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
> OF THE USE
> > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
> DAMAGE.
> > + */
> > +
> > +#ifndef _RTE_GRO_H_
> > +#define _RTE_GRO_H_
> > +
> > +#ifdef __cplusplus
> > +extern "C" {
> > +#endif
> > +
> > +/**
> > + * the max packets number that rte_gro_reassemble_burst can
> > + * process in each invocation.
> > + */
> > +#define RTE_GRO_MAX_BURST_ITEM_NUM 128UL
> > +
> > +/* max number of supported GRO types */
> > +#define RTE_GRO_TYPE_MAX_NUM 64
> > +#define RTE_GRO_TYPE_SUPPORT_NUM 0 /**< current supported
> GRO num */
> > +
> > +
> > +struct rte_gro_param {
> > + uint64_t desired_gro_types; /**< desired GRO types */
>
> Make it gro_types for simplicity.
Thanks, I will change it.
>
> > + uint32_t max_packet_size; /**< max length of merged packets
> */
>
> Refer to the tcp4 gro implementation, this is the max size for tcp
> payload. But in principle, the 65535-byte limitation (including TCP
> header) is because IP header lenght is 2-byte long.
Yes, it's a bug. The length of (IP header + TCP header + payload)
should be less than 64KB. But max_packet_size is uint32_t.
>
> What does it means for other GRO engines then? I think these should be
> decided by each gro engine. And applications don't have to change them.
Make sense. We shouldn't expose it applications. I will remove this parameter.
About the above bug, I will fix it by limiting the max length of a TCP/IPv4 packet
to (l2_header + 64KB).
>
> > + uint16_t max_flow_num; /**< max flow number */
> > + uint16_t max_item_per_flow; /**< max packet number per flow
> */
> > +
> > + /* socket index where the Ethernet port connects to */
>
> The comment needs to be refined. We have different socket idx for port,
> for pmd thread. We just explain how this will be used: "socket index for
> allocating gro related data structure".
Thanks, I will change it.
>
> > + uint16_t socket_id;
> > + /* max TTL for a packet in the GRO table, measured in nanosecond
> */
> > + uint64_t max_timeout_cycles;
>
> We don't need to set it in lightweight mode. Please add this into the
> comment.
Thanks, I will add this comment.
>
> > +};
> > +
> > +/**
> > + * This function create a GRO table, which is used to merge packets in
> > + * rte_gro_reassemble.
> > + *
> > + * @param param
> > + * applications use it to pass needed parameters to create a GRO table.
> > + * @return
> > + * if create successfully, return a pointer which points to the GRO
> > + * table. Otherwise, return NULL.
> > + */
> > +void *rte_gro_tbl_create(
> > + const struct rte_gro_param *param);
>
> Merge above two lines into one.
Thanks, I will modify it.
>
> > +/**
> > + * This function destroys a GRO table.
> > + */
> > +void rte_gro_tbl_destroy(void *tbl);
> > +
> > +/**
> > + * This is one of the main reassembly APIs, which merges numbers of
> > + * packets at a time. It assumes that all inputted packets are with
> > + * correct checksums. That is, applications should guarantee all
> > + * inputted packets are correct. Besides, it doesn't re-calculate
> > + * checksums for merged packets. If inputted packets are IP fragmented,
> > + * this function assumes them are complete (i.e. with L4 header). After
> > + * finishing processing, it returns all GROed packets to applications
> > + * immediately.
> > + *
> > + * @param pkts
> > + * a pointer array which points to the packets to reassemble. Besides,
> > + * it keeps packet addresses for GROed packets.
> > + * @param nb_pkts
> > + * the number of packets to reassemble.
> > + * @param param
> > + * applications use it to tell rte_gro_reassemble_burst what rules
> > + * are demanded.
> > + * @return
> > + * the number of packets after been GROed.
> > + */
> > +uint16_t rte_gro_reassemble_burst(struct rte_mbuf **pkts,
> > + uint16_t nb_pkts,
> > + const struct rte_gro_param *param);
>
> Fix the alignment.
>
> > +
> > +/**
> > + * Reassembly function, which tries to merge inputted packets with
> > + * the packets in a given GRO table. This function assumes all inputted
> > + * packet is with correct checksums. And it won't update checksums if
> > + * two packets are merged. Besides, if inputted packets are IP
> > + * fragmented, this function assumes they are complete packets (i.e.
> > + * with L4 header).
> > + *
> > + * If the inputted packets don't have data or are with unsupported GRO
> > + * types, they won't be processed and are returned to applications.
> > + * Otherwise, the inputted packets are either merged or inserted into
> > + * the table. If applications want get packets in the table, they need
> > + * to call flush API.
> > + *
> > + * @param pkts
> > + * packet to reassemble. Besides, after this function finishes, it
> > + * keeps the unprocessed packets (i.e. without data or unsupported
> > + * GRO types).
> > + * @param nb_pkts
> > + * the number of packets to reassemble.
> > + * @param tbl
> > + * a pointer points to a GRO table.
> > + * @return
> > + * return the number of unprocessed packets (i.e. without data or
> > + * unsupported GRO types). If all packets are processed (merged or
> > + * inserted into the table), return 0.
> > + */
> > +uint16_t rte_gro_reassemble(struct rte_mbuf **pkts,
> > + uint16_t nb_pkts,
> > + void *tbl);
> > +
> > +/**
> > + * This function flushes the timeout packets from reassembly tables of
> > + * desired GRO types. The max number of flushed timeout packets is the
> > + * element number of the array which is used to keep the flushed packets.
> > + *
> > + * Besides, this function won't re-calculate checksums for merged
> > + * packets in the tables. That is, the returned packets may be with
> > + * wrong checksums.
> > + *
> > + * @param tbl
> > + * a pointer points to a GRO table object.
> > + * @param desired_gro_types
> > + * rte_gro_timeout_flush only processes packets which belong to the
> > + * GRO types specified by desired_gro_types.
> > + * @param out
> > + * a pointer array that is used to keep flushed timeout packets.
> > + * @param nb_out
> > + * the element number of out. It's also the max number of timeout
> > + * packets that can be flushed finally.
> > + * @return
> > + * the number of flushed packets. If no packets are flushed, return 0.
> > + */
> > +uint16_t rte_gro_timeout_flush(void *tbl,
> > + uint64_t desired_gro_types,
> > + struct rte_mbuf **out,
> > + uint16_t max_nb_out);
> > +
> > +/**
> > + * This function returns the number of packets in a given GRO table.
> > + * @param tbl
> > + * pointer points to a GRO table.
> > + * @return
> > + * the number of packets in the table.
> > + */
> > +uint64_t rte_gro_tbl_item_num(void *tbl);
> > +
> > +#ifdef __cplusplus
> > +}
> > +#endif
> > +
> > +#endif
> > diff --git a/lib/librte_gro/rte_gro_version.map
> b/lib/librte_gro/rte_gro_version.map
> > new file mode 100644
> > index 0000000..358fb9d
> > --- /dev/null
> > +++ b/lib/librte_gro/rte_gro_version.map
> > @@ -0,0 +1,12 @@
> > +DPDK_17.08 {
> > + global:
> > +
> > + rte_gro_tbl_create;
> > + rte_gro_tbl_destroy;
>
> As stated earlier, here are the API names I suggested:
> rte_gro_ctl_create()/rte_gro_ctl_destroy()/rte_gro_get_count().
Thanks, I will change the names.
>
> > + rte_gro_reassemble_burst;
> > + rte_gro_reassemble;
> > + rte_gro_timeout_flush;
> > + rte_gro_tbl_item_num;
> > +
> > + local: *;
> > +};
> > diff --git a/mk/rte.app.mk b/mk/rte.app.mk
> > index bcaf1b3..fc3776d 100644
> > --- a/mk/rte.app.mk
> > +++ b/mk/rte.app.mk
> > @@ -98,6 +98,7 @@ _LDLIBS-$(CONFIG_RTE_LIBRTE_RING) += -
> lrte_ring
> > _LDLIBS-$(CONFIG_RTE_LIBRTE_EAL) += -lrte_eal
> > _LDLIBS-$(CONFIG_RTE_LIBRTE_CMDLINE) += -lrte_cmdline
> > _LDLIBS-$(CONFIG_RTE_LIBRTE_REORDER) += -lrte_reorder
> > +_LDLIBS-$(CONFIG_RTE_LIBRTE_GRO) += -lrte_gro
> >
> > ifeq ($(CONFIG_RTE_EXEC_ENV_LINUXAPP),y)
> > _LDLIBS-$(CONFIG_RTE_LIBRTE_KNI) += -lrte_kni
More information about the dev
mailing list