From: "Yan, Zhirun" <zhirun.yan@intel.com>
To: Vamsi Attunuru <vattunuru@marvell.com>,
"dev@dpdk.org" <dev@dpdk.org>,
"thomas@monjalon.net" <thomas@monjalon.net>,
"jerinj@marvell.com" <jerinj@marvell.com>
Cc: "ndabilpuram@marvell.com" <ndabilpuram@marvell.com>,
"Liang, Cunming" <cunming.liang@intel.com>,
"Wang, Haiyue" <haiyue.wang@intel.com>,
"Sunil Kumar Kori" <skori@marvell.com>
Subject: RE: [PATCH v2 4/4] app: add testgraph application
Date: Fri, 12 May 2023 11:08:07 +0000 [thread overview]
Message-ID: <SN7PR11MB6775BFEE43A4AE9A22CC30E885759@SN7PR11MB6775.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20230425131516.3308612-5-vattunuru@marvell.com>
> -----Original Message-----
> From: Vamsi Attunuru <vattunuru@marvell.com>
> Sent: Tuesday, April 25, 2023 9:15 PM
> To: dev@dpdk.org; thomas@monjalon.net; jerinj@marvell.com
> Cc: vattunuru@marvell.com; ndabilpuram@marvell.com
> Subject: [PATCH v2 4/4] app: add testgraph application
>
> Patch adds test-graph application to validate graph
> and node libraries.
>
> Signed-off-by: Vamsi Attunuru <vattunuru@marvell.com>
> ---
> app/meson.build | 1 +
> app/test-graph/cmdline.c | 211 +++++
> app/test-graph/cmdline_graph.c | 294 +++++++
> app/test-graph/cmdline_graph.h | 19 +
> app/test-graph/meson.build | 14 +
> app/test-graph/parameters.c | 157 ++++
> app/test-graph/testgraph.c | 1426 ++++++++++++++++++++++++++++++++
> app/test-graph/testgraph.h | 91 ++
> doc/guides/tools/index.rst | 1 +
> doc/guides/tools/testgraph.rst | 131 +++
> 10 files changed, 2345 insertions(+)
>
> diff --git a/app/meson.build b/app/meson.build
> index 74d2420f67..6c7b24e604 100644
> --- a/app/meson.build
> +++ b/app/meson.build
> @@ -22,6 +22,7 @@ apps = [
> 'test-eventdev',
> 'test-fib',
> 'test-flow-perf',
> + 'test-graph',
> 'test-gpudev',
> 'test-mldev',
> 'test-pipeline',
> diff --git a/app/test-graph/cmdline.c b/app/test-graph/cmdline.c
> new file mode 100644
> index 0000000000..d9474d827a
> --- /dev/null
> +++ b/app/test-graph/cmdline.c
> @@ -0,0 +1,211 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#include <stdlib.h>
> +
> +#include <cmdline.h>
> +#include <cmdline_parse.h>
> +#include <cmdline_parse_etheraddr.h>
> +#include <cmdline_parse_ipaddr.h>
> +#include <cmdline_parse_num.h>
> +#include <cmdline_parse_string.h>
> +#include <cmdline_rdline.h>
> +#include <cmdline_socket.h>
> +
> +#include "cmdline_graph.h"
> +#include "testgraph.h"
> +
> +static struct cmdline *testgraph_cl;
> +static cmdline_parse_ctx_t *main_ctx;
> +
> +/* *** Help command with introduction. *** */
> +struct cmd_help_brief_result {
> + cmdline_fixed_string_t help;
> +};
> +
> +static void
> +cmd_help_brief_parsed(__rte_unused void *parsed_result, struct cmdline *cl,
> __rte_unused void *data)
> +{
> + cmdline_printf(cl,
> + "\n"
> + "Help is available for the following sections:\n\n"
> + " help control : Start and stop graph walk.\n"
> + " help display : Displaying port, stats and config
> "
> + "information.\n"
> + " help config : Configuration information.\n"
> + " help all : All of the above sections.\n\n");
> +}
> +
> +static cmdline_parse_token_string_t cmd_help_brief_help =
> + TOKEN_STRING_INITIALIZER(struct cmd_help_brief_result, help, "help");
> +
> +static cmdline_parse_inst_t cmd_help_brief = {
> + .f = cmd_help_brief_parsed,
> + .data = NULL,
> + .help_str = "help: Show help",
> + .tokens = {
> + (void *)&cmd_help_brief_help,
> + NULL,
> + },
> +};
> +
> +/* *** Help command with help sections. *** */
> +struct cmd_help_long_result {
> + cmdline_fixed_string_t help;
> + cmdline_fixed_string_t section;
> +};
> +
> +static void
> +cmd_help_long_parsed(void *parsed_result, struct cmdline *cl, __rte_unused
> void *data)
> +{
> + int show_all = 0;
> + struct cmd_help_long_result *res = parsed_result;
> +
> + if (!strcmp(res->section, "all"))
> + show_all = 1;
> +
> + if (show_all || !strcmp(res->section, "control")) {
> +
> + cmdline_printf(cl, "\n"
> + "Control forwarding:\n"
> + "-------------------\n\n"
> +
> + "start graph_walk\n"
> + " Start graph_walk on worker threads.\n\n"
> +
> + "stop graph_walk\n"
> + " Stop worker threads from running
> graph_walk.\n\n"
> +
> + "quit\n"
> + " Quit to prompt.\n\n");
> + }
> +
> + if (show_all || !strcmp(res->section, "display")) {
> +
> + cmdline_printf(cl,
> + "\n"
> + "Display:\n"
> + "--------\n\n"
> +
> + "show node_list\n"
> + " Display the list of supported nodes.\n\n"
> +
> + "show graph_stats\n"
> + " Display the node statistics of graph cluster.\n\n");
> + }
> +
> + if (show_all || !strcmp(res->section, "config")) {
> + cmdline_printf(cl, "\n"
> + "Configuration:\n"
> + "--------------\n"
> + "set lcore_config
> (port_id0,rxq0,lcore_idX),..."
> + ".....,(port_idX,rxqX,lcoreidY)\n"
> + " Set lcore configuration.\n\n"
> +
> + "create_graph
> (node0_name,node1_name,...,nodeX_name)\n"
> + " Create graph instances using the provided
> node details.\n\n"
> +
> + "destroy_graph\n"
> + " Destroy the graph instances.\n\n");
> + }
> +}
> +
> +static cmdline_parse_token_string_t cmd_help_long_help =
> + TOKEN_STRING_INITIALIZER(struct cmd_help_long_result, help, "help");
> +
> +static cmdline_parse_token_string_t cmd_help_long_section =
> TOKEN_STRING_INITIALIZER(
> + struct cmd_help_long_result, section, "all#control#display#config");
> +
> +static cmdline_parse_inst_t cmd_help_long = {
> + .f = cmd_help_long_parsed,
> + .data = NULL,
> + .help_str = "help all|control|display|config: "
> + "Show help",
> + .tokens = {
> + (void *)&cmd_help_long_help,
> + (void *)&cmd_help_long_section,
> + NULL,
> + },
> +};
> +
> +/* *** QUIT *** */
> +struct cmd_quit_result {
> + cmdline_fixed_string_t quit;
> +};
> +
> +static void
> +cmd_quit_parsed(__rte_unused void *parsed_result, struct cmdline *cl,
> __rte_unused void *data)
> +{
> + cmdline_quit(cl);
> +}
> +
> +static cmdline_parse_token_string_t cmd_quit_quit =
> + TOKEN_STRING_INITIALIZER(struct cmd_quit_result, quit, "quit");
> +
> +static cmdline_parse_inst_t cmd_quit = {
> + .f = cmd_quit_parsed,
> + .data = NULL,
> + .help_str = "quit: Exit application",
> + .tokens = {
> + (void *)&cmd_quit_quit,
> + NULL,
> + },
> +};
> +
> +/* list of instructions */
> +static cmdline_parse_ctx_t builtin_ctx[] = {
> + (cmdline_parse_inst_t *)&cmd_help_brief,
> + (cmdline_parse_inst_t *)&cmd_help_long,
> + (cmdline_parse_inst_t *)&cmd_quit,
> + (cmdline_parse_inst_t *)&cmd_show_node_list,
> + (cmdline_parse_inst_t *)&cmd_set_lcore_config,
> + (cmdline_parse_inst_t *)&cmd_create_graph,
> + (cmdline_parse_inst_t *)&cmd_destroy_graph,
> + (cmdline_parse_inst_t *)&cmd_start_graph_walk,
> + (cmdline_parse_inst_t *)&cmd_stop_graph_walk,
> + (cmdline_parse_inst_t *)&cmd_show_graph_stats,
> + NULL,
> +};
> +
> +int
> +init_cmdline(void)
> +{
> + unsigned int count;
> + unsigned int i;
> +
> + count = 0;
> + for (i = 0; builtin_ctx[i] != NULL; i++)
> + count++;
> +
> + /* cmdline expects a NULL terminated array */
> + main_ctx = calloc(count + 1, sizeof(main_ctx[0]));
> + if (main_ctx == NULL)
> + return -1;
> +
> + count = 0;
> + for (i = 0; builtin_ctx[i] != NULL; i++, count++)
> + main_ctx[count] = builtin_ctx[i];
> +
> + return 0;
> +}
> +
> +void
> +prompt_exit(void)
> +{
> + cmdline_quit(testgraph_cl);
> +}
> +
> +/* prompt function, called from main on MAIN lcore */
> +void
> +prompt(void)
> +{
> + testgraph_cl = cmdline_stdin_new(main_ctx, "testgraph> ");
> + if (testgraph_cl == NULL) {
> + fprintf(stderr, "Failed to create stdin based cmdline context\n");
> + return;
> + }
> +
> + cmdline_interact(testgraph_cl);
> + cmdline_stdin_exit(testgraph_cl);
> +}
> diff --git a/app/test-graph/cmdline_graph.c b/app/test-graph/cmdline_graph.c
> new file mode 100644
> index 0000000000..d66149a224
> --- /dev/null
> +++ b/app/test-graph/cmdline_graph.c
> @@ -0,0 +1,294 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#include <stdlib.h>
> +#include <string.h>
> +
> +#include <cmdline_parse.h>
> +#include <cmdline_parse_num.h>
> +#include <cmdline_parse_string.h>
> +
> +#include "cmdline_graph.h"
> +#include "testgraph.h"
> +
> +/* *** Show supported node details *** */
> +struct cmd_show_node_list_result {
> + cmdline_fixed_string_t show;
> + cmdline_fixed_string_t node_list;
> +};
> +
> +static cmdline_parse_token_string_t cmd_show_node_list_show =
> + TOKEN_STRING_INITIALIZER(struct cmd_show_node_list_result, show,
> "show");
> +static cmdline_parse_token_string_t cmd_show_node_list_node_list =
> + TOKEN_STRING_INITIALIZER(struct cmd_show_node_list_result,
> node_list, "node_list");
> +
> +static void
> +cmd_show_node_list_parsed(__rte_unused void *parsed_result, __rte_unused
> struct cmdline *cl,
> + __rte_unused void *data)
> +{
> + rte_node_t node_cnt = rte_node_max_count();
> + rte_node_t id;
> +
> + printf("\n**** Supported Graph Nodes ****\n");
> + for (id = 0; id < node_cnt; id++)
> + printf("%s\n", rte_node_id_to_name(id));
> +
> + printf("********************************\n");
> +}
> +
> +cmdline_parse_inst_t cmd_show_node_list = {
> + .f = cmd_show_node_list_parsed,
> + .data = NULL,
> + .help_str = "show node_list",
> + .tokens = {
> + (void *)&cmd_show_node_list_show,
> + (void *)&cmd_show_node_list_node_list,
> + NULL,
> + },
> +};
> +
> +/* *** Set lcore config *** */
> +struct cmd_set_lcore_config_result {
> + cmdline_fixed_string_t set;
> + cmdline_fixed_string_t lcore_config;
> + cmdline_multi_string_t token_string;
> +};
> +
> +static cmdline_parse_token_string_t cmd_set_lcore_config_set =
> + TOKEN_STRING_INITIALIZER(struct cmd_set_lcore_config_result, set,
> "set");
> +static cmdline_parse_token_string_t cmd_set_lcore_config_lcore_config =
> + TOKEN_STRING_INITIALIZER(struct cmd_set_lcore_config_result,
> lcore_config, "lcore_config");
> +static cmdline_parse_token_string_t cmd_set_lcore_config_token_string =
> TOKEN_STRING_INITIALIZER(
> + struct cmd_set_lcore_config_result, token_string,
> TOKEN_STRING_MULTI);
> +
> +static void
> +cmd_set_lcore_config_parsed(void *parsed_result, __rte_unused struct
> cmdline *cl,
> + __rte_unused void *data)
> +{
> + struct cmd_set_lcore_config_result *res = parsed_result;
> + const char *t_str = res->token_string;
> + int ret;
> +
> + /* Parse string */
> + ret = parse_config(t_str);
> + if (ret) {
> + printf(" lcore_config string parse error\n");
> + return;
> + }
> +
> + validate_config();
> +}
> +
> +cmdline_parse_inst_t cmd_set_lcore_config = {
> + .f = cmd_set_lcore_config_parsed,
> + .data = NULL,
> + .help_str = "set lcore_config "
> + "(port,queue,lcore),[(port,queue,lcore) ...
> (port,queue,lcore)]",
> + .tokens = {
> + (void *)&cmd_set_lcore_config_set,
> + (void *)&cmd_set_lcore_config_lcore_config,
> + (void *)&cmd_set_lcore_config_token_string,
> + NULL,
> + },
> +};
> +
> +/* *** Create graph *** */
> +struct cmd_create_graph_result {
> + cmdline_fixed_string_t create_graph;
> + cmdline_multi_string_t token_string;
> +};
> +
> +static cmdline_parse_token_string_t cmd_create_graph_create_graph =
> + TOKEN_STRING_INITIALIZER(struct cmd_create_graph_result,
> create_graph, "create_graph");
> +static cmdline_parse_token_string_t cmd_create_graph_token_string =
> + TOKEN_STRING_INITIALIZER(struct cmd_create_graph_result,
> token_string, TOKEN_STRING_MULTI);
> +
> +static void
> +cmd_create_graph_parsed(void *parsed_result, __rte_unused struct cmdline
> *cl,
> + __rte_unused void *data)
> +{
> + struct cmd_create_graph_result *res = parsed_result;
> + const char *t_str = res->token_string;
> + uint64_t valid_nodes = 0;
> + int ret;
> +
> + ret = parse_node_patterns(t_str);
> + if (ret) {
> + printf("parse_node_patterns failed\n");
> + cleanup_node_pattern();
> + return;
> + }
> +
> + ret = validate_node_names(&valid_nodes);
> + if (ret) {
> + printf("validate_node_names() failed\n");
> + cleanup_node_pattern();
> + return;
> + }
Hi Vamsi,
The global node_pattern should be cleaned in destroy stage or make sure it populated from cmd line completely.
In Interactive-mode, user may create and destroy a graph and then create a new one.
For this case, the node name is changed by the previous creation, especially for rx/tx node.(ethdev_rx -> ethdev_rx-0-0)
> +
> + nb_conf = ethdev_ports_setup();
> +
> + ethdev_rxq_configure();
> + ethdev_txq_configure();
> +
> + ret = configure_graph_nodes(valid_nodes);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "configure_graph_nodes: err=%d\n",
> ret);
> +
> + ret = create_graph(valid_nodes);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "create_graph: err=%d\n", ret);
> +
> + stats = create_graph_cluster_stats();
> + if (stats == NULL)
> + rte_exit(EXIT_FAILURE, "create_graph_cluster_stats() failed\n");
> +
> + check_all_ports_link_status(enabled_port_mask);
> + start_eth_ports();
> +}
Could this create cmd be break down for graph_config and graph_create?
In graph_config stage, we can set/verify graph topo, chose graph-walk models and update node ctx, etc.
> +
> +cmdline_parse_inst_t cmd_create_graph = {
> + .f = cmd_create_graph_parsed,
> + .data = NULL,
> + .help_str = "create_graph "
> + "[node_name0,node_name1,node_name2 ... node_nameX]",
> + .tokens = {
> + (void *)&cmd_create_graph_create_graph,
> + (void *)&cmd_create_graph_token_string,
> + NULL,
> + },
> +};
> +
> +/**** Destroy graph ****/
> +struct cmd_destroy_graph_result {
> + cmdline_fixed_string_t destroy_graph;
> +};
> +
> +static cmdline_parse_token_string_t cmd_destroy_graph_destroy_graph =
> + TOKEN_STRING_INITIALIZER(struct cmd_destroy_graph_result,
> destroy_graph, "destroy_graph");
> +
> +static void
> +cmd_destroy_graph_parsed(__rte_unused void *parsed_result, __rte_unused
> struct cmdline *cl,
> + __rte_unused void *data)
> +{
> + uint32_t lcore_id;
> +
> + run_graph_walk = false;
> + graph_walk_quit = true;
> +
> + RTE_LCORE_FOREACH_WORKER(lcore_id)
> + rte_eal_wait_lcore(lcore_id);
> +
> + destroy_graph();
> + stop_eth_ports();
> +}
> +
> +cmdline_parse_inst_t cmd_destroy_graph = {
> + .f = cmd_destroy_graph_parsed,
> + .data = NULL,
> + .help_str = "destroy_graph",
> + .tokens = {
> + (void *)&cmd_destroy_graph_destroy_graph,
> + NULL,
> + },
> +};
> +
> +/**** Start graph_walk ****/
> +struct cmd_start_graph_walk_result {
> + cmdline_fixed_string_t start;
> + cmdline_fixed_string_t graph_walk;
> +};
> +
> +static cmdline_parse_token_string_t cmd_start_graph_walk_start =
> + TOKEN_STRING_INITIALIZER(struct cmd_start_graph_walk_result, start,
> "start");
> +static cmdline_parse_token_string_t cmd_start_graph_walk_graph_walk =
> + TOKEN_STRING_INITIALIZER(struct cmd_start_graph_walk_result,
> graph_walk, "graph_walk");
> +
> +static void
> +cmd_start_graph_walk_parsed(__rte_unused void *parsed_result,
> __rte_unused struct cmdline *cl,
> + __rte_unused void *data)
> +{
> + static bool launch_graph_walk;
> +
> + if (!launch_graph_walk) {
> + rte_eal_mp_remote_launch(graph_main_loop, NULL,
> SKIP_MAIN);
As run_graph_walk is used to control graph-walk, Why not put the launch in create stage?
> + launch_graph_walk = true;
> + }
> +
> + run_graph_walk = true;
> +}
> +
> +cmdline_parse_inst_t cmd_start_graph_walk = {
> + .f = cmd_start_graph_walk_parsed,
> + .data = NULL,
> + .help_str = "start graph_walk",
> + .tokens = {
> + (void *)&cmd_start_graph_walk_start,
> + (void *)&cmd_start_graph_walk_graph_walk,
> + NULL,
> + },
> +};
> +
> +/**** Stop graph_walk ****/
> +struct cmd_stop_graph_walk_result {
> + cmdline_fixed_string_t stop;
> + cmdline_fixed_string_t graph_walk;
> +};
> +
> +static cmdline_parse_token_string_t cmd_stop_graph_walk_stop =
> + TOKEN_STRING_INITIALIZER(struct cmd_stop_graph_walk_result, stop,
> "stop");
> +static cmdline_parse_token_string_t cmd_stop_graph_walk_graph_walk =
> + TOKEN_STRING_INITIALIZER(struct cmd_stop_graph_walk_result,
> graph_walk, "graph_walk");
> +
> +static void
> +cmd_stop_graph_walk_parsed(__rte_unused void *parsed_result,
> __rte_unused struct cmdline *cl,
> + __rte_unused void *data)
> +{
> + run_graph_walk = false;
> +}
> +
> +cmdline_parse_inst_t cmd_stop_graph_walk = {
> + .f = cmd_stop_graph_walk_parsed,
> + .data = NULL,
> + .help_str = "stop graph_walk",
> + .tokens = {
> + (void *)&cmd_stop_graph_walk_stop,
> + (void *)&cmd_stop_graph_walk_graph_walk,
> + NULL,
> + },
> +};
> +
> +/**** Show graph_stats ****/
> +struct cmd_show_graph_stats_result {
> + cmdline_fixed_string_t show;
> + cmdline_fixed_string_t graph_stats;
> +};
> +
> +static cmdline_parse_token_string_t cmd_show_graph_stats_show =
> + TOKEN_STRING_INITIALIZER(struct cmd_show_graph_stats_result,
> show, "show");
> +static cmdline_parse_token_string_t cmd_show_graph_stats_graph_stats =
> + TOKEN_STRING_INITIALIZER(struct cmd_show_graph_stats_result,
> graph_stats, "graph_stats");
> +
> +static void
> +cmd_show_graph_stats_parsed(__rte_unused void *parsed_result,
> __rte_unused struct cmdline *cl,
> + __rte_unused void *data)
> +{
> + if (rte_graph_has_stats_feature()) {
> + if (stats)
> + rte_graph_cluster_stats_get(stats, 0);
> + } else {
> + printf(" graph stats feature not enabled in rte_config.\n");
> + }
> +}
> +
> +cmdline_parse_inst_t cmd_show_graph_stats = {
> + .f = cmd_show_graph_stats_parsed,
> + .data = NULL,
> + .help_str = "show graph_stats",
> + .tokens = {
> + (void *)&cmd_show_graph_stats_show,
> + (void *)&cmd_show_graph_stats_graph_stats,
> + NULL,
> + },
> +};
> diff --git a/app/test-graph/cmdline_graph.h b/app/test-graph/cmdline_graph.h
> new file mode 100644
> index 0000000000..2846ff5425
> --- /dev/null
> +++ b/app/test-graph/cmdline_graph.h
> @@ -0,0 +1,19 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#ifndef _CMDLINE_GRAPH_H_
> +#define _CMDLINE_GRAPH_H_
> +
> +extern cmdline_parse_inst_t cmd_show_node_list;
> +extern cmdline_parse_inst_t cmd_set_lcore_config;
> +
> +extern cmdline_parse_inst_t cmd_create_graph;
> +extern cmdline_parse_inst_t cmd_destroy_graph;
> +
> +extern cmdline_parse_inst_t cmd_start_graph_walk;
> +extern cmdline_parse_inst_t cmd_stop_graph_walk;
> +
> +extern cmdline_parse_inst_t cmd_show_graph_stats;
> +
> +#endif /* _CMDLINE_GRAPH_H_ */
> diff --git a/app/test-graph/meson.build b/app/test-graph/meson.build
> new file mode 100644
> index 0000000000..d93802a975
> --- /dev/null
> +++ b/app/test-graph/meson.build
> @@ -0,0 +1,14 @@
> +# SPDX-License-Identifier: BSD-3-Clause
> +# Copyright(C) 2023 Marvell International Ltd.
> +
> +# override default name to drop the hyphen
> +name = 'test-graph'
> +cflags += '-Wno-deprecated-declarations'
> +sources = files(
> + 'cmdline.c',
> + 'cmdline_graph.c',
> + 'parameters.c',
> + 'testgraph.c',
> +)
> +
> +deps += ['ethdev', 'cmdline', 'graph', 'node', 'eal']
> diff --git a/app/test-graph/parameters.c b/app/test-graph/parameters.c
> new file mode 100644
> index 0000000000..b990ca4a1c
> --- /dev/null
> +++ b/app/test-graph/parameters.c
> @@ -0,0 +1,157 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#include <getopt.h>
> +#include <stdlib.h>
> +
> +#include "testgraph.h"
> +
> +static const char short_options[] = "p:" /* portmask */
> + "P" /* promiscuous */
> + "i" /* interactive */
> + ;
> +
> +#define CMD_LINE_OPT_CONFIG "config"
> +#define CMD_LINE_OPT_NODE_PATTERN "node-pattern"
> +#define CMD_LINE_OPT_INTERACTIVE "interactive"
> +#define CMD_LINE_OPT_NO_NUMA "no-numa"
> +#define CMD_LINE_OPT_PER_PORT_POOL "per-port-pool"
> +enum {
> + /* Long options mapped to a short option */
> +
> + /* First long only option value must be >= 256, so that we won't
> + * conflict with short options
> + */
> + CMD_LINE_OPT_MIN_NUM = 256,
> + CMD_LINE_OPT_CONFIG_NUM,
> + CMD_LINE_OPT_NODE_PATTERN_NUM,
> + CMD_LINE_OPT_INTERACTIVE_NUM,
> + CMD_LINE_OPT_NO_NUMA_NUM,
> + CMD_LINE_OPT_PARSE_PER_PORT_POOL,
> +};
> +
> +static const struct option lgopts[] = {
> + {CMD_LINE_OPT_CONFIG, 1, 0, CMD_LINE_OPT_CONFIG_NUM},
> + {CMD_LINE_OPT_NODE_PATTERN, 1, 0,
> CMD_LINE_OPT_NODE_PATTERN_NUM},
> + {CMD_LINE_OPT_INTERACTIVE, 0, 0,
> CMD_LINE_OPT_INTERACTIVE_NUM},
> + {CMD_LINE_OPT_NO_NUMA, 0, 0, CMD_LINE_OPT_NO_NUMA_NUM},
> + {CMD_LINE_OPT_PER_PORT_POOL, 0, 0,
> CMD_LINE_OPT_PARSE_PER_PORT_POOL},
> + {NULL, 0, 0, 0},
> +};
> +
> +/* Display usage */
> +static void
> +print_usage(const char *prgname)
> +{
> + fprintf(stderr,
> + "%s [EAL options] --"
> + " -p PORTMASK"
> + " [-P]"
> + " [-i]"
> + " --config (port,queue,lcore)[,(port,queue,lcore)]"
> + " --node-pattern (node_name0,node_name1[,node_nameX)]"
> + " [--no-numa]"
> + " [--per-port-pool]"
> + " [--interactive]"
> +
> + " -p PORTMASK: Hexadecimal bitmask of ports to configure\n"
> + " -P : Enable promiscuous mode\n"
> + " -i : Enter interactive mode\n"
> + " --config (port,queue,lcore): Rx queue configuration\n"
> + " --node-pattern (node_names): node patterns to create
> graph\n"
> + " --no-numa: Disable numa awareness\n"
> + " --per-port-pool: Use separate buffer pool per port\n"
> + " --interactive: Enter interactive mode\n",
> + prgname);
> +}
> +
> +static int
> +parse_portmask(const char *portmask)
> +{
> + char *end = NULL;
> + unsigned long pm;
> +
> + /* Parse hexadecimal string */
> + pm = strtoul(portmask, &end, 16);
> + if ((portmask[0] == '\0') || (end == NULL) || (*end != '\0'))
> + return 0;
> +
> + return pm;
> +}
> +
> +/* Parse the argument given in the command line of the application */
> +int
> +parse_cmdline_args(int argc, char **argv)
> +{
> + char *prgname = argv[0];
> + int option_index;
> + char **argvopt;
> + int opt, ret;
> +
> + argvopt = argv;
> +
> + /* Error or normal output strings. */
> + while ((opt = getopt_long(argc, argvopt, short_options, lgopts,
> &option_index)) != EOF) {
> +
> + switch (opt) {
> + /* Portmask */
> + case 'p':
> + enabled_port_mask = parse_portmask(optarg);
> + if (enabled_port_mask == 0) {
> + fprintf(stderr, "Invalid portmask\n");
> + print_usage(prgname);
> + return -1;
> + }
> + break;
> +
> + case 'P':
> + promiscuous_on = 1;
> + break;
> +
> + /* Long options */
> + case CMD_LINE_OPT_CONFIG_NUM:
> + ret = parse_config(optarg);
> + if (ret) {
> + fprintf(stderr, "Invalid config\n");
> + print_usage(prgname);
> + return -1;
> + }
> + break;
> + case CMD_LINE_OPT_NODE_PATTERN_NUM:
> + ret = parse_node_patterns(optarg);
> + if (ret) {
> + fprintf(stderr, "Invalid node_patterns\n");
> + print_usage(prgname);
> + return -1;
> + }
> + break;
> +
> + case CMD_LINE_OPT_INTERACTIVE_NUM:
> + case 'i':
> + printf("Interactive-mode selected\n");
> + interactive = 1;
> + break;
> +
> + case CMD_LINE_OPT_NO_NUMA_NUM:
> + numa_on = 0;
> + break;
> +
> + case CMD_LINE_OPT_PARSE_PER_PORT_POOL:
> + printf("Per port buffer pool is enabled\n");
> + per_port_pool = 1;
> + break;
> +
> + default:
> + print_usage(prgname);
> + return -1;
> + }
> + }
> +
> + if (optind >= 0)
> + argv[optind - 1] = prgname;
> + ret = optind - 1;
> + optind = 1; /* Reset getopt lib */
> +
> + return ret;
> +}
> diff --git a/app/test-graph/testgraph.c b/app/test-graph/testgraph.c
> new file mode 100644
> index 0000000000..aff921acf2
> --- /dev/null
> +++ b/app/test-graph/testgraph.c
> @@ -0,0 +1,1426 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#include <fnmatch.h>
> +#include <signal.h>
> +#include <stdlib.h>
> +
> +#include <rte_bus.h>
> +#include <rte_byteorder.h>
> +#include <rte_common.h>
> +#include <rte_dev.h>
> +#include <rte_eal.h>
> +#include <rte_ethdev.h>
> +#include <rte_ether.h>
> +#include <rte_graph_worker.h>
> +#include <rte_launch.h>
> +#include <rte_lcore.h>
> +#include <rte_log.h>
> +#include <rte_malloc.h>
> +#include <rte_mempool.h>
> +#include <rte_per_lcore.h>
> +#include <ethdev_rx_priv.h>
> +#include <ethdev_tx_priv.h>
> +#include <punt_kernel_priv.h>
> +#include <kernel_recv_priv.h>
> +
> +#include "testgraph.h"
> +
> +/* Log type */
> +#define RTE_LOGTYPE_TEST_GRAPH RTE_LOGTYPE_USER1
> +
> +/*
> + * Configurable number of RX/TX ring descriptors
> + */
> +#define RX_DESC_DEFAULT 1024
> +#define TX_DESC_DEFAULT 1024
> +
> +#define MAX_TX_QUEUE_PER_PORT RTE_MAX_ETHPORTS
> +#define MAX_RX_QUEUE_PER_PORT 128
> +
> +#define MAX_RX_QUEUE_PER_LCORE 16
> +
> +#define NB_SOCKETS 8
> +
> +/* Static global variables used within this file. */
> +uint16_t nb_rxd = RX_DESC_DEFAULT;
> +uint16_t nb_txd = TX_DESC_DEFAULT;
> +
> +static volatile bool force_quit;
> +volatile bool graph_walk_quit;
> +volatile bool run_graph_walk = true;
> +
> +char node_pattern[MAX_NODE_PATTERNS][RTE_NODE_NAMESIZE] = {0};
> +const char **node_patterns;
> +uint8_t num_patterns;
> +
> +uint8_t interactive; /**< interactive mode is off by default */
> +int promiscuous_on; /**< Ports set in promiscuous mode off by default. */
> +int numa_on = 1; /**< NUMA is enabled by default. */
> +int per_port_pool; /**< Use separate buffer pools per port, disabled by default
> */
> +int testgraph_logtype; /**< Log type for testgraph logs */
> +
> +struct rte_graph_cluster_stats *stats;
> +
> +uint32_t enabled_port_mask; /**< Mask of enabled ports */
> +
> +uint32_t nb_conf;
> +
> +struct lcore_rx_queue {
> + uint16_t port_id;
> + uint8_t queue_id;
> + char node_name[RTE_NODE_NAMESIZE];
> +};
> +
> +/* Lcore conf */
> +struct lcore_conf {
> + uint16_t n_rx_queue;
> + struct lcore_rx_queue rx_queue_list[MAX_RX_QUEUE_PER_LCORE];
> + char punt_kernel_node_name[RTE_NODE_NAMESIZE];
> + char kernel_recv_node_name[RTE_NODE_NAMESIZE];
> +
> + struct rte_graph *graph;
> + char name[RTE_GRAPH_NAMESIZE];
> + rte_graph_t graph_id;
> +} __rte_cache_aligned;
> +
> +static struct lcore_conf lcore_conf[RTE_MAX_LCORE];
> +
> +struct lcore_params lcore_params_array[MAX_LCORE_PARAMS];
> +static struct lcore_params lcore_params_array_default[] = {
> + {0, 0, 2}, {1, 0, 2}, {2, 0, 2}, {0, 1, 3}, {1, 1, 3},
> + {2, 1, 3}, {0, 2, 4}, {1, 2, 4}, {2, 2, 4},
> +};
> +
> +struct lcore_params *lcore_params = lcore_params_array_default;
> +uint16_t nb_lcore_params = RTE_DIM(lcore_params_array_default);
> +
> +static struct rte_eth_conf port_conf = {
> + .rxmode = {
> + .mq_mode = RTE_ETH_MQ_RX_RSS,
> + },
> + .rx_adv_conf = {
> + .rss_conf = {
> + .rss_key = NULL,
> + .rss_hf = RTE_ETH_RSS_IP,
> + },
> + },
> + .txmode = {
> + .mq_mode = RTE_ETH_MQ_TX_NONE,
> + },
> +};
> +
> +static const struct node_list test_node_list[] = {{{"ethdev_rx", "ethdev_tx"}, 0,
> 2},
> + {{"ethdev_rx", "punt_kernel"},
> 0, 2},
> + {{"kernel_recv", "ethdev_tx"},
> 0, 2} };
> +
Here we support 3 topo, could you support the original l3fwd nodes first?
> +static const struct node_list supported_nodes[] = {{{"ethdev_rx"},
> TEST_GRAPH_ETHDEV_RX_NODE, 1},
> + {{"ethdev_tx"},
> TEST_GRAPH_ETHDEV_TX_NODE, 1},
> + {{"punt_kernel"},
> TEST_GRAPH_PUNT_KERNEL_NODE, 1},
> + {{"kernel_recv"},
> TEST_GRAPH_KERNEL_RECV_NODE, 1},
> + {{"ip4_lookup"},
> TEST_GRAPH_IP4_LOOKUP_NODE, 1},
> + {{"ip4_rewrite"},
> TEST_GRAPH_IP4_REWRITE_NODE, 1},
> + {{"pkt_cls"},
> TEST_GRAPH_PKT_CLS_NODE, 1},
> + {{"pkt_drop"},
> TEST_GRAPH_PKT_DROP_NODE, 1},
> + {{"NULL"},
> TEST_GRAPH_NULL_NODE, 1} };
> +
> +static struct rte_mempool
> *pktmbuf_pool[RTE_MAX_ETHPORTS][NB_SOCKETS];
> +
> +struct rte_node_ethdev_config ethdev_conf[RTE_MAX_ETHPORTS];
> +
> +static int
> +check_lcore_params(void)
> +{
> + uint8_t queue, lcore;
> + int socketid;
> + uint16_t i;
> +
> + for (i = 0; i < nb_lcore_params; ++i) {
> + queue = lcore_params[i].queue_id;
> + if (queue >= MAX_RX_QUEUE_PER_PORT) {
> + printf("Invalid queue number: %hhu\n", queue);
> + return -1;
> + }
> + lcore = lcore_params[i].lcore_id;
> + if (!rte_lcore_is_enabled(lcore)) {
> + printf("Error: lcore %hhu is not enabled in lcore mask\n",
> lcore);
> + return -1;
> + }
> +
> + if (lcore == rte_get_main_lcore()) {
> + printf("Error: lcore %u is main lcore\n", lcore);
> + return -1;
> + }
> + socketid = rte_lcore_to_socket_id(lcore);
> + if ((socketid != 0) && (numa_on == 0)) {
> + printf("Warning: lcore %hhu is on socket %d with numa
> off\n", lcore,
> + socketid);
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int
> +check_port_config(void)
> +{
> + uint16_t portid;
> + uint16_t i;
> +
> + for (i = 0; i < nb_lcore_params; ++i) {
> + portid = lcore_params[i].port_id;
> + if ((enabled_port_mask & (1 << portid)) == 0) {
> + printf("Port %u is not enabled in port mask\n", portid);
> + return -1;
> + }
> + if (!rte_eth_dev_is_valid_port(portid)) {
> + printf("Port %u is not present on the board\n", portid);
> + return -1;
> + }
> + }
> +
> + return 0;
> +}
> +
> +static uint8_t
> +get_port_n_rx_queues(const uint16_t port)
> +{
> + int queue = -1;
> + uint16_t i;
> +
> + for (i = 0; i < nb_lcore_params; ++i) {
> + if (lcore_params[i].port_id == port) {
> + if (lcore_params[i].queue_id == queue + 1)
> + queue = lcore_params[i].queue_id;
> + else
> + rte_exit(EXIT_FAILURE, "Queue ids of the
> port %d must be"
> + " in sequence and must start with 0\n",
> + lcore_params[i].port_id);
> + }
> + }
> +
> + return (uint8_t)(++queue);
> +}
> +
> +static int
> +init_lcore_rx_queues(void)
> +{
> + uint16_t i, nb_rx_queue;
> + uint8_t lcore;
> +
> + for (i = 0; i < nb_lcore_params; ++i) {
> + lcore = lcore_params[i].lcore_id;
> + nb_rx_queue = lcore_conf[lcore].n_rx_queue;
> + if (nb_rx_queue >= MAX_RX_QUEUE_PER_LCORE) {
> + printf("Error: too many queues (%u) for lcore: %u\n",
> + (unsigned int)nb_rx_queue + 1, (unsigned int)lcore);
> + return -1;
> + }
> +
> + lcore_conf[lcore].rx_queue_list[nb_rx_queue].port_id =
> lcore_params[i].port_id;
> + lcore_conf[lcore].rx_queue_list[nb_rx_queue].queue_id =
> lcore_params[i].queue_id;
> + lcore_conf[lcore].n_rx_queue++;
> + }
> +
> + return 0;
> +}
> +
> +int
> +validate_config(void)
> +{
> + int rc = -1;
> +
> + if (check_lcore_params() < 0) {
> + printf("check_lcore_params() failed\n");
> + goto exit;
> + }
> +
> + if (init_lcore_rx_queues() < 0) {
> + printf("init_lcore_rx_queues() failed\n");
> + goto exit;
> + }
> +
> + if (check_port_config() < 0) {
> + printf("check_port_config() failed\n");
> + goto exit;
> + }
> +
> + return 0;
> +
> +exit:
> + return rc;
> +}
> +
> +#define MEMPOOL_CACHE_SIZE 256
> +
> +/*
> + * This expression is used to calculate the number of mbufs needed
> + * depending on user input, taking into account memory for rx and
> + * tx hardware rings, cache per lcore and mtable per port per lcore.
> + * RTE_MAX is used to ensure that NB_MBUF never goes below a minimum
> + * value of 8192
> + */
> +#define NB_MBUF(nports) \
> + RTE_MAX((nports * nb_rx_queue * nb_rxd + nports * nb_lcores *
> RTE_GRAPH_BURST_SIZE + \
> + nports * nb_tx_queue * nb_txd + nb_lcores *
> MEMPOOL_CACHE_SIZE), \
> + 8192u)
> +
> +static int
> +init_mem(uint16_t portid, uint32_t nb_mbuf)
> +{
> + uint32_t lcore_id;
> + int socketid;
> + char s[64];
> +
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> +
> + if (numa_on)
> + socketid = rte_lcore_to_socket_id(lcore_id);
> + else
> + socketid = 0;
> +
> + if (socketid >= NB_SOCKETS) {
> + rte_exit(EXIT_FAILURE, "Socket %d of lcore %u is out of
> range %d\n",
> + socketid, lcore_id, NB_SOCKETS);
> + }
> +
> + if (pktmbuf_pool[portid][socketid] == NULL) {
> + snprintf(s, sizeof(s), "mbuf_pool_%d:%d", portid,
> socketid);
> + /* Create a pool with priv size of a cacheline */
> + pktmbuf_pool[portid][socketid] =
> rte_pktmbuf_pool_create(
> + s, nb_mbuf, MEMPOOL_CACHE_SIZE,
> RTE_CACHE_LINE_SIZE,
> + RTE_MBUF_DEFAULT_BUF_SIZE, socketid);
> + if (pktmbuf_pool[portid][socketid] == NULL)
> + rte_exit(EXIT_FAILURE, "Cannot init mbuf pool
> on socket %d\n",
> + socketid);
> + else
> + printf("Allocated mbuf pool on socket %d\n",
> socketid);
> + }
> + }
> +
> + return 0;
> +}
> +
> +void
> +check_all_ports_link_status(uint32_t port_mask)
> +{
> +#define CHECK_INTERVAL 100 /* 100ms */
> +#define MAX_CHECK_TIME 90 /* 9s (90 * 100ms) in total */
> + uint8_t count, all_ports_up, print_flag = 0;
> + struct rte_eth_link link;
> + uint16_t portid;
> + int ret;
> + char link_status_text[RTE_ETH_LINK_MAX_STR_LEN];
> +
> + printf("\nChecking link status");
> + fflush(stdout);
> + for (count = 0; count <= MAX_CHECK_TIME; count++) {
> + if (force_quit)
> + return;
> + all_ports_up = 1;
> + RTE_ETH_FOREACH_DEV(portid) {
> + if (force_quit)
> + return;
> + if ((port_mask & (1 << portid)) == 0)
> + continue;
> + memset(&link, 0, sizeof(link));
> + ret = rte_eth_link_get_nowait(portid, &link);
> + if (ret < 0) {
> + all_ports_up = 0;
> + if (print_flag == 1)
> + printf("Port %u link get failed: %s\n",
> portid,
> + rte_strerror(-ret));
> + continue;
> + }
> + /* Print link status if flag set */
> + if (print_flag == 1) {
> + rte_eth_link_to_str(link_status_text,
> sizeof(link_status_text),
> + &link);
> + printf("Port %d %s\n", portid, link_status_text);
> + continue;
> + }
> + /* Clear all_ports_up flag if any link down */
> + if (link.link_status == RTE_ETH_LINK_DOWN) {
> + all_ports_up = 0;
> + break;
> + }
> + }
> + /* After finally printing all link status, get out */
> + if (print_flag == 1)
> + break;
> +
> + if (all_ports_up == 0) {
> + printf(".");
> + fflush(stdout);
> + rte_delay_ms(CHECK_INTERVAL);
> + }
> +
> + /* Set the print_flag if all ports up or timeout */
> + if (all_ports_up == 1 || count == (MAX_CHECK_TIME - 1)) {
> + print_flag = 1;
> + printf("Done\n");
> + }
> + }
> +}
> +
> +static void
> +signal_handler(int signum)
> +{
> + if (signum == SIGINT || signum == SIGTERM) {
> + printf("\n\nSignal %d received, preparing to exit...\n", signum);
> + force_quit = true;
> + }
> + prompt_exit();
> +}
> +
> +int
> +graph_main_loop(void *conf)
> +{
> + struct lcore_conf *qconf;
> + struct rte_graph *graph;
> + uint32_t lcore_id;
> +
> + RTE_SET_USED(conf);
> +
> + lcore_id = rte_lcore_id();
> + qconf = &lcore_conf[lcore_id];
> + graph = qconf->graph;
> +
> + if (!graph) {
> + RTE_LOG(INFO, TEST_GRAPH, "Lcore %u has nothing to do\n",
> lcore_id);
> + return 0;
> + }
> +
> + RTE_LOG(INFO, TEST_GRAPH, "Entering main loop on lcore %u,
> graph %s(%p)\n", lcore_id,
> + qconf->name, graph);
> +
> + while (likely(!force_quit && !graph_walk_quit)) {
> + if (likely(run_graph_walk))
> + rte_graph_walk(graph);
> + }
graph_walk_quit could be changed to true in destroy stage, we should set it false in create stage also.
Otherwise, it will walk only once for interactive-mode.
I am agree to use different switch for different stage to start/stop, create/destroy and termination.
force_quit + graph_walk_quit + run_graph_walk looks little overcomplicated. Could it only use
force_quit to pause the walk in worker?
> +
> + return 0;
> +}
> +
> +struct rte_graph_cluster_stats *
> +create_graph_cluster_stats(void)
> +{
> + struct rte_graph_cluster_stats_param s_param;
> + struct rte_graph_cluster_stats *stat;
> + const char *pattern = "worker_*";
> +
> + /* Prepare stats object */
> + memset(&s_param, 0, sizeof(s_param));
> + s_param.f = stdout;
> + s_param.socket_id = SOCKET_ID_ANY;
> + s_param.graph_patterns = &pattern;
> + s_param.nb_graph_patterns = 1;
> +
> + stat = rte_graph_cluster_stats_create(&s_param);
> + if (stat == NULL)
> + rte_exit(EXIT_FAILURE, "Unable to create stats object\n");
> +
> + return stat;
> +}
> +
> +static void
> +print_stats(void)
> +{
> + const char topLeft[] = {27, '[', '1', ';', '1', 'H', '\0'};
> + const char clr[] = {27, '[', '2', 'J', '\0'};
> +
> + while (!force_quit) {
> + /* Clear screen and move to top left */
> + printf("%s%s", clr, topLeft);
> + rte_graph_cluster_stats_get(stats, 0);
> + rte_delay_ms(1E3);
> + }
> +
> + rte_graph_cluster_stats_destroy(stats);
> +}
> +
> +int
> +parse_config(const char *q_arg)
> +{
> + enum fieldnames { FLD_PORT = 0, FLD_QUEUE, FLD_LCORE,
> _NUM_FLD };
> + unsigned long int_fld[_NUM_FLD];
> + const char *p, *p0 = q_arg;
> + char *str_fld[_NUM_FLD];
> + uint32_t size;
> + char s[256];
> + char *end;
> + int i;
> +
> + nb_lcore_params = 0;
> +
> + while ((p = strchr(p0, '(')) != NULL) {
> + ++p;
> + p0 = strchr(p, ')');
> + if (p0 == NULL)
> + goto exit;
> +
> + size = p0 - p;
> + if (size >= sizeof(s))
> + goto exit;
> +
> + memcpy(s, p, size);
> + s[size] = '\0';
> + if (rte_strsplit(s, sizeof(s), str_fld, _NUM_FLD, ',') != _NUM_FLD)
> + goto exit;
> + for (i = 0; i < _NUM_FLD; i++) {
> + errno = 0;
> + int_fld[i] = strtoul(str_fld[i], &end, 0);
> + if (errno != 0 || end == str_fld[i])
> + goto exit;
> + }
> +
> + if (nb_lcore_params >= MAX_LCORE_PARAMS) {
> + printf("Exceeded max number of lcore params: %hu\n",
> nb_lcore_params);
> + goto exit;
> + }
> +
> + if (int_fld[FLD_PORT] >= RTE_MAX_ETHPORTS ||
> int_fld[FLD_LCORE] >= RTE_MAX_LCORE) {
> + printf("Invalid port/lcore id\n");
> + goto exit;
> + }
> +
> + lcore_params_array[nb_lcore_params].port_id =
> (uint8_t)int_fld[FLD_PORT];
> + lcore_params_array[nb_lcore_params].queue_id =
> (uint8_t)int_fld[FLD_QUEUE];
> + lcore_params_array[nb_lcore_params].lcore_id =
> (uint8_t)int_fld[FLD_LCORE];
> + ++nb_lcore_params;
> + }
> + lcore_params = lcore_params_array;
> +
> + return 0;
> +exit:
> + /* Revert to default config */
> + lcore_params = lcore_params_array_default;
> + nb_lcore_params = RTE_DIM(lcore_params_array_default);
> +
> + return -1;
> +}
> +
> +int
> +parse_node_patterns(const char *q_arg)
> +{
> + const char *p, *p0 = q_arg;
> + int ret = -EINVAL;
> + uint32_t size;
> +
> + num_patterns = 0;
> +
> + p = strchr(p0, '(');
> + if (p != NULL) {
> + ++p;
> + while ((p0 = strchr(p, ',')) != NULL) {
> + size = p0 - p;
> + if (size >= RTE_NODE_NAMESIZE)
> + goto exit;
> +
> + if (num_patterns >= MAX_NODE_PATTERNS) {
> + printf("Too many nodes passed.\n");
> + goto exit;
> + }
> +
> + memcpy(node_pattern[num_patterns++], p, size);
Here, if not make node_pattern clean, node_pattern[0](size,MAX) will be tainted by previous create.
> + p = p0 + 1;
> + }
> +
> + p0 = strchr(p, ')');
> + if (p0 != NULL) {
> + size = p0 - p;
> + if (size >= RTE_NODE_NAMESIZE)
> + goto exit;
> +
> + if (num_patterns >= MAX_NODE_PATTERNS) {
> + printf("Too many nodes passed.\n");
> + goto exit;
> + }
> +
> + memcpy(node_pattern[num_patterns++], p, size);
> + } else {
> + goto exit;
> + }
> + } else {
> + goto exit;
> + }
> +
> + return 0;
> +exit:
> + return ret;
> +}
> +
> +static void
> +set_default_node_pattern(void)
> +{
> + uint16_t idx;
> +
> + for (idx = 0; idx < test_node_list[0].size; idx++)
> + strcpy(node_pattern[num_patterns++],
> test_node_list[0].nodes[idx]);
> +}
> +
> +int
> +validate_node_names(uint64_t *valid_nodes)
Actually, this func limits the node name to follow test_node_list.
It's OK if there are few nodes/edges.
But how to describe much nodes/edges like
graph_dot {
A->B;
A->C;
B->D;
C->D;
}
If update edges implicitly, there is no need to specific the node sequence. Both of (A,B,C,D) and (A,C,B,D) are correct.
> +{
> + rte_node_t node_cnt = rte_node_max_count();
> + bool pattern_matched = false;
> + rte_node_t id = 0;
> + int ret = -EINVAL;
> + uint16_t idx, i, j;
> +
> + for (idx = 0; idx < num_patterns; idx++) {
> + for (id = 0; id < node_cnt; id++) {
> + if (strncmp(node_pattern[idx],
> rte_node_id_to_name(id),
> + RTE_GRAPH_NAMESIZE) == 0)
> + break;
> + }
> + if (node_cnt == id) {
> + printf("Invalid node name passed\n");
> + return ret;
> + }
> + }
> +
> + for (i = 0; i < RTE_DIM(test_node_list); i++) {
> + idx = 0;
> + if (test_node_list[i].size == num_patterns) {
> + for (j = 0; j < num_patterns; j++) {
> + if (strncmp(node_pattern[j],
> test_node_list[i].nodes[j],
> + RTE_GRAPH_NAMESIZE) == 0)
> + idx++;
> + }
> + if (idx == num_patterns)
> + pattern_matched = true;
> + }
> + }
> +
> + if (!pattern_matched) {
> + printf("Unsupported node pattern passed\n\n");
> + printf("Test supported node patterns are:\n");
> + for (i = 0; i < RTE_DIM(test_node_list); i++) {
> + printf("(");
> + for (j = 0; j < (test_node_list[i].size - 1); j++)
> + printf("%s,", test_node_list[i].nodes[j]);
> + printf("%s", test_node_list[i].nodes[j]);
> + printf(")\n");
> + }
> +
> + return ret;
> + }
> +
> + for (i = 0; i < RTE_DIM(supported_nodes); i++) {
> + for (j = 0; j < num_patterns; j++) {
> + if (strncmp(node_pattern[j],
> supported_nodes[i].nodes[0],
> + RTE_GRAPH_NAMESIZE) == 0) {
> + *valid_nodes |= supported_nodes[i].test_id;
> + break;
> + }
> + }
> + }
> +
> + return 0;
> +}
> +
> +void
> +cleanup_node_pattern(void)
> +{
> + while (num_patterns) {
> + memset(node_pattern[num_patterns - 1], 0,
> RTE_GRAPH_NAMESIZE);
> + num_patterns--;
> + }
> +}
> +
> +static int
> +ethdev_tx_node_configure(void)
> +{
> + struct ethdev_tx_node_main *tx_node_data;
> + struct rte_node_register *tx_node;
> + char name[RTE_NODE_NAMESIZE];
> + uint16_t port_id;
> + uint32_t id;
> +
> + tx_node_data = ethdev_tx_node_data_get();
> + tx_node = ethdev_tx_node_get();
> +
> + RTE_ETH_FOREACH_DEV(port_id) {
> + /* Skip ports that are not enabled */
> + if ((enabled_port_mask & (1 << port_id)) == 0) {
> + printf("\nSkipping disabled port %d\n", port_id);
> + continue;
> + }
> +
> + if (!rte_eth_dev_is_valid_port(port_id))
> + return -EINVAL;
> +
> + /* Create a per port tx node from base node */
> + snprintf(name, sizeof(name), "%u", port_id);
> + id = rte_node_clone(tx_node->id, name);
> + tx_node_data->nodes[port_id] = id;
> +
> + printf("ethdev:: Tx node %s-%s: is at %u\n", tx_node->name,
> name, id);
> + }
> +
> + return 0;
> +}
> +
> +static int
> +punt_kernel_node_configure(void)
> +{
> + struct rte_node_register *punt_node;
> + char name[RTE_NODE_NAMESIZE];
> + struct lcore_conf *qconf;
> + uint32_t lcore_id;
> + uint32_t id;
> +
> + punt_node = punt_kernel_node_get();
> +
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> +
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + /* Create a per lcore punt_kernel node from base node */
> + snprintf(name, sizeof(name), "%u", lcore_id);
> + id = rte_node_clone(punt_node->id, name);
> + strcpy(qconf->punt_kernel_node_name,
> rte_node_id_to_name(id));
> +
> + printf("punt_kernel node %s-%s: is at %u\n", punt_node->name,
> name, id);
> + }
> +
> + return 0;
> +}
> +
> +static int
> +kernel_recv_node_configure(void)
> +{
> + struct rte_node_register *recv_node;
> + char name[RTE_NODE_NAMESIZE];
> + struct lcore_conf *qconf;
> + uint32_t lcore_id;
> + uint32_t id;
> +
> + recv_node = kernel_recv_node_get();
> +
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> +
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + /* Create a per lcore kernel_recv node from base node */
> + snprintf(name, sizeof(name), "%u", lcore_id);
> + id = rte_node_clone(recv_node->id, name);
> + strcpy(qconf->kernel_recv_node_name,
> rte_node_id_to_name(id));
> +
> + printf("kernel_recv node %s-%s: is at %u\n", recv_node->name,
> name, id);
> + }
> +
> + return 0;
> +}
> +
> +static int
> +ethdev_rx_node_configure(struct rte_node_ethdev_config *conf, uint16_t
> nb_confs)
> +{
> + char name[RTE_NODE_NAMESIZE];
> + uint16_t i, j, port_id;
> + uint32_t id;
> +
> + for (i = 0; i < nb_confs; i++) {
> + port_id = conf[i].port_id;
> +
> + if (!rte_eth_dev_is_valid_port(port_id))
> + return -EINVAL;
> +
> + /* Create node for each rx port queue pair */
> + for (j = 0; j < conf[i].num_rx_queues; j++) {
> + struct ethdev_rx_node_main *rx_node_data;
> + struct rte_node_register *rx_node;
> + ethdev_rx_node_elem_t *elem;
> +
> + rx_node_data = ethdev_rx_get_node_data_get();
> + rx_node = ethdev_rx_node_get();
> + snprintf(name, sizeof(name), "%u-%u", port_id, j);
> + /* Clone a new rx node with same edges as parent */
> + id = rte_node_clone(rx_node->id, name);
> + if (id == RTE_NODE_ID_INVALID)
> + return -EIO;
> +
> + /* Add it to list of ethdev rx nodes for lookup */
> + elem = malloc(sizeof(ethdev_rx_node_elem_t));
> + if (elem == NULL)
> + return -ENOMEM;
> + memset(elem, 0, sizeof(ethdev_rx_node_elem_t));
> + elem->ctx.port_id = port_id;
> + elem->ctx.queue_id = j;
> + elem->nid = id;
> + elem->next = rx_node_data->head;
> + rx_node_data->head = elem;
> +
> + printf("ethdev:: Rx node %s-%s: is at %u\n", rx_node-
> >name, name, id);
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int
> +update_ethdev_rx_node_next(rte_node_t id, const char *edge_name)
> +{
> + struct ethdev_rx_node_main *rx_node_data;
> + ethdev_rx_node_elem_t *elem;
> + char *next_nodes[16];
> + rte_edge_t count;
> + uint16_t i;
> +
> + count = rte_node_edge_count(id);
> + rte_node_edge_get(id, next_nodes);
> +
> + for (i = 0; i < count; i++) {
> + if (fnmatch(edge_name, next_nodes[i], 0) == 0) {
> + rx_node_data = ethdev_rx_get_node_data_get();
> + elem = rx_node_data->head;
> + while (elem->next != rx_node_data->head) {
> + if (elem->nid == id)
> + break;
> + elem = elem->next;
> + }
> +
> + if (elem->nid == id)
> + elem->ctx.cls_next = i;
> + break;
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int
> +link_ethdev_rx_to_tx_node(uint32_t lcore_id)
> +{
> + const char * const pattern[] = {"ethdev_tx-*"};
> + uint16_t queue, queue_id, port_id;
> + char name[RTE_NODE_NAMESIZE];
> + const char *next_node = name;
> + struct lcore_conf *qconf;
> + uint32_t rx_id;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + for (queue = 0; queue < qconf->n_rx_queue; ++queue) {
> + port_id = qconf->rx_queue_list[queue].port_id;
> + queue_id = qconf->rx_queue_list[queue].queue_id;
> +
> + snprintf(name, sizeof(name), "ethdev_rx-%u-%u", port_id,
> queue_id);
> + rx_id = rte_node_from_name(name);
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], name);
> +
> + /* Prepare the actual name of the cloned node */
> + snprintf(name, sizeof(name), "ethdev_tx-%u", port_id);
> +
> + /* Update ethdev_rx node edges */
> + rte_node_edge_update(rx_id, RTE_EDGE_ID_INVALID,
> &next_node, 1);
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], name);
> +
> + /* Update node_next details */
> + update_ethdev_rx_node_next(rx_id, pattern[0]);
> + }
> +
> + return 0;
> +}
> +
> +static int
> +link_ethdev_rx_to_punt_kernel_node(uint32_t lcore_id)
> +{
> + const char * const pattern[] = {"punt_kernel-*"};
> + uint16_t queueid, portid, queue;
> + char name[RTE_NODE_NAMESIZE];
> + const char *next_node = name;
> + struct lcore_conf *qconf;
> + rte_node_t rx_id;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + for (queue = 0; queue < qconf->n_rx_queue; ++queue) {
> +
> + portid = qconf->rx_queue_list[queue].port_id;
> + queueid = qconf->rx_queue_list[queue].queue_id;
> +
> + snprintf(name, sizeof(name), "ethdev_rx-%u-%u", portid,
> queueid);
> + rx_id = rte_node_from_name(name);
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], name);
> +
> + next_node = qconf->punt_kernel_node_name;
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], next_node);
> +
> + /* Update ethdev_rx node edges */
> + rte_node_edge_update(rx_id, RTE_EDGE_ID_INVALID,
> &next_node, 1);
> +
> + /* Update node_next details */
> + update_ethdev_rx_node_next(rx_id, pattern[0]);
> + }
> +
> + return 0;
> +}
> +
> +static int
> +update_kernel_recv_node_next(rte_node_t id, const char *edge_name)
> +{
> + struct kernel_recv_node_main *rx_node_data;
> + kernel_recv_node_elem_t *elem;
> + char *next_nodes[16];
> + rte_edge_t count;
> + uint16_t i;
> +
> + count = rte_node_edge_count(id);
> + rte_node_edge_get(id, next_nodes);
> +
> + for (i = 0; i < count; i++) {
> + if (fnmatch(edge_name, next_nodes[i], 0) == 0) {
> + rx_node_data = kernel_recv_node_data_get();
> + elem = rx_node_data->head;
> + while (elem->next != rx_node_data->head) {
> + if (elem->nid == id)
> + break;
> + elem = elem->next;
> + }
> +
> + if (elem->nid == id)
> + elem->ctx.recv_info->cls_next = i;
> + break;
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int
> +link_kernel_recv_to_ethdev_tx_node(uint32_t lcore_id)
> +{
> + const char * const pattern[] = {"ethdev_tx-*"};
> + char name[RTE_NODE_NAMESIZE];
> + const char *next_node = name;
> + struct lcore_conf *qconf;
> + uint16_t port_id;
> + rte_node_t id;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + id = rte_node_from_name(qconf->kernel_recv_node_name);
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], qconf-
> >kernel_recv_node_name);
> +
> + port_id = lcore_id;
> +
> + if ((enabled_port_mask & (1 << port_id)) == 0) {
> + /* Use available port_id */
> + RTE_ETH_FOREACH_DEV(port_id) {
> + if ((enabled_port_mask & (1 << port_id)) != 0)
> + break;
> + }
> + }
> +
> + if (!rte_eth_dev_is_valid_port(port_id)) {
> + printf("Port %u is not present on the board\n", port_id);
> + return -1;
> + }
> +
> + /* Prepare the actual name of the cloned node */
> + snprintf(name, sizeof(name), "ethdev_tx-%u", port_id);
> +
> + /* Update ethdev_rx node edges */
> + rte_node_edge_update(id, RTE_EDGE_ID_INVALID, &next_node, 1);
> +
> + /* Fill node pattern */
> + strcpy(node_pattern[num_patterns++], name);
> +
> + /* Update node_next details */
> + update_kernel_recv_node_next(id, pattern[0]);
> +
> + return 0;
> +}
These explicitly update node and link could be put in a independent stage, like graph_config.
Also, I wonder if update link could be more generic. Cause they have some similar steps.
BTY, In this patch, it could support rx->tx first.
Then add a new patch to support other topo(kernel_recv, punt_kernel and l3fwd).
And we could enable some help function in another patch.
It could make the patch small.
> +
> +uint32_t
> +ethdev_ports_setup(void)
> +{
> + struct rte_eth_dev_info dev_info;
> + uint32_t nb_tx_queue, nb_lcores;
> + uint32_t nb_ports, nb_conf = 0;
> + uint8_t nb_rx_queue;
> + uint16_t portid;
> + int ret;
> +
> + nb_ports = rte_eth_dev_count_avail();
> + nb_lcores = rte_lcore_count();
> +
> + RTE_ETH_FOREACH_DEV(portid) {
> + struct rte_eth_conf local_port_conf = port_conf;
> +
> + /* Skip ports that are not enabled */
> + if ((enabled_port_mask & (1 << portid)) == 0) {
> + printf("\nSkipping disabled port %d\n", portid);
> + continue;
> + }
> +
> + /* Init port */
> + printf("Initializing port %d ... ", portid);
> + fflush(stdout);
> +
> + nb_rx_queue = get_port_n_rx_queues(portid);
> + nb_tx_queue = nb_lcores;
> + if (nb_tx_queue > MAX_TX_QUEUE_PER_PORT)
> + nb_tx_queue = MAX_TX_QUEUE_PER_PORT;
> + printf("Creating queues: nb_rxq=%d nb_txq=%u...\n",
> nb_rx_queue, nb_tx_queue);
> +
> + rte_eth_dev_info_get(portid, &dev_info);
> +
> + if (dev_info.tx_offload_capa &
> RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE)
> + local_port_conf.txmode.offloads |=
> RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE;
> +
> + local_port_conf.rx_adv_conf.rss_conf.rss_hf &=
> dev_info.flow_type_rss_offloads;
> + if (local_port_conf.rx_adv_conf.rss_conf.rss_hf !=
> + port_conf.rx_adv_conf.rss_conf.rss_hf) {
> + printf("Port %u modified RSS hash function based on
> hardware support,"
> + "requested:%#" PRIx64 " configured:%#" PRIx64
> "\n",
> + portid, port_conf.rx_adv_conf.rss_conf.rss_hf,
> + local_port_conf.rx_adv_conf.rss_conf.rss_hf);
> + }
> +
> + ret = rte_eth_dev_configure(portid, nb_rx_queue, nb_tx_queue,
> &local_port_conf);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE, "Cannot configure device:
> err=%d, port=%d\n", ret,
> + portid);
> +
> + ret = rte_eth_dev_adjust_nb_rx_tx_desc(portid, &nb_rxd,
> &nb_txd);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE,
> + "Cannot adjust number of descriptors:
> err=%d,"
> + "port=%d\n",
> + ret, portid);
> +
> + /* Init memory */
> + if (!per_port_pool)
> + ret = init_mem(0, NB_MBUF(nb_ports));
> + else
> + ret = init_mem(portid, NB_MBUF(1));
> +
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE, "init_mem() failed\n");
> +
> + /* Setup ethdev node config */
> + ethdev_conf[nb_conf].port_id = portid;
> + ethdev_conf[nb_conf].num_rx_queues = nb_rx_queue;
> + ethdev_conf[nb_conf].num_tx_queues = nb_tx_queue;
> + if (!per_port_pool)
> + ethdev_conf[nb_conf].mp = pktmbuf_pool[0];
> +
> + else
> + ethdev_conf[nb_conf].mp = pktmbuf_pool[portid];
> + ethdev_conf[nb_conf].mp_count = NB_SOCKETS;
> +
> + nb_conf++;
> + }
> +
> + return nb_conf;
> +}
> +
> +void
> +ethdev_rxq_configure(void)
> +{
> + struct rte_eth_dev_info dev_info;
> + uint16_t queueid, portid;
> + struct lcore_conf *qconf;
> + uint8_t queue, socketid;
> + uint32_t lcore_id;
> + int ret;
> +
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> + qconf = &lcore_conf[lcore_id];
> + printf("\nInitializing rx queues on lcore %u ... ", lcore_id);
> + fflush(stdout);
> +
> + /* Init RX queues */
> + for (queue = 0; queue < qconf->n_rx_queue; ++queue) {
> + struct rte_eth_rxconf rxq_conf;
> +
> + portid = qconf->rx_queue_list[queue].port_id;
> + queueid = qconf->rx_queue_list[queue].queue_id;
> +
> + if (numa_on)
> + socketid =
> (uint8_t)rte_lcore_to_socket_id(lcore_id);
> + else
> + socketid = 0;
> +
> + printf("rxq=%d,%d,%d ", portid, queueid, socketid);
> + fflush(stdout);
> +
> + rte_eth_dev_info_get(portid, &dev_info);
> + rxq_conf = dev_info.default_rxconf;
> + rxq_conf.offloads = port_conf.rxmode.offloads;
> + if (!per_port_pool)
> + ret = rte_eth_rx_queue_setup(portid, queueid,
> nb_rxd, socketid,
> + &rxq_conf,
> pktmbuf_pool[0][socketid]);
> + else
> + ret = rte_eth_rx_queue_setup(portid, queueid,
> nb_rxd, socketid,
> + &rxq_conf,
> +
> pktmbuf_pool[portid][socketid]);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE,
> "rte_eth_rx_queue_setup: err=%d, port=%d\n",
> + ret, portid);
> +
> + snprintf(qconf->rx_queue_list[queue].node_name,
> RTE_NODE_NAMESIZE,
> + "ethdev_rx-%u-%u", portid, queueid);
> + }
> + }
> + printf("\n");
> +}
> +
> +void
> +ethdev_txq_configure(void)
> +{
> + struct rte_eth_dev_info dev_info;
> + struct rte_eth_txconf *txconf;
> + uint16_t queueid, portid;
> + uint32_t lcore_id;
> + uint8_t socketid;
> + int ret;
> +
> + RTE_ETH_FOREACH_DEV(portid) {
> + struct rte_eth_conf local_port_conf = port_conf;
> +
> + /* Skip ports that are not enabled */
> + if ((enabled_port_mask & (1 << portid)) == 0) {
> + printf("\nSkipping disabled port %d\n", portid);
> + continue;
> + }
> +
> + rte_eth_dev_info_get(portid, &dev_info);
> +
> + /* Init one TX queue per (lcore,port) pair */
> + queueid = 0;
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> +
> + if (numa_on)
> + socketid =
> (uint8_t)rte_lcore_to_socket_id(lcore_id);
> + else
> + socketid = 0;
> +
> + printf("txq=%u,%d,%d ", lcore_id, queueid, socketid);
> + fflush(stdout);
> +
> + txconf = &dev_info.default_txconf;
> + txconf->offloads = local_port_conf.txmode.offloads;
> + ret = rte_eth_tx_queue_setup(portid, queueid, nb_txd,
> socketid, txconf);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE,
> "rte_eth_tx_queue_setup: err=%d, port=%d\n",
> + ret, portid);
> + queueid++;
> + }
> + }
> + printf("\n");
> +}
> +
> +int
> +configure_graph_nodes(uint64_t valid_nodes)
> +{
> + int ret = 0;
> +
> + if (valid_nodes & TEST_GRAPH_ETHDEV_RX_NODE) {
> + ret = ethdev_rx_node_configure(ethdev_conf, nb_conf);
> + if (ret) {
> + printf("ethdev_rx_node_configure: err=%d\n", ret);
> + goto exit;
> + }
> + }
> +
> + if (valid_nodes & TEST_GRAPH_ETHDEV_TX_NODE) {
> + ret = ethdev_tx_node_configure();
> + if (ret) {
> + printf("ethdev_tx_node_configure: err=%d\n", ret);
> + goto exit;
> + }
> + }
> +
> + if (valid_nodes & TEST_GRAPH_PUNT_KERNEL_NODE) {
> + ret = punt_kernel_node_configure();
> + if (ret) {
> + printf("punt_kernel_node_configure: err=%d\n", ret);
> + goto exit;
> + }
> + }
> +
> + if (valid_nodes & TEST_GRAPH_KERNEL_RECV_NODE) {
> + ret = kernel_recv_node_configure();
> + if (ret) {
> + printf("kernel_recv_node_configure: err=%d\n", ret);
> + goto exit;
> + }
> + }
> +
> +exit:
> + return ret;
> +}
> +
> +static int
> +link_graph_nodes(uint64_t valid_nodes, uint32_t lcore_id)
> +{
> + int ret = 0;
> +
> + num_patterns = 0;
> +
> + if (valid_nodes == (TEST_GRAPH_ETHDEV_TX_NODE |
> TEST_GRAPH_ETHDEV_RX_NODE)) {
> + ret = link_ethdev_rx_to_tx_node(lcore_id);
> + if (ret) {
> + printf("link_ethdev_rx_to_tx_node: err=%d\n", ret);
> + goto exit;
> + }
> + } else if (valid_nodes == (TEST_GRAPH_ETHDEV_RX_NODE |
> TEST_GRAPH_PUNT_KERNEL_NODE)) {
> + link_ethdev_rx_to_punt_kernel_node(lcore_id);
> + } else if (valid_nodes == (TEST_GRAPH_ETHDEV_TX_NODE |
> TEST_GRAPH_KERNEL_RECV_NODE)) {
> + link_kernel_recv_to_ethdev_tx_node(lcore_id);
> + } else {
> + printf("Invalid node map\n");
> + ret = -EINVAL;
> + }
> +
> +exit:
> + return ret;
> +}
> +
> +void
> +start_eth_ports(void)
> +{
> + uint16_t portid;
> + int ret;
> +
> + RTE_ETH_FOREACH_DEV(portid) {
> + if ((enabled_port_mask & (1 << portid)) == 0)
> + continue;
> +
> + ret = rte_eth_dev_start(portid);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE, "rte_eth_dev_start: err=%d,
> port=%d\n", ret, portid);
> +
> + if (promiscuous_on)
> + rte_eth_promiscuous_enable(portid);
> + }
> +}
> +
> +void
> +stop_eth_ports(void)
> +{
> + uint16_t portid;
> + int ret;
> +
> + RTE_ETH_FOREACH_DEV(portid) {
> + if ((enabled_port_mask & (1 << portid)) == 0)
> + continue;
> + printf("Closing port %d...", portid);
> + ret = rte_eth_dev_stop(portid);
> + if (ret != 0)
> + printf("Failed to stop port %u: %s\n", portid,
> rte_strerror(-ret));
> + rte_eth_dev_close(portid);
close should be in exit stage or when application is killed.
> + }
> +}
> +
> +int
> +create_graph(uint64_t valid_nodes)
> +{
> + struct rte_graph_param graph_conf;
> + struct lcore_conf *qconf;
> + uint32_t lcore_id;
> + int ret;
> +
> + node_patterns = malloc(MAX_NODE_PATTERNS *
> sizeof(*node_patterns));
> + if (!node_patterns)
> + return -ENOMEM;
> +
> + memset(&graph_conf, 0, sizeof(graph_conf));
> + graph_conf.node_patterns = node_patterns;
> +
> + for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> + rte_graph_t graph_id;
> + rte_edge_t i;
> +
> + if (rte_lcore_is_enabled(lcore_id) == 0)
> + continue;
> +
> + qconf = &lcore_conf[lcore_id];
> +
> + /* Skip graph creation if no source exists */
> + if (!qconf->n_rx_queue)
> + continue;
> +
> + ret = link_graph_nodes(valid_nodes, lcore_id);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "link_graph_nodes(): failed\n");
> +
> + for (i = 0; i < num_patterns; i++) {
> + graph_conf.node_patterns[i] = node_pattern[i];
> + printf("%s,", graph_conf.node_patterns[i]);
> + }
> + printf("\n");
> +
> + graph_conf.nb_node_patterns = num_patterns;
> +
> + graph_conf.socket_id = rte_lcore_to_socket_id(lcore_id);
> +
> + snprintf(qconf->name, sizeof(qconf->name), "worker_%u",
> lcore_id);
> +
> + graph_id = rte_graph_create(qconf->name, &graph_conf);
> + if (graph_id == RTE_GRAPH_ID_INVALID)
> + rte_exit(EXIT_FAILURE, "rte_graph_create(): graph_id
> invalid for lcore%u\n",
> + lcore_id);
> +
> + qconf->graph_id = graph_id;
> + qconf->graph = rte_graph_lookup(qconf->name);
> +
> + if (!qconf->graph)
> + rte_exit(EXIT_FAILURE, "rte_graph_lookup(): graph %s
> not found\n",
> + qconf->name);
> + }
> +
> + return 0;
> +}
> +
> +int
> +destroy_graph(void)
> +{
> + uint32_t lcore_id;
> + rte_graph_t id;
> + int ret;
> +
> + RTE_LCORE_FOREACH_WORKER(lcore_id)
> + {
> + if (lcore_conf[lcore_id].graph) {
> + id = rte_graph_from_name(lcore_conf[lcore_id].name);
> + if (rte_graph_destroy(id)) {
> + printf("graph_id %u destroy failed.\n", id);
> + ret = -1;
> + }
> + }
> + }
> +
> + if (node_patterns)
> + free(node_patterns);
Cause this cmd could be called twice in interactive-mode, we should set it to NULL after free.
> +
> + return ret;
> +}
> +
> +int
> +main(int argc, char **argv)
> +{
> + uint64_t valid_nodes;
> + uint32_t lcore_id;
> + int ret;
> +
> + graph_walk_quit = false;
> + force_quit = false;
> + interactive = 0;
> +
> + node_patterns = NULL;
> +
> + signal(SIGINT, signal_handler);
> + signal(SIGTERM, signal_handler);
> +
> + testgraph_logtype = rte_log_register("testgraph");
> + if (testgraph_logtype < 0)
> + rte_exit(EXIT_FAILURE, "Cannot register log type");
> +
> + set_default_node_pattern();
> +
> + rte_log_set_level(testgraph_logtype, RTE_LOG_DEBUG);
> +
> + ret = rte_eal_init(argc, argv);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE, "Cannot init EAL: %s\n",
> rte_strerror(rte_errno));
> + argc -= ret;
> + argv += ret;
> +
> + if (argc > 1) {
> + ret = parse_cmdline_args(argc, argv);
> + if (ret < 0)
> + rte_exit(EXIT_FAILURE, "Invalid command line
> parameters\n");
> + }
> +
> +#ifdef RTE_LIB_CMDLINE
> + if (init_cmdline() != 0)
> + rte_exit(EXIT_FAILURE, "Could not initialise cmdline
> context.\n");
> +
> + if (interactive == 1) {
> + prompt();
> + } else
> +#endif
> + {
> + if (validate_config() < 0)
> + rte_exit(EXIT_FAILURE, "Config validation failed.\n");
> +
> + ret = validate_node_names(&valid_nodes);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "validate_node_names:
> err=%d\n", ret);
> +
> + nb_conf = ethdev_ports_setup();
> +
> + ethdev_rxq_configure();
> +
> + ethdev_txq_configure();
> +
> + ret = configure_graph_nodes(valid_nodes);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "configure_graph_nodes:
> err=%d\n", ret);
> +
> + ret = create_graph(valid_nodes);
> + if (ret)
> + rte_exit(EXIT_FAILURE, "create_graph: err=%d\n", ret);
> +
> + stats = create_graph_cluster_stats();
> + if (stats == NULL)
> + rte_exit(EXIT_FAILURE, "create_graph_cluster_stats()
> failed\n");
> +
> + check_all_ports_link_status(enabled_port_mask);
> +
> + start_eth_ports();
> +
> + /* Launch per-lcore init on every worker lcore */
> + rte_eal_mp_remote_launch(graph_main_loop, NULL,
> SKIP_MAIN);
> +
> + /* Accumulate and print stats on main until exit */
> + if (rte_graph_has_stats_feature())
> + print_stats();
> +
> + /* Wait for worker cores to exit */
> + RTE_LCORE_FOREACH_WORKER(lcore_id) {
> + ret = rte_eal_wait_lcore(lcore_id);
> + if (ret < 0)
> + break;
> + }
> +
> + ret = destroy_graph();
> +
> + stop_eth_ports();
> + }
> +
> + /* clean up the EAL */
> + ret = rte_eal_cleanup();
> + if (ret != 0)
> + rte_exit(EXIT_FAILURE, "EAL cleanup failed: %s\n", strerror(-
> ret));
> +
> + return EXIT_SUCCESS;
> +}
> diff --git a/app/test-graph/testgraph.h b/app/test-graph/testgraph.h
> new file mode 100644
> index 0000000000..ea1f8ef4fa
> --- /dev/null
> +++ b/app/test-graph/testgraph.h
> @@ -0,0 +1,91 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(C) 2023 Marvell International Ltd.
> + */
> +
> +#ifndef _TESTGRAPH_H_
> +#define _TESTGRAPH_H_
> +
> +#include <stdbool.h>
> +
> +#include <rte_ethdev.h>
> +#include <rte_graph.h>
> +#include <rte_node_eth_api.h>
> +
> +#include <cmdline.h>
> +#include <cmdline_parse.h>
> +
> +
> +#define MAX_LCORE_PARAMS 1024
> +#define MAX_NODE_PATTERNS 128
> +
> +#ifndef BIT_ULL
> +#define BIT_ULL(nr) (1ULL << (nr))
> +#endif
> +
> +#define TEST_GRAPH_ETHDEV_RX_NODE BIT_ULL(0)
> +#define TEST_GRAPH_ETHDEV_TX_NODE BIT_ULL(1)
> +#define TEST_GRAPH_PUNT_KERNEL_NODE BIT_ULL(2)
> +#define TEST_GRAPH_KERNEL_RECV_NODE BIT_ULL(3)
> +#define TEST_GRAPH_IP4_LOOKUP_NODE BIT_ULL(4)
> +#define TEST_GRAPH_IP4_REWRITE_NODE BIT_ULL(5)
> +#define TEST_GRAPH_PKT_CLS_NODE BIT_ULL(6)
> +#define TEST_GRAPH_PKT_DROP_NODE BIT_ULL(7)
> +#define TEST_GRAPH_NULL_NODE BIT_ULL(8)
> +
> +static volatile bool force_quit;
> +
> +extern struct rte_node_ethdev_config ethdev_conf[RTE_MAX_ETHPORTS];
> +extern uint32_t enabled_port_mask;
> +extern uint32_t nb_conf;
> +
> +extern int promiscuous_on; /**< Ports set in promiscuous mode off by
> default. */
> +extern uint8_t interactive; /**< interactive mode is disabled by default.
> */
> +extern uint32_t enabled_port_mask; /**< Mask of enabled ports */
> +extern int numa_on; /**< NUMA is enabled by default. */
> +extern int per_port_pool;
> +
> +extern volatile bool graph_walk_quit;
> +extern volatile bool run_graph_walk;
> +extern struct rte_graph_cluster_stats *stats;
> +
> +extern char node_pattern[MAX_NODE_PATTERNS][RTE_NODE_NAMESIZE];
> +extern uint8_t num_patterns;
> +
> +struct node_list {
> + const char *nodes[MAX_NODE_PATTERNS];
> + uint64_t test_id;
> + uint8_t size;
> +};
> +
> +struct lcore_params {
> + uint16_t port_id;
> + uint8_t queue_id;
> + uint8_t lcore_id;
> +} __rte_cache_aligned;
> +
> +void prompt(void);
> +void prompt_exit(void);
> +int init_cmdline(void);
> +int validate_config(void);
> +int parse_cmdline_args(int argc, char **argv);
> +uint32_t ethdev_ports_setup(void);
> +void ethdev_rxq_configure(void);
> +void ethdev_txq_configure(void);
> +void start_eth_ports(void);
> +void stop_eth_ports(void);
> +int create_graph(uint64_t valid_nodes);
> +int destroy_graph(void);
> +int graph_main_loop(void *conf);
> +struct rte_graph_cluster_stats *create_graph_cluster_stats(void);
> +void check_all_ports_link_status(uint32_t port_mask);
> +int configure_graph_nodes(uint64_t valid_nodes);
> +
> +int parse_config(const char *q_arg);
> +int parse_node_patterns(const char *q_arg);
> +int validate_node_names(uint64_t *valid_nodes);
> +void cleanup_node_pattern(void);
> +
> +#define TESTGRAPH_LOG(level, fmt, args...) \
> + rte_log(RTE_LOG_##level, testgraph_logtype, "testgraph: " fmt, ##args)
> +
> +#endif /* _TESTGRAPH_H_ */
> diff --git a/doc/guides/tools/index.rst b/doc/guides/tools/index.rst
> index 6f84fc31ff..f18c508fa2 100644
> --- a/doc/guides/tools/index.rst
> +++ b/doc/guides/tools/index.rst
> @@ -20,6 +20,7 @@ DPDK Tools User Guides
> cryptoperf
> comp_perf
> testeventdev
> + testgraph
> testregex
> testmldev
> dts
> diff --git a/doc/guides/tools/testgraph.rst b/doc/guides/tools/testgraph.rst
> new file mode 100644
> index 0000000000..3c1e058724
> --- /dev/null
> +++ b/doc/guides/tools/testgraph.rst
> @@ -0,0 +1,131 @@
> +.. SPDX-License-Identifier: BSD-3-Clause
> + Copyright(C) 2023 Marvell International Ltd.
> +
> +dpdk-test-graph Application
> +===========================
> +
> +The ``dpdk-test-graph`` tool is a Data Plane Development Kit (DPDK) application
> that allows
> +exercising various graph library features. This application has a generic
> framework to add
> +new test configurations and expand test coverage to verify the functionality of
> graph nodes
> +and observe the graph cluster statistics.
> +
> +Running the Application
> +-----------------------
> +
> +The application has a number of command line options:
> +
> +.. code-block:: console
> +
> + dpdk-test-eventdev [EAL Options] -- [application options]
> +
> +EAL Options
> +~~~~~~~~~~~
> +
> +The following are the EAL command-line options that can be used in
> conjunction
> +with the ``dpdk-test-graph`` application.
> +See the DPDK Getting Started Guides for more information on these options.
> +
> +* ``-c <COREMASK>`` or ``-l <CORELIST>``
> +
> + Set the hexadecimal bitmask of the cores to run on. The corelist is a
> + list of cores to use.
> +
> +Application Options
> +~~~~~~~~~~~~~~~~~~~
> +
> +The following are the application command-line options:
> +
> +* ``-p <n>``
> +
> + Set the ethdev port mask.
> +
> +* ``-P``
> +
> + Set the ethdev ports in promiscuous mode.
> +
> +* ``--config <config>``
> +
> + Set the Rxq configuration.
> + (i.e. ``--config (port_id,rxq,lcore_id)[,(port_id,rxq,lcore_id)]``).
> +
> +* ``--node-pattern <n>``
> +
> + Set the node patterns to use in graph creation.
> + (i.e. ``--node-pattern (node_name0,node_name1[,node_nameX])``).
> +
> +* ``--per-port-pool``
> +
> + Use separate buffer pool per port.
> +
> +* ``--no-numa``
> +
> + Disable numa awareness.
> +
> +* ``--interactive``
> +
> + Switch to interactive mode.
> +
> +Running the Tool
> +~~~~~~~~~~~~~~~~
> +
> +Here is the sample command line to run simple iofwd test::
> +
> + ./dpdk-test-graph -a 0002:03:00.0 -a 0002:04:00.0 -c 0xF -- -p 0x3 -P \
> + --config "(0,0,2),(1,0,2)" --node-pattern "(ethdev_rx,ethdev_tx)"
> +
> +Below is a sample command line to punt rx packets to kernel::
> +
> + ./dpdk-test-graph -a 0002:03:00.0 -a 0002:04:00.0 -c 0xF -- -p 0x3 -P \
> + --config "(0,0,2),(1,0,2)" --node-pattern "(ethdev_rx,punt_kernel)"
> +
> +Interactive mode
> +~~~~~~~~~~~~~~~~
> +
> +Tool uses ``--interactive`` command line option to enter interactive mode and
> use cmdline options
> +to setup the required node configurations, create graph and than start
> graph_walk.
> +
> +
> +testgraph> help
> +
> +Help is available for the following sections:
> +
> + help control : Start and stop graph walk.
> + help display : Displaying port, stats and config information.
> + help config : Configuration information.
> + help all : All of the above sections.
> +
> +testgraph> help all
> +
> +Control forwarding:
> +
> +start graph_walk
> + Start graph_walk on worker threads.
> +
> +stop graph_walk
> + Stop worker threads from running graph_walk.
> +
> +quit
> + Quit to prompt.
> +
> +
> +Display:
> +
> +show node_list
> + Display the list of supported nodes.
> +
> +show graph_stats
> + Display the node statistics of graph cluster.
> +
> +
> +Configuration:
> +
> +set lcore_config (port_id0,rxq0,lcore_idX),........,(port_idX,rxqX,lcoreidY)
> + Set lcore configuration.
> +
> +create_graph (node0_name,node1_name,...,nodeX_name)
> + Create graph instances using the provided node details.
> +
> +destroy_graph
> + Destroy the graph instances.
> +
> +testgraph>
> --
> 2.25.1
next prev parent reply other threads:[~2023-05-12 11:08 UTC|newest]
Thread overview: 182+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-21 6:02 [PATCH 0/4] app: introduce " Vamsi Attunuru
2023-04-21 6:02 ` [PATCH 1/4] node: add pkt punt to kernel node Vamsi Attunuru
2023-05-29 17:29 ` Jerin Jacob
2023-05-31 12:36 ` [EXT] " Vamsi Krishna Attunuru
2023-04-21 6:02 ` [PATCH 2/4] node: add a node to receive pkts from kernel Vamsi Attunuru
2023-05-29 17:39 ` Jerin Jacob
2023-06-01 2:40 ` [EXT] " Vamsi Krishna Attunuru
2023-04-21 6:02 ` [PATCH 3/4] node: remove hardcoded node next details Vamsi Attunuru
2023-04-21 6:02 ` [PATCH 4/4] app: add testgraph application Vamsi Attunuru
2023-05-09 6:09 ` Jerin Jacob
2023-04-25 13:15 ` [PATCH v2 0/4] app: introduce " Vamsi Attunuru
2023-04-25 13:15 ` [PATCH v2 1/4] node: add pkt punt to kernel node Vamsi Attunuru
2023-05-30 8:16 ` Nithin Dabilpuram
2023-05-31 12:35 ` [EXT] " Vamsi Krishna Attunuru
2023-04-25 13:15 ` [PATCH v2 2/4] node: add a node to receive pkts from kernel Vamsi Attunuru
2023-04-25 13:15 ` [PATCH v2 3/4] node: remove hardcoded node next details Vamsi Attunuru
2023-04-25 13:15 ` [PATCH v2 4/4] app: add testgraph application Vamsi Attunuru
2023-05-09 2:34 ` [EXT] " Sunil Kumar Kori
2023-05-09 3:39 ` Vamsi Krishna Attunuru
2023-05-09 8:53 ` Sunil Kumar Kori
2023-05-09 12:24 ` Vamsi Krishna Attunuru
2023-05-12 11:08 ` Yan, Zhirun [this message]
2023-05-22 7:07 ` Vamsi Krishna Attunuru
2023-05-30 7:34 ` Jerin Jacob
2023-06-01 2:44 ` [EXT] " Vamsi Krishna Attunuru
2023-07-20 15:52 ` Rakesh Kudurumalla
2023-07-21 6:48 ` Jerin Jacob
2023-07-21 7:01 ` Sunil Kumar Kori
2023-07-21 7:39 ` Rakesh Kudurumalla
2023-09-08 11:00 ` Sunil Kumar Kori
2023-09-08 10:49 ` [PATCH v3 1/1] app/graph: add example for different usecases skori
2023-09-09 1:18 ` [EXT] " Nithin Kumar Dabilpuram
2023-09-19 16:04 ` [PATCH v4 01/14] app/graph: add application framework to read CLI skori
2023-09-19 16:04 ` [PATCH v4 02/14] app/graph: add telnet connectivity framework skori
2023-09-20 4:34 ` Jerin Jacob
2023-09-20 8:14 ` Bruce Richardson
2023-09-19 16:04 ` [PATCH v4 03/14] app/graph: add parser utility APIs skori
2023-09-19 16:04 ` [PATCH v4 04/14] app/graph: add mempool command line interfaces skori
2023-09-19 16:04 ` [PATCH v4 05/14] app/graph: add ethdev " skori
2023-09-19 16:04 ` [PATCH v4 06/14] app/graph: add ipv4_lookup " skori
2023-09-19 16:04 ` [PATCH v4 07/14] app/graph: add ipv6_lookup " skori
2023-09-19 16:04 ` [PATCH v4 08/14] app/graph: add neigh " skori
2023-09-19 16:04 ` [PATCH v4 09/14] app/graph: add ethdev_rx " skori
2023-09-19 16:04 ` [PATCH v4 10/14] app/graph: add graph " skori
2023-09-19 16:04 ` [PATCH v4 11/14] app/graph: add CLI option to enable graph stats skori
2023-09-19 16:04 ` [PATCH v4 12/14] app/graph: add l3fwd usecase skori
2023-09-19 16:04 ` [PATCH v4 13/14] doc: add graph application user guide skori
2023-09-20 4:28 ` Jerin Jacob
2023-09-19 16:04 ` [PATCH v4 14/14] maintainers: add maintainers for graph app skori
2023-09-20 4:40 ` [PATCH v4 01/14] app/graph: add application framework to read CLI Jerin Jacob
2023-09-21 10:08 ` [PATCH v5 00/12] add CLI based graph application skori
2023-09-21 10:08 ` [PATCH v5 01/12] app/graph: add application framework to read CLI skori
2023-09-21 10:08 ` [PATCH v5 02/12] app/graph: add telnet connectivity framework skori
2023-09-21 10:08 ` [PATCH v5 03/12] app/graph: add parser utility APIs skori
2023-09-21 10:08 ` [PATCH v5 04/12] app/graph: add mempool command line interfaces skori
2023-09-21 10:08 ` [PATCH v5 05/12] app/graph: add ethdev " skori
2023-09-21 10:08 ` [PATCH v5 06/12] app/graph: add ipv4_lookup " skori
2023-09-21 10:08 ` [PATCH v5 07/12] app/graph: add ipv6_lookup " skori
2023-09-21 10:08 ` [PATCH v5 08/12] app/graph: add neigh " skori
2023-09-21 10:08 ` [PATCH v5 09/12] app/graph: add ethdev_rx " skori
2023-09-21 10:08 ` [PATCH v5 10/12] app/graph: add graph " skori
2023-09-21 10:08 ` [PATCH v5 11/12] app/graph: add CLI option to enable graph stats skori
2023-09-21 10:08 ` [PATCH v5 12/12] app/graph: add l3fwd usecase skori
2023-09-26 10:57 ` [PATCH v6 00/12] add CLI based graph application skori
2023-09-26 10:57 ` [PATCH v6 01/12] app/graph: add application framework to read CLI skori
2023-09-26 10:57 ` [PATCH v6 02/12] app/graph: add telnet connectivity framework skori
2023-09-26 10:57 ` [PATCH v6 03/12] app/graph: add parser utility APIs skori
2023-09-26 10:57 ` [PATCH v6 04/12] app/graph: add mempool command line interfaces skori
2023-09-26 10:57 ` [PATCH v6 05/12] app/graph: add ethdev " skori
2023-09-26 10:57 ` [PATCH v6 06/12] app/graph: add ipv4_lookup " skori
2023-09-26 10:57 ` [PATCH v6 07/12] app/graph: add ipv6_lookup " skori
2023-09-26 10:57 ` [PATCH v6 08/12] app/graph: add neigh " skori
2023-09-26 10:57 ` [PATCH v6 09/12] app/graph: add ethdev_rx " skori
2023-09-26 10:57 ` [PATCH v6 10/12] app/graph: add graph " skori
2023-09-26 10:57 ` [PATCH v6 11/12] app/graph: add CLI option to enable graph stats skori
2023-09-26 10:57 ` [PATCH v6 12/12] app/graph: add l3fwd use case skori
2023-09-27 11:54 ` [PATCH v7 00/12] add CLI based graph application skori
2023-09-27 11:54 ` [PATCH v7 01/12] app/graph: add application framework to read CLI skori
2023-09-27 11:54 ` [PATCH v7 02/12] app/graph: add telnet connectivity framework skori
2023-09-27 11:54 ` [PATCH v7 03/12] app/graph: add parser utility APIs skori
2023-09-27 11:54 ` [PATCH v7 04/12] app/graph: add mempool command line interfaces skori
2023-09-27 11:54 ` [PATCH v7 05/12] app/graph: add ethdev " skori
2023-09-27 11:54 ` [PATCH v7 06/12] app/graph: add ipv4_lookup " skori
2023-09-27 11:54 ` [PATCH v7 07/12] app/graph: add ipv6_lookup " skori
2023-09-27 11:54 ` [PATCH v7 08/12] app/graph: add neigh " skori
2023-09-27 11:54 ` [PATCH v7 09/12] app/graph: add ethdev_rx " skori
2023-09-27 11:54 ` [PATCH v7 10/12] app/graph: add graph " skori
2023-09-27 11:54 ` [PATCH v7 11/12] app/graph: add CLI option to enable graph stats skori
2023-09-27 11:54 ` [PATCH v7 12/12] app/graph: add l3fwd use case skori
2023-09-29 9:58 ` [PATCH v8 00/12] add CLI based graph application skori
2023-09-29 9:58 ` [PATCH v8 01/12] app/graph: add application framework to read CLI skori
2023-10-16 9:00 ` Jerin Jacob
2023-10-17 6:19 ` [EXT] " Sunil Kumar Kori
2023-10-18 6:33 ` [PATCH v9 00/12] add CLI based graph application skori
2023-10-18 6:33 ` [PATCH v9 01/12] app/graph: support application CLI framework skori
2023-10-18 6:33 ` [PATCH v9 02/12] app/graph: support telnet connectivity framework skori
2023-10-18 6:33 ` [PATCH v9 03/12] app/graph: support parser utility APIs skori
2023-10-18 6:33 ` [PATCH v9 04/12] app/graph: support mempool command line interfaces skori
2023-10-18 6:33 ` [PATCH v9 05/12] app/graph: support ethdev " skori
2023-10-18 6:33 ` [PATCH v9 06/12] app/graph: support IPv4 lookup " skori
2023-10-18 6:33 ` [PATCH v9 07/12] app/graph: support IPv6 " skori
2023-10-18 6:33 ` [PATCH v9 08/12] app/graph: support neigh " skori
2023-10-18 6:33 ` [PATCH v9 09/12] app/graph: support ethdev Rx " skori
2023-10-18 6:33 ` [PATCH v9 10/12] app/graph: support graph " skori
2023-10-18 6:33 ` [PATCH v9 11/12] app/graph: support CLI option to enable graph stats skori
2023-10-18 6:33 ` [PATCH v9 12/12] app/graph: support l3fwd use case skori
2023-10-18 10:38 ` Jerin Jacob
2023-10-19 10:49 ` [PATCH v10 00/12] add CLI based graph application skori
2023-10-19 10:49 ` [PATCH v10 01/12] app/graph: support application CLI framework skori
2023-10-19 10:49 ` [PATCH v10 02/12] app/graph: support telnet connectivity framework skori
2023-10-19 10:49 ` [PATCH v10 03/12] app/graph: support parser utility APIs skori
2023-10-19 10:49 ` [PATCH v10 04/12] app/graph: support mempool command line interfaces skori
2023-10-19 10:49 ` [PATCH v10 05/12] app/graph: support ethdev " skori
2023-10-19 10:49 ` [PATCH v10 06/12] app/graph: support IPv4 lookup " skori
2023-10-19 10:49 ` [PATCH v10 07/12] app/graph: support IPv6 " skori
2023-10-19 10:49 ` [PATCH v10 08/12] app/graph: support neigh " skori
2023-10-19 10:49 ` [PATCH v10 09/12] app/graph: support ethdev Rx " skori
2023-10-19 10:49 ` [PATCH v10 10/12] app/graph: support graph " skori
2023-10-19 10:49 ` [PATCH v10 11/12] app/graph: support CLI option to enable graph stats skori
2023-10-19 10:50 ` [PATCH v10 12/12] app/graph: support l3fwd use case skori
2023-10-19 12:28 ` [EXT] " Jerin Jacob Kollanukkaran
2023-10-23 7:06 ` Nithin Dabilpuram
2023-10-19 17:29 ` [PATCH v11 00/12] add CLI based graph application skori
2023-10-19 17:30 ` [PATCH v11 01/12] app/graph: support application CLI framework skori
2023-10-23 7:03 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 02/12] app/graph: support telnet connectivity framework skori
2023-10-23 7:03 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 03/12] app/graph: support parser utility APIs skori
2023-10-23 7:03 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 04/12] app/graph: support mempool command line interfaces skori
2023-10-23 7:04 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 05/12] app/graph: support ethdev " skori
2023-10-23 7:04 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 06/12] app/graph: support IPv4 lookup " skori
2023-10-23 7:04 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 07/12] app/graph: support IPv6 " skori
2023-10-23 7:04 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 08/12] app/graph: support neigh " skori
2023-10-23 7:05 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 09/12] app/graph: support ethdev Rx " skori
2023-10-23 7:05 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 10/12] app/graph: support graph " skori
2023-10-23 7:06 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 11/12] app/graph: support CLI option to enable graph stats skori
2023-10-23 7:06 ` Nithin Dabilpuram
2023-10-19 17:30 ` [PATCH v11 12/12] app/graph: support l3fwd use case skori
2023-10-23 13:03 ` [PATCH v11 00/12] add CLI based graph application Sunil Kumar Kori
2023-11-03 16:44 ` Thomas Monjalon
2023-09-29 9:58 ` [PATCH v8 02/12] app/graph: add telnet connectivity framework skori
2023-10-16 9:04 ` Jerin Jacob
2023-10-17 6:21 ` [EXT] " Sunil Kumar Kori
2023-09-29 9:58 ` [PATCH v8 03/12] app/graph: add parser utility APIs skori
2023-10-16 9:05 ` Jerin Jacob
2023-09-29 9:58 ` [PATCH v8 04/12] app/graph: add mempool command line interfaces skori
2023-10-16 9:09 ` Jerin Jacob
2023-10-17 6:22 ` [EXT] " Sunil Kumar Kori
2023-09-29 9:58 ` [PATCH v8 05/12] app/graph: add ethdev " skori
2023-10-16 13:20 ` Jerin Jacob
2023-10-16 14:10 ` Bruce Richardson
2023-10-17 6:26 ` [EXT] " Sunil Kumar Kori
2023-09-29 9:58 ` [PATCH v8 06/12] app/graph: add ipv4_lookup " skori
2023-10-16 15:47 ` Jerin Jacob
2023-10-17 6:30 ` [EXT] " Sunil Kumar Kori
2023-09-29 9:58 ` [PATCH v8 07/12] app/graph: add ipv6_lookup " skori
2023-09-29 9:58 ` [PATCH v8 08/12] app/graph: add neigh " skori
2023-09-29 9:58 ` [PATCH v8 09/12] app/graph: add ethdev_rx " skori
2023-09-29 9:58 ` [PATCH v8 10/12] app/graph: add graph " skori
2023-09-29 9:58 ` [PATCH v8 11/12] app/graph: add CLI option to enable graph stats skori
2023-10-16 15:55 ` Jerin Jacob
2023-09-29 9:58 ` [PATCH v8 12/12] app/graph: add l3fwd use case skori
2023-10-16 16:01 ` [PATCH v8 00/12] add CLI based graph application Jerin Jacob
2023-10-16 16:27 ` Stephen Hemminger
2023-06-02 16:22 ` [PATCH v3 0/3] node: Introduce kernel_rx & kernel_tx nodes Vamsi Attunuru
2023-06-02 16:22 ` [PATCH v3 1/3] node/kernel_tx: support packet transmit to kernel Vamsi Attunuru
2023-06-05 12:47 ` Nithin Dabilpuram
2023-06-12 19:32 ` Thomas Monjalon
2023-06-02 16:22 ` [PATCH v3 2/3] node/kernel_rx: support receiving packets from kernel Vamsi Attunuru
2023-06-05 12:50 ` Nithin Dabilpuram
2023-06-02 16:22 ` [PATCH v3 3/3] node/ethdev_rx: remove hardcoded node next details Vamsi Attunuru
2023-06-05 12:51 ` Nithin Dabilpuram
2023-06-12 16:12 ` [PATCH v3 0/3] node: Introduce kernel_rx & kernel_tx nodes Vamsi Krishna Attunuru
2023-06-12 19:31 ` Thomas Monjalon
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=SN7PR11MB6775BFEE43A4AE9A22CC30E885759@SN7PR11MB6775.namprd11.prod.outlook.com \
--to=zhirun.yan@intel.com \
--cc=cunming.liang@intel.com \
--cc=dev@dpdk.org \
--cc=haiyue.wang@intel.com \
--cc=jerinj@marvell.com \
--cc=ndabilpuram@marvell.com \
--cc=skori@marvell.com \
--cc=thomas@monjalon.net \
--cc=vattunuru@marvell.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).