* [dpdk-dev] [PATCH] app/test-fib: add test-fib application
@ 2019-12-11 18:45 Vladimir Medvedkin
2020-02-16 18:30 ` Thomas Monjalon
0 siblings, 1 reply; 2+ messages in thread
From: Vladimir Medvedkin @ 2019-12-11 18:45 UTC (permalink / raw)
To: dev; +Cc: konstantin.ananyev
Introduce new application to provide user to evaluate and perform
custom functional and performance tests for FIB library.
Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
---
app/test-fib/Makefile | 21 +
app/test-fib/main.c | 1262 ++++++++++++++++++++++++++++++++++++++++++++++
app/test-fib/meson.build | 6 +
3 files changed, 1289 insertions(+)
create mode 100644 app/test-fib/Makefile
create mode 100644 app/test-fib/main.c
create mode 100644 app/test-fib/meson.build
diff --git a/app/test-fib/Makefile b/app/test-fib/Makefile
new file mode 100644
index 0000000..78b45fe
--- /dev/null
+++ b/app/test-fib/Makefile
@@ -0,0 +1,21 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2010-2014 Intel Corporation
+
+include $(RTE_SDK)/mk/rte.vars.mk
+
+ifeq ($(CONFIG_RTE_LIBRTE_FIB),y)
+
+APP = testfib
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+CFLAGS += -DALLOW_EXPERIMENTAL_API
+
+LDFLAGS += -lresolv
+
+# all source are stored in SRCS-y
+SRCS-y := main.c
+
+include $(RTE_SDK)/mk/rte.app.mk
+
+endif
diff --git a/app/test-fib/main.c b/app/test-fib/main.c
new file mode 100644
index 0000000..b9ae0dc
--- /dev/null
+++ b/app/test-fib/main.c
@@ -0,0 +1,1262 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2019 Intel Corporation
+ */
+
+#include <rte_string_fns.h>
+#include <getopt.h>
+#include <string.h>
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <netinet/in.h>
+#include <arpa/inet.h>
+
+#include <rte_cycles.h>
+#include <rte_errno.h>
+#include <rte_ip.h>
+#include <rte_random.h>
+#include <rte_malloc.h>
+#include <rte_lpm.h>
+#include <rte_lpm6.h>
+#include <rte_fib.h>
+#include <rte_fib6.h>
+
+#define PRINT_USAGE_START "%s [EAL options] --\n"
+
+#define GET_CB_FIELD(in, fd, base, lim, dlm) do { \
+ unsigned long val; \
+ char *end_fld; \
+ errno = 0; \
+ val = strtoul((in), &end_fld, (base)); \
+ if (errno != 0 || end_fld[0] != (dlm) || val > (lim)) \
+ return -EINVAL; \
+ (fd) = (typeof(fd))val; \
+ (in) = end_fld + 1; \
+} while (0)
+
+#define DEF_ROUTES_NUM 0x10000
+#define DEF_LOOKUP_IPS_NUM 0x100000
+#define BURST_SZ 64
+#define DEFAULT_LPM_TBL8 100000U
+
+#define CMP_FLAG (1 << 0)
+#define CMP_ALL_FLAG (1 << 1)
+#define IPV6_FLAG (1 << 2)
+#define FIB_RIB_TYPE (1 << 3)
+#define FIB_V4_DIR_TYPE (1 << 4)
+#define FIB_V6_TRIE_TYPE (1 << 4)
+#define FIB_TYPE_MASK (FIB_RIB_TYPE|FIB_V4_DIR_TYPE|FIB_V6_TRIE_TYPE)
+#define SHUFFLE_FLAG (1 << 7)
+#define DRY_RUN_FLAG (1 << 8)
+
+static char *distrib_string;
+static char line[LINE_MAX];
+
+enum {
+ RT_PREFIX,
+ RT_NEXTHOP,
+ RT_NUM
+};
+
+#ifndef NIPQUAD
+#define NIPQUAD_FMT "%u.%u.%u.%u"
+#define NIPQUAD(addr) \
+ (unsigned)((unsigned char *)&addr)[3], \
+ (unsigned)((unsigned char *)&addr)[2], \
+ (unsigned)((unsigned char *)&addr)[1], \
+ (unsigned)((unsigned char *)&addr)[0]
+
+#define NIPQUAD6_FMT "%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x"
+#define NIPQUAD6(addr) \
+ ((uint8_t *)addr)[0] << 8 | \
+ ((uint8_t *)addr)[1], \
+ ((uint8_t *)addr)[2] << 8 | \
+ ((uint8_t *)addr)[3], \
+ ((uint8_t *)addr)[4] << 8 | \
+ ((uint8_t *)addr)[5], \
+ ((uint8_t *)addr)[6] << 8 | \
+ ((uint8_t *)addr)[7], \
+ ((uint8_t *)addr)[8] << 8 | \
+ ((uint8_t *)addr)[9], \
+ ((uint8_t *)addr)[10] << 8 | \
+ ((uint8_t *)addr)[11], \
+ ((uint8_t *)addr)[12] << 8 | \
+ ((uint8_t *)addr)[13], \
+ ((uint8_t *)addr)[14] << 8 | \
+ ((uint8_t *)addr)[15]
+#endif
+
+static struct {
+ const char *prgname;
+ const char *routes_file;
+ const char *lookup_ips_file;
+ const char *routes_file_s;
+ const char *lookup_ips_file_s;
+ void *rt;
+ void *lookup_tbl;
+ uint32_t nb_routes;
+ uint32_t nb_lookup_ips;
+ uint32_t nb_lookup_ips_rnd;
+ uint32_t nb_routes_per_depth[128 + 1];
+ uint32_t flags;
+ uint32_t tbl8;
+ uint8_t ent_sz;
+ uint8_t rnd_lookup_ips_ratio;
+ uint8_t print_fract;
+} config = {
+ .routes_file = NULL,
+ .lookup_ips_file = NULL,
+ .nb_routes = DEF_ROUTES_NUM,
+ .nb_lookup_ips = DEF_LOOKUP_IPS_NUM,
+ .nb_lookup_ips_rnd = 0,
+ .nb_routes_per_depth = {0},
+ .flags = FIB_V4_DIR_TYPE,
+ .tbl8 = DEFAULT_LPM_TBL8,
+ .ent_sz = 4,
+ .rnd_lookup_ips_ratio = 0,
+ .print_fract = 10
+};
+
+struct rt_rule_4 {
+ uint32_t addr;
+ uint8_t depth;
+ uint64_t nh;
+};
+
+struct rt_rule_6 {
+ uint8_t addr[16];
+ uint8_t depth;
+ uint64_t nh;
+};
+
+static uint64_t
+get_rnd_rng(uint64_t l, uint64_t u)
+{
+ if (l == u)
+ return l;
+ else
+ return (rte_rand() % (u - l) + l);
+}
+
+static __rte_always_inline __attribute__((pure)) uint8_t
+bits_in_nh(uint8_t nh_sz)
+{
+ return 8 * (1 << nh_sz);
+}
+
+static __rte_always_inline __attribute__((pure)) uint64_t
+get_max_nh(uint8_t nh_sz)
+{
+ /* min between fib and lpm6 which is 21 bits */
+ return RTE_MIN(((1ULL << (bits_in_nh(nh_sz) - 1)) - 1),
+ (1ULL << 21) - 1);
+}
+
+static int
+get_fib_type(void)
+{
+ if (config.flags & IPV6_FLAG) {
+ if ((config.flags & FIB_TYPE_MASK) == FIB_V6_TRIE_TYPE)
+ return RTE_FIB6_TRIE;
+ else
+ return RTE_FIB6_DUMMY;
+ } else {
+ if ((config.flags & FIB_TYPE_MASK) == FIB_V4_DIR_TYPE)
+ return RTE_FIB_DIR24_8;
+ if ((config.flags & FIB_TYPE_MASK) == FIB_RIB_TYPE)
+ return RTE_FIB_DUMMY;
+ }
+ return -1;
+}
+
+static int
+complete_distrib(uint8_t depth_lim, const uint32_t n, uint8_t rpd[],
+ uint32_t nrpd[])
+{
+ uint8_t depth;
+ uint32_t nr = 0;
+ uint8_t m = 0;
+
+ /*
+ * complete number of routes for every depth
+ * that was configured with ratio
+ */
+ for (depth = 0; depth <= depth_lim; depth++) {
+ if (rpd[depth] != 0) {
+ if (rpd[depth] == UINT8_MAX)
+ config.nb_routes_per_depth[depth] =
+ nrpd[depth];
+ else
+ config.nb_routes_per_depth[depth] =
+ (n * rpd[depth]) / 100;
+
+ nr += config.nb_routes_per_depth[depth];
+ m++;
+ }
+ }
+
+ if (nr > n) {
+ printf("Too much configured routes\n");
+ return -1;
+ }
+
+ /*complete number of routes for every unspecified depths*/
+ for (depth = 0; depth <= depth_lim; depth++) {
+ if (rpd[depth] == 0) {
+ /*we don't need more than two /1 routes*/
+ uint64_t max_routes_per_depth =
+ 1ULL << RTE_MIN(depth, 63);
+ uint32_t avg_routes_left = (n - nr) /
+ (depth_lim + 1 - m++);
+ config.nb_routes_per_depth[depth] =
+ RTE_MIN(max_routes_per_depth, avg_routes_left);
+ nr += config.nb_routes_per_depth[depth];
+ }
+ }
+
+ return 0;
+}
+
+static int
+parse_distrib(uint8_t depth_lim, const uint32_t n)
+{
+ uint8_t rpd[128 + 1] = {0}; /*routes ratios per depth including /0 */
+ uint32_t nrpd[128 + 1] = {0}; /* number of routes per depth */
+ uint32_t n_routes;
+ uint8_t depth, ratio, ratio_acc = 0;
+ char *in;
+
+ in = strtok(distrib_string, ",");
+
+ /*parse configures routes percentage ratios*/
+ while (in != NULL) {
+ GET_CB_FIELD(in, depth, 0, UINT8_MAX, ':');
+ if (in[strlen(in) - 1] == '%') {
+ in[strlen(in) - 1] = 0;
+ GET_CB_FIELD(in, ratio, 0, UINT8_MAX, '\0');
+ if (depth > depth_lim) {
+ printf("Depth /%d is bigger than maximum "
+ "allowed depth /%d for this AF\n",
+ depth, depth_lim);
+ return -EINVAL;
+ }
+ if (ratio > 100) {
+ printf("Ratio for depth /%d is bigger "
+ "than 100%%\n", depth);
+ return -EINVAL;
+ }
+ if ((depth < 64) && ((n * ratio) / 100) >
+ (1ULL << depth)) {
+ printf("Configured ratio %d%% for depth /%d "
+ "has %d different routes, but maximum "
+ "is %lu\n", ratio, depth,
+ ((n * ratio) / 100), (1UL << depth));
+ return -EINVAL;
+ }
+ rpd[depth] = ratio;
+ /*configured zero routes for a given depth*/
+ if (ratio == 0)
+ rpd[depth] = UINT8_MAX;
+ /*sum of all percentage ratios*/
+ ratio_acc += ratio;
+ } else {
+ GET_CB_FIELD(in, n_routes, 0, UINT32_MAX, '\0');
+ rpd[depth] = UINT8_MAX;
+ nrpd[depth] = n_routes;
+ }
+
+ /*number of configured depths in*/
+ in = strtok(NULL, ",");
+ }
+
+ if (ratio_acc > 100) {
+ printf("Total ratio's sum is bigger than 100%%\n");
+ return -EINVAL;
+ }
+
+ return complete_distrib(depth_lim, n, rpd, nrpd);
+}
+
+static void
+shuffle_rt_4(struct rt_rule_4 *rt, int n)
+{
+ struct rt_rule_4 tmp;
+ int i, j;
+
+ for (i = 0; i < n; i++) {
+ j = rte_rand() % n;
+ tmp.addr = rt[i].addr;
+ tmp.depth = rt[i].depth;
+ tmp.nh = rt[i].nh;
+
+ rt[i].addr = rt[j].addr;
+ rt[i].depth = rt[j].depth;
+ rt[i].nh = rt[j].nh;
+
+ rt[j].addr = tmp.addr;
+ rt[j].depth = tmp.depth;
+ rt[j].nh = tmp.nh;
+ }
+}
+
+static void
+shuffle_rt_6(struct rt_rule_6 *rt, int n)
+{
+ struct rt_rule_6 tmp;
+ int i, j;
+
+ for (i = 0; i < n; i++) {
+ j = rte_rand() % n;
+ memcpy(tmp.addr, rt[i].addr, 16);
+ tmp.depth = rt[i].depth;
+ tmp.nh = rt[i].nh;
+
+ memcpy(rt[i].addr, rt[j].addr, 16);
+ rt[i].depth = rt[j].depth;
+ rt[i].nh = rt[j].nh;
+
+ memcpy(rt[j].addr, tmp.addr, 16);
+ rt[j].depth = tmp.depth;
+ rt[j].nh = tmp.nh;
+ }
+}
+
+static void
+gen_random_rt_4(struct rt_rule_4 *rt, int nh_sz)
+{
+ uint32_t i, j, k = 0;
+
+ if (config.nb_routes_per_depth[0] != 0) {
+ rt[k].addr = 0;
+ rt[k].depth = 0;
+ rt[k++].nh = rte_rand() & get_max_nh(nh_sz);
+ }
+
+ for (i = 1; i <= 32; i++) {
+ double edge = 0;
+ double step;
+ step = (double)(1ULL << i) / config.nb_routes_per_depth[i];
+ for (j = 0; j < config.nb_routes_per_depth[i];
+ j++, k++, edge += step) {
+ uint64_t rnd_val = get_rnd_rng((uint64_t)edge,
+ (uint64_t)(edge + step));
+ rt[k].addr = rnd_val << (32 - i);
+ rt[k].depth = i;
+ rt[k].nh = rte_rand() & get_max_nh(nh_sz);
+ }
+ }
+}
+
+static void
+complete_v6_addr(uint32_t *addr, uint32_t rnd, int n)
+{
+ int i;
+
+ for (i = 0; i < n; i++)
+ addr[i] = rte_rand();
+ addr[i++] = rnd;
+ for (; i < 4; i++)
+ addr[i] = 0;
+}
+
+static void
+gen_random_rt_6(struct rt_rule_6 *rt, int nh_sz)
+{
+ uint32_t i, j, k = 0;
+
+ if (config.nb_routes_per_depth[0] != 0) {
+ memset(rt[k].addr, 0, 16);
+ rt[k].depth = 0;
+ rt[k++].nh = rte_rand() & get_max_nh(nh_sz);
+ }
+
+ for (int a = 0; a < 4; a++) {
+ for (i = 1; i <= 32; i++) {
+ uint32_t rnd;
+ double edge = 0;
+ double step = (double)(1ULL << i) /
+ config.nb_routes_per_depth[(a * 32) + i];
+ for (j = 0; j < config.nb_routes_per_depth[a * 32 + i];
+ j++, k++, edge += step) {
+ uint64_t rnd_val = get_rnd_rng((uint64_t)edge,
+ (uint64_t)(edge + step));
+ rnd = rte_cpu_to_be_32(rnd_val << (32 - i));
+ complete_v6_addr((uint32_t *)rt[k].addr,
+ rnd, a);
+ rt[k].depth = (a * 32) + i;
+ rt[k].nh = rte_rand() & get_max_nh(nh_sz);
+ }
+ }
+ }
+}
+
+static inline void
+set_rnd_ipv6(uint8_t *addr, uint8_t *route, int depth)
+{
+ int i;
+
+ for (i = 0; i < 16; i++)
+ addr[i] = rte_rand();
+
+ for (i = 0; i < 16; i++) {
+ if (depth >= 8)
+ addr[i] = route[i];
+ else if (depth > 0) {
+ addr[i] &= (uint16_t)UINT8_MAX >> depth;
+ addr[i] |= route[i] & UINT8_MAX << (8 - depth);
+ } else
+ return;
+ depth -= 8;
+ }
+}
+
+static void
+gen_rnd_lookup_tbl(int af)
+{
+ uint32_t *tbl4 = config.lookup_tbl;
+ uint8_t *tbl6 = config.lookup_tbl;
+ struct rt_rule_4 *rt4 = (struct rt_rule_4 *)config.rt;
+ struct rt_rule_6 *rt6 = (struct rt_rule_6 *)config.rt;
+ uint32_t i, j;
+
+ if (af == AF_INET) {
+ for (i = 0, j = 0; i < config.nb_lookup_ips;
+ i++, j = (j + 1) % config.nb_routes) {
+ if ((rte_rand() % 100) < config.rnd_lookup_ips_ratio) {
+ tbl4[i] = rte_rand();
+ config.nb_lookup_ips_rnd++;
+ } else
+ tbl4[i] = rt4[j].addr | (rte_rand() &
+ ((1ULL << (32 - rt4[j].depth)) - 1));
+ }
+ } else {
+ for (i = 0, j = 0; i < config.nb_lookup_ips;
+ i++, j = (j + 1) % config.nb_routes) {
+ if ((rte_rand() % 100) < config.rnd_lookup_ips_ratio) {
+ set_rnd_ipv6(&tbl6[i * 16], rt6[j].addr, 0);
+ config.nb_lookup_ips_rnd++;
+ } else {
+ set_rnd_ipv6(&tbl6[i * 16], rt6[j].addr,
+ rt6[j].depth);
+ }
+ }
+ }
+}
+
+static int
+parse_rt_4(FILE *f)
+{
+ int ret, i, j = 0;
+ char *s, *sp, *in[RT_NUM];
+ static const char *dlm = " \t\n";
+ int string_tok_nb = RTE_DIM(in);
+ struct rt_rule_4 *rt;
+
+ rt = (struct rt_rule_4 *)config.rt;
+
+ while (fgets(line, sizeof(line), f) != NULL) {
+ s = line;
+ for (i = 0; i != string_tok_nb; i++) {
+ in[i] = strtok_r(s, dlm, &sp);
+ if (in[i] == NULL)
+ return -EINVAL;
+ s = NULL;
+ }
+
+ ret = inet_net_pton(AF_INET, in[RT_PREFIX], &rt[j].addr,
+ sizeof(rt[j].addr));
+ if (ret == -1)
+ return -errno;
+
+ rt[j].addr = rte_be_to_cpu_32(rt[j].addr);
+ rt[j].depth = ret;
+ config.nb_routes_per_depth[ret]++;
+ GET_CB_FIELD(in[RT_NEXTHOP], rt[j].nh, 0,
+ UINT32_MAX, 0);
+ j++;
+ }
+ return 0;
+}
+
+static int
+__inet_net_pton6(char *prefix, uint8_t *addr)
+{
+ const char *dlm = "/";
+ char *s, *sp;
+ int ret, depth;
+
+ if ((prefix == NULL) || (addr == NULL))
+ return -EINVAL;
+
+ s = strtok_r(prefix, dlm, &sp);
+ if (s == NULL)
+ return -EINVAL;
+
+ ret = inet_pton(AF_INET6, s, addr);
+ if (ret != 1)
+ return -errno;
+
+ s = strtok_r(NULL, dlm, &sp);
+ GET_CB_FIELD(s, depth, 0, 128, 0);
+
+ return depth;
+}
+
+static int
+parse_rt_6(FILE *f)
+{
+ int ret, i, j = 0;
+ char *s, *sp, *in[RT_NUM];
+ static const char *dlm = " \t\n";
+ int string_tok_nb = RTE_DIM(in);
+ struct rt_rule_6 *rt;
+
+ rt = (struct rt_rule_6 *)config.rt;
+
+ while (fgets(line, sizeof(line), f) != NULL) {
+ s = line;
+ for (i = 0; i != string_tok_nb; i++) {
+ in[i] = strtok_r(s, dlm, &sp);
+ if (in[i] == NULL)
+ return -EINVAL;
+ s = NULL;
+ }
+
+ ret = __inet_net_pton6(in[RT_PREFIX], rt[j].addr);
+ if (ret < 0)
+ return ret;
+
+ rt[j].depth = ret;
+ config.nb_routes_per_depth[ret]++;
+ GET_CB_FIELD(in[RT_NEXTHOP], rt[j].nh, 0,
+ UINT32_MAX, 0);
+ j++;
+ }
+
+ return 0;
+}
+
+static int
+parse_lookup(FILE *f, int af)
+{
+ int ret, i = 0;
+ uint8_t *tbl = (uint8_t *)config.lookup_tbl;
+ int step = (af == AF_INET) ? 4 : 16;
+ char *s;
+
+ while (fgets(line, sizeof(line), f) != NULL) {
+ s = strtok(line, " \t\n");
+ ret = inet_pton(af, s, &tbl[i]);
+ if (ret != 1)
+ return -EINVAL;
+ i += step;
+ }
+ return 0;
+}
+
+static int
+dump_lookup(int af)
+{
+ FILE *f;
+ uint32_t *tbl4 = config.lookup_tbl;
+ uint8_t *tbl6 = config.lookup_tbl;
+ uint32_t i;
+
+ f = fopen(config.lookup_ips_file_s, "w");
+ if (f == NULL) {
+ printf("Can not open file %s\n", config.lookup_ips_file_s);
+ return -1;
+ }
+
+ if (af == AF_INET) {
+ for (i = 0; i < config.nb_lookup_ips; i++)
+ fprintf(f, NIPQUAD_FMT"\n", NIPQUAD(tbl4[i]));
+ } else {
+ for (i = 0; i < config.nb_lookup_ips; i++)
+ fprintf(f, NIPQUAD6_FMT"\n", NIPQUAD6(&tbl6[i * 16]));
+ }
+ fclose(f);
+ return 0;
+}
+
+static void
+print_config(void)
+{
+ uint8_t depth_lim;
+ char dlm;
+ int i;
+
+ depth_lim = ((config.flags & IPV6_FLAG) == IPV6_FLAG) ? 128 : 32;
+
+ fprintf(stdout,
+ "Routes total: %u\n"
+ "Routes distribution:\n", config.nb_routes);
+
+ for (i = 1; i <= depth_lim; i++) {
+ fprintf(stdout,
+ "depth /%d:%u", i, config.nb_routes_per_depth[i]);
+ if (i % 4 == 0)
+ dlm = '\n';
+ else
+ dlm = '\t';
+ fprintf(stdout, "%c", dlm);
+ }
+
+ fprintf(stdout,
+ "Lookup tuples: %u\n"
+ "Configured ratios of random ips for lookup: %u\n"
+ "Random lookup ips: %u\n",
+ config.nb_lookup_ips, config.rnd_lookup_ips_ratio,
+ config.nb_lookup_ips_rnd);
+}
+
+static void
+print_usage(void)
+{
+ fprintf(stdout,
+ PRINT_USAGE_START
+ "[-f <routes file>]\n"
+ "[-t <ip's file for lookup>]\n"
+ "[-n <number of routes (if -f is not specified)>]\n"
+ "[-l <number of ip's for lookup (if -t is not specified)>]\n"
+ "[-d <\",\" separated \"depth:n%%\"routes depth distribution"
+ "(if -f is not specified)>]\n"
+ "[-r <percentage ratio of random ip's to lookup"
+ "(if -t is not specified)>]\n"
+ "[-c <do comarison with LPM library>]\n"
+ "[-6 <do tests with ipv6 (default ipv4)>]\n"
+ "[-s <shuffle randomly generated routes>]\n"
+ "[-a <check nexthops for all ipv4 address space"
+ "(only valid with -c)>]\n"
+ "[-b <fib algorithm>]\n\tavailible options for ipv4\n"
+ "\t\trib - RIB based FIB\n"
+ "\t\tdir - DIR24_8 based FIB\n"
+ "\tavailible options for ipv6:\n"
+ "\t\trib - RIB based FIB\n"
+ "\t\ttrie - TRIE based FIB\n"
+ "defaults are: dir for ipv4 and trie for ipv6\n"
+ "[-e <entry size (valid only for dir and trie fib types): "
+ "1/2/4/8 (default 4)>]\n"
+ "[-g <number of tbl8's for dir24_8 or trie FIBs>]\n"
+ "[-w <path to the file to dump routing table>]\n"
+ "[-u <path to the file to dump ip's for lookup>]\n",
+ config.prgname);
+}
+
+static int
+check_config(void)
+{
+ if ((config.routes_file == NULL) && (config.lookup_ips_file != NULL)) {
+ printf("-t option only valid with -f option\n");
+ return -1;
+ }
+
+ if ((config.flags & CMP_ALL_FLAG) && (config.flags & IPV6_FLAG)) {
+ printf("-a flag is only valid for ipv4\n");
+ return -1;
+ }
+
+ if ((config.flags & CMP_ALL_FLAG) &&
+ ((config.flags & CMP_FLAG) != CMP_FLAG)) {
+ printf("-a flag is valid only with -c flag\n");
+ return -1;
+ }
+
+ if (!((config.ent_sz == 1) || (config.ent_sz == 2) ||
+ (config.ent_sz == 4) || (config.ent_sz == 8))) {
+ printf("wrong -e option %d, can be 1 or 2 or 4 or 8\n",
+ config.ent_sz);
+ return -1;
+ }
+
+ if ((config.ent_sz == 1) && (config.flags & IPV6_FLAG)) {
+ printf("-e 1 is valid only for ipv4\n");
+ return -1;
+ }
+ return 0;
+}
+
+static void
+parse_opts(int argc, char **argv)
+{
+ int opt;
+ char *endptr;
+
+ while ((opt = getopt(argc, argv, "f:t:n:d:l:r:c6ab:e:g:w:u:s")) !=
+ -1) {
+ switch (opt) {
+ case 'f':
+ config.routes_file = optarg;
+ break;
+ case 't':
+ config.lookup_ips_file = optarg;
+ break;
+ case 'w':
+ config.routes_file_s = optarg;
+ config.flags |= DRY_RUN_FLAG;
+ break;
+ case 'u':
+ config.lookup_ips_file_s = optarg;
+ config.flags |= DRY_RUN_FLAG;
+ break;
+ case 'n':
+ errno = 0;
+ config.nb_routes = strtoul(optarg, &endptr, 10);
+ if ((errno != 0) || (config.nb_routes == 0)) {
+ print_usage();
+ rte_exit(-EINVAL, "Invalid option -n\n");
+ }
+ break;
+ case 'd':
+ distrib_string = optarg;
+ break;
+ case 'l':
+ errno = 0;
+ config.nb_lookup_ips = strtoul(optarg, &endptr, 10);
+ if ((errno != 0) || (config.nb_lookup_ips == 0)) {
+ print_usage();
+ rte_exit(-EINVAL, "Invalid option -l\n");
+ }
+ break;
+ case 'r':
+ errno = 0;
+ config.rnd_lookup_ips_ratio =
+ strtoul(optarg, &endptr, 10);
+ if ((errno != 0) ||
+ (config.rnd_lookup_ips_ratio == 0) ||
+ (config.rnd_lookup_ips_ratio >= 100)) {
+ print_usage();
+ rte_exit(-EINVAL, "Invalid option -r\n");
+ }
+ break;
+ case 's':
+ config.flags |= SHUFFLE_FLAG;
+ break;
+ case 'c':
+ config.flags |= CMP_FLAG;
+ break;
+ case '6':
+ config.flags |= IPV6_FLAG;
+ break;
+ case 'a':
+ config.flags |= CMP_ALL_FLAG;
+ break;
+ case 'b':
+ if (strcmp(optarg, "rib") == 0) {
+ config.flags &= ~FIB_TYPE_MASK;
+ config.flags |= FIB_RIB_TYPE;
+ } else if (strcmp(optarg, "dir") == 0) {
+ config.flags &= ~FIB_TYPE_MASK;
+ config.flags |= FIB_V4_DIR_TYPE;
+ } else if (strcmp(optarg, "trie") == 0) {
+ config.flags &= ~FIB_TYPE_MASK;
+ config.flags |= FIB_V6_TRIE_TYPE;
+ } else
+ rte_exit(-EINVAL, "Invalid option -b\n");
+ break;
+ case 'e':
+ errno = 0;
+ config.ent_sz = strtoul(optarg, &endptr, 10);
+ if (errno != 0) {
+ print_usage();
+ rte_exit(-EINVAL, "Invalid option -e\n");
+ }
+ break;
+ case 'g':
+ errno = 0;
+ config.tbl8 = strtoul(optarg, &endptr, 10);
+ if ((errno != 0) || (config.tbl8 == 0)) {
+ print_usage();
+ rte_exit(-EINVAL, "Invalid option -g\n");
+ }
+ break;
+ default:
+ print_usage();
+ rte_exit(-EINVAL, "Invalid options\n");
+ }
+ }
+}
+
+static int
+dump_rt_4(struct rt_rule_4 *rt)
+{
+ FILE *f;
+ uint32_t i;
+
+ f = fopen(config.routes_file_s, "w");
+ if (f == NULL) {
+ printf("Can not open file %s\n", config.routes_file_s);
+ return -1;
+ }
+
+ for (i = 0; i < config.nb_routes; i++)
+ fprintf(f, NIPQUAD_FMT"/%d %lu\n", NIPQUAD(rt[i].addr),
+ rt[i].depth, rt[i].nh);
+
+ fclose(f);
+ return 0;
+}
+
+static inline void
+print_depth_err(void)
+{
+ printf("LPM does not support /0 prefix length (defaul route), use "
+ "-d 0:0 option or remove /0 prefix from routes file\n");
+}
+
+static int
+run_v4(void)
+{
+ uint64_t start, acc;
+ uint64_t def_nh = 0;
+ struct rte_fib *fib;
+ struct rte_fib_conf conf = {0};
+ struct rt_rule_4 *rt;
+ uint32_t i, j, k;
+ int ret = 0;
+ struct rte_lpm *lpm = NULL;
+ struct rte_lpm_config lpm_conf;
+ uint32_t *tbl4 = config.lookup_tbl;
+ uint64_t fib_nh[BURST_SZ];
+ uint32_t lpm_nh[BURST_SZ];
+
+ rt = (struct rt_rule_4 *)config.rt;
+
+ if (config.flags & DRY_RUN_FLAG) {
+ if (config.routes_file_s != NULL)
+ ret = dump_rt_4(rt);
+ if (ret != 0)
+ return ret;
+ if (config.lookup_ips_file_s != NULL)
+ ret = dump_lookup(AF_INET);
+ return ret;
+ }
+
+ conf.type = get_fib_type();
+ conf.default_nh = def_nh;
+ conf.max_routes = config.nb_routes * 2;
+ if (conf.type == RTE_FIB_DIR24_8) {
+ conf.dir24_8.nh_sz = __builtin_ctz(config.ent_sz);
+ conf.dir24_8.num_tbl8 = RTE_MIN(config.tbl8,
+ get_max_nh(conf.dir24_8.nh_sz));
+ }
+
+ fib = rte_fib_create("test", -1, &conf);
+ if (fib == NULL) {
+ printf("Can not alloc FIB, err %d\n", rte_errno);
+ return -rte_errno;
+ }
+
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++) {
+ ret = rte_fib_add(fib, rt[i + j].addr, rt[i + j].depth,
+ rt[i + j].nh);
+ if (unlikely(ret != 0)) {
+ printf("Can not add a route to FIB, err %d\n",
+ ret);
+ return -ret;
+ }
+ }
+ printf("AVG FIB add %lu\n", (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+
+ if (config.flags & CMP_FLAG) {
+ lpm_conf.max_rules = config.nb_routes * 2;
+ lpm_conf.number_tbl8s = RTE_MAX(conf.dir24_8.num_tbl8,
+ config.tbl8);
+
+ lpm = rte_lpm_create("test_lpm", -1, &lpm_conf);
+ if (lpm == NULL) {
+ printf("Can not alloc LPM, err %d\n", rte_errno);
+ return -rte_errno;
+ }
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++) {
+ ret = rte_lpm_add(lpm, rt[i + j].addr,
+ rt[i + j].depth, rt[i + j].nh);
+ if (ret != 0) {
+ if (rt[i + j].depth == 0)
+ print_depth_err();
+ printf("Can not add a route to LPM, "
+ "err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG LPM add %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+ }
+
+ acc = 0;
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ start = rte_rdtsc_precise();
+ ret = rte_fib_lookup_bulk(fib, tbl4 + i, fib_nh, BURST_SZ);
+ acc += rte_rdtsc_precise() - start;
+ if (ret != 0) {
+ printf("FIB lookup fails, err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG FIB lookup %.1f\n", (double)acc / (double)i);
+
+ if (config.flags & CMP_FLAG) {
+ acc = 0;
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ start = rte_rdtsc_precise();
+ ret = rte_lpm_lookup_bulk(lpm, tbl4 + i, lpm_nh,
+ BURST_SZ);
+ acc += rte_rdtsc_precise() - start;
+ if (ret != 0) {
+ printf("LPM lookup fails, err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG LPM lookup %.1f\n", (double)acc / (double)i);
+
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ rte_fib_lookup_bulk(fib, tbl4 + i, fib_nh, BURST_SZ);
+ rte_lpm_lookup_bulk(lpm, tbl4 + i, lpm_nh, BURST_SZ);
+ for (j = 0; j < BURST_SZ; j++) {
+ struct rte_lpm_tbl_entry *tbl;
+ tbl = (struct rte_lpm_tbl_entry *)&lpm_nh[j];
+ if ((fib_nh[j] != tbl->next_hop) &&
+ !((tbl->valid == 0) &&
+ (fib_nh[j] == def_nh))) {
+ printf("FAIL\n");
+ return -1;
+ }
+ }
+ }
+ printf("FIB and LPM lookup returns same values\n");
+ }
+
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++)
+ rte_fib_delete(fib, rt[i + j].addr, rt[i + j].depth);
+
+ printf("AVG FIB delete %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+
+ if (config.flags & CMP_FLAG) {
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++)
+ rte_lpm_delete(lpm, rt[i + j].addr,
+ rt[i + j].depth);
+
+ printf("AVG LPM delete %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+ }
+
+ return 0;
+}
+
+static int
+dump_rt_6(struct rt_rule_6 *rt)
+{
+ FILE *f;
+ uint32_t i;
+
+ f = fopen(config.routes_file_s, "w");
+ if (f == NULL) {
+ printf("Can not open file %s\n", config.routes_file_s);
+ return -1;
+ }
+
+ for (i = 0; i < config.nb_routes; i++) {
+ fprintf(f, NIPQUAD6_FMT"/%d %lu\n", NIPQUAD6(rt[i].addr),
+ rt[i].depth, rt[i].nh);
+
+ }
+ fclose(f);
+ return 0;
+}
+
+static int
+run_v6(void)
+{
+ uint64_t start, acc;
+ uint64_t def_nh = 0;
+ struct rte_fib6 *fib;
+ struct rte_fib6_conf conf = {0};
+ struct rt_rule_6 *rt;
+ uint32_t i, j, k;
+ int ret = 0;
+ struct rte_lpm6 *lpm = NULL;
+ struct rte_lpm6_config lpm_conf;
+ uint8_t *tbl6;
+ uint64_t fib_nh[BURST_SZ];
+ int32_t lpm_nh[BURST_SZ];
+
+ rt = (struct rt_rule_6 *)config.rt;
+ tbl6 = config.lookup_tbl;
+
+ if (config.flags & DRY_RUN_FLAG) {
+ if (config.routes_file_s != NULL)
+ ret = dump_rt_6(rt);
+ if (ret != 0)
+ return ret;
+ if (config.lookup_ips_file_s != NULL)
+ ret = dump_lookup(AF_INET6);
+ return ret;
+ }
+
+ conf.type = get_fib_type();
+ conf.default_nh = def_nh;
+ conf.max_routes = config.nb_routes * 2;
+ if (conf.type == RTE_FIB6_TRIE) {
+ conf.trie.nh_sz = __builtin_ctz(config.ent_sz);
+ conf.trie.num_tbl8 = RTE_MIN(config.tbl8,
+ get_max_nh(conf.trie.nh_sz));
+ }
+
+ fib = rte_fib6_create("test", -1, &conf);
+ if (fib == NULL) {
+ printf("Can not alloc FIB, err %d\n", rte_errno);
+ return -rte_errno;
+ }
+
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++) {
+ ret = rte_fib6_add(fib, rt[i + j].addr,
+ rt[i + j].depth, rt[i + j].nh);
+ if (unlikely(ret != 0)) {
+ printf("Can not add a route to FIB, err %d\n",
+ ret);
+ return -ret;
+ }
+ }
+ printf("AVG FIB add %lu\n", (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+
+ if (config.flags & CMP_FLAG) {
+ lpm_conf.max_rules = config.nb_routes * 2;
+ lpm_conf.number_tbl8s = RTE_MAX(conf.trie.num_tbl8,
+ config.tbl8);
+
+ lpm = rte_lpm6_create("test_lpm", -1, &lpm_conf);
+ if (lpm == NULL) {
+ printf("Can not alloc LPM, err %d\n", rte_errno);
+ return -rte_errno;
+ }
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++) {
+ ret = rte_lpm6_add(lpm, rt[i + j].addr,
+ rt[i + j].depth, rt[i + j].nh);
+ if (ret != 0) {
+ if (rt[i + j].depth == 0)
+ print_depth_err();
+ printf("Can not add a route to LPM, "
+ "err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG LPM add %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+ }
+
+ acc = 0;
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ start = rte_rdtsc_precise();
+ ret = rte_fib6_lookup_bulk(fib, (uint8_t (*)[16])(tbl6 + i*16),
+ fib_nh, BURST_SZ);
+ acc += rte_rdtsc_precise() - start;
+ if (ret != 0) {
+ printf("FIB lookup fails, err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG FIB lookup %.1f\n", (double)acc / (double)i);
+
+ if (config.flags & CMP_FLAG) {
+ acc = 0;
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ start = rte_rdtsc_precise();
+ ret = rte_lpm6_lookup_bulk_func(lpm,
+ (uint8_t (*)[16])(tbl6 + i*16),
+ lpm_nh, BURST_SZ);
+ acc += rte_rdtsc_precise() - start;
+ if (ret != 0) {
+ printf("LPM lookup fails, err %d\n", ret);
+ return -ret;
+ }
+ }
+ printf("AVG LPM lookup %.1f\n", (double)acc / (double)i);
+
+ for (i = 0; i < config.nb_lookup_ips; i += BURST_SZ) {
+ rte_fib6_lookup_bulk(fib,
+ (uint8_t (*)[16])(tbl6 + i*16),
+ fib_nh, BURST_SZ);
+ rte_lpm6_lookup_bulk_func(lpm,
+ (uint8_t (*)[16])(tbl6 + i*16),
+ lpm_nh, BURST_SZ);
+ for (j = 0; j < BURST_SZ; j++) {
+ if ((fib_nh[j] != (uint32_t)lpm_nh[j]) &&
+ !((lpm_nh[j] == -1) &&
+ (fib_nh[j] == def_nh))) {
+ printf("FAIL\n");
+ return -1;
+ }
+ }
+ }
+ printf("FIB and LPM lookup returns same values\n");
+ }
+
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++)
+ rte_fib6_delete(fib, rt[i + j].addr, rt[i + j].depth);
+
+ printf("AVG FIB delete %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+
+ if (config.flags & CMP_FLAG) {
+ for (k = config.print_fract, i = 0; k > 0; k--) {
+ start = rte_rdtsc_precise();
+ for (j = 0; j < (config.nb_routes - i) / k; j++)
+ rte_lpm6_delete(lpm, rt[i + j].addr,
+ rt[i + j].depth);
+
+ printf("AVG LPM delete %lu\n",
+ (rte_rdtsc_precise() - start) / j);
+ i += j;
+ }
+ }
+ return 0;
+}
+
+int
+main(int argc, char **argv)
+{
+ int ret, af, rt_ent_sz, lookup_ent_sz;
+ FILE *fr, *fl;
+ uint8_t depth_lim;
+
+ ret = rte_eal_init(argc, argv);
+ if (ret < 0)
+ rte_panic("Cannot init EAL\n");
+
+ argc -= ret;
+ argv += ret;
+
+ config.prgname = argv[0];
+
+ parse_opts(argc, argv);
+
+ ret = check_config();
+ if (ret != 0)
+ rte_exit(-ret, "Bad configuration\n");
+
+ af = ((config.flags & IPV6_FLAG) == 0) ? AF_INET : AF_INET6;
+ depth_lim = (af == AF_INET) ? 32 : 128;
+ rt_ent_sz = (af == AF_INET) ? sizeof(struct rt_rule_4) :
+ sizeof(struct rt_rule_6);
+ lookup_ent_sz = (af == AF_INET) ? 4 : 16;
+
+ /* Count number of rules in file*/
+ if (config.routes_file != NULL) {
+ fr = fopen(config.routes_file, "r");
+ if (fr == NULL)
+ rte_exit(-errno, "Can not open file with routes %s\n",
+ config.routes_file);
+
+ config.nb_routes = 0;
+ while (fgets(line, sizeof(line), fr) != NULL)
+ config.nb_routes++;
+ rewind(fr);
+ }
+
+ /* Count number of ip's in file*/
+ if (config.lookup_ips_file != NULL) {
+ fl = fopen(config.lookup_ips_file, "r");
+ if (fl == NULL)
+ rte_exit(-errno, "Can not open file with ip's %s\n",
+ config.lookup_ips_file);
+
+ config.nb_lookup_ips = 0;
+ while (fgets(line, sizeof(line), fl) != NULL)
+ config.nb_lookup_ips++;
+ rewind(fl);
+ }
+
+ /* Alloc routes table*/
+ config.rt = rte_malloc(NULL, rt_ent_sz * config.nb_routes, 0);
+ if (config.rt == NULL)
+ rte_exit(-ENOMEM, "Can not alloc rt\n");
+
+ /* Alloc table with ip's for lookup*/
+ config.lookup_tbl = rte_malloc(NULL, lookup_ent_sz *
+ config.nb_lookup_ips, 0);
+ if (config.lookup_tbl == NULL)
+ rte_exit(-ENOMEM, "Can not alloc lookup table\n");
+
+ /* Fill routes table */
+ if (config.routes_file == NULL) {
+ if (distrib_string != NULL)
+ ret = parse_distrib(depth_lim, config.nb_routes);
+ else {
+ uint8_t rpd[129] = {0};
+ uint32_t nrpd[129] = {0};
+ ret = complete_distrib(depth_lim, config.nb_routes,
+ rpd, nrpd);
+ }
+ if (ret != 0)
+ rte_exit(-ret,
+ "Bad routes distribution configuration\n");
+ if (af == AF_INET) {
+ gen_random_rt_4(config.rt,
+ __builtin_ctz(config.ent_sz));
+ if (config.flags & SHUFFLE_FLAG)
+ shuffle_rt_4(config.rt, config.nb_routes);
+ } else {
+ gen_random_rt_6(config.rt,
+ __builtin_ctz(config.ent_sz));
+ if (config.flags & SHUFFLE_FLAG)
+ shuffle_rt_6(config.rt, config.nb_routes);
+ }
+ } else {
+ if (af == AF_INET)
+ ret = parse_rt_4(fr);
+ else
+ ret = parse_rt_6(fr);
+
+ if (ret != 0) {
+ rte_exit(-ret, "failed to parse routes file %s\n",
+ config.routes_file);
+ }
+ }
+
+ /* Fill lookup table with ip's*/
+ if (config.lookup_ips_file == NULL)
+ gen_rnd_lookup_tbl(af);
+ else {
+ ret = parse_lookup(fl, af);
+ if (ret != 0)
+ rte_exit(-ret, "failed to parse lookup file\n");
+ }
+
+ print_config();
+
+ if (af == AF_INET)
+ ret = run_v4();
+ else
+ ret = run_v6();
+
+ return ret;
+}
diff --git a/app/test-fib/meson.build b/app/test-fib/meson.build
new file mode 100644
index 0000000..4db0010
--- /dev/null
+++ b/app/test-fib/meson.build
@@ -0,0 +1,6 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2019 Intel Corporation
+
+allow_experimental_apis = true
+sources = files('main.c')
+deps += ['fib', 'lpm', 'net']
--
2.7.4
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [dpdk-dev] [PATCH] app/test-fib: add test-fib application
2019-12-11 18:45 [dpdk-dev] [PATCH] app/test-fib: add test-fib application Vladimir Medvedkin
@ 2020-02-16 18:30 ` Thomas Monjalon
0 siblings, 0 replies; 2+ messages in thread
From: Thomas Monjalon @ 2020-02-16 18:30 UTC (permalink / raw)
To: Vladimir Medvedkin; +Cc: dev, konstantin.ananyev
11/12/2019 19:45, Vladimir Medvedkin:
> Introduce new application to provide user to evaluate and perform
> custom functional and performance tests for FIB library.
>
> Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
> ---
> app/test-fib/Makefile | 21 +
> app/test-fib/main.c | 1262 ++++++++++++++++++++++++++++++++++++++++++++++
> app/test-fib/meson.build | 6 +
> 3 files changed, 1289 insertions(+)
Added files in MAINTAINERS.
Nobody reviewed this app during the release cycle unfortunately,
but I assume it's fined.
Applied, thanks
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-02-16 18:30 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-12-11 18:45 [dpdk-dev] [PATCH] app/test-fib: add test-fib application Vladimir Medvedkin
2020-02-16 18:30 ` Thomas Monjalon
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).