From bdcb53df8ca1e24732204d519aa87888b603c280 Mon Sep 17 00:00:00 2001 From: Kamil Wiatrowski Date: Tue, 8 Aug 2017 15:04:46 +0100 Subject: [PATCH] intel_pmu plugin: implement core groups option Having this option will allow to configure monitoring of PMU events only on specific CPUs thus decrease number of opened descriptors significantly and avoid 'Too many open files' errors. New "Cores" option is similar to that available in intel_rdt plugin. New tool utils_config_cores is added to parse "Cores" option. Change-Id: I8f792e1f2560c4cf19aee101fdb07c925d682778 Signed-off-by: Kamil Wiatrowski --- Makefile.am | 12 +- src/collectd.conf.in | 1 + src/collectd.conf.pod | 18 ++ src/intel_pmu.c | 187 +++++++++++++++------ src/utils_config_cores.c | 371 ++++++++++++++++++++++++++++++++++++++++++ src/utils_config_cores.h | 131 +++++++++++++++ src/utils_config_cores_test.c | 257 +++++++++++++++++++++++++++++ 7 files changed, 929 insertions(+), 48 deletions(-) create mode 100644 src/utils_config_cores.c create mode 100644 src/utils_config_cores.h create mode 100644 src/utils_config_cores_test.c diff --git a/Makefile.am b/Makefile.am index 91e4b46b..2547a4ed 100644 --- a/Makefile.am +++ b/Makefile.am @@ -141,7 +141,8 @@ check_PROGRAMS = \ test_utils_subst \ test_utils_time \ test_utils_vl_lookup \ - test_libcollectd_network_parse + test_libcollectd_network_parse \ + test_utils_config_cores TESTS = $(check_PROGRAMS) @@ -326,6 +327,11 @@ test_utils_subst_SOURCES = \ src/daemon/utils_subst.h test_utils_subst_LDADD = libplugin_mock.la +test_utils_config_cores_SOURCES = \ + src/utils_config_cores_test.c \ + src/testing.h +test_utils_config_cores_LDADD = libplugin_mock.la + libavltree_la_SOURCES = \ src/daemon/utils_avltree.c \ src/daemon/utils_avltree.h @@ -913,7 +919,9 @@ endif if BUILD_PLUGIN_INTEL_PMU pkglib_LTLIBRARIES += intel_pmu.la -intel_pmu_la_SOURCES = src/intel_pmu.c +intel_pmu_la_SOURCES = src/intel_pmu.c \ + src/utils_config_cores.h \ + src/utils_config_cores.c intel_pmu_la_CPPFLAGS = $(AM_CPPFLAGS) $(BUILD_WITH_LIBJEVENTS_CPPFLAGS) intel_pmu_la_LDFLAGS = $(PLUGIN_LDFLAGS) $(BUILD_WITH_LIBJEVENTS_LDFLAGS) intel_pmu_la_LIBADD = $(BUILD_WITH_LIBJEVENTS_LIBS) diff --git a/src/collectd.conf.in b/src/collectd.conf.in index 4efa29e4..5eb17c4d 100644 --- a/src/collectd.conf.in +++ b/src/collectd.conf.in @@ -665,6 +665,7 @@ # ReportSoftwareEvents true # EventList "/var/cache/pmu/GenuineIntel-6-2D-core.json" # HardwareEvents "L2_RQSTS.CODE_RD_HIT,L2_RQSTS.CODE_RD_MISS" "L2_RQSTS.ALL_CODE_RD" +# Cores "[0-3]" # # diff --git a/src/collectd.conf.pod b/src/collectd.conf.pod index 0e7a6046..880cb997 100644 --- a/src/collectd.conf.pod +++ b/src/collectd.conf.pod @@ -3175,6 +3175,7 @@ B ReportSoftwareEvents true EventList "/var/cache/pmu/GenuineIntel-6-2D-core.json" HardwareEvents "L2_RQSTS.CODE_RD_HIT,L2_RQSTS.CODE_RD_MISS" "L2_RQSTS.ALL_CODE_RD" + Cores "0-3" "4,6" "[12-15]" B @@ -3246,6 +3247,23 @@ event_download.py script to download event list for current CPU. This field is a list of event names or groups of comma separated event names. This option requires B option to be configured. +=item B I + +All events are reported on a per core basis. Monitoring of the events can be +configured for a group of cores (aggregated statistics). This field defines +groups of cores on which to monitor supported events. The field is represented +as list of strings with core group values. Each string represents a list of +cores in a group. If a group is enclosed in square brackets each core is added +individually to a separate group (that is statistics are not aggregated). +Allowed formats are: + 0,1,2,3 + 0-10,20-18 + 1,3,5-8,10,0x10-12 + [4-15,32-63] + +If an empty string is provided as value for this field default cores +configuration is applied - that is separate group is created for each core. + =back =head2 Plugin C diff --git a/src/intel_pmu.c b/src/intel_pmu.c index e0d796f9..a7ed7baf 100644 --- a/src/intel_pmu.c +++ b/src/intel_pmu.c @@ -23,11 +23,14 @@ * * Authors: * Serhiy Pshyk + * Kamil Wiatrowski **/ #include "collectd.h" #include "common.h" +#include "utils_config_cores.h" + #include #include @@ -70,6 +73,7 @@ struct intel_pmu_ctx_s { char event_list_fn[PATH_MAX]; char **hw_events; size_t hw_events_count; + core_groups_list_t cores; struct eventlist *event_list; }; typedef struct intel_pmu_ctx_s intel_pmu_ctx_t; @@ -196,8 +200,57 @@ static void pmu_dump_config(void) { } } +static void pmu_dump_cgroups(void) { + + DEBUG(PMU_PLUGIN ": Core groups:"); + + for (size_t i = 0; i < g_ctx.cores.num_cgroups; i++) { + core_group_t *cgroup = g_ctx.cores.cgroups + i; + const size_t cores_size = cgroup->num_cores * 4 + 1; + char *cores = calloc(cores_size, sizeof(*cores)); + if (cores == NULL) { + DEBUG(PMU_PLUGIN ": Failed to allocate string to list cores."); + return; + } + for (size_t j = 0; j < cgroup->num_cores; j++) + snprintf(cores + strlen(cores), cores_size - strlen(cores), " %d", + cgroup->cores[j]); + + DEBUG(PMU_PLUGIN ": group[%zu]", i); + DEBUG(PMU_PLUGIN ": description: %s", cgroup->desc); + DEBUG(PMU_PLUGIN ": cores count: %zu", cgroup->num_cores); + DEBUG(PMU_PLUGIN ": cores :%s", cores); + sfree(cores); + } +} + #endif /* COLLECT_DEBUG */ +static int pmu_validate_cgroups(core_group_t *cgroups, size_t len, + int max_cores) { + /* i - group index, j - core index */ + for (size_t i = 0; i < len; i++) { + for (size_t j = 0; j < cgroups[i].num_cores; j++) { + int core = (int) cgroups[i].cores[j]; + + /* Core index cannot exceed number of cores in system, + note that max_cores include both online and offline CPUs. */ + if (core >= max_cores) { + ERROR(PMU_PLUGIN ": Core %d is not valid, max core index: %d.", core, + max_cores - 1); + return -1; + } + } + /* Check if cores are set in remaining groups */ + for (size_t k = i+1; k < len; k++) + if (config_cores_cmp_cgroups(&cgroups[i], &cgroups[k]) != 0) { + ERROR(PMU_PLUGIN ": Same cores cannot be set in different groups."); + return -1; + } + } + return 0; +} + static int pmu_config_hw_events(oconfig_item_t *ci) { if (strcasecmp("HardwareEvents", ci->key) != 0) { @@ -252,6 +305,8 @@ static int pmu_config(oconfig_item_t *ci) { ret = pmu_config_hw_events(child); } else if (strcasecmp("ReportSoftwareEvents", child->key) == 0) { ret = cf_util_get_boolean(child, &g_ctx.sw_events); + } else if (strcasecmp("Cores", child->key) == 0) { + ret = config_cores_parse(child, &g_ctx.cores); } else { ERROR(PMU_PLUGIN ": Unknown configuration parameter \"%s\".", child->key); ret = -1; @@ -270,7 +325,7 @@ static int pmu_config(oconfig_item_t *ci) { return 0; } -static void pmu_submit_counter(int cpu, char *event, counter_t value, +static void pmu_submit_counter(char *cgroup, char *event, counter_t value, meta_data_t *meta) { value_list_t vl = VALUE_LIST_INIT; @@ -278,12 +333,9 @@ static void pmu_submit_counter(int cpu, char *event, counter_t value, vl.values_len = 1; sstrncpy(vl.plugin, PMU_PLUGIN, sizeof(vl.plugin)); - if (cpu == -1) { - snprintf(vl.plugin_instance, sizeof(vl.plugin_instance), "all"); - } else { + sstrncpy(vl.plugin_instance, cgroup, sizeof(vl.plugin_instance)); + if (meta) vl.meta = meta; - snprintf(vl.plugin_instance, sizeof(vl.plugin_instance), "%d", cpu); - } sstrncpy(vl.type, "counter", sizeof(vl.type)); sstrncpy(vl.type_instance, event, sizeof(vl.type_instance)); @@ -316,49 +368,65 @@ static void pmu_dispatch_data(void) { struct event *e; for (e = g_ctx.event_list->eventlist; e; e = e->next) { - uint64_t all_value = 0; - int event_enabled = 0; - for (int i = 0; i < g_ctx.event_list->num_cpus; i++) { - - if (e->efd[i].fd < 0) - continue; - - event_enabled++; - - /* If there are more events than counters, the kernel uses time - * multiplexing. With multiplexing, at the end of the run, - * the counter is scaled basing on total time enabled vs time running. - * final_count = raw_count * time_enabled/time_running - */ - uint64_t value = event_scaled_value(e, i); - all_value += value; - - /* get meta data with information about scaling */ - meta_data_t *meta = pmu_meta_data_create(&e->efd[i]); - - /* dispatch per CPU value */ - pmu_submit_counter(i, e->event, value, meta); - - meta_data_destroy(meta); - } + for (size_t i = 0; i < g_ctx.cores.num_cgroups; i++) { + core_group_t *cgroup = g_ctx.cores.cgroups + i; + uint64_t cgroup_value = 0; + int event_enabled_cgroup = 0; + meta_data_t *meta = NULL; + + for (size_t j = 0; j < cgroup->num_cores; j++) { + int core = (int) cgroup->cores[j]; + if (e->efd[core].fd < 0) + continue; + + event_enabled_cgroup++; + + /* If there are more events than counters, the kernel uses time + * multiplexing. With multiplexing, at the end of the run, + * the counter is scaled basing on total time enabled vs time running. + * final_count = raw_count * time_enabled/time_running + */ + uint64_t value = event_scaled_value(e, core); + cgroup_value += value; + + /* get meta data with information about scaling */ + if (cgroup->num_cores == 1) + meta = pmu_meta_data_create(&e->efd[core]); + } - if (event_enabled > 0) { - DEBUG(PMU_PLUGIN ": %-20s %'10lu", e->event, all_value); - /* dispatch all CPU value */ - pmu_submit_counter(-1, e->event, all_value, NULL); + if (event_enabled_cgroup > 0) { + DEBUG(PMU_PLUGIN ": %s/%s = %lu", e->event, cgroup->desc, cgroup_value); + /* dispatch per core group value */ + pmu_submit_counter(cgroup->desc, e->event, cgroup_value, meta); + meta_data_destroy(meta); + } } } } static int pmu_read(__attribute__((unused)) user_data_t *ud) { int ret; + struct event *e; DEBUG(PMU_PLUGIN ": %s:%d", __FUNCTION__, __LINE__); - ret = read_all_events(g_ctx.event_list); - if (ret != 0) { - ERROR(PMU_PLUGIN ": Failed to read values of all events."); - return ret; + /* read all events only for configured cores */ + for (e = g_ctx.event_list->eventlist; e; e = e->next) { + for (size_t i = 0; i < g_ctx.cores.num_cgroups; i++) { + core_group_t *cgroup = g_ctx.cores.cgroups + i; + for (size_t j = 0; j < cgroup->num_cores; j++) { + int core = (int) cgroup->cores[j]; + if (e->efd[core].fd < 0) + continue; + + ret = read_event(e, core); + if (ret != 0) { + ERROR(PMU_PLUGIN ": Failed to read value of %s/%d event.", e->event, + core); + return ret; + } + } + } } pmu_dispatch_data(); @@ -473,13 +541,18 @@ static int pmu_setup_events(struct eventlist *el, bool measure_all, for (e = el->eventlist; e; e = e->next) { - for (int i = 0; i < el->num_cpus; i++) { - if (setup_event(e, i, leader, measure_all, measure_pid) < 0) { - WARNING(PMU_PLUGIN ": perf event '%s' is not available (cpu=%d).", - e->event, i); - } else { - /* success if at least one event was set */ - ret = 0; + for (size_t i = 0; i < g_ctx.cores.num_cgroups; i++) { + core_group_t *cgroup = g_ctx.cores.cgroups + i; + for (size_t j = 0; j < cgroup->num_cores; j++) { + int core = (int) cgroup->cores[j]; + + if (setup_event(e, core, leader, measure_all, measure_pid) < 0) { + WARNING(PMU_PLUGIN ": perf event '%s' is not available (cpu=%d).", + e->event, core); + } else { + /* success if at least one event was set */ + ret = 0; + } } } @@ -503,6 +576,24 @@ static int pmu_init(void) { return -ENOMEM; } + if (g_ctx.cores.num_cgroups == 0) { + ret = config_cores_default(g_ctx.event_list->num_cpus, &g_ctx.cores); + if (ret != 0) { + ERROR(PMU_PLUGIN ": Failed to set default core groups."); + goto init_error; + } + } else { + ret = pmu_validate_cgroups(g_ctx.cores.cgroups, g_ctx.cores.num_cgroups, + g_ctx.event_list->num_cpus); + if (ret != 0) { + ERROR(PMU_PLUGIN ": Invalid core groups configuration."); + goto init_error; + } + } +#if COLLECT_DEBUG + pmu_dump_cgroups(); +#endif + if (g_ctx.hw_cache_events) { ret = pmu_add_events(g_ctx.event_list, PERF_TYPE_HW_CACHE, g_hw_cache_events, @@ -578,6 +669,8 @@ init_error: sfree(g_ctx.hw_events); g_ctx.hw_events_count = 0; + config_cores_cleanup(&g_ctx.cores); + return ret; } @@ -593,6 +686,8 @@ static int pmu_shutdown(void) { sfree(g_ctx.hw_events); g_ctx.hw_events_count = 0; + config_cores_cleanup(&g_ctx.cores); + return 0; } diff --git a/src/utils_config_cores.c b/src/utils_config_cores.c new file mode 100644 index 00000000..1daa76db --- /dev/null +++ b/src/utils_config_cores.c @@ -0,0 +1,371 @@ +/** + * collectd - src/utils_config_cores.c + * + * Copyright(c) 2017 Intel Corporation. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + * Authors: + * Kamil Wiatrowski + **/ + +#include "collectd.h" + +#include "common.h" + +#include "utils_config_cores.h" + +#define UTIL_NAME "utils_config_cores" + +#define MAX_SOCKETS 8 +#define MAX_SOCKET_CORES 64 +#define MAX_CORES (MAX_SOCKET_CORES * MAX_SOCKETS) + +static inline _Bool is_in_list(unsigned val, const unsigned *list, size_t len) { + for (size_t i = 0; i < len; i++) + if (list[i] == val) + return 1; + return 0; +} + +static int str_to_uint(const char *s, unsigned *n) { + if (s == NULL || n == NULL) + return -EINVAL; + char *endptr = NULL; + + *n = (unsigned) strtoul(s, &endptr, 0); + if (*s == '\0' || *endptr != '\0') { + ERROR(UTIL_NAME ": Failed to parse '%s' into unsigned number", s); + return -EINVAL; + } + + return 0; +} + +/* + * NAME + * str_list_to_nums + * + * DESCRIPTION + * Converts string of characters representing list of numbers into array of + * numbers. Allowed formats are: + * 0,1,2,3 + * 0-10,20-18 + * 1,3,5-8,10,0x10-12 + * + * Numbers can be in decimal or hexadecimal format. + * + * PARAMETERS + * `s' String representing list of unsigned numbers. + * `nums' Array to put converted numeric values into. + * `nums_len' Maximum number of elements that nums can accommodate. + * + * RETURN VALUE + * Number of elements placed into nums. + */ +static size_t str_list_to_nums(char *s, unsigned *nums, size_t nums_len) { + char *saveptr; + char *token; + size_t idx = 0; + + while((token = strtok_r(s, ",", &saveptr))) { + char *pos; + unsigned start, end = 0; + s = NULL; + + while (isspace(*token)) + token++; + if (*token == '\0') + continue; + + pos = strchr(token, '-'); + if (pos) { + *pos = '\0'; + } + + if (str_to_uint(token, &start)) + return 0; + + if (pos) { + if (str_to_uint(pos + 1, &end)) + return 0; + } else { + end = start; + } + + if (start > end) { + unsigned swap = start; + start = end; + end = swap; + } + + for (unsigned i = start; i <= end; i++) { + if (is_in_list(i, nums, idx)) + continue; + if (idx >= nums_len) { + WARNING(UTIL_NAME ": exceeded the cores number limit: %zu", nums_len); + return idx; + } + nums[idx] = i; + idx++; + } + } + return idx; +} + +/* + * NAME + * check_core_grouping + * + * DESCRIPTION + * Look for [...] brackets in *in string and if found copy the + * part between brackets into *out string and set grouped to 0. + * Otherwise grouped is set to 1 and input is copied without leading + * whitespaces. + * + * PARAMETERS + * `out' Output string to store result. + * `in' Input string to be parsed and copied. + * `out_size' Maximum number of elements that out can accommodate. + * `grouped' Set by function depending if cores should be grouped or not. + * + * RETURN VALUE + * Zero upon success or non-zero if an error occurred. + */ +static int check_core_grouping(char *out, const char *in, size_t out_size, + _Bool *grouped) { + const char *start = in; + char *end; + while (isspace(*start)) + ++start; + if (start[0] == '[') { + *grouped = 0; + ++start; + end = strchr(start, ']'); + if (end == NULL) { + ERROR(UTIL_NAME ": Missing closing bracket ] in option %s.", in); + return -EINVAL; + } + if ((end - start) >= out_size) { + ERROR(UTIL_NAME ": Out buffer is too small."); + return -EINVAL; + } + sstrncpy(out, start, end - start + 1); + DEBUG(UTIL_NAME ": Mask for individual (not aggregated) cores: %s", out); + } else { + *grouped = 1; + sstrncpy(out, start, out_size); + } + return 0; +} + +int config_cores_parse(const oconfig_item_t *ci, core_groups_list_t *cgl) { + if (ci == NULL || cgl == NULL) + return -EINVAL; + if (ci->values_num == 0 || ci->values_num > MAX_CORES) + return -EINVAL; + core_group_t cgroups[MAX_CORES] = {{0}}; + size_t cg_idx = 0; /* index for cgroups array */ + int ret = 0; + + for (int i = 0; i < ci->values_num; i++) { + if (ci->values[i].type != OCONFIG_TYPE_STRING) { + WARNING(UTIL_NAME ": The %s option requires string arguments.", ci->key); + return -EINVAL; + } + } + + if (ci->values_num == 1 && ci->values[0].value.string && + strlen(ci->values[0].value.string) == 0) + return 0; + + for (int i = 0; i < ci->values_num; i++) { + size_t n; + _Bool grouped = 1; + char str[DATA_MAX_NAME_LEN]; + unsigned cores[MAX_CORES] = {0}; + + if (cg_idx >= STATIC_ARRAY_SIZE(cgroups)) { + ERROR(UTIL_NAME ": Configuration exceeds maximum number of cores: %zu", + STATIC_ARRAY_SIZE(cgroups)); + ret = -EINVAL; + goto parse_error; + } + if ((ci->values[i].value.string == NULL) || + (strlen(ci->values[i].value.string) == 0)) { + ERROR(UTIL_NAME ": Failed to parse parameters for %s option.", ci->key); + ret = -EINVAL; + goto parse_error; + } + + ret = check_core_grouping(str, ci->values[i].value.string, sizeof(str), + &grouped); + if (ret != 0) { + ERROR(UTIL_NAME ": Failed to parse config option [%d] %s.", i, + ci->values[i].value.string); + goto parse_error; + } + n = str_list_to_nums(str, cores, STATIC_ARRAY_SIZE(cores)); + if (n == 0) { + ERROR(UTIL_NAME ": Failed to parse config option [%d] %s.", i, + ci->values[i].value.string); + ret = -EINVAL; + goto parse_error; + } + + if (grouped) { + cgroups[cg_idx].desc = strdup(ci->values[i].value.string); + if (cgroups[cg_idx].desc == NULL) { + ERROR(UTIL_NAME ": Failed to allocate description."); + ret = -ENOMEM; + goto parse_error; + } + + cgroups[cg_idx].cores = calloc(n, sizeof(*cgroups[cg_idx].cores)); + if (cgroups[cg_idx].cores == NULL) { + ERROR(UTIL_NAME ": Failed to allocate cores for cgroup."); + ret = -ENOMEM; + goto parse_error; + } + + for (size_t j = 0; j < n; j++) + cgroups[cg_idx].cores[j] = cores[j]; + + cgroups[cg_idx].num_cores = n; + cg_idx++; + } else { + for (size_t j = 0; j < n && cg_idx < STATIC_ARRAY_SIZE(cgroups); j++) { + char desc[DATA_MAX_NAME_LEN]; + snprintf(desc, sizeof(desc), "%u", cores[j]); + + cgroups[cg_idx].desc = strdup(desc); + if (cgroups[cg_idx].desc == NULL) { + ERROR(UTIL_NAME ": Failed to allocate desc for core %u.", cores[j]); + ret = -ENOMEM; + goto parse_error; + } + + cgroups[cg_idx].cores = calloc(1, sizeof(*(cgroups[cg_idx].cores))); + if (cgroups[cg_idx].cores == NULL) { + ERROR(UTIL_NAME ": Failed to allocate cgroup for core %u.", cores[j]); + ret = -ENOMEM; + goto parse_error; + } + cgroups[cg_idx].num_cores = 1; + cgroups[cg_idx].cores[0] = cores[j]; + cg_idx++; + } + } + } + + cgl->cgroups = calloc(cg_idx, sizeof(*cgl->cgroups)); + if (cgl->cgroups == NULL) { + ERROR(UTIL_NAME ": Failed to allocate core groups."); + ret = -ENOMEM; + goto parse_error; + } + + cgl->num_cgroups = cg_idx; + for (size_t i = 0; i < cg_idx; i++) + cgl->cgroups[i] = cgroups[i]; + + return 0; + +parse_error: + + cg_idx = 0; + while(cg_idx < STATIC_ARRAY_SIZE(cgroups) && cgroups[cg_idx].desc != NULL) { + sfree(cgroups[cg_idx].desc); + sfree(cgroups[cg_idx].cores); + cg_idx++; + } + return ret; +} + +int config_cores_default(int num_cores, core_groups_list_t *cgl) { + if (cgl == NULL || num_cores < 0 || num_cores > MAX_CORES) + return -EINVAL; + + cgl->cgroups = calloc(num_cores, sizeof(*(cgl->cgroups))); + if (cgl->cgroups == NULL) { + ERROR(UTIL_NAME ": Failed to allocate memory for core groups."); + return -ENOMEM; + } + cgl->num_cgroups = num_cores; + + for (int i = 0; i < num_cores; i++) { + char desc[DATA_MAX_NAME_LEN]; + snprintf(desc, sizeof(desc), "%d", i); + + cgl->cgroups[i].cores = calloc(1, sizeof(*(cgl->cgroups[i].cores))); + if (cgl->cgroups[i].cores == NULL) { + ERROR(UTIL_NAME ": Failed to allocate default cores for cgroup %d.", i); + config_cores_cleanup(cgl); + return -ENOMEM; + } + cgl->cgroups[i].num_cores = 1; + cgl->cgroups[i].cores[0] = i; + + cgl->cgroups[i].desc = strdup(desc); + if (cgl->cgroups[i].desc == NULL) { + ERROR(UTIL_NAME ": Failed to allocate description for cgroup %d.", i); + config_cores_cleanup(cgl); + return -ENOMEM; + } + } + return 0; +} + +void config_cores_cleanup(core_groups_list_t *cgl) { + if (cgl == NULL) + return; + for (size_t i = 0; i < cgl->num_cgroups; i++) { + sfree(cgl->cgroups[i].desc); + sfree(cgl->cgroups[i].cores); + } + sfree(cgl->cgroups); + cgl->num_cgroups = 0; +} + +int config_cores_cmp_cgroups(const core_group_t *cg_a, + const core_group_t *cg_b) { + size_t found = 0; + + assert(cg_a != NULL); + assert(cg_b != NULL); + + const size_t sz_a = cg_a->num_cores; + const size_t sz_b = cg_b->num_cores; + const unsigned *tab_a = cg_a->cores; + const unsigned *tab_b = cg_b->cores; + + for (size_t i = 0; i < sz_a; i++) + if (is_in_list(tab_a[i], tab_b, sz_b)) + found++; + + /* if no cores are the same */ + if (!found) + return 0; + /* if group contains same cores */ + if (sz_a == sz_b && sz_b == found) + return 1; + /* if not all cores are the same */ + return -1; +} + diff --git a/src/utils_config_cores.h b/src/utils_config_cores.h new file mode 100644 index 00000000..de79dde2 --- /dev/null +++ b/src/utils_config_cores.h @@ -0,0 +1,131 @@ +/** + * collectd - src/utils_config_cores.h + * + * Copyright(c) 2017 Intel Corporation. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + * Authors: + * Kamil Wiatrowski + **/ + +#ifndef UTILS_CONFIG_CORES_H +#define UTILS_CONFIG_CORES_H 1 + +struct core_group_s { + char *desc; + unsigned *cores; + size_t num_cores; +}; +typedef struct core_group_s core_group_t; + +struct core_groups_list_s { + core_group_t *cgroups; + size_t num_cgroups; +}; +typedef struct core_groups_list_s core_groups_list_t; + +/* + * NAME + * config_cores_parse + * + * DESCRIPTION + * Convert strings from config item into list of core groups. + * + * PARAMETERS + * `ci' Pointer to config item. + * `cgl' Pointer to core groups list to be filled. + * + * RETURN VALUE + * Zero upon success or non-zero if an error occurred. + * + * NOTES + * In case of an error, *cgl is not modified. + * Numbers can be in decimal or hexadecimal format. + * The memory allocated for *cgroups in list needs to be freed + * with config_cores_cleanup. + * + * EXAMPLES + * If config is "0-3" "[4-15]" it means that cores 0-3 are aggregated + * into one group and cores 4 to 15 are stored individualily in + * separate groups. Examples of allowed formats: + * "0,3,4" "10-15" - cores collected into two groups + * "0" "0x3" "7" - 3 cores, each in individual group + * "[32-63]" - 32 cores, each in individual group + * + * For empty string "" *cgl is not modified and zero is returned. + */ +int config_cores_parse(const oconfig_item_t *ci, core_groups_list_t *cgl); + +/* + * NAME + * config_cores_default + * + * DESCRIPTION + * Set number of cores starting from zero into individual + * core groups in *cgl list. + * + * PARAMETERS + * `num_cores' Number of cores to be configured. + * `cgl' Pointer to core groups list. + * + * RETURN VALUE + * Zero upon success or non-zero if an error occurred. + * + * NOTES + * The memory allocated for *cgroups in list needs to be freed + * with config_cores_cleanup. In case of error the memory is + * freed by the function itself. + */ +int config_cores_default(int num_cores, core_groups_list_t *cgl); + +/* + * NAME + * config_cores_cleanup + * + * DESCRIPTION + * Free the memory allocated for cgroups and set + * num_cgroups to zero. + * + * PARAMETERS + * `cgl' Pointer to core groups list. + */ +void config_cores_cleanup(core_groups_list_t *cgl); + +/* + * NAME + * config_cores_cmp_cgroups + * + * DESCRIPTION + * Function to compare cores in 2 core groups. + * + * PARAMETERS + * `cg_a' Pointer to core group a. + * `cg_b' Pointer to core group b. + * + * RETURN VALUE + * 1 if both groups contain the same cores + * 0 if none of their cores match + * -1 if some but not all cores match + */ +int config_cores_cmp_cgroups(const core_group_t *cg_a, + const core_group_t *cg_b); + +#endif /* UTILS_CONFIG_CORES_H */ + diff --git a/src/utils_config_cores_test.c b/src/utils_config_cores_test.c new file mode 100644 index 00000000..8dc4d531 --- /dev/null +++ b/src/utils_config_cores_test.c @@ -0,0 +1,257 @@ +/** + * collectd - src/utils_config_cores_test.c + * + * Copyright(c) 2017 Intel Corporation. All rights reserved. + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + * Authors: + * Kamil Wiatrowski + **/ + +#include "testing.h" +#include "utils_config_cores.c" /* sic */ + +oconfig_value_t test_cfg_values[] = { + { {"0"}, OCONFIG_TYPE_STRING }, + { {"1-2"}, OCONFIG_TYPE_STRING }, + { {"[3-4]"}, OCONFIG_TYPE_STRING }}; + +oconfig_item_t test_cfg = { + "Cores", + test_cfg_values, + STATIC_ARRAY_SIZE(test_cfg_values), + NULL, + NULL, + 0 }; + +static int compare_with_test_config(core_groups_list_t *cgl) { + if (cgl->num_cgroups == 4 && + cgl->cgroups[0].num_cores == 1 && + strcmp("0", cgl->cgroups[0].desc) == 0 && + cgl->cgroups[0].cores[0] == 0 && + cgl->cgroups[1].num_cores == 2 && + strcmp("1-2", cgl->cgroups[1].desc) == 0 && + cgl->cgroups[1].cores[0] == 1 && + cgl->cgroups[1].cores[1] == 2 && + cgl->cgroups[2].num_cores == 1 && + strcmp("3", cgl->cgroups[2].desc) == 0 && + cgl->cgroups[2].cores[0] == 3 && + cgl->cgroups[3].num_cores == 1 && + strcmp("4", cgl->cgroups[3].desc) == 0 && + cgl->cgroups[3].cores[0] == 4) + return 0; + + return -1; +} + +DEF_TEST(string_to_uint) { + int ret = 0; + char *s = "13", *s1 = "0xd", *s2 = "g"; + unsigned n = 0; + + ret = str_to_uint(s, &n); + EXPECT_EQ_INT(0, ret); + EXPECT_EQ_INT(13, n); + + ret = str_to_uint(s1, &n); + EXPECT_EQ_INT(0, ret); + EXPECT_EQ_INT(13, n); + + ret = str_to_uint(s2, &n); + OK(ret < 0); + + ret = str_to_uint(NULL, &n); + OK(ret < 0); + return 0; +} + +DEF_TEST(cores_list_to_numbers) { + size_t n = 0; + unsigned nums[MAX_CORES]; + char str[64] = ""; + + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(0, n); + + strncpy(str, "1", STATIC_ARRAY_SIZE(str)); + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(1, n); + EXPECT_EQ_INT(1, nums[0]); + + strncpy(str, "0,2-3", STATIC_ARRAY_SIZE(str)); + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(3, n); + EXPECT_EQ_INT(0, nums[0]); + EXPECT_EQ_INT(2, nums[1]); + EXPECT_EQ_INT(3, nums[2]); + + strncpy(str, "11-0xa", STATIC_ARRAY_SIZE(str)); + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(2, n); + EXPECT_EQ_INT(10, nums[0]); + EXPECT_EQ_INT(11, nums[1]); + + snprintf(str, sizeof(str), "0-%d", (MAX_CORES - 1)); + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(MAX_CORES, n); + EXPECT_EQ_INT(0, nums[0]); + EXPECT_EQ_INT(MAX_CORES-1, nums[MAX_CORES-1]); + + /* Should return 0 for incorrect syntax. */ + strncpy(str, "5g", STATIC_ARRAY_SIZE(str)); + n = str_list_to_nums(str, nums, STATIC_ARRAY_SIZE(nums)); + EXPECT_EQ_INT(0, n); + return 0; +} + +DEF_TEST(check_grouped_cores) { + int ret = 0; + _Bool grouped; + char src[64] = "[5-15]"; + char dest[64]; + + ret = check_core_grouping(dest, src, sizeof(dest), &grouped); + EXPECT_EQ_INT(0, ret); + EXPECT_EQ_INT(0, grouped); + EXPECT_EQ_STR("5-15", dest); + + strncpy(src, " 5-15", STATIC_ARRAY_SIZE(src)); + ret = check_core_grouping(dest, src, sizeof(dest), &grouped); + EXPECT_EQ_INT(0, ret); + EXPECT_EQ_INT(1, grouped); + EXPECT_EQ_STR("5-15", dest); + return 0; +} + +DEF_TEST(cores_option_parse) { + int ret = 0; + core_groups_list_t cgl = {0}; + + ret = config_cores_parse(&test_cfg, &cgl); + EXPECT_EQ_INT(0, ret); + CHECK_NOT_NULL(cgl.cgroups); + EXPECT_EQ_INT(0, compare_with_test_config(&cgl)); + + config_cores_cleanup(&cgl); + return 0; +} + +DEF_TEST(cores_option_parse_fail) { + int ret = 0; + core_groups_list_t cgl = {0}; + /* Wrong value, missing closing bracket ] */ + oconfig_value_t values = { {"[0-15"}, OCONFIG_TYPE_STRING }; + oconfig_item_t cfg = { "Cores", &values, 1, NULL, NULL, 0 }; + + ret = config_cores_parse(&cfg, &cgl); + EXPECT_EQ_INT(-EINVAL, ret); + EXPECT_EQ_INT(0, cgl.num_cgroups); + OK(NULL == cgl.cgroups); + return 0; +} + +DEF_TEST(cores_default_list) { + int ret = 0; + core_groups_list_t cgl = {0}; + + ret = config_cores_default(2, &cgl); + EXPECT_EQ_INT(0, ret); + EXPECT_EQ_INT(2, cgl.num_cgroups); + CHECK_NOT_NULL(cgl.cgroups); + + CHECK_NOT_NULL(cgl.cgroups[0].cores); + CHECK_NOT_NULL(cgl.cgroups[0].desc); + EXPECT_EQ_STR("0", cgl.cgroups[0].desc); + EXPECT_EQ_INT(1, cgl.cgroups[0].num_cores); + EXPECT_EQ_INT(0, cgl.cgroups[0].cores[0]); + + CHECK_NOT_NULL(cgl.cgroups[1].cores); + CHECK_NOT_NULL(cgl.cgroups[1].desc); + EXPECT_EQ_STR("1", cgl.cgroups[1].desc); + EXPECT_EQ_INT(1, cgl.cgroups[1].num_cores); + EXPECT_EQ_INT(1, cgl.cgroups[1].cores[0]); + + config_cores_cleanup(&cgl); + return 0; +} + +DEF_TEST(cores_default_list_fail) { + int ret = 0; + core_groups_list_t cgl = {0}; + + ret = config_cores_default(-1, &cgl); + OK(ret < 0); + ret = config_cores_default(MAX_CORES+1, &cgl); + OK(ret < 0); + ret = config_cores_default(1, NULL); + OK(ret < 0); + return 0; +} + +DEF_TEST(cores_group_cleanup) { + core_groups_list_t cgl; + cgl.cgroups = calloc(1, sizeof(*cgl.cgroups)); + CHECK_NOT_NULL(cgl.cgroups); + cgl.num_cgroups = 1; + cgl.cgroups[0].desc = strdup("1"); + cgl.cgroups[0].cores = calloc(1, sizeof(*cgl.cgroups[0].cores)); + CHECK_NOT_NULL(cgl.cgroups[0].cores); + cgl.cgroups[0].cores[0] = 1; + cgl.cgroups[0].num_cores = 1; + + config_cores_cleanup(&cgl); + OK(NULL == cgl.cgroups); + EXPECT_EQ_INT(0, cgl.num_cgroups); + return 0; +} + +DEF_TEST(cores_group_cmp) { + unsigned cores_mock[] = {0,1,2}; + core_group_t group_mock = { "0,1,2", cores_mock, 3 }; + unsigned cores_mock_2[] = {2,3}; + core_group_t group_mock_2 = { "2,3", cores_mock_2, 2 }; + + int ret = config_cores_cmp_cgroups(&group_mock, &group_mock); + EXPECT_EQ_INT(1, ret); + + ret = config_cores_cmp_cgroups(&group_mock, &group_mock_2); + EXPECT_EQ_INT(-1, ret); + + cores_mock_2[0] = 4; + ret = config_cores_cmp_cgroups(&group_mock, &group_mock_2); + EXPECT_EQ_INT(0, ret); + return 0; +} + +int main(void) { + RUN_TEST(string_to_uint); + RUN_TEST(cores_list_to_numbers); + RUN_TEST(check_grouped_cores); + + RUN_TEST(cores_group_cleanup); + RUN_TEST(cores_option_parse); + RUN_TEST(cores_option_parse_fail); + RUN_TEST(cores_default_list); + RUN_TEST(cores_default_list_fail); + + RUN_TEST(cores_group_cmp); + + END_TEST; +} -- 2.11.0