From 5d1646d90e1f2cceb9f0828f4b28318cd0ec7744 Mon Sep 17 00:00:00 2001 From: Daniel Baumann Date: Sat, 27 Apr 2024 12:05:51 +0200 Subject: Adding upstream version 5.10.209. Signed-off-by: Daniel Baumann --- tools/perf/util/stat-shadow.c | 1201 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 1201 insertions(+) create mode 100644 tools/perf/util/stat-shadow.c (limited to 'tools/perf/util/stat-shadow.c') diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c new file mode 100644 index 000000000..901265127 --- /dev/null +++ b/tools/perf/util/stat-shadow.c @@ -0,0 +1,1201 @@ +// SPDX-License-Identifier: GPL-2.0 +#include +#include "evsel.h" +#include "stat.h" +#include "color.h" +#include "pmu.h" +#include "rblist.h" +#include "evlist.h" +#include "expr.h" +#include "metricgroup.h" +#include + +/* + * AGGR_GLOBAL: Use CPU 0 + * AGGR_SOCKET: Use first CPU of socket + * AGGR_DIE: Use first CPU of die + * AGGR_CORE: Use first CPU of core + * AGGR_NONE: Use matching CPU + * AGGR_THREAD: Not supported? + */ + +struct runtime_stat rt_stat; +struct stats walltime_nsecs_stats; + +struct saved_value { + struct rb_node rb_node; + struct evsel *evsel; + enum stat_type type; + int ctx; + int cpu; + struct runtime_stat *stat; + struct stats stats; + u64 metric_total; + int metric_other; +}; + +static int saved_value_cmp(struct rb_node *rb_node, const void *entry) +{ + struct saved_value *a = container_of(rb_node, + struct saved_value, + rb_node); + const struct saved_value *b = entry; + + if (a->cpu != b->cpu) + return a->cpu - b->cpu; + + /* + * Previously the rbtree was used to link generic metrics. + * The keys were evsel/cpu. Now the rbtree is extended to support + * per-thread shadow stats. For shadow stats case, the keys + * are cpu/type/ctx/stat (evsel is NULL). For generic metrics + * case, the keys are still evsel/cpu (type/ctx/stat are 0 or NULL). + */ + if (a->type != b->type) + return a->type - b->type; + + if (a->ctx != b->ctx) + return a->ctx - b->ctx; + + if (a->evsel == NULL && b->evsel == NULL) { + if (a->stat == b->stat) + return 0; + + if ((char *)a->stat < (char *)b->stat) + return -1; + + return 1; + } + + if (a->evsel == b->evsel) + return 0; + if ((char *)a->evsel < (char *)b->evsel) + return -1; + return +1; +} + +static struct rb_node *saved_value_new(struct rblist *rblist __maybe_unused, + const void *entry) +{ + struct saved_value *nd = malloc(sizeof(struct saved_value)); + + if (!nd) + return NULL; + memcpy(nd, entry, sizeof(struct saved_value)); + return &nd->rb_node; +} + +static void saved_value_delete(struct rblist *rblist __maybe_unused, + struct rb_node *rb_node) +{ + struct saved_value *v; + + BUG_ON(!rb_node); + v = container_of(rb_node, struct saved_value, rb_node); + free(v); +} + +static struct saved_value *saved_value_lookup(struct evsel *evsel, + int cpu, + bool create, + enum stat_type type, + int ctx, + struct runtime_stat *st) +{ + struct rblist *rblist; + struct rb_node *nd; + struct saved_value dm = { + .cpu = cpu, + .evsel = evsel, + .type = type, + .ctx = ctx, + .stat = st, + }; + + rblist = &st->value_list; + + nd = rblist__find(rblist, &dm); + if (nd) + return container_of(nd, struct saved_value, rb_node); + if (create) { + rblist__add_node(rblist, &dm); + nd = rblist__find(rblist, &dm); + if (nd) + return container_of(nd, struct saved_value, rb_node); + } + return NULL; +} + +void runtime_stat__init(struct runtime_stat *st) +{ + struct rblist *rblist = &st->value_list; + + rblist__init(rblist); + rblist->node_cmp = saved_value_cmp; + rblist->node_new = saved_value_new; + rblist->node_delete = saved_value_delete; +} + +void runtime_stat__exit(struct runtime_stat *st) +{ + rblist__exit(&st->value_list); +} + +void perf_stat__init_shadow_stats(void) +{ + runtime_stat__init(&rt_stat); +} + +static int evsel_context(struct evsel *evsel) +{ + int ctx = 0; + + if (evsel->core.attr.exclude_kernel) + ctx |= CTX_BIT_KERNEL; + if (evsel->core.attr.exclude_user) + ctx |= CTX_BIT_USER; + if (evsel->core.attr.exclude_hv) + ctx |= CTX_BIT_HV; + if (evsel->core.attr.exclude_host) + ctx |= CTX_BIT_HOST; + if (evsel->core.attr.exclude_idle) + ctx |= CTX_BIT_IDLE; + + return ctx; +} + +static void reset_stat(struct runtime_stat *st) +{ + struct rblist *rblist; + struct rb_node *pos, *next; + + rblist = &st->value_list; + next = rb_first_cached(&rblist->entries); + while (next) { + pos = next; + next = rb_next(pos); + memset(&container_of(pos, struct saved_value, rb_node)->stats, + 0, + sizeof(struct stats)); + } +} + +void perf_stat__reset_shadow_stats(void) +{ + reset_stat(&rt_stat); + memset(&walltime_nsecs_stats, 0, sizeof(walltime_nsecs_stats)); +} + +void perf_stat__reset_shadow_per_stat(struct runtime_stat *st) +{ + reset_stat(st); +} + +static void update_runtime_stat(struct runtime_stat *st, + enum stat_type type, + int ctx, int cpu, u64 count) +{ + struct saved_value *v = saved_value_lookup(NULL, cpu, true, + type, ctx, st); + + if (v) + update_stats(&v->stats, count); +} + +/* + * Update various tracking values we maintain to print + * more semantic information such as miss/hit ratios, + * instruction rates, etc: + */ +void perf_stat__update_shadow_stats(struct evsel *counter, u64 count, + int cpu, struct runtime_stat *st) +{ + int ctx = evsel_context(counter); + u64 count_ns = count; + struct saved_value *v; + + count *= counter->scale; + + if (evsel__is_clock(counter)) + update_runtime_stat(st, STAT_NSECS, 0, cpu, count_ns); + else if (evsel__match(counter, HARDWARE, HW_CPU_CYCLES)) + update_runtime_stat(st, STAT_CYCLES, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, CYCLES_IN_TX)) + update_runtime_stat(st, STAT_CYCLES_IN_TX, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TRANSACTION_START)) + update_runtime_stat(st, STAT_TRANSACTION, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, ELISION_START)) + update_runtime_stat(st, STAT_ELISION, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_TOTAL_SLOTS)) + update_runtime_stat(st, STAT_TOPDOWN_TOTAL_SLOTS, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_SLOTS_ISSUED)) + update_runtime_stat(st, STAT_TOPDOWN_SLOTS_ISSUED, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_SLOTS_RETIRED)) + update_runtime_stat(st, STAT_TOPDOWN_SLOTS_RETIRED, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_FETCH_BUBBLES)) + update_runtime_stat(st, STAT_TOPDOWN_FETCH_BUBBLES, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_RECOVERY_BUBBLES)) + update_runtime_stat(st, STAT_TOPDOWN_RECOVERY_BUBBLES, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_RETIRING)) + update_runtime_stat(st, STAT_TOPDOWN_RETIRING, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_BAD_SPEC)) + update_runtime_stat(st, STAT_TOPDOWN_BAD_SPEC, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_FE_BOUND)) + update_runtime_stat(st, STAT_TOPDOWN_FE_BOUND, + ctx, cpu, count); + else if (perf_stat_evsel__is(counter, TOPDOWN_BE_BOUND)) + update_runtime_stat(st, STAT_TOPDOWN_BE_BOUND, + ctx, cpu, count); + else if (evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND)) + update_runtime_stat(st, STAT_STALLED_CYCLES_FRONT, + ctx, cpu, count); + else if (evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND)) + update_runtime_stat(st, STAT_STALLED_CYCLES_BACK, + ctx, cpu, count); + else if (evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS)) + update_runtime_stat(st, STAT_BRANCHES, ctx, cpu, count); + else if (evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES)) + update_runtime_stat(st, STAT_CACHEREFS, ctx, cpu, count); + else if (evsel__match(counter, HW_CACHE, HW_CACHE_L1D)) + update_runtime_stat(st, STAT_L1_DCACHE, ctx, cpu, count); + else if (evsel__match(counter, HW_CACHE, HW_CACHE_L1I)) + update_runtime_stat(st, STAT_L1_ICACHE, ctx, cpu, count); + else if (evsel__match(counter, HW_CACHE, HW_CACHE_LL)) + update_runtime_stat(st, STAT_LL_CACHE, ctx, cpu, count); + else if (evsel__match(counter, HW_CACHE, HW_CACHE_DTLB)) + update_runtime_stat(st, STAT_DTLB_CACHE, ctx, cpu, count); + else if (evsel__match(counter, HW_CACHE, HW_CACHE_ITLB)) + update_runtime_stat(st, STAT_ITLB_CACHE, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, SMI_NUM)) + update_runtime_stat(st, STAT_SMI_NUM, ctx, cpu, count); + else if (perf_stat_evsel__is(counter, APERF)) + update_runtime_stat(st, STAT_APERF, ctx, cpu, count); + + if (counter->collect_stat) { + v = saved_value_lookup(counter, cpu, true, STAT_NONE, 0, st); + update_stats(&v->stats, count); + if (counter->metric_leader) + v->metric_total += count; + } else if (counter->metric_leader) { + v = saved_value_lookup(counter->metric_leader, + cpu, true, STAT_NONE, 0, st); + v->metric_total += count; + v->metric_other++; + } +} + +/* used for get_ratio_color() */ +enum grc_type { + GRC_STALLED_CYCLES_FE, + GRC_STALLED_CYCLES_BE, + GRC_CACHE_MISSES, + GRC_MAX_NR +}; + +static const char *get_ratio_color(enum grc_type type, double ratio) +{ + static const double grc_table[GRC_MAX_NR][3] = { + [GRC_STALLED_CYCLES_FE] = { 50.0, 30.0, 10.0 }, + [GRC_STALLED_CYCLES_BE] = { 75.0, 50.0, 20.0 }, + [GRC_CACHE_MISSES] = { 20.0, 10.0, 5.0 }, + }; + const char *color = PERF_COLOR_NORMAL; + + if (ratio > grc_table[type][0]) + color = PERF_COLOR_RED; + else if (ratio > grc_table[type][1]) + color = PERF_COLOR_MAGENTA; + else if (ratio > grc_table[type][2]) + color = PERF_COLOR_YELLOW; + + return color; +} + +static struct evsel *perf_stat__find_event(struct evlist *evsel_list, + const char *name) +{ + struct evsel *c2; + + evlist__for_each_entry (evsel_list, c2) { + if (!strcasecmp(c2->name, name) && !c2->collect_stat) + return c2; + } + return NULL; +} + +/* Mark MetricExpr target events and link events using them to them. */ +void perf_stat__collect_metric_expr(struct evlist *evsel_list) +{ + struct evsel *counter, *leader, **metric_events, *oc; + bool found; + struct expr_parse_ctx ctx; + struct hashmap_entry *cur; + size_t bkt; + int i; + + expr__ctx_init(&ctx); + evlist__for_each_entry(evsel_list, counter) { + bool invalid = false; + + leader = counter->leader; + if (!counter->metric_expr) + continue; + + expr__ctx_clear(&ctx); + metric_events = counter->metric_events; + if (!metric_events) { + if (expr__find_other(counter->metric_expr, + counter->name, + &ctx, 1) < 0) + continue; + + metric_events = calloc(sizeof(struct evsel *), + hashmap__size(&ctx.ids) + 1); + if (!metric_events) { + expr__ctx_clear(&ctx); + return; + } + counter->metric_events = metric_events; + } + + i = 0; + hashmap__for_each_entry((&ctx.ids), cur, bkt) { + const char *metric_name = (const char *)cur->key; + + found = false; + if (leader) { + /* Search in group */ + for_each_group_member (oc, leader) { + if (!strcasecmp(oc->name, + metric_name) && + !oc->collect_stat) { + found = true; + break; + } + } + } + if (!found) { + /* Search ignoring groups */ + oc = perf_stat__find_event(evsel_list, + metric_name); + } + if (!oc) { + /* Deduping one is good enough to handle duplicated PMUs. */ + static char *printed; + + /* + * Adding events automatically would be difficult, because + * it would risk creating groups that are not schedulable. + * perf stat doesn't understand all the scheduling constraints + * of events. So we ask the user instead to add the missing + * events. + */ + if (!printed || + strcasecmp(printed, metric_name)) { + fprintf(stderr, + "Add %s event to groups to get metric expression for %s\n", + metric_name, + counter->name); + printed = strdup(metric_name); + } + invalid = true; + continue; + } + metric_events[i++] = oc; + oc->collect_stat = true; + } + metric_events[i] = NULL; + if (invalid) { + free(metric_events); + counter->metric_events = NULL; + counter->metric_expr = NULL; + } + } + expr__ctx_clear(&ctx); +} + +static double runtime_stat_avg(struct runtime_stat *st, + enum stat_type type, int ctx, int cpu) +{ + struct saved_value *v; + + v = saved_value_lookup(NULL, cpu, false, type, ctx, st); + if (!v) + return 0.0; + + return avg_stats(&v->stats); +} + +static double runtime_stat_n(struct runtime_stat *st, + enum stat_type type, int ctx, int cpu) +{ + struct saved_value *v; + + v = saved_value_lookup(NULL, cpu, false, type, ctx, st); + if (!v) + return 0.0; + + return v->stats.n; +} + +static void print_stalled_cycles_frontend(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_STALLED_CYCLES_FE, ratio); + + if (ratio) + out->print_metric(config, out->ctx, color, "%7.2f%%", "frontend cycles idle", + ratio); + else + out->print_metric(config, out->ctx, NULL, NULL, "frontend cycles idle", 0); +} + +static void print_stalled_cycles_backend(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_STALLED_CYCLES_BE, ratio); + + out->print_metric(config, out->ctx, color, "%7.2f%%", "backend cycles idle", ratio); +} + +static void print_branch_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_BRANCHES, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all branches", ratio); +} + +static void print_l1_dcache_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) + +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_L1_DCACHE, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all L1-dcache accesses", ratio); +} + +static void print_l1_icache_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) + +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_L1_ICACHE, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all L1-icache accesses", ratio); +} + +static void print_dtlb_cache_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_DTLB_CACHE, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all dTLB cache accesses", ratio); +} + +static void print_itlb_cache_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_ITLB_CACHE, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all iTLB cache accesses", ratio); +} + +static void print_ll_cache_misses(struct perf_stat_config *config, + int cpu, + struct evsel *evsel, + double avg, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double total, ratio = 0.0; + const char *color; + int ctx = evsel_context(evsel); + + total = runtime_stat_avg(st, STAT_LL_CACHE, ctx, cpu); + + if (total) + ratio = avg / total * 100.0; + + color = get_ratio_color(GRC_CACHE_MISSES, ratio); + out->print_metric(config, out->ctx, color, "%7.2f%%", "of all LL-cache accesses", ratio); +} + +/* + * High level "TopDown" CPU core pipe line bottleneck break down. + * + * Basic concept following + * Yasin, A Top Down Method for Performance analysis and Counter architecture + * ISPASS14 + * + * The CPU pipeline is divided into 4 areas that can be bottlenecks: + * + * Frontend -> Backend -> Retiring + * BadSpeculation in addition means out of order execution that is thrown away + * (for example branch mispredictions) + * Frontend is instruction decoding. + * Backend is execution, like computation and accessing data in memory + * Retiring is good execution that is not directly bottlenecked + * + * The formulas are computed in slots. + * A slot is an entry in the pipeline each for the pipeline width + * (for example a 4-wide pipeline has 4 slots for each cycle) + * + * Formulas: + * BadSpeculation = ((SlotsIssued - SlotsRetired) + RecoveryBubbles) / + * TotalSlots + * Retiring = SlotsRetired / TotalSlots + * FrontendBound = FetchBubbles / TotalSlots + * BackendBound = 1.0 - BadSpeculation - Retiring - FrontendBound + * + * The kernel provides the mapping to the low level CPU events and any scaling + * needed for the CPU pipeline width, for example: + * + * TotalSlots = Cycles * 4 + * + * The scaling factor is communicated in the sysfs unit. + * + * In some cases the CPU may not be able to measure all the formulas due to + * missing events. In this case multiple formulas are combined, as possible. + * + * Full TopDown supports more levels to sub-divide each area: for example + * BackendBound into computing bound and memory bound. For now we only + * support Level 1 TopDown. + */ + +static double sanitize_val(double x) +{ + if (x < 0 && x >= -0.02) + return 0.0; + return x; +} + +static double td_total_slots(int ctx, int cpu, struct runtime_stat *st) +{ + return runtime_stat_avg(st, STAT_TOPDOWN_TOTAL_SLOTS, ctx, cpu); +} + +static double td_bad_spec(int ctx, int cpu, struct runtime_stat *st) +{ + double bad_spec = 0; + double total_slots; + double total; + + total = runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_ISSUED, ctx, cpu) - + runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_RETIRED, ctx, cpu) + + runtime_stat_avg(st, STAT_TOPDOWN_RECOVERY_BUBBLES, ctx, cpu); + + total_slots = td_total_slots(ctx, cpu, st); + if (total_slots) + bad_spec = total / total_slots; + return sanitize_val(bad_spec); +} + +static double td_retiring(int ctx, int cpu, struct runtime_stat *st) +{ + double retiring = 0; + double total_slots = td_total_slots(ctx, cpu, st); + double ret_slots = runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_RETIRED, + ctx, cpu); + + if (total_slots) + retiring = ret_slots / total_slots; + return retiring; +} + +static double td_fe_bound(int ctx, int cpu, struct runtime_stat *st) +{ + double fe_bound = 0; + double total_slots = td_total_slots(ctx, cpu, st); + double fetch_bub = runtime_stat_avg(st, STAT_TOPDOWN_FETCH_BUBBLES, + ctx, cpu); + + if (total_slots) + fe_bound = fetch_bub / total_slots; + return fe_bound; +} + +static double td_be_bound(int ctx, int cpu, struct runtime_stat *st) +{ + double sum = (td_fe_bound(ctx, cpu, st) + + td_bad_spec(ctx, cpu, st) + + td_retiring(ctx, cpu, st)); + if (sum == 0) + return 0; + return sanitize_val(1.0 - sum); +} + +/* + * Kernel reports metrics multiplied with slots. To get back + * the ratios we need to recreate the sum. + */ + +static double td_metric_ratio(int ctx, int cpu, + enum stat_type type, + struct runtime_stat *stat) +{ + double sum = runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, ctx, cpu) + + runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, ctx, cpu) + + runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, ctx, cpu) + + runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, ctx, cpu); + double d = runtime_stat_avg(stat, type, ctx, cpu); + + if (sum) + return d / sum; + return 0; +} + +/* + * ... but only if most of the values are actually available. + * We allow two missing. + */ + +static bool full_td(int ctx, int cpu, + struct runtime_stat *stat) +{ + int c = 0; + + if (runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, ctx, cpu) > 0) + c++; + if (runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, ctx, cpu) > 0) + c++; + if (runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, ctx, cpu) > 0) + c++; + if (runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, ctx, cpu) > 0) + c++; + return c >= 2; +} + +static void print_smi_cost(struct perf_stat_config *config, + int cpu, struct evsel *evsel, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + double smi_num, aperf, cycles, cost = 0.0; + int ctx = evsel_context(evsel); + const char *color = NULL; + + smi_num = runtime_stat_avg(st, STAT_SMI_NUM, ctx, cpu); + aperf = runtime_stat_avg(st, STAT_APERF, ctx, cpu); + cycles = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + + if ((cycles == 0) || (aperf == 0)) + return; + + if (smi_num) + cost = (aperf - cycles) / aperf * 100.00; + + if (cost > 10) + color = PERF_COLOR_RED; + out->print_metric(config, out->ctx, color, "%8.1f%%", "SMI cycles%", cost); + out->print_metric(config, out->ctx, NULL, "%4.0f", "SMI#", smi_num); +} + +static int prepare_metric(struct evsel **metric_events, + struct metric_ref *metric_refs, + struct expr_parse_ctx *pctx, + int cpu, + struct runtime_stat *st) +{ + double scale; + char *n, *pn; + int i, j, ret; + + expr__ctx_init(pctx); + for (i = 0; metric_events[i]; i++) { + struct saved_value *v; + struct stats *stats; + u64 metric_total = 0; + + if (!strcmp(metric_events[i]->name, "duration_time")) { + stats = &walltime_nsecs_stats; + scale = 1e-9; + } else { + v = saved_value_lookup(metric_events[i], cpu, false, + STAT_NONE, 0, st); + if (!v) + break; + stats = &v->stats; + scale = 1.0; + + if (v->metric_other) + metric_total = v->metric_total; + } + + n = strdup(metric_events[i]->name); + if (!n) + return -ENOMEM; + /* + * This display code with --no-merge adds [cpu] postfixes. + * These are not supported by the parser. Remove everything + * after the space. + */ + pn = strchr(n, ' '); + if (pn) + *pn = 0; + + if (metric_total) + expr__add_id_val(pctx, n, metric_total); + else + expr__add_id_val(pctx, n, avg_stats(stats)*scale); + } + + for (j = 0; metric_refs && metric_refs[j].metric_name; j++) { + ret = expr__add_ref(pctx, &metric_refs[j]); + if (ret) + return ret; + } + + return i; +} + +static void generic_metric(struct perf_stat_config *config, + const char *metric_expr, + struct evsel **metric_events, + struct metric_ref *metric_refs, + char *name, + const char *metric_name, + const char *metric_unit, + int runtime, + int cpu, + struct perf_stat_output_ctx *out, + struct runtime_stat *st) +{ + print_metric_t print_metric = out->print_metric; + struct expr_parse_ctx pctx; + double ratio, scale; + int i; + void *ctxp = out->ctx; + + i = prepare_metric(metric_events, metric_refs, &pctx, cpu, st); + if (i < 0) + return; + + if (!metric_events[i]) { + if (expr__parse(&ratio, &pctx, metric_expr, runtime) == 0) { + char *unit; + char metric_bf[64]; + + if (metric_unit && metric_name) { + if (perf_pmu__convert_scale(metric_unit, + &unit, &scale) >= 0) { + ratio *= scale; + } + if (strstr(metric_expr, "?")) + scnprintf(metric_bf, sizeof(metric_bf), + "%s %s_%d", unit, metric_name, runtime); + else + scnprintf(metric_bf, sizeof(metric_bf), + "%s %s", unit, metric_name); + + print_metric(config, ctxp, NULL, "%8.1f", + metric_bf, ratio); + } else { + print_metric(config, ctxp, NULL, "%8.2f", + metric_name ? + metric_name : + out->force_header ? name : "", + ratio); + } + } else { + print_metric(config, ctxp, NULL, NULL, + out->force_header ? + (metric_name ? metric_name : name) : "", 0); + } + } else { + print_metric(config, ctxp, NULL, NULL, + out->force_header ? + (metric_name ? metric_name : name) : "", 0); + } + + expr__ctx_clear(&pctx); +} + +double test_generic_metric(struct metric_expr *mexp, int cpu, struct runtime_stat *st) +{ + struct expr_parse_ctx pctx; + double ratio = 0.0; + + if (prepare_metric(mexp->metric_events, mexp->metric_refs, &pctx, cpu, st) < 0) + goto out; + + if (expr__parse(&ratio, &pctx, mexp->metric_expr, 1)) + ratio = 0.0; + +out: + expr__ctx_clear(&pctx); + return ratio; +} + +void perf_stat__print_shadow_stats(struct perf_stat_config *config, + struct evsel *evsel, + double avg, int cpu, + struct perf_stat_output_ctx *out, + struct rblist *metric_events, + struct runtime_stat *st) +{ + void *ctxp = out->ctx; + print_metric_t print_metric = out->print_metric; + double total, ratio = 0.0, total2; + const char *color = NULL; + int ctx = evsel_context(evsel); + struct metric_event *me; + int num = 1; + + if (evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) { + total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + + if (total) { + ratio = avg / total; + print_metric(config, ctxp, NULL, "%7.2f ", + "insn per cycle", ratio); + } else { + print_metric(config, ctxp, NULL, NULL, "insn per cycle", 0); + } + + total = runtime_stat_avg(st, STAT_STALLED_CYCLES_FRONT, + ctx, cpu); + + total = max(total, runtime_stat_avg(st, + STAT_STALLED_CYCLES_BACK, + ctx, cpu)); + + if (total && avg) { + out->new_line(config, ctxp); + ratio = total / avg; + print_metric(config, ctxp, NULL, "%7.2f ", + "stalled cycles per insn", + ratio); + } + } else if (evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES)) { + if (runtime_stat_n(st, STAT_BRANCHES, ctx, cpu) != 0) + print_branch_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all branches", 0); + } else if ( + evsel->core.attr.type == PERF_TYPE_HW_CACHE && + evsel->core.attr.config == ( PERF_COUNT_HW_CACHE_L1D | + ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { + + if (runtime_stat_n(st, STAT_L1_DCACHE, ctx, cpu) != 0) + print_l1_dcache_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all L1-dcache accesses", 0); + } else if ( + evsel->core.attr.type == PERF_TYPE_HW_CACHE && + evsel->core.attr.config == ( PERF_COUNT_HW_CACHE_L1I | + ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { + + if (runtime_stat_n(st, STAT_L1_ICACHE, ctx, cpu) != 0) + print_l1_icache_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all L1-icache accesses", 0); + } else if ( + evsel->core.attr.type == PERF_TYPE_HW_CACHE && + evsel->core.attr.config == ( PERF_COUNT_HW_CACHE_DTLB | + ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { + + if (runtime_stat_n(st, STAT_DTLB_CACHE, ctx, cpu) != 0) + print_dtlb_cache_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all dTLB cache accesses", 0); + } else if ( + evsel->core.attr.type == PERF_TYPE_HW_CACHE && + evsel->core.attr.config == ( PERF_COUNT_HW_CACHE_ITLB | + ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { + + if (runtime_stat_n(st, STAT_ITLB_CACHE, ctx, cpu) != 0) + print_itlb_cache_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all iTLB cache accesses", 0); + } else if ( + evsel->core.attr.type == PERF_TYPE_HW_CACHE && + evsel->core.attr.config == ( PERF_COUNT_HW_CACHE_LL | + ((PERF_COUNT_HW_CACHE_OP_READ) << 8) | + ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) { + + if (runtime_stat_n(st, STAT_LL_CACHE, ctx, cpu) != 0) + print_ll_cache_misses(config, cpu, evsel, avg, out, st); + else + print_metric(config, ctxp, NULL, NULL, "of all LL-cache accesses", 0); + } else if (evsel__match(evsel, HARDWARE, HW_CACHE_MISSES)) { + total = runtime_stat_avg(st, STAT_CACHEREFS, ctx, cpu); + + if (total) + ratio = avg * 100 / total; + + if (runtime_stat_n(st, STAT_CACHEREFS, ctx, cpu) != 0) + print_metric(config, ctxp, NULL, "%8.3f %%", + "of all cache refs", ratio); + else + print_metric(config, ctxp, NULL, NULL, "of all cache refs", 0); + } else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_FRONTEND)) { + print_stalled_cycles_frontend(config, cpu, evsel, avg, out, st); + } else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_BACKEND)) { + print_stalled_cycles_backend(config, cpu, evsel, avg, out, st); + } else if (evsel__match(evsel, HARDWARE, HW_CPU_CYCLES)) { + total = runtime_stat_avg(st, STAT_NSECS, 0, cpu); + + if (total) { + ratio = avg / total; + print_metric(config, ctxp, NULL, "%8.3f", "GHz", ratio); + } else { + print_metric(config, ctxp, NULL, NULL, "Ghz", 0); + } + } else if (perf_stat_evsel__is(evsel, CYCLES_IN_TX)) { + total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + + if (total) + print_metric(config, ctxp, NULL, + "%7.2f%%", "transactional cycles", + 100.0 * (avg / total)); + else + print_metric(config, ctxp, NULL, NULL, "transactional cycles", + 0); + } else if (perf_stat_evsel__is(evsel, CYCLES_IN_TX_CP)) { + total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu); + total2 = runtime_stat_avg(st, STAT_CYCLES_IN_TX, ctx, cpu); + + if (total2 < avg) + total2 = avg; + if (total) + print_metric(config, ctxp, NULL, "%7.2f%%", "aborted cycles", + 100.0 * ((total2-avg) / total)); + else + print_metric(config, ctxp, NULL, NULL, "aborted cycles", 0); + } else if (perf_stat_evsel__is(evsel, TRANSACTION_START)) { + total = runtime_stat_avg(st, STAT_CYCLES_IN_TX, + ctx, cpu); + + if (avg) + ratio = total / avg; + + if (runtime_stat_n(st, STAT_CYCLES_IN_TX, ctx, cpu) != 0) + print_metric(config, ctxp, NULL, "%8.0f", + "cycles / transaction", ratio); + else + print_metric(config, ctxp, NULL, NULL, "cycles / transaction", + 0); + } else if (perf_stat_evsel__is(evsel, ELISION_START)) { + total = runtime_stat_avg(st, STAT_CYCLES_IN_TX, + ctx, cpu); + + if (avg) + ratio = total / avg; + + print_metric(config, ctxp, NULL, "%8.0f", "cycles / elision", ratio); + } else if (evsel__is_clock(evsel)) { + if ((ratio = avg_stats(&walltime_nsecs_stats)) != 0) + print_metric(config, ctxp, NULL, "%8.3f", "CPUs utilized", + avg / (ratio * evsel->scale)); + else + print_metric(config, ctxp, NULL, NULL, "CPUs utilized", 0); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_FETCH_BUBBLES)) { + double fe_bound = td_fe_bound(ctx, cpu, st); + + if (fe_bound > 0.2) + color = PERF_COLOR_RED; + print_metric(config, ctxp, color, "%8.1f%%", "frontend bound", + fe_bound * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_SLOTS_RETIRED)) { + double retiring = td_retiring(ctx, cpu, st); + + if (retiring > 0.7) + color = PERF_COLOR_GREEN; + print_metric(config, ctxp, color, "%8.1f%%", "retiring", + retiring * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_RECOVERY_BUBBLES)) { + double bad_spec = td_bad_spec(ctx, cpu, st); + + if (bad_spec > 0.1) + color = PERF_COLOR_RED; + print_metric(config, ctxp, color, "%8.1f%%", "bad speculation", + bad_spec * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_SLOTS_ISSUED)) { + double be_bound = td_be_bound(ctx, cpu, st); + const char *name = "backend bound"; + static int have_recovery_bubbles = -1; + + /* In case the CPU does not support topdown-recovery-bubbles */ + if (have_recovery_bubbles < 0) + have_recovery_bubbles = pmu_have_event("cpu", + "topdown-recovery-bubbles"); + if (!have_recovery_bubbles) + name = "backend bound/bad spec"; + + if (be_bound > 0.2) + color = PERF_COLOR_RED; + if (td_total_slots(ctx, cpu, st) > 0) + print_metric(config, ctxp, color, "%8.1f%%", name, + be_bound * 100.); + else + print_metric(config, ctxp, NULL, NULL, name, 0); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_RETIRING) && + full_td(ctx, cpu, st)) { + double retiring = td_metric_ratio(ctx, cpu, + STAT_TOPDOWN_RETIRING, st); + + if (retiring > 0.7) + color = PERF_COLOR_GREEN; + print_metric(config, ctxp, color, "%8.1f%%", "retiring", + retiring * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_FE_BOUND) && + full_td(ctx, cpu, st)) { + double fe_bound = td_metric_ratio(ctx, cpu, + STAT_TOPDOWN_FE_BOUND, st); + + if (fe_bound > 0.2) + color = PERF_COLOR_RED; + print_metric(config, ctxp, color, "%8.1f%%", "frontend bound", + fe_bound * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_BE_BOUND) && + full_td(ctx, cpu, st)) { + double be_bound = td_metric_ratio(ctx, cpu, + STAT_TOPDOWN_BE_BOUND, st); + + if (be_bound > 0.2) + color = PERF_COLOR_RED; + print_metric(config, ctxp, color, "%8.1f%%", "backend bound", + be_bound * 100.); + } else if (perf_stat_evsel__is(evsel, TOPDOWN_BAD_SPEC) && + full_td(ctx, cpu, st)) { + double bad_spec = td_metric_ratio(ctx, cpu, + STAT_TOPDOWN_BAD_SPEC, st); + + if (bad_spec > 0.1) + color = PERF_COLOR_RED; + print_metric(config, ctxp, color, "%8.1f%%", "bad speculation", + bad_spec * 100.); + } else if (evsel->metric_expr) { + generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL, + evsel->name, evsel->metric_name, NULL, 1, cpu, out, st); + } else if (runtime_stat_n(st, STAT_NSECS, 0, cpu) != 0) { + char unit = 'M'; + char unit_buf[10]; + + total = runtime_stat_avg(st, STAT_NSECS, 0, cpu); + + if (total) + ratio = 1000.0 * avg / total; + if (ratio < 0.001) { + ratio *= 1000; + unit = 'K'; + } + snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit); + print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio); + } else if (perf_stat_evsel__is(evsel, SMI_NUM)) { + print_smi_cost(config, cpu, evsel, out, st); + } else { + num = 0; + } + + if ((me = metricgroup__lookup(metric_events, evsel, false)) != NULL) { + struct metric_expr *mexp; + + list_for_each_entry (mexp, &me->head, nd) { + if (num++ > 0) + out->new_line(config, ctxp); + generic_metric(config, mexp->metric_expr, mexp->metric_events, + mexp->metric_refs, evsel->name, mexp->metric_name, + mexp->metric_unit, mexp->runtime, cpu, out, st); + } + } + if (num == 0) + print_metric(config, ctxp, NULL, NULL, NULL, 0); +} -- cgit v1.2.3