diff options
Diffstat (limited to 'tools/perf/util/evlist.c')
-rw-r--r-- | tools/perf/util/evlist.c | 1422 |
1 files changed, 1019 insertions, 403 deletions
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index 1548237b6558..6612b00949e7 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -15,14 +15,20 @@ #include "target.h" #include "evlist.h" #include "evsel.h" +#include "record.h" #include "debug.h" #include "units.h" +#include "bpf_counter.h" #include <internal/lib.h> // page_size #include "affinity.h" #include "../perf.h" #include "asm/bug.h" #include "bpf-event.h" #include "util/string2.h" +#include "util/perf_api_probe.h" +#include "util/evsel_fprintf.h" +#include "util/evlist-hybrid.h" +#include "util/pmu.h" #include <signal.h> #include <unistd.h> #include <sched.h> @@ -34,12 +40,15 @@ #include <fcntl.h> #include <sys/ioctl.h> #include <sys/mman.h> +#include <sys/prctl.h> +#include <sys/timerfd.h> #include <linux/bitops.h> #include <linux/hash.h> #include <linux/log2.h> #include <linux/err.h> #include <linux/string.h> +#include <linux/time64.h> #include <linux/zalloc.h> #include <perf/evlist.h> #include <perf/evsel.h> @@ -62,6 +71,9 @@ void evlist__init(struct evlist *evlist, struct perf_cpu_map *cpus, perf_evlist__set_maps(&evlist->core, cpus, threads); evlist->workload.pid = -1; evlist->bkw_mmap_state = BKW_MMAP_NOTREADY; + evlist->ctl_fd.fd = -1; + evlist->ctl_fd.ack = -1; + evlist->ctl_fd.pos = -1; } struct evlist *evlist__new(void) @@ -74,11 +86,11 @@ struct evlist *evlist__new(void) return evlist; } -struct evlist *perf_evlist__new_default(void) +struct evlist *evlist__new_default(void) { struct evlist *evlist = evlist__new(); - if (evlist && perf_evlist__add_default(evlist)) { + if (evlist && evlist__add_default(evlist)) { evlist__delete(evlist); evlist = NULL; } @@ -86,11 +98,11 @@ struct evlist *perf_evlist__new_default(void) return evlist; } -struct evlist *perf_evlist__new_dummy(void) +struct evlist *evlist__new_dummy(void) { struct evlist *evlist = evlist__new(); - if (evlist && perf_evlist__add_dummy(evlist)) { + if (evlist && evlist__add_dummy(evlist)) { evlist__delete(evlist); evlist = NULL; } @@ -99,13 +111,13 @@ struct evlist *perf_evlist__new_dummy(void) } /** - * perf_evlist__set_id_pos - set the positions of event ids. + * evlist__set_id_pos - set the positions of event ids. * @evlist: selected event list * * Events with compatible sample types all have the same id_pos * and is_pos. For convenience, put a copy on evlist. */ -void perf_evlist__set_id_pos(struct evlist *evlist) +void evlist__set_id_pos(struct evlist *evlist) { struct evsel *first = evlist__first(evlist); @@ -113,14 +125,14 @@ void perf_evlist__set_id_pos(struct evlist *evlist) evlist->is_pos = first->is_pos; } -static void perf_evlist__update_id_pos(struct evlist *evlist) +static void evlist__update_id_pos(struct evlist *evlist) { struct evsel *evsel; evlist__for_each_entry(evlist, evsel) - perf_evsel__calc_id_pos(evsel); + evsel__calc_id_pos(evsel); - perf_evlist__set_id_pos(evlist); + evlist__set_id_pos(evlist); } static void evlist__purge(struct evlist *evlist) @@ -138,6 +150,7 @@ static void evlist__purge(struct evlist *evlist) void evlist__exit(struct evlist *evlist) { + event_enable_timer__exit(&evlist->eet); zfree(&evlist->mmap); zfree(&evlist->overwrite_mmap); perf_evlist__exit(&evlist->core); @@ -157,14 +170,12 @@ void evlist__delete(struct evlist *evlist) void evlist__add(struct evlist *evlist, struct evsel *entry) { - entry->evlist = evlist; - entry->idx = evlist->core.nr_entries; - entry->tracking = !entry->idx; - perf_evlist__add(&evlist->core, &entry->core); + entry->evlist = evlist; + entry->tracking = !entry->core.idx; if (evlist->core.nr_entries == 1) - perf_evlist__set_id_pos(evlist); + evlist__set_id_pos(evlist); } void evlist__remove(struct evlist *evlist, struct evsel *evsel) @@ -173,27 +184,36 @@ void evlist__remove(struct evlist *evlist, struct evsel *evsel) perf_evlist__remove(&evlist->core, &evsel->core); } -void perf_evlist__splice_list_tail(struct evlist *evlist, - struct list_head *list) +void evlist__splice_list_tail(struct evlist *evlist, struct list_head *list) { - struct evsel *evsel, *temp; + while (!list_empty(list)) { + struct evsel *evsel, *temp, *leader = NULL; - __evlist__for_each_entry_safe(list, temp, evsel) { - list_del_init(&evsel->core.node); - evlist__add(evlist, evsel); + __evlist__for_each_entry_safe(list, temp, evsel) { + list_del_init(&evsel->core.node); + evlist__add(evlist, evsel); + leader = evsel; + break; + } + + __evlist__for_each_entry_safe(list, temp, evsel) { + if (evsel__has_leader(evsel, leader)) { + list_del_init(&evsel->core.node); + evlist__add(evlist, evsel); + } + } } } int __evlist__set_tracepoints_handlers(struct evlist *evlist, const struct evsel_str_handler *assocs, size_t nr_assocs) { - struct evsel *evsel; size_t i; int err; for (i = 0; i < nr_assocs; i++) { // Adding a handler for an event not in this evlist, just ignore it. - evsel = perf_evlist__find_tracepoint_by_name(evlist, assocs[i].name); + struct evsel *evsel = evlist__find_tracepoint_by_name(evlist, assocs[i].name); if (evsel == NULL) continue; @@ -208,32 +228,17 @@ out: return err; } -void __perf_evlist__set_leader(struct list_head *list) -{ - struct evsel *evsel, *leader; - - leader = list_entry(list->next, struct evsel, core.node); - evsel = list_entry(list->prev, struct evsel, core.node); - - leader->core.nr_members = evsel->idx - leader->idx + 1; - - __evlist__for_each_entry(list, evsel) { - evsel->leader = leader; - } -} - -void perf_evlist__set_leader(struct evlist *evlist) +void evlist__set_leader(struct evlist *evlist) { - if (evlist->core.nr_entries) { - evlist->nr_groups = evlist->core.nr_entries > 1 ? 1 : 0; - __perf_evlist__set_leader(&evlist->core.entries); - } + perf_evlist__set_leader(&evlist->core); } -int __perf_evlist__add_default(struct evlist *evlist, bool precise) +int __evlist__add_default(struct evlist *evlist, bool precise) { - struct evsel *evsel = perf_evsel__new_cycles(precise); + struct evsel *evsel; + evsel = evsel__new_cycles(precise, PERF_TYPE_HARDWARE, + PERF_COUNT_HW_CPU_CYCLES); if (evsel == NULL) return -ENOMEM; @@ -241,14 +246,20 @@ int __perf_evlist__add_default(struct evlist *evlist, bool precise) return 0; } -int perf_evlist__add_dummy(struct evlist *evlist) +static struct evsel *evlist__dummy_event(struct evlist *evlist) { struct perf_event_attr attr = { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_DUMMY, .size = sizeof(attr), /* to capture ABI version */ }; - struct evsel *evsel = perf_evsel__new_idx(&attr, evlist->core.nr_entries); + + return evsel__new_idx(&attr, evlist->core.nr_entries); +} + +int evlist__add_dummy(struct evlist *evlist) +{ + struct evsel *evsel = evlist__dummy_event(evlist); if (evsel == NULL) return -ENOMEM; @@ -257,21 +268,57 @@ int perf_evlist__add_dummy(struct evlist *evlist) return 0; } -static int evlist__add_attrs(struct evlist *evlist, - struct perf_event_attr *attrs, size_t nr_attrs) +struct evsel *evlist__add_aux_dummy(struct evlist *evlist, bool system_wide) +{ + struct evsel *evsel = evlist__dummy_event(evlist); + + if (!evsel) + return NULL; + + evsel->core.attr.exclude_kernel = 1; + evsel->core.attr.exclude_guest = 1; + evsel->core.attr.exclude_hv = 1; + evsel->core.attr.freq = 0; + evsel->core.attr.sample_period = 1; + evsel->core.system_wide = system_wide; + evsel->no_aux_samples = true; + evsel->name = strdup("dummy:u"); + + evlist__add(evlist, evsel); + return evsel; +} + +struct evsel *evlist__add_sched_switch(struct evlist *evlist, bool system_wide) +{ + struct evsel *evsel = evsel__newtp_idx("sched", "sched_switch", 0); + + if (IS_ERR(evsel)) + return evsel; + + evsel__set_sample_bit(evsel, CPU); + evsel__set_sample_bit(evsel, TIME); + + evsel->core.system_wide = system_wide; + evsel->no_aux_samples = true; + + evlist__add(evlist, evsel); + return evsel; +}; + +int evlist__add_attrs(struct evlist *evlist, struct perf_event_attr *attrs, size_t nr_attrs) { struct evsel *evsel, *n; LIST_HEAD(head); size_t i; for (i = 0; i < nr_attrs; i++) { - evsel = perf_evsel__new_idx(attrs + i, evlist->core.nr_entries + i); + evsel = evsel__new_idx(attrs + i, evlist->core.nr_entries + i); if (evsel == NULL) goto out_delete_partial_list; list_add_tail(&evsel->core.node, &head); } - perf_evlist__splice_list_tail(evlist, &head); + evlist__splice_list_tail(evlist, &head); return 0; @@ -281,8 +328,7 @@ out_delete_partial_list: return -1; } -int __perf_evlist__add_default_attrs(struct evlist *evlist, - struct perf_event_attr *attrs, size_t nr_attrs) +int __evlist__add_default_attrs(struct evlist *evlist, struct perf_event_attr *attrs, size_t nr_attrs) { size_t i; @@ -292,8 +338,17 @@ int __perf_evlist__add_default_attrs(struct evlist *evlist, return evlist__add_attrs(evlist, attrs, nr_attrs); } -struct evsel * -perf_evlist__find_tracepoint_by_id(struct evlist *evlist, int id) +__weak int arch_evlist__add_default_attrs(struct evlist *evlist, + struct perf_event_attr *attrs, + size_t nr_attrs) +{ + if (!nr_attrs) + return 0; + + return __evlist__add_default_attrs(evlist, attrs, nr_attrs); +} + +struct evsel *evlist__find_tracepoint_by_id(struct evlist *evlist, int id) { struct evsel *evsel; @@ -306,9 +361,7 @@ perf_evlist__find_tracepoint_by_id(struct evlist *evlist, int id) return NULL; } -struct evsel * -perf_evlist__find_tracepoint_by_name(struct evlist *evlist, - const char *name) +struct evsel *evlist__find_tracepoint_by_name(struct evlist *evlist, const char *name) { struct evsel *evsel; @@ -321,10 +374,9 @@ perf_evlist__find_tracepoint_by_name(struct evlist *evlist, return NULL; } -int perf_evlist__add_newtp(struct evlist *evlist, - const char *sys, const char *name, void *handler) +int evlist__add_newtp(struct evlist *evlist, const char *sys, const char *name, void *handler) { - struct evsel *evsel = perf_evsel__newtp(sys, name); + struct evsel *evsel = evsel__newtp(sys, name); if (IS_ERR(evsel)) return -1; @@ -334,161 +386,231 @@ int perf_evlist__add_newtp(struct evlist *evlist, return 0; } -static int perf_evlist__nr_threads(struct evlist *evlist, - struct evsel *evsel) +struct evlist_cpu_iterator evlist__cpu_begin(struct evlist *evlist, struct affinity *affinity) { - if (evsel->core.system_wide) - return 1; - else - return perf_thread_map__nr(evlist->core.threads); + struct evlist_cpu_iterator itr = { + .container = evlist, + .evsel = NULL, + .cpu_map_idx = 0, + .evlist_cpu_map_idx = 0, + .evlist_cpu_map_nr = perf_cpu_map__nr(evlist->core.all_cpus), + .cpu = (struct perf_cpu){ .cpu = -1}, + .affinity = affinity, + }; + + if (evlist__empty(evlist)) { + /* Ensure the empty list doesn't iterate. */ + itr.evlist_cpu_map_idx = itr.evlist_cpu_map_nr; + } else { + itr.evsel = evlist__first(evlist); + if (itr.affinity) { + itr.cpu = perf_cpu_map__cpu(evlist->core.all_cpus, 0); + affinity__set(itr.affinity, itr.cpu.cpu); + itr.cpu_map_idx = perf_cpu_map__idx(itr.evsel->core.cpus, itr.cpu); + /* + * If this CPU isn't in the evsel's cpu map then advance + * through the list. + */ + if (itr.cpu_map_idx == -1) + evlist_cpu_iterator__next(&itr); + } + } + return itr; +} + +void evlist_cpu_iterator__next(struct evlist_cpu_iterator *evlist_cpu_itr) +{ + while (evlist_cpu_itr->evsel != evlist__last(evlist_cpu_itr->container)) { + evlist_cpu_itr->evsel = evsel__next(evlist_cpu_itr->evsel); + evlist_cpu_itr->cpu_map_idx = + perf_cpu_map__idx(evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); + if (evlist_cpu_itr->cpu_map_idx != -1) + return; + } + evlist_cpu_itr->evlist_cpu_map_idx++; + if (evlist_cpu_itr->evlist_cpu_map_idx < evlist_cpu_itr->evlist_cpu_map_nr) { + evlist_cpu_itr->evsel = evlist__first(evlist_cpu_itr->container); + evlist_cpu_itr->cpu = + perf_cpu_map__cpu(evlist_cpu_itr->container->core.all_cpus, + evlist_cpu_itr->evlist_cpu_map_idx); + if (evlist_cpu_itr->affinity) + affinity__set(evlist_cpu_itr->affinity, evlist_cpu_itr->cpu.cpu); + evlist_cpu_itr->cpu_map_idx = + perf_cpu_map__idx(evlist_cpu_itr->evsel->core.cpus, + evlist_cpu_itr->cpu); + /* + * If this CPU isn't in the evsel's cpu map then advance through + * the list. + */ + if (evlist_cpu_itr->cpu_map_idx == -1) + evlist_cpu_iterator__next(evlist_cpu_itr); + } } -void evlist__cpu_iter_start(struct evlist *evlist) +bool evlist_cpu_iterator__end(const struct evlist_cpu_iterator *evlist_cpu_itr) { - struct evsel *pos; - - /* - * Reset the per evsel cpu_iter. This is needed because - * each evsel's cpumap may have a different index space, - * and some operations need the index to modify - * the FD xyarray (e.g. open, close) - */ - evlist__for_each_entry(evlist, pos) - pos->cpu_iter = 0; + return evlist_cpu_itr->evlist_cpu_map_idx >= evlist_cpu_itr->evlist_cpu_map_nr; } -bool evsel__cpu_iter_skip_no_inc(struct evsel *ev, int cpu) +static int evsel__strcmp(struct evsel *pos, char *evsel_name) { - if (ev->cpu_iter >= ev->core.cpus->nr) - return true; - if (cpu >= 0 && ev->core.cpus->map[ev->cpu_iter] != cpu) - return true; - return false; + if (!evsel_name) + return 0; + if (evsel__is_dummy_event(pos)) + return 1; + return strcmp(pos->name, evsel_name); } -bool evsel__cpu_iter_skip(struct evsel *ev, int cpu) +static int evlist__is_enabled(struct evlist *evlist) { - if (!evsel__cpu_iter_skip_no_inc(ev, cpu)) { - ev->cpu_iter++; - return false; + struct evsel *pos; + + evlist__for_each_entry(evlist, pos) { + if (!evsel__is_group_leader(pos) || !pos->core.fd) + continue; + /* If at least one event is enabled, evlist is enabled. */ + if (!pos->disabled) + return true; } - return true; + return false; } -void evlist__disable(struct evlist *evlist) +static void __evlist__disable(struct evlist *evlist, char *evsel_name, bool excl_dummy) { struct evsel *pos; - struct affinity affinity; - int cpu, i; - - if (affinity__setup(&affinity) < 0) - return; - - evlist__for_each_cpu(evlist, i, cpu) { - affinity__set(&affinity, cpu); - - evlist__for_each_entry(evlist, pos) { - if (evsel__cpu_iter_skip(pos, cpu)) + struct evlist_cpu_iterator evlist_cpu_itr; + struct affinity saved_affinity, *affinity = NULL; + bool has_imm = false; + + // See explanation in evlist__close() + if (!cpu_map__is_dummy(evlist->core.user_requested_cpus)) { + if (affinity__setup(&saved_affinity) < 0) + return; + affinity = &saved_affinity; + } + + /* Disable 'immediate' events last */ + for (int imm = 0; imm <= 1; imm++) { + evlist__for_each_cpu(evlist_cpu_itr, evlist, affinity) { + pos = evlist_cpu_itr.evsel; + if (evsel__strcmp(pos, evsel_name)) + continue; + if (pos->disabled || !evsel__is_group_leader(pos) || !pos->core.fd) continue; - if (pos->disabled || !perf_evsel__is_group_leader(pos) || !pos->core.fd) + if (excl_dummy && evsel__is_dummy_event(pos)) continue; - evsel__disable_cpu(pos, pos->cpu_iter - 1); + if (pos->immediate) + has_imm = true; + if (pos->immediate != imm) + continue; + evsel__disable_cpu(pos, evlist_cpu_itr.cpu_map_idx); } + if (!has_imm) + break; } - affinity__cleanup(&affinity); + + affinity__cleanup(affinity); evlist__for_each_entry(evlist, pos) { - if (!perf_evsel__is_group_leader(pos) || !pos->core.fd) + if (evsel__strcmp(pos, evsel_name)) + continue; + if (!evsel__is_group_leader(pos) || !pos->core.fd) + continue; + if (excl_dummy && evsel__is_dummy_event(pos)) continue; pos->disabled = true; } - evlist->enabled = false; + /* + * If we disabled only single event, we need to check + * the enabled state of the evlist manually. + */ + if (evsel_name) + evlist->enabled = evlist__is_enabled(evlist); + else + evlist->enabled = false; } -void evlist__enable(struct evlist *evlist) +void evlist__disable(struct evlist *evlist) { - struct evsel *pos; - struct affinity affinity; - int cpu, i; + __evlist__disable(evlist, NULL, false); +} - if (affinity__setup(&affinity) < 0) - return; +void evlist__disable_non_dummy(struct evlist *evlist) +{ + __evlist__disable(evlist, NULL, true); +} - evlist__for_each_cpu(evlist, i, cpu) { - affinity__set(&affinity, cpu); +void evlist__disable_evsel(struct evlist *evlist, char *evsel_name) +{ + __evlist__disable(evlist, evsel_name, false); +} - evlist__for_each_entry(evlist, pos) { - if (evsel__cpu_iter_skip(pos, cpu)) - continue; - if (!perf_evsel__is_group_leader(pos) || !pos->core.fd) - continue; - evsel__enable_cpu(pos, pos->cpu_iter - 1); - } +static void __evlist__enable(struct evlist *evlist, char *evsel_name, bool excl_dummy) +{ + struct evsel *pos; + struct evlist_cpu_iterator evlist_cpu_itr; + struct affinity saved_affinity, *affinity = NULL; + + // See explanation in evlist__close() + if (!cpu_map__is_dummy(evlist->core.user_requested_cpus)) { + if (affinity__setup(&saved_affinity) < 0) + return; + affinity = &saved_affinity; } - affinity__cleanup(&affinity); + + evlist__for_each_cpu(evlist_cpu_itr, evlist, affinity) { + pos = evlist_cpu_itr.evsel; + if (evsel__strcmp(pos, evsel_name)) + continue; + if (!evsel__is_group_leader(pos) || !pos->core.fd) + continue; + if (excl_dummy && evsel__is_dummy_event(pos)) + continue; + evsel__enable_cpu(pos, evlist_cpu_itr.cpu_map_idx); + } + affinity__cleanup(affinity); evlist__for_each_entry(evlist, pos) { - if (!perf_evsel__is_group_leader(pos) || !pos->core.fd) + if (evsel__strcmp(pos, evsel_name)) + continue; + if (!evsel__is_group_leader(pos) || !pos->core.fd) + continue; + if (excl_dummy && evsel__is_dummy_event(pos)) continue; pos->disabled = false; } + /* + * Even single event sets the 'enabled' for evlist, + * so the toggle can work properly and toggle to + * 'disabled' state. + */ evlist->enabled = true; } -void perf_evlist__toggle_enable(struct evlist *evlist) +void evlist__enable(struct evlist *evlist) { - (evlist->enabled ? evlist__disable : evlist__enable)(evlist); + __evlist__enable(evlist, NULL, false); } -static int perf_evlist__enable_event_cpu(struct evlist *evlist, - struct evsel *evsel, int cpu) +void evlist__enable_non_dummy(struct evlist *evlist) { - int thread; - int nr_threads = perf_evlist__nr_threads(evlist, evsel); - - if (!evsel->core.fd) - return -EINVAL; - - for (thread = 0; thread < nr_threads; thread++) { - int err = ioctl(FD(evsel, cpu, thread), PERF_EVENT_IOC_ENABLE, 0); - if (err) - return err; - } - return 0; + __evlist__enable(evlist, NULL, true); } -static int perf_evlist__enable_event_thread(struct evlist *evlist, - struct evsel *evsel, - int thread) +void evlist__enable_evsel(struct evlist *evlist, char *evsel_name) { - int cpu; - int nr_cpus = perf_cpu_map__nr(evlist->core.cpus); - - if (!evsel->core.fd) - return -EINVAL; - - for (cpu = 0; cpu < nr_cpus; cpu++) { - int err = ioctl(FD(evsel, cpu, thread), PERF_EVENT_IOC_ENABLE, 0); - if (err) - return err; - } - return 0; + __evlist__enable(evlist, evsel_name, false); } -int perf_evlist__enable_event_idx(struct evlist *evlist, - struct evsel *evsel, int idx) +void evlist__toggle_enable(struct evlist *evlist) { - bool per_cpu_mmaps = !perf_cpu_map__empty(evlist->core.cpus); - - if (per_cpu_mmaps) - return perf_evlist__enable_event_cpu(evlist, evsel, idx); - else - return perf_evlist__enable_event_thread(evlist, evsel, idx); + (evlist->enabled ? evlist__disable : evlist__enable)(evlist); } int evlist__add_pollfd(struct evlist *evlist, int fd) { - return perf_evlist__add_pollfd(&evlist->core, fd, NULL, POLLIN); + return perf_evlist__add_pollfd(&evlist->core, fd, NULL, POLLIN, fdarray_flag__default); } int evlist__filter_pollfd(struct evlist *evlist, short revents_and_mask) @@ -496,12 +618,21 @@ int evlist__filter_pollfd(struct evlist *evlist, short revents_and_mask) return perf_evlist__filter_pollfd(&evlist->core, revents_and_mask); } +#ifdef HAVE_EVENTFD_SUPPORT +int evlist__add_wakeup_eventfd(struct evlist *evlist, int fd) +{ + return perf_evlist__add_pollfd(&evlist->core, fd, NULL, POLLIN, + fdarray_flag__nonfilterable | + fdarray_flag__non_perf_event); +} +#endif + int evlist__poll(struct evlist *evlist, int timeout) { return perf_evlist__poll(&evlist->core, timeout); } -struct perf_sample_id *perf_evlist__id2sid(struct evlist *evlist, u64 id) +struct perf_sample_id *evlist__id2sid(struct evlist *evlist, u64 id) { struct hlist_head *head; struct perf_sample_id *sid; @@ -517,40 +648,38 @@ struct perf_sample_id *perf_evlist__id2sid(struct evlist *evlist, u64 id) return NULL; } -struct evsel *perf_evlist__id2evsel(struct evlist *evlist, u64 id) +struct evsel *evlist__id2evsel(struct evlist *evlist, u64 id) { struct perf_sample_id *sid; if (evlist->core.nr_entries == 1 || !id) return evlist__first(evlist); - sid = perf_evlist__id2sid(evlist, id); + sid = evlist__id2sid(evlist, id); if (sid) return container_of(sid->evsel, struct evsel, core); - if (!perf_evlist__sample_id_all(evlist)) + if (!evlist__sample_id_all(evlist)) return evlist__first(evlist); return NULL; } -struct evsel *perf_evlist__id2evsel_strict(struct evlist *evlist, - u64 id) +struct evsel *evlist__id2evsel_strict(struct evlist *evlist, u64 id) { struct perf_sample_id *sid; if (!id) return NULL; - sid = perf_evlist__id2sid(evlist, id); + sid = evlist__id2sid(evlist, id); if (sid) return container_of(sid->evsel, struct evsel, core); return NULL; } -static int perf_evlist__event2id(struct evlist *evlist, - union perf_event *event, u64 *id) +static int evlist__event2id(struct evlist *evlist, union perf_event *event, u64 *id) { const __u64 *array = event->sample.array; ssize_t n; @@ -570,8 +699,7 @@ static int perf_evlist__event2id(struct evlist *evlist, return 0; } -struct evsel *perf_evlist__event2evsel(struct evlist *evlist, - union perf_event *event) +struct evsel *evlist__event2evsel(struct evlist *evlist, union perf_event *event) { struct evsel *first = evlist__first(evlist); struct hlist_head *head; @@ -586,7 +714,7 @@ struct evsel *perf_evlist__event2evsel(struct evlist *evlist, event->header.type != PERF_RECORD_SAMPLE) return first; - if (perf_evlist__event2id(evlist, event, &id)) + if (evlist__event2id(evlist, event, &id)) return NULL; /* Synthesized events have an id of zero */ @@ -603,7 +731,7 @@ struct evsel *perf_evlist__event2evsel(struct evlist *evlist, return NULL; } -static int perf_evlist__set_paused(struct evlist *evlist, bool value) +static int evlist__set_paused(struct evlist *evlist, bool value) { int i; @@ -623,14 +751,14 @@ static int perf_evlist__set_paused(struct evlist *evlist, bool value) return 0; } -static int perf_evlist__pause(struct evlist *evlist) +static int evlist__pause(struct evlist *evlist) { - return perf_evlist__set_paused(evlist, true); + return evlist__set_paused(evlist, true); } -static int perf_evlist__resume(struct evlist *evlist) +static int evlist__resume(struct evlist *evlist) { - return perf_evlist__set_paused(evlist, false); + return evlist__set_paused(evlist, false); } static void evlist__munmap_nofree(struct evlist *evlist) @@ -690,13 +818,15 @@ static struct mmap *evlist__alloc_mmap(struct evlist *evlist, static void perf_evlist__mmap_cb_idx(struct perf_evlist *_evlist, + struct perf_evsel *_evsel, struct perf_mmap_param *_mp, - int idx, bool per_cpu) + int idx) { struct evlist *evlist = container_of(_evlist, struct evlist, core); struct mmap_params *mp = container_of(_mp, struct mmap_params, core); + struct evsel *evsel = container_of(_evsel, struct evsel, core); - auxtrace_mmap_params__set_idx(&mp->auxtrace_mp, evlist, idx, per_cpu); + auxtrace_mmap_params__set_idx(&mp->auxtrace_mp, evlist, evsel, idx); } static struct perf_mmap* @@ -715,7 +845,7 @@ perf_evlist__mmap_cb_get(struct perf_evlist *_evlist, bool overwrite, int idx) if (overwrite) { evlist->overwrite_mmap = maps; if (evlist->bkw_mmap_state == BKW_MMAP_NOTREADY) - perf_evlist__toggle_bkw_mmap(evlist, BKW_MMAP_RUNNING); + evlist__toggle_bkw_mmap(evlist, BKW_MMAP_RUNNING); } else { evlist->mmap = maps; } @@ -726,7 +856,7 @@ perf_evlist__mmap_cb_get(struct perf_evlist *_evlist, bool overwrite, int idx) static int perf_evlist__mmap_cb_mmap(struct perf_mmap *_map, struct perf_mmap_param *_mp, - int output, int cpu) + int output, struct perf_cpu cpu) { struct mmap *map = container_of(_map, struct mmap, core); struct mmap_params *mp = container_of(_mp, struct mmap_params, core); @@ -815,7 +945,7 @@ static long parse_pages_arg(const char *str, unsigned long min, return pages; } -int __perf_evlist__parse_mmap_pages(unsigned int *mmap_pages, const char *str) +int __evlist__parse_mmap_pages(unsigned int *mmap_pages, const char *str) { unsigned long max = UINT_MAX; long pages; @@ -833,10 +963,9 @@ int __perf_evlist__parse_mmap_pages(unsigned int *mmap_pages, const char *str) return 0; } -int perf_evlist__parse_mmap_pages(const struct option *opt, const char *str, - int unset __maybe_unused) +int evlist__parse_mmap_pages(const struct option *opt, const char *str, int unset __maybe_unused) { - return __perf_evlist__parse_mmap_pages(opt->value, str); + return __evlist__parse_mmap_pages(opt->value, str); } /** @@ -892,7 +1021,7 @@ int evlist__mmap(struct evlist *evlist, unsigned int pages) return evlist__mmap_ex(evlist, pages, 0, false, 0, PERF_AFFINITY_SYS, 1, 0); } -int perf_evlist__create_maps(struct evlist *evlist, struct target *target) +int evlist__create_maps(struct evlist *evlist, struct target *target) { bool all_threads = (target->per_thread && target->system_wide); struct perf_cpu_map *cpus; @@ -930,10 +1059,14 @@ int perf_evlist__create_maps(struct evlist *evlist, struct target *target) if (!cpus) goto out_delete_threads; - evlist->core.has_user_cpus = !!target->cpu_list; + evlist->core.has_user_cpus = !!target->cpu_list && !target->hybrid; perf_evlist__set_maps(&evlist->core, cpus, threads); + /* as evlist now has references, put count here */ + perf_cpu_map__put(cpus); + perf_thread_map__put(threads); + return 0; out_delete_threads: @@ -941,25 +1074,7 @@ out_delete_threads: return -1; } -void __perf_evlist__set_sample_bit(struct evlist *evlist, - enum perf_event_sample_format bit) -{ - struct evsel *evsel; - - evlist__for_each_entry(evlist, evsel) - __perf_evsel__set_sample_bit(evsel, bit); -} - -void __perf_evlist__reset_sample_bit(struct evlist *evlist, - enum perf_event_sample_format bit) -{ - struct evsel *evsel; - - evlist__for_each_entry(evlist, evsel) - __perf_evsel__reset_sample_bit(evsel, bit); -} - -int perf_evlist__apply_filters(struct evlist *evlist, struct evsel **err_evsel) +int evlist__apply_filters(struct evlist *evlist, struct evsel **err_evsel) { struct evsel *evsel; int err = 0; @@ -982,7 +1097,7 @@ int perf_evlist__apply_filters(struct evlist *evlist, struct evsel **err_evsel) return err; } -int perf_evlist__set_tp_filter(struct evlist *evlist, const char *filter) +int evlist__set_tp_filter(struct evlist *evlist, const char *filter) { struct evsel *evsel; int err = 0; @@ -994,7 +1109,7 @@ int perf_evlist__set_tp_filter(struct evlist *evlist, const char *filter) if (evsel->core.attr.type != PERF_TYPE_TRACEPOINT) continue; - err = perf_evsel__set_filter(evsel, filter); + err = evsel__set_filter(evsel, filter); if (err) break; } @@ -1002,7 +1117,7 @@ int perf_evlist__set_tp_filter(struct evlist *evlist, const char *filter) return err; } -int perf_evlist__append_tp_filter(struct evlist *evlist, const char *filter) +int evlist__append_tp_filter(struct evlist *evlist, const char *filter) { struct evsel *evsel; int err = 0; @@ -1014,7 +1129,7 @@ int perf_evlist__append_tp_filter(struct evlist *evlist, const char *filter) if (evsel->core.attr.type != PERF_TYPE_TRACEPOINT) continue; - err = perf_evsel__append_tp_filter(evsel, filter); + err = evsel__append_tp_filter(evsel, filter); if (err) break; } @@ -1048,35 +1163,35 @@ out_free: return NULL; } -int perf_evlist__set_tp_filter_pids(struct evlist *evlist, size_t npids, pid_t *pids) +int evlist__set_tp_filter_pids(struct evlist *evlist, size_t npids, pid_t *pids) { char *filter = asprintf__tp_filter_pids(npids, pids); - int ret = perf_evlist__set_tp_filter(evlist, filter); + int ret = evlist__set_tp_filter(evlist, filter); free(filter); return ret; } -int perf_evlist__set_tp_filter_pid(struct evlist *evlist, pid_t pid) +int evlist__set_tp_filter_pid(struct evlist *evlist, pid_t pid) { - return perf_evlist__set_tp_filter_pids(evlist, 1, &pid); + return evlist__set_tp_filter_pids(evlist, 1, &pid); } -int perf_evlist__append_tp_filter_pids(struct evlist *evlist, size_t npids, pid_t *pids) +int evlist__append_tp_filter_pids(struct evlist *evlist, size_t npids, pid_t *pids) { char *filter = asprintf__tp_filter_pids(npids, pids); - int ret = perf_evlist__append_tp_filter(evlist, filter); + int ret = evlist__append_tp_filter(evlist, filter); free(filter); return ret; } -int perf_evlist__append_tp_filter_pid(struct evlist *evlist, pid_t pid) +int evlist__append_tp_filter_pid(struct evlist *evlist, pid_t pid) { - return perf_evlist__append_tp_filter_pids(evlist, 1, &pid); + return evlist__append_tp_filter_pids(evlist, 1, &pid); } -bool perf_evlist__valid_sample_type(struct evlist *evlist) +bool evlist__valid_sample_type(struct evlist *evlist) { struct evsel *pos; @@ -1095,7 +1210,7 @@ bool perf_evlist__valid_sample_type(struct evlist *evlist) return true; } -u64 __perf_evlist__combined_sample_type(struct evlist *evlist) +u64 __evlist__combined_sample_type(struct evlist *evlist) { struct evsel *evsel; @@ -1108,13 +1223,13 @@ u64 __perf_evlist__combined_sample_type(struct evlist *evlist) return evlist->combined_sample_type; } -u64 perf_evlist__combined_sample_type(struct evlist *evlist) +u64 evlist__combined_sample_type(struct evlist *evlist) { evlist->combined_sample_type = 0; - return __perf_evlist__combined_sample_type(evlist); + return __evlist__combined_sample_type(evlist); } -u64 perf_evlist__combined_branch_type(struct evlist *evlist) +u64 evlist__combined_branch_type(struct evlist *evlist) { struct evsel *evsel; u64 branch_type = 0; @@ -1124,18 +1239,20 @@ u64 perf_evlist__combined_branch_type(struct evlist *evlist) return branch_type; } -bool perf_evlist__valid_read_format(struct evlist *evlist) +bool evlist__valid_read_format(struct evlist *evlist) { struct evsel *first = evlist__first(evlist), *pos = first; u64 read_format = first->core.attr.read_format; u64 sample_type = first->core.attr.sample_type; evlist__for_each_entry(evlist, pos) { - if (read_format != pos->core.attr.read_format) - return false; + if (read_format != pos->core.attr.read_format) { + pr_debug("Read format differs %#" PRIx64 " vs %#" PRIx64 "\n", + read_format, (u64)pos->core.attr.read_format); + } } - /* PERF_SAMPLE_READ imples PERF_FORMAT_ID. */ + /* PERF_SAMPLE_READ implies PERF_FORMAT_ID. */ if ((sample_type & PERF_SAMPLE_READ) && !(read_format & PERF_FORMAT_ID)) { return false; @@ -1144,40 +1261,14 @@ bool perf_evlist__valid_read_format(struct evlist *evlist) return true; } -u16 perf_evlist__id_hdr_size(struct evlist *evlist) +u16 evlist__id_hdr_size(struct evlist *evlist) { struct evsel *first = evlist__first(evlist); - struct perf_sample *data; - u64 sample_type; - u16 size = 0; - if (!first->core.attr.sample_id_all) - goto out; - - sample_type = first->core.attr.sample_type; - - if (sample_type & PERF_SAMPLE_TID) - size += sizeof(data->tid) * 2; - - if (sample_type & PERF_SAMPLE_TIME) - size += sizeof(data->time); - - if (sample_type & PERF_SAMPLE_ID) - size += sizeof(data->id); - - if (sample_type & PERF_SAMPLE_STREAM_ID) - size += sizeof(data->stream_id); - - if (sample_type & PERF_SAMPLE_CPU) - size += sizeof(data->cpu) * 2; - - if (sample_type & PERF_SAMPLE_IDENTIFIER) - size += sizeof(data->id); -out: - return size; + return first->core.attr.sample_id_all ? evsel__id_hdr_size(first) : 0; } -bool perf_evlist__valid_sample_id_all(struct evlist *evlist) +bool evlist__valid_sample_id_all(struct evlist *evlist) { struct evsel *first = evlist__first(evlist), *pos = first; @@ -1189,14 +1280,13 @@ bool perf_evlist__valid_sample_id_all(struct evlist *evlist) return true; } -bool perf_evlist__sample_id_all(struct evlist *evlist) +bool evlist__sample_id_all(struct evlist *evlist) { struct evsel *first = evlist__first(evlist); return first->core.attr.sample_id_all; } -void perf_evlist__set_selected(struct evlist *evlist, - struct evsel *evsel) +void evlist__set_selected(struct evlist *evlist, struct evsel *evsel) { evlist->selected = evsel; } @@ -1204,14 +1294,15 @@ void perf_evlist__set_selected(struct evlist *evlist, void evlist__close(struct evlist *evlist) { struct evsel *evsel; + struct evlist_cpu_iterator evlist_cpu_itr; struct affinity affinity; - int cpu, i; /* - * With perf record core.cpus is usually NULL. + * With perf record core.user_requested_cpus is usually NULL. * Use the old method to handle this for now. */ - if (!evlist->core.cpus) { + if (!evlist->core.user_requested_cpus || + cpu_map__is_dummy(evlist->core.user_requested_cpus)) { evlist__for_each_entry_reverse(evlist, evsel) evsel__close(evsel); return; @@ -1219,27 +1310,24 @@ void evlist__close(struct evlist *evlist) if (affinity__setup(&affinity) < 0) return; - evlist__for_each_cpu(evlist, i, cpu) { - affinity__set(&affinity, cpu); - evlist__for_each_entry_reverse(evlist, evsel) { - if (evsel__cpu_iter_skip(evsel, cpu)) - continue; - perf_evsel__close_cpu(&evsel->core, evsel->cpu_iter - 1); - } + evlist__for_each_cpu(evlist_cpu_itr, evlist, &affinity) { + perf_evsel__close_cpu(&evlist_cpu_itr.evsel->core, + evlist_cpu_itr.cpu_map_idx); } + affinity__cleanup(&affinity); evlist__for_each_entry_reverse(evlist, evsel) { perf_evsel__free_fd(&evsel->core); perf_evsel__free_id(&evsel->core); } + perf_evlist__reset_id_hash(&evlist->core); } -static int perf_evlist__create_syswide_maps(struct evlist *evlist) +static int evlist__create_syswide_maps(struct evlist *evlist) { struct perf_cpu_map *cpus; struct perf_thread_map *threads; - int err = -ENOMEM; /* * Try reading /sys/devices/system/cpu/online to get @@ -1259,11 +1347,12 @@ static int perf_evlist__create_syswide_maps(struct evlist *evlist) goto out_put; perf_evlist__set_maps(&evlist->core, cpus, threads); -out: - return err; + + perf_thread_map__put(threads); out_put: perf_cpu_map__put(cpus); - goto out; +out: + return -ENOMEM; } int evlist__open(struct evlist *evlist) @@ -1275,13 +1364,13 @@ int evlist__open(struct evlist *evlist) * Default: one fd per CPU, all threads, aka systemwide * as sys_perf_event_open(cpu = -1, thread = -1) is EINVAL */ - if (evlist->core.threads == NULL && evlist->core.cpus == NULL) { - err = perf_evlist__create_syswide_maps(evlist); + if (evlist->core.threads == NULL && evlist->core.user_requested_cpus == NULL) { + err = evlist__create_syswide_maps(evlist); if (err < 0) goto out_err; } - perf_evlist__update_id_pos(evlist); + evlist__update_id_pos(evlist); evlist__for_each_entry(evlist, evsel) { err = evsel__open(evsel, evsel->core.cpus, evsel->core.threads); @@ -1296,9 +1385,8 @@ out_err: return err; } -int perf_evlist__prepare_workload(struct evlist *evlist, struct target *target, - const char *argv[], bool pipe_output, - void (*exec_error)(int signo, siginfo_t *info, void *ucontext)) +int evlist__prepare_workload(struct evlist *evlist, struct target *target, const char *argv[], + bool pipe_output, void (*exec_error)(int signo, siginfo_t *info, void *ucontext)) { int child_ready_pipe[2], go_pipe[2]; char bf; @@ -1332,6 +1420,13 @@ int perf_evlist__prepare_workload(struct evlist *evlist, struct target *target, fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); /* + * Change the name of this process not to confuse --exclude-perf users + * that sees 'perf' in the window up to the execvp() and thinks that + * perf samples are not being excluded. + */ + prctl(PR_SET_NAME, "perf-exec"); + + /* * Tell the parent we're ready to go */ close(child_ready_pipe[1]); @@ -1343,7 +1438,7 @@ int perf_evlist__prepare_workload(struct evlist *evlist, struct target *target, /* * The parent will ask for the execvp() to be performed by * writing exactly one byte, in workload.cork_fd, usually via - * perf_evlist__start_workload(). + * evlist__start_workload(). * * For cancelling the workload without actually running it, * the parent will just close workload.cork_fd, without writing @@ -1410,7 +1505,7 @@ out_close_ready_pipe: return -1; } -int perf_evlist__start_workload(struct evlist *evlist) +int evlist__start_workload(struct evlist *evlist) { if (evlist->workload.cork_fd > 0) { char bf = 0; @@ -1429,29 +1524,37 @@ int perf_evlist__start_workload(struct evlist *evlist) return 0; } -int perf_evlist__parse_sample(struct evlist *evlist, union perf_event *event, - struct perf_sample *sample) +int evlist__parse_sample(struct evlist *evlist, union perf_event *event, struct perf_sample *sample) { - struct evsel *evsel = perf_evlist__event2evsel(evlist, event); + struct evsel *evsel = evlist__event2evsel(evlist, event); + int ret; if (!evsel) return -EFAULT; - return perf_evsel__parse_sample(evsel, event, sample); + ret = evsel__parse_sample(evsel, event, sample); + if (ret) + return ret; + if (perf_guest && sample->id) { + struct perf_sample_id *sid = evlist__id2sid(evlist, sample->id); + + if (sid) { + sample->machine_pid = sid->machine_pid; + sample->vcpu = sid->vcpu.cpu; + } + } + return 0; } -int perf_evlist__parse_sample_timestamp(struct evlist *evlist, - union perf_event *event, - u64 *timestamp) +int evlist__parse_sample_timestamp(struct evlist *evlist, union perf_event *event, u64 *timestamp) { - struct evsel *evsel = perf_evlist__event2evsel(evlist, event); + struct evsel *evsel = evlist__event2evsel(evlist, event); if (!evsel) return -EFAULT; - return perf_evsel__parse_sample_timestamp(evsel, event, timestamp); + return evsel__parse_sample_timestamp(evsel, event, timestamp); } -int perf_evlist__strerror_open(struct evlist *evlist, - int err, char *buf, size_t size) +int evlist__strerror_open(struct evlist *evlist, int err, char *buf, size_t size) { int printed, value; char sbuf[STRERR_BUFSIZE], *emsg = str_error_r(err, sbuf, sizeof(sbuf)); @@ -1504,7 +1607,7 @@ out_default: return 0; } -int perf_evlist__strerror_mmap(struct evlist *evlist, int err, char *buf, size_t size) +int evlist__strerror_mmap(struct evlist *evlist, int err, char *buf, size_t size) { char sbuf[STRERR_BUFSIZE], *emsg = str_error_r(err, sbuf, sizeof(sbuf)); int pages_attempted = evlist->core.mmap_len / 1024, pages_max_per_user, printed = 0; @@ -1535,8 +1638,7 @@ int perf_evlist__strerror_mmap(struct evlist *evlist, int err, char *buf, size_t return 0; } -void perf_evlist__to_front(struct evlist *evlist, - struct evsel *move_evsel) +void evlist__to_front(struct evlist *evlist, struct evsel *move_evsel) { struct evsel *evsel, *n; LIST_HEAD(move); @@ -1545,15 +1647,26 @@ void perf_evlist__to_front(struct evlist *evlist, return; evlist__for_each_entry_safe(evlist, n, evsel) { - if (evsel->leader == move_evsel->leader) + if (evsel__leader(evsel) == evsel__leader(move_evsel)) list_move_tail(&evsel->core.node, &move); } list_splice(&move, &evlist->core.entries); } -void perf_evlist__set_tracking_event(struct evlist *evlist, - struct evsel *tracking_evsel) +struct evsel *evlist__get_tracking_event(struct evlist *evlist) +{ + struct evsel *evsel; + + evlist__for_each_entry(evlist, evsel) { + if (evsel->tracking) + return evsel; + } + + return evlist__first(evlist); +} + +void evlist__set_tracking_event(struct evlist *evlist, struct evsel *tracking_evsel) { struct evsel *evsel; @@ -1568,9 +1681,7 @@ void perf_evlist__set_tracking_event(struct evlist *evlist, tracking_evsel->tracking = true; } -struct evsel * -perf_evlist__find_evsel_by_str(struct evlist *evlist, - const char *str) +struct evsel *evlist__find_evsel_by_str(struct evlist *evlist, const char *str) { struct evsel *evsel; @@ -1584,8 +1695,7 @@ perf_evlist__find_evsel_by_str(struct evlist *evlist, return NULL; } -void perf_evlist__toggle_bkw_mmap(struct evlist *evlist, - enum bkw_mmap_state state) +void evlist__toggle_bkw_mmap(struct evlist *evlist, enum bkw_mmap_state state) { enum bkw_mmap_state old_state = evlist->bkw_mmap_state; enum action { @@ -1628,10 +1738,10 @@ void perf_evlist__toggle_bkw_mmap(struct evlist *evlist, switch (action) { case PAUSE: - perf_evlist__pause(evlist); + evlist__pause(evlist); break; case RESUME: - perf_evlist__resume(evlist); + evlist__resume(evlist); break; case NONE: default: @@ -1642,7 +1752,7 @@ state_err: return; } -bool perf_evlist__exclude_kernel(struct evlist *evlist) +bool evlist__exclude_kernel(struct evlist *evlist) { struct evsel *evsel; @@ -1659,24 +1769,23 @@ bool perf_evlist__exclude_kernel(struct evlist *evlist) * the group display. Set the artificial group and set the leader's * forced_leader flag to notify the display code. */ -void perf_evlist__force_leader(struct evlist *evlist) +void evlist__force_leader(struct evlist *evlist) { - if (!evlist->nr_groups) { + if (!evlist->core.nr_groups) { struct evsel *leader = evlist__first(evlist); - perf_evlist__set_leader(evlist); + evlist__set_leader(evlist); leader->forced_leader = true; } } -struct evsel *perf_evlist__reset_weak_group(struct evlist *evsel_list, - struct evsel *evsel, - bool close) +struct evsel *evlist__reset_weak_group(struct evlist *evsel_list, struct evsel *evsel, bool close) { struct evsel *c2, *leader; bool is_open = true; - leader = evsel->leader; + leader = evsel__leader(evsel); + pr_debug("Weak group for %s/%d failed\n", leader->name, leader->core.nr_members); @@ -1687,11 +1796,16 @@ struct evsel *perf_evlist__reset_weak_group(struct evlist *evsel_list, evlist__for_each_entry(evsel_list, c2) { if (c2 == evsel) is_open = false; - if (c2->leader == leader) { + if (evsel__has_leader(c2, leader)) { if (is_open && close) perf_evsel__close(&c2->core); - c2->leader = c2; - c2->core.nr_members = 0; + /* + * We want to close all members of the group and reopen + * them. Some events, like Intel topdown, require being + * in a group and so keep these in the group. + */ + evsel__remove_from_group(c2, leader); + /* * Set this for all former members of the group * to indicate they get reopened. @@ -1699,135 +1813,637 @@ struct evsel *perf_evlist__reset_weak_group(struct evlist *evsel_list, c2->reset_group = true; } } + /* Reset the leader count if all entries were removed. */ + if (leader->core.nr_members == 1) + leader->core.nr_members = 0; return leader; } -int perf_evlist__add_sb_event(struct evlist **evlist, - struct perf_event_attr *attr, - perf_evsel__sb_cb_t cb, - void *data) +static int evlist__parse_control_fifo(const char *str, int *ctl_fd, int *ctl_fd_ack, bool *ctl_fd_close) { - struct evsel *evsel; - bool new_evlist = (*evlist) == NULL; + char *s, *p; + int ret = 0, fd; - if (*evlist == NULL) - *evlist = evlist__new(); - if (*evlist == NULL) - return -1; + if (strncmp(str, "fifo:", 5)) + return -EINVAL; + + str += 5; + if (!*str || *str == ',') + return -EINVAL; + + s = strdup(str); + if (!s) + return -ENOMEM; + + p = strchr(s, ','); + if (p) + *p = '\0'; - if (!attr->sample_id_all) { - pr_warning("enabling sample_id_all for all side band events\n"); - attr->sample_id_all = 1; + /* + * O_RDWR avoids POLLHUPs which is necessary to allow the other + * end of a FIFO to be repeatedly opened and closed. + */ + fd = open(s, O_RDWR | O_NONBLOCK | O_CLOEXEC); + if (fd < 0) { + pr_err("Failed to open '%s'\n", s); + ret = -errno; + goto out_free; + } + *ctl_fd = fd; + *ctl_fd_close = true; + + if (p && *++p) { + /* O_RDWR | O_NONBLOCK means the other end need not be open */ + fd = open(p, O_RDWR | O_NONBLOCK | O_CLOEXEC); + if (fd < 0) { + pr_err("Failed to open '%s'\n", p); + ret = -errno; + goto out_free; + } + *ctl_fd_ack = fd; } - evsel = perf_evsel__new_idx(attr, (*evlist)->core.nr_entries); - if (!evsel) - goto out_err; +out_free: + free(s); + return ret; +} + +int evlist__parse_control(const char *str, int *ctl_fd, int *ctl_fd_ack, bool *ctl_fd_close) +{ + char *comma = NULL, *endptr = NULL; + + *ctl_fd_close = false; + + if (strncmp(str, "fd:", 3)) + return evlist__parse_control_fifo(str, ctl_fd, ctl_fd_ack, ctl_fd_close); + + *ctl_fd = strtoul(&str[3], &endptr, 0); + if (endptr == &str[3]) + return -EINVAL; + + comma = strchr(str, ','); + if (comma) { + if (endptr != comma) + return -EINVAL; + + *ctl_fd_ack = strtoul(comma + 1, &endptr, 0); + if (endptr == comma + 1 || *endptr != '\0') + return -EINVAL; + } - evsel->side_band.cb = cb; - evsel->side_band.data = data; - evlist__add(*evlist, evsel); return 0; +} -out_err: - if (new_evlist) { - evlist__delete(*evlist); - *evlist = NULL; +void evlist__close_control(int ctl_fd, int ctl_fd_ack, bool *ctl_fd_close) +{ + if (*ctl_fd_close) { + *ctl_fd_close = false; + close(ctl_fd); + if (ctl_fd_ack >= 0) + close(ctl_fd_ack); } - return -1; } -static void *perf_evlist__poll_thread(void *arg) +int evlist__initialize_ctlfd(struct evlist *evlist, int fd, int ack) { - struct evlist *evlist = arg; - bool draining = false; - int i, done = 0; - /* - * In order to read symbols from other namespaces perf to needs to call - * setns(2). This isn't permitted if the struct_fs has multiple users. - * unshare(2) the fs so that we may continue to setns into namespaces - * that we're observing when, for instance, reading the build-ids at - * the end of a 'perf record' session. - */ - unshare(CLONE_FS); + if (fd == -1) { + pr_debug("Control descriptor is not initialized\n"); + return 0; + } + + evlist->ctl_fd.pos = perf_evlist__add_pollfd(&evlist->core, fd, NULL, POLLIN, + fdarray_flag__nonfilterable | + fdarray_flag__non_perf_event); + if (evlist->ctl_fd.pos < 0) { + evlist->ctl_fd.pos = -1; + pr_err("Failed to add ctl fd entry: %m\n"); + return -1; + } + + evlist->ctl_fd.fd = fd; + evlist->ctl_fd.ack = ack; + + return 0; +} - while (!done) { - bool got_data = false; +bool evlist__ctlfd_initialized(struct evlist *evlist) +{ + return evlist->ctl_fd.pos >= 0; +} - if (evlist->thread.done) - draining = true; +int evlist__finalize_ctlfd(struct evlist *evlist) +{ + struct pollfd *entries = evlist->core.pollfd.entries; - if (!draining) - evlist__poll(evlist, 1000); + if (!evlist__ctlfd_initialized(evlist)) + return 0; + + entries[evlist->ctl_fd.pos].fd = -1; + entries[evlist->ctl_fd.pos].events = 0; + entries[evlist->ctl_fd.pos].revents = 0; - for (i = 0; i < evlist->core.nr_mmaps; i++) { - struct mmap *map = &evlist->mmap[i]; - union perf_event *event; + evlist->ctl_fd.pos = -1; + evlist->ctl_fd.ack = -1; + evlist->ctl_fd.fd = -1; - if (perf_mmap__read_init(&map->core)) + return 0; +} + +static int evlist__ctlfd_recv(struct evlist *evlist, enum evlist_ctl_cmd *cmd, + char *cmd_data, size_t data_size) +{ + int err; + char c; + size_t bytes_read = 0; + + *cmd = EVLIST_CTL_CMD_UNSUPPORTED; + memset(cmd_data, 0, data_size); + data_size--; + + do { + err = read(evlist->ctl_fd.fd, &c, 1); + if (err > 0) { + if (c == '\n' || c == '\0') + break; + cmd_data[bytes_read++] = c; + if (bytes_read == data_size) + break; + continue; + } else if (err == -1) { + if (errno == EINTR) continue; - while ((event = perf_mmap__read_event(&map->core)) != NULL) { - struct evsel *evsel = perf_evlist__event2evsel(evlist, event); + if (errno == EAGAIN || errno == EWOULDBLOCK) + err = 0; + else + pr_err("Failed to read from ctlfd %d: %m\n", evlist->ctl_fd.fd); + } + break; + } while (1); + + pr_debug("Message from ctl_fd: \"%s%s\"\n", cmd_data, + bytes_read == data_size ? "" : c == '\n' ? "\\n" : "\\0"); + + if (bytes_read > 0) { + if (!strncmp(cmd_data, EVLIST_CTL_CMD_ENABLE_TAG, + (sizeof(EVLIST_CTL_CMD_ENABLE_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_ENABLE; + } else if (!strncmp(cmd_data, EVLIST_CTL_CMD_DISABLE_TAG, + (sizeof(EVLIST_CTL_CMD_DISABLE_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_DISABLE; + } else if (!strncmp(cmd_data, EVLIST_CTL_CMD_SNAPSHOT_TAG, + (sizeof(EVLIST_CTL_CMD_SNAPSHOT_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_SNAPSHOT; + pr_debug("is snapshot\n"); + } else if (!strncmp(cmd_data, EVLIST_CTL_CMD_EVLIST_TAG, + (sizeof(EVLIST_CTL_CMD_EVLIST_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_EVLIST; + } else if (!strncmp(cmd_data, EVLIST_CTL_CMD_STOP_TAG, + (sizeof(EVLIST_CTL_CMD_STOP_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_STOP; + } else if (!strncmp(cmd_data, EVLIST_CTL_CMD_PING_TAG, + (sizeof(EVLIST_CTL_CMD_PING_TAG)-1))) { + *cmd = EVLIST_CTL_CMD_PING; + } + } + + return bytes_read ? (int)bytes_read : err; +} + +int evlist__ctlfd_ack(struct evlist *evlist) +{ + int err; + + if (evlist->ctl_fd.ack == -1) + return 0; + + err = write(evlist->ctl_fd.ack, EVLIST_CTL_CMD_ACK_TAG, + sizeof(EVLIST_CTL_CMD_ACK_TAG)); + if (err == -1) + pr_err("failed to write to ctl_ack_fd %d: %m\n", evlist->ctl_fd.ack); + + return err; +} + +static int get_cmd_arg(char *cmd_data, size_t cmd_size, char **arg) +{ + char *data = cmd_data + cmd_size; + + /* no argument */ + if (!*data) + return 0; + + /* there's argument */ + if (*data == ' ') { + *arg = data + 1; + return 1; + } + + /* malformed */ + return -1; +} + +static int evlist__ctlfd_enable(struct evlist *evlist, char *cmd_data, bool enable) +{ + struct evsel *evsel; + char *name; + int err; + + err = get_cmd_arg(cmd_data, + enable ? sizeof(EVLIST_CTL_CMD_ENABLE_TAG) - 1 : + sizeof(EVLIST_CTL_CMD_DISABLE_TAG) - 1, + &name); + if (err < 0) { + pr_info("failed: wrong command\n"); + return -1; + } + + if (err) { + evsel = evlist__find_evsel_by_str(evlist, name); + if (evsel) { + if (enable) + evlist__enable_evsel(evlist, name); + else + evlist__disable_evsel(evlist, name); + pr_info("Event %s %s\n", evsel->name, + enable ? "enabled" : "disabled"); + } else { + pr_info("failed: can't find '%s' event\n", name); + } + } else { + if (enable) { + evlist__enable(evlist); + pr_info(EVLIST_ENABLED_MSG); + } else { + evlist__disable(evlist); + pr_info(EVLIST_DISABLED_MSG); + } + } + + return 0; +} + +static int evlist__ctlfd_list(struct evlist *evlist, char *cmd_data) +{ + struct perf_attr_details details = { .verbose = false, }; + struct evsel *evsel; + char *arg; + int err; + + err = get_cmd_arg(cmd_data, + sizeof(EVLIST_CTL_CMD_EVLIST_TAG) - 1, + &arg); + if (err < 0) { + pr_info("failed: wrong command\n"); + return -1; + } + + if (err) { + if (!strcmp(arg, "-v")) { + details.verbose = true; + } else if (!strcmp(arg, "-g")) { + details.event_group = true; + } else if (!strcmp(arg, "-F")) { + details.freq = true; + } else { + pr_info("failed: wrong command\n"); + return -1; + } + } - if (evsel && evsel->side_band.cb) - evsel->side_band.cb(event, evsel->side_band.data); - else - pr_warning("cannot locate proper evsel for the side band event\n"); + evlist__for_each_entry(evlist, evsel) + evsel__fprintf(evsel, &details, stderr); - perf_mmap__consume(&map->core); - got_data = true; + return 0; +} + +int evlist__ctlfd_process(struct evlist *evlist, enum evlist_ctl_cmd *cmd) +{ + int err = 0; + char cmd_data[EVLIST_CTL_CMD_MAX_LEN]; + int ctlfd_pos = evlist->ctl_fd.pos; + struct pollfd *entries = evlist->core.pollfd.entries; + + if (!evlist__ctlfd_initialized(evlist) || !entries[ctlfd_pos].revents) + return 0; + + if (entries[ctlfd_pos].revents & POLLIN) { + err = evlist__ctlfd_recv(evlist, cmd, cmd_data, + EVLIST_CTL_CMD_MAX_LEN); + if (err > 0) { + switch (*cmd) { + case EVLIST_CTL_CMD_ENABLE: + case EVLIST_CTL_CMD_DISABLE: + err = evlist__ctlfd_enable(evlist, cmd_data, + *cmd == EVLIST_CTL_CMD_ENABLE); + break; + case EVLIST_CTL_CMD_EVLIST: + err = evlist__ctlfd_list(evlist, cmd_data); + break; + case EVLIST_CTL_CMD_SNAPSHOT: + case EVLIST_CTL_CMD_STOP: + case EVLIST_CTL_CMD_PING: + break; + case EVLIST_CTL_CMD_ACK: + case EVLIST_CTL_CMD_UNSUPPORTED: + default: + pr_debug("ctlfd: unsupported %d\n", *cmd); + break; } - perf_mmap__read_done(&map->core); + if (!(*cmd == EVLIST_CTL_CMD_ACK || *cmd == EVLIST_CTL_CMD_UNSUPPORTED || + *cmd == EVLIST_CTL_CMD_SNAPSHOT)) + evlist__ctlfd_ack(evlist); } + } - if (draining && !got_data) - break; + if (entries[ctlfd_pos].revents & (POLLHUP | POLLERR)) + evlist__finalize_ctlfd(evlist); + else + entries[ctlfd_pos].revents = 0; + + return err; +} + +/** + * struct event_enable_time - perf record -D/--delay single time range. + * @start: start of time range to enable events in milliseconds + * @end: end of time range to enable events in milliseconds + * + * N.B. this structure is also accessed as an array of int. + */ +struct event_enable_time { + int start; + int end; +}; + +static int parse_event_enable_time(const char *str, struct event_enable_time *range, bool first) +{ + const char *fmt = first ? "%u - %u %n" : " , %u - %u %n"; + int ret, start, end, n; + + ret = sscanf(str, fmt, &start, &end, &n); + if (ret != 2 || end <= start) + return -EINVAL; + if (range) { + range->start = start; + range->end = end; } - return NULL; + return n; +} + +static ssize_t parse_event_enable_times(const char *str, struct event_enable_time *range) +{ + int incr = !!range; + bool first = true; + ssize_t ret, cnt; + + for (cnt = 0; *str; cnt++) { + ret = parse_event_enable_time(str, range, first); + if (ret < 0) + return ret; + /* Check no overlap */ + if (!first && range && range->start <= range[-1].end) + return -EINVAL; + str += ret; + range += incr; + first = false; + } + return cnt; } -int perf_evlist__start_sb_thread(struct evlist *evlist, - struct target *target) +/** + * struct event_enable_timer - control structure for perf record -D/--delay. + * @evlist: event list + * @times: time ranges that events are enabled (N.B. this is also accessed as an + * array of int) + * @times_cnt: number of time ranges + * @timerfd: timer file descriptor + * @pollfd_pos: position in @evlist array of file descriptors to poll (fdarray) + * @times_step: current position in (int *)@times)[], + * refer event_enable_timer__process() + * + * Note, this structure is only used when there are time ranges, not when there + * is only an initial delay. + */ +struct event_enable_timer { + struct evlist *evlist; + struct event_enable_time *times; + size_t times_cnt; + int timerfd; + int pollfd_pos; + size_t times_step; +}; + +static int str_to_delay(const char *str) +{ + char *endptr; + long d; + + d = strtol(str, &endptr, 10); + if (*endptr || d > INT_MAX || d < -1) + return 0; + return d; +} + +int evlist__parse_event_enable_time(struct evlist *evlist, struct record_opts *opts, + const char *str, int unset) { - struct evsel *counter; + enum fdarray_flags flags = fdarray_flag__nonfilterable | fdarray_flag__non_perf_event; + struct event_enable_timer *eet; + ssize_t times_cnt; + ssize_t ret; + int err; - if (!evlist) + if (unset) return 0; - if (perf_evlist__create_maps(evlist, target)) - goto out_delete_evlist; + opts->initial_delay = str_to_delay(str); + if (opts->initial_delay) + return 0; + + ret = parse_event_enable_times(str, NULL); + if (ret < 0) + return ret; + + times_cnt = ret; + if (times_cnt == 0) + return -EINVAL; + + eet = zalloc(sizeof(*eet)); + if (!eet) + return -ENOMEM; - evlist__for_each_entry(evlist, counter) { - if (evsel__open(counter, evlist->core.cpus, - evlist->core.threads) < 0) - goto out_delete_evlist; + eet->times = calloc(times_cnt, sizeof(*eet->times)); + if (!eet->times) { + err = -ENOMEM; + goto free_eet; } - if (evlist__mmap(evlist, UINT_MAX)) - goto out_delete_evlist; + if (parse_event_enable_times(str, eet->times) != times_cnt) { + err = -EINVAL; + goto free_eet_times; + } + + eet->times_cnt = times_cnt; + + eet->timerfd = timerfd_create(CLOCK_MONOTONIC, TFD_CLOEXEC); + if (eet->timerfd == -1) { + err = -errno; + pr_err("timerfd_create failed: %s\n", strerror(errno)); + goto free_eet_times; + } - evlist__for_each_entry(evlist, counter) { - if (evsel__enable(counter)) - goto out_delete_evlist; + eet->pollfd_pos = perf_evlist__add_pollfd(&evlist->core, eet->timerfd, NULL, POLLIN, flags); + if (eet->pollfd_pos < 0) { + err = eet->pollfd_pos; + goto close_timerfd; } - evlist->thread.done = 0; - if (pthread_create(&evlist->thread.th, NULL, perf_evlist__poll_thread, evlist)) - goto out_delete_evlist; + eet->evlist = evlist; + evlist->eet = eet; + opts->initial_delay = eet->times[0].start; return 0; -out_delete_evlist: - evlist__delete(evlist); - evlist = NULL; - return -1; +close_timerfd: + close(eet->timerfd); +free_eet_times: + free(eet->times); +free_eet: + free(eet); + return err; } -void perf_evlist__stop_sb_thread(struct evlist *evlist) +static int event_enable_timer__set_timer(struct event_enable_timer *eet, int ms) { - if (!evlist) + struct itimerspec its = { + .it_value.tv_sec = ms / MSEC_PER_SEC, + .it_value.tv_nsec = (ms % MSEC_PER_SEC) * NSEC_PER_MSEC, + }; + int err = 0; + + if (timerfd_settime(eet->timerfd, 0, &its, NULL) < 0) { + err = -errno; + pr_err("timerfd_settime failed: %s\n", strerror(errno)); + } + return err; +} + +int event_enable_timer__start(struct event_enable_timer *eet) +{ + int ms; + + if (!eet) + return 0; + + ms = eet->times[0].end - eet->times[0].start; + eet->times_step = 1; + + return event_enable_timer__set_timer(eet, ms); +} + +int event_enable_timer__process(struct event_enable_timer *eet) +{ + struct pollfd *entries; + short revents; + + if (!eet) + return 0; + + entries = eet->evlist->core.pollfd.entries; + revents = entries[eet->pollfd_pos].revents; + entries[eet->pollfd_pos].revents = 0; + + if (revents & POLLIN) { + size_t step = eet->times_step; + size_t pos = step / 2; + + if (step & 1) { + evlist__disable_non_dummy(eet->evlist); + pr_info(EVLIST_DISABLED_MSG); + if (pos >= eet->times_cnt - 1) { + /* Disarm timer */ + event_enable_timer__set_timer(eet, 0); + return 1; /* Stop */ + } + } else { + evlist__enable_non_dummy(eet->evlist); + pr_info(EVLIST_ENABLED_MSG); + } + + step += 1; + pos = step / 2; + + if (pos < eet->times_cnt) { + int *times = (int *)eet->times; /* Accessing 'times' as array of int */ + int ms = times[step] - times[step - 1]; + + eet->times_step = step; + return event_enable_timer__set_timer(eet, ms); + } + } + + return 0; +} + +void event_enable_timer__exit(struct event_enable_timer **ep) +{ + if (!ep || !*ep) return; - evlist->thread.done = 1; - pthread_join(evlist->thread.th, NULL); - evlist__delete(evlist); + free((*ep)->times); + zfree(ep); +} + +struct evsel *evlist__find_evsel(struct evlist *evlist, int idx) +{ + struct evsel *evsel; + + evlist__for_each_entry(evlist, evsel) { + if (evsel->core.idx == idx) + return evsel; + } + return NULL; +} + +int evlist__scnprintf_evsels(struct evlist *evlist, size_t size, char *bf) +{ + struct evsel *evsel; + int printed = 0; + + evlist__for_each_entry(evlist, evsel) { + if (evsel__is_dummy_event(evsel)) + continue; + if (size > (strlen(evsel__name(evsel)) + (printed ? 2 : 1))) { + printed += scnprintf(bf + printed, size - printed, "%s%s", printed ? "," : "", evsel__name(evsel)); + } else { + printed += scnprintf(bf + printed, size - printed, "%s...", printed ? "," : ""); + break; + } + } + + return printed; +} + +void evlist__check_mem_load_aux(struct evlist *evlist) +{ + struct evsel *leader, *evsel, *pos; + + /* + * For some platforms, the 'mem-loads' event is required to use + * together with 'mem-loads-aux' within a group and 'mem-loads-aux' + * must be the group leader. Now we disable this group before reporting + * because 'mem-loads-aux' is just an auxiliary event. It doesn't carry + * any valid memory load information. + */ + evlist__for_each_entry(evlist, evsel) { + leader = evsel__leader(evsel); + if (leader == evsel) + continue; + + if (leader->name && strstr(leader->name, "mem-loads-aux")) { + for_each_group_evsel(pos, leader) { + evsel__set_leader(pos, pos); + pos->core.nr_members = 0; + } + } + } } |