diff options
Diffstat (limited to 'tools/perf/util/session.c')
| -rw-r--r-- | tools/perf/util/session.c | 1806 |
1 files changed, 1282 insertions, 524 deletions
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index c422cd67631..64a186edc7b 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -1,386 +1,531 @@ -#define _FILE_OFFSET_BITS 64 - #include <linux/kernel.h> +#include <traceevent/event-parse.h> #include <byteswap.h> #include <unistd.h> #include <sys/types.h> #include <sys/mman.h> +#include "evlist.h" +#include "evsel.h" #include "session.h" +#include "tool.h" #include "sort.h" #include "util.h" +#include "cpumap.h" +#include "perf_regs.h" +#include "vdso.h" -static int perf_session__open(struct perf_session *self, bool force) +static int perf_session__open(struct perf_session *session) { - struct stat input_stat; - - if (!strcmp(self->filename, "-")) { - self->fd_pipe = true; - self->fd = STDIN_FILENO; + struct perf_data_file *file = session->file; - if (perf_header__read(self, self->fd) < 0) - pr_err("incompatible file format"); - - return 0; - } - - self->fd = open(self->filename, O_RDONLY); - if (self->fd < 0) { - pr_err("failed to open file: %s", self->filename); - if (!strcmp(self->filename, "perf.data")) - pr_err(" (try 'perf record' first)"); - pr_err("\n"); - return -errno; + if (perf_session__read_header(session) < 0) { + pr_err("incompatible file format (rerun with -v to learn more)"); + return -1; } - if (fstat(self->fd, &input_stat) < 0) - goto out_close; + if (perf_data_file__is_pipe(file)) + return 0; - if (!force && input_stat.st_uid && (input_stat.st_uid != geteuid())) { - pr_err("file %s not owned by current user or root\n", - self->filename); - goto out_close; + if (!perf_evlist__valid_sample_type(session->evlist)) { + pr_err("non matching sample_type"); + return -1; } - if (!input_stat.st_size) { - pr_info("zero-sized file (%s), nothing to do!\n", - self->filename); - goto out_close; + if (!perf_evlist__valid_sample_id_all(session->evlist)) { + pr_err("non matching sample_id_all"); + return -1; } - if (perf_header__read(self, self->fd) < 0) { - pr_err("incompatible file format"); - goto out_close; + if (!perf_evlist__valid_read_format(session->evlist)) { + pr_err("non matching read_format"); + return -1; } - self->size = input_stat.st_size; return 0; - -out_close: - close(self->fd); - self->fd = -1; - return -1; } -void perf_session__update_sample_type(struct perf_session *self) +void perf_session__set_id_hdr_size(struct perf_session *session) { - self->sample_type = perf_header__sample_type(&self->header); + u16 id_hdr_size = perf_evlist__id_hdr_size(session->evlist); + + machines__set_id_hdr_size(&session->machines, id_hdr_size); } -int perf_session__create_kernel_maps(struct perf_session *self) +int perf_session__create_kernel_maps(struct perf_session *session) { - int ret = machine__create_kernel_maps(&self->host_machine); + int ret = machine__create_kernel_maps(&session->machines.host); if (ret >= 0) - ret = machines__create_guest_kernel_maps(&self->machines); + ret = machines__create_guest_kernel_maps(&session->machines); return ret; } -struct perf_session *perf_session__new(const char *filename, int mode, bool force, bool repipe) +static void perf_session__destroy_kernel_maps(struct perf_session *session) { - size_t len = filename ? strlen(filename) + 1 : 0; - struct perf_session *self = zalloc(sizeof(*self) + len); + machines__destroy_kernel_maps(&session->machines); +} + +struct perf_session *perf_session__new(struct perf_data_file *file, + bool repipe, struct perf_tool *tool) +{ + struct perf_session *session = zalloc(sizeof(*session)); - if (self == NULL) + if (!session) goto out; - if (perf_header__init(&self->header) < 0) - goto out_free; - - memcpy(self->filename, filename, len); - self->threads = RB_ROOT; - INIT_LIST_HEAD(&self->dead_threads); - self->hists_tree = RB_ROOT; - self->last_match = NULL; - self->mmap_window = 32; - self->cwd = NULL; - self->cwdlen = 0; - self->machines = RB_ROOT; - self->repipe = repipe; - INIT_LIST_HEAD(&self->ordered_samples.samples_head); - machine__init(&self->host_machine, "", HOST_KERNEL_ID); - - if (mode == O_RDONLY) { - if (perf_session__open(self, force) < 0) + session->repipe = repipe; + INIT_LIST_HEAD(&session->ordered_samples.samples); + INIT_LIST_HEAD(&session->ordered_samples.sample_cache); + INIT_LIST_HEAD(&session->ordered_samples.to_free); + machines__init(&session->machines); + + if (file) { + if (perf_data_file__open(file)) goto out_delete; - } else if (mode == O_WRONLY) { + + session->file = file; + + if (perf_data_file__is_read(file)) { + if (perf_session__open(session) < 0) + goto out_close; + + perf_session__set_id_hdr_size(session); + } + } + + if (!file || perf_data_file__is_write(file)) { /* * In O_RDONLY mode this will be performed when reading the - * kernel MMAP event, in event__process_mmap(). + * kernel MMAP event, in perf_event__process_mmap(). */ - if (perf_session__create_kernel_maps(self) < 0) + if (perf_session__create_kernel_maps(session) < 0) goto out_delete; } - perf_session__update_sample_type(self); -out: - return self; -out_free: - free(self); - return NULL; -out_delete: - perf_session__delete(self); + if (tool && tool->ordering_requires_timestamps && + tool->ordered_samples && !perf_evlist__sample_id_all(session->evlist)) { + dump_printf("WARNING: No sample_id_all support, falling back to unordered processing\n"); + tool->ordered_samples = false; + } + + return session; + + out_close: + perf_data_file__close(file); + out_delete: + perf_session__delete(session); + out: return NULL; } -void perf_session__delete(struct perf_session *self) +static void perf_session__delete_dead_threads(struct perf_session *session) { - perf_header__exit(&self->header); - close(self->fd); - free(self->cwd); - free(self); + machine__delete_dead_threads(&session->machines.host); } -void perf_session__remove_thread(struct perf_session *self, struct thread *th) +static void perf_session__delete_threads(struct perf_session *session) { - rb_erase(&th->rb_node, &self->threads); - /* - * We may have references to this thread, for instance in some hist_entry - * instances, so just move them to a separate list. - */ - list_add_tail(&th->node, &self->dead_threads); + machine__delete_threads(&session->machines.host); } -static bool symbol__match_parent_regex(struct symbol *sym) +static void perf_session_env__delete(struct perf_session_env *env) { - if (sym->name && !regexec(&parent_regex, sym->name, 0, NULL, 0)) - return 1; - - return 0; + zfree(&env->hostname); + zfree(&env->os_release); + zfree(&env->version); + zfree(&env->arch); + zfree(&env->cpu_desc); + zfree(&env->cpuid); + + zfree(&env->cmdline); + zfree(&env->sibling_cores); + zfree(&env->sibling_threads); + zfree(&env->numa_nodes); + zfree(&env->pmu_mappings); } -struct map_symbol *perf_session__resolve_callchain(struct perf_session *self, - struct thread *thread, - struct ip_callchain *chain, - struct symbol **parent) +void perf_session__delete(struct perf_session *session) { - u8 cpumode = PERF_RECORD_MISC_USER; - unsigned int i; - struct map_symbol *syms = calloc(chain->nr, sizeof(*syms)); - - if (!syms) - return NULL; + perf_session__destroy_kernel_maps(session); + perf_session__delete_dead_threads(session); + perf_session__delete_threads(session); + perf_session_env__delete(&session->header.env); + machines__exit(&session->machines); + if (session->file) + perf_data_file__close(session->file); + free(session); + vdso__exit(); +} - for (i = 0; i < chain->nr; i++) { - u64 ip = chain->ips[i]; - struct addr_location al; - - if (ip >= PERF_CONTEXT_MAX) { - switch (ip) { - case PERF_CONTEXT_HV: - cpumode = PERF_RECORD_MISC_HYPERVISOR; break; - case PERF_CONTEXT_KERNEL: - cpumode = PERF_RECORD_MISC_KERNEL; break; - case PERF_CONTEXT_USER: - cpumode = PERF_RECORD_MISC_USER; break; - default: - break; - } - continue; - } +static int process_event_synth_tracing_data_stub(struct perf_tool *tool + __maybe_unused, + union perf_event *event + __maybe_unused, + struct perf_session *session + __maybe_unused) +{ + dump_printf(": unhandled!\n"); + return 0; +} - al.filtered = false; - thread__find_addr_location(thread, self, cpumode, - MAP__FUNCTION, thread->pid, ip, &al, NULL); - if (al.sym != NULL) { - if (sort__has_parent && !*parent && - symbol__match_parent_regex(al.sym)) - *parent = al.sym; - if (!symbol_conf.use_callchain) - break; - syms[i].map = al.map; - syms[i].sym = al.sym; - } - } +static int process_event_synth_attr_stub(struct perf_tool *tool __maybe_unused, + union perf_event *event __maybe_unused, + struct perf_evlist **pevlist + __maybe_unused) +{ + dump_printf(": unhandled!\n"); + return 0; +} - return syms; +static int process_event_sample_stub(struct perf_tool *tool __maybe_unused, + union perf_event *event __maybe_unused, + struct perf_sample *sample __maybe_unused, + struct perf_evsel *evsel __maybe_unused, + struct machine *machine __maybe_unused) +{ + dump_printf(": unhandled!\n"); + return 0; } -static int process_event_stub(event_t *event __used, - struct perf_session *session __used) +static int process_event_stub(struct perf_tool *tool __maybe_unused, + union perf_event *event __maybe_unused, + struct perf_sample *sample __maybe_unused, + struct machine *machine __maybe_unused) { dump_printf(": unhandled!\n"); return 0; } -static int process_finished_round_stub(event_t *event __used, - struct perf_session *session __used, - struct perf_event_ops *ops __used) +static int process_finished_round_stub(struct perf_tool *tool __maybe_unused, + union perf_event *event __maybe_unused, + struct perf_session *perf_session + __maybe_unused) { dump_printf(": unhandled!\n"); return 0; } -static int process_finished_round(event_t *event, - struct perf_session *session, - struct perf_event_ops *ops); - -static void perf_event_ops__fill_defaults(struct perf_event_ops *handler) -{ - if (handler->sample == NULL) - handler->sample = process_event_stub; - if (handler->mmap == NULL) - handler->mmap = process_event_stub; - if (handler->comm == NULL) - handler->comm = process_event_stub; - if (handler->fork == NULL) - handler->fork = process_event_stub; - if (handler->exit == NULL) - handler->exit = process_event_stub; - if (handler->lost == NULL) - handler->lost = process_event_stub; - if (handler->read == NULL) - handler->read = process_event_stub; - if (handler->throttle == NULL) - handler->throttle = process_event_stub; - if (handler->unthrottle == NULL) - handler->unthrottle = process_event_stub; - if (handler->attr == NULL) - handler->attr = process_event_stub; - if (handler->event_type == NULL) - handler->event_type = process_event_stub; - if (handler->tracing_data == NULL) - handler->tracing_data = process_event_stub; - if (handler->build_id == NULL) - handler->build_id = process_event_stub; - if (handler->finished_round == NULL) { - if (handler->ordered_samples) - handler->finished_round = process_finished_round; +static int process_finished_round(struct perf_tool *tool, + union perf_event *event, + struct perf_session *session); + +void perf_tool__fill_defaults(struct perf_tool *tool) +{ + if (tool->sample == NULL) + tool->sample = process_event_sample_stub; + if (tool->mmap == NULL) + tool->mmap = process_event_stub; + if (tool->mmap2 == NULL) + tool->mmap2 = process_event_stub; + if (tool->comm == NULL) + tool->comm = process_event_stub; + if (tool->fork == NULL) + tool->fork = process_event_stub; + if (tool->exit == NULL) + tool->exit = process_event_stub; + if (tool->lost == NULL) + tool->lost = perf_event__process_lost; + if (tool->read == NULL) + tool->read = process_event_sample_stub; + if (tool->throttle == NULL) + tool->throttle = process_event_stub; + if (tool->unthrottle == NULL) + tool->unthrottle = process_event_stub; + if (tool->attr == NULL) + tool->attr = process_event_synth_attr_stub; + if (tool->tracing_data == NULL) + tool->tracing_data = process_event_synth_tracing_data_stub; + if (tool->build_id == NULL) + tool->build_id = process_finished_round_stub; + if (tool->finished_round == NULL) { + if (tool->ordered_samples) + tool->finished_round = process_finished_round; else - handler->finished_round = process_finished_round_stub; + tool->finished_round = process_finished_round_stub; } } + +static void swap_sample_id_all(union perf_event *event, void *data) +{ + void *end = (void *) event + event->header.size; + int size = end - data; + + BUG_ON(size % sizeof(u64)); + mem_bswap_64(data, size); +} -void mem_bswap_64(void *src, int byte_size) +static void perf_event__all64_swap(union perf_event *event, + bool sample_id_all __maybe_unused) { - u64 *m = src; + struct perf_event_header *hdr = &event->header; + mem_bswap_64(hdr + 1, event->header.size - sizeof(*hdr)); +} + +static void perf_event__comm_swap(union perf_event *event, bool sample_id_all) +{ + event->comm.pid = bswap_32(event->comm.pid); + event->comm.tid = bswap_32(event->comm.tid); - while (byte_size > 0) { - *m = bswap_64(*m); - byte_size -= sizeof(u64); - ++m; + if (sample_id_all) { + void *data = &event->comm.comm; + + data += PERF_ALIGN(strlen(data) + 1, sizeof(u64)); + swap_sample_id_all(event, data); } } -static void event__all64_swap(event_t *self) +static void perf_event__mmap_swap(union perf_event *event, + bool sample_id_all) { - struct perf_event_header *hdr = &self->header; - mem_bswap_64(hdr + 1, self->header.size - sizeof(*hdr)); + event->mmap.pid = bswap_32(event->mmap.pid); + event->mmap.tid = bswap_32(event->mmap.tid); + event->mmap.start = bswap_64(event->mmap.start); + event->mmap.len = bswap_64(event->mmap.len); + event->mmap.pgoff = bswap_64(event->mmap.pgoff); + + if (sample_id_all) { + void *data = &event->mmap.filename; + + data += PERF_ALIGN(strlen(data) + 1, sizeof(u64)); + swap_sample_id_all(event, data); + } } -static void event__comm_swap(event_t *self) +static void perf_event__mmap2_swap(union perf_event *event, + bool sample_id_all) +{ + event->mmap2.pid = bswap_32(event->mmap2.pid); + event->mmap2.tid = bswap_32(event->mmap2.tid); + event->mmap2.start = bswap_64(event->mmap2.start); + event->mmap2.len = bswap_64(event->mmap2.len); + event->mmap2.pgoff = bswap_64(event->mmap2.pgoff); + event->mmap2.maj = bswap_32(event->mmap2.maj); + event->mmap2.min = bswap_32(event->mmap2.min); + event->mmap2.ino = bswap_64(event->mmap2.ino); + + if (sample_id_all) { + void *data = &event->mmap2.filename; + + data += PERF_ALIGN(strlen(data) + 1, sizeof(u64)); + swap_sample_id_all(event, data); + } +} +static void perf_event__task_swap(union perf_event *event, bool sample_id_all) { - self->comm.pid = bswap_32(self->comm.pid); - self->comm.tid = bswap_32(self->comm.tid); + event->fork.pid = bswap_32(event->fork.pid); + event->fork.tid = bswap_32(event->fork.tid); + event->fork.ppid = bswap_32(event->fork.ppid); + event->fork.ptid = bswap_32(event->fork.ptid); + event->fork.time = bswap_64(event->fork.time); + + if (sample_id_all) + swap_sample_id_all(event, &event->fork + 1); } -static void event__mmap_swap(event_t *self) +static void perf_event__read_swap(union perf_event *event, bool sample_id_all) { - self->mmap.pid = bswap_32(self->mmap.pid); - self->mmap.tid = bswap_32(self->mmap.tid); - self->mmap.start = bswap_64(self->mmap.start); - self->mmap.len = bswap_64(self->mmap.len); - self->mmap.pgoff = bswap_64(self->mmap.pgoff); + event->read.pid = bswap_32(event->read.pid); + event->read.tid = bswap_32(event->read.tid); + event->read.value = bswap_64(event->read.value); + event->read.time_enabled = bswap_64(event->read.time_enabled); + event->read.time_running = bswap_64(event->read.time_running); + event->read.id = bswap_64(event->read.id); + + if (sample_id_all) + swap_sample_id_all(event, &event->read + 1); } -static void event__task_swap(event_t *self) +static void perf_event__throttle_swap(union perf_event *event, + bool sample_id_all) { - self->fork.pid = bswap_32(self->fork.pid); - self->fork.tid = bswap_32(self->fork.tid); - self->fork.ppid = bswap_32(self->fork.ppid); - self->fork.ptid = bswap_32(self->fork.ptid); - self->fork.time = bswap_64(self->fork.time); + event->throttle.time = bswap_64(event->throttle.time); + event->throttle.id = bswap_64(event->throttle.id); + event->throttle.stream_id = bswap_64(event->throttle.stream_id); + + if (sample_id_all) + swap_sample_id_all(event, &event->throttle + 1); } -static void event__read_swap(event_t *self) +static u8 revbyte(u8 b) { - self->read.pid = bswap_32(self->read.pid); - self->read.tid = bswap_32(self->read.tid); - self->read.value = bswap_64(self->read.value); - self->read.time_enabled = bswap_64(self->read.time_enabled); - self->read.time_running = bswap_64(self->read.time_running); - self->read.id = bswap_64(self->read.id); + int rev = (b >> 4) | ((b & 0xf) << 4); + rev = ((rev & 0xcc) >> 2) | ((rev & 0x33) << 2); + rev = ((rev & 0xaa) >> 1) | ((rev & 0x55) << 1); + return (u8) rev; } -static void event__attr_swap(event_t *self) +/* + * XXX this is hack in attempt to carry flags bitfield + * throught endian village. ABI says: + * + * Bit-fields are allocated from right to left (least to most significant) + * on little-endian implementations and from left to right (most to least + * significant) on big-endian implementations. + * + * The above seems to be byte specific, so we need to reverse each + * byte of the bitfield. 'Internet' also says this might be implementation + * specific and we probably need proper fix and carry perf_event_attr + * bitfield flags in separate data file FEAT_ section. Thought this seems + * to work for now. + */ +static void swap_bitfield(u8 *p, unsigned len) { - size_t size; + unsigned i; - self->attr.attr.type = bswap_32(self->attr.attr.type); - self->attr.attr.size = bswap_32(self->attr.attr.size); - self->attr.attr.config = bswap_64(self->attr.attr.config); - self->attr.attr.sample_period = bswap_64(self->attr.attr.sample_period); - self->attr.attr.sample_type = bswap_64(self->attr.attr.sample_type); - self->attr.attr.read_format = bswap_64(self->attr.attr.read_format); - self->attr.attr.wakeup_events = bswap_32(self->attr.attr.wakeup_events); - self->attr.attr.bp_type = bswap_32(self->attr.attr.bp_type); - self->attr.attr.bp_addr = bswap_64(self->attr.attr.bp_addr); - self->attr.attr.bp_len = bswap_64(self->attr.attr.bp_len); + for (i = 0; i < len; i++) { + *p = revbyte(*p); + p++; + } +} - size = self->header.size; - size -= (void *)&self->attr.id - (void *)self; - mem_bswap_64(self->attr.id, size); +/* exported for swapping attributes in file header */ +void perf_event__attr_swap(struct perf_event_attr *attr) +{ + attr->type = bswap_32(attr->type); + attr->size = bswap_32(attr->size); + attr->config = bswap_64(attr->config); + attr->sample_period = bswap_64(attr->sample_period); + attr->sample_type = bswap_64(attr->sample_type); + attr->read_format = bswap_64(attr->read_format); + attr->wakeup_events = bswap_32(attr->wakeup_events); + attr->bp_type = bswap_32(attr->bp_type); + attr->bp_addr = bswap_64(attr->bp_addr); + attr->bp_len = bswap_64(attr->bp_len); + attr->branch_sample_type = bswap_64(attr->branch_sample_type); + attr->sample_regs_user = bswap_64(attr->sample_regs_user); + attr->sample_stack_user = bswap_32(attr->sample_stack_user); + + swap_bitfield((u8 *) (&attr->read_format + 1), sizeof(u64)); } -static void event__event_type_swap(event_t *self) +static void perf_event__hdr_attr_swap(union perf_event *event, + bool sample_id_all __maybe_unused) { - self->event_type.event_type.event_id = - bswap_64(self->event_type.event_type.event_id); + size_t size; + + perf_event__attr_swap(&event->attr.attr); + + size = event->header.size; + size -= (void *)&event->attr.id - (void *)event; + mem_bswap_64(event->attr.id, size); } -static void event__tracing_data_swap(event_t *self) +static void perf_event__event_type_swap(union perf_event *event, + bool sample_id_all __maybe_unused) { - self->tracing_data.size = bswap_32(self->tracing_data.size); + event->event_type.event_type.event_id = + bswap_64(event->event_type.event_type.event_id); } -typedef void (*event__swap_op)(event_t *self); +static void perf_event__tracing_data_swap(union perf_event *event, + bool sample_id_all __maybe_unused) +{ + event->tracing_data.size = bswap_32(event->tracing_data.size); +} -static event__swap_op event__swap_ops[] = { - [PERF_RECORD_MMAP] = event__mmap_swap, - [PERF_RECORD_COMM] = event__comm_swap, - [PERF_RECORD_FORK] = event__task_swap, - [PERF_RECORD_EXIT] = event__task_swap, - [PERF_RECORD_LOST] = event__all64_swap, - [PERF_RECORD_READ] = event__read_swap, - [PERF_RECORD_SAMPLE] = event__all64_swap, - [PERF_RECORD_HEADER_ATTR] = event__attr_swap, - [PERF_RECORD_HEADER_EVENT_TYPE] = event__event_type_swap, - [PERF_RECORD_HEADER_TRACING_DATA] = event__tracing_data_swap, - [PERF_RECORD_HEADER_BUILD_ID] = NULL, - [PERF_RECORD_HEADER_MAX] = NULL, +typedef void (*perf_event__swap_op)(union perf_event *event, + bool sample_id_all); + +static perf_event__swap_op perf_event__swap_ops[] = { + [PERF_RECORD_MMAP] = perf_event__mmap_swap, + [PERF_RECORD_MMAP2] = perf_event__mmap2_swap, + [PERF_RECORD_COMM] = perf_event__comm_swap, + [PERF_RECORD_FORK] = perf_event__task_swap, + [PERF_RECORD_EXIT] = perf_event__task_swap, + [PERF_RECORD_LOST] = perf_event__all64_swap, + [PERF_RECORD_READ] = perf_event__read_swap, + [PERF_RECORD_THROTTLE] = perf_event__throttle_swap, + [PERF_RECORD_UNTHROTTLE] = perf_event__throttle_swap, + [PERF_RECORD_SAMPLE] = perf_event__all64_swap, + [PERF_RECORD_HEADER_ATTR] = perf_event__hdr_attr_swap, + [PERF_RECORD_HEADER_EVENT_TYPE] = perf_event__event_type_swap, + [PERF_RECORD_HEADER_TRACING_DATA] = perf_event__tracing_data_swap, + [PERF_RECORD_HEADER_BUILD_ID] = NULL, + [PERF_RECORD_HEADER_MAX] = NULL, }; struct sample_queue { u64 timestamp; - struct sample_event *event; + u64 file_offset; + union perf_event *event; struct list_head list; }; -static void flush_sample_queue(struct perf_session *s, - struct perf_event_ops *ops) +static void perf_session_free_sample_buffers(struct perf_session *session) +{ + struct ordered_samples *os = &session->ordered_samples; + + while (!list_empty(&os->to_free)) { + struct sample_queue *sq; + + sq = list_entry(os->to_free.next, struct sample_queue, list); + list_del(&sq->list); + free(sq); + } +} + +static int perf_session_deliver_event(struct perf_session *session, + union perf_event *event, + struct perf_sample *sample, + struct perf_tool *tool, + u64 file_offset); + +static int flush_sample_queue(struct perf_session *s, + struct perf_tool *tool) { - struct list_head *head = &s->ordered_samples.samples_head; - u64 limit = s->ordered_samples.next_flush; + struct ordered_samples *os = &s->ordered_samples; + struct list_head *head = &os->samples; struct sample_queue *tmp, *iter; + struct perf_sample sample; + u64 limit = os->next_flush; + u64 last_ts = os->last_sample ? os->last_sample->timestamp : 0ULL; + bool show_progress = limit == ULLONG_MAX; + struct ui_progress prog; + int ret; + + if (!tool->ordered_samples || !limit) + return 0; - if (!ops->ordered_samples || !limit) - return; + if (show_progress) + ui_progress__init(&prog, os->nr_samples, "Processing time ordered events..."); list_for_each_entry_safe(iter, tmp, head, list) { + if (session_done()) + return 0; + if (iter->timestamp > limit) - return; + break; + + ret = perf_evlist__parse_sample(s->evlist, iter->event, &sample); + if (ret) + pr_err("Can't parse sample, err = %d\n", ret); + else { + ret = perf_session_deliver_event(s, iter->event, &sample, tool, + iter->file_offset); + if (ret) + return ret; + } - if (iter == s->ordered_samples.last_inserted) - s->ordered_samples.last_inserted = NULL; + os->last_flush = iter->timestamp; + list_del(&iter->list); + list_add(&iter->list, &os->sample_cache); - ops->sample((event_t *)iter->event, s); + if (show_progress) + ui_progress__update(&prog, 1); + } - s->ordered_samples.last_flush = iter->timestamp; - list_del(&iter->list); - free(iter->event); - free(iter); + if (list_empty(head)) { + os->last_sample = NULL; + } else if (last_ts <= limit) { + os->last_sample = + list_entry(head->prev, struct sample_queue, list); } + + os->nr_samples = 0; + + return 0; } /* @@ -422,200 +567,530 @@ static void flush_sample_queue(struct perf_session *s, * Flush every events below timestamp 7 * etc... */ -static int process_finished_round(event_t *event __used, - struct perf_session *session, - struct perf_event_ops *ops) +static int process_finished_round(struct perf_tool *tool, + union perf_event *event __maybe_unused, + struct perf_session *session) { - flush_sample_queue(session, ops); - session->ordered_samples.next_flush = session->ordered_samples.max_timestamp; + int ret = flush_sample_queue(session, tool); + if (!ret) + session->ordered_samples.next_flush = session->ordered_samples.max_timestamp; - return 0; + return ret; } -static void __queue_sample_end(struct sample_queue *new, struct list_head *head) +/* The queue is ordered by time */ +static void __queue_event(struct sample_queue *new, struct perf_session *s) { - struct sample_queue *iter; + struct ordered_samples *os = &s->ordered_samples; + struct sample_queue *sample = os->last_sample; + u64 timestamp = new->timestamp; + struct list_head *p; - list_for_each_entry_reverse(iter, head, list) { - if (iter->timestamp < new->timestamp) { - list_add(&new->list, &iter->list); - return; - } + ++os->nr_samples; + os->last_sample = new; + + if (!sample) { + list_add(&new->list, &os->samples); + os->max_timestamp = timestamp; + return; } - list_add(&new->list, head); + /* + * last_sample might point to some random place in the list as it's + * the last queued event. We expect that the new event is close to + * this. + */ + if (sample->timestamp <= timestamp) { + while (sample->timestamp <= timestamp) { + p = sample->list.next; + if (p == &os->samples) { + list_add_tail(&new->list, &os->samples); + os->max_timestamp = timestamp; + return; + } + sample = list_entry(p, struct sample_queue, list); + } + list_add_tail(&new->list, &sample->list); + } else { + while (sample->timestamp > timestamp) { + p = sample->list.prev; + if (p == &os->samples) { + list_add(&new->list, &os->samples); + return; + } + sample = list_entry(p, struct sample_queue, list); + } + list_add(&new->list, &sample->list); + } } -static void __queue_sample_before(struct sample_queue *new, - struct sample_queue *iter, - struct list_head *head) +#define MAX_SAMPLE_BUFFER (64 * 1024 / sizeof(struct sample_queue)) + +int perf_session_queue_event(struct perf_session *s, union perf_event *event, + struct perf_sample *sample, u64 file_offset) { - list_for_each_entry_continue_reverse(iter, head, list) { - if (iter->timestamp < new->timestamp) { - list_add(&new->list, &iter->list); - return; - } + struct ordered_samples *os = &s->ordered_samples; + struct list_head *sc = &os->sample_cache; + u64 timestamp = sample->time; + struct sample_queue *new; + + if (!timestamp || timestamp == ~0ULL) + return -ETIME; + + if (timestamp < s->ordered_samples.last_flush) { + printf("Warning: Timestamp below last timeslice flush\n"); + return -EINVAL; } - list_add(&new->list, head); + if (!list_empty(sc)) { + new = list_entry(sc->next, struct sample_queue, list); + list_del(&new->list); + } else if (os->sample_buffer) { + new = os->sample_buffer + os->sample_buffer_idx; + if (++os->sample_buffer_idx == MAX_SAMPLE_BUFFER) + os->sample_buffer = NULL; + } else { + os->sample_buffer = malloc(MAX_SAMPLE_BUFFER * sizeof(*new)); + if (!os->sample_buffer) + return -ENOMEM; + list_add(&os->sample_buffer->list, &os->to_free); + os->sample_buffer_idx = 2; + new = os->sample_buffer + 1; + } + + new->timestamp = timestamp; + new->file_offset = file_offset; + new->event = event; + + __queue_event(new, s); + + return 0; } -static void __queue_sample_after(struct sample_queue *new, - struct sample_queue *iter, - struct list_head *head) +static void callchain__printf(struct perf_sample *sample) { - list_for_each_entry_continue(iter, head, list) { - if (iter->timestamp > new->timestamp) { - list_add_tail(&new->list, &iter->list); - return; - } + unsigned int i; + + printf("... chain: nr:%" PRIu64 "\n", sample->callchain->nr); + + for (i = 0; i < sample->callchain->nr; i++) + printf("..... %2d: %016" PRIx64 "\n", + i, sample->callchain->ips[i]); +} + +static void branch_stack__printf(struct perf_sample *sample) +{ + uint64_t i; + + printf("... branch stack: nr:%" PRIu64 "\n", sample->branch_stack->nr); + + for (i = 0; i < sample->branch_stack->nr; i++) + printf("..... %2"PRIu64": %016" PRIx64 " -> %016" PRIx64 "\n", + i, sample->branch_stack->entries[i].from, + sample->branch_stack->entries[i].to); +} + +static void regs_dump__printf(u64 mask, u64 *regs) +{ + unsigned rid, i = 0; + + for_each_set_bit(rid, (unsigned long *) &mask, sizeof(mask) * 8) { + u64 val = regs[i++]; + + printf(".... %-5s 0x%" PRIx64 "\n", + perf_reg_name(rid), val); } - list_add_tail(&new->list, head); } -/* The queue is ordered by time */ -static void __queue_sample_event(struct sample_queue *new, - struct perf_session *s) +static void regs_user__printf(struct perf_sample *sample) { - struct sample_queue *last_inserted = s->ordered_samples.last_inserted; - struct list_head *head = &s->ordered_samples.samples_head; + struct regs_dump *user_regs = &sample->user_regs; + if (user_regs->regs) { + u64 mask = user_regs->mask; + printf("... user regs: mask 0x%" PRIx64 "\n", mask); + regs_dump__printf(mask, user_regs->regs); + } +} - if (!last_inserted) { - __queue_sample_end(new, head); +static void stack_user__printf(struct stack_dump *dump) +{ + printf("... ustack: size %" PRIu64 ", offset 0x%x\n", + dump->size, dump->offset); +} + +static void perf_session__print_tstamp(struct perf_session *session, + union perf_event *event, + struct perf_sample *sample) +{ + u64 sample_type = __perf_evlist__combined_sample_type(session->evlist); + + if (event->header.type != PERF_RECORD_SAMPLE && + !perf_evlist__sample_id_all(session->evlist)) { + fputs("-1 -1 ", stdout); return; } - /* - * Most of the time the current event has a timestamp - * very close to the last event inserted, unless we just switched - * to another event buffer. Having a sorting based on a list and - * on the last inserted event that is close to the current one is - * probably more efficient than an rbtree based sorting. - */ - if (last_inserted->timestamp >= new->timestamp) - __queue_sample_before(new, last_inserted, head); - else - __queue_sample_after(new, last_inserted, head); + if ((sample_type & PERF_SAMPLE_CPU)) + printf("%u ", sample->cpu); + + if (sample_type & PERF_SAMPLE_TIME) + printf("%" PRIu64 " ", sample->time); } -static int queue_sample_event(event_t *event, struct sample_data *data, - struct perf_session *s) +static void sample_read__printf(struct perf_sample *sample, u64 read_format) { - u64 timestamp = data->time; - struct sample_queue *new; + printf("... sample_read:\n"); + if (read_format & PERF_FORMAT_TOTAL_TIME_ENABLED) + printf("...... time enabled %016" PRIx64 "\n", + sample->read.time_enabled); - if (timestamp < s->ordered_samples.last_flush) { - printf("Warning: Timestamp below last timeslice flush\n"); - return -EINVAL; - } + if (read_format & PERF_FORMAT_TOTAL_TIME_RUNNING) + printf("...... time running %016" PRIx64 "\n", + sample->read.time_running); - new = malloc(sizeof(*new)); - if (!new) - return -ENOMEM; + if (read_format & PERF_FORMAT_GROUP) { + u64 i; - new->timestamp = timestamp; + printf(".... group nr %" PRIu64 "\n", sample->read.group.nr); - new->event = malloc(event->header.size); - if (!new->event) { - free(new); - return -ENOMEM; - } + for (i = 0; i < sample->read.group.nr; i++) { + struct sample_read_value *value; + + value = &sample->read.group.values[i]; + printf("..... id %016" PRIx64 + ", value %016" PRIx64 "\n", + value->id, value->value); + } + } else + printf("..... id %016" PRIx64 ", value %016" PRIx64 "\n", + sample->read.one.id, sample->read.one.value); +} - memcpy(new->event, event, event->header.size); +static void dump_event(struct perf_session *session, union perf_event *event, + u64 file_offset, struct perf_sample *sample) +{ + if (!dump_trace) + return; - __queue_sample_event(new, s); - s->ordered_samples.last_inserted = new; + printf("\n%#" PRIx64 " [%#x]: event: %d\n", + file_offset, event->header.size, event->header.type); - if (new->timestamp > s->ordered_samples.max_timestamp) - s->ordered_samples.max_timestamp = new->timestamp; + trace_event(event); - return 0; + if (sample) + perf_session__print_tstamp(session, event, sample); + + printf("%#" PRIx64 " [%#x]: PERF_RECORD_%s", file_offset, + event->header.size, perf_event__name(event->header.type)); } -static int perf_session__process_sample(event_t *event, struct perf_session *s, - struct perf_event_ops *ops) +static void dump_sample(struct perf_evsel *evsel, union perf_event *event, + struct perf_sample *sample) { - struct sample_data data; + u64 sample_type; - if (!ops->ordered_samples) - return ops->sample(event, s); + if (!dump_trace) + return; - bzero(&data, sizeof(struct sample_data)); - event__parse_sample(event, s->sample_type, &data); + printf("(IP, 0x%x): %d/%d: %#" PRIx64 " period: %" PRIu64 " addr: %#" PRIx64 "\n", + event->header.misc, sample->pid, sample->tid, sample->ip, + sample->period, sample->addr); - queue_sample_event(event, &data, s); + sample_type = evsel->attr.sample_type; - return 0; + if (sample_type & PERF_SAMPLE_CALLCHAIN) + callchain__printf(sample); + + if (sample_type & PERF_SAMPLE_BRANCH_STACK) + branch_stack__printf(sample); + + if (sample_type & PERF_SAMPLE_REGS_USER) + regs_user__printf(sample); + + if (sample_type & PERF_SAMPLE_STACK_USER) + stack_user__printf(&sample->user_stack); + + if (sample_type & PERF_SAMPLE_WEIGHT) + printf("... weight: %" PRIu64 "\n", sample->weight); + + if (sample_type & PERF_SAMPLE_DATA_SRC) + printf(" . data_src: 0x%"PRIx64"\n", sample->data_src); + + if (sample_type & PERF_SAMPLE_TRANSACTION) + printf("... transaction: %" PRIx64 "\n", sample->transaction); + + if (sample_type & PERF_SAMPLE_READ) + sample_read__printf(sample, evsel->attr.read_format); } -static int perf_session__process_event(struct perf_session *self, - event_t *event, - struct perf_event_ops *ops, - u64 offset, u64 head) +static struct machine * + perf_session__find_machine_for_cpumode(struct perf_session *session, + union perf_event *event, + struct perf_sample *sample) { - trace_event(event); + const u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; + struct machine *machine; + + if (perf_guest && + ((cpumode == PERF_RECORD_MISC_GUEST_KERNEL) || + (cpumode == PERF_RECORD_MISC_GUEST_USER))) { + u32 pid; + + if (event->header.type == PERF_RECORD_MMAP + || event->header.type == PERF_RECORD_MMAP2) + pid = event->mmap.pid; + else + pid = sample->pid; + + machine = perf_session__find_machine(session, pid); + if (!machine) + machine = perf_session__findnew_machine(session, + DEFAULT_GUEST_KERNEL_ID); + return machine; + } + + return &session->machines.host; +} + +static int deliver_sample_value(struct perf_session *session, + struct perf_tool *tool, + union perf_event *event, + struct perf_sample *sample, + struct sample_read_value *v, + struct machine *machine) +{ + struct perf_sample_id *sid; + + sid = perf_evlist__id2sid(session->evlist, v->id); + if (sid) { + sample->id = v->id; + sample->period = v->value - sid->period; + sid->period = v->value; + } - if (event->header.type < PERF_RECORD_HEADER_MAX) { - dump_printf("%#Lx [%#x]: PERF_RECORD_%s", - offset + head, event->header.size, - event__name[event->header.type]); - hists__inc_nr_events(&self->hists, event->header.type); + if (!sid || sid->evsel == NULL) { + ++session->stats.nr_unknown_id; + return 0; } - if (self->header.needs_swap && event__swap_ops[event->header.type]) - event__swap_ops[event->header.type](event); + return tool->sample(tool, event, sample, sid->evsel, machine); +} + +static int deliver_sample_group(struct perf_session *session, + struct perf_tool *tool, + union perf_event *event, + struct perf_sample *sample, + struct machine *machine) +{ + int ret = -EINVAL; + u64 i; + + for (i = 0; i < sample->read.group.nr; i++) { + ret = deliver_sample_value(session, tool, event, sample, + &sample->read.group.values[i], + machine); + if (ret) + break; + } + + return ret; +} + +static int +perf_session__deliver_sample(struct perf_session *session, + struct perf_tool *tool, + union perf_event *event, + struct perf_sample *sample, + struct perf_evsel *evsel, + struct machine *machine) +{ + /* We know evsel != NULL. */ + u64 sample_type = evsel->attr.sample_type; + u64 read_format = evsel->attr.read_format; + + /* Standard sample delievery. */ + if (!(sample_type & PERF_SAMPLE_READ)) + return tool->sample(tool, event, sample, evsel, machine); + + /* For PERF_SAMPLE_READ we have either single or group mode. */ + if (read_format & PERF_FORMAT_GROUP) + return deliver_sample_group(session, tool, event, sample, + machine); + else + return deliver_sample_value(session, tool, event, sample, + &sample->read.one, machine); +} + +static int perf_session_deliver_event(struct perf_session *session, + union perf_event *event, + struct perf_sample *sample, + struct perf_tool *tool, + u64 file_offset) +{ + struct perf_evsel *evsel; + struct machine *machine; + + dump_event(session, event, file_offset, sample); + + evsel = perf_evlist__id2evsel(session->evlist, sample->id); + if (evsel != NULL && event->header.type != PERF_RECORD_SAMPLE) { + /* + * XXX We're leaving PERF_RECORD_SAMPLE unnacounted here + * because the tools right now may apply filters, discarding + * some of the samples. For consistency, in the future we + * should have something like nr_filtered_samples and remove + * the sample->period from total_sample_period, etc, KISS for + * now tho. + * + * Also testing against NULL allows us to handle files without + * attr.sample_id_all and/or without PERF_SAMPLE_ID. In the + * future probably it'll be a good idea to restrict event + * processing via perf_session to files with both set. + */ + hists__inc_nr_events(&evsel->hists, event->header.type); + } + + machine = perf_session__find_machine_for_cpumode(session, event, + sample); switch (event->header.type) { case PERF_RECORD_SAMPLE: - return perf_session__process_sample(event, self, ops); + dump_sample(evsel, event, sample); + if (evsel == NULL) { + ++session->stats.nr_unknown_id; + return 0; + } + if (machine == NULL) { + ++session->stats.nr_unprocessable_samples; + return 0; + } + return perf_session__deliver_sample(session, tool, event, + sample, evsel, machine); case PERF_RECORD_MMAP: - return ops->mmap(event, self); + return tool->mmap(tool, event, sample, machine); + case PERF_RECORD_MMAP2: + return tool->mmap2(tool, event, sample, machine); case PERF_RECORD_COMM: - return ops->comm(event, self); + return tool->comm(tool, event, sample, machine); case PERF_RECORD_FORK: - return ops->fork(event, self); + return tool->fork(tool, event, sample, machine); case PERF_RECORD_EXIT: - return ops->exit(event, self); + return tool->exit(tool, event, sample, machine); case PERF_RECORD_LOST: - return ops->lost(event, self); + if (tool->lost == perf_event__process_lost) + session->stats.total_lost += event->lost.lost; + return tool->lost(tool, event, sample, machine); case PERF_RECORD_READ: - return ops->read(event, self); + return tool->read(tool, event, sample, evsel, machine); case PERF_RECORD_THROTTLE: - return ops->throttle(event, self); + return tool->throttle(tool, event, sample, machine); case PERF_RECORD_UNTHROTTLE: - return ops->unthrottle(event, self); + return tool->unthrottle(tool, event, sample, machine); + default: + ++session->stats.nr_unknown_events; + return -1; + } +} + +static int perf_session__process_user_event(struct perf_session *session, union perf_event *event, + struct perf_tool *tool, u64 file_offset) +{ + int fd = perf_data_file__fd(session->file); + int err; + + dump_event(session, event, file_offset, NULL); + + /* These events are processed right away */ + switch (event->header.type) { case PERF_RECORD_HEADER_ATTR: - return ops->attr(event, self); + err = tool->attr(tool, event, &session->evlist); + if (err == 0) + perf_session__set_id_hdr_size(session); + return err; case PERF_RECORD_HEADER_EVENT_TYPE: - return ops->event_type(event, self); + /* + * Depreceated, but we need to handle it for sake + * of old data files create in pipe mode. + */ + return 0; case PERF_RECORD_HEADER_TRACING_DATA: /* setup for reading amidst mmap */ - lseek(self->fd, offset + head, SEEK_SET); - return ops->tracing_data(event, self); + lseek(fd, file_offset, SEEK_SET); + return tool->tracing_data(tool, event, session); case PERF_RECORD_HEADER_BUILD_ID: - return ops->build_id(event, self); + return tool->build_id(tool, event, session); case PERF_RECORD_FINISHED_ROUND: - return ops->finished_round(event, self, ops); + return tool->finished_round(tool, event, session); default: - ++self->hists.stats.nr_unknown_events; - return -1; + return -EINVAL; } } -void perf_event_header__bswap(struct perf_event_header *self) +static void event_swap(union perf_event *event, bool sample_id_all) { - self->type = bswap_32(self->type); - self->misc = bswap_16(self->misc); - self->size = bswap_16(self->size); + perf_event__swap_op swap; + + swap = perf_event__swap_ops[event->header.type]; + if (swap) + swap(event, sample_id_all); } -static struct thread *perf_session__register_idle_thread(struct perf_session *self) +static int perf_session__process_event(struct perf_session *session, + union perf_event *event, + struct perf_tool *tool, + u64 file_offset) { - struct thread *thread = perf_session__findnew(self, 0); + struct perf_sample sample; + int ret; + + if (session->header.needs_swap) + event_swap(event, perf_evlist__sample_id_all(session->evlist)); - if (thread == NULL || thread__set_comm(thread, "swapper")) { + if (event->header.type >= PERF_RECORD_HEADER_MAX) + return -EINVAL; + + events_stats__inc(&session->stats, event->header.type); + + if (event->header.type >= PERF_RECORD_USER_TYPE_START) + return perf_session__process_user_event(session, event, tool, file_offset); + + /* + * For all kernel events we get the sample data + */ + ret = perf_evlist__parse_sample(session->evlist, event, &sample); + if (ret) + return ret; + + if (tool->ordered_samples) { + ret = perf_session_queue_event(session, event, &sample, + file_offset); + if (ret != -ETIME) + return ret; + } + + return perf_session_deliver_event(session, event, &sample, tool, + file_offset); +} + +void perf_event_header__bswap(struct perf_event_header *hdr) +{ + hdr->type = bswap_32(hdr->type); + hdr->misc = bswap_16(hdr->misc); + hdr->size = bswap_16(hdr->size); +} + +struct thread *perf_session__findnew(struct perf_session *session, pid_t pid) +{ + return machine__findnew_thread(&session->machines.host, 0, pid); +} + +static struct thread *perf_session__register_idle_thread(struct perf_session *session) +{ + struct thread *thread = perf_session__findnew(session, 0); + + if (thread == NULL || thread__set_comm(thread, "swapper", 0)) { pr_err("problem inserting idle task.\n"); thread = NULL; } @@ -623,41 +1098,71 @@ static struct thread *perf_session__register_idle_thread(struct perf_session *se return thread; } -int do_read(int fd, void *buf, size_t size) +static void perf_session__warn_about_errors(const struct perf_session *session, + const struct perf_tool *tool) { - void *buf_start = buf; - - while (size) { - int ret = read(fd, buf, size); + if (tool->lost == perf_event__process_lost && + session->stats.nr_events[PERF_RECORD_LOST] != 0) { + ui__warning("Processed %d events and lost %d chunks!\n\n" + "Check IO/CPU overload!\n\n", + session->stats.nr_events[0], + session->stats.nr_events[PERF_RECORD_LOST]); + } - if (ret <= 0) - return ret; + if (session->stats.nr_unknown_events != 0) { + ui__warning("Found %u unknown events!\n\n" + "Is this an older tool processing a perf.data " + "file generated by a more recent tool?\n\n" + "If that is not the case, consider " + "reporting to linux-kernel@vger.kernel.org.\n\n", + session->stats.nr_unknown_events); + } - size -= ret; - buf += ret; + if (session->stats.nr_unknown_id != 0) { + ui__warning("%u samples with id not present in the header\n", + session->stats.nr_unknown_id); } - return buf - buf_start; + if (session->stats.nr_invalid_chains != 0) { + ui__warning("Found invalid callchains!\n\n" + "%u out of %u events were discarded for this reason.\n\n" + "Consider reporting to linux-kernel@vger.kernel.org.\n\n", + session->stats.nr_invalid_chains, + session->stats.nr_events[PERF_RECORD_SAMPLE]); + } + + if (session->stats.nr_unprocessable_samples != 0) { + ui__warning("%u unprocessable samples recorded.\n" + "Do you have a KVM guest running and not using 'perf kvm'?\n", + session->stats.nr_unprocessable_samples); + } } -#define session_done() (*(volatile int *)(&session_done)) volatile int session_done; -static int __perf_session__process_pipe_events(struct perf_session *self, - struct perf_event_ops *ops) +static int __perf_session__process_pipe_events(struct perf_session *session, + struct perf_tool *tool) { - event_t event; - uint32_t size; + int fd = perf_data_file__fd(session->file); + union perf_event *event; + uint32_t size, cur_size = 0; + void *buf = NULL; int skip = 0; u64 head; - int err; + ssize_t err; void *p; - perf_event_ops__fill_defaults(ops); + perf_tool__fill_defaults(tool); head = 0; + cur_size = sizeof(union perf_event); + + buf = malloc(cur_size); + if (!buf) + return -errno; more: - err = do_read(self->fd, &event, sizeof(struct perf_event_header)); + event = buf; + err = readn(fd, event, sizeof(struct perf_event_header)); if (err <= 0) { if (err == 0) goto done; @@ -666,19 +1171,30 @@ more: goto out_err; } - if (self->header.needs_swap) - perf_event_header__bswap(&event.header); + if (session->header.needs_swap) + perf_event_header__bswap(&event->header); - size = event.header.size; - if (size == 0) - size = 8; + size = event->header.size; + if (size < sizeof(struct perf_event_header)) { + pr_err("bad event header size\n"); + goto out_err; + } - p = &event; + if (size > cur_size) { + void *new = realloc(buf, size); + if (!new) { + pr_err("failed to allocate memory to read event\n"); + goto out_err; + } + buf = new; + cur_size = size; + event = buf; + } + p = event; p += sizeof(struct perf_event_header); if (size - sizeof(struct perf_event_header)) { - err = do_read(self->fd, p, - size - sizeof(struct perf_event_header)); + err = readn(fd, p, size - sizeof(struct perf_event_header)); if (err <= 0) { if (err == 0) { pr_err("unexpected end of event stream\n"); @@ -690,187 +1206,202 @@ more: } } - if (size == 0 || - (skip = perf_session__process_event(self, &event, ops, - 0, head)) < 0) { - dump_printf("%#Lx [%#x]: skipping unknown header type: %d\n", - head, event.header.size, event.header.type); - /* - * assume we lost track of the stream, check alignment, and - * increment a single u64 in the hope to catch on again 'soon'. - */ - if (unlikely(head & 7)) - head &= ~7ULL; - - size = 8; + if ((skip = perf_session__process_event(session, event, tool, head)) < 0) { + pr_err("%#" PRIx64 " [%#x]: failed to process type: %d\n", + head, event->header.size, event->header.type); + err = -EINVAL; + goto out_err; } head += size; - dump_printf("\n%#Lx [%#x]: event: %d\n", - head, event.header.size, event.header.type); - if (skip > 0) head += skip; if (!session_done()) goto more; done: - err = 0; + /* do the final flush for ordered samples */ + session->ordered_samples.next_flush = ULLONG_MAX; + err = flush_sample_queue(session, tool); out_err: + free(buf); + perf_session__warn_about_errors(session, tool); + perf_session_free_sample_buffers(session); return err; } -int __perf_session__process_events(struct perf_session *self, +static union perf_event * +fetch_mmaped_event(struct perf_session *session, + u64 head, size_t mmap_size, char *buf) +{ + union perf_event *event; + + /* + * Ensure we have enough space remaining to read + * the size of the event in the headers. + */ + if (head + sizeof(event->header) > mmap_size) + return NULL; + + event = (union perf_event *)(buf + head); + + if (session->header.needs_swap) + perf_event_header__bswap(&event->header); + + if (head + event->header.size > mmap_size) { + /* We're not fetching the event so swap back again */ + if (session->header.needs_swap) + perf_event_header__bswap(&event->header); + return NULL; + } + + return event; +} + +/* + * On 64bit we can mmap the data file in one go. No need for tiny mmap + * slices. On 32bit we use 32MB. + */ +#if BITS_PER_LONG == 64 +#define MMAP_SIZE ULLONG_MAX +#define NUM_MMAPS 1 +#else +#define MMAP_SIZE (32 * 1024 * 1024ULL) +#define NUM_MMAPS 128 +#endif + +int __perf_session__process_events(struct perf_session *session, u64 data_offset, u64 data_size, - u64 file_size, struct perf_event_ops *ops) + u64 file_size, struct perf_tool *tool) { - int err, mmap_prot, mmap_flags; - u64 head, shift; - u64 offset = 0; - size_t page_size; - event_t *event; + int fd = perf_data_file__fd(session->file); + u64 head, page_offset, file_offset, file_pos; + int err, mmap_prot, mmap_flags, map_idx = 0; + size_t mmap_size; + char *buf, *mmaps[NUM_MMAPS]; + union perf_event *event; uint32_t size; - char *buf; - struct ui_progress *progress = ui_progress__new("Processing events...", - self->size); - if (progress == NULL) - return -1; + struct ui_progress prog; - perf_event_ops__fill_defaults(ops); + perf_tool__fill_defaults(tool); - page_size = sysconf(_SC_PAGESIZE); + page_offset = page_size * (data_offset / page_size); + file_offset = page_offset; + head = data_offset - page_offset; - head = data_offset; - shift = page_size * (head / page_size); - offset += shift; - head -= shift; + if (data_size && (data_offset + data_size < file_size)) + file_size = data_offset + data_size; + + ui_progress__init(&prog, file_size, "Processing events..."); + + mmap_size = MMAP_SIZE; + if (mmap_size > file_size) + mmap_size = file_size; + + memset(mmaps, 0, sizeof(mmaps)); mmap_prot = PROT_READ; mmap_flags = MAP_SHARED; - if (self->header.needs_swap) { + if (session->header.needs_swap) { mmap_prot |= PROT_WRITE; mmap_flags = MAP_PRIVATE; } remap: - buf = mmap(NULL, page_size * self->mmap_window, mmap_prot, - mmap_flags, self->fd, offset); + buf = mmap(NULL, mmap_size, mmap_prot, mmap_flags, fd, + file_offset); if (buf == MAP_FAILED) { pr_err("failed to mmap file\n"); err = -errno; goto out_err; } + mmaps[map_idx] = buf; + map_idx = (map_idx + 1) & (ARRAY_SIZE(mmaps) - 1); + file_pos = file_offset + head; more: - event = (event_t *)(buf + head); - ui_progress__update(progress, offset); - - if (self->header.needs_swap) - perf_event_header__bswap(&event->header); - size = event->header.size; - if (size == 0) - size = 8; - - if (head + event->header.size >= page_size * self->mmap_window) { - int munmap_ret; - - shift = page_size * (head / page_size); - - munmap_ret = munmap(buf, page_size * self->mmap_window); - assert(munmap_ret == 0); + event = fetch_mmaped_event(session, head, mmap_size, buf); + if (!event) { + if (mmaps[map_idx]) { + munmap(mmaps[map_idx], mmap_size); + mmaps[map_idx] = NULL; + } - offset += shift; - head -= shift; + page_offset = page_size * (head / page_size); + file_offset += page_offset; + head -= page_offset; goto remap; } size = event->header.size; - dump_printf("\n%#Lx [%#x]: event: %d\n", - offset + head, event->header.size, event->header.type); - - if (size == 0 || - perf_session__process_event(self, event, ops, offset, head) < 0) { - dump_printf("%#Lx [%#x]: skipping unknown header type: %d\n", - offset + head, event->header.size, - event->header.type); - /* - * assume we lost track of the stream, check alignment, and - * increment a single u64 in the hope to catch on again 'soon'. - */ - if (unlikely(head & 7)) - head &= ~7ULL; - - size = 8; + if (size < sizeof(struct perf_event_header) || + perf_session__process_event(session, event, tool, file_pos) < 0) { + pr_err("%#" PRIx64 " [%#x]: failed to process type: %d\n", + file_offset + head, event->header.size, + event->header.type); + err = -EINVAL; + goto out_err; } head += size; + file_pos += size; + + ui_progress__update(&prog, size); - if (offset + head >= data_offset + data_size) - goto done; + if (session_done()) + goto out; - if (offset + head < file_size) + if (file_pos < file_size) goto more; -done: - err = 0; + +out: /* do the final flush for ordered samples */ - self->ordered_samples.next_flush = ULLONG_MAX; - flush_sample_queue(self, ops); + session->ordered_samples.next_flush = ULLONG_MAX; + err = flush_sample_queue(session, tool); out_err: - ui_progress__delete(progress); + ui_progress__finish(); + perf_session__warn_about_errors(session, tool); + perf_session_free_sample_buffers(session); return err; } -int perf_session__process_events(struct perf_session *self, - struct perf_event_ops *ops) +int perf_session__process_events(struct perf_session *session, + struct perf_tool *tool) { + u64 size = perf_data_file__size(session->file); int err; - if (perf_session__register_idle_thread(self) == NULL) + if (perf_session__register_idle_thread(session) == NULL) return -ENOMEM; - if (!symbol_conf.full_paths) { - char bf[PATH_MAX]; - - if (getcwd(bf, sizeof(bf)) == NULL) { - err = -errno; -out_getcwd_err: - pr_err("failed to get the current directory\n"); - goto out_err; - } - self->cwd = strdup(bf); - if (self->cwd == NULL) { - err = -ENOMEM; - goto out_getcwd_err; - } - self->cwdlen = strlen(self->cwd); - } - - if (!self->fd_pipe) - err = __perf_session__process_events(self, - self->header.data_offset, - self->header.data_size, - self->size, ops); + if (!perf_data_file__is_pipe(session->file)) + err = __perf_session__process_events(session, + session->header.data_offset, + session->header.data_size, + size, tool); else - err = __perf_session__process_pipe_events(self, ops); -out_err: + err = __perf_session__process_pipe_events(session, tool); + return err; } -bool perf_session__has_traces(struct perf_session *self, const char *msg) +bool perf_session__has_traces(struct perf_session *session, const char *msg) { - if (!(self->sample_type & PERF_SAMPLE_RAW)) { - pr_err("No trace sample to read. Did you call 'perf %s'?\n", msg); - return false; + struct perf_evsel *evsel; + + evlist__for_each(session->evlist, evsel) { + if (evsel->attr.type == PERF_TYPE_TRACEPOINT) + return true; } - return true; + pr_err("No trace sample to read. Did you call 'perf %s'?\n", msg); + return false; } -int perf_session__set_kallsyms_ref_reloc_sym(struct map **maps, - const char *symbol_name, - u64 addr) +int maps__set_kallsyms_ref_reloc_sym(struct map **maps, + const char *symbol_name, u64 addr) { char *bracket; enum map_type i; @@ -900,16 +1431,243 @@ int perf_session__set_kallsyms_ref_reloc_sym(struct map **maps, return 0; } -size_t perf_session__fprintf_dsos(struct perf_session *self, FILE *fp) +size_t perf_session__fprintf_dsos(struct perf_session *session, FILE *fp) +{ + return machines__fprintf_dsos(&session->machines, fp); +} + +size_t perf_session__fprintf_dsos_buildid(struct perf_session *session, FILE *fp, + bool (skip)(struct dso *dso, int parm), int parm) +{ + return machines__fprintf_dsos_buildid(&session->machines, fp, skip, parm); +} + +size_t perf_session__fprintf_nr_events(struct perf_session *session, FILE *fp) +{ + struct perf_evsel *pos; + size_t ret = fprintf(fp, "Aggregated stats:\n"); + + ret += events_stats__fprintf(&session->stats, fp); + + evlist__for_each(session->evlist, pos) { + ret += fprintf(fp, "%s stats:\n", perf_evsel__name(pos)); + ret += events_stats__fprintf(&pos->hists.stats, fp); + } + + return ret; +} + +size_t perf_session__fprintf(struct perf_session *session, FILE *fp) +{ + /* + * FIXME: Here we have to actually print all the machines in this + * session, not just the host... + */ + return machine__fprintf(&session->machines.host, fp); +} + +struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session, + unsigned int type) +{ + struct perf_evsel *pos; + + evlist__for_each(session->evlist, pos) { + if (pos->attr.type == type) + return pos; + } + return NULL; +} + +void perf_evsel__print_ip(struct perf_evsel *evsel, struct perf_sample *sample, + struct addr_location *al, + unsigned int print_opts, unsigned int stack_depth) +{ + struct callchain_cursor_node *node; + int print_ip = print_opts & PRINT_IP_OPT_IP; + int print_sym = print_opts & PRINT_IP_OPT_SYM; + int print_dso = print_opts & PRINT_IP_OPT_DSO; + int print_symoffset = print_opts & PRINT_IP_OPT_SYMOFFSET; + int print_oneline = print_opts & PRINT_IP_OPT_ONELINE; + int print_srcline = print_opts & PRINT_IP_OPT_SRCLINE; + char s = print_oneline ? ' ' : '\t'; + + if (symbol_conf.use_callchain && sample->callchain) { + struct addr_location node_al; + + if (machine__resolve_callchain(al->machine, evsel, al->thread, + sample, NULL, NULL, + PERF_MAX_STACK_DEPTH) != 0) { + if (verbose) + error("Failed to resolve callchain. Skipping\n"); + return; + } + callchain_cursor_commit(&callchain_cursor); + + if (print_symoffset) + node_al = *al; + + while (stack_depth) { + u64 addr = 0; + + node = callchain_cursor_current(&callchain_cursor); + if (!node) + break; + + if (node->sym && node->sym->ignore) + goto next; + + if (print_ip) + printf("%c%16" PRIx64, s, node->ip); + + if (node->map) + addr = node->map->map_ip(node->map, node->ip); + + if (print_sym) { + printf(" "); + if (print_symoffset) { + node_al.addr = addr; + node_al.map = node->map; + symbol__fprintf_symname_offs(node->sym, &node_al, stdout); + } else + symbol__fprintf_symname(node->sym, stdout); + } + + if (print_dso) { + printf(" ("); + map__fprintf_dsoname(node->map, stdout); + printf(")"); + } + + if (print_srcline) + map__fprintf_srcline(node->map, addr, "\n ", + stdout); + + if (!print_oneline) + printf("\n"); + + stack_depth--; +next: + callchain_cursor_advance(&callchain_cursor); + } + + } else { + if (al->sym && al->sym->ignore) + return; + + if (print_ip) + printf("%16" PRIx64, sample->ip); + + if (print_sym) { + printf(" "); + if (print_symoffset) + symbol__fprintf_symname_offs(al->sym, al, + stdout); + else + symbol__fprintf_symname(al->sym, stdout); + } + + if (print_dso) { + printf(" ("); + map__fprintf_dsoname(al->map, stdout); + printf(")"); + } + + if (print_srcline) + map__fprintf_srcline(al->map, al->addr, "\n ", stdout); + } +} + +int perf_session__cpu_bitmap(struct perf_session *session, + const char *cpu_list, unsigned long *cpu_bitmap) +{ + int i, err = -1; + struct cpu_map *map; + + for (i = 0; i < PERF_TYPE_MAX; ++i) { + struct perf_evsel *evsel; + + evsel = perf_session__find_first_evtype(session, i); + if (!evsel) + continue; + + if (!(evsel->attr.sample_type & PERF_SAMPLE_CPU)) { + pr_err("File does not contain CPU events. " + "Remove -c option to proceed.\n"); + return -1; + } + } + + map = cpu_map__new(cpu_list); + if (map == NULL) { + pr_err("Invalid cpu_list\n"); + return -1; + } + + for (i = 0; i < map->nr; i++) { + int cpu = map->map[i]; + + if (cpu >= MAX_NR_CPUS) { + pr_err("Requested CPU %d too large. " + "Consider raising MAX_NR_CPUS\n", cpu); + goto out_delete_map; + } + + set_bit(cpu, cpu_bitmap); + } + + err = 0; + +out_delete_map: + cpu_map__delete(map); + return err; +} + +void perf_session__fprintf_info(struct perf_session *session, FILE *fp, + bool full) { - return __dsos__fprintf(&self->host_machine.kernel_dsos, fp) + - __dsos__fprintf(&self->host_machine.user_dsos, fp) + - machines__fprintf_dsos(&self->machines, fp); + struct stat st; + int fd, ret; + + if (session == NULL || fp == NULL) + return; + + fd = perf_data_file__fd(session->file); + + ret = fstat(fd, &st); + if (ret == -1) + return; + + fprintf(fp, "# ========\n"); + fprintf(fp, "# captured on: %s", ctime(&st.st_ctime)); + perf_header__fprintf_info(session, fp, full); + fprintf(fp, "# ========\n#\n"); } -size_t perf_session__fprintf_dsos_buildid(struct perf_session *self, FILE *fp, - bool with_hits) + +int __perf_session__set_tracepoints_handlers(struct perf_session *session, + const struct perf_evsel_str_handler *assocs, + size_t nr_assocs) { - size_t ret = machine__fprintf_dsos_buildid(&self->host_machine, fp, with_hits); - return ret + machines__fprintf_dsos_buildid(&self->machines, fp, with_hits); + struct perf_evsel *evsel; + size_t i; + int err; + + for (i = 0; i < nr_assocs; i++) { + /* + * Adding a handler for an event not in the session, + * just ignore it. + */ + evsel = perf_evlist__find_tracepoint_by_name(session->evlist, assocs[i].name); + if (evsel == NULL) + continue; + + err = -EEXIST; + if (evsel->handler != NULL) + goto out; + evsel->handler = assocs[i].handler; + } + + err = 0; +out: + return err; } |