diff options
Diffstat (limited to 'tools/perf/builtin-record.c')
| -rw-r--r-- | tools/perf/builtin-record.c | 1339 |
1 files changed, 685 insertions, 654 deletions
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 93bd2ff001f..378b85b731a 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -5,8 +5,6 @@ * (or a CPU, or a PID) into the perf.data output file - for * later analysis via perf report. */ -#define _FILE_OFFSET_BITS 64 - #include "builtin.h" #include "perf.h" @@ -18,150 +16,68 @@ #include "util/header.h" #include "util/event.h" +#include "util/evlist.h" +#include "util/evsel.h" #include "util/debug.h" #include "util/session.h" +#include "util/tool.h" #include "util/symbol.h" #include "util/cpumap.h" +#include "util/thread_map.h" +#include "util/data.h" #include <unistd.h> #include <sched.h> #include <sys/mman.h> -enum write_mode_t { - WRITE_FORCE, - WRITE_APPEND -}; -static int *fd[MAX_NR_CPUS][MAX_COUNTERS]; - -static u64 user_interval = ULLONG_MAX; -static u64 default_interval = 0; - -static int nr_cpus = 0; -static unsigned int page_size; -static unsigned int mmap_pages = 128; -static unsigned int user_freq = UINT_MAX; -static int freq = 1000; -static int output; -static int pipe_output = 0; -static const char *output_name = "perf.data"; -static int group = 0; -static int realtime_prio = 0; -static bool raw_samples = false; -static bool system_wide = false; -static pid_t target_pid = -1; -static pid_t target_tid = -1; -static pid_t *all_tids = NULL; -static int thread_num = 0; -static pid_t child_pid = -1; -static bool no_inherit = false; -static enum write_mode_t write_mode = WRITE_FORCE; -static bool call_graph = false; -static bool inherit_stat = false; -static bool no_samples = false; -static bool sample_address = false; -static bool no_buildid = false; - -static long samples = 0; -static u64 bytes_written = 0; - -static struct pollfd *event_array; - -static int nr_poll = 0; -static int nr_cpu = 0; - -static int file_new = 1; -static off_t post_processing_offset; - -static struct perf_session *session; -static const char *cpu_list; - -struct mmap_data { - int counter; - void *base; - unsigned int mask; - unsigned int prev; +struct record { + struct perf_tool tool; + struct record_opts opts; + u64 bytes_written; + struct perf_data_file file; + struct perf_evlist *evlist; + struct perf_session *session; + const char *progname; + int realtime_prio; + bool no_buildid; + bool no_buildid_cache; + long samples; }; -static struct mmap_data mmap_array[MAX_NR_CPUS]; - -static unsigned long mmap_read_head(struct mmap_data *md) -{ - struct perf_event_mmap_page *pc = md->base; - long head; - - head = pc->data_head; - rmb(); - - return head; -} - -static void mmap_write_tail(struct mmap_data *md, unsigned long tail) +static int record__write(struct record *rec, void *bf, size_t size) { - struct perf_event_mmap_page *pc = md->base; - - /* - * ensure all reads are done before we write the tail out. - */ - /* mb(); */ - pc->data_tail = tail; -} - -static void advance_output(size_t size) -{ - bytes_written += size; -} - -static void write_output(void *buf, size_t size) -{ - while (size) { - int ret = write(output, buf, size); - - if (ret < 0) - die("failed to write"); - - size -= ret; - buf += ret; - - bytes_written += ret; + if (perf_data_file__write(rec->session->file, bf, size) < 0) { + pr_err("failed to write perf data, error: %m\n"); + return -1; } + + rec->bytes_written += size; + return 0; } -static int process_synthesized_event(event_t *event, - struct perf_session *self __used) +static int process_synthesized_event(struct perf_tool *tool, + union perf_event *event, + struct perf_sample *sample __maybe_unused, + struct machine *machine __maybe_unused) { - write_output(event, event->header.size); - return 0; + struct record *rec = container_of(tool, struct record, tool); + return record__write(rec, event, event->header.size); } -static void mmap_read(struct mmap_data *md) +static int record__mmap_read(struct record *rec, struct perf_mmap *md) { - unsigned int head = mmap_read_head(md); + unsigned int head = perf_mmap__read_head(md); unsigned int old = md->prev; unsigned char *data = md->base + page_size; unsigned long size; void *buf; - int diff; + int rc = 0; - /* - * If we're further behind than half the buffer, there's a chance - * the writer will bite our tail and mess up the samples under us. - * - * If we somehow ended up ahead of the head, we got messed up. - * - * In either case, truncate and restart at head. - */ - diff = head - old; - if (diff < 0) { - fprintf(stderr, "WARNING: failed to keep up with mmap data\n"); - /* - * head points to a known good entry, start there. - */ - old = head; - } + if (old == head) + return 0; - if (old != head) - samples++; + rec->samples++; size = head - old; @@ -170,288 +86,126 @@ static void mmap_read(struct mmap_data *md) size = md->mask + 1 - (old & md->mask); old += size; - write_output(buf, size); + if (record__write(rec, buf, size) < 0) { + rc = -1; + goto out; + } } buf = &data[old & md->mask]; size = head - old; old += size; - write_output(buf, size); + if (record__write(rec, buf, size) < 0) { + rc = -1; + goto out; + } md->prev = old; - mmap_write_tail(md, old); + perf_mmap__write_tail(md, old); + +out: + return rc; } static volatile int done = 0; static volatile int signr = -1; +static volatile int child_finished = 0; static void sig_handler(int sig) { + if (sig == SIGCHLD) + child_finished = 1; + else + signr = sig; + done = 1; - signr = sig; } -static void sig_atexit(void) +static void record__sig_exit(void) { - if (child_pid > 0) - kill(child_pid, SIGTERM); - if (signr == -1) return; signal(signr, SIG_DFL); - kill(getpid(), signr); + raise(signr); } -static int group_fd; - -static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int nr) +static int record__open(struct record *rec) { - struct perf_header_attr *h_attr; + char msg[512]; + struct perf_evsel *pos; + struct perf_evlist *evlist = rec->evlist; + struct perf_session *session = rec->session; + struct record_opts *opts = &rec->opts; + int rc = 0; - if (nr < session->header.attrs) { - h_attr = session->header.attr[nr]; - } else { - h_attr = perf_header_attr__new(a); - if (h_attr != NULL) - if (perf_header__add_attr(&session->header, h_attr) < 0) { - perf_header_attr__delete(h_attr); - h_attr = NULL; - } - } - - return h_attr; -} - -static void create_counter(int counter, int cpu) -{ - char *filter = filters[counter]; - struct perf_event_attr *attr = attrs + counter; - struct perf_header_attr *h_attr; - int track = !counter; /* only the first counter needs these */ - int thread_index; - int ret; - struct { - u64 count; - u64 time_enabled; - u64 time_running; - u64 id; - } read_data; - - attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | - PERF_FORMAT_TOTAL_TIME_RUNNING | - PERF_FORMAT_ID; - - attr->sample_type |= PERF_SAMPLE_IP | PERF_SAMPLE_TID; - - if (nr_counters > 1) - attr->sample_type |= PERF_SAMPLE_ID; - - /* - * We default some events to a 1 default interval. But keep - * it a weak assumption overridable by the user. - */ - if (!attr->sample_period || (user_freq != UINT_MAX && - user_interval != ULLONG_MAX)) { - if (freq) { - attr->sample_type |= PERF_SAMPLE_PERIOD; - attr->freq = 1; - attr->sample_freq = freq; - } else { - attr->sample_period = default_interval; - } - } - - if (no_samples) - attr->sample_freq = 0; + perf_evlist__config(evlist, opts); - if (inherit_stat) - attr->inherit_stat = 1; - - if (sample_address) { - attr->sample_type |= PERF_SAMPLE_ADDR; - attr->mmap_data = track; - } - - if (call_graph) - attr->sample_type |= PERF_SAMPLE_CALLCHAIN; - - if (system_wide) - attr->sample_type |= PERF_SAMPLE_CPU; - - if (raw_samples) { - attr->sample_type |= PERF_SAMPLE_TIME; - attr->sample_type |= PERF_SAMPLE_RAW; - attr->sample_type |= PERF_SAMPLE_CPU; - } - - attr->mmap = track; - attr->comm = track; - attr->inherit = !no_inherit; - if (target_pid == -1 && target_tid == -1 && !system_wide) { - attr->disabled = 1; - attr->enable_on_exec = 1; - } - - for (thread_index = 0; thread_index < thread_num; thread_index++) { + evlist__for_each(evlist, pos) { try_again: - fd[nr_cpu][counter][thread_index] = sys_perf_event_open(attr, - all_tids[thread_index], cpu, group_fd, 0); - - if (fd[nr_cpu][counter][thread_index] < 0) { - int err = errno; - - if (err == EPERM || err == EACCES) - die("Permission error - are you root?\n" - "\t Consider tweaking" - " /proc/sys/kernel/perf_event_paranoid.\n"); - else if (err == ENODEV && cpu_list) { - die("No such device - did you specify" - " an out-of-range profile CPU?\n"); - } - - /* - * If it's cycles then fall back to hrtimer - * based cpu-clock-tick sw counter, which - * is always available even if no PMU support: - */ - if (attr->type == PERF_TYPE_HARDWARE - && attr->config == PERF_COUNT_HW_CPU_CYCLES) { - + if (perf_evsel__open(pos, evlist->cpus, evlist->threads) < 0) { + if (perf_evsel__fallback(pos, errno, msg, sizeof(msg))) { if (verbose) - warning(" ... trying to fall back to cpu-clock-ticks\n"); - attr->type = PERF_TYPE_SOFTWARE; - attr->config = PERF_COUNT_SW_CPU_CLOCK; + ui__warning("%s\n", msg); goto try_again; } - printf("\n"); - error("perfcounter syscall returned with %d (%s)\n", - fd[nr_cpu][counter][thread_index], strerror(err)); - -#if defined(__i386__) || defined(__x86_64__) - if (attr->type == PERF_TYPE_HARDWARE && err == EOPNOTSUPP) - die("No hardware sampling interrupt available." - " No APIC? If so then you can boot the kernel" - " with the \"lapic\" boot parameter to" - " force-enable it.\n"); -#endif - - die("No CONFIG_PERF_EVENTS=y kernel support configured?\n"); - exit(-1); - } - - h_attr = get_header_attr(attr, counter); - if (h_attr == NULL) - die("nomem\n"); - - if (!file_new) { - if (memcmp(&h_attr->attr, attr, sizeof(*attr))) { - fprintf(stderr, "incompatible append\n"); - exit(-1); - } - } - - if (read(fd[nr_cpu][counter][thread_index], &read_data, sizeof(read_data)) == -1) { - perror("Unable to read perf file descriptor"); - exit(-1); - } - if (perf_header_attr__add_id(h_attr, read_data.id) < 0) { - pr_warning("Not enough memory to add id\n"); - exit(-1); + rc = -errno; + perf_evsel__open_strerror(pos, &opts->target, + errno, msg, sizeof(msg)); + ui__error("%s\n", msg); + goto out; } + } - assert(fd[nr_cpu][counter][thread_index] >= 0); - fcntl(fd[nr_cpu][counter][thread_index], F_SETFL, O_NONBLOCK); + if (perf_evlist__apply_filters(evlist)) { + error("failed to set filter with %d (%s)\n", errno, + strerror(errno)); + rc = -1; + goto out; + } - /* - * First counter acts as the group leader: - */ - if (group && group_fd == -1) - group_fd = fd[nr_cpu][counter][thread_index]; - - if (counter || thread_index) { - ret = ioctl(fd[nr_cpu][counter][thread_index], - PERF_EVENT_IOC_SET_OUTPUT, - fd[nr_cpu][0][0]); - if (ret) { - error("failed to set output: %d (%s)\n", errno, - strerror(errno)); - exit(-1); - } + if (perf_evlist__mmap(evlist, opts->mmap_pages, false) < 0) { + if (errno == EPERM) { + pr_err("Permission error mapping pages.\n" + "Consider increasing " + "/proc/sys/kernel/perf_event_mlock_kb,\n" + "or try again with a smaller value of -m/--mmap_pages.\n" + "(current value: %u)\n", opts->mmap_pages); + rc = -errno; } else { - mmap_array[nr_cpu].counter = counter; - mmap_array[nr_cpu].prev = 0; - mmap_array[nr_cpu].mask = mmap_pages*page_size - 1; - mmap_array[nr_cpu].base = mmap(NULL, (mmap_pages+1)*page_size, - PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter][thread_index], 0); - if (mmap_array[nr_cpu].base == MAP_FAILED) { - error("failed to mmap with %d (%s)\n", errno, strerror(errno)); - exit(-1); - } - - event_array[nr_poll].fd = fd[nr_cpu][counter][thread_index]; - event_array[nr_poll].events = POLLIN; - nr_poll++; - } - - if (filter != NULL) { - ret = ioctl(fd[nr_cpu][counter][thread_index], - PERF_EVENT_IOC_SET_FILTER, filter); - if (ret) { - error("failed to set filter with %d (%s)\n", errno, - strerror(errno)); - exit(-1); - } + pr_err("failed to mmap with %d (%s)\n", errno, strerror(errno)); + rc = -errno; } + goto out; } -} - -static void open_counters(int cpu) -{ - int counter; - - group_fd = -1; - for (counter = 0; counter < nr_counters; counter++) - create_counter(counter, cpu); - nr_cpu++; + session->evlist = evlist; + perf_session__set_id_hdr_size(session); +out: + return rc; } -static int process_buildids(void) +static int process_buildids(struct record *rec) { - u64 size = lseek(output, 0, SEEK_CUR); + struct perf_data_file *file = &rec->file; + struct perf_session *session = rec->session; + u64 start = session->header.data_offset; + u64 size = lseek(file->fd, 0, SEEK_CUR); if (size == 0) return 0; - session->fd = output; - return __perf_session__process_events(session, post_processing_offset, - size - post_processing_offset, + return __perf_session__process_events(session, start, + size - start, size, &build_id__mark_dso_hit_ops); } -static void atexit_header(void) -{ - if (!pipe_output) { - session->header.data_size += bytes_written; - - process_buildids(); - perf_header__write(&session->header, output, true); - perf_session__delete(session); - symbol__exit(); - } -} - -static void event__synthesize_guest_os(struct machine *machine, void *data) +static void perf_event__synthesize_guest_os(struct machine *machine, void *data) { int err; - struct perf_session *psession = data; - - if (machine__is_host(machine)) - return; - + struct perf_tool *tool = data; /* *As for guest kernel when processing subcommand record&report, *we arrange module mmap prior to guest kernel mmap and trigger @@ -460,8 +214,8 @@ static void event__synthesize_guest_os(struct machine *machine, void *data) *method is used to avoid symbol missing when the first addr is *in module instead of in guest kernel. */ - err = event__synthesize_modules(process_synthesized_event, - psession, machine); + err = perf_event__synthesize_modules(tool, process_synthesized_event, + machine); if (err < 0) pr_err("Couldn't record guest kernel [%d]'s reference" " relocation symbol.\n", machine->pid); @@ -470,11 +224,8 @@ static void event__synthesize_guest_os(struct machine *machine, void *data) * We use _stext for guest kernel because guest kernel's /proc/kallsyms * have no _text sometimes. */ - err = event__synthesize_kernel_mmap(process_synthesized_event, - psession, machine, "_text"); - if (err < 0) - err = event__synthesize_kernel_mmap(process_synthesized_event, - psession, machine, "_stext"); + err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event, + machine); if (err < 0) pr_err("Couldn't record guest kernel [%d]'s reference" " relocation symbol.\n", machine->pid); @@ -485,187 +236,140 @@ static struct perf_event_header finished_round_event = { .type = PERF_RECORD_FINISHED_ROUND, }; -static void mmap_read_all(void) +static int record__mmap_read_all(struct record *rec) { int i; + int rc = 0; - for (i = 0; i < nr_cpu; i++) { - if (mmap_array[i].base) - mmap_read(&mmap_array[i]); + for (i = 0; i < rec->evlist->nr_mmaps; i++) { + if (rec->evlist->mmap[i].base) { + if (record__mmap_read(rec, &rec->evlist->mmap[i]) != 0) { + rc = -1; + goto out; + } + } } - if (perf_header__has_feat(&session->header, HEADER_TRACE_INFO)) - write_output(&finished_round_event, sizeof(finished_round_event)); + if (perf_header__has_feat(&rec->session->header, HEADER_TRACING_DATA)) + rc = record__write(rec, &finished_round_event, sizeof(finished_round_event)); + +out: + return rc; } -static int __cmd_record(int argc, const char **argv) +static void record__init_features(struct record *rec) +{ + struct perf_session *session = rec->session; + int feat; + + for (feat = HEADER_FIRST_FEATURE; feat < HEADER_LAST_FEATURE; feat++) + perf_header__set_feat(&session->header, feat); + + if (rec->no_buildid) + perf_header__clear_feat(&session->header, HEADER_BUILD_ID); + + if (!have_tracepoints(&rec->evlist->entries)) + perf_header__clear_feat(&session->header, HEADER_TRACING_DATA); + + if (!rec->opts.branch_stack) + perf_header__clear_feat(&session->header, HEADER_BRANCH_STACK); +} + +static volatile int workload_exec_errno; + +/* + * perf_evlist__prepare_workload will send a SIGUSR1 + * if the fork fails, since we asked by setting its + * want_signal to true. + */ +static void workload_exec_failed_signal(int signo __maybe_unused, + siginfo_t *info, + void *ucontext __maybe_unused) +{ + workload_exec_errno = info->si_value.sival_int; + done = 1; + child_finished = 1; +} + +static int __cmd_record(struct record *rec, int argc, const char **argv) { - int i, counter; - struct stat st; - int flags; int err; + int status = 0; unsigned long waking = 0; - int child_ready_pipe[2], go_pipe[2]; const bool forks = argc > 0; - char buf; struct machine *machine; + struct perf_tool *tool = &rec->tool; + struct record_opts *opts = &rec->opts; + struct perf_data_file *file = &rec->file; + struct perf_session *session; + bool disabled = false; - page_size = sysconf(_SC_PAGE_SIZE); + rec->progname = argv[0]; - atexit(sig_atexit); + atexit(record__sig_exit); signal(SIGCHLD, sig_handler); signal(SIGINT, sig_handler); + signal(SIGTERM, sig_handler); - if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) { - perror("failed to create pipes"); - exit(-1); - } - - if (!strcmp(output_name, "-")) - pipe_output = 1; - else if (!stat(output_name, &st) && st.st_size) { - if (write_mode == WRITE_FORCE) { - char oldname[PATH_MAX]; - snprintf(oldname, sizeof(oldname), "%s.old", - output_name); - unlink(oldname); - rename(output_name, oldname); - } - } else if (write_mode == WRITE_APPEND) { - write_mode = WRITE_FORCE; - } - - flags = O_CREAT|O_RDWR; - if (write_mode == WRITE_APPEND) - file_new = 0; - else - flags |= O_TRUNC; - - if (pipe_output) - output = STDOUT_FILENO; - else - output = open(output_name, flags, S_IRUSR | S_IWUSR); - if (output < 0) { - perror("failed to create output file"); - exit(-1); - } - - session = perf_session__new(output_name, O_WRONLY, - write_mode == WRITE_FORCE, false); + session = perf_session__new(file, false, NULL); if (session == NULL) { - pr_err("Not enough memory for reading perf file header\n"); + pr_err("Perf session creation failed.\n"); return -1; } - if (!file_new) { - err = perf_header__read(session, output); - if (err < 0) - goto out_delete_session; - } + rec->session = session; - if (have_tracepoints(attrs, nr_counters)) - perf_header__set_feat(&session->header, HEADER_TRACE_INFO); - - /* - * perf_session__delete(session) will be called at atexit_header() - */ - atexit(atexit_header); + record__init_features(rec); if (forks) { - child_pid = fork(); - if (child_pid < 0) { - perror("failed to fork"); - exit(-1); - } - - if (!child_pid) { - if (pipe_output) - dup2(2, 1); - close(child_ready_pipe[0]); - close(go_pipe[1]); - fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); - - /* - * Do a dummy execvp to get the PLT entry resolved, - * so we avoid the resolver overhead on the real - * execvp call. - */ - execvp("", (char **)argv); - - /* - * Tell the parent we're ready to go - */ - close(child_ready_pipe[1]); - - /* - * Wait until the parent tells us to go. - */ - if (read(go_pipe[0], &buf, 1) == -1) - perror("unable to read pipe"); - - execvp(argv[0], (char **)argv); - - perror(argv[0]); - exit(-1); - } - - if (!system_wide && target_tid == -1 && target_pid == -1) - all_tids[0] = child_pid; - - close(child_ready_pipe[1]); - close(go_pipe[0]); - /* - * wait for child to settle - */ - if (read(child_ready_pipe[0], &buf, 1) == -1) { - perror("unable to read pipe"); - exit(-1); + err = perf_evlist__prepare_workload(rec->evlist, &opts->target, + argv, file->is_pipe, + workload_exec_failed_signal); + if (err < 0) { + pr_err("Couldn't run the workload!\n"); + status = err; + goto out_delete_session; } - close(child_ready_pipe[0]); } - nr_cpus = read_cpu_map(cpu_list); - if (nr_cpus < 1) { - perror("failed to collect number of CPUs"); - return -1; + if (record__open(rec) != 0) { + err = -1; + goto out_child; } - if (!system_wide && no_inherit && !cpu_list) { - open_counters(-1); - } else { - for (i = 0; i < nr_cpus; i++) - open_counters(cpumap[i]); - } + if (!rec->evlist->nr_groups) + perf_header__clear_feat(&session->header, HEADER_GROUP_DESC); - if (pipe_output) { - err = perf_header__write_pipe(output); + if (file->is_pipe) { + err = perf_header__write_pipe(file->fd); if (err < 0) - return err; - } else if (file_new) { - err = perf_header__write(&session->header, output, false); + goto out_child; + } else { + err = perf_session__write_header(session, rec->evlist, + file->fd, false); if (err < 0) - return err; + goto out_child; } - post_processing_offset = lseek(output, 0, SEEK_CUR); + if (!rec->no_buildid + && !perf_header__has_feat(&session->header, HEADER_BUILD_ID)) { + pr_err("Couldn't generate buildids. " + "Use --no-buildid to profile anyway.\n"); + err = -1; + goto out_child; + } - if (pipe_output) { - err = event__synthesize_attrs(&session->header, - process_synthesized_event, - session); - if (err < 0) { - pr_err("Couldn't synthesize attrs.\n"); - return err; - } + machine = &session->machines.host; - err = event__synthesize_event_types(process_synthesized_event, - session); + if (file->is_pipe) { + err = perf_event__synthesize_attrs(tool, session, + process_synthesized_event); if (err < 0) { - pr_err("Couldn't synthesize event_types.\n"); - return err; + pr_err("Couldn't synthesize attrs.\n"); + goto out_child; } - if (have_tracepoints(attrs, nr_counters)) { + if (have_tracepoints(&rec->evlist->entries)) { /* * FIXME err <= 0 here actually means that * there were no tracepoints so its not really @@ -674,112 +378,384 @@ static int __cmd_record(int argc, const char **argv) * return this more properly and also * propagate errors that now are calling die() */ - err = event__synthesize_tracing_data(output, attrs, - nr_counters, - process_synthesized_event, - session); + err = perf_event__synthesize_tracing_data(tool, file->fd, rec->evlist, + process_synthesized_event); if (err <= 0) { pr_err("Couldn't record tracing data.\n"); - return err; + goto out_child; } - advance_output(err); + rec->bytes_written += err; } } - machine = perf_session__find_host_machine(session); - if (!machine) { - pr_err("Couldn't find native kernel information.\n"); - return -1; - } + err = perf_event__synthesize_kernel_mmap(tool, process_synthesized_event, + machine); + if (err < 0) + pr_err("Couldn't record kernel reference relocation symbol\n" + "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n" + "Check /proc/kallsyms permission or run as root.\n"); - err = event__synthesize_kernel_mmap(process_synthesized_event, - session, machine, "_text"); + err = perf_event__synthesize_modules(tool, process_synthesized_event, + machine); if (err < 0) - err = event__synthesize_kernel_mmap(process_synthesized_event, - session, machine, "_stext"); - if (err < 0) { - pr_err("Couldn't record kernel reference relocation symbol.\n"); - return err; - } + pr_err("Couldn't record kernel module information.\n" + "Symbol resolution may be skewed if relocation was used (e.g. kexec).\n" + "Check /proc/modules permission or run as root.\n"); - err = event__synthesize_modules(process_synthesized_event, - session, machine); - if (err < 0) { - pr_err("Couldn't record kernel reference relocation symbol.\n"); - return err; + if (perf_guest) { + machines__process_guests(&session->machines, + perf_event__synthesize_guest_os, tool); } - if (perf_guest) - perf_session__process_machines(session, event__synthesize_guest_os); - if (!system_wide) - event__synthesize_thread(target_tid, process_synthesized_event, - session); - else - event__synthesize_threads(process_synthesized_event, session); + err = __machine__synthesize_threads(machine, tool, &opts->target, rec->evlist->threads, + process_synthesized_event, opts->sample_address); + if (err != 0) + goto out_child; - if (realtime_prio) { + if (rec->realtime_prio) { struct sched_param param; - param.sched_priority = realtime_prio; + param.sched_priority = rec->realtime_prio; if (sched_setscheduler(0, SCHED_FIFO, ¶m)) { pr_err("Could not set realtime priority.\n"); - exit(-1); + err = -1; + goto out_child; } } /* + * When perf is starting the traced process, all the events + * (apart from group members) have enable_on_exec=1 set, + * so don't spoil it by prematurely enabling them. + */ + if (!target__none(&opts->target) && !opts->initial_delay) + perf_evlist__enable(rec->evlist); + + /* * Let the child rip */ if (forks) - close(go_pipe[1]); + perf_evlist__start_workload(rec->evlist); + + if (opts->initial_delay) { + usleep(opts->initial_delay * 1000); + perf_evlist__enable(rec->evlist); + } for (;;) { - int hits = samples; - int thread; + int hits = rec->samples; - mmap_read_all(); + if (record__mmap_read_all(rec) < 0) { + err = -1; + goto out_child; + } - if (hits == samples) { + if (hits == rec->samples) { if (done) break; - err = poll(event_array, nr_poll, -1); + err = poll(rec->evlist->pollfd, rec->evlist->nr_fds, -1); + /* + * Propagate error, only if there's any. Ignore positive + * number of returned events and interrupt error. + */ + if (err > 0 || (err < 0 && errno == EINTR)) + err = 0; waking++; } - if (done) { - for (i = 0; i < nr_cpu; i++) { - for (counter = 0; - counter < nr_counters; - counter++) { - for (thread = 0; - thread < thread_num; - thread++) - ioctl(fd[i][counter][thread], - PERF_EVENT_IOC_DISABLE); - } - } + /* + * When perf is starting the traced process, at the end events + * die with the process and we wait for that. Thus no need to + * disable events in this case. + */ + if (done && !disabled && !target__none(&opts->target)) { + perf_evlist__disable(rec->evlist); + disabled = true; } } - if (quiet) - return 0; + if (forks && workload_exec_errno) { + char msg[512]; + const char *emsg = strerror_r(workload_exec_errno, msg, sizeof(msg)); + pr_err("Workload failed: %s\n", emsg); + err = -1; + goto out_child; + } - fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking); + if (!quiet) { + fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking); + + /* + * Approximate RIP event size: 24 bytes. + */ + fprintf(stderr, + "[ perf record: Captured and wrote %.3f MB %s (~%" PRIu64 " samples) ]\n", + (double)rec->bytes_written / 1024.0 / 1024.0, + file->path, + rec->bytes_written / 24); + } + +out_child: + if (forks) { + int exit_status; + + if (!child_finished) + kill(rec->evlist->workload.pid, SIGTERM); + + wait(&exit_status); + + if (err < 0) + status = err; + else if (WIFEXITED(exit_status)) + status = WEXITSTATUS(exit_status); + else if (WIFSIGNALED(exit_status)) + signr = WTERMSIG(exit_status); + } else + status = err; + + if (!err && !file->is_pipe) { + rec->session->header.data_size += rec->bytes_written; + + if (!rec->no_buildid) + process_buildids(rec); + perf_session__write_header(rec->session, rec->evlist, + file->fd, true); + } + +out_delete_session: + perf_session__delete(session); + return status; +} + +#define BRANCH_OPT(n, m) \ + { .name = n, .mode = (m) } + +#define BRANCH_END { .name = NULL } + +struct branch_mode { + const char *name; + int mode; +}; + +static const struct branch_mode branch_modes[] = { + BRANCH_OPT("u", PERF_SAMPLE_BRANCH_USER), + BRANCH_OPT("k", PERF_SAMPLE_BRANCH_KERNEL), + BRANCH_OPT("hv", PERF_SAMPLE_BRANCH_HV), + BRANCH_OPT("any", PERF_SAMPLE_BRANCH_ANY), + BRANCH_OPT("any_call", PERF_SAMPLE_BRANCH_ANY_CALL), + BRANCH_OPT("any_ret", PERF_SAMPLE_BRANCH_ANY_RETURN), + BRANCH_OPT("ind_call", PERF_SAMPLE_BRANCH_IND_CALL), + BRANCH_OPT("abort_tx", PERF_SAMPLE_BRANCH_ABORT_TX), + BRANCH_OPT("in_tx", PERF_SAMPLE_BRANCH_IN_TX), + BRANCH_OPT("no_tx", PERF_SAMPLE_BRANCH_NO_TX), + BRANCH_OPT("cond", PERF_SAMPLE_BRANCH_COND), + BRANCH_END +}; + +static int +parse_branch_stack(const struct option *opt, const char *str, int unset) +{ +#define ONLY_PLM \ + (PERF_SAMPLE_BRANCH_USER |\ + PERF_SAMPLE_BRANCH_KERNEL |\ + PERF_SAMPLE_BRANCH_HV) + + uint64_t *mode = (uint64_t *)opt->value; + const struct branch_mode *br; + char *s, *os = NULL, *p; + int ret = -1; + + if (unset) + return 0; /* - * Approximate RIP event size: 24 bytes. + * cannot set it twice, -b + --branch-filter for instance */ - fprintf(stderr, - "[ perf record: Captured and wrote %.3f MB %s (~%lld samples) ]\n", - (double)bytes_written / 1024.0 / 1024.0, - output_name, - bytes_written / 24); + if (*mode) + return -1; + + /* str may be NULL in case no arg is passed to -b */ + if (str) { + /* because str is read-only */ + s = os = strdup(str); + if (!s) + return -1; + + for (;;) { + p = strchr(s, ','); + if (p) + *p = '\0'; + + for (br = branch_modes; br->name; br++) { + if (!strcasecmp(s, br->name)) + break; + } + if (!br->name) { + ui__warning("unknown branch filter %s," + " check man page\n", s); + goto error; + } + + *mode |= br->mode; + + if (!p) + break; + + s = p + 1; + } + } + ret = 0; + /* default to any branch */ + if ((*mode & ~ONLY_PLM) == 0) { + *mode = PERF_SAMPLE_BRANCH_ANY; + } +error: + free(os); + return ret; +} + +#ifdef HAVE_DWARF_UNWIND_SUPPORT +static int get_stack_size(char *str, unsigned long *_size) +{ + char *endptr; + unsigned long size; + unsigned long max_size = round_down(USHRT_MAX, sizeof(u64)); + + size = strtoul(str, &endptr, 0); + + do { + if (*endptr) + break; + + size = round_up(size, sizeof(u64)); + if (!size || size > max_size) + break; + + *_size = size; + return 0; + + } while (0); + + pr_err("callchain: Incorrect stack dump size (max %ld): %s\n", + max_size, str); + return -1; +} +#endif /* HAVE_DWARF_UNWIND_SUPPORT */ + +int record_parse_callchain(const char *arg, struct record_opts *opts) +{ + char *tok, *name, *saveptr = NULL; + char *buf; + int ret = -1; + + /* We need buffer that we know we can write to. */ + buf = malloc(strlen(arg) + 1); + if (!buf) + return -ENOMEM; + + strcpy(buf, arg); + + tok = strtok_r((char *)buf, ",", &saveptr); + name = tok ? : (char *)buf; + + do { + /* Framepointer style */ + if (!strncmp(name, "fp", sizeof("fp"))) { + if (!strtok_r(NULL, ",", &saveptr)) { + opts->call_graph = CALLCHAIN_FP; + ret = 0; + } else + pr_err("callchain: No more arguments " + "needed for -g fp\n"); + break; + +#ifdef HAVE_DWARF_UNWIND_SUPPORT + /* Dwarf style */ + } else if (!strncmp(name, "dwarf", sizeof("dwarf"))) { + const unsigned long default_stack_dump_size = 8192; + + ret = 0; + opts->call_graph = CALLCHAIN_DWARF; + opts->stack_dump_size = default_stack_dump_size; + + tok = strtok_r(NULL, ",", &saveptr); + if (tok) { + unsigned long size = 0; + + ret = get_stack_size(tok, &size); + opts->stack_dump_size = size; + } +#endif /* HAVE_DWARF_UNWIND_SUPPORT */ + } else { + pr_err("callchain: Unknown --call-graph option " + "value: %s\n", arg); + break; + } + + } while (0); + + free(buf); + return ret; +} + +static void callchain_debug(struct record_opts *opts) +{ + static const char *str[CALLCHAIN_MAX] = { "NONE", "FP", "DWARF" }; + + pr_debug("callchain: type %s\n", str[opts->call_graph]); + + if (opts->call_graph == CALLCHAIN_DWARF) + pr_debug("callchain: stack dump size %d\n", + opts->stack_dump_size); +} + +int record_parse_callchain_opt(const struct option *opt, + const char *arg, + int unset) +{ + struct record_opts *opts = opt->value; + int ret; + + opts->call_graph_enabled = !unset; + + /* --no-call-graph */ + if (unset) { + opts->call_graph = CALLCHAIN_NONE; + pr_debug("callchain: disabled\n"); + return 0; + } + + ret = record_parse_callchain(arg, opts); + if (!ret) + callchain_debug(opts); + + return ret; +} + +int record_callchain_opt(const struct option *opt, + const char *arg __maybe_unused, + int unset __maybe_unused) +{ + struct record_opts *opts = opt->value; + + opts->call_graph_enabled = !unset; + + if (opts->call_graph == CALLCHAIN_NONE) + opts->call_graph = CALLCHAIN_FP; + + callchain_debug(opts); return 0; +} -out_delete_session: - perf_session__delete(session); - return err; +static int perf_record_config(const char *var, const char *value, void *cb) +{ + struct record *rec = cb; + + if (!strcmp(var, "record.call-graph")) + return record_parse_callchain(value, &rec->opts); + + return perf_default_config(var, value, cb); } static const char * const record_usage[] = { @@ -788,142 +764,197 @@ static const char * const record_usage[] = { NULL }; -static bool force, append_file; +/* + * XXX Ideally would be local to cmd_record() and passed to a record__new + * because we need to have access to it in record__exit, that is called + * after cmd_record() exits, but since record_options need to be accessible to + * builtin-script, leave it here. + * + * At least we don't ouch it in all the other functions here directly. + * + * Just say no to tons of global variables, sigh. + */ +static struct record record = { + .opts = { + .mmap_pages = UINT_MAX, + .user_freq = UINT_MAX, + .user_interval = ULLONG_MAX, + .freq = 4000, + .target = { + .uses_mmap = true, + .default_per_cpu = true, + }, + }, +}; + +#define CALLCHAIN_HELP "setup and enables call-graph (stack chain/backtrace) recording: " + +#ifdef HAVE_DWARF_UNWIND_SUPPORT +const char record_callchain_help[] = CALLCHAIN_HELP "fp dwarf"; +#else +const char record_callchain_help[] = CALLCHAIN_HELP "fp"; +#endif +/* + * XXX Will stay a global variable till we fix builtin-script.c to stop messing + * with it and switch to use the library functions in perf_evlist that came + * from builtin-record.c, i.e. use record_opts, + * perf_evlist__prepare_workload, etc instead of fork+exec'in 'perf record', + * using pipes, etc. + */ const struct option record_options[] = { - OPT_CALLBACK('e', "event", NULL, "event", + OPT_CALLBACK('e', "event", &record.evlist, "event", "event selector. use 'perf list' to list available events", - parse_events), - OPT_CALLBACK(0, "filter", NULL, "filter", + parse_events_option), + OPT_CALLBACK(0, "filter", &record.evlist, "filter", "event filter", parse_filter), - OPT_INTEGER('p', "pid", &target_pid, + OPT_STRING('p', "pid", &record.opts.target.pid, "pid", "record events on existing process id"), - OPT_INTEGER('t', "tid", &target_tid, + OPT_STRING('t', "tid", &record.opts.target.tid, "tid", "record events on existing thread id"), - OPT_INTEGER('r', "realtime", &realtime_prio, + OPT_INTEGER('r', "realtime", &record.realtime_prio, "collect data with this RT SCHED_FIFO priority"), - OPT_BOOLEAN('R', "raw-samples", &raw_samples, + OPT_BOOLEAN(0, "no-buffering", &record.opts.no_buffering, + "collect data without buffering"), + OPT_BOOLEAN('R', "raw-samples", &record.opts.raw_samples, "collect raw sample records from all opened counters"), - OPT_BOOLEAN('a', "all-cpus", &system_wide, + OPT_BOOLEAN('a', "all-cpus", &record.opts.target.system_wide, "system-wide collection from all CPUs"), - OPT_BOOLEAN('A', "append", &append_file, - "append to the output file to do incremental profiling"), - OPT_STRING('C', "cpu", &cpu_list, "cpu", + OPT_STRING('C', "cpu", &record.opts.target.cpu_list, "cpu", "list of cpus to monitor"), - OPT_BOOLEAN('f', "force", &force, - "overwrite existing data file (deprecated)"), - OPT_U64('c', "count", &user_interval, "event period to sample"), - OPT_STRING('o', "output", &output_name, "file", + OPT_U64('c', "count", &record.opts.user_interval, "event period to sample"), + OPT_STRING('o', "output", &record.file.path, "file", "output file name"), - OPT_BOOLEAN('i', "no-inherit", &no_inherit, - "child tasks do not inherit counters"), - OPT_UINTEGER('F', "freq", &user_freq, "profile at this frequency"), - OPT_UINTEGER('m', "mmap-pages", &mmap_pages, "number of mmap data pages"), - OPT_BOOLEAN('g', "call-graph", &call_graph, - "do call-graph (stack chain/backtrace) recording"), + OPT_BOOLEAN_SET('i', "no-inherit", &record.opts.no_inherit, + &record.opts.no_inherit_set, + "child tasks do not inherit counters"), + OPT_UINTEGER('F', "freq", &record.opts.user_freq, "profile at this frequency"), + OPT_CALLBACK('m', "mmap-pages", &record.opts.mmap_pages, "pages", + "number of mmap data pages", + perf_evlist__parse_mmap_pages), + OPT_BOOLEAN(0, "group", &record.opts.group, + "put the counters into a counter group"), + OPT_CALLBACK_NOOPT('g', NULL, &record.opts, + NULL, "enables call-graph recording" , + &record_callchain_opt), + OPT_CALLBACK(0, "call-graph", &record.opts, + "mode[,dump_size]", record_callchain_help, + &record_parse_callchain_opt), OPT_INCR('v', "verbose", &verbose, "be more verbose (show counter open errors, etc)"), OPT_BOOLEAN('q', "quiet", &quiet, "don't print any message"), - OPT_BOOLEAN('s', "stat", &inherit_stat, + OPT_BOOLEAN('s', "stat", &record.opts.inherit_stat, "per thread counts"), - OPT_BOOLEAN('d', "data", &sample_address, + OPT_BOOLEAN('d', "data", &record.opts.sample_address, "Sample addresses"), - OPT_BOOLEAN('n', "no-samples", &no_samples, + OPT_BOOLEAN('T', "timestamp", &record.opts.sample_time, "Sample timestamps"), + OPT_BOOLEAN('P', "period", &record.opts.period, "Sample period"), + OPT_BOOLEAN('n', "no-samples", &record.opts.no_samples, "don't sample"), - OPT_BOOLEAN('N', "no-buildid-cache", &no_buildid, + OPT_BOOLEAN('N', "no-buildid-cache", &record.no_buildid_cache, "do not update the buildid cache"), + OPT_BOOLEAN('B', "no-buildid", &record.no_buildid, + "do not collect buildids in perf.data"), + OPT_CALLBACK('G', "cgroup", &record.evlist, "name", + "monitor event in cgroup name only", + parse_cgroups), + OPT_UINTEGER('D', "delay", &record.opts.initial_delay, + "ms to wait before starting measurement after program start"), + OPT_STRING('u', "uid", &record.opts.target.uid_str, "user", + "user to profile"), + + OPT_CALLBACK_NOOPT('b', "branch-any", &record.opts.branch_stack, + "branch any", "sample any taken branches", + parse_branch_stack), + + OPT_CALLBACK('j', "branch-filter", &record.opts.branch_stack, + "branch filter mask", "branch stack filter modes", + parse_branch_stack), + OPT_BOOLEAN('W', "weight", &record.opts.sample_weight, + "sample by weight (on special events only)"), + OPT_BOOLEAN(0, "transaction", &record.opts.sample_transaction, + "sample transaction flags (special events only)"), + OPT_BOOLEAN(0, "per-thread", &record.opts.target.per_thread, + "use per-thread mmaps"), OPT_END() }; -int cmd_record(int argc, const char **argv, const char *prefix __used) +int cmd_record(int argc, const char **argv, const char *prefix __maybe_unused) { - int i, j, err = -ENOMEM; + int err = -ENOMEM; + struct record *rec = &record; + char errbuf[BUFSIZ]; + + rec->evlist = perf_evlist__new(); + if (rec->evlist == NULL) + return -ENOMEM; + + perf_config(perf_record_config, rec); argc = parse_options(argc, argv, record_options, record_usage, PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc && target_pid == -1 && target_tid == -1 && - !system_wide && !cpu_list) + if (!argc && target__none(&rec->opts.target)) usage_with_options(record_usage, record_options); - if (force && append_file) { - fprintf(stderr, "Can't overwrite and append at the same time." - " You need to choose between -f and -A"); + if (nr_cgroups && !rec->opts.target.system_wide) { + ui__error("cgroup monitoring only available in" + " system-wide mode\n"); usage_with_options(record_usage, record_options); - } else if (append_file) { - write_mode = WRITE_APPEND; - } else { - write_mode = WRITE_FORCE; } symbol__init(); - if (no_buildid) + + if (symbol_conf.kptr_restrict) + pr_warning( +"WARNING: Kernel address maps (/proc/{kallsyms,modules}) are restricted,\n" +"check /proc/sys/kernel/kptr_restrict.\n\n" +"Samples in kernel functions may not be resolved if a suitable vmlinux\n" +"file is not found in the buildid cache or in the vmlinux path.\n\n" +"Samples in kernel modules won't be resolved at all.\n\n" +"If some relocation was applied (e.g. kexec) symbols may be misresolved\n" +"even with a suitable vmlinux or kallsyms file.\n\n"); + + if (rec->no_buildid_cache || rec->no_buildid) disable_buildid_cache(); - if (!nr_counters) { - nr_counters = 1; - attrs[0].type = PERF_TYPE_HARDWARE; - attrs[0].config = PERF_COUNT_HW_CPU_CYCLES; + if (rec->evlist->nr_entries == 0 && + perf_evlist__add_default(rec->evlist) < 0) { + pr_err("Not enough memory for event selector list\n"); + goto out_symbol_exit; } - if (target_pid != -1) { - target_tid = target_pid; - thread_num = find_all_tid(target_pid, &all_tids); - if (thread_num <= 0) { - fprintf(stderr, "Can't find all threads of pid %d\n", - target_pid); - usage_with_options(record_usage, record_options); - } - } else { - all_tids=malloc(sizeof(pid_t)); - if (!all_tids) - goto out_symbol_exit; + if (rec->opts.target.tid && !rec->opts.no_inherit_set) + rec->opts.no_inherit = true; - all_tids[0] = target_tid; - thread_num = 1; + err = target__validate(&rec->opts.target); + if (err) { + target__strerror(&rec->opts.target, err, errbuf, BUFSIZ); + ui__warning("%s", errbuf); } - for (i = 0; i < MAX_NR_CPUS; i++) { - for (j = 0; j < MAX_COUNTERS; j++) { - fd[i][j] = malloc(sizeof(int)*thread_num); - if (!fd[i][j]) - goto out_free_fd; - } - } - event_array = malloc( - sizeof(struct pollfd)*MAX_NR_CPUS*MAX_COUNTERS*thread_num); - if (!event_array) - goto out_free_fd; + err = target__parse_uid(&rec->opts.target); + if (err) { + int saved_errno = errno; - if (user_interval != ULLONG_MAX) - default_interval = user_interval; - if (user_freq != UINT_MAX) - freq = user_freq; + target__strerror(&rec->opts.target, err, errbuf, BUFSIZ); + ui__error("%s", errbuf); - /* - * User specified count overrides default frequency. - */ - if (default_interval) - freq = 0; - else if (freq) { - default_interval = freq; - } else { - fprintf(stderr, "frequency and count are zero, aborting\n"); - err = -EINVAL; - goto out_free_event_array; + err = -saved_errno; + goto out_symbol_exit; } - err = __cmd_record(argc, argv); + err = -ENOMEM; + if (perf_evlist__create_maps(rec->evlist, &rec->opts.target) < 0) + usage_with_options(record_usage, record_options); -out_free_event_array: - free(event_array); -out_free_fd: - for (i = 0; i < MAX_NR_CPUS; i++) { - for (j = 0; j < MAX_COUNTERS; j++) - free(fd[i][j]); + if (record_opts__config(&rec->opts)) { + err = -EINVAL; + goto out_symbol_exit; } - free(all_tids); - all_tids = NULL; + + err = __cmd_record(&record, argc, argv); out_symbol_exit: + perf_evlist__delete(rec->evlist); symbol__exit(); return err; } |
