Messages in this thread | | | Subject | Re: [PATCH v1 4/8] perf stat: implement resume and pause control commands handling | From | Alexey Budankov <> | Date | Thu, 2 Apr 2020 18:06:47 +0300 |
| |
On 02.04.2020 17:17, Jiri Olsa wrote: > On Fri, Mar 27, 2020 at 11:48:31AM +0300, Alexey Budankov wrote: > > SNIP > >> >> - if (interval) { >> - ts.tv_sec = interval / USEC_PER_MSEC; >> - ts.tv_nsec = (interval % USEC_PER_MSEC) * NSEC_PER_MSEC; >> - } else if (timeout) { >> - ts.tv_sec = timeout / USEC_PER_MSEC; >> - ts.tv_nsec = (timeout % USEC_PER_MSEC) * NSEC_PER_MSEC; >> - } else { >> - ts.tv_sec = 1; >> - ts.tv_nsec = 0; >> - } >> - >> if (forks) { >> if (perf_evlist__prepare_workload(evsel_list, &target, argv, is_pipe, >> workload_exec_failed_signal) < 0) { >> @@ -688,18 +745,10 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx) >> perf_evlist__start_workload(evsel_list); >> enable_counters(); >> >> - if (interval || timeout) { >> - while (!waitpid(child_pid, &status, WNOHANG)) { >> - nanosleep(&ts, NULL); >> - if (timeout) >> - break; >> - process_interval(); >> - if (interval_count && !(--times)) >> - break; >> - } >> - } >> - if (child_pid != -1) >> + if (stat_config.ctl_fd == -1 && !interval && !timeout) >> wait4(child_pid, &status, 0, &stat_config.ru_data); >> + else >> + status = handle_events(child_pid, &stat_config); >> >> if (workload_exec_errno) { >> const char *emsg = str_error_r(workload_exec_errno, msg, sizeof(msg)); >> @@ -711,18 +760,7 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx) >> psignal(WTERMSIG(status), argv[0]); >> } else { >> enable_counters(); >> - while (!done) { >> - nanosleep(&ts, NULL); >> - if (!is_target_alive(&target, evsel_list->core.threads)) >> - break; >> - if (timeout) >> - break; >> - if (interval) { >> - process_interval(); >> - if (interval_count && !(--times)) >> - break; >> - } >> - } > > could you please separate the change to have: > - factor out the above too loops to the new function > - adding control fds handling to that function > > so it's more obvious we don't break anything
Accepted.
~Alexey
| |