History

tracing: Call on_each_cpu() when adding or removing single pids from set_event_pid

For the case where pids are already in set_event_pid, and one is added or removed then each CPU should be checked to make sure that the new or old pid is on or not on a CPU. For example: # echo 123 >> set_event_pid or # echo '!123' >> set_event_pid Link: http://lkml.kernel.org/r/20151030061643.GA19480@cac

Steven Rostedt (Red Hat) committed 11 years ago

Commit 799fd441 parent fb662281 file · +7 −6Browse files

Unified · Split

kernel/trace/trace_events.c+7−6
@@ -1729,14 +1729,15 @@ ftrace_event_pid_write(struct file *filp, const char __user *ubuf, · 1728 unmodified lines
17291729 tr, INT_MAX);
17301730 register_trace_prio_sched_wakeup(event_filter_pid_sched_wakeup_probe_post,
17311731 tr, 0);
1732−
1733− /*
1734− * Ignoring of pids is done at task switch. But we have to
1735− * check for those tasks that are currently running.
1736− */
1737− on_each_cpu(ignore_task_cpu, tr, 1);
17381732 }
17391733
1734+ /*
1735+ * Ignoring of pids is done at task switch. But we have to
1736+ * check for those tasks that are currently running.
1737+ * Always do this in case a pid was appended or removed.
1738+ */
1739+ on_each_cpu(ignore_task_cpu, tr, 1);
1740+
17401741 mutex_unlock(&event_mutex);
17411742
17421743 ret = read;