+/* SPDX-License-Identifier: GPL-2.0 */
#undef TRACE_SYSTEM
#define TRACE_SYSTEM sched
#if !defined(LTTNG_TRACE_SCHED_H) || defined(TRACE_HEADER_MULTI_READ)
#define LTTNG_TRACE_SCHED_H
-#include "../../../probes/lttng-tracepoint-event.h"
+#include <probes/lttng-tracepoint-event.h>
#include <linux/sched.h>
+#include <linux/pid_namespace.h>
#include <linux/binfmts.h>
#include <linux/version.h>
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,9,0))
#include <linux/sched/rt.h>
#endif
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,19,0))
+#define lttng_proc_inum ns.inum
+#else
+#define lttng_proc_inum proc_inum
+#endif
+
+#define LTTNG_MAX_PID_NS_LEVEL 32
+
#ifndef _TRACE_SCHED_DEF_
#define _TRACE_SCHED_DEF_
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,19,0))
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,15,0))
+
+static inline long __trace_sched_switch_state(bool preempt, struct task_struct *p)
+{
+ unsigned int state;
+
+#ifdef CONFIG_SCHED_DEBUG
+ BUG_ON(p != current);
+#endif /* CONFIG_SCHED_DEBUG */
+
+ /*
+ * Preemption ignores task state, therefore preempted tasks are always
+ * RUNNING (we will not have dequeued if state != RUNNING).
+ */
+ if (preempt)
+ return TASK_REPORT_MAX;
+
+ /*
+ * task_state_index() uses fls() and returns a value from 0-8 range.
+ * Decrement it by 1 (except TASK_RUNNING state i.e 0) before using
+ * it for left shift operation to get the correct task->state
+ * mapping.
+ */
+ state = task_state_index(p);
+
+ return state ? (1 << (state - 1)) : state;
+}
+
+#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(4,14,0))
+
+static inline long __trace_sched_switch_state(bool preempt, struct task_struct *p)
+{
+ unsigned int state;
+
+#ifdef CONFIG_SCHED_DEBUG
+ BUG_ON(p != current);
+#endif /* CONFIG_SCHED_DEBUG */
+
+ /*
+ * Preemption ignores task state, therefore preempted tasks are always
+ * RUNNING (we will not have dequeued if state != RUNNING).
+ */
+ if (preempt)
+ return TASK_REPORT_MAX;
+
+ /*
+ * __get_task_state() uses fls() and returns a value from 0-8 range.
+ * Decrement it by 1 (except TASK_RUNNING state i.e 0) before using
+ * it for left shift operation to get the correct task->state
+ * mapping.
+ */
+ state = __get_task_state(p);
+
+ return state ? (1 << (state - 1)) : state;
+}
+
+#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(4,4,0))
+
+static inline long __trace_sched_switch_state(bool preempt, struct task_struct *p)
+{
+#ifdef CONFIG_SCHED_DEBUG
+ BUG_ON(p != current);
+#endif /* CONFIG_SCHED_DEBUG */
+ /*
+ * Preemption ignores task state, therefore preempted tasks are always RUNNING
+ * (we will not have dequeued if state != RUNNING).
+ */
+ return preempt ? TASK_RUNNING | TASK_STATE_MAX : p->state;
+}
+
+#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(3,19,0))
static inline long __trace_sched_switch_state(struct task_struct *p)
{
return state;
}
-#elif (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
+#else
static inline long __trace_sched_switch_state(struct task_struct *p)
{
/*
* Tracepoint for waking up a task:
*/
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,3,0) || \
+ LTTNG_RT_KERNEL_RANGE(4,1,10,11, 4,2,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,18,27,26, 3,19,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,14,61,63, 3,15,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,12,54,73, 3,13,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,10,97,106, 3,11,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,4,110,139, 3,5,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,2,77,111, 3,3,0,0))
+LTTNG_TRACEPOINT_EVENT_CLASS(sched_wakeup_template,
+
+ TP_PROTO(struct task_struct *p),
+
+ TP_ARGS(p),
+
+ TP_FIELDS(
+ ctf_array_text(char, comm, p->comm, TASK_COMM_LEN)
+ ctf_integer(pid_t, tid, p->pid)
+ ctf_integer(int, prio, p->prio - MAX_RT_PRIO)
+ ctf_integer(int, target_cpu, task_cpu(p))
+ )
+)
+#else /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,3,0)) */
LTTNG_TRACEPOINT_EVENT_CLASS(sched_wakeup_template,
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
TP_PROTO(struct task_struct *p, int success),
TP_ARGS(p, success),
-#else
- TP_PROTO(struct rq *rq, struct task_struct *p, int success),
-
- TP_ARGS(rq, p, success),
-#endif
TP_FIELDS(
ctf_array_text(char, comm, p->comm, TASK_COMM_LEN)
ctf_integer(pid_t, tid, p->pid)
- ctf_integer(int, prio, p->prio)
+ ctf_integer(int, prio, p->prio - MAX_RT_PRIO)
ctf_integer(int, success, success)
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,32))
ctf_integer(int, target_cpu, task_cpu(p))
-#endif
)
)
+#endif /* #else #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,3,0)) */
+
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,3,0) || \
+ LTTNG_RT_KERNEL_RANGE(4,1,10,11, 4,2,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,18,27,26, 3,19,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,14,61,63, 3,15,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,12,54,73, 3,13,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,10,97,106, 3,11,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,4,110,139, 3,5,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(3,2,77,111, 3,3,0,0))
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
+/*
+ * Tracepoint called when waking a task; this tracepoint is guaranteed to be
+ * called from the waking context.
+ */
+LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_wakeup_template, sched_waking,
+ TP_PROTO(struct task_struct *p),
+ TP_ARGS(p))
+/*
+ * Tracepoint called when the task is actually woken; p->state == TASK_RUNNNG.
+ * It it not always called from the waking context.
+ */
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_wakeup_template, sched_wakeup,
- TP_PROTO(struct task_struct *p, int success),
- TP_ARGS(p, success))
+ TP_PROTO(struct task_struct *p),
+ TP_ARGS(p))
/*
* Tracepoint for waking up a new task:
*/
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_wakeup_template, sched_wakeup_new,
- TP_PROTO(struct task_struct *p, int success),
- TP_ARGS(p, success))
+ TP_PROTO(struct task_struct *p),
+ TP_ARGS(p))
-#else /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
+#else
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_wakeup_template, sched_wakeup,
- TP_PROTO(struct rq *rq, struct task_struct *p, int success),
- TP_ARGS(rq, p, success))
+ TP_PROTO(struct task_struct *p, int success),
+ TP_ARGS(p, success))
/*
* Tracepoint for waking up a new task:
*/
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_wakeup_template, sched_wakeup_new,
- TP_PROTO(struct rq *rq, struct task_struct *p, int success),
- TP_ARGS(rq, p, success))
+ TP_PROTO(struct task_struct *p, int success),
+ TP_ARGS(p, success))
-#endif /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
+#endif /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,3,0)) */
/*
* Tracepoint for task switches, performed by the scheduler:
*/
LTTNG_TRACEPOINT_EVENT(sched_switch,
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
- TP_PROTO(struct task_struct *prev,
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,4,0))
+ TP_PROTO(bool preempt,
+ struct task_struct *prev,
struct task_struct *next),
- TP_ARGS(prev, next),
-#else /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
- TP_PROTO(struct rq *rq, struct task_struct *prev,
+ TP_ARGS(preempt, prev, next),
+#else
+ TP_PROTO(struct task_struct *prev,
struct task_struct *next),
- TP_ARGS(rq, prev, next),
-#endif /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
+ TP_ARGS(prev, next),
+#endif /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,4,0)) */
TP_FIELDS(
ctf_array_text(char, prev_comm, prev->comm, TASK_COMM_LEN)
ctf_integer(pid_t, prev_tid, prev->pid)
ctf_integer(int, prev_prio, prev->prio - MAX_RT_PRIO)
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
- ctf_integer(long, prev_state, __trace_sched_switch_state(prev))
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,4,0))
+ ctf_integer(long, prev_state, __trace_sched_switch_state(preempt, prev))
#else
- ctf_integer(long, prev_state, prev->state)
+ ctf_integer(long, prev_state, __trace_sched_switch_state(prev))
#endif
ctf_array_text(char, next_comm, next->comm, TASK_COMM_LEN)
ctf_integer(pid_t, next_tid, next->pid)
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_process_template, sched_process_free,
TP_PROTO(struct task_struct *p),
TP_ARGS(p))
-
+
/*
* Tracepoint for a task exiting:
/*
* Tracepoint for waiting on task to unschedule:
*/
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35))
LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_process_template, sched_wait_task,
TP_PROTO(struct task_struct *p),
TP_ARGS(p))
-#else /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
-LTTNG_TRACEPOINT_EVENT_INSTANCE(sched_process_template, sched_wait_task,
- TP_PROTO(struct rq *rq, struct task_struct *p),
- TP_ARGS(rq, p))
-#endif /* #if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,35)) */
/*
* Tracepoint for a waiting task:
* == child_pid, while creation of a thread yields to child_tid !=
* child_pid.
*/
-LTTNG_TRACEPOINT_EVENT(sched_process_fork,
+LTTNG_TRACEPOINT_EVENT_CODE(sched_process_fork,
TP_PROTO(struct task_struct *parent, struct task_struct *child),
TP_ARGS(parent, child),
+ TP_locvar(
+ pid_t vtids[LTTNG_MAX_PID_NS_LEVEL];
+ unsigned int ns_level;
+ ),
+
+ TP_code_pre(
+ if (child) {
+ struct pid *child_pid;
+ unsigned int i;
+
+ child_pid = task_pid(child);
+ tp_locvar->ns_level =
+ min_t(unsigned int, child_pid->level + 1,
+ LTTNG_MAX_PID_NS_LEVEL);
+ for (i = 0; i < tp_locvar->ns_level; i++)
+ tp_locvar->vtids[i] = child_pid->numbers[i].nr;
+ }
+ ),
+
TP_FIELDS(
ctf_array_text(char, parent_comm, parent->comm, TASK_COMM_LEN)
ctf_integer(pid_t, parent_tid, parent->pid)
ctf_integer(pid_t, parent_pid, parent->tgid)
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,8,0))
+ ctf_integer(unsigned int, parent_ns_inum,
+ ({
+ unsigned int parent_ns_inum = 0;
+
+ if (parent) {
+ struct pid_namespace *pid_ns;
+
+ pid_ns = task_active_pid_ns(parent);
+ if (pid_ns)
+ parent_ns_inum =
+ pid_ns->lttng_proc_inum;
+ }
+ parent_ns_inum;
+ }))
+#endif
ctf_array_text(char, child_comm, child->comm, TASK_COMM_LEN)
ctf_integer(pid_t, child_tid, child->pid)
+ ctf_sequence(pid_t, vtids, tp_locvar->vtids, u8, tp_locvar->ns_level)
ctf_integer(pid_t, child_pid, child->tgid)
- )
-)
-
-#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,33))
-/*
- * Tracepoint for sending a signal:
- */
-LTTNG_TRACEPOINT_EVENT(sched_signal_send,
-
- TP_PROTO(int sig, struct task_struct *p),
-
- TP_ARGS(sig, p),
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,8,0))
+ ctf_integer(unsigned int, child_ns_inum,
+ ({
+ unsigned int child_ns_inum = 0;
+
+ if (child) {
+ struct pid_namespace *pid_ns;
+
+ pid_ns = task_active_pid_ns(child);
+ if (pid_ns)
+ child_ns_inum =
+ pid_ns->lttng_proc_inum;
+ }
+ child_ns_inum;
+ }))
+#endif
+ ),
- TP_FIELDS(
- ctf_integer(int, sig, sig)
- ctf_array_text(char, comm, p->comm, TASK_COMM_LEN)
- ctf_integer(pid_t, tid, p->pid)
- )
+ TP_code_post()
)
-#endif
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,4,0))
/*
)
#endif
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,32))
/*
* XXX the below sched_stat tracepoints only apply to SCHED_OTHER/BATCH/IDLE
* adding sched_stat support to SCHED_FIFO/RR would be welcome.
ctf_integer(u64, vruntime, vruntime)
)
)
-#endif
-#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,37))
+#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,12,0) || \
+ LTTNG_RT_KERNEL_RANGE(4,9,27,18, 4,10,0,0) || \
+ LTTNG_RT_KERNEL_RANGE(4,11,5,1, 4,12,0,0))
+/*
+ * Tracepoint for showing priority inheritance modifying a tasks
+ * priority.
+ */
+LTTNG_TRACEPOINT_EVENT(sched_pi_setprio,
+
+ TP_PROTO(struct task_struct *tsk, struct task_struct *pi_task),
+
+ TP_ARGS(tsk, pi_task),
+
+ TP_FIELDS(
+ ctf_array_text(char, comm, tsk->comm, TASK_COMM_LEN)
+ ctf_integer(pid_t, tid, tsk->pid)
+ ctf_integer(int, oldprio, tsk->prio - MAX_RT_PRIO)
+ ctf_integer(int, newprio, pi_task ? pi_task->prio - MAX_RT_PRIO : tsk->prio - MAX_RT_PRIO)
+ )
+)
+#else
/*
* Tracepoint for showing priority inheritance modifying a tasks
* priority.
#endif /* LTTNG_TRACE_SCHED_H */
/* This part must be outside protection */
-#include "../../../probes/define_trace.h"
+#include <probes/define_trace.h>