lkml.org 
[lkml]   [2009]   [May]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 3/3] ftrace: add tracepoint for itimer


On Fri, 22 May 2009, Xiao Guangrong wrote:

> This patch can trace itimer start/expire/cancle event
>
> Example ftrace output:
> <...>-4183 [001] 66533.730163: itimer_start: task=main which=0 it_value=13000000000 it_interval=0
> <idle>-0 [001] 66546.698154: itimer_expire: task=main which=0
> <...>-4183 [000] 66546.698533: itimer_cancel: task=main which=0
>
> Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
> Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
> ---
> include/trace/events/timer.h | 66 ++++++++++++++++++++++++++++++++++++++++++
> kernel/itimer.c | 17 +++++++++--
> kernel/posix-cpu-timers.c | 3 ++
> 3 files changed, 83 insertions(+), 3 deletions(-)
>
> diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h
> index e5dc9d5..767c703 100644
> --- a/include/trace/events/timer.h
> +++ b/include/trace/events/timer.h
> @@ -175,6 +175,72 @@ TRACE_EVENT(hrtimer_cancel,
> TP_printk("timer=%p func=%pf", __entry->timer, __entry->function)
> );
>
> +TRACE_EVENT(itimer_start,
> +
> + TP_PROTO(int which, struct itimerval *value),
> +
> + TP_ARGS(which, value),
> +
> + TP_STRUCT__entry(
> + __field( int, which )
> + __field( unsigned long long, it_value )
> + __field( unsigned long long, it_interval )
> + __string( comm, current->comm )
> + ),
> +
> + TP_fast_assign(
> + __entry->which = which;
> + __entry->it_value = (which == ITIMER_REAL ?
> + ktime_to_ns(timeval_to_ktime(value->it_value)) :
> + timeval_to_cputime(&value->it_value));
> + __entry->it_interval = (which == ITIMER_REAL ?
> + ktime_to_ns(timeval_to_ktime(value->it_interval)) :
> + timeval_to_cputime(&value->it_interval));

Wouldn't it be faster to do these time conversions in the printk instead
of the trace point?

> + __assign_str(comm, current->comm);
> + ),
> +
> + TP_printk("task=%s which=%d it_value=%llu it_interval=%llu", __get_str(comm),
> + __entry->which, __entry->it_value, __entry->it_interval)
> +);
> +
> +TRACE_EVENT(itimer_expire,
> +
> + TP_PROTO(int which, struct task_struct *task),
> +
> + TP_ARGS(which, task),
> +
> + TP_STRUCT__entry(
> + __field( int , which )
> + __string( comm, task->comm )
> + ),
> +
> + TP_fast_assign(
> + __entry->which = which;
> + __assign_str(comm, task->comm);
> + ),
> +
> + TP_printk("task=%s which=%d", __get_str(comm), __entry->which)
> +);
> +
> +TRACE_EVENT(itimer_cancel,
> +
> + TP_PROTO(int which),
> +
> + TP_ARGS(which),
> +
> + TP_STRUCT__entry(
> + __field( int , which )
> + __string( comm, current->comm )
> + ),
> +
> + TP_fast_assign(
> + __entry->which = which;
> + __assign_str(comm, current->comm);
> + ),
> +
> + TP_printk("task=%s which=%d", __get_str(comm), __entry->which)
> +);
> +
> #endif /* _TRACE_TIMER_H */
>
> /* This part must be outside protection */
> diff --git a/kernel/itimer.c b/kernel/itimer.c
> index 58762f7..1d2dd64 100644
> --- a/kernel/itimer.c
> +++ b/kernel/itimer.c
> @@ -12,6 +12,7 @@
> #include <linux/time.h>
> #include <linux/posix-timers.h>
> #include <linux/hrtimer.h>
> +#include <trace/events/timer.h>
>
> #include <asm/uaccess.h>
>
> @@ -123,6 +124,7 @@ enum hrtimer_restart it_real_fn(struct hrtimer *timer)
> struct signal_struct *sig =
> container_of(timer, struct signal_struct, real_timer);
>
> + trace_itimer_expire(ITIMER_REAL, pid_task(sig->leader_pid, PIDTYPE_PID));
> kill_pid_info(SIGALRM, SEND_SIG_PRIV, sig->leader_pid);
>
> return HRTIMER_NORESTART;
> @@ -168,8 +170,11 @@ again:
> tsk->signal->it_real_incr =
> timeval_to_ktime(value->it_interval);
> hrtimer_start(timer, expires, HRTIMER_MODE_REL);
> - } else
> + trace_itimer_start(ITIMER_REAL, value);
> + } else{
> tsk->signal->it_real_incr.tv64 = 0;
> + trace_itimer_cancel(ITIMER_REAL);
> + }
>
> spin_unlock_irq(&tsk->sighand->siglock);
> break;
> @@ -186,7 +191,10 @@ again:
> jiffies_to_cputime(1));
> set_process_cpu_timer(tsk, CPUCLOCK_VIRT,
> &nval, &cval);
> - }
> + trace_itimer_start(ITIMER_VIRTUAL, value);
> + }else
> + trace_itimer_cancel(ITIMER_VIRTUAL);
> +
> tsk->signal->it_virt_expires = nval;
> tsk->signal->it_virt_incr = ninterval;
> spin_unlock_irq(&tsk->sighand->siglock);
> @@ -208,7 +216,10 @@ again:
> jiffies_to_cputime(1));
> set_process_cpu_timer(tsk, CPUCLOCK_PROF,
> &nval, &cval);
> - }
> + trace_itimer_start(ITIMER_PROF, value);
> + }else
> + trace_itimer_cancel(ITIMER_PROF);
> +
> tsk->signal->it_prof_expires = nval;
> tsk->signal->it_prof_incr = ninterval;
> spin_unlock_irq(&tsk->sighand->siglock);
> diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
> index bece7c0..edb4e8c 100644
> --- a/kernel/posix-cpu-timers.c
> +++ b/kernel/posix-cpu-timers.c
> @@ -8,6 +8,7 @@
> #include <linux/math64.h>
> #include <asm/uaccess.h>
> #include <linux/kernel_stat.h>
> +#include <trace/events/timer.h>
>
> /*
> * Called after updating RLIMIT_CPU to set timer expiration if necessary.
> @@ -1160,6 +1161,7 @@ static void check_process_timers(struct task_struct *tsk,
> sig->it_prof_expires = cputime_add(
> sig->it_prof_expires, ptime);
> }
> + trace_itimer_expire(ITIMER_PROF, tsk);
> __group_send_sig_info(SIGPROF, SEND_SIG_PRIV, tsk);
> }
> if (!cputime_eq(sig->it_prof_expires, cputime_zero) &&
> @@ -1176,6 +1178,7 @@ static void check_process_timers(struct task_struct *tsk,
> sig->it_virt_expires = cputime_add(
> sig->it_virt_expires, utime);
> }
> + trace_itimer_expire(ITIMER_VIRTUAL, tsk);
> __group_send_sig_info(SIGVTALRM, SEND_SIG_PRIV, tsk);
> }
> if (!cputime_eq(sig->it_virt_expires, cputime_zero) &&
> --
> 1.6.1.2
>
>


\
 
 \ /
  Last update: 2009-05-26 18:43    [W:0.055 / U:0.144 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site