Re: [PATCH 3/3] ftrace: add tracepoint for itimer

From: Xiao Guangrong
Date: Wed May 27 2009 - 03:45:17 EST



Steven Rostedt wrote:
>
> 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@xxxxxxxxxx>
>> Signed-off-by: Xiao Guangrong <xiaoguangrong@xxxxxxxxxxxxxx>
>> ---
>> 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?
>

Hi Steven:

Thank for you review, I will correct it.
:-)

>> + __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
>>
>>
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/