On Wed, 13 Nov 2019, Steven Rostedt wrote:
On Wed, 13 Nov 2019 23:28:47 +0100 (CET) Thomas Gleixner email@example.com wrote:
On Fri, 8 Nov 2019, Arnd Bergmann wrote:
- TP_PROTO(int which, const struct itimerval *const value,
- TP_PROTO(int which, const struct itimerspec64 *const value, unsigned long long expires),
TP_ARGS(which, value, expires), @@ -321,12 +321,12 @@ TRACE_EVENT(itimer_state, __entry->which = which; __entry->expires = expires; __entry->value_sec = value->it_value.tv_sec;
__entry->value_usec = value->it_value.tv_usec;
- __entry->interval_sec = value->it_interval.tv_sec;
__entry->value_usec = value->it_value.tv_nsec / NSEC_PER_USEC;
__entry->interval_usec = value->it_interval.tv_usec;
__entry->interval_usec = value->it_interval.tv_nsec / NSEC_PER_USEC;
Hmm, having a division in a tracepoint is clearly suboptimal.
Right, we should move the division into the TP_printk()
__entry->interval_nsec = alue->it_interval.tv_nsec;
- TP_printk("which=%d expires=%llu it_value=%ld.%ld it_interval=%ld.%ld",
- TP_printk("which=%d expires=%llu it_value=%ld.%06ld it_interval=%ld.%06ld",
We print only 6 digits after the . so that would be even correct w/o a division. But it probably does not matter much.
Well, we still need the division in the printk, otherwise it will print more than 6. That's just the minimum and it will print the full number.
That's fine. The print is not really timing critical, the tracepoint very much so.