2009-02-26 18:47:11 +01:00
|
|
|
/*
|
|
|
|
* tracing clocks
|
|
|
|
*
|
|
|
|
* Copyright (C) 2009 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
|
|
|
|
*
|
|
|
|
* Implements 3 trace clock variants, with differing scalability/precision
|
|
|
|
* tradeoffs:
|
|
|
|
*
|
|
|
|
* - local: CPU-local trace clock
|
|
|
|
* - medium: scalable global clock with some jitter
|
|
|
|
* - global: globally monotonic, serialized clock
|
|
|
|
*
|
|
|
|
* Tracer plugins will chose a default from these clocks.
|
|
|
|
*/
|
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/hardirq.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/percpu.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/ktime.h>
|
2009-03-22 18:11:11 +01:00
|
|
|
#include <linux/trace_clock.h>
|
2009-02-26 18:47:11 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* trace_clock_local(): the simplest and least coherent tracing clock.
|
|
|
|
*
|
|
|
|
* Useful for tracing that does not cross to other CPUs nor
|
|
|
|
* does it go through idle events.
|
|
|
|
*/
|
|
|
|
u64 notrace trace_clock_local(void)
|
|
|
|
{
|
2009-03-10 19:03:43 +01:00
|
|
|
unsigned long flags;
|
|
|
|
u64 clock;
|
|
|
|
|
2009-02-26 18:47:11 +01:00
|
|
|
/*
|
|
|
|
* sched_clock() is an architecture implemented, fast, scalable,
|
|
|
|
* lockless clock. It is not guaranteed to be coherent across
|
|
|
|
* CPUs, nor across CPU idle events.
|
|
|
|
*/
|
2009-03-10 19:03:43 +01:00
|
|
|
raw_local_irq_save(flags);
|
|
|
|
clock = sched_clock();
|
|
|
|
raw_local_irq_restore(flags);
|
|
|
|
|
|
|
|
return clock;
|
2009-02-26 18:47:11 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* trace_clock(): 'inbetween' trace clock. Not completely serialized,
|
|
|
|
* but not completely incorrect when crossing CPUs either.
|
|
|
|
*
|
|
|
|
* This is based on cpu_clock(), which will allow at most ~1 jiffy of
|
|
|
|
* jitter between CPUs. So it's a pretty scalable clock, but there
|
|
|
|
* can be offsets in the trace data.
|
|
|
|
*/
|
|
|
|
u64 notrace trace_clock(void)
|
|
|
|
{
|
|
|
|
return cpu_clock(raw_smp_processor_id());
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* trace_clock_global(): special globally coherent trace clock
|
|
|
|
*
|
|
|
|
* It has higher overhead than the other trace clocks but is still
|
|
|
|
* an order of magnitude faster than GTOD derived hardware clocks.
|
|
|
|
*
|
|
|
|
* Used by plugins that need globally coherent timestamps.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static u64 prev_trace_clock_time;
|
|
|
|
|
|
|
|
static raw_spinlock_t trace_clock_lock ____cacheline_aligned_in_smp =
|
|
|
|
(raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED;
|
|
|
|
|
|
|
|
u64 notrace trace_clock_global(void)
|
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
int this_cpu;
|
|
|
|
u64 now;
|
|
|
|
|
|
|
|
raw_local_irq_save(flags);
|
|
|
|
|
|
|
|
this_cpu = raw_smp_processor_id();
|
|
|
|
now = cpu_clock(this_cpu);
|
|
|
|
/*
|
|
|
|
* If in an NMI context then dont risk lockups and return the
|
|
|
|
* cpu_clock() time:
|
|
|
|
*/
|
|
|
|
if (unlikely(in_nmi()))
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
__raw_spin_lock(&trace_clock_lock);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* TODO: if this happens often then maybe we should reset
|
|
|
|
* my_scd->clock to prev_trace_clock_time+1, to make sure
|
|
|
|
* we start ticking with the local clock from now on?
|
|
|
|
*/
|
|
|
|
if ((s64)(now - prev_trace_clock_time) < 0)
|
|
|
|
now = prev_trace_clock_time + 1;
|
|
|
|
|
|
|
|
prev_trace_clock_time = now;
|
|
|
|
|
|
|
|
__raw_spin_unlock(&trace_clock_lock);
|
|
|
|
|
|
|
|
out:
|
|
|
|
raw_local_irq_restore(flags);
|
|
|
|
|
|
|
|
return now;
|
|
|
|
}
|