sched/cputime: Increment kcpustat directly on irqtime account

The irqtime is accounted is nsecs and stored in
cpu_irq_time.hardirq_time and cpu_irq_time.softirq_time. Once the
accumulated amount reaches a new jiffy, this one gets accounted to the
kcpustat.

This was necessary when kcpustat was stored in cputime_t, which could at
worst have jiffies granularity. But now kcpustat is stored in nsecs
so this whole discretization game with temporary irqtime storage has
become unnecessary.

We can now directly account the irqtime to the kcpustat.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Link: http://lkml.kernel.org/r/1485832191-26889-17-git-send-email-fweisbec@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 0bdef50d88bc4bc5b1d46bcff3f5b646b1ed777f..bee6c97b1e830cab5f92e6861a93f9e06ee90801 100644 (file)
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -44,6 +44,7 @@ void disable_sched_clock_irqtime(void)
 void irqtime_account_irq(struct task_struct *curr)
 {
        struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
+       u64 *cpustat = kcpustat_this_cpu->cpustat;
        s64 delta;
        int cpu;
 
@@ -61,49 +62,35 @@ void irqtime_account_irq(struct task_struct *curr)
         * in that case, so as not to confuse scheduler with a special task
         * that do not consume any time, but still wants to run.
         */
-       if (hardirq_count())
-               irqtime->hardirq_time += delta;
-       else if (in_serving_softirq() && curr != this_cpu_ksoftirqd())
-               irqtime->softirq_time += delta;
+       if (hardirq_count()) {
+               cpustat[CPUTIME_IRQ] += delta;
+               irqtime->tick_delta += delta;
+       } else if (in_serving_softirq() && curr != this_cpu_ksoftirqd()) {
+               cpustat[CPUTIME_SOFTIRQ] += delta;
+               irqtime->tick_delta += delta;
+       }
 
        u64_stats_update_end(&irqtime->sync);
 }
 EXPORT_SYMBOL_GPL(irqtime_account_irq);
 
-static cputime_t irqtime_account_update(u64 irqtime, int idx, cputime_t maxtime)
+static cputime_t irqtime_tick_accounted(cputime_t maxtime)
 {
-       u64 *cpustat = kcpustat_this_cpu->cpustat;
-       cputime_t irq_cputime;
-
-       irq_cputime = nsecs_to_cputime64(irqtime - cpustat[idx]);
-       irq_cputime = min(irq_cputime, maxtime);
-       cpustat[idx] += cputime_to_nsecs(irq_cputime);
-
-       return irq_cputime;
-}
+       struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
+       cputime_t delta;
 
-static cputime_t irqtime_account_hi_update(cputime_t maxtime)
-{
-       return irqtime_account_update(__this_cpu_read(cpu_irqtime.hardirq_time),
-                                     CPUTIME_IRQ, maxtime);
-}
+       delta = nsecs_to_cputime(irqtime->tick_delta);
+       delta = min(delta, maxtime);
+       irqtime->tick_delta -= cputime_to_nsecs(delta);
 
-static cputime_t irqtime_account_si_update(cputime_t maxtime)
-{
-       return irqtime_account_update(__this_cpu_read(cpu_irqtime.softirq_time),
-                                     CPUTIME_SOFTIRQ, maxtime);
+       return delta;
 }
 
 #else /* CONFIG_IRQ_TIME_ACCOUNTING */
 
 #define sched_clock_irqtime    (0)
 
-static cputime_t irqtime_account_hi_update(cputime_t dummy)
-{
-       return 0;
-}
-
-static cputime_t irqtime_account_si_update(cputime_t dummy)
+static cputime_t irqtime_tick_accounted(cputime_t dummy)
 {
        return 0;
 }
@@ -280,10 +267,7 @@ static inline cputime_t account_other_time(cputime_t max)
        accounted = steal_account_process_time(max);
 
        if (accounted < max)
-               accounted += irqtime_account_hi_update(max - accounted);
-
-       if (accounted < max)
-               accounted += irqtime_account_si_update(max - accounted);
+               accounted += irqtime_tick_accounted(max - accounted);
 
        return accounted;
 }

diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index 6eeae7ebd99b74d6c9d6820e4e2ad2bcf8b3a7d8..8ff5cc539e8a103b50493280da550e632af810ca 100644 (file)
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -4,6 +4,7 @@
 #include <linux/sched/rt.h>
 #include <linux/u64_stats_sync.h>
 #include <linux/sched/deadline.h>
+#include <linux/kernel_stat.h>
 #include <linux/binfmts.h>
 #include <linux/mutex.h>
 #include <linux/spinlock.h>
@@ -1827,8 +1828,7 @@ static inline void nohz_balance_exit_idle(unsigned int cpu) { }
 
 #ifdef CONFIG_IRQ_TIME_ACCOUNTING
 struct irqtime {
-       u64                     hardirq_time;
-       u64                     softirq_time;
+       u64                     tick_delta;
        u64                     irq_start_time;
        struct u64_stats_sync   sync;
 };
@@ -1838,12 +1838,13 @@ DECLARE_PER_CPU(struct irqtime, cpu_irqtime);
 static inline u64 irq_time_read(int cpu)
 {
        struct irqtime *irqtime = &per_cpu(cpu_irqtime, cpu);
+       u64 *cpustat = kcpustat_cpu(cpu).cpustat;
        unsigned int seq;
        u64 total;
 
        do {
                seq = __u64_stats_fetch_begin(&irqtime->sync);
-               total = irqtime->softirq_time + irqtime->hardirq_time;
+               total = cpustat[CPUTIME_SOFTIRQ] + cpustat[CPUTIME_IRQ];
        } while (__u64_stats_fetch_retry(&irqtime->sync, seq));
 
        return total;