tick/sched: Release seqcount before invoking calc_load_global()

calc_load_global() does not need the sequence count protection.

[ tglx: Split it up properly and added comments ]

Signed-off-by: Yunfeng Ye <yeyunfeng@huawei.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/20201117132006.660902274@linutronix.de
This commit is contained in:
Yunfeng Ye 2020-11-17 14:19:48 +01:00 committed by Thomas Gleixner
parent 7a35bf2a6a
commit 896b969e67

View File

@ -20,6 +20,7 @@
#include <linux/sched/clock.h>
#include <linux/sched/stat.h>
#include <linux/sched/nohz.h>
#include <linux/sched/loadavg.h>
#include <linux/module.h>
#include <linux/irq_work.h>
#include <linux/posix-timers.h>
@ -107,7 +108,8 @@ static void tick_do_update_jiffies64(ktime_t now)
tick_period);
}
do_timer(ticks);
/* Advance jiffies to complete the jiffies_seq protected job */
jiffies_64 += ticks;
/*
* Keep the tick_next_period variable up to date. WRITE_ONCE()
@ -116,7 +118,15 @@ static void tick_do_update_jiffies64(ktime_t now)
WRITE_ONCE(tick_next_period,
ktime_add(last_jiffies_update, tick_period));
/*
* Release the sequence count. calc_global_load() below is not
* protected by it, but jiffies_lock needs to be held to prevent
* concurrent invocations.
*/
write_seqcount_end(&jiffies_seq);
calc_global_load();
raw_spin_unlock(&jiffies_lock);
update_wall_time();
}