Record guest as a VTIME state instead of guessing it from VTIME_SYS and
PF_VCPU. This is going to simplify the cputime read side especially as
its state machine is going to further expand in order to fully support
kcpustat on nohz_full.
Signed-off-by: Frederic Weisbecker <frede...@kernel.org>
Cc: Yauheni Kaliuta <yauheni.kali...@redhat.com>
Cc: Thomas Gleixner <t...@linutronix.de>
Cc: Rik van Riel <r...@redhat.com>
Cc: Peter Zijlstra <pet...@infradead.org>
Cc: Wanpeng Li <wanpen...@tencent.com>
Cc: Ingo Molnar <mi...@kernel.org>
---
include/linux/sched.h | 2 ++
kernel/sched/cputime.c | 18 +++++++++++-------
2 files changed, 13 insertions(+), 7 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 6d13938..d458d65 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -271,6 +271,8 @@ enum vtime_state {
VTIME_SYS,
/* Task runs in userspace in a CPU with VTIME active: */
VTIME_USER,
+ /* Task runs as guests in a CPU with VTIME active: */
+ VTIME_GUEST,
};
struct vtime {
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index a9f42cc..f64afd7 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -732,7 +732,7 @@ static void __vtime_account_kernel(struct task_struct *tsk,
struct vtime *vtime)
{
/* We might have scheduled out from guest path */
- if (tsk->flags & PF_VCPU)
+ if (vtime->state == VTIME_GUEST)
vtime_account_guest(tsk, vtime);
else
vtime_account_system(tsk, vtime);
@@ -787,6 +787,7 @@ void vtime_guest_enter(struct task_struct *tsk)
write_seqcount_begin(&vtime->seqcount);
vtime_account_system(tsk, vtime);
tsk->flags |= PF_VCPU;
+ vtime->state = VTIME_GUEST;
write_seqcount_end(&vtime->seqcount);
}
EXPORT_SYMBOL_GPL(vtime_guest_enter);
@@ -798,6 +799,7 @@ void vtime_guest_exit(struct task_struct *tsk)
write_seqcount_begin(&vtime->seqcount);
vtime_account_guest(tsk, vtime);
tsk->flags &= ~PF_VCPU;
+ vtime->state = VTIME_SYS;
write_seqcount_end(&vtime->seqcount);
}
EXPORT_SYMBOL_GPL(vtime_guest_exit);
@@ -825,6 +827,8 @@ void vtime_task_switch_generic(struct task_struct *prev)
write_seqcount_begin(&vtime->seqcount);
if (is_idle_task(current))
vtime->state = VTIME_IDLE;
+ else if (current->flags & PF_VCPU)
+ vtime->state = VTIME_GUEST;
else
vtime->state = VTIME_SYS;
vtime->starttime = sched_clock();
@@ -859,7 +863,7 @@ u64 task_gtime(struct task_struct *t)
seq = read_seqcount_begin(&vtime->seqcount);
gtime = t->gtime;
- if (vtime->state == VTIME_SYS && t->flags & PF_VCPU)
+ if (vtime->state == VTIME_GUEST)
gtime += vtime->gtime + vtime_delta(vtime);
} while (read_seqcount_retry(&vtime->seqcount, seq));
@@ -897,13 +901,13 @@ void task_cputime(struct task_struct *t, u64 *utime, u64
*stime)
delta = vtime_delta(vtime);
/*
- * Task runs either in user or kernel space, add pending nohz
time to
- * the right place.
+ * Task runs either in user (including guest) or kernel space,
+ * add pending nohz time to the right place.
*/
- if (vtime->state == VTIME_USER || t->flags & PF_VCPU)
- *utime += vtime->utime + delta;
- else if (vtime->state == VTIME_SYS)
+ if (vtime->state == VTIME_SYS)
*stime += vtime->stime + delta;
+ else
+ *utime += vtime->utime + delta;
} while (read_seqcount_retry(&vtime->seqcount, seq));
}
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_GEN */
--
2.7.4
