Re: [PATCH 1/6] sched/cputime: Support other fields on kcpustat_field()

[Ingo Molnar <mingo@kernel.org>]

* Frederic Weisbecker <frederic@kernel.org> wrote:

> Provide support for user, nice, guest and guest_nice fields through
> kcpustat_field().
> 
> Whether we account the delta to a nice or not nice field is decided on
> top of the nice value snapshot taken at the time we call kcpustat_field().
> If the nice value of the task has been changed since the last vtime
> update, we may have inacurrate distribution of the nice VS unnice
> cputime.
> 
> However this is considered as a minor issue compared to the proper fix
> that would involve interrupting the target on nice updates, which is
> undesired on nohz_full CPUs.
> 
> Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
> Cc: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Rik van Riel <riel@surriel.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Wanpeng Li <wanpengli@tencent.com>
> Cc: Ingo Molnar <mingo@kernel.org>
> ---
>  kernel/sched/cputime.c | 53 +++++++++++++++++++++++++++++++++---------
>  1 file changed, 42 insertions(+), 11 deletions(-)
> 
> diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
> index e0cd20693ef5..b2cf544e2109 100644
> --- a/kernel/sched/cputime.c
> +++ b/kernel/sched/cputime.c
> @@ -912,11 +912,21 @@ void task_cputime(struct task_struct *t, u64 *utime, u64 *stime)
>  	} while (read_seqcount_retry(&vtime->seqcount, seq));
>  }
>  
> +static u64 kcpustat_user_vtime(struct vtime *vtime)
> +{
> +	if (vtime->state == VTIME_USER)
> +		return vtime->utime + vtime_delta(vtime);
> +	else if (vtime->state == VTIME_GUEST)
> +		return vtime->gtime + vtime_delta(vtime);
> +	return 0;
> +}
> +
>  static int kcpustat_field_vtime(u64 *cpustat,
> -				struct vtime *vtime,
> +				struct task_struct *tsk,
>  				enum cpu_usage_stat usage,
>  				int cpu, u64 *val)
>  {
> +	struct vtime *vtime = &tsk->vtime;
>  	unsigned int seq;
>  	int err;
>  
> @@ -946,9 +956,36 @@ static int kcpustat_field_vtime(u64 *cpustat,
>  
>  		*val = cpustat[usage];
>  
> -		if (vtime->state == VTIME_SYS)
> -			*val += vtime->stime + vtime_delta(vtime);
> -
> +		/*
> +		 * Nice VS unnice cputime accounting may be inaccurate if
> +		 * the nice value has changed since the last vtime update.
> +		 * But proper fix would involve interrupting target on nice
> +		 * updates which is a no go on nohz_full.

Well, we actually already interrupt the target in both sys_nice() and 
sys_setpriority() etc. syscall variants: we call set_user_nice() which 
calls resched_curr() and the task is sent an IPI and runs through a 
reschedule.

But ... I do agree that this kind of granularity of nice/non-nice 
accounting doesn't really matter in practice: the changing of nice values 
is a relatively low frequency operation on most systems.

But nevertheless the comment should probably be updated to reflect this.

Thanks,

	Ingo