Re: [PATCH v1] ringbuffer: Don't choose the process with adj equal OOM_SCORE_ADJ_MIN

[Steven Rostedt]

On Sun, 8 Apr 2018 13:54:59 +0800
Zhaoyang Huang <huangzhaoyang@xxxxxxxxx> wrote:

> On Sun, Apr 8, 2018 at 11:48 AM, Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
> > On Sun,  8 Apr 2018 10:16:23 +0800
> > Zhaoyang Huang <huangzhaoyang@xxxxxxxxx> wrote:
> >  
> >> Don't choose the process with adj == OOM_SCORE_ADJ_MIN which
> >> over-allocating pages for ring buffers.  
> >
> > Why?
> >
> > -- Steve  
> because in oom_evaluate_task, the process with adj == OOM_SCORE_ADJ_MIN will
> be suppressed by oom_badness, but with applying your latest patch,
> such process will
> be selected by oom_task_origin
> 
>         if (oom_task_origin(task)) {
>                 points = ULONG_MAX;
>                 goto select;
>         }
> 
>         points = oom_badness(task, NULL, oc->nodemask, oc->totalpages);
>         if (!points || points < oc->chosen_points)
>                 goto next;

And what's wrong with that?

-- Steve