On Wed, 2010-11-03 at 18:50 -0700, David Rientjes wrote:
quoted text
> On Thu, 4 Nov 2010, Figo.zhang wrote:
> 
> > > > CAP_SYS_RESOURCE also had better get 3% bonus for protection.
> > > >
> > > 
> > > 
> > > Would you like to elaborate as to why?
> > > 
> > > 
> > 
> > process with CAP_SYS_RESOURCE capibility which have system resource
> > limits, like journaling resource on ext3/4 filesystem, RTC clock. so it
> > also the same treatment as process with CAP_SYS_ADMIN.
> > 
> 
> NACK, there's no justification that these tasks should be given a 3% 
> memory bonus in the oom killer heuristic; in fact, since they can allocate 
> without limits it is more important to target these tasks if they are 
> using an egregious amount of memory.  CAP_SYS_RESOURCE threads have the 
> ability to lower their own oom_score_adj values, thus, they should protect 
> themselves if necessary like everything else.

In your new heuristic, you also get CAP_SYS_RESOURCE to protection.
see fs/proc/base.c, line 1167:
	if (oom_score_adj < task->signal->oom_score_adj &&
			!capable(CAP_SYS_RESOURCE)) {
		err = -EACCES;
		goto err_sighand;
	}

so i want to protect some process like normal process not
CAP_SYS_RESOUCE, i set a small oom_score_adj , if new oom_score_adj is
small than now and it is not limited resource, it will not adjust, that
seems not right?





--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/