Re: threads-max observe limits

[Michal Hocko]

On Thu 19-09-19 14:33:24, Eric W. Biederman wrote:
> Michal Hocko <mhocko@xxxxxxxxxx> writes:
> 
> > On Tue 17-09-19 12:26:18, Eric W. Biederman wrote:
[...]
> >> Michal is it a very small effect your customers are seeing?
> >> Is it another bug somewhere else?
> >
> > I am still trying to get more information. Reportedly they see a
> > different auto tuned limit between two kernel versions which results in
> > an applicaton complaining. As already mentioned this might be a side
> > effect of something else and this is not yet fully analyzed. My main
> > point for bringing up this discussion is ...
> 
> Please this sounds like the kind of issue that will reveal something
> deeper about what is going on.

Yes, I have pushed for that information even before starting discussion
here. I haven't heard from the customer yet unfortunatelly.
 
> >> b) Not being able to bump threads_max to the physical limit of
> >>    the machine is very clearly a regression.
> >
> > ... exactly this part. The changelog of the respective patch doesn't
> > really exaplain why it is needed except of "it sounds like a good idea
> > to be consistent".
> 
> I suggest doing a partial revert to just:
> 
> diff --git a/kernel/fork.c b/kernel/fork.c
> index 7a74ade4e7d6..de8264ea34a7 100644
> --- a/kernel/fork.c
> +++ b/kernel/fork.c
> @@ -2943,7 +2943,7 @@ int sysctl_max_threads(struct ctl_table *table, int write,
>  	if (ret || !write)
>  		return ret;
>  
> -	set_max_threads(threads);
> +	max_threads = threads;
>  
>  	return 0;
>  }
> 
> proc_dointvec_minmax limiting the values to MIN_THREADS and MAX_THREADS
> is justifiable.  Those are the minimum and maximum values the kernel can
> function with.

MAX_THREADS limit makes perfect sense because crossing it reflects a
real constrain for correctness. MIN_THREADS, on the other hand, is only
the low gate for auto tuning to not come with an unbootable system. I do
not think we should jump into the admin call on the lower bound. There
might be a good reason to restrict the number of threads to 1. So here
is what I would like to go with.