Re: [PATCH] psi: reduce min window size to 50ms

[Michal Hocko]

On Fri 24-02-23 13:07:57, Suren Baghdasaryan wrote:
> On Fri, Feb 24, 2023 at 4:47 AM Michal Hocko <mhocko@xxxxxxxx> wrote:
> >
> > On Tue 14-02-23 11:34:30, Suren Baghdasaryan wrote:
> > [...]
> > > Your suggestion to have this limit configurable sounds like obvious
> > > solution. I would like to get some opinions from other maintainers.
> > > Johannes, WDYT? CC'ing Michal to chime in as well since this is mostly
> > > related to memory stalls.
> >
> > I do not think that making this configurable helps much. Many users will
> > be bound to distribution config and also it would be hard to experiment
> > with a recompile cycle every time. This seems just too impractical.
> >
> > Is there any reason why we shouldn't allow any timeout? Shorter
> > timeouts could be restricted to a priviledged context to avoid an easy
> > way to swamp system by too frequent polling.
> 
> Hmm, ok. Maybe then we just ensure that only privileged users can set
> triggers and remove the min limit (use a >0 check)?

This could break existing userspace which is not privileged. I would
just go with CAP_SYS_NICE or similar with small (sub min) timeouts.

> > Btw. it seems that there is is only a limit on a single trigger per fd
> > but no limits per user so it doesn't sound too hard to end up with too
> > much polling even with a larger timeouts. To me it seems like we need to
> > contain the polling thread to be bound by the cpu controller.
> 
> Hmm. We have one "psimon" thread per cgroup (+1 system-level one) and
> poll_min_period for each thread is chosen as the min() of polling
> periods between triggers created in that group. So, a bad trigger that
> causes overly aggressive polling and polling thread being throttled,
> might affect other triggers in that cgroup.

Yes, and why that would be a problem?
-- 
Michal Hocko
SUSE Labs