On Tue 23-01-18 14:22:07, David Rientjes wrote: > On Tue, 23 Jan 2018, Michal Hocko wrote: > > > > It can't, because the current patchset locks the system into a single > > > selection criteria that is unnecessary and the mount option would become a > > > no-op after the policy per subtree becomes configurable by the user as > > > part of the hierarchy itself. > > > > This is simply not true! OOM victim selection has changed in the > > past and will be always a subject to changes in future. Current > > implementation doesn't provide any externally controlable selection > > policy and therefore the default can be assumed. Whatever that default > > means now or in future. The only contract added here is the kill full > > memcg if selected and that can be implemented on _any_ selection policy. > > > > The current implementation of memory.oom_group is based on top of a > selection implementation that is broken in three ways I have listed for > months: This doesn't lead to anywhere. You are not presenting any new arguments and you are ignoring feedback you have received so far. We have tried really hard. Considering different _independent_ people presented more or less consistent view on these points I think you should deeply reconsider how you take that feedback. > - allows users to intentionally/unintentionally evade the oom killer, > requires not locking the selection implementation for the entire > system, requires subtree control to prevent, makes a mount option > obsolete, and breaks existing users who would use the implementation > based on 4.16 if this were merged, > > - unfairly compares the root mem cgroup vs leaf mem cgroup such that > users must structure their hierarchy only for 4.16 in such a way > that _all_ processes are under hierarchical control and have no > power to create sub cgroups because of the point above and > completely breaks any user of oom_score_adj in a completely > undocumented and unspecified way, such that fixing that breakage > would also break any existing users who would use the implementation > based on 4.16 if this were merged, and > > - does not allow userspace to protect important cgroups, which can be > built on top. For the last time. This all can be done on top of the proposed solution without breaking the proposed user API. I am really _convinced_ that you underestimate how complex it is to provide a sane selection policy API and it will take _months_ to settle on something. Existing OOM APIs are a sad story and I definitly do not want to repeat same mistakes from the past. -- Michal Hocko SUSE Labs -- To unsubscribe from this list: send the line "unsubscribe cgroups" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html