On Wed 22-05-19 08:13:57, Chris Wilson wrote: > Quoting Michal Hocko (2019-05-22 07:34:42) > > On Wed 22-05-19 06:06:31, Tetsuo Handa wrote: > > [...] > > > Since OOM notifier will be called after shrinkers are attempted, > > > can i915 move from OOM notifier to shrinker? > > > > That would be indeed preferable. OOM notifier is an API from hell. > > We were^W are still trying to make the shrinker nonblocking to avoid > incurring horrible latencies for light direct reclaim. The consequence > of avoiding heavy work in the shrinker is that we moved it to the oom > notifier as being the last chance we have to return all (can be literally > all) the system memory. > > The alternative to using a separate oom notifier would be more > reclaim/shrinker phases? do_shrink_slab already knows the reclaim priority. So I guess we can push it to shrinkers via shrink_control so they can act depending on the reclaim pressure. -- Michal Hocko SUSE Labs _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
