On Mon 05-02-24 20:01:40, T.J. Mercier wrote: > On Mon, Feb 5, 2024 at 1:16 PM Michal Hocko <mhocko@xxxxxxxx> wrote: > > > > On Mon 05-02-24 12:47:47, T.J. Mercier wrote: > > > On Mon, Feb 5, 2024 at 12:36 PM Michal Hocko <mhocko@xxxxxxxx> wrote: > > [...] > > > > This of something like > > > > timeout $TIMEOUT echo $TARGET > $MEMCG_PATH/memory.reclaim > > > > where timeout acts as a stop gap if the reclaim cannot finish in > > > > TIMEOUT. > > > > > > Yeah I get the desired behavior, but using sc->nr_reclaimed to achieve > > > it is what's bothering me. > > > > I am not really happy about this subtlety. If we have a better way then > > let's do it. Better in its own patch, though. > > > > > It's already wired up that way though, so if you want to make this > > > change now then I can try to test for the difference using really > > > large reclaim targets. > > > > Yes, please. If you want it a separate patch then no objection from me > > of course. If you do no like the nr_to_reclaim bailout then maybe we can > > go with a simple break out flag in scan_control. > > > > Thanks! > > It's a bit difficult to test under the too_many_isolated check, so I > moved the fatal_signal_pending check outside and tried with that. > Performing full reclaim on the /uid_0 cgroup with a 250ms delay before > SIGKILL, I got an average of 16ms better latency with > sc->nr_to_reclaim across 20 runs ignoring one 1s outlier with > SWAP_CLUSTER_MAX. This will obviously scale with the number of memcgs in the hierarchy but you are right that too_many_isolated makes the whole fatal_signal_pending check rather inefficient. I haven't missed that. The reclaim path is rather convoluted so this will likely be more complex than I anticipated. I will think about that some more. In order to not delay your patch, please repost with suggested updates to the changelog. This needs addressing IMO but I do not think this is critical at this stage. -- Michal Hocko SUSE Labs
