On Tue, Nov 15, 2022 at 01:32:55PM +0100, Jan Kara wrote: > Currently, drop_caches are reclaiming node-by-node, looping on each node > until reclaim could not make progress. This can however leave quite some > slab entries (such as filesystem inodes) unreclaimed if objects say on > node 1 keep objects on node 0 pinned. So move the "loop until no > progress" loop to the node-by-node iteration to retry reclaim also on > other nodes if reclaim on some nodes made progress. This fixes problem > when drop_caches was not reclaiming lots of otherwise perfectly fine to > reclaim inodes. > > Reported-by: You Zhou <you.zhou@xxxxxxxxx> > Reported-and-tested-by: Pengfei Xu <pengfei.xu@xxxxxxxxx> > Signed-off-by: Jan Kara <jack@xxxxxxx> Reviewed-by: Shakeel Butt <shakeelb@xxxxxxxxxx>
