The quilt patch titled Subject: mm: memcg: use larger batches for proactive reclaim has been removed from the -mm tree. Its filename was mm-memcg-use-larger-batches-for-proactive-reclaim.patch This patch was dropped because it was merged into the mm-stable branch of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm ------------------------------------------------------ From: "T.J. Mercier" <tjmercier@xxxxxxxxxx> Subject: mm: memcg: use larger batches for proactive reclaim Date: Fri, 2 Feb 2024 23:38:54 +0000 Before 388536ac291 ("mm:vmscan: fix inaccurate reclaim during proactive reclaim") we passed the number of pages for the reclaim request directly to try_to_free_mem_cgroup_pages, which could lead to significant overreclaim. After 0388536ac291 the number of pages was limited to a maximum 32 (SWAP_CLUSTER_MAX) to reduce the amount of overreclaim. However such a small batch size caused a regression in reclaim performance due to many more reclaim start/stop cycles inside memory_reclaim. The restart cost is amortized over more pages with larger batch sizes, and becomes a significant component of the runtime if the batch size is too small. Reclaim tries to balance nr_to_reclaim fidelity with fairness across nodes and cgroups over which the pages are spread. As such, the bigger the request, the bigger the absolute overreclaim error. Historic in-kernel users of reclaim have used fixed, small sized requests to approach an appropriate reclaim rate over time. When we reclaim a user request of arbitrary size, use decaying batch sizes to manage error while maintaining reasonable throughput. MGLRU enabled - memcg LRU used root - full reclaim pages/sec time (sec) pre-0388536ac291 : 68047 10.46 post-0388536ac291 : 13742 inf (reclaim-reclaimed)/4 : 67352 10.51 MGLRU enabled - memcg LRU not used /uid_0 - 1G reclaim pages/sec time (sec) overreclaim (MiB) pre-0388536ac291 : 258822 1.12 107.8 post-0388536ac291 : 105174 2.49 3.5 (reclaim-reclaimed)/4 : 233396 1.12 -7.4 MGLRU enabled - memcg LRU not used /uid_0 - full reclaim pages/sec time (sec) pre-0388536ac291 : 72334 7.09 post-0388536ac291 : 38105 14.45 (reclaim-reclaimed)/4 : 72914 6.96 [tjmercier@xxxxxxxxxx: v4] Link: https://lkml.kernel.org/r/20240206175251.3364296-1-tjmercier@xxxxxxxxxx Link: https://lkml.kernel.org/r/20240202233855.1236422-1-tjmercier@xxxxxxxxxx Fixes: 0388536ac291 ("mm:vmscan: fix inaccurate reclaim during proactive reclaim") Signed-off-by: T.J. Mercier <tjmercier@xxxxxxxxxx> Reviewed-by: Yosry Ahmed <yosryahmed@xxxxxxxxxx> Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx> Reviewed-by: Michal Koutny <mkoutny@xxxxxxxx> Acked-by: Shakeel Butt <shakeelb@xxxxxxxxxx> Acked-by: Michal Hocko <mhocko@xxxxxxxx> Cc: Roman Gushchin <roman.gushchin@xxxxxxxxx> Cc: Shakeel Butt <shakeelb@xxxxxxxxxx> Cc: Muchun Song <songmuchun@xxxxxxxxxxxxx> Cc: Efly Young <yangyifei03@xxxxxxxxxxxx> Cc: Yu Zhao <yuzhao@xxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- mm/memcontrol.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) --- a/mm/memcontrol.c~mm-memcg-use-larger-batches-for-proactive-reclaim +++ a/mm/memcontrol.c @@ -6981,6 +6981,8 @@ static ssize_t memory_reclaim(struct ker reclaim_options = MEMCG_RECLAIM_MAY_SWAP | MEMCG_RECLAIM_PROACTIVE; while (nr_reclaimed < nr_to_reclaim) { + /* Will converge on zero, but reclaim enforces a minimum */ + unsigned long batch_size = (nr_to_reclaim - nr_reclaimed) / 4; unsigned long reclaimed; if (signal_pending(current)) @@ -6995,8 +6997,7 @@ static ssize_t memory_reclaim(struct ker lru_add_drain_all(); reclaimed = try_to_free_mem_cgroup_pages(memcg, - min(nr_to_reclaim - nr_reclaimed, SWAP_CLUSTER_MAX), - GFP_KERNEL, reclaim_options); + batch_size, GFP_KERNEL, reclaim_options); if (!reclaimed && !nr_retries--) return -EAGAIN; _ Patches currently in -mm which might be from tjmercier@xxxxxxxxxx are