All of lore.kernel.org
 help / color / mirror / Atom feed
* [nacked] mm-throttle-and-inc-min_seq-when-both-page-types-reach-min_nr_gens.patch removed from -mm tree
@ 2024-10-10 20:48 Andrew Morton
  0 siblings, 0 replies; only message in thread
From: Andrew Morton @ 2024-10-10 20:48 UTC (permalink / raw)
  To: mm-commits, yuzhao, zhaoyang.huang, akpm


The quilt patch titled
     Subject: mm: throttle and inc min_seq when both page types reach MIN_NR_GENS
has been removed from the -mm tree.  Its filename was
     mm-throttle-and-inc-min_seq-when-both-page-types-reach-min_nr_gens.patch

This patch was dropped because it was nacked

------------------------------------------------------
From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
Subject: mm: throttle and inc min_seq when both page types reach MIN_NR_GENS
Date: Wed, 9 Oct 2024 15:49:53 +0800

The test case of [1] leads to system hang which caused by a local watchdog
thread starved over 20s on a 5.5GB RAM ANDROID15(v6.6) system.  This
commit solve the issue by have the reclaimer be throttled and increase
min_seq if both page types reach MIN_NR_GENS, which may introduce a
livelock of switching type with holding lruvec->lru_lock.

[1]
launch below script 8 times simutanously which allocates 1GB virtual
memory and access it from user space by each thread.
$ costmem -c1024000 -b12800 -o0 &

Link: https://lkml.kernel.org/r/20241009074953.608591-1-zhaoyang.huang@unisoc.com
Signed-off-by: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
Cc: Yu Zhao <yuzhao@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 mm/vmscan.c |   16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

--- a/mm/vmscan.c~mm-throttle-and-inc-min_seq-when-both-page-types-reach-min_nr_gens
+++ a/mm/vmscan.c
@@ -4387,11 +4387,23 @@ static int scan_folios(struct lruvec *lr
 	int remaining = MAX_LRU_BATCH;
 	struct lru_gen_folio *lrugen = &lruvec->lrugen;
 	struct mem_cgroup *memcg = lruvec_memcg(lruvec);
+	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
 
 	VM_WARN_ON_ONCE(!list_empty(list));
 
-	if (get_nr_gens(lruvec, type) == MIN_NR_GENS)
-		return 0;
+	if (get_nr_gens(lruvec, type) == MIN_NR_GENS) {
+		/*
+		 * throttle for a while and then increase the min_seq since
+		 * both page types reach the limit.
+		 */
+		if (get_nr_gens(lruvec, !type) == MIN_NR_GENS) {
+			spin_unlock_irq(&lruvec->lru_lock);
+			reclaim_throttle(pgdat, VMSCAN_THROTTLE_ISOLATED);
+			spin_lock_irq(&lruvec->lru_lock);
+			try_to_inc_min_seq(lruvec, get_swappiness(lruvec, sc));
+		} else
+			return 0;
+	}
 
 	gen = lru_gen_from_seq(lrugen->min_seq[type]);
 
_

Patches currently in -mm which might be from zhaoyang.huang@unisoc.com are

mm-migrate-lru_refs_mask-bits-in-folio_migrate_flags.patch
mm-optimization-on-page-allocation-when-cma-enabled.patch


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2024-10-10 20:48 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-10-10 20:48 [nacked] mm-throttle-and-inc-min_seq-when-both-page-types-reach-min_nr_gens.patch removed from -mm tree Andrew Morton

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.