Skip to contentSkip to conversation

mm/vmscan: reduce lru_lock contention via vmstat-derived scan-balance cost

RFC New Usama Arif 1 participant 2 messages Jun 26

Subsystems (7)
lore
Share & tools
lore

Usama Arif

author
#0 lore raw

cover mm/vmscan: reduce lru_lock contention via vmstat-derived scan-balance cost

The anon/file scan balance heuristic in get_scan_count() is fed by two scalars in struct lruvec (anon_cost, file_cost) that every reclaim producer updates under lruvec->lru_lock. The cost-recording work itself is trivial, but it both contends for and contributes to contention on lru_lock - which is often a contention point on memory-pressured workloads. Specifically:

- shrink_inactive_list() re-acquires lru_lock at function exit just to call lru_note_cost_unlock_irq().
- shrink_active_list() does the same after rotation accounting.
- workingset_refault() takes folio_lruvec_lock_irq() purely to record the refault cost.
- prepare_scan_control() snapshots anon_cost/file_cost under lru_lock.
- lru_note_cost_unlock_irq() itself walks parent_lruvec() and re-acquires lru_lock on every ancestor, multiplying the cost of every update by memcg-hierarchy depth.

This patch removes those producer-side acquisitions entirely. The producer-local inputs (PGROTATE_*, PGRECLAIM_PAGEOUT_*) become per-LRU vmstat counters; WORKINGSET_RESTORE_* already captures the refault input. prepare_scan_control() reads the raw cost signal lock-free from those vmstats and folds the delta into a per-lruvec accumulator. A dedicated per-lruvec cost_lock — not touched by isolate_lru_folios(), move_folios_to_lru(), or folio_add_lru() — serialises the accumulator RMW and the lrusize/4 halving check. Hierarchy aggregation is implicit in rstat propagation, so the parent_lruvec() walk and the lru_reparent_memcg() cost-splice both disappear.

Trade-offs:
- Signal freshness is slightly worse: cost reads see rstat- aggregated values that can lag until periodic / reader-triggered flushing. Decay timing is also coarser since multiple producer events may be batched into one read-side halving check. The cost signal is a heuristic feeding the anon-vs-file split, it's not a precise control loop — it's deliberately smoothed by the lrusize/4 halving. Producing/consuming it with a tiny lag should not be perceptible.
- Per-lruvec footprint grows by 2 unsigned longs + a spinlock, its a small cost.

== Numbers ==

Tested on a 176-core, 256 GB host. The benchmark drives sustained swap-out/refault inside a tight memcg using vm-scalability/usemem:

  usemem -n 16 --prealloc --prefault --random $((256*1024*1024))

run inside a two-level memcg with memory.max=512M on the leaf (4 GB anon working set has to fit in 512 MB -> continuous shrink_inactive_list + workingset_refault). A 16 GB swap file is used. Measurement is a 30 s perf lock record -a window over otherwise-idle hardware.

Workload rates are identical on both kernels (the bench drives the same memory pressure):

                          baseline    patched      delta
  pgscan_direct  / s      172,662     171,817      ~0%
  pgsteal_direct / s       67,162      66,306      ~0%
  workingset_refault_anon / s
                           40,696      39,830      ~0%

perf lock contention (total wait per 30 s window):

  Lock Name                Before      After     % change
  shrink_lruvec+0x770     722.84 ms    0         -100% (eliminated)
        (= lru_note_cost_unlock_irq)
  workingset_refault+0x167 385.26 ms   0         -100% (eliminated)
        (= lru_note_cost_refault)
  shrink_node+0x4ad       689.43 ms    26.95 ms  -96%
  shrink_active_list      208.34 ms    15.97 ms  -92%
  lru_add_drain_cpu+0x34    1.96 s    917.71 ms  -53%

Total LRU lock wait ~4.23 s ~1.66 s -61%

The two specific contention sites the patch removes (shrink_lruvec+0x770 = lru_note_cost_unlock_irq; workingset_refault+0x167 = lru_note_cost_refault) are completely absent from the patched perf-lock-contention output. Secondary reductions in shrink_node, shrink_active_list, lru_add_drain_cpu and pgrefill/pgactivate look like knock-on effects from removing the cost-recording overhead and the parent_lruvec walk.

The remaining ~1.66 s of LRU lock wait on the patched kernel is dominated by the per-CPU pagevec drain (lru_add_drain_cpu) and the main reclaim path in shrink_lruvec.

The numbers above can be reproduced using the script in [1].

== Alternatives considered ==

1. cost_lock for both producer and consumer (no vmstat indirection):
Keep the producer loop, just swap lru_lock for a new per-lruvec cost_lock. Decouples cost from LRU manipulation, but producers still synchronously contend on cost_lock, the parent_lruvec() walk is still required (O(memcg-depth) acquisitions per recording, now on cost_lock), and lru_reparent_memcg() still needs explicit cost-splice. We can do much better and this series removes the producer lock entirely and gets hierarchy propagation for "free" via rstat.

2. Attempt to switch to using MGLRU's scan model:
MGLRU has no anon_cost/file_cost at all. It replaces the cost heuristic with generation-based aging: per-LRU sequence numbers (min_seq/max_seq) age folios into generations, and the older-generation type is the one to scan. So lru_note_cost_unlock_irq() / lru_note_cost_refault() are simply not called when lru_gen_enabled() — by design it sidesteps every concern this patch addresses.
But MGLRU is not a substitute for fixing classic LRU:
- It relies on a lot of things including per-lruvec generation lists, bloom filters, mm_struct walk infrastructure, working-set protection tiers and a whole sysfs interface. Replacing classic LRU's cost recording with the MGLRU model would mean dragging in all of that.
- It changes scan-balance semantics, not just the locking, so it's a heuristic change we would need to evaluate separately. There are known regressions (database/anon-heavy workloads sensitive to swappiness, or file-cache-dominated workloads where MGLRU's bloom-filter protection differs from classic refault tracking).
This patch preserves classic-LRU semantics.

3. Atomic cost counter:
lrusize/4 halving has no clean atomic form, and the parent walk still has to run explicitly. Reusing vmstats gives per-CPU aggregation AND rstat hierarchy propagation for free.

4. Drop cost_lock from the existing patch and reuse lru_lock in the consumer (prepare_scan_control()):
Saves 1 lock space per lruvec but re-couples the cost path to LRU manipulation, though just from the consumer side this time. prepare_scan_control() runs at the start of every shrink_lruvec() cycle, so under sustained memory pressure it would take lru_lock on the hot path and block isolate_lru_folios() / move_folios_to_lru() / folio_add_lru() i.e. when reclaim is in flight. A dedicated cost_lock is never taken by anyone except the consumer cost calucation.

[1] https://gist.github.com/uarif1/a4eb33a86c5b2d7bbc55b42f0956e884

Usama Arif (1):
mm/vmscan: reduce lru_lock contention via vmstat-derived scan-balance
cost

 include/linux/mmzone.h | 11 +++++--
 include/linux/swap.h   |  3 --
 mm/memcontrol-v1.c     |  4 +--
 mm/memcontrol.c        |  4 +++
 mm/mmzone.c            |  1 +
 mm/swap.c              | 69 ------------------------------------------
 mm/vmscan.c            | 64 +++++++++++++++++++++++++++++++++------
 mm/vmstat.c            |  4 +++
 mm/workingset.c        |  5 ---
 9 files changed, 74 insertions(+), 91 deletions(-)
Signature · 1 line
2.53.0-Meta
author
#1 lore raw

RFC 1/1 mm/vmscan: reduce lru_lock contention via vmstat-derived scan-balance cost

The anon/file scan balance in get_scan_count() is driven by two scalars in struct lruvec, anon_cost and file_cost, accumulated by every reclaim producer under lruvec->lru_lock. The acquisition sites for cost work specifically are:

- shrink_inactive_list() re-takes lru_lock at function exit purely to call lru_note_cost_unlock_irq() with (nr_pageout, nr_scanned - nr_reclaimed). One acquisition per inactive shrink.
- shrink_active_list() does the same with (0, nr_rotated). One acquisition per active shrink.
- workingset_refault() takes the lock via folio_lruvec_lock_irq() purely to record the refault cost. One acquisition per refault.
- prepare_scan_control() takes lru_lock just to snapshot the two

    scalars into sc->{anon,file}_cost.

- lru_note_cost_unlock_irq() itself walks parent_lruvec and re-acquires lru_lock on each ancestor to propagate the update, adding O(memcg-depth) acquisitions per producer call.

This hurts because lru_lock is already a heavy contention point on memory-heavy workloads: every isolate_lru_folios(), move_folios_to_lru() and folio_add_lru() takes it. The cost work itself is trivial (two scalar bumps and one comparison), but it contends with and causes contention for actual LRU manipulation. The parent_lruvec() also walks multiplies cost-update overhead by memcg hierarchy depth.

Replace the producer-side accumulators with a read-side accumulator fed from per-LRU vmstat counters. The old producer formula was:

cost = nr_io * SWAP_CLUSTER_MAX + nr_rotated

Add explicit node_stat counters for the producer-local inputs:

PGRECLAIM_PAGEOUT_{ANON,FILE} - reclaim-driven pageout submissions
(formerly stat.nr_pageout, weighted by SWAP_CLUSTER_MAX).
PGROTATE_{ANON,FILE} - reclaim-driven rotations, bumped from
both shrink_inactive_list (by nr_scanned - nr_reclaimed) and shrink_active_list (by nr_rotated), unweighted.

WORKINGSET_RESTORE_{ANON,FILE} already captures the refault IO that lru_note_cost_refault() used to bill.

In prepare_scan_control() the raw cost signal is recomputed lock-free of lru_lock from monotonic counters:

now = (PGRECLAIM_PAGEOUT_X + WORKINGSET_RESTORE_X) * SWAP_CLUSTER_MAX
+ PGROTATE_X

The delta against a per-lruvec prev_cost[] snapshot is folded into cost_accum[]. The lrusize/4 halving threshold is preserved, but the decay check now happens at the read site instead of on every producer update.

A dedicated per-lruvec spinlock, cost_lock, serialises the prev_cost RMW, the accumulator update, and the halving check against concurrent reclaimers in the same memcg+node.

Hierarchy aggregation is now implicit in the vmstat accounting. The producer-side parent_lruvec() walk and lru_reparent_memcg() cost splice existed only because anon_cost/file_cost were private lruvec fields. With the cost expressed as lruvec vmstats, rstat propagates the underlying counters through the memcg hierarchy and prepare_scan_control() consumes the same ratelimited rstat view as the surrounding reclaim heuristics.

---
 include/linux/mmzone.h | 11 +++++--
 include/linux/swap.h   |  3 --
 mm/memcontrol-v1.c     |  4 +--
 mm/memcontrol.c        |  4 +++
 mm/mmzone.c            |  1 +
 mm/swap.c              | 69 ------------------------------------------
 mm/vmscan.c            | 64 +++++++++++++++++++++++++++++++++------
 mm/vmstat.c            |  4 +++
 mm/workingset.c        |  5 ---
 9 files changed, 74 insertions(+), 91 deletions(-)
include/linux/mmzone.h +9 -2
index ca2712187147..0627622a5184 100644
@@ -323,6 +323,10 @@ enum node_stat_item {
 	PGSCAN_PROACTIVE,
 	PGSCAN_ANON,
 	PGSCAN_FILE,
+	PGRECLAIM_PAGEOUT_ANON,
+	PGRECLAIM_PAGEOUT_FILE,
+	PGROTATE_ANON,
+	PGROTATE_FILE,
 	PGREFILL,
 #ifdef CONFIG_HUGETLB_PAGE
 	NR_HUGETLB,
@@ -763,9 +767,12 @@ struct lruvec {
 	 * These track the cost of reclaiming one LRU - file or anon -
 	 * over the other. As the observed cost of reclaiming one LRU
 	 * increases, the reclaim scan balance tips toward the other.
+	 * Updated and decayed at prepare_scan_control() time; cost_lock
+	 * serialises that update.
 	 */
-	unsigned long			anon_cost;
-	unsigned long			file_cost;
+	unsigned long			prev_cost[ANON_AND_FILE];
+	unsigned long			cost_accum[ANON_AND_FILE];
+	spinlock_t			cost_lock;
 	/* Non-resident age, driven by LRU movement */
 	atomic_long_t			nonresident_age;
 	/* Refaults at the time of last reclaim cycle */
include/linux/swap.h +0 -3
index 6d72778e6cc3..d35a4761ebd7 100644
@@ -309,9 +309,6 @@ extern unsigned long totalreserve_pages;
 
 
 /* linux/mm/swap.c */
-void lru_note_cost_unlock_irq(struct lruvec *lruvec, bool file,
-		unsigned int nr_io, unsigned int nr_rotated);
-void lru_note_cost_refault(struct folio *);
 void folio_add_lru(struct folio *);
 void folio_add_lru_vma(struct folio *, struct vm_area_struct *);
 void mark_page_accessed(struct page *);
mm/memcontrol-v1.c +2 -2
index 765069211567..c7a52bb68f4c 100644
@@ -1988,8 +1988,8 @@ void memcg1_stat_format(struct mem_cgroup *memcg, struct seq_buf *s)
 		for_each_online_pgdat(pgdat) {
 			mz = memcg->nodeinfo[pgdat->node_id];
 
-			anon_cost += mz->lruvec.anon_cost;
-			file_cost += mz->lruvec.file_cost;
+			anon_cost += mz->lruvec.cost_accum[WORKINGSET_ANON];
+			file_cost += mz->lruvec.cost_accum[WORKINGSET_FILE];
 		}
 		seq_buf_printf(s, "anon_cost %lu\n", anon_cost);
 		seq_buf_printf(s, "file_cost %lu\n", file_cost);
mm/memcontrol.c +4 -0
index 56cd4af08232..3c068ebefd97 100644
@@ -419,6 +419,10 @@ static const unsigned int memcg_node_stat_items[] = {
 	PGSCAN_PROACTIVE,
 	PGSCAN_ANON,
 	PGSCAN_FILE,
+	PGRECLAIM_PAGEOUT_ANON,
+	PGRECLAIM_PAGEOUT_FILE,
+	PGROTATE_ANON,
+	PGROTATE_FILE,
 	PGREFILL,
 #ifdef CONFIG_HUGETLB_PAGE
 	NR_HUGETLB,
mm/mmzone.c +1 -0
index 0c8f181d9d50..17139db4d291 100644
@@ -78,6 +78,7 @@ void lruvec_init(struct lruvec *lruvec)
 
 	memset(lruvec, 0, sizeof(struct lruvec));
 	spin_lock_init(&lruvec->lru_lock);
+	spin_lock_init(&lruvec->cost_lock);
 	zswap_lruvec_state_init(lruvec);
 
 	for_each_lru(lru)
mm/swap.c +0 -69

81 lines not on this page · raw message

mm/vmscan.c +54 -10

85 lines not on this page · raw message

mm/vmstat.c +4 -0
index f534972f517d..2dfed5e9b64d 100644
@@ -1289,6 +1289,10 @@ const char * const vmstat_text[] = {
 	[I(PGSCAN_PROACTIVE)]			= "pgscan_proactive",
 	[I(PGSCAN_ANON)]			= "pgscan_anon",
 	[I(PGSCAN_FILE)]			= "pgscan_file",
+	[I(PGRECLAIM_PAGEOUT_ANON)]		= "pgreclaim_pageout_anon",
+	[I(PGRECLAIM_PAGEOUT_FILE)]		= "pgreclaim_pageout_file",
+	[I(PGROTATE_ANON)]			= "pgrotate_anon",
+	[I(PGROTATE_FILE)]			= "pgrotate_file",
 	[I(PGREFILL)]				= "pgrefill",
 #ifdef CONFIG_HUGETLB_PAGE
 	[I(NR_HUGETLB)]				= "nr_hugetlb",
mm/workingset.c +0 -5
index f351798e723a..7ac2b88c80ae 100644
@@ -584,11 +584,6 @@ void workingset_refault(struct folio *folio, void *shadow)
 	/* Folio was active prior to eviction */
 	if (workingset) {
 		folio_set_workingset(folio);
-		/*
-		 * XXX: Move to folio_add_lru() when it supports new vs
-		 * putback
-		 */
-		lru_note_cost_refault(folio);
 		mod_lruvec_state(lruvec, WORKINGSET_RESTORE_BASE + file, nr);
 	}
 out:
Signature · 1 line
2.53.0-Meta