[RESEND PATCH v2 10/12] mm: memcontrol: scale stat_threshold for byted-sized vmstat

[Muchun Song <songmuchun@bytedance.com>]

Some vmstat counters are being accounted in bytes not pages, so the
stat_threshold should also scale to bytes.

The vmstat counters are already long type for memcg (can reference
to struct lruvec_stat). For the global per-node vmstat counters
also can scale to long. But the maximum vmstat threshold is 125,
so the type of s32 is enough.

Signed-off-by: Muchun Song <songmuchun@bytedance.com>
---
 include/linux/mmzone.h | 17 ++++++-----------
 include/linux/vmstat.h |  1 -
 mm/vmstat.c            | 24 +++++++++++++-----------
 3 files changed, 19 insertions(+), 23 deletions(-)

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index 1f9c83778629..d53328551225 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -216,17 +216,12 @@ enum node_stat_item {
  */
 static __always_inline bool vmstat_item_in_bytes(int idx)
 {
-	/*
-	 * Global and per-node slab counters track slab pages.
-	 * It's expected that changes are multiples of PAGE_SIZE.
-	 * Internally values are stored in pages.
-	 *
-	 * Per-memcg and per-lruvec counters track memory, consumed
-	 * by individual slab objects. These counters are actually
-	 * byte-precise.
-	 */
 	return (idx == NR_SLAB_RECLAIMABLE_B ||
-		idx == NR_SLAB_UNRECLAIMABLE_B);
+		idx == NR_SLAB_UNRECLAIMABLE_B ||
+#if IS_ENABLED(CONFIG_SHADOW_CALL_STACK)
+		idx == NR_KERNEL_SCS_B ||
+#endif
+		idx == NR_KERNEL_STACK_B);
 }
 
 /*
@@ -340,7 +335,7 @@ struct per_cpu_pageset {
 
 struct per_cpu_nodestat {
 	s8 stat_threshold;
-	s8 vm_node_stat_diff[NR_VM_NODE_STAT_ITEMS];
+	s32 vm_node_stat_diff[NR_VM_NODE_STAT_ITEMS];
 };
 
 #endif /* !__GENERATING_BOUNDS.H */
diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h
index fd1a3d5d4926..afd84dc2398c 100644
--- a/include/linux/vmstat.h
+++ b/include/linux/vmstat.h
@@ -211,7 +211,6 @@ static inline unsigned long global_node_page_state(enum node_stat_item item)
 {
 	long x = atomic_long_read(&vm_node_stat[item]);
 
-	VM_WARN_ON_ONCE(vmstat_item_in_bytes(item));
 #ifdef CONFIG_SMP
 	if (x < 0)
 		x = 0;
diff --git a/mm/vmstat.c b/mm/vmstat.c
index 7fb0c7cb9516..25751b1d8e2e 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -341,13 +341,15 @@ void __mod_node_page_state(struct pglist_data *pgdat, enum node_stat_item item,
 				long delta)
 {
 	struct per_cpu_nodestat __percpu *pcp = pgdat->per_cpu_nodestats;
-	s8 __percpu *p = pcp->vm_node_stat_diff + item;
+	s32 __percpu *p = pcp->vm_node_stat_diff + item;
 	long x;
 	long t;
 
 	x = delta + __this_cpu_read(*p);
 
 	t = __this_cpu_read(pcp->stat_threshold);
+	if (vmstat_item_in_bytes(item))
+		t <<= PAGE_SHIFT;
 
 	if (unlikely(abs(x) > t)) {
 		node_page_state_add(x, pgdat, item);
@@ -399,15 +401,15 @@ void __inc_zone_state(struct zone *zone, enum zone_stat_item item)
 void __inc_node_state(struct pglist_data *pgdat, enum node_stat_item item)
 {
 	struct per_cpu_nodestat __percpu *pcp = pgdat->per_cpu_nodestats;
-	s8 __percpu *p = pcp->vm_node_stat_diff + item;
-	s8 v, t;
+	s32 __percpu *p = pcp->vm_node_stat_diff + item;
+	s32 v, t;
 
 	VM_WARN_ON_ONCE(vmstat_item_in_bytes(item));
 
 	v = __this_cpu_inc_return(*p);
 	t = __this_cpu_read(pcp->stat_threshold);
 	if (unlikely(v > t)) {
-		s8 overstep = t >> 1;
+		s32 overstep = t >> 1;
 
 		node_page_state_add(v + overstep, pgdat, item);
 		__this_cpu_write(*p, -overstep);
@@ -445,8 +447,8 @@ void __dec_zone_state(struct zone *zone, enum zone_stat_item item)
 void __dec_node_state(struct pglist_data *pgdat, enum node_stat_item item)
 {
 	struct per_cpu_nodestat __percpu *pcp = pgdat->per_cpu_nodestats;
-	s8 __percpu *p = pcp->vm_node_stat_diff + item;
-	s8 v, t;
+	s32 __percpu *p = pcp->vm_node_stat_diff + item;
+	s32 v, t;
 
 	VM_WARN_ON_ONCE(vmstat_item_in_bytes(item));
 
@@ -546,7 +548,7 @@ static inline void mod_node_state(struct pglist_data *pgdat,
        enum node_stat_item item, int delta, int overstep_mode)
 {
 	struct per_cpu_nodestat __percpu *pcp = pgdat->per_cpu_nodestats;
-	s8 __percpu *p = pcp->vm_node_stat_diff + item;
+	s32 __percpu *p = pcp->vm_node_stat_diff + item;
 	long o, n, t, z;
 
 	do {
@@ -563,6 +565,8 @@ static inline void mod_node_state(struct pglist_data *pgdat,
 		 * for all cpus in a node.
 		 */
 		t = this_cpu_read(pcp->stat_threshold);
+		if (vmstat_item_in_bytes(item))
+			t <<= PAGE_SHIFT;
 
 		o = this_cpu_read(*p);
 		n = delta + o;
@@ -829,7 +833,7 @@ static int refresh_cpu_vm_stats(bool do_pagesets)
 		struct per_cpu_nodestat __percpu *p = pgdat->per_cpu_nodestats;
 
 		for (i = 0; i < NR_VM_NODE_STAT_ITEMS; i++) {
-			int v;
+			s32 v;
 
 			v = this_cpu_xchg(p->vm_node_stat_diff[i], 0);
 			if (v) {
@@ -899,7 +903,7 @@ void cpu_vm_stats_fold(int cpu)
 
 		for (i = 0; i < NR_VM_NODE_STAT_ITEMS; i++)
 			if (p->vm_node_stat_diff[i]) {
-				int v;
+				s32 v;
 
 				v = p->vm_node_stat_diff[i];
 				p->vm_node_stat_diff[i] = 0;
@@ -1017,8 +1021,6 @@ unsigned long node_page_state(struct pglist_data *pgdat,
 {
 	long x = atomic_long_read(&pgdat->vm_stat[item]);
 
-	VM_WARN_ON_ONCE(vmstat_item_in_bytes(item));
-
 #ifdef CONFIG_SMP
 	if (x < 0)
 		x = 0;
-- 
2.11.0