Skip to contentSkip to conversation

mm: added node_reclaim successes to VM event counters

PATCH Applied Matthew Cassell 1 participant 1 message Jul 22, 2024

Subsystems (4)
Memory management MM core MGLRU MM reclaim
lore
Share & tools
lore

Matthew Cassell

author
#0 lore raw

PATCH 1/1 mm: added node_reclaim successes to VM event counters

/proc/vmstat currently shows the number of node_reclaim() failures when vm.zone_reclaim_mode is set appropriately. It would be convenient to have the number of successes right next to zone_reclaim_failed (similar to compaction and migration).

While just a trivially addition to the vmstat file. It was helpful during benchmarking to not have to probe node_reclaim() to observe the success/failure ratio.

Signed-off-by Matthew Cassell <[email protected]>
---
 include/linux/vm_event_item.h | 1 +
 mm/vmscan.c                   | 4 +++-
 mm/vmstat.c                   | 1 +
 3 files changed, 5 insertions(+), 1 deletion(-)
include/linux/vm_event_item.h +1 -0
index 747943bc8cc2..6fff8ed3b1fd 100644
@@ -50,6 +50,7 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT,
 		PGSTEAL_ANON,
 		PGSTEAL_FILE,
 #ifdef CONFIG_NUMA
+		PGSCAN_ZONE_RECLAIM_SUCCESS,
 		PGSCAN_ZONE_RECLAIM_FAILED,
 #endif
 		PGINODESTEAL, SLABS_SCANNED, KSWAPD_INODESTEAL,
mm/vmscan.c +3 -1
index 2e34de9cd0d4..321f259d713a 100644
@@ -7496,7 +7496,9 @@ int node_reclaim(struct pglist_data *pgdat, gfp_t gfp_mask, unsigned int order)
 	ret = __node_reclaim(pgdat, gfp_mask, order);
 	clear_bit(PGDAT_RECLAIM_LOCKED, &pgdat->flags);
 
-	if (!ret)
+	if (ret)
+		count_vm_event(PGSCAN_ZONE_RECLAIM_SUCCESS);
+	else
 		count_vm_event(PGSCAN_ZONE_RECLAIM_FAILED);
 
 	return ret;
mm/vmstat.c +1 -0
index 8507c497218b..27e02b188da8 100644
@@ -1295,6 +1295,7 @@ const char * const vmstat_text[] = {
 	"pgsteal_file",
 
 #ifdef CONFIG_NUMA
+	"zone_reclaim_success",
 	"zone_reclaim_failed",
 #endif
 	"pginodesteal",
Signature · 1 line
2.34.1