mm: don't wait on congested zones in balance_pgdat()

From: Zlatko Calusic <zlatko.calusic@iskon.hr>

Commit 92df3a7 ("mm: vmscan: throttle reclaim if encountering too
many dirty pages under writeback") introduced waiting on congested zones
based on a sane algorithm in shrink_inactive_list().

What this means is that there's no more need for throttling and
additional heuristics in balance_pgdat().  So, let's remove it and tidy
up the code.

Signed-off-by: Zlatko Calusic <zlatko.calusic@iskon.hr>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/vm_event_item.h

@@ -36,7 +36,6 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT,
 #endif
 		PGINODESTEAL, SLABS_SCANNED, KSWAPD_INODESTEAL,
 		KSWAPD_LOW_WMARK_HIT_QUICKLY, KSWAPD_HIGH_WMARK_HIT_QUICKLY,
-		KSWAPD_SKIP_CONGESTION_WAIT,
 		PAGEOUTRUN, ALLOCSTALL, PGROTATED,
 #ifdef CONFIG_NUMA_BALANCING
 		NUMA_PTE_UPDATES,

mm/vmscan.c

@@ -2617,7 +2617,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 							int *classzone_idx)
 {
 	bool pgdat_is_balanced = false;
-	struct zone *unbalanced_zone;
 	int i;
 	int end_zone = 0;	/* Inclusive.  0 = ZONE_DMA */
 	unsigned long total_scanned;
@@ -2648,9 +2647,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 
 	do {
 		unsigned long lru_pages = 0;
-		int has_under_min_watermark_zone = 0;
-
-		unbalanced_zone = NULL;
 
 		/*
 		 * Scan in the highmem->dma direction for the highest
@@ -2790,17 +2786,7 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 				continue;
 			}
 
-			if (!zone_balanced(zone, testorder, 0, end_zone)) {
-				unbalanced_zone = zone;
-				/*
-				 * We are still under min water mark.  This
-				 * means that we have a GFP_ATOMIC allocation
-				 * failure risk. Hurry up!
-				 */
-				if (!zone_watermark_ok_safe(zone, order,
-					    min_wmark_pages(zone), end_zone, 0))
-					has_under_min_watermark_zone = 1;
-			} else {
+			if (zone_balanced(zone, testorder, 0, end_zone))
 				/*
 				 * If a zone reaches its high watermark,
 				 * consider it to be no longer congested. It's
@@ -2809,8 +2795,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 				 * speculatively avoid congestion waits
 				 */
 				zone_clear_flag(zone, ZONE_CONGESTED);
-			}
-
 		}
 
 		/*
@@ -2827,17 +2811,6 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order,
 			break;		/* kswapd: all done */
 		}
 
-		/*
-		 * OK, kswapd is getting into trouble.  Take a nap, then take
-		 * another pass across the zones.
-		 */
-		if (total_scanned && (sc.priority < DEF_PRIORITY - 2)) {
-			if (has_under_min_watermark_zone)
-				count_vm_event(KSWAPD_SKIP_CONGESTION_WAIT);
-			else if (unbalanced_zone)
-				wait_iff_congested(unbalanced_zone, BLK_RW_ASYNC, HZ/10);
-		}
-
 		/*
 		 * We do this so kswapd doesn't build up large priorities for
 		 * example when it is freeing in parallel with allocators. It

mm/vmstat.c

@@ -770,7 +770,6 @@ const char * const vmstat_text[] = {
 	"kswapd_inodesteal",
 	"kswapd_low_wmark_hit_quickly",
 	"kswapd_high_wmark_hit_quickly",
-	"kswapd_skip_congestion_wait",
 	"pageoutrun",
 	"allocstall",