Re: [PATCH 1/3] mm, reclaim: make should_continue_reclaim perform dryrun detection

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 8/5/19 10:42 AM, Vlastimil Babka wrote:
> On 8/3/19 12:39 AM, Mike Kravetz wrote:
>> From: Hillf Danton <hdanton@xxxxxxxx>
>>
>> Address the issue of should_continue_reclaim continuing true too often
>> for __GFP_RETRY_MAYFAIL attempts when !nr_reclaimed and nr_scanned.
>> This could happen during hugetlb page allocation causing stalls for
>> minutes or hours.
>>
>> We can stop reclaiming pages if compaction reports it can make a progress.
>> A code reshuffle is needed to do that.
> 
>> And it has side-effects, however,
>> with allocation latencies in other cases but that would come at the cost
>> of potential premature reclaim which has consequences of itself.
> 
> Based on Mel's longer explanation, can we clarify the wording here? e.g.:
> 
> There might be side-effect for other high-order allocations that would
> potentially benefit from more reclaim before compaction for them to be
> faster and less likely to stall, but the consequences of
> premature/over-reclaim are considered worse.
> 
>> We can also bail out of reclaiming pages if we know that there are not
>> enough inactive lru pages left to satisfy the costly allocation.
>>
>> We can give up reclaiming pages too if we see dryrun occur, with the
>> certainty of plenty of inactive pages. IOW with dryrun detected, we are
>> sure we have reclaimed as many pages as we could.
>>
>> Cc: Mike Kravetz <mike.kravetz@xxxxxxxxxx>
>> Cc: Mel Gorman <mgorman@xxxxxxx>
>> Cc: Michal Hocko <mhocko@xxxxxxxxxx>
>> Cc: Vlastimil Babka <vbabka@xxxxxxx>
>> Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
>> Signed-off-by: Hillf Danton <hdanton@xxxxxxxx>
>> Tested-by: Mike Kravetz <mike.kravetz@xxxxxxxxxx>
>> Acked-by: Mel Gorman <mgorman@xxxxxxx>
> 
> Acked-by: Vlastimil Babka <vbabka@xxxxxxx>
> I will send some followup cleanup.

How about this?
----8<----
>From 0040b32462587171ad22395a56699cc036ad483f Mon Sep 17 00:00:00 2001
From: Vlastimil Babka <vbabka@xxxxxxx>
Date: Mon, 5 Aug 2019 12:49:40 +0200
Subject: [PATCH] mm, reclaim: cleanup should_continue_reclaim()

After commit "mm, reclaim: make should_continue_reclaim perform dryrun
detection", closer look at the function shows, that nr_reclaimed == 0 means
the function will always return false. And since non-zero nr_reclaimed implies
non_zero nr_scanned, testing nr_scanned serves no purpose, and so does the
testing for __GFP_RETRY_MAYFAIL.

This patch thus cleans up the function to test only !nr_reclaimed upfront, and
remove the __GFP_RETRY_MAYFAIL test and nr_scanned parameter completely.
Comment is also updated, explaining that approximating "full LRU list has been
scanned" with nr_scanned == 0 didn't really work.

Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
---
 mm/vmscan.c | 43 ++++++++++++++-----------------------------
 1 file changed, 14 insertions(+), 29 deletions(-)

diff --git a/mm/vmscan.c b/mm/vmscan.c
index ad498b76e492..db3c9e06a888 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2582,7 +2582,6 @@ static bool in_reclaim_compaction(struct scan_control *sc)
  */
 static inline bool should_continue_reclaim(struct pglist_data *pgdat,
 					unsigned long nr_reclaimed,
-					unsigned long nr_scanned,
 					struct scan_control *sc)
 {
 	unsigned long pages_for_compaction;
@@ -2593,28 +2592,18 @@ static inline bool should_continue_reclaim(struct pglist_data *pgdat,
 	if (!in_reclaim_compaction(sc))
 		return false;
 
-	/* Consider stopping depending on scan and reclaim activity */
-	if (sc->gfp_mask & __GFP_RETRY_MAYFAIL) {
-		/*
-		 * For __GFP_RETRY_MAYFAIL allocations, stop reclaiming if the
-		 * full LRU list has been scanned and we are still failing
-		 * to reclaim pages. This full LRU scan is potentially
-		 * expensive but a __GFP_RETRY_MAYFAIL caller really wants to succeed
-		 */
-		if (!nr_reclaimed && !nr_scanned)
-			return false;
-	} else {
-		/*
-		 * For non-__GFP_RETRY_MAYFAIL allocations which can presumably
-		 * fail without consequence, stop if we failed to reclaim
-		 * any pages from the last SWAP_CLUSTER_MAX number of
-		 * pages that were scanned. This will return to the
-		 * caller faster at the risk reclaim/compaction and
-		 * the resulting allocation attempt fails
-		 */
-		if (!nr_reclaimed)
-			return false;
-	}
+	/*
+	 * Stop if we failed to reclaim any pages from the last SWAP_CLUSTER_MAX
+	 * number of pages that were scanned. This will return to the caller
+	 * with the risk reclaim/compaction and the resulting allocation attempt
+	 * fails. In the past we have tried harder for __GFP_RETRY_MAYFAIL
+	 * allocations through requiring that the full LRU list has been scanned
+	 * first, by assuming that zero delta of sc->nr_scanned means full LRU
+	 * scan, but that approximation was wrong, and there were corner cases
+	 * where always a non-zero amount of pages were scanned.
+	 */
+	if (!nr_reclaimed)
+		return false;
 
 	/* If compaction would go ahead or the allocation would succeed, stop */
 	for (z = 0; z <= sc->reclaim_idx; z++) {
@@ -2641,11 +2630,7 @@ static inline bool should_continue_reclaim(struct pglist_data *pgdat,
 	if (get_nr_swap_pages() > 0)
 		inactive_lru_pages += node_page_state(pgdat, NR_INACTIVE_ANON);
 
-	return inactive_lru_pages > pages_for_compaction &&
-		/*
-		 * avoid dryrun with plenty of inactive pages
-		 */
-		nr_scanned && nr_reclaimed;
+	return inactive_lru_pages > pages_for_compaction;
 }
 
 static bool pgdat_memcg_congested(pg_data_t *pgdat, struct mem_cgroup *memcg)
@@ -2810,7 +2795,7 @@ static bool shrink_node(pg_data_t *pgdat, struct scan_control *sc)
 			wait_iff_congested(BLK_RW_ASYNC, HZ/10);
 
 	} while (should_continue_reclaim(pgdat, sc->nr_reclaimed - nr_reclaimed,
-					 sc->nr_scanned - nr_scanned, sc));
+					 sc));
 
 	/*
 	 * Kswapd gives up on balancing particular nodes after too
-- 
2.22.0





[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux