linux-kernel - Re: mmotm woes, mainly compaction

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <570CB9CE.1070408@suse.cz>
Date:	Tue, 12 Apr 2016 11:03:10 +0200
From:	Vlastimil Babka <vbabka@...e.cz>
To:	Hugh Dickins <hughd@...gle.com>, Michal Hocko <mhocko@...nel.org>,
	Andrew Morton <akpm@...ux-foundation.org>
Cc:	Michal Hocko <mhocko@...e.com>, linux-kernel@...r.kernel.org,
	linux-mm@...ck.org
Subject: Re: mmotm woes, mainly compaction

On 04/12/2016 09:18 AM, Hugh Dickins wrote:
> 3. /proc/sys/vm/stat_refresh warns nr_isolated_anon and nr_isolated_file
>     go increasingly negative under compaction: which would add delay when
>     should be none, or no delay when should delay.  putback_movable_pages()
>     decrements the NR_ISOLATED counts which acct_isolated() increments,
>     so isolate_migratepages_block() needs to acct before putback in that
>     special case, and isolate_migratepages_range() can always do the acct
>     itself, leaving migratepages putback to caller like most other places.

The isolate_migratepages_block() part is mmotm-specific, so I'll split
it out in this patch. Thanks for catching it and the lack of reset for
cc->nr_migratepages which wasn't mentioned in changelog so I added it.
 
> 5. It's easier to track the life of cc->migratepages if we don't assign
>     it to a migratelist variable.

This is also included here.

This is a -fix for:
mm-compaction-skip-blocks-where-isolation-fails-in-async-direct-compaction.patch

----8<----
>From 59a0075b6cf85045aa2dc5cee1f27797bcd0b3d2 Mon Sep 17 00:00:00 2001
From: Hugh Dickins <hughd@...gle.com>
Date: Tue, 12 Apr 2016 10:51:20 +0200
Subject: [PATCH] mm, compaction: prevent nr_isolated_* from going negative

/proc/sys/vm/stat_refresh warns nr_isolated_anon and nr_isolated_file
go increasingly negative under compaction: which would add delay when
should be none, or no delay when should delay.  putback_movable_pages()
decrements the NR_ISOLATED counts which acct_isolated() increments,
so isolate_migratepages_block() needs to acct before putback in that
special case. It's also useful to reset cc->nr_migratepages after putback
so we don't needlessly return too early on the COMPACT_CLUSTER_MAX check.

Also it's easier to track the life of cc->migratepages if we don't assign
it to a migratelist variable.
---
 mm/compaction.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/mm/compaction.c b/mm/compaction.c
index 67f886ecd773..ab649fba3d88 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -638,7 +638,6 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
 {
 	struct zone *zone = cc->zone;
 	unsigned long nr_scanned = 0, nr_isolated = 0;
-	struct list_head *migratelist = &cc->migratepages;
 	struct lruvec *lruvec;
 	unsigned long flags = 0;
 	bool locked = false;
@@ -817,7 +816,7 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
 		del_page_from_lru_list(page, lruvec, page_lru(page));
 
 isolate_success:
-		list_add(&page->lru, migratelist);
+		list_add(&page->lru, &cc->migratepages);
 		cc->nr_migratepages++;
 		nr_isolated++;
 
@@ -851,9 +850,11 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn,
 				spin_unlock_irqrestore(&zone->lru_lock,	flags);
 				locked = false;
 			}
-			putback_movable_pages(migratelist);
-			nr_isolated = 0;
+			acct_isolated(zone, cc);
+			putback_movable_pages(&cc->migratepages);
+			cc->nr_migratepages = 0;
 			cc->last_migrated_pfn = 0;
+			nr_isolated = 0;
 		}
 
 		if (low_pfn < next_skip_pfn) {
-- 
2.8.1