Thread (19 messages) 19 messages, 4 authors, 2011-07-19

Re: [PATCH 4/4] mm: compaction: Abort compaction if too many pages are isolated and caller is asynchronous

From: Minchan Kim <hidden>
Date: 2011-06-07 15:50:39
Also in: lkml

On Tue, Jun 07, 2011 at 04:07:05PM +0100, Mel Gorman wrote:
quoted hunk ↗ jump to hunk
Asynchronous compaction is used when promoting to huge pages. This is
all very nice but if there are a number of processes in compacting
memory, a large number of pages can be isolated. An "asynchronous"
process can stall for long periods of time as a result with a user
reporting that firefox can stall for 10s of seconds. This patch aborts
asynchronous compaction if too many pages are isolated as it's better to
fail a hugepage promotion than stall a process.

Reported-and-tested-by: Ury Stankevich <redacted>
Signed-off-by: Mel Gorman <mgorman@suse.de>
---
 mm/compaction.c |   28 +++++++++++++++++++++++-----
 1 files changed, 23 insertions(+), 5 deletions(-)
diff --git a/mm/compaction.c b/mm/compaction.c
index 5c744ab..cb28580 100644
--- a/mm/compaction.c
+++ b/mm/compaction.c
@@ -251,11 +251,18 @@ static bool too_many_isolated(struct zone *zone)
 	return isolated > (inactive + active) / 2;
 }
 
+/* possible outcome of isolate_migratepages */
+typedef enum {
+	ISOLATE_ABORT,		/* Abort compaction now */
+	ISOLATE_NONE,		/* No pages isolated, continue scanning */
+	ISOLATE_SUCCESS,	/* Pages isolated, migrate */
+} isolate_migrate_t;
+
 /*
  * Isolate all pages that can be migrated from the block pointed to by
  * the migrate scanner within compact_control.
  */
-static unsigned long isolate_migratepages(struct zone *zone,
+static isolate_migrate_t isolate_migratepages(struct zone *zone,
 					struct compact_control *cc)
 {
 	unsigned long low_pfn, end_pfn;
@@ -272,7 +279,7 @@ static unsigned long isolate_migratepages(struct zone *zone,
 	/* Do not cross the free scanner or scan within a memory hole */
 	if (end_pfn > cc->free_pfn || !pfn_valid(low_pfn)) {
 		cc->migrate_pfn = end_pfn;
-		return 0;
+		return ISOLATE_NONE;
 	}
 
 	/*
@@ -281,10 +288,14 @@ static unsigned long isolate_migratepages(struct zone *zone,
 	 * delay for some time until fewer pages are isolated
 	 */
 	while (unlikely(too_many_isolated(zone))) {
+		/* async migration should just abort */
+		if (!cc->sync)
+			return ISOLATE_ABORT;
+
 		congestion_wait(BLK_RW_ASYNC, HZ/10);
 
 		if (fatal_signal_pending(current))
-			return 0;
+			return ISOLATE_ABORT;
 	}
 
 	/* Time to isolate some pages for migration */
@@ -369,7 +380,7 @@ static unsigned long isolate_migratepages(struct zone *zone,
 
 	trace_mm_compaction_isolate_migratepages(nr_scanned, nr_isolated);
 
-	return cc->nr_migratepages;
+	return ISOLATE_SUCCESS;
 }
 
 /*
@@ -533,8 +544,14 @@ static int compact_zone(struct zone *zone, struct compact_control *cc)
 		unsigned long nr_migrate, nr_remaining;
 		int err;
 
-		if (!isolate_migratepages(zone, cc))
+		switch (isolate_migratepages(zone, cc)) {
+		case ISOLATE_ABORT:
In this case, you change old behavior slightly.
In old case, we return COMPACT_PARTIAL to cancel migration.
But this patch makes to return COMPACT_SUCCESS.
At present, return value of compact_zone is only used by __alloc_pages_direct_compact
and it only consider COMPACT_SKIPPED so it would be not a problem.
But I think it would be better to return COMPACT_PARTIAL instead of COMPACT_CONTINUE
for consistency with compact_finished and right semantic for the future user of compact_zone.

-- 
Kind regards
Minchan Kim

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help