Split IRQ-off and zone->lock while freeing pages from PCP list OpenNu…

…voton#1

Split the IRQ-off section while accessing the PCP list from zone->lock
while freeing pages.
Introcude  isolate_pcp_pages() which separates the pages from the PCP
list onto a temporary list and then free the temporary list via
free_pcppages_bulk().

Signed-off-by: Peter Zijlstra <[email protected]>
Signed-off-by: Sebastian Andrzej Siewior <[email protected]>

mm/page_alloc.c

@@ -1331,7 +1331,7 @@ static inline void prefetch_buddy(struct page *page)
 }
 
 /*
- * Frees a number of pages from the PCP lists
+ * Frees a number of pages which have been collected from the pcp lists.
  * Assumes all pages on list are in same zone, and of same order.
  * count is the number of pages to free.
  *
@@ -1342,14 +1342,40 @@ static inline void prefetch_buddy(struct page *page)
  * pinned" detection logic.
  */
 static void free_pcppages_bulk(struct zone *zone, int count,
-					struct per_cpu_pages *pcp)
+			       struct list_head *head)
+{
+	bool isolated_pageblocks;
+	struct page *page, *tmp;
+	unsigned long flags;
+
+	spin_lock_irqsave(&zone->lock, flags);
+	isolated_pageblocks = has_isolate_pageblock(zone);
+
+	/*
+	 * Use safe version since after __free_one_page(),
+	 * page->lru.next will not point to original list.
+	 */
+	list_for_each_entry_safe(page, tmp, head, lru) {
+		int mt = get_pcppage_migratetype(page);
+		/* MIGRATE_ISOLATE page should not go to pcplists */
+		VM_BUG_ON_PAGE(is_migrate_isolate(mt), page);
+		/* Pageblock could have been isolated meanwhile */
+		if (unlikely(isolated_pageblocks))
+			mt = get_pageblock_migratetype(page);
+
+		__free_one_page(page, page_to_pfn(page), zone, 0, mt, FPI_NONE);
+		trace_mm_page_pcpu_drain(page, 0, mt);
+	}
+	spin_unlock_irqrestore(&zone->lock, flags);
+}
+
+static void isolate_pcp_pages(int count, struct per_cpu_pages *pcp,
+			      struct list_head *dst)
 {
 	int migratetype = 0;
 	int batch_free = 0;
 	int prefetch_nr = 0;
-	bool isolated_pageblocks;
-	struct page *page, *tmp;
-	LIST_HEAD(head);
+	struct page *page;
 
 	/*
 	 * Ensure proper count is passed which otherwise would stuck in the
@@ -1386,7 +1412,7 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 			if (bulkfree_pcp_prepare(page))
 				continue;
 
-			list_add_tail(&page->lru, &head);
+			list_add_tail(&page->lru, dst);
 
 			/*
 			 * We are going to put the page back to the global
@@ -1401,26 +1427,6 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 				prefetch_buddy(page);
 		} while (--count && --batch_free && !list_empty(list));
 	}
-
-	spin_lock(&zone->lock);
-	isolated_pageblocks = has_isolate_pageblock(zone);
-
-	/*
-	 * Use safe version since after __free_one_page(),
-	 * page->lru.next will not point to original list.
-	 */
-	list_for_each_entry_safe(page, tmp, &head, lru) {
-		int mt = get_pcppage_migratetype(page);
-		/* MIGRATE_ISOLATE page should not go to pcplists */
-		VM_BUG_ON_PAGE(is_migrate_isolate(mt), page);
-		/* Pageblock could have been isolated meanwhile */
-		if (unlikely(isolated_pageblocks))
-			mt = get_pageblock_migratetype(page);
-
-		__free_one_page(page, page_to_pfn(page), zone, 0, mt, FPI_NONE);
-		trace_mm_page_pcpu_drain(page, 0, mt);
-	}
-	spin_unlock(&zone->lock);
 }
 
 static void free_one_page(struct zone *zone,
@@ -2938,13 +2944,18 @@ void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp)
 {
 	unsigned long flags;
 	int to_drain, batch;
+	LIST_HEAD(dst);
 
 	local_irq_save(flags);
 	batch = READ_ONCE(pcp->batch);
 	to_drain = min(pcp->count, batch);
 	if (to_drain > 0)
-		free_pcppages_bulk(zone, to_drain, pcp);
+		isolate_pcp_pages(to_drain, pcp, &dst);
+
 	local_irq_restore(flags);
+
+	if (to_drain > 0)
+		free_pcppages_bulk(zone, to_drain, &dst);
 }
 #endif
 
@@ -2960,14 +2971,21 @@ static void drain_pages_zone(unsigned int cpu, struct zone *zone)
 	unsigned long flags;
 	struct per_cpu_pageset *pset;
 	struct per_cpu_pages *pcp;
+	LIST_HEAD(dst);
+	int count;
 
 	local_irq_save(flags);
 	pset = per_cpu_ptr(zone->pageset, cpu);
 
 	pcp = &pset->pcp;
-	if (pcp->count)
-		free_pcppages_bulk(zone, pcp->count, pcp);
+	count = pcp->count;
+	if (count)
+		isolate_pcp_pages(count, pcp, &dst);
+
 	local_irq_restore(flags);
+
+	if (count)
+		free_pcppages_bulk(zone, count, &dst);
 }
 
 /*
@@ -3196,7 +3214,10 @@ static void free_unref_page_commit(struct page *page, unsigned long pfn)
 	pcp->count++;
 	if (pcp->count >= pcp->high) {
 		unsigned long batch = READ_ONCE(pcp->batch);
-		free_pcppages_bulk(zone, batch, pcp);
+		LIST_HEAD(dst);
+
+		isolate_pcp_pages(batch, pcp, &dst);
+		free_pcppages_bulk(zone, batch, &dst);
 	}
 }