mm/page_alloc: allow high-order pages to be stored on the per-cpu lists

The per-cpu page allocator (PCP) only stores order-0 pages.  This means
that all THP and "cheap" high-order allocations including SLUB contends on
the zone->lock.  This patch extends the PCP allocator to store THP and
"cheap" high-order pages.  Note that struct per_cpu_pages increases in
size to 256 bytes (4 cache lines) on x86-64.

Note that this is not necessarily a universal performance win because of
how it is implemented.  High-order pages can cause pcp->high to be
exceeded prematurely for lower-orders so for example, a large number of
THP pages being freed could release order-0 pages from the PCP lists.
Hence, much depends on the allocation/free pattern as observed by a single
CPU to determine if caching helps or hurts a particular workload.

That said, basic performance testing passed.  The following is a netperf
UDP_STREAM test which hits the relevant patches as some of the network
allocations are high-order.

netperf-udp
                                 5.13.0-rc2             5.13.0-rc2
                           mm-pcpburst-v3r4   mm-pcphighorder-v1r7
Hmean     send-64         261.46 (   0.00%)      266.30 *   1.85%*
Hmean     send-128        516.35 (   0.00%)      536.78 *   3.96%*
Hmean     send-256       1014.13 (   0.00%)     1034.63 *   2.02%*
Hmean     send-1024      3907.65 (   0.00%)     4046.11 *   3.54%*
Hmean     send-2048      7492.93 (   0.00%)     7754.85 *   3.50%*
Hmean     send-3312     11410.04 (   0.00%)    11772.32 *   3.18%*
Hmean     send-4096     13521.95 (   0.00%)    13912.34 *   2.89%*
Hmean     send-8192     21660.50 (   0.00%)    22730.72 *   4.94%*
Hmean     send-16384    31902.32 (   0.00%)    32637.50 *   2.30%*

Functionally, a patch like this is necessary to make bulk allocation of
high-order pages work with similar performance to order-0 bulk
allocations.  The bulk allocator is not updated in this series as it would
have to be determined by bulk allocation users how they want to track the
order of pages allocated with the bulk allocator.

Link: https://lkml.kernel.org/r/20210611135753.GC30378@techsingularity.net
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

1 files changed, 123 insertions, 46 deletions

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 0e441f1677f3..34f097ecfe08 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -674,10 +674,53 @@ out:
 	add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE);
 }
 
+static inline unsigned int order_to_pindex(int migratetype, int order)
+{
+	int base = order;
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	if (order > PAGE_ALLOC_COSTLY_ORDER) {
+		VM_BUG_ON(order != pageblock_order);
+		base = PAGE_ALLOC_COSTLY_ORDER + 1;
+	}
+#else
+	VM_BUG_ON(order > PAGE_ALLOC_COSTLY_ORDER);
+#endif
+
+	return (MIGRATE_PCPTYPES * base) + migratetype;
+}
+
+static inline int pindex_to_order(unsigned int pindex)
+{
+	int order = pindex / MIGRATE_PCPTYPES;
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	if (order > PAGE_ALLOC_COSTLY_ORDER) {
+		order = pageblock_order;
+		VM_BUG_ON(order != pageblock_order);
+	}
+#else
+	VM_BUG_ON(order > PAGE_ALLOC_COSTLY_ORDER);
+#endif
+
+	return order;
+}
+
+static inline bool pcp_allowed_order(unsigned int order)
+{
+	if (order <= PAGE_ALLOC_COSTLY_ORDER)
+		return true;
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	if (order == pageblock_order)
+		return true;
+#endif
+	return false;
+}
+
 static inline void free_the_page(struct page *page, unsigned int order)
 {
-	if (order == 0)		/* Via pcp? */
-		free_unref_page(page);
+	if (pcp_allowed_order(order))		/* Via pcp? */
+		free_unref_page(page, order);
 	else
 		__free_pages_ok(page, order, FPI_NONE);
 }
@@ -700,7 +743,7 @@ static inline void free_the_page(struct page *page, unsigned int order)
 void free_compound_page(struct page *page)
 {
 	mem_cgroup_uncharge(page);
-	__free_pages_ok(page, compound_order(page), FPI_NONE);
+	free_the_page(page, compound_order(page));
 }
 
 void prep_compound_page(struct page *page, unsigned int order)
@@ -1350,9 +1393,9 @@ static __always_inline bool free_pages_prepare(struct page *page,
  * to pcp lists. With debug_pagealloc also enabled, they are also rechecked when
  * moved from pcp lists to free lists.
  */
-static bool free_pcp_prepare(struct page *page)
+static bool free_pcp_prepare(struct page *page, unsigned int order)
 {
-	return free_pages_prepare(page, 0, true, FPI_NONE);
+	return free_pages_prepare(page, order, true, FPI_NONE);
 }
 
 static bool bulkfree_pcp_prepare(struct page *page)
@@ -1369,12 +1412,12 @@ static bool bulkfree_pcp_prepare(struct page *page)
  * debug_pagealloc enabled, they are checked also immediately when being freed
  * to the pcp lists.
  */
-static bool free_pcp_prepare(struct page *page)
+static bool free_pcp_prepare(struct page *page, unsigned int order)
 {
 	if (debug_pagealloc_enabled_static())
-		return free_pages_prepare(page, 0, true, FPI_NONE);
+		return free_pages_prepare(page, order, true, FPI_NONE);
 	else
-		return free_pages_prepare(page, 0, false, FPI_NONE);
+		return free_pages_prepare(page, order, false, FPI_NONE);
 }
 
 static bool bulkfree_pcp_prepare(struct page *page)
@@ -1406,8 +1449,10 @@ static inline void prefetch_buddy(struct page *page)
 static void free_pcppages_bulk(struct zone *zone, int count,
 					struct per_cpu_pages *pcp)
 {
-	int migratetype = 0;
+	int pindex = 0;
 	int batch_free = 0;
+	int nr_freed = 0;
+	unsigned int order;
 	int prefetch_nr = READ_ONCE(pcp->batch);
 	bool isolated_pageblocks;
 	struct page *page, *tmp;
@@ -1418,7 +1463,7 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 	 * below while (list_empty(list)) loop.
 	 */
 	count = min(pcp->count, count);
-	while (count) {
+	while (count > 0) {
 		struct list_head *list;
 
 		/*
@@ -1430,24 +1475,31 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 		 */
 		do {
 			batch_free++;
-			if (++migratetype == MIGRATE_PCPTYPES)
-				migratetype = 0;
-			list = &pcp->lists[migratetype];
+			if (++pindex == NR_PCP_LISTS)
+				pindex = 0;
+			list = &pcp->lists[pindex];
 		} while (list_empty(list));
 
 		/* This is the only non-empty list. Free them all. */
-		if (batch_free == MIGRATE_PCPTYPES)
+		if (batch_free == NR_PCP_LISTS)
 			batch_free = count;
 
+		order = pindex_to_order(pindex);
+		BUILD_BUG_ON(MAX_ORDER >= (1<<NR_PCP_ORDER_WIDTH));
 		do {
 			page = list_last_entry(list, struct page, lru);
 			/* must delete to avoid corrupting pcp list */
 			list_del(&page->lru);
-			pcp->count--;
+			nr_freed += 1 << order;
+			count -= 1 << order;
 
 			if (bulkfree_pcp_prepare(page))
 				continue;
 
+			/* Encode order with the migratetype */
+			page->index <<= NR_PCP_ORDER_WIDTH;
+			page->index |= order;
+
 			list_add_tail(&page->lru, &head);
 
 			/*
@@ -1463,8 +1515,9 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 				prefetch_buddy(page);
 				prefetch_nr--;
 			}
-		} while (--count && --batch_free && !list_empty(list));
+		} while (count > 0 && --batch_free && !list_empty(list));
 	}
+	pcp->count -= nr_freed;
 
 	/*
 	 * local_lock_irq held so equivalent to spin_lock_irqsave for
@@ -1479,14 +1532,19 @@ static void free_pcppages_bulk(struct zone *zone, int count,
 	 */
 	list_for_each_entry_safe(page, tmp, &head, lru) {
 		int mt = get_pcppage_migratetype(page);
+
+		/* mt has been encoded with the order (see above) */
+		order = mt & NR_PCP_ORDER_MASK;
+		mt >>= NR_PCP_ORDER_WIDTH;
+
 		/* 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);
+		__free_one_page(page, page_to_pfn(page), zone, order, mt, FPI_NONE);
+		trace_mm_page_pcpu_drain(page, order, mt);
 	}
 	spin_unlock(&zone->lock);
 }
@@ -3263,11 +3321,12 @@ void mark_free_pages(struct zone *zone)
 }
 #endif /* CONFIG_PM */
 
-static bool free_unref_page_prepare(struct page *page, unsigned long pfn)
+static bool free_unref_page_prepare(struct page *page, unsigned long pfn,
+							unsigned int order)
 {
 	int migratetype;
 
-	if (!free_pcp_prepare(page))
+	if (!free_pcp_prepare(page, order))
 		return false;
 
 	migratetype = get_pfnblock_migratetype(page, pfn);
@@ -3317,16 +3376,18 @@ static int nr_pcp_high(struct per_cpu_pages *pcp, struct zone *zone)
 }
 
 static void free_unref_page_commit(struct page *page, unsigned long pfn,
-				   int migratetype)
+				   int migratetype, unsigned int order)
 {
 	struct zone *zone = page_zone(page);
 	struct per_cpu_pages *pcp;
 	int high;
+	int pindex;
 
 	__count_vm_event(PGFREE);
 	pcp = this_cpu_ptr(zone->per_cpu_pageset);
-	list_add(&page->lru, &pcp->lists[migratetype]);
-	pcp->count++;
+	pindex = order_to_pindex(migratetype, order);
+	list_add(&page->lru, &pcp->lists[pindex]);
+	pcp->count += 1 << order;
 	high = nr_pcp_high(pcp, zone);
 	if (pcp->count >= high) {
 		int batch = READ_ONCE(pcp->batch);
@@ -3336,15 +3397,15 @@ static void free_unref_page_commit(struct page *page, unsigned long pfn,
 }
 
 /*
- * Free a 0-order page
+ * Free a pcp page
  */
-void free_unref_page(struct page *page)
+void free_unref_page(struct page *page, unsigned int order)
 {
 	unsigned long flags;
 	unsigned long pfn = page_to_pfn(page);
 	int migratetype;
 
-	if (!free_unref_page_prepare(page, pfn))
+	if (!free_unref_page_prepare(page, pfn, order))
 		return;
 
 	/*
@@ -3357,14 +3418,14 @@ void free_unref_page(struct page *page)
 	migratetype = get_pcppage_migratetype(page);
 	if (unlikely(migratetype >= MIGRATE_PCPTYPES)) {
 		if (unlikely(is_migrate_isolate(migratetype))) {
-			free_one_page(page_zone(page), page, pfn, 0, migratetype, FPI_NONE);
+			free_one_page(page_zone(page), page, pfn, order, migratetype, FPI_NONE);
 			return;
 		}
 		migratetype = MIGRATE_MOVABLE;
 	}
 
 	local_lock_irqsave(&pagesets.lock, flags);
-	free_unref_page_commit(page, pfn, migratetype);
+	free_unref_page_commit(page, pfn, migratetype, order);
 	local_unlock_irqrestore(&pagesets.lock, flags);
 }
 
@@ -3381,7 +3442,7 @@ void free_unref_page_list(struct list_head *list)
 	/* Prepare pages for freeing */
 	list_for_each_entry_safe(page, next, list, lru) {
 		pfn = page_to_pfn(page);
-		if (!free_unref_page_prepare(page, pfn))
+		if (!free_unref_page_prepare(page, pfn, 0))
 			list_del(&page->lru);
 
 		/*
@@ -3413,7 +3474,7 @@ void free_unref_page_list(struct list_head *list)
 		set_page_private(page, 0);
 		migratetype = get_pcppage_migratetype(page);
 		trace_mm_page_free_batched(page);
-		free_unref_page_commit(page, pfn, migratetype);
+		free_unref_page_commit(page, pfn, migratetype, 0);
 
 		/*
 		 * Guard against excessive IRQ disabled times when we get
@@ -3549,7 +3610,8 @@ static inline void zone_statistics(struct zone *preferred_zone, struct zone *z,
 
 /* Remove page from the per-cpu list, caller must protect the list */
 static inline
-struct page *__rmqueue_pcplist(struct zone *zone, int migratetype,
+struct page *__rmqueue_pcplist(struct zone *zone, unsigned int order,
+			int migratetype,
 			unsigned int alloc_flags,
 			struct per_cpu_pages *pcp,
 			struct list_head *list)
@@ -3558,16 +3620,30 @@ struct page *__rmqueue_pcplist(struct zone *zone, int migratetype,
 
 	do {
 		if (list_empty(list)) {
-			pcp->count += rmqueue_bulk(zone, 0,
-					READ_ONCE(pcp->batch), list,
+			int batch = READ_ONCE(pcp->batch);
+			int alloced;
+
+			/*
+			 * Scale batch relative to order if batch implies
+			 * free pages can be stored on the PCP. Batch can
+			 * be 1 for small zones or for boot pagesets which
+			 * should never store free pages as the pages may
+			 * belong to arbitrary zones.
+			 */
+			if (batch > 1)
+				batch = max(batch >> order, 2);
+			alloced = rmqueue_bulk(zone, order,
+					batch, list,
 					migratetype, alloc_flags);
+
+			pcp->count += alloced << order;
 			if (unlikely(list_empty(list)))
 				return NULL;
 		}
 
 		page = list_first_entry(list, struct page, lru);
 		list_del(&page->lru);
-		pcp->count--;
+		pcp->count -= 1 << order;
 	} while (check_new_pcp(page));
 
 	return page;
@@ -3575,8 +3651,9 @@ struct page *__rmqueue_pcplist(struct zone *zone, int migratetype,
 
 /* Lock and remove page from the per-cpu list */
 static struct page *rmqueue_pcplist(struct zone *preferred_zone,
-			struct zone *zone, gfp_t gfp_flags,
-			int migratetype, unsigned int alloc_flags)
+			struct zone *zone, unsigned int order,
+			gfp_t gfp_flags, int migratetype,
+			unsigned int alloc_flags)
 {
 	struct per_cpu_pages *pcp;
 	struct list_head *list;
@@ -3592,8 +3669,8 @@ static struct page *rmqueue_pcplist(struct zone *preferred_zone,
 	 */
 	pcp = this_cpu_ptr(zone->per_cpu_pageset);
 	pcp->free_factor >>= 1;
-	list = &pcp->lists[migratetype];
-	page = __rmqueue_pcplist(zone,  migratetype, alloc_flags, pcp, list);
+	list = &pcp->lists[order_to_pindex(migratetype, order)];
+	page = __rmqueue_pcplist(zone, order, migratetype, alloc_flags, pcp, list);
 	local_unlock_irqrestore(&pagesets.lock, flags);
 	if (page) {
 		__count_zid_vm_events(PGALLOC, page_zonenum(page), 1);
@@ -3614,15 +3691,15 @@ struct page *rmqueue(struct zone *preferred_zone,
 	unsigned long flags;
 	struct page *page;
 
-	if (likely(order == 0)) {
+	if (likely(pcp_allowed_order(order))) {
 		/*
 		 * MIGRATE_MOVABLE pcplist could have the pages on CMA area and
 		 * we need to skip it when CMA area isn't allowed.
 		 */
 		if (!IS_ENABLED(CONFIG_CMA) || alloc_flags & ALLOC_CMA ||
 				migratetype != MIGRATE_MOVABLE) {
-			page = rmqueue_pcplist(preferred_zone, zone, gfp_flags,
-					migratetype, alloc_flags);
+			page = rmqueue_pcplist(preferred_zone, zone, order,
+					gfp_flags, migratetype, alloc_flags);
 			goto out;
 		}
 	}
@@ -5201,7 +5278,7 @@ unsigned long __alloc_pages_bulk(gfp_t gfp, int preferred_nid,
 	/* Attempt the batch allocation */
 	local_lock_irqsave(&pagesets.lock, flags);
 	pcp = this_cpu_ptr(zone->per_cpu_pageset);
-	pcp_list = &pcp->lists[ac.migratetype];
+	pcp_list = &pcp->lists[order_to_pindex(ac.migratetype, 0)];
 
 	while (nr_populated < nr_pages) {
 
@@ -5211,7 +5288,7 @@ unsigned long __alloc_pages_bulk(gfp_t gfp, int preferred_nid,
 			continue;
 		}
 
-		page = __rmqueue_pcplist(zone, ac.migratetype, alloc_flags,
+		page = __rmqueue_pcplist(zone, 0, ac.migratetype, alloc_flags,
 								pcp, pcp_list);
 		if (unlikely(!page)) {
 			/* Try and get at least one page */
@@ -6778,13 +6855,13 @@ static void pageset_update(struct per_cpu_pages *pcp, unsigned long high,
 
 static void per_cpu_pages_init(struct per_cpu_pages *pcp, struct per_cpu_zonestat *pzstats)
 {
-	int migratetype;
+	int pindex;
 
 	memset(pcp, 0, sizeof(*pcp));
 	memset(pzstats, 0, sizeof(*pzstats));
 
-	for (migratetype = 0; migratetype < MIGRATE_PCPTYPES; migratetype++)
-		INIT_LIST_HEAD(&pcp->lists[migratetype]);
+	for (pindex = 0; pindex < NR_PCP_LISTS; pindex++)
+		INIT_LIST_HEAD(&pcp->lists[pindex]);
 
 	/*
 	 * Set batch and high values safe for a boot pageset. A true percpu