mm: accurately calculate zone->managed_pages for highmem zones

Commit "mm: introduce new field 'managed_pages' to struct zone" assumes
that all highmem pages will be freed into the buddy system by function
mem_init().  But that's not always true, some architectures may reserve
some highmem pages during boot.  For example PPC may allocate highmem
pages for giagant HugeTLB pages, and several architectures have code to
check PageReserved flag to exclude highmem pages allocated during boot
when freeing highmem pages into the buddy system.

So treat highmem pages in the same way as normal pages, that is to:
1) reset zone->managed_pages to zero in mem_init().
2) recalculate managed_pages when freeing pages into the buddy system.

Signed-off-by: Jiang Liu <jiang.liu@huawei.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Joonsoo Kim <js1304@gmail.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: <sworddragon2@aol.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Chris Metcalf <cmetcalf@tilera.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Jianguo Wu <wujianguo@huawei.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Tang Chen <tangchen@cn.fujitsu.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Wen Congyang <wency@cn.fujitsu.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

3 files changed, 35 insertions, 28 deletions

diff --git a/mm/bootmem.c b/mm/bootmem.c
index 2b0bcb0..eb792323 100644
--- a/mm/bootmem.c
+++ b/mm/bootmem.c
@@ -241,20 +241,26 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata)
 	return count;
 }
 
-static void reset_node_lowmem_managed_pages(pg_data_t *pgdat)
+static int reset_managed_pages_done __initdata;
+
+static inline void __init reset_node_managed_pages(pg_data_t *pgdat)
 {
 	struct zone *z;
 
-	/*
-	 * In free_area_init_core(), highmem zone's managed_pages is set to
-	 * present_pages, and bootmem allocator doesn't allocate from highmem
-	 * zones. So there's no need to recalculate managed_pages because all
-	 * highmem pages will be managed by the buddy system. Here highmem
-	 * zone also includes highmem movable zone.
-	 */
+	if (reset_managed_pages_done)
+		return;
+
 	for (z = pgdat->node_zones; z < pgdat->node_zones + MAX_NR_ZONES; z++)
-		if (!is_highmem(z))
-			z->managed_pages = 0;
+		z->managed_pages = 0;
+}
+
+void __init reset_all_zones_managed_pages(void)
+{
+	struct pglist_data *pgdat;
+
+	for_each_online_pgdat(pgdat)
+		reset_node_managed_pages(pgdat);
+	reset_managed_pages_done = 1;
 }
 
 /**
@@ -266,7 +272,7 @@ static void reset_node_lowmem_managed_pages(pg_data_t *pgdat)
 unsigned long __init free_all_bootmem_node(pg_data_t *pgdat)
 {
 	register_page_bootmem_info_node(pgdat);
-	reset_node_lowmem_managed_pages(pgdat);
+	reset_node_managed_pages(pgdat);
 	return free_all_bootmem_core(pgdat->bdata);
 }
 
@@ -279,10 +285,8 @@ unsigned long __init free_all_bootmem(void)
 {
 	unsigned long total_pages = 0;
 	bootmem_data_t *bdata;
-	struct pglist_data *pgdat;
 
-	for_each_online_pgdat(pgdat)
-		reset_node_lowmem_managed_pages(pgdat);
+	reset_all_zones_managed_pages();
 
 	list_for_each_entry(bdata, &bdata_list, list)
 		total_pages += free_all_bootmem_core(bdata);
diff --git a/mm/nobootmem.c b/mm/nobootmem.c
index bdd3fa2..0ae8d91 100644
--- a/mm/nobootmem.c
+++ b/mm/nobootmem.c
@@ -137,20 +137,25 @@ static unsigned long __init free_low_memory_core_early(void)
 	return count;
 }
 
-static void reset_node_lowmem_managed_pages(pg_data_t *pgdat)
+static int reset_managed_pages_done __initdata;
+
+static inline void __init reset_node_managed_pages(pg_data_t *pgdat)
 {
 	struct zone *z;
 
-	/*
-	 * In free_area_init_core(), highmem zone's managed_pages is set to
-	 * present_pages, and bootmem allocator doesn't allocate from highmem
-	 * zones. So there's no need to recalculate managed_pages because all
-	 * highmem pages will be managed by the buddy system. Here highmem
-	 * zone also includes highmem movable zone.
-	 */
+	if (reset_managed_pages_done)
+		return;
 	for (z = pgdat->node_zones; z < pgdat->node_zones + MAX_NR_ZONES; z++)
-		if (!is_highmem(z))
-			z->managed_pages = 0;
+		z->managed_pages = 0;
+}
+
+void __init reset_all_zones_managed_pages(void)
+{
+	struct pglist_data *pgdat;
+
+	for_each_online_pgdat(pgdat)
+		reset_node_managed_pages(pgdat);
+	reset_managed_pages_done = 1;
 }
 
 /**
@@ -160,10 +165,7 @@ static void reset_node_lowmem_managed_pages(pg_data_t *pgdat)
  */
 unsigned long __init free_all_bootmem(void)
 {
-	struct pglist_data *pgdat;
-
-	for_each_online_pgdat(pgdat)
-		reset_node_lowmem_managed_pages(pgdat);
+	reset_all_zones_managed_pages();
 
 	/*
 	 * We need to use MAX_NUMNODES instead of NODE_DATA(0)->node_id
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index f22542f..22438eba 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -5232,6 +5232,7 @@ void free_highmem_page(struct page *page)
 {
 	__free_reserved_page(page);
 	totalram_pages++;
+	page_zone(page)->managed_pages++;
 	totalhigh_pages++;
 }
 #endif