mm/hotplug: correctly setup fallback zonelists when creating new pgdat

When hotadd_new_pgdat() is called to create new pgdat for a new node, a
fallback zonelist should be created for the new node.  There's code to try
to achieve that in hotadd_new_pgdat() as below:

	/*
	 * The node we allocated has no zone fallback lists. For avoiding
	 * to access not-initialized zonelist, build here.
	 */
	mutex_lock(&zonelists_mutex);
	build_all_zonelists(pgdat, NULL);
	mutex_unlock(&zonelists_mutex);

But it doesn't work as expected.  When hotadd_new_pgdat() is called, the
new node is still in offline state because node_set_online(nid) hasn't
been called yet.  And build_all_zonelists() only builds zonelists for
online nodes as:

        for_each_online_node(nid) {
                pg_data_t *pgdat = NODE_DATA(nid);

                build_zonelists(pgdat);
                build_zonelist_cache(pgdat);
        }

Though we hope to create zonelist for the new pgdat, but it doesn't.  So
add a new parameter "pgdat" the build_all_zonelists() to build pgdat for
the new pgdat too.

Signed-off-by: Jiang Liu <liuj97@gmail.com>
Signed-off-by: Xishi Qiu <qiuxishi@huawei.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Tony Luck <tony.luck@intel.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Keping Chen <chenkeping@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index f64afa5929fe4af94fa9def985d3bb9bfee694b0..98f079bcf399cab2156ba8f555f319ae4dd0ddb5 100644 (file)
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -721,7 +721,7 @@ typedef struct pglist_data {
 #include <linux/memory_hotplug.h>
 
 extern struct mutex zonelists_mutex;
-void build_all_zonelists(void *data);
+void build_all_zonelists(pg_data_t *pgdat, struct zone *zone);
 void wakeup_kswapd(struct zone *zone, int order, enum zone_type classzone_idx);
 bool zone_watermark_ok(struct zone *z, int order, unsigned long mark,
                int classzone_idx, int alloc_flags);

diff --git a/init/main.c b/init/main.c
index 95316a1b4a76abcc4763422eacdd7f4c646499ab..e60679de61c3a050640554966d34542fcc13a7bd 100644 (file)
--- a/init/main.c
+++ b/init/main.c
@@ -506,7 +506,7 @@ asmlinkage void __init start_kernel(void)
        setup_per_cpu_areas();
        smp_prepare_boot_cpu(); /* arch-specific boot-cpu hooks */
 
-       build_all_zonelists(NULL);
+       build_all_zonelists(NULL, NULL);
        page_alloc_init();
 
        printk(KERN_NOTICE "Kernel command line: %s\n", boot_command_line);

diff --git a/kernel/cpu.c b/kernel/cpu.c
index a4eb5227a19e482eaf2b821c94de845228381bdd..14d32588cccdb3c6d2d64dcd451b5f2efa8e9a19 100644 (file)
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -416,7 +416,7 @@ int __cpuinit cpu_up(unsigned int cpu)
 
        if (pgdat->node_zonelists->_zonerefs->zone == NULL) {
                mutex_lock(&zonelists_mutex);
-               build_all_zonelists(NULL);
+               build_all_zonelists(NULL, NULL);
                mutex_unlock(&zonelists_mutex);
        }
 #endif

diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index 427bb291dd0fdeeb6db93c86c66df6a6f92c9b28..b8731040b9f9292f566f50d9f59b1112f87f104c 100644 (file)
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -513,7 +513,7 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages)
        zone->present_pages += onlined_pages;
        zone->zone_pgdat->node_present_pages += onlined_pages;
        if (need_zonelists_rebuild)
-               build_all_zonelists(zone);
+               build_all_zonelists(NULL, zone);
        else
                zone_pcp_update(zone);
 
@@ -562,7 +562,7 @@ static pg_data_t __ref *hotadd_new_pgdat(int nid, u64 start)
         * to access not-initialized zonelist, build here.
         */
        mutex_lock(&zonelists_mutex);
-       build_all_zonelists(NULL);
+       build_all_zonelists(pgdat, NULL);
        mutex_unlock(&zonelists_mutex);
 
        return pgdat;

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 6c7e3bd93a856d9d1d46ae4ca6062f298be6e43f..9ad6866ac49c0425d815ddb747a04cbece93506c 100644 (file)
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -3032,7 +3032,7 @@ int numa_zonelist_order_handler(ctl_table *table, int write,
                        user_zonelist_order = oldval;
                } else if (oldval != user_zonelist_order) {
                        mutex_lock(&zonelists_mutex);
-                       build_all_zonelists(NULL);
+                       build_all_zonelists(NULL, NULL);
                        mutex_unlock(&zonelists_mutex);
                }
        }
@@ -3415,10 +3415,17 @@ static __init_refok int __build_all_zonelists(void *data)
 {
        int nid;
        int cpu;
+       pg_data_t *self = data;
 
 #ifdef CONFIG_NUMA
        memset(node_load, 0, sizeof(node_load));
 #endif
+
+       if (self && !node_online(self->node_id)) {
+               build_zonelists(self);
+               build_zonelist_cache(self);
+       }
+
        for_each_online_node(nid) {
                pg_data_t *pgdat = NODE_DATA(nid);
 
@@ -3463,7 +3470,7 @@ static __init_refok int __build_all_zonelists(void *data)
  * Called with zonelists_mutex held always
  * unless system_state == SYSTEM_BOOTING.
  */
-void __ref build_all_zonelists(void *data)
+void __ref build_all_zonelists(pg_data_t *pgdat, struct zone *zone)
 {
        set_zonelist_order();
 
@@ -3475,10 +3482,10 @@ void __ref build_all_zonelists(void *data)
                /* we have to stop all cpus to guarantee there is no user
                   of zonelist */
 #ifdef CONFIG_MEMORY_HOTPLUG
-               if (data)
-                       setup_zone_pageset((struct zone *)data);
+               if (zone)
+                       setup_zone_pageset(zone);
 #endif
-               stop_machine(__build_all_zonelists, NULL, NULL);
+               stop_machine(__build_all_zonelists, pgdat, NULL);
                /* cpuset refresh routine should be here */
        }
        vm_total_pages = nr_free_pagecache_pages();