Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
[linux-2.6.git] / mm / memory_hotplug.c
index 0fb3302..c46887b 100644 (file)
 #include <linux/highmem.h>
 #include <linux/vmalloc.h>
 #include <linux/ioport.h>
-#include <linux/cpuset.h>
 #include <linux/delay.h>
 #include <linux/migrate.h>
 #include <linux/page-isolation.h>
+#include <linux/pfn.h>
+#include <linux/suspend.h>
+#include <linux/mm_inline.h>
+#include <linux/firmware-map.h>
 
 #include <asm/tlbflush.h>
 
+#include "internal.h"
+
+DEFINE_MUTEX(mem_hotplug_mutex);
+
+void lock_memory_hotplug(void)
+{
+       mutex_lock(&mem_hotplug_mutex);
+
+       /* for exclusive hibernation if CONFIG_HIBERNATION=y */
+       lock_system_sleep();
+}
+
+void unlock_memory_hotplug(void)
+{
+       unlock_system_sleep();
+       mutex_unlock(&mem_hotplug_mutex);
+}
+
+
 /* add this memory to iomem resource */
 static struct resource *register_memory_resource(u64 start, u64 size)
 {
@@ -58,29 +80,168 @@ static void release_memory_resource(struct resource *res)
        return;
 }
 
-
 #ifdef CONFIG_MEMORY_HOTPLUG_SPARSE
-static int __add_zone(struct zone *zone, unsigned long phys_start_pfn)
+#ifndef CONFIG_SPARSEMEM_VMEMMAP
+static void get_page_bootmem(unsigned long info,  struct page *page,
+                            unsigned long type)
+{
+       page->lru.next = (struct list_head *) type;
+       SetPagePrivate(page);
+       set_page_private(page, info);
+       atomic_inc(&page->_count);
+}
+
+/* reference to __meminit __free_pages_bootmem is valid
+ * so use __ref to tell modpost not to generate a warning */
+void __ref put_page_bootmem(struct page *page)
+{
+       unsigned long type;
+
+       type = (unsigned long) page->lru.next;
+       BUG_ON(type < MEMORY_HOTPLUG_MIN_BOOTMEM_TYPE ||
+              type > MEMORY_HOTPLUG_MAX_BOOTMEM_TYPE);
+
+       if (atomic_dec_return(&page->_count) == 1) {
+               ClearPagePrivate(page);
+               set_page_private(page, 0);
+               INIT_LIST_HEAD(&page->lru);
+               __free_pages_bootmem(page, 0);
+       }
+
+}
+
+static void register_page_bootmem_info_section(unsigned long start_pfn)
+{
+       unsigned long *usemap, mapsize, section_nr, i;
+       struct mem_section *ms;
+       struct page *page, *memmap;
+
+       if (!pfn_valid(start_pfn))
+               return;
+
+       section_nr = pfn_to_section_nr(start_pfn);
+       ms = __nr_to_section(section_nr);
+
+       /* Get section's memmap address */
+       memmap = sparse_decode_mem_map(ms->section_mem_map, section_nr);
+
+       /*
+        * Get page for the memmap's phys address
+        * XXX: need more consideration for sparse_vmemmap...
+        */
+       page = virt_to_page(memmap);
+       mapsize = sizeof(struct page) * PAGES_PER_SECTION;
+       mapsize = PAGE_ALIGN(mapsize) >> PAGE_SHIFT;
+
+       /* remember memmap's page */
+       for (i = 0; i < mapsize; i++, page++)
+               get_page_bootmem(section_nr, page, SECTION_INFO);
+
+       usemap = __nr_to_section(section_nr)->pageblock_flags;
+       page = virt_to_page(usemap);
+
+       mapsize = PAGE_ALIGN(usemap_size()) >> PAGE_SHIFT;
+
+       for (i = 0; i < mapsize; i++, page++)
+               get_page_bootmem(section_nr, page, MIX_SECTION_INFO);
+
+}
+
+void register_page_bootmem_info_node(struct pglist_data *pgdat)
+{
+       unsigned long i, pfn, end_pfn, nr_pages;
+       int node = pgdat->node_id;
+       struct page *page;
+       struct zone *zone;
+
+       nr_pages = PAGE_ALIGN(sizeof(struct pglist_data)) >> PAGE_SHIFT;
+       page = virt_to_page(pgdat);
+
+       for (i = 0; i < nr_pages; i++, page++)
+               get_page_bootmem(node, page, NODE_INFO);
+
+       zone = &pgdat->node_zones[0];
+       for (; zone < pgdat->node_zones + MAX_NR_ZONES - 1; zone++) {
+               if (zone->wait_table) {
+                       nr_pages = zone->wait_table_hash_nr_entries
+                               * sizeof(wait_queue_head_t);
+                       nr_pages = PAGE_ALIGN(nr_pages) >> PAGE_SHIFT;
+                       page = virt_to_page(zone->wait_table);
+
+                       for (i = 0; i < nr_pages; i++, page++)
+                               get_page_bootmem(node, page, NODE_INFO);
+               }
+       }
+
+       pfn = pgdat->node_start_pfn;
+       end_pfn = pfn + pgdat->node_spanned_pages;
+
+       /* register_section info */
+       for (; pfn < end_pfn; pfn += PAGES_PER_SECTION)
+               register_page_bootmem_info_section(pfn);
+
+}
+#endif /* !CONFIG_SPARSEMEM_VMEMMAP */
+
+static void grow_zone_span(struct zone *zone, unsigned long start_pfn,
+                          unsigned long end_pfn)
+{
+       unsigned long old_zone_end_pfn;
+
+       zone_span_writelock(zone);
+
+       old_zone_end_pfn = zone->zone_start_pfn + zone->spanned_pages;
+       if (start_pfn < zone->zone_start_pfn)
+               zone->zone_start_pfn = start_pfn;
+
+       zone->spanned_pages = max(old_zone_end_pfn, end_pfn) -
+                               zone->zone_start_pfn;
+
+       zone_span_writeunlock(zone);
+}
+
+static void grow_pgdat_span(struct pglist_data *pgdat, unsigned long start_pfn,
+                           unsigned long end_pfn)
+{
+       unsigned long old_pgdat_end_pfn =
+               pgdat->node_start_pfn + pgdat->node_spanned_pages;
+
+       if (start_pfn < pgdat->node_start_pfn)
+               pgdat->node_start_pfn = start_pfn;
+
+       pgdat->node_spanned_pages = max(old_pgdat_end_pfn, end_pfn) -
+                                       pgdat->node_start_pfn;
+}
+
+static int __meminit __add_zone(struct zone *zone, unsigned long phys_start_pfn)
 {
        struct pglist_data *pgdat = zone->zone_pgdat;
        int nr_pages = PAGES_PER_SECTION;
        int nid = pgdat->node_id;
        int zone_type;
+       unsigned long flags;
 
        zone_type = zone - pgdat->node_zones;
        if (!zone->wait_table) {
-               int ret = 0;
+               int ret;
+
                ret = init_currently_empty_zone(zone, phys_start_pfn,
                                                nr_pages, MEMMAP_HOTPLUG);
-               if (ret < 0)
+               if (ret)
                        return ret;
        }
+       pgdat_resize_lock(zone->zone_pgdat, &flags);
+       grow_zone_span(zone, phys_start_pfn, phys_start_pfn + nr_pages);
+       grow_pgdat_span(zone->zone_pgdat, phys_start_pfn,
+                       phys_start_pfn + nr_pages);
+       pgdat_resize_unlock(zone->zone_pgdat, &flags);
        memmap_init_zone(nr_pages, nid, zone_type,
                         phys_start_pfn, MEMMAP_HOTPLUG);
        return 0;
 }
 
-static int __add_section(struct zone *zone, unsigned long phys_start_pfn)
+static int __meminit __add_section(int nid, struct zone *zone,
+                                       unsigned long phys_start_pfn)
 {
        int nr_pages = PAGES_PER_SECTION;
        int ret;
@@ -98,8 +259,38 @@ static int __add_section(struct zone *zone, unsigned long phys_start_pfn)
        if (ret < 0)
                return ret;
 
-       return register_new_memory(__pfn_to_section(phys_start_pfn));
+       return register_new_memory(nid, __pfn_to_section(phys_start_pfn));
+}
+
+#ifdef CONFIG_SPARSEMEM_VMEMMAP
+static int __remove_section(struct zone *zone, struct mem_section *ms)
+{
+       /*
+        * XXX: Freeing memmap with vmemmap is not implement yet.
+        *      This should be removed later.
+        */
+       return -EBUSY;
+}
+#else
+static int __remove_section(struct zone *zone, struct mem_section *ms)
+{
+       unsigned long flags;
+       struct pglist_data *pgdat = zone->zone_pgdat;
+       int ret = -EINVAL;
+
+       if (!valid_section(ms))
+               return ret;
+
+       ret = unregister_memory_section(ms);
+       if (ret)
+               return ret;
+
+       pgdat_resize_lock(pgdat, &flags);
+       sparse_remove_one_section(zone, ms);
+       pgdat_resize_unlock(pgdat, &flags);
+       return 0;
 }
+#endif
 
 /*
  * Reasonably generic function for adding memory.  It is
@@ -107,8 +298,8 @@ static int __add_section(struct zone *zone, unsigned long phys_start_pfn)
  * call this function after deciding the zone to which to
  * add the new pages.
  */
-int __add_pages(struct zone *zone, unsigned long phys_start_pfn,
-                unsigned long nr_pages)
+int __ref __add_pages(int nid, struct zone *zone, unsigned long phys_start_pfn,
+                       unsigned long nr_pages)
 {
        unsigned long i;
        int err = 0;
@@ -118,7 +309,7 @@ int __add_pages(struct zone *zone, unsigned long phys_start_pfn,
        end_sec = pfn_to_section_nr(phys_start_pfn + nr_pages - 1);
 
        for (i = start_sec; i <= end_sec; i++) {
-               err = __add_section(zone, i << PFN_SECTION_SHIFT);
+               err = __add_section(nid, zone, i << PFN_SECTION_SHIFT);
 
                /*
                 * EEXIST is finally dealt with by ioresource collision
@@ -134,34 +325,58 @@ int __add_pages(struct zone *zone, unsigned long phys_start_pfn,
 }
 EXPORT_SYMBOL_GPL(__add_pages);
 
-static void grow_zone_span(struct zone *zone,
-               unsigned long start_pfn, unsigned long end_pfn)
+/**
+ * __remove_pages() - remove sections of pages from a zone
+ * @zone: zone from which pages need to be removed
+ * @phys_start_pfn: starting pageframe (must be aligned to start of a section)
+ * @nr_pages: number of pages to remove (must be multiple of section size)
+ *
+ * Generic helper function to remove section mappings and sysfs entries
+ * for the section of the memory we are removing. Caller needs to make
+ * sure that pages are marked reserved and zones are adjust properly by
+ * calling offline_pages().
+ */
+int __remove_pages(struct zone *zone, unsigned long phys_start_pfn,
+                unsigned long nr_pages)
 {
-       unsigned long old_zone_end_pfn;
-
-       zone_span_writelock(zone);
+       unsigned long i, ret = 0;
+       int sections_to_remove;
 
-       old_zone_end_pfn = zone->zone_start_pfn + zone->spanned_pages;
-       if (start_pfn < zone->zone_start_pfn)
-               zone->zone_start_pfn = start_pfn;
-
-       zone->spanned_pages = max(old_zone_end_pfn, end_pfn) -
-                               zone->zone_start_pfn;
-
-       zone_span_writeunlock(zone);
+       /*
+        * We can only remove entire sections
+        */
+       BUG_ON(phys_start_pfn & ~PAGE_SECTION_MASK);
+       BUG_ON(nr_pages % PAGES_PER_SECTION);
+
+       sections_to_remove = nr_pages / PAGES_PER_SECTION;
+       for (i = 0; i < sections_to_remove; i++) {
+               unsigned long pfn = phys_start_pfn + i*PAGES_PER_SECTION;
+               release_mem_region(pfn << PAGE_SHIFT,
+                                  PAGES_PER_SECTION << PAGE_SHIFT);
+               ret = __remove_section(zone, __pfn_to_section(pfn));
+               if (ret)
+                       break;
+       }
+       return ret;
 }
+EXPORT_SYMBOL_GPL(__remove_pages);
 
-static void grow_pgdat_span(struct pglist_data *pgdat,
-               unsigned long start_pfn, unsigned long end_pfn)
+void online_page(struct page *page)
 {
-       unsigned long old_pgdat_end_pfn =
-               pgdat->node_start_pfn + pgdat->node_spanned_pages;
+       unsigned long pfn = page_to_pfn(page);
 
-       if (start_pfn < pgdat->node_start_pfn)
-               pgdat->node_start_pfn = start_pfn;
+       totalram_pages++;
+       if (pfn >= num_physpages)
+               num_physpages = pfn + 1;
 
-       pgdat->node_spanned_pages = max(old_pgdat_end_pfn, end_pfn) -
-                                       pgdat->node_start_pfn;
+#ifdef CONFIG_HIGHMEM
+       if (PageHighMem(page))
+               totalhigh_pages++;
+#endif
+
+       ClearPageReserved(page);
+       init_page_count(page);
+       __free_page(page);
 }
 
 static int online_pages_range(unsigned long start_pfn, unsigned long nr_pages,
@@ -181,9 +396,8 @@ static int online_pages_range(unsigned long start_pfn, unsigned long nr_pages,
 }
 
 
-int online_pages(unsigned long pfn, unsigned long nr_pages)
+int __ref online_pages(unsigned long pfn, unsigned long nr_pages)
 {
-       unsigned long flags;
        unsigned long onlined_pages = 0;
        struct zone *zone;
        int need_zonelists_rebuild = 0;
@@ -191,6 +405,7 @@ int online_pages(unsigned long pfn, unsigned long nr_pages)
        int ret;
        struct memory_notify arg;
 
+       lock_memory_hotplug();
        arg.start_pfn = pfn;
        arg.nr_pages = nr_pages;
        arg.status_change_nid = -1;
@@ -203,6 +418,7 @@ int online_pages(unsigned long pfn, unsigned long nr_pages)
        ret = notifier_to_errno(ret);
        if (ret) {
                memory_notify(MEM_CANCEL_ONLINE, &arg);
+               unlock_memory_hotplug();
                return ret;
        }
        /*
@@ -211,43 +427,56 @@ int online_pages(unsigned long pfn, unsigned long nr_pages)
         * memory_block->state_mutex.
         */
        zone = page_zone(pfn_to_page(pfn));
-       pgdat_resize_lock(zone->zone_pgdat, &flags);
-       grow_zone_span(zone, pfn, pfn + nr_pages);
-       grow_pgdat_span(zone->zone_pgdat, pfn, pfn + nr_pages);
-       pgdat_resize_unlock(zone->zone_pgdat, &flags);
-
        /*
         * If this zone is not populated, then it is not in zonelist.
         * This means the page allocator ignores this zone.
         * So, zonelist must be updated after online.
         */
+       mutex_lock(&zonelists_mutex);
        if (!populated_zone(zone))
                need_zonelists_rebuild = 1;
 
-       walk_memory_resource(pfn, nr_pages, &onlined_pages,
+       ret = walk_system_ram_range(pfn, nr_pages, &onlined_pages,
                online_pages_range);
+       if (ret) {
+               mutex_unlock(&zonelists_mutex);
+               printk(KERN_DEBUG "online_pages %lx at %lx failed\n",
+                       nr_pages, pfn);
+               memory_notify(MEM_CANCEL_ONLINE, &arg);
+               unlock_memory_hotplug();
+               return ret;
+       }
+
        zone->present_pages += onlined_pages;
        zone->zone_pgdat->node_present_pages += onlined_pages;
+       if (need_zonelists_rebuild)
+               build_all_zonelists(zone);
+       else
+               zone_pcp_update(zone);
+
+       mutex_unlock(&zonelists_mutex);
+
+       init_per_zone_wmark_min();
 
-       setup_per_zone_pages_min();
        if (onlined_pages) {
                kswapd_run(zone_to_nid(zone));
                node_set_state(zone_to_nid(zone), N_HIGH_MEMORY);
        }
 
-       if (need_zonelists_rebuild)
-               build_all_zonelists();
        vm_total_pages = nr_free_pagecache_pages();
+
        writeback_set_ratelimit();
 
        if (onlined_pages)
                memory_notify(MEM_ONLINE, &arg);
+       unlock_memory_hotplug();
 
        return 0;
 }
 #endif /* CONFIG_MEMORY_HOTPLUG_SPARSE */
 
-static pg_data_t *hotadd_new_pgdat(int nid, u64 start)
+/* we are OK calling __meminit stuff here - we have CONFIG_MEMORY_HOTPLUG */
+static pg_data_t __ref *hotadd_new_pgdat(int nid, u64 start)
 {
        struct pglist_data *pgdat;
        unsigned long zones_size[MAX_NR_ZONES] = {0};
@@ -263,7 +492,15 @@ static pg_data_t *hotadd_new_pgdat(int nid, u64 start)
        /* we can use NODE_DATA(nid) from here */
 
        /* init node's zones as empty zones, we don't have any present pages.*/
-       free_area_init_node(nid, pgdat, zones_size, start_pfn, zholes_size);
+       free_area_init_node(nid, zones_size, start_pfn, zholes_size);
+
+       /*
+        * 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(NULL);
+       mutex_unlock(&zonelists_mutex);
 
        return pgdat;
 }
@@ -276,21 +513,49 @@ static void rollback_node_hotadd(int nid, pg_data_t *pgdat)
 }
 
 
-int add_memory(int nid, u64 start, u64 size)
+/*
+ * called by cpu_up() to online a node without onlined memory.
+ */
+int mem_online_node(int nid)
+{
+       pg_data_t       *pgdat;
+       int     ret;
+
+       lock_memory_hotplug();
+       pgdat = hotadd_new_pgdat(nid, 0);
+       if (!pgdat) {
+               ret = -ENOMEM;
+               goto out;
+       }
+       node_set_online(nid);
+       ret = register_one_node(nid);
+       BUG_ON(ret);
+
+out:
+       unlock_memory_hotplug();
+       return ret;
+}
+
+/* we are OK calling __meminit stuff here - we have CONFIG_MEMORY_HOTPLUG */
+int __ref add_memory(int nid, u64 start, u64 size)
 {
        pg_data_t *pgdat = NULL;
        int new_pgdat = 0;
        struct resource *res;
        int ret;
 
+       lock_memory_hotplug();
+
        res = register_memory_resource(start, size);
+       ret = -EEXIST;
        if (!res)
-               return -EEXIST;
+               goto out;
 
        if (!node_online(nid)) {
                pgdat = hotadd_new_pgdat(nid, start);
+               ret = -ENOMEM;
                if (!pgdat)
-                       return -ENOMEM;
+                       goto out;
                new_pgdat = 1;
        }
 
@@ -303,8 +568,6 @@ int add_memory(int nid, u64 start, u64 size)
        /* we online node here. we can't roll back from here. */
        node_set_online(nid);
 
-       cpuset_track_online_nodes();
-
        if (new_pgdat) {
                ret = register_one_node(nid);
                /*
@@ -315,7 +578,11 @@ int add_memory(int nid, u64 start, u64 size)
                BUG_ON(ret);
        }
 
-       return ret;
+       /* create new memmap entry */
+       firmware_map_add_hotplug(start, start + size, "System RAM");
+
+       goto out;
+
 error:
        /* rollback pgdat allocation and others */
        if (new_pgdat)
@@ -323,12 +590,61 @@ error:
        if (res)
                release_memory_resource(res);
 
+out:
+       unlock_memory_hotplug();
        return ret;
 }
 EXPORT_SYMBOL_GPL(add_memory);
 
 #ifdef CONFIG_MEMORY_HOTREMOVE
 /*
+ * A free page on the buddy free lists (not the per-cpu lists) has PageBuddy
+ * set and the size of the free page is given by page_order(). Using this,
+ * the function determines if the pageblock contains only free pages.
+ * Due to buddy contraints, a free page at least the size of a pageblock will
+ * be located at the start of the pageblock
+ */
+static inline int pageblock_free(struct page *page)
+{
+       return PageBuddy(page) && page_order(page) >= pageblock_order;
+}
+
+/* Return the start of the next active pageblock after a given page */
+static struct page *next_active_pageblock(struct page *page)
+{
+       /* Ensure the starting page is pageblock-aligned */
+       BUG_ON(page_to_pfn(page) & (pageblock_nr_pages - 1));
+
+       /* If the entire pageblock is free, move to the end of free page */
+       if (pageblock_free(page)) {
+               int order;
+               /* be careful. we don't have locks, page_order can be changed.*/
+               order = page_order(page);
+               if ((order < MAX_ORDER) && (order >= pageblock_order))
+                       return page + (1 << order);
+       }
+
+       return page + pageblock_nr_pages;
+}
+
+/* Checks if this range of memory is likely to be hot-removable. */
+int is_mem_section_removable(unsigned long start_pfn, unsigned long nr_pages)
+{
+       struct page *page = pfn_to_page(start_pfn);
+       struct page *end_page = page + nr_pages;
+
+       /* Check the starting page of each pageblock within the range */
+       for (; page < end_page; page = next_active_pageblock(page)) {
+               if (!is_pageblock_removable_nolock(page))
+                       return 0;
+               cond_resched();
+       }
+
+       /* All pageblocks in the memory block are likely to be hot-removable */
+       return 1;
+}
+
+/*
  * Confirm all pages in a range [start, end) is belongs to the same zone.
  */
 static int test_pages_in_a_zone(unsigned long start_pfn, unsigned long end_pfn)
@@ -358,7 +674,7 @@ static int test_pages_in_a_zone(unsigned long start_pfn, unsigned long end_pfn)
  * Scanning pfn is much easier than scanning lru list.
  * Scan pfn from start to end and Find LRU page.
  */
-int scan_lru_pages(unsigned long start, unsigned long end)
+static unsigned long scan_lru_pages(unsigned long start, unsigned long end)
 {
        unsigned long pfn;
        struct page *page;
@@ -373,15 +689,12 @@ int scan_lru_pages(unsigned long start, unsigned long end)
 }
 
 static struct page *
-hotremove_migrate_alloc(struct page *page,
-                       unsigned long private,
-                       int **x)
+hotremove_migrate_alloc(struct page *page, unsigned long private, int **x)
 {
-       /* This should be improoooooved!! */
-       return alloc_page(GFP_HIGHUSER_PAGECACHE);
+       /* This should be improooooved!! */
+       return alloc_page(GFP_HIGHUSER_MOVABLE);
 }
 
-
 #define NR_OFFLINE_AT_ONCE_PAGES       (256)
 static int
 do_migrate_range(unsigned long start_pfn, unsigned long end_pfn)
@@ -397,39 +710,47 @@ do_migrate_range(unsigned long start_pfn, unsigned long end_pfn)
                if (!pfn_valid(pfn))
                        continue;
                page = pfn_to_page(pfn);
-               if (!page_count(page))
+               if (!get_page_unless_zero(page))
                        continue;
                /*
                 * We can skip free pages. And we can only deal with pages on
                 * LRU.
                 */
-               ret = isolate_lru_page(page, &source);
+               ret = isolate_lru_page(page);
                if (!ret) { /* Success */
+                       put_page(page);
+                       list_add_tail(&page->lru, &source);
                        move_pages--;
+                       inc_zone_page_state(page, NR_ISOLATED_ANON +
+                                           page_is_file_cache(page));
+
                } else {
-                       /* Becasue we don't have big zone->lock. we should
-                          check this again here. */
-                       if (page_count(page))
-                               not_managed++;
 #ifdef CONFIG_DEBUG_VM
-                       printk(KERN_INFO "removing from LRU failed"
-                                        " %lx/%d/%lx\n",
-                               pfn, page_count(page), page->flags);
+                       printk(KERN_ALERT "removing pfn %lx from LRU failed\n",
+                              pfn);
+                       dump_page(page);
 #endif
+                       put_page(page);
+                       /* Because we don't have big zone->lock. we should
+                          check this again here. */
+                       if (page_count(page)) {
+                               not_managed++;
+                               ret = -EBUSY;
+                               break;
+                       }
                }
        }
-       ret = -EBUSY;
-       if (not_managed) {
-               if (!list_empty(&source))
+       if (!list_empty(&source)) {
+               if (not_managed) {
+                       putback_lru_pages(&source);
+                       goto out;
+               }
+               /* this function returns # of failed pages */
+               ret = migrate_pages(&source, hotremove_migrate_alloc, 0,
+                                                               true, true);
+               if (ret)
                        putback_lru_pages(&source);
-               goto out;
        }
-       ret = 0;
-       if (list_empty(&source))
-               goto out;
-       /* this function returns # of failed pages */
-       ret = migrate_pages(&source, hotremove_migrate_alloc, 0);
-
 out:
        return ret;
 }
@@ -448,7 +769,7 @@ offline_isolated_pages_cb(unsigned long start, unsigned long nr_pages,
 static void
 offline_isolated_pages(unsigned long start_pfn, unsigned long end_pfn)
 {
-       walk_memory_resource(start_pfn, end_pfn - start_pfn, NULL,
+       walk_system_ram_range(start_pfn, end_pfn - start_pfn, NULL,
                                offline_isolated_pages_cb);
 }
 
@@ -474,14 +795,14 @@ check_pages_isolated(unsigned long start_pfn, unsigned long end_pfn)
        long offlined = 0;
        int ret;
 
-       ret = walk_memory_resource(start_pfn, end_pfn - start_pfn, &offlined,
+       ret = walk_system_ram_range(start_pfn, end_pfn - start_pfn, &offlined,
                        check_pages_isolated_cb);
        if (ret < 0)
                offlined = (long)ret;
        return offlined;
 }
 
-int offline_pages(unsigned long start_pfn,
+static int __ref offline_pages(unsigned long start_pfn,
                  unsigned long end_pfn, unsigned long timeout)
 {
        unsigned long pfn, nr_pages, expire;
@@ -501,6 +822,8 @@ int offline_pages(unsigned long start_pfn,
        if (!test_pages_in_a_zone(start_pfn, end_pfn))
                return -EINVAL;
 
+       lock_memory_hotplug();
+
        zone = page_zone(pfn_to_page(start_pfn));
        node = zone_to_nid(zone);
        nr_pages = end_pfn - start_pfn;
@@ -508,7 +831,7 @@ int offline_pages(unsigned long start_pfn,
        /* set above range as isolated */
        ret = start_isolate_page_range(start_pfn, end_pfn);
        if (ret)
-               return ret;
+               goto out;
 
        arg.start_pfn = start_pfn;
        arg.nr_pages = nr_pages;
@@ -536,7 +859,6 @@ repeat:
        ret = 0;
        if (drain) {
                lru_add_drain_all();
-               flush_scheduled_work();
                cond_resched();
                drain_all_pages();
        }
@@ -558,7 +880,6 @@ repeat:
        }
        /* drain all zone's lru pagevec, this is asyncronous... */
        lru_add_drain_all();
-       flush_scheduled_work();
        yield();
        /* drain pcp pages , this is synchrouns. */
        drain_all_pages();
@@ -578,12 +899,19 @@ repeat:
        zone->present_pages -= offlined_pages;
        zone->zone_pgdat->node_present_pages -= offlined_pages;
        totalram_pages -= offlined_pages;
-       num_physpages -= offlined_pages;
+
+       init_per_zone_wmark_min();
+
+       if (!node_present_pages(node)) {
+               node_clear_state(node, N_HIGH_MEMORY);
+               kswapd_stop(node);
+       }
 
        vm_total_pages = nr_free_pagecache_pages();
        writeback_set_ratelimit();
 
        memory_notify(MEM_OFFLINE, &arg);
+       unlock_memory_hotplug();
        return 0;
 
 failed_removal:
@@ -593,12 +921,23 @@ failed_removal:
        /* pushback to free area */
        undo_isolate_page_range(start_pfn, end_pfn);
 
+out:
+       unlock_memory_hotplug();
        return ret;
 }
+
+int remove_memory(u64 start, u64 size)
+{
+       unsigned long start_pfn, end_pfn;
+
+       start_pfn = PFN_DOWN(start);
+       end_pfn = start_pfn + PFN_DOWN(size);
+       return offline_pages(start_pfn, end_pfn, 120 * HZ);
+}
 #else
 int remove_memory(u64 start, u64 size)
 {
        return -EINVAL;
 }
-EXPORT_SYMBOL_GPL(remove_memory);
 #endif /* CONFIG_MEMORY_HOTREMOVE */
+EXPORT_SYMBOL_GPL(remove_memory);