mm: migrate: consolidate mem_cgroup_migrate() calls
authorJohannes Weiner <hannes@cmpxchg.org>
Tue, 15 Mar 2016 21:57:54 +0000 (14:57 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Tue, 15 Mar 2016 23:55:16 +0000 (16:55 -0700)
Rather than scattering mem_cgroup_migrate() calls all over the place,
have a single call from a safe place where every migration operation
eventually ends up in - migrate_page_copy().

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Suggested-by: Hugh Dickins <hughd@google.com>
Acked-by: Vladimir Davydov <vdavydov@virtuozzo.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Mateusz Guzik <mguzik@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky.work@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/migrate.c

index 848327d4a7edad7c4bd97dcb7166eebdd0839c69..568284ec75d419863866f43065003883b8d81939 100644 (file)
@@ -331,8 +331,6 @@ int migrate_page_move_mapping(struct address_space *mapping,
                if (PageSwapBacked(page))
                        SetPageSwapBacked(newpage);
 
-               mem_cgroup_migrate(page, newpage);
-
                return MIGRATEPAGE_SUCCESS;
        }
 
@@ -428,8 +426,6 @@ int migrate_page_move_mapping(struct address_space *mapping,
        }
        local_irq_enable();
 
-       mem_cgroup_migrate(page, newpage);
-
        return MIGRATEPAGE_SUCCESS;
 }
 
@@ -471,8 +467,6 @@ int migrate_huge_page_move_mapping(struct address_space *mapping,
 
        spin_unlock_irq(&mapping->tree_lock);
 
-       mem_cgroup_migrate(page, newpage);
-
        return MIGRATEPAGE_SUCCESS;
 }
 
@@ -586,6 +580,8 @@ void migrate_page_copy(struct page *newpage, struct page *page)
                end_page_writeback(newpage);
 
        copy_page_owner(page, newpage);
+
+       mem_cgroup_migrate(page, newpage);
 }
 
 /************************************************************
@@ -1846,7 +1842,6 @@ fail_putback:
        }
 
        mlock_migrate_page(new_page, page);
-       mem_cgroup_migrate(page, new_page);
        page_remove_rmap(page, true);
        set_page_owner_migrate_reason(new_page, MR_NUMA_MISPLACED);