Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1 | /* |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2 | * Memory merging support. |
| 3 | * |
| 4 | * This code enables dynamic sharing of identical pages found in different |
| 5 | * memory areas, even if they are not shared by fork() |
| 6 | * |
Izik Eidus | 36b2528 | 2009-09-21 17:02:06 -0700 | [diff] [blame] | 7 | * Copyright (C) 2008-2009 Red Hat, Inc. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 8 | * Authors: |
| 9 | * Izik Eidus |
| 10 | * Andrea Arcangeli |
| 11 | * Chris Wright |
Izik Eidus | 36b2528 | 2009-09-21 17:02:06 -0700 | [diff] [blame] | 12 | * Hugh Dickins |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 13 | * |
| 14 | * This work is licensed under the terms of the GNU GPL, version 2. |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 15 | */ |
| 16 | |
| 17 | #include <linux/errno.h> |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 18 | #include <linux/mm.h> |
| 19 | #include <linux/fs.h> |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 20 | #include <linux/mman.h> |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 21 | #include <linux/sched.h> |
| 22 | #include <linux/rwsem.h> |
| 23 | #include <linux/pagemap.h> |
| 24 | #include <linux/rmap.h> |
| 25 | #include <linux/spinlock.h> |
| 26 | #include <linux/jhash.h> |
| 27 | #include <linux/delay.h> |
| 28 | #include <linux/kthread.h> |
| 29 | #include <linux/wait.h> |
| 30 | #include <linux/slab.h> |
| 31 | #include <linux/rbtree.h> |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 32 | #include <linux/memory.h> |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 33 | #include <linux/mmu_notifier.h> |
Izik Eidus | 2c6854f | 2009-09-23 15:56:04 -0700 | [diff] [blame] | 34 | #include <linux/swap.h> |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 35 | #include <linux/ksm.h> |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 36 | #include <linux/hash.h> |
Andrea Arcangeli | 878aee7 | 2011-01-13 15:47:10 -0800 | [diff] [blame] | 37 | #include <linux/freezer.h> |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 38 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 39 | #include <asm/tlbflush.h> |
Hugh Dickins | 73848b4 | 2009-12-14 17:59:22 -0800 | [diff] [blame] | 40 | #include "internal.h" |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 41 | |
| 42 | /* |
| 43 | * A few notes about the KSM scanning process, |
| 44 | * to make it easier to understand the data structures below: |
| 45 | * |
| 46 | * In order to reduce excessive scanning, KSM sorts the memory pages by their |
| 47 | * contents into a data structure that holds pointers to the pages' locations. |
| 48 | * |
| 49 | * Since the contents of the pages may change at any moment, KSM cannot just |
| 50 | * insert the pages into a normal sorted tree and expect it to find anything. |
| 51 | * Therefore KSM uses two data structures - the stable and the unstable tree. |
| 52 | * |
| 53 | * The stable tree holds pointers to all the merged pages (ksm pages), sorted |
| 54 | * by their contents. Because each such page is write-protected, searching on |
| 55 | * this tree is fully assured to be working (except when pages are unmapped), |
| 56 | * and therefore this tree is called the stable tree. |
| 57 | * |
| 58 | * In addition to the stable tree, KSM uses a second data structure called the |
| 59 | * unstable tree: this tree holds pointers to pages which have been found to |
| 60 | * be "unchanged for a period of time". The unstable tree sorts these pages |
| 61 | * by their contents, but since they are not write-protected, KSM cannot rely |
| 62 | * upon the unstable tree to work correctly - the unstable tree is liable to |
| 63 | * be corrupted as its contents are modified, and so it is called unstable. |
| 64 | * |
| 65 | * KSM solves this problem by several techniques: |
| 66 | * |
| 67 | * 1) The unstable tree is flushed every time KSM completes scanning all |
| 68 | * memory areas, and then the tree is rebuilt again from the beginning. |
| 69 | * 2) KSM will only insert into the unstable tree, pages whose hash value |
| 70 | * has not changed since the previous scan of all memory areas. |
| 71 | * 3) The unstable tree is a RedBlack Tree - so its balancing is based on the |
| 72 | * colors of the nodes and not on their contents, assuring that even when |
| 73 | * the tree gets "corrupted" it won't get out of balance, so scanning time |
| 74 | * remains the same (also, searching and inserting nodes in an rbtree uses |
| 75 | * the same algorithm, so we have no overhead when we flush and rebuild). |
| 76 | * 4) KSM never flushes the stable tree, which means that even if it were to |
| 77 | * take 10 attempts to find a page in the unstable tree, once it is found, |
| 78 | * it is secured in the stable tree. (When we scan a new page, we first |
| 79 | * compare it against the stable tree, and then against the unstable tree.) |
| 80 | */ |
| 81 | |
| 82 | /** |
| 83 | * struct mm_slot - ksm information per mm that is being scanned |
| 84 | * @link: link to the mm_slots hash list |
| 85 | * @mm_list: link into the mm_slots list, rooted in ksm_mm_head |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 86 | * @rmap_list: head for this mm_slot's singly-linked list of rmap_items |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 87 | * @mm: the mm that this information is valid for |
| 88 | */ |
| 89 | struct mm_slot { |
| 90 | struct hlist_node link; |
| 91 | struct list_head mm_list; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 92 | struct rmap_item *rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 93 | struct mm_struct *mm; |
| 94 | }; |
| 95 | |
| 96 | /** |
| 97 | * struct ksm_scan - cursor for scanning |
| 98 | * @mm_slot: the current mm_slot we are scanning |
| 99 | * @address: the next address inside that to be scanned |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 100 | * @rmap_list: link to the next rmap to be scanned in the rmap_list |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 101 | * @seqnr: count of completed full scans (needed when removing unstable node) |
| 102 | * |
| 103 | * There is only the one ksm_scan instance of this cursor structure. |
| 104 | */ |
| 105 | struct ksm_scan { |
| 106 | struct mm_slot *mm_slot; |
| 107 | unsigned long address; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 108 | struct rmap_item **rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 109 | unsigned long seqnr; |
| 110 | }; |
| 111 | |
| 112 | /** |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 113 | * struct stable_node - node of the stable rbtree |
| 114 | * @node: rb node of this ksm page in the stable tree |
| 115 | * @hlist: hlist head of rmap_items using this ksm page |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 116 | * @kpfn: page frame number of this ksm page |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 117 | */ |
| 118 | struct stable_node { |
| 119 | struct rb_node node; |
| 120 | struct hlist_head hlist; |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 121 | unsigned long kpfn; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 122 | }; |
| 123 | |
| 124 | /** |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 125 | * struct rmap_item - reverse mapping item for virtual addresses |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 126 | * @rmap_list: next rmap_item in mm_slot's singly-linked rmap_list |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 127 | * @anon_vma: pointer to anon_vma for this mm,address, when in stable tree |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 128 | * @mm: the memory structure this rmap_item is pointing into |
| 129 | * @address: the virtual address this rmap_item tracks (+ flags in low bits) |
| 130 | * @oldchecksum: previous checksum of the page at that virtual address |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 131 | * @node: rb node of this rmap_item in the unstable tree |
| 132 | * @head: pointer to stable_node heading this list in the stable tree |
| 133 | * @hlist: link into hlist of rmap_items hanging off that stable_node |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 134 | */ |
| 135 | struct rmap_item { |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 136 | struct rmap_item *rmap_list; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 137 | struct anon_vma *anon_vma; /* when stable */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 138 | struct mm_struct *mm; |
| 139 | unsigned long address; /* + low bits used for flags below */ |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 140 | unsigned int oldchecksum; /* when unstable */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 141 | union { |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 142 | struct rb_node node; /* when node of unstable tree */ |
| 143 | struct { /* when listed from stable tree */ |
| 144 | struct stable_node *head; |
| 145 | struct hlist_node hlist; |
| 146 | }; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 147 | }; |
| 148 | }; |
| 149 | |
| 150 | #define SEQNR_MASK 0x0ff /* low bits of unstable tree seqnr */ |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 151 | #define UNSTABLE_FLAG 0x100 /* is a node of the unstable tree */ |
| 152 | #define STABLE_FLAG 0x200 /* is listed from the stable tree */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 153 | |
| 154 | /* The stable and unstable tree heads */ |
| 155 | static struct rb_root root_stable_tree = RB_ROOT; |
| 156 | static struct rb_root root_unstable_tree = RB_ROOT; |
| 157 | |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 158 | #define MM_SLOTS_HASH_SHIFT 10 |
| 159 | #define MM_SLOTS_HASH_HEADS (1 << MM_SLOTS_HASH_SHIFT) |
| 160 | static struct hlist_head mm_slots_hash[MM_SLOTS_HASH_HEADS]; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 161 | |
| 162 | static struct mm_slot ksm_mm_head = { |
| 163 | .mm_list = LIST_HEAD_INIT(ksm_mm_head.mm_list), |
| 164 | }; |
| 165 | static struct ksm_scan ksm_scan = { |
| 166 | .mm_slot = &ksm_mm_head, |
| 167 | }; |
| 168 | |
| 169 | static struct kmem_cache *rmap_item_cache; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 170 | static struct kmem_cache *stable_node_cache; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 171 | static struct kmem_cache *mm_slot_cache; |
| 172 | |
| 173 | /* The number of nodes in the stable tree */ |
Hugh Dickins | b402826 | 2009-09-21 17:02:09 -0700 | [diff] [blame] | 174 | static unsigned long ksm_pages_shared; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 175 | |
Hugh Dickins | e178dfd | 2009-09-21 17:02:10 -0700 | [diff] [blame] | 176 | /* The number of page slots additionally sharing those nodes */ |
Hugh Dickins | b402826 | 2009-09-21 17:02:09 -0700 | [diff] [blame] | 177 | static unsigned long ksm_pages_sharing; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 178 | |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 179 | /* The number of nodes in the unstable tree */ |
| 180 | static unsigned long ksm_pages_unshared; |
| 181 | |
| 182 | /* The number of rmap_items in use: to calculate pages_volatile */ |
| 183 | static unsigned long ksm_rmap_items; |
| 184 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 185 | /* Number of pages ksmd should scan in one batch */ |
Izik Eidus | 2c6854f | 2009-09-23 15:56:04 -0700 | [diff] [blame] | 186 | static unsigned int ksm_thread_pages_to_scan = 100; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 187 | |
| 188 | /* Milliseconds ksmd should sleep between batches */ |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 189 | static unsigned int ksm_thread_sleep_millisecs = 20; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 190 | |
| 191 | #define KSM_RUN_STOP 0 |
| 192 | #define KSM_RUN_MERGE 1 |
| 193 | #define KSM_RUN_UNMERGE 2 |
Izik Eidus | 2c6854f | 2009-09-23 15:56:04 -0700 | [diff] [blame] | 194 | static unsigned int ksm_run = KSM_RUN_STOP; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 195 | |
| 196 | static DECLARE_WAIT_QUEUE_HEAD(ksm_thread_wait); |
| 197 | static DEFINE_MUTEX(ksm_thread_mutex); |
| 198 | static DEFINE_SPINLOCK(ksm_mmlist_lock); |
| 199 | |
| 200 | #define KSM_KMEM_CACHE(__struct, __flags) kmem_cache_create("ksm_"#__struct,\ |
| 201 | sizeof(struct __struct), __alignof__(struct __struct),\ |
| 202 | (__flags), NULL) |
| 203 | |
| 204 | static int __init ksm_slab_init(void) |
| 205 | { |
| 206 | rmap_item_cache = KSM_KMEM_CACHE(rmap_item, 0); |
| 207 | if (!rmap_item_cache) |
| 208 | goto out; |
| 209 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 210 | stable_node_cache = KSM_KMEM_CACHE(stable_node, 0); |
| 211 | if (!stable_node_cache) |
| 212 | goto out_free1; |
| 213 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 214 | mm_slot_cache = KSM_KMEM_CACHE(mm_slot, 0); |
| 215 | if (!mm_slot_cache) |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 216 | goto out_free2; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 217 | |
| 218 | return 0; |
| 219 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 220 | out_free2: |
| 221 | kmem_cache_destroy(stable_node_cache); |
| 222 | out_free1: |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 223 | kmem_cache_destroy(rmap_item_cache); |
| 224 | out: |
| 225 | return -ENOMEM; |
| 226 | } |
| 227 | |
| 228 | static void __init ksm_slab_free(void) |
| 229 | { |
| 230 | kmem_cache_destroy(mm_slot_cache); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 231 | kmem_cache_destroy(stable_node_cache); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 232 | kmem_cache_destroy(rmap_item_cache); |
| 233 | mm_slot_cache = NULL; |
| 234 | } |
| 235 | |
| 236 | static inline struct rmap_item *alloc_rmap_item(void) |
| 237 | { |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 238 | struct rmap_item *rmap_item; |
| 239 | |
| 240 | rmap_item = kmem_cache_zalloc(rmap_item_cache, GFP_KERNEL); |
| 241 | if (rmap_item) |
| 242 | ksm_rmap_items++; |
| 243 | return rmap_item; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 244 | } |
| 245 | |
| 246 | static inline void free_rmap_item(struct rmap_item *rmap_item) |
| 247 | { |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 248 | ksm_rmap_items--; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 249 | rmap_item->mm = NULL; /* debug safety */ |
| 250 | kmem_cache_free(rmap_item_cache, rmap_item); |
| 251 | } |
| 252 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 253 | static inline struct stable_node *alloc_stable_node(void) |
| 254 | { |
| 255 | return kmem_cache_alloc(stable_node_cache, GFP_KERNEL); |
| 256 | } |
| 257 | |
| 258 | static inline void free_stable_node(struct stable_node *stable_node) |
| 259 | { |
| 260 | kmem_cache_free(stable_node_cache, stable_node); |
| 261 | } |
| 262 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 263 | static inline struct mm_slot *alloc_mm_slot(void) |
| 264 | { |
| 265 | if (!mm_slot_cache) /* initialization failed */ |
| 266 | return NULL; |
| 267 | return kmem_cache_zalloc(mm_slot_cache, GFP_KERNEL); |
| 268 | } |
| 269 | |
| 270 | static inline void free_mm_slot(struct mm_slot *mm_slot) |
| 271 | { |
| 272 | kmem_cache_free(mm_slot_cache, mm_slot); |
| 273 | } |
| 274 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 275 | static struct mm_slot *get_mm_slot(struct mm_struct *mm) |
| 276 | { |
| 277 | struct mm_slot *mm_slot; |
| 278 | struct hlist_head *bucket; |
| 279 | struct hlist_node *node; |
| 280 | |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 281 | bucket = &mm_slots_hash[hash_ptr(mm, MM_SLOTS_HASH_SHIFT)]; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 282 | hlist_for_each_entry(mm_slot, node, bucket, link) { |
| 283 | if (mm == mm_slot->mm) |
| 284 | return mm_slot; |
| 285 | } |
| 286 | return NULL; |
| 287 | } |
| 288 | |
| 289 | static void insert_to_mm_slots_hash(struct mm_struct *mm, |
| 290 | struct mm_slot *mm_slot) |
| 291 | { |
| 292 | struct hlist_head *bucket; |
| 293 | |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 294 | bucket = &mm_slots_hash[hash_ptr(mm, MM_SLOTS_HASH_SHIFT)]; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 295 | mm_slot->mm = mm; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 296 | hlist_add_head(&mm_slot->link, bucket); |
| 297 | } |
| 298 | |
| 299 | static inline int in_stable_tree(struct rmap_item *rmap_item) |
| 300 | { |
| 301 | return rmap_item->address & STABLE_FLAG; |
| 302 | } |
| 303 | |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 304 | static void hold_anon_vma(struct rmap_item *rmap_item, |
| 305 | struct anon_vma *anon_vma) |
| 306 | { |
| 307 | rmap_item->anon_vma = anon_vma; |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 308 | get_anon_vma(anon_vma); |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 309 | } |
| 310 | |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 311 | static void ksm_drop_anon_vma(struct rmap_item *rmap_item) |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 312 | { |
| 313 | struct anon_vma *anon_vma = rmap_item->anon_vma; |
| 314 | |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 315 | drop_anon_vma(anon_vma); |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 316 | } |
| 317 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 318 | /* |
Hugh Dickins | a913e18 | 2009-09-21 17:02:26 -0700 | [diff] [blame] | 319 | * ksmd, and unmerge_and_remove_all_rmap_items(), must not touch an mm's |
| 320 | * page tables after it has passed through ksm_exit() - which, if necessary, |
| 321 | * takes mmap_sem briefly to serialize against them. ksm_exit() does not set |
| 322 | * a special flag: they can just back out as soon as mm_users goes to zero. |
| 323 | * ksm_test_exit() is used throughout to make this test for exit: in some |
| 324 | * places for correctness, in some places just to avoid unnecessary work. |
| 325 | */ |
| 326 | static inline bool ksm_test_exit(struct mm_struct *mm) |
| 327 | { |
| 328 | return atomic_read(&mm->mm_users) == 0; |
| 329 | } |
| 330 | |
| 331 | /* |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 332 | * We use break_ksm to break COW on a ksm page: it's a stripped down |
| 333 | * |
| 334 | * if (get_user_pages(current, mm, addr, 1, 1, 1, &page, NULL) == 1) |
| 335 | * put_page(page); |
| 336 | * |
| 337 | * but taking great care only to touch a ksm page, in a VM_MERGEABLE vma, |
| 338 | * in case the application has unmapped and remapped mm,addr meanwhile. |
| 339 | * Could a ksm page appear anywhere else? Actually yes, in a VM_PFNMAP |
| 340 | * mmap of /dev/mem or /dev/kmem, where we would not want to touch it. |
| 341 | */ |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 342 | static int break_ksm(struct vm_area_struct *vma, unsigned long addr) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 343 | { |
| 344 | struct page *page; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 345 | int ret = 0; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 346 | |
| 347 | do { |
| 348 | cond_resched(); |
| 349 | page = follow_page(vma, addr, FOLL_GET); |
Dan Carpenter | 22eccdd | 2010-04-23 13:18:10 -0400 | [diff] [blame] | 350 | if (IS_ERR_OR_NULL(page)) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 351 | break; |
| 352 | if (PageKsm(page)) |
| 353 | ret = handle_mm_fault(vma->vm_mm, vma, addr, |
| 354 | FAULT_FLAG_WRITE); |
| 355 | else |
| 356 | ret = VM_FAULT_WRITE; |
| 357 | put_page(page); |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 358 | } while (!(ret & (VM_FAULT_WRITE | VM_FAULT_SIGBUS | VM_FAULT_OOM))); |
| 359 | /* |
| 360 | * We must loop because handle_mm_fault() may back out if there's |
| 361 | * any difficulty e.g. if pte accessed bit gets updated concurrently. |
| 362 | * |
| 363 | * VM_FAULT_WRITE is what we have been hoping for: it indicates that |
| 364 | * COW has been broken, even if the vma does not permit VM_WRITE; |
| 365 | * but note that a concurrent fault might break PageKsm for us. |
| 366 | * |
| 367 | * VM_FAULT_SIGBUS could occur if we race with truncation of the |
| 368 | * backing file, which also invalidates anonymous pages: that's |
| 369 | * okay, that truncation will have unmapped the PageKsm for us. |
| 370 | * |
| 371 | * VM_FAULT_OOM: at the time of writing (late July 2009), setting |
| 372 | * aside mem_cgroup limits, VM_FAULT_OOM would only be set if the |
| 373 | * current task has TIF_MEMDIE set, and will be OOM killed on return |
| 374 | * to user; and ksmd, having no mm, would never be chosen for that. |
| 375 | * |
| 376 | * But if the mm is in a limited mem_cgroup, then the fault may fail |
| 377 | * with VM_FAULT_OOM even if the current task is not TIF_MEMDIE; and |
| 378 | * even ksmd can fail in this way - though it's usually breaking ksm |
| 379 | * just to undo a merge it made a moment before, so unlikely to oom. |
| 380 | * |
| 381 | * That's a pity: we might therefore have more kernel pages allocated |
| 382 | * than we're counting as nodes in the stable tree; but ksm_do_scan |
| 383 | * will retry to break_cow on each pass, so should recover the page |
| 384 | * in due course. The important thing is to not let VM_MERGEABLE |
| 385 | * be cleared while any such pages might remain in the area. |
| 386 | */ |
| 387 | return (ret & VM_FAULT_OOM) ? -ENOMEM : 0; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 388 | } |
| 389 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 390 | static void break_cow(struct rmap_item *rmap_item) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 391 | { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 392 | struct mm_struct *mm = rmap_item->mm; |
| 393 | unsigned long addr = rmap_item->address; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 394 | struct vm_area_struct *vma; |
| 395 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 396 | /* |
| 397 | * It is not an accident that whenever we want to break COW |
| 398 | * to undo, we also need to drop a reference to the anon_vma. |
| 399 | */ |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 400 | ksm_drop_anon_vma(rmap_item); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 401 | |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 402 | down_read(&mm->mmap_sem); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 403 | if (ksm_test_exit(mm)) |
| 404 | goto out; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 405 | vma = find_vma(mm, addr); |
| 406 | if (!vma || vma->vm_start > addr) |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 407 | goto out; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 408 | if (!(vma->vm_flags & VM_MERGEABLE) || !vma->anon_vma) |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 409 | goto out; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 410 | break_ksm(vma, addr); |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 411 | out: |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 412 | up_read(&mm->mmap_sem); |
| 413 | } |
| 414 | |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 415 | static struct page *page_trans_compound_anon(struct page *page) |
| 416 | { |
| 417 | if (PageTransCompound(page)) { |
Andrea Arcangeli | 22e5c47 | 2011-01-13 15:47:20 -0800 | [diff] [blame] | 418 | struct page *head = compound_trans_head(page); |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 419 | /* |
Andrea Arcangeli | 22e5c47 | 2011-01-13 15:47:20 -0800 | [diff] [blame] | 420 | * head may actually be splitted and freed from under |
| 421 | * us but it's ok here. |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 422 | */ |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 423 | if (PageAnon(head)) |
| 424 | return head; |
| 425 | } |
| 426 | return NULL; |
| 427 | } |
| 428 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 429 | static struct page *get_mergeable_page(struct rmap_item *rmap_item) |
| 430 | { |
| 431 | struct mm_struct *mm = rmap_item->mm; |
| 432 | unsigned long addr = rmap_item->address; |
| 433 | struct vm_area_struct *vma; |
| 434 | struct page *page; |
| 435 | |
| 436 | down_read(&mm->mmap_sem); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 437 | if (ksm_test_exit(mm)) |
| 438 | goto out; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 439 | vma = find_vma(mm, addr); |
| 440 | if (!vma || vma->vm_start > addr) |
| 441 | goto out; |
| 442 | if (!(vma->vm_flags & VM_MERGEABLE) || !vma->anon_vma) |
| 443 | goto out; |
| 444 | |
| 445 | page = follow_page(vma, addr, FOLL_GET); |
Dan Carpenter | 22eccdd | 2010-04-23 13:18:10 -0400 | [diff] [blame] | 446 | if (IS_ERR_OR_NULL(page)) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 447 | goto out; |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 448 | if (PageAnon(page) || page_trans_compound_anon(page)) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 449 | flush_anon_page(vma, page, addr); |
| 450 | flush_dcache_page(page); |
| 451 | } else { |
| 452 | put_page(page); |
| 453 | out: page = NULL; |
| 454 | } |
| 455 | up_read(&mm->mmap_sem); |
| 456 | return page; |
| 457 | } |
| 458 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 459 | static void remove_node_from_stable_tree(struct stable_node *stable_node) |
| 460 | { |
| 461 | struct rmap_item *rmap_item; |
| 462 | struct hlist_node *hlist; |
| 463 | |
| 464 | hlist_for_each_entry(rmap_item, hlist, &stable_node->hlist, hlist) { |
| 465 | if (rmap_item->hlist.next) |
| 466 | ksm_pages_sharing--; |
| 467 | else |
| 468 | ksm_pages_shared--; |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 469 | ksm_drop_anon_vma(rmap_item); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 470 | rmap_item->address &= PAGE_MASK; |
| 471 | cond_resched(); |
| 472 | } |
| 473 | |
| 474 | rb_erase(&stable_node->node, &root_stable_tree); |
| 475 | free_stable_node(stable_node); |
| 476 | } |
| 477 | |
| 478 | /* |
| 479 | * get_ksm_page: checks if the page indicated by the stable node |
| 480 | * is still its ksm page, despite having held no reference to it. |
| 481 | * In which case we can trust the content of the page, and it |
| 482 | * returns the gotten page; but if the page has now been zapped, |
| 483 | * remove the stale node from the stable tree and return NULL. |
| 484 | * |
| 485 | * You would expect the stable_node to hold a reference to the ksm page. |
| 486 | * But if it increments the page's count, swapping out has to wait for |
| 487 | * ksmd to come around again before it can free the page, which may take |
| 488 | * seconds or even minutes: much too unresponsive. So instead we use a |
| 489 | * "keyhole reference": access to the ksm page from the stable node peeps |
| 490 | * out through its keyhole to see if that page still holds the right key, |
| 491 | * pointing back to this stable node. This relies on freeing a PageAnon |
| 492 | * page to reset its page->mapping to NULL, and relies on no other use of |
| 493 | * a page to put something that might look like our key in page->mapping. |
| 494 | * |
| 495 | * include/linux/pagemap.h page_cache_get_speculative() is a good reference, |
| 496 | * but this is different - made simpler by ksm_thread_mutex being held, but |
| 497 | * interesting for assuming that no other use of the struct page could ever |
| 498 | * put our expected_mapping into page->mapping (or a field of the union which |
| 499 | * coincides with page->mapping). The RCU calls are not for KSM at all, but |
| 500 | * to keep the page_count protocol described with page_cache_get_speculative. |
| 501 | * |
| 502 | * Note: it is possible that get_ksm_page() will return NULL one moment, |
| 503 | * then page the next, if the page is in between page_freeze_refs() and |
| 504 | * page_unfreeze_refs(): this shouldn't be a problem anywhere, the page |
| 505 | * is on its way to being freed; but it is an anomaly to bear in mind. |
| 506 | */ |
| 507 | static struct page *get_ksm_page(struct stable_node *stable_node) |
| 508 | { |
| 509 | struct page *page; |
| 510 | void *expected_mapping; |
| 511 | |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 512 | page = pfn_to_page(stable_node->kpfn); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 513 | expected_mapping = (void *)stable_node + |
| 514 | (PAGE_MAPPING_ANON | PAGE_MAPPING_KSM); |
| 515 | rcu_read_lock(); |
| 516 | if (page->mapping != expected_mapping) |
| 517 | goto stale; |
| 518 | if (!get_page_unless_zero(page)) |
| 519 | goto stale; |
| 520 | if (page->mapping != expected_mapping) { |
| 521 | put_page(page); |
| 522 | goto stale; |
| 523 | } |
| 524 | rcu_read_unlock(); |
| 525 | return page; |
| 526 | stale: |
| 527 | rcu_read_unlock(); |
| 528 | remove_node_from_stable_tree(stable_node); |
| 529 | return NULL; |
| 530 | } |
| 531 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 532 | /* |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 533 | * Removing rmap_item from stable or unstable tree. |
| 534 | * This function will clean the information from the stable/unstable tree. |
| 535 | */ |
| 536 | static void remove_rmap_item_from_tree(struct rmap_item *rmap_item) |
| 537 | { |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 538 | if (rmap_item->address & STABLE_FLAG) { |
| 539 | struct stable_node *stable_node; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 540 | struct page *page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 541 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 542 | stable_node = rmap_item->head; |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 543 | page = get_ksm_page(stable_node); |
| 544 | if (!page) |
| 545 | goto out; |
| 546 | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 547 | lock_page(page); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 548 | hlist_del(&rmap_item->hlist); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 549 | unlock_page(page); |
| 550 | put_page(page); |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 551 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 552 | if (stable_node->hlist.first) |
| 553 | ksm_pages_sharing--; |
| 554 | else |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 555 | ksm_pages_shared--; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 556 | |
Rik van Riel | 7654506 | 2010-08-09 17:18:41 -0700 | [diff] [blame] | 557 | ksm_drop_anon_vma(rmap_item); |
Hugh Dickins | 93d1771 | 2009-12-14 17:59:16 -0800 | [diff] [blame] | 558 | rmap_item->address &= PAGE_MASK; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 559 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 560 | } else if (rmap_item->address & UNSTABLE_FLAG) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 561 | unsigned char age; |
| 562 | /* |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 563 | * Usually ksmd can and must skip the rb_erase, because |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 564 | * root_unstable_tree was already reset to RB_ROOT. |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 565 | * But be careful when an mm is exiting: do the rb_erase |
| 566 | * if this rmap_item was inserted by this scan, rather |
| 567 | * than left over from before. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 568 | */ |
| 569 | age = (unsigned char)(ksm_scan.seqnr - rmap_item->address); |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 570 | BUG_ON(age > 1); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 571 | if (!age) |
| 572 | rb_erase(&rmap_item->node, &root_unstable_tree); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 573 | |
Hugh Dickins | 93d1771 | 2009-12-14 17:59:16 -0800 | [diff] [blame] | 574 | ksm_pages_unshared--; |
| 575 | rmap_item->address &= PAGE_MASK; |
| 576 | } |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 577 | out: |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 578 | cond_resched(); /* we're called from many long loops */ |
| 579 | } |
| 580 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 581 | static void remove_trailing_rmap_items(struct mm_slot *mm_slot, |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 582 | struct rmap_item **rmap_list) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 583 | { |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 584 | while (*rmap_list) { |
| 585 | struct rmap_item *rmap_item = *rmap_list; |
| 586 | *rmap_list = rmap_item->rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 587 | remove_rmap_item_from_tree(rmap_item); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 588 | free_rmap_item(rmap_item); |
| 589 | } |
| 590 | } |
| 591 | |
| 592 | /* |
| 593 | * Though it's very tempting to unmerge in_stable_tree(rmap_item)s rather |
| 594 | * than check every pte of a given vma, the locking doesn't quite work for |
| 595 | * that - an rmap_item is assigned to the stable tree after inserting ksm |
| 596 | * page and upping mmap_sem. Nor does it fit with the way we skip dup'ing |
| 597 | * rmap_items from parent to child at fork time (so as not to waste time |
| 598 | * if exit comes before the next scan reaches it). |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 599 | * |
| 600 | * Similarly, although we'd like to remove rmap_items (so updating counts |
| 601 | * and freeing memory) when unmerging an area, it's easier to leave that |
| 602 | * to the next pass of ksmd - consider, for example, how ksmd might be |
| 603 | * in cmp_and_merge_page on one of the rmap_items we would be removing. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 604 | */ |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 605 | static int unmerge_ksm_pages(struct vm_area_struct *vma, |
| 606 | unsigned long start, unsigned long end) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 607 | { |
| 608 | unsigned long addr; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 609 | int err = 0; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 610 | |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 611 | for (addr = start; addr < end && !err; addr += PAGE_SIZE) { |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 612 | if (ksm_test_exit(vma->vm_mm)) |
| 613 | break; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 614 | if (signal_pending(current)) |
| 615 | err = -ERESTARTSYS; |
| 616 | else |
| 617 | err = break_ksm(vma, addr); |
| 618 | } |
| 619 | return err; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 620 | } |
| 621 | |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 622 | #ifdef CONFIG_SYSFS |
| 623 | /* |
| 624 | * Only called through the sysfs control interface: |
| 625 | */ |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 626 | static int unmerge_and_remove_all_rmap_items(void) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 627 | { |
| 628 | struct mm_slot *mm_slot; |
| 629 | struct mm_struct *mm; |
| 630 | struct vm_area_struct *vma; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 631 | int err = 0; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 632 | |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 633 | spin_lock(&ksm_mmlist_lock); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 634 | ksm_scan.mm_slot = list_entry(ksm_mm_head.mm_list.next, |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 635 | struct mm_slot, mm_list); |
| 636 | spin_unlock(&ksm_mmlist_lock); |
| 637 | |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 638 | for (mm_slot = ksm_scan.mm_slot; |
| 639 | mm_slot != &ksm_mm_head; mm_slot = ksm_scan.mm_slot) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 640 | mm = mm_slot->mm; |
| 641 | down_read(&mm->mmap_sem); |
| 642 | for (vma = mm->mmap; vma; vma = vma->vm_next) { |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 643 | if (ksm_test_exit(mm)) |
| 644 | break; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 645 | if (!(vma->vm_flags & VM_MERGEABLE) || !vma->anon_vma) |
| 646 | continue; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 647 | err = unmerge_ksm_pages(vma, |
| 648 | vma->vm_start, vma->vm_end); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 649 | if (err) |
| 650 | goto error; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 651 | } |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 652 | |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 653 | remove_trailing_rmap_items(mm_slot, &mm_slot->rmap_list); |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 654 | |
| 655 | spin_lock(&ksm_mmlist_lock); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 656 | ksm_scan.mm_slot = list_entry(mm_slot->mm_list.next, |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 657 | struct mm_slot, mm_list); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 658 | if (ksm_test_exit(mm)) { |
| 659 | hlist_del(&mm_slot->link); |
| 660 | list_del(&mm_slot->mm_list); |
| 661 | spin_unlock(&ksm_mmlist_lock); |
| 662 | |
| 663 | free_mm_slot(mm_slot); |
| 664 | clear_bit(MMF_VM_MERGEABLE, &mm->flags); |
| 665 | up_read(&mm->mmap_sem); |
| 666 | mmdrop(mm); |
| 667 | } else { |
| 668 | spin_unlock(&ksm_mmlist_lock); |
| 669 | up_read(&mm->mmap_sem); |
| 670 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 671 | } |
| 672 | |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 673 | ksm_scan.seqnr = 0; |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 674 | return 0; |
| 675 | |
| 676 | error: |
| 677 | up_read(&mm->mmap_sem); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 678 | spin_lock(&ksm_mmlist_lock); |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 679 | ksm_scan.mm_slot = &ksm_mm_head; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 680 | spin_unlock(&ksm_mmlist_lock); |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 681 | return err; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 682 | } |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 683 | #endif /* CONFIG_SYSFS */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 684 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 685 | static u32 calc_checksum(struct page *page) |
| 686 | { |
| 687 | u32 checksum; |
| 688 | void *addr = kmap_atomic(page, KM_USER0); |
| 689 | checksum = jhash2(addr, PAGE_SIZE / 4, 17); |
| 690 | kunmap_atomic(addr, KM_USER0); |
| 691 | return checksum; |
| 692 | } |
| 693 | |
| 694 | static int memcmp_pages(struct page *page1, struct page *page2) |
| 695 | { |
| 696 | char *addr1, *addr2; |
| 697 | int ret; |
| 698 | |
| 699 | addr1 = kmap_atomic(page1, KM_USER0); |
| 700 | addr2 = kmap_atomic(page2, KM_USER1); |
| 701 | ret = memcmp(addr1, addr2, PAGE_SIZE); |
| 702 | kunmap_atomic(addr2, KM_USER1); |
| 703 | kunmap_atomic(addr1, KM_USER0); |
| 704 | return ret; |
| 705 | } |
| 706 | |
| 707 | static inline int pages_identical(struct page *page1, struct page *page2) |
| 708 | { |
| 709 | return !memcmp_pages(page1, page2); |
| 710 | } |
| 711 | |
| 712 | static int write_protect_page(struct vm_area_struct *vma, struct page *page, |
| 713 | pte_t *orig_pte) |
| 714 | { |
| 715 | struct mm_struct *mm = vma->vm_mm; |
| 716 | unsigned long addr; |
| 717 | pte_t *ptep; |
| 718 | spinlock_t *ptl; |
| 719 | int swapped; |
| 720 | int err = -EFAULT; |
| 721 | |
| 722 | addr = page_address_in_vma(page, vma); |
| 723 | if (addr == -EFAULT) |
| 724 | goto out; |
| 725 | |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 726 | BUG_ON(PageTransCompound(page)); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 727 | ptep = page_check_address(page, mm, addr, &ptl, 0); |
| 728 | if (!ptep) |
| 729 | goto out; |
| 730 | |
Hugh Dickins | 4e31635 | 2010-10-02 17:49:08 -0700 | [diff] [blame] | 731 | if (pte_write(*ptep) || pte_dirty(*ptep)) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 732 | pte_t entry; |
| 733 | |
| 734 | swapped = PageSwapCache(page); |
| 735 | flush_cache_page(vma, addr, page_to_pfn(page)); |
| 736 | /* |
| 737 | * Ok this is tricky, when get_user_pages_fast() run it doesnt |
| 738 | * take any lock, therefore the check that we are going to make |
| 739 | * with the pagecount against the mapcount is racey and |
| 740 | * O_DIRECT can happen right after the check. |
| 741 | * So we clear the pte and flush the tlb before the check |
| 742 | * this assure us that no O_DIRECT can happen after the check |
| 743 | * or in the middle of the check. |
| 744 | */ |
| 745 | entry = ptep_clear_flush(vma, addr, ptep); |
| 746 | /* |
| 747 | * Check that no O_DIRECT or similar I/O is in progress on the |
| 748 | * page |
| 749 | */ |
Hugh Dickins | 31e855e | 2009-12-14 17:59:17 -0800 | [diff] [blame] | 750 | if (page_mapcount(page) + 1 + swapped != page_count(page)) { |
Robin Holt | cb53237 | 2010-03-23 13:35:26 -0700 | [diff] [blame] | 751 | set_pte_at(mm, addr, ptep, entry); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 752 | goto out_unlock; |
| 753 | } |
Hugh Dickins | 4e31635 | 2010-10-02 17:49:08 -0700 | [diff] [blame] | 754 | if (pte_dirty(entry)) |
| 755 | set_page_dirty(page); |
| 756 | entry = pte_mkclean(pte_wrprotect(entry)); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 757 | set_pte_at_notify(mm, addr, ptep, entry); |
| 758 | } |
| 759 | *orig_pte = *ptep; |
| 760 | err = 0; |
| 761 | |
| 762 | out_unlock: |
| 763 | pte_unmap_unlock(ptep, ptl); |
| 764 | out: |
| 765 | return err; |
| 766 | } |
| 767 | |
| 768 | /** |
| 769 | * replace_page - replace page in vma by new ksm page |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 770 | * @vma: vma that holds the pte pointing to page |
| 771 | * @page: the page we are replacing by kpage |
| 772 | * @kpage: the ksm page we replace page by |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 773 | * @orig_pte: the original value of the pte |
| 774 | * |
| 775 | * Returns 0 on success, -EFAULT on failure. |
| 776 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 777 | static int replace_page(struct vm_area_struct *vma, struct page *page, |
| 778 | struct page *kpage, pte_t orig_pte) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 779 | { |
| 780 | struct mm_struct *mm = vma->vm_mm; |
| 781 | pgd_t *pgd; |
| 782 | pud_t *pud; |
| 783 | pmd_t *pmd; |
| 784 | pte_t *ptep; |
| 785 | spinlock_t *ptl; |
| 786 | unsigned long addr; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 787 | int err = -EFAULT; |
| 788 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 789 | addr = page_address_in_vma(page, vma); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 790 | if (addr == -EFAULT) |
| 791 | goto out; |
| 792 | |
| 793 | pgd = pgd_offset(mm, addr); |
| 794 | if (!pgd_present(*pgd)) |
| 795 | goto out; |
| 796 | |
| 797 | pud = pud_offset(pgd, addr); |
| 798 | if (!pud_present(*pud)) |
| 799 | goto out; |
| 800 | |
| 801 | pmd = pmd_offset(pud, addr); |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 802 | BUG_ON(pmd_trans_huge(*pmd)); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 803 | if (!pmd_present(*pmd)) |
| 804 | goto out; |
| 805 | |
| 806 | ptep = pte_offset_map_lock(mm, pmd, addr, &ptl); |
| 807 | if (!pte_same(*ptep, orig_pte)) { |
| 808 | pte_unmap_unlock(ptep, ptl); |
| 809 | goto out; |
| 810 | } |
| 811 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 812 | get_page(kpage); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 813 | page_add_anon_rmap(kpage, vma, addr); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 814 | |
| 815 | flush_cache_page(vma, addr, pte_pfn(*ptep)); |
| 816 | ptep_clear_flush(vma, addr, ptep); |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 817 | set_pte_at_notify(mm, addr, ptep, mk_pte(kpage, vma->vm_page_prot)); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 818 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 819 | page_remove_rmap(page); |
Hugh Dickins | ae52a2a | 2011-01-13 15:46:28 -0800 | [diff] [blame] | 820 | if (!page_mapped(page)) |
| 821 | try_to_free_swap(page); |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 822 | put_page(page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 823 | |
| 824 | pte_unmap_unlock(ptep, ptl); |
| 825 | err = 0; |
| 826 | out: |
| 827 | return err; |
| 828 | } |
| 829 | |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 830 | static int page_trans_compound_anon_split(struct page *page) |
| 831 | { |
| 832 | int ret = 0; |
| 833 | struct page *transhuge_head = page_trans_compound_anon(page); |
| 834 | if (transhuge_head) { |
| 835 | /* Get the reference on the head to split it. */ |
| 836 | if (get_page_unless_zero(transhuge_head)) { |
| 837 | /* |
| 838 | * Recheck we got the reference while the head |
| 839 | * was still anonymous. |
| 840 | */ |
| 841 | if (PageAnon(transhuge_head)) |
| 842 | ret = split_huge_page(transhuge_head); |
| 843 | else |
| 844 | /* |
| 845 | * Retry later if split_huge_page run |
| 846 | * from under us. |
| 847 | */ |
| 848 | ret = 1; |
| 849 | put_page(transhuge_head); |
| 850 | } else |
| 851 | /* Retry later if split_huge_page run from under us. */ |
| 852 | ret = 1; |
| 853 | } |
| 854 | return ret; |
| 855 | } |
| 856 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 857 | /* |
| 858 | * try_to_merge_one_page - take two pages and merge them into one |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 859 | * @vma: the vma that holds the pte pointing to page |
| 860 | * @page: the PageAnon page that we want to replace with kpage |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 861 | * @kpage: the PageKsm page that we want to map instead of page, |
| 862 | * or NULL the first time when we want to use page as kpage. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 863 | * |
| 864 | * This function returns 0 if the pages were merged, -EFAULT otherwise. |
| 865 | */ |
| 866 | static int try_to_merge_one_page(struct vm_area_struct *vma, |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 867 | struct page *page, struct page *kpage) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 868 | { |
| 869 | pte_t orig_pte = __pte(0); |
| 870 | int err = -EFAULT; |
| 871 | |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 872 | if (page == kpage) /* ksm page forked */ |
| 873 | return 0; |
| 874 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 875 | if (!(vma->vm_flags & VM_MERGEABLE)) |
| 876 | goto out; |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 877 | if (PageTransCompound(page) && page_trans_compound_anon_split(page)) |
| 878 | goto out; |
| 879 | BUG_ON(PageTransCompound(page)); |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 880 | if (!PageAnon(page)) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 881 | goto out; |
| 882 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 883 | /* |
| 884 | * We need the page lock to read a stable PageSwapCache in |
| 885 | * write_protect_page(). We use trylock_page() instead of |
| 886 | * lock_page() because we don't want to wait here - we |
| 887 | * prefer to continue scanning and merging different pages, |
| 888 | * then come back to this page when it is unlocked. |
| 889 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 890 | if (!trylock_page(page)) |
Hugh Dickins | 31e855e | 2009-12-14 17:59:17 -0800 | [diff] [blame] | 891 | goto out; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 892 | /* |
| 893 | * If this anonymous page is mapped only here, its pte may need |
| 894 | * to be write-protected. If it's mapped elsewhere, all of its |
| 895 | * ptes are necessarily already write-protected. But in either |
| 896 | * case, we need to lock and check page_count is not raised. |
| 897 | */ |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 898 | if (write_protect_page(vma, page, &orig_pte) == 0) { |
| 899 | if (!kpage) { |
| 900 | /* |
| 901 | * While we hold page lock, upgrade page from |
| 902 | * PageAnon+anon_vma to PageKsm+NULL stable_node: |
| 903 | * stable_tree_insert() will update stable_node. |
| 904 | */ |
| 905 | set_page_stable_node(page, NULL); |
| 906 | mark_page_accessed(page); |
| 907 | err = 0; |
| 908 | } else if (pages_identical(page, kpage)) |
| 909 | err = replace_page(vma, page, kpage, orig_pte); |
| 910 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 911 | |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 912 | if ((vma->vm_flags & VM_LOCKED) && kpage && !err) { |
Hugh Dickins | 73848b4 | 2009-12-14 17:59:22 -0800 | [diff] [blame] | 913 | munlock_vma_page(page); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 914 | if (!PageMlocked(kpage)) { |
| 915 | unlock_page(page); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 916 | lock_page(kpage); |
| 917 | mlock_vma_page(kpage); |
| 918 | page = kpage; /* for final unlock */ |
| 919 | } |
| 920 | } |
Hugh Dickins | 73848b4 | 2009-12-14 17:59:22 -0800 | [diff] [blame] | 921 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 922 | unlock_page(page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 923 | out: |
| 924 | return err; |
| 925 | } |
| 926 | |
| 927 | /* |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 928 | * try_to_merge_with_ksm_page - like try_to_merge_two_pages, |
| 929 | * but no new kernel page is allocated: kpage must already be a ksm page. |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 930 | * |
| 931 | * This function returns 0 if the pages were merged, -EFAULT otherwise. |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 932 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 933 | static int try_to_merge_with_ksm_page(struct rmap_item *rmap_item, |
| 934 | struct page *page, struct page *kpage) |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 935 | { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 936 | struct mm_struct *mm = rmap_item->mm; |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 937 | struct vm_area_struct *vma; |
| 938 | int err = -EFAULT; |
| 939 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 940 | down_read(&mm->mmap_sem); |
| 941 | if (ksm_test_exit(mm)) |
| 942 | goto out; |
| 943 | vma = find_vma(mm, rmap_item->address); |
| 944 | if (!vma || vma->vm_start > rmap_item->address) |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 945 | goto out; |
| 946 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 947 | err = try_to_merge_one_page(vma, page, kpage); |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 948 | if (err) |
| 949 | goto out; |
| 950 | |
| 951 | /* Must get reference to anon_vma while still holding mmap_sem */ |
| 952 | hold_anon_vma(rmap_item, vma->anon_vma); |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 953 | out: |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 954 | up_read(&mm->mmap_sem); |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 955 | return err; |
| 956 | } |
| 957 | |
| 958 | /* |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 959 | * try_to_merge_two_pages - take two identical pages and prepare them |
| 960 | * to be merged into one page. |
| 961 | * |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 962 | * This function returns the kpage if we successfully merged two identical |
| 963 | * pages into one ksm page, NULL otherwise. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 964 | * |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 965 | * Note that this function upgrades page to ksm page: if one of the pages |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 966 | * is already a ksm page, try_to_merge_with_ksm_page should be used. |
| 967 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 968 | static struct page *try_to_merge_two_pages(struct rmap_item *rmap_item, |
| 969 | struct page *page, |
| 970 | struct rmap_item *tree_rmap_item, |
| 971 | struct page *tree_page) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 972 | { |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 973 | int err; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 974 | |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 975 | err = try_to_merge_with_ksm_page(rmap_item, page, NULL); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 976 | if (!err) { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 977 | err = try_to_merge_with_ksm_page(tree_rmap_item, |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 978 | tree_page, page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 979 | /* |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 980 | * If that fails, we have a ksm page with only one pte |
| 981 | * pointing to it: so break it. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 982 | */ |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 983 | if (err) |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 984 | break_cow(rmap_item); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 985 | } |
Hugh Dickins | 80e1482 | 2009-12-14 17:59:29 -0800 | [diff] [blame] | 986 | return err ? NULL : page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 987 | } |
| 988 | |
| 989 | /* |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 990 | * stable_tree_search - search for page inside the stable tree |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 991 | * |
| 992 | * This function checks if there is a page inside the stable tree |
| 993 | * with identical content to the page that we are scanning right now. |
| 994 | * |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 995 | * This function returns the stable tree node of identical content if found, |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 996 | * NULL otherwise. |
| 997 | */ |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 998 | static struct page *stable_tree_search(struct page *page) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 999 | { |
| 1000 | struct rb_node *node = root_stable_tree.rb_node; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1001 | struct stable_node *stable_node; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1002 | |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1003 | stable_node = page_stable_node(page); |
| 1004 | if (stable_node) { /* ksm page forked */ |
| 1005 | get_page(page); |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1006 | return page; |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1007 | } |
| 1008 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1009 | while (node) { |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1010 | struct page *tree_page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1011 | int ret; |
| 1012 | |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1013 | cond_resched(); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1014 | stable_node = rb_entry(node, struct stable_node, node); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1015 | tree_page = get_ksm_page(stable_node); |
| 1016 | if (!tree_page) |
| 1017 | return NULL; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1018 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1019 | ret = memcmp_pages(page, tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1020 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1021 | if (ret < 0) { |
| 1022 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1023 | node = node->rb_left; |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1024 | } else if (ret > 0) { |
| 1025 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1026 | node = node->rb_right; |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1027 | } else |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1028 | return tree_page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1029 | } |
| 1030 | |
| 1031 | return NULL; |
| 1032 | } |
| 1033 | |
| 1034 | /* |
| 1035 | * stable_tree_insert - insert rmap_item pointing to new ksm page |
| 1036 | * into the stable tree. |
| 1037 | * |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1038 | * This function returns the stable tree node just allocated on success, |
| 1039 | * NULL otherwise. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1040 | */ |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1041 | static struct stable_node *stable_tree_insert(struct page *kpage) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1042 | { |
| 1043 | struct rb_node **new = &root_stable_tree.rb_node; |
| 1044 | struct rb_node *parent = NULL; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1045 | struct stable_node *stable_node; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1046 | |
| 1047 | while (*new) { |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1048 | struct page *tree_page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1049 | int ret; |
| 1050 | |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1051 | cond_resched(); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1052 | stable_node = rb_entry(*new, struct stable_node, node); |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1053 | tree_page = get_ksm_page(stable_node); |
| 1054 | if (!tree_page) |
| 1055 | return NULL; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1056 | |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1057 | ret = memcmp_pages(kpage, tree_page); |
| 1058 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1059 | |
| 1060 | parent = *new; |
| 1061 | if (ret < 0) |
| 1062 | new = &parent->rb_left; |
| 1063 | else if (ret > 0) |
| 1064 | new = &parent->rb_right; |
| 1065 | else { |
| 1066 | /* |
| 1067 | * It is not a bug that stable_tree_search() didn't |
| 1068 | * find this node: because at that time our page was |
| 1069 | * not yet write-protected, so may have changed since. |
| 1070 | */ |
| 1071 | return NULL; |
| 1072 | } |
| 1073 | } |
| 1074 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1075 | stable_node = alloc_stable_node(); |
| 1076 | if (!stable_node) |
| 1077 | return NULL; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1078 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1079 | rb_link_node(&stable_node->node, parent, new); |
| 1080 | rb_insert_color(&stable_node->node, &root_stable_tree); |
| 1081 | |
| 1082 | INIT_HLIST_HEAD(&stable_node->hlist); |
| 1083 | |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1084 | stable_node->kpfn = page_to_pfn(kpage); |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1085 | set_page_stable_node(kpage, stable_node); |
| 1086 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1087 | return stable_node; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1088 | } |
| 1089 | |
| 1090 | /* |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1091 | * unstable_tree_search_insert - search for identical page, |
| 1092 | * else insert rmap_item into the unstable tree. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1093 | * |
| 1094 | * This function searches for a page in the unstable tree identical to the |
| 1095 | * page currently being scanned; and if no identical page is found in the |
| 1096 | * tree, we insert rmap_item as a new object into the unstable tree. |
| 1097 | * |
| 1098 | * This function returns pointer to rmap_item found to be identical |
| 1099 | * to the currently scanned page, NULL otherwise. |
| 1100 | * |
| 1101 | * This function does both searching and inserting, because they share |
| 1102 | * the same walking algorithm in an rbtree. |
| 1103 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1104 | static |
| 1105 | struct rmap_item *unstable_tree_search_insert(struct rmap_item *rmap_item, |
| 1106 | struct page *page, |
| 1107 | struct page **tree_pagep) |
| 1108 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1109 | { |
| 1110 | struct rb_node **new = &root_unstable_tree.rb_node; |
| 1111 | struct rb_node *parent = NULL; |
| 1112 | |
| 1113 | while (*new) { |
| 1114 | struct rmap_item *tree_rmap_item; |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1115 | struct page *tree_page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1116 | int ret; |
| 1117 | |
Hugh Dickins | d178f27 | 2009-11-09 15:58:23 +0000 | [diff] [blame] | 1118 | cond_resched(); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1119 | tree_rmap_item = rb_entry(*new, struct rmap_item, node); |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1120 | tree_page = get_mergeable_page(tree_rmap_item); |
Dan Carpenter | 22eccdd | 2010-04-23 13:18:10 -0400 | [diff] [blame] | 1121 | if (IS_ERR_OR_NULL(tree_page)) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1122 | return NULL; |
| 1123 | |
| 1124 | /* |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1125 | * Don't substitute a ksm page for a forked page. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1126 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1127 | if (page == tree_page) { |
| 1128 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1129 | return NULL; |
| 1130 | } |
| 1131 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1132 | ret = memcmp_pages(page, tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1133 | |
| 1134 | parent = *new; |
| 1135 | if (ret < 0) { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1136 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1137 | new = &parent->rb_left; |
| 1138 | } else if (ret > 0) { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1139 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1140 | new = &parent->rb_right; |
| 1141 | } else { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1142 | *tree_pagep = tree_page; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1143 | return tree_rmap_item; |
| 1144 | } |
| 1145 | } |
| 1146 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1147 | rmap_item->address |= UNSTABLE_FLAG; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1148 | rmap_item->address |= (ksm_scan.seqnr & SEQNR_MASK); |
| 1149 | rb_link_node(&rmap_item->node, parent, new); |
| 1150 | rb_insert_color(&rmap_item->node, &root_unstable_tree); |
| 1151 | |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 1152 | ksm_pages_unshared++; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1153 | return NULL; |
| 1154 | } |
| 1155 | |
| 1156 | /* |
| 1157 | * stable_tree_append - add another rmap_item to the linked list of |
| 1158 | * rmap_items hanging off a given node of the stable tree, all sharing |
| 1159 | * the same ksm page. |
| 1160 | */ |
| 1161 | static void stable_tree_append(struct rmap_item *rmap_item, |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1162 | struct stable_node *stable_node) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1163 | { |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1164 | rmap_item->head = stable_node; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1165 | rmap_item->address |= STABLE_FLAG; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1166 | hlist_add_head(&rmap_item->hlist, &stable_node->hlist); |
Hugh Dickins | e178dfd | 2009-09-21 17:02:10 -0700 | [diff] [blame] | 1167 | |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1168 | if (rmap_item->hlist.next) |
| 1169 | ksm_pages_sharing++; |
| 1170 | else |
| 1171 | ksm_pages_shared++; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1172 | } |
| 1173 | |
| 1174 | /* |
Hugh Dickins | 81464e30 | 2009-09-21 17:02:15 -0700 | [diff] [blame] | 1175 | * cmp_and_merge_page - first see if page can be merged into the stable tree; |
| 1176 | * if not, compare checksum to previous and if it's the same, see if page can |
| 1177 | * be inserted into the unstable tree, or merged with a page already there and |
| 1178 | * both transferred to the stable tree. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1179 | * |
| 1180 | * @page: the page that we are searching identical page to. |
| 1181 | * @rmap_item: the reverse mapping into the virtual address of this page |
| 1182 | */ |
| 1183 | static void cmp_and_merge_page(struct page *page, struct rmap_item *rmap_item) |
| 1184 | { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1185 | struct rmap_item *tree_rmap_item; |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1186 | struct page *tree_page = NULL; |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1187 | struct stable_node *stable_node; |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1188 | struct page *kpage; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1189 | unsigned int checksum; |
| 1190 | int err; |
| 1191 | |
Hugh Dickins | 93d1771 | 2009-12-14 17:59:16 -0800 | [diff] [blame] | 1192 | remove_rmap_item_from_tree(rmap_item); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1193 | |
| 1194 | /* We first start with searching the page inside the stable tree */ |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1195 | kpage = stable_tree_search(page); |
| 1196 | if (kpage) { |
Hugh Dickins | 08beca4 | 2009-12-14 17:59:21 -0800 | [diff] [blame] | 1197 | err = try_to_merge_with_ksm_page(rmap_item, page, kpage); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1198 | if (!err) { |
| 1199 | /* |
| 1200 | * The page was successfully merged: |
| 1201 | * add its rmap_item to the stable tree. |
| 1202 | */ |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1203 | lock_page(kpage); |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1204 | stable_tree_append(rmap_item, page_stable_node(kpage)); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1205 | unlock_page(kpage); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1206 | } |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1207 | put_page(kpage); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1208 | return; |
| 1209 | } |
| 1210 | |
| 1211 | /* |
Hugh Dickins | 4035c07 | 2009-12-14 17:59:27 -0800 | [diff] [blame] | 1212 | * If the hash value of the page has changed from the last time |
| 1213 | * we calculated it, this page is changing frequently: therefore we |
| 1214 | * don't want to insert it in the unstable tree, and we don't want |
| 1215 | * to waste our time searching for something identical to it there. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1216 | */ |
| 1217 | checksum = calc_checksum(page); |
| 1218 | if (rmap_item->oldchecksum != checksum) { |
| 1219 | rmap_item->oldchecksum = checksum; |
| 1220 | return; |
| 1221 | } |
| 1222 | |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1223 | tree_rmap_item = |
| 1224 | unstable_tree_search_insert(rmap_item, page, &tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1225 | if (tree_rmap_item) { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1226 | kpage = try_to_merge_two_pages(rmap_item, page, |
| 1227 | tree_rmap_item, tree_page); |
| 1228 | put_page(tree_page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1229 | /* |
| 1230 | * As soon as we merge this page, we want to remove the |
| 1231 | * rmap_item of the page we have merged with from the unstable |
| 1232 | * tree, and insert it instead as new node in the stable tree. |
| 1233 | */ |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1234 | if (kpage) { |
Hugh Dickins | 93d1771 | 2009-12-14 17:59:16 -0800 | [diff] [blame] | 1235 | remove_rmap_item_from_tree(tree_rmap_item); |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 1236 | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1237 | lock_page(kpage); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1238 | stable_node = stable_tree_insert(kpage); |
| 1239 | if (stable_node) { |
| 1240 | stable_tree_append(tree_rmap_item, stable_node); |
| 1241 | stable_tree_append(rmap_item, stable_node); |
| 1242 | } |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1243 | unlock_page(kpage); |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1244 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1245 | /* |
| 1246 | * If we fail to insert the page into the stable tree, |
| 1247 | * we will have 2 virtual addresses that are pointing |
| 1248 | * to a ksm page left outside the stable tree, |
| 1249 | * in which case we need to break_cow on both. |
| 1250 | */ |
Hugh Dickins | 7b6ba2c | 2009-12-14 17:59:20 -0800 | [diff] [blame] | 1251 | if (!stable_node) { |
Hugh Dickins | 8dd3557 | 2009-12-14 17:59:18 -0800 | [diff] [blame] | 1252 | break_cow(tree_rmap_item); |
| 1253 | break_cow(rmap_item); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1254 | } |
| 1255 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1256 | } |
| 1257 | } |
| 1258 | |
| 1259 | static struct rmap_item *get_next_rmap_item(struct mm_slot *mm_slot, |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1260 | struct rmap_item **rmap_list, |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1261 | unsigned long addr) |
| 1262 | { |
| 1263 | struct rmap_item *rmap_item; |
| 1264 | |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1265 | while (*rmap_list) { |
| 1266 | rmap_item = *rmap_list; |
Hugh Dickins | 93d1771 | 2009-12-14 17:59:16 -0800 | [diff] [blame] | 1267 | if ((rmap_item->address & PAGE_MASK) == addr) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1268 | return rmap_item; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1269 | if (rmap_item->address > addr) |
| 1270 | break; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1271 | *rmap_list = rmap_item->rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1272 | remove_rmap_item_from_tree(rmap_item); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1273 | free_rmap_item(rmap_item); |
| 1274 | } |
| 1275 | |
| 1276 | rmap_item = alloc_rmap_item(); |
| 1277 | if (rmap_item) { |
| 1278 | /* It has already been zeroed */ |
| 1279 | rmap_item->mm = mm_slot->mm; |
| 1280 | rmap_item->address = addr; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1281 | rmap_item->rmap_list = *rmap_list; |
| 1282 | *rmap_list = rmap_item; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1283 | } |
| 1284 | return rmap_item; |
| 1285 | } |
| 1286 | |
| 1287 | static struct rmap_item *scan_get_next_rmap_item(struct page **page) |
| 1288 | { |
| 1289 | struct mm_struct *mm; |
| 1290 | struct mm_slot *slot; |
| 1291 | struct vm_area_struct *vma; |
| 1292 | struct rmap_item *rmap_item; |
| 1293 | |
| 1294 | if (list_empty(&ksm_mm_head.mm_list)) |
| 1295 | return NULL; |
| 1296 | |
| 1297 | slot = ksm_scan.mm_slot; |
| 1298 | if (slot == &ksm_mm_head) { |
Hugh Dickins | 2919bfd | 2011-01-13 15:47:29 -0800 | [diff] [blame] | 1299 | /* |
| 1300 | * A number of pages can hang around indefinitely on per-cpu |
| 1301 | * pagevecs, raised page count preventing write_protect_page |
| 1302 | * from merging them. Though it doesn't really matter much, |
| 1303 | * it is puzzling to see some stuck in pages_volatile until |
| 1304 | * other activity jostles them out, and they also prevented |
| 1305 | * LTP's KSM test from succeeding deterministically; so drain |
| 1306 | * them here (here rather than on entry to ksm_do_scan(), |
| 1307 | * so we don't IPI too often when pages_to_scan is set low). |
| 1308 | */ |
| 1309 | lru_add_drain_all(); |
| 1310 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1311 | root_unstable_tree = RB_ROOT; |
| 1312 | |
| 1313 | spin_lock(&ksm_mmlist_lock); |
| 1314 | slot = list_entry(slot->mm_list.next, struct mm_slot, mm_list); |
| 1315 | ksm_scan.mm_slot = slot; |
| 1316 | spin_unlock(&ksm_mmlist_lock); |
| 1317 | next_mm: |
| 1318 | ksm_scan.address = 0; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1319 | ksm_scan.rmap_list = &slot->rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1320 | } |
| 1321 | |
| 1322 | mm = slot->mm; |
| 1323 | down_read(&mm->mmap_sem); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1324 | if (ksm_test_exit(mm)) |
| 1325 | vma = NULL; |
| 1326 | else |
| 1327 | vma = find_vma(mm, ksm_scan.address); |
| 1328 | |
| 1329 | for (; vma; vma = vma->vm_next) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1330 | if (!(vma->vm_flags & VM_MERGEABLE)) |
| 1331 | continue; |
| 1332 | if (ksm_scan.address < vma->vm_start) |
| 1333 | ksm_scan.address = vma->vm_start; |
| 1334 | if (!vma->anon_vma) |
| 1335 | ksm_scan.address = vma->vm_end; |
| 1336 | |
| 1337 | while (ksm_scan.address < vma->vm_end) { |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1338 | if (ksm_test_exit(mm)) |
| 1339 | break; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1340 | *page = follow_page(vma, ksm_scan.address, FOLL_GET); |
Andrea Arcangeli | 21ae5b0 | 2011-01-13 15:47:00 -0800 | [diff] [blame] | 1341 | if (IS_ERR_OR_NULL(*page)) { |
| 1342 | ksm_scan.address += PAGE_SIZE; |
| 1343 | cond_resched(); |
| 1344 | continue; |
| 1345 | } |
Andrea Arcangeli | 29ad768 | 2011-01-13 15:47:19 -0800 | [diff] [blame] | 1346 | if (PageAnon(*page) || |
| 1347 | page_trans_compound_anon(*page)) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1348 | flush_anon_page(vma, *page, ksm_scan.address); |
| 1349 | flush_dcache_page(*page); |
| 1350 | rmap_item = get_next_rmap_item(slot, |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1351 | ksm_scan.rmap_list, ksm_scan.address); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1352 | if (rmap_item) { |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1353 | ksm_scan.rmap_list = |
| 1354 | &rmap_item->rmap_list; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1355 | ksm_scan.address += PAGE_SIZE; |
| 1356 | } else |
| 1357 | put_page(*page); |
| 1358 | up_read(&mm->mmap_sem); |
| 1359 | return rmap_item; |
| 1360 | } |
Andrea Arcangeli | 21ae5b0 | 2011-01-13 15:47:00 -0800 | [diff] [blame] | 1361 | put_page(*page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1362 | ksm_scan.address += PAGE_SIZE; |
| 1363 | cond_resched(); |
| 1364 | } |
| 1365 | } |
| 1366 | |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1367 | if (ksm_test_exit(mm)) { |
| 1368 | ksm_scan.address = 0; |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1369 | ksm_scan.rmap_list = &slot->rmap_list; |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1370 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1371 | /* |
| 1372 | * Nuke all the rmap_items that are above this current rmap: |
| 1373 | * because there were no VM_MERGEABLE vmas with such addresses. |
| 1374 | */ |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1375 | remove_trailing_rmap_items(slot, ksm_scan.rmap_list); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1376 | |
| 1377 | spin_lock(&ksm_mmlist_lock); |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1378 | ksm_scan.mm_slot = list_entry(slot->mm_list.next, |
| 1379 | struct mm_slot, mm_list); |
| 1380 | if (ksm_scan.address == 0) { |
| 1381 | /* |
| 1382 | * We've completed a full scan of all vmas, holding mmap_sem |
| 1383 | * throughout, and found no VM_MERGEABLE: so do the same as |
| 1384 | * __ksm_exit does to remove this mm from all our lists now. |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1385 | * This applies either when cleaning up after __ksm_exit |
| 1386 | * (but beware: we can reach here even before __ksm_exit), |
| 1387 | * or when all VM_MERGEABLE areas have been unmapped (and |
| 1388 | * mmap_sem then protects against race with MADV_MERGEABLE). |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1389 | */ |
| 1390 | hlist_del(&slot->link); |
| 1391 | list_del(&slot->mm_list); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1392 | spin_unlock(&ksm_mmlist_lock); |
| 1393 | |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1394 | free_mm_slot(slot); |
| 1395 | clear_bit(MMF_VM_MERGEABLE, &mm->flags); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1396 | up_read(&mm->mmap_sem); |
| 1397 | mmdrop(mm); |
| 1398 | } else { |
| 1399 | spin_unlock(&ksm_mmlist_lock); |
| 1400 | up_read(&mm->mmap_sem); |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1401 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1402 | |
| 1403 | /* Repeat until we've completed scanning the whole list */ |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1404 | slot = ksm_scan.mm_slot; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1405 | if (slot != &ksm_mm_head) |
| 1406 | goto next_mm; |
| 1407 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1408 | ksm_scan.seqnr++; |
| 1409 | return NULL; |
| 1410 | } |
| 1411 | |
| 1412 | /** |
| 1413 | * ksm_do_scan - the ksm scanner main worker function. |
| 1414 | * @scan_npages - number of pages we want to scan before we return. |
| 1415 | */ |
| 1416 | static void ksm_do_scan(unsigned int scan_npages) |
| 1417 | { |
| 1418 | struct rmap_item *rmap_item; |
Dan Carpenter | 22eccdd | 2010-04-23 13:18:10 -0400 | [diff] [blame] | 1419 | struct page *uninitialized_var(page); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1420 | |
Andrea Arcangeli | 878aee7 | 2011-01-13 15:47:10 -0800 | [diff] [blame] | 1421 | while (scan_npages-- && likely(!freezing(current))) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1422 | cond_resched(); |
| 1423 | rmap_item = scan_get_next_rmap_item(&page); |
| 1424 | if (!rmap_item) |
| 1425 | return; |
| 1426 | if (!PageKsm(page) || !in_stable_tree(rmap_item)) |
| 1427 | cmp_and_merge_page(page, rmap_item); |
| 1428 | put_page(page); |
| 1429 | } |
| 1430 | } |
| 1431 | |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1432 | static int ksmd_should_run(void) |
| 1433 | { |
| 1434 | return (ksm_run & KSM_RUN_MERGE) && !list_empty(&ksm_mm_head.mm_list); |
| 1435 | } |
| 1436 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1437 | static int ksm_scan_thread(void *nothing) |
| 1438 | { |
Andrea Arcangeli | 878aee7 | 2011-01-13 15:47:10 -0800 | [diff] [blame] | 1439 | set_freezable(); |
Izik Eidus | 339aa62 | 2009-09-21 17:02:07 -0700 | [diff] [blame] | 1440 | set_user_nice(current, 5); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1441 | |
| 1442 | while (!kthread_should_stop()) { |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1443 | mutex_lock(&ksm_thread_mutex); |
| 1444 | if (ksmd_should_run()) |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1445 | ksm_do_scan(ksm_thread_pages_to_scan); |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1446 | mutex_unlock(&ksm_thread_mutex); |
| 1447 | |
Andrea Arcangeli | 878aee7 | 2011-01-13 15:47:10 -0800 | [diff] [blame] | 1448 | try_to_freeze(); |
| 1449 | |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1450 | if (ksmd_should_run()) { |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1451 | schedule_timeout_interruptible( |
| 1452 | msecs_to_jiffies(ksm_thread_sleep_millisecs)); |
| 1453 | } else { |
Andrea Arcangeli | 878aee7 | 2011-01-13 15:47:10 -0800 | [diff] [blame] | 1454 | wait_event_freezable(ksm_thread_wait, |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1455 | ksmd_should_run() || kthread_should_stop()); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1456 | } |
| 1457 | } |
| 1458 | return 0; |
| 1459 | } |
| 1460 | |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1461 | int ksm_madvise(struct vm_area_struct *vma, unsigned long start, |
| 1462 | unsigned long end, int advice, unsigned long *vm_flags) |
| 1463 | { |
| 1464 | struct mm_struct *mm = vma->vm_mm; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1465 | int err; |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1466 | |
| 1467 | switch (advice) { |
| 1468 | case MADV_MERGEABLE: |
| 1469 | /* |
| 1470 | * Be somewhat over-protective for now! |
| 1471 | */ |
| 1472 | if (*vm_flags & (VM_MERGEABLE | VM_SHARED | VM_MAYSHARE | |
| 1473 | VM_PFNMAP | VM_IO | VM_DONTEXPAND | |
| 1474 | VM_RESERVED | VM_HUGETLB | VM_INSERTPAGE | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1475 | VM_NONLINEAR | VM_MIXEDMAP | VM_SAO)) |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1476 | return 0; /* just ignore the advice */ |
| 1477 | |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1478 | if (!test_bit(MMF_VM_MERGEABLE, &mm->flags)) { |
| 1479 | err = __ksm_enter(mm); |
| 1480 | if (err) |
| 1481 | return err; |
| 1482 | } |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1483 | |
| 1484 | *vm_flags |= VM_MERGEABLE; |
| 1485 | break; |
| 1486 | |
| 1487 | case MADV_UNMERGEABLE: |
| 1488 | if (!(*vm_flags & VM_MERGEABLE)) |
| 1489 | return 0; /* just ignore the advice */ |
| 1490 | |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1491 | if (vma->anon_vma) { |
| 1492 | err = unmerge_ksm_pages(vma, start, end); |
| 1493 | if (err) |
| 1494 | return err; |
| 1495 | } |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1496 | |
| 1497 | *vm_flags &= ~VM_MERGEABLE; |
| 1498 | break; |
| 1499 | } |
| 1500 | |
| 1501 | return 0; |
| 1502 | } |
| 1503 | |
| 1504 | int __ksm_enter(struct mm_struct *mm) |
| 1505 | { |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1506 | struct mm_slot *mm_slot; |
| 1507 | int needs_wakeup; |
| 1508 | |
| 1509 | mm_slot = alloc_mm_slot(); |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1510 | if (!mm_slot) |
| 1511 | return -ENOMEM; |
| 1512 | |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1513 | /* Check ksm_run too? Would need tighter locking */ |
| 1514 | needs_wakeup = list_empty(&ksm_mm_head.mm_list); |
| 1515 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1516 | spin_lock(&ksm_mmlist_lock); |
| 1517 | insert_to_mm_slots_hash(mm, mm_slot); |
| 1518 | /* |
| 1519 | * Insert just behind the scanning cursor, to let the area settle |
| 1520 | * down a little; when fork is followed by immediate exec, we don't |
| 1521 | * want ksmd to waste time setting up and tearing down an rmap_list. |
| 1522 | */ |
| 1523 | list_add_tail(&mm_slot->mm_list, &ksm_scan.mm_slot->mm_list); |
| 1524 | spin_unlock(&ksm_mmlist_lock); |
| 1525 | |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1526 | set_bit(MMF_VM_MERGEABLE, &mm->flags); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1527 | atomic_inc(&mm->mm_count); |
Hugh Dickins | 6e15838 | 2009-09-21 17:02:14 -0700 | [diff] [blame] | 1528 | |
| 1529 | if (needs_wakeup) |
| 1530 | wake_up_interruptible(&ksm_thread_wait); |
| 1531 | |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1532 | return 0; |
| 1533 | } |
| 1534 | |
Andrea Arcangeli | 1c2fb7a | 2009-09-21 17:02:22 -0700 | [diff] [blame] | 1535 | void __ksm_exit(struct mm_struct *mm) |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1536 | { |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1537 | struct mm_slot *mm_slot; |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1538 | int easy_to_free = 0; |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1539 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1540 | /* |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1541 | * This process is exiting: if it's straightforward (as is the |
| 1542 | * case when ksmd was never running), free mm_slot immediately. |
| 1543 | * But if it's at the cursor or has rmap_items linked to it, use |
| 1544 | * mmap_sem to synchronize with any break_cows before pagetables |
| 1545 | * are freed, and leave the mm_slot on the list for ksmd to free. |
| 1546 | * Beware: ksm may already have noticed it exiting and freed the slot. |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1547 | */ |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1548 | |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1549 | spin_lock(&ksm_mmlist_lock); |
| 1550 | mm_slot = get_mm_slot(mm); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1551 | if (mm_slot && ksm_scan.mm_slot != mm_slot) { |
Hugh Dickins | 6514d51 | 2009-12-14 17:59:19 -0800 | [diff] [blame] | 1552 | if (!mm_slot->rmap_list) { |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1553 | hlist_del(&mm_slot->link); |
| 1554 | list_del(&mm_slot->mm_list); |
| 1555 | easy_to_free = 1; |
| 1556 | } else { |
| 1557 | list_move(&mm_slot->mm_list, |
| 1558 | &ksm_scan.mm_slot->mm_list); |
| 1559 | } |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1560 | } |
Hugh Dickins | cd551f9 | 2009-09-21 17:02:17 -0700 | [diff] [blame] | 1561 | spin_unlock(&ksm_mmlist_lock); |
| 1562 | |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1563 | if (easy_to_free) { |
| 1564 | free_mm_slot(mm_slot); |
| 1565 | clear_bit(MMF_VM_MERGEABLE, &mm->flags); |
| 1566 | mmdrop(mm); |
| 1567 | } else if (mm_slot) { |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1568 | down_write(&mm->mmap_sem); |
| 1569 | up_write(&mm->mmap_sem); |
Hugh Dickins | 9ba6929 | 2009-09-21 17:02:20 -0700 | [diff] [blame] | 1570 | } |
Hugh Dickins | f8af4da | 2009-09-21 17:01:57 -0700 | [diff] [blame] | 1571 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1572 | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1573 | struct page *ksm_does_need_to_copy(struct page *page, |
| 1574 | struct vm_area_struct *vma, unsigned long address) |
| 1575 | { |
| 1576 | struct page *new_page; |
| 1577 | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1578 | new_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, address); |
| 1579 | if (new_page) { |
| 1580 | copy_user_highpage(new_page, page, address, vma); |
| 1581 | |
| 1582 | SetPageDirty(new_page); |
| 1583 | __SetPageUptodate(new_page); |
| 1584 | SetPageSwapBacked(new_page); |
| 1585 | __set_page_locked(new_page); |
| 1586 | |
| 1587 | if (page_evictable(new_page, vma)) |
| 1588 | lru_cache_add_lru(new_page, LRU_ACTIVE_ANON); |
| 1589 | else |
| 1590 | add_page_to_unevictable_list(new_page); |
| 1591 | } |
| 1592 | |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1593 | return new_page; |
| 1594 | } |
| 1595 | |
| 1596 | int page_referenced_ksm(struct page *page, struct mem_cgroup *memcg, |
| 1597 | unsigned long *vm_flags) |
| 1598 | { |
| 1599 | struct stable_node *stable_node; |
| 1600 | struct rmap_item *rmap_item; |
| 1601 | struct hlist_node *hlist; |
| 1602 | unsigned int mapcount = page_mapcount(page); |
| 1603 | int referenced = 0; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1604 | int search_new_forks = 0; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1605 | |
| 1606 | VM_BUG_ON(!PageKsm(page)); |
| 1607 | VM_BUG_ON(!PageLocked(page)); |
| 1608 | |
| 1609 | stable_node = page_stable_node(page); |
| 1610 | if (!stable_node) |
| 1611 | return 0; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1612 | again: |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1613 | hlist_for_each_entry(rmap_item, hlist, &stable_node->hlist, hlist) { |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1614 | struct anon_vma *anon_vma = rmap_item->anon_vma; |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1615 | struct anon_vma_chain *vmac; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1616 | struct vm_area_struct *vma; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1617 | |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1618 | anon_vma_lock(anon_vma); |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1619 | list_for_each_entry(vmac, &anon_vma->head, same_anon_vma) { |
| 1620 | vma = vmac->vma; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1621 | if (rmap_item->address < vma->vm_start || |
| 1622 | rmap_item->address >= vma->vm_end) |
| 1623 | continue; |
| 1624 | /* |
| 1625 | * Initially we examine only the vma which covers this |
| 1626 | * rmap_item; but later, if there is still work to do, |
| 1627 | * we examine covering vmas in other mms: in case they |
| 1628 | * were forked from the original since ksmd passed. |
| 1629 | */ |
| 1630 | if ((rmap_item->mm == vma->vm_mm) == search_new_forks) |
| 1631 | continue; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1632 | |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1633 | if (memcg && !mm_match_cgroup(vma->vm_mm, memcg)) |
| 1634 | continue; |
| 1635 | |
| 1636 | referenced += page_referenced_one(page, vma, |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1637 | rmap_item->address, &mapcount, vm_flags); |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1638 | if (!search_new_forks || !mapcount) |
| 1639 | break; |
| 1640 | } |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1641 | anon_vma_unlock(anon_vma); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1642 | if (!mapcount) |
| 1643 | goto out; |
| 1644 | } |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1645 | if (!search_new_forks++) |
| 1646 | goto again; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1647 | out: |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1648 | return referenced; |
| 1649 | } |
| 1650 | |
| 1651 | int try_to_unmap_ksm(struct page *page, enum ttu_flags flags) |
| 1652 | { |
| 1653 | struct stable_node *stable_node; |
| 1654 | struct hlist_node *hlist; |
| 1655 | struct rmap_item *rmap_item; |
| 1656 | int ret = SWAP_AGAIN; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1657 | int search_new_forks = 0; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1658 | |
| 1659 | VM_BUG_ON(!PageKsm(page)); |
| 1660 | VM_BUG_ON(!PageLocked(page)); |
| 1661 | |
| 1662 | stable_node = page_stable_node(page); |
| 1663 | if (!stable_node) |
| 1664 | return SWAP_FAIL; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1665 | again: |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1666 | hlist_for_each_entry(rmap_item, hlist, &stable_node->hlist, hlist) { |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1667 | struct anon_vma *anon_vma = rmap_item->anon_vma; |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1668 | struct anon_vma_chain *vmac; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1669 | struct vm_area_struct *vma; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1670 | |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1671 | anon_vma_lock(anon_vma); |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1672 | list_for_each_entry(vmac, &anon_vma->head, same_anon_vma) { |
| 1673 | vma = vmac->vma; |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1674 | if (rmap_item->address < vma->vm_start || |
| 1675 | rmap_item->address >= vma->vm_end) |
| 1676 | continue; |
| 1677 | /* |
| 1678 | * Initially we examine only the vma which covers this |
| 1679 | * rmap_item; but later, if there is still work to do, |
| 1680 | * we examine covering vmas in other mms: in case they |
| 1681 | * were forked from the original since ksmd passed. |
| 1682 | */ |
| 1683 | if ((rmap_item->mm == vma->vm_mm) == search_new_forks) |
| 1684 | continue; |
| 1685 | |
| 1686 | ret = try_to_unmap_one(page, vma, |
| 1687 | rmap_item->address, flags); |
| 1688 | if (ret != SWAP_AGAIN || !page_mapped(page)) { |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1689 | anon_vma_unlock(anon_vma); |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1690 | goto out; |
| 1691 | } |
| 1692 | } |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1693 | anon_vma_unlock(anon_vma); |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1694 | } |
Hugh Dickins | db114b8 | 2009-12-14 17:59:25 -0800 | [diff] [blame] | 1695 | if (!search_new_forks++) |
| 1696 | goto again; |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1697 | out: |
Hugh Dickins | 5ad6468 | 2009-12-14 17:59:24 -0800 | [diff] [blame] | 1698 | return ret; |
| 1699 | } |
| 1700 | |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1701 | #ifdef CONFIG_MIGRATION |
| 1702 | int rmap_walk_ksm(struct page *page, int (*rmap_one)(struct page *, |
| 1703 | struct vm_area_struct *, unsigned long, void *), void *arg) |
| 1704 | { |
| 1705 | struct stable_node *stable_node; |
| 1706 | struct hlist_node *hlist; |
| 1707 | struct rmap_item *rmap_item; |
| 1708 | int ret = SWAP_AGAIN; |
| 1709 | int search_new_forks = 0; |
| 1710 | |
| 1711 | VM_BUG_ON(!PageKsm(page)); |
| 1712 | VM_BUG_ON(!PageLocked(page)); |
| 1713 | |
| 1714 | stable_node = page_stable_node(page); |
| 1715 | if (!stable_node) |
| 1716 | return ret; |
| 1717 | again: |
| 1718 | hlist_for_each_entry(rmap_item, hlist, &stable_node->hlist, hlist) { |
| 1719 | struct anon_vma *anon_vma = rmap_item->anon_vma; |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1720 | struct anon_vma_chain *vmac; |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1721 | struct vm_area_struct *vma; |
| 1722 | |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1723 | anon_vma_lock(anon_vma); |
Rik van Riel | 5beb493 | 2010-03-05 13:42:07 -0800 | [diff] [blame] | 1724 | list_for_each_entry(vmac, &anon_vma->head, same_anon_vma) { |
| 1725 | vma = vmac->vma; |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1726 | if (rmap_item->address < vma->vm_start || |
| 1727 | rmap_item->address >= vma->vm_end) |
| 1728 | continue; |
| 1729 | /* |
| 1730 | * Initially we examine only the vma which covers this |
| 1731 | * rmap_item; but later, if there is still work to do, |
| 1732 | * we examine covering vmas in other mms: in case they |
| 1733 | * were forked from the original since ksmd passed. |
| 1734 | */ |
| 1735 | if ((rmap_item->mm == vma->vm_mm) == search_new_forks) |
| 1736 | continue; |
| 1737 | |
| 1738 | ret = rmap_one(page, vma, rmap_item->address, arg); |
| 1739 | if (ret != SWAP_AGAIN) { |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1740 | anon_vma_unlock(anon_vma); |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1741 | goto out; |
| 1742 | } |
| 1743 | } |
Rik van Riel | cba48b9 | 2010-08-09 17:18:38 -0700 | [diff] [blame] | 1744 | anon_vma_unlock(anon_vma); |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1745 | } |
| 1746 | if (!search_new_forks++) |
| 1747 | goto again; |
| 1748 | out: |
| 1749 | return ret; |
| 1750 | } |
| 1751 | |
| 1752 | void ksm_migrate_page(struct page *newpage, struct page *oldpage) |
| 1753 | { |
| 1754 | struct stable_node *stable_node; |
| 1755 | |
| 1756 | VM_BUG_ON(!PageLocked(oldpage)); |
| 1757 | VM_BUG_ON(!PageLocked(newpage)); |
| 1758 | VM_BUG_ON(newpage->mapping != oldpage->mapping); |
| 1759 | |
| 1760 | stable_node = page_stable_node(newpage); |
| 1761 | if (stable_node) { |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1762 | VM_BUG_ON(stable_node->kpfn != page_to_pfn(oldpage)); |
| 1763 | stable_node->kpfn = page_to_pfn(newpage); |
Hugh Dickins | e9995ef | 2009-12-14 17:59:31 -0800 | [diff] [blame] | 1764 | } |
| 1765 | } |
| 1766 | #endif /* CONFIG_MIGRATION */ |
| 1767 | |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1768 | #ifdef CONFIG_MEMORY_HOTREMOVE |
| 1769 | static struct stable_node *ksm_check_stable_tree(unsigned long start_pfn, |
| 1770 | unsigned long end_pfn) |
| 1771 | { |
| 1772 | struct rb_node *node; |
| 1773 | |
| 1774 | for (node = rb_first(&root_stable_tree); node; node = rb_next(node)) { |
| 1775 | struct stable_node *stable_node; |
| 1776 | |
| 1777 | stable_node = rb_entry(node, struct stable_node, node); |
| 1778 | if (stable_node->kpfn >= start_pfn && |
| 1779 | stable_node->kpfn < end_pfn) |
| 1780 | return stable_node; |
| 1781 | } |
| 1782 | return NULL; |
| 1783 | } |
| 1784 | |
| 1785 | static int ksm_memory_callback(struct notifier_block *self, |
| 1786 | unsigned long action, void *arg) |
| 1787 | { |
| 1788 | struct memory_notify *mn = arg; |
| 1789 | struct stable_node *stable_node; |
| 1790 | |
| 1791 | switch (action) { |
| 1792 | case MEM_GOING_OFFLINE: |
| 1793 | /* |
| 1794 | * Keep it very simple for now: just lock out ksmd and |
| 1795 | * MADV_UNMERGEABLE while any memory is going offline. |
KOSAKI Motohiro | a0b0f58 | 2010-12-02 14:31:20 -0800 | [diff] [blame] | 1796 | * mutex_lock_nested() is necessary because lockdep was alarmed |
| 1797 | * that here we take ksm_thread_mutex inside notifier chain |
| 1798 | * mutex, and later take notifier chain mutex inside |
| 1799 | * ksm_thread_mutex to unlock it. But that's safe because both |
| 1800 | * are inside mem_hotplug_mutex. |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1801 | */ |
KOSAKI Motohiro | a0b0f58 | 2010-12-02 14:31:20 -0800 | [diff] [blame] | 1802 | mutex_lock_nested(&ksm_thread_mutex, SINGLE_DEPTH_NESTING); |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 1803 | break; |
| 1804 | |
| 1805 | case MEM_OFFLINE: |
| 1806 | /* |
| 1807 | * Most of the work is done by page migration; but there might |
| 1808 | * be a few stable_nodes left over, still pointing to struct |
| 1809 | * pages which have been offlined: prune those from the tree. |
| 1810 | */ |
| 1811 | while ((stable_node = ksm_check_stable_tree(mn->start_pfn, |
| 1812 | mn->start_pfn + mn->nr_pages)) != NULL) |
| 1813 | remove_node_from_stable_tree(stable_node); |
| 1814 | /* fallthrough */ |
| 1815 | |
| 1816 | case MEM_CANCEL_OFFLINE: |
| 1817 | mutex_unlock(&ksm_thread_mutex); |
| 1818 | break; |
| 1819 | } |
| 1820 | return NOTIFY_OK; |
| 1821 | } |
| 1822 | #endif /* CONFIG_MEMORY_HOTREMOVE */ |
| 1823 | |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 1824 | #ifdef CONFIG_SYSFS |
| 1825 | /* |
| 1826 | * This all compiles without CONFIG_SYSFS, but is a waste of space. |
| 1827 | */ |
| 1828 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1829 | #define KSM_ATTR_RO(_name) \ |
| 1830 | static struct kobj_attribute _name##_attr = __ATTR_RO(_name) |
| 1831 | #define KSM_ATTR(_name) \ |
| 1832 | static struct kobj_attribute _name##_attr = \ |
| 1833 | __ATTR(_name, 0644, _name##_show, _name##_store) |
| 1834 | |
| 1835 | static ssize_t sleep_millisecs_show(struct kobject *kobj, |
| 1836 | struct kobj_attribute *attr, char *buf) |
| 1837 | { |
| 1838 | return sprintf(buf, "%u\n", ksm_thread_sleep_millisecs); |
| 1839 | } |
| 1840 | |
| 1841 | static ssize_t sleep_millisecs_store(struct kobject *kobj, |
| 1842 | struct kobj_attribute *attr, |
| 1843 | const char *buf, size_t count) |
| 1844 | { |
| 1845 | unsigned long msecs; |
| 1846 | int err; |
| 1847 | |
| 1848 | err = strict_strtoul(buf, 10, &msecs); |
| 1849 | if (err || msecs > UINT_MAX) |
| 1850 | return -EINVAL; |
| 1851 | |
| 1852 | ksm_thread_sleep_millisecs = msecs; |
| 1853 | |
| 1854 | return count; |
| 1855 | } |
| 1856 | KSM_ATTR(sleep_millisecs); |
| 1857 | |
| 1858 | static ssize_t pages_to_scan_show(struct kobject *kobj, |
| 1859 | struct kobj_attribute *attr, char *buf) |
| 1860 | { |
| 1861 | return sprintf(buf, "%u\n", ksm_thread_pages_to_scan); |
| 1862 | } |
| 1863 | |
| 1864 | static ssize_t pages_to_scan_store(struct kobject *kobj, |
| 1865 | struct kobj_attribute *attr, |
| 1866 | const char *buf, size_t count) |
| 1867 | { |
| 1868 | int err; |
| 1869 | unsigned long nr_pages; |
| 1870 | |
| 1871 | err = strict_strtoul(buf, 10, &nr_pages); |
| 1872 | if (err || nr_pages > UINT_MAX) |
| 1873 | return -EINVAL; |
| 1874 | |
| 1875 | ksm_thread_pages_to_scan = nr_pages; |
| 1876 | |
| 1877 | return count; |
| 1878 | } |
| 1879 | KSM_ATTR(pages_to_scan); |
| 1880 | |
| 1881 | static ssize_t run_show(struct kobject *kobj, struct kobj_attribute *attr, |
| 1882 | char *buf) |
| 1883 | { |
| 1884 | return sprintf(buf, "%u\n", ksm_run); |
| 1885 | } |
| 1886 | |
| 1887 | static ssize_t run_store(struct kobject *kobj, struct kobj_attribute *attr, |
| 1888 | const char *buf, size_t count) |
| 1889 | { |
| 1890 | int err; |
| 1891 | unsigned long flags; |
| 1892 | |
| 1893 | err = strict_strtoul(buf, 10, &flags); |
| 1894 | if (err || flags > UINT_MAX) |
| 1895 | return -EINVAL; |
| 1896 | if (flags > KSM_RUN_UNMERGE) |
| 1897 | return -EINVAL; |
| 1898 | |
| 1899 | /* |
| 1900 | * KSM_RUN_MERGE sets ksmd running, and 0 stops it running. |
| 1901 | * KSM_RUN_UNMERGE stops it running and unmerges all rmap_items, |
Hugh Dickins | d0f209f | 2009-12-14 17:59:34 -0800 | [diff] [blame] | 1902 | * breaking COW to free the pages_shared (but leaves mm_slots |
| 1903 | * on the list for when ksmd may be set running again). |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1904 | */ |
| 1905 | |
| 1906 | mutex_lock(&ksm_thread_mutex); |
| 1907 | if (ksm_run != flags) { |
| 1908 | ksm_run = flags; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1909 | if (flags & KSM_RUN_UNMERGE) { |
Hugh Dickins | 35451be | 2009-09-21 17:02:27 -0700 | [diff] [blame] | 1910 | current->flags |= PF_OOM_ORIGIN; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1911 | err = unmerge_and_remove_all_rmap_items(); |
Hugh Dickins | 35451be | 2009-09-21 17:02:27 -0700 | [diff] [blame] | 1912 | current->flags &= ~PF_OOM_ORIGIN; |
Hugh Dickins | d952b79 | 2009-09-21 17:02:16 -0700 | [diff] [blame] | 1913 | if (err) { |
| 1914 | ksm_run = KSM_RUN_STOP; |
| 1915 | count = err; |
| 1916 | } |
| 1917 | } |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1918 | } |
| 1919 | mutex_unlock(&ksm_thread_mutex); |
| 1920 | |
| 1921 | if (flags & KSM_RUN_MERGE) |
| 1922 | wake_up_interruptible(&ksm_thread_wait); |
| 1923 | |
| 1924 | return count; |
| 1925 | } |
| 1926 | KSM_ATTR(run); |
| 1927 | |
Hugh Dickins | b402826 | 2009-09-21 17:02:09 -0700 | [diff] [blame] | 1928 | static ssize_t pages_shared_show(struct kobject *kobj, |
| 1929 | struct kobj_attribute *attr, char *buf) |
| 1930 | { |
| 1931 | return sprintf(buf, "%lu\n", ksm_pages_shared); |
| 1932 | } |
| 1933 | KSM_ATTR_RO(pages_shared); |
| 1934 | |
| 1935 | static ssize_t pages_sharing_show(struct kobject *kobj, |
| 1936 | struct kobj_attribute *attr, char *buf) |
| 1937 | { |
Hugh Dickins | e178dfd | 2009-09-21 17:02:10 -0700 | [diff] [blame] | 1938 | return sprintf(buf, "%lu\n", ksm_pages_sharing); |
Hugh Dickins | b402826 | 2009-09-21 17:02:09 -0700 | [diff] [blame] | 1939 | } |
| 1940 | KSM_ATTR_RO(pages_sharing); |
| 1941 | |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 1942 | static ssize_t pages_unshared_show(struct kobject *kobj, |
| 1943 | struct kobj_attribute *attr, char *buf) |
| 1944 | { |
| 1945 | return sprintf(buf, "%lu\n", ksm_pages_unshared); |
| 1946 | } |
| 1947 | KSM_ATTR_RO(pages_unshared); |
| 1948 | |
| 1949 | static ssize_t pages_volatile_show(struct kobject *kobj, |
| 1950 | struct kobj_attribute *attr, char *buf) |
| 1951 | { |
| 1952 | long ksm_pages_volatile; |
| 1953 | |
| 1954 | ksm_pages_volatile = ksm_rmap_items - ksm_pages_shared |
| 1955 | - ksm_pages_sharing - ksm_pages_unshared; |
| 1956 | /* |
| 1957 | * It was not worth any locking to calculate that statistic, |
| 1958 | * but it might therefore sometimes be negative: conceal that. |
| 1959 | */ |
| 1960 | if (ksm_pages_volatile < 0) |
| 1961 | ksm_pages_volatile = 0; |
| 1962 | return sprintf(buf, "%ld\n", ksm_pages_volatile); |
| 1963 | } |
| 1964 | KSM_ATTR_RO(pages_volatile); |
| 1965 | |
| 1966 | static ssize_t full_scans_show(struct kobject *kobj, |
| 1967 | struct kobj_attribute *attr, char *buf) |
| 1968 | { |
| 1969 | return sprintf(buf, "%lu\n", ksm_scan.seqnr); |
| 1970 | } |
| 1971 | KSM_ATTR_RO(full_scans); |
| 1972 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1973 | static struct attribute *ksm_attrs[] = { |
| 1974 | &sleep_millisecs_attr.attr, |
| 1975 | &pages_to_scan_attr.attr, |
| 1976 | &run_attr.attr, |
Hugh Dickins | b402826 | 2009-09-21 17:02:09 -0700 | [diff] [blame] | 1977 | &pages_shared_attr.attr, |
| 1978 | &pages_sharing_attr.attr, |
Hugh Dickins | 473b0ce | 2009-09-21 17:02:11 -0700 | [diff] [blame] | 1979 | &pages_unshared_attr.attr, |
| 1980 | &pages_volatile_attr.attr, |
| 1981 | &full_scans_attr.attr, |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1982 | NULL, |
| 1983 | }; |
| 1984 | |
| 1985 | static struct attribute_group ksm_attr_group = { |
| 1986 | .attrs = ksm_attrs, |
| 1987 | .name = "ksm", |
| 1988 | }; |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 1989 | #endif /* CONFIG_SYSFS */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 1990 | |
| 1991 | static int __init ksm_init(void) |
| 1992 | { |
| 1993 | struct task_struct *ksm_thread; |
| 1994 | int err; |
| 1995 | |
| 1996 | err = ksm_slab_init(); |
| 1997 | if (err) |
| 1998 | goto out; |
| 1999 | |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2000 | ksm_thread = kthread_run(ksm_scan_thread, NULL, "ksmd"); |
| 2001 | if (IS_ERR(ksm_thread)) { |
| 2002 | printk(KERN_ERR "ksm: creating kthread failed\n"); |
| 2003 | err = PTR_ERR(ksm_thread); |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 2004 | goto out_free; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2005 | } |
| 2006 | |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 2007 | #ifdef CONFIG_SYSFS |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2008 | err = sysfs_create_group(mm_kobj, &ksm_attr_group); |
| 2009 | if (err) { |
| 2010 | printk(KERN_ERR "ksm: register sysfs failed\n"); |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 2011 | kthread_stop(ksm_thread); |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 2012 | goto out_free; |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2013 | } |
Hugh Dickins | c73602a | 2009-10-07 16:32:22 -0700 | [diff] [blame] | 2014 | #else |
| 2015 | ksm_run = KSM_RUN_MERGE; /* no way for user to start it */ |
| 2016 | |
Hugh Dickins | 2ffd867 | 2009-09-21 17:02:23 -0700 | [diff] [blame] | 2017 | #endif /* CONFIG_SYSFS */ |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2018 | |
Hugh Dickins | 62b61f6 | 2009-12-14 17:59:33 -0800 | [diff] [blame] | 2019 | #ifdef CONFIG_MEMORY_HOTREMOVE |
| 2020 | /* |
| 2021 | * Choose a high priority since the callback takes ksm_thread_mutex: |
| 2022 | * later callbacks could only be taking locks which nest within that. |
| 2023 | */ |
| 2024 | hotplug_memory_notifier(ksm_memory_callback, 100); |
| 2025 | #endif |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2026 | return 0; |
| 2027 | |
Lai Jiangshan | d9f8984 | 2010-08-09 17:20:02 -0700 | [diff] [blame] | 2028 | out_free: |
Izik Eidus | 31dbd01 | 2009-09-21 17:02:03 -0700 | [diff] [blame] | 2029 | ksm_slab_free(); |
| 2030 | out: |
| 2031 | return err; |
| 2032 | } |
| 2033 | module_init(ksm_init) |