Blame view
mm/zsmalloc.c
60.3 KB
61989a80f staging: zsmalloc... |
1 2 3 4 |
/* * zsmalloc memory allocator * * Copyright (C) 2011 Nitin Gupta |
31fc00bb7 zsmalloc: add cop... |
5 |
* Copyright (C) 2012, 2013 Minchan Kim |
61989a80f staging: zsmalloc... |
6 7 8 9 10 11 12 |
* * This code is released using a dual license strategy: BSD/GPL * You can choose the license that better fits your requirements. * * Released under the terms of 3-clause BSD License * Released under the terms of GNU General Public License Version 2.0 */ |
2db51dae5 staging: zsmalloc... |
13 |
/* |
2db51dae5 staging: zsmalloc... |
14 15 16 17 |
* Following is how we use various fields and flags of underlying * struct page(s) to form a zspage. * * Usage of struct page fields: |
3783689a1 zsmalloc: introdu... |
18 |
* page->private: points to zspage |
48b4800a1 zsmalloc: page mi... |
19 20 21 |
* page->freelist(index): links together all component pages of a zspage * For the huge page, this is always 0, so we use this field * to store handle. |
fd8544639 mm/zsmalloc: keep... |
22 |
* page->units: first object offset in a subpage of zspage |
2db51dae5 staging: zsmalloc... |
23 24 25 26 |
* * Usage of struct page flags: * PG_private: identifies the first component page * PG_private2: identifies the last component page |
48b4800a1 zsmalloc: page mi... |
27 |
* PG_owner_priv_1: indentifies the huge component page |
2db51dae5 staging: zsmalloc... |
28 29 |
* */ |
4abaac9b7 update "mm/zsmall... |
30 |
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt |
61989a80f staging: zsmalloc... |
31 32 |
#include <linux/module.h> #include <linux/kernel.h> |
312fcae22 zsmalloc: support... |
33 |
#include <linux/sched.h> |
61989a80f staging: zsmalloc... |
34 35 36 |
#include <linux/bitops.h> #include <linux/errno.h> #include <linux/highmem.h> |
61989a80f staging: zsmalloc... |
37 38 39 40 41 42 |
#include <linux/string.h> #include <linux/slab.h> #include <asm/tlbflush.h> #include <asm/pgtable.h> #include <linux/cpumask.h> #include <linux/cpu.h> |
0cbb613fa staging: fix powe... |
43 |
#include <linux/vmalloc.h> |
759b26b29 zsmalloc: use pre... |
44 |
#include <linux/preempt.h> |
0959c63f1 zsmalloc: collaps... |
45 46 |
#include <linux/spinlock.h> #include <linux/types.h> |
0f050d997 mm/zsmalloc: add ... |
47 |
#include <linux/debugfs.h> |
bcf1647d0 zsmalloc: move it... |
48 |
#include <linux/zsmalloc.h> |
c795779df mm/zpool: zbud/zs... |
49 |
#include <linux/zpool.h> |
48b4800a1 zsmalloc: page mi... |
50 |
#include <linux/mount.h> |
dd4123f32 mm: fix build war... |
51 |
#include <linux/migrate.h> |
48b4800a1 zsmalloc: page mi... |
52 53 54 |
#include <linux/pagemap.h> #define ZSPAGE_MAGIC 0x58 |
0959c63f1 zsmalloc: collaps... |
55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 |
/* * This must be power of 2 and greater than of equal to sizeof(link_free). * These two conditions ensure that any 'struct link_free' itself doesn't * span more than 1 page which avoids complex case of mapping 2 pages simply * to restore link_free pointer values. */ #define ZS_ALIGN 8 /* * A single 'zspage' is composed of up to 2^N discontiguous 0-order (single) * pages. ZS_MAX_ZSPAGE_ORDER defines upper limit on N. */ #define ZS_MAX_ZSPAGE_ORDER 2 #define ZS_MAX_PAGES_PER_ZSPAGE (_AC(1, UL) << ZS_MAX_ZSPAGE_ORDER) |
2e40e163a zsmalloc: decoupl... |
70 |
#define ZS_HANDLE_SIZE (sizeof(unsigned long)) |
0959c63f1 zsmalloc: collaps... |
71 72 |
/* * Object location (<PFN>, <obj_idx>) is encoded as |
c3e3e88ad zsmalloc: add mor... |
73 |
* as single (unsigned long) handle value. |
0959c63f1 zsmalloc: collaps... |
74 |
* |
bfd093f5e zsmalloc: use fre... |
75 |
* Note that object index <obj_idx> starts from 0. |
0959c63f1 zsmalloc: collaps... |
76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 |
* * This is made more complicated by various memory models and PAE. */ #ifndef MAX_PHYSMEM_BITS #ifdef CONFIG_HIGHMEM64G #define MAX_PHYSMEM_BITS 36 #else /* !CONFIG_HIGHMEM64G */ /* * If this definition of MAX_PHYSMEM_BITS is used, OBJ_INDEX_BITS will just * be PAGE_SHIFT */ #define MAX_PHYSMEM_BITS BITS_PER_LONG #endif #endif #define _PFN_BITS (MAX_PHYSMEM_BITS - PAGE_SHIFT) |
312fcae22 zsmalloc: support... |
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 |
/* * Memory for allocating for handle keeps object position by * encoding <page, obj_idx> and the encoded value has a room * in least bit(ie, look at obj_to_location). * We use the bit to synchronize between object access by * user and migration. */ #define HANDLE_PIN_BIT 0 /* * Head in allocated object should have OBJ_ALLOCATED_TAG * to identify the object was allocated or not. * It's okay to add the status bit in the least bit because * header keeps handle which is 4byte-aligned address so we * have room for two bit at least. */ #define OBJ_ALLOCATED_TAG 1 #define OBJ_TAG_BITS 1 #define OBJ_INDEX_BITS (BITS_PER_LONG - _PFN_BITS - OBJ_TAG_BITS) |
0959c63f1 zsmalloc: collaps... |
112 113 114 115 116 117 |
#define OBJ_INDEX_MASK ((_AC(1, UL) << OBJ_INDEX_BITS) - 1) #define MAX(a, b) ((a) >= (b) ? (a) : (b)) /* ZS_MIN_ALLOC_SIZE must be multiple of ZS_ALIGN */ #define ZS_MIN_ALLOC_SIZE \ MAX(32, (ZS_MAX_PAGES_PER_ZSPAGE << PAGE_SHIFT >> OBJ_INDEX_BITS)) |
2e40e163a zsmalloc: decoupl... |
118 |
/* each chunk includes extra space to keep handle */ |
7b60a6852 zsmalloc: record ... |
119 |
#define ZS_MAX_ALLOC_SIZE PAGE_SIZE |
0959c63f1 zsmalloc: collaps... |
120 121 |
/* |
7eb52512a zsmalloc: fixup t... |
122 |
* On systems with 4K page size, this gives 255 size classes! There is a |
0959c63f1 zsmalloc: collaps... |
123 124 125 126 127 128 129 130 131 132 133 |
* trader-off here: * - Large number of size classes is potentially wasteful as free page are * spread across these classes * - Small number of size classes causes large internal fragmentation * - Probably its better to use specific size classes (empirically * determined). NOTE: all those class sizes must be set as multiple of * ZS_ALIGN to make sure link_free itself never has to span 2 pages. * * ZS_MIN_ALLOC_SIZE and ZS_SIZE_CLASS_DELTA must be multiple of ZS_ALIGN * (reason above) */ |
3783689a1 zsmalloc: introdu... |
134 |
#define ZS_SIZE_CLASS_DELTA (PAGE_SIZE >> CLASS_BITS) |
0959c63f1 zsmalloc: collaps... |
135 |
|
0959c63f1 zsmalloc: collaps... |
136 |
enum fullness_group { |
0959c63f1 zsmalloc: collaps... |
137 |
ZS_EMPTY, |
48b4800a1 zsmalloc: page mi... |
138 139 140 141 |
ZS_ALMOST_EMPTY, ZS_ALMOST_FULL, ZS_FULL, NR_ZS_FULLNESS, |
0959c63f1 zsmalloc: collaps... |
142 |
}; |
0f050d997 mm/zsmalloc: add ... |
143 |
enum zs_stat_type { |
48b4800a1 zsmalloc: page mi... |
144 145 146 147 |
CLASS_EMPTY, CLASS_ALMOST_EMPTY, CLASS_ALMOST_FULL, CLASS_FULL, |
0f050d997 mm/zsmalloc: add ... |
148 149 |
OBJ_ALLOCATED, OBJ_USED, |
48b4800a1 zsmalloc: page mi... |
150 |
NR_ZS_STAT_TYPE, |
0f050d997 mm/zsmalloc: add ... |
151 |
}; |
0f050d997 mm/zsmalloc: add ... |
152 153 154 |
struct zs_size_stat { unsigned long objs[NR_ZS_STAT_TYPE]; }; |
572445941 zsmalloc: always ... |
155 156 |
#ifdef CONFIG_ZSMALLOC_STAT static struct dentry *zs_stat_root; |
0f050d997 mm/zsmalloc: add ... |
157 |
#endif |
48b4800a1 zsmalloc: page mi... |
158 159 160 |
#ifdef CONFIG_COMPACTION static struct vfsmount *zsmalloc_mnt; #endif |
0959c63f1 zsmalloc: collaps... |
161 |
/* |
40f9fb8cf mm/zsmalloc: supp... |
162 163 164 165 166 |
* number of size_classes */ static int zs_size_classes; /* |
0959c63f1 zsmalloc: collaps... |
167 168 169 170 |
* We assign a page to ZS_ALMOST_EMPTY fullness group when: * n <= N / f, where * n = number of allocated objects * N = total number of objects zspage can store |
6dd9737e3 mm/zsmalloc.c: co... |
171 |
* f = fullness_threshold_frac |
0959c63f1 zsmalloc: collaps... |
172 173 174 175 176 177 178 179 180 181 182 |
* * Similarly, we assign zspage to: * ZS_ALMOST_FULL when n > N / f * ZS_EMPTY when n == 0 * ZS_FULL when n == N * * (see: fix_fullness_group()) */ static const int fullness_threshold_frac = 4; struct size_class { |
572445941 zsmalloc: always ... |
183 |
spinlock_t lock; |
48b4800a1 zsmalloc: page mi... |
184 |
struct list_head fullness_list[NR_ZS_FULLNESS]; |
0959c63f1 zsmalloc: collaps... |
185 186 187 188 189 |
/* * Size of objects stored in this class. Must be multiple * of ZS_ALIGN. */ int size; |
1fc6e27d7 zsmalloc: keep ma... |
190 |
int objs_per_zspage; |
7dfa46122 zsmalloc: reorgan... |
191 192 |
/* Number of PAGE_SIZE sized pages to combine to form a 'zspage' */ int pages_per_zspage; |
48b4800a1 zsmalloc: page mi... |
193 194 195 |
unsigned int index; struct zs_size_stat stats; |
0959c63f1 zsmalloc: collaps... |
196 |
}; |
48b4800a1 zsmalloc: page mi... |
197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
/* huge object: pages_per_zspage == 1 && maxobj_per_zspage == 1 */ static void SetPageHugeObject(struct page *page) { SetPageOwnerPriv1(page); } static void ClearPageHugeObject(struct page *page) { ClearPageOwnerPriv1(page); } static int PageHugeObject(struct page *page) { return PageOwnerPriv1(page); } |
0959c63f1 zsmalloc: collaps... |
212 213 |
/* * Placed within free objects to form a singly linked list. |
3783689a1 zsmalloc: introdu... |
214 |
* For every zspage, zspage->freeobj gives head of this list. |
0959c63f1 zsmalloc: collaps... |
215 216 217 218 |
* * This must be power of 2 and less than or equal to ZS_ALIGN */ struct link_free { |
2e40e163a zsmalloc: decoupl... |
219 220 |
union { /* |
bfd093f5e zsmalloc: use fre... |
221 |
* Free object index; |
2e40e163a zsmalloc: decoupl... |
222 223 |
* It's valid for non-allocated object */ |
bfd093f5e zsmalloc: use fre... |
224 |
unsigned long next; |
2e40e163a zsmalloc: decoupl... |
225 226 227 228 229 |
/* * Handle of allocated object. */ unsigned long handle; }; |
0959c63f1 zsmalloc: collaps... |
230 231 232 |
}; struct zs_pool { |
6f3526d6d mm: zsmalloc: con... |
233 |
const char *name; |
0f050d997 mm/zsmalloc: add ... |
234 |
|
40f9fb8cf mm/zsmalloc: supp... |
235 |
struct size_class **size_class; |
2e40e163a zsmalloc: decoupl... |
236 |
struct kmem_cache *handle_cachep; |
3783689a1 zsmalloc: introdu... |
237 |
struct kmem_cache *zspage_cachep; |
0959c63f1 zsmalloc: collaps... |
238 |
|
13de8933c zsmalloc: move pa... |
239 |
atomic_long_t pages_allocated; |
0f050d997 mm/zsmalloc: add ... |
240 |
|
7d3f39382 zsmalloc/zram: in... |
241 |
struct zs_pool_stats stats; |
ab9d306d9 zsmalloc: use shr... |
242 243 244 245 246 247 248 249 |
/* Compact classes */ struct shrinker shrinker; /* * To signify that register_shrinker() was successful * and unregister_shrinker() will not Oops. */ bool shrinker_enabled; |
0f050d997 mm/zsmalloc: add ... |
250 251 252 |
#ifdef CONFIG_ZSMALLOC_STAT struct dentry *stat_dentry; #endif |
48b4800a1 zsmalloc: page mi... |
253 254 255 256 |
#ifdef CONFIG_COMPACTION struct inode *inode; struct work_struct free_work; #endif |
0959c63f1 zsmalloc: collaps... |
257 |
}; |
61989a80f staging: zsmalloc... |
258 259 260 261 262 |
/* * A zspage's class index and fullness group * are encoded in its (first)page->mapping */ |
3783689a1 zsmalloc: introdu... |
263 264 |
#define FULLNESS_BITS 2 #define CLASS_BITS 8 |
48b4800a1 zsmalloc: page mi... |
265 266 |
#define ISOLATED_BITS 3 #define MAGIC_VAL_BITS 8 |
4f42047bb zsmalloc: use acc... |
267 |
|
3783689a1 zsmalloc: introdu... |
268 269 270 271 |
struct zspage { struct { unsigned int fullness:FULLNESS_BITS; unsigned int class:CLASS_BITS; |
48b4800a1 zsmalloc: page mi... |
272 273 |
unsigned int isolated:ISOLATED_BITS; unsigned int magic:MAGIC_VAL_BITS; |
3783689a1 zsmalloc: introdu... |
274 275 |
}; unsigned int inuse; |
bfd093f5e zsmalloc: use fre... |
276 |
unsigned int freeobj; |
3783689a1 zsmalloc: introdu... |
277 278 |
struct page *first_page; struct list_head list; /* fullness list */ |
48b4800a1 zsmalloc: page mi... |
279 280 281 |
#ifdef CONFIG_COMPACTION rwlock_t lock; #endif |
3783689a1 zsmalloc: introdu... |
282 |
}; |
61989a80f staging: zsmalloc... |
283 |
|
f553646a6 staging: zsmalloc... |
284 |
struct mapping_area { |
1b945aeef zsmalloc: add Kco... |
285 |
#ifdef CONFIG_PGTABLE_MAPPING |
f553646a6 staging: zsmalloc... |
286 287 288 289 290 291 292 |
struct vm_struct *vm; /* vm area for mapping object that span pages */ #else char *vm_buf; /* copy buffer for objects that span pages */ #endif char *vm_addr; /* address of kmap_atomic()'ed pages */ enum zs_mapmode vm_mm; /* mapping mode */ }; |
48b4800a1 zsmalloc: page mi... |
293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 |
#ifdef CONFIG_COMPACTION static int zs_register_migration(struct zs_pool *pool); static void zs_unregister_migration(struct zs_pool *pool); static void migrate_lock_init(struct zspage *zspage); static void migrate_read_lock(struct zspage *zspage); static void migrate_read_unlock(struct zspage *zspage); static void kick_deferred_free(struct zs_pool *pool); static void init_deferred_free(struct zs_pool *pool); static void SetZsPageMovable(struct zs_pool *pool, struct zspage *zspage); #else static int zsmalloc_mount(void) { return 0; } static void zsmalloc_unmount(void) {} static int zs_register_migration(struct zs_pool *pool) { return 0; } static void zs_unregister_migration(struct zs_pool *pool) {} static void migrate_lock_init(struct zspage *zspage) {} static void migrate_read_lock(struct zspage *zspage) {} static void migrate_read_unlock(struct zspage *zspage) {} static void kick_deferred_free(struct zs_pool *pool) {} static void init_deferred_free(struct zs_pool *pool) {} static void SetZsPageMovable(struct zs_pool *pool, struct zspage *zspage) {} #endif |
3783689a1 zsmalloc: introdu... |
314 |
static int create_cache(struct zs_pool *pool) |
2e40e163a zsmalloc: decoupl... |
315 316 317 |
{ pool->handle_cachep = kmem_cache_create("zs_handle", ZS_HANDLE_SIZE, 0, 0, NULL); |
3783689a1 zsmalloc: introdu... |
318 319 320 321 322 323 324 325 326 327 328 329 |
if (!pool->handle_cachep) return 1; pool->zspage_cachep = kmem_cache_create("zspage", sizeof(struct zspage), 0, 0, NULL); if (!pool->zspage_cachep) { kmem_cache_destroy(pool->handle_cachep); pool->handle_cachep = NULL; return 1; } return 0; |
2e40e163a zsmalloc: decoupl... |
330 |
} |
3783689a1 zsmalloc: introdu... |
331 |
static void destroy_cache(struct zs_pool *pool) |
2e40e163a zsmalloc: decoupl... |
332 |
{ |
cd10add00 zsmalloc: remove ... |
333 |
kmem_cache_destroy(pool->handle_cachep); |
3783689a1 zsmalloc: introdu... |
334 |
kmem_cache_destroy(pool->zspage_cachep); |
2e40e163a zsmalloc: decoupl... |
335 |
} |
3783689a1 zsmalloc: introdu... |
336 |
static unsigned long cache_alloc_handle(struct zs_pool *pool, gfp_t gfp) |
2e40e163a zsmalloc: decoupl... |
337 338 |
{ return (unsigned long)kmem_cache_alloc(pool->handle_cachep, |
48b4800a1 zsmalloc: page mi... |
339 |
gfp & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); |
2e40e163a zsmalloc: decoupl... |
340 |
} |
3783689a1 zsmalloc: introdu... |
341 |
static void cache_free_handle(struct zs_pool *pool, unsigned long handle) |
2e40e163a zsmalloc: decoupl... |
342 343 344 |
{ kmem_cache_free(pool->handle_cachep, (void *)handle); } |
3783689a1 zsmalloc: introdu... |
345 346 |
static struct zspage *cache_alloc_zspage(struct zs_pool *pool, gfp_t flags) { |
48b4800a1 zsmalloc: page mi... |
347 348 |
return kmem_cache_alloc(pool->zspage_cachep, flags & ~(__GFP_HIGHMEM|__GFP_MOVABLE)); |
3783689a1 zsmalloc: introdu... |
349 350 351 352 353 354 |
}; static void cache_free_zspage(struct zs_pool *pool, struct zspage *zspage) { kmem_cache_free(pool->zspage_cachep, zspage); } |
2e40e163a zsmalloc: decoupl... |
355 356 |
static void record_obj(unsigned long handle, unsigned long obj) { |
c102f07ca zsmalloc: fix mig... |
357 358 359 360 361 362 |
/* * lsb of @obj represents handle lock while other bits * represent object value the handle is pointing so * updating shouldn't do store tearing. */ WRITE_ONCE(*(unsigned long *)handle, obj); |
2e40e163a zsmalloc: decoupl... |
363 |
} |
c795779df mm/zpool: zbud/zs... |
364 365 366 |
/* zpool driver */ #ifdef CONFIG_ZPOOL |
6f3526d6d mm: zsmalloc: con... |
367 |
static void *zs_zpool_create(const char *name, gfp_t gfp, |
786727799 mm: zpool: consti... |
368 |
const struct zpool_ops *zpool_ops, |
479305fd7 zpool: remove zpo... |
369 |
struct zpool *zpool) |
c795779df mm/zpool: zbud/zs... |
370 |
{ |
d0d8da2dc zsmalloc: require... |
371 372 373 374 375 376 |
/* * Ignore global gfp flags: zs_malloc() may be invoked from * different contexts and its caller must provide a valid * gfp mask. */ return zs_create_pool(name); |
c795779df mm/zpool: zbud/zs... |
377 378 379 380 381 382 383 384 385 386 |
} static void zs_zpool_destroy(void *pool) { zs_destroy_pool(pool); } static int zs_zpool_malloc(void *pool, size_t size, gfp_t gfp, unsigned long *handle) { |
d0d8da2dc zsmalloc: require... |
387 |
*handle = zs_malloc(pool, size, gfp); |
c795779df mm/zpool: zbud/zs... |
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 |
return *handle ? 0 : -1; } static void zs_zpool_free(void *pool, unsigned long handle) { zs_free(pool, handle); } static int zs_zpool_shrink(void *pool, unsigned int pages, unsigned int *reclaimed) { return -EINVAL; } static void *zs_zpool_map(void *pool, unsigned long handle, enum zpool_mapmode mm) { enum zs_mapmode zs_mm; switch (mm) { case ZPOOL_MM_RO: zs_mm = ZS_MM_RO; break; case ZPOOL_MM_WO: zs_mm = ZS_MM_WO; break; case ZPOOL_MM_RW: /* fallthru */ default: zs_mm = ZS_MM_RW; break; } return zs_map_object(pool, handle, zs_mm); } static void zs_zpool_unmap(void *pool, unsigned long handle) { zs_unmap_object(pool, handle); } static u64 zs_zpool_total_size(void *pool) { |
722cdc172 zsmalloc: change ... |
428 |
return zs_get_total_pages(pool) << PAGE_SHIFT; |
c795779df mm/zpool: zbud/zs... |
429 430 431 432 433 434 435 436 437 438 439 440 441 442 |
} static struct zpool_driver zs_zpool_driver = { .type = "zsmalloc", .owner = THIS_MODULE, .create = zs_zpool_create, .destroy = zs_zpool_destroy, .malloc = zs_zpool_malloc, .free = zs_zpool_free, .shrink = zs_zpool_shrink, .map = zs_zpool_map, .unmap = zs_zpool_unmap, .total_size = zs_zpool_total_size, }; |
137f8cff5 mm/zpool: use pre... |
443 |
MODULE_ALIAS("zpool-zsmalloc"); |
c795779df mm/zpool: zbud/zs... |
444 |
#endif /* CONFIG_ZPOOL */ |
61989a80f staging: zsmalloc... |
445 446 |
/* per-cpu VM mapping areas for zspage accesses that cross page boundaries */ static DEFINE_PER_CPU(struct mapping_area, zs_map_area); |
48b4800a1 zsmalloc: page mi... |
447 448 449 450 |
static bool is_zspage_isolated(struct zspage *zspage) { return zspage->isolated; } |
61989a80f staging: zsmalloc... |
451 452 |
static int is_first_page(struct page *page) { |
a27545bf0 zsmalloc: use Pag... |
453 |
return PagePrivate(page); |
61989a80f staging: zsmalloc... |
454 |
} |
48b4800a1 zsmalloc: page mi... |
455 |
/* Protected by class->lock */ |
3783689a1 zsmalloc: introdu... |
456 |
static inline int get_zspage_inuse(struct zspage *zspage) |
4f42047bb zsmalloc: use acc... |
457 |
{ |
3783689a1 zsmalloc: introdu... |
458 |
return zspage->inuse; |
4f42047bb zsmalloc: use acc... |
459 |
} |
3783689a1 zsmalloc: introdu... |
460 |
static inline void set_zspage_inuse(struct zspage *zspage, int val) |
4f42047bb zsmalloc: use acc... |
461 |
{ |
3783689a1 zsmalloc: introdu... |
462 |
zspage->inuse = val; |
4f42047bb zsmalloc: use acc... |
463 |
} |
3783689a1 zsmalloc: introdu... |
464 |
static inline void mod_zspage_inuse(struct zspage *zspage, int val) |
4f42047bb zsmalloc: use acc... |
465 |
{ |
3783689a1 zsmalloc: introdu... |
466 |
zspage->inuse += val; |
4f42047bb zsmalloc: use acc... |
467 |
} |
48b4800a1 zsmalloc: page mi... |
468 |
static inline struct page *get_first_page(struct zspage *zspage) |
4f42047bb zsmalloc: use acc... |
469 |
{ |
48b4800a1 zsmalloc: page mi... |
470 |
struct page *first_page = zspage->first_page; |
3783689a1 zsmalloc: introdu... |
471 |
|
48b4800a1 zsmalloc: page mi... |
472 473 |
VM_BUG_ON_PAGE(!is_first_page(first_page), first_page); return first_page; |
4f42047bb zsmalloc: use acc... |
474 |
} |
48b4800a1 zsmalloc: page mi... |
475 |
static inline int get_first_obj_offset(struct page *page) |
4f42047bb zsmalloc: use acc... |
476 |
{ |
48b4800a1 zsmalloc: page mi... |
477 478 |
return page->units; } |
3783689a1 zsmalloc: introdu... |
479 |
|
48b4800a1 zsmalloc: page mi... |
480 481 482 |
static inline void set_first_obj_offset(struct page *page, int offset) { page->units = offset; |
4f42047bb zsmalloc: use acc... |
483 |
} |
bfd093f5e zsmalloc: use fre... |
484 |
static inline unsigned int get_freeobj(struct zspage *zspage) |
4f42047bb zsmalloc: use acc... |
485 |
{ |
bfd093f5e zsmalloc: use fre... |
486 |
return zspage->freeobj; |
4f42047bb zsmalloc: use acc... |
487 |
} |
bfd093f5e zsmalloc: use fre... |
488 |
static inline void set_freeobj(struct zspage *zspage, unsigned int obj) |
4f42047bb zsmalloc: use acc... |
489 |
{ |
bfd093f5e zsmalloc: use fre... |
490 |
zspage->freeobj = obj; |
4f42047bb zsmalloc: use acc... |
491 |
} |
3783689a1 zsmalloc: introdu... |
492 |
static void get_zspage_mapping(struct zspage *zspage, |
a42094676 zsmalloc: use fir... |
493 |
unsigned int *class_idx, |
61989a80f staging: zsmalloc... |
494 495 |
enum fullness_group *fullness) { |
48b4800a1 zsmalloc: page mi... |
496 |
BUG_ON(zspage->magic != ZSPAGE_MAGIC); |
3783689a1 zsmalloc: introdu... |
497 498 |
*fullness = zspage->fullness; *class_idx = zspage->class; |
61989a80f staging: zsmalloc... |
499 |
} |
3783689a1 zsmalloc: introdu... |
500 |
static void set_zspage_mapping(struct zspage *zspage, |
a42094676 zsmalloc: use fir... |
501 |
unsigned int class_idx, |
61989a80f staging: zsmalloc... |
502 503 |
enum fullness_group fullness) { |
3783689a1 zsmalloc: introdu... |
504 505 |
zspage->class = class_idx; zspage->fullness = fullness; |
61989a80f staging: zsmalloc... |
506 |
} |
c3e3e88ad zsmalloc: add mor... |
507 508 509 510 511 512 513 |
/* * zsmalloc divides the pool into various size classes where each * class maintains a list of zspages where each zspage is divided * into equal sized chunks. Each allocation falls into one of these * classes depending on its size. This function returns index of the * size class which has chunk size big enough to hold the give size. */ |
61989a80f staging: zsmalloc... |
514 515 516 517 518 519 520 |
static int get_size_class_index(int size) { int idx = 0; if (likely(size > ZS_MIN_ALLOC_SIZE)) idx = DIV_ROUND_UP(size - ZS_MIN_ALLOC_SIZE, ZS_SIZE_CLASS_DELTA); |
7b60a6852 zsmalloc: record ... |
521 |
return min(zs_size_classes - 1, idx); |
61989a80f staging: zsmalloc... |
522 |
} |
248ca1b05 zsmalloc: add ful... |
523 524 525 |
static inline void zs_stat_inc(struct size_class *class, enum zs_stat_type type, unsigned long cnt) { |
48b4800a1 zsmalloc: page mi... |
526 |
class->stats.objs[type] += cnt; |
248ca1b05 zsmalloc: add ful... |
527 528 529 530 531 |
} static inline void zs_stat_dec(struct size_class *class, enum zs_stat_type type, unsigned long cnt) { |
48b4800a1 zsmalloc: page mi... |
532 |
class->stats.objs[type] -= cnt; |
248ca1b05 zsmalloc: add ful... |
533 534 535 536 537 |
} static inline unsigned long zs_stat_get(struct size_class *class, enum zs_stat_type type) { |
48b4800a1 zsmalloc: page mi... |
538 |
return class->stats.objs[type]; |
248ca1b05 zsmalloc: add ful... |
539 |
} |
572445941 zsmalloc: always ... |
540 |
#ifdef CONFIG_ZSMALLOC_STAT |
4abaac9b7 update "mm/zsmall... |
541 |
static void __init zs_stat_init(void) |
248ca1b05 zsmalloc: add ful... |
542 |
{ |
4abaac9b7 update "mm/zsmall... |
543 544 545 546 547 |
if (!debugfs_initialized()) { pr_warn("debugfs not available, stat dir not created "); return; } |
248ca1b05 zsmalloc: add ful... |
548 549 550 |
zs_stat_root = debugfs_create_dir("zsmalloc", NULL); if (!zs_stat_root) |
4abaac9b7 update "mm/zsmall... |
551 552 |
pr_warn("debugfs 'zsmalloc' stat dir creation failed "); |
248ca1b05 zsmalloc: add ful... |
553 554 555 556 557 558 |
} static void __exit zs_stat_exit(void) { debugfs_remove_recursive(zs_stat_root); } |
1120ed548 mm/zsmalloc: add ... |
559 |
static unsigned long zs_can_compact(struct size_class *class); |
248ca1b05 zsmalloc: add ful... |
560 561 562 563 564 565 566 |
static int zs_stats_size_show(struct seq_file *s, void *v) { int i; struct zs_pool *pool = s->private; struct size_class *class; int objs_per_zspage; unsigned long class_almost_full, class_almost_empty; |
1120ed548 mm/zsmalloc: add ... |
567 |
unsigned long obj_allocated, obj_used, pages_used, freeable; |
248ca1b05 zsmalloc: add ful... |
568 569 |
unsigned long total_class_almost_full = 0, total_class_almost_empty = 0; unsigned long total_objs = 0, total_used_objs = 0, total_pages = 0; |
1120ed548 mm/zsmalloc: add ... |
570 |
unsigned long total_freeable = 0; |
248ca1b05 zsmalloc: add ful... |
571 |
|
1120ed548 mm/zsmalloc: add ... |
572 573 |
seq_printf(s, " %5s %5s %11s %12s %13s %10s %10s %16s %8s ", |
248ca1b05 zsmalloc: add ful... |
574 575 |
"class", "size", "almost_full", "almost_empty", "obj_allocated", "obj_used", "pages_used", |
1120ed548 mm/zsmalloc: add ... |
576 |
"pages_per_zspage", "freeable"); |
248ca1b05 zsmalloc: add ful... |
577 578 579 580 581 582 583 584 585 586 587 588 |
for (i = 0; i < zs_size_classes; i++) { class = pool->size_class[i]; if (class->index != i) continue; spin_lock(&class->lock); class_almost_full = zs_stat_get(class, CLASS_ALMOST_FULL); class_almost_empty = zs_stat_get(class, CLASS_ALMOST_EMPTY); obj_allocated = zs_stat_get(class, OBJ_ALLOCATED); obj_used = zs_stat_get(class, OBJ_USED); |
1120ed548 mm/zsmalloc: add ... |
589 |
freeable = zs_can_compact(class); |
248ca1b05 zsmalloc: add ful... |
590 |
spin_unlock(&class->lock); |
b4fd07a08 mm/zsmalloc: use ... |
591 |
objs_per_zspage = class->objs_per_zspage; |
248ca1b05 zsmalloc: add ful... |
592 593 |
pages_used = obj_allocated / objs_per_zspage * class->pages_per_zspage; |
1120ed548 mm/zsmalloc: add ... |
594 595 596 |
seq_printf(s, " %5u %5u %11lu %12lu %13lu" " %10lu %10lu %16d %8lu ", |
248ca1b05 zsmalloc: add ful... |
597 598 |
i, class->size, class_almost_full, class_almost_empty, obj_allocated, obj_used, pages_used, |
1120ed548 mm/zsmalloc: add ... |
599 |
class->pages_per_zspage, freeable); |
248ca1b05 zsmalloc: add ful... |
600 601 602 603 604 605 |
total_class_almost_full += class_almost_full; total_class_almost_empty += class_almost_empty; total_objs += obj_allocated; total_used_objs += obj_used; total_pages += pages_used; |
1120ed548 mm/zsmalloc: add ... |
606 |
total_freeable += freeable; |
248ca1b05 zsmalloc: add ful... |
607 608 609 610 |
} seq_puts(s, " "); |
1120ed548 mm/zsmalloc: add ... |
611 612 |
seq_printf(s, " %5s %5s %11lu %12lu %13lu %10lu %10lu %16s %8lu ", |
248ca1b05 zsmalloc: add ful... |
613 614 |
"Total", "", total_class_almost_full, total_class_almost_empty, total_objs, |
1120ed548 mm/zsmalloc: add ... |
615 |
total_used_objs, total_pages, "", total_freeable); |
248ca1b05 zsmalloc: add ful... |
616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 |
return 0; } static int zs_stats_size_open(struct inode *inode, struct file *file) { return single_open(file, zs_stats_size_show, inode->i_private); } static const struct file_operations zs_stat_size_ops = { .open = zs_stats_size_open, .read = seq_read, .llseek = seq_lseek, .release = single_release, }; |
d34f61572 mm/zsmalloc: don'... |
631 |
static void zs_pool_stat_create(struct zs_pool *pool, const char *name) |
248ca1b05 zsmalloc: add ful... |
632 633 |
{ struct dentry *entry; |
4abaac9b7 update "mm/zsmall... |
634 635 636 |
if (!zs_stat_root) { pr_warn("no root stat dir, not creating <%s> stat dir ", name); |
d34f61572 mm/zsmalloc: don'... |
637 |
return; |
4abaac9b7 update "mm/zsmall... |
638 |
} |
248ca1b05 zsmalloc: add ful... |
639 640 641 642 643 |
entry = debugfs_create_dir(name, zs_stat_root); if (!entry) { pr_warn("debugfs dir <%s> creation failed ", name); |
d34f61572 mm/zsmalloc: don'... |
644 |
return; |
248ca1b05 zsmalloc: add ful... |
645 646 647 648 649 650 651 652 653 |
} pool->stat_dentry = entry; entry = debugfs_create_file("classes", S_IFREG | S_IRUGO, pool->stat_dentry, pool, &zs_stat_size_ops); if (!entry) { pr_warn("%s: debugfs file entry <%s> creation failed ", name, "classes"); |
4abaac9b7 update "mm/zsmall... |
654 655 |
debugfs_remove_recursive(pool->stat_dentry); pool->stat_dentry = NULL; |
248ca1b05 zsmalloc: add ful... |
656 |
} |
248ca1b05 zsmalloc: add ful... |
657 658 659 660 661 662 663 664 |
} static void zs_pool_stat_destroy(struct zs_pool *pool) { debugfs_remove_recursive(pool->stat_dentry); } #else /* CONFIG_ZSMALLOC_STAT */ |
4abaac9b7 update "mm/zsmall... |
665 |
static void __init zs_stat_init(void) |
248ca1b05 zsmalloc: add ful... |
666 |
{ |
248ca1b05 zsmalloc: add ful... |
667 668 669 670 671 |
} static void __exit zs_stat_exit(void) { } |
d34f61572 mm/zsmalloc: don'... |
672 |
static inline void zs_pool_stat_create(struct zs_pool *pool, const char *name) |
248ca1b05 zsmalloc: add ful... |
673 |
{ |
248ca1b05 zsmalloc: add ful... |
674 675 676 677 678 |
} static inline void zs_pool_stat_destroy(struct zs_pool *pool) { } |
248ca1b05 zsmalloc: add ful... |
679 |
#endif |
48b4800a1 zsmalloc: page mi... |
680 |
|
c3e3e88ad zsmalloc: add mor... |
681 682 683 684 685 686 687 |
/* * For each size class, zspages are divided into different groups * depending on how "full" they are. This was done so that we could * easily find empty or nearly empty zspages when we try to shrink * the pool (not yet implemented). This function returns fullness * status of the given page. */ |
1fc6e27d7 zsmalloc: keep ma... |
688 |
static enum fullness_group get_fullness_group(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
689 |
struct zspage *zspage) |
61989a80f staging: zsmalloc... |
690 |
{ |
1fc6e27d7 zsmalloc: keep ma... |
691 |
int inuse, objs_per_zspage; |
61989a80f staging: zsmalloc... |
692 |
enum fullness_group fg; |
830e4bc5b zsmalloc: clean u... |
693 |
|
3783689a1 zsmalloc: introdu... |
694 |
inuse = get_zspage_inuse(zspage); |
1fc6e27d7 zsmalloc: keep ma... |
695 |
objs_per_zspage = class->objs_per_zspage; |
61989a80f staging: zsmalloc... |
696 697 698 |
if (inuse == 0) fg = ZS_EMPTY; |
1fc6e27d7 zsmalloc: keep ma... |
699 |
else if (inuse == objs_per_zspage) |
61989a80f staging: zsmalloc... |
700 |
fg = ZS_FULL; |
1fc6e27d7 zsmalloc: keep ma... |
701 |
else if (inuse <= 3 * objs_per_zspage / fullness_threshold_frac) |
61989a80f staging: zsmalloc... |
702 703 704 705 706 707 |
fg = ZS_ALMOST_EMPTY; else fg = ZS_ALMOST_FULL; return fg; } |
c3e3e88ad zsmalloc: add mor... |
708 709 710 711 712 713 |
/* * Each size class maintains various freelists and zspages are assigned * to one of these freelists based on the number of live objects they * have. This functions inserts the given zspage into the freelist * identified by <class, fullness_group>. */ |
251cbb951 zsmalloc: reorder... |
714 |
static void insert_zspage(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
715 716 |
struct zspage *zspage, enum fullness_group fullness) |
61989a80f staging: zsmalloc... |
717 |
{ |
3783689a1 zsmalloc: introdu... |
718 |
struct zspage *head; |
61989a80f staging: zsmalloc... |
719 |
|
48b4800a1 zsmalloc: page mi... |
720 |
zs_stat_inc(class, fullness, 1); |
3783689a1 zsmalloc: introdu... |
721 722 |
head = list_first_entry_or_null(&class->fullness_list[fullness], struct zspage, list); |
58f171174 zsmalloc: partial... |
723 |
/* |
3783689a1 zsmalloc: introdu... |
724 725 |
* We want to see more ZS_FULL pages and less almost empty/full. * Put pages with higher ->inuse first. |
58f171174 zsmalloc: partial... |
726 |
*/ |
3783689a1 zsmalloc: introdu... |
727 728 729 730 731 732 733 |
if (head) { if (get_zspage_inuse(zspage) < get_zspage_inuse(head)) { list_add(&zspage->list, &head->list); return; } } list_add(&zspage->list, &class->fullness_list[fullness]); |
61989a80f staging: zsmalloc... |
734 |
} |
c3e3e88ad zsmalloc: add mor... |
735 736 737 738 |
/* * This function removes the given zspage from the freelist identified * by <class, fullness_group>. */ |
251cbb951 zsmalloc: reorder... |
739 |
static void remove_zspage(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
740 741 |
struct zspage *zspage, enum fullness_group fullness) |
61989a80f staging: zsmalloc... |
742 |
{ |
3783689a1 zsmalloc: introdu... |
743 |
VM_BUG_ON(list_empty(&class->fullness_list[fullness])); |
48b4800a1 zsmalloc: page mi... |
744 |
VM_BUG_ON(is_zspage_isolated(zspage)); |
61989a80f staging: zsmalloc... |
745 |
|
3783689a1 zsmalloc: introdu... |
746 |
list_del_init(&zspage->list); |
48b4800a1 zsmalloc: page mi... |
747 |
zs_stat_dec(class, fullness, 1); |
61989a80f staging: zsmalloc... |
748 |
} |
c3e3e88ad zsmalloc: add mor... |
749 750 751 752 753 754 755 756 757 |
/* * Each size class maintains zspages in different fullness groups depending * on the number of live objects they contain. When allocating or freeing * objects, the fullness status of the page can change, say, from ALMOST_FULL * to ALMOST_EMPTY when freeing an object. This function checks if such * a status change has occurred for the given page and accordingly moves the * page from the freelist of the old fullness group to that of the new * fullness group. */ |
c78062612 zsmalloc: factor ... |
758 |
static enum fullness_group fix_fullness_group(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
759 |
struct zspage *zspage) |
61989a80f staging: zsmalloc... |
760 761 |
{ int class_idx; |
61989a80f staging: zsmalloc... |
762 |
enum fullness_group currfg, newfg; |
3783689a1 zsmalloc: introdu... |
763 764 |
get_zspage_mapping(zspage, &class_idx, &currfg); newfg = get_fullness_group(class, zspage); |
61989a80f staging: zsmalloc... |
765 766 |
if (newfg == currfg) goto out; |
48b4800a1 zsmalloc: page mi... |
767 768 769 770 |
if (!is_zspage_isolated(zspage)) { remove_zspage(class, zspage, currfg); insert_zspage(class, zspage, newfg); } |
3783689a1 zsmalloc: introdu... |
771 |
set_zspage_mapping(zspage, class_idx, newfg); |
61989a80f staging: zsmalloc... |
772 773 774 775 776 777 778 779 780 781 |
out: return newfg; } /* * We have to decide on how many pages to link together * to form a zspage for each size class. This is important * to reduce wastage due to unusable space left at end of * each zspage which is given as: |
888fa374e mm/zsmalloc.c: fi... |
782 783 |
* wastage = Zp % class_size * usage = Zp - wastage |
61989a80f staging: zsmalloc... |
784 785 786 787 788 789 |
* where Zp = zspage size = k * PAGE_SIZE where k = 1, 2, ... * * For example, for size class of 3/8 * PAGE_SIZE, we should * link together 3 PAGE_SIZE sized pages to form a zspage * since then we can perfectly fit in 8 such objects. */ |
2e3b61547 staging: zsmalloc... |
790 |
static int get_pages_per_zspage(int class_size) |
61989a80f staging: zsmalloc... |
791 792 793 794 |
{ int i, max_usedpc = 0; /* zspage order which gives maximum used size per KB */ int max_usedpc_order = 1; |
84d4faaba staging: zsmalloc... |
795 |
for (i = 1; i <= ZS_MAX_PAGES_PER_ZSPAGE; i++) { |
61989a80f staging: zsmalloc... |
796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 |
int zspage_size; int waste, usedpc; zspage_size = i * PAGE_SIZE; waste = zspage_size % class_size; usedpc = (zspage_size - waste) * 100 / zspage_size; if (usedpc > max_usedpc) { max_usedpc = usedpc; max_usedpc_order = i; } } return max_usedpc_order; } |
3783689a1 zsmalloc: introdu... |
811 |
static struct zspage *get_zspage(struct page *page) |
61989a80f staging: zsmalloc... |
812 |
{ |
48b4800a1 zsmalloc: page mi... |
813 814 815 816 |
struct zspage *zspage = (struct zspage *)page->private; BUG_ON(zspage->magic != ZSPAGE_MAGIC); return zspage; |
61989a80f staging: zsmalloc... |
817 818 819 820 |
} static struct page *get_next_page(struct page *page) { |
48b4800a1 zsmalloc: page mi... |
821 822 823 824 |
if (unlikely(PageHugeObject(page))) return NULL; return page->freelist; |
61989a80f staging: zsmalloc... |
825 |
} |
bfd093f5e zsmalloc: use fre... |
826 827 828 829 |
/** * obj_to_location - get (<page>, <obj_idx>) from encoded object value * @page: page object resides in zspage * @obj_idx: object index |
67296874e staging: zsmalloc... |
830 |
*/ |
bfd093f5e zsmalloc: use fre... |
831 832 |
static void obj_to_location(unsigned long obj, struct page **page, unsigned int *obj_idx) |
61989a80f staging: zsmalloc... |
833 |
{ |
bfd093f5e zsmalloc: use fre... |
834 835 836 837 |
obj >>= OBJ_TAG_BITS; *page = pfn_to_page(obj >> OBJ_INDEX_BITS); *obj_idx = (obj & OBJ_INDEX_MASK); } |
61989a80f staging: zsmalloc... |
838 |
|
bfd093f5e zsmalloc: use fre... |
839 840 841 842 843 844 845 846 |
/** * location_to_obj - get obj value encoded from (<page>, <obj_idx>) * @page: page object resides in zspage * @obj_idx: object index */ static unsigned long location_to_obj(struct page *page, unsigned int obj_idx) { unsigned long obj; |
61989a80f staging: zsmalloc... |
847 |
|
312fcae22 zsmalloc: support... |
848 |
obj = page_to_pfn(page) << OBJ_INDEX_BITS; |
bfd093f5e zsmalloc: use fre... |
849 |
obj |= obj_idx & OBJ_INDEX_MASK; |
312fcae22 zsmalloc: support... |
850 |
obj <<= OBJ_TAG_BITS; |
61989a80f staging: zsmalloc... |
851 |
|
bfd093f5e zsmalloc: use fre... |
852 |
return obj; |
61989a80f staging: zsmalloc... |
853 |
} |
2e40e163a zsmalloc: decoupl... |
854 855 856 857 |
static unsigned long handle_to_obj(unsigned long handle) { return *(unsigned long *)handle; } |
48b4800a1 zsmalloc: page mi... |
858 |
static unsigned long obj_to_head(struct page *page, void *obj) |
312fcae22 zsmalloc: support... |
859 |
{ |
48b4800a1 zsmalloc: page mi... |
860 |
if (unlikely(PageHugeObject(page))) { |
830e4bc5b zsmalloc: clean u... |
861 |
VM_BUG_ON_PAGE(!is_first_page(page), page); |
3783689a1 zsmalloc: introdu... |
862 |
return page->index; |
7b60a6852 zsmalloc: record ... |
863 864 |
} else return *(unsigned long *)obj; |
312fcae22 zsmalloc: support... |
865 |
} |
48b4800a1 zsmalloc: page mi... |
866 867 868 869 |
static inline int testpin_tag(unsigned long handle) { return bit_spin_is_locked(HANDLE_PIN_BIT, (unsigned long *)handle); } |
312fcae22 zsmalloc: support... |
870 871 |
static inline int trypin_tag(unsigned long handle) { |
1b8320b62 zsmalloc: use bit... |
872 |
return bit_spin_trylock(HANDLE_PIN_BIT, (unsigned long *)handle); |
312fcae22 zsmalloc: support... |
873 874 875 876 |
} static void pin_tag(unsigned long handle) { |
1b8320b62 zsmalloc: use bit... |
877 |
bit_spin_lock(HANDLE_PIN_BIT, (unsigned long *)handle); |
312fcae22 zsmalloc: support... |
878 879 880 881 |
} static void unpin_tag(unsigned long handle) { |
1b8320b62 zsmalloc: use bit... |
882 |
bit_spin_unlock(HANDLE_PIN_BIT, (unsigned long *)handle); |
312fcae22 zsmalloc: support... |
883 |
} |
f4477e90b staging: zsmalloc... |
884 885 |
static void reset_page(struct page *page) { |
48b4800a1 zsmalloc: page mi... |
886 |
__ClearPageMovable(page); |
18fd06bf7 mm/zsmalloc: use ... |
887 888 |
ClearPagePrivate(page); ClearPagePrivate2(page); |
f4477e90b staging: zsmalloc... |
889 |
set_page_private(page, 0); |
48b4800a1 zsmalloc: page mi... |
890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 |
page_mapcount_reset(page); ClearPageHugeObject(page); page->freelist = NULL; } /* * To prevent zspage destroy during migration, zspage freeing should * hold locks of all pages in the zspage. */ void lock_zspage(struct zspage *zspage) { struct page *page = get_first_page(zspage); do { lock_page(page); } while ((page = get_next_page(page)) != NULL); } int trylock_zspage(struct zspage *zspage) { struct page *cursor, *fail; for (cursor = get_first_page(zspage); cursor != NULL; cursor = get_next_page(cursor)) { if (!trylock_page(cursor)) { fail = cursor; goto unlock; } } return 1; unlock: for (cursor = get_first_page(zspage); cursor != fail; cursor = get_next_page(cursor)) unlock_page(cursor); return 0; |
f4477e90b staging: zsmalloc... |
927 |
} |
48b4800a1 zsmalloc: page mi... |
928 929 |
static void __free_zspage(struct zs_pool *pool, struct size_class *class, struct zspage *zspage) |
61989a80f staging: zsmalloc... |
930 |
{ |
3783689a1 zsmalloc: introdu... |
931 |
struct page *page, *next; |
48b4800a1 zsmalloc: page mi... |
932 933 934 935 936 937 |
enum fullness_group fg; unsigned int class_idx; get_zspage_mapping(zspage, &class_idx, &fg); assert_spin_locked(&class->lock); |
61989a80f staging: zsmalloc... |
938 |
|
3783689a1 zsmalloc: introdu... |
939 |
VM_BUG_ON(get_zspage_inuse(zspage)); |
48b4800a1 zsmalloc: page mi... |
940 |
VM_BUG_ON(fg != ZS_EMPTY); |
61989a80f staging: zsmalloc... |
941 |
|
48b4800a1 zsmalloc: page mi... |
942 |
next = page = get_first_page(zspage); |
3783689a1 zsmalloc: introdu... |
943 |
do { |
48b4800a1 zsmalloc: page mi... |
944 945 |
VM_BUG_ON_PAGE(!PageLocked(page), page); next = get_next_page(page); |
3783689a1 zsmalloc: introdu... |
946 |
reset_page(page); |
48b4800a1 zsmalloc: page mi... |
947 |
unlock_page(page); |
91537fee0 mm: add NR_ZSMALL... |
948 |
dec_zone_page_state(page, NR_ZSPAGES); |
3783689a1 zsmalloc: introdu... |
949 950 951 |
put_page(page); page = next; } while (page != NULL); |
61989a80f staging: zsmalloc... |
952 |
|
3783689a1 zsmalloc: introdu... |
953 |
cache_free_zspage(pool, zspage); |
48b4800a1 zsmalloc: page mi... |
954 |
|
b4fd07a08 mm/zsmalloc: use ... |
955 |
zs_stat_dec(class, OBJ_ALLOCATED, class->objs_per_zspage); |
48b4800a1 zsmalloc: page mi... |
956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 |
atomic_long_sub(class->pages_per_zspage, &pool->pages_allocated); } static void free_zspage(struct zs_pool *pool, struct size_class *class, struct zspage *zspage) { VM_BUG_ON(get_zspage_inuse(zspage)); VM_BUG_ON(list_empty(&zspage->list)); if (!trylock_zspage(zspage)) { kick_deferred_free(pool); return; } remove_zspage(class, zspage, ZS_EMPTY); __free_zspage(pool, class, zspage); |
61989a80f staging: zsmalloc... |
973 974 975 |
} /* Initialize a newly allocated zspage */ |
3783689a1 zsmalloc: introdu... |
976 |
static void init_zspage(struct size_class *class, struct zspage *zspage) |
61989a80f staging: zsmalloc... |
977 |
{ |
bfd093f5e zsmalloc: use fre... |
978 |
unsigned int freeobj = 1; |
61989a80f staging: zsmalloc... |
979 |
unsigned long off = 0; |
48b4800a1 zsmalloc: page mi... |
980 |
struct page *page = get_first_page(zspage); |
830e4bc5b zsmalloc: clean u... |
981 |
|
61989a80f staging: zsmalloc... |
982 983 984 |
while (page) { struct page *next_page; struct link_free *link; |
af4ee5e97 zsmalloc: correct... |
985 |
void *vaddr; |
61989a80f staging: zsmalloc... |
986 |
|
3783689a1 zsmalloc: introdu... |
987 |
set_first_obj_offset(page, off); |
61989a80f staging: zsmalloc... |
988 |
|
af4ee5e97 zsmalloc: correct... |
989 990 |
vaddr = kmap_atomic(page); link = (struct link_free *)vaddr + off / sizeof(*link); |
5538c5623 zsmalloc: simplif... |
991 992 |
while ((off += class->size) < PAGE_SIZE) { |
3b1d9ca65 zsmalloc: use OBJ... |
993 |
link->next = freeobj++ << OBJ_TAG_BITS; |
5538c5623 zsmalloc: simplif... |
994 |
link += class->size / sizeof(*link); |
61989a80f staging: zsmalloc... |
995 996 997 998 999 1000 1001 1002 |
} /* * We now come to the last (full or partial) object on this * page, which must point to the first object on the next * page (if present) */ next_page = get_next_page(page); |
bfd093f5e zsmalloc: use fre... |
1003 |
if (next_page) { |
3b1d9ca65 zsmalloc: use OBJ... |
1004 |
link->next = freeobj++ << OBJ_TAG_BITS; |
bfd093f5e zsmalloc: use fre... |
1005 1006 |
} else { /* |
3b1d9ca65 zsmalloc: use OBJ... |
1007 |
* Reset OBJ_TAG_BITS bit to last link to tell |
bfd093f5e zsmalloc: use fre... |
1008 1009 |
* whether it's allocated object or not. */ |
3b1d9ca65 zsmalloc: use OBJ... |
1010 |
link->next = -1 << OBJ_TAG_BITS; |
bfd093f5e zsmalloc: use fre... |
1011 |
} |
af4ee5e97 zsmalloc: correct... |
1012 |
kunmap_atomic(vaddr); |
61989a80f staging: zsmalloc... |
1013 |
page = next_page; |
5538c5623 zsmalloc: simplif... |
1014 |
off %= PAGE_SIZE; |
61989a80f staging: zsmalloc... |
1015 |
} |
bdb0af7ca zsmalloc: factor ... |
1016 |
|
bfd093f5e zsmalloc: use fre... |
1017 |
set_freeobj(zspage, 0); |
61989a80f staging: zsmalloc... |
1018 |
} |
48b4800a1 zsmalloc: page mi... |
1019 1020 |
static void create_page_chain(struct size_class *class, struct zspage *zspage, struct page *pages[]) |
61989a80f staging: zsmalloc... |
1021 |
{ |
bdb0af7ca zsmalloc: factor ... |
1022 1023 1024 |
int i; struct page *page; struct page *prev_page = NULL; |
48b4800a1 zsmalloc: page mi... |
1025 |
int nr_pages = class->pages_per_zspage; |
61989a80f staging: zsmalloc... |
1026 1027 1028 |
/* * Allocate individual pages and link them together as: |
48b4800a1 zsmalloc: page mi... |
1029 |
* 1. all pages are linked together using page->freelist |
3783689a1 zsmalloc: introdu... |
1030 |
* 2. each sub-page point to zspage using page->private |
61989a80f staging: zsmalloc... |
1031 |
* |
3783689a1 zsmalloc: introdu... |
1032 1033 |
* we set PG_private to identify the first page (i.e. no other sub-page * has this flag set) and PG_private_2 to identify the last page. |
61989a80f staging: zsmalloc... |
1034 |
*/ |
bdb0af7ca zsmalloc: factor ... |
1035 1036 |
for (i = 0; i < nr_pages; i++) { page = pages[i]; |
3783689a1 zsmalloc: introdu... |
1037 |
set_page_private(page, (unsigned long)zspage); |
48b4800a1 zsmalloc: page mi... |
1038 |
page->freelist = NULL; |
bdb0af7ca zsmalloc: factor ... |
1039 |
if (i == 0) { |
3783689a1 zsmalloc: introdu... |
1040 |
zspage->first_page = page; |
a27545bf0 zsmalloc: use Pag... |
1041 |
SetPagePrivate(page); |
48b4800a1 zsmalloc: page mi... |
1042 1043 1044 |
if (unlikely(class->objs_per_zspage == 1 && class->pages_per_zspage == 1)) SetPageHugeObject(page); |
3783689a1 zsmalloc: introdu... |
1045 |
} else { |
48b4800a1 zsmalloc: page mi... |
1046 |
prev_page->freelist = page; |
61989a80f staging: zsmalloc... |
1047 |
} |
48b4800a1 zsmalloc: page mi... |
1048 |
if (i == nr_pages - 1) |
a27545bf0 zsmalloc: use Pag... |
1049 |
SetPagePrivate2(page); |
61989a80f staging: zsmalloc... |
1050 1051 |
prev_page = page; } |
bdb0af7ca zsmalloc: factor ... |
1052 |
} |
61989a80f staging: zsmalloc... |
1053 |
|
bdb0af7ca zsmalloc: factor ... |
1054 1055 1056 |
/* * Allocate a zspage for the given size class */ |
3783689a1 zsmalloc: introdu... |
1057 1058 1059 |
static struct zspage *alloc_zspage(struct zs_pool *pool, struct size_class *class, gfp_t gfp) |
bdb0af7ca zsmalloc: factor ... |
1060 1061 |
{ int i; |
bdb0af7ca zsmalloc: factor ... |
1062 |
struct page *pages[ZS_MAX_PAGES_PER_ZSPAGE]; |
3783689a1 zsmalloc: introdu... |
1063 1064 1065 1066 1067 1068 |
struct zspage *zspage = cache_alloc_zspage(pool, gfp); if (!zspage) return NULL; memset(zspage, 0, sizeof(struct zspage)); |
48b4800a1 zsmalloc: page mi... |
1069 1070 |
zspage->magic = ZSPAGE_MAGIC; migrate_lock_init(zspage); |
61989a80f staging: zsmalloc... |
1071 |
|
bdb0af7ca zsmalloc: factor ... |
1072 1073 |
for (i = 0; i < class->pages_per_zspage; i++) { struct page *page; |
61989a80f staging: zsmalloc... |
1074 |
|
3783689a1 zsmalloc: introdu... |
1075 |
page = alloc_page(gfp); |
bdb0af7ca zsmalloc: factor ... |
1076 |
if (!page) { |
91537fee0 mm: add NR_ZSMALL... |
1077 1078 |
while (--i >= 0) { dec_zone_page_state(pages[i], NR_ZSPAGES); |
bdb0af7ca zsmalloc: factor ... |
1079 |
__free_page(pages[i]); |
91537fee0 mm: add NR_ZSMALL... |
1080 |
} |
3783689a1 zsmalloc: introdu... |
1081 |
cache_free_zspage(pool, zspage); |
bdb0af7ca zsmalloc: factor ... |
1082 1083 |
return NULL; } |
91537fee0 mm: add NR_ZSMALL... |
1084 1085 |
inc_zone_page_state(page, NR_ZSPAGES); |
bdb0af7ca zsmalloc: factor ... |
1086 |
pages[i] = page; |
61989a80f staging: zsmalloc... |
1087 |
} |
48b4800a1 zsmalloc: page mi... |
1088 |
create_page_chain(class, zspage, pages); |
3783689a1 zsmalloc: introdu... |
1089 |
init_zspage(class, zspage); |
bdb0af7ca zsmalloc: factor ... |
1090 |
|
3783689a1 zsmalloc: introdu... |
1091 |
return zspage; |
61989a80f staging: zsmalloc... |
1092 |
} |
3783689a1 zsmalloc: introdu... |
1093 |
static struct zspage *find_get_zspage(struct size_class *class) |
61989a80f staging: zsmalloc... |
1094 1095 |
{ int i; |
3783689a1 zsmalloc: introdu... |
1096 |
struct zspage *zspage; |
61989a80f staging: zsmalloc... |
1097 |
|
48b4800a1 zsmalloc: page mi... |
1098 |
for (i = ZS_ALMOST_FULL; i >= ZS_EMPTY; i--) { |
3783689a1 zsmalloc: introdu... |
1099 1100 1101 |
zspage = list_first_entry_or_null(&class->fullness_list[i], struct zspage, list); if (zspage) |
61989a80f staging: zsmalloc... |
1102 1103 |
break; } |
3783689a1 zsmalloc: introdu... |
1104 |
return zspage; |
61989a80f staging: zsmalloc... |
1105 |
} |
1b945aeef zsmalloc: add Kco... |
1106 |
#ifdef CONFIG_PGTABLE_MAPPING |
f553646a6 staging: zsmalloc... |
1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 |
static inline int __zs_cpu_up(struct mapping_area *area) { /* * Make sure we don't leak memory if a cpu UP notification * and zs_init() race and both call zs_cpu_up() on the same cpu */ if (area->vm) return 0; area->vm = alloc_vm_area(PAGE_SIZE * 2, NULL); if (!area->vm) return -ENOMEM; return 0; } static inline void __zs_cpu_down(struct mapping_area *area) { if (area->vm) free_vm_area(area->vm); area->vm = NULL; } static inline void *__zs_map_object(struct mapping_area *area, struct page *pages[2], int off, int size) { |
f6f8ed473 mm/vmalloc.c: cle... |
1131 |
BUG_ON(map_vm_area(area->vm, PAGE_KERNEL, pages)); |
f553646a6 staging: zsmalloc... |
1132 1133 1134 1135 1136 1137 1138 1139 |
area->vm_addr = area->vm->addr; return area->vm_addr + off; } static inline void __zs_unmap_object(struct mapping_area *area, struct page *pages[2], int off, int size) { unsigned long addr = (unsigned long)area->vm_addr; |
f553646a6 staging: zsmalloc... |
1140 |
|
d95abbbb2 staging: zsmalloc... |
1141 |
unmap_kernel_range(addr, PAGE_SIZE * 2); |
f553646a6 staging: zsmalloc... |
1142 |
} |
1b945aeef zsmalloc: add Kco... |
1143 |
#else /* CONFIG_PGTABLE_MAPPING */ |
f553646a6 staging: zsmalloc... |
1144 1145 1146 1147 1148 1149 1150 1151 1152 |
static inline int __zs_cpu_up(struct mapping_area *area) { /* * Make sure we don't leak memory if a cpu UP notification * and zs_init() race and both call zs_cpu_up() on the same cpu */ if (area->vm_buf) return 0; |
40f9fb8cf mm/zsmalloc: supp... |
1153 |
area->vm_buf = kmalloc(ZS_MAX_ALLOC_SIZE, GFP_KERNEL); |
f553646a6 staging: zsmalloc... |
1154 1155 1156 1157 1158 1159 1160 |
if (!area->vm_buf) return -ENOMEM; return 0; } static inline void __zs_cpu_down(struct mapping_area *area) { |
40f9fb8cf mm/zsmalloc: supp... |
1161 |
kfree(area->vm_buf); |
f553646a6 staging: zsmalloc... |
1162 1163 1164 1165 1166 |
area->vm_buf = NULL; } static void *__zs_map_object(struct mapping_area *area, struct page *pages[2], int off, int size) |
5f601902c staging: zsmalloc... |
1167 |
{ |
5f601902c staging: zsmalloc... |
1168 1169 |
int sizes[2]; void *addr; |
f553646a6 staging: zsmalloc... |
1170 |
char *buf = area->vm_buf; |
5f601902c staging: zsmalloc... |
1171 |
|
f553646a6 staging: zsmalloc... |
1172 1173 1174 1175 1176 1177 |
/* disable page faults to match kmap_atomic() return conditions */ pagefault_disable(); /* no read fastpath */ if (area->vm_mm == ZS_MM_WO) goto out; |
5f601902c staging: zsmalloc... |
1178 1179 1180 |
sizes[0] = PAGE_SIZE - off; sizes[1] = size - sizes[0]; |
5f601902c staging: zsmalloc... |
1181 1182 1183 1184 1185 1186 1187 |
/* copy object to per-cpu buffer */ addr = kmap_atomic(pages[0]); memcpy(buf, addr + off, sizes[0]); kunmap_atomic(addr); addr = kmap_atomic(pages[1]); memcpy(buf + sizes[0], addr, sizes[1]); kunmap_atomic(addr); |
f553646a6 staging: zsmalloc... |
1188 1189 |
out: return area->vm_buf; |
5f601902c staging: zsmalloc... |
1190 |
} |
f553646a6 staging: zsmalloc... |
1191 1192 |
static void __zs_unmap_object(struct mapping_area *area, struct page *pages[2], int off, int size) |
5f601902c staging: zsmalloc... |
1193 |
{ |
5f601902c staging: zsmalloc... |
1194 1195 |
int sizes[2]; void *addr; |
2e40e163a zsmalloc: decoupl... |
1196 |
char *buf; |
5f601902c staging: zsmalloc... |
1197 |
|
f553646a6 staging: zsmalloc... |
1198 1199 1200 |
/* no write fastpath */ if (area->vm_mm == ZS_MM_RO) goto out; |
5f601902c staging: zsmalloc... |
1201 |
|
7b60a6852 zsmalloc: record ... |
1202 |
buf = area->vm_buf; |
a82cbf071 zsmalloc: drop un... |
1203 1204 1205 |
buf = buf + ZS_HANDLE_SIZE; size -= ZS_HANDLE_SIZE; off += ZS_HANDLE_SIZE; |
2e40e163a zsmalloc: decoupl... |
1206 |
|
5f601902c staging: zsmalloc... |
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 |
sizes[0] = PAGE_SIZE - off; sizes[1] = size - sizes[0]; /* copy per-cpu buffer to object */ addr = kmap_atomic(pages[0]); memcpy(addr + off, buf, sizes[0]); kunmap_atomic(addr); addr = kmap_atomic(pages[1]); memcpy(addr, buf + sizes[0], sizes[1]); kunmap_atomic(addr); |
f553646a6 staging: zsmalloc... |
1217 1218 1219 1220 |
out: /* enable page faults to match kunmap_atomic() return conditions */ pagefault_enable(); |
5f601902c staging: zsmalloc... |
1221 |
} |
61989a80f staging: zsmalloc... |
1222 |
|
1b945aeef zsmalloc: add Kco... |
1223 |
#endif /* CONFIG_PGTABLE_MAPPING */ |
f553646a6 staging: zsmalloc... |
1224 |
|
61989a80f staging: zsmalloc... |
1225 1226 1227 |
static int zs_cpu_notifier(struct notifier_block *nb, unsigned long action, void *pcpu) { |
f553646a6 staging: zsmalloc... |
1228 |
int ret, cpu = (long)pcpu; |
61989a80f staging: zsmalloc... |
1229 1230 1231 1232 1233 |
struct mapping_area *area; switch (action) { case CPU_UP_PREPARE: area = &per_cpu(zs_map_area, cpu); |
f553646a6 staging: zsmalloc... |
1234 1235 1236 |
ret = __zs_cpu_up(area); if (ret) return notifier_from_errno(ret); |
61989a80f staging: zsmalloc... |
1237 1238 1239 1240 |
break; case CPU_DEAD: case CPU_UP_CANCELED: area = &per_cpu(zs_map_area, cpu); |
f553646a6 staging: zsmalloc... |
1241 |
__zs_cpu_down(area); |
61989a80f staging: zsmalloc... |
1242 1243 1244 1245 1246 1247 1248 1249 1250 |
break; } return NOTIFY_OK; } static struct notifier_block zs_cpu_nb = { .notifier_call = zs_cpu_notifier }; |
b1b00a5b8 zsmalloc: fix zs_... |
1251 |
static int zs_register_cpu_notifier(void) |
61989a80f staging: zsmalloc... |
1252 |
{ |
b1b00a5b8 zsmalloc: fix zs_... |
1253 |
int cpu, uninitialized_var(ret); |
61989a80f staging: zsmalloc... |
1254 |
|
f0e71fcd0 zsmalloc: Fix CPU... |
1255 1256 1257 |
cpu_notifier_register_begin(); __register_cpu_notifier(&zs_cpu_nb); |
61989a80f staging: zsmalloc... |
1258 1259 |
for_each_online_cpu(cpu) { ret = zs_cpu_notifier(NULL, CPU_UP_PREPARE, (void *)(long)cpu); |
b1b00a5b8 zsmalloc: fix zs_... |
1260 1261 |
if (notifier_to_errno(ret)) break; |
61989a80f staging: zsmalloc... |
1262 |
} |
f0e71fcd0 zsmalloc: Fix CPU... |
1263 1264 |
cpu_notifier_register_done(); |
b1b00a5b8 zsmalloc: fix zs_... |
1265 1266 |
return notifier_to_errno(ret); } |
f0e71fcd0 zsmalloc: Fix CPU... |
1267 |
|
66cdef663 mm/zsmalloc: adju... |
1268 |
static void zs_unregister_cpu_notifier(void) |
40f9fb8cf mm/zsmalloc: supp... |
1269 |
{ |
66cdef663 mm/zsmalloc: adju... |
1270 |
int cpu; |
40f9fb8cf mm/zsmalloc: supp... |
1271 |
|
66cdef663 mm/zsmalloc: adju... |
1272 |
cpu_notifier_register_begin(); |
40f9fb8cf mm/zsmalloc: supp... |
1273 |
|
66cdef663 mm/zsmalloc: adju... |
1274 1275 1276 |
for_each_online_cpu(cpu) zs_cpu_notifier(NULL, CPU_DEAD, (void *)(long)cpu); __unregister_cpu_notifier(&zs_cpu_nb); |
40f9fb8cf mm/zsmalloc: supp... |
1277 |
|
66cdef663 mm/zsmalloc: adju... |
1278 |
cpu_notifier_register_done(); |
b1b00a5b8 zsmalloc: fix zs_... |
1279 |
} |
35b3445e9 mm/zsmalloc: add ... |
1280 |
static void __init init_zs_size_classes(void) |
b1b00a5b8 zsmalloc: fix zs_... |
1281 |
{ |
66cdef663 mm/zsmalloc: adju... |
1282 |
int nr; |
c795779df mm/zpool: zbud/zs... |
1283 |
|
66cdef663 mm/zsmalloc: adju... |
1284 1285 1286 |
nr = (ZS_MAX_ALLOC_SIZE - ZS_MIN_ALLOC_SIZE) / ZS_SIZE_CLASS_DELTA + 1; if ((ZS_MAX_ALLOC_SIZE - ZS_MIN_ALLOC_SIZE) % ZS_SIZE_CLASS_DELTA) nr += 1; |
40f9fb8cf mm/zsmalloc: supp... |
1287 |
|
66cdef663 mm/zsmalloc: adju... |
1288 |
zs_size_classes = nr; |
61989a80f staging: zsmalloc... |
1289 |
} |
64d90465f mm/zsmalloc: avoi... |
1290 1291 |
static bool can_merge(struct size_class *prev, int pages_per_zspage, int objs_per_zspage) |
9eec4cd53 zsmalloc: merge s... |
1292 |
{ |
64d90465f mm/zsmalloc: avoi... |
1293 1294 1295 |
if (prev->pages_per_zspage == pages_per_zspage && prev->objs_per_zspage == objs_per_zspage) return true; |
9eec4cd53 zsmalloc: merge s... |
1296 |
|
64d90465f mm/zsmalloc: avoi... |
1297 |
return false; |
9eec4cd53 zsmalloc: merge s... |
1298 |
} |
3783689a1 zsmalloc: introdu... |
1299 |
static bool zspage_full(struct size_class *class, struct zspage *zspage) |
312fcae22 zsmalloc: support... |
1300 |
{ |
3783689a1 zsmalloc: introdu... |
1301 |
return get_zspage_inuse(zspage) == class->objs_per_zspage; |
312fcae22 zsmalloc: support... |
1302 |
} |
66cdef663 mm/zsmalloc: adju... |
1303 1304 1305 1306 1307 |
unsigned long zs_get_total_pages(struct zs_pool *pool) { return atomic_long_read(&pool->pages_allocated); } EXPORT_SYMBOL_GPL(zs_get_total_pages); |
4bbc0bc06 staging: zsmalloc... |
1308 |
/** |
66cdef663 mm/zsmalloc: adju... |
1309 1310 1311 |
* zs_map_object - get address of allocated object from handle. * @pool: pool from which the object was allocated * @handle: handle returned from zs_malloc |
4bbc0bc06 staging: zsmalloc... |
1312 |
* |
66cdef663 mm/zsmalloc: adju... |
1313 1314 1315 |
* Before using an object allocated from zs_malloc, it must be mapped using * this function. When done with the object, it must be unmapped using * zs_unmap_object. |
4bbc0bc06 staging: zsmalloc... |
1316 |
* |
66cdef663 mm/zsmalloc: adju... |
1317 1318 1319 1320 |
* Only one object can be mapped per cpu at a time. There is no protection * against nested mappings. * * This function returns with preemption and page faults disabled. |
4bbc0bc06 staging: zsmalloc... |
1321 |
*/ |
66cdef663 mm/zsmalloc: adju... |
1322 1323 |
void *zs_map_object(struct zs_pool *pool, unsigned long handle, enum zs_mapmode mm) |
61989a80f staging: zsmalloc... |
1324 |
{ |
3783689a1 zsmalloc: introdu... |
1325 |
struct zspage *zspage; |
66cdef663 mm/zsmalloc: adju... |
1326 |
struct page *page; |
bfd093f5e zsmalloc: use fre... |
1327 1328 |
unsigned long obj, off; unsigned int obj_idx; |
61989a80f staging: zsmalloc... |
1329 |
|
66cdef663 mm/zsmalloc: adju... |
1330 1331 1332 1333 1334 |
unsigned int class_idx; enum fullness_group fg; struct size_class *class; struct mapping_area *area; struct page *pages[2]; |
2e40e163a zsmalloc: decoupl... |
1335 |
void *ret; |
61989a80f staging: zsmalloc... |
1336 |
|
9eec4cd53 zsmalloc: merge s... |
1337 |
/* |
66cdef663 mm/zsmalloc: adju... |
1338 1339 1340 |
* Because we use per-cpu mapping areas shared among the * pools/users, we can't allow mapping in interrupt context * because it can corrupt another users mappings. |
9eec4cd53 zsmalloc: merge s... |
1341 |
*/ |
830e4bc5b zsmalloc: clean u... |
1342 |
WARN_ON_ONCE(in_interrupt()); |
61989a80f staging: zsmalloc... |
1343 |
|
312fcae22 zsmalloc: support... |
1344 1345 |
/* From now on, migration cannot move the object */ pin_tag(handle); |
2e40e163a zsmalloc: decoupl... |
1346 1347 |
obj = handle_to_obj(handle); obj_to_location(obj, &page, &obj_idx); |
3783689a1 zsmalloc: introdu... |
1348 |
zspage = get_zspage(page); |
48b4800a1 zsmalloc: page mi... |
1349 1350 1351 |
/* migration cannot move any subpage in this zspage */ migrate_read_lock(zspage); |
3783689a1 zsmalloc: introdu... |
1352 |
get_zspage_mapping(zspage, &class_idx, &fg); |
66cdef663 mm/zsmalloc: adju... |
1353 |
class = pool->size_class[class_idx]; |
bfd093f5e zsmalloc: use fre... |
1354 |
off = (class->size * obj_idx) & ~PAGE_MASK; |
df8b5bb99 mm/zsmalloc: avoi... |
1355 |
|
66cdef663 mm/zsmalloc: adju... |
1356 1357 1358 1359 1360 |
area = &get_cpu_var(zs_map_area); area->vm_mm = mm; if (off + class->size <= PAGE_SIZE) { /* this object is contained entirely within a page */ area->vm_addr = kmap_atomic(page); |
2e40e163a zsmalloc: decoupl... |
1361 1362 |
ret = area->vm_addr + off; goto out; |
61989a80f staging: zsmalloc... |
1363 |
} |
66cdef663 mm/zsmalloc: adju... |
1364 1365 1366 1367 |
/* this object spans two pages */ pages[0] = page; pages[1] = get_next_page(page); BUG_ON(!pages[1]); |
9eec4cd53 zsmalloc: merge s... |
1368 |
|
2e40e163a zsmalloc: decoupl... |
1369 1370 |
ret = __zs_map_object(area, pages, off, class->size); out: |
48b4800a1 zsmalloc: page mi... |
1371 |
if (likely(!PageHugeObject(page))) |
7b60a6852 zsmalloc: record ... |
1372 1373 1374 |
ret += ZS_HANDLE_SIZE; return ret; |
61989a80f staging: zsmalloc... |
1375 |
} |
66cdef663 mm/zsmalloc: adju... |
1376 |
EXPORT_SYMBOL_GPL(zs_map_object); |
61989a80f staging: zsmalloc... |
1377 |
|
66cdef663 mm/zsmalloc: adju... |
1378 |
void zs_unmap_object(struct zs_pool *pool, unsigned long handle) |
61989a80f staging: zsmalloc... |
1379 |
{ |
3783689a1 zsmalloc: introdu... |
1380 |
struct zspage *zspage; |
66cdef663 mm/zsmalloc: adju... |
1381 |
struct page *page; |
bfd093f5e zsmalloc: use fre... |
1382 1383 |
unsigned long obj, off; unsigned int obj_idx; |
61989a80f staging: zsmalloc... |
1384 |
|
66cdef663 mm/zsmalloc: adju... |
1385 1386 1387 1388 |
unsigned int class_idx; enum fullness_group fg; struct size_class *class; struct mapping_area *area; |
9eec4cd53 zsmalloc: merge s... |
1389 |
|
2e40e163a zsmalloc: decoupl... |
1390 1391 |
obj = handle_to_obj(handle); obj_to_location(obj, &page, &obj_idx); |
3783689a1 zsmalloc: introdu... |
1392 1393 |
zspage = get_zspage(page); get_zspage_mapping(zspage, &class_idx, &fg); |
66cdef663 mm/zsmalloc: adju... |
1394 |
class = pool->size_class[class_idx]; |
bfd093f5e zsmalloc: use fre... |
1395 |
off = (class->size * obj_idx) & ~PAGE_MASK; |
61989a80f staging: zsmalloc... |
1396 |
|
66cdef663 mm/zsmalloc: adju... |
1397 1398 1399 1400 1401 |
area = this_cpu_ptr(&zs_map_area); if (off + class->size <= PAGE_SIZE) kunmap_atomic(area->vm_addr); else { struct page *pages[2]; |
40f9fb8cf mm/zsmalloc: supp... |
1402 |
|
66cdef663 mm/zsmalloc: adju... |
1403 1404 1405 1406 1407 1408 1409 |
pages[0] = page; pages[1] = get_next_page(page); BUG_ON(!pages[1]); __zs_unmap_object(area, pages, off, class->size); } put_cpu_var(zs_map_area); |
48b4800a1 zsmalloc: page mi... |
1410 1411 |
migrate_read_unlock(zspage); |
312fcae22 zsmalloc: support... |
1412 |
unpin_tag(handle); |
61989a80f staging: zsmalloc... |
1413 |
} |
66cdef663 mm/zsmalloc: adju... |
1414 |
EXPORT_SYMBOL_GPL(zs_unmap_object); |
61989a80f staging: zsmalloc... |
1415 |
|
251cbb951 zsmalloc: reorder... |
1416 |
static unsigned long obj_malloc(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
1417 |
struct zspage *zspage, unsigned long handle) |
c78062612 zsmalloc: factor ... |
1418 |
{ |
bfd093f5e zsmalloc: use fre... |
1419 |
int i, nr_page, offset; |
c78062612 zsmalloc: factor ... |
1420 1421 1422 1423 |
unsigned long obj; struct link_free *link; struct page *m_page; |
bfd093f5e zsmalloc: use fre... |
1424 |
unsigned long m_offset; |
c78062612 zsmalloc: factor ... |
1425 |
void *vaddr; |
312fcae22 zsmalloc: support... |
1426 |
handle |= OBJ_ALLOCATED_TAG; |
3783689a1 zsmalloc: introdu... |
1427 |
obj = get_freeobj(zspage); |
bfd093f5e zsmalloc: use fre... |
1428 1429 1430 1431 1432 1433 1434 1435 |
offset = obj * class->size; nr_page = offset >> PAGE_SHIFT; m_offset = offset & ~PAGE_MASK; m_page = get_first_page(zspage); for (i = 0; i < nr_page; i++) m_page = get_next_page(m_page); |
c78062612 zsmalloc: factor ... |
1436 1437 1438 |
vaddr = kmap_atomic(m_page); link = (struct link_free *)vaddr + m_offset / sizeof(*link); |
3b1d9ca65 zsmalloc: use OBJ... |
1439 |
set_freeobj(zspage, link->next >> OBJ_TAG_BITS); |
48b4800a1 zsmalloc: page mi... |
1440 |
if (likely(!PageHugeObject(m_page))) |
7b60a6852 zsmalloc: record ... |
1441 1442 1443 |
/* record handle in the header of allocated chunk */ link->handle = handle; else |
3783689a1 zsmalloc: introdu... |
1444 1445 |
/* record handle to page->index */ zspage->first_page->index = handle; |
c78062612 zsmalloc: factor ... |
1446 |
kunmap_atomic(vaddr); |
3783689a1 zsmalloc: introdu... |
1447 |
mod_zspage_inuse(zspage, 1); |
c78062612 zsmalloc: factor ... |
1448 |
zs_stat_inc(class, OBJ_USED, 1); |
bfd093f5e zsmalloc: use fre... |
1449 |
obj = location_to_obj(m_page, obj); |
c78062612 zsmalloc: factor ... |
1450 1451 |
return obj; } |
61989a80f staging: zsmalloc... |
1452 1453 1454 1455 |
/** * zs_malloc - Allocate block of given size from pool. * @pool: pool to allocate from * @size: size of block to allocate |
fd8544639 mm/zsmalloc: keep... |
1456 |
* @gfp: gfp flags when allocating object |
61989a80f staging: zsmalloc... |
1457 |
* |
00a61d861 staging: zsmalloc... |
1458 |
* On success, handle to the allocated object is returned, |
c23443483 staging: zsmalloc... |
1459 |
* otherwise 0. |
61989a80f staging: zsmalloc... |
1460 1461 |
* Allocation requests with size > ZS_MAX_ALLOC_SIZE will fail. */ |
d0d8da2dc zsmalloc: require... |
1462 |
unsigned long zs_malloc(struct zs_pool *pool, size_t size, gfp_t gfp) |
61989a80f staging: zsmalloc... |
1463 |
{ |
2e40e163a zsmalloc: decoupl... |
1464 |
unsigned long handle, obj; |
61989a80f staging: zsmalloc... |
1465 |
struct size_class *class; |
48b4800a1 zsmalloc: page mi... |
1466 |
enum fullness_group newfg; |
3783689a1 zsmalloc: introdu... |
1467 |
struct zspage *zspage; |
61989a80f staging: zsmalloc... |
1468 |
|
7b60a6852 zsmalloc: record ... |
1469 |
if (unlikely(!size || size > ZS_MAX_ALLOC_SIZE)) |
2e40e163a zsmalloc: decoupl... |
1470 |
return 0; |
3783689a1 zsmalloc: introdu... |
1471 |
handle = cache_alloc_handle(pool, gfp); |
2e40e163a zsmalloc: decoupl... |
1472 |
if (!handle) |
c23443483 staging: zsmalloc... |
1473 |
return 0; |
61989a80f staging: zsmalloc... |
1474 |
|
2e40e163a zsmalloc: decoupl... |
1475 1476 |
/* extra space in chunk to keep the handle */ size += ZS_HANDLE_SIZE; |
9eec4cd53 zsmalloc: merge s... |
1477 |
class = pool->size_class[get_size_class_index(size)]; |
61989a80f staging: zsmalloc... |
1478 1479 |
spin_lock(&class->lock); |
3783689a1 zsmalloc: introdu... |
1480 |
zspage = find_get_zspage(class); |
48b4800a1 zsmalloc: page mi... |
1481 1482 1483 1484 1485 |
if (likely(zspage)) { obj = obj_malloc(class, zspage, handle); /* Now move the zspage to another fullness group, if required */ fix_fullness_group(class, zspage); record_obj(handle, obj); |
61989a80f staging: zsmalloc... |
1486 |
spin_unlock(&class->lock); |
61989a80f staging: zsmalloc... |
1487 |
|
48b4800a1 zsmalloc: page mi... |
1488 1489 |
return handle; } |
0f050d997 mm/zsmalloc: add ... |
1490 |
|
48b4800a1 zsmalloc: page mi... |
1491 1492 1493 1494 1495 1496 |
spin_unlock(&class->lock); zspage = alloc_zspage(pool, class, gfp); if (!zspage) { cache_free_handle(pool, handle); return 0; |
61989a80f staging: zsmalloc... |
1497 |
} |
48b4800a1 zsmalloc: page mi... |
1498 |
spin_lock(&class->lock); |
3783689a1 zsmalloc: introdu... |
1499 |
obj = obj_malloc(class, zspage, handle); |
48b4800a1 zsmalloc: page mi... |
1500 1501 1502 |
newfg = get_fullness_group(class, zspage); insert_zspage(class, zspage, newfg); set_zspage_mapping(zspage, class->index, newfg); |
2e40e163a zsmalloc: decoupl... |
1503 |
record_obj(handle, obj); |
48b4800a1 zsmalloc: page mi... |
1504 1505 |
atomic_long_add(class->pages_per_zspage, &pool->pages_allocated); |
b4fd07a08 mm/zsmalloc: use ... |
1506 |
zs_stat_inc(class, OBJ_ALLOCATED, class->objs_per_zspage); |
48b4800a1 zsmalloc: page mi... |
1507 1508 1509 |
/* We completely set up zspage so mark them as movable */ SetZsPageMovable(pool, zspage); |
61989a80f staging: zsmalloc... |
1510 |
spin_unlock(&class->lock); |
2e40e163a zsmalloc: decoupl... |
1511 |
return handle; |
61989a80f staging: zsmalloc... |
1512 1513 |
} EXPORT_SYMBOL_GPL(zs_malloc); |
1ee471658 zsmalloc: remove ... |
1514 |
static void obj_free(struct size_class *class, unsigned long obj) |
61989a80f staging: zsmalloc... |
1515 1516 |
{ struct link_free *link; |
3783689a1 zsmalloc: introdu... |
1517 1518 |
struct zspage *zspage; struct page *f_page; |
bfd093f5e zsmalloc: use fre... |
1519 1520 |
unsigned long f_offset; unsigned int f_objidx; |
af4ee5e97 zsmalloc: correct... |
1521 |
void *vaddr; |
61989a80f staging: zsmalloc... |
1522 |
|
312fcae22 zsmalloc: support... |
1523 |
obj &= ~OBJ_ALLOCATED_TAG; |
2e40e163a zsmalloc: decoupl... |
1524 |
obj_to_location(obj, &f_page, &f_objidx); |
bfd093f5e zsmalloc: use fre... |
1525 |
f_offset = (class->size * f_objidx) & ~PAGE_MASK; |
3783689a1 zsmalloc: introdu... |
1526 |
zspage = get_zspage(f_page); |
61989a80f staging: zsmalloc... |
1527 |
|
c78062612 zsmalloc: factor ... |
1528 |
vaddr = kmap_atomic(f_page); |
61989a80f staging: zsmalloc... |
1529 1530 |
/* Insert this object in containing zspage's freelist */ |
af4ee5e97 zsmalloc: correct... |
1531 |
link = (struct link_free *)(vaddr + f_offset); |
3b1d9ca65 zsmalloc: use OBJ... |
1532 |
link->next = get_freeobj(zspage) << OBJ_TAG_BITS; |
af4ee5e97 zsmalloc: correct... |
1533 |
kunmap_atomic(vaddr); |
bfd093f5e zsmalloc: use fre... |
1534 |
set_freeobj(zspage, f_objidx); |
3783689a1 zsmalloc: introdu... |
1535 |
mod_zspage_inuse(zspage, -1); |
0f050d997 mm/zsmalloc: add ... |
1536 |
zs_stat_dec(class, OBJ_USED, 1); |
c78062612 zsmalloc: factor ... |
1537 1538 1539 1540 |
} void zs_free(struct zs_pool *pool, unsigned long handle) { |
3783689a1 zsmalloc: introdu... |
1541 1542 |
struct zspage *zspage; struct page *f_page; |
bfd093f5e zsmalloc: use fre... |
1543 1544 |
unsigned long obj; unsigned int f_objidx; |
c78062612 zsmalloc: factor ... |
1545 1546 1547 |
int class_idx; struct size_class *class; enum fullness_group fullness; |
48b4800a1 zsmalloc: page mi... |
1548 |
bool isolated; |
c78062612 zsmalloc: factor ... |
1549 1550 1551 |
if (unlikely(!handle)) return; |
312fcae22 zsmalloc: support... |
1552 |
pin_tag(handle); |
c78062612 zsmalloc: factor ... |
1553 |
obj = handle_to_obj(handle); |
c78062612 zsmalloc: factor ... |
1554 |
obj_to_location(obj, &f_page, &f_objidx); |
3783689a1 zsmalloc: introdu... |
1555 |
zspage = get_zspage(f_page); |
c78062612 zsmalloc: factor ... |
1556 |
|
48b4800a1 zsmalloc: page mi... |
1557 |
migrate_read_lock(zspage); |
3783689a1 zsmalloc: introdu... |
1558 |
get_zspage_mapping(zspage, &class_idx, &fullness); |
c78062612 zsmalloc: factor ... |
1559 1560 1561 |
class = pool->size_class[class_idx]; spin_lock(&class->lock); |
1ee471658 zsmalloc: remove ... |
1562 |
obj_free(class, obj); |
3783689a1 zsmalloc: introdu... |
1563 |
fullness = fix_fullness_group(class, zspage); |
48b4800a1 zsmalloc: page mi... |
1564 1565 1566 |
if (fullness != ZS_EMPTY) { migrate_read_unlock(zspage); goto out; |
312fcae22 zsmalloc: support... |
1567 |
} |
48b4800a1 zsmalloc: page mi... |
1568 1569 1570 1571 1572 1573 1574 |
isolated = is_zspage_isolated(zspage); migrate_read_unlock(zspage); /* If zspage is isolated, zs_page_putback will free the zspage */ if (likely(!isolated)) free_zspage(pool, class, zspage); out: |
61989a80f staging: zsmalloc... |
1575 |
spin_unlock(&class->lock); |
312fcae22 zsmalloc: support... |
1576 |
unpin_tag(handle); |
3783689a1 zsmalloc: introdu... |
1577 |
cache_free_handle(pool, handle); |
312fcae22 zsmalloc: support... |
1578 1579 |
} EXPORT_SYMBOL_GPL(zs_free); |
251cbb951 zsmalloc: reorder... |
1580 1581 |
static void zs_object_copy(struct size_class *class, unsigned long dst, unsigned long src) |
312fcae22 zsmalloc: support... |
1582 1583 |
{ struct page *s_page, *d_page; |
bfd093f5e zsmalloc: use fre... |
1584 |
unsigned int s_objidx, d_objidx; |
312fcae22 zsmalloc: support... |
1585 1586 1587 1588 1589 1590 1591 1592 1593 |
unsigned long s_off, d_off; void *s_addr, *d_addr; int s_size, d_size, size; int written = 0; s_size = d_size = class->size; obj_to_location(src, &s_page, &s_objidx); obj_to_location(dst, &d_page, &d_objidx); |
bfd093f5e zsmalloc: use fre... |
1594 1595 |
s_off = (class->size * s_objidx) & ~PAGE_MASK; d_off = (class->size * d_objidx) & ~PAGE_MASK; |
312fcae22 zsmalloc: support... |
1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 |
if (s_off + class->size > PAGE_SIZE) s_size = PAGE_SIZE - s_off; if (d_off + class->size > PAGE_SIZE) d_size = PAGE_SIZE - d_off; s_addr = kmap_atomic(s_page); d_addr = kmap_atomic(d_page); while (1) { size = min(s_size, d_size); memcpy(d_addr + d_off, s_addr + s_off, size); written += size; if (written == class->size) break; |
495819ead zsmalloc: micro-o... |
1613 1614 1615 1616 1617 1618 |
s_off += size; s_size -= size; d_off += size; d_size -= size; if (s_off >= PAGE_SIZE) { |
312fcae22 zsmalloc: support... |
1619 1620 1621 |
kunmap_atomic(d_addr); kunmap_atomic(s_addr); s_page = get_next_page(s_page); |
312fcae22 zsmalloc: support... |
1622 1623 1624 1625 |
s_addr = kmap_atomic(s_page); d_addr = kmap_atomic(d_page); s_size = class->size - written; s_off = 0; |
312fcae22 zsmalloc: support... |
1626 |
} |
495819ead zsmalloc: micro-o... |
1627 |
if (d_off >= PAGE_SIZE) { |
312fcae22 zsmalloc: support... |
1628 1629 |
kunmap_atomic(d_addr); d_page = get_next_page(d_page); |
312fcae22 zsmalloc: support... |
1630 1631 1632 |
d_addr = kmap_atomic(d_page); d_size = class->size - written; d_off = 0; |
312fcae22 zsmalloc: support... |
1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 |
} } kunmap_atomic(d_addr); kunmap_atomic(s_addr); } /* * Find alloced object in zspage from index object and * return handle. */ |
251cbb951 zsmalloc: reorder... |
1644 |
static unsigned long find_alloced_obj(struct size_class *class, |
cf675acb7 mm/zsmalloc: take... |
1645 |
struct page *page, int *obj_idx) |
312fcae22 zsmalloc: support... |
1646 1647 1648 |
{ unsigned long head; int offset = 0; |
cf675acb7 mm/zsmalloc: take... |
1649 |
int index = *obj_idx; |
312fcae22 zsmalloc: support... |
1650 1651 |
unsigned long handle = 0; void *addr = kmap_atomic(page); |
3783689a1 zsmalloc: introdu... |
1652 |
offset = get_first_obj_offset(page); |
312fcae22 zsmalloc: support... |
1653 1654 1655 |
offset += class->size * index; while (offset < PAGE_SIZE) { |
48b4800a1 zsmalloc: page mi... |
1656 |
head = obj_to_head(page, addr + offset); |
312fcae22 zsmalloc: support... |
1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 |
if (head & OBJ_ALLOCATED_TAG) { handle = head & ~OBJ_ALLOCATED_TAG; if (trypin_tag(handle)) break; handle = 0; } offset += class->size; index++; } kunmap_atomic(addr); |
cf675acb7 mm/zsmalloc: take... |
1669 1670 |
*obj_idx = index; |
312fcae22 zsmalloc: support... |
1671 1672 1673 1674 |
return handle; } struct zs_compact_control { |
3783689a1 zsmalloc: introdu... |
1675 |
/* Source spage for migration which could be a subpage of zspage */ |
312fcae22 zsmalloc: support... |
1676 1677 1678 1679 1680 1681 |
struct page *s_page; /* Destination page for migration which should be a first page * of zspage. */ struct page *d_page; /* Starting object index within @s_page which used for live object * in the subpage. */ |
41b88e14c mm/zsmalloc: use ... |
1682 |
int obj_idx; |
312fcae22 zsmalloc: support... |
1683 1684 1685 1686 1687 1688 1689 1690 1691 |
}; static int migrate_zspage(struct zs_pool *pool, struct size_class *class, struct zs_compact_control *cc) { unsigned long used_obj, free_obj; unsigned long handle; struct page *s_page = cc->s_page; struct page *d_page = cc->d_page; |
41b88e14c mm/zsmalloc: use ... |
1692 |
int obj_idx = cc->obj_idx; |
312fcae22 zsmalloc: support... |
1693 1694 1695 |
int ret = 0; while (1) { |
cf675acb7 mm/zsmalloc: take... |
1696 |
handle = find_alloced_obj(class, s_page, &obj_idx); |
312fcae22 zsmalloc: support... |
1697 1698 1699 1700 |
if (!handle) { s_page = get_next_page(s_page); if (!s_page) break; |
41b88e14c mm/zsmalloc: use ... |
1701 |
obj_idx = 0; |
312fcae22 zsmalloc: support... |
1702 1703 1704 1705 |
continue; } /* Stop if there is no more space */ |
3783689a1 zsmalloc: introdu... |
1706 |
if (zspage_full(class, get_zspage(d_page))) { |
312fcae22 zsmalloc: support... |
1707 1708 1709 1710 1711 1712 |
unpin_tag(handle); ret = -ENOMEM; break; } used_obj = handle_to_obj(handle); |
3783689a1 zsmalloc: introdu... |
1713 |
free_obj = obj_malloc(class, get_zspage(d_page), handle); |
251cbb951 zsmalloc: reorder... |
1714 |
zs_object_copy(class, free_obj, used_obj); |
41b88e14c mm/zsmalloc: use ... |
1715 |
obj_idx++; |
c102f07ca zsmalloc: fix mig... |
1716 1717 1718 1719 1720 1721 1722 |
/* * record_obj updates handle's value to free_obj and it will * invalidate lock bit(ie, HANDLE_PIN_BIT) of handle, which * breaks synchronization using pin_tag(e,g, zs_free) so * let's keep the lock bit. */ free_obj |= BIT(HANDLE_PIN_BIT); |
312fcae22 zsmalloc: support... |
1723 1724 |
record_obj(handle, free_obj); unpin_tag(handle); |
1ee471658 zsmalloc: remove ... |
1725 |
obj_free(class, used_obj); |
312fcae22 zsmalloc: support... |
1726 1727 1728 1729 |
} /* Remember last position in this iteration */ cc->s_page = s_page; |
41b88e14c mm/zsmalloc: use ... |
1730 |
cc->obj_idx = obj_idx; |
312fcae22 zsmalloc: support... |
1731 1732 1733 |
return ret; } |
3783689a1 zsmalloc: introdu... |
1734 |
static struct zspage *isolate_zspage(struct size_class *class, bool source) |
312fcae22 zsmalloc: support... |
1735 1736 |
{ int i; |
3783689a1 zsmalloc: introdu... |
1737 1738 |
struct zspage *zspage; enum fullness_group fg[2] = {ZS_ALMOST_EMPTY, ZS_ALMOST_FULL}; |
312fcae22 zsmalloc: support... |
1739 |
|
3783689a1 zsmalloc: introdu... |
1740 1741 1742 1743 1744 1745 1746 1747 1748 |
if (!source) { fg[0] = ZS_ALMOST_FULL; fg[1] = ZS_ALMOST_EMPTY; } for (i = 0; i < 2; i++) { zspage = list_first_entry_or_null(&class->fullness_list[fg[i]], struct zspage, list); if (zspage) { |
48b4800a1 zsmalloc: page mi... |
1749 |
VM_BUG_ON(is_zspage_isolated(zspage)); |
3783689a1 zsmalloc: introdu... |
1750 1751 |
remove_zspage(class, zspage, fg[i]); return zspage; |
312fcae22 zsmalloc: support... |
1752 1753 |
} } |
3783689a1 zsmalloc: introdu... |
1754 |
return zspage; |
312fcae22 zsmalloc: support... |
1755 |
} |
860c707dc zsmalloc: account... |
1756 |
/* |
3783689a1 zsmalloc: introdu... |
1757 |
* putback_zspage - add @zspage into right class's fullness list |
860c707dc zsmalloc: account... |
1758 |
* @class: destination class |
3783689a1 zsmalloc: introdu... |
1759 |
* @zspage: target page |
860c707dc zsmalloc: account... |
1760 |
* |
3783689a1 zsmalloc: introdu... |
1761 |
* Return @zspage's fullness_group |
860c707dc zsmalloc: account... |
1762 |
*/ |
4aa409cab zsmalloc: separat... |
1763 |
static enum fullness_group putback_zspage(struct size_class *class, |
3783689a1 zsmalloc: introdu... |
1764 |
struct zspage *zspage) |
312fcae22 zsmalloc: support... |
1765 |
{ |
312fcae22 zsmalloc: support... |
1766 |
enum fullness_group fullness; |
48b4800a1 zsmalloc: page mi... |
1767 |
VM_BUG_ON(is_zspage_isolated(zspage)); |
3783689a1 zsmalloc: introdu... |
1768 1769 1770 |
fullness = get_fullness_group(class, zspage); insert_zspage(class, zspage, fullness); set_zspage_mapping(zspage, class->index, fullness); |
839373e64 zsmalloc: remove ... |
1771 |
|
860c707dc zsmalloc: account... |
1772 |
return fullness; |
61989a80f staging: zsmalloc... |
1773 |
} |
312fcae22 zsmalloc: support... |
1774 |
|
48b4800a1 zsmalloc: page mi... |
1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 |
#ifdef CONFIG_COMPACTION static struct dentry *zs_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data) { static const struct dentry_operations ops = { .d_dname = simple_dname, }; return mount_pseudo(fs_type, "zsmalloc:", NULL, &ops, ZSMALLOC_MAGIC); } static struct file_system_type zsmalloc_fs = { .name = "zsmalloc", .mount = zs_mount, .kill_sb = kill_anon_super, }; static int zsmalloc_mount(void) { int ret = 0; zsmalloc_mnt = kern_mount(&zsmalloc_fs); if (IS_ERR(zsmalloc_mnt)) ret = PTR_ERR(zsmalloc_mnt); return ret; } static void zsmalloc_unmount(void) { kern_unmount(zsmalloc_mnt); } static void migrate_lock_init(struct zspage *zspage) { rwlock_init(&zspage->lock); } static void migrate_read_lock(struct zspage *zspage) { read_lock(&zspage->lock); } static void migrate_read_unlock(struct zspage *zspage) { read_unlock(&zspage->lock); } static void migrate_write_lock(struct zspage *zspage) { write_lock(&zspage->lock); } static void migrate_write_unlock(struct zspage *zspage) { write_unlock(&zspage->lock); } /* Number of isolated subpage for *page migration* in this zspage */ static void inc_zspage_isolation(struct zspage *zspage) { zspage->isolated++; } static void dec_zspage_isolation(struct zspage *zspage) { zspage->isolated--; } static void replace_sub_page(struct size_class *class, struct zspage *zspage, struct page *newpage, struct page *oldpage) { struct page *page; struct page *pages[ZS_MAX_PAGES_PER_ZSPAGE] = {NULL, }; int idx = 0; page = get_first_page(zspage); do { if (page == oldpage) pages[idx] = newpage; else pages[idx] = page; idx++; } while ((page = get_next_page(page)) != NULL); create_page_chain(class, zspage, pages); set_first_obj_offset(newpage, get_first_obj_offset(oldpage)); if (unlikely(PageHugeObject(oldpage))) newpage->index = oldpage->index; __SetPageMovable(newpage, page_mapping(oldpage)); } bool zs_page_isolate(struct page *page, isolate_mode_t mode) { struct zs_pool *pool; struct size_class *class; int class_idx; enum fullness_group fullness; struct zspage *zspage; struct address_space *mapping; /* * Page is locked so zspage couldn't be destroyed. For detail, look at * lock_zspage in free_zspage. */ VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(PageIsolated(page), page); zspage = get_zspage(page); /* * Without class lock, fullness could be stale while class_idx is okay * because class_idx is constant unless page is freed so we should get * fullness again under class lock. */ get_zspage_mapping(zspage, &class_idx, &fullness); mapping = page_mapping(page); pool = mapping->private_data; class = pool->size_class[class_idx]; spin_lock(&class->lock); if (get_zspage_inuse(zspage) == 0) { spin_unlock(&class->lock); return false; } /* zspage is isolated for object migration */ if (list_empty(&zspage->list) && !is_zspage_isolated(zspage)) { spin_unlock(&class->lock); return false; } /* * If this is first time isolation for the zspage, isolate zspage from * size_class to prevent further object allocation from the zspage. */ if (!list_empty(&zspage->list) && !is_zspage_isolated(zspage)) { get_zspage_mapping(zspage, &class_idx, &fullness); remove_zspage(class, zspage, fullness); } inc_zspage_isolation(zspage); spin_unlock(&class->lock); return true; } int zs_page_migrate(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode) { struct zs_pool *pool; struct size_class *class; int class_idx; enum fullness_group fullness; struct zspage *zspage; struct page *dummy; void *s_addr, *d_addr, *addr; int offset, pos; unsigned long handle, head; unsigned long old_obj, new_obj; unsigned int obj_idx; int ret = -EAGAIN; VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(!PageIsolated(page), page); zspage = get_zspage(page); /* Concurrent compactor cannot migrate any subpage in zspage */ migrate_write_lock(zspage); get_zspage_mapping(zspage, &class_idx, &fullness); pool = mapping->private_data; class = pool->size_class[class_idx]; offset = get_first_obj_offset(page); spin_lock(&class->lock); if (!get_zspage_inuse(zspage)) { ret = -EBUSY; goto unlock_class; } pos = offset; s_addr = kmap_atomic(page); while (pos < PAGE_SIZE) { head = obj_to_head(page, s_addr + pos); if (head & OBJ_ALLOCATED_TAG) { handle = head & ~OBJ_ALLOCATED_TAG; if (!trypin_tag(handle)) goto unpin_objects; } pos += class->size; } /* * Here, any user cannot access all objects in the zspage so let's move. */ d_addr = kmap_atomic(newpage); memcpy(d_addr, s_addr, PAGE_SIZE); kunmap_atomic(d_addr); for (addr = s_addr + offset; addr < s_addr + pos; addr += class->size) { head = obj_to_head(page, addr); if (head & OBJ_ALLOCATED_TAG) { handle = head & ~OBJ_ALLOCATED_TAG; if (!testpin_tag(handle)) BUG(); old_obj = handle_to_obj(handle); obj_to_location(old_obj, &dummy, &obj_idx); new_obj = (unsigned long)location_to_obj(newpage, obj_idx); new_obj |= BIT(HANDLE_PIN_BIT); record_obj(handle, new_obj); } } replace_sub_page(class, zspage, newpage, page); get_page(newpage); dec_zspage_isolation(zspage); /* * Page migration is done so let's putback isolated zspage to * the list if @page is final isolated subpage in the zspage. */ if (!is_zspage_isolated(zspage)) putback_zspage(class, zspage); reset_page(page); put_page(page); page = newpage; |
dd4123f32 mm: fix build war... |
2007 |
ret = MIGRATEPAGE_SUCCESS; |
48b4800a1 zsmalloc: page mi... |
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 |
unpin_objects: for (addr = s_addr + offset; addr < s_addr + pos; addr += class->size) { head = obj_to_head(page, addr); if (head & OBJ_ALLOCATED_TAG) { handle = head & ~OBJ_ALLOCATED_TAG; if (!testpin_tag(handle)) BUG(); unpin_tag(handle); } } kunmap_atomic(s_addr); unlock_class: spin_unlock(&class->lock); migrate_write_unlock(zspage); return ret; } void zs_page_putback(struct page *page) { struct zs_pool *pool; struct size_class *class; int class_idx; enum fullness_group fg; struct address_space *mapping; struct zspage *zspage; VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(!PageIsolated(page), page); zspage = get_zspage(page); get_zspage_mapping(zspage, &class_idx, &fg); mapping = page_mapping(page); pool = mapping->private_data; class = pool->size_class[class_idx]; spin_lock(&class->lock); dec_zspage_isolation(zspage); if (!is_zspage_isolated(zspage)) { fg = putback_zspage(class, zspage); /* * Due to page_lock, we cannot free zspage immediately * so let's defer. */ if (fg == ZS_EMPTY) schedule_work(&pool->free_work); } spin_unlock(&class->lock); } const struct address_space_operations zsmalloc_aops = { .isolate_page = zs_page_isolate, .migratepage = zs_page_migrate, .putback_page = zs_page_putback, }; static int zs_register_migration(struct zs_pool *pool) { pool->inode = alloc_anon_inode(zsmalloc_mnt->mnt_sb); if (IS_ERR(pool->inode)) { pool->inode = NULL; return 1; } pool->inode->i_mapping->private_data = pool; pool->inode->i_mapping->a_ops = &zsmalloc_aops; return 0; } static void zs_unregister_migration(struct zs_pool *pool) { flush_work(&pool->free_work); |
c3491eca3 zsmalloc: Delete ... |
2081 |
iput(pool->inode); |
48b4800a1 zsmalloc: page mi... |
2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 |
} /* * Caller should hold page_lock of all pages in the zspage * In here, we cannot use zspage meta data. */ static void async_free_zspage(struct work_struct *work) { int i; struct size_class *class; unsigned int class_idx; enum fullness_group fullness; struct zspage *zspage, *tmp; LIST_HEAD(free_pages); struct zs_pool *pool = container_of(work, struct zs_pool, free_work); for (i = 0; i < zs_size_classes; i++) { class = pool->size_class[i]; if (class->index != i) continue; spin_lock(&class->lock); list_splice_init(&class->fullness_list[ZS_EMPTY], &free_pages); spin_unlock(&class->lock); } list_for_each_entry_safe(zspage, tmp, &free_pages, list) { list_del(&zspage->list); lock_zspage(zspage); get_zspage_mapping(zspage, &class_idx, &fullness); VM_BUG_ON(fullness != ZS_EMPTY); class = pool->size_class[class_idx]; spin_lock(&class->lock); __free_zspage(pool, pool->size_class[class_idx], zspage); spin_unlock(&class->lock); } }; static void kick_deferred_free(struct zs_pool *pool) { schedule_work(&pool->free_work); } static void init_deferred_free(struct zs_pool *pool) { INIT_WORK(&pool->free_work, async_free_zspage); } static void SetZsPageMovable(struct zs_pool *pool, struct zspage *zspage) { struct page *page = get_first_page(zspage); do { WARN_ON(!trylock_page(page)); __SetPageMovable(page, pool->inode->i_mapping); unlock_page(page); } while ((page = get_next_page(page)) != NULL); } #endif |
04f05909e zsmalloc: introdu... |
2144 2145 2146 2147 |
/* * * Based on the number of unused allocated objects calculate * and return the number of pages that we can free. |
04f05909e zsmalloc: introdu... |
2148 2149 2150 2151 |
*/ static unsigned long zs_can_compact(struct size_class *class) { unsigned long obj_wasted; |
44f43e99f zsmalloc: fix zs_... |
2152 2153 |
unsigned long obj_allocated = zs_stat_get(class, OBJ_ALLOCATED); unsigned long obj_used = zs_stat_get(class, OBJ_USED); |
04f05909e zsmalloc: introdu... |
2154 |
|
44f43e99f zsmalloc: fix zs_... |
2155 2156 |
if (obj_allocated <= obj_used) return 0; |
04f05909e zsmalloc: introdu... |
2157 |
|
44f43e99f zsmalloc: fix zs_... |
2158 |
obj_wasted = obj_allocated - obj_used; |
b4fd07a08 mm/zsmalloc: use ... |
2159 |
obj_wasted /= class->objs_per_zspage; |
04f05909e zsmalloc: introdu... |
2160 |
|
6cbf16b3b zsmalloc: use cla... |
2161 |
return obj_wasted * class->pages_per_zspage; |
04f05909e zsmalloc: introdu... |
2162 |
} |
7d3f39382 zsmalloc/zram: in... |
2163 |
static void __zs_compact(struct zs_pool *pool, struct size_class *class) |
312fcae22 zsmalloc: support... |
2164 |
{ |
312fcae22 zsmalloc: support... |
2165 |
struct zs_compact_control cc; |
3783689a1 zsmalloc: introdu... |
2166 2167 |
struct zspage *src_zspage; struct zspage *dst_zspage = NULL; |
312fcae22 zsmalloc: support... |
2168 |
|
312fcae22 zsmalloc: support... |
2169 |
spin_lock(&class->lock); |
3783689a1 zsmalloc: introdu... |
2170 |
while ((src_zspage = isolate_zspage(class, true))) { |
312fcae22 zsmalloc: support... |
2171 |
|
04f05909e zsmalloc: introdu... |
2172 2173 |
if (!zs_can_compact(class)) break; |
41b88e14c mm/zsmalloc: use ... |
2174 |
cc.obj_idx = 0; |
48b4800a1 zsmalloc: page mi... |
2175 |
cc.s_page = get_first_page(src_zspage); |
312fcae22 zsmalloc: support... |
2176 |
|
3783689a1 zsmalloc: introdu... |
2177 |
while ((dst_zspage = isolate_zspage(class, false))) { |
48b4800a1 zsmalloc: page mi... |
2178 |
cc.d_page = get_first_page(dst_zspage); |
312fcae22 zsmalloc: support... |
2179 |
/* |
0dc63d488 zsmalloc: cosmeti... |
2180 2181 |
* If there is no more space in dst_page, resched * and see if anyone had allocated another zspage. |
312fcae22 zsmalloc: support... |
2182 2183 2184 |
*/ if (!migrate_zspage(pool, class, &cc)) break; |
4aa409cab zsmalloc: separat... |
2185 |
putback_zspage(class, dst_zspage); |
312fcae22 zsmalloc: support... |
2186 2187 2188 |
} /* Stop if we couldn't find slot */ |
3783689a1 zsmalloc: introdu... |
2189 |
if (dst_zspage == NULL) |
312fcae22 zsmalloc: support... |
2190 |
break; |
4aa409cab zsmalloc: separat... |
2191 2192 |
putback_zspage(class, dst_zspage); if (putback_zspage(class, src_zspage) == ZS_EMPTY) { |
48b4800a1 zsmalloc: page mi... |
2193 |
free_zspage(pool, class, src_zspage); |
6cbf16b3b zsmalloc: use cla... |
2194 |
pool->stats.pages_compacted += class->pages_per_zspage; |
4aa409cab zsmalloc: separat... |
2195 |
} |
312fcae22 zsmalloc: support... |
2196 |
spin_unlock(&class->lock); |
312fcae22 zsmalloc: support... |
2197 2198 2199 |
cond_resched(); spin_lock(&class->lock); } |
3783689a1 zsmalloc: introdu... |
2200 |
if (src_zspage) |
4aa409cab zsmalloc: separat... |
2201 |
putback_zspage(class, src_zspage); |
312fcae22 zsmalloc: support... |
2202 |
|
7d3f39382 zsmalloc/zram: in... |
2203 |
spin_unlock(&class->lock); |
312fcae22 zsmalloc: support... |
2204 2205 2206 2207 2208 |
} unsigned long zs_compact(struct zs_pool *pool) { int i; |
312fcae22 zsmalloc: support... |
2209 2210 2211 2212 2213 2214 2215 2216 |
struct size_class *class; for (i = zs_size_classes - 1; i >= 0; i--) { class = pool->size_class[i]; if (!class) continue; if (class->index != i) continue; |
7d3f39382 zsmalloc/zram: in... |
2217 |
__zs_compact(pool, class); |
312fcae22 zsmalloc: support... |
2218 |
} |
860c707dc zsmalloc: account... |
2219 |
return pool->stats.pages_compacted; |
312fcae22 zsmalloc: support... |
2220 2221 |
} EXPORT_SYMBOL_GPL(zs_compact); |
61989a80f staging: zsmalloc... |
2222 |
|
7d3f39382 zsmalloc/zram: in... |
2223 2224 2225 2226 2227 |
void zs_pool_stats(struct zs_pool *pool, struct zs_pool_stats *stats) { memcpy(stats, &pool->stats, sizeof(struct zs_pool_stats)); } EXPORT_SYMBOL_GPL(zs_pool_stats); |
ab9d306d9 zsmalloc: use shr... |
2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 |
static unsigned long zs_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc) { unsigned long pages_freed; struct zs_pool *pool = container_of(shrinker, struct zs_pool, shrinker); pages_freed = pool->stats.pages_compacted; /* * Compact classes and calculate compaction delta. * Can run concurrently with a manually triggered * (by user) compaction. */ pages_freed = zs_compact(pool) - pages_freed; return pages_freed ? pages_freed : SHRINK_STOP; } static unsigned long zs_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) { int i; struct size_class *class; unsigned long pages_to_free = 0; struct zs_pool *pool = container_of(shrinker, struct zs_pool, shrinker); |
ab9d306d9 zsmalloc: use shr... |
2254 2255 2256 2257 2258 2259 |
for (i = zs_size_classes - 1; i >= 0; i--) { class = pool->size_class[i]; if (!class) continue; if (class->index != i) continue; |
ab9d306d9 zsmalloc: use shr... |
2260 |
pages_to_free += zs_can_compact(class); |
ab9d306d9 zsmalloc: use shr... |
2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 |
} return pages_to_free; } static void zs_unregister_shrinker(struct zs_pool *pool) { if (pool->shrinker_enabled) { unregister_shrinker(&pool->shrinker); pool->shrinker_enabled = false; } } static int zs_register_shrinker(struct zs_pool *pool) { pool->shrinker.scan_objects = zs_shrinker_scan; pool->shrinker.count_objects = zs_shrinker_count; pool->shrinker.batch = 0; pool->shrinker.seeks = DEFAULT_SEEKS; return register_shrinker(&pool->shrinker); } |
00a61d861 staging: zsmalloc... |
2283 |
/** |
66cdef663 mm/zsmalloc: adju... |
2284 |
* zs_create_pool - Creates an allocation pool to work from. |
fd8544639 mm/zsmalloc: keep... |
2285 |
* @name: pool name to be created |
166cfda75 staging: zsmalloc... |
2286 |
* |
66cdef663 mm/zsmalloc: adju... |
2287 2288 |
* This function must be called before anything when using * the zsmalloc allocator. |
166cfda75 staging: zsmalloc... |
2289 |
* |
66cdef663 mm/zsmalloc: adju... |
2290 2291 |
* On success, a pointer to the newly created pool is returned, * otherwise NULL. |
396b7fd6f staging/zsmalloc:... |
2292 |
*/ |
d0d8da2dc zsmalloc: require... |
2293 |
struct zs_pool *zs_create_pool(const char *name) |
61989a80f staging: zsmalloc... |
2294 |
{ |
66cdef663 mm/zsmalloc: adju... |
2295 2296 2297 |
int i; struct zs_pool *pool; struct size_class *prev_class = NULL; |
61989a80f staging: zsmalloc... |
2298 |
|
66cdef663 mm/zsmalloc: adju... |
2299 2300 2301 |
pool = kzalloc(sizeof(*pool), GFP_KERNEL); if (!pool) return NULL; |
61989a80f staging: zsmalloc... |
2302 |
|
48b4800a1 zsmalloc: page mi... |
2303 |
init_deferred_free(pool); |
66cdef663 mm/zsmalloc: adju... |
2304 2305 2306 2307 2308 2309 |
pool->size_class = kcalloc(zs_size_classes, sizeof(struct size_class *), GFP_KERNEL); if (!pool->size_class) { kfree(pool); return NULL; } |
61989a80f staging: zsmalloc... |
2310 |
|
2e40e163a zsmalloc: decoupl... |
2311 2312 2313 |
pool->name = kstrdup(name, GFP_KERNEL); if (!pool->name) goto err; |
3783689a1 zsmalloc: introdu... |
2314 |
if (create_cache(pool)) |
2e40e163a zsmalloc: decoupl... |
2315 |
goto err; |
c60369f01 staging: zsmalloc... |
2316 |
/* |
66cdef663 mm/zsmalloc: adju... |
2317 2318 |
* Iterate reversly, because, size of size_class that we want to use * for merging should be larger or equal to current size. |
c60369f01 staging: zsmalloc... |
2319 |
*/ |
66cdef663 mm/zsmalloc: adju... |
2320 2321 2322 |
for (i = zs_size_classes - 1; i >= 0; i--) { int size; int pages_per_zspage; |
64d90465f mm/zsmalloc: avoi... |
2323 |
int objs_per_zspage; |
66cdef663 mm/zsmalloc: adju... |
2324 |
struct size_class *class; |
3783689a1 zsmalloc: introdu... |
2325 |
int fullness = 0; |
c60369f01 staging: zsmalloc... |
2326 |
|
66cdef663 mm/zsmalloc: adju... |
2327 2328 2329 2330 |
size = ZS_MIN_ALLOC_SIZE + i * ZS_SIZE_CLASS_DELTA; if (size > ZS_MAX_ALLOC_SIZE) size = ZS_MAX_ALLOC_SIZE; pages_per_zspage = get_pages_per_zspage(size); |
64d90465f mm/zsmalloc: avoi... |
2331 |
objs_per_zspage = pages_per_zspage * PAGE_SIZE / size; |
61989a80f staging: zsmalloc... |
2332 |
|
66cdef663 mm/zsmalloc: adju... |
2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 |
/* * size_class is used for normal zsmalloc operation such * as alloc/free for that size. Although it is natural that we * have one size_class for each size, there is a chance that we * can get more memory utilization if we use one size_class for * many different sizes whose size_class have same * characteristics. So, we makes size_class point to * previous size_class if possible. */ if (prev_class) { |
64d90465f mm/zsmalloc: avoi... |
2343 |
if (can_merge(prev_class, pages_per_zspage, objs_per_zspage)) { |
66cdef663 mm/zsmalloc: adju... |
2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 |
pool->size_class[i] = prev_class; continue; } } class = kzalloc(sizeof(struct size_class), GFP_KERNEL); if (!class) goto err; class->size = size; class->index = i; class->pages_per_zspage = pages_per_zspage; |
64d90465f mm/zsmalloc: avoi... |
2356 |
class->objs_per_zspage = objs_per_zspage; |
66cdef663 mm/zsmalloc: adju... |
2357 2358 |
spin_lock_init(&class->lock); pool->size_class[i] = class; |
48b4800a1 zsmalloc: page mi... |
2359 2360 |
for (fullness = ZS_EMPTY; fullness < NR_ZS_FULLNESS; fullness++) |
3783689a1 zsmalloc: introdu... |
2361 |
INIT_LIST_HEAD(&class->fullness_list[fullness]); |
66cdef663 mm/zsmalloc: adju... |
2362 2363 |
prev_class = class; |
61989a80f staging: zsmalloc... |
2364 |
} |
d34f61572 mm/zsmalloc: don'... |
2365 2366 |
/* debug only, don't abort if it fails */ zs_pool_stat_create(pool, name); |
0f050d997 mm/zsmalloc: add ... |
2367 |
|
48b4800a1 zsmalloc: page mi... |
2368 2369 |
if (zs_register_migration(pool)) goto err; |
ab9d306d9 zsmalloc: use shr... |
2370 2371 2372 2373 2374 2375 |
/* * Not critical, we still can use the pool * and user can trigger compaction manually. */ if (zs_register_shrinker(pool) == 0) pool->shrinker_enabled = true; |
66cdef663 mm/zsmalloc: adju... |
2376 2377 2378 2379 2380 |
return pool; err: zs_destroy_pool(pool); return NULL; |
61989a80f staging: zsmalloc... |
2381 |
} |
66cdef663 mm/zsmalloc: adju... |
2382 |
EXPORT_SYMBOL_GPL(zs_create_pool); |
61989a80f staging: zsmalloc... |
2383 |
|
66cdef663 mm/zsmalloc: adju... |
2384 |
void zs_destroy_pool(struct zs_pool *pool) |
61989a80f staging: zsmalloc... |
2385 |
{ |
66cdef663 mm/zsmalloc: adju... |
2386 |
int i; |
61989a80f staging: zsmalloc... |
2387 |
|
ab9d306d9 zsmalloc: use shr... |
2388 |
zs_unregister_shrinker(pool); |
48b4800a1 zsmalloc: page mi... |
2389 |
zs_unregister_migration(pool); |
0f050d997 mm/zsmalloc: add ... |
2390 |
zs_pool_stat_destroy(pool); |
66cdef663 mm/zsmalloc: adju... |
2391 2392 2393 |
for (i = 0; i < zs_size_classes; i++) { int fg; struct size_class *class = pool->size_class[i]; |
61989a80f staging: zsmalloc... |
2394 |
|
66cdef663 mm/zsmalloc: adju... |
2395 2396 |
if (!class) continue; |
61989a80f staging: zsmalloc... |
2397 |
|
66cdef663 mm/zsmalloc: adju... |
2398 2399 |
if (class->index != i) continue; |
61989a80f staging: zsmalloc... |
2400 |
|
48b4800a1 zsmalloc: page mi... |
2401 |
for (fg = ZS_EMPTY; fg < NR_ZS_FULLNESS; fg++) { |
3783689a1 zsmalloc: introdu... |
2402 |
if (!list_empty(&class->fullness_list[fg])) { |
66cdef663 mm/zsmalloc: adju... |
2403 2404 2405 2406 2407 2408 2409 |
pr_info("Freeing non-empty class with size %db, fullness group %d ", class->size, fg); } } kfree(class); } |
f553646a6 staging: zsmalloc... |
2410 |
|
3783689a1 zsmalloc: introdu... |
2411 |
destroy_cache(pool); |
66cdef663 mm/zsmalloc: adju... |
2412 |
kfree(pool->size_class); |
0f050d997 mm/zsmalloc: add ... |
2413 |
kfree(pool->name); |
66cdef663 mm/zsmalloc: adju... |
2414 2415 2416 |
kfree(pool); } EXPORT_SYMBOL_GPL(zs_destroy_pool); |
b74185108 staging: zsmalloc... |
2417 |
|
66cdef663 mm/zsmalloc: adju... |
2418 2419 |
static int __init zs_init(void) { |
48b4800a1 zsmalloc: page mi... |
2420 2421 2422 2423 2424 2425 2426 |
int ret; ret = zsmalloc_mount(); if (ret) goto out; ret = zs_register_cpu_notifier(); |
66cdef663 mm/zsmalloc: adju... |
2427 |
|
0f050d997 mm/zsmalloc: add ... |
2428 2429 |
if (ret) goto notifier_fail; |
66cdef663 mm/zsmalloc: adju... |
2430 2431 2432 2433 2434 2435 |
init_zs_size_classes(); #ifdef CONFIG_ZPOOL zpool_register_driver(&zs_zpool_driver); #endif |
0f050d997 mm/zsmalloc: add ... |
2436 |
|
4abaac9b7 update "mm/zsmall... |
2437 |
zs_stat_init(); |
66cdef663 mm/zsmalloc: adju... |
2438 |
return 0; |
0f050d997 mm/zsmalloc: add ... |
2439 |
|
0f050d997 mm/zsmalloc: add ... |
2440 2441 |
notifier_fail: zs_unregister_cpu_notifier(); |
48b4800a1 zsmalloc: page mi... |
2442 2443 |
zsmalloc_unmount(); out: |
0f050d997 mm/zsmalloc: add ... |
2444 |
return ret; |
61989a80f staging: zsmalloc... |
2445 |
} |
61989a80f staging: zsmalloc... |
2446 |
|
66cdef663 mm/zsmalloc: adju... |
2447 |
static void __exit zs_exit(void) |
61989a80f staging: zsmalloc... |
2448 |
{ |
66cdef663 mm/zsmalloc: adju... |
2449 2450 2451 |
#ifdef CONFIG_ZPOOL zpool_unregister_driver(&zs_zpool_driver); #endif |
48b4800a1 zsmalloc: page mi... |
2452 |
zsmalloc_unmount(); |
66cdef663 mm/zsmalloc: adju... |
2453 |
zs_unregister_cpu_notifier(); |
0f050d997 mm/zsmalloc: add ... |
2454 2455 |
zs_stat_exit(); |
61989a80f staging: zsmalloc... |
2456 |
} |
069f101fa staging: zsmalloc... |
2457 2458 2459 2460 2461 2462 |
module_init(zs_init); module_exit(zs_exit); MODULE_LICENSE("Dual BSD/GPL"); MODULE_AUTHOR("Nitin Gupta <ngupta@vflare.org>"); |