Blame view
mm/z3fold.c
47 KB
09c434b8a
|
1 |
// SPDX-License-Identifier: GPL-2.0-only |
9a001fc19
|
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 |
/* * z3fold.c * * Author: Vitaly Wool <vitaly.wool@konsulko.com> * Copyright (C) 2016, Sony Mobile Communications Inc. * * This implementation is based on zbud written by Seth Jennings. * * z3fold is an special purpose allocator for storing compressed pages. It * can store up to three compressed pages per page which improves the * compression ratio of zbud while retaining its main concepts (e. g. always * storing an integral number of objects per page) and simplicity. * It still has simple and deterministic reclaim properties that make it * preferable to a higher density approach (with no requirement on integral * number of object per page) when reclaim is used. * * As in zbud, pages are divided into "chunks". The size of the chunks is * fixed at compile time and is determined by NCHUNKS_ORDER below. * * z3fold doesn't export any API and is meant to be used via zpool API. */ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include <linux/atomic.h> |
d30561c56
|
27 |
#include <linux/sched.h> |
1f862989b
|
28 |
#include <linux/cpumask.h> |
9a001fc19
|
29 30 31 |
#include <linux/list.h> #include <linux/mm.h> #include <linux/module.h> |
1f862989b
|
32 33 34 35 |
#include <linux/page-flags.h> #include <linux/migrate.h> #include <linux/node.h> #include <linux/compaction.h> |
d30561c56
|
36 |
#include <linux/percpu.h> |
1f862989b
|
37 |
#include <linux/mount.h> |
ea8157ab2
|
38 |
#include <linux/pseudo_fs.h> |
1f862989b
|
39 |
#include <linux/fs.h> |
9a001fc19
|
40 |
#include <linux/preempt.h> |
d30561c56
|
41 |
#include <linux/workqueue.h> |
9a001fc19
|
42 43 44 |
#include <linux/slab.h> #include <linux/spinlock.h> #include <linux/zpool.h> |
ea8157ab2
|
45 |
#include <linux/magic.h> |
af4798a5b
|
46 |
#include <linux/kmemleak.h> |
9a001fc19
|
47 |
|
7c2b8baa6
|
48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
/* * NCHUNKS_ORDER determines the internal allocation granularity, effectively * adjusting internal fragmentation. It also determines the number of * freelists maintained in each pool. NCHUNKS_ORDER of 6 means that the * allocation granularity will be in chunks of size PAGE_SIZE/64. Some chunks * in the beginning of an allocated page are occupied by z3fold header, so * NCHUNKS will be calculated to 63 (or 62 in case CONFIG_DEBUG_SPINLOCK=y), * which shows the max number of free chunks in z3fold page, also there will * be 63, or 62, respectively, freelists per pool. */ #define NCHUNKS_ORDER 6 #define CHUNK_SHIFT (PAGE_SHIFT - NCHUNKS_ORDER) #define CHUNK_SIZE (1 << CHUNK_SHIFT) #define ZHDR_SIZE_ALIGNED round_up(sizeof(struct z3fold_header), CHUNK_SIZE) #define ZHDR_CHUNKS (ZHDR_SIZE_ALIGNED >> CHUNK_SHIFT) #define TOTAL_CHUNKS (PAGE_SIZE >> CHUNK_SHIFT) |
e3c0db4fe
|
65 |
#define NCHUNKS (TOTAL_CHUNKS - ZHDR_CHUNKS) |
7c2b8baa6
|
66 67 68 69 |
#define BUDDY_MASK (0x3) #define BUDDY_SHIFT 2 #define SLOTS_ALIGN (0x40) |
9a001fc19
|
70 71 72 |
/***************** * Structures *****************/ |
ede93213a
|
73 74 75 76 77 78 79 80 81 82 |
struct z3fold_pool; struct z3fold_ops { int (*evict)(struct z3fold_pool *pool, unsigned long handle); }; enum buddy { HEADLESS = 0, FIRST, MIDDLE, LAST, |
7c2b8baa6
|
83 84 85 86 87 88 89 90 91 |
BUDDIES_MAX = LAST }; struct z3fold_buddy_slots { /* * we are using BUDDY_MASK in handle_to_buddy etc. so there should * be enough slots to hold all possible variants */ unsigned long slot[BUDDY_MASK + 1]; |
fc5488651
|
92 |
unsigned long pool; /* back link */ |
4a3ac9311
|
93 |
rwlock_t lock; |
ede93213a
|
94 |
}; |
7c2b8baa6
|
95 |
#define HANDLE_FLAG_MASK (0x03) |
ede93213a
|
96 97 |
/* |
d30561c56
|
98 |
* struct z3fold_header - z3fold page metadata occupying first chunks of each |
ede93213a
|
99 |
* z3fold page, except for HEADLESS pages |
d30561c56
|
100 101 |
* @buddy: links the z3fold page into the relevant list in the * pool |
2f1e5e4d8
|
102 |
* @page_lock: per-page lock |
d30561c56
|
103 104 |
* @refcount: reference count for the z3fold page * @work: work_struct for page layout optimization |
7c2b8baa6
|
105 |
* @slots: pointer to the structure holding buddy slots |
bb9a374df
|
106 |
* @pool: pointer to the containing pool |
d30561c56
|
107 |
* @cpu: CPU which this page "belongs" to |
ede93213a
|
108 109 110 111 |
* @first_chunks: the size of the first buddy in chunks, 0 if free * @middle_chunks: the size of the middle buddy in chunks, 0 if free * @last_chunks: the size of the last buddy in chunks, 0 if free * @first_num: the starting number (for the first handle) |
1f862989b
|
112 |
* @mapped_count: the number of objects currently mapped |
ede93213a
|
113 114 115 |
*/ struct z3fold_header { struct list_head buddy; |
2f1e5e4d8
|
116 |
spinlock_t page_lock; |
5a27aa822
|
117 |
struct kref refcount; |
d30561c56
|
118 |
struct work_struct work; |
7c2b8baa6
|
119 |
struct z3fold_buddy_slots *slots; |
bb9a374df
|
120 |
struct z3fold_pool *pool; |
d30561c56
|
121 |
short cpu; |
ede93213a
|
122 123 124 125 126 |
unsigned short first_chunks; unsigned short middle_chunks; unsigned short last_chunks; unsigned short start_middle; unsigned short first_num:2; |
1f862989b
|
127 |
unsigned short mapped_count:2; |
4a3ac9311
|
128 |
unsigned short foreign_handles:2; |
ede93213a
|
129 |
}; |
9a001fc19
|
130 131 |
/** * struct z3fold_pool - stores metadata for each z3fold pool |
d30561c56
|
132 133 134 135 136 137 |
* @name: pool name * @lock: protects pool unbuddied/lru lists * @stale_lock: protects pool stale page list * @unbuddied: per-cpu array of lists tracking z3fold pages that contain 2- * buddies; the list each z3fold page is added to depends on * the size of its free region. |
9a001fc19
|
138 139 |
* @lru: list tracking the z3fold pages in LRU order by most recently * added buddy. |
d30561c56
|
140 |
* @stale: list of pages marked for freeing |
9a001fc19
|
141 |
* @pages_nr: number of z3fold pages in the pool. |
7c2b8baa6
|
142 |
* @c_handle: cache for z3fold_buddy_slots allocation |
9a001fc19
|
143 144 |
* @ops: pointer to a structure of user defined operations specified at * pool creation time. |
30522175d
|
145 146 |
* @zpool: zpool driver * @zpool_ops: zpool operations structure with an evict callback |
d30561c56
|
147 148 149 |
* @compact_wq: workqueue for page layout background optimization * @release_wq: workqueue for safe page release * @work: work_struct for safe page release |
1f862989b
|
150 |
* @inode: inode for z3fold pseudo filesystem |
9a001fc19
|
151 152 153 154 155 |
* * This structure is allocated at pool creation time and maintains metadata * pertaining to a particular z3fold pool. */ struct z3fold_pool { |
d30561c56
|
156 |
const char *name; |
9a001fc19
|
157 |
spinlock_t lock; |
d30561c56
|
158 159 |
spinlock_t stale_lock; struct list_head *unbuddied; |
9a001fc19
|
160 |
struct list_head lru; |
d30561c56
|
161 |
struct list_head stale; |
12d59ae67
|
162 |
atomic64_t pages_nr; |
7c2b8baa6
|
163 |
struct kmem_cache *c_handle; |
9a001fc19
|
164 165 166 |
const struct z3fold_ops *ops; struct zpool *zpool; const struct zpool_ops *zpool_ops; |
d30561c56
|
167 168 169 |
struct workqueue_struct *compact_wq; struct workqueue_struct *release_wq; struct work_struct work; |
1f862989b
|
170 |
struct inode *inode; |
9a001fc19
|
171 |
}; |
9a001fc19
|
172 173 174 175 |
/* * Internal z3fold page flags */ enum z3fold_page_flags { |
5a27aa822
|
176 |
PAGE_HEADLESS = 0, |
9a001fc19
|
177 |
MIDDLE_CHUNK_MAPPED, |
d30561c56
|
178 |
NEEDS_COMPACTING, |
6098d7e13
|
179 |
PAGE_STALE, |
ca0246bb9
|
180 |
PAGE_CLAIMED, /* by either reclaim or free */ |
9a001fc19
|
181 |
}; |
4a3ac9311
|
182 |
/* |
dcf5aedb2
|
183 184 185 186 187 188 189 |
* handle flags, go under HANDLE_FLAG_MASK */ enum z3fold_handle_flags { HANDLES_NOFREE = 0, }; /* |
4a3ac9311
|
190 191 192 193 |
* Forward declarations */ static struct z3fold_header *__z3fold_alloc(struct z3fold_pool *, size_t, bool); static void compact_page_work(struct work_struct *w); |
9a001fc19
|
194 195 196 197 198 199 200 201 202 203 204 205 |
/***************** * Helpers *****************/ /* Converts an allocation size in bytes to size in z3fold chunks */ static int size_to_chunks(size_t size) { return (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT; } #define for_each_unbuddied_list(_iter, _begin) \ for ((_iter) = (_begin); (_iter) < NCHUNKS; (_iter)++) |
bb9f6f63f
|
206 207 |
static inline struct z3fold_buddy_slots *alloc_slots(struct z3fold_pool *pool, gfp_t gfp) |
7c2b8baa6
|
208 |
{ |
f1549cb5a
|
209 |
struct z3fold_buddy_slots *slots; |
f94afee99
|
210 |
slots = kmem_cache_zalloc(pool->c_handle, |
f1549cb5a
|
211 |
(gfp & ~(__GFP_HIGHMEM | __GFP_MOVABLE))); |
7c2b8baa6
|
212 213 |
if (slots) { |
af4798a5b
|
214 215 |
/* It will be freed separately in free_handle(). */ kmemleak_not_leak(slots); |
7c2b8baa6
|
216 |
slots->pool = (unsigned long)pool; |
4a3ac9311
|
217 |
rwlock_init(&slots->lock); |
7c2b8baa6
|
218 219 220 221 222 223 224 225 226 227 228 229 230 231 |
} return slots; } static inline struct z3fold_pool *slots_to_pool(struct z3fold_buddy_slots *s) { return (struct z3fold_pool *)(s->pool & ~HANDLE_FLAG_MASK); } static inline struct z3fold_buddy_slots *handle_to_slots(unsigned long handle) { return (struct z3fold_buddy_slots *)(handle & ~(SLOTS_ALIGN - 1)); } |
4a3ac9311
|
232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 |
/* Lock a z3fold page */ static inline void z3fold_page_lock(struct z3fold_header *zhdr) { spin_lock(&zhdr->page_lock); } /* Try to lock a z3fold page */ static inline int z3fold_page_trylock(struct z3fold_header *zhdr) { return spin_trylock(&zhdr->page_lock); } /* Unlock a z3fold page */ static inline void z3fold_page_unlock(struct z3fold_header *zhdr) { spin_unlock(&zhdr->page_lock); } |
767cc6c55
|
249 250 |
/* return locked z3fold page if it's not headless */ static inline struct z3fold_header *get_z3fold_header(unsigned long handle) |
4a3ac9311
|
251 252 253 254 255 256 257 258 259 260 261 262 263 |
{ struct z3fold_buddy_slots *slots; struct z3fold_header *zhdr; int locked = 0; if (!(handle & (1 << PAGE_HEADLESS))) { slots = handle_to_slots(handle); do { unsigned long addr; read_lock(&slots->lock); addr = *(unsigned long *)handle; zhdr = (struct z3fold_header *)(addr & PAGE_MASK); |
767cc6c55
|
264 |
locked = z3fold_page_trylock(zhdr); |
4a3ac9311
|
265 266 267 268 |
read_unlock(&slots->lock); if (locked) break; cpu_relax(); |
767cc6c55
|
269 |
} while (true); |
4a3ac9311
|
270 271 272 273 274 275 |
} else { zhdr = (struct z3fold_header *)(handle & PAGE_MASK); } return zhdr; } |
4a3ac9311
|
276 277 278 279 280 281 282 |
static inline void put_z3fold_header(struct z3fold_header *zhdr) { struct page *page = virt_to_page(zhdr); if (!test_bit(PAGE_HEADLESS, &page->private)) z3fold_page_unlock(zhdr); } |
fc5488651
|
283 |
static inline void free_handle(unsigned long handle, struct z3fold_header *zhdr) |
7c2b8baa6
|
284 285 286 287 288 289 290 |
{ struct z3fold_buddy_slots *slots; int i; bool is_free; if (handle & (1 << PAGE_HEADLESS)) return; |
4a3ac9311
|
291 292 |
if (WARN_ON(*(unsigned long *)handle == 0)) return; |
7c2b8baa6
|
293 |
slots = handle_to_slots(handle); |
4a3ac9311
|
294 295 |
write_lock(&slots->lock); *(unsigned long *)handle = 0; |
dcf5aedb2
|
296 297 298 299 300 |
if (test_bit(HANDLES_NOFREE, &slots->pool)) { write_unlock(&slots->lock); return; /* simple case, nothing else to do */ } |
fc5488651
|
301 302 |
if (zhdr->slots != slots) zhdr->foreign_handles--; |
4a3ac9311
|
303 |
|
7c2b8baa6
|
304 305 306 307 308 309 310 |
is_free = true; for (i = 0; i <= BUDDY_MASK; i++) { if (slots->slot[i]) { is_free = false; break; } } |
d8f117abb
|
311 |
write_unlock(&slots->lock); |
7c2b8baa6
|
312 313 314 |
if (is_free) { struct z3fold_pool *pool = slots_to_pool(slots); |
fc5488651
|
315 316 |
if (zhdr->slots == slots) zhdr->slots = NULL; |
7c2b8baa6
|
317 318 319 |
kmem_cache_free(pool->c_handle, slots); } } |
ea8157ab2
|
320 |
static int z3fold_init_fs_context(struct fs_context *fc) |
1f862989b
|
321 |
{ |
ea8157ab2
|
322 |
return init_pseudo(fc, Z3FOLD_MAGIC) ? 0 : -ENOMEM; |
1f862989b
|
323 324 325 326 |
} static struct file_system_type z3fold_fs = { .name = "z3fold", |
ea8157ab2
|
327 |
.init_fs_context = z3fold_init_fs_context, |
1f862989b
|
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 |
.kill_sb = kill_anon_super, }; static struct vfsmount *z3fold_mnt; static int z3fold_mount(void) { int ret = 0; z3fold_mnt = kern_mount(&z3fold_fs); if (IS_ERR(z3fold_mnt)) ret = PTR_ERR(z3fold_mnt); return ret; } static void z3fold_unmount(void) { kern_unmount(z3fold_mnt); } static const struct address_space_operations z3fold_aops; static int z3fold_register_migration(struct z3fold_pool *pool) { pool->inode = alloc_anon_inode(z3fold_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 = &z3fold_aops; return 0; } static void z3fold_unregister_migration(struct z3fold_pool *pool) { if (pool->inode) iput(pool->inode); |
cb152a1a9
|
366 |
} |
1f862989b
|
367 |
|
9a001fc19
|
368 |
/* Initializes the z3fold header of a newly allocated z3fold page */ |
63398413c
|
369 |
static struct z3fold_header *init_z3fold_page(struct page *page, bool headless, |
bb9f6f63f
|
370 |
struct z3fold_pool *pool, gfp_t gfp) |
9a001fc19
|
371 372 |
{ struct z3fold_header *zhdr = page_address(page); |
63398413c
|
373 |
struct z3fold_buddy_slots *slots; |
9a001fc19
|
374 375 |
INIT_LIST_HEAD(&page->lru); |
9a001fc19
|
376 377 |
clear_bit(PAGE_HEADLESS, &page->private); clear_bit(MIDDLE_CHUNK_MAPPED, &page->private); |
d30561c56
|
378 379 |
clear_bit(NEEDS_COMPACTING, &page->private); clear_bit(PAGE_STALE, &page->private); |
ca0246bb9
|
380 |
clear_bit(PAGE_CLAIMED, &page->private); |
63398413c
|
381 382 383 384 385 386 |
if (headless) return zhdr; slots = alloc_slots(pool, gfp); if (!slots) return NULL; |
9a001fc19
|
387 |
|
c457cd96f
|
388 |
memset(zhdr, 0, sizeof(*zhdr)); |
2f1e5e4d8
|
389 |
spin_lock_init(&zhdr->page_lock); |
5a27aa822
|
390 |
kref_init(&zhdr->refcount); |
d30561c56
|
391 |
zhdr->cpu = -1; |
7c2b8baa6
|
392 |
zhdr->slots = slots; |
bb9a374df
|
393 |
zhdr->pool = pool; |
9a001fc19
|
394 |
INIT_LIST_HEAD(&zhdr->buddy); |
d30561c56
|
395 |
INIT_WORK(&zhdr->work, compact_page_work); |
9a001fc19
|
396 397 398 399 |
return zhdr; } /* Resets the struct page fields and frees the page */ |
1f862989b
|
400 |
static void free_z3fold_page(struct page *page, bool headless) |
9a001fc19
|
401 |
{ |
1f862989b
|
402 403 404 405 406 407 |
if (!headless) { lock_page(page); __ClearPageMovable(page); unlock_page(page); } ClearPagePrivate(page); |
5a27aa822
|
408 409 |
__free_page(page); } |
7c2b8baa6
|
410 411 412 413 414 |
/* Helper function to build the index */ static inline int __idx(struct z3fold_header *zhdr, enum buddy bud) { return (bud + zhdr->first_num) & BUDDY_MASK; } |
9a001fc19
|
415 416 417 418 |
/* * Encodes the handle of a particular buddy within a z3fold page * Pool lock should be held as this function accesses first_num */ |
3f9d2b576
|
419 420 421 |
static unsigned long __encode_handle(struct z3fold_header *zhdr, struct z3fold_buddy_slots *slots, enum buddy bud) |
9a001fc19
|
422 |
{ |
7c2b8baa6
|
423 424 |
unsigned long h = (unsigned long)zhdr; int idx = 0; |
9a001fc19
|
425 |
|
7c2b8baa6
|
426 427 428 429 430 431 432 433 434 435 436 437 |
/* * For a headless page, its handle is its pointer with the extra * PAGE_HEADLESS bit set */ if (bud == HEADLESS) return h | (1 << PAGE_HEADLESS); /* otherwise, return pointer to encoded handle */ idx = __idx(zhdr, bud); h += idx; if (bud == LAST) h |= (zhdr->last_chunks << BUDDY_SHIFT); |
4a3ac9311
|
438 |
write_lock(&slots->lock); |
7c2b8baa6
|
439 |
slots->slot[idx] = h; |
4a3ac9311
|
440 |
write_unlock(&slots->lock); |
7c2b8baa6
|
441 |
return (unsigned long)&slots->slot[idx]; |
9a001fc19
|
442 |
} |
3f9d2b576
|
443 444 445 446 |
static unsigned long encode_handle(struct z3fold_header *zhdr, enum buddy bud) { return __encode_handle(zhdr, zhdr->slots, bud); } |
ca0246bb9
|
447 448 449 |
/* only for LAST bud, returns zero otherwise */ static unsigned short handle_to_chunks(unsigned long handle) { |
4a3ac9311
|
450 451 |
struct z3fold_buddy_slots *slots = handle_to_slots(handle); unsigned long addr; |
7c2b8baa6
|
452 |
|
4a3ac9311
|
453 454 455 |
read_lock(&slots->lock); addr = *(unsigned long *)handle; read_unlock(&slots->lock); |
7c2b8baa6
|
456 |
return (addr & ~PAGE_MASK) >> BUDDY_SHIFT; |
ca0246bb9
|
457 |
} |
f201ebd87
|
458 459 460 461 462 |
/* * (handle & BUDDY_MASK) < zhdr->first_num is possible in encode_handle * but that doesn't matter. because the masking will result in the * correct buddy number. */ |
9a001fc19
|
463 464 |
static enum buddy handle_to_buddy(unsigned long handle) { |
7c2b8baa6
|
465 |
struct z3fold_header *zhdr; |
4a3ac9311
|
466 |
struct z3fold_buddy_slots *slots = handle_to_slots(handle); |
7c2b8baa6
|
467 |
unsigned long addr; |
4a3ac9311
|
468 |
read_lock(&slots->lock); |
7c2b8baa6
|
469 470 |
WARN_ON(handle & (1 << PAGE_HEADLESS)); addr = *(unsigned long *)handle; |
4a3ac9311
|
471 |
read_unlock(&slots->lock); |
7c2b8baa6
|
472 473 |
zhdr = (struct z3fold_header *)(addr & PAGE_MASK); return (addr - zhdr->first_num) & BUDDY_MASK; |
9a001fc19
|
474 |
} |
9050cce10
|
475 476 |
static inline struct z3fold_pool *zhdr_to_pool(struct z3fold_header *zhdr) { |
bb9a374df
|
477 |
return zhdr->pool; |
9050cce10
|
478 |
} |
d30561c56
|
479 480 481 |
static void __release_z3fold_page(struct z3fold_header *zhdr, bool locked) { struct page *page = virt_to_page(zhdr); |
9050cce10
|
482 |
struct z3fold_pool *pool = zhdr_to_pool(zhdr); |
d30561c56
|
483 484 485 |
WARN_ON(!list_empty(&zhdr->buddy)); set_bit(PAGE_STALE, &page->private); |
355293574
|
486 |
clear_bit(NEEDS_COMPACTING, &page->private); |
d30561c56
|
487 488 |
spin_lock(&pool->lock); if (!list_empty(&page->lru)) |
1f862989b
|
489 |
list_del_init(&page->lru); |
d30561c56
|
490 |
spin_unlock(&pool->lock); |
4a3ac9311
|
491 |
|
d30561c56
|
492 493 |
if (locked) z3fold_page_unlock(zhdr); |
4a3ac9311
|
494 |
|
d30561c56
|
495 496 497 498 499 |
spin_lock(&pool->stale_lock); list_add(&zhdr->buddy, &pool->stale); queue_work(pool->release_wq, &pool->work); spin_unlock(&pool->stale_lock); } |
70ad3196a
|
500 |
static void release_z3fold_page(struct kref *ref) |
d30561c56
|
501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 |
{ struct z3fold_header *zhdr = container_of(ref, struct z3fold_header, refcount); __release_z3fold_page(zhdr, false); } static void release_z3fold_page_locked(struct kref *ref) { struct z3fold_header *zhdr = container_of(ref, struct z3fold_header, refcount); WARN_ON(z3fold_page_trylock(zhdr)); __release_z3fold_page(zhdr, true); } static void release_z3fold_page_locked_list(struct kref *ref) { struct z3fold_header *zhdr = container_of(ref, struct z3fold_header, refcount); |
9050cce10
|
519 |
struct z3fold_pool *pool = zhdr_to_pool(zhdr); |
4a3ac9311
|
520 |
|
9050cce10
|
521 |
spin_lock(&pool->lock); |
d30561c56
|
522 |
list_del_init(&zhdr->buddy); |
9050cce10
|
523 |
spin_unlock(&pool->lock); |
d30561c56
|
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 |
WARN_ON(z3fold_page_trylock(zhdr)); __release_z3fold_page(zhdr, true); } static void free_pages_work(struct work_struct *w) { struct z3fold_pool *pool = container_of(w, struct z3fold_pool, work); spin_lock(&pool->stale_lock); while (!list_empty(&pool->stale)) { struct z3fold_header *zhdr = list_first_entry(&pool->stale, struct z3fold_header, buddy); struct page *page = virt_to_page(zhdr); list_del(&zhdr->buddy); if (WARN_ON(!test_bit(PAGE_STALE, &page->private))) continue; |
d30561c56
|
542 543 |
spin_unlock(&pool->stale_lock); cancel_work_sync(&zhdr->work); |
1f862989b
|
544 |
free_z3fold_page(page, false); |
d30561c56
|
545 546 547 548 549 |
cond_resched(); spin_lock(&pool->stale_lock); } spin_unlock(&pool->stale_lock); } |
9a001fc19
|
550 551 552 553 554 555 556 557 558 559 560 561 562 563 |
/* * Returns the number of free chunks in a z3fold page. * NB: can't be used with HEADLESS pages. */ static int num_free_chunks(struct z3fold_header *zhdr) { int nfree; /* * If there is a middle object, pick up the bigger free space * either before or after it. Otherwise just subtract the number * of chunks occupied by the first and the last objects. */ if (zhdr->middle_chunks != 0) { int nfree_before = zhdr->first_chunks ? |
ede93213a
|
564 |
0 : zhdr->start_middle - ZHDR_CHUNKS; |
9a001fc19
|
565 |
int nfree_after = zhdr->last_chunks ? |
ede93213a
|
566 567 |
0 : TOTAL_CHUNKS - (zhdr->start_middle + zhdr->middle_chunks); |
9a001fc19
|
568 569 570 571 572 |
nfree = max(nfree_before, nfree_after); } else nfree = NCHUNKS - zhdr->first_chunks - zhdr->last_chunks; return nfree; } |
9050cce10
|
573 574 575 576 577 578 |
/* Add to the appropriate unbuddied list */ static inline void add_to_unbuddied(struct z3fold_pool *pool, struct z3fold_header *zhdr) { if (zhdr->first_chunks == 0 || zhdr->last_chunks == 0 || zhdr->middle_chunks == 0) { |
135f97fd0
|
579 |
struct list_head *unbuddied; |
9050cce10
|
580 |
int freechunks = num_free_chunks(zhdr); |
135f97fd0
|
581 582 583 |
migrate_disable(); unbuddied = this_cpu_ptr(pool->unbuddied); |
9050cce10
|
584 585 586 587 |
spin_lock(&pool->lock); list_add(&zhdr->buddy, &unbuddied[freechunks]); spin_unlock(&pool->lock); zhdr->cpu = smp_processor_id(); |
135f97fd0
|
588 |
migrate_enable(); |
9050cce10
|
589 590 |
} } |
dcf5aedb2
|
591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 |
static inline enum buddy get_free_buddy(struct z3fold_header *zhdr, int chunks) { enum buddy bud = HEADLESS; if (zhdr->middle_chunks) { if (!zhdr->first_chunks && chunks <= zhdr->start_middle - ZHDR_CHUNKS) bud = FIRST; else if (!zhdr->last_chunks) bud = LAST; } else { if (!zhdr->first_chunks) bud = FIRST; else if (!zhdr->last_chunks) bud = LAST; else bud = MIDDLE; } return bud; } |
ede93213a
|
612 613 614 615 616 617 618 619 |
static inline void *mchunk_memmove(struct z3fold_header *zhdr, unsigned short dst_chunk) { void *beg = zhdr; return memmove(beg + (dst_chunk << CHUNK_SHIFT), beg + (zhdr->start_middle << CHUNK_SHIFT), zhdr->middle_chunks << CHUNK_SHIFT); } |
4a3ac9311
|
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 |
static inline bool buddy_single(struct z3fold_header *zhdr) { return !((zhdr->first_chunks && zhdr->middle_chunks) || (zhdr->first_chunks && zhdr->last_chunks) || (zhdr->middle_chunks && zhdr->last_chunks)); } static struct z3fold_header *compact_single_buddy(struct z3fold_header *zhdr) { struct z3fold_pool *pool = zhdr_to_pool(zhdr); void *p = zhdr; unsigned long old_handle = 0; size_t sz = 0; struct z3fold_header *new_zhdr = NULL; int first_idx = __idx(zhdr, FIRST); int middle_idx = __idx(zhdr, MIDDLE); int last_idx = __idx(zhdr, LAST); unsigned short *moved_chunks = NULL; /* * No need to protect slots here -- all the slots are "local" and * the page lock is already taken */ if (zhdr->first_chunks && zhdr->slots->slot[first_idx]) { p += ZHDR_SIZE_ALIGNED; sz = zhdr->first_chunks << CHUNK_SHIFT; old_handle = (unsigned long)&zhdr->slots->slot[first_idx]; moved_chunks = &zhdr->first_chunks; } else if (zhdr->middle_chunks && zhdr->slots->slot[middle_idx]) { p += zhdr->start_middle << CHUNK_SHIFT; sz = zhdr->middle_chunks << CHUNK_SHIFT; old_handle = (unsigned long)&zhdr->slots->slot[middle_idx]; moved_chunks = &zhdr->middle_chunks; } else if (zhdr->last_chunks && zhdr->slots->slot[last_idx]) { p += PAGE_SIZE - (zhdr->last_chunks << CHUNK_SHIFT); sz = zhdr->last_chunks << CHUNK_SHIFT; old_handle = (unsigned long)&zhdr->slots->slot[last_idx]; moved_chunks = &zhdr->last_chunks; } if (sz > 0) { enum buddy new_bud = HEADLESS; short chunks = size_to_chunks(sz); void *q; new_zhdr = __z3fold_alloc(pool, sz, false); if (!new_zhdr) return NULL; if (WARN_ON(new_zhdr == zhdr)) goto out_fail; |
dcf5aedb2
|
671 |
new_bud = get_free_buddy(new_zhdr, chunks); |
4a3ac9311
|
672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 |
q = new_zhdr; switch (new_bud) { case FIRST: new_zhdr->first_chunks = chunks; q += ZHDR_SIZE_ALIGNED; break; case MIDDLE: new_zhdr->middle_chunks = chunks; new_zhdr->start_middle = new_zhdr->first_chunks + ZHDR_CHUNKS; q += new_zhdr->start_middle << CHUNK_SHIFT; break; case LAST: new_zhdr->last_chunks = chunks; q += PAGE_SIZE - (new_zhdr->last_chunks << CHUNK_SHIFT); break; default: goto out_fail; } new_zhdr->foreign_handles++; memcpy(q, p, sz); write_lock(&zhdr->slots->lock); *(unsigned long *)old_handle = (unsigned long)new_zhdr + __idx(new_zhdr, new_bud); if (new_bud == LAST) *(unsigned long *)old_handle |= (new_zhdr->last_chunks << BUDDY_SHIFT); write_unlock(&zhdr->slots->lock); add_to_unbuddied(pool, new_zhdr); z3fold_page_unlock(new_zhdr); *moved_chunks = 0; } return new_zhdr; out_fail: if (new_zhdr) { if (kref_put(&new_zhdr->refcount, release_z3fold_page_locked)) atomic64_dec(&pool->pages_nr); else { add_to_unbuddied(pool, new_zhdr); z3fold_page_unlock(new_zhdr); } } return NULL; } |
1b096e5ae
|
720 |
#define BIG_CHUNK_GAP 3 |
9a001fc19
|
721 722 723 724 |
/* Has to be called with lock held */ static int z3fold_compact_page(struct z3fold_header *zhdr) { struct page *page = virt_to_page(zhdr); |
9a001fc19
|
725 |
|
ede93213a
|
726 727 |
if (test_bit(MIDDLE_CHUNK_MAPPED, &page->private)) return 0; /* can't move middle chunk, it's used */ |
9a001fc19
|
728 |
|
1f862989b
|
729 730 |
if (unlikely(PageIsolated(page))) return 0; |
ede93213a
|
731 732 733 734 735 736 |
if (zhdr->middle_chunks == 0) return 0; /* nothing to compact */ if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) { /* move to the beginning */ mchunk_memmove(zhdr, ZHDR_CHUNKS); |
9a001fc19
|
737 738 739 740 |
zhdr->first_chunks = zhdr->middle_chunks; zhdr->middle_chunks = 0; zhdr->start_middle = 0; zhdr->first_num++; |
1b096e5ae
|
741 |
return 1; |
9a001fc19
|
742 |
} |
1b096e5ae
|
743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 |
/* * moving data is expensive, so let's only do that if * there's substantial gain (at least BIG_CHUNK_GAP chunks) */ if (zhdr->first_chunks != 0 && zhdr->last_chunks == 0 && zhdr->start_middle - (zhdr->first_chunks + ZHDR_CHUNKS) >= BIG_CHUNK_GAP) { mchunk_memmove(zhdr, zhdr->first_chunks + ZHDR_CHUNKS); zhdr->start_middle = zhdr->first_chunks + ZHDR_CHUNKS; return 1; } else if (zhdr->last_chunks != 0 && zhdr->first_chunks == 0 && TOTAL_CHUNKS - (zhdr->last_chunks + zhdr->start_middle + zhdr->middle_chunks) >= BIG_CHUNK_GAP) { unsigned short new_start = TOTAL_CHUNKS - zhdr->last_chunks - zhdr->middle_chunks; mchunk_memmove(zhdr, new_start); zhdr->start_middle = new_start; return 1; } return 0; |
9a001fc19
|
766 |
} |
d30561c56
|
767 768 |
static void do_compact_page(struct z3fold_header *zhdr, bool locked) { |
9050cce10
|
769 |
struct z3fold_pool *pool = zhdr_to_pool(zhdr); |
d30561c56
|
770 |
struct page *page; |
d30561c56
|
771 772 773 774 775 776 |
page = virt_to_page(zhdr); if (locked) WARN_ON(z3fold_page_trylock(zhdr)); else z3fold_page_lock(zhdr); |
5d03a6613
|
777 |
if (WARN_ON(!test_and_clear_bit(NEEDS_COMPACTING, &page->private))) { |
d30561c56
|
778 779 780 781 782 783 |
z3fold_page_unlock(zhdr); return; } spin_lock(&pool->lock); list_del_init(&zhdr->buddy); spin_unlock(&pool->lock); |
5d03a6613
|
784 785 786 787 |
if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { atomic64_dec(&pool->pages_nr); return; } |
dcf5aedb2
|
788 789 |
if (test_bit(PAGE_STALE, &page->private) || test_and_set_bit(PAGE_CLAIMED, &page->private)) { |
1f862989b
|
790 791 792 |
z3fold_page_unlock(zhdr); return; } |
4a3ac9311
|
793 794 795 796 |
if (!zhdr->foreign_handles && buddy_single(zhdr) && zhdr->mapped_count == 0 && compact_single_buddy(zhdr)) { if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) atomic64_dec(&pool->pages_nr); |
dcf5aedb2
|
797 798 |
else { clear_bit(PAGE_CLAIMED, &page->private); |
4a3ac9311
|
799 |
z3fold_page_unlock(zhdr); |
dcf5aedb2
|
800 |
} |
4a3ac9311
|
801 802 |
return; } |
d30561c56
|
803 |
z3fold_compact_page(zhdr); |
9050cce10
|
804 |
add_to_unbuddied(pool, zhdr); |
dcf5aedb2
|
805 |
clear_bit(PAGE_CLAIMED, &page->private); |
d30561c56
|
806 807 808 809 810 811 812 813 814 815 |
z3fold_page_unlock(zhdr); } static void compact_page_work(struct work_struct *w) { struct z3fold_header *zhdr = container_of(w, struct z3fold_header, work); do_compact_page(zhdr, false); } |
9050cce10
|
816 817 818 819 820 821 822 823 824 825 |
/* returns _locked_ z3fold page header or NULL */ static inline struct z3fold_header *__z3fold_alloc(struct z3fold_pool *pool, size_t size, bool can_sleep) { struct z3fold_header *zhdr = NULL; struct page *page; struct list_head *unbuddied; int chunks = size_to_chunks(size), i; lookup: |
135f97fd0
|
826 |
migrate_disable(); |
9050cce10
|
827 |
/* First, try to find an unbuddied z3fold page. */ |
135f97fd0
|
828 |
unbuddied = this_cpu_ptr(pool->unbuddied); |
9050cce10
|
829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 |
for_each_unbuddied_list(i, chunks) { struct list_head *l = &unbuddied[i]; zhdr = list_first_entry_or_null(READ_ONCE(l), struct z3fold_header, buddy); if (!zhdr) continue; /* Re-check under lock. */ spin_lock(&pool->lock); l = &unbuddied[i]; if (unlikely(zhdr != list_first_entry(READ_ONCE(l), struct z3fold_header, buddy)) || !z3fold_page_trylock(zhdr)) { spin_unlock(&pool->lock); zhdr = NULL; |
135f97fd0
|
846 |
migrate_enable(); |
9050cce10
|
847 848 849 850 851 852 853 854 855 |
if (can_sleep) cond_resched(); goto lookup; } list_del_init(&zhdr->buddy); zhdr->cpu = -1; spin_unlock(&pool->lock); page = virt_to_page(zhdr); |
4a3ac9311
|
856 857 |
if (test_bit(NEEDS_COMPACTING, &page->private) || test_bit(PAGE_CLAIMED, &page->private)) { |
9050cce10
|
858 859 |
z3fold_page_unlock(zhdr); zhdr = NULL; |
135f97fd0
|
860 |
migrate_enable(); |
9050cce10
|
861 862 863 864 865 866 867 868 869 870 871 872 873 874 |
if (can_sleep) cond_resched(); goto lookup; } /* * this page could not be removed from its unbuddied * list while pool lock was held, and then we've taken * page lock so kref_put could not be called before * we got here, so it's safe to just call kref_get() */ kref_get(&zhdr->refcount); break; } |
135f97fd0
|
875 |
migrate_enable(); |
9050cce10
|
876 |
|
351618b20
|
877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 |
if (!zhdr) { int cpu; /* look for _exact_ match on other cpus' lists */ for_each_online_cpu(cpu) { struct list_head *l; unbuddied = per_cpu_ptr(pool->unbuddied, cpu); spin_lock(&pool->lock); l = &unbuddied[chunks]; zhdr = list_first_entry_or_null(READ_ONCE(l), struct z3fold_header, buddy); if (!zhdr || !z3fold_page_trylock(zhdr)) { spin_unlock(&pool->lock); zhdr = NULL; continue; } list_del_init(&zhdr->buddy); zhdr->cpu = -1; spin_unlock(&pool->lock); page = virt_to_page(zhdr); |
4a3ac9311
|
901 902 |
if (test_bit(NEEDS_COMPACTING, &page->private) || test_bit(PAGE_CLAIMED, &page->private)) { |
351618b20
|
903 904 905 906 907 908 909 910 911 912 |
z3fold_page_unlock(zhdr); zhdr = NULL; if (can_sleep) cond_resched(); continue; } kref_get(&zhdr->refcount); break; } } |
fc5488651
|
913 914 915 |
if (zhdr && !zhdr->slots) zhdr->slots = alloc_slots(pool, can_sleep ? GFP_NOIO : GFP_ATOMIC); |
9050cce10
|
916 917 |
return zhdr; } |
d30561c56
|
918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 |
/* * API Functions */ /** * z3fold_create_pool() - create a new z3fold pool * @name: pool name * @gfp: gfp flags when allocating the z3fold pool structure * @ops: user-defined operations for the z3fold pool * * Return: pointer to the new z3fold pool or NULL if the metadata allocation * failed. */ static struct z3fold_pool *z3fold_create_pool(const char *name, gfp_t gfp, const struct z3fold_ops *ops) { struct z3fold_pool *pool = NULL; int i, cpu; pool = kzalloc(sizeof(struct z3fold_pool), gfp); if (!pool) goto out; |
7c2b8baa6
|
941 942 943 944 945 |
pool->c_handle = kmem_cache_create("z3fold_handle", sizeof(struct z3fold_buddy_slots), SLOTS_ALIGN, 0, NULL); if (!pool->c_handle) goto out_c; |
d30561c56
|
946 947 |
spin_lock_init(&pool->lock); spin_lock_init(&pool->stale_lock); |
e891f60e2
|
948 949 |
pool->unbuddied = __alloc_percpu(sizeof(struct list_head) * NCHUNKS, __alignof__(struct list_head)); |
1ec6995d1
|
950 951 |
if (!pool->unbuddied) goto out_pool; |
d30561c56
|
952 953 954 955 956 957 958 959 960 961 962 963 |
for_each_possible_cpu(cpu) { struct list_head *unbuddied = per_cpu_ptr(pool->unbuddied, cpu); for_each_unbuddied_list(i, 0) INIT_LIST_HEAD(&unbuddied[i]); } INIT_LIST_HEAD(&pool->lru); INIT_LIST_HEAD(&pool->stale); atomic64_set(&pool->pages_nr, 0); pool->name = name; pool->compact_wq = create_singlethread_workqueue(pool->name); if (!pool->compact_wq) |
1ec6995d1
|
964 |
goto out_unbuddied; |
d30561c56
|
965 966 967 |
pool->release_wq = create_singlethread_workqueue(pool->name); if (!pool->release_wq) goto out_wq; |
1f862989b
|
968 969 |
if (z3fold_register_migration(pool)) goto out_rwq; |
d30561c56
|
970 971 972 |
INIT_WORK(&pool->work, free_pages_work); pool->ops = ops; return pool; |
1f862989b
|
973 974 |
out_rwq: destroy_workqueue(pool->release_wq); |
d30561c56
|
975 976 |
out_wq: destroy_workqueue(pool->compact_wq); |
1ec6995d1
|
977 978 979 |
out_unbuddied: free_percpu(pool->unbuddied); out_pool: |
7c2b8baa6
|
980 981 |
kmem_cache_destroy(pool->c_handle); out_c: |
d30561c56
|
982 |
kfree(pool); |
1ec6995d1
|
983 |
out: |
d30561c56
|
984 985 986 987 988 989 990 991 992 993 994 |
return NULL; } /** * z3fold_destroy_pool() - destroys an existing z3fold pool * @pool: the z3fold pool to be destroyed * * The pool should be emptied before this function is called. */ static void z3fold_destroy_pool(struct z3fold_pool *pool) { |
7c2b8baa6
|
995 |
kmem_cache_destroy(pool->c_handle); |
6051d3bd3
|
996 997 998 999 1000 |
/* * We need to destroy pool->compact_wq before pool->release_wq, * as any pending work on pool->compact_wq will call * queue_work(pool->release_wq, &pool->work). |
b997052bc
|
1001 1002 1003 |
* * There are still outstanding pages until both workqueues are drained, * so we cannot unregister migration until then. |
6051d3bd3
|
1004 |
*/ |
d30561c56
|
1005 |
destroy_workqueue(pool->compact_wq); |
6051d3bd3
|
1006 |
destroy_workqueue(pool->release_wq); |
b997052bc
|
1007 |
z3fold_unregister_migration(pool); |
dac0d1cfd
|
1008 |
free_percpu(pool->unbuddied); |
d30561c56
|
1009 1010 |
kfree(pool); } |
9a001fc19
|
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 |
/** * z3fold_alloc() - allocates a region of a given size * @pool: z3fold pool from which to allocate * @size: size in bytes of the desired allocation * @gfp: gfp flags used if the pool needs to grow * @handle: handle of the new allocation * * This function will attempt to find a free region in the pool large enough to * satisfy the allocation request. A search of the unbuddied lists is * performed first. If no suitable free region is found, then a new page is * allocated and added to the pool to satisfy the request. * * gfp should not set __GFP_HIGHMEM as highmem pages cannot be used * as z3fold pool pages. * * Return: 0 if success and handle is set, otherwise -EINVAL if the size or * gfp arguments are invalid or -ENOMEM if the pool was unable to allocate * a new page. */ static int z3fold_alloc(struct z3fold_pool *pool, size_t size, gfp_t gfp, unsigned long *handle) { |
9050cce10
|
1033 |
int chunks = size_to_chunks(size); |
9a001fc19
|
1034 |
struct z3fold_header *zhdr = NULL; |
d30561c56
|
1035 |
struct page *page = NULL; |
9a001fc19
|
1036 |
enum buddy bud; |
8a97ea546
|
1037 |
bool can_sleep = gfpflags_allow_blocking(gfp); |
9a001fc19
|
1038 |
|
f1549cb5a
|
1039 |
if (!size) |
9a001fc19
|
1040 1041 1042 1043 1044 1045 1046 1047 |
return -EINVAL; if (size > PAGE_SIZE) return -ENOSPC; if (size > PAGE_SIZE - ZHDR_SIZE_ALIGNED - CHUNK_SIZE) bud = HEADLESS; else { |
9050cce10
|
1048 1049 |
retry: zhdr = __z3fold_alloc(pool, size, can_sleep); |
d30561c56
|
1050 |
if (zhdr) { |
dcf5aedb2
|
1051 1052 |
bud = get_free_buddy(zhdr, chunks); if (bud == HEADLESS) { |
5a27aa822
|
1053 |
if (kref_put(&zhdr->refcount, |
d30561c56
|
1054 |
release_z3fold_page_locked)) |
5a27aa822
|
1055 |
atomic64_dec(&pool->pages_nr); |
d30561c56
|
1056 1057 |
else z3fold_page_unlock(zhdr); |
2f1e5e4d8
|
1058 1059 1060 |
pr_err("No free chunks in unbuddied "); WARN_ON(1); |
9050cce10
|
1061 |
goto retry; |
9a001fc19
|
1062 |
} |
9050cce10
|
1063 |
page = virt_to_page(zhdr); |
2f1e5e4d8
|
1064 |
goto found; |
9a001fc19
|
1065 1066 |
} bud = FIRST; |
9a001fc19
|
1067 |
} |
5c9bab592
|
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 |
page = NULL; if (can_sleep) { spin_lock(&pool->stale_lock); zhdr = list_first_entry_or_null(&pool->stale, struct z3fold_header, buddy); /* * Before allocating a page, let's see if we can take one from * the stale pages list. cancel_work_sync() can sleep so we * limit this case to the contexts where we can sleep */ if (zhdr) { list_del(&zhdr->buddy); spin_unlock(&pool->stale_lock); |
d30561c56
|
1081 |
cancel_work_sync(&zhdr->work); |
5c9bab592
|
1082 1083 1084 1085 |
page = virt_to_page(zhdr); } else { spin_unlock(&pool->stale_lock); } |
d30561c56
|
1086 |
} |
5c9bab592
|
1087 1088 |
if (!page) page = alloc_page(gfp); |
d30561c56
|
1089 |
|
9a001fc19
|
1090 1091 |
if (!page) return -ENOMEM; |
2f1e5e4d8
|
1092 |
|
63398413c
|
1093 |
zhdr = init_z3fold_page(page, bud == HEADLESS, pool, gfp); |
9050cce10
|
1094 1095 1096 1097 1098 |
if (!zhdr) { __free_page(page); return -ENOMEM; } atomic64_inc(&pool->pages_nr); |
9a001fc19
|
1099 1100 1101 1102 1103 |
if (bud == HEADLESS) { set_bit(PAGE_HEADLESS, &page->private); goto headless; } |
810481a24
|
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 |
if (can_sleep) { lock_page(page); __SetPageMovable(page, pool->inode->i_mapping); unlock_page(page); } else { if (trylock_page(page)) { __SetPageMovable(page, pool->inode->i_mapping); unlock_page(page); } } |
2f1e5e4d8
|
1114 |
z3fold_page_lock(zhdr); |
9a001fc19
|
1115 1116 1117 1118 1119 1120 1121 1122 |
found: if (bud == FIRST) zhdr->first_chunks = chunks; else if (bud == LAST) zhdr->last_chunks = chunks; else { zhdr->middle_chunks = chunks; |
ede93213a
|
1123 |
zhdr->start_middle = zhdr->first_chunks + ZHDR_CHUNKS; |
9a001fc19
|
1124 |
} |
9050cce10
|
1125 |
add_to_unbuddied(pool, zhdr); |
9a001fc19
|
1126 1127 |
headless: |
d30561c56
|
1128 |
spin_lock(&pool->lock); |
9a001fc19
|
1129 1130 1131 1132 1133 1134 1135 1136 |
/* Add/move z3fold page to beginning of LRU */ if (!list_empty(&page->lru)) list_del(&page->lru); list_add(&page->lru, &pool->lru); *handle = encode_handle(zhdr, bud); spin_unlock(&pool->lock); |
2f1e5e4d8
|
1137 1138 |
if (bud != HEADLESS) z3fold_page_unlock(zhdr); |
9a001fc19
|
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 |
return 0; } /** * z3fold_free() - frees the allocation associated with the given handle * @pool: pool in which the allocation resided * @handle: handle associated with the allocation returned by z3fold_alloc() * * In the case that the z3fold page in which the allocation resides is under * reclaim, as indicated by the PG_reclaim flag being set, this function * only sets the first|last_chunks to 0. The page is actually freed * once both buddies are evicted (see z3fold_reclaim_page() below). */ static void z3fold_free(struct z3fold_pool *pool, unsigned long handle) { struct z3fold_header *zhdr; |
9a001fc19
|
1156 1157 |
struct page *page; enum buddy bud; |
5b6807de1
|
1158 |
bool page_claimed; |
9a001fc19
|
1159 |
|
4a3ac9311
|
1160 |
zhdr = get_z3fold_header(handle); |
9a001fc19
|
1161 |
page = virt_to_page(zhdr); |
5b6807de1
|
1162 |
page_claimed = test_and_set_bit(PAGE_CLAIMED, &page->private); |
9a001fc19
|
1163 1164 |
if (test_bit(PAGE_HEADLESS, &page->private)) { |
ca0246bb9
|
1165 1166 1167 1168 1169 |
/* if a headless page is under reclaim, just leave. * NB: we use test_and_set_bit for a reason: if the bit * has not been set before, we release this page * immediately so we don't care about its value any more. */ |
5b6807de1
|
1170 |
if (!page_claimed) { |
ca0246bb9
|
1171 1172 1173 |
spin_lock(&pool->lock); list_del(&page->lru); spin_unlock(&pool->lock); |
4a3ac9311
|
1174 |
put_z3fold_header(zhdr); |
1f862989b
|
1175 |
free_z3fold_page(page, true); |
ca0246bb9
|
1176 |
atomic64_dec(&pool->pages_nr); |
9a001fc19
|
1177 |
} |
ca0246bb9
|
1178 |
return; |
9a001fc19
|
1179 |
} |
ca0246bb9
|
1180 |
/* Non-headless case */ |
ca0246bb9
|
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 |
bud = handle_to_buddy(handle); switch (bud) { case FIRST: zhdr->first_chunks = 0; break; case MIDDLE: zhdr->middle_chunks = 0; break; case LAST: zhdr->last_chunks = 0; break; default: pr_err("%s: unknown bud %d ", __func__, bud); WARN_ON(1); |
4a3ac9311
|
1197 |
put_z3fold_header(zhdr); |
d30561c56
|
1198 1199 |
return; } |
4a3ac9311
|
1200 |
if (!page_claimed) |
fc5488651
|
1201 |
free_handle(handle, zhdr); |
d30561c56
|
1202 1203 1204 1205 |
if (kref_put(&zhdr->refcount, release_z3fold_page_locked_list)) { atomic64_dec(&pool->pages_nr); return; } |
5b6807de1
|
1206 1207 |
if (page_claimed) { /* the page has not been claimed by us */ |
6098d7e13
|
1208 1209 1210 |
z3fold_page_unlock(zhdr); return; } |
dcf5aedb2
|
1211 |
if (test_and_set_bit(NEEDS_COMPACTING, &page->private)) { |
4a3ac9311
|
1212 |
put_z3fold_header(zhdr); |
5b6807de1
|
1213 |
clear_bit(PAGE_CLAIMED, &page->private); |
d30561c56
|
1214 1215 1216 |
return; } if (zhdr->cpu < 0 || !cpu_online(zhdr->cpu)) { |
2f1e5e4d8
|
1217 |
spin_lock(&pool->lock); |
d30561c56
|
1218 |
list_del_init(&zhdr->buddy); |
2f1e5e4d8
|
1219 |
spin_unlock(&pool->lock); |
d30561c56
|
1220 |
zhdr->cpu = -1; |
5d03a6613
|
1221 |
kref_get(&zhdr->refcount); |
5b6807de1
|
1222 |
clear_bit(PAGE_CLAIMED, &page->private); |
4a3ac9311
|
1223 |
do_compact_page(zhdr, true); |
d30561c56
|
1224 |
return; |
9a001fc19
|
1225 |
} |
5d03a6613
|
1226 |
kref_get(&zhdr->refcount); |
5b6807de1
|
1227 |
clear_bit(PAGE_CLAIMED, &page->private); |
4a3ac9311
|
1228 1229 |
queue_work_on(zhdr->cpu, pool->compact_wq, &zhdr->work); put_z3fold_header(zhdr); |
9a001fc19
|
1230 1231 1232 1233 1234 |
} /** * z3fold_reclaim_page() - evicts allocations from a pool page and frees it * @pool: pool from which a page will attempt to be evicted |
f144c390f
|
1235 |
* @retries: number of pages on the LRU list for which eviction will |
9a001fc19
|
1236 1237 1238 1239 1240 1241 1242 1243 1244 |
* be attempted before failing * * z3fold reclaim is different from normal system reclaim in that it is done * from the bottom, up. This is because only the bottom layer, z3fold, has * information on how the allocations are organized within each z3fold page. * This has the potential to create interesting locking situations between * z3fold and the user, however. * * To avoid these, this is how z3fold_reclaim_page() should be called: |
f144c390f
|
1245 |
* |
9a001fc19
|
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 |
* The user detects a page should be reclaimed and calls z3fold_reclaim_page(). * z3fold_reclaim_page() will remove a z3fold page from the pool LRU list and * call the user-defined eviction handler with the pool and handle as * arguments. * * If the handle can not be evicted, the eviction handler should return * non-zero. z3fold_reclaim_page() will add the z3fold page back to the * appropriate list and try the next z3fold page on the LRU up to * a user defined number of retries. * * If the handle is successfully evicted, the eviction handler should * return 0 _and_ should have called z3fold_free() on the handle. z3fold_free() * contains logic to delay freeing the page if the page is under reclaim, * as indicated by the setting of the PG_reclaim flag on the underlying page. * * If all buddies in the z3fold page are successfully evicted, then the * z3fold page can be freed. * * Returns: 0 if page is successfully freed, otherwise -EINVAL if there are * no pages to evict or an eviction handler is not registered, -EAGAIN if * the retry limit was hit. */ static int z3fold_reclaim_page(struct z3fold_pool *pool, unsigned int retries) { |
4a3ac9311
|
1270 |
int i, ret = -1; |
d30561c56
|
1271 1272 1273 |
struct z3fold_header *zhdr = NULL; struct page *page = NULL; struct list_head *pos; |
9a001fc19
|
1274 |
unsigned long first_handle = 0, middle_handle = 0, last_handle = 0; |
dcf5aedb2
|
1275 1276 1277 1278 |
struct z3fold_buddy_slots slots __attribute__((aligned(SLOTS_ALIGN))); rwlock_init(&slots.lock); slots.pool = (unsigned long)pool | (1 << HANDLES_NOFREE); |
9a001fc19
|
1279 1280 |
spin_lock(&pool->lock); |
2f1e5e4d8
|
1281 |
if (!pool->ops || !pool->ops->evict || retries == 0) { |
9a001fc19
|
1282 1283 1284 1285 |
spin_unlock(&pool->lock); return -EINVAL; } for (i = 0; i < retries; i++) { |
2f1e5e4d8
|
1286 1287 1288 1289 |
if (list_empty(&pool->lru)) { spin_unlock(&pool->lock); return -EINVAL; } |
d30561c56
|
1290 1291 |
list_for_each_prev(pos, &pool->lru) { page = list_entry(pos, struct page, lru); |
ca0246bb9
|
1292 |
|
3f9d2b576
|
1293 |
zhdr = page_address(page); |
6d679578f
|
1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 |
if (test_bit(PAGE_HEADLESS, &page->private)) { /* * For non-headless pages, we wait to do this * until we have the page lock to avoid racing * with __z3fold_alloc(). Headless pages don't * have a lock (and __z3fold_alloc() will never * see them), but we still need to test and set * PAGE_CLAIMED to avoid racing with * z3fold_free(), so just do it now before * leaving the loop. */ if (test_and_set_bit(PAGE_CLAIMED, &page->private)) continue; |
d30561c56
|
1307 |
break; |
6d679578f
|
1308 |
} |
d30561c56
|
1309 |
|
dcf5aedb2
|
1310 1311 1312 1313 |
if (kref_get_unless_zero(&zhdr->refcount) == 0) { zhdr = NULL; break; } |
ca0246bb9
|
1314 |
if (!z3fold_page_trylock(zhdr)) { |
dcf5aedb2
|
1315 1316 1317 |
if (kref_put(&zhdr->refcount, release_z3fold_page)) atomic64_dec(&pool->pages_nr); |
ca0246bb9
|
1318 |
zhdr = NULL; |
d30561c56
|
1319 |
continue; /* can't evict at this point */ |
ca0246bb9
|
1320 |
} |
dcf5aedb2
|
1321 1322 1323 1324 1325 1326 1327 1328 |
/* test_and_set_bit is of course atomic, but we still * need to do it under page lock, otherwise checking * that bit in __z3fold_alloc wouldn't make sense */ if (zhdr->foreign_handles || test_and_set_bit(PAGE_CLAIMED, &page->private)) { if (kref_put(&zhdr->refcount, |
28473d91f
|
1329 |
release_z3fold_page_locked)) |
dcf5aedb2
|
1330 1331 1332 |
atomic64_dec(&pool->pages_nr); else z3fold_page_unlock(zhdr); |
4a3ac9311
|
1333 1334 1335 |
zhdr = NULL; continue; /* can't evict such page */ } |
d30561c56
|
1336 1337 |
list_del_init(&zhdr->buddy); zhdr->cpu = -1; |
6098d7e13
|
1338 |
break; |
d30561c56
|
1339 |
} |
ca0246bb9
|
1340 1341 |
if (!zhdr) break; |
5a27aa822
|
1342 |
list_del_init(&page->lru); |
d30561c56
|
1343 |
spin_unlock(&pool->lock); |
9a001fc19
|
1344 |
|
9a001fc19
|
1345 |
if (!test_bit(PAGE_HEADLESS, &page->private)) { |
9a001fc19
|
1346 |
/* |
3f9d2b576
|
1347 1348 1349 1350 |
* We need encode the handles before unlocking, and * use our local slots structure because z3fold_free * can zero out zhdr->slots and we can't do much * about that |
9a001fc19
|
1351 1352 1353 1354 |
*/ first_handle = 0; last_handle = 0; middle_handle = 0; |
dcf5aedb2
|
1355 |
memset(slots.slot, 0, sizeof(slots.slot)); |
9a001fc19
|
1356 |
if (zhdr->first_chunks) |
dcf5aedb2
|
1357 1358 |
first_handle = __encode_handle(zhdr, &slots, FIRST); |
9a001fc19
|
1359 |
if (zhdr->middle_chunks) |
dcf5aedb2
|
1360 1361 |
middle_handle = __encode_handle(zhdr, &slots, MIDDLE); |
9a001fc19
|
1362 |
if (zhdr->last_chunks) |
dcf5aedb2
|
1363 1364 |
last_handle = __encode_handle(zhdr, &slots, LAST); |
d30561c56
|
1365 1366 1367 1368 |
/* * it's safe to unlock here because we hold a * reference to this page */ |
2f1e5e4d8
|
1369 |
z3fold_page_unlock(zhdr); |
9a001fc19
|
1370 |
} else { |
4a3ac9311
|
1371 |
first_handle = encode_handle(zhdr, HEADLESS); |
9a001fc19
|
1372 1373 |
last_handle = middle_handle = 0; } |
9a001fc19
|
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 |
/* Issue the eviction callback(s) */ if (middle_handle) { ret = pool->ops->evict(pool, middle_handle); if (ret) goto next; } if (first_handle) { ret = pool->ops->evict(pool, first_handle); if (ret) goto next; } if (last_handle) { ret = pool->ops->evict(pool, last_handle); if (ret) goto next; } next: |
5a27aa822
|
1391 1392 |
if (test_bit(PAGE_HEADLESS, &page->private)) { if (ret == 0) { |
1f862989b
|
1393 |
free_z3fold_page(page, true); |
ca0246bb9
|
1394 |
atomic64_dec(&pool->pages_nr); |
5a27aa822
|
1395 |
return 0; |
5a27aa822
|
1396 |
} |
6098d7e13
|
1397 1398 1399 |
spin_lock(&pool->lock); list_add(&page->lru, &pool->lru); spin_unlock(&pool->lock); |
3f9d2b576
|
1400 |
clear_bit(PAGE_CLAIMED, &page->private); |
6098d7e13
|
1401 |
} else { |
dcf5aedb2
|
1402 |
struct z3fold_buddy_slots *slots = zhdr->slots; |
6098d7e13
|
1403 |
z3fold_page_lock(zhdr); |
6098d7e13
|
1404 1405 |
if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { |
dcf5aedb2
|
1406 |
kmem_cache_free(pool->c_handle, slots); |
6098d7e13
|
1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 |
atomic64_dec(&pool->pages_nr); return 0; } /* * if we are here, the page is still not completely * free. Take the global pool lock then to be able * to add it back to the lru list */ spin_lock(&pool->lock); list_add(&page->lru, &pool->lru); |
d5567c9df
|
1417 |
spin_unlock(&pool->lock); |
6098d7e13
|
1418 |
z3fold_page_unlock(zhdr); |
3f9d2b576
|
1419 |
clear_bit(PAGE_CLAIMED, &page->private); |
5a27aa822
|
1420 |
} |
2f1e5e4d8
|
1421 |
|
6098d7e13
|
1422 1423 |
/* We started off locked to we need to lock the pool back */ spin_lock(&pool->lock); |
9a001fc19
|
1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 |
} spin_unlock(&pool->lock); return -EAGAIN; } /** * z3fold_map() - maps the allocation associated with the given handle * @pool: pool in which the allocation resides * @handle: handle associated with the allocation to be mapped * * Extracts the buddy number from handle and constructs the pointer to the * correct starting chunk within the page. * * Returns: a pointer to the mapped allocation */ static void *z3fold_map(struct z3fold_pool *pool, unsigned long handle) { struct z3fold_header *zhdr; struct page *page; void *addr; enum buddy buddy; |
4a3ac9311
|
1445 |
zhdr = get_z3fold_header(handle); |
9a001fc19
|
1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 |
addr = zhdr; page = virt_to_page(zhdr); if (test_bit(PAGE_HEADLESS, &page->private)) goto out; buddy = handle_to_buddy(handle); switch (buddy) { case FIRST: addr += ZHDR_SIZE_ALIGNED; break; case MIDDLE: addr += zhdr->start_middle << CHUNK_SHIFT; set_bit(MIDDLE_CHUNK_MAPPED, &page->private); break; case LAST: |
ca0246bb9
|
1462 |
addr += PAGE_SIZE - (handle_to_chunks(handle) << CHUNK_SHIFT); |
9a001fc19
|
1463 1464 1465 1466 1467 1468 1469 1470 |
break; default: pr_err("unknown buddy id %d ", buddy); WARN_ON(1); addr = NULL; break; } |
2f1e5e4d8
|
1471 |
|
1f862989b
|
1472 1473 |
if (addr) zhdr->mapped_count++; |
9a001fc19
|
1474 |
out: |
4a3ac9311
|
1475 |
put_z3fold_header(zhdr); |
9a001fc19
|
1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 |
return addr; } /** * z3fold_unmap() - unmaps the allocation associated with the given handle * @pool: pool in which the allocation resides * @handle: handle associated with the allocation to be unmapped */ static void z3fold_unmap(struct z3fold_pool *pool, unsigned long handle) { struct z3fold_header *zhdr; struct page *page; enum buddy buddy; |
4a3ac9311
|
1489 |
zhdr = get_z3fold_header(handle); |
9a001fc19
|
1490 |
page = virt_to_page(zhdr); |
2f1e5e4d8
|
1491 |
if (test_bit(PAGE_HEADLESS, &page->private)) |
9a001fc19
|
1492 |
return; |
9a001fc19
|
1493 1494 1495 1496 |
buddy = handle_to_buddy(handle); if (buddy == MIDDLE) clear_bit(MIDDLE_CHUNK_MAPPED, &page->private); |
1f862989b
|
1497 |
zhdr->mapped_count--; |
4a3ac9311
|
1498 |
put_z3fold_header(zhdr); |
9a001fc19
|
1499 1500 1501 1502 1503 1504 |
} /** * z3fold_get_pool_size() - gets the z3fold pool size in pages * @pool: pool whose size is being queried * |
12d59ae67
|
1505 |
* Returns: size in pages of the given pool. |
9a001fc19
|
1506 1507 1508 |
*/ static u64 z3fold_get_pool_size(struct z3fold_pool *pool) { |
12d59ae67
|
1509 |
return atomic64_read(&pool->pages_nr); |
9a001fc19
|
1510 |
} |
1f862989b
|
1511 1512 1513 1514 1515 1516 1517 |
static bool z3fold_page_isolate(struct page *page, isolate_mode_t mode) { struct z3fold_header *zhdr; struct z3fold_pool *pool; VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(PageIsolated(page), page); |
dcf5aedb2
|
1518 |
if (test_bit(PAGE_HEADLESS, &page->private)) |
1f862989b
|
1519 1520 1521 1522 1523 1524 1525 |
return false; zhdr = page_address(page); z3fold_page_lock(zhdr); if (test_bit(NEEDS_COMPACTING, &page->private) || test_bit(PAGE_STALE, &page->private)) goto out; |
4a3ac9311
|
1526 1527 |
if (zhdr->mapped_count != 0 || zhdr->foreign_handles != 0) goto out; |
dcf5aedb2
|
1528 1529 |
if (test_and_set_bit(PAGE_CLAIMED, &page->private)) goto out; |
1f862989b
|
1530 |
pool = zhdr_to_pool(zhdr); |
4a3ac9311
|
1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 |
spin_lock(&pool->lock); if (!list_empty(&zhdr->buddy)) list_del_init(&zhdr->buddy); if (!list_empty(&page->lru)) list_del_init(&page->lru); spin_unlock(&pool->lock); kref_get(&zhdr->refcount); z3fold_page_unlock(zhdr); return true; |
1f862989b
|
1541 |
|
1f862989b
|
1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 |
out: z3fold_page_unlock(zhdr); return false; } static int z3fold_page_migrate(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode) { struct z3fold_header *zhdr, *new_zhdr; struct z3fold_pool *pool; struct address_space *new_mapping; VM_BUG_ON_PAGE(!PageMovable(page), page); VM_BUG_ON_PAGE(!PageIsolated(page), page); |
dcf5aedb2
|
1556 |
VM_BUG_ON_PAGE(!test_bit(PAGE_CLAIMED, &page->private), page); |
810481a24
|
1557 |
VM_BUG_ON_PAGE(!PageLocked(newpage), newpage); |
1f862989b
|
1558 1559 1560 |
zhdr = page_address(page); pool = zhdr_to_pool(zhdr); |
dcf5aedb2
|
1561 |
if (!z3fold_page_trylock(zhdr)) |
1f862989b
|
1562 |
return -EAGAIN; |
4a3ac9311
|
1563 |
if (zhdr->mapped_count != 0 || zhdr->foreign_handles != 0) { |
1f862989b
|
1564 |
z3fold_page_unlock(zhdr); |
dcf5aedb2
|
1565 |
clear_bit(PAGE_CLAIMED, &page->private); |
1f862989b
|
1566 1567 |
return -EBUSY; } |
c92d2f385
|
1568 1569 1570 1571 |
if (work_pending(&zhdr->work)) { z3fold_page_unlock(zhdr); return -EAGAIN; } |
1f862989b
|
1572 1573 1574 1575 1576 1577 |
new_zhdr = page_address(newpage); memcpy(new_zhdr, zhdr, PAGE_SIZE); newpage->private = page->private; page->private = 0; z3fold_page_unlock(zhdr); spin_lock_init(&new_zhdr->page_lock); |
c92d2f385
|
1578 1579 1580 1581 1582 1583 |
INIT_WORK(&new_zhdr->work, compact_page_work); /* * z3fold_page_isolate() ensures that new_zhdr->buddy is empty, * so we only have to reinitialize it. */ INIT_LIST_HEAD(&new_zhdr->buddy); |
1f862989b
|
1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 |
new_mapping = page_mapping(page); __ClearPageMovable(page); ClearPagePrivate(page); get_page(newpage); z3fold_page_lock(new_zhdr); if (new_zhdr->first_chunks) encode_handle(new_zhdr, FIRST); if (new_zhdr->last_chunks) encode_handle(new_zhdr, LAST); if (new_zhdr->middle_chunks) encode_handle(new_zhdr, MIDDLE); set_bit(NEEDS_COMPACTING, &newpage->private); new_zhdr->cpu = smp_processor_id(); spin_lock(&pool->lock); list_add(&newpage->lru, &pool->lru); spin_unlock(&pool->lock); __SetPageMovable(newpage, new_mapping); z3fold_page_unlock(new_zhdr); queue_work_on(new_zhdr->cpu, pool->compact_wq, &new_zhdr->work); page_mapcount_reset(page); |
dcf5aedb2
|
1607 |
clear_bit(PAGE_CLAIMED, &page->private); |
1f862989b
|
1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 |
put_page(page); return 0; } static void z3fold_page_putback(struct page *page) { struct z3fold_header *zhdr; struct z3fold_pool *pool; zhdr = page_address(page); pool = zhdr_to_pool(zhdr); z3fold_page_lock(zhdr); if (!list_empty(&zhdr->buddy)) list_del_init(&zhdr->buddy); INIT_LIST_HEAD(&page->lru); if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) { atomic64_dec(&pool->pages_nr); return; } spin_lock(&pool->lock); list_add(&page->lru, &pool->lru); spin_unlock(&pool->lock); |
dcf5aedb2
|
1631 |
clear_bit(PAGE_CLAIMED, &page->private); |
1f862989b
|
1632 1633 1634 1635 1636 1637 1638 1639 |
z3fold_page_unlock(zhdr); } static const struct address_space_operations z3fold_aops = { .isolate_page = z3fold_page_isolate, .migratepage = z3fold_page_migrate, .putback_page = z3fold_page_putback, }; |
9a001fc19
|
1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 |
/***************** * zpool ****************/ static int z3fold_zpool_evict(struct z3fold_pool *pool, unsigned long handle) { if (pool->zpool && pool->zpool_ops && pool->zpool_ops->evict) return pool->zpool_ops->evict(pool->zpool, handle); else return -ENOENT; } static const struct z3fold_ops z3fold_zpool_ops = { .evict = z3fold_zpool_evict }; static void *z3fold_zpool_create(const char *name, gfp_t gfp, const struct zpool_ops *zpool_ops, struct zpool *zpool) { struct z3fold_pool *pool; |
d30561c56
|
1661 1662 |
pool = z3fold_create_pool(name, gfp, zpool_ops ? &z3fold_zpool_ops : NULL); |
9a001fc19
|
1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 |
if (pool) { pool->zpool = zpool; pool->zpool_ops = zpool_ops; } return pool; } static void z3fold_zpool_destroy(void *pool) { z3fold_destroy_pool(pool); } static int z3fold_zpool_malloc(void *pool, size_t size, gfp_t gfp, unsigned long *handle) { return z3fold_alloc(pool, size, gfp, handle); } static void z3fold_zpool_free(void *pool, unsigned long handle) { z3fold_free(pool, handle); } static int z3fold_zpool_shrink(void *pool, unsigned int pages, unsigned int *reclaimed) { unsigned int total = 0; int ret = -EINVAL; while (total < pages) { ret = z3fold_reclaim_page(pool, 8); if (ret < 0) break; total++; } if (reclaimed) *reclaimed = total; return ret; } static void *z3fold_zpool_map(void *pool, unsigned long handle, enum zpool_mapmode mm) { return z3fold_map(pool, handle); } static void z3fold_zpool_unmap(void *pool, unsigned long handle) { z3fold_unmap(pool, handle); } static u64 z3fold_zpool_total_size(void *pool) { return z3fold_get_pool_size(pool) * PAGE_SIZE; } static struct zpool_driver z3fold_zpool_driver = { .type = "z3fold", |
e818e820c
|
1721 |
.sleep_mapped = true, |
9a001fc19
|
1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 |
.owner = THIS_MODULE, .create = z3fold_zpool_create, .destroy = z3fold_zpool_destroy, .malloc = z3fold_zpool_malloc, .free = z3fold_zpool_free, .shrink = z3fold_zpool_shrink, .map = z3fold_zpool_map, .unmap = z3fold_zpool_unmap, .total_size = z3fold_zpool_total_size, }; MODULE_ALIAS("zpool-z3fold"); static int __init init_z3fold(void) { |
1f862989b
|
1737 |
int ret; |
014284a08
|
1738 1739 1740 1741 1742 |
/* * Make sure the z3fold header is not larger than the page size and * there has remaining spaces for its buddy. */ BUILD_BUG_ON(ZHDR_SIZE_ALIGNED > PAGE_SIZE - CHUNK_SIZE); |
1f862989b
|
1743 1744 1745 |
ret = z3fold_mount(); if (ret) return ret; |
9a001fc19
|
1746 1747 1748 1749 1750 1751 1752 |
zpool_register_driver(&z3fold_zpool_driver); return 0; } static void __exit exit_z3fold(void) { |
1f862989b
|
1753 |
z3fold_unmount(); |
9a001fc19
|
1754 1755 1756 1757 1758 1759 1760 1761 1762 |
zpool_unregister_driver(&z3fold_zpool_driver); } module_init(init_z3fold); module_exit(exit_z3fold); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Vitaly Wool <vitalywool@gmail.com>"); MODULE_DESCRIPTION("3-Fold Allocator for Compressed Pages"); |