^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * zbud.c
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Copyright (C) 2013, Seth Jennings, IBM
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) * Concepts based on zcache internal zbud allocator by Dan Magenheimer.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) * zbud is an special purpose allocator for storing compressed pages. Contrary
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * to what its name may suggest, zbud is not a buddy allocator, but rather an
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) * allocator that "buddies" two compressed pages together in a single memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) * page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * While this design limits storage density, it has simple and deterministic
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * reclaim properties that make it preferable to a higher density approach when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) * reclaim will be used.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) * zbud works by storing compressed pages, or "zpages", together in pairs in a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) * single memory page called a "zbud page". The first buddy is "left
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) * justified" at the beginning of the zbud page, and the last buddy is "right
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) * justified" at the end of the zbud page. The benefit is that if either
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) * buddy is freed, the freed buddy space, coalesced with whatever slack space
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) * that existed between the buddies, results in the largest possible free region
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) * within the zbud page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) * zbud also provides an attractive lower bound on density. The ratio of zpages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) * to zbud pages can not be less than 1. This ensures that zbud can never "do
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) * harm" by using more pages to store zpages than the uncompressed zpages would
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * have used on their own.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) * zbud pages are divided into "chunks". The size of the chunks is fixed at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) * compile time and determined by NCHUNKS_ORDER below. Dividing zbud pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) * into chunks allows organizing unbuddied zbud pages into a manageable number
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) * of unbuddied lists according to the number of free chunks available in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) * zbud page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * The zbud API differs from that of conventional allocators in that the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) * allocation function, zbud_alloc(), returns an opaque handle to the user,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) * not a dereferenceable pointer. The user must map the handle using
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) * zbud_map() in order to get a usable pointer by which to access the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) * allocation data and unmap the handle with zbud_unmap() when operations
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) * on the allocation data are complete.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) #include <linux/atomic.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) #include <linux/list.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) #include <linux/module.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) #include <linux/preempt.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) #include <linux/spinlock.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) #include <linux/zbud.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) #include <linux/zpool.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) * Structures
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) *****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) * NCHUNKS_ORDER determines the internal allocation granularity, effectively
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) * adjusting internal fragmentation. It also determines the number of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) * freelists maintained in each pool. NCHUNKS_ORDER of 6 means that the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) * allocation granularity will be in chunks of size PAGE_SIZE/64. As one chunk
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) * in allocated page is occupied by zbud header, NCHUNKS will be calculated to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) * 63 which shows the max number of free chunks in zbud page, also there will be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) * 63 freelists per pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) #define NCHUNKS_ORDER 6
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) #define CHUNK_SHIFT (PAGE_SHIFT - NCHUNKS_ORDER)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) #define CHUNK_SIZE (1 << CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) #define ZHDR_SIZE_ALIGNED CHUNK_SIZE
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) #define NCHUNKS ((PAGE_SIZE - ZHDR_SIZE_ALIGNED) >> CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) * struct zbud_pool - stores metadata for each zbud pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) * @lock: protects all pool fields and first|last_chunk fields of any
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) * zbud page in the pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) * @unbuddied: array of lists tracking zbud pages that only contain one buddy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) * the lists each zbud page is added to depends on the size of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) * its free region.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) * @buddied: list tracking the zbud pages that contain two buddies;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) * these zbud pages are full
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) * @lru: list tracking the zbud pages in LRU order by most recently
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) * added buddy.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) * @pages_nr: number of zbud pages in the pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) * @ops: pointer to a structure of user defined operations specified at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) * pool creation time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) * This structure is allocated at pool creation time and maintains metadata
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) * pertaining to a particular zbud pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) struct zbud_pool {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) spinlock_t lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) struct list_head unbuddied[NCHUNKS];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) struct list_head buddied;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) struct list_head lru;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) u64 pages_nr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) const struct zbud_ops *ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) #ifdef CONFIG_ZPOOL
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) struct zpool *zpool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) const struct zpool_ops *zpool_ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) * struct zbud_header - zbud page metadata occupying the first chunk of each
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) * zbud page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) * @buddy: links the zbud page into the unbuddied/buddied lists in the pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) * @lru: links the zbud page into the lru list in the pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) * @first_chunks: the size of the first buddy in chunks, 0 if free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) * @last_chunks: the size of the last buddy in chunks, 0 if free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) struct zbud_header {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) struct list_head buddy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) struct list_head lru;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) unsigned int first_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) unsigned int last_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) bool under_reclaim;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) * zpool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) ****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) #ifdef CONFIG_ZPOOL
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) static int zbud_zpool_evict(struct zbud_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) if (pool->zpool && pool->zpool_ops && pool->zpool_ops->evict)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) return pool->zpool_ops->evict(pool->zpool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) return -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) static const struct zbud_ops zbud_zpool_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) .evict = zbud_zpool_evict
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) static void *zbud_zpool_create(const char *name, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) const struct zpool_ops *zpool_ops,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) struct zpool *zpool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) struct zbud_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) pool = zbud_create_pool(gfp, zpool_ops ? &zbud_zpool_ops : NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) if (pool) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) pool->zpool = zpool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) pool->zpool_ops = zpool_ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) return pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) static void zbud_zpool_destroy(void *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) zbud_destroy_pool(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) static int zbud_zpool_malloc(void *pool, size_t size, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) unsigned long *handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) return zbud_alloc(pool, size, gfp, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) static void zbud_zpool_free(void *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) zbud_free(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) static int zbud_zpool_shrink(void *pool, unsigned int pages,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) unsigned int *reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) unsigned int total = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) int ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) while (total < pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) ret = zbud_reclaim_page(pool, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) total++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) if (reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) *reclaimed = total;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) static void *zbud_zpool_map(void *pool, unsigned long handle,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) enum zpool_mapmode mm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) return zbud_map(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) static void zbud_zpool_unmap(void *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) zbud_unmap(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) static u64 zbud_zpool_total_size(void *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) return zbud_get_pool_size(pool) * PAGE_SIZE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) static struct zpool_driver zbud_zpool_driver = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) .type = "zbud",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) .owner = THIS_MODULE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) .create = zbud_zpool_create,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) .destroy = zbud_zpool_destroy,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) .malloc = zbud_zpool_malloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) .free = zbud_zpool_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) .shrink = zbud_zpool_shrink,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) .map = zbud_zpool_map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) .unmap = zbud_zpool_unmap,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) .total_size = zbud_zpool_total_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) MODULE_ALIAS("zpool-zbud");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) #endif /* CONFIG_ZPOOL */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) * Helpers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) *****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) /* Just to make the code easier to read */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) enum buddy {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) FIRST,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) LAST
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) /* Converts an allocation size in bytes to size in zbud chunks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) static int size_to_chunks(size_t size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) return (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) #define for_each_unbuddied_list(_iter, _begin) \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) for ((_iter) = (_begin); (_iter) < NCHUNKS; (_iter)++)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) /* Initializes the zbud header of a newly allocated zbud page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) static struct zbud_header *init_zbud_page(struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) struct zbud_header *zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) zhdr->first_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) zhdr->last_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) INIT_LIST_HEAD(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) INIT_LIST_HEAD(&zhdr->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) zhdr->under_reclaim = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) return zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) /* Resets the struct page fields and frees the page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) static void free_zbud_page(struct zbud_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) __free_page(virt_to_page(zhdr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) * Encodes the handle of a particular buddy within a zbud page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) * Pool lock should be held as this function accesses first|last_chunks
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) static unsigned long encode_handle(struct zbud_header *zhdr, enum buddy bud)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) unsigned long handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) * For now, the encoded handle is actually just the pointer to the data
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) * but this might not always be the case. A little information hiding.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) * Add CHUNK_SIZE to the handle if it is the first allocation to jump
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) * over the zbud header in the first chunk.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) handle = (unsigned long)zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) if (bud == FIRST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) /* skip over zbud header */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) handle += ZHDR_SIZE_ALIGNED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) else /* bud == LAST */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) handle += PAGE_SIZE - (zhdr->last_chunks << CHUNK_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) return handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) /* Returns the zbud page where a given handle is stored */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) static struct zbud_header *handle_to_zbud_header(unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) return (struct zbud_header *)(handle & PAGE_MASK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) /* Returns the number of free chunks in a zbud page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) static int num_free_chunks(struct zbud_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) * Rather than branch for different situations, just use the fact that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) * free buddies have a length of zero to simplify everything.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) return NCHUNKS - zhdr->first_chunks - zhdr->last_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) * API Functions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) *****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) * zbud_create_pool() - create a new zbud pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) * @gfp: gfp flags when allocating the zbud pool structure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) * @ops: user-defined operations for the zbud pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) * Return: pointer to the new zbud pool or NULL if the metadata allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) * failed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) struct zbud_pool *zbud_create_pool(gfp_t gfp, const struct zbud_ops *ops)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) struct zbud_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) pool = kzalloc(sizeof(struct zbud_pool), gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) if (!pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) spin_lock_init(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) for_each_unbuddied_list(i, 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) INIT_LIST_HEAD(&pool->unbuddied[i]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) INIT_LIST_HEAD(&pool->buddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) INIT_LIST_HEAD(&pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) pool->pages_nr = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) pool->ops = ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) return pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) * zbud_destroy_pool() - destroys an existing zbud pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) * @pool: the zbud pool to be destroyed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) * The pool should be emptied before this function is called.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) void zbud_destroy_pool(struct zbud_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) kfree(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) * zbud_alloc() - allocates a region of a given size
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) * @pool: zbud pool from which to allocate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) * @size: size in bytes of the desired allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) * @gfp: gfp flags used if the pool needs to grow
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) * @handle: handle of the new allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) * This function will attempt to find a free region in the pool large enough to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) * satisfy the allocation request. A search of the unbuddied lists is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) * performed first. If no suitable free region is found, then a new page is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) * allocated and added to the pool to satisfy the request.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) * gfp should not set __GFP_HIGHMEM as highmem pages cannot be used
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) * as zbud pool pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) * Return: 0 if success and handle is set, otherwise -EINVAL if the size or
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) * gfp arguments are invalid or -ENOMEM if the pool was unable to allocate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) * a new page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) int zbud_alloc(struct zbud_pool *pool, size_t size, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) unsigned long *handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) int chunks, i, freechunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) struct zbud_header *zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) enum buddy bud;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) if (!size || (gfp & __GFP_HIGHMEM))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) if (size > PAGE_SIZE - ZHDR_SIZE_ALIGNED - CHUNK_SIZE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) return -ENOSPC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) chunks = size_to_chunks(size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) /* First, try to find an unbuddied zbud page. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) for_each_unbuddied_list(i, chunks) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) if (!list_empty(&pool->unbuddied[i])) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) zhdr = list_first_entry(&pool->unbuddied[i],
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) struct zbud_header, buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) list_del(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) if (zhdr->first_chunks == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) bud = FIRST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) bud = LAST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) goto found;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) /* Couldn't find unbuddied zbud page, create new one */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) page = alloc_page(gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) pool->pages_nr++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) zhdr = init_zbud_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) bud = FIRST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) found:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) if (bud == FIRST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) zhdr->first_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) zhdr->last_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) if (zhdr->first_chunks == 0 || zhdr->last_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) /* Add to unbuddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) freechunks = num_free_chunks(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) list_add(&zhdr->buddy, &pool->unbuddied[freechunks]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) /* Add to buddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) list_add(&zhdr->buddy, &pool->buddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) /* Add/move zbud page to beginning of LRU */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) if (!list_empty(&zhdr->lru))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) list_del(&zhdr->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) list_add(&zhdr->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) *handle = encode_handle(zhdr, bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) * zbud_free() - frees the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) * @pool: pool in which the allocation resided
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) * @handle: handle associated with the allocation returned by zbud_alloc()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) * In the case that the zbud page in which the allocation resides is under
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) * reclaim, as indicated by the PG_reclaim flag being set, this function
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) * only sets the first|last_chunks to 0. The page is actually freed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) * once both buddies are evicted (see zbud_reclaim_page() below).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) void zbud_free(struct zbud_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) struct zbud_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) int freechunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) zhdr = handle_to_zbud_header(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) /* If first buddy, handle will be page aligned */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) if ((handle - ZHDR_SIZE_ALIGNED) & ~PAGE_MASK)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) zhdr->last_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) zhdr->first_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) if (zhdr->under_reclaim) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) /* zbud page is under reclaim, reclaim will free */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) /* Remove from existing buddy list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) list_del(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) /* zbud page is empty, free */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) list_del(&zhdr->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) free_zbud_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) pool->pages_nr--;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) /* Add to unbuddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) freechunks = num_free_chunks(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) list_add(&zhdr->buddy, &pool->unbuddied[freechunks]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) * zbud_reclaim_page() - evicts allocations from a pool page and frees it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) * @pool: pool from which a page will attempt to be evicted
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) * @retries: number of pages on the LRU list for which eviction will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) * be attempted before failing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) * zbud reclaim is different from normal system reclaim in that the reclaim is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) * done from the bottom, up. This is because only the bottom layer, zbud, has
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) * information on how the allocations are organized within each zbud page. This
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) * has the potential to create interesting locking situations between zbud and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) * the user, however.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) * To avoid these, this is how zbud_reclaim_page() should be called:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) * The user detects a page should be reclaimed and calls zbud_reclaim_page().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) * zbud_reclaim_page() will remove a zbud page from the pool LRU list and call
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) * the user-defined eviction handler with the pool and handle as arguments.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) * If the handle can not be evicted, the eviction handler should return
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) * non-zero. zbud_reclaim_page() will add the zbud page back to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) * appropriate list and try the next zbud page on the LRU up to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) * a user defined number of retries.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) * If the handle is successfully evicted, the eviction handler should
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) * return 0 _and_ should have called zbud_free() on the handle. zbud_free()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) * contains logic to delay freeing the page if the page is under reclaim,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) * as indicated by the setting of the PG_reclaim flag on the underlying page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) * If all buddies in the zbud page are successfully evicted, then the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) * zbud page can be freed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497) * Returns: 0 if page is successfully freed, otherwise -EINVAL if there are
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) * no pages to evict or an eviction handler is not registered, -EAGAIN if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) * the retry limit was hit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) int zbud_reclaim_page(struct zbud_pool *pool, unsigned int retries)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) int i, ret, freechunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) struct zbud_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) unsigned long first_handle = 0, last_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508) if (!pool->ops || !pool->ops->evict || list_empty(&pool->lru) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) retries == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) for (i = 0; i < retries; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) zhdr = list_last_entry(&pool->lru, struct zbud_header, lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) list_del(&zhdr->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) list_del(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) /* Protect zbud page against free */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) zhdr->under_reclaim = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) * We need encode the handles before unlocking, since we can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) * race with free that will set (first|last)_chunks to 0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) first_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) last_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) if (zhdr->first_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) first_handle = encode_handle(zhdr, FIRST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) if (zhdr->last_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) last_handle = encode_handle(zhdr, LAST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) /* Issue the eviction callback(s) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532) if (first_handle) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) ret = pool->ops->evict(pool, first_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) goto next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) if (last_handle) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) ret = pool->ops->evict(pool, last_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) goto next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) next:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) zhdr->under_reclaim = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545) if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) * Both buddies are now free, free the zbud page and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) * return success.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) free_zbud_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) pool->pages_nr--;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) } else if (zhdr->first_chunks == 0 ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) zhdr->last_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) /* add to unbuddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) freechunks = num_free_chunks(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) list_add(&zhdr->buddy, &pool->unbuddied[freechunks]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) /* add to buddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) list_add(&zhdr->buddy, &pool->buddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564) /* add to beginning of LRU */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565) list_add(&zhdr->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) return -EAGAIN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) * zbud_map() - maps the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) * @pool: pool in which the allocation resides
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574) * @handle: handle associated with the allocation to be mapped
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) * While trivial for zbud, the mapping functions for others allocators
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577) * implementing this allocation API could have more complex information encoded
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) * in the handle and could create temporary mappings to make the data
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) * accessible to the user.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) * Returns: a pointer to the mapped allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583) void *zbud_map(struct zbud_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) return (void *)(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) * zbud_unmap() - maps the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) * @pool: pool in which the allocation resides
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) * @handle: handle associated with the allocation to be unmapped
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593) void zbud_unmap(struct zbud_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598) * zbud_get_pool_size() - gets the zbud pool size in pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) * @pool: pool whose size is being queried
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) * Returns: size in pages of the given pool. The pool lock need not be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) * taken to access pages_nr.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604) u64 zbud_get_pool_size(struct zbud_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) return pool->pages_nr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) static int __init init_zbud(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) /* Make sure the zbud header will fit in one chunk */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) BUILD_BUG_ON(sizeof(struct zbud_header) > ZHDR_SIZE_ALIGNED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) pr_info("loaded\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) #ifdef CONFIG_ZPOOL
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) zpool_register_driver(&zbud_zpool_driver);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622) static void __exit exit_zbud(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) #ifdef CONFIG_ZPOOL
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) zpool_unregister_driver(&zbud_zpool_driver);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) pr_info("unloaded\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) module_init(init_zbud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) module_exit(exit_zbud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) MODULE_LICENSE("GPL");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) MODULE_AUTHOR("Seth Jennings <sjennings@variantweb.net>");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636) MODULE_DESCRIPTION("Buddy Allocator for Compressed Pages");
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags