^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) * z3fold.c
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Author: Vitaly Wool <vitaly.wool@konsulko.com>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * Copyright (C) 2016, Sony Mobile Communications Inc.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) * This implementation is based on zbud written by Seth Jennings.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * z3fold is an special purpose allocator for storing compressed pages. It
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) * can store up to three compressed pages per page which improves the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) * compression ratio of zbud while retaining its main concepts (e. g. always
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) * storing an integral number of objects per page) and simplicity.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * It still has simple and deterministic reclaim properties that make it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * preferable to a higher density approach (with no requirement on integral
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) * number of object per page) when reclaim is used.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) * As in zbud, pages are divided into "chunks". The size of the chunks is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) * fixed at compile time and is determined by NCHUNKS_ORDER below.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) * z3fold doesn't export any API and is meant to be used via zpool API.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) #include <linux/atomic.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) #include <linux/sched.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) #include <linux/cpumask.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) #include <linux/list.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) #include <linux/module.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) #include <linux/page-flags.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) #include <linux/migrate.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) #include <linux/node.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) #include <linux/compaction.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) #include <linux/percpu.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) #include <linux/mount.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) #include <linux/pseudo_fs.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) #include <linux/fs.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) #include <linux/preempt.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) #include <linux/workqueue.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) #include <linux/spinlock.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) #include <linux/zpool.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) #include <linux/magic.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) #include <linux/kmemleak.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) * NCHUNKS_ORDER determines the internal allocation granularity, effectively
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) * adjusting internal fragmentation. It also determines the number of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) * freelists maintained in each pool. NCHUNKS_ORDER of 6 means that the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) * allocation granularity will be in chunks of size PAGE_SIZE/64. Some chunks
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) * in the beginning of an allocated page are occupied by z3fold header, so
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) * NCHUNKS will be calculated to 63 (or 62 in case CONFIG_DEBUG_SPINLOCK=y),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) * which shows the max number of free chunks in z3fold page, also there will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) * be 63, or 62, respectively, freelists per pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) #define NCHUNKS_ORDER 6
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) #define CHUNK_SHIFT (PAGE_SHIFT - NCHUNKS_ORDER)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) #define CHUNK_SIZE (1 << CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) #define ZHDR_SIZE_ALIGNED round_up(sizeof(struct z3fold_header), CHUNK_SIZE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) #define ZHDR_CHUNKS (ZHDR_SIZE_ALIGNED >> CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) #define TOTAL_CHUNKS (PAGE_SIZE >> CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) #define NCHUNKS ((PAGE_SIZE - ZHDR_SIZE_ALIGNED) >> CHUNK_SHIFT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) #define BUDDY_MASK (0x3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) #define BUDDY_SHIFT 2
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) #define SLOTS_ALIGN (0x40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) * Structures
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) *****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) struct z3fold_pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) struct z3fold_ops {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) int (*evict)(struct z3fold_pool *pool, unsigned long handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) enum buddy {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) HEADLESS = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) FIRST,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) MIDDLE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) LAST,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) BUDDIES_MAX = LAST
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) struct z3fold_buddy_slots {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) * we are using BUDDY_MASK in handle_to_buddy etc. so there should
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) * be enough slots to hold all possible variants
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) unsigned long slot[BUDDY_MASK + 1];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) unsigned long pool; /* back link */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) rwlock_t lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) #define HANDLE_FLAG_MASK (0x03)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) * struct z3fold_header - z3fold page metadata occupying first chunks of each
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) * z3fold page, except for HEADLESS pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) * @buddy: links the z3fold page into the relevant list in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) * pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) * @page_lock: per-page lock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) * @refcount: reference count for the z3fold page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * @work: work_struct for page layout optimization
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) * @slots: pointer to the structure holding buddy slots
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) * @pool: pointer to the containing pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) * @cpu: CPU which this page "belongs" to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) * @first_chunks: the size of the first buddy in chunks, 0 if free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) * @middle_chunks: the size of the middle buddy in chunks, 0 if free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) * @last_chunks: the size of the last buddy in chunks, 0 if free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) * @first_num: the starting number (for the first handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) * @mapped_count: the number of objects currently mapped
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) struct z3fold_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) spinlock_t page_lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) struct kref refcount;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) struct work_struct work;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) struct z3fold_buddy_slots *slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) struct z3fold_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) short cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) unsigned short first_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) unsigned short middle_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) unsigned short last_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) unsigned short start_middle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) unsigned short first_num:2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) unsigned short mapped_count:2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) unsigned short foreign_handles:2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) * struct z3fold_pool - stores metadata for each z3fold pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) * @name: pool name
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) * @lock: protects pool unbuddied/lru lists
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) * @stale_lock: protects pool stale page list
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) * @unbuddied: per-cpu array of lists tracking z3fold pages that contain 2-
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) * buddies; the list each z3fold page is added to depends on
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) * the size of its free region.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) * @lru: list tracking the z3fold pages in LRU order by most recently
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) * added buddy.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) * @stale: list of pages marked for freeing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) * @pages_nr: number of z3fold pages in the pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) * @c_handle: cache for z3fold_buddy_slots allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) * @ops: pointer to a structure of user defined operations specified at
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) * pool creation time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) * @compact_wq: workqueue for page layout background optimization
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) * @release_wq: workqueue for safe page release
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) * @work: work_struct for safe page release
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) * @inode: inode for z3fold pseudo filesystem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) * This structure is allocated at pool creation time and maintains metadata
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) * pertaining to a particular z3fold pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) struct z3fold_pool {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) const char *name;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) spinlock_t lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) spinlock_t stale_lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) struct list_head *unbuddied;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) struct list_head lru;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) struct list_head stale;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) atomic64_t pages_nr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) struct kmem_cache *c_handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) const struct z3fold_ops *ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) struct zpool *zpool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) const struct zpool_ops *zpool_ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) struct workqueue_struct *compact_wq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) struct workqueue_struct *release_wq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) struct work_struct work;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) struct inode *inode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) * Internal z3fold page flags
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) enum z3fold_page_flags {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) PAGE_HEADLESS = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) MIDDLE_CHUNK_MAPPED,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) NEEDS_COMPACTING,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) PAGE_STALE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) PAGE_CLAIMED, /* by either reclaim or free */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) * handle flags, go under HANDLE_FLAG_MASK
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) enum z3fold_handle_flags {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) HANDLES_NOFREE = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) * Forward declarations
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) static struct z3fold_header *__z3fold_alloc(struct z3fold_pool *, size_t, bool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) static void compact_page_work(struct work_struct *w);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) * Helpers
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) *****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) /* Converts an allocation size in bytes to size in z3fold chunks */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) static int size_to_chunks(size_t size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) return (size + CHUNK_SIZE - 1) >> CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) #define for_each_unbuddied_list(_iter, _begin) \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) for ((_iter) = (_begin); (_iter) < NCHUNKS; (_iter)++)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) static inline struct z3fold_buddy_slots *alloc_slots(struct z3fold_pool *pool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) gfp_t gfp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) struct z3fold_buddy_slots *slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) slots = kmem_cache_zalloc(pool->c_handle,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) (gfp & ~(__GFP_HIGHMEM | __GFP_MOVABLE)));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) if (slots) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) /* It will be freed separately in free_handle(). */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) kmemleak_not_leak(slots);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) slots->pool = (unsigned long)pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) rwlock_init(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) return slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) static inline struct z3fold_pool *slots_to_pool(struct z3fold_buddy_slots *s)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) return (struct z3fold_pool *)(s->pool & ~HANDLE_FLAG_MASK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) static inline struct z3fold_buddy_slots *handle_to_slots(unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) return (struct z3fold_buddy_slots *)(handle & ~(SLOTS_ALIGN - 1));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) /* Lock a z3fold page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) static inline void z3fold_page_lock(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) spin_lock(&zhdr->page_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) /* Try to lock a z3fold page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) static inline int z3fold_page_trylock(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) return spin_trylock(&zhdr->page_lock);
^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) /* Unlock a z3fold page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) static inline void z3fold_page_unlock(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) spin_unlock(&zhdr->page_lock);
^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) static inline struct z3fold_header *__get_z3fold_header(unsigned long handle,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) bool lock)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) struct z3fold_buddy_slots *slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) int locked = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) if (!(handle & (1 << PAGE_HEADLESS))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) slots = handle_to_slots(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) unsigned long addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) read_lock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) addr = *(unsigned long *)handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) zhdr = (struct z3fold_header *)(addr & PAGE_MASK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) if (lock)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) locked = z3fold_page_trylock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) read_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) if (locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) cpu_relax();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) } while (lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) zhdr = (struct z3fold_header *)(handle & PAGE_MASK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) return zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) /* Returns the z3fold page where a given handle is stored */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) static inline struct z3fold_header *handle_to_z3fold_header(unsigned long h)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) return __get_z3fold_header(h, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) /* return locked z3fold page if it's not headless */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) static inline struct z3fold_header *get_z3fold_header(unsigned long h)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) return __get_z3fold_header(h, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) static inline void put_z3fold_header(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) struct page *page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) if (!test_bit(PAGE_HEADLESS, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) static inline void free_handle(unsigned long handle, struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) struct z3fold_buddy_slots *slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) bool is_free;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) if (handle & (1 << PAGE_HEADLESS))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) if (WARN_ON(*(unsigned long *)handle == 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) slots = handle_to_slots(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) write_lock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) *(unsigned long *)handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) if (test_bit(HANDLES_NOFREE, &slots->pool)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) write_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) return; /* simple case, nothing else to do */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) if (zhdr->slots != slots)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) zhdr->foreign_handles--;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) is_free = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) for (i = 0; i <= BUDDY_MASK; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) if (slots->slot[i]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) is_free = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) write_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) if (is_free) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) struct z3fold_pool *pool = slots_to_pool(slots);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) if (zhdr->slots == slots)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) zhdr->slots = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) kmem_cache_free(pool->c_handle, slots);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) static int z3fold_init_fs_context(struct fs_context *fc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) return init_pseudo(fc, Z3FOLD_MAGIC) ? 0 : -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) static struct file_system_type z3fold_fs = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) .name = "z3fold",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) .init_fs_context = z3fold_init_fs_context,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) .kill_sb = kill_anon_super,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) static struct vfsmount *z3fold_mnt;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) static int z3fold_mount(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) z3fold_mnt = kern_mount(&z3fold_fs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) if (IS_ERR(z3fold_mnt))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) ret = PTR_ERR(z3fold_mnt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) static void z3fold_unmount(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) kern_unmount(z3fold_mnt);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) static const struct address_space_operations z3fold_aops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) static int z3fold_register_migration(struct z3fold_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) pool->inode = alloc_anon_inode(z3fold_mnt->mnt_sb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) if (IS_ERR(pool->inode)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) pool->inode = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) pool->inode->i_mapping->private_data = pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) pool->inode->i_mapping->a_ops = &z3fold_aops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) static void z3fold_unregister_migration(struct z3fold_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) if (pool->inode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) iput(pool->inode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) /* Initializes the z3fold header of a newly allocated z3fold page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) static struct z3fold_header *init_z3fold_page(struct page *page, bool headless,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) struct z3fold_pool *pool, gfp_t gfp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) struct z3fold_header *zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) struct z3fold_buddy_slots *slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) INIT_LIST_HEAD(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) clear_bit(PAGE_HEADLESS, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) clear_bit(MIDDLE_CHUNK_MAPPED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) clear_bit(NEEDS_COMPACTING, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407) clear_bit(PAGE_STALE, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) if (headless)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) return zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) slots = alloc_slots(pool, gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) if (!slots)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) spin_lock_init(&zhdr->page_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) kref_init(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) zhdr->first_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) zhdr->middle_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) zhdr->last_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) zhdr->first_num = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) zhdr->start_middle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) zhdr->cpu = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) zhdr->foreign_handles = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) zhdr->mapped_count = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) zhdr->slots = slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) zhdr->pool = pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) INIT_LIST_HEAD(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) INIT_WORK(&zhdr->work, compact_page_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) return zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) /* Resets the struct page fields and frees the page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) static void free_z3fold_page(struct page *page, bool headless)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) if (!headless) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) lock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) __ClearPageMovable(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) ClearPagePrivate(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) __free_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) /* Helper function to build the index */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) static inline int __idx(struct z3fold_header *zhdr, enum buddy bud)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) return (bud + zhdr->first_num) & BUDDY_MASK;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) * Encodes the handle of a particular buddy within a z3fold page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) * Pool lock should be held as this function accesses first_num
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) static unsigned long __encode_handle(struct z3fold_header *zhdr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) struct z3fold_buddy_slots *slots,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) enum buddy bud)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) unsigned long h = (unsigned long)zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) int idx = 0;
^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) * For a headless page, its handle is its pointer with the extra
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) * PAGE_HEADLESS bit set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) if (bud == HEADLESS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) return h | (1 << PAGE_HEADLESS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) /* otherwise, return pointer to encoded handle */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) idx = __idx(zhdr, bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) h += idx;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) if (bud == LAST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) h |= (zhdr->last_chunks << BUDDY_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) write_lock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) slots->slot[idx] = h;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) write_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) return (unsigned long)&slots->slot[idx];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) static unsigned long encode_handle(struct z3fold_header *zhdr, enum buddy bud)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) return __encode_handle(zhdr, zhdr->slots, bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) /* only for LAST bud, returns zero otherwise */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) static unsigned short handle_to_chunks(unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) struct z3fold_buddy_slots *slots = handle_to_slots(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) unsigned long addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) read_lock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) addr = *(unsigned long *)handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) read_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) return (addr & ~PAGE_MASK) >> BUDDY_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) * (handle & BUDDY_MASK) < zhdr->first_num is possible in encode_handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) * but that doesn't matter. because the masking will result in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) * correct buddy number.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) static enum buddy handle_to_buddy(unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) struct z3fold_buddy_slots *slots = handle_to_slots(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) unsigned long addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) read_lock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) WARN_ON(handle & (1 << PAGE_HEADLESS));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) addr = *(unsigned long *)handle;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) read_unlock(&slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) zhdr = (struct z3fold_header *)(addr & PAGE_MASK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) return (addr - zhdr->first_num) & BUDDY_MASK;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) static inline struct z3fold_pool *zhdr_to_pool(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) return zhdr->pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) static void __release_z3fold_page(struct z3fold_header *zhdr, bool locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) struct page *page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) struct z3fold_pool *pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) WARN_ON(!list_empty(&zhdr->buddy));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) set_bit(PAGE_STALE, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) clear_bit(NEEDS_COMPACTING, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) if (!list_empty(&page->lru))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532) list_del_init(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) if (locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) spin_lock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) list_add(&zhdr->buddy, &pool->stale);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) queue_work(pool->release_wq, &pool->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) spin_unlock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) static void __attribute__((__unused__))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545) release_z3fold_page(struct kref *ref)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) struct z3fold_header *zhdr = container_of(ref, struct z3fold_header,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549) __release_z3fold_page(zhdr, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) static void release_z3fold_page_locked(struct kref *ref)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) struct z3fold_header *zhdr = container_of(ref, struct z3fold_header,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) WARN_ON(z3fold_page_trylock(zhdr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) __release_z3fold_page(zhdr, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) static void release_z3fold_page_locked_list(struct kref *ref)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562) struct z3fold_header *zhdr = container_of(ref, struct z3fold_header,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564) struct z3fold_pool *pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570) WARN_ON(z3fold_page_trylock(zhdr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571) __release_z3fold_page(zhdr, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574) static void free_pages_work(struct work_struct *w)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) struct z3fold_pool *pool = container_of(w, struct z3fold_pool, work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) spin_lock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) while (!list_empty(&pool->stale)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580) struct z3fold_header *zhdr = list_first_entry(&pool->stale,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) struct z3fold_header, buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) struct page *page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584) list_del(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) if (WARN_ON(!test_bit(PAGE_STALE, &page->private)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587) spin_unlock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588) cancel_work_sync(&zhdr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) free_z3fold_page(page, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) spin_lock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593) spin_unlock(&pool->stale_lock);
^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) * Returns the number of free chunks in a z3fold page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598) * NB: can't be used with HEADLESS pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) static int num_free_chunks(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) int nfree;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604) * If there is a middle object, pick up the bigger free space
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) * either before or after it. Otherwise just subtract the number
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) * of chunks occupied by the first and the last objects.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) if (zhdr->middle_chunks != 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) int nfree_before = zhdr->first_chunks ?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) 0 : zhdr->start_middle - ZHDR_CHUNKS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) int nfree_after = zhdr->last_chunks ?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) 0 : TOTAL_CHUNKS -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) (zhdr->start_middle + zhdr->middle_chunks);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) nfree = max(nfree_before, nfree_after);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) } else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) nfree = NCHUNKS - zhdr->first_chunks - zhdr->last_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) return nfree;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) /* Add to the appropriate unbuddied list */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621) static inline void add_to_unbuddied(struct z3fold_pool *pool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622) struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) if (zhdr->first_chunks == 0 || zhdr->last_chunks == 0 ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) zhdr->middle_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) struct list_head *unbuddied = get_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) int freechunks = num_free_chunks(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630) list_add(&zhdr->buddy, &unbuddied[freechunks]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) zhdr->cpu = smp_processor_id();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633) put_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) static inline enum buddy get_free_buddy(struct z3fold_header *zhdr, int chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639) enum buddy bud = HEADLESS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641) if (zhdr->middle_chunks) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) if (!zhdr->first_chunks &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643) chunks <= zhdr->start_middle - ZHDR_CHUNKS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) bud = FIRST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645) else if (!zhdr->last_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646) bud = LAST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648) if (!zhdr->first_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) bud = FIRST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650) else if (!zhdr->last_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651) bud = LAST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653) bud = MIDDLE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) return bud;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) static inline void *mchunk_memmove(struct z3fold_header *zhdr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660) unsigned short dst_chunk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) void *beg = zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) return memmove(beg + (dst_chunk << CHUNK_SHIFT),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) beg + (zhdr->start_middle << CHUNK_SHIFT),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665) zhdr->middle_chunks << CHUNK_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668) static inline bool buddy_single(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) return !((zhdr->first_chunks && zhdr->middle_chunks) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671) (zhdr->first_chunks && zhdr->last_chunks) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) (zhdr->middle_chunks && zhdr->last_chunks));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675) static struct z3fold_header *compact_single_buddy(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) struct z3fold_pool *pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678) void *p = zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) unsigned long old_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) size_t sz = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681) struct z3fold_header *new_zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682) int first_idx = __idx(zhdr, FIRST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) int middle_idx = __idx(zhdr, MIDDLE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) int last_idx = __idx(zhdr, LAST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685) unsigned short *moved_chunks = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688) * No need to protect slots here -- all the slots are "local" and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) * the page lock is already taken
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691) if (zhdr->first_chunks && zhdr->slots->slot[first_idx]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) p += ZHDR_SIZE_ALIGNED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) sz = zhdr->first_chunks << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694) old_handle = (unsigned long)&zhdr->slots->slot[first_idx];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) moved_chunks = &zhdr->first_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696) } else if (zhdr->middle_chunks && zhdr->slots->slot[middle_idx]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) p += zhdr->start_middle << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698) sz = zhdr->middle_chunks << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699) old_handle = (unsigned long)&zhdr->slots->slot[middle_idx];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) moved_chunks = &zhdr->middle_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) } else if (zhdr->last_chunks && zhdr->slots->slot[last_idx]) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702) p += PAGE_SIZE - (zhdr->last_chunks << CHUNK_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) sz = zhdr->last_chunks << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704) old_handle = (unsigned long)&zhdr->slots->slot[last_idx];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705) moved_chunks = &zhdr->last_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708) if (sz > 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) enum buddy new_bud = HEADLESS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710) short chunks = size_to_chunks(sz);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711) void *q;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713) new_zhdr = __z3fold_alloc(pool, sz, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) if (!new_zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) if (WARN_ON(new_zhdr == zhdr))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718) goto out_fail;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720) new_bud = get_free_buddy(new_zhdr, chunks);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721) q = new_zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) switch (new_bud) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723) case FIRST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) new_zhdr->first_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725) q += ZHDR_SIZE_ALIGNED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727) case MIDDLE:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) new_zhdr->middle_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729) new_zhdr->start_middle =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730) new_zhdr->first_chunks + ZHDR_CHUNKS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731) q += new_zhdr->start_middle << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733) case LAST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) new_zhdr->last_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) q += PAGE_SIZE - (new_zhdr->last_chunks << CHUNK_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) goto out_fail;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) new_zhdr->foreign_handles++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741) memcpy(q, p, sz);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) write_lock(&zhdr->slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743) *(unsigned long *)old_handle = (unsigned long)new_zhdr +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744) __idx(new_zhdr, new_bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) if (new_bud == LAST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) *(unsigned long *)old_handle |=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747) (new_zhdr->last_chunks << BUDDY_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) write_unlock(&zhdr->slots->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749) add_to_unbuddied(pool, new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750) z3fold_page_unlock(new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752) *moved_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) return new_zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) out_fail:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758) if (new_zhdr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759) if (kref_put(&new_zhdr->refcount, release_z3fold_page_locked))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762) add_to_unbuddied(pool, new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) z3fold_page_unlock(new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770) #define BIG_CHUNK_GAP 3
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) /* Has to be called with lock held */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) static int z3fold_compact_page(struct z3fold_header *zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774) struct page *page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) if (test_bit(MIDDLE_CHUNK_MAPPED, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777) return 0; /* can't move middle chunk, it's used */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779) if (unlikely(PageIsolated(page)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782) if (zhdr->middle_chunks == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783) return 0; /* nothing to compact */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785) if (zhdr->first_chunks == 0 && zhdr->last_chunks == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786) /* move to the beginning */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787) mchunk_memmove(zhdr, ZHDR_CHUNKS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) zhdr->first_chunks = zhdr->middle_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789) zhdr->middle_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790) zhdr->start_middle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) zhdr->first_num++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796) * moving data is expensive, so let's only do that if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) * there's substantial gain (at least BIG_CHUNK_GAP chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) if (zhdr->first_chunks != 0 && zhdr->last_chunks == 0 &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800) zhdr->start_middle - (zhdr->first_chunks + ZHDR_CHUNKS) >=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) BIG_CHUNK_GAP) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) mchunk_memmove(zhdr, zhdr->first_chunks + ZHDR_CHUNKS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) zhdr->start_middle = zhdr->first_chunks + ZHDR_CHUNKS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805) } else if (zhdr->last_chunks != 0 && zhdr->first_chunks == 0 &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806) TOTAL_CHUNKS - (zhdr->last_chunks + zhdr->start_middle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807) + zhdr->middle_chunks) >=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) BIG_CHUNK_GAP) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) unsigned short new_start = TOTAL_CHUNKS - zhdr->last_chunks -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) zhdr->middle_chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811) mchunk_memmove(zhdr, new_start);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812) zhdr->start_middle = new_start;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) static void do_compact_page(struct z3fold_header *zhdr, bool locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) struct z3fold_pool *pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) if (locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826) WARN_ON(z3fold_page_trylock(zhdr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) z3fold_page_lock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829) if (WARN_ON(!test_and_clear_bit(NEEDS_COMPACTING, &page->private))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837) if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) if (test_bit(PAGE_STALE, &page->private) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843) test_and_set_bit(PAGE_CLAIMED, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) if (!zhdr->foreign_handles && buddy_single(zhdr) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849) zhdr->mapped_count == 0 && compact_single_buddy(zhdr)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) if (kref_put(&zhdr->refcount, release_z3fold_page_locked))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859) z3fold_compact_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) add_to_unbuddied(pool, zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865) static void compact_page_work(struct work_struct *w)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) struct z3fold_header *zhdr = container_of(w, struct z3fold_header,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) do_compact_page(zhdr, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873) /* returns _locked_ z3fold page header or NULL */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874) static inline struct z3fold_header *__z3fold_alloc(struct z3fold_pool *pool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) size_t size, bool can_sleep)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877) struct z3fold_header *zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879) struct list_head *unbuddied;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880) int chunks = size_to_chunks(size), i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) lookup:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) /* First, try to find an unbuddied z3fold page. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) unbuddied = get_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885) for_each_unbuddied_list(i, chunks) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886) struct list_head *l = &unbuddied[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) zhdr = list_first_entry_or_null(READ_ONCE(l),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889) struct z3fold_header, buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) if (!zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894) /* Re-check under lock. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896) l = &unbuddied[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) if (unlikely(zhdr != list_first_entry(READ_ONCE(l),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898) struct z3fold_header, buddy)) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) !z3fold_page_trylock(zhdr)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902) put_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903) if (can_sleep)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905) goto lookup;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908) zhdr->cpu = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912) if (test_bit(NEEDS_COMPACTING, &page->private) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) test_bit(PAGE_CLAIMED, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) put_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) if (can_sleep)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919) goto lookup;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) * this page could not be removed from its unbuddied
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) * list while pool lock was held, and then we've taken
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) * page lock so kref_put could not be called before
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926) * we got here, so it's safe to just call kref_get()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928) kref_get(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931) put_cpu_ptr(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) if (!zhdr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934) int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936) /* look for _exact_ match on other cpus' lists */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937) for_each_online_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938) struct list_head *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940) unbuddied = per_cpu_ptr(pool->unbuddied, cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) l = &unbuddied[chunks];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944) zhdr = list_first_entry_or_null(READ_ONCE(l),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945) struct z3fold_header, buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) if (!zhdr || !z3fold_page_trylock(zhdr)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953) zhdr->cpu = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) if (test_bit(NEEDS_COMPACTING, &page->private) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958) test_bit(PAGE_CLAIMED, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961) if (can_sleep)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965) kref_get(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970) if (zhdr && !zhdr->slots)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971) zhdr->slots = alloc_slots(pool,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) can_sleep ? GFP_NOIO : GFP_ATOMIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973) return zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) * API Functions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) * z3fold_create_pool() - create a new z3fold pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982) * @name: pool name
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983) * @gfp: gfp flags when allocating the z3fold pool structure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984) * @ops: user-defined operations for the z3fold pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) * Return: pointer to the new z3fold pool or NULL if the metadata allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) * failed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) static struct z3fold_pool *z3fold_create_pool(const char *name, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990) const struct z3fold_ops *ops)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992) struct z3fold_pool *pool = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993) int i, cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995) pool = kzalloc(sizeof(struct z3fold_pool), gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) if (!pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) pool->c_handle = kmem_cache_create("z3fold_handle",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999) sizeof(struct z3fold_buddy_slots),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000) SLOTS_ALIGN, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001) if (!pool->c_handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002) goto out_c;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003) spin_lock_init(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004) spin_lock_init(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) pool->unbuddied = __alloc_percpu(sizeof(struct list_head)*NCHUNKS, 2);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006) if (!pool->unbuddied)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) goto out_pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009) struct list_head *unbuddied =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) per_cpu_ptr(pool->unbuddied, cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) for_each_unbuddied_list(i, 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) INIT_LIST_HEAD(&unbuddied[i]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014) INIT_LIST_HEAD(&pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) INIT_LIST_HEAD(&pool->stale);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016) atomic64_set(&pool->pages_nr, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017) pool->name = name;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) pool->compact_wq = create_singlethread_workqueue(pool->name);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) if (!pool->compact_wq)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020) goto out_unbuddied;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) pool->release_wq = create_singlethread_workqueue(pool->name);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022) if (!pool->release_wq)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023) goto out_wq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) if (z3fold_register_migration(pool))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025) goto out_rwq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026) INIT_WORK(&pool->work, free_pages_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027) pool->ops = ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028) return pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030) out_rwq:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031) destroy_workqueue(pool->release_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032) out_wq:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) destroy_workqueue(pool->compact_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034) out_unbuddied:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035) free_percpu(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036) out_pool:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037) kmem_cache_destroy(pool->c_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038) out_c:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) kfree(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045) * z3fold_destroy_pool() - destroys an existing z3fold pool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) * @pool: the z3fold pool to be destroyed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048) * The pool should be emptied before this function is called.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050) static void z3fold_destroy_pool(struct z3fold_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052) kmem_cache_destroy(pool->c_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055) * We need to destroy pool->compact_wq before pool->release_wq,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056) * as any pending work on pool->compact_wq will call
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057) * queue_work(pool->release_wq, &pool->work).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) * There are still outstanding pages until both workqueues are drained,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060) * so we cannot unregister migration until then.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063) destroy_workqueue(pool->compact_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064) destroy_workqueue(pool->release_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065) z3fold_unregister_migration(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066) free_percpu(pool->unbuddied);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067) kfree(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071) * z3fold_alloc() - allocates a region of a given size
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) * @pool: z3fold pool from which to allocate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073) * @size: size in bytes of the desired allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074) * @gfp: gfp flags used if the pool needs to grow
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) * @handle: handle of the new allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) * This function will attempt to find a free region in the pool large enough to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078) * satisfy the allocation request. A search of the unbuddied lists is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079) * performed first. If no suitable free region is found, then a new page is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080) * allocated and added to the pool to satisfy the request.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) * gfp should not set __GFP_HIGHMEM as highmem pages cannot be used
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083) * as z3fold pool pages.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) * Return: 0 if success and handle is set, otherwise -EINVAL if the size or
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086) * gfp arguments are invalid or -ENOMEM if the pool was unable to allocate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087) * a new page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) static int z3fold_alloc(struct z3fold_pool *pool, size_t size, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090) unsigned long *handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092) int chunks = size_to_chunks(size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093) struct z3fold_header *zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) struct page *page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095) enum buddy bud;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) bool can_sleep = gfpflags_allow_blocking(gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098) if (!size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) if (size > PAGE_SIZE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102) return -ENOSPC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) if (size > PAGE_SIZE - ZHDR_SIZE_ALIGNED - CHUNK_SIZE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105) bud = HEADLESS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) retry:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108) zhdr = __z3fold_alloc(pool, size, can_sleep);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109) if (zhdr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) bud = get_free_buddy(zhdr, chunks);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111) if (bud == HEADLESS) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) if (kref_put(&zhdr->refcount,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) release_z3fold_page_locked))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) pr_err("No free chunks in unbuddied\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118) WARN_ON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) goto retry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) goto found;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124) bud = FIRST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128) if (can_sleep) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) spin_lock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130) zhdr = list_first_entry_or_null(&pool->stale,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131) struct z3fold_header, buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133) * Before allocating a page, let's see if we can take one from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134) * the stale pages list. cancel_work_sync() can sleep so we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135) * limit this case to the contexts where we can sleep
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137) if (zhdr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138) list_del(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) spin_unlock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140) cancel_work_sync(&zhdr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143) spin_unlock(&pool->stale_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147) page = alloc_page(gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) if (!page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152) zhdr = init_z3fold_page(page, bud == HEADLESS, pool, gfp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153) if (!zhdr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) __free_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) atomic64_inc(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) if (bud == HEADLESS) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160) set_bit(PAGE_HEADLESS, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161) goto headless;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) if (can_sleep) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164) lock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165) __SetPageMovable(page, pool->inode->i_mapping);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) if (trylock_page(page)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169) __SetPageMovable(page, pool->inode->i_mapping);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) unlock_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) z3fold_page_lock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) found:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176) if (bud == FIRST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177) zhdr->first_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178) else if (bud == LAST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) zhdr->last_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) zhdr->middle_chunks = chunks;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182) zhdr->start_middle = zhdr->first_chunks + ZHDR_CHUNKS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) add_to_unbuddied(pool, zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) headless:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188) /* Add/move z3fold page to beginning of LRU */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) if (!list_empty(&page->lru))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) list_del(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192) list_add(&page->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) *handle = encode_handle(zhdr, bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1195) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1196) if (bud != HEADLESS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1197) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1198)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1199) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1200) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1201)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1202) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1203) * z3fold_free() - frees the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1204) * @pool: pool in which the allocation resided
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1205) * @handle: handle associated with the allocation returned by z3fold_alloc()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1206) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1207) * In the case that the z3fold page in which the allocation resides is under
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1208) * reclaim, as indicated by the PG_reclaim flag being set, this function
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1209) * only sets the first|last_chunks to 0. The page is actually freed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1210) * once both buddies are evicted (see z3fold_reclaim_page() below).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1211) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1212) static void z3fold_free(struct z3fold_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1213) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1214) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1215) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1216) enum buddy bud;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1217) bool page_claimed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1218)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1219) zhdr = get_z3fold_header(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1220) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1221) page_claimed = test_and_set_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1222)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1223) if (test_bit(PAGE_HEADLESS, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1224) /* if a headless page is under reclaim, just leave.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1225) * NB: we use test_and_set_bit for a reason: if the bit
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1226) * has not been set before, we release this page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1227) * immediately so we don't care about its value any more.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1228) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1229) if (!page_claimed) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1230) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1231) list_del(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1232) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1233) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1234) free_z3fold_page(page, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1235) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1236) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1237) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1238) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1239)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1240) /* Non-headless case */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1241) bud = handle_to_buddy(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1242)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1243) switch (bud) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1244) case FIRST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1245) zhdr->first_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1246) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1247) case MIDDLE:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1248) zhdr->middle_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1249) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1250) case LAST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1251) zhdr->last_chunks = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1252) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1253) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1254) pr_err("%s: unknown bud %d\n", __func__, bud);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1255) WARN_ON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1256) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1257) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1258) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1259)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1260) if (!page_claimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1261) free_handle(handle, zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1262) if (kref_put(&zhdr->refcount, release_z3fold_page_locked_list)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1263) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1264) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1265) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1266) if (page_claimed) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1267) /* the page has not been claimed by us */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1268) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1269) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1270) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1271) if (test_and_set_bit(NEEDS_COMPACTING, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1272) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1273) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1274) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1275) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1276) if (zhdr->cpu < 0 || !cpu_online(zhdr->cpu)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1277) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1278) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1279) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1280) zhdr->cpu = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1281) kref_get(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1282) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1283) do_compact_page(zhdr, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1284) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1285) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1286) kref_get(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1287) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1288) queue_work_on(zhdr->cpu, pool->compact_wq, &zhdr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1289) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1290) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1291)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1292) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1293) * z3fold_reclaim_page() - evicts allocations from a pool page and frees it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1294) * @pool: pool from which a page will attempt to be evicted
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1295) * @retries: number of pages on the LRU list for which eviction will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1296) * be attempted before failing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1297) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1298) * z3fold reclaim is different from normal system reclaim in that it is done
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1299) * from the bottom, up. This is because only the bottom layer, z3fold, has
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1300) * information on how the allocations are organized within each z3fold page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1301) * This has the potential to create interesting locking situations between
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1302) * z3fold and the user, however.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1303) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1304) * To avoid these, this is how z3fold_reclaim_page() should be called:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1305) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1306) * The user detects a page should be reclaimed and calls z3fold_reclaim_page().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1307) * z3fold_reclaim_page() will remove a z3fold page from the pool LRU list and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1308) * call the user-defined eviction handler with the pool and handle as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1309) * arguments.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1310) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1311) * If the handle can not be evicted, the eviction handler should return
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1312) * non-zero. z3fold_reclaim_page() will add the z3fold page back to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1313) * appropriate list and try the next z3fold page on the LRU up to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1314) * a user defined number of retries.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1315) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1316) * If the handle is successfully evicted, the eviction handler should
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1317) * return 0 _and_ should have called z3fold_free() on the handle. z3fold_free()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1318) * contains logic to delay freeing the page if the page is under reclaim,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1319) * as indicated by the setting of the PG_reclaim flag on the underlying page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1320) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1321) * If all buddies in the z3fold page are successfully evicted, then the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1322) * z3fold page can be freed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1323) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1324) * Returns: 0 if page is successfully freed, otherwise -EINVAL if there are
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1325) * no pages to evict or an eviction handler is not registered, -EAGAIN if
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1326) * the retry limit was hit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1327) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1328) static int z3fold_reclaim_page(struct z3fold_pool *pool, unsigned int retries)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1329) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1330) int i, ret = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1331) struct z3fold_header *zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1332) struct page *page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1333) struct list_head *pos;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1334) unsigned long first_handle = 0, middle_handle = 0, last_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1335) struct z3fold_buddy_slots slots __attribute__((aligned(SLOTS_ALIGN)));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1336)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1337) rwlock_init(&slots.lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1338) slots.pool = (unsigned long)pool | (1 << HANDLES_NOFREE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1339)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1340) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1341) if (!pool->ops || !pool->ops->evict || retries == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1342) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1343) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1344) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1345) for (i = 0; i < retries; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1346) if (list_empty(&pool->lru)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1347) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1348) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1349) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1350) list_for_each_prev(pos, &pool->lru) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1351) page = list_entry(pos, struct page, lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1352)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1353) zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1354) if (test_bit(PAGE_HEADLESS, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1355) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1356) * For non-headless pages, we wait to do this
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1357) * until we have the page lock to avoid racing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1358) * with __z3fold_alloc(). Headless pages don't
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1359) * have a lock (and __z3fold_alloc() will never
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1360) * see them), but we still need to test and set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1361) * PAGE_CLAIMED to avoid racing with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1362) * z3fold_free(), so just do it now before
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1363) * leaving the loop.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1364) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1365) if (test_and_set_bit(PAGE_CLAIMED, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1366) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1367)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1368) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1369) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1370)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1371) if (kref_get_unless_zero(&zhdr->refcount) == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1372) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1373) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1374) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1375) if (!z3fold_page_trylock(zhdr)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1376) if (kref_put(&zhdr->refcount,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1377) release_z3fold_page))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1378) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1379) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1380) continue; /* can't evict at this point */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1381) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1382)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1383) /* test_and_set_bit is of course atomic, but we still
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1384) * need to do it under page lock, otherwise checking
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1385) * that bit in __z3fold_alloc wouldn't make sense
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1386) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1387) if (zhdr->foreign_handles ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1388) test_and_set_bit(PAGE_CLAIMED, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1389) if (kref_put(&zhdr->refcount,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1390) release_z3fold_page_locked))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1391) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1392) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1393) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1394) zhdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1395) continue; /* can't evict such page */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1396) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1397) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1398) zhdr->cpu = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1399) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1400) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1401)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1402) if (!zhdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1403) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1404)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1405) list_del_init(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1406) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1407)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1408) if (!test_bit(PAGE_HEADLESS, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1409) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1410) * We need encode the handles before unlocking, and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1411) * use our local slots structure because z3fold_free
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1412) * can zero out zhdr->slots and we can't do much
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1413) * about that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1414) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1415) first_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1416) last_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1417) middle_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1418) memset(slots.slot, 0, sizeof(slots.slot));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1419) if (zhdr->first_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1420) first_handle = __encode_handle(zhdr, &slots,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1421) FIRST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1422) if (zhdr->middle_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1423) middle_handle = __encode_handle(zhdr, &slots,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1424) MIDDLE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1425) if (zhdr->last_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1426) last_handle = __encode_handle(zhdr, &slots,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1427) LAST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1428) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1429) * it's safe to unlock here because we hold a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1430) * reference to this page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1431) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1432) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1433) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1434) first_handle = encode_handle(zhdr, HEADLESS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1435) last_handle = middle_handle = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1436) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1437) /* Issue the eviction callback(s) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1438) if (middle_handle) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1439) ret = pool->ops->evict(pool, middle_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1440) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1441) goto next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1442) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1443) if (first_handle) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1444) ret = pool->ops->evict(pool, first_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1445) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1446) goto next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1447) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1448) if (last_handle) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1449) ret = pool->ops->evict(pool, last_handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1450) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1451) goto next;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1452) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1453) next:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1454) if (test_bit(PAGE_HEADLESS, &page->private)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1455) if (ret == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1456) free_z3fold_page(page, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1457) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1458) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1459) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1460) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1461) list_add(&page->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1462) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1463) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1464) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1465) struct z3fold_buddy_slots *slots = zhdr->slots;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1466) z3fold_page_lock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1467) if (kref_put(&zhdr->refcount,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1468) release_z3fold_page_locked)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1469) kmem_cache_free(pool->c_handle, slots);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1470) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1471) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1472) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1473) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1474) * if we are here, the page is still not completely
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1475) * free. Take the global pool lock then to be able
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1476) * to add it back to the lru list
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1477) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1478) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1479) list_add(&page->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1480) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1481) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1482) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1483) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1484)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1485) /* We started off locked to we need to lock the pool back */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1486) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1487) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1488) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1489) return -EAGAIN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1490) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1491)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1492) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1493) * z3fold_map() - maps the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1494) * @pool: pool in which the allocation resides
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1495) * @handle: handle associated with the allocation to be mapped
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1496) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1497) * Extracts the buddy number from handle and constructs the pointer to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1498) * correct starting chunk within the page.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1499) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1500) * Returns: a pointer to the mapped allocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1501) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1502) static void *z3fold_map(struct z3fold_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1503) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1504) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1505) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1506) void *addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1507) enum buddy buddy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1508)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1509) zhdr = get_z3fold_header(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1510) addr = zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1511) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1512)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1513) if (test_bit(PAGE_HEADLESS, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1514) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1515)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1516) buddy = handle_to_buddy(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1517) switch (buddy) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1518) case FIRST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1519) addr += ZHDR_SIZE_ALIGNED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1520) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1521) case MIDDLE:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1522) addr += zhdr->start_middle << CHUNK_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1523) set_bit(MIDDLE_CHUNK_MAPPED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1524) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1525) case LAST:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1526) addr += PAGE_SIZE - (handle_to_chunks(handle) << CHUNK_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1527) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1528) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1529) pr_err("unknown buddy id %d\n", buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1530) WARN_ON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1531) addr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1532) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1533) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1535) if (addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1536) zhdr->mapped_count++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1537) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1538) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1539) return addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1540) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1541)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1542) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1543) * z3fold_unmap() - unmaps the allocation associated with the given handle
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1544) * @pool: pool in which the allocation resides
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1545) * @handle: handle associated with the allocation to be unmapped
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1546) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1547) static void z3fold_unmap(struct z3fold_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1548) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1549) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1550) struct page *page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1551) enum buddy buddy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1552)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1553) zhdr = get_z3fold_header(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1554) page = virt_to_page(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1555)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1556) if (test_bit(PAGE_HEADLESS, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1557) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1558)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1559) buddy = handle_to_buddy(handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1560) if (buddy == MIDDLE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1561) clear_bit(MIDDLE_CHUNK_MAPPED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1562) zhdr->mapped_count--;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1563) put_z3fold_header(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1564) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1565)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1566) /**
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1567) * z3fold_get_pool_size() - gets the z3fold pool size in pages
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1568) * @pool: pool whose size is being queried
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1569) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1570) * Returns: size in pages of the given pool.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1571) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1572) static u64 z3fold_get_pool_size(struct z3fold_pool *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1573) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1574) return atomic64_read(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1575) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1576)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1577) static bool z3fold_page_isolate(struct page *page, isolate_mode_t mode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1578) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1579) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1580) struct z3fold_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1581)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1582) VM_BUG_ON_PAGE(!PageMovable(page), page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1583) VM_BUG_ON_PAGE(PageIsolated(page), page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1584)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1585) if (test_bit(PAGE_HEADLESS, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1586) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1587)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1588) zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1589) z3fold_page_lock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1590) if (test_bit(NEEDS_COMPACTING, &page->private) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1591) test_bit(PAGE_STALE, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1592) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1593)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1594) if (zhdr->mapped_count != 0 || zhdr->foreign_handles != 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1595) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1596)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1597) if (test_and_set_bit(PAGE_CLAIMED, &page->private))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1598) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1599) pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1600) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1601) if (!list_empty(&zhdr->buddy))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1602) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1603) if (!list_empty(&page->lru))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1604) list_del_init(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1605) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1606)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1607) kref_get(&zhdr->refcount);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1608) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1609) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1610)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1611) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1612) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1613) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1614) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1615)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1616) static int z3fold_page_migrate(struct address_space *mapping, struct page *newpage,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1617) struct page *page, enum migrate_mode mode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1618) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1619) struct z3fold_header *zhdr, *new_zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1620) struct z3fold_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1621) struct address_space *new_mapping;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1622)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1623) VM_BUG_ON_PAGE(!PageMovable(page), page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1624) VM_BUG_ON_PAGE(!PageIsolated(page), page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1625) VM_BUG_ON_PAGE(!test_bit(PAGE_CLAIMED, &page->private), page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1626) VM_BUG_ON_PAGE(!PageLocked(newpage), newpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1627)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1628) zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1629) pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1630)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1631) if (!z3fold_page_trylock(zhdr))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1632) return -EAGAIN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1633) if (zhdr->mapped_count != 0 || zhdr->foreign_handles != 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1634) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1635) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1636) return -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1637) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1638) if (work_pending(&zhdr->work)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1639) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1640) return -EAGAIN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1641) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1642) new_zhdr = page_address(newpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1643) memcpy(new_zhdr, zhdr, PAGE_SIZE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1644) newpage->private = page->private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1645) page->private = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1646) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1647) spin_lock_init(&new_zhdr->page_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1648) INIT_WORK(&new_zhdr->work, compact_page_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1649) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1650) * z3fold_page_isolate() ensures that new_zhdr->buddy is empty,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1651) * so we only have to reinitialize it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1652) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1653) INIT_LIST_HEAD(&new_zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1654) new_mapping = page_mapping(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1655) __ClearPageMovable(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1656) ClearPagePrivate(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1657)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1658) get_page(newpage);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1659) z3fold_page_lock(new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1660) if (new_zhdr->first_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1661) encode_handle(new_zhdr, FIRST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1662) if (new_zhdr->last_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1663) encode_handle(new_zhdr, LAST);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1664) if (new_zhdr->middle_chunks)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1665) encode_handle(new_zhdr, MIDDLE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1666) set_bit(NEEDS_COMPACTING, &newpage->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1667) new_zhdr->cpu = smp_processor_id();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1668) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1669) list_add(&newpage->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1670) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1671) __SetPageMovable(newpage, new_mapping);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1672) z3fold_page_unlock(new_zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1673)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1674) queue_work_on(new_zhdr->cpu, pool->compact_wq, &new_zhdr->work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1675)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1676) page_mapcount_reset(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1677) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1678) put_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1679) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1680) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1681)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1682) static void z3fold_page_putback(struct page *page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1683) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1684) struct z3fold_header *zhdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1685) struct z3fold_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1686)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1687) zhdr = page_address(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1688) pool = zhdr_to_pool(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1689)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1690) z3fold_page_lock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1691) if (!list_empty(&zhdr->buddy))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1692) list_del_init(&zhdr->buddy);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1693) INIT_LIST_HEAD(&page->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1694) if (kref_put(&zhdr->refcount, release_z3fold_page_locked)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1695) atomic64_dec(&pool->pages_nr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1696) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1697) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1698) spin_lock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1699) list_add(&page->lru, &pool->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1700) spin_unlock(&pool->lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1701) clear_bit(PAGE_CLAIMED, &page->private);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1702) z3fold_page_unlock(zhdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1703) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1704)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1705) static const struct address_space_operations z3fold_aops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1706) .isolate_page = z3fold_page_isolate,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1707) .migratepage = z3fold_page_migrate,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1708) .putback_page = z3fold_page_putback,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1709) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1710)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1711) /*****************
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1712) * zpool
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1713) ****************/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1714)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1715) static int z3fold_zpool_evict(struct z3fold_pool *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1716) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1717) if (pool->zpool && pool->zpool_ops && pool->zpool_ops->evict)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1718) return pool->zpool_ops->evict(pool->zpool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1719) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1720) return -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1721) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1722)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1723) static const struct z3fold_ops z3fold_zpool_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1724) .evict = z3fold_zpool_evict
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1725) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1726)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1727) static void *z3fold_zpool_create(const char *name, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1728) const struct zpool_ops *zpool_ops,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1729) struct zpool *zpool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1730) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1731) struct z3fold_pool *pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1732)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1733) pool = z3fold_create_pool(name, gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1734) zpool_ops ? &z3fold_zpool_ops : NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1735) if (pool) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1736) pool->zpool = zpool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1737) pool->zpool_ops = zpool_ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1738) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1739) return pool;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1740) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1741)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1742) static void z3fold_zpool_destroy(void *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1743) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1744) z3fold_destroy_pool(pool);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1745) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1746)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1747) static int z3fold_zpool_malloc(void *pool, size_t size, gfp_t gfp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1748) unsigned long *handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1749) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1750) return z3fold_alloc(pool, size, gfp, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1751) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1752) static void z3fold_zpool_free(void *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1753) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1754) z3fold_free(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1755) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1756)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1757) static int z3fold_zpool_shrink(void *pool, unsigned int pages,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1758) unsigned int *reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1759) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1760) unsigned int total = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1761) int ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1762)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1763) while (total < pages) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1764) ret = z3fold_reclaim_page(pool, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1765) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1766) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1767) total++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1768) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1769)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1770) if (reclaimed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1771) *reclaimed = total;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1772)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1773) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1774) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1775)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1776) static void *z3fold_zpool_map(void *pool, unsigned long handle,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1777) enum zpool_mapmode mm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1778) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1779) return z3fold_map(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1780) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1781) static void z3fold_zpool_unmap(void *pool, unsigned long handle)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1782) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1783) z3fold_unmap(pool, handle);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1784) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1785)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1786) static u64 z3fold_zpool_total_size(void *pool)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1787) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1788) return z3fold_get_pool_size(pool) * PAGE_SIZE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1789) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1790)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1791) static struct zpool_driver z3fold_zpool_driver = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1792) .type = "z3fold",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1793) .owner = THIS_MODULE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1794) .create = z3fold_zpool_create,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1795) .destroy = z3fold_zpool_destroy,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1796) .malloc = z3fold_zpool_malloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1797) .free = z3fold_zpool_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1798) .shrink = z3fold_zpool_shrink,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1799) .map = z3fold_zpool_map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1800) .unmap = z3fold_zpool_unmap,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1801) .total_size = z3fold_zpool_total_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1802) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1803)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1804) MODULE_ALIAS("zpool-z3fold");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1805)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1806) static int __init init_z3fold(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1807) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1808) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1809)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1810) /* Make sure the z3fold header is not larger than the page size */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1811) BUILD_BUG_ON(ZHDR_SIZE_ALIGNED > PAGE_SIZE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1812) ret = z3fold_mount();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1813) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1814) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1815)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1816) zpool_register_driver(&z3fold_zpool_driver);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1817)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1818) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1819) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1820)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1821) static void __exit exit_z3fold(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1822) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1823) z3fold_unmount();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1824) zpool_unregister_driver(&z3fold_zpool_driver);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1825) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1826)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1827) module_init(init_z3fold);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1828) module_exit(exit_z3fold);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1829)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1830) MODULE_LICENSE("GPL");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1831) MODULE_AUTHOR("Vitaly Wool <vitalywool@gmail.com>");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1832) MODULE_DESCRIPTION("3-Fold 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