Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (C) 2002 Sistina Software (UK) Limited. |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 3 | * Copyright (C) 2006 Red Hat GmbH |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 4 | * |
| 5 | * This file is released under the GPL. |
| 6 | * |
| 7 | * Kcopyd provides a simple interface for copying an area of one |
| 8 | * block-device to one or more other block-devices, with an asynchronous |
| 9 | * completion notification. |
| 10 | */ |
| 11 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 12 | #include <linux/types.h> |
Arun Sharma | 60063497 | 2011-07-26 16:09:06 -0700 | [diff] [blame] | 13 | #include <linux/atomic.h> |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 14 | #include <linux/blkdev.h> |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 15 | #include <linux/fs.h> |
| 16 | #include <linux/init.h> |
| 17 | #include <linux/list.h> |
| 18 | #include <linux/mempool.h> |
| 19 | #include <linux/module.h> |
| 20 | #include <linux/pagemap.h> |
| 21 | #include <linux/slab.h> |
| 22 | #include <linux/vmalloc.h> |
| 23 | #include <linux/workqueue.h> |
Arjan van de Ven | 48c9c27 | 2006-03-27 01:18:20 -0800 | [diff] [blame] | 24 | #include <linux/mutex.h> |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 25 | #include <linux/delay.h> |
Mikulas Patocka | 586e80e | 2008-10-21 17:44:59 +0100 | [diff] [blame] | 26 | #include <linux/device-mapper.h> |
Alasdair G Kergon | a765e20 | 2008-04-24 22:02:01 +0100 | [diff] [blame] | 27 | #include <linux/dm-kcopyd.h> |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 28 | |
Mike Snitzer | 4cc9613 | 2016-05-12 16:28:10 -0400 | [diff] [blame] | 29 | #include "dm-core.h" |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 30 | |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 31 | #define SUB_JOB_SIZE 128 |
| 32 | #define SPLIT_COUNT 8 |
| 33 | #define MIN_JOBS 8 |
Mikulas Patocka | 5f43ba2 | 2011-05-29 13:03:11 +0100 | [diff] [blame] | 34 | #define RESERVE_PAGES (DIV_ROUND_UP(SUB_JOB_SIZE << SECTOR_SHIFT, PAGE_SIZE)) |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 35 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 36 | /*----------------------------------------------------------------- |
| 37 | * Each kcopyd client has its own little pool of preallocated |
| 38 | * pages for kcopyd io. |
| 39 | *---------------------------------------------------------------*/ |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 40 | struct dm_kcopyd_client { |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 41 | struct page_list *pages; |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 42 | unsigned nr_reserved_pages; |
| 43 | unsigned nr_free_pages; |
Alasdair G Kergon | 138728dc | 2006-03-27 01:17:50 -0800 | [diff] [blame] | 44 | |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 45 | struct dm_io_client *io_client; |
| 46 | |
Alasdair G Kergon | 138728dc | 2006-03-27 01:17:50 -0800 | [diff] [blame] | 47 | wait_queue_head_t destroyq; |
| 48 | atomic_t nr_jobs; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 49 | |
Mikulas Patocka | 08d8757 | 2008-04-24 21:43:46 +0100 | [diff] [blame] | 50 | mempool_t *job_pool; |
| 51 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 52 | struct workqueue_struct *kcopyd_wq; |
| 53 | struct work_struct kcopyd_work; |
| 54 | |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 55 | struct dm_kcopyd_throttle *throttle; |
| 56 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 57 | /* |
| 58 | * We maintain three lists of jobs: |
| 59 | * |
| 60 | * i) jobs waiting for pages |
| 61 | * ii) jobs that have pages, and are waiting for the io to be issued. |
| 62 | * iii) jobs that have completed. |
| 63 | * |
| 64 | * All three of these are protected by job_lock. |
| 65 | */ |
| 66 | spinlock_t job_lock; |
| 67 | struct list_head complete_jobs; |
| 68 | struct list_head io_jobs; |
| 69 | struct list_head pages_jobs; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 70 | }; |
| 71 | |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 72 | static struct page_list zero_page_list; |
| 73 | |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 74 | static DEFINE_SPINLOCK(throttle_spinlock); |
| 75 | |
| 76 | /* |
| 77 | * IO/IDLE accounting slowly decays after (1 << ACCOUNT_INTERVAL_SHIFT) period. |
| 78 | * When total_period >= (1 << ACCOUNT_INTERVAL_SHIFT) the counters are divided |
| 79 | * by 2. |
| 80 | */ |
| 81 | #define ACCOUNT_INTERVAL_SHIFT SHIFT_HZ |
| 82 | |
| 83 | /* |
| 84 | * Sleep this number of milliseconds. |
| 85 | * |
| 86 | * The value was decided experimentally. |
| 87 | * Smaller values seem to cause an increased copy rate above the limit. |
| 88 | * The reason for this is unknown but possibly due to jiffies rounding errors |
| 89 | * or read/write cache inside the disk. |
| 90 | */ |
| 91 | #define SLEEP_MSEC 100 |
| 92 | |
| 93 | /* |
| 94 | * Maximum number of sleep events. There is a theoretical livelock if more |
| 95 | * kcopyd clients do work simultaneously which this limit avoids. |
| 96 | */ |
| 97 | #define MAX_SLEEPS 10 |
| 98 | |
| 99 | static void io_job_start(struct dm_kcopyd_throttle *t) |
| 100 | { |
| 101 | unsigned throttle, now, difference; |
| 102 | int slept = 0, skew; |
| 103 | |
| 104 | if (unlikely(!t)) |
| 105 | return; |
| 106 | |
| 107 | try_again: |
| 108 | spin_lock_irq(&throttle_spinlock); |
| 109 | |
Mark Rutland | 6aa7de0 | 2017-10-23 14:07:29 -0700 | [diff] [blame] | 110 | throttle = READ_ONCE(t->throttle); |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 111 | |
| 112 | if (likely(throttle >= 100)) |
| 113 | goto skip_limit; |
| 114 | |
| 115 | now = jiffies; |
| 116 | difference = now - t->last_jiffies; |
| 117 | t->last_jiffies = now; |
| 118 | if (t->num_io_jobs) |
| 119 | t->io_period += difference; |
| 120 | t->total_period += difference; |
| 121 | |
| 122 | /* |
| 123 | * Maintain sane values if we got a temporary overflow. |
| 124 | */ |
| 125 | if (unlikely(t->io_period > t->total_period)) |
| 126 | t->io_period = t->total_period; |
| 127 | |
| 128 | if (unlikely(t->total_period >= (1 << ACCOUNT_INTERVAL_SHIFT))) { |
| 129 | int shift = fls(t->total_period >> ACCOUNT_INTERVAL_SHIFT); |
| 130 | t->total_period >>= shift; |
| 131 | t->io_period >>= shift; |
| 132 | } |
| 133 | |
| 134 | skew = t->io_period - throttle * t->total_period / 100; |
| 135 | |
| 136 | if (unlikely(skew > 0) && slept < MAX_SLEEPS) { |
| 137 | slept++; |
| 138 | spin_unlock_irq(&throttle_spinlock); |
| 139 | msleep(SLEEP_MSEC); |
| 140 | goto try_again; |
| 141 | } |
| 142 | |
| 143 | skip_limit: |
| 144 | t->num_io_jobs++; |
| 145 | |
| 146 | spin_unlock_irq(&throttle_spinlock); |
| 147 | } |
| 148 | |
| 149 | static void io_job_finish(struct dm_kcopyd_throttle *t) |
| 150 | { |
| 151 | unsigned long flags; |
| 152 | |
| 153 | if (unlikely(!t)) |
| 154 | return; |
| 155 | |
| 156 | spin_lock_irqsave(&throttle_spinlock, flags); |
| 157 | |
| 158 | t->num_io_jobs--; |
| 159 | |
Mark Rutland | 6aa7de0 | 2017-10-23 14:07:29 -0700 | [diff] [blame] | 160 | if (likely(READ_ONCE(t->throttle) >= 100)) |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 161 | goto skip_limit; |
| 162 | |
| 163 | if (!t->num_io_jobs) { |
| 164 | unsigned now, difference; |
| 165 | |
| 166 | now = jiffies; |
| 167 | difference = now - t->last_jiffies; |
| 168 | t->last_jiffies = now; |
| 169 | |
| 170 | t->io_period += difference; |
| 171 | t->total_period += difference; |
| 172 | |
| 173 | /* |
| 174 | * Maintain sane values if we got a temporary overflow. |
| 175 | */ |
| 176 | if (unlikely(t->io_period > t->total_period)) |
| 177 | t->io_period = t->total_period; |
| 178 | } |
| 179 | |
| 180 | skip_limit: |
| 181 | spin_unlock_irqrestore(&throttle_spinlock, flags); |
| 182 | } |
| 183 | |
| 184 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 185 | static void wake(struct dm_kcopyd_client *kc) |
| 186 | { |
| 187 | queue_work(kc->kcopyd_wq, &kc->kcopyd_work); |
| 188 | } |
| 189 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 190 | /* |
| 191 | * Obtain one page for the use of kcopyd. |
| 192 | */ |
Mikulas Patocka | f99b55e | 2011-05-29 13:03:04 +0100 | [diff] [blame] | 193 | static struct page_list *alloc_pl(gfp_t gfp) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 194 | { |
| 195 | struct page_list *pl; |
| 196 | |
Mikulas Patocka | f99b55e | 2011-05-29 13:03:04 +0100 | [diff] [blame] | 197 | pl = kmalloc(sizeof(*pl), gfp); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 198 | if (!pl) |
| 199 | return NULL; |
| 200 | |
Mikulas Patocka | f99b55e | 2011-05-29 13:03:04 +0100 | [diff] [blame] | 201 | pl->page = alloc_page(gfp); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 202 | if (!pl->page) { |
| 203 | kfree(pl); |
| 204 | return NULL; |
| 205 | } |
| 206 | |
| 207 | return pl; |
| 208 | } |
| 209 | |
| 210 | static void free_pl(struct page_list *pl) |
| 211 | { |
| 212 | __free_page(pl->page); |
| 213 | kfree(pl); |
| 214 | } |
| 215 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 216 | /* |
| 217 | * Add the provided pages to a client's free page list, releasing |
| 218 | * back to the system any beyond the reserved_pages limit. |
| 219 | */ |
| 220 | static void kcopyd_put_pages(struct dm_kcopyd_client *kc, struct page_list *pl) |
| 221 | { |
| 222 | struct page_list *next; |
| 223 | |
| 224 | do { |
| 225 | next = pl->next; |
| 226 | |
| 227 | if (kc->nr_free_pages >= kc->nr_reserved_pages) |
| 228 | free_pl(pl); |
| 229 | else { |
| 230 | pl->next = kc->pages; |
| 231 | kc->pages = pl; |
| 232 | kc->nr_free_pages++; |
| 233 | } |
| 234 | |
| 235 | pl = next; |
| 236 | } while (pl); |
| 237 | } |
| 238 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 239 | static int kcopyd_get_pages(struct dm_kcopyd_client *kc, |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 240 | unsigned int nr, struct page_list **pages) |
| 241 | { |
| 242 | struct page_list *pl; |
| 243 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 244 | *pages = NULL; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 245 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 246 | do { |
Mel Gorman | d0164ad | 2015-11-06 16:28:21 -0800 | [diff] [blame] | 247 | pl = alloc_pl(__GFP_NOWARN | __GFP_NORETRY | __GFP_KSWAPD_RECLAIM); |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 248 | if (unlikely(!pl)) { |
| 249 | /* Use reserved pages */ |
| 250 | pl = kc->pages; |
| 251 | if (unlikely(!pl)) |
| 252 | goto out_of_memory; |
| 253 | kc->pages = pl->next; |
| 254 | kc->nr_free_pages--; |
| 255 | } |
| 256 | pl->next = *pages; |
| 257 | *pages = pl; |
| 258 | } while (--nr); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 259 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 260 | return 0; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 261 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 262 | out_of_memory: |
| 263 | if (*pages) |
| 264 | kcopyd_put_pages(kc, *pages); |
| 265 | return -ENOMEM; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 266 | } |
| 267 | |
| 268 | /* |
| 269 | * These three functions resize the page pool. |
| 270 | */ |
| 271 | static void drop_pages(struct page_list *pl) |
| 272 | { |
| 273 | struct page_list *next; |
| 274 | |
| 275 | while (pl) { |
| 276 | next = pl->next; |
| 277 | free_pl(pl); |
| 278 | pl = next; |
| 279 | } |
| 280 | } |
| 281 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 282 | /* |
| 283 | * Allocate and reserve nr_pages for the use of a specific client. |
| 284 | */ |
| 285 | static int client_reserve_pages(struct dm_kcopyd_client *kc, unsigned nr_pages) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 286 | { |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 287 | unsigned i; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 288 | struct page_list *pl = NULL, *next; |
| 289 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 290 | for (i = 0; i < nr_pages; i++) { |
Mikulas Patocka | f99b55e | 2011-05-29 13:03:04 +0100 | [diff] [blame] | 291 | next = alloc_pl(GFP_KERNEL); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 292 | if (!next) { |
| 293 | if (pl) |
| 294 | drop_pages(pl); |
| 295 | return -ENOMEM; |
| 296 | } |
| 297 | next->next = pl; |
| 298 | pl = next; |
| 299 | } |
| 300 | |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 301 | kc->nr_reserved_pages += nr_pages; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 302 | kcopyd_put_pages(kc, pl); |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 303 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 304 | return 0; |
| 305 | } |
| 306 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 307 | static void client_free_pages(struct dm_kcopyd_client *kc) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 308 | { |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 309 | BUG_ON(kc->nr_free_pages != kc->nr_reserved_pages); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 310 | drop_pages(kc->pages); |
| 311 | kc->pages = NULL; |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 312 | kc->nr_free_pages = kc->nr_reserved_pages = 0; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 313 | } |
| 314 | |
| 315 | /*----------------------------------------------------------------- |
| 316 | * kcopyd_jobs need to be allocated by the *clients* of kcopyd, |
| 317 | * for this reason we use a mempool to prevent the client from |
| 318 | * ever having to do io (which could cause a deadlock). |
| 319 | *---------------------------------------------------------------*/ |
| 320 | struct kcopyd_job { |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 321 | struct dm_kcopyd_client *kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 322 | struct list_head list; |
| 323 | unsigned long flags; |
| 324 | |
| 325 | /* |
| 326 | * Error state of the job. |
| 327 | */ |
| 328 | int read_err; |
Alasdair G Kergon | 4cdc1d1 | 2008-03-28 14:16:10 -0700 | [diff] [blame] | 329 | unsigned long write_err; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 330 | |
| 331 | /* |
| 332 | * Either READ or WRITE |
| 333 | */ |
| 334 | int rw; |
Heinz Mauelshagen | 22a1ceb | 2008-04-24 21:43:17 +0100 | [diff] [blame] | 335 | struct dm_io_region source; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 336 | |
| 337 | /* |
| 338 | * The destinations for the transfer. |
| 339 | */ |
| 340 | unsigned int num_dests; |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 341 | struct dm_io_region dests[DM_KCOPYD_MAX_REGIONS]; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 342 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 343 | struct page_list *pages; |
| 344 | |
| 345 | /* |
| 346 | * Set this to ensure you are notified when the job has |
| 347 | * completed. 'context' is for callback to use. |
| 348 | */ |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 349 | dm_kcopyd_notify_fn fn; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 350 | void *context; |
| 351 | |
| 352 | /* |
| 353 | * These fields are only used if the job has been split |
| 354 | * into more manageable parts. |
| 355 | */ |
Matthias Kaehlcke | def5b5b | 2007-10-19 22:38:52 +0100 | [diff] [blame] | 356 | struct mutex lock; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 357 | atomic_t sub_jobs; |
| 358 | sector_t progress; |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 359 | sector_t write_offset; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 360 | |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 361 | struct kcopyd_job *master_job; |
| 362 | }; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 363 | |
Christoph Lameter | e18b890 | 2006-12-06 20:33:20 -0800 | [diff] [blame] | 364 | static struct kmem_cache *_job_cache; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 365 | |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 366 | int __init dm_kcopyd_init(void) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 367 | { |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 368 | _job_cache = kmem_cache_create("kcopyd_job", |
| 369 | sizeof(struct kcopyd_job) * (SPLIT_COUNT + 1), |
| 370 | __alignof__(struct kcopyd_job), 0, NULL); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 371 | if (!_job_cache) |
| 372 | return -ENOMEM; |
| 373 | |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 374 | zero_page_list.next = &zero_page_list; |
| 375 | zero_page_list.page = ZERO_PAGE(0); |
| 376 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 377 | return 0; |
| 378 | } |
| 379 | |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 380 | void dm_kcopyd_exit(void) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 381 | { |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 382 | kmem_cache_destroy(_job_cache); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 383 | _job_cache = NULL; |
| 384 | } |
| 385 | |
| 386 | /* |
| 387 | * Functions to push and pop a job onto the head of a given job |
| 388 | * list. |
| 389 | */ |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 390 | static struct kcopyd_job *pop_io_job(struct list_head *jobs, |
| 391 | struct dm_kcopyd_client *kc) |
| 392 | { |
| 393 | struct kcopyd_job *job; |
| 394 | |
| 395 | /* |
| 396 | * For I/O jobs, pop any read, any write without sequential write |
| 397 | * constraint and sequential writes that are at the right position. |
| 398 | */ |
| 399 | list_for_each_entry(job, jobs, list) { |
| 400 | if (job->rw == READ || !test_bit(DM_KCOPYD_WRITE_SEQ, &job->flags)) { |
| 401 | list_del(&job->list); |
| 402 | return job; |
| 403 | } |
| 404 | |
| 405 | if (job->write_offset == job->master_job->write_offset) { |
| 406 | job->master_job->write_offset += job->source.count; |
| 407 | list_del(&job->list); |
| 408 | return job; |
| 409 | } |
| 410 | } |
| 411 | |
| 412 | return NULL; |
| 413 | } |
| 414 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 415 | static struct kcopyd_job *pop(struct list_head *jobs, |
| 416 | struct dm_kcopyd_client *kc) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 417 | { |
| 418 | struct kcopyd_job *job = NULL; |
| 419 | unsigned long flags; |
| 420 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 421 | spin_lock_irqsave(&kc->job_lock, flags); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 422 | |
| 423 | if (!list_empty(jobs)) { |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 424 | if (jobs == &kc->io_jobs) |
| 425 | job = pop_io_job(jobs, kc); |
| 426 | else { |
| 427 | job = list_entry(jobs->next, struct kcopyd_job, list); |
| 428 | list_del(&job->list); |
| 429 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 430 | } |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 431 | spin_unlock_irqrestore(&kc->job_lock, flags); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 432 | |
| 433 | return job; |
| 434 | } |
| 435 | |
Alasdair G Kergon | 028867a | 2007-07-12 17:26:32 +0100 | [diff] [blame] | 436 | static void push(struct list_head *jobs, struct kcopyd_job *job) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 437 | { |
| 438 | unsigned long flags; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 439 | struct dm_kcopyd_client *kc = job->kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 440 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 441 | spin_lock_irqsave(&kc->job_lock, flags); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 442 | list_add_tail(&job->list, jobs); |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 443 | spin_unlock_irqrestore(&kc->job_lock, flags); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 444 | } |
| 445 | |
Kazuo Ito | b673c3a | 2008-10-21 17:44:50 +0100 | [diff] [blame] | 446 | |
| 447 | static void push_head(struct list_head *jobs, struct kcopyd_job *job) |
| 448 | { |
| 449 | unsigned long flags; |
| 450 | struct dm_kcopyd_client *kc = job->kc; |
| 451 | |
| 452 | spin_lock_irqsave(&kc->job_lock, flags); |
| 453 | list_add(&job->list, jobs); |
| 454 | spin_unlock_irqrestore(&kc->job_lock, flags); |
| 455 | } |
| 456 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 457 | /* |
| 458 | * These three functions process 1 item from the corresponding |
| 459 | * job list. |
| 460 | * |
| 461 | * They return: |
| 462 | * < 0: error |
| 463 | * 0: success |
| 464 | * > 0: can't process yet. |
| 465 | */ |
| 466 | static int run_complete_job(struct kcopyd_job *job) |
| 467 | { |
| 468 | void *context = job->context; |
| 469 | int read_err = job->read_err; |
Alasdair G Kergon | 4cdc1d1 | 2008-03-28 14:16:10 -0700 | [diff] [blame] | 470 | unsigned long write_err = job->write_err; |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 471 | dm_kcopyd_notify_fn fn = job->fn; |
| 472 | struct dm_kcopyd_client *kc = job->kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 473 | |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 474 | if (job->pages && job->pages != &zero_page_list) |
Mikulas Patocka | 7383085 | 2009-04-09 00:27:16 +0100 | [diff] [blame] | 475 | kcopyd_put_pages(kc, job->pages); |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 476 | /* |
| 477 | * If this is the master job, the sub jobs have already |
| 478 | * completed so we can free everything. |
| 479 | */ |
| 480 | if (job->master_job == job) |
| 481 | mempool_free(job, kc->job_pool); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 482 | fn(read_err, write_err, context); |
Alasdair G Kergon | 138728dc | 2006-03-27 01:17:50 -0800 | [diff] [blame] | 483 | |
| 484 | if (atomic_dec_and_test(&kc->nr_jobs)) |
| 485 | wake_up(&kc->destroyq); |
| 486 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 487 | return 0; |
| 488 | } |
| 489 | |
| 490 | static void complete_io(unsigned long error, void *context) |
| 491 | { |
| 492 | struct kcopyd_job *job = (struct kcopyd_job *) context; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 493 | struct dm_kcopyd_client *kc = job->kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 494 | |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 495 | io_job_finish(kc->throttle); |
| 496 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 497 | if (error) { |
Mike Christie | 5111166 | 2016-06-05 14:31:46 -0500 | [diff] [blame] | 498 | if (op_is_write(job->rw)) |
Jonathan Brassow | ce503f5 | 2006-06-26 00:27:30 -0700 | [diff] [blame] | 499 | job->write_err |= error; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 500 | else |
| 501 | job->read_err = 1; |
| 502 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 503 | if (!test_bit(DM_KCOPYD_IGNORE_ERROR, &job->flags)) { |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 504 | push(&kc->complete_jobs, job); |
| 505 | wake(kc); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 506 | return; |
| 507 | } |
| 508 | } |
| 509 | |
Mike Christie | 5111166 | 2016-06-05 14:31:46 -0500 | [diff] [blame] | 510 | if (op_is_write(job->rw)) |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 511 | push(&kc->complete_jobs, job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 512 | |
| 513 | else { |
| 514 | job->rw = WRITE; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 515 | push(&kc->io_jobs, job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 516 | } |
| 517 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 518 | wake(kc); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 519 | } |
| 520 | |
| 521 | /* |
| 522 | * Request io on as many buffer heads as we can currently get for |
| 523 | * a particular job. |
| 524 | */ |
| 525 | static int run_io_job(struct kcopyd_job *job) |
| 526 | { |
| 527 | int r; |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 528 | struct dm_io_request io_req = { |
Mike Christie | e604714 | 2016-06-05 14:32:04 -0500 | [diff] [blame] | 529 | .bi_op = job->rw, |
| 530 | .bi_op_flags = 0, |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 531 | .mem.type = DM_IO_PAGE_LIST, |
| 532 | .mem.ptr.pl = job->pages, |
Mikulas Patocka | 4622afb | 2011-08-02 12:32:02 +0100 | [diff] [blame] | 533 | .mem.offset = 0, |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 534 | .notify.fn = complete_io, |
| 535 | .notify.context = job, |
| 536 | .client = job->kc->io_client, |
| 537 | }; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 538 | |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 539 | /* |
| 540 | * If we need to write sequentially and some reads or writes failed, |
| 541 | * no point in continuing. |
| 542 | */ |
| 543 | if (test_bit(DM_KCOPYD_WRITE_SEQ, &job->flags) && |
| 544 | job->master_job->write_err) |
| 545 | return -EIO; |
| 546 | |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 547 | io_job_start(job->kc->throttle); |
| 548 | |
Jens Axboe | 7eaceac | 2011-03-10 08:52:07 +0100 | [diff] [blame] | 549 | if (job->rw == READ) |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 550 | r = dm_io(&io_req, 1, &job->source, NULL); |
Jens Axboe | 721a960 | 2011-03-09 11:56:30 +0100 | [diff] [blame] | 551 | else |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 552 | r = dm_io(&io_req, job->num_dests, job->dests, NULL); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 553 | |
| 554 | return r; |
| 555 | } |
| 556 | |
| 557 | static int run_pages_job(struct kcopyd_job *job) |
| 558 | { |
| 559 | int r; |
Mikulas Patocka | 5bf45a3 | 2011-08-02 12:32:02 +0100 | [diff] [blame] | 560 | unsigned nr_pages = dm_div_up(job->dests[0].count, PAGE_SIZE >> 9); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 561 | |
Mikulas Patocka | 5bf45a3 | 2011-08-02 12:32:02 +0100 | [diff] [blame] | 562 | r = kcopyd_get_pages(job->kc, nr_pages, &job->pages); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 563 | if (!r) { |
| 564 | /* this job is ready for io */ |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 565 | push(&job->kc->io_jobs, job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 566 | return 0; |
| 567 | } |
| 568 | |
| 569 | if (r == -ENOMEM) |
| 570 | /* can't complete now */ |
| 571 | return 1; |
| 572 | |
| 573 | return r; |
| 574 | } |
| 575 | |
| 576 | /* |
| 577 | * Run through a list for as long as possible. Returns the count |
| 578 | * of successful jobs. |
| 579 | */ |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 580 | static int process_jobs(struct list_head *jobs, struct dm_kcopyd_client *kc, |
| 581 | int (*fn) (struct kcopyd_job *)) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 582 | { |
| 583 | struct kcopyd_job *job; |
| 584 | int r, count = 0; |
| 585 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 586 | while ((job = pop(jobs, kc))) { |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 587 | |
| 588 | r = fn(job); |
| 589 | |
| 590 | if (r < 0) { |
| 591 | /* error this rogue job */ |
Mike Christie | 5111166 | 2016-06-05 14:31:46 -0500 | [diff] [blame] | 592 | if (op_is_write(job->rw)) |
Alasdair G Kergon | 4cdc1d1 | 2008-03-28 14:16:10 -0700 | [diff] [blame] | 593 | job->write_err = (unsigned long) -1L; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 594 | else |
| 595 | job->read_err = 1; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 596 | push(&kc->complete_jobs, job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 597 | break; |
| 598 | } |
| 599 | |
| 600 | if (r > 0) { |
| 601 | /* |
| 602 | * We couldn't service this job ATM, so |
| 603 | * push this job back onto the list. |
| 604 | */ |
Kazuo Ito | b673c3a | 2008-10-21 17:44:50 +0100 | [diff] [blame] | 605 | push_head(jobs, job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 606 | break; |
| 607 | } |
| 608 | |
| 609 | count++; |
| 610 | } |
| 611 | |
| 612 | return count; |
| 613 | } |
| 614 | |
| 615 | /* |
| 616 | * kcopyd does this every time it's woken up. |
| 617 | */ |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 618 | static void do_work(struct work_struct *work) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 619 | { |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 620 | struct dm_kcopyd_client *kc = container_of(work, |
| 621 | struct dm_kcopyd_client, kcopyd_work); |
Jens Axboe | 7eaceac | 2011-03-10 08:52:07 +0100 | [diff] [blame] | 622 | struct blk_plug plug; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 623 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 624 | /* |
| 625 | * The order that these are called is *very* important. |
| 626 | * complete jobs can free some pages for pages jobs. |
| 627 | * Pages jobs when successful will jump onto the io jobs |
| 628 | * list. io jobs call wake when they complete and it all |
| 629 | * starts again. |
| 630 | */ |
Jens Axboe | 7eaceac | 2011-03-10 08:52:07 +0100 | [diff] [blame] | 631 | blk_start_plug(&plug); |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 632 | process_jobs(&kc->complete_jobs, kc, run_complete_job); |
| 633 | process_jobs(&kc->pages_jobs, kc, run_pages_job); |
| 634 | process_jobs(&kc->io_jobs, kc, run_io_job); |
Jens Axboe | 7eaceac | 2011-03-10 08:52:07 +0100 | [diff] [blame] | 635 | blk_finish_plug(&plug); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 636 | } |
| 637 | |
| 638 | /* |
| 639 | * If we are copying a small region we just dispatch a single job |
| 640 | * to do the copy, otherwise the io has to be split up into many |
| 641 | * jobs. |
| 642 | */ |
| 643 | static void dispatch_job(struct kcopyd_job *job) |
| 644 | { |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 645 | struct dm_kcopyd_client *kc = job->kc; |
| 646 | atomic_inc(&kc->nr_jobs); |
Mikulas Patocka | 9ca170a | 2009-12-10 23:52:13 +0000 | [diff] [blame] | 647 | if (unlikely(!job->source.count)) |
| 648 | push(&kc->complete_jobs, job); |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 649 | else if (job->pages == &zero_page_list) |
| 650 | push(&kc->io_jobs, job); |
Mikulas Patocka | 9ca170a | 2009-12-10 23:52:13 +0000 | [diff] [blame] | 651 | else |
| 652 | push(&kc->pages_jobs, job); |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 653 | wake(kc); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 654 | } |
| 655 | |
Alasdair G Kergon | 4cdc1d1 | 2008-03-28 14:16:10 -0700 | [diff] [blame] | 656 | static void segment_complete(int read_err, unsigned long write_err, |
| 657 | void *context) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 658 | { |
| 659 | /* FIXME: tidy this function */ |
| 660 | sector_t progress = 0; |
| 661 | sector_t count = 0; |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 662 | struct kcopyd_job *sub_job = (struct kcopyd_job *) context; |
| 663 | struct kcopyd_job *job = sub_job->master_job; |
Mikulas Patocka | 7383085 | 2009-04-09 00:27:16 +0100 | [diff] [blame] | 664 | struct dm_kcopyd_client *kc = job->kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 665 | |
Matthias Kaehlcke | def5b5b | 2007-10-19 22:38:52 +0100 | [diff] [blame] | 666 | mutex_lock(&job->lock); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 667 | |
| 668 | /* update the error */ |
| 669 | if (read_err) |
| 670 | job->read_err = 1; |
| 671 | |
| 672 | if (write_err) |
Jonathan Brassow | ce503f5 | 2006-06-26 00:27:30 -0700 | [diff] [blame] | 673 | job->write_err |= write_err; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 674 | |
| 675 | /* |
| 676 | * Only dispatch more work if there hasn't been an error. |
| 677 | */ |
| 678 | if ((!job->read_err && !job->write_err) || |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 679 | test_bit(DM_KCOPYD_IGNORE_ERROR, &job->flags)) { |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 680 | /* get the next chunk of work */ |
| 681 | progress = job->progress; |
| 682 | count = job->source.count - progress; |
| 683 | if (count) { |
| 684 | if (count > SUB_JOB_SIZE) |
| 685 | count = SUB_JOB_SIZE; |
| 686 | |
| 687 | job->progress += count; |
| 688 | } |
| 689 | } |
Matthias Kaehlcke | def5b5b | 2007-10-19 22:38:52 +0100 | [diff] [blame] | 690 | mutex_unlock(&job->lock); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 691 | |
| 692 | if (count) { |
| 693 | int i; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 694 | |
| 695 | *sub_job = *job; |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 696 | sub_job->write_offset = progress; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 697 | sub_job->source.sector += progress; |
| 698 | sub_job->source.count = count; |
| 699 | |
| 700 | for (i = 0; i < job->num_dests; i++) { |
| 701 | sub_job->dests[i].sector += progress; |
| 702 | sub_job->dests[i].count = count; |
| 703 | } |
| 704 | |
| 705 | sub_job->fn = segment_complete; |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 706 | sub_job->context = sub_job; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 707 | dispatch_job(sub_job); |
| 708 | |
| 709 | } else if (atomic_dec_and_test(&job->sub_jobs)) { |
| 710 | |
| 711 | /* |
Mikulas Patocka | 340cd44 | 2009-04-09 00:27:17 +0100 | [diff] [blame] | 712 | * Queue the completion callback to the kcopyd thread. |
| 713 | * |
| 714 | * Some callers assume that all the completions are called |
| 715 | * from a single thread and don't race with each other. |
| 716 | * |
| 717 | * We must not call the callback directly here because this |
| 718 | * code may not be executing in the thread. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 719 | */ |
Mikulas Patocka | 340cd44 | 2009-04-09 00:27:17 +0100 | [diff] [blame] | 720 | push(&kc->complete_jobs, job); |
| 721 | wake(kc); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 722 | } |
| 723 | } |
| 724 | |
| 725 | /* |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 726 | * Create some sub jobs to share the work between them. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 727 | */ |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 728 | static void split_job(struct kcopyd_job *master_job) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 729 | { |
| 730 | int i; |
| 731 | |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 732 | atomic_inc(&master_job->kc->nr_jobs); |
Mikulas Patocka | 340cd44 | 2009-04-09 00:27:17 +0100 | [diff] [blame] | 733 | |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 734 | atomic_set(&master_job->sub_jobs, SPLIT_COUNT); |
| 735 | for (i = 0; i < SPLIT_COUNT; i++) { |
| 736 | master_job[i + 1].master_job = master_job; |
| 737 | segment_complete(0, 0u, &master_job[i + 1]); |
| 738 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 739 | } |
| 740 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 741 | int dm_kcopyd_copy(struct dm_kcopyd_client *kc, struct dm_io_region *from, |
| 742 | unsigned int num_dests, struct dm_io_region *dests, |
| 743 | unsigned int flags, dm_kcopyd_notify_fn fn, void *context) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 744 | { |
| 745 | struct kcopyd_job *job; |
Mike Snitzer | 70d6c40 | 2012-12-21 20:23:37 +0000 | [diff] [blame] | 746 | int i; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 747 | |
| 748 | /* |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 749 | * Allocate an array of jobs consisting of one master job |
| 750 | * followed by SPLIT_COUNT sub jobs. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 751 | */ |
Mikulas Patocka | 08d8757 | 2008-04-24 21:43:46 +0100 | [diff] [blame] | 752 | job = mempool_alloc(kc->job_pool, GFP_NOIO); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 753 | |
| 754 | /* |
| 755 | * set up for the read. |
| 756 | */ |
| 757 | job->kc = kc; |
| 758 | job->flags = flags; |
| 759 | job->read_err = 0; |
| 760 | job->write_err = 0; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 761 | |
| 762 | job->num_dests = num_dests; |
| 763 | memcpy(&job->dests, dests, sizeof(*dests) * num_dests); |
| 764 | |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 765 | /* |
| 766 | * If one of the destination is a host-managed zoned block device, |
| 767 | * we need to write sequentially. If one of the destination is a |
| 768 | * host-aware device, then leave it to the caller to choose what to do. |
| 769 | */ |
| 770 | if (!test_bit(DM_KCOPYD_WRITE_SEQ, &job->flags)) { |
| 771 | for (i = 0; i < job->num_dests; i++) { |
| 772 | if (bdev_zoned_model(dests[i].bdev) == BLK_ZONED_HM) { |
| 773 | set_bit(DM_KCOPYD_WRITE_SEQ, &job->flags); |
| 774 | break; |
| 775 | } |
| 776 | } |
| 777 | } |
| 778 | |
| 779 | /* |
| 780 | * If we need to write sequentially, errors cannot be ignored. |
| 781 | */ |
| 782 | if (test_bit(DM_KCOPYD_WRITE_SEQ, &job->flags) && |
| 783 | test_bit(DM_KCOPYD_IGNORE_ERROR, &job->flags)) |
| 784 | clear_bit(DM_KCOPYD_IGNORE_ERROR, &job->flags); |
| 785 | |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 786 | if (from) { |
| 787 | job->source = *from; |
| 788 | job->pages = NULL; |
| 789 | job->rw = READ; |
| 790 | } else { |
| 791 | memset(&job->source, 0, sizeof job->source); |
| 792 | job->source.count = job->dests[0].count; |
| 793 | job->pages = &zero_page_list; |
Mike Snitzer | 70d6c40 | 2012-12-21 20:23:37 +0000 | [diff] [blame] | 794 | |
| 795 | /* |
Christoph Hellwig | 615ec94 | 2017-04-05 19:21:06 +0200 | [diff] [blame] | 796 | * Use WRITE ZEROES to optimize zeroing if all dests support it. |
Mike Snitzer | 70d6c40 | 2012-12-21 20:23:37 +0000 | [diff] [blame] | 797 | */ |
Christoph Hellwig | 615ec94 | 2017-04-05 19:21:06 +0200 | [diff] [blame] | 798 | job->rw = REQ_OP_WRITE_ZEROES; |
Mike Snitzer | 70d6c40 | 2012-12-21 20:23:37 +0000 | [diff] [blame] | 799 | for (i = 0; i < job->num_dests; i++) |
Christoph Hellwig | 615ec94 | 2017-04-05 19:21:06 +0200 | [diff] [blame] | 800 | if (!bdev_write_zeroes_sectors(job->dests[i].bdev)) { |
Mike Snitzer | 70d6c40 | 2012-12-21 20:23:37 +0000 | [diff] [blame] | 801 | job->rw = WRITE; |
| 802 | break; |
| 803 | } |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 804 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 805 | |
| 806 | job->fn = fn; |
| 807 | job->context = context; |
Mikulas Patocka | c6ea41f | 2011-05-29 13:03:00 +0100 | [diff] [blame] | 808 | job->master_job = job; |
Damien Le Moal | b73c67c | 2017-05-08 16:40:51 -0700 | [diff] [blame] | 809 | job->write_offset = 0; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 810 | |
Mikulas Patocka | a705a34 | 2011-05-29 13:02:58 +0100 | [diff] [blame] | 811 | if (job->source.count <= SUB_JOB_SIZE) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 812 | dispatch_job(job); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 813 | else { |
Matthias Kaehlcke | def5b5b | 2007-10-19 22:38:52 +0100 | [diff] [blame] | 814 | mutex_init(&job->lock); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 815 | job->progress = 0; |
| 816 | split_job(job); |
| 817 | } |
| 818 | |
| 819 | return 0; |
| 820 | } |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 821 | EXPORT_SYMBOL(dm_kcopyd_copy); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 822 | |
Mikulas Patocka | 7f06965 | 2011-10-31 20:18:58 +0000 | [diff] [blame] | 823 | int dm_kcopyd_zero(struct dm_kcopyd_client *kc, |
| 824 | unsigned num_dests, struct dm_io_region *dests, |
| 825 | unsigned flags, dm_kcopyd_notify_fn fn, void *context) |
| 826 | { |
| 827 | return dm_kcopyd_copy(kc, NULL, num_dests, dests, flags, fn, context); |
| 828 | } |
| 829 | EXPORT_SYMBOL(dm_kcopyd_zero); |
| 830 | |
Mikulas Patocka | a6e50b4 | 2011-08-02 12:32:04 +0100 | [diff] [blame] | 831 | void *dm_kcopyd_prepare_callback(struct dm_kcopyd_client *kc, |
| 832 | dm_kcopyd_notify_fn fn, void *context) |
| 833 | { |
| 834 | struct kcopyd_job *job; |
| 835 | |
| 836 | job = mempool_alloc(kc->job_pool, GFP_NOIO); |
| 837 | |
| 838 | memset(job, 0, sizeof(struct kcopyd_job)); |
| 839 | job->kc = kc; |
| 840 | job->fn = fn; |
| 841 | job->context = context; |
Alasdair G Kergon | d136f2e | 2011-10-23 20:55:17 +0100 | [diff] [blame] | 842 | job->master_job = job; |
Mikulas Patocka | a6e50b4 | 2011-08-02 12:32:04 +0100 | [diff] [blame] | 843 | |
| 844 | atomic_inc(&kc->nr_jobs); |
| 845 | |
| 846 | return job; |
| 847 | } |
| 848 | EXPORT_SYMBOL(dm_kcopyd_prepare_callback); |
| 849 | |
| 850 | void dm_kcopyd_do_callback(void *j, int read_err, unsigned long write_err) |
| 851 | { |
| 852 | struct kcopyd_job *job = j; |
| 853 | struct dm_kcopyd_client *kc = job->kc; |
| 854 | |
| 855 | job->read_err = read_err; |
| 856 | job->write_err = write_err; |
| 857 | |
| 858 | push(&kc->complete_jobs, job); |
| 859 | wake(kc); |
| 860 | } |
| 861 | EXPORT_SYMBOL(dm_kcopyd_do_callback); |
| 862 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 863 | /* |
| 864 | * Cancels a kcopyd job, eg. someone might be deactivating a |
| 865 | * mirror. |
| 866 | */ |
Adrian Bunk | 0b56306 | 2006-01-06 00:20:08 -0800 | [diff] [blame] | 867 | #if 0 |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 868 | int kcopyd_cancel(struct kcopyd_job *job, int block) |
| 869 | { |
| 870 | /* FIXME: finish */ |
| 871 | return -1; |
| 872 | } |
Adrian Bunk | 0b56306 | 2006-01-06 00:20:08 -0800 | [diff] [blame] | 873 | #endif /* 0 */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 874 | |
| 875 | /*----------------------------------------------------------------- |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 876 | * Client setup |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 877 | *---------------------------------------------------------------*/ |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 878 | struct dm_kcopyd_client *dm_kcopyd_client_create(struct dm_kcopyd_throttle *throttle) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 879 | { |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 880 | int r = -ENOMEM; |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 881 | struct dm_kcopyd_client *kc; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 882 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 883 | kc = kmalloc(sizeof(*kc), GFP_KERNEL); |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 884 | if (!kc) |
Mikulas Patocka | fa34ce7 | 2011-05-29 13:03:13 +0100 | [diff] [blame] | 885 | return ERR_PTR(-ENOMEM); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 886 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 887 | spin_lock_init(&kc->job_lock); |
| 888 | INIT_LIST_HEAD(&kc->complete_jobs); |
| 889 | INIT_LIST_HEAD(&kc->io_jobs); |
| 890 | INIT_LIST_HEAD(&kc->pages_jobs); |
Mikulas Patocka | df5d2e9 | 2013-03-01 22:45:49 +0000 | [diff] [blame] | 891 | kc->throttle = throttle; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 892 | |
Mikulas Patocka | 08d8757 | 2008-04-24 21:43:46 +0100 | [diff] [blame] | 893 | kc->job_pool = mempool_create_slab_pool(MIN_JOBS, _job_cache); |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 894 | if (!kc->job_pool) |
| 895 | goto bad_slab; |
Mikulas Patocka | 08d8757 | 2008-04-24 21:43:46 +0100 | [diff] [blame] | 896 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 897 | INIT_WORK(&kc->kcopyd_work, do_work); |
Tejun Heo | 670368a | 2013-07-30 08:40:21 -0400 | [diff] [blame] | 898 | kc->kcopyd_wq = alloc_workqueue("kcopyd", WQ_MEM_RECLAIM, 0); |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 899 | if (!kc->kcopyd_wq) |
| 900 | goto bad_workqueue; |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 901 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 902 | kc->pages = NULL; |
Mikulas Patocka | d047145 | 2011-05-29 13:03:07 +0100 | [diff] [blame] | 903 | kc->nr_reserved_pages = kc->nr_free_pages = 0; |
Mikulas Patocka | 5f43ba2 | 2011-05-29 13:03:11 +0100 | [diff] [blame] | 904 | r = client_reserve_pages(kc, RESERVE_PAGES); |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 905 | if (r) |
| 906 | goto bad_client_pages; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 907 | |
Mikulas Patocka | bda8efe | 2011-05-29 13:03:09 +0100 | [diff] [blame] | 908 | kc->io_client = dm_io_client_create(); |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 909 | if (IS_ERR(kc->io_client)) { |
| 910 | r = PTR_ERR(kc->io_client); |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 911 | goto bad_io_client; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 912 | } |
| 913 | |
Alasdair G Kergon | 138728dc | 2006-03-27 01:17:50 -0800 | [diff] [blame] | 914 | init_waitqueue_head(&kc->destroyq); |
| 915 | atomic_set(&kc->nr_jobs, 0); |
| 916 | |
Mikulas Patocka | fa34ce7 | 2011-05-29 13:03:13 +0100 | [diff] [blame] | 917 | return kc; |
Mikulas Patocka | 945fa4d | 2008-04-24 21:43:49 +0100 | [diff] [blame] | 918 | |
| 919 | bad_io_client: |
| 920 | client_free_pages(kc); |
| 921 | bad_client_pages: |
| 922 | destroy_workqueue(kc->kcopyd_wq); |
| 923 | bad_workqueue: |
| 924 | mempool_destroy(kc->job_pool); |
| 925 | bad_slab: |
| 926 | kfree(kc); |
| 927 | |
Mikulas Patocka | fa34ce7 | 2011-05-29 13:03:13 +0100 | [diff] [blame] | 928 | return ERR_PTR(r); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 929 | } |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 930 | EXPORT_SYMBOL(dm_kcopyd_client_create); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 931 | |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 932 | void dm_kcopyd_client_destroy(struct dm_kcopyd_client *kc) |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 933 | { |
Alasdair G Kergon | 138728dc | 2006-03-27 01:17:50 -0800 | [diff] [blame] | 934 | /* Wait for completion of all jobs submitted by this client. */ |
| 935 | wait_event(kc->destroyq, !atomic_read(&kc->nr_jobs)); |
| 936 | |
Mikulas Patocka | 8c0cbc2 | 2008-04-24 21:43:44 +0100 | [diff] [blame] | 937 | BUG_ON(!list_empty(&kc->complete_jobs)); |
| 938 | BUG_ON(!list_empty(&kc->io_jobs)); |
| 939 | BUG_ON(!list_empty(&kc->pages_jobs)); |
| 940 | destroy_workqueue(kc->kcopyd_wq); |
Milan Broz | 373a392 | 2007-05-09 02:33:02 -0700 | [diff] [blame] | 941 | dm_io_client_destroy(kc->io_client); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 942 | client_free_pages(kc); |
Mikulas Patocka | 08d8757 | 2008-04-24 21:43:46 +0100 | [diff] [blame] | 943 | mempool_destroy(kc->job_pool); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 944 | kfree(kc); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 945 | } |
Heinz Mauelshagen | eb69aca | 2008-04-24 21:43:19 +0100 | [diff] [blame] | 946 | EXPORT_SYMBOL(dm_kcopyd_client_destroy); |