git.sesse.net Git - bcachefs-tools-debian/blob - libbcachefs/btree_write_buffer.c

   1 // SPDX-License-Identifier: GPL-2.0
   2
   3 #include "bcachefs.h"
   4 #include "btree_locking.h"
   5 #include "btree_update.h"
   6 #include "btree_update_interior.h"
   7 #include "btree_write_buffer.h"
   8 #include "error.h"
   9 #include "journal.h"
  10 #include "journal_reclaim.h"
  11
  12 #include <linux/sort.h>
  13
  14 static int btree_write_buffered_key_cmp(const void *_l, const void *_r)
  15 {
  16         const struct btree_write_buffered_key *l = _l;
  17         const struct btree_write_buffered_key *r = _r;
  18
  19         return  cmp_int(l->btree, r->btree) ?:
  20                 bpos_cmp(l->k.k.p, r->k.k.p) ?:
  21                 cmp_int(l->journal_seq, r->journal_seq) ?:
  22                 cmp_int(l->journal_offset, r->journal_offset);
  23 }
  24
  25 static int btree_write_buffered_journal_cmp(const void *_l, const void *_r)
  26 {
  27         const struct btree_write_buffered_key *l = _l;
  28         const struct btree_write_buffered_key *r = _r;
  29
  30         return  cmp_int(l->journal_seq, r->journal_seq);
  31 }
  32
  33 static int bch2_btree_write_buffer_flush_one(struct btree_trans *trans,
  34                                              struct btree_iter *iter,
  35                                              struct btree_write_buffered_key *wb,
  36                                              unsigned commit_flags,
  37                                              bool *write_locked,
  38                                              size_t *fast)
  39 {
  40         struct bch_fs *c = trans->c;
  41         struct btree_path *path;
  42         int ret;
  43
  44         ret = bch2_btree_iter_traverse(iter);
  45         if (ret)
  46                 return ret;
  47
  48         path = iter->path;
  49
  50         if (!*write_locked) {
  51                 ret = bch2_btree_node_lock_write(trans, path, &path->l[0].b->c);
  52                 if (ret)
  53                         return ret;
  54
  55                 bch2_btree_node_prep_for_write(trans, path, path->l[0].b);
  56                 *write_locked = true;
  57         }
  58
  59         if (!bch2_btree_node_insert_fits(c, path->l[0].b, wb->k.k.u64s)) {
  60                 bch2_btree_node_unlock_write(trans, path, path->l[0].b);
  61                 *write_locked = false;
  62                 goto trans_commit;
  63         }
  64
  65         bch2_btree_insert_key_leaf(trans, path, &wb->k, wb->journal_seq);
  66         (*fast)++;
  67
  68         if (path->ref > 1) {
  69                 /*
  70                  * We can't clone a path that has write locks: if the path is
  71                  * shared, unlock before set_pos(), traverse():
  72                  */
  73                 bch2_btree_node_unlock_write(trans, path, path->l[0].b);
  74                 *write_locked = false;
  75         }
  76         return 0;
  77 trans_commit:
  78         return  bch2_trans_update_seq(trans, wb->journal_seq, iter, &wb->k,
  79                                       BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?:
  80                 bch2_trans_commit(trans, NULL, NULL,
  81                                   commit_flags|
  82                                   BTREE_INSERT_NOCHECK_RW|
  83                                   BTREE_INSERT_NOFAIL|
  84                                   BTREE_INSERT_JOURNAL_RECLAIM);
  85 }
  86
  87 static union btree_write_buffer_state btree_write_buffer_switch(struct btree_write_buffer *wb)
  88 {
  89         union btree_write_buffer_state old, new;
  90         u64 v = READ_ONCE(wb->state.v);
  91
  92         do {
  93                 old.v = new.v = v;
  94
  95                 new.nr = 0;
  96                 new.idx++;
  97         } while ((v = atomic64_cmpxchg_acquire(&wb->state.counter, old.v, new.v)) != old.v);
  98
  99         while (old.idx == 0 ? wb->state.ref0 : wb->state.ref1)
 100                 cpu_relax();
 101
 102         smp_mb();
 103
 104         return old;
 105 }
 106
 107 /*
 108  * Update a btree with a write buffered key using the journal seq of the
 109  * original write buffer insert.
 110  *
 111  * It is not safe to rejournal the key once it has been inserted into the write
 112  * buffer because that may break recovery ordering. For example, the key may
 113  * have already been modified in the active write buffer in a seq that comes
 114  * before the current transaction. If we were to journal this key again and
 115  * crash, recovery would process updates in the wrong order.
 116  */
 117 static int
 118 btree_write_buffered_insert(struct btree_trans *trans,
 119                           struct btree_write_buffered_key *wb)
 120 {
 121         struct btree_iter iter;
 122         int ret;
 123
 124         bch2_trans_iter_init(trans, &iter, wb->btree, bkey_start_pos(&wb->k.k),
 125                              BTREE_ITER_CACHED|BTREE_ITER_INTENT);
 126
 127         ret   = bch2_btree_iter_traverse(&iter) ?:
 128                 bch2_trans_update_seq(trans, wb->journal_seq, &iter, &wb->k,
 129                                       BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
 130         bch2_trans_iter_exit(trans, &iter);
 131         return ret;
 132 }
 133
 134 int __bch2_btree_write_buffer_flush(struct btree_trans *trans, unsigned commit_flags,
 135                                     bool locked)
 136 {
 137         struct bch_fs *c = trans->c;
 138         struct journal *j = &c->journal;
 139         struct btree_write_buffer *wb = &c->btree_write_buffer;
 140         struct journal_entry_pin pin;
 141         struct btree_write_buffered_key *i, *keys;
 142         struct btree_iter iter = { NULL };
 143         size_t nr = 0, skipped = 0, fast = 0, slowpath = 0;
 144         bool write_locked = false;
 145         union btree_write_buffer_state s;
 146         int ret = 0;
 147
 148         memset(&pin, 0, sizeof(pin));
 149
 150         if (!locked && !mutex_trylock(&wb->flush_lock))
 151                 return 0;
 152
 153         bch2_journal_pin_copy(j, &pin, &wb->journal_pin, NULL);
 154         bch2_journal_pin_drop(j, &wb->journal_pin);
 155
 156         s = btree_write_buffer_switch(wb);
 157         keys = wb->keys[s.idx];
 158         nr = s.nr;
 159
 160         if (race_fault())
 161                 goto slowpath;
 162
 163         /*
 164          * We first sort so that we can detect and skip redundant updates, and
 165          * then we attempt to flush in sorted btree order, as this is most
 166          * efficient.
 167          *
 168          * However, since we're not flushing in the order they appear in the
 169          * journal we won't be able to drop our journal pin until everything is
 170          * flushed - which means this could deadlock the journal if we weren't
 171          * passing BTREE_INSERT_JOURNAL_RECLAIM. This causes the update to fail
 172          * if it would block taking a journal reservation.
 173          *
 174          * If that happens, simply skip the key so we can optimistically insert
 175          * as many keys as possible in the fast path.
 176          */
 177         sort(keys, nr, sizeof(keys[0]),
 178              btree_write_buffered_key_cmp, NULL);
 179
 180         for (i = keys; i < keys + nr; i++) {
 181                 if (i + 1 < keys + nr &&
 182                     i[0].btree == i[1].btree &&
 183                     bpos_eq(i[0].k.k.p, i[1].k.k.p)) {
 184                         skipped++;
 185                         i->journal_seq = 0;
 186                         continue;
 187                 }
 188
 189                 if (write_locked &&
 190                     (iter.path->btree_id != i->btree ||
 191                      bpos_gt(i->k.k.p, iter.path->l[0].b->key.k.p))) {
 192                         bch2_btree_node_unlock_write(trans, iter.path, iter.path->l[0].b);
 193                         write_locked = false;
 194                 }
 195
 196                 if (!iter.path || iter.path->btree_id != i->btree) {
 197                         bch2_trans_iter_exit(trans, &iter);
 198                         bch2_trans_iter_init(trans, &iter, i->btree, i->k.k.p,
 199                                              BTREE_ITER_INTENT|BTREE_ITER_ALL_SNAPSHOTS);
 200                 }
 201
 202                 bch2_btree_iter_set_pos(&iter, i->k.k.p);
 203                 iter.path->preserve = false;
 204
 205                 do {
 206                         ret = bch2_btree_write_buffer_flush_one(trans, &iter, i,
 207                                                 commit_flags, &write_locked, &fast);
 208                         if (!write_locked)
 209                                 bch2_trans_begin(trans);
 210                 } while (bch2_err_matches(ret, BCH_ERR_transaction_restart));
 211
 212                 if (ret == -BCH_ERR_journal_reclaim_would_deadlock) {
 213                         slowpath++;
 214                         continue;
 215                 }
 216                 if (ret)
 217                         break;
 218
 219                 i->journal_seq = 0;
 220         }
 221
 222         if (write_locked)
 223                 bch2_btree_node_unlock_write(trans, iter.path, iter.path->l[0].b);
 224         bch2_trans_iter_exit(trans, &iter);
 225
 226         trace_write_buffer_flush(trans, nr, skipped, fast, wb->size);
 227
 228         if (slowpath)
 229                 goto slowpath;
 230
 231         bch2_fs_fatal_err_on(ret, c, "%s: insert error %s", __func__, bch2_err_str(ret));
 232 out:
 233         bch2_journal_pin_drop(j, &pin);
 234         mutex_unlock(&wb->flush_lock);
 235         return ret;
 236 slowpath:
 237         trace_write_buffer_flush_slowpath(trans, i - keys, nr);
 238
 239         /*
 240          * Now sort the rest by journal seq and bump the journal pin as we go.
 241          * The slowpath zapped the seq of keys that were successfully flushed so
 242          * we can skip those here.
 243          */
 244         sort(keys, nr, sizeof(keys[0]),
 245              btree_write_buffered_journal_cmp,
 246              NULL);
 247
 248         commit_flags &= ~BCH_WATERMARK_MASK;
 249         commit_flags |= BCH_WATERMARK_reclaim;
 250
 251         for (i = keys; i < keys + nr; i++) {
 252                 if (!i->journal_seq)
 253                         continue;
 254
 255                 if (i->journal_seq > pin.seq) {
 256                         struct journal_entry_pin pin2;
 257
 258                         memset(&pin2, 0, sizeof(pin2));
 259
 260                         bch2_journal_pin_add(j, i->journal_seq, &pin2, NULL);
 261                         bch2_journal_pin_drop(j, &pin);
 262                         bch2_journal_pin_copy(j, &pin, &pin2, NULL);
 263                         bch2_journal_pin_drop(j, &pin2);
 264                 }
 265
 266                 ret = commit_do(trans, NULL, NULL,
 267                                 commit_flags|
 268                                 BTREE_INSERT_NOFAIL|
 269                                 BTREE_INSERT_JOURNAL_RECLAIM,
 270                                 btree_write_buffered_insert(trans, i));
 271                 if (bch2_fs_fatal_err_on(ret, c, "%s: insert error %s", __func__, bch2_err_str(ret)))
 272                         break;
 273         }
 274
 275         goto out;
 276 }
 277
 278 int bch2_btree_write_buffer_flush_sync(struct btree_trans *trans)
 279 {
 280         bch2_trans_unlock(trans);
 281         mutex_lock(&trans->c->btree_write_buffer.flush_lock);
 282         return __bch2_btree_write_buffer_flush(trans, 0, true);
 283 }
 284
 285 int bch2_btree_write_buffer_flush(struct btree_trans *trans)
 286 {
 287         return __bch2_btree_write_buffer_flush(trans, 0, false);
 288 }
 289
 290 static int bch2_btree_write_buffer_journal_flush(struct journal *j,
 291                                 struct journal_entry_pin *_pin, u64 seq)
 292 {
 293         struct bch_fs *c = container_of(j, struct bch_fs, journal);
 294         struct btree_write_buffer *wb = &c->btree_write_buffer;
 295
 296         mutex_lock(&wb->flush_lock);
 297
 298         return bch2_trans_run(c,
 299                         __bch2_btree_write_buffer_flush(trans, BTREE_INSERT_NOCHECK_RW, true));
 300 }
 301
 302 static inline u64 btree_write_buffer_ref(int idx)
 303 {
 304         return ((union btree_write_buffer_state) {
 305                 .ref0 = idx == 0,
 306                 .ref1 = idx == 1,
 307         }).v;
 308 }
 309
 310 int bch2_btree_insert_keys_write_buffer(struct btree_trans *trans)
 311 {
 312         struct bch_fs *c = trans->c;
 313         struct btree_write_buffer *wb = &c->btree_write_buffer;
 314         struct btree_write_buffered_key *i;
 315         union btree_write_buffer_state old, new;
 316         int ret = 0;
 317         u64 v;
 318
 319         trans_for_each_wb_update(trans, i) {
 320                 EBUG_ON(i->k.k.u64s > BTREE_WRITE_BUFERED_U64s_MAX);
 321
 322                 i->journal_seq          = trans->journal_res.seq;
 323                 i->journal_offset       = trans->journal_res.offset;
 324         }
 325
 326         preempt_disable();
 327         v = READ_ONCE(wb->state.v);
 328         do {
 329                 old.v = new.v = v;
 330
 331                 new.v += btree_write_buffer_ref(new.idx);
 332                 new.nr += trans->nr_wb_updates;
 333                 if (new.nr > wb->size) {
 334                         ret = -BCH_ERR_btree_insert_need_flush_buffer;
 335                         goto out;
 336                 }
 337         } while ((v = atomic64_cmpxchg_acquire(&wb->state.counter, old.v, new.v)) != old.v);
 338
 339         memcpy(wb->keys[new.idx] + old.nr,
 340                trans->wb_updates,
 341                sizeof(trans->wb_updates[0]) * trans->nr_wb_updates);
 342
 343         bch2_journal_pin_add(&c->journal, trans->journal_res.seq, &wb->journal_pin,
 344                              bch2_btree_write_buffer_journal_flush);
 345
 346         atomic64_sub_return_release(btree_write_buffer_ref(new.idx), &wb->state.counter);
 347 out:
 348         preempt_enable();
 349         return ret;
 350 }
 351
 352 void bch2_fs_btree_write_buffer_exit(struct bch_fs *c)
 353 {
 354         struct btree_write_buffer *wb = &c->btree_write_buffer;
 355
 356         BUG_ON(wb->state.nr && !bch2_journal_error(&c->journal));
 357
 358         kvfree(wb->keys[1]);
 359         kvfree(wb->keys[0]);
 360 }
 361
 362 int bch2_fs_btree_write_buffer_init(struct bch_fs *c)
 363 {
 364         struct btree_write_buffer *wb = &c->btree_write_buffer;
 365
 366         mutex_init(&wb->flush_lock);
 367         wb->size = c->opts.btree_write_buffer_size;
 368
 369         wb->keys[0] = kvmalloc_array(wb->size, sizeof(*wb->keys[0]), GFP_KERNEL);
 370         wb->keys[1] = kvmalloc_array(wb->size, sizeof(*wb->keys[1]), GFP_KERNEL);
 371         if (!wb->keys[0] || !wb->keys[1])
 372                 return -BCH_ERR_ENOMEM_fs_btree_write_buffer_init;
 373
 374         return 0;
 375 }