-b9bd69421f7364ca4ff11c827fd0e171a8b826ea
+d464ec667b2b9de097e39d1505b45aafd87a9552
return ret;
}
-static inline u64 swab40(u64 x)
-{
- return (((x & 0x00000000ffULL) << 32)|
- ((x & 0x000000ff00ULL) << 16)|
- ((x & 0x0000ff0000ULL) >> 0)|
- ((x & 0x00ff000000ULL) >> 16)|
- ((x & 0xff00000000ULL) >> 32));
-}
-
void bch2_alloc_v4_swab(struct bkey_s k)
{
struct bch_alloc_v4 *a = bkey_s_to_alloc_v4(k).v;
a->io_time[1] = swab64(a->io_time[1]);
a->stripe = swab32(a->stripe);
a->nr_external_backpointers = swab32(a->nr_external_backpointers);
+ a->fragmentation_lru = swab64(a->fragmentation_lru);
bps = alloc_v4_backpointers(a);
for (bp = bps; bp < bps + BCH_ALLOC_V4_NR_BACKPOINTERS(a); bp++) {
{
struct bkey_s_backpointer bp = bkey_s_to_backpointer(k);
- bp.v->bucket_offset = swab32(bp.v->bucket_offset);
+ bp.v->bucket_offset = swab40(bp.v->bucket_offset);
bp.v->bucket_len = swab32(bp.v->bucket_len);
bch2_bpos_swab(&bp.v->pos);
}
#include "buckets.h"
#include "super.h"
+static inline u64 swab40(u64 x)
+{
+ return (((x & 0x00000000ffULL) << 32)|
+ ((x & 0x000000ff00ULL) << 16)|
+ ((x & 0x0000ff0000ULL) >> 0)|
+ ((x & 0x00ff000000ULL) >> 16)|
+ ((x & 0xff00000000ULL) >> 32));
+}
+
int bch2_backpointer_invalid(struct bch_fs *, struct bkey_s_c k,
enum bkey_invalid_flags, struct printbuf *);
void bch2_backpointer_to_text(struct printbuf *, const struct bch_backpointer *);
#define bkey_lr_packed(_l, _r) \
((_l)->format + ((_r)->format << 1))
-#define bkey_copy(_dst, _src) \
-do { \
- BUILD_BUG_ON(!type_is(_dst, struct bkey_i *) && \
- !type_is(_dst, struct bkey_packed *)); \
- BUILD_BUG_ON(!type_is(_src, struct bkey_i *) && \
- !type_is(_src, struct bkey_packed *)); \
- EBUG_ON((u64 *) (_dst) > (u64 *) (_src) && \
- (u64 *) (_dst) < (u64 *) (_src) + \
- ((struct bkey *) (_src))->u64s); \
- \
- memcpy_u64s_small((_dst), (_src), \
- ((struct bkey *) (_src))->u64s); \
-} while (0)
+static inline void bkey_p_copy(struct bkey_packed *dst, const struct bkey_packed *src)
+{
+ memcpy_u64s_small(dst, src, src->u64s);
+}
+
+static inline void bkey_copy(struct bkey_i *dst, const struct bkey_i *src)
+{
+ memcpy_u64s_small(dst, src, src->k.u64s);
+}
struct btree;
while ((k = sort_iter_peek(iter))) {
if (!bkey_deleted(k) &&
!should_drop_next_key(iter)) {
- bkey_copy(out, k);
+ bkey_p_copy(out, k);
btree_keys_account_key_add(&nr, 0, out);
out = bkey_p_next(out);
}
continue;
if (!transform)
- bkey_copy(out, in);
+ bkey_p_copy(out, in);
else if (bch2_bkey_transform(out_f, out, bkey_packed(in)
? in_f : &bch2_bkey_format_current, in))
out->format = KEY_FORMAT_LOCAL_BTREE;
memcpy_u64s_small(out, in, bkeyp_key_u64s(f, in));
set_bkeyp_val_u64s(f, out, 0);
} else {
- bkey_copy(out, in);
+ bkey_p_copy(out, in);
}
out->needs_whiteout |= needs_whiteout;
out = bkey_p_next(out);
k = new_whiteouts;
while (ptrs != ptrs_end) {
- bkey_copy(k, *ptrs);
+ bkey_p_copy(k, *ptrs);
k = bkey_p_next(k);
ptrs++;
}
n = bkey_p_next(k);
if (!bkey_deleted(k)) {
- bkey_copy(out, k);
+ bkey_p_copy(out, k);
out = bkey_p_next(out);
} else {
BUG_ON(k->needs_whiteout);
k.needs_whiteout = true;
b->whiteout_u64s += k.u64s;
- bkey_copy(unwritten_whiteouts_start(c, b), &k);
+ bkey_p_copy(unwritten_whiteouts_start(c, b), &k);
}
/*
#define _BCACHEFS_ERRCODE_H
#define BCH_ERRCODES() \
+ x(ERANGE, ERANGE_option_too_small) \
+ x(ERANGE, ERANGE_option_too_big) \
x(ENOMEM, ENOMEM_stripe_buf) \
x(ENOMEM, ENOMEM_replicas_table) \
x(ENOMEM, ENOMEM_cpu_replicas) \
return ret;
}
+/*
+ * Determine when a writepage io is full. We have to limit writepage bios to a
+ * single page per bvec (i.e. 1MB with 4k pages) because that is the limit to
+ * what the bounce path in bch2_write_extent() can handle. In theory we could
+ * loosen this restriction for non-bounce I/O, but we don't have that context
+ * here. Ideally, we can up this limit and make it configurable in the future
+ * when the bounce path can be enhanced to accommodate larger source bios.
+ */
+static inline bool bch_io_full(struct bch_writepage_io *io, unsigned len)
+{
+ struct bio *bio = &io->op.wbio.bio;
+ return bio_full(bio, len) ||
+ (bio->bi_iter.bi_size + len > BIO_MAX_VECS * PAGE_SIZE);
+}
+
static void bch2_writepage_io_done(struct bch_write_op *op)
{
struct bch_writepage_io *io =
if (w->io &&
(w->io->op.res.nr_replicas != nr_replicas_this_write ||
- bio_full(&w->io->op.wbio.bio, sectors << 9) ||
- w->io->op.wbio.bio.bi_iter.bi_size + (sectors << 9) >=
- (BIO_MAX_VECS * PAGE_SIZE) ||
+ bch_io_full(w->io, sectors << 9) ||
bio_end_sector(&w->io->op.wbio.bio) != sector))
bch2_writepage_do_io(w);
.inum = inode->ei_inode.bi_dir,
};
- if (!parent_inum.inum)
- return NULL;
-
return d_obtain_alias(bch2_vfs_inode_get(c, parent_inum));
}
return ret;
}
-int bch2_inode_write(struct btree_trans *trans,
+int bch2_inode_write_flags(struct btree_trans *trans,
struct btree_iter *iter,
- struct bch_inode_unpacked *inode)
+ struct bch_inode_unpacked *inode,
+ enum btree_update_flags flags)
{
struct bkey_inode_buf *inode_p;
bch2_inode_pack_inlined(inode_p, inode);
inode_p->inode.k.p.snapshot = iter->snapshot;
- return bch2_trans_update(trans, iter, &inode_p->inode.k_i, 0);
+ return bch2_trans_update(trans, iter, &inode_p->inode.k_i, flags);
}
struct bkey_i *bch2_inode_to_v3(struct btree_trans *trans, struct bkey_i *k)
return ret ?: -BCH_ERR_transaction_restart_nested;
}
-static int may_delete_deleted_inode(struct btree_trans *trans, struct bpos pos)
+static int may_delete_deleted_inode(struct btree_trans *trans,
+ struct btree_iter *iter,
+ struct bpos pos,
+ bool *need_another_pass)
{
struct bch_fs *c = trans->c;
- struct btree_iter iter;
+ struct btree_iter inode_iter;
struct bkey_s_c k;
struct bch_inode_unpacked inode;
int ret;
- if (bch2_snapshot_is_internal_node(c, pos.snapshot))
- return 0;
-
- if (!fsck_err_on(c->sb.clean, c,
- deleted_inode_but_clean,
- "filesystem marked as clean but have deleted inode %llu:%u",
- pos.offset, pos.snapshot))
- return 0;
-
- k = bch2_bkey_get_iter(trans, &iter, BTREE_ID_inodes, pos, BTREE_ITER_CACHED);
+ k = bch2_bkey_get_iter(trans, &inode_iter, BTREE_ID_inodes, pos, BTREE_ITER_CACHED);
ret = bkey_err(k);
if (ret)
return ret;
pos.offset, pos.snapshot))
goto delete;
+ if (c->sb.clean &&
+ !fsck_err(c,
+ deleted_inode_but_clean,
+ "filesystem marked as clean but have deleted inode %llu:%u",
+ pos.offset, pos.snapshot))
+ return 0;
+
+ if (bch2_snapshot_is_internal_node(c, pos.snapshot)) {
+ struct bpos new_min_pos;
+
+ ret = bch2_propagate_key_to_snapshot_leaves(trans, inode_iter.btree_id, k, &new_min_pos);
+ if (ret)
+ goto err;
+
+ inode.bi_flags &= ~BCH_INODE_unlinked;
+
+ ret = bch2_inode_write_flags(trans, &inode_iter, &inode,
+ BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE);
+ bch_err_msg(c, ret, "clearing inode unlinked flag");
+ if (ret)
+ return ret;
+
+ /*
+ * We'll need another write buffer flush to pick up the new
+ * unlinked inodes in the snapshot leaves:
+ */
+ *need_another_pass = true;
+ return 0;
+ }
+
return 1;
err:
fsck_err:
struct btree_trans *trans = bch2_trans_get(c);
struct btree_iter iter;
struct bkey_s_c k;
+ bool need_another_pass;
int ret;
+again:
+ need_another_pass = false;
ret = bch2_btree_write_buffer_flush_sync(trans);
if (ret)
*/
for_each_btree_key(trans, iter, BTREE_ID_deleted_inodes, POS_MIN,
BTREE_ITER_PREFETCH|BTREE_ITER_ALL_SNAPSHOTS, k, ret) {
- ret = lockrestart_do(trans, may_delete_deleted_inode(trans, k.k->p));
+ ret = lockrestart_do(trans, may_delete_deleted_inode(trans, &iter, k.k->p,
+ &need_another_pass));
if (ret < 0)
break;
bch2_fs_lazy_rw(c);
}
+ bch_verbose(c, "deleting unlinked inode %llu:%u", k.k->p.offset, k.k->p.snapshot);
+
ret = bch2_inode_rm_snapshot(trans, k.k->p.offset, k.k->p.snapshot);
if (ret && !bch2_err_matches(ret, BCH_ERR_transaction_restart))
break;
}
}
bch2_trans_iter_exit(trans, &iter);
+
+ if (!ret && need_another_pass)
+ goto again;
err:
bch2_trans_put(trans);
#define _BCACHEFS_INODE_H
#include "bkey.h"
+#include "bkey_methods.h"
#include "opts.h"
enum bkey_invalid_flags;
int bch2_inode_peek(struct btree_trans *, struct btree_iter *,
struct bch_inode_unpacked *, subvol_inum, unsigned);
-int bch2_inode_write(struct btree_trans *, struct btree_iter *,
- struct bch_inode_unpacked *);
+
+int bch2_inode_write_flags(struct btree_trans *, struct btree_iter *,
+ struct bch_inode_unpacked *, enum btree_update_flags);
+
+static inline int bch2_inode_write(struct btree_trans *trans,
+ struct btree_iter *iter,
+ struct bch_inode_unpacked *inode)
+{
+ return bch2_inode_write_flags(trans, iter, inode, 0);
+}
void bch2_inode_init_early(struct bch_fs *,
struct bch_inode_unpacked *);
if (err)
prt_printf(err, "%s: too small (min %llu)",
opt->attr.name, opt->min);
- return -ERANGE;
+ return -BCH_ERR_ERANGE_option_too_small;
}
if (opt->max && v >= opt->max) {
if (err)
prt_printf(err, "%s: too big (max %llu)",
opt->attr.name, opt->max);
- return -ERANGE;
+ return -BCH_ERR_ERANGE_option_too_big;
}
if ((opt->flags & OPT_SB_FIELD_SECTORS) && (v & 511)) {
if (*i) {
last_journal_entry = &(*i)->j;
(*i)->ignore = false;
+ /*
+ * This was probably a NO_FLUSH entry,
+ * so last_seq was garbage - but we know
+ * we're only using a single journal
+ * entry, set it here:
+ */
+ (*i)->j.last_seq = (*i)->j.seq;
break;
}
}
if (opt_defined(*opts, sb))
goto err;
- printk(KERN_ERR "bcachefs (%s): error reading default superblock: %s",
+ printk(KERN_ERR "bcachefs (%s): error reading default superblock: %s\n",
path, err.buf);
printbuf_reset(&err);
ret = bch2_sb_validate(sb, &err, READ);
if (ret) {
- printk(KERN_ERR "bcachefs (%s): error validating superblock: %s",
+ printk(KERN_ERR "bcachefs (%s): error validating superblock: %s\n",
path, err.buf);
goto err_no_print;
}
printbuf_exit(&err);
return ret;
err:
- printk(KERN_ERR "bcachefs (%s): error reading superblock: %s",
+ printk(KERN_ERR "bcachefs (%s): error reading superblock: %s\n",
path, err.buf);
err_no_print:
bch2_free_super(sb);