linux/fs/bcachefs/extents.c

1547 lines
40 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2010 Kent Overstreet <kent.overstreet@gmail.com>
*
* Code for managing the extent btree and dynamically updating the writeback
* dirty sector count.
*/
#include "bcachefs.h"
#include "bkey_methods.h"
#include "btree_cache.h"
#include "btree_gc.h"
#include "btree_io.h"
#include "btree_iter.h"
#include "buckets.h"
#include "checksum.h"
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
#include "compress.h"
#include "debug.h"
#include "disk_groups.h"
#include "error.h"
#include "extents.h"
#include "inode.h"
#include "journal.h"
#include "replicas.h"
#include "super.h"
#include "super-io.h"
#include "trace.h"
#include "util.h"
static unsigned bch2_crc_field_size_max[] = {
[BCH_EXTENT_ENTRY_crc32] = CRC32_SIZE_MAX,
[BCH_EXTENT_ENTRY_crc64] = CRC64_SIZE_MAX,
[BCH_EXTENT_ENTRY_crc128] = CRC128_SIZE_MAX,
};
static void bch2_extent_crc_pack(union bch_extent_crc *,
struct bch_extent_crc_unpacked,
enum bch_extent_entry_type);
static struct bch_dev_io_failures *dev_io_failures(struct bch_io_failures *f,
unsigned dev)
{
struct bch_dev_io_failures *i;
for (i = f->devs; i < f->devs + f->nr; i++)
if (i->dev == dev)
return i;
return NULL;
}
void bch2_mark_io_failure(struct bch_io_failures *failed,
struct extent_ptr_decoded *p)
{
struct bch_dev_io_failures *f = dev_io_failures(failed, p->ptr.dev);
if (!f) {
BUG_ON(failed->nr >= ARRAY_SIZE(failed->devs));
f = &failed->devs[failed->nr++];
f->dev = p->ptr.dev;
f->idx = p->idx;
f->nr_failed = 1;
f->nr_retries = 0;
} else if (p->idx != f->idx) {
f->idx = p->idx;
f->nr_failed = 1;
f->nr_retries = 0;
} else {
f->nr_failed++;
}
}
static inline u64 dev_latency(struct bch_fs *c, unsigned dev)
{
struct bch_dev *ca = bch2_dev_rcu(c, dev);
return ca ? atomic64_read(&ca->cur_latency[READ]) : S64_MAX;
}
/*
* returns true if p1 is better than p2:
*/
static inline bool ptr_better(struct bch_fs *c,
const struct extent_ptr_decoded p1,
const struct extent_ptr_decoded p2)
{
if (likely(!p1.idx && !p2.idx)) {
u64 l1 = dev_latency(c, p1.ptr.dev);
u64 l2 = dev_latency(c, p2.ptr.dev);
/* Pick at random, biased in favor of the faster device: */
return bch2_rand_range(l1 + l2) > l1;
}
if (bch2_force_reconstruct_read)
return p1.idx > p2.idx;
return p1.idx < p2.idx;
}
/*
* This picks a non-stale pointer, preferably from a device other than @avoid.
* Avoid can be NULL, meaning pick any. If there are no non-stale pointers to
* other devices, it will still pick a pointer from avoid.
*/
int bch2_bkey_pick_read_device(struct bch_fs *c, struct bkey_s_c k,
struct bch_io_failures *failed,
struct extent_ptr_decoded *pick)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
struct bch_dev_io_failures *f;
int ret = 0;
if (k.k->type == KEY_TYPE_error)
return -EIO;
rcu_read_lock();
bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
/*
* Unwritten extent: no need to actually read, treat it as a
* hole and return 0s:
*/
if (p.ptr.unwritten) {
ret = 0;
break;
}
/*
* If there are any dirty pointers it's an error if we can't
* read:
*/
if (!ret && !p.ptr.cached)
ret = -EIO;
struct bch_dev *ca = bch2_dev_rcu(c, p.ptr.dev);
if (p.ptr.cached && (!ca || dev_ptr_stale_rcu(ca, &p.ptr)))
continue;
f = failed ? dev_io_failures(failed, p.ptr.dev) : NULL;
if (f)
p.idx = f->nr_failed < f->nr_retries
? f->idx
: f->idx + 1;
if (!p.idx && !ca)
p.idx++;
if (!p.idx && p.has_ec && bch2_force_reconstruct_read)
p.idx++;
if (!p.idx && !bch2_dev_is_readable(ca))
p.idx++;
if (p.idx >= (unsigned) p.has_ec + 1)
continue;
if (ret > 0 && !ptr_better(c, p, *pick))
continue;
*pick = p;
ret = 1;
}
rcu_read_unlock();
return ret;
}
/* KEY_TYPE_btree_ptr: */
int bch2_btree_ptr_invalid(struct bch_fs *c, struct bkey_s_c k,
enum bch_validate_flags flags,
struct printbuf *err)
{
int ret = 0;
bkey_fsck_err_on(bkey_val_u64s(k.k) > BCH_REPLICAS_MAX, c, err,
btree_ptr_val_too_big,
"value too big (%zu > %u)", bkey_val_u64s(k.k), BCH_REPLICAS_MAX);
ret = bch2_bkey_ptrs_invalid(c, k, flags, err);
fsck_err:
return ret;
}
void bch2_btree_ptr_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
bch2_bkey_ptrs_to_text(out, c, k);
}
int bch2_btree_ptr_v2_invalid(struct bch_fs *c, struct bkey_s_c k,
enum bch_validate_flags flags,
struct printbuf *err)
{
struct bkey_s_c_btree_ptr_v2 bp = bkey_s_c_to_btree_ptr_v2(k);
int ret = 0;
bkey_fsck_err_on(bkey_val_u64s(k.k) > BKEY_BTREE_PTR_VAL_U64s_MAX,
c, err, btree_ptr_v2_val_too_big,
"value too big (%zu > %zu)",
bkey_val_u64s(k.k), BKEY_BTREE_PTR_VAL_U64s_MAX);
bkey_fsck_err_on(bpos_ge(bp.v->min_key, bp.k->p),
c, err, btree_ptr_v2_min_key_bad,
"min_key > key");
if (flags & BCH_VALIDATE_write)
bkey_fsck_err_on(!bp.v->sectors_written,
c, err, btree_ptr_v2_written_0,
"sectors_written == 0");
ret = bch2_bkey_ptrs_invalid(c, k, flags, err);
fsck_err:
return ret;
}
void bch2_btree_ptr_v2_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
struct bkey_s_c_btree_ptr_v2 bp = bkey_s_c_to_btree_ptr_v2(k);
prt_printf(out, "seq %llx written %u min_key %s",
le64_to_cpu(bp.v->seq),
le16_to_cpu(bp.v->sectors_written),
BTREE_PTR_RANGE_UPDATED(bp.v) ? "R " : "");
bch2_bpos_to_text(out, bp.v->min_key);
prt_printf(out, " ");
bch2_bkey_ptrs_to_text(out, c, k);
}
void bch2_btree_ptr_v2_compat(enum btree_id btree_id, unsigned version,
unsigned big_endian, int write,
struct bkey_s k)
{
struct bkey_s_btree_ptr_v2 bp = bkey_s_to_btree_ptr_v2(k);
compat_bpos(0, btree_id, version, big_endian, write, &bp.v->min_key);
if (version < bcachefs_metadata_version_inode_btree_change &&
btree_id_is_extents(btree_id) &&
!bkey_eq(bp.v->min_key, POS_MIN))
bp.v->min_key = write
? bpos_nosnap_predecessor(bp.v->min_key)
: bpos_nosnap_successor(bp.v->min_key);
}
/* KEY_TYPE_extent: */
bool bch2_extent_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_c r)
{
struct bkey_ptrs l_ptrs = bch2_bkey_ptrs(l);
struct bkey_ptrs_c r_ptrs = bch2_bkey_ptrs_c(r);
union bch_extent_entry *en_l;
const union bch_extent_entry *en_r;
struct extent_ptr_decoded lp, rp;
bool use_right_ptr;
en_l = l_ptrs.start;
en_r = r_ptrs.start;
while (en_l < l_ptrs.end && en_r < r_ptrs.end) {
if (extent_entry_type(en_l) != extent_entry_type(en_r))
return false;
en_l = extent_entry_next(en_l);
en_r = extent_entry_next(en_r);
}
if (en_l < l_ptrs.end || en_r < r_ptrs.end)
return false;
en_l = l_ptrs.start;
en_r = r_ptrs.start;
lp.crc = bch2_extent_crc_unpack(l.k, NULL);
rp.crc = bch2_extent_crc_unpack(r.k, NULL);
while (__bkey_ptr_next_decode(l.k, l_ptrs.end, lp, en_l) &&
__bkey_ptr_next_decode(r.k, r_ptrs.end, rp, en_r)) {
if (lp.ptr.offset + lp.crc.offset + lp.crc.live_size !=
rp.ptr.offset + rp.crc.offset ||
lp.ptr.dev != rp.ptr.dev ||
lp.ptr.gen != rp.ptr.gen ||
lp.ptr.unwritten != rp.ptr.unwritten ||
lp.has_ec != rp.has_ec)
return false;
/* Extents may not straddle buckets: */
rcu_read_lock();
struct bch_dev *ca = bch2_dev_rcu(c, lp.ptr.dev);
bool same_bucket = ca && PTR_BUCKET_NR(ca, &lp.ptr) == PTR_BUCKET_NR(ca, &rp.ptr);
rcu_read_unlock();
if (!same_bucket)
return false;
if (lp.has_ec != rp.has_ec ||
(lp.has_ec &&
(lp.ec.block != rp.ec.block ||
lp.ec.redundancy != rp.ec.redundancy ||
lp.ec.idx != rp.ec.idx)))
return false;
if (lp.crc.compression_type != rp.crc.compression_type ||
lp.crc.nonce != rp.crc.nonce)
return false;
if (lp.crc.offset + lp.crc.live_size + rp.crc.live_size <=
lp.crc.uncompressed_size) {
/* can use left extent's crc entry */
} else if (lp.crc.live_size <= rp.crc.offset) {
/* can use right extent's crc entry */
} else {
/* check if checksums can be merged: */
if (lp.crc.csum_type != rp.crc.csum_type ||
lp.crc.nonce != rp.crc.nonce ||
crc_is_compressed(lp.crc) ||
!bch2_checksum_mergeable(lp.crc.csum_type))
return false;
if (lp.crc.offset + lp.crc.live_size != lp.crc.compressed_size ||
rp.crc.offset)
return false;
if (lp.crc.csum_type &&
lp.crc.uncompressed_size +
rp.crc.uncompressed_size > (c->opts.encoded_extent_max >> 9))
return false;
}
en_l = extent_entry_next(en_l);
en_r = extent_entry_next(en_r);
}
en_l = l_ptrs.start;
en_r = r_ptrs.start;
while (en_l < l_ptrs.end && en_r < r_ptrs.end) {
if (extent_entry_is_crc(en_l)) {
struct bch_extent_crc_unpacked crc_l = bch2_extent_crc_unpack(l.k, entry_to_crc(en_l));
struct bch_extent_crc_unpacked crc_r = bch2_extent_crc_unpack(r.k, entry_to_crc(en_r));
if (crc_l.uncompressed_size + crc_r.uncompressed_size >
bch2_crc_field_size_max[extent_entry_type(en_l)])
return false;
}
en_l = extent_entry_next(en_l);
en_r = extent_entry_next(en_r);
}
use_right_ptr = false;
en_l = l_ptrs.start;
en_r = r_ptrs.start;
while (en_l < l_ptrs.end) {
if (extent_entry_type(en_l) == BCH_EXTENT_ENTRY_ptr &&
use_right_ptr)
en_l->ptr = en_r->ptr;
if (extent_entry_is_crc(en_l)) {
struct bch_extent_crc_unpacked crc_l =
bch2_extent_crc_unpack(l.k, entry_to_crc(en_l));
struct bch_extent_crc_unpacked crc_r =
bch2_extent_crc_unpack(r.k, entry_to_crc(en_r));
use_right_ptr = false;
if (crc_l.offset + crc_l.live_size + crc_r.live_size <=
crc_l.uncompressed_size) {
/* can use left extent's crc entry */
} else if (crc_l.live_size <= crc_r.offset) {
/* can use right extent's crc entry */
crc_r.offset -= crc_l.live_size;
bch2_extent_crc_pack(entry_to_crc(en_l), crc_r,
extent_entry_type(en_l));
use_right_ptr = true;
} else {
crc_l.csum = bch2_checksum_merge(crc_l.csum_type,
crc_l.csum,
crc_r.csum,
crc_r.uncompressed_size << 9);
crc_l.uncompressed_size += crc_r.uncompressed_size;
crc_l.compressed_size += crc_r.compressed_size;
bch2_extent_crc_pack(entry_to_crc(en_l), crc_l,
extent_entry_type(en_l));
}
}
en_l = extent_entry_next(en_l);
en_r = extent_entry_next(en_r);
}
bch2_key_resize(l.k, l.k->size + r.k->size);
return true;
}
/* KEY_TYPE_reservation: */
int bch2_reservation_invalid(struct bch_fs *c, struct bkey_s_c k,
enum bch_validate_flags flags,
struct printbuf *err)
{
struct bkey_s_c_reservation r = bkey_s_c_to_reservation(k);
int ret = 0;
bkey_fsck_err_on(!r.v->nr_replicas || r.v->nr_replicas > BCH_REPLICAS_MAX, c, err,
reservation_key_nr_replicas_invalid,
"invalid nr_replicas (%u)", r.v->nr_replicas);
fsck_err:
return ret;
}
void bch2_reservation_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
struct bkey_s_c_reservation r = bkey_s_c_to_reservation(k);
prt_printf(out, "generation %u replicas %u",
le32_to_cpu(r.v->generation),
r.v->nr_replicas);
}
bool bch2_reservation_merge(struct bch_fs *c, struct bkey_s _l, struct bkey_s_c _r)
{
struct bkey_s_reservation l = bkey_s_to_reservation(_l);
struct bkey_s_c_reservation r = bkey_s_c_to_reservation(_r);
if (l.v->generation != r.v->generation ||
l.v->nr_replicas != r.v->nr_replicas)
return false;
bch2_key_resize(l.k, l.k->size + r.k->size);
return true;
}
/* Extent checksum entries: */
/* returns true if not equal */
static inline bool bch2_crc_unpacked_cmp(struct bch_extent_crc_unpacked l,
struct bch_extent_crc_unpacked r)
{
return (l.csum_type != r.csum_type ||
l.compression_type != r.compression_type ||
l.compressed_size != r.compressed_size ||
l.uncompressed_size != r.uncompressed_size ||
l.offset != r.offset ||
l.live_size != r.live_size ||
l.nonce != r.nonce ||
bch2_crc_cmp(l.csum, r.csum));
}
static inline bool can_narrow_crc(struct bch_extent_crc_unpacked u,
struct bch_extent_crc_unpacked n)
{
return !crc_is_compressed(u) &&
u.csum_type &&
u.uncompressed_size > u.live_size &&
bch2_csum_type_is_encryption(u.csum_type) ==
bch2_csum_type_is_encryption(n.csum_type);
}
bool bch2_can_narrow_extent_crcs(struct bkey_s_c k,
struct bch_extent_crc_unpacked n)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
struct bch_extent_crc_unpacked crc;
const union bch_extent_entry *i;
if (!n.csum_type)
return false;
bkey_for_each_crc(k.k, ptrs, crc, i)
if (can_narrow_crc(crc, n))
return true;
return false;
}
/*
* We're writing another replica for this extent, so while we've got the data in
* memory we'll be computing a new checksum for the currently live data.
*
* If there are other replicas we aren't moving, and they are checksummed but
* not compressed, we can modify them to point to only the data that is
* currently live (so that readers won't have to bounce) while we've got the
* checksum we need:
*/
bool bch2_bkey_narrow_crcs(struct bkey_i *k, struct bch_extent_crc_unpacked n)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(bkey_i_to_s(k));
struct bch_extent_crc_unpacked u;
struct extent_ptr_decoded p;
union bch_extent_entry *i;
bool ret = false;
/* Find a checksum entry that covers only live data: */
if (!n.csum_type) {
bkey_for_each_crc(&k->k, ptrs, u, i)
if (!crc_is_compressed(u) &&
u.csum_type &&
u.live_size == u.uncompressed_size) {
n = u;
goto found;
}
return false;
}
found:
BUG_ON(crc_is_compressed(n));
BUG_ON(n.offset);
BUG_ON(n.live_size != k->k.size);
restart_narrow_pointers:
ptrs = bch2_bkey_ptrs(bkey_i_to_s(k));
bkey_for_each_ptr_decode(&k->k, ptrs, p, i)
if (can_narrow_crc(p.crc, n)) {
bch2_bkey_drop_ptr_noerror(bkey_i_to_s(k), &i->ptr);
p.ptr.offset += p.crc.offset;
p.crc = n;
bch2_extent_ptr_decoded_append(k, &p);
ret = true;
goto restart_narrow_pointers;
}
return ret;
}
static void bch2_extent_crc_pack(union bch_extent_crc *dst,
struct bch_extent_crc_unpacked src,
enum bch_extent_entry_type type)
{
#define set_common_fields(_dst, _src) \
_dst.type = 1 << type; \
_dst.csum_type = _src.csum_type, \
_dst.compression_type = _src.compression_type, \
_dst._compressed_size = _src.compressed_size - 1, \
_dst._uncompressed_size = _src.uncompressed_size - 1, \
_dst.offset = _src.offset
switch (type) {
case BCH_EXTENT_ENTRY_crc32:
set_common_fields(dst->crc32, src);
dst->crc32.csum = (u32 __force) *((__le32 *) &src.csum.lo);
break;
case BCH_EXTENT_ENTRY_crc64:
set_common_fields(dst->crc64, src);
dst->crc64.nonce = src.nonce;
dst->crc64.csum_lo = (u64 __force) src.csum.lo;
dst->crc64.csum_hi = (u64 __force) *((__le16 *) &src.csum.hi);
break;
case BCH_EXTENT_ENTRY_crc128:
set_common_fields(dst->crc128, src);
dst->crc128.nonce = src.nonce;
dst->crc128.csum = src.csum;
break;
default:
BUG();
}
#undef set_common_fields
}
void bch2_extent_crc_append(struct bkey_i *k,
struct bch_extent_crc_unpacked new)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(bkey_i_to_s(k));
union bch_extent_crc *crc = (void *) ptrs.end;
enum bch_extent_entry_type type;
if (bch_crc_bytes[new.csum_type] <= 4 &&
new.uncompressed_size <= CRC32_SIZE_MAX &&
new.nonce <= CRC32_NONCE_MAX)
type = BCH_EXTENT_ENTRY_crc32;
else if (bch_crc_bytes[new.csum_type] <= 10 &&
new.uncompressed_size <= CRC64_SIZE_MAX &&
new.nonce <= CRC64_NONCE_MAX)
type = BCH_EXTENT_ENTRY_crc64;
else if (bch_crc_bytes[new.csum_type] <= 16 &&
new.uncompressed_size <= CRC128_SIZE_MAX &&
new.nonce <= CRC128_NONCE_MAX)
type = BCH_EXTENT_ENTRY_crc128;
else
BUG();
bch2_extent_crc_pack(crc, new, type);
k->k.u64s += extent_entry_u64s(ptrs.end);
EBUG_ON(bkey_val_u64s(&k->k) > BKEY_EXTENT_VAL_U64s_MAX);
}
/* Generic code for keys with pointers: */
unsigned bch2_bkey_nr_ptrs(struct bkey_s_c k)
{
return bch2_bkey_devs(k).nr;
}
unsigned bch2_bkey_nr_ptrs_allocated(struct bkey_s_c k)
{
return k.k->type == KEY_TYPE_reservation
? bkey_s_c_to_reservation(k).v->nr_replicas
: bch2_bkey_dirty_devs(k).nr;
}
unsigned bch2_bkey_nr_ptrs_fully_allocated(struct bkey_s_c k)
{
unsigned ret = 0;
if (k.k->type == KEY_TYPE_reservation) {
ret = bkey_s_c_to_reservation(k).v->nr_replicas;
} else {
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
bkey_for_each_ptr_decode(k.k, ptrs, p, entry)
ret += !p.ptr.cached && !crc_is_compressed(p.crc);
}
return ret;
}
unsigned bch2_bkey_sectors_compressed(struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
unsigned ret = 0;
bkey_for_each_ptr_decode(k.k, ptrs, p, entry)
if (!p.ptr.cached && crc_is_compressed(p.crc))
ret += p.crc.compressed_size;
return ret;
}
bool bch2_bkey_is_incompressible(struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct bch_extent_crc_unpacked crc;
bkey_for_each_crc(k.k, ptrs, crc, entry)
if (crc.compression_type == BCH_COMPRESSION_TYPE_incompressible)
return true;
return false;
}
unsigned bch2_bkey_replicas(struct bch_fs *c, struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p = { 0 };
unsigned replicas = 0;
bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
if (p.ptr.cached)
continue;
if (p.has_ec)
replicas += p.ec.redundancy;
replicas++;
}
return replicas;
}
static inline unsigned __extent_ptr_durability(struct bch_dev *ca, struct extent_ptr_decoded *p)
{
bcachefs: Nocow support This adds support for nocow mode, where we do writes in-place when possible. Patch components: - New boolean filesystem and inode option, nocow: note that when nocow is enabled, data checksumming and compression are implicitly disabled - To prevent in-place writes from racing with data moves (data_update.c) or bucket reuse (i.e. a bucket being reused and re-allocated while a nocow write is in flight, we have a new locking mechanism. Buckets can be locked for either data update or data move, using a fixed size hash table of two_state_shared locks. We don't have any chaining, meaning updates and moves to different buckets that hash to the same lock will wait unnecessarily - we'll want to watch for this becoming an issue. - The allocator path also needs to check for in-place writes in flight to a given bucket before giving it out: thus we add another counter to bucket_alloc_state so we can track this. - Fsync now may need to issue cache flushes to block devices instead of flushing the journal. We add a device bitmask to bch_inode_info, ei_devs_need_flush, which tracks devices that need to have flushes issued - note that this will lead to unnecessary flushes when other codepaths have already issued flushes, we may want to replace this with a sequence number. - New nocow write path: look up extents, and if they're writable write to them - otherwise fall back to the normal COW write path. XXX: switch to sequence numbers instead of bitmask for devs needing journal flush XXX: ei_quota_lock being a mutex means bch2_nocow_write_done() needs to run in process context - see if we can improve this Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2022-11-02 21:12:00 +00:00
if (p->ptr.cached)
return 0;
return p->has_ec
? p->ec.redundancy + 1
: ca->mi.durability;
}
unsigned bch2_extent_ptr_desired_durability(struct bch_fs *c, struct extent_ptr_decoded *p)
{
struct bch_dev *ca = bch2_dev_rcu(c, p->ptr.dev);
return ca ? __extent_ptr_durability(ca, p) : 0;
}
unsigned bch2_extent_ptr_durability(struct bch_fs *c, struct extent_ptr_decoded *p)
{
struct bch_dev *ca = bch2_dev_rcu(c, p->ptr.dev);
if (!ca || ca->mi.state == BCH_MEMBER_STATE_failed)
return 0;
return __extent_ptr_durability(ca, p);
}
unsigned bch2_bkey_durability(struct bch_fs *c, struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
unsigned durability = 0;
rcu_read_lock();
bkey_for_each_ptr_decode(k.k, ptrs, p, entry)
durability += bch2_extent_ptr_durability(c, &p);
rcu_read_unlock();
return durability;
}
static unsigned bch2_bkey_durability_safe(struct bch_fs *c, struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
unsigned durability = 0;
rcu_read_lock();
bkey_for_each_ptr_decode(k.k, ptrs, p, entry)
if (p.ptr.dev < c->sb.nr_devices && c->devs[p.ptr.dev])
durability += bch2_extent_ptr_durability(c, &p);
rcu_read_unlock();
return durability;
}
void bch2_bkey_extent_entry_drop(struct bkey_i *k, union bch_extent_entry *entry)
{
union bch_extent_entry *end = bkey_val_end(bkey_i_to_s(k));
union bch_extent_entry *next = extent_entry_next(entry);
memmove_u64s(entry, next, (u64 *) end - (u64 *) next);
k->k.u64s -= extent_entry_u64s(entry);
}
void bch2_extent_ptr_decoded_append(struct bkey_i *k,
struct extent_ptr_decoded *p)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(bkey_i_to_s(k));
struct bch_extent_crc_unpacked crc =
bch2_extent_crc_unpack(&k->k, NULL);
union bch_extent_entry *pos;
if (!bch2_crc_unpacked_cmp(crc, p->crc)) {
pos = ptrs.start;
goto found;
}
bkey_for_each_crc(&k->k, ptrs, crc, pos)
if (!bch2_crc_unpacked_cmp(crc, p->crc)) {
pos = extent_entry_next(pos);
goto found;
}
bch2_extent_crc_append(k, p->crc);
pos = bkey_val_end(bkey_i_to_s(k));
found:
p->ptr.type = 1 << BCH_EXTENT_ENTRY_ptr;
__extent_entry_insert(k, pos, to_entry(&p->ptr));
if (p->has_ec) {
p->ec.type = 1 << BCH_EXTENT_ENTRY_stripe_ptr;
__extent_entry_insert(k, pos, to_entry(&p->ec));
}
}
static union bch_extent_entry *extent_entry_prev(struct bkey_ptrs ptrs,
union bch_extent_entry *entry)
{
union bch_extent_entry *i = ptrs.start;
if (i == entry)
return NULL;
while (extent_entry_next(i) != entry)
i = extent_entry_next(i);
return i;
}
/*
* Returns pointer to the next entry after the one being dropped:
*/
union bch_extent_entry *bch2_bkey_drop_ptr_noerror(struct bkey_s k,
struct bch_extent_ptr *ptr)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(k);
union bch_extent_entry *entry = to_entry(ptr), *next;
union bch_extent_entry *ret = entry;
bool drop_crc = true;
EBUG_ON(ptr < &ptrs.start->ptr ||
ptr >= &ptrs.end->ptr);
EBUG_ON(ptr->type != 1 << BCH_EXTENT_ENTRY_ptr);
for (next = extent_entry_next(entry);
next != ptrs.end;
next = extent_entry_next(next)) {
if (extent_entry_is_crc(next)) {
break;
} else if (extent_entry_is_ptr(next)) {
drop_crc = false;
break;
}
}
extent_entry_drop(k, entry);
while ((entry = extent_entry_prev(ptrs, entry))) {
if (extent_entry_is_ptr(entry))
break;
if ((extent_entry_is_crc(entry) && drop_crc) ||
extent_entry_is_stripe_ptr(entry)) {
ret = (void *) ret - extent_entry_bytes(entry);
extent_entry_drop(k, entry);
}
}
return ret;
}
union bch_extent_entry *bch2_bkey_drop_ptr(struct bkey_s k,
struct bch_extent_ptr *ptr)
{
bool have_dirty = bch2_bkey_dirty_devs(k.s_c).nr;
union bch_extent_entry *ret =
bch2_bkey_drop_ptr_noerror(k, ptr);
/*
* If we deleted all the dirty pointers and there's still cached
* pointers, we could set the cached pointers to dirty if they're not
* stale - but to do that correctly we'd need to grab an open_bucket
* reference so that we don't race with bucket reuse:
*/
if (have_dirty &&
!bch2_bkey_dirty_devs(k.s_c).nr) {
k.k->type = KEY_TYPE_error;
set_bkey_val_u64s(k.k, 0);
ret = NULL;
} else if (!bch2_bkey_nr_ptrs(k.s_c)) {
k.k->type = KEY_TYPE_deleted;
set_bkey_val_u64s(k.k, 0);
ret = NULL;
}
return ret;
}
void bch2_bkey_drop_device(struct bkey_s k, unsigned dev)
{
bch2_bkey_drop_ptrs(k, ptr, ptr->dev == dev);
}
void bch2_bkey_drop_device_noerror(struct bkey_s k, unsigned dev)
{
struct bch_extent_ptr *ptr = bch2_bkey_has_device(k, dev);
if (ptr)
bch2_bkey_drop_ptr_noerror(k, ptr);
}
const struct bch_extent_ptr *bch2_bkey_has_device_c(struct bkey_s_c k, unsigned dev)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
bkey_for_each_ptr(ptrs, ptr)
if (ptr->dev == dev)
return ptr;
return NULL;
}
bool bch2_bkey_has_target(struct bch_fs *c, struct bkey_s_c k, unsigned target)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
struct bch_dev *ca;
bool ret = false;
rcu_read_lock();
bkey_for_each_ptr(ptrs, ptr)
if (bch2_dev_in_target(c, ptr->dev, target) &&
(ca = bch2_dev_rcu(c, ptr->dev)) &&
(!ptr->cached ||
!dev_ptr_stale_rcu(ca, ptr))) {
ret = true;
break;
}
rcu_read_unlock();
return ret;
}
bool bch2_bkey_matches_ptr(struct bch_fs *c, struct bkey_s_c k,
struct bch_extent_ptr m, u64 offset)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
bkey_for_each_ptr_decode(k.k, ptrs, p, entry)
if (p.ptr.dev == m.dev &&
p.ptr.gen == m.gen &&
(s64) p.ptr.offset + p.crc.offset - bkey_start_offset(k.k) ==
(s64) m.offset - offset)
return true;
return false;
}
/*
* Returns true if two extents refer to the same data:
*/
bool bch2_extents_match(struct bkey_s_c k1, struct bkey_s_c k2)
{
bcachefs: Nocow support This adds support for nocow mode, where we do writes in-place when possible. Patch components: - New boolean filesystem and inode option, nocow: note that when nocow is enabled, data checksumming and compression are implicitly disabled - To prevent in-place writes from racing with data moves (data_update.c) or bucket reuse (i.e. a bucket being reused and re-allocated while a nocow write is in flight, we have a new locking mechanism. Buckets can be locked for either data update or data move, using a fixed size hash table of two_state_shared locks. We don't have any chaining, meaning updates and moves to different buckets that hash to the same lock will wait unnecessarily - we'll want to watch for this becoming an issue. - The allocator path also needs to check for in-place writes in flight to a given bucket before giving it out: thus we add another counter to bucket_alloc_state so we can track this. - Fsync now may need to issue cache flushes to block devices instead of flushing the journal. We add a device bitmask to bch_inode_info, ei_devs_need_flush, which tracks devices that need to have flushes issued - note that this will lead to unnecessary flushes when other codepaths have already issued flushes, we may want to replace this with a sequence number. - New nocow write path: look up extents, and if they're writable write to them - otherwise fall back to the normal COW write path. XXX: switch to sequence numbers instead of bitmask for devs needing journal flush XXX: ei_quota_lock being a mutex means bch2_nocow_write_done() needs to run in process context - see if we can improve this Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2022-11-02 21:12:00 +00:00
if (k1.k->type != k2.k->type)
return false;
bcachefs: Nocow support This adds support for nocow mode, where we do writes in-place when possible. Patch components: - New boolean filesystem and inode option, nocow: note that when nocow is enabled, data checksumming and compression are implicitly disabled - To prevent in-place writes from racing with data moves (data_update.c) or bucket reuse (i.e. a bucket being reused and re-allocated while a nocow write is in flight, we have a new locking mechanism. Buckets can be locked for either data update or data move, using a fixed size hash table of two_state_shared locks. We don't have any chaining, meaning updates and moves to different buckets that hash to the same lock will wait unnecessarily - we'll want to watch for this becoming an issue. - The allocator path also needs to check for in-place writes in flight to a given bucket before giving it out: thus we add another counter to bucket_alloc_state so we can track this. - Fsync now may need to issue cache flushes to block devices instead of flushing the journal. We add a device bitmask to bch_inode_info, ei_devs_need_flush, which tracks devices that need to have flushes issued - note that this will lead to unnecessary flushes when other codepaths have already issued flushes, we may want to replace this with a sequence number. - New nocow write path: look up extents, and if they're writable write to them - otherwise fall back to the normal COW write path. XXX: switch to sequence numbers instead of bitmask for devs needing journal flush XXX: ei_quota_lock being a mutex means bch2_nocow_write_done() needs to run in process context - see if we can improve this Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2022-11-02 21:12:00 +00:00
if (bkey_extent_is_direct_data(k1.k)) {
struct bkey_ptrs_c ptrs1 = bch2_bkey_ptrs_c(k1);
struct bkey_ptrs_c ptrs2 = bch2_bkey_ptrs_c(k2);
const union bch_extent_entry *entry1, *entry2;
struct extent_ptr_decoded p1, p2;
if (bkey_extent_is_unwritten(k1) != bkey_extent_is_unwritten(k2))
return false;
bkey_for_each_ptr_decode(k1.k, ptrs1, p1, entry1)
bkey_for_each_ptr_decode(k2.k, ptrs2, p2, entry2)
if (p1.ptr.dev == p2.ptr.dev &&
p1.ptr.gen == p2.ptr.gen &&
(s64) p1.ptr.offset + p1.crc.offset - bkey_start_offset(k1.k) ==
(s64) p2.ptr.offset + p2.crc.offset - bkey_start_offset(k2.k))
return true;
bcachefs: Nocow support This adds support for nocow mode, where we do writes in-place when possible. Patch components: - New boolean filesystem and inode option, nocow: note that when nocow is enabled, data checksumming and compression are implicitly disabled - To prevent in-place writes from racing with data moves (data_update.c) or bucket reuse (i.e. a bucket being reused and re-allocated while a nocow write is in flight, we have a new locking mechanism. Buckets can be locked for either data update or data move, using a fixed size hash table of two_state_shared locks. We don't have any chaining, meaning updates and moves to different buckets that hash to the same lock will wait unnecessarily - we'll want to watch for this becoming an issue. - The allocator path also needs to check for in-place writes in flight to a given bucket before giving it out: thus we add another counter to bucket_alloc_state so we can track this. - Fsync now may need to issue cache flushes to block devices instead of flushing the journal. We add a device bitmask to bch_inode_info, ei_devs_need_flush, which tracks devices that need to have flushes issued - note that this will lead to unnecessary flushes when other codepaths have already issued flushes, we may want to replace this with a sequence number. - New nocow write path: look up extents, and if they're writable write to them - otherwise fall back to the normal COW write path. XXX: switch to sequence numbers instead of bitmask for devs needing journal flush XXX: ei_quota_lock being a mutex means bch2_nocow_write_done() needs to run in process context - see if we can improve this Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2022-11-02 21:12:00 +00:00
return false;
} else {
/* KEY_TYPE_deleted, etc. */
return true;
}
}
struct bch_extent_ptr *
bch2_extent_has_ptr(struct bkey_s_c k1, struct extent_ptr_decoded p1, struct bkey_s k2)
{
struct bkey_ptrs ptrs2 = bch2_bkey_ptrs(k2);
union bch_extent_entry *entry2;
struct extent_ptr_decoded p2;
bkey_for_each_ptr_decode(k2.k, ptrs2, p2, entry2)
if (p1.ptr.dev == p2.ptr.dev &&
p1.ptr.gen == p2.ptr.gen &&
(s64) p1.ptr.offset + p1.crc.offset - bkey_start_offset(k1.k) ==
(s64) p2.ptr.offset + p2.crc.offset - bkey_start_offset(k2.k))
return &entry2->ptr;
return NULL;
}
void bch2_extent_ptr_set_cached(struct bkey_s k, struct bch_extent_ptr *ptr)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(k);
union bch_extent_entry *entry;
union bch_extent_entry *ec = NULL;
bkey_extent_entry_for_each(ptrs, entry) {
if (&entry->ptr == ptr) {
ptr->cached = true;
if (ec)
extent_entry_drop(k, ec);
return;
}
if (extent_entry_is_stripe_ptr(entry))
ec = entry;
else if (extent_entry_is_ptr(entry))
ec = NULL;
}
BUG();
}
/*
* bch_extent_normalize - clean up an extent, dropping stale pointers etc.
*
* Returns true if @k should be dropped entirely
*
* For existing keys, only called when btree nodes are being rewritten, not when
* they're merely being compacted/resorted in memory.
*/
bool bch2_extent_normalize(struct bch_fs *c, struct bkey_s k)
{
struct bch_dev *ca;
rcu_read_lock();
bch2_bkey_drop_ptrs(k, ptr,
ptr->cached &&
(ca = bch2_dev_rcu(c, ptr->dev)) &&
dev_ptr_stale_rcu(ca, ptr) > 0);
rcu_read_unlock();
return bkey_deleted(k.k);
}
void bch2_extent_ptr_to_text(struct printbuf *out, struct bch_fs *c, const struct bch_extent_ptr *ptr)
{
out->atomic++;
rcu_read_lock();
struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev);
if (!ca) {
prt_printf(out, "ptr: %u:%llu gen %u%s", ptr->dev,
(u64) ptr->offset, ptr->gen,
ptr->cached ? " cached" : "");
} else {
u32 offset;
u64 b = sector_to_bucket_and_offset(ca, ptr->offset, &offset);
prt_printf(out, "ptr: %u:%llu:%u gen %u",
ptr->dev, b, offset, ptr->gen);
if (ptr->cached)
prt_str(out, " cached");
if (ptr->unwritten)
prt_str(out, " unwritten");
int stale = dev_ptr_stale_rcu(ca, ptr);
if (stale > 0)
prt_printf(out, " stale");
else if (stale)
prt_printf(out, " invalid");
}
rcu_read_unlock();
--out->atomic;
}
void bch2_bkey_ptrs_to_text(struct printbuf *out, struct bch_fs *c,
struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
bool first = true;
if (c)
prt_printf(out, "durability: %u ", bch2_bkey_durability_safe(c, k));
bkey_extent_entry_for_each(ptrs, entry) {
if (!first)
prt_printf(out, " ");
switch (__extent_entry_type(entry)) {
case BCH_EXTENT_ENTRY_ptr:
bch2_extent_ptr_to_text(out, c, entry_to_ptr(entry));
break;
case BCH_EXTENT_ENTRY_crc32:
case BCH_EXTENT_ENTRY_crc64:
case BCH_EXTENT_ENTRY_crc128: {
struct bch_extent_crc_unpacked crc =
bch2_extent_crc_unpack(k.k, entry_to_crc(entry));
prt_printf(out, "crc: c_size %u size %u offset %u nonce %u csum ",
crc.compressed_size,
crc.uncompressed_size,
crc.offset, crc.nonce);
bch2_prt_csum_type(out, crc.csum_type);
prt_str(out, " compress ");
bch2_prt_compression_type(out, crc.compression_type);
break;
}
case BCH_EXTENT_ENTRY_stripe_ptr: {
const struct bch_extent_stripe_ptr *ec = &entry->stripe_ptr;
prt_printf(out, "ec: idx %llu block %u",
(u64) ec->idx, ec->block);
break;
}
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
case BCH_EXTENT_ENTRY_rebalance: {
const struct bch_extent_rebalance *r = &entry->rebalance;
prt_str(out, "rebalance: target ");
if (c)
bch2_target_to_text(out, c, r->target);
else
prt_printf(out, "%u", r->target);
prt_str(out, " compression ");
bch2_compression_opt_to_text(out, r->compression);
break;
}
default:
prt_printf(out, "(invalid extent entry %.16llx)", *((u64 *) entry));
return;
}
first = false;
}
}
static int extent_ptr_invalid(struct bch_fs *c,
struct bkey_s_c k,
enum bch_validate_flags flags,
const struct bch_extent_ptr *ptr,
unsigned size_ondisk,
bool metadata,
struct printbuf *err)
{
int ret = 0;
rcu_read_lock();
struct bch_dev *ca = bch2_dev_rcu(c, ptr->dev);
if (!ca) {
rcu_read_unlock();
return 0;
}
u32 bucket_offset;
u64 bucket = sector_to_bucket_and_offset(ca, ptr->offset, &bucket_offset);
unsigned first_bucket = ca->mi.first_bucket;
u64 nbuckets = ca->mi.nbuckets;
unsigned bucket_size = ca->mi.bucket_size;
rcu_read_unlock();
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
bkey_for_each_ptr(ptrs, ptr2)
bkey_fsck_err_on(ptr != ptr2 && ptr->dev == ptr2->dev, c, err,
ptr_to_duplicate_device,
"multiple pointers to same device (%u)", ptr->dev);
bkey_fsck_err_on(bucket >= nbuckets, c, err,
ptr_after_last_bucket,
"pointer past last bucket (%llu > %llu)", bucket, nbuckets);
bkey_fsck_err_on(bucket < first_bucket, c, err,
ptr_before_first_bucket,
"pointer before first bucket (%llu < %u)", bucket, first_bucket);
bkey_fsck_err_on(bucket_offset + size_ondisk > bucket_size, c, err,
ptr_spans_multiple_buckets,
"pointer spans multiple buckets (%u + %u > %u)",
bucket_offset, size_ondisk, bucket_size);
fsck_err:
return ret;
}
int bch2_bkey_ptrs_invalid(struct bch_fs *c, struct bkey_s_c k,
enum bch_validate_flags flags,
struct printbuf *err)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
struct bch_extent_crc_unpacked crc;
unsigned size_ondisk = k.k->size;
unsigned nonce = UINT_MAX;
unsigned nr_ptrs = 0;
bool have_written = false, have_unwritten = false, have_ec = false, crc_since_last_ptr = false;
int ret = 0;
if (bkey_is_btree_ptr(k.k))
size_ondisk = btree_sectors(c);
bkey_extent_entry_for_each(ptrs, entry) {
bkey_fsck_err_on(__extent_entry_type(entry) >= BCH_EXTENT_ENTRY_MAX, c, err,
extent_ptrs_invalid_entry,
"invalid extent entry type (got %u, max %u)",
__extent_entry_type(entry), BCH_EXTENT_ENTRY_MAX);
bkey_fsck_err_on(bkey_is_btree_ptr(k.k) &&
!extent_entry_is_ptr(entry), c, err,
btree_ptr_has_non_ptr,
"has non ptr field");
switch (extent_entry_type(entry)) {
case BCH_EXTENT_ENTRY_ptr:
ret = extent_ptr_invalid(c, k, flags, &entry->ptr,
size_ondisk, false, err);
if (ret)
return ret;
bkey_fsck_err_on(entry->ptr.cached && have_ec, c, err,
ptr_cached_and_erasure_coded,
"cached, erasure coded ptr");
if (!entry->ptr.unwritten)
have_written = true;
else
have_unwritten = true;
have_ec = false;
crc_since_last_ptr = false;
nr_ptrs++;
break;
case BCH_EXTENT_ENTRY_crc32:
case BCH_EXTENT_ENTRY_crc64:
case BCH_EXTENT_ENTRY_crc128:
crc = bch2_extent_crc_unpack(k.k, entry_to_crc(entry));
bkey_fsck_err_on(crc.offset + crc.live_size > crc.uncompressed_size, c, err,
ptr_crc_uncompressed_size_too_small,
"checksum offset + key size > uncompressed size");
bkey_fsck_err_on(!bch2_checksum_type_valid(c, crc.csum_type), c, err,
ptr_crc_csum_type_unknown,
"invalid checksum type");
bkey_fsck_err_on(crc.compression_type >= BCH_COMPRESSION_TYPE_NR, c, err,
ptr_crc_compression_type_unknown,
"invalid compression type");
if (bch2_csum_type_is_encryption(crc.csum_type)) {
if (nonce == UINT_MAX)
nonce = crc.offset + crc.nonce;
else if (nonce != crc.offset + crc.nonce)
bkey_fsck_err(c, err, ptr_crc_nonce_mismatch,
"incorrect nonce");
}
bkey_fsck_err_on(crc_since_last_ptr, c, err,
ptr_crc_redundant,
"redundant crc entry");
crc_since_last_ptr = true;
bkey_fsck_err_on(crc_is_encoded(crc) &&
(crc.uncompressed_size > c->opts.encoded_extent_max >> 9) &&
(flags & (BCH_VALIDATE_write|BCH_VALIDATE_commit)), c, err,
ptr_crc_uncompressed_size_too_big,
"too large encoded extent");
size_ondisk = crc.compressed_size;
break;
case BCH_EXTENT_ENTRY_stripe_ptr:
bkey_fsck_err_on(have_ec, c, err,
ptr_stripe_redundant,
"redundant stripe entry");
have_ec = true;
break;
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
case BCH_EXTENT_ENTRY_rebalance: {
const struct bch_extent_rebalance *r = &entry->rebalance;
if (!bch2_compression_opt_valid(r->compression)) {
struct bch_compression_opt opt = __bch2_compression_decode(r->compression);
prt_printf(err, "invalid compression opt %u:%u",
opt.type, opt.level);
return -BCH_ERR_invalid_bkey;
}
break;
}
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
}
}
bkey_fsck_err_on(!nr_ptrs, c, err,
extent_ptrs_no_ptrs,
"no ptrs");
bkey_fsck_err_on(nr_ptrs > BCH_BKEY_PTRS_MAX, c, err,
extent_ptrs_too_many_ptrs,
"too many ptrs: %u > %u", nr_ptrs, BCH_BKEY_PTRS_MAX);
bkey_fsck_err_on(have_written && have_unwritten, c, err,
extent_ptrs_written_and_unwritten,
"extent with unwritten and written ptrs");
bkey_fsck_err_on(k.k->type != KEY_TYPE_extent && have_unwritten, c, err,
extent_ptrs_unwritten,
"has unwritten ptrs");
bkey_fsck_err_on(crc_since_last_ptr, c, err,
extent_ptrs_redundant_crc,
"redundant crc entry");
bkey_fsck_err_on(have_ec, c, err,
extent_ptrs_redundant_stripe,
"redundant stripe entry");
fsck_err:
return ret;
}
void bch2_ptr_swab(struct bkey_s k)
{
struct bkey_ptrs ptrs = bch2_bkey_ptrs(k);
union bch_extent_entry *entry;
u64 *d;
for (d = (u64 *) ptrs.start;
d != (u64 *) ptrs.end;
d++)
*d = swab64(*d);
for (entry = ptrs.start;
entry < ptrs.end;
entry = extent_entry_next(entry)) {
switch (extent_entry_type(entry)) {
case BCH_EXTENT_ENTRY_ptr:
break;
case BCH_EXTENT_ENTRY_crc32:
entry->crc32.csum = swab32(entry->crc32.csum);
break;
case BCH_EXTENT_ENTRY_crc64:
entry->crc64.csum_hi = swab16(entry->crc64.csum_hi);
entry->crc64.csum_lo = swab64(entry->crc64.csum_lo);
break;
case BCH_EXTENT_ENTRY_crc128:
entry->crc128.csum.hi = (__force __le64)
swab64((__force u64) entry->crc128.csum.hi);
entry->crc128.csum.lo = (__force __le64)
swab64((__force u64) entry->crc128.csum.lo);
break;
case BCH_EXTENT_ENTRY_stripe_ptr:
break;
case BCH_EXTENT_ENTRY_rebalance:
break;
}
}
}
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
const struct bch_extent_rebalance *bch2_bkey_rebalance_opts(struct bkey_s_c k)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
const union bch_extent_entry *entry;
bkey_extent_entry_for_each(ptrs, entry)
if (__extent_entry_type(entry) == BCH_EXTENT_ENTRY_rebalance)
return &entry->rebalance;
return NULL;
}
unsigned bch2_bkey_ptrs_need_rebalance(struct bch_fs *c, struct bkey_s_c k,
unsigned target, unsigned compression)
{
struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
unsigned rewrite_ptrs = 0;
if (compression) {
unsigned compression_type = bch2_compression_opt_to_type(compression);
const union bch_extent_entry *entry;
struct extent_ptr_decoded p;
unsigned i = 0;
bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
if (p.crc.compression_type == BCH_COMPRESSION_TYPE_incompressible ||
p.ptr.unwritten) {
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
rewrite_ptrs = 0;
goto incompressible;
}
if (!p.ptr.cached && p.crc.compression_type != compression_type)
rewrite_ptrs |= 1U << i;
i++;
}
}
incompressible:
if (target && bch2_target_accepts_data(c, BCH_DATA_user, target)) {
unsigned i = 0;
bkey_for_each_ptr(ptrs, ptr) {
if (!ptr->cached && !bch2_dev_in_target(c, ptr->dev, target))
rewrite_ptrs |= 1U << i;
i++;
}
}
return rewrite_ptrs;
}
bool bch2_bkey_needs_rebalance(struct bch_fs *c, struct bkey_s_c k)
{
const struct bch_extent_rebalance *r = bch2_bkey_rebalance_opts(k);
/*
* If it's an indirect extent, we don't delete the rebalance entry when
* done so that we know what options were applied - check if it still
* needs work done:
*/
if (r &&
k.k->type == KEY_TYPE_reflink_v &&
!bch2_bkey_ptrs_need_rebalance(c, k, r->target, r->compression))
r = NULL;
return r != NULL;
}
int bch2_bkey_set_needs_rebalance(struct bch_fs *c, struct bkey_i *_k,
struct bch_io_opts *opts)
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
{
struct bkey_s k = bkey_i_to_s(_k);
struct bch_extent_rebalance *r;
unsigned target = opts->background_target;
unsigned compression = background_compression(*opts);
bcachefs: rebalance_work This adds a new btree, rebalance_work, to eliminate scanning required for finding extents that need work done on them in the background - i.e. for the background_target and background_compression options. rebalance_work is a bitset btree, where a KEY_TYPE_set corresponds to an extent in the extents or reflink btree at the same pos. A new extent field is added, bch_extent_rebalance, which indicates that this extent has work that needs to be done in the background - and which options to use. This allows per-inode options to be propagated to indirect extents - at least in some circumstances. In this patch, changing IO options on a file will not propagate the new options to indirect extents pointed to by that file. Updating (setting/clearing) the rebalance_work btree is done by the extent trigger, which looks at the bch_extent_rebalance field. Scanning is still requrired after changing IO path options - either just for a given inode, or for the whole filesystem. We indicate that scanning is required by adding a KEY_TYPE_cookie key to the rebalance_work btree: the cookie counter is so that we can detect that scanning is still required when an option has been flipped mid-way through an existing scan. Future possible work: - Propagate options to indirect extents when being changed - Add other IO path options - nr_replicas, ec, to rebalance_work so they can be applied in the background when they change - Add a counter, for bcachefs fs usage output, showing the pending amount of rebalance work: we'll probably want to do this after the disk space accounting rewrite (moving it to a new btree) Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
2023-10-20 17:33:14 +00:00
bool needs_rebalance;
if (!bkey_extent_is_direct_data(k.k))
return 0;
/* get existing rebalance entry: */
r = (struct bch_extent_rebalance *) bch2_bkey_rebalance_opts(k.s_c);
if (r) {
if (k.k->type == KEY_TYPE_reflink_v) {
/*
* indirect extents: existing options take precedence,
* so that we don't move extents back and forth if
* they're referenced by different inodes with different
* options:
*/
if (r->target)
target = r->target;
if (r->compression)
compression = r->compression;
}
r->target = target;
r->compression = compression;
}
needs_rebalance = bch2_bkey_ptrs_need_rebalance(c, k.s_c, target, compression);
if (needs_rebalance && !r) {
union bch_extent_entry *new = bkey_val_end(k);
new->rebalance.type = 1U << BCH_EXTENT_ENTRY_rebalance;
new->rebalance.compression = compression;
new->rebalance.target = target;
new->rebalance.unused = 0;
k.k->u64s += extent_entry_u64s(new);
} else if (!needs_rebalance && r && k.k->type != KEY_TYPE_reflink_v) {
/*
* For indirect extents, don't delete the rebalance entry when
* we're finished so that we know we specifically moved it or
* compressed it to its current location/compression type
*/
extent_entry_drop(k, (union bch_extent_entry *) r);
}
return 0;
}
/* Generic extent code: */
int bch2_cut_front_s(struct bpos where, struct bkey_s k)
{
unsigned new_val_u64s = bkey_val_u64s(k.k);
int val_u64s_delta;
u64 sub;
if (bkey_le(where, bkey_start_pos(k.k)))
return 0;
EBUG_ON(bkey_gt(where, k.k->p));
sub = where.offset - bkey_start_offset(k.k);
k.k->size -= sub;
if (!k.k->size) {
k.k->type = KEY_TYPE_deleted;
new_val_u64s = 0;
}
switch (k.k->type) {
case KEY_TYPE_extent:
case KEY_TYPE_reflink_v: {
struct bkey_ptrs ptrs = bch2_bkey_ptrs(k);
union bch_extent_entry *entry;
bool seen_crc = false;
bkey_extent_entry_for_each(ptrs, entry) {
switch (extent_entry_type(entry)) {
case BCH_EXTENT_ENTRY_ptr:
if (!seen_crc)
entry->ptr.offset += sub;
break;
case BCH_EXTENT_ENTRY_crc32:
entry->crc32.offset += sub;
break;
case BCH_EXTENT_ENTRY_crc64:
entry->crc64.offset += sub;
break;
case BCH_EXTENT_ENTRY_crc128:
entry->crc128.offset += sub;
break;
case BCH_EXTENT_ENTRY_stripe_ptr:
break;
case BCH_EXTENT_ENTRY_rebalance:
break;
}
if (extent_entry_is_crc(entry))
seen_crc = true;
}
break;
}
case KEY_TYPE_reflink_p: {
struct bkey_s_reflink_p p = bkey_s_to_reflink_p(k);
le64_add_cpu(&p.v->idx, sub);
break;
}
case KEY_TYPE_inline_data:
case KEY_TYPE_indirect_inline_data: {
void *p = bkey_inline_data_p(k);
unsigned bytes = bkey_inline_data_bytes(k.k);
sub = min_t(u64, sub << 9, bytes);
memmove(p, p + sub, bytes - sub);
new_val_u64s -= sub >> 3;
break;
}
}
val_u64s_delta = bkey_val_u64s(k.k) - new_val_u64s;
BUG_ON(val_u64s_delta < 0);
set_bkey_val_u64s(k.k, new_val_u64s);
memset(bkey_val_end(k), 0, val_u64s_delta * sizeof(u64));
return -val_u64s_delta;
}
int bch2_cut_back_s(struct bpos where, struct bkey_s k)
{
unsigned new_val_u64s = bkey_val_u64s(k.k);
int val_u64s_delta;
u64 len = 0;
if (bkey_ge(where, k.k->p))
return 0;
EBUG_ON(bkey_lt(where, bkey_start_pos(k.k)));
len = where.offset - bkey_start_offset(k.k);
k.k->p.offset = where.offset;
k.k->size = len;
if (!len) {
k.k->type = KEY_TYPE_deleted;
new_val_u64s = 0;
}
switch (k.k->type) {
case KEY_TYPE_inline_data:
case KEY_TYPE_indirect_inline_data:
new_val_u64s = (bkey_inline_data_offset(k.k) +
min(bkey_inline_data_bytes(k.k), k.k->size << 9)) >> 3;
break;
}
val_u64s_delta = bkey_val_u64s(k.k) - new_val_u64s;
BUG_ON(val_u64s_delta < 0);
set_bkey_val_u64s(k.k, new_val_u64s);
memset(bkey_val_end(k), 0, val_u64s_delta * sizeof(u64));
return -val_u64s_delta;
}