Adding upstream version 6.8.9.upstream/6.8.9

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 70 insertions, 935 deletions

diff --git a/fs/bcachefs/bcachefs_format.h b/fs/bcachefs/bcachefs_format.h
index fe78e87603..0668b682a2 100644
--- a/fs/bcachefs/bcachefs_format.h
+++ b/fs/bcachefs/bcachefs_format.h
@@ -307,6 +307,13 @@ struct bkey_i {
 	struct bch_val	v;
 };
 
+#define POS_KEY(_pos)							\
+((struct bkey) {							\
+	.u64s		= BKEY_U64s,					\
+	.format		= KEY_FORMAT_CURRENT,				\
+	.p		= _pos,						\
+})
+
 #define KEY(_inode, _offset, _size)					\
 ((struct bkey) {							\
 	.u64s		= BKEY_U64s,					\
@@ -410,600 +417,12 @@ struct bch_set {
 	struct bch_val		v;
 };
 
-/* Extents */
-
-/*
- * In extent bkeys, the value is a list of pointers (bch_extent_ptr), optionally
- * preceded by checksum/compression information (bch_extent_crc32 or
- * bch_extent_crc64).
- *
- * One major determining factor in the format of extents is how we handle and
- * represent extents that have been partially overwritten and thus trimmed:
- *
- * If an extent is not checksummed or compressed, when the extent is trimmed we
- * don't have to remember the extent we originally allocated and wrote: we can
- * merely adjust ptr->offset to point to the start of the data that is currently
- * live. The size field in struct bkey records the current (live) size of the
- * extent, and is also used to mean "size of region on disk that we point to" in
- * this case.
- *
- * Thus an extent that is not checksummed or compressed will consist only of a
- * list of bch_extent_ptrs, with none of the fields in
- * bch_extent_crc32/bch_extent_crc64.
- *
- * When an extent is checksummed or compressed, it's not possible to read only
- * the data that is currently live: we have to read the entire extent that was
- * originally written, and then return only the part of the extent that is
- * currently live.
- *
- * Thus, in addition to the current size of the extent in struct bkey, we need
- * to store the size of the originally allocated space - this is the
- * compressed_size and uncompressed_size fields in bch_extent_crc32/64. Also,
- * when the extent is trimmed, instead of modifying the offset field of the
- * pointer, we keep a second smaller offset field - "offset into the original
- * extent of the currently live region".
- *
- * The other major determining factor is replication and data migration:
- *
- * Each pointer may have its own bch_extent_crc32/64. When doing a replicated
- * write, we will initially write all the replicas in the same format, with the
- * same checksum type and compression format - however, when copygc runs later (or
- * tiering/cache promotion, anything that moves data), it is not in general
- * going to rewrite all the pointers at once - one of the replicas may be in a
- * bucket on one device that has very little fragmentation while another lives
- * in a bucket that has become heavily fragmented, and thus is being rewritten
- * sooner than the rest.
- *
- * Thus it will only move a subset of the pointers (or in the case of
- * tiering/cache promotion perhaps add a single pointer without dropping any
- * current pointers), and if the extent has been partially overwritten it must
- * write only the currently live portion (or copygc would not be able to reduce
- * fragmentation!) - which necessitates a different bch_extent_crc format for
- * the new pointer.
- *
- * But in the interests of space efficiency, we don't want to store one
- * bch_extent_crc for each pointer if we don't have to.
- *
- * Thus, a bch_extent consists of bch_extent_crc32s, bch_extent_crc64s, and
- * bch_extent_ptrs appended arbitrarily one after the other. We determine the
- * type of a given entry with a scheme similar to utf8 (except we're encoding a
- * type, not a size), encoding the type in the position of the first set bit:
- *
- * bch_extent_crc32	- 0b1
- * bch_extent_ptr	- 0b10
- * bch_extent_crc64	- 0b100
- *
- * We do it this way because bch_extent_crc32 is _very_ constrained on bits (and
- * bch_extent_crc64 is the least constrained).
- *
- * Then, each bch_extent_crc32/64 applies to the pointers that follow after it,
- * until the next bch_extent_crc32/64.
- *
- * If there are no bch_extent_crcs preceding a bch_extent_ptr, then that pointer
- * is neither checksummed nor compressed.
- */
-
 /* 128 bits, sufficient for cryptographic MACs: */
 struct bch_csum {
 	__le64			lo;
 	__le64			hi;
 } __packed __aligned(8);
 
-#define BCH_EXTENT_ENTRY_TYPES()		\
-	x(ptr,			0)		\
-	x(crc32,		1)		\
-	x(crc64,		2)		\
-	x(crc128,		3)		\
-	x(stripe_ptr,		4)		\
-	x(rebalance,		5)
-#define BCH_EXTENT_ENTRY_MAX	6
-
-enum bch_extent_entry_type {
-#define x(f, n) BCH_EXTENT_ENTRY_##f = n,
-	BCH_EXTENT_ENTRY_TYPES()
-#undef x
-};
-
-/* Compressed/uncompressed size are stored biased by 1: */
-struct bch_extent_crc32 {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u32			type:2,
-				_compressed_size:7,
-				_uncompressed_size:7,
-				offset:7,
-				_unused:1,
-				csum_type:4,
-				compression_type:4;
-	__u32			csum;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u32			csum;
-	__u32			compression_type:4,
-				csum_type:4,
-				_unused:1,
-				offset:7,
-				_uncompressed_size:7,
-				_compressed_size:7,
-				type:2;
-#endif
-} __packed __aligned(8);
-
-#define CRC32_SIZE_MAX		(1U << 7)
-#define CRC32_NONCE_MAX		0
-
-struct bch_extent_crc64 {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u64			type:3,
-				_compressed_size:9,
-				_uncompressed_size:9,
-				offset:9,
-				nonce:10,
-				csum_type:4,
-				compression_type:4,
-				csum_hi:16;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u64			csum_hi:16,
-				compression_type:4,
-				csum_type:4,
-				nonce:10,
-				offset:9,
-				_uncompressed_size:9,
-				_compressed_size:9,
-				type:3;
-#endif
-	__u64			csum_lo;
-} __packed __aligned(8);
-
-#define CRC64_SIZE_MAX		(1U << 9)
-#define CRC64_NONCE_MAX		((1U << 10) - 1)
-
-struct bch_extent_crc128 {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u64			type:4,
-				_compressed_size:13,
-				_uncompressed_size:13,
-				offset:13,
-				nonce:13,
-				csum_type:4,
-				compression_type:4;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u64			compression_type:4,
-				csum_type:4,
-				nonce:13,
-				offset:13,
-				_uncompressed_size:13,
-				_compressed_size:13,
-				type:4;
-#endif
-	struct bch_csum		csum;
-} __packed __aligned(8);
-
-#define CRC128_SIZE_MAX		(1U << 13)
-#define CRC128_NONCE_MAX	((1U << 13) - 1)
-
-/*
- * @reservation - pointer hasn't been written to, just reserved
- */
-struct bch_extent_ptr {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u64			type:1,
-				cached:1,
-				unused:1,
-				unwritten:1,
-				offset:44, /* 8 petabytes */
-				dev:8,
-				gen:8;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u64			gen:8,
-				dev:8,
-				offset:44,
-				unwritten:1,
-				unused:1,
-				cached:1,
-				type:1;
-#endif
-} __packed __aligned(8);
-
-struct bch_extent_stripe_ptr {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u64			type:5,
-				block:8,
-				redundancy:4,
-				idx:47;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u64			idx:47,
-				redundancy:4,
-				block:8,
-				type:5;
-#endif
-};
-
-struct bch_extent_rebalance {
-#if defined(__LITTLE_ENDIAN_BITFIELD)
-	__u64			type:6,
-				unused:34,
-				compression:8, /* enum bch_compression_opt */
-				target:16;
-#elif defined (__BIG_ENDIAN_BITFIELD)
-	__u64			target:16,
-				compression:8,
-				unused:34,
-				type:6;
-#endif
-};
-
-union bch_extent_entry {
-#if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ ||  __BITS_PER_LONG == 64
-	unsigned long			type;
-#elif __BITS_PER_LONG == 32
-	struct {
-		unsigned long		pad;
-		unsigned long		type;
-	};
-#else
-#error edit for your odd byteorder.
-#endif
-
-#define x(f, n) struct bch_extent_##f	f;
-	BCH_EXTENT_ENTRY_TYPES()
-#undef x
-};
-
-struct bch_btree_ptr {
-	struct bch_val		v;
-
-	__u64			_data[0];
-	struct bch_extent_ptr	start[];
-} __packed __aligned(8);
-
-struct bch_btree_ptr_v2 {
-	struct bch_val		v;
-
-	__u64			mem_ptr;
-	__le64			seq;
-	__le16			sectors_written;
-	__le16			flags;
-	struct bpos		min_key;
-	__u64			_data[0];
-	struct bch_extent_ptr	start[];
-} __packed __aligned(8);
-
-LE16_BITMASK(BTREE_PTR_RANGE_UPDATED,	struct bch_btree_ptr_v2, flags, 0, 1);
-
-struct bch_extent {
-	struct bch_val		v;
-
-	__u64			_data[0];
-	union bch_extent_entry	start[];
-} __packed __aligned(8);
-
-struct bch_reservation {
-	struct bch_val		v;
-
-	__le32			generation;
-	__u8			nr_replicas;
-	__u8			pad[3];
-} __packed __aligned(8);
-
-/* Maximum size (in u64s) a single pointer could be: */
-#define BKEY_EXTENT_PTR_U64s_MAX\
-	((sizeof(struct bch_extent_crc128) +			\
-	  sizeof(struct bch_extent_ptr)) / sizeof(__u64))
-
-/* Maximum possible size of an entire extent value: */
-#define BKEY_EXTENT_VAL_U64s_MAX				\
-	(1 + BKEY_EXTENT_PTR_U64s_MAX * (BCH_REPLICAS_MAX + 1))
-
-/* * Maximum possible size of an entire extent, key + value: */
-#define BKEY_EXTENT_U64s_MAX		(BKEY_U64s + BKEY_EXTENT_VAL_U64s_MAX)
-
-/* Btree pointers don't carry around checksums: */
-#define BKEY_BTREE_PTR_VAL_U64s_MAX				\
-	((sizeof(struct bch_btree_ptr_v2) +			\
-	  sizeof(struct bch_extent_ptr) * BCH_REPLICAS_MAX) / sizeof(__u64))
-#define BKEY_BTREE_PTR_U64s_MAX					\
-	(BKEY_U64s + BKEY_BTREE_PTR_VAL_U64s_MAX)
-
-/* Inodes */
-
-#define BLOCKDEV_INODE_MAX	4096
-
-#define BCACHEFS_ROOT_INO	4096
-
-struct bch_inode {
-	struct bch_val		v;
-
-	__le64			bi_hash_seed;
-	__le32			bi_flags;
-	__le16			bi_mode;
-	__u8			fields[];
-} __packed __aligned(8);
-
-struct bch_inode_v2 {
-	struct bch_val		v;
-
-	__le64			bi_journal_seq;
-	__le64			bi_hash_seed;
-	__le64			bi_flags;
-	__le16			bi_mode;
-	__u8			fields[];
-} __packed __aligned(8);
-
-struct bch_inode_v3 {
-	struct bch_val		v;
-
-	__le64			bi_journal_seq;
-	__le64			bi_hash_seed;
-	__le64			bi_flags;
-	__le64			bi_sectors;
-	__le64			bi_size;
-	__le64			bi_version;
-	__u8			fields[];
-} __packed __aligned(8);
-
-#define INODEv3_FIELDS_START_INITIAL	6
-#define INODEv3_FIELDS_START_CUR	(offsetof(struct bch_inode_v3, fields) / sizeof(__u64))
-
-struct bch_inode_generation {
-	struct bch_val		v;
-
-	__le32			bi_generation;
-	__le32			pad;
-} __packed __aligned(8);
-
-/*
- * bi_subvol and bi_parent_subvol are only set for subvolume roots:
- */
-
-#define BCH_INODE_FIELDS_v2()			\
-	x(bi_atime,			96)	\
-	x(bi_ctime,			96)	\
-	x(bi_mtime,			96)	\
-	x(bi_otime,			96)	\
-	x(bi_size,			64)	\
-	x(bi_sectors,			64)	\
-	x(bi_uid,			32)	\
-	x(bi_gid,			32)	\
-	x(bi_nlink,			32)	\
-	x(bi_generation,		32)	\
-	x(bi_dev,			32)	\
-	x(bi_data_checksum,		8)	\
-	x(bi_compression,		8)	\
-	x(bi_project,			32)	\
-	x(bi_background_compression,	8)	\
-	x(bi_data_replicas,		8)	\
-	x(bi_promote_target,		16)	\
-	x(bi_foreground_target,		16)	\
-	x(bi_background_target,		16)	\
-	x(bi_erasure_code,		16)	\
-	x(bi_fields_set,		16)	\
-	x(bi_dir,			64)	\
-	x(bi_dir_offset,		64)	\
-	x(bi_subvol,			32)	\
-	x(bi_parent_subvol,		32)
-
-#define BCH_INODE_FIELDS_v3()			\
-	x(bi_atime,			96)	\
-	x(bi_ctime,			96)	\
-	x(bi_mtime,			96)	\
-	x(bi_otime,			96)	\
-	x(bi_uid,			32)	\
-	x(bi_gid,			32)	\
-	x(bi_nlink,			32)	\
-	x(bi_generation,		32)	\
-	x(bi_dev,			32)	\
-	x(bi_data_checksum,		8)	\
-	x(bi_compression,		8)	\
-	x(bi_project,			32)	\
-	x(bi_background_compression,	8)	\
-	x(bi_data_replicas,		8)	\
-	x(bi_promote_target,		16)	\
-	x(bi_foreground_target,		16)	\
-	x(bi_background_target,		16)	\
-	x(bi_erasure_code,		16)	\
-	x(bi_fields_set,		16)	\
-	x(bi_dir,			64)	\
-	x(bi_dir_offset,		64)	\
-	x(bi_subvol,			32)	\
-	x(bi_parent_subvol,		32)	\
-	x(bi_nocow,			8)
-
-/* subset of BCH_INODE_FIELDS */
-#define BCH_INODE_OPTS()			\
-	x(data_checksum,		8)	\
-	x(compression,			8)	\
-	x(project,			32)	\
-	x(background_compression,	8)	\
-	x(data_replicas,		8)	\
-	x(promote_target,		16)	\
-	x(foreground_target,		16)	\
-	x(background_target,		16)	\
-	x(erasure_code,			16)	\
-	x(nocow,			8)
-
-enum inode_opt_id {
-#define x(name, ...)				\
-	Inode_opt_##name,
-	BCH_INODE_OPTS()
-#undef  x
-	Inode_opt_nr,
-};
-
-#define BCH_INODE_FLAGS()			\
-	x(sync,				0)	\
-	x(immutable,			1)	\
-	x(append,			2)	\
-	x(nodump,			3)	\
-	x(noatime,			4)	\
-	x(i_size_dirty,			5)	\
-	x(i_sectors_dirty,		6)	\
-	x(unlinked,			7)	\
-	x(backptr_untrusted,		8)
-
-/* bits 20+ reserved for packed fields below: */
-
-enum bch_inode_flags {
-#define x(t, n)	BCH_INODE_##t = 1U << n,
-	BCH_INODE_FLAGS()
-#undef x
-};
-
-enum __bch_inode_flags {
-#define x(t, n)	__BCH_INODE_##t = n,
-	BCH_INODE_FLAGS()
-#undef x
-};
-
-LE32_BITMASK(INODE_STR_HASH,	struct bch_inode, bi_flags, 20, 24);
-LE32_BITMASK(INODE_NR_FIELDS,	struct bch_inode, bi_flags, 24, 31);
-LE32_BITMASK(INODE_NEW_VARINT,	struct bch_inode, bi_flags, 31, 32);
-
-LE64_BITMASK(INODEv2_STR_HASH,	struct bch_inode_v2, bi_flags, 20, 24);
-LE64_BITMASK(INODEv2_NR_FIELDS,	struct bch_inode_v2, bi_flags, 24, 31);
-
-LE64_BITMASK(INODEv3_STR_HASH,	struct bch_inode_v3, bi_flags, 20, 24);
-LE64_BITMASK(INODEv3_NR_FIELDS,	struct bch_inode_v3, bi_flags, 24, 31);
-
-LE64_BITMASK(INODEv3_FIELDS_START,
-				struct bch_inode_v3, bi_flags, 31, 36);
-LE64_BITMASK(INODEv3_MODE,	struct bch_inode_v3, bi_flags, 36, 52);
-
-/* Dirents */
-
-/*
- * Dirents (and xattrs) have to implement string lookups; since our b-tree
- * doesn't support arbitrary length strings for the key, we instead index by a
- * 64 bit hash (currently truncated sha1) of the string, stored in the offset
- * field of the key - using linear probing to resolve hash collisions. This also
- * provides us with the readdir cookie posix requires.
- *
- * Linear probing requires us to use whiteouts for deletions, in the event of a
- * collision:
- */
-
-struct bch_dirent {
-	struct bch_val		v;
-
-	/* Target inode number: */
-	union {
-	__le64			d_inum;
-	struct {		/* DT_SUBVOL */
-	__le32			d_child_subvol;
-	__le32			d_parent_subvol;
-	};
-	};
-
-	/*
-	 * Copy of mode bits 12-15 from the target inode - so userspace can get
-	 * the filetype without having to do a stat()
-	 */
-	__u8			d_type;
-
-	__u8			d_name[];
-} __packed __aligned(8);
-
-#define DT_SUBVOL	16
-#define BCH_DT_MAX	17
-
-#define BCH_NAME_MAX	512
-
-/* Xattrs */
-
-#define KEY_TYPE_XATTR_INDEX_USER			0
-#define KEY_TYPE_XATTR_INDEX_POSIX_ACL_ACCESS	1
-#define KEY_TYPE_XATTR_INDEX_POSIX_ACL_DEFAULT	2
-#define KEY_TYPE_XATTR_INDEX_TRUSTED			3
-#define KEY_TYPE_XATTR_INDEX_SECURITY	        4
-
-struct bch_xattr {
-	struct bch_val		v;
-	__u8			x_type;
-	__u8			x_name_len;
-	__le16			x_val_len;
-	__u8			x_name[];
-} __packed __aligned(8);
-
-/* Bucket/allocation information: */
-
-struct bch_alloc {
-	struct bch_val		v;
-	__u8			fields;
-	__u8			gen;
-	__u8			data[];
-} __packed __aligned(8);
-
-#define BCH_ALLOC_FIELDS_V1()			\
-	x(read_time,		16)		\
-	x(write_time,		16)		\
-	x(data_type,		8)		\
-	x(dirty_sectors,	16)		\
-	x(cached_sectors,	16)		\
-	x(oldest_gen,		8)		\
-	x(stripe,		32)		\
-	x(stripe_redundancy,	8)
-
-enum {
-#define x(name, _bits) BCH_ALLOC_FIELD_V1_##name,
-	BCH_ALLOC_FIELDS_V1()
-#undef x
-};
-
-struct bch_alloc_v2 {
-	struct bch_val		v;
-	__u8			nr_fields;
-	__u8			gen;
-	__u8			oldest_gen;
-	__u8			data_type;
-	__u8			data[];
-} __packed __aligned(8);
-
-#define BCH_ALLOC_FIELDS_V2()			\
-	x(read_time,		64)		\
-	x(write_time,		64)		\
-	x(dirty_sectors,	32)		\
-	x(cached_sectors,	32)		\
-	x(stripe,		32)		\
-	x(stripe_redundancy,	8)
-
-struct bch_alloc_v3 {
-	struct bch_val		v;
-	__le64			journal_seq;
-	__le32			flags;
-	__u8			nr_fields;
-	__u8			gen;
-	__u8			oldest_gen;
-	__u8			data_type;
-	__u8			data[];
-} __packed __aligned(8);
-
-LE32_BITMASK(BCH_ALLOC_V3_NEED_DISCARD,struct bch_alloc_v3, flags,  0,  1)
-LE32_BITMASK(BCH_ALLOC_V3_NEED_INC_GEN,struct bch_alloc_v3, flags,  1,  2)
-
-struct bch_alloc_v4 {
-	struct bch_val		v;
-	__u64			journal_seq;
-	__u32			flags;
-	__u8			gen;
-	__u8			oldest_gen;
-	__u8			data_type;
-	__u8			stripe_redundancy;
-	__u32			dirty_sectors;
-	__u32			cached_sectors;
-	__u64			io_time[2];
-	__u32			stripe;
-	__u32			nr_external_backpointers;
-	__u64			fragmentation_lru;
-} __packed __aligned(8);
-
-#define BCH_ALLOC_V4_U64s_V0	6
-#define BCH_ALLOC_V4_U64s	(sizeof(struct bch_alloc_v4) / sizeof(__u64))
-
-BITMASK(BCH_ALLOC_V4_NEED_DISCARD,	struct bch_alloc_v4, flags,  0,  1)
-BITMASK(BCH_ALLOC_V4_NEED_INC_GEN,	struct bch_alloc_v4, flags,  1,  2)
-BITMASK(BCH_ALLOC_V4_BACKPOINTERS_START,struct bch_alloc_v4, flags,  2,  8)
-BITMASK(BCH_ALLOC_V4_NR_BACKPOINTERS,	struct bch_alloc_v4, flags,  8,  14)
-
-#define BCH_ALLOC_V4_NR_BACKPOINTERS_MAX	40
-
 struct bch_backpointer {
 	struct bch_val		v;
 	__u8			btree_id;
@@ -1014,154 +433,6 @@ struct bch_backpointer {
 	struct bpos		pos;
 } __packed __aligned(8);
 
-#define KEY_TYPE_BUCKET_GENS_BITS	8
-#define KEY_TYPE_BUCKET_GENS_NR		(1U << KEY_TYPE_BUCKET_GENS_BITS)
-#define KEY_TYPE_BUCKET_GENS_MASK	(KEY_TYPE_BUCKET_GENS_NR - 1)
-
-struct bch_bucket_gens {
-	struct bch_val		v;
-	u8			gens[KEY_TYPE_BUCKET_GENS_NR];
-} __packed __aligned(8);
-
-/* Quotas: */
-
-enum quota_types {
-	QTYP_USR		= 0,
-	QTYP_GRP		= 1,
-	QTYP_PRJ		= 2,
-	QTYP_NR			= 3,
-};
-
-enum quota_counters {
-	Q_SPC			= 0,
-	Q_INO			= 1,
-	Q_COUNTERS		= 2,
-};
-
-struct bch_quota_counter {
-	__le64			hardlimit;
-	__le64			softlimit;
-};
-
-struct bch_quota {
-	struct bch_val		v;
-	struct bch_quota_counter c[Q_COUNTERS];
-} __packed __aligned(8);
-
-/* Erasure coding */
-
-struct bch_stripe {
-	struct bch_val		v;
-	__le16			sectors;
-	__u8			algorithm;
-	__u8			nr_blocks;
-	__u8			nr_redundant;
-
-	__u8			csum_granularity_bits;
-	__u8			csum_type;
-	__u8			pad;
-
-	struct bch_extent_ptr	ptrs[];
-} __packed __aligned(8);
-
-/* Reflink: */
-
-struct bch_reflink_p {
-	struct bch_val		v;
-	__le64			idx;
-	/*
-	 * A reflink pointer might point to an indirect extent which is then
-	 * later split (by copygc or rebalance). If we only pointed to part of
-	 * the original indirect extent, and then one of the fragments is
-	 * outside the range we point to, we'd leak a refcount: so when creating
-	 * reflink pointers, we need to store pad values to remember the full
-	 * range we were taking a reference on.
-	 */
-	__le32			front_pad;
-	__le32			back_pad;
-} __packed __aligned(8);
-
-struct bch_reflink_v {
-	struct bch_val		v;
-	__le64			refcount;
-	union bch_extent_entry	start[0];
-	__u64			_data[];
-} __packed __aligned(8);
-
-struct bch_indirect_inline_data {
-	struct bch_val		v;
-	__le64			refcount;
-	u8			data[];
-};
-
-/* Inline data */
-
-struct bch_inline_data {
-	struct bch_val		v;
-	u8			data[];
-};
-
-/* Subvolumes: */
-
-#define SUBVOL_POS_MIN		POS(0, 1)
-#define SUBVOL_POS_MAX		POS(0, S32_MAX)
-#define BCACHEFS_ROOT_SUBVOL	1
-
-struct bch_subvolume {
-	struct bch_val		v;
-	__le32			flags;
-	__le32			snapshot;
-	__le64			inode;
-	/*
-	 * Snapshot subvolumes form a tree, separate from the snapshot nodes
-	 * tree - if this subvolume is a snapshot, this is the ID of the
-	 * subvolume it was created from:
-	 */
-	__le32			parent;
-	__le32			pad;
-	bch_le128		otime;
-};
-
-LE32_BITMASK(BCH_SUBVOLUME_RO,		struct bch_subvolume, flags,  0,  1)
-/*
- * We need to know whether a subvolume is a snapshot so we can know whether we
- * can delete it (or whether it should just be rm -rf'd)
- */
-LE32_BITMASK(BCH_SUBVOLUME_SNAP,	struct bch_subvolume, flags,  1,  2)
-LE32_BITMASK(BCH_SUBVOLUME_UNLINKED,	struct bch_subvolume, flags,  2,  3)
-
-/* Snapshots */
-
-struct bch_snapshot {
-	struct bch_val		v;
-	__le32			flags;
-	__le32			parent;
-	__le32			children[2];
-	__le32			subvol;
-	/* corresponds to a bch_snapshot_tree in BTREE_ID_snapshot_trees */
-	__le32			tree;
-	__le32			depth;
-	__le32			skip[3];
-};
-
-LE32_BITMASK(BCH_SNAPSHOT_DELETED,	struct bch_snapshot, flags,  0,  1)
-
-/* True if a subvolume points to this snapshot node: */
-LE32_BITMASK(BCH_SNAPSHOT_SUBVOL,	struct bch_snapshot, flags,  1,  2)
-
-/*
- * Snapshot trees:
- *
- * The snapshot_trees btree gives us persistent indentifier for each tree of
- * bch_snapshot nodes, and allow us to record and easily find the root/master
- * subvolume that other snapshots were created from:
- */
-struct bch_snapshot_tree {
-	struct bch_val		v;
-	__le32			master_subvol;
-	__le32			root_snapshot;
-};
-
 /* LRU btree: */
 
 struct bch_lru {
@@ -1171,33 +442,6 @@ struct bch_lru {
 
 #define LRU_ID_STRIPES		(1U << 16)
 
-/* Logged operations btree: */
-
-struct bch_logged_op_truncate {
-	struct bch_val		v;
-	__le32			subvol;
-	__le32			pad;
-	__le64			inum;
-	__le64			new_i_size;
-};
-
-enum logged_op_finsert_state {
-	LOGGED_OP_FINSERT_start,
-	LOGGED_OP_FINSERT_shift_extents,
-	LOGGED_OP_FINSERT_finish,
-};
-
-struct bch_logged_op_finsert {
-	struct bch_val		v;
-	__u8			state;
-	__u8			pad[3];
-	__le32			subvol;
-	__le64			inum;
-	__le64			dst_offset;
-	__le64			src_offset;
-	__le64			pos;
-};
-
 /* Optional/variable size superblock sections: */
 
 struct bch_sb_field {
@@ -1223,6 +467,19 @@ struct bch_sb_field {
 	x(ext,				13)	\
 	x(downgrade,			14)
 
+#include "alloc_background_format.h"
+#include "extents_format.h"
+#include "reflink_format.h"
+#include "ec_format.h"
+#include "inode_format.h"
+#include "dirent_format.h"
+#include "xattr_format.h"
+#include "quota_format.h"
+#include "logged_ops_format.h"
+#include "snapshot_format.h"
+#include "subvolume_format.h"
+#include "sb-counters_format.h"
+
 enum bch_sb_field_type {
 #define x(f, nr)	BCH_SB_FIELD_##f = nr,
 	BCH_SB_FIELDS()
@@ -1296,6 +553,7 @@ struct bch_member {
 	__le64			errors[BCH_MEMBER_ERROR_NR];
 	__le64			errors_at_reset[BCH_MEMBER_ERROR_NR];
 	__le64			errors_reset_time;
+	__le64			seq;
 };
 
 #define BCH_MEMBER_V1_BYTES	56
@@ -1442,7 +700,7 @@ struct bch_sb_field_replicas_v0 {
 	struct bch_replicas_entry_v0 entries[];
 } __packed __aligned(8);
 
-struct bch_replicas_entry {
+struct bch_replicas_entry_v1 {
 	__u8			data_type;
 	__u8			nr_devs;
 	__u8			nr_required;
@@ -1454,24 +712,7 @@ struct bch_replicas_entry {
 
 struct bch_sb_field_replicas {
 	struct bch_sb_field	field;
-	struct bch_replicas_entry entries[];
-} __packed __aligned(8);
-
-/* BCH_SB_FIELD_quota: */
-
-struct bch_sb_quota_counter {
-	__le32				timelimit;
-	__le32				warnlimit;
-};
-
-struct bch_sb_quota_type {
-	__le64				flags;
-	struct bch_sb_quota_counter	c[Q_COUNTERS];
-};
-
-struct bch_sb_field_quota {
-	struct bch_sb_field		field;
-	struct bch_sb_quota_type	q[QTYP_NR];
+	struct bch_replicas_entry_v1 entries[];
 } __packed __aligned(8);
 
 /* BCH_SB_FIELD_disk_groups: */
@@ -1492,99 +733,6 @@ struct bch_sb_field_disk_groups {
 	struct bch_disk_group	entries[];
 } __packed __aligned(8);
 
-/* BCH_SB_FIELD_counters */
-
-#define BCH_PERSISTENT_COUNTERS()				\
-	x(io_read,					0)	\
-	x(io_write,					1)	\
-	x(io_move,					2)	\
-	x(bucket_invalidate,				3)	\
-	x(bucket_discard,				4)	\
-	x(bucket_alloc,					5)	\
-	x(bucket_alloc_fail,				6)	\
-	x(btree_cache_scan,				7)	\
-	x(btree_cache_reap,				8)	\
-	x(btree_cache_cannibalize,			9)	\
-	x(btree_cache_cannibalize_lock,			10)	\
-	x(btree_cache_cannibalize_lock_fail,		11)	\
-	x(btree_cache_cannibalize_unlock,		12)	\
-	x(btree_node_write,				13)	\
-	x(btree_node_read,				14)	\
-	x(btree_node_compact,				15)	\
-	x(btree_node_merge,				16)	\
-	x(btree_node_split,				17)	\
-	x(btree_node_rewrite,				18)	\
-	x(btree_node_alloc,				19)	\
-	x(btree_node_free,				20)	\
-	x(btree_node_set_root,				21)	\
-	x(btree_path_relock_fail,			22)	\
-	x(btree_path_upgrade_fail,			23)	\
-	x(btree_reserve_get_fail,			24)	\
-	x(journal_entry_full,				25)	\
-	x(journal_full,					26)	\
-	x(journal_reclaim_finish,			27)	\
-	x(journal_reclaim_start,			28)	\
-	x(journal_write,				29)	\
-	x(read_promote,					30)	\
-	x(read_bounce,					31)	\
-	x(read_split,					33)	\
-	x(read_retry,					32)	\
-	x(read_reuse_race,				34)	\
-	x(move_extent_read,				35)	\
-	x(move_extent_write,				36)	\
-	x(move_extent_finish,				37)	\
-	x(move_extent_fail,				38)	\
-	x(move_extent_start_fail,			39)	\
-	x(copygc,					40)	\
-	x(copygc_wait,					41)	\
-	x(gc_gens_end,					42)	\
-	x(gc_gens_start,				43)	\
-	x(trans_blocked_journal_reclaim,		44)	\
-	x(trans_restart_btree_node_reused,		45)	\
-	x(trans_restart_btree_node_split,		46)	\
-	x(trans_restart_fault_inject,			47)	\
-	x(trans_restart_iter_upgrade,			48)	\
-	x(trans_restart_journal_preres_get,		49)	\
-	x(trans_restart_journal_reclaim,		50)	\
-	x(trans_restart_journal_res_get,		51)	\
-	x(trans_restart_key_cache_key_realloced,	52)	\
-	x(trans_restart_key_cache_raced,		53)	\
-	x(trans_restart_mark_replicas,			54)	\
-	x(trans_restart_mem_realloced,			55)	\
-	x(trans_restart_memory_allocation_failure,	56)	\
-	x(trans_restart_relock,				57)	\
-	x(trans_restart_relock_after_fill,		58)	\
-	x(trans_restart_relock_key_cache_fill,		59)	\
-	x(trans_restart_relock_next_node,		60)	\
-	x(trans_restart_relock_parent_for_fill,		61)	\
-	x(trans_restart_relock_path,			62)	\
-	x(trans_restart_relock_path_intent,		63)	\
-	x(trans_restart_too_many_iters,			64)	\
-	x(trans_restart_traverse,			65)	\
-	x(trans_restart_upgrade,			66)	\
-	x(trans_restart_would_deadlock,			67)	\
-	x(trans_restart_would_deadlock_write,		68)	\
-	x(trans_restart_injected,			69)	\
-	x(trans_restart_key_cache_upgrade,		70)	\
-	x(trans_traverse_all,				71)	\
-	x(transaction_commit,				72)	\
-	x(write_super,					73)	\
-	x(trans_restart_would_deadlock_recursion_limit,	74)	\
-	x(trans_restart_write_buffer_flush,		75)	\
-	x(trans_restart_split_race,			76)
-
-enum bch_persistent_counters {
-#define x(t, n, ...) BCH_COUNTER_##t,
-	BCH_PERSISTENT_COUNTERS()
-#undef x
-	BCH_COUNTER_NR
-};
-
-struct bch_sb_field_counters {
-	struct bch_sb_field	field;
-	__le64			d[];
-};
-
 /*
  * On clean shutdown, store btree roots and current journal sequence number in
  * the superblock:
@@ -1662,69 +810,41 @@ struct bch_sb_field_downgrade {
 #define BCH_VERSION_MINOR(_v)		((__u16) ((_v) & ~(~0U << 10)))
 #define BCH_VERSION(_major, _minor)	(((_major) << 10)|(_minor) << 0)
 
-#define RECOVERY_PASS_ALL_FSCK		(1ULL << 63)
-
 /*
  * field 1:		version name
  * field 2:		BCH_VERSION(major, minor)
  * field 3:		recovery passess required on upgrade
  */
 #define BCH_METADATA_VERSIONS()						\
-	x(bkey_renumber,		BCH_VERSION(0, 10),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(inode_btree_change,		BCH_VERSION(0, 11),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(snapshot,			BCH_VERSION(0, 12),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(inode_backpointers,		BCH_VERSION(0, 13),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(btree_ptr_sectors_written,	BCH_VERSION(0, 14),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(snapshot_2,			BCH_VERSION(0, 15),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_fs_upgrade_for_subvolumes)|		\
-	  BIT_ULL(BCH_RECOVERY_PASS_initialize_subvolumes)|		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(reflink_p_fix,		BCH_VERSION(0, 16),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_fix_reflink_p))			\
-	x(subvol_dirent,		BCH_VERSION(0, 17),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(inode_v2,			BCH_VERSION(0, 18),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(freespace,			BCH_VERSION(0, 19),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(alloc_v4,			BCH_VERSION(0, 20),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(new_data_types,		BCH_VERSION(0, 21),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(backpointers,			BCH_VERSION(0, 22),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(inode_v3,			BCH_VERSION(0, 23),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(unwritten_extents,		BCH_VERSION(0, 24),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(bucket_gens,			BCH_VERSION(0, 25),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_bucket_gens_init)|			\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(lru_v2,			BCH_VERSION(0, 26),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(fragmentation_lru,		BCH_VERSION(0, 27),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(no_bps_in_alloc_keys,		BCH_VERSION(0, 28),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(snapshot_trees,		BCH_VERSION(0, 29),		\
-	  RECOVERY_PASS_ALL_FSCK)					\
-	x(major_minor,			BCH_VERSION(1,  0),		\
-	  0)								\
-	x(snapshot_skiplists,		BCH_VERSION(1,  1),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_check_snapshots))			\
-	x(deleted_inodes,		BCH_VERSION(1,  2),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_check_inodes))			\
-	x(rebalance_work,		BCH_VERSION(1,  3),		\
-	  BIT_ULL(BCH_RECOVERY_PASS_set_fs_needs_rebalance))
+	x(bkey_renumber,		BCH_VERSION(0, 10))		\
+	x(inode_btree_change,		BCH_VERSION(0, 11))		\
+	x(snapshot,			BCH_VERSION(0, 12))		\
+	x(inode_backpointers,		BCH_VERSION(0, 13))		\
+	x(btree_ptr_sectors_written,	BCH_VERSION(0, 14))		\
+	x(snapshot_2,			BCH_VERSION(0, 15))		\
+	x(reflink_p_fix,		BCH_VERSION(0, 16))		\
+	x(subvol_dirent,		BCH_VERSION(0, 17))		\
+	x(inode_v2,			BCH_VERSION(0, 18))		\
+	x(freespace,			BCH_VERSION(0, 19))		\
+	x(alloc_v4,			BCH_VERSION(0, 20))		\
+	x(new_data_types,		BCH_VERSION(0, 21))		\
+	x(backpointers,			BCH_VERSION(0, 22))		\
+	x(inode_v3,			BCH_VERSION(0, 23))		\
+	x(unwritten_extents,		BCH_VERSION(0, 24))		\
+	x(bucket_gens,			BCH_VERSION(0, 25))		\
+	x(lru_v2,			BCH_VERSION(0, 26))		\
+	x(fragmentation_lru,		BCH_VERSION(0, 27))		\
+	x(no_bps_in_alloc_keys,		BCH_VERSION(0, 28))		\
+	x(snapshot_trees,		BCH_VERSION(0, 29))		\
+	x(major_minor,			BCH_VERSION(1,  0))		\
+	x(snapshot_skiplists,		BCH_VERSION(1,  1))		\
+	x(deleted_inodes,		BCH_VERSION(1,  2))		\
+	x(rebalance_work,		BCH_VERSION(1,  3))		\
+	x(member_seq,			BCH_VERSION(1,  4))
 
 enum bcachefs_metadata_version {
 	bcachefs_metadata_version_min = 9,
-#define x(t, n, upgrade_passes)	bcachefs_metadata_version_##t = n,
+#define x(t, n)	bcachefs_metadata_version_##t = n,
 	BCH_METADATA_VERSIONS()
 #undef x
 	bcachefs_metadata_version_max
@@ -1786,7 +906,8 @@ struct bch_sb {
 	__le32			time_base_hi;
 	__le32			time_precision;
 
-	__le64			flags[8];
+	__le64			flags[7];
+	__le64			write_time;
 	__le64			features[2];
 	__le64			compat[2];
 
@@ -2153,7 +1274,8 @@ static inline __u64 __bset_magic(struct bch_sb *sb)
 	x(clock,		7)		\
 	x(dev_usage,		8)		\
 	x(log,			9)		\
-	x(overwrite,		10)
+	x(overwrite,		10)		\
+	x(write_buffer_keys,	11)
 
 enum {
 #define x(f, nr)	BCH_JSET_ENTRY_##f	= nr,
@@ -2162,6 +1284,19 @@ enum {
 	BCH_JSET_ENTRY_NR
 };
 
+static inline bool jset_entry_is_key(struct jset_entry *e)
+{
+	switch (e->type) {
+	case BCH_JSET_ENTRY_btree_keys:
+	case BCH_JSET_ENTRY_btree_root:
+	case BCH_JSET_ENTRY_overwrite:
+	case BCH_JSET_ENTRY_write_buffer_keys:
+		return true;
+	}
+
+	return false;
+}
+
 /*
  * Journal sequence numbers can be blacklisted: bsets record the max sequence
  * number of all the journal entries they contain updates for, so that on
@@ -2203,7 +1338,7 @@ struct jset_entry_usage {
 struct jset_entry_data_usage {
 	struct jset_entry	entry;
 	__le64			v;
-	struct bch_replicas_entry r;
+	struct bch_replicas_entry_v1 r;
 } __packed;
 
 struct jset_entry_clock {
@@ -2224,8 +1359,8 @@ struct jset_entry_dev_usage {
 	__le32			dev;
 	__u32			pad;
 
-	__le64			buckets_ec;
-	__le64			_buckets_unavailable; /* No longer used */
+	__le64			_buckets_ec;		/* No longer used */
+	__le64			_buckets_unavailable;	/* No longer used */
 
 	struct jset_entry_dev_usage_type d[];
 };
@@ -2239,7 +1374,7 @@ static inline unsigned jset_entry_dev_usage_nr_types(struct jset_entry_dev_usage
 struct jset_entry_log {
 	struct jset_entry	entry;
 	u8			d[];
-} __packed;
+} __packed __aligned(8);
 
 /*
  * On disk format for a journal entry: