Merge tag 'bcachefs-2025-04-24' of git://evilpiepirate.org/bcachefs

Pull bcachefs fixes from Kent Overstreet:

 - Case insensitive directories now work

 - Ciemap now correctly reports on unwritten pagecache data

 - bcachefs tools 1.25.1 was incorrectly picking unaligned bucket sizes;
   fix journal and write path bugs this uncovered

And assorted smaller fixes...

* tag 'bcachefs-2025-04-24' of git://evilpiepirate.org/bcachefs: (24 commits)
  bcachefs: Rework fiemap transaction restart handling
  bcachefs: add fiemap delalloc extent detection
  bcachefs: refactor fiemap processing into extent helper and struct
  bcachefs: track current fiemap offset in start variable
  bcachefs: drop duplicate fiemap sync flag
  bcachefs: Fix btree_iter_peek_prev() at end of inode
  bcachefs: Make btree_iter_peek_prev() assert more precise
  bcachefs: Unit test fixes
  bcachefs: Print mount opts earlier
  bcachefs: unlink: casefold d_invalidate
  bcachefs: Fix casefold lookups
  bcachefs: Casefold is now a regular opts.h option
  bcachefs: Implement fileattr_(get|set)
  bcachefs: Allocator now copes with unaligned buckets
  bcachefs: Start copygc, rebalance threads earlier
  bcachefs: Refactor bch2_run_recovery_passes()
  bcachefs: bch2_copygc_wakeup()
  bcachefs: Fix ref leak in write_super()
  bcachefs: Change __journal_entry_close() assert to ERO
  bcachefs: Ensure journal space is block size aligned
  ...

Author: torvalds

fs/bcachefs/alloc_foreground.c

@@ -1425,6 +1425,8 @@ int bch2_alloc_sectors_start_trans(struct btree_trans *trans,
 	open_bucket_for_each(c, &wp->ptrs, ob, i)
 		wp->sectors_free = min(wp->sectors_free, ob->sectors_free);
 
+	wp->sectors_free = rounddown(wp->sectors_free, block_sectors(c));
+
 	BUG_ON(!wp->sectors_free || wp->sectors_free == UINT_MAX);
 
 	return 0;

fs/bcachefs/alloc_foreground.h

@@ -110,7 +110,9 @@ static inline void bch2_alloc_sectors_done_inlined(struct bch_fs *c, struct writ
 	unsigned i;
 
 	open_bucket_for_each(c, &wp->ptrs, ob, i)
-		ob_push(c, !ob->sectors_free ? &ptrs : &keep, ob);
+		ob_push(c, ob->sectors_free < block_sectors(c)
+			? &ptrs
+			: &keep, ob);
 	wp->ptrs = keep;
 
 	mutex_unlock(&wp->lock);

fs/bcachefs/bcachefs_format.h

@@ -366,6 +366,10 @@ static inline void bkey_init(struct bkey *k)
 #define __BKEY_PADDED(key, pad)					\
 	struct bkey_i key; __u64 key ## _pad[pad]
 
+enum bch_bkey_type_flags {
+	BKEY_TYPE_strict_btree_checks	= BIT(0),
+};
+
 /*
  * - DELETED keys are used internally to mark keys that should be ignored but
  *   override keys in composition order.  Their version number is ignored.
@@ -383,46 +387,46 @@ static inline void bkey_init(struct bkey *k)
  *
  * - WHITEOUT: for hash table btrees
  */
-#define BCH_BKEY_TYPES()				\
-	x(deleted,		0)			\
-	x(whiteout,		1)			\
-	x(error,		2)			\
-	x(cookie,		3)			\
-	x(hash_whiteout,	4)			\
-	x(btree_ptr,		5)			\
-	x(extent,		6)			\
-	x(reservation,		7)			\
-	x(inode,		8)			\
-	x(inode_generation,	9)			\
-	x(dirent,		10)			\
-	x(xattr,		11)			\
-	x(alloc,		12)			\
-	x(quota,		13)			\
-	x(stripe,		14)			\
-	x(reflink_p,		15)			\
-	x(reflink_v,		16)			\
-	x(inline_data,		17)			\
-	x(btree_ptr_v2,		18)			\
-	x(indirect_inline_data,	19)			\
-	x(alloc_v2,		20)			\
-	x(subvolume,		21)			\
-	x(snapshot,		22)			\
-	x(inode_v2,		23)			\
-	x(alloc_v3,		24)			\
-	x(set,			25)			\
-	x(lru,			26)			\
-	x(alloc_v4,		27)			\
-	x(backpointer,		28)			\
-	x(inode_v3,		29)			\
-	x(bucket_gens,		30)			\
-	x(snapshot_tree,	31)			\
-	x(logged_op_truncate,	32)			\
-	x(logged_op_finsert,	33)			\
-	x(accounting,		34)			\
-	x(inode_alloc_cursor,	35)
+#define BCH_BKEY_TYPES()						\
+	x(deleted,		0,	0)				\
+	x(whiteout,		1,	0)				\
+	x(error,		2,	0)				\
+	x(cookie,		3,	0)				\
+	x(hash_whiteout,	4,	BKEY_TYPE_strict_btree_checks)	\
+	x(btree_ptr,		5,	BKEY_TYPE_strict_btree_checks)	\
+	x(extent,		6,	BKEY_TYPE_strict_btree_checks)	\
+	x(reservation,		7,	BKEY_TYPE_strict_btree_checks)	\
+	x(inode,		8,	BKEY_TYPE_strict_btree_checks)	\
+	x(inode_generation,	9,	BKEY_TYPE_strict_btree_checks)	\
+	x(dirent,		10,	BKEY_TYPE_strict_btree_checks)	\
+	x(xattr,		11,	BKEY_TYPE_strict_btree_checks)	\
+	x(alloc,		12,	BKEY_TYPE_strict_btree_checks)	\
+	x(quota,		13,	BKEY_TYPE_strict_btree_checks)	\
+	x(stripe,		14,	BKEY_TYPE_strict_btree_checks)	\
+	x(reflink_p,		15,	BKEY_TYPE_strict_btree_checks)	\
+	x(reflink_v,		16,	BKEY_TYPE_strict_btree_checks)	\
+	x(inline_data,		17,	BKEY_TYPE_strict_btree_checks)	\
+	x(btree_ptr_v2,		18,	BKEY_TYPE_strict_btree_checks)	\
+	x(indirect_inline_data,	19,	BKEY_TYPE_strict_btree_checks)	\
+	x(alloc_v2,		20,	BKEY_TYPE_strict_btree_checks)	\
+	x(subvolume,		21,	BKEY_TYPE_strict_btree_checks)	\
+	x(snapshot,		22,	BKEY_TYPE_strict_btree_checks)	\
+	x(inode_v2,		23,	BKEY_TYPE_strict_btree_checks)	\
+	x(alloc_v3,		24,	BKEY_TYPE_strict_btree_checks)	\
+	x(set,			25,	0)				\
+	x(lru,			26,	BKEY_TYPE_strict_btree_checks)	\
+	x(alloc_v4,		27,	BKEY_TYPE_strict_btree_checks)	\
+	x(backpointer,		28,	BKEY_TYPE_strict_btree_checks)	\
+	x(inode_v3,		29,	BKEY_TYPE_strict_btree_checks)	\
+	x(bucket_gens,		30,	BKEY_TYPE_strict_btree_checks)	\
+	x(snapshot_tree,	31,	BKEY_TYPE_strict_btree_checks)	\
+	x(logged_op_truncate,	32,	BKEY_TYPE_strict_btree_checks)	\
+	x(logged_op_finsert,	33,	BKEY_TYPE_strict_btree_checks)	\
+	x(accounting,		34,	BKEY_TYPE_strict_btree_checks)	\
+	x(inode_alloc_cursor,	35,	BKEY_TYPE_strict_btree_checks)
 
 enum bch_bkey_type {
-#define x(name, nr) KEY_TYPE_##name	= nr,
+#define x(name, nr, ...) KEY_TYPE_##name	= nr,
 	BCH_BKEY_TYPES()
 #undef x
 	KEY_TYPE_MAX,
@@ -863,6 +867,7 @@ LE64_BITMASK(BCH_SB_VERSION_INCOMPAT_ALLOWED,
 LE64_BITMASK(BCH_SB_SHARD_INUMS_NBITS,	struct bch_sb, flags[6],  0,  4);
 LE64_BITMASK(BCH_SB_WRITE_ERROR_TIMEOUT,struct bch_sb, flags[6],  4, 14);
 LE64_BITMASK(BCH_SB_CSUM_ERR_RETRY_NR,	struct bch_sb, flags[6], 14, 20);
+LE64_BITMASK(BCH_SB_CASEFOLD,		struct bch_sb, flags[6], 22, 23);
 
 static inline __u64 BCH_SB_COMPRESSION_TYPE(const struct bch_sb *sb)
 {

fs/bcachefs/bkey_methods.c

@@ -21,7 +21,7 @@
 #include "xattr.h"
 
 const char * const bch2_bkey_types[] = {
-#define x(name, nr) #name,
+#define x(name, nr, ...) #name,
 	BCH_BKEY_TYPES()
 #undef x
 	NULL
@@ -115,7 +115,7 @@ static bool key_type_set_merge(struct bch_fs *c, struct bkey_s l, struct bkey_s_
 })
 
 const struct bkey_ops bch2_bkey_ops[] = {
-#define x(name, nr) [KEY_TYPE_##name]	= bch2_bkey_ops_##name,
+#define x(name, nr, ...) [KEY_TYPE_##name]	= bch2_bkey_ops_##name,
 	BCH_BKEY_TYPES()
 #undef x
 };
@@ -155,6 +155,12 @@ static u64 bch2_key_types_allowed[] = {
 #undef x
 };
 
+static const enum bch_bkey_type_flags bch2_bkey_type_flags[] = {
+#define x(name, nr, flags)	[KEY_TYPE_##name] = flags,
+	BCH_BKEY_TYPES()
+#undef x
+};
+
 const char *bch2_btree_node_type_str(enum btree_node_type type)
 {
 	return type == BKEY_TYPE_btree ? "internal btree node" : bch2_btree_id_str(type - 1);
@@ -177,8 +183,18 @@ int __bch2_bkey_validate(struct bch_fs *c, struct bkey_s_c k,
 	if (type >= BKEY_TYPE_NR)
 		return 0;
 
-	bkey_fsck_err_on(k.k->type < KEY_TYPE_MAX &&
-			 (type == BKEY_TYPE_btree || (from.flags & BCH_VALIDATE_commit)) &&
+	enum bch_bkey_type_flags bkey_flags = k.k->type < KEY_TYPE_MAX
+		? bch2_bkey_type_flags[k.k->type]
+		: 0;
+
+	bool strict_key_type_allowed =
+		(from.flags & BCH_VALIDATE_commit) ||
+		type == BKEY_TYPE_btree ||
+		(from.btree < BTREE_ID_NR &&
+		 (bkey_flags & BKEY_TYPE_strict_btree_checks));
+
+	bkey_fsck_err_on(strict_key_type_allowed &&
+			 k.k->type < KEY_TYPE_MAX &&
 			 !(bch2_key_types_allowed[type] & BIT_ULL(k.k->type)),
 			 c, bkey_invalid_type_for_btree,
 			 "invalid key type for btree %s (%s)",

fs/bcachefs/btree_iter.c

@@ -2577,7 +2577,10 @@ struct bkey_s_c bch2_btree_iter_peek_prev_min(struct btree_trans *trans, struct
 					      struct bpos end)
 {
 	if ((iter->flags & (BTREE_ITER_is_extents|BTREE_ITER_filter_snapshots)) &&
-	   !bkey_eq(iter->pos, POS_MAX)) {
+	   !bkey_eq(iter->pos, POS_MAX) &&
+	   !((iter->flags & BTREE_ITER_is_extents) &&
+	     iter->pos.offset == U64_MAX)) {
+
 		/*
 		 * bkey_start_pos(), for extents, is not monotonically
 		 * increasing until after filtering for snapshots:
@@ -2602,7 +2605,7 @@ struct bkey_s_c bch2_btree_iter_peek_prev_min(struct btree_trans *trans, struct
 
 	bch2_trans_verify_not_unlocked_or_in_restart(trans);
 	bch2_btree_iter_verify_entry_exit(iter);
-	EBUG_ON((iter->flags & BTREE_ITER_filter_snapshots) && bpos_eq(end, POS_MIN));
+	EBUG_ON((iter->flags & BTREE_ITER_filter_snapshots) && iter->pos.inode != end.inode);
 
 	int ret = trans_maybe_inject_restart(trans, _RET_IP_);
 	if (unlikely(ret)) {

fs/bcachefs/dirent.c

@@ -13,8 +13,8 @@
 
 #include <linux/dcache.h>
 
-static int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *info,
-				const struct qstr *str, struct qstr *out_cf)
+int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *info,
+		  const struct qstr *str, struct qstr *out_cf)
 {
 	*out_cf = (struct qstr) QSTR_INIT(NULL, 0);
 
@@ -35,18 +35,6 @@ static int bch2_casefold(struct btree_trans *trans, const struct bch_hash_info *
 #endif
 }
 
-static inline int bch2_maybe_casefold(struct btree_trans *trans,
-				      const struct bch_hash_info *info,
-				      const struct qstr *str, struct qstr *out_cf)
-{
-	if (likely(!info->cf_encoding)) {
-		*out_cf = *str;
-		return 0;
-	} else {
-		return bch2_casefold(trans, info, str, out_cf);
-	}
-}
-
 static unsigned bch2_dirent_name_bytes(struct bkey_s_c_dirent d)
 {
 	if (bkey_val_bytes(d.k) < offsetof(struct bch_dirent, d_name))

fs/bcachefs/dirent.h

@@ -23,6 +23,21 @@ struct bch_fs;
 struct bch_hash_info;
 struct bch_inode_info;
 
+int bch2_casefold(struct btree_trans *, const struct bch_hash_info *,
+		  const struct qstr *, struct qstr *);
+
+static inline int bch2_maybe_casefold(struct btree_trans *trans,
+				      const struct bch_hash_info *info,
+				      const struct qstr *str, struct qstr *out_cf)
+{
+	if (likely(!info->cf_encoding)) {
+		*out_cf = *str;
+		return 0;
+	} else {
+		return bch2_casefold(trans, info, str, out_cf);
+	}
+}
+
 struct qstr bch2_dirent_get_name(struct bkey_s_c_dirent d);
 
 static inline unsigned dirent_val_u64s(unsigned len, unsigned cf_len)

fs/bcachefs/error.c

@@ -272,9 +272,6 @@ static struct fsck_err_state *fsck_err_get(struct bch_fs *c,
 {
 	struct fsck_err_state *s;
 
-	if (!test_bit(BCH_FS_fsck_running, &c->flags))
-		return NULL;
-
 	list_for_each_entry(s, &c->fsck_error_msgs, list)
 		if (s->id == id) {
 			/*
@@ -639,14 +636,14 @@ int __bch2_bkey_fsck_err(struct bch_fs *c,
 	return ret;
 }
 
-void bch2_flush_fsck_errs(struct bch_fs *c)
+static void __bch2_flush_fsck_errs(struct bch_fs *c, bool print)
 {
 	struct fsck_err_state *s, *n;
 
 	mutex_lock(&c->fsck_error_msgs_lock);
 
 	list_for_each_entry_safe(s, n, &c->fsck_error_msgs, list) {
-		if (s->ratelimited && s->last_msg)
+		if (print && s->ratelimited && s->last_msg)
 			bch_err(c, "Saw %llu errors like:\n  %s", s->nr, s->last_msg);
 
 		list_del(&s->list);
@@ -657,6 +654,16 @@ void bch2_flush_fsck_errs(struct bch_fs *c)
 	mutex_unlock(&c->fsck_error_msgs_lock);
 }
 
+void bch2_flush_fsck_errs(struct bch_fs *c)
+{
+	__bch2_flush_fsck_errs(c, true);
+}
+
+void bch2_free_fsck_errs(struct bch_fs *c)
+{
+	__bch2_flush_fsck_errs(c, false);
+}
+
 int bch2_inum_offset_err_msg_trans(struct btree_trans *trans, struct printbuf *out,
 				    subvol_inum inum, u64 offset)
 {

fs/bcachefs/error.h

@@ -93,6 +93,7 @@ int __bch2_fsck_err(struct bch_fs *, struct btree_trans *,
 			_flags, BCH_FSCK_ERR_##_err_type, __VA_ARGS__)
 
 void bch2_flush_fsck_errs(struct bch_fs *);
+void bch2_free_fsck_errs(struct bch_fs *);
 
 #define fsck_err_wrap(_do)						\
 ({									\