Merge tag 'bcachefs-2024-10-22' of https://github.com/koverstreet/bca…

…chefs

Pull bcachefs fixes from Kent Overstreet:
 "Lots of hotfixes:

   - transaction restart injection has been shaking out a few things

   - fix a data corruption in the buffered write path on -ENOSPC, found
     by xfstests generic/299

   - Some small show_options fixes

   - Repair mismatches in inode hash type, seed: different snapshot
     versions of an inode must have the same hash/type seed, used for
     directory entries and xattrs. We were checking the hash seed, but
     not the type, and a user contributed a filesystem where the hash
     type on one inode had somehow been flipped; these fixes allow his
     filesystem to repair.

     Additionally, the hash type flip made some directory entries
     invisible, which were then recreated by userspace; so the hash
     check code now checks for duplicate non dangling dirents, and
     renames one of them if necessary.

   - Don't use wait_event_interruptible() in recovery: this fixes some
     filesystems failing to mount with -ERESTARTSYS

   - Workaround for kvmalloc not supporting > INT_MAX allocations,
     causing an -ENOMEM when allocating the sorted array of journal
     keys: this allows a 75 TB filesystem to mount

   - Make sure bch_inode_unpacked.bi_snapshot is set in the old inode
     compat path: this alllows Marcin's filesystem (in use since before
     6.7) to repair and mount"

* tag 'bcachefs-2024-10-22' of https://github.com/koverstreet/bcachefs: (26 commits)
  bcachefs: Set bch_inode_unpacked.bi_snapshot in old inode path
  bcachefs: Mark more errors as AUTOFIX
  bcachefs: Workaround for kvmalloc() not supporting > INT_MAX allocations
  bcachefs: Don't use wait_event_interruptible() in recovery
  bcachefs: Fix __bch2_fsck_err() warning
  bcachefs: fsck: Improve hash_check_key()
  bcachefs: bch2_hash_set_or_get_in_snapshot()
  bcachefs: Repair mismatches in inode hash seed, type
  bcachefs: Add hash seed, type to inode_to_text()
  bcachefs: INODE_STR_HASH() for bch_inode_unpacked
  bcachefs: Run in-kernel offline fsck without ratelimit errors
  bcachefs: skip mount option handle for empty string.
  bcachefs: fix incorrect show_options results
  bcachefs: Fix data corruption on -ENOSPC in buffered write path
  bcachefs: bch2_folio_reservation_get_partial() is now better behaved
  bcachefs: fix disk reservation accounting in bch2_folio_reservation_get()
  bcachefS: ec: fix data type on stripe deletion
  bcachefs: Don't use commit_do() unnecessarily
  bcachefs: handle restarts in bch2_bucket_io_time_reset()
  bcachefs: fix restart handling in __bch2_resume_logged_op_finsert()
  ...

fs/bcachefs/alloc_background.c

@@ -1977,7 +1977,7 @@ static void bch2_do_discards_fast_work(struct work_struct *work)
 					     ca->mi.bucket_size,
 					     GFP_KERNEL);
 
-		int ret = bch2_trans_do(c, NULL, NULL,
+		int ret = bch2_trans_commit_do(c, NULL, NULL,
 			BCH_WATERMARK_btree|
 			BCH_TRANS_COMMIT_no_enospc,
 			bch2_clear_bucket_needs_discard(trans, POS(ca->dev_idx, bucket)));
@@ -2137,14 +2137,15 @@ static void bch2_do_invalidates_work(struct work_struct *work)
 
 		struct bkey_s_c k = next_lru_key(trans, &iter, ca, &wrapped);
 		ret = bkey_err(k);
-		if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
-			continue;
 		if (ret)
-			break;
+			goto restart_err;
 		if (!k.k)
 			break;
 
 		ret = invalidate_one_bucket(trans, &iter, k, &nr_to_invalidate);
+restart_err:
+		if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
+			continue;
 		if (ret)
 			break;
 
@@ -2350,24 +2351,19 @@ int bch2_dev_remove_alloc(struct bch_fs *c, struct bch_dev *ca)
 
 /* Bucket IO clocks: */
 
-int bch2_bucket_io_time_reset(struct btree_trans *trans, unsigned dev,
-			      size_t bucket_nr, int rw)
+static int __bch2_bucket_io_time_reset(struct btree_trans *trans, unsigned dev,
+				size_t bucket_nr, int rw)
 {
 	struct bch_fs *c = trans->c;
-	struct btree_iter iter;
-	struct bkey_i_alloc_v4 *a;
-	u64 now;
-	int ret = 0;
-
-	if (bch2_trans_relock(trans))
-		bch2_trans_begin(trans);
 
-	a = bch2_trans_start_alloc_update_noupdate(trans, &iter, POS(dev, bucket_nr));
-	ret = PTR_ERR_OR_ZERO(a);
+	struct btree_iter iter;
+	struct bkey_i_alloc_v4 *a =
+		bch2_trans_start_alloc_update_noupdate(trans, &iter, POS(dev, bucket_nr));
+	int ret = PTR_ERR_OR_ZERO(a);
 	if (ret)
 		return ret;
 
-	now = bch2_current_io_time(c, rw);
+	u64 now = bch2_current_io_time(c, rw);
 	if (a->v.io_time[rw] == now)
 		goto out;
 
@@ -2380,6 +2376,15 @@ int bch2_bucket_io_time_reset(struct btree_trans *trans, unsigned dev,
 	return ret;
 }
 
+int bch2_bucket_io_time_reset(struct btree_trans *trans, unsigned dev,
+			      size_t bucket_nr, int rw)
+{
+	if (bch2_trans_relock(trans))
+		bch2_trans_begin(trans);
+
+	return nested_lockrestart_do(trans, __bch2_bucket_io_time_reset(trans, dev, bucket_nr, rw));
+}
+
 /* Startup/shutdown (ro/rw): */
 
 void bch2_recalc_capacity(struct bch_fs *c)

fs/bcachefs/alloc_foreground.c

@@ -684,7 +684,7 @@ struct open_bucket *bch2_bucket_alloc(struct bch_fs *c, struct bch_dev *ca,
 	struct bch_dev_usage usage;
 	struct open_bucket *ob;
 
-	bch2_trans_do(c, NULL, NULL, 0,
+	bch2_trans_do(c,
 		      PTR_ERR_OR_ZERO(ob = bch2_bucket_alloc_trans(trans, ca, watermark,
 							data_type, cl, false, &usage)));
 	return ob;

fs/bcachefs/btree_gc.c

@@ -820,12 +820,22 @@ static int bch2_alloc_write_key(struct btree_trans *trans,
 	 * fix that here:
 	 */
 	alloc_data_type_set(&gc, gc.data_type);
-
 	if (gc.data_type != old_gc.data_type ||
 	    gc.dirty_sectors != old_gc.dirty_sectors) {
 		ret = bch2_alloc_key_to_dev_counters(trans, ca, &old_gc, &gc, BTREE_TRIGGER_gc);
 		if (ret)
 			return ret;
+
+		/*
+		 * Ugly: alloc_key_to_dev_counters(..., BTREE_TRIGGER_gc) is not
+		 * safe w.r.t. transaction restarts, so fixup the gc_bucket so
+		 * we don't run it twice:
+		 */
+		percpu_down_read(&c->mark_lock);
+		struct bucket *gc_m = gc_bucket(ca, iter->pos.offset);
+		gc_m->data_type = gc.data_type;
+		gc_m->dirty_sectors = gc.dirty_sectors;
+		percpu_up_read(&c->mark_lock);
 	}
 
 	if (fsck_err_on(new.data_type != gc.data_type,

fs/bcachefs/btree_io.c

@@ -1871,7 +1871,7 @@ static void btree_node_write_work(struct work_struct *work)
 
 		}
 	} else {
-		ret = bch2_trans_do(c, NULL, NULL, 0,
+		ret = bch2_trans_do(c,
 			bch2_btree_node_update_key_get_iter(trans, b, &wbio->key,
 					BCH_WATERMARK_interior_updates|
 					BCH_TRANS_COMMIT_journal_reclaim|

fs/bcachefs/btree_iter.h

@@ -912,6 +912,8 @@ struct bkey_s_c bch2_btree_iter_peek_and_restart_outlined(struct btree_iter *);
 	_ret;								\
 })
 
+#define bch2_trans_do(_c, _do)	bch2_trans_run(_c, lockrestart_do(trans, _do))
+
 struct btree_trans *__bch2_trans_get(struct bch_fs *, unsigned);
 void bch2_trans_put(struct btree_trans *);
 

fs/bcachefs/btree_update.c

@@ -668,7 +668,7 @@ int bch2_btree_insert(struct bch_fs *c, enum btree_id id, struct bkey_i *k,
 		      struct disk_reservation *disk_res, int flags,
 		      enum btree_iter_update_trigger_flags iter_flags)
 {
-	return bch2_trans_do(c, disk_res, NULL, flags,
+	return bch2_trans_commit_do(c, disk_res, NULL, flags,
 			     bch2_btree_insert_trans(trans, id, k, iter_flags));
 }
 
@@ -865,7 +865,7 @@ __bch2_fs_log_msg(struct bch_fs *c, unsigned commit_flags, const char *fmt,
 		memcpy(l->d, buf.buf, buf.pos);
 		c->journal.early_journal_entries.nr += jset_u64s(u64s);
 	} else {
-		ret = bch2_trans_do(c, NULL, NULL,
+		ret = bch2_trans_commit_do(c, NULL, NULL,
 			BCH_TRANS_COMMIT_lazy_rw|commit_flags,
 			__bch2_trans_log_msg(trans, &buf, u64s));
 	}

fs/bcachefs/btree_update.h

@@ -192,7 +192,7 @@ static inline int bch2_trans_commit(struct btree_trans *trans,
 	nested_lockrestart_do(_trans, _do ?: bch2_trans_commit(_trans, (_disk_res),\
 					(_journal_seq), (_flags)))
 
-#define bch2_trans_do(_c, _disk_res, _journal_seq, _flags, _do)		\
+#define bch2_trans_commit_do(_c, _disk_res, _journal_seq, _flags, _do)		\
 	bch2_trans_run(_c, commit_do(trans, _disk_res, _journal_seq, _flags, _do))
 
 #define trans_for_each_update(_trans, _i)				\

fs/bcachefs/btree_update_interior.c

@@ -2239,10 +2239,8 @@ static void async_btree_node_rewrite_work(struct work_struct *work)
 	struct async_btree_rewrite *a =
 		container_of(work, struct async_btree_rewrite, work);
 	struct bch_fs *c = a->c;
-	int ret;
 
-	ret = bch2_trans_do(c, NULL, NULL, 0,
-		      async_btree_node_rewrite_trans(trans, a));
+	int ret = bch2_trans_do(c, async_btree_node_rewrite_trans(trans, a));
 	bch_err_fn_ratelimited(c, ret);
 	bch2_write_ref_put(c, BCH_WRITE_REF_node_rewrite);
 	kfree(a);

fs/bcachefs/buckets.c

@@ -1160,11 +1160,11 @@ int bch2_trans_mark_dev_sbs(struct bch_fs *c)
 #define SECTORS_CACHE	1024
 
 int __bch2_disk_reservation_add(struct bch_fs *c, struct disk_reservation *res,
-			      u64 sectors, int flags)
+				u64 sectors, enum bch_reservation_flags flags)
 {
 	struct bch_fs_pcpu *pcpu;
 	u64 old, get;
-	s64 sectors_available;
+	u64 sectors_available;
 	int ret;
 
 	percpu_down_read(&c->mark_lock);
@@ -1202,6 +1202,9 @@ int __bch2_disk_reservation_add(struct bch_fs *c, struct disk_reservation *res,
 	percpu_u64_set(&c->pcpu->sectors_available, 0);
 	sectors_available = avail_factor(__bch2_fs_usage_read_short(c).free);
 
+	if (sectors_available && (flags & BCH_DISK_RESERVATION_PARTIAL))
+		sectors = min(sectors, sectors_available);
+
 	if (sectors <= sectors_available ||
 	    (flags & BCH_DISK_RESERVATION_NOFAIL)) {
 		atomic64_set(&c->sectors_available,

fs/bcachefs/buckets.h

@@ -344,14 +344,16 @@ static inline void bch2_disk_reservation_put(struct bch_fs *c,
 	}
 }
 
-#define BCH_DISK_RESERVATION_NOFAIL		(1 << 0)
+enum bch_reservation_flags {
+	BCH_DISK_RESERVATION_NOFAIL	= 1 << 0,
+	BCH_DISK_RESERVATION_PARTIAL	= 1 << 1,
+};
 
-int __bch2_disk_reservation_add(struct bch_fs *,
-				struct disk_reservation *,
-				u64, int);
+int __bch2_disk_reservation_add(struct bch_fs *, struct disk_reservation *,
+				u64, enum bch_reservation_flags);
 
 static inline int bch2_disk_reservation_add(struct bch_fs *c, struct disk_reservation *res,
-					    u64 sectors, int flags)
+					    u64 sectors, enum bch_reservation_flags flags)
 {
 #ifdef __KERNEL__
 	u64 old, new;

fs/bcachefs/chardev.c

@@ -225,6 +225,7 @@ static long bch2_ioctl_fsck_offline(struct bch_ioctl_fsck_offline __user *user_a
 
 	opt_set(thr->opts, stdio, (u64)(unsigned long)&thr->thr.stdio);
 	opt_set(thr->opts, read_only, 1);
+	opt_set(thr->opts, ratelimit_errors, 0);
 
 	/* We need request_key() to be called before we punt to kthread: */
 	opt_set(thr->opts, nostart, true);

fs/bcachefs/darray.c

@@ -2,14 +2,27 @@
 
 #include <linux/log2.h>
 #include <linux/slab.h>
+#include <linux/vmalloc.h>
 #include "darray.h"
 
 int __bch2_darray_resize_noprof(darray_char *d, size_t element_size, size_t new_size, gfp_t gfp)
 {
 	if (new_size > d->size) {
 		new_size = roundup_pow_of_two(new_size);
 
-		void *data = kvmalloc_array_noprof(new_size, element_size, gfp);
+		/*
+		 * This is a workaround: kvmalloc() doesn't support > INT_MAX
+		 * allocations, but vmalloc() does.
+		 * The limit needs to be lifted from kvmalloc, and when it does
+		 * we'll go back to just using that.
+		 */
+		size_t bytes;
+		if (unlikely(check_mul_overflow(new_size, element_size, &bytes)))
+			return -ENOMEM;
+
+		void *data = likely(bytes < INT_MAX)
+			? kvmalloc_noprof(bytes, gfp)
+			: vmalloc_noprof(bytes);
 		if (!data)
 			return -ENOMEM;
 

fs/bcachefs/dirent.c

@@ -250,13 +250,6 @@ int bch2_dirent_create(struct btree_trans *trans, subvol_inum dir,
 	return ret;
 }
 
-static void dirent_copy_target(struct bkey_i_dirent *dst,
-			       struct bkey_s_c_dirent src)
-{
-	dst->v.d_inum = src.v->d_inum;
-	dst->v.d_type = src.v->d_type;
-}
-
 int bch2_dirent_read_target(struct btree_trans *trans, subvol_inum dir,
 			    struct bkey_s_c_dirent d, subvol_inum *target)
 {

fs/bcachefs/dirent.h

@@ -34,6 +34,13 @@ static inline unsigned dirent_val_u64s(unsigned len)
 int bch2_dirent_read_target(struct btree_trans *, subvol_inum,
 			    struct bkey_s_c_dirent, subvol_inum *);
 
+static inline void dirent_copy_target(struct bkey_i_dirent *dst,
+				      struct bkey_s_c_dirent src)
+{
+	dst->v.d_inum = src.v->d_inum;
+	dst->v.d_type = src.v->d_type;
+}
+
 int bch2_dirent_create_snapshot(struct btree_trans *, u32, u64, u32,
 			const struct bch_hash_info *, u8,
 			const struct qstr *, u64, u64 *,

fs/bcachefs/disk_accounting.c

@@ -856,8 +856,10 @@ int bch2_dev_usage_init(struct bch_dev *ca, bool gc)
 	};
 	u64 v[3] = { ca->mi.nbuckets - ca->mi.first_bucket, 0, 0 };
 
-	int ret = bch2_trans_do(c, NULL, NULL, 0,
-			bch2_disk_accounting_mod(trans, &acc, v, ARRAY_SIZE(v), gc));
+	int ret = bch2_trans_do(c, ({
+		bch2_disk_accounting_mod(trans, &acc, v, ARRAY_SIZE(v), gc) ?:
+		(!gc ? bch2_trans_commit(trans, NULL, NULL, 0) : 0);
+	}));
 	bch_err_fn(c, ret);
 	return ret;
 }

fs/bcachefs/ec.c

@@ -266,12 +266,12 @@ static int __mark_stripe_bucket(struct btree_trans *trans,
 	if (!deleting) {
 		a->stripe		= s.k->p.offset;
 		a->stripe_redundancy	= s.v->nr_redundant;
+		alloc_data_type_set(a, data_type);
 	} else {
 		a->stripe		= 0;
 		a->stripe_redundancy	= 0;
+		alloc_data_type_set(a, BCH_DATA_user);
 	}
-
-	alloc_data_type_set(a, data_type);
 err:
 	printbuf_exit(&buf);
 	return ret;
@@ -1186,7 +1186,7 @@ static void ec_stripe_delete_work(struct work_struct *work)
 		if (!idx)
 			break;
 
-		int ret = bch2_trans_do(c, NULL, NULL, BCH_TRANS_COMMIT_no_enospc,
+		int ret = bch2_trans_commit_do(c, NULL, NULL, BCH_TRANS_COMMIT_no_enospc,
 					ec_stripe_delete(trans, idx));
 		bch_err_fn(c, ret);
 		if (ret)
@@ -1519,14 +1519,14 @@ static void ec_stripe_create(struct ec_stripe_new *s)
 		goto err;
 	}
 
-	ret = bch2_trans_do(c, &s->res, NULL,
-			    BCH_TRANS_COMMIT_no_check_rw|
-			    BCH_TRANS_COMMIT_no_enospc,
-			    ec_stripe_key_update(trans,
-					s->have_existing_stripe
-					? bkey_i_to_stripe(&s->existing_stripe.key)
-					: NULL,
-					bkey_i_to_stripe(&s->new_stripe.key)));
+	ret = bch2_trans_commit_do(c, &s->res, NULL,
+		BCH_TRANS_COMMIT_no_check_rw|
+		BCH_TRANS_COMMIT_no_enospc,
+		ec_stripe_key_update(trans,
+				     s->have_existing_stripe
+				     ? bkey_i_to_stripe(&s->existing_stripe.key)
+				     : NULL,
+				     bkey_i_to_stripe(&s->new_stripe.key)));
 	bch_err_msg(c, ret, "creating stripe key");
 	if (ret) {
 		goto err;

fs/bcachefs/error.c

@@ -251,7 +251,10 @@ int __bch2_fsck_err(struct bch_fs *c,
 	 *   delete the key)
 	 * - and we don't need to warn if we're not prompting
 	 */
-	WARN_ON(!(flags & FSCK_AUTOFIX) && !trans && bch2_current_has_btree_trans(c));
+	WARN_ON((flags & FSCK_CAN_FIX) &&
+		!(flags & FSCK_AUTOFIX) &&
+		!trans &&
+		bch2_current_has_btree_trans(c));
 
 	if ((flags & FSCK_CAN_FIX) &&
 	    test_bit(err, c->sb.errors_silent))

fs/bcachefs/fs-io-buffered.c

@@ -856,6 +856,12 @@ static int __bch2_buffered_write(struct bch_inode_info *inode,
 				folios_trunc(&fs, fi);
 				end = min(end, folio_end_pos(darray_last(fs)));
 			} else {
+				if (!folio_test_uptodate(f)) {
+					ret = bch2_read_single_folio(f, mapping);
+					if (ret)
+						goto out;
+				}
+
 				folios_trunc(&fs, fi + 1);
 				end = f_pos + f_reserved;
 			}