Update bcachefs sources to 710cd382bf bcachefs: Fix for leaking of reflinked extents

.bcachefs_revision

@@ -1 +1 @@
-07c2895cb3372c0e7b406ab13264de80c7ff19eb
+710cd382bf5f50ab8114a7cc22d78b5b2f574529

libbcachefs/bcachefs_format.h

@@ -913,18 +913,24 @@ struct bch_stripe {
 struct bch_reflink_p {
 	struct bch_val		v;
 	__le64			idx;
-
-	__le32			reservation_generation;
-	__u8			nr_replicas;
-	__u8			pad[3];
-};
+	/*
+	 * 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;
+} __attribute__((packed, aligned(8)));
 
 struct bch_reflink_v {
 	struct bch_val		v;
 	__le64			refcount;
 	union bch_extent_entry	start[0];
 	__u64			_data[0];
-};
+} __attribute__((packed, aligned(8)));
 
 struct bch_indirect_inline_data {
 	struct bch_val		v;

libbcachefs/btree_gc.c

@@ -1767,7 +1767,6 @@ static int bch2_gc_btree_gens(struct bch_fs *c, enum btree_id btree_id)
 			bch2_bkey_buf_reassemble(&sk, c, k);
 			bch2_extent_normalize(c, bkey_i_to_s(sk.k));
 
-
 			commit_err =
 				bch2_trans_update(&trans, &iter, sk.k, 0) ?:
 				bch2_trans_commit(&trans, NULL, NULL,

libbcachefs/buckets.c

@@ -1143,8 +1143,10 @@ static int bch2_mark_reflink_p(struct bch_fs *c,
 	struct bkey_s_c_reflink_p p = bkey_s_c_to_reflink_p(k);
 	struct reflink_gc *ref;
 	size_t l, r, m;
-	u64 idx = le64_to_cpu(p.v->idx);
-	unsigned sectors = p.k->size;
+	u64 idx = le64_to_cpu(p.v->idx) - le32_to_cpu(p.v->front_pad);
+	u64 sectors = (u64) le32_to_cpu(p.v->front_pad) +
+			    le32_to_cpu(p.v->back_pad) +
+			    p.k->size;
 	s64 ret = 0;
 
 	BUG_ON((flags & (BTREE_TRIGGER_INSERT|BTREE_TRIGGER_OVERWRITE)) ==
@@ -1720,12 +1722,33 @@ static int __bch2_trans_mark_reflink_p(struct btree_trans *trans,
 		bch2_fs_inconsistent(c,
 			"%llu:%llu len %u points to nonexistent indirect extent %llu",
 			p.k->p.inode, p.k->p.offset, p.k->size, idx);
-		bch2_inconsistent_error(c);
 		ret = -EIO;
 		goto err;
 	}
 
-	BUG_ON(!*refcount && (flags & BTREE_TRIGGER_OVERWRITE));
+	if (!*refcount && (flags & BTREE_TRIGGER_OVERWRITE)) {
+		bch2_fs_inconsistent(c,
+			"%llu:%llu len %u idx %llu indirect extent refcount underflow",
+			p.k->p.inode, p.k->p.offset, p.k->size, idx);
+		ret = -EIO;
+		goto err;
+	}
+
+	if (flags & BTREE_TRIGGER_INSERT) {
+		struct bch_reflink_p *v = (struct bch_reflink_p *) p.v;
+		u64 pad;
+
+		pad = max_t(s64, le32_to_cpu(v->front_pad),
+			    le64_to_cpu(v->idx) - bkey_start_offset(k.k));
+		BUG_ON(pad > U32_MAX);
+		v->front_pad = cpu_to_le32(pad);
+
+		pad = max_t(s64, le32_to_cpu(v->back_pad),
+			    k.k->p.offset - p.k->size - le64_to_cpu(v->idx));
+		BUG_ON(pad > U32_MAX);
+		v->back_pad = cpu_to_le32(pad);
+	}
+
 	le64_add_cpu(refcount, add);
 
 	if (!*refcount) {
@@ -1748,10 +1771,20 @@ static int bch2_trans_mark_reflink_p(struct btree_trans *trans,
 				     struct bkey_s_c k, unsigned flags)
 {
 	struct bkey_s_c_reflink_p p = bkey_s_c_to_reflink_p(k);
-	u64 idx = le64_to_cpu(p.v->idx);
-	unsigned sectors = p.k->size;
+	u64 idx, sectors;
 	s64 ret = 0;
 
+	if (flags & BTREE_TRIGGER_INSERT) {
+		struct bch_reflink_p *v = (struct bch_reflink_p *) p.v;
+
+		v->front_pad = v->back_pad = 0;
+	}
+
+	idx = le64_to_cpu(p.v->idx) - le32_to_cpu(p.v->front_pad);
+	sectors = (u64) le32_to_cpu(p.v->front_pad) +
+			le32_to_cpu(p.v->back_pad) +
+			p.k->size;
+
 	while (sectors) {
 		ret = __bch2_trans_mark_reflink_p(trans, p, idx, flags);
 		if (ret < 0)

libbcachefs/extents.c

@@ -480,7 +480,7 @@ restart_narrow_pointers:
 
 	bkey_for_each_ptr_decode(&k->k, ptrs, p, i)
 		if (can_narrow_crc(p.crc, n)) {
-			bch2_bkey_drop_ptr(bkey_i_to_s(k), &i->ptr);
+			__bch2_bkey_drop_ptr(bkey_i_to_s(k), &i->ptr);
 			p.ptr.offset += p.crc.offset;
 			p.crc = n;
 			bch2_extent_ptr_decoded_append(k, &p);
@@ -785,41 +785,85 @@ static union bch_extent_entry *extent_entry_prev(struct bkey_ptrs ptrs,
 	return i;
 }
 
-union bch_extent_entry *bch2_bkey_drop_ptr(struct bkey_s k,
+static void extent_entry_drop(struct bkey_s k, union bch_extent_entry *entry)
+{
+	union bch_extent_entry *next = extent_entry_next(entry);
+
+	/* stripes have ptrs, but their layout doesn't work with this code */
+	BUG_ON(k.k->type == KEY_TYPE_stripe);
+
+	memmove_u64s_down(entry, next,
+			  (u64 *) bkey_val_end(k) - (u64 *) next);
+	k.k->u64s -= (u64 *) next - (u64 *) entry;
+}
+
+/*
+ * Returns pointer to the next entry after the one being dropped:
+ */
+union bch_extent_entry *__bch2_bkey_drop_ptr(struct bkey_s k,
 					     struct bch_extent_ptr *ptr)
 {
 	struct bkey_ptrs ptrs = bch2_bkey_ptrs(k);
-	union bch_extent_entry *dst, *src, *prev;
+	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);
 
-	src = extent_entry_next(to_entry(ptr));
-	if (src != ptrs.end &&
-	    !extent_entry_is_crc(src))
+	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;
-
-	dst = to_entry(ptr);
-	while ((prev = extent_entry_prev(ptrs, dst))) {
-		if (extent_entry_is_ptr(prev))
-			break;
-
-		if (extent_entry_is_crc(prev)) {
-			if (drop_crc)
-				dst = prev;
 			break;
 		}
-
-		dst = prev;
 	}
 
-	memmove_u64s_down(dst, src,
-			  (u64 *) ptrs.end - (u64 *) src);
-	k.k->u64s -= (u64 *) src - (u64 *) dst;
+	extent_entry_drop(k, entry);
 
-	return dst;
+	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(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)
@@ -889,10 +933,6 @@ bool bch2_extent_normalize(struct bch_fs *c, struct bkey_s k)
 		ptr->cached &&
 		ptr_stale(bch_dev_bkey_exists(c, ptr->dev), ptr));
 
-	/* will only happen if all pointers were cached: */
-	if (!bch2_bkey_nr_ptrs(k.s_c))
-		k.k->type = KEY_TYPE_deleted;
-
 	return bkey_deleted(k.k);
 }
 

libbcachefs/extents.h

@@ -78,12 +78,12 @@ static inline size_t extent_entry_u64s(const union bch_extent_entry *entry)
 
 static inline bool extent_entry_is_ptr(const union bch_extent_entry *e)
 {
-	switch (extent_entry_type(e)) {
-	case BCH_EXTENT_ENTRY_ptr:
-		return true;
-	default:
-		return false;
+	return extent_entry_type(e) == BCH_EXTENT_ENTRY_ptr;
 }
+
+static inline bool extent_entry_is_stripe_ptr(const union bch_extent_entry *e)
+{
+	return extent_entry_type(e) == BCH_EXTENT_ENTRY_stripe_ptr;
 }
 
 static inline bool extent_entry_is_crc(const union bch_extent_entry *e)
@@ -578,6 +578,8 @@ void bch2_bkey_extent_entry_drop(struct bkey_i *, union bch_extent_entry *);
 void bch2_bkey_append_ptr(struct bkey_i *, struct bch_extent_ptr);
 void bch2_extent_ptr_decoded_append(struct bkey_i *,
 				    struct extent_ptr_decoded *);
+union bch_extent_entry *__bch2_bkey_drop_ptr(struct bkey_s,
+					     struct bch_extent_ptr *);
 union bch_extent_entry *bch2_bkey_drop_ptr(struct bkey_s,
 					   struct bch_extent_ptr *);
 

libbcachefs/io.c

@@ -782,6 +782,8 @@ static struct bio *bch2_write_bio_alloc(struct bch_fs *c,
 				       ? ((unsigned long) buf & (PAGE_SIZE - 1))
 				       : 0), PAGE_SIZE);
 
+	pages = min(pages, BIO_MAX_VECS);
+
 	bio = bio_alloc_bioset(GFP_NOIO, pages, &c->bio_write);
 	wbio			= wbio_init(bio);
 	wbio->put_bio		= true;

libbcachefs/migrate.c

@@ -73,6 +73,15 @@ static int __bch2_dev_usrdata_drop(struct bch_fs *c, unsigned dev_idx, int flags
 		 */
 		bch2_extent_normalize(c, bkey_i_to_s(sk.k));
 
+		/*
+		 * Since we're not inserting through an extent iterator
+		 * (BTREE_ITER_ALL_SNAPSHOTS iterators aren't extent iterators),
+		 * we aren't using the extent overwrite path to delete, we're
+		 * just using the normal key deletion path:
+		 */
+		if (bkey_deleted(&sk.k->k))
+			sk.k->k.size = 0;
+
 		ret   = bch2_btree_iter_traverse(&iter) ?:
 			bch2_trans_update(&trans, &iter, sk.k,
 					  BTREE_UPDATE_INTERNAL_SNAPSHOT_NODE) ?:

libbcachefs/move.c

@@ -196,7 +196,7 @@ static int bch2_migrate_index_update(struct bch_write_op *op)
 				extent_for_each_ptr(extent_i_to_s(new), new_ptr)
 					new_ptr->cached = true;
 
-			bch2_bkey_drop_ptr(bkey_i_to_s(insert), old_ptr);
+			__bch2_bkey_drop_ptr(bkey_i_to_s(insert), old_ptr);
 		}
 
 		extent_for_each_ptr_decode(extent_i_to_s(new), p, entry) {

libbcachefs/reflink.c

@@ -32,6 +32,9 @@ const char *bch2_reflink_p_invalid(const struct bch_fs *c, struct bkey_s_c k)
 	if (bkey_val_bytes(p.k) != sizeof(*p.v))
 		return "incorrect value size";
 
+	if (le64_to_cpu(p.v->idx) < le32_to_cpu(p.v->front_pad))
+		return "idx < front_pad";
+
 	return NULL;
 }
 

libbcachefs/util.c

@@ -525,7 +525,11 @@ int bch2_bio_alloc_pages(struct bio *bio, size_t size, gfp_t gfp_mask)
 		if (!page)
 			return -ENOMEM;
 
-		BUG_ON(!bio_add_page(bio, page, len, 0));
+		if (unlikely(!bio_add_page(bio, page, len, 0))) {
+			__free_page(page);
+			break;
+		}
+
 		size -= len;
 	}