Merge with e5ebdf1e41 bcachefs: Add a few tracepoints

Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>

18 files changed, 141 insertions, 363 deletions

diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h
index 289d7ae4e98c..fa9593764f0c 100644
--- a/fs/bcachefs/bcachefs.h
+++ b/fs/bcachefs/bcachefs.h
@@ -732,7 +732,7 @@ struct bch_fs {
 	ZSTD_parameters		zstd_params;
 
 	struct crypto_shash	*sha256;
-	struct crypto_skcipher	*chacha20;
+	struct crypto_sync_skcipher *chacha20;
 	struct crypto_shash	*poly1305;
 
 	atomic64_t		key_version;
diff --git a/fs/bcachefs/btree_cache.h b/fs/bcachefs/btree_cache.h
index 132cc95a4c02..98cca30778ea 100644
--- a/fs/bcachefs/btree_cache.h
+++ b/fs/bcachefs/btree_cache.h
@@ -94,7 +94,7 @@ static inline unsigned btree_blocks(struct bch_fs *c)
 	return c->opts.btree_node_size >> c->block_bits;
 }
 
-#define BTREE_SPLIT_THRESHOLD(c)		(btree_max_u64s(c) * 3 / 4)
+#define BTREE_SPLIT_THRESHOLD(c)		(btree_max_u64s(c) * 2 / 3)
 
 #define BTREE_FOREGROUND_MERGE_THRESHOLD(c)	(btree_max_u64s(c) * 1 / 3)
 #define BTREE_FOREGROUND_MERGE_HYSTERESIS(c)			\
diff --git a/fs/bcachefs/btree_io.c b/fs/bcachefs/btree_io.c
index 61d4c98d3476..ac8b98861aae 100644
--- a/fs/bcachefs/btree_io.c
+++ b/fs/bcachefs/btree_io.c
@@ -620,7 +620,7 @@ static void bset_encrypt(struct bch_fs *c, struct bset *i, unsigned offset)
 		bch2_encrypt(c, BSET_CSUM_TYPE(i), nonce, &bn->flags,
 			     bytes);
 
-		nonce = nonce_add(nonce, round_up(bytes, CHACHA20_BLOCK_SIZE));
+		nonce = nonce_add(nonce, round_up(bytes, CHACHA_BLOCK_SIZE));
 	}
 
 	bch2_encrypt(c, BSET_CSUM_TYPE(i), nonce, i->_data,
diff --git a/fs/bcachefs/btree_update_interior.c b/fs/bcachefs/btree_update_interior.c
index 82b66a667e35..98782d55b846 100644
--- a/fs/bcachefs/btree_update_interior.c
+++ b/fs/bcachefs/btree_update_interior.c
@@ -609,33 +609,11 @@ static void bch2_btree_update_free(struct btree_update *as)
 	mutex_unlock(&c->btree_interior_update_lock);
 }
 
-static void btree_update_nodes_reachable(struct btree_update *as, u64 seq)
-{
-	struct bch_fs *c = as->c;
-
-	while (as->nr_new_nodes) {
-		struct btree *b = as->new_nodes[--as->nr_new_nodes];
-
-		BUG_ON(b->will_make_reachable != (unsigned long) as);
-		b->will_make_reachable = 0;
-
-		/*
-		 * b->will_make_reachable prevented it from being written, so
-		 * write it now if it needs to be written:
-		 */
-		btree_node_lock_type(c, b, SIX_LOCK_read);
-		bch2_btree_node_write_cond(c, b, btree_node_need_write(b));
-		six_unlock_read(&b->lock);
-	}
-
-	while (as->nr_pending)
-		bch2_btree_node_free_ondisk(c, &as->pending[--as->nr_pending],
-					    seq);
-}
-
 static void btree_update_nodes_written(struct closure *cl)
 {
 	struct btree_update *as = container_of(cl, struct btree_update, cl);
+	struct btree *new_nodes[BTREE_MAX_DEPTH * 2 + GC_MERGE_NODES];
+	unsigned nr_new_nodes;
 	struct journal_res res = { 0 };
 	struct bch_fs *c = as->c;
 	struct btree *b;
@@ -650,6 +628,7 @@ static void btree_update_nodes_written(struct closure *cl)
 	mutex_lock(&c->btree_interior_update_lock);
 	as->nodes_written = true;
 again:
+	nr_new_nodes = 0;
 	as = list_first_entry_or_null(&c->btree_interior_updates_unwritten,
 				      struct btree_update, unwritten_list);
 	if (!as || !as->nodes_written) {
@@ -738,8 +717,23 @@ free_update:
 		six_unlock_intent(&b->lock);
 	}
 
-	if (!ret)
-		btree_update_nodes_reachable(as, res.seq);
+	if (!ret) {
+		nr_new_nodes = as->nr_new_nodes;
+		memcpy(new_nodes,
+		       as->new_nodes,
+		       as->nr_new_nodes * sizeof(struct btree *));
+
+		while (as->nr_new_nodes) {
+			struct btree *b = as->new_nodes[--as->nr_new_nodes];
+
+			BUG_ON(b->will_make_reachable != (unsigned long) as);
+			b->will_make_reachable = 0;
+		}
+
+		while (as->nr_pending)
+			bch2_btree_node_free_ondisk(c,
+				&as->pending[--as->nr_pending], res.seq);
+	}
 
 	__bch2_btree_update_free(as);
 	/*
@@ -747,6 +741,20 @@ free_update:
 	 * nodes to be writeable:
 	 */
 	closure_wake_up(&c->btree_interior_update_wait);
+
+	/*
+	 * Can't take btree node locks while holding btree_interior_update_lock:
+	 * */
+	mutex_unlock(&c->btree_interior_update_lock);
+
+	while (nr_new_nodes) {
+		struct btree *b = new_nodes[--nr_new_nodes];
+		btree_node_lock_type(c, b, SIX_LOCK_read);
+		bch2_btree_node_write_cond(c, b, btree_node_need_write(b));
+		six_unlock_read(&b->lock);
+	}
+
+	mutex_lock(&c->btree_interior_update_lock);
 	goto again;
 }
 
@@ -963,11 +971,13 @@ bch2_btree_update_start(struct btree_trans *trans, enum btree_id id,
 				      BTREE_UPDATE_JOURNAL_RES,
 				      JOURNAL_RES_GET_NONBLOCK);
 	if (ret == -EAGAIN) {
+		if (flags & BTREE_INSERT_NOUNLOCK)
+			return -EINTR;
+
 		bch2_trans_unlock(trans);
 
 		ret = bch2_journal_preres_get(&c->journal, &journal_preres,
-					      BTREE_UPDATE_JOURNAL_RES,
-					      JOURNAL_RES_GET_NONBLOCK);
+					      BTREE_UPDATE_JOURNAL_RES, 0);
 		if (ret)
 			return ERR_PTR(ret);
 
@@ -978,8 +988,10 @@ bch2_btree_update_start(struct btree_trans *trans, enum btree_id id,
 	}
 
 	reserve = bch2_btree_reserve_get(c, nr_nodes, flags, cl);
-	if (IS_ERR(reserve))
+	if (IS_ERR(reserve)) {
+		bch2_journal_preres_put(&c->journal, &journal_preres);
 		return ERR_CAST(reserve);
+	}
 
 	as = mempool_alloc(&c->btree_interior_update_pool, GFP_NOIO);
 	memset(as, 0, sizeof(*as));
@@ -1543,8 +1555,10 @@ int bch2_btree_split_leaf(struct bch_fs *c, struct btree_iter *iter,
 	/* Hack, because gc and splitting nodes doesn't mix yet: */
 	if (!(flags & BTREE_INSERT_GC_LOCK_HELD) &&
 	    !down_read_trylock(&c->gc_lock)) {
-		if (flags & BTREE_INSERT_NOUNLOCK)
+		if (flags & BTREE_INSERT_NOUNLOCK) {
+			trace_transaction_restart_ip(trans->ip, _THIS_IP_);
 			return -EINTR;
+		}
 
 		bch2_trans_unlock(trans);
 		down_read(&c->gc_lock);
@@ -1572,6 +1586,8 @@ int bch2_btree_split_leaf(struct bch_fs *c, struct btree_iter *iter,
 			BUG_ON(flags & BTREE_INSERT_NOUNLOCK);
 			bch2_trans_unlock(trans);
 			ret = -EINTR;
+
+			trace_transaction_restart_ip(trans->ip, _THIS_IP_);
 		}
 		goto out;
 	}
@@ -1677,6 +1693,7 @@ retry:
 
 	as = bch2_btree_update_start(trans, iter->btree_id,
 			 btree_update_reserve_required(c, parent) + 1,
+			 flags|
 			 BTREE_INSERT_NOFAIL|
 			 BTREE_INSERT_USE_RESERVE,
 			 !(flags & BTREE_INSERT_NOUNLOCK) ? &cl : NULL);
diff --git a/fs/bcachefs/checksum.c b/fs/bcachefs/checksum.c
index 9ef2cb9b9904..6f1afa4a3119 100644
--- a/fs/bcachefs/checksum.c
+++ b/fs/bcachefs/checksum.c
@@ -10,7 +10,7 @@
 #include <linux/random.h>
 #include <linux/scatterlist.h>
 #include <crypto/algapi.h>
-#include <crypto/chacha20.h>
+#include <crypto/chacha.h>
 #include <crypto/hash.h>
 #include <crypto/poly1305.h>
 #include <keys/user-type.h>
@@ -67,21 +67,21 @@ static u64 bch2_checksum_update(unsigned type, u64 crc, const void *data, size_t
 	}
 }
 
-static inline void do_encrypt_sg(struct crypto_skcipher *tfm,
+static inline void do_encrypt_sg(struct crypto_sync_skcipher *tfm,
 				 struct nonce nonce,
 				 struct scatterlist *sg, size_t len)
 {
-	SKCIPHER_REQUEST_ON_STACK(req, tfm);
+	SYNC_SKCIPHER_REQUEST_ON_STACK(req, tfm);
 	int ret;
 
-	skcipher_request_set_tfm(req, tfm);
+	skcipher_request_set_sync_tfm(req, tfm);
 	skcipher_request_set_crypt(req, sg, sg, len, nonce.d);
 
 	ret = crypto_skcipher_encrypt(req);
 	BUG_ON(ret);
 }
 
-static inline void do_encrypt(struct crypto_skcipher *tfm,
+static inline void do_encrypt(struct crypto_sync_skcipher *tfm,
 			      struct nonce nonce,
 			      void *buf, size_t len)
 {
@@ -94,8 +94,8 @@ static inline void do_encrypt(struct crypto_skcipher *tfm,
 int bch2_chacha_encrypt_key(struct bch_key *key, struct nonce nonce,
 			    void *buf, size_t len)
 {
-	struct crypto_skcipher *chacha20 =
-		crypto_alloc_skcipher("chacha20", 0, 0);
+	struct crypto_sync_skcipher *chacha20 =
+		crypto_alloc_sync_skcipher("chacha20", 0, 0);
 	int ret;
 
 	if (!chacha20) {
@@ -103,7 +103,8 @@ int bch2_chacha_encrypt_key(struct bch_key *key, struct nonce nonce,
 		return PTR_ERR(chacha20);
 	}
 
-	ret = crypto_skcipher_setkey(chacha20, (void *) key, sizeof(*key));
+	ret = crypto_skcipher_setkey(&chacha20->base,
+				     (void *) key, sizeof(*key));
 	if (ret) {
 		pr_err("crypto_skcipher_setkey() error: %i", ret);
 		goto err;
@@ -111,7 +112,7 @@ int bch2_chacha_encrypt_key(struct bch_key *key, struct nonce nonce,
 
 	do_encrypt(chacha20, nonce, buf, len);
 err:
-	crypto_free_skcipher(chacha20);
+	crypto_free_sync_skcipher(chacha20);
 	return ret;
 }
 
@@ -198,7 +199,7 @@ static struct bch_csum __bch2_checksum_bio(struct bch_fs *c, unsigned type,
 			kunmap_atomic(p);
 		}
 #else
-		__bio_for_each_contig_segment(bv, bio, *iter, *iter)
+		__bio_for_each_bvec(bv, bio, *iter, *iter)
 			crc = bch2_checksum_update(type, crc,
 				page_address(bv.bv_page) + bv.bv_offset,
 				bv.bv_len);
@@ -223,7 +224,7 @@ static struct bch_csum __bch2_checksum_bio(struct bch_fs *c, unsigned type,
 			kunmap_atomic(p);
 		}
 #else
-		__bio_for_each_contig_segment(bv, bio, *iter, *iter)
+		__bio_for_each_bvec(bv, bio, *iter, *iter)
 			crypto_shash_update(desc,
 				page_address(bv.bv_page) + bv.bv_offset,
 				bv.bv_len);
@@ -462,7 +463,7 @@ err:
 static int bch2_alloc_ciphers(struct bch_fs *c)
 {
 	if (!c->chacha20)
-		c->chacha20 = crypto_alloc_skcipher("chacha20", 0, 0);
+		c->chacha20 = crypto_alloc_sync_skcipher("chacha20", 0, 0);
 	if (IS_ERR(c->chacha20)) {
 		bch_err(c, "error requesting chacha20 module: %li",
 			PTR_ERR(c->chacha20));
@@ -545,7 +546,7 @@ int bch2_enable_encryption(struct bch_fs *c, bool keyed)
 			goto err;
 	}
 
-	ret = crypto_skcipher_setkey(c->chacha20,
+	ret = crypto_skcipher_setkey(&c->chacha20->base,
 			(void *) &key.key, sizeof(key.key));
 	if (ret)
 		goto err;
@@ -573,7 +574,7 @@ void bch2_fs_encryption_exit(struct bch_fs *c)
 	if (!IS_ERR_OR_NULL(c->poly1305))
 		crypto_free_shash(c->poly1305);
 	if (!IS_ERR_OR_NULL(c->chacha20))
-		crypto_free_skcipher(c->chacha20);
+		crypto_free_sync_skcipher(c->chacha20);
 	if (!IS_ERR_OR_NULL(c->sha256))
 		crypto_free_shash(c->sha256);
 }
@@ -605,7 +606,7 @@ int bch2_fs_encryption_init(struct bch_fs *c)
 	if (ret)
 		goto out;
 
-	ret = crypto_skcipher_setkey(c->chacha20,
+	ret = crypto_skcipher_setkey(&c->chacha20->base,
 			(void *) &key.key, sizeof(key.key));
 	if (ret)
 		goto out;
diff --git a/fs/bcachefs/checksum.h b/fs/bcachefs/checksum.h
index 833537cc8fd0..24dee8039d57 100644
--- a/fs/bcachefs/checksum.h
+++ b/fs/bcachefs/checksum.h
@@ -7,7 +7,7 @@
 #include "super-io.h"
 
 #include <linux/crc64.h>
-#include <crypto/chacha20.h>
+#include <crypto/chacha.h>
 
 static inline bool bch2_checksum_mergeable(unsigned type)
 {
@@ -138,9 +138,9 @@ static inline bool bch2_crc_cmp(struct bch_csum l, struct bch_csum r)
 /* for skipping ahead and encrypting/decrypting at an offset: */
 static inline struct nonce nonce_add(struct nonce nonce, unsigned offset)
 {
-	EBUG_ON(offset & (CHACHA20_BLOCK_SIZE - 1));
+	EBUG_ON(offset & (CHACHA_BLOCK_SIZE - 1));
 
-	le32_add_cpu(&nonce.d[0], offset / CHACHA20_BLOCK_SIZE);
+	le32_add_cpu(&nonce.d[0], offset / CHACHA_BLOCK_SIZE);
 	return nonce;
 }
 
diff --git a/fs/bcachefs/compress.c b/fs/bcachefs/compress.c
index 4a163b6d59da..0713286d7999 100644
--- a/fs/bcachefs/compress.c
+++ b/fs/bcachefs/compress.c
@@ -55,7 +55,7 @@ static struct bbuf __bio_map_or_bounce(struct bch_fs *c, struct bio *bio,
 	BUG_ON(bvec_iter_sectors(start) > c->sb.encoded_extent_max);
 
 #ifndef CONFIG_HIGHMEM
-	__bio_for_each_contig_segment(bv, bio, iter, start) {
+	__bio_for_each_bvec(bv, bio, iter, start) {
 		if (bv.bv_len == start.bi_size)
 			return (struct bbuf) {
 				.b = page_address(bv.bv_page) + bv.bv_offset,
diff --git a/fs/bcachefs/extents.c b/fs/bcachefs/extents.c
index 3f66457d2272..d1a4ab04fbbf 100644
--- a/fs/bcachefs/extents.c
+++ b/fs/bcachefs/extents.c
@@ -180,7 +180,7 @@ void bch2_btree_ptr_debugcheck(struct bch_fs *c, struct bkey_s_c k)
 		return;
 
 	bch2_fs_inconsistent_on(!test_bit(BCH_FS_REBUILD_REPLICAS, &c->flags) &&
-		!bch2_bkey_replicas_marked(c, k, false), c,
+		!bch2_bkey_replicas_marked_locked(c, k, false), c,
 		"btree key bad (replicas not marked in superblock):\n%s",
 		(bch2_bkey_val_to_text(&PBUF(buf), c, k), buf));
 
diff --git a/fs/bcachefs/fs-io.c b/fs/bcachefs/fs-io.c
index 707ce27eace6..0aa3afade4ea 100644
--- a/fs/bcachefs/fs-io.c
+++ b/fs/bcachefs/fs-io.c
@@ -602,7 +602,7 @@ int bch2_migrate_page(struct address_space *mapping, struct page *newpage,
 	EBUG_ON(!PageLocked(page));
 	EBUG_ON(!PageLocked(newpage));
 
-	ret = migrate_page_move_mapping(mapping, newpage, page, NULL, mode, 0);
+	ret = migrate_page_move_mapping(mapping, newpage, page, 0);
 	if (ret != MIGRATEPAGE_SUCCESS)
 		return ret;
 
@@ -627,10 +627,10 @@ int bch2_migrate_page(struct address_space *mapping, struct page *newpage,
 
 static void bch2_readpages_end_io(struct bio *bio)
 {
+	struct bvec_iter_all iter;
 	struct bio_vec *bv;
-	unsigned i;
 
-	bio_for_each_segment_all(bv, bio, i) {
+	bio_for_each_segment_all(bv, bio, iter) {
 		struct page *page = bv->bv_page;
 
 		if (!bio->bi_status) {
@@ -782,11 +782,8 @@ static void readpage_bio_extend(struct readpages_iter *iter,
 			if (!get_more)
 				break;
 
-			rcu_read_lock();
-			page = radix_tree_lookup(&iter->mapping->i_pages, page_offset);
-			rcu_read_unlock();
-
-			if (page && !radix_tree_exceptional_entry(page))
+			page = xa_load(&iter->mapping->i_pages, page_offset);
+			if (page && !xa_is_value(page))
 				break;
 
 			page = __page_cache_alloc(readahead_gfp_mask(iter->mapping));
@@ -1037,32 +1034,33 @@ static void bch2_writepage_io_done(struct closure *cl)
 					struct bch_writepage_io, cl);
 	struct bch_fs *c = io->op.c;
 	struct bio *bio = &io->op.wbio.bio;
+	struct bvec_iter_all iter;
 	struct bio_vec *bvec;
-	unsigned i, j;
+	unsigned i;
 
 	if (io->op.error) {
-		bio_for_each_segment_all(bvec, bio, i) {
+		bio_for_each_segment_all(bvec, bio, iter) {
 			struct bch_page_state *s;
 
 			SetPageError(bvec->bv_page);
-			mapping_set_error(io->inode->v.i_mapping, -EIO);
+			mapping_set_error(bvec->bv_page->mapping, -EIO);
 
 			s = __bch2_page_state(bvec->bv_page);
 			spin_lock(&s->lock);
-			for (j = 0; j < PAGE_SECTORS; j++)
-				s->s[j].nr_replicas = 0;
+			for (i = 0; i < PAGE_SECTORS; i++)
+				s->s[i].nr_replicas = 0;
 			spin_unlock(&s->lock);
 		}
 	}
 
 	if (io->op.flags & BCH_WRITE_WROTE_DATA_INLINE) {
-		bio_for_each_segment_all(bvec, bio, i) {
+		bio_for_each_segment_all(bvec, bio, iter) {
 			struct bch_page_state *s;
 
 			s = __bch2_page_state(bvec->bv_page);
 			spin_lock(&s->lock);
-			for (j = 0; j < PAGE_SECTORS; j++)
-				s->s[j].nr_replicas = 0;
+			for (i = 0; i < PAGE_SECTORS; i++)
+				s->s[i].nr_replicas = 0;
 			spin_unlock(&s->lock);
 		}
 	}
@@ -1086,7 +1084,7 @@ static void bch2_writepage_io_done(struct closure *cl)
 	 */
 	i_sectors_acct(c, io->inode, NULL, io->op.i_sectors_delta);
 
-	bio_for_each_segment_all(bvec, bio, i) {
+	bio_for_each_segment_all(bvec, bio, iter) {
 		struct bch_page_state *s = __bch2_page_state(bvec->bv_page);
 
 		if (atomic_dec_and_test(&s->write_count))
@@ -1240,7 +1238,7 @@ do_io:
 
 		if (w->io &&
 		    (w->io->op.res.nr_replicas != nr_replicas_this_write ||
-		     bio_full(&w->io->op.wbio.bio) ||
+		     bio_full(&w->io->op.wbio.bio, PAGE_SIZE) ||
 		     w->io->op.wbio.bio.bi_iter.bi_size >= (256U << 20) ||
 		     bio_end_sector(&w->io->op.wbio.bio) != sector))
 			bch2_writepage_do_io(w);
@@ -1805,8 +1803,9 @@ static long bch2_dio_write_loop(struct dio_write *dio)
 	struct address_space *mapping = req->ki_filp->f_mapping;
 	struct bch_inode_info *inode = file_bch_inode(req->ki_filp);
 	struct bio *bio = &dio->op.wbio.bio;
+	struct bvec_iter_all iter;
 	struct bio_vec *bv;
-	unsigned i, unaligned;
+	unsigned unaligned;
 	u64 new_i_size;
 	bool sync = dio->sync;
 	long ret;
@@ -1838,7 +1837,7 @@ static long bch2_dio_write_loop(struct dio_write *dio)
 			 * bio_iov_iter_get_pages was only able to get <
 			 * blocksize worth of pages:
 			 */
-			bio_for_each_segment_all(bv, bio, i)
+			bio_for_each_segment_all(bv, bio, iter)
 				put_page(bv->bv_page);
 			ret = -EFAULT;
 			goto err;
@@ -1886,7 +1885,7 @@ loop:
 			i_size_write(&inode->v, new_i_size);
 		spin_unlock(&inode->v.i_lock);
 
-		bio_for_each_segment_all(bv, bio, i)
+		bio_for_each_segment_all(bv, bio, iter)
 			put_page(bv->bv_page);
 		if (!dio->iter.count || dio->op.error)
 			break;
@@ -2823,235 +2822,6 @@ static void mark_range_unallocated(struct bch_inode_info *inode,
 	} while (index <= end_index);
 }
 
-static int generic_access_check_limits(struct file *file, loff_t pos,
-				       loff_t *count)
-{
-	struct inode *inode = file->f_mapping->host;
-	loff_t max_size = inode->i_sb->s_maxbytes;
-
-	if (!(file->f_flags & O_LARGEFILE))
-		max_size = MAX_NON_LFS;
-
-	if (unlikely(pos >= max_size))
-		return -EFBIG;
-	*count = min(*count, max_size - pos);
-	return 0;
-}
-
-static int generic_write_check_limits(struct file *file, loff_t pos,
-				      loff_t *count)
-{
-	loff_t limit = rlimit(RLIMIT_FSIZE);
-
-	if (limit != RLIM_INFINITY) {
-		if (pos >= limit) {
-			send_sig(SIGXFSZ, current, 0);
-			return -EFBIG;
-		}
-		*count = min(*count, limit - pos);
-	}
-
-	return generic_access_check_limits(file, pos, count);
-}
-
-static int generic_remap_checks(struct file *file_in, loff_t pos_in,
-			 struct file *file_out, loff_t pos_out,
-			 loff_t *req_count, unsigned int remap_flags)
-{
-	struct inode *inode_in = file_in->f_mapping->host;
-	struct inode *inode_out = file_out->f_mapping->host;
-	uint64_t count = *req_count;
-	uint64_t bcount;
-	loff_t size_in, size_out;
-	loff_t bs = inode_out->i_sb->s_blocksize;
-	int ret;
-
-	/* The start of both ranges must be aligned to an fs block. */
-	if (!IS_ALIGNED(pos_in, bs) || !IS_ALIGNED(pos_out, bs))
-		return -EINVAL;
-
-	/* Ensure offsets don't wrap. */
-	if (pos_in + count < pos_in || pos_out + count < pos_out)
-		return -EINVAL;
-
-	size_in = i_size_read(inode_in);
-	size_out = i_size_read(inode_out);
-
-	/* Dedupe requires both ranges to be within EOF. */
-	if ((remap_flags & REMAP_FILE_DEDUP) &&
-	    (pos_in >= size_in || pos_in + count > size_in ||
-	     pos_out >= size_out || pos_out + count > size_out))
-		return -EINVAL;
-
-	/* Ensure the infile range is within the infile. */
-	if (pos_in >= size_in)
-		return -EINVAL;
-	count = min(count, size_in - (uint64_t)pos_in);
-
-	ret = generic_access_check_limits(file_in, pos_in, &count);
-	if (ret)
-		return ret;
-
-	ret = generic_write_check_limits(file_out, pos_out, &count);
-	if (ret)
-		return ret;
-
-	/*
-	 * If the user wanted us to link to the infile's EOF, round up to the
-	 * next block boundary for this check.
-	 *
-	 * Otherwise, make sure the count is also block-aligned, having
-	 * already confirmed the starting offsets' block alignment.
-	 */
-	if (pos_in + count == size_in) {
-		bcount = ALIGN(size_in, bs) - pos_in;
-	} else {
-		if (!IS_ALIGNED(count, bs))
-			count = ALIGN_DOWN(count, bs);
-		bcount = count;
-	}
-
-	/* Don't allow overlapped cloning within the same file. */
-	if (inode_in == inode_out &&
-	    pos_out + bcount > pos_in &&
-	    pos_out < pos_in + bcount)
-		return -EINVAL;
-
-	/*
-	 * We shortened the request but the caller can't deal with that, so
-	 * bounce the request back to userspace.
-	 */
-	if (*req_count != count && !(remap_flags & REMAP_FILE_CAN_SHORTEN))
-		return -EINVAL;
-
-	*req_count = count;
-	return 0;
-}
-
-static int generic_remap_check_len(struct inode *inode_in,
-				   struct inode *inode_out,
-				   loff_t pos_out,
-				   loff_t *len,
-				   unsigned int remap_flags)
-{
-	u64 blkmask = i_blocksize(inode_in) - 1;
-	loff_t new_len = *len;
-
-	if ((*len & blkmask) == 0)
-		return 0;
-
-	if ((remap_flags & REMAP_FILE_DEDUP) ||
-	    pos_out + *len < i_size_read(inode_out))
-		new_len &= ~blkmask;
-
-	if (new_len == *len)
-		return 0;
-
-	if (remap_flags & REMAP_FILE_CAN_SHORTEN) {
-		*len = new_len;
-		return 0;
-	}
-
-	return (remap_flags & REMAP_FILE_DEDUP) ? -EBADE : -EINVAL;
-}
-
-static int generic_remap_file_range_prep(struct file *file_in, loff_t pos_in,
-				  struct file *file_out, loff_t pos_out,
-				  loff_t *len, unsigned int remap_flags)
-{
-	struct inode *inode_in = file_inode(file_in);
-	struct inode *inode_out = file_inode(file_out);
-	bool same_inode = (inode_in == inode_out);
-	int ret;
-
-	/* Don't touch certain kinds of inodes */
-	if (IS_IMMUTABLE(inode_out))
-		return -EPERM;
-
-	if (IS_SWAPFILE(inode_in) || IS_SWAPFILE(inode_out))
-		return -ETXTBSY;
-
-	/* Don't reflink dirs, pipes, sockets... */
-	if (S_ISDIR(inode_in->i_mode) || S_ISDIR(inode_out->i_mode))
-		return -EISDIR;
-	if (!S_ISREG(inode_in->i_mode) || !S_ISREG(inode_out->i_mode))
-		return -EINVAL;
-
-	/* Zero length dedupe exits immediately; reflink goes to EOF. */
-	if (*len == 0) {
-		loff_t isize = i_size_read(inode_in);
-
-		if ((remap_flags & REMAP_FILE_DEDUP) || pos_in == isize)
-			return 0;
-		if (pos_in > isize)
-			return -EINVAL;
-		*len = isize - pos_in;
-		if (*len == 0)
-			return 0;
-	}
-
-	/* Check that we don't violate system file offset limits. */
-	ret = generic_remap_checks(file_in, pos_in, file_out, pos_out, len,
-			remap_flags);
-	if (ret)
-		return ret;
-
-	/* Wait for the completion of any pending IOs on both files */
-	inode_dio_wait(inode_in);
-	if (!same_inode)
-		inode_dio_wait(inode_out);
-
-	ret = filemap_write_and_wait_range(inode_in->i_mapping,
-			pos_in, pos_in + *len - 1);
-	if (ret)
-		return ret;
-
-	ret = filemap_write_and_wait_range(inode_out->i_mapping,
-			pos_out, pos_out + *len - 1);
-	if (ret)
-		return ret;
-
-	/*
-	 * Check that the extents are the same.
-	 */
-	if (remap_flags & REMAP_FILE_DEDUP) {
-		bool		is_same = false;
-
-		ret = vfs_dedupe_file_range_compare(inode_in, pos_in,
-				inode_out, pos_out, *len, &is_same);
-		if (ret)
-			return ret;
-		if (!is_same)
-			return -EBADE;
-	}
-
-	ret = generic_remap_check_len(inode_in, inode_out, pos_out, len,
-			remap_flags);
-	if (ret)
-		return ret;
-
-	/* If can't alter the file contents, we're done. */
-	if (!(remap_flags & REMAP_FILE_DEDUP)) {
-		/* Update the timestamps, since we can alter file contents. */
-		if (!(file_out->f_mode & FMODE_NOCMTIME)) {
-			ret = file_update_time(file_out);
-			if (ret)
-				return ret;
-		}
-
-		/*
-		 * Clear the security bits if the process is not being run by
-		 * root.  This keeps people from modifying setuid and setgid
-		 * binaries.
-		 */
-		ret = file_remove_privs(file_out);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
 loff_t bch2_remap_file_range(struct file *file_src, loff_t pos_src,
 			     struct file *file_dst, loff_t pos_dst,
 			     loff_t len, unsigned remap_flags)
@@ -3244,7 +3014,7 @@ static loff_t page_hole_offset(struct address_space *mapping, loff_t offset)
 	loff_t ret = -1;
 
 	page = find_lock_entry(mapping, index);
-	if (!page || radix_tree_exception(page))
+	if (!page || xa_is_value(page))
 		return offset;
 
 	pg_offset = __page_hole_offset(page, offset & (PAGE_SIZE - 1));
diff --git a/fs/bcachefs/fs-io.h b/fs/bcachefs/fs-io.h
index 1b593ea707d5..7063556d289b 100644
--- a/fs/bcachefs/fs-io.h
+++ b/fs/bcachefs/fs-io.h
@@ -35,10 +35,6 @@ int bch2_fsync(struct file *, loff_t, loff_t, int);
 int bch2_truncate(struct bch_inode_info *, struct iattr *);
 long bch2_fallocate_dispatch(struct file *, int, loff_t, loff_t);
 
-#define REMAP_FILE_ADVISORY		(0)
-#define REMAP_FILE_DEDUP		(1 << 0)
-#define REMAP_FILE_CAN_SHORTEN		(1 << 1)
-
 loff_t bch2_remap_file_range(struct file *, loff_t, struct file *,
 			     loff_t, loff_t, unsigned);
 
diff --git a/fs/bcachefs/fs.c b/fs/bcachefs/fs.c
index b1445bdc3e9d..1c89a1b2c2d0 100644
--- a/fs/bcachefs/fs.c
+++ b/fs/bcachefs/fs.c
@@ -966,15 +966,6 @@ static int bch2_vfs_readdir(struct file *file, struct dir_context *ctx)
 	return bch2_readdir(c, inode->v.i_ino, ctx);
 }
 
-static int bch2_clone_file_range(struct file *file_src, loff_t pos_src,
-				 struct file *file_dst, loff_t pos_dst,
-				 u64 len)
-{
-	return bch2_remap_file_range(file_src, pos_src,
-				     file_dst, pos_dst,
-				     len, 0);
-}
-
 static const struct file_operations bch_file_operations = {
 	.llseek		= bch2_llseek,
 	.read_iter	= bch2_read_iter,
@@ -992,7 +983,7 @@ static const struct file_operations bch_file_operations = {
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= bch2_compat_fs_ioctl,
 #endif
-	.clone_file_range = bch2_clone_file_range,
+	.remap_file_range = bch2_remap_file_range,
 };
 
 static const struct inode_operations bch_file_inode_operations = {
@@ -1523,7 +1514,7 @@ static struct dentry *bch2_mount(struct file_system_type *fs_type,
 
 	sb->s_bdi->congested_fn		= bch2_congested;
 	sb->s_bdi->congested_data	= c;
-	sb->s_bdi->ra_pages		= VM_MAX_READAHEAD * 1024 / PAGE_SIZE;
+	sb->s_bdi->ra_pages		= VM_READAHEAD_PAGES;
 
 	for_each_online_member(ca, c, i) {
 		struct block_device *bdev = ca->disk_sb.bdev;
diff --git a/fs/bcachefs/io.c b/fs/bcachefs/io.c
index ff3dd31c134c..19059702428a 100644
--- a/fs/bcachefs/io.c
+++ b/fs/bcachefs/io.c
@@ -124,10 +124,10 @@ void bch2_latency_acct(struct bch_dev *ca, u64 submit_time, int rw)
 
 void bch2_bio_free_pages_pool(struct bch_fs *c, struct bio *bio)
 {
+	struct bvec_iter_all iter;
 	struct bio_vec *bv;
-	unsigned i;
 
-	bio_for_each_segment_all(bv, bio, i)
+	bio_for_each_segment_all(bv, bio, iter)
 		if (bv->bv_page != ZERO_PAGE(0))
 			mempool_free(bv->bv_page, &c->bio_bounce_pages);
 	bio->bi_vcnt = 0;
diff --git a/fs/bcachefs/journal.h b/fs/bcachefs/journal.h
index ec61137df00a..1ba8b62b9f2b 100644
--- a/fs/bcachefs/journal.h
+++ b/fs/bcachefs/journal.h
@@ -344,7 +344,9 @@ static inline int bch2_journal_res_get(struct journal *j, struct journal_res *re
 		return ret;
 out:
 	if (!(flags & JOURNAL_RES_GET_CHECK)) {
-		lock_acquire_shared(&j->res_map, 0, 0, NULL, _THIS_IP_);
+		lock_acquire_shared(&j->res_map, 0,
+				    (flags & JOURNAL_RES_GET_NONBLOCK) != 0,
+				    NULL, _THIS_IP_);
 		EBUG_ON(!res->ref);
 	}
 	return 0;
diff --git a/fs/bcachefs/journal_reclaim.c b/fs/bcachefs/journal_reclaim.c
index db3afd908474..d34434f62454 100644
--- a/fs/bcachefs/journal_reclaim.c
+++ b/fs/bcachefs/journal_reclaim.c
@@ -322,14 +322,12 @@ void bch2_journal_pin_drop(struct journal *j,
 	spin_unlock(&j->lock);
 }
 
-void __bch2_journal_pin_add(struct journal *j, u64 seq,
+static void bch2_journal_pin_add_locked(struct journal *j, u64 seq,
 			    struct journal_entry_pin *pin,
 			    journal_pin_flush_fn flush_fn)
 {
 	struct journal_entry_pin_list *pin_list = journal_seq_pin(j, seq);
 
-	spin_lock(&j->lock);
-
 	__journal_pin_drop(j, pin);
 
 	BUG_ON(!atomic_read(&pin_list->count));
@@ -339,7 +337,14 @@ void __bch2_journal_pin_add(struct journal *j, u64 seq,
 	pin->flush	= flush_fn;
 
 	list_add(&pin->list, flush_fn ? &pin_list->list : &pin_list->flushed);
+}
 
+void __bch2_journal_pin_add(struct journal *j, u64 seq,
+			    struct journal_entry_pin *pin,
+			    journal_pin_flush_fn flush_fn)
+{
+	spin_lock(&j->lock);
+	bch2_journal_pin_add_locked(j, seq, pin, flush_fn);
 	spin_unlock(&j->lock);
 
 	/*
@@ -354,9 +359,13 @@ void bch2_journal_pin_copy(struct journal *j,
 			   struct journal_entry_pin *src,
 			   journal_pin_flush_fn flush_fn)
 {
+	spin_lock(&j->lock);
+
 	if (journal_pin_active(src) &&
 	    (!journal_pin_active(dst) || src->seq < dst->seq))
-		__bch2_journal_pin_add(j, src->seq, dst, flush_fn);
+		bch2_journal_pin_add_locked(j, src->seq, dst, flush_fn);
+
+	spin_unlock(&j->lock);
 }
 
 /**
diff --git a/fs/bcachefs/move.c b/fs/bcachefs/move.c
index 728c7554779f..4afda95f4017 100644
--- a/fs/bcachefs/move.c
+++ b/fs/bcachefs/move.c
@@ -306,12 +306,12 @@ static void move_free(struct closure *cl)
 {
 	struct moving_io *io = container_of(cl, struct moving_io, cl);
 	struct moving_context *ctxt = io->write.ctxt;
+	struct bvec_iter_all iter;
 	struct bio_vec *bv;
-	unsigned i;
 
 	bch2_disk_reservation_put(io->write.op.c, &io->write.op.res);
 
-	bio_for_each_segment_all(bv, &io->write.op.wbio.bio, i)
+	bio_for_each_segment_all(bv, &io->write.op.wbio.bio, iter)
 		if (bv->bv_page)
 			__free_page(bv->bv_page);
 
diff --git a/fs/bcachefs/rebalance.c b/fs/bcachefs/rebalance.c
index ab1934325948..e15a2b1dc5d0 100644
--- a/fs/bcachefs/rebalance.c
+++ b/fs/bcachefs/rebalance.c
@@ -204,17 +204,21 @@ static int bch2_rebalance_thread(void *arg)
 			prev_run_time;
 
 		if (w.dev_most_full_percent < 20 && throttle > 0) {
-			r->state = REBALANCE_THROTTLED;
 			r->throttled_until_iotime = io_start +
 				div_u64(w.dev_most_full_capacity *
 					(20 - w.dev_most_full_percent),
 					50);
-			r->throttled_until_cputime = start + throttle;
 
-			bch2_kthread_io_clock_wait(clock,
-				r->throttled_until_iotime,
-				throttle);
-			continue;
+			if (atomic_long_read(&clock->now) + clock->max_slop <
+			    r->throttled_until_iotime) {
+				r->throttled_until_cputime = start + throttle;
+				r->state = REBALANCE_THROTTLED;
+
+				bch2_kthread_io_clock_wait(clock,
+					r->throttled_until_iotime,
+					throttle);
+				continue;
+			}
 		}
 
 		/* minimum 1 mb/sec: */
diff --git a/fs/bcachefs/util.h b/fs/bcachefs/util.h
index 2b19a0038045..0128daba5970 100644
--- a/fs/bcachefs/util.h
+++ b/fs/bcachefs/util.h
@@ -664,35 +664,6 @@ static inline void memset_u64s_tail(void *s, int c, unsigned bytes)
 	memset(s + bytes, c, rem);
 }
 
-static inline struct bio_vec next_contig_bvec(struct bio *bio,
-					      struct bvec_iter *iter)
-{
-	struct bio_vec bv = bio_iter_iovec(bio, *iter);
-
-	bio_advance_iter(bio, iter, bv.bv_len);
-#ifndef CONFIG_HIGHMEM
-	while (iter->bi_size) {
-		struct bio_vec next = bio_iter_iovec(bio, *iter);
-
-		if (page_address(bv.bv_page) + bv.bv_offset + bv.bv_len !=
-		    page_address(next.bv_page) + next.bv_offset)
-			break;
-
-		bv.bv_len += next.bv_len;
-		bio_advance_iter(bio, iter, next.bv_len);
-	}
-#endif
-	return bv;
-}
-
-#define __bio_for_each_contig_segment(bv, bio, iter, start)		\
-	for (iter = (start);						\
-	     (iter).bi_size &&						\
-		((bv = next_contig_bvec((bio), &(iter))), 1);)
-
-#define bio_for_each_contig_segment(bv, bio, iter)			\
-	__bio_for_each_contig_segment(bv, bio, iter, (bio)->bi_iter)
-
 void sort_cmp_size(void *base, size_t num, size_t size,
 	  int (*cmp_func)(const void *, const void *, size_t),
 	  void (*swap_func)(void *, void *, size_t));
diff --git a/include/trace/events/bcachefs.h b/include/trace/events/bcachefs.h
index d7e898b02491..01a9cc736cab 100644
--- a/include/trace/events/bcachefs.h
+++ b/include/trace/events/bcachefs.h
@@ -499,6 +499,23 @@ TRACE_EVENT(copygc,
 		__entry->buckets_moved, __entry->buckets_not_moved)
 );
 
+TRACE_EVENT(transaction_restart_ip,
+	TP_PROTO(unsigned long caller, unsigned long ip),
+	TP_ARGS(caller, ip),
+
+	TP_STRUCT__entry(
+		__field(unsigned long,		caller	)
+		__field(unsigned long,		ip	)
+	),
+
+	TP_fast_assign(
+		__entry->caller	= caller;
+		__entry->ip	= ip;
+	),
+
+	TP_printk("%pF %pF", (void *) __entry->caller, (void *) __entry->ip)
+);
+
 DECLARE_EVENT_CLASS(transaction_restart,
 	TP_PROTO(unsigned long ip),
 	TP_ARGS(ip),