bcachefs: Clean up btree write buffer write ref handling

__bch2_btree_write_buffer_flush() now assumes a write ref is already
held (as called by the transaction commit path); and the wrappers
bch2_write_buffer_flush() and flush_sync() take an explicit write ref.

This means internally the write buffer code can always use
BTREE_INSERT_NOCHECK_RW, instead of in the previous code passing flags
around and hoping the NOCHECK_RW flag was always carried around
correctly.

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

fs/bcachefs/bcachefs.h

@@ -672,7 +672,8 @@ struct btree_trans_buf {
 	x(invalidate)							\
 	x(delete_dead_snapshots)					\
 	x(snapshot_delete_pagecache)					\
-	x(sysfs)
+	x(sysfs)							\
+	x(btree_write_buffer)
 
 enum bch_write_ref {
 #define x(n) BCH_WRITE_REF_##n,

fs/bcachefs/btree_trans_commit.c

@@ -961,8 +961,7 @@ int bch2_trans_commit_error(struct btree_trans *trans, unsigned flags,
 
 			if (wb->state.nr > wb->size * 3 / 4) {
 				bch2_trans_begin(trans);
-				ret = __bch2_btree_write_buffer_flush(trans,
-						flags|BCH_TRANS_COMMIT_no_check_rw, true);
+				ret = __bch2_btree_write_buffer_flush(trans, true);
 				if (!ret) {
 					trace_and_count(c, trans_restart_write_buffer_flush, trans, _THIS_IP_);
 					ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_write_buffer_flush);
@@ -1077,8 +1076,7 @@ int __bch2_trans_commit(struct btree_trans *trans, unsigned flags)
 		bch2_trans_begin(trans);
 		bch2_trans_unlock(trans);
 
-		ret = __bch2_btree_write_buffer_flush(trans,
-					flags|BCH_TRANS_COMMIT_no_check_rw, true);
+		ret = __bch2_btree_write_buffer_flush(trans, true);
 		if (!ret) {
 			trace_and_count(c, trans_restart_write_buffer_flush, trans, _THIS_IP_);
 			ret = btree_trans_restart(trans, BCH_ERR_transaction_restart_write_buffer_flush);

fs/bcachefs/btree_write_buffer.c

@@ -137,8 +137,7 @@ btree_write_buffered_insert(struct btree_trans *trans,
 	return ret;
 }
 
-int __bch2_btree_write_buffer_flush(struct btree_trans *trans, unsigned commit_flags,
-				    bool locked)
+int __bch2_btree_write_buffer_flush(struct btree_trans *trans, bool locked)
 {
 	struct bch_fs *c = trans->c;
 	struct journal *j = &c->journal;
@@ -210,8 +209,8 @@ int __bch2_btree_write_buffer_flush(struct btree_trans *trans, unsigned commit_f
 		iter.path->preserve = false;
 
 		do {
-			ret = bch2_btree_write_buffer_flush_one(trans, &iter, i,
-						commit_flags, &write_locked, &fast);
+			ret = bch2_btree_write_buffer_flush_one(trans, &iter, i, 0,
+								&write_locked, &fast);
 			if (!write_locked)
 				bch2_trans_begin(trans);
 		} while (bch2_err_matches(ret, BCH_ERR_transaction_restart));
@@ -252,9 +251,6 @@ slowpath:
 	     btree_write_buffered_journal_cmp,
 	     NULL);
 
-	commit_flags &= ~BCH_WATERMARK_MASK;
-	commit_flags |= BCH_WATERMARK_reclaim;
-
 	for (i = keys; i < keys + nr; i++) {
 		if (!i->journal_seq)
 			continue;
@@ -263,7 +259,8 @@ slowpath:
 			      bch2_btree_write_buffer_journal_flush);
 
 		ret = commit_do(trans, NULL, NULL,
-				commit_flags|
+				BCH_WATERMARK_reclaim|
+				BCH_TRANS_COMMIT_no_check_rw|
 				BCH_TRANS_COMMIT_no_enospc|
 				BCH_TRANS_COMMIT_no_journal_res|
 				BCH_TRANS_COMMIT_journal_reclaim,
@@ -279,16 +276,33 @@ int bch2_btree_write_buffer_flush_sync(struct btree_trans *trans)
 {
 	struct bch_fs *c = trans->c;
 
+	if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_btree_write_buffer))
+		return -BCH_ERR_erofs_no_writes;
+
 	trace_and_count(c, write_buffer_flush_sync, trans, _RET_IP_);
 
 	bch2_trans_unlock(trans);
-	mutex_lock(&c->btree_write_buffer.flush_lock);
-	return __bch2_btree_write_buffer_flush(trans, 0, true);
+	mutex_lock(&trans->c->btree_write_buffer.flush_lock);
+	int ret = __bch2_btree_write_buffer_flush(trans, true);
+	bch2_write_ref_put(c, BCH_WRITE_REF_btree_write_buffer);
+	return ret;
+}
+
+int bch2_btree_write_buffer_flush_nocheck_rw(struct btree_trans *trans)
+{
+	return __bch2_btree_write_buffer_flush(trans, false);
 }
 
 int bch2_btree_write_buffer_flush(struct btree_trans *trans)
 {
-	return __bch2_btree_write_buffer_flush(trans, 0, false);
+	struct bch_fs *c = trans->c;
+
+	if (!bch2_write_ref_tryget(c, BCH_WRITE_REF_btree_write_buffer))
+		return -BCH_ERR_erofs_no_writes;
+
+	int ret = bch2_btree_write_buffer_flush_nocheck_rw(trans);
+	bch2_write_ref_put(c, BCH_WRITE_REF_btree_write_buffer);
+	return ret;
 }
 
 static int bch2_btree_write_buffer_journal_flush(struct journal *j,
@@ -300,7 +314,7 @@ static int bch2_btree_write_buffer_journal_flush(struct journal *j,
 	mutex_lock(&wb->flush_lock);
 
 	return bch2_trans_run(c,
-			__bch2_btree_write_buffer_flush(trans, BCH_TRANS_COMMIT_no_check_rw, true));
+			__bch2_btree_write_buffer_flush(trans, true));
 }
 
 static inline u64 btree_write_buffer_ref(int idx)

fs/bcachefs/btree_write_buffer.h

@@ -2,7 +2,8 @@
 #ifndef _BCACHEFS_BTREE_WRITE_BUFFER_H
 #define _BCACHEFS_BTREE_WRITE_BUFFER_H
 
-int __bch2_btree_write_buffer_flush(struct btree_trans *, unsigned, bool);
+int bch2_btree_write_buffer_flush_nocheck_rw(struct btree_trans *);
+int __bch2_btree_write_buffer_flush(struct btree_trans *, bool);
 int bch2_btree_write_buffer_flush_sync(struct btree_trans *);
 int bch2_btree_write_buffer_flush(struct btree_trans *);
 

fs/bcachefs/ec.c

@@ -1005,7 +1005,7 @@ static int ec_stripe_update_extents(struct bch_fs *c, struct ec_stripe_buf *s)
 	unsigned i, nr_data = v->nr_blocks - v->nr_redundant;
 	int ret = 0;
 
-	ret = bch2_btree_write_buffer_flush(trans);
+	ret = bch2_btree_write_buffer_flush_nocheck_rw(trans);
 	if (ret)
 		goto err;
 

fs/bcachefs/inode.c

@@ -1162,10 +1162,6 @@ int bch2_delete_dead_inodes(struct bch_fs *c)
 again:
 	need_another_pass = false;
 
-	ret = bch2_btree_write_buffer_flush_sync(trans);
-	if (ret)
-		goto err;
-
 	/*
 	 * Weird transaction restart handling here because on successful delete,
 	 * bch2_inode_rm_snapshot() will return a nested transaction restart,
@@ -1196,8 +1192,12 @@ again:
 	}
 	bch2_trans_iter_exit(trans, &iter);
 
-	if (!ret && need_another_pass)
+	if (!ret && need_another_pass) {
+		ret = bch2_btree_write_buffer_flush_sync(trans);
+		if (ret)
+			goto err;
 		goto again;
+	}
 err:
 	bch2_trans_put(trans);
 

fs/bcachefs/movinggc.c

@@ -154,6 +154,9 @@ static int bch2_copygc_get_buckets(struct moving_context *ctxt,
 	move_buckets_wait(ctxt, buckets_in_flight, false);
 
 	ret = bch2_btree_write_buffer_flush(trans);
+	if (bch2_err_matches(ret, EROFS))
+		return ret;
+
 	if (bch2_fs_fatal_err_on(ret, c, "%s: error %s from bch2_btree_write_buffer_flush()",
 				 __func__, bch2_err_str(ret)))
 		return ret;