Merge branch 'for-linus' of git://git.kernel.dk/linux-block

Pull block fixes from Jens Axboe:
 "Five fixes for this series:

   - a fix from me to ensure that blk-mq drivers that terminate IO in
     their ->queue_rq() handler by returning QUEUE_ERROR don't stall
     with a scheduler enabled.

   - four nbd fixes from Josef and Ratna, fixing various problems that
     are critical enough to go in for this cycle. They have been well
     tested"

* 'for-linus' of git://git.kernel.dk/linux-block:
  nbd: replace kill_bdev() with __invalidate_device()
  nbd: set queue timeout properly
  nbd: set rq->errors to actual error code
  nbd: handle ERESTARTSYS properly
  blk-mq: include errors in did_work calculation

2 files changed, 107 insertions, 36 deletions

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 08a49c69738b..6b6e7bc041db 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -969,7 +969,7 @@ bool blk_mq_dispatch_rq_list(struct blk_mq_hw_ctx *hctx, struct list_head *list)
 	struct request *rq;
 	LIST_HEAD(driver_list);
 	struct list_head *dptr;
-	int queued, ret = BLK_MQ_RQ_QUEUE_OK;
+	int errors, queued, ret = BLK_MQ_RQ_QUEUE_OK;
 
 	/*
 	 * Start off with dptr being NULL, so we start the first request
@@ -980,7 +980,7 @@ bool blk_mq_dispatch_rq_list(struct blk_mq_hw_ctx *hctx, struct list_head *list)
 	/*
 	 * Now process all the entries, sending them to the driver.
 	 */
-	queued = 0;
+	errors = queued = 0;
 	while (!list_empty(list)) {
 		struct blk_mq_queue_data bd;
 
@@ -1037,6 +1037,7 @@ bool blk_mq_dispatch_rq_list(struct blk_mq_hw_ctx *hctx, struct list_head *list)
 		default:
 			pr_err("blk-mq: bad return on queue: %d\n", ret);
 		case BLK_MQ_RQ_QUEUE_ERROR:
+			errors++;
 			rq->errors = -EIO;
 			blk_mq_end_request(rq, rq->errors);
 			break;
@@ -1088,7 +1089,7 @@ bool blk_mq_dispatch_rq_list(struct blk_mq_hw_ctx *hctx, struct list_head *list)
 			blk_mq_run_hw_queue(hctx, true);
 	}
 
-	return queued != 0;
+	return (queued + errors) != 0;
 }
 
 static void __blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx)
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 7e4287bc19e5..d8a23561b4cb 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -47,6 +47,8 @@ static DEFINE_MUTEX(nbd_index_mutex);
 struct nbd_sock {
 	struct socket *sock;
 	struct mutex tx_lock;
+	struct request *pending;
+	int sent;
 };
 
 #define NBD_TIMEDOUT			0
@@ -124,7 +126,8 @@ static const char *nbdcmd_to_ascii(int cmd)
 
 static int nbd_size_clear(struct nbd_device *nbd, struct block_device *bdev)
 {
-	bd_set_size(bdev, 0);
+	if (bdev->bd_openers <= 1)
+		bd_set_size(bdev, 0);
 	set_capacity(nbd->disk, 0);
 	kobject_uevent(&nbd_to_dev(nbd)->kobj, KOBJ_CHANGE);
 
@@ -190,7 +193,7 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
 
 	dev_err(nbd_to_dev(nbd), "Connection timed out, shutting down connection\n");
 	set_bit(NBD_TIMEDOUT, &nbd->runtime_flags);
-	req->errors++;
+	req->errors = -EIO;
 
 	mutex_lock(&nbd->config_lock);
 	sock_shutdown(nbd);
@@ -202,7 +205,7 @@ static enum blk_eh_timer_return nbd_xmit_timeout(struct request *req,
  *  Send or receive packet.
  */
 static int sock_xmit(struct nbd_device *nbd, int index, int send,
-		     struct iov_iter *iter, int msg_flags)
+		     struct iov_iter *iter, int msg_flags, int *sent)
 {
 	struct socket *sock = nbd->socks[index]->sock;
 	int result;
@@ -237,6 +240,8 @@ static int sock_xmit(struct nbd_device *nbd, int index, int send,
 				result = -EPIPE; /* short read */
 			break;
 		}
+		if (sent)
+			*sent += result;
 	} while (msg_data_left(&msg));
 
 	tsk_restore_flags(current, pflags, PF_MEMALLOC);
@@ -248,6 +253,7 @@ static int sock_xmit(struct nbd_device *nbd, int index, int send,
 static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 {
 	struct request *req = blk_mq_rq_from_pdu(cmd);
+	struct nbd_sock *nsock = nbd->socks[index];
 	int result;
 	struct nbd_request request = {.magic = htonl(NBD_REQUEST_MAGIC)};
 	struct kvec iov = {.iov_base = &request, .iov_len = sizeof(request)};
@@ -256,6 +262,7 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 	struct bio *bio;
 	u32 type;
 	u32 tag = blk_mq_unique_tag(req);
+	int sent = nsock->sent, skip = 0;
 
 	iov_iter_kvec(&from, WRITE | ITER_KVEC, &iov, 1, sizeof(request));
 
@@ -283,6 +290,17 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 		return -EIO;
 	}
 
+	/* We did a partial send previously, and we at least sent the whole
+	 * request struct, so just go and send the rest of the pages in the
+	 * request.
+	 */
+	if (sent) {
+		if (sent >= sizeof(request)) {
+			skip = sent - sizeof(request);
+			goto send_pages;
+		}
+		iov_iter_advance(&from, sent);
+	}
 	request.type = htonl(type);
 	if (type != NBD_CMD_FLUSH) {
 		request.from = cpu_to_be64((u64)blk_rq_pos(req) << 9);
@@ -294,15 +312,27 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 		cmd, nbdcmd_to_ascii(type),
 		(unsigned long long)blk_rq_pos(req) << 9, blk_rq_bytes(req));
 	result = sock_xmit(nbd, index, 1, &from,
-			(type == NBD_CMD_WRITE) ? MSG_MORE : 0);
+			(type == NBD_CMD_WRITE) ? MSG_MORE : 0, &sent);
 	if (result <= 0) {
+		if (result == -ERESTARTSYS) {
+			/* If we havne't sent anything we can just return BUSY,
+			 * however if we have sent something we need to make
+			 * sure we only allow this req to be sent until we are
+			 * completely done.
+			 */
+			if (sent) {
+				nsock->pending = req;
+				nsock->sent = sent;
+			}
+			return BLK_MQ_RQ_QUEUE_BUSY;
+		}
 		dev_err_ratelimited(disk_to_dev(nbd->disk),
 			"Send control failed (result %d)\n", result);
 		return -EIO;
 	}
-
+send_pages:
 	if (type != NBD_CMD_WRITE)
-		return 0;
+		goto out;
 
 	bio = req->bio;
 	while (bio) {
@@ -318,8 +348,25 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 				cmd, bvec.bv_len);
 			iov_iter_bvec(&from, ITER_BVEC | WRITE,
 				      &bvec, 1, bvec.bv_len);
-			result = sock_xmit(nbd, index, 1, &from, flags);
+			if (skip) {
+				if (skip >= iov_iter_count(&from)) {
+					skip -= iov_iter_count(&from);
+					continue;
+				}
+				iov_iter_advance(&from, skip);
+				skip = 0;
+			}
+			result = sock_xmit(nbd, index, 1, &from, flags, &sent);
 			if (result <= 0) {
+				if (result == -ERESTARTSYS) {
+					/* We've already sent the header, we
+					 * have no choice but to set pending and
+					 * return BUSY.
+					 */
+					nsock->pending = req;
+					nsock->sent = sent;
+					return BLK_MQ_RQ_QUEUE_BUSY;
+				}
 				dev_err(disk_to_dev(nbd->disk),
 					"Send data failed (result %d)\n",
 					result);
@@ -336,6 +383,9 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 		}
 		bio = next;
 	}
+out:
+	nsock->pending = NULL;
+	nsock->sent = 0;
 	return 0;
 }
 
@@ -353,7 +403,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 
 	reply.magic = 0;
 	iov_iter_kvec(&to, READ | ITER_KVEC, &iov, 1, sizeof(reply));
-	result = sock_xmit(nbd, index, 0, &to, MSG_WAITALL);
+	result = sock_xmit(nbd, index, 0, &to, MSG_WAITALL, NULL);
 	if (result <= 0) {
 		if (!test_bit(NBD_DISCONNECTED, &nbd->runtime_flags) &&
 		    !test_bit(NBD_DISCONNECT_REQUESTED, &nbd->runtime_flags))
@@ -383,7 +433,7 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 	if (ntohl(reply.error)) {
 		dev_err(disk_to_dev(nbd->disk), "Other side returned error (%d)\n",
 			ntohl(reply.error));
-		req->errors++;
+		req->errors = -EIO;
 		return cmd;
 	}
 
@@ -395,11 +445,11 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 		rq_for_each_segment(bvec, req, iter) {
 			iov_iter_bvec(&to, ITER_BVEC | READ,
 				      &bvec, 1, bvec.bv_len);
-			result = sock_xmit(nbd, index, 0, &to, MSG_WAITALL);
+			result = sock_xmit(nbd, index, 0, &to, MSG_WAITALL, NULL);
 			if (result <= 0) {
 				dev_err(disk_to_dev(nbd->disk), "Receive data failed (result %d)\n",
 					result);
-				req->errors++;
+				req->errors = -EIO;
 				return cmd;
 			}
 			dev_dbg(nbd_to_dev(nbd), "request %p: got %d bytes data\n",
@@ -469,7 +519,7 @@ static void nbd_clear_req(struct request *req, void *data, bool reserved)
 	if (!blk_mq_request_started(req))
 		return;
 	cmd = blk_mq_rq_to_pdu(req);
-	req->errors++;
+	req->errors = -EIO;
 	nbd_end_request(cmd);
 }
 
@@ -482,22 +532,23 @@ static void nbd_clear_que(struct nbd_device *nbd)
 }
 
 
-static void nbd_handle_cmd(struct nbd_cmd *cmd, int index)
+static int nbd_handle_cmd(struct nbd_cmd *cmd, int index)
 {
 	struct request *req = blk_mq_rq_from_pdu(cmd);
 	struct nbd_device *nbd = cmd->nbd;
 	struct nbd_sock *nsock;
+	int ret;
 
 	if (index >= nbd->num_connections) {
 		dev_err_ratelimited(disk_to_dev(nbd->disk),
 				    "Attempted send on invalid socket\n");
-		goto error_out;
+		return -EINVAL;
 	}
 
 	if (test_bit(NBD_DISCONNECTED, &nbd->runtime_flags)) {
 		dev_err_ratelimited(disk_to_dev(nbd->disk),
 				    "Attempted send on closed socket\n");
-		goto error_out;
+		return -EINVAL;
 	}
 
 	req->errors = 0;
@@ -508,29 +559,30 @@ static void nbd_handle_cmd(struct nbd_cmd *cmd, int index)
 		mutex_unlock(&nsock->tx_lock);
 		dev_err_ratelimited(disk_to_dev(nbd->disk),
 				    "Attempted send on closed socket\n");
-		goto error_out;
+		return -EINVAL;
 	}
 
-	if (nbd_send_cmd(nbd, cmd, index) != 0) {
-		dev_err_ratelimited(disk_to_dev(nbd->disk),
-				    "Request send failed\n");
-		req->errors++;
-		nbd_end_request(cmd);
+	/* Handle the case that we have a pending request that was partially
+	 * transmitted that _has_ to be serviced first.  We need to call requeue
+	 * here so that it gets put _after_ the request that is already on the
+	 * dispatch list.
+	 */
+	if (unlikely(nsock->pending && nsock->pending != req)) {
+		blk_mq_requeue_request(req, true);
+		ret = 0;
+		goto out;
 	}
-
+	ret = nbd_send_cmd(nbd, cmd, index);
+out:
 	mutex_unlock(&nsock->tx_lock);
-
-	return;
-
-error_out:
-	req->errors++;
-	nbd_end_request(cmd);
+	return ret;
 }
 
 static int nbd_queue_rq(struct blk_mq_hw_ctx *hctx,
 			const struct blk_mq_queue_data *bd)
 {
 	struct nbd_cmd *cmd = blk_mq_rq_to_pdu(bd->rq);
+	int ret;
 
 	/*
 	 * Since we look at the bio's to send the request over the network we
@@ -543,10 +595,20 @@ static int nbd_queue_rq(struct blk_mq_hw_ctx *hctx,
 	 */
 	init_completion(&cmd->send_complete);
 	blk_mq_start_request(bd->rq);
-	nbd_handle_cmd(cmd, hctx->queue_num);
+
+	/* We can be called directly from the user space process, which means we
+	 * could possibly have signals pending so our sendmsg will fail.  In
+	 * this case we need to return that we are busy, otherwise error out as
+	 * appropriate.
+	 */
+	ret = nbd_handle_cmd(cmd, hctx->queue_num);
+	if (ret < 0)
+		ret = BLK_MQ_RQ_QUEUE_ERROR;
+	if (!ret)
+		ret = BLK_MQ_RQ_QUEUE_OK;
 	complete(&cmd->send_complete);
 
-	return BLK_MQ_RQ_QUEUE_OK;
+	return ret;
 }
 
 static int nbd_add_socket(struct nbd_device *nbd, struct block_device *bdev,
@@ -581,6 +643,8 @@ static int nbd_add_socket(struct nbd_device *nbd, struct block_device *bdev,
 
 	mutex_init(&nsock->tx_lock);
 	nsock->sock = sock;
+	nsock->pending = NULL;
+	nsock->sent = 0;
 	socks[nbd->num_connections++] = nsock;
 
 	if (max_part)
@@ -602,6 +666,8 @@ static void nbd_reset(struct nbd_device *nbd)
 
 static void nbd_bdev_reset(struct block_device *bdev)
 {
+	if (bdev->bd_openers > 1)
+		return;
 	set_device_ro(bdev, false);
 	bdev->bd_inode->i_size = 0;
 	if (max_part > 0) {
@@ -634,7 +700,7 @@ static void send_disconnects(struct nbd_device *nbd)
 
 	for (i = 0; i < nbd->num_connections; i++) {
 		iov_iter_kvec(&from, WRITE | ITER_KVEC, &iov, 1, sizeof(request));
-		ret = sock_xmit(nbd, i, 1, &from, 0);
+		ret = sock_xmit(nbd, i, 1, &from, 0, NULL);
 		if (ret <= 0)
 			dev_err(disk_to_dev(nbd->disk),
 				"Send disconnect failed %d\n", ret);
@@ -665,7 +731,8 @@ static int nbd_clear_sock(struct nbd_device *nbd, struct block_device *bdev)
 {
 	sock_shutdown(nbd);
 	nbd_clear_que(nbd);
-	kill_bdev(bdev);
+
+	__invalidate_device(bdev, true);
 	nbd_bdev_reset(bdev);
 	/*
 	 * We want to give the run thread a chance to wait for everybody
@@ -781,7 +848,10 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 		nbd_size_set(nbd, bdev, nbd->blksize, arg);
 		return 0;
 	case NBD_SET_TIMEOUT:
-		nbd->tag_set.timeout = arg * HZ;
+		if (arg) {
+			nbd->tag_set.timeout = arg * HZ;
+			blk_queue_rq_timeout(nbd->disk->queue, arg * HZ);
+		}
 		return 0;
 
 	case NBD_SET_FLAGS: