LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH 1/3] nbd: fix connection timed out error after reconnecting to server
@ 2019-05-24  9:43 Yao Liu
  2019-05-24  9:43 ` [PATCH 2/3] nbd: notify userland even if nbd has already disconnected Yao Liu
                   ` (2 more replies)
  0 siblings, 3 replies; 15+ messages in thread
From: Yao Liu @ 2019-05-24  9:43 UTC (permalink / raw)
  To: Josef Bacik, Jens Axboe; +Cc: linux-block, nbd, linux-kernel

Some I/O requests that have been sent succussfully but have not yet been
replied won't be resubmitted after reconnecting because of server restart,
so we add a list to track them.

Signed-off-by: Yao Liu <yotta.liu@ucloud.cn>
---
 drivers/block/nbd.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 46 insertions(+)

diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 053958a..ca69d6e 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -113,6 +113,8 @@ struct nbd_device {
 	struct list_head list;
 	struct task_struct *task_recv;
 	struct task_struct *task_setup;
+	struct mutex outstanding_lock;
+	struct list_head outstanding_queue;
 };
 
 #define NBD_CMD_REQUEUED	1
@@ -125,6 +127,7 @@ struct nbd_cmd {
 	blk_status_t status;
 	unsigned long flags;
 	u32 cmd_cookie;
+	struct list_head outstanding_entry;
 };
 
 #if IS_ENABLED(CONFIG_DEBUG_FS)
@@ -619,6 +622,24 @@ static int nbd_send_cmd(struct nbd_device *nbd, struct nbd_cmd *cmd, int index)
 	return 0;
 }
 
+static struct nbd_cmd *nbd_get_cmd(struct nbd_device *nbd,
+					struct nbd_cmd *xcmd)
+{
+	struct nbd_cmd *cmd, *tmp;
+
+	mutex_lock(&nbd->outstanding_lock);
+	list_for_each_entry_safe(cmd, tmp, &nbd->outstanding_queue, outstanding_entry) {
+		if (cmd != xcmd)
+			continue;
+		list_del_init(&cmd->outstanding_entry);
+		mutex_unlock(&nbd->outstanding_lock);
+		return cmd;
+	}
+	mutex_unlock(&nbd->outstanding_lock);
+
+	return ERR_PTR(-ENOENT);
+}
+
 /* NULL returned = something went wrong, inform userspace */
 static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 {
@@ -714,12 +735,30 @@ static struct nbd_cmd *nbd_read_stat(struct nbd_device *nbd, int index)
 				req, bvec.bv_len);
 		}
 	}
+	cmd = nbd_get_cmd(nbd, cmd);
+	if (IS_ERR(cmd)) {
+		dev_err(disk_to_dev(nbd->disk), "Unexpected reply (%d) %p which not in outstanding queue\n",
+			tag, req);
+		ret = -ENOENT;
+	}
 out:
 	trace_nbd_payload_received(req, handle);
 	mutex_unlock(&cmd->lock);
 	return ret ? ERR_PTR(ret) : cmd;
 }
 
+static void nbd_requeue_outstanding(struct nbd_device *nbd)
+{
+	struct nbd_cmd *cmd, *tmp;
+
+	mutex_lock(&nbd->outstanding_lock);
+	list_for_each_entry_safe(cmd, tmp, &nbd->outstanding_queue, outstanding_entry) {
+		nbd_requeue_cmd(cmd);
+		list_del_init(&cmd->outstanding_entry);
+	}
+	mutex_unlock(&nbd->outstanding_lock);
+}
+
 static void recv_work(struct work_struct *work)
 {
 	struct recv_thread_args *args = container_of(work,
@@ -742,6 +781,7 @@ static void recv_work(struct work_struct *work)
 
 		blk_mq_complete_request(blk_mq_rq_from_pdu(cmd));
 	}
+	nbd_requeue_outstanding(nbd);
 	atomic_dec(&config->recv_threads);
 	wake_up(&config->recv_wq);
 	nbd_config_put(nbd);
@@ -892,6 +932,10 @@ static int nbd_handle_cmd(struct nbd_cmd *cmd, int index)
 		nbd_mark_nsock_dead(nbd, nsock, 1);
 		nbd_requeue_cmd(cmd);
 		ret = 0;
+	} else if (ret == 0) {
+		mutex_lock(&nbd->outstanding_lock);
+		list_add_tail(&cmd->outstanding_entry, &nbd->outstanding_queue);
+		mutex_unlock(&nbd->outstanding_lock);
 	}
 out:
 	mutex_unlock(&nsock->tx_lock);
@@ -1615,6 +1659,8 @@ static int nbd_dev_add(int index)
 	refcount_set(&nbd->config_refs, 0);
 	refcount_set(&nbd->refs, 1);
 	INIT_LIST_HEAD(&nbd->list);
+	mutex_init(&nbd->outstanding_lock);
+	INIT_LIST_HEAD(&nbd->outstanding_queue);
 	disk->major = NBD_MAJOR;
 	disk->first_minor = index << part_shift;
 	disk->fops = &nbd_fops;
-- 
1.8.3.1


^ permalink raw reply	[flat|nested] 15+ messages in thread

end of thread, other threads:[~2019-05-29 13:49 UTC | newest]

Thread overview: 15+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-05-24  9:43 [PATCH 1/3] nbd: fix connection timed out error after reconnecting to server Yao Liu
2019-05-24  9:43 ` [PATCH 2/3] nbd: notify userland even if nbd has already disconnected Yao Liu
2019-05-24 13:08   ` Josef Bacik
2019-05-27 18:23     ` Yao Liu
2019-05-28 16:36       ` Mike Christie
2019-05-28 20:05         ` Yao Liu
2019-05-28 16:54       ` Josef Bacik
2019-05-24  9:43 ` [PATCH 3/3] nbd: mark sock as dead even if it's the last one Yao Liu
2019-05-24 13:17   ` Josef Bacik
2019-05-27 18:29     ` Yao Liu
2019-05-24 13:07 ` [PATCH 1/3] nbd: fix connection timed out error after reconnecting to server Josef Bacik
2019-05-27 18:07   ` Yao Liu
2019-05-28 16:57     ` Josef Bacik
2019-05-28 19:04       ` Yao Liu
2019-05-29 13:49         ` Josef Bacik

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).