From: Sagi Grimberg <sagi@xxxxxxxxxxx>
[ Upstream commit adc99fd378398f4c58798a1c57889872967d56a6 ]
If the controller died exactly when we are receiving icresp
we hang because icresp may never return. Make sure to set a
high finite limit.
Fixes: 3f2304f8c6d6 ("nvme-tcp: add NVMe over TCP host driver")
Signed-off-by: Sagi Grimberg <sagi@xxxxxxxxxxx>
Signed-off-by: Christoph Hellwig <hch@xxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
drivers/nvme/host/tcp.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index 4862fa962011d..26461bf3fdcc3 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -1392,6 +1392,9 @@ static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl,
}
}
+ /* Set 10 seconds timeout for icresp recvmsg */
+ queue->sock->sk->sk_rcvtimeo = 10 * HZ;
+
queue->sock->sk->sk_allocation = GFP_ATOMIC;
nvme_tcp_set_queue_io_cpu(queue);
queue->request = NULL;
--
2.25.1
