[PATCH V2] nvme-fabrics: fix I/O connect error handling

[Chaitanya Kulkarni <kch@nvidia.com>]

In nvmf_connect_io_queue(), if connect I/O command fails, we log the
error and continue for authentication. This overrides error captured
from __nvme_submit_sync_cmd(), causing wrong return value.

Add goto out_free_data after logging connect error to fix the issue.

Fixes: f50fff73d620c ("nvme: implement In-Band authentication")
Signed-off-by: Chaitanya Kulkarni <kch@nvidia.com>
---
V2: add fixes tag (Christoph)

 drivers/nvme/host/fabrics.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index 3499acbf6a82..495c171daead 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -534,6 +534,7 @@ int nvmf_connect_io_queue(struct nvme_ctrl *ctrl, u16 qid)
 	if (ret) {
 		nvmf_log_connect_error(ctrl, ret, le32_to_cpu(res.u32),
 				       &cmd, data);
+		goto out_free_data;
 	}
 	result = le32_to_cpu(res.u32);
 	if (result & (NVME_CONNECT_AUTHREQ_ATR | NVME_CONNECT_AUTHREQ_ASCR)) {
-- 
2.40.0