On 2018/9/20 0:14, Jason Gunthorpe wrote:

On Wed, Sep 19, 2018 at 08:28:38PM +0800, YueHaibing wrote:

quoted

rdma_set_src_addr_rcu should check copy_src_l2_addr fails,
rather than always return 0. Also copy_src_l2_addr should return
'ret' as its return value while rdma_translate_ip fails.

Fixes: c31d4b2ddf07 ("RDMA/core: Protect against changing dst->dev during destination resolve")
Signed-off-by: YueHaibing <redacted>
Reviewed-by: Parav Pandit <redacted>
---
 drivers/infiniband/core/addr.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

Applied to for-next, thanks

quoted

diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c
index 7a0356c..8a31b11 100644
--- a/drivers/infiniband/core/addr.c
+++ b/drivers/infiniband/core/addr.c

@@ -468,7 +468,7 @@ static int addr_resolve_neigh(const struct dst_entry *dst,
 	return ret;
 }
 
-static void copy_src_l2_addr(struct rdma_dev_addr *dev_addr,
+static int copy_src_l2_addr(struct rdma_dev_addr *dev_addr,
 			     const struct sockaddr *dst_in,
 			     const struct dst_entry *dst,
 			     const struct net_device *ndev)

I fixed the indenting as Parav pointed out

quoted

@@ -492,6 +492,8 @@ static void copy_src_l2_addr(struct rdma_dev_addr *dev_addr,
 						RDMA_NETWORK_IPV6;
 	else
 		dev_addr->network = RDMA_NETWORK_IB;
+
+	return ret;
 }
 
 static int rdma_set_src_addr_rcu(struct rdma_dev_addr *dev_addr,

@@ -499,6 +501,7 @@ static int rdma_set_src_addr_rcu(struct rdma_dev_addr *dev_addr,
 				 const struct sockaddr *dst_in,
 				 const struct dst_entry *dst)
 {
+	int ret;
 	struct net_device *ndev = READ_ONCE(dst->dev);
 
 	*ndev_flags = ndev->flags;

@@ -515,8 +518,8 @@ static int rdma_set_src_addr_rcu(struct rdma_dev_addr *dev_addr,
 			return -ENODEV;
 	}
 
-	copy_src_l2_addr(dev_addr, dst_in, dst, ndev);
-	return 0;
+	ret = copy_src_l2_addr(dev_addr, dst_in, dst, ndev);
+	return ret;
 }

And made this just 'return copy_src_l2_addr' without the ret.

Thank you.

Jason

.