[PATCH 1/1] staging: gdm72xx: fix an skb memory leak

[Javier Martinez Canillas]

The NLMSG_PUT() macro contains a hidden goto that jumps to the
nlmsg_failure label. Since the sk_buff was allocated before the macro,
jumping to the nlmsg_failure label leaks the memory allocated for it.

Calling kfree() before returning would fix it, but is better to avoid
using this error prone macro and use nlmsg_put() instead.

Also, use nlmsg_data() instead of NLMSG_DATA() to check type.

Signed-off-by: Javier Martinez Canillas <javier at dowhile0.org>
---

It just was compiled tested since I don't have the hardware

 drivers/staging/gdm72xx/netlink_k.c |   10 +++++++---
 1 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/staging/gdm72xx/netlink_k.c b/drivers/staging/gdm72xx/netlink_k.c
index 9fa432d..064815b 100644
--- a/drivers/staging/gdm72xx/netlink_k.c
+++ b/drivers/staging/gdm72xx/netlink_k.c
@@ -126,8 +126,13 @@ int netlink_send(struct sock *sock, int group, u16 type, void *msg, int len)
 	}
 
 	seq++;
-	nlh = NLMSG_PUT(skb, 0, seq, type, len);
-	memcpy(NLMSG_DATA(nlh), msg, len);
+	nlh = nlmsg_put(skb, 0, seq, type, len, 0);
+	if (!nlh) {
+		kfree_skb(skb);
+		return -EMSGSIZE;
+	}
+
+	memcpy(nlmsg_data(nlh), msg, len);
 
 	NETLINK_CB(skb).pid = 0;
 	NETLINK_CB(skb).dst_group = 0;
@@ -144,6 +149,5 @@ int netlink_send(struct sock *sock, int group, u16 type, void *msg, int len)
 		ret = 0;
 	}
 
-nlmsg_failure:
 	return ret;
 }
-- 
1.7.7.6