Re: [PATCH net-next-2.6 v3] can: Unify droping of invalid tx skbs and netdev stats

Oliver Hartkopp wrote:

David Miller wrote:

quoted

From: Oliver Hartkopp <redacted>
Date: Tue, 12 Jan 2010 00:19:21 +0100

quoted

 static netdev_tx_t vcan_tx(struct sk_buff *skb, struct net_device *dev)
 {
+	struct can_frame *cf = (struct can_frame *)skb->data;
 	struct net_device_stats *stats = &dev->stats;
 	int loop;
 
+	if (unlikely(skb->len != sizeof(*cf) || cf->can_dlc > 8)) {
+		kfree_skb(skb);
+		stats->tx_dropped++;
+		return NETDEV_TX_OK;
+	}
+

 ...

quoted

+/* Drop a given socketbuffer if it does not contain a valid CAN frame. */
+static inline int can_dropped_invalid_skb(struct net_device *dev,
+					  struct sk_buff *skb)
+{
+	const struct can_frame *cf = (struct can_frame *)skb->data;
+
+	if (unlikely(skb->len != sizeof(*cf) || cf->can_dlc > 8)) {
+		kfree_skb(skb);
+		dev->stats.tx_dropped++;
+		return 1;
+	}
+
+	return 0;
+}
+
 struct net_device *alloc_candev(int sizeof_priv, unsigned int echo_skb_max);

Why are you not using the new helper function in vcan_tx()?

I just wanted the vcan driver keep off including "include/linux/can/dev.h"
which is intended for 'real' CAN hardware.

As the vcan software devices do not need the bitrate-setting and skb echo
handling from the driver library for real CAN devices, this inline function
would be the only reason to include ".../dev.h"

But i don't have a strong preference to do it like this. Do you think i
should change it to used the defined inline function?

Yes, for the sake of consistency.

Wolfgang.