[ofa-general] [PATCH 4/10] IB/ipoib: Add LSO support to ipoib

[Or Gerlitz]

Eli Cohen wrote:
> --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
> +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
> @@ -418,14 +449,35 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb,
>  {
> +	int hlen;
> +	void *phead;
> +
> +	if (!skb_is_gso(skb)) {
> +		if (unlikely(skb->len > priv->mcast_mtu + IPOIB_ENCAP_LEN)) {
> +			ipoib_warn(priv, "packet len %d (> %d) too long to send, dropping\n",
> +				   skb->len, priv->mcast_mtu + IPOIB_ENCAP_LEN);
> +			++dev->stats.tx_dropped;
> +			++dev->stats.tx_errors;
> +			ipoib_cm_skb_too_long(dev, skb, priv->mcast_mtu);
> +			return;
> +		}
> +		phead = 0;
> +		hlen = 0;
> +	} else {
> +		/*
> +		 * LSO header is limited to max 60 bytes
> +		 */
> +		if (unlikely((ip_hdr(skb)->ihl + tcp_hdr(skb)->doff) > 15)) {
> +			ipoib_warn(priv, "ip(%d) and tcp(%d) headers too long, dropping skb\n",
> +				   ip_hdr(skb)->ihl << 2, tcp_hdr(skb)->doff << 2);
> +			goto drop;
> +		}
>   
is the 60 bytes being a limitation of the connectX HW, the Linux kernel 
stack or some "lso spec"?
> +		hlen = ((ip_hdr(skb)->ihl + tcp_hdr(skb)->doff) << 2) + IPOIB_ENCAP_LEN;
> +		phead = skb->data;
>   
Looking in e1000 tso code (eg 
http://lxr.linux.no/linux/drivers/net/e1000/e1000_main.c#L2884) I see 
that the header len is gotten by
skb_transport_offset(skb) + tcp_hdrlen(skb), is that what 
((ip_hdr(skb)->ihl + tcp_hdr(skb)->doff) << 2) gives? 
> @@ -470,6 +521,12 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb,
>  			netif_stop_queue(dev);
>  		}
>  	}
> +	return;
> +
> +drop:
> +	++dev->stats.tx_errors;
> +	dev_kfree_skb_any(skb);
> +	return;
>   
shouldn't the tx_dropped counter be incremented here?

Or.