Messages in this thread | | | From | David Laight <> | Subject | RE: [PATCH v6 2/4] net: moxa: replace build_skb() with netdev_alloc_skb_ip_align() / memcpy() | Date | Tue, 26 Aug 2014 09:10:19 +0000 |
| |
From: Arnd Bergmann > On Monday 25 August 2014 16:22:22 Jonas Jensen wrote: > > @@ -226,13 +226,15 @@ static int moxart_rx_poll(struct napi_struct *napi, int budget) > > if (len > RX_BUF_SIZE) > > len = RX_BUF_SIZE; > > > > - skb = build_skb(priv->rx_buf[rx_head], priv->rx_buf_size); > > + skb = netdev_alloc_skb_ip_align(ndev, len); > > + > > if (unlikely(!skb)) { > > - net_dbg_ratelimited("build_skb failed\n"); > > + net_dbg_ratelimited("netdev_alloc_skb_ip_align failed\n"); > > priv->stats.rx_dropped++; > > priv->stats.rx_errors++; > > } > > > > + memcpy(skb->data, priv->rx_buf[rx_head], len);
Is this memcpy() aligned? If the hardware can receive to a 4n+2 offset it is probably better to copy the two bytes before the frame data and to round the length of to a whole number of words.
> > skb_put(skb, len); > > skb->protocol = eth_type_trans(skb, ndev); > > napi_gro_receive(&priv->napi, skb); > > While this seems correct, I wonder why you don't do the normal approach of > dequeuing the skb from the chain and adding a newly allocated skb to it to > save the memcpy.
Because the receive buffer area isn't made of skbs. Post-allocating the skb also reduces the 'true size' of the skb.
David
| |