Nuke HAS_IP_COPYSUM for net drivers

PATCH In review Roger Luethi 2 participants 2 messages May 25, 2004 – May 27, 2004

Subsystems (2)

Networking Network drivers

lore

Share & tools

lore

Roger Luethi

author

2004-05-25 19:11 UTC #0 lore raw

PATCH 1/1 Nuke HAS_IP_COPYSUM for net drivers

Entirely untested, but Obviously Correct(TM). HAS_IP_COPYSUM has been utterly meaningless for a long time.

Respective patch for via-rhine follows once I know the state of its pending patches (which I wouldn't mind getting feedback on (hint, hint!)).

Roger

drivers/net/epic100.c +0 -5

diff -urp linux-2.6.6/drivers/net/epic100.c linux-2.6.6-patch/drivers/net/epic100.c
@@ -1215,13 +1215,8 @@ static int epic_rx(struct net_device *de
 							    ep->rx_ring[entry].bufaddr,
 							    ep->rx_buf_sz,
 							    PCI_DMA_FROMDEVICE);
-#if 1 /* HAS_IP_COPYSUM */
 				eth_copy_and_sum(skb, ep->rx_skbuff[entry]->tail, pkt_len, 0);
 				skb_put(skb, pkt_len);
-#else
-				memcpy(skb_put(skb, pkt_len), ep->rx_skbuff[entry]->tail,
-					   pkt_len);
-#endif
 				pci_dma_sync_single_for_device(ep->pci_dev,
 							       ep->rx_ring[entry].bufaddr,
 							       ep->rx_buf_sz,

drivers/net/natsemi.c +0 -5

diff -urp linux-2.6.6/drivers/net/natsemi.c linux-2.6.6-patch/drivers/net/natsemi.c
@@ -1798,14 +1798,9 @@ static void netdev_rx(struct net_device 
 					np->rx_dma[entry],
 					buflen,
 					PCI_DMA_FROMDEVICE);
-#if HAS_IP_COPYSUM
 				eth_copy_and_sum(skb,
 					np->rx_skbuff[entry]->tail, pkt_len, 0);
 				skb_put(skb, pkt_len);
-#else
-				memcpy(skb_put(skb, pkt_len),
-					np->rx_skbuff[entry]->tail, pkt_len);
-#endif
 				pci_dma_sync_single_for_device(np->pci_dev,
 					np->rx_dma[entry],
 					buflen,

drivers/net/tulip/winbond-840.c +0 -6

diff -urp linux-2.6.6/drivers/net/tulip/winbond-840.c linux-2.6.6-patch/drivers/net/tulip/winbond-840.c
@@ -1292,14 +1292,8 @@ static int netdev_rx(struct net_device *
 				pci_dma_sync_single_for_cpu(np->pci_dev,np->rx_addr[entry],
 							    np->rx_skbuff[entry]->len,
 							    PCI_DMA_FROMDEVICE);
-				/* Call copy + cksum if available. */
-#if HAS_IP_COPYSUM
 				eth_copy_and_sum(skb, np->rx_skbuff[entry]->tail, pkt_len, 0);
 				skb_put(skb, pkt_len);
-#else
-				memcpy(skb_put(skb, pkt_len), np->rx_skbuff[entry]->tail,
-					   pkt_len);
-#endif
 				pci_dma_sync_single_for_device(np->pci_dev,np->rx_addr[entry],
 							       np->rx_skbuff[entry]->len,
 							       PCI_DMA_FROMDEVICE);

drivers/net/yellowfin.c +0 -5

diff -urp linux-2.6.6/drivers/net/yellowfin.c linux-2.6.6-patch/drivers/net/yellowfin.c
@@ -1201,13 +1201,8 @@ static int yellowfin_rx(struct net_devic
 					break;
 				skb->dev = dev;
 				skb_reserve(skb, 2);	/* 16 byte align the IP header */
-#if HAS_IP_COPYSUM
 				eth_copy_and_sum(skb, rx_skb->tail, pkt_len, 0);
 				skb_put(skb, pkt_len);
-#else
-				memcpy(skb_put(skb, pkt_len), 
-					rx_skb->tail, pkt_len);
-#endif
 				pci_dma_sync_single_for_device(yp->pci_dev, desc->addr,
 											   yp->rx_buf_sz,
 											   PCI_DMA_FROMDEVICE);

Jeff Garzik

2004-05-27 17:36 UTC #1 lore raw

Quoted 2 lines · Roger Luethi

Entirely untested, but Obviously Correct(TM). HAS_IP_COPYSUM has been utterly meaningless for a long time.

I will apply to netdev-2.6 queue, which means it will get testing and review via Andrew's -mm tree.

After 2.6.7 is released, I'll push upstream.

Quoted 3 lines

Respective patch for via-rhine follows once I know the state of its pending patches (which I wouldn't mind getting feedback on (hint, hint!)).

Working through the backlog today, yell at me if you haven't heard anything by the end of the day :)

Jeff