sky2: optimize transmit completion

Don't reference the list element in hardware transmit ring on transmit
completion. The list element is updated by hardware, therefore
it causes a cache miss. Do book keeping in software structure.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 7c1880d8b31ff6300973dfc4dae4c10cd86e94f7..a3e40d70c07167fdb87d4502fb5a639b24117edc 100644 (file)
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -1001,8 +1001,11 @@ static void sky2_prefetch_init(struct sky2_hw *hw, u32 qaddr,
 static inline struct sky2_tx_le *get_tx_le(struct sky2_port *sky2, u16 *slot)
 {
        struct sky2_tx_le *le = sky2->tx_le + *slot;
+       struct tx_ring_info *re = sky2->tx_ring + *slot;
 
        *slot = RING_NEXT(*slot, sky2->tx_ring_size);
+       re->flags = 0;
+       re->skb = NULL;
        le->ctrl = 0;
        return le;
 }
@@ -1021,12 +1024,6 @@ static void tx_init(struct sky2_port *sky2)
        sky2->tx_last_upper = 0;
 }
 
-static inline struct tx_ring_info *tx_le_re(struct sky2_port *sky2,
-                                           struct sky2_tx_le *le)
-{
-       return sky2->tx_ring + (le - sky2->tx_le);
-}
-
 /* Update chip's next pointer */
 static inline void sky2_put_idx(struct sky2_hw *hw, unsigned q, u16 idx)
 {
@@ -1563,6 +1560,19 @@ static unsigned tx_le_req(const struct sk_buff *skb)
        return count;
 }
 
+static void sky2_tx_unmap(struct pci_dev *pdev,
+                         const struct tx_ring_info *re)
+{
+       if (re->flags & TX_MAP_SINGLE)
+               pci_unmap_single(pdev, pci_unmap_addr(re, mapaddr),
+                                pci_unmap_len(re, maplen),
+                                PCI_DMA_TODEVICE);
+       else if (re->flags & TX_MAP_PAGE)
+               pci_unmap_page(pdev, pci_unmap_addr(re, mapaddr),
+                              pci_unmap_len(re, maplen),
+                              PCI_DMA_TODEVICE);
+}
+
 /*
  * Put one packet in ring for transmit.
  * A single packet can generate multiple list elements, and
@@ -1667,16 +1677,17 @@ static int sky2_xmit_frame(struct sk_buff *skb, struct net_device *dev)
                }
        }
 
+       re = sky2->tx_ring + slot;
+       re->flags = TX_MAP_SINGLE;
+       pci_unmap_addr_set(re, mapaddr, mapping);
+       pci_unmap_len_set(re, maplen, len);
+
        le = get_tx_le(sky2, &slot);
        le->addr = cpu_to_le32(lower_32_bits(mapping));
        le->length = cpu_to_le16(len);
        le->ctrl = ctrl;
        le->opcode = mss ? (OP_LARGESEND | HW_OWNER) : (OP_PACKET | HW_OWNER);
 
-       re = tx_le_re(sky2, le);
-       re->skb = skb;
-       pci_unmap_addr_set(re, mapaddr, mapping);
-       pci_unmap_len_set(re, maplen, len);
 
        for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
                const skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
@@ -1695,18 +1706,19 @@ static int sky2_xmit_frame(struct sk_buff *skb, struct net_device *dev)
                        le->opcode = OP_ADDR64 | HW_OWNER;
                }
 
+               re = sky2->tx_ring + slot;
+               re->flags = TX_MAP_PAGE;
+               pci_unmap_addr_set(re, mapaddr, mapping);
+               pci_unmap_len_set(re, maplen, frag->size);
+
                le = get_tx_le(sky2, &slot);
                le->addr = cpu_to_le32(lower_32_bits(mapping));
                le->length = cpu_to_le16(frag->size);
                le->ctrl = ctrl;
                le->opcode = OP_BUFFER | HW_OWNER;
-
-               re = tx_le_re(sky2, le);
-               re->skb = skb;
-               pci_unmap_addr_set(re, mapaddr, mapping);
-               pci_unmap_len_set(re, maplen, frag->size);
        }
 
+       re->skb = skb;
        le->ctrl |= EOP;
 
        sky2->tx_prod = slot;
@@ -1720,23 +1732,9 @@ static int sky2_xmit_frame(struct sk_buff *skb, struct net_device *dev)
 
 mapping_unwind:
        for (i = sky2->tx_prod; i != slot; i = RING_NEXT(i, sky2->tx_ring_size)) {
-               le = sky2->tx_le + i;
                re = sky2->tx_ring + i;
 
-               switch(le->opcode & ~HW_OWNER) {
-               case OP_LARGESEND:
-               case OP_PACKET:
-                       pci_unmap_single(hw->pdev,
-                                        pci_unmap_addr(re, mapaddr),
-                                        pci_unmap_len(re, maplen),
-                                        PCI_DMA_TODEVICE);
-                       break;
-               case OP_BUFFER:
-                       pci_unmap_page(hw->pdev, pci_unmap_addr(re, mapaddr),
-                                      pci_unmap_len(re, maplen),
-                                      PCI_DMA_TODEVICE);
-                       break;
-               }
+               sky2_tx_unmap(hw->pdev, re);
        }
 
 mapping_error:
@@ -1759,34 +1757,18 @@ mapping_error:
 static void sky2_tx_complete(struct sky2_port *sky2, u16 done)
 {
        struct net_device *dev = sky2->netdev;
-       struct pci_dev *pdev = sky2->hw->pdev;
        unsigned idx;
 
        BUG_ON(done >= sky2->tx_ring_size);
 
        for (idx = sky2->tx_cons; idx != done;
             idx = RING_NEXT(idx, sky2->tx_ring_size)) {
-               struct sky2_tx_le *le = sky2->tx_le + idx;
                struct tx_ring_info *re = sky2->tx_ring + idx;
+               struct sk_buff *skb = re->skb;
 
-               switch(le->opcode & ~HW_OWNER) {
-               case OP_LARGESEND:
-               case OP_PACKET:
-                       pci_unmap_single(pdev,
-                                        pci_unmap_addr(re, mapaddr),
-                                        pci_unmap_len(re, maplen),
-                                        PCI_DMA_TODEVICE);
-                       break;
-               case OP_BUFFER:
-                       pci_unmap_page(pdev, pci_unmap_addr(re, mapaddr),
-                                      pci_unmap_len(re, maplen),
-                                      PCI_DMA_TODEVICE);
-                       break;
-               }
-
-               if (le->ctrl & EOP) {
-                       struct sk_buff *skb = re->skb;
+               sky2_tx_unmap(sky2->hw->pdev, re);
 
+               if (skb) {
                        if (unlikely(netif_msg_tx_done(sky2)))
                                printk(KERN_DEBUG "%s: tx done %u\n",
                                       dev->name, idx);

diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h
index 2c262f763f93f0172c8e7f5094f5dd33c10322dc..9d07a466aec16707282f37f08c6b4026b3cab695 100644 (file)
--- a/drivers/net/sky2.h
+++ b/drivers/net/sky2.h
@@ -1984,6 +1984,9 @@ struct sky2_status_le {
 
 struct tx_ring_info {
        struct sk_buff  *skb;
+       unsigned long flags;
+#define TX_MAP_SINGLE   0x0001
+#define TX_MAP_PAGE     000002
        DECLARE_PCI_UNMAP_ADDR(mapaddr);
        DECLARE_PCI_UNMAP_LEN(maplen);
 };