Merge branch 'netdev_page_frags'

Alexander Duyck says:

====================
Refactor netdev page frags and move them into mm/

This patch series addresses several things.

First I found an issue in the performance of the pfmemalloc check from
build_skb.  To work around it I have provided a cached copy of pfmemalloc
to be used in __netdev_alloc_skb and __napi_alloc_skb.

Second I moved the page fragment allocation logic into the mm tree and
added functionality for freeing page fragments.  I had to fix igb before I
could do this as it was using a reference to NETDEV_FRAG_PAGE_MAX_SIZE
incorrectly.

Finally I went through and replaced all of the duplicate code that was
calling put_page and replaced it with calls to skb_free_frag.

With these changes in place a simple receive and drop test increased from a
packet rate of 8.9Mpps to 9.8Mpps.  The gains breakdown as follows:

8.9Mpps	Before			9.8Mpps	After
------------------------	------------------------
7.8%	put_compound_page	9.1%	__free_page_frag
3.9%	skb_free_head
1.1%	put_page

4.9%	build_skb		3.8%	__napi_alloc_skb
2.5%	__alloc_rx_skb
1.9%	__napi_alloc_skb
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

7 files changed, 15 insertions, 25 deletions

diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
index a8bb8f664d3d..b10d1744e5ae 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
@@ -662,7 +662,7 @@ static int bnx2x_fill_frag_skb(struct bnx2x *bp, struct bnx2x_fastpath *fp,
 static void bnx2x_frag_free(const struct bnx2x_fastpath *fp, void *data)
 {
 	if (fp->rx_frag_size)
-		put_page(virt_to_head_page(data));
+		skb_free_frag(data);
 	else
 		kfree(data);
 }
diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c
index 069952fa5d64..73c934cf6c61 100644
--- a/drivers/net/ethernet/broadcom/tg3.c
+++ b/drivers/net/ethernet/broadcom/tg3.c
@@ -6618,7 +6618,7 @@ static void tg3_tx(struct tg3_napi *tnapi)
 static void tg3_frag_free(bool is_frag, void *data)
 {
 	if (is_frag)
-		put_page(virt_to_head_page(data));
+		skb_free_frag(data);
 	else
 		kfree(data);
 }
diff --git a/drivers/net/ethernet/hisilicon/hip04_eth.c b/drivers/net/ethernet/hisilicon/hip04_eth.c
index 3b39fdddeb57..d49bee38cd31 100644
--- a/drivers/net/ethernet/hisilicon/hip04_eth.c
+++ b/drivers/net/ethernet/hisilicon/hip04_eth.c
@@ -798,7 +798,7 @@ static void hip04_free_ring(struct net_device *ndev, struct device *d)
 
 	for (i = 0; i < RX_DESC_NUM; i++)
 		if (priv->rx_buf[i])
-			put_page(virt_to_head_page(priv->rx_buf[i]));
+			skb_free_frag(priv->rx_buf[i]);
 
 	for (i = 0; i < TX_DESC_NUM; i++)
 		if (priv->tx_skb[i])
diff --git a/drivers/net/ethernet/intel/e1000/e1000_main.c b/drivers/net/ethernet/intel/e1000/e1000_main.c
index 983eb4e6f7aa..74dc15055971 100644
--- a/drivers/net/ethernet/intel/e1000/e1000_main.c
+++ b/drivers/net/ethernet/intel/e1000/e1000_main.c
@@ -2079,11 +2079,6 @@ static void *e1000_alloc_frag(const struct e1000_adapter *a)
 	return data;
 }
 
-static void e1000_free_frag(const void *data)
-{
-	put_page(virt_to_head_page(data));
-}
-
 /**
  * e1000_clean_rx_ring - Free Rx Buffers per Queue
  * @adapter: board private structure
@@ -2107,7 +2102,7 @@ static void e1000_clean_rx_ring(struct e1000_adapter *adapter,
 						 adapter->rx_buffer_len,
 						 DMA_FROM_DEVICE);
 			if (buffer_info->rxbuf.data) {
-				e1000_free_frag(buffer_info->rxbuf.data);
+				skb_free_frag(buffer_info->rxbuf.data);
 				buffer_info->rxbuf.data = NULL;
 			}
 		} else if (adapter->clean_rx == e1000_clean_jumbo_rx_irq) {
@@ -4594,28 +4589,28 @@ static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
 			data = e1000_alloc_frag(adapter);
 			/* Failed allocation, critical failure */
 			if (!data) {
-				e1000_free_frag(olddata);
+				skb_free_frag(olddata);
 				adapter->alloc_rx_buff_failed++;
 				break;
 			}
 
 			if (!e1000_check_64k_bound(adapter, data, bufsz)) {
 				/* give up */
-				e1000_free_frag(data);
-				e1000_free_frag(olddata);
+				skb_free_frag(data);
+				skb_free_frag(olddata);
 				adapter->alloc_rx_buff_failed++;
 				break;
 			}
 
 			/* Use new allocation */
-			e1000_free_frag(olddata);
+			skb_free_frag(olddata);
 		}
 		buffer_info->dma = dma_map_single(&pdev->dev,
 						  data,
 						  adapter->rx_buffer_len,
 						  DMA_FROM_DEVICE);
 		if (dma_mapping_error(&pdev->dev, buffer_info->dma)) {
-			e1000_free_frag(data);
+			skb_free_frag(data);
 			buffer_info->dma = 0;
 			adapter->alloc_rx_buff_failed++;
 			break;
@@ -4637,7 +4632,7 @@ static void e1000_alloc_rx_buffers(struct e1000_adapter *adapter,
 					 adapter->rx_buffer_len,
 					 DMA_FROM_DEVICE);
 
-			e1000_free_frag(data);
+			skb_free_frag(data);
 			buffer_info->rxbuf.data = NULL;
 			buffer_info->dma = 0;
 
diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c
index e63664649029..8a45ed7506c5 100644
--- a/drivers/net/ethernet/intel/igb/igb_main.c
+++ b/drivers/net/ethernet/intel/igb/igb_main.c
@@ -4974,6 +4974,7 @@ netdev_tx_t igb_xmit_frame_ring(struct sk_buff *skb,
 	struct igb_tx_buffer *first;
 	int tso;
 	u32 tx_flags = 0;
+	unsigned short f;
 	u16 count = TXD_USE_COUNT(skb_headlen(skb));
 	__be16 protocol = vlan_get_protocol(skb);
 	u8 hdr_len = 0;
@@ -4984,14 +4985,8 @@ netdev_tx_t igb_xmit_frame_ring(struct sk_buff *skb,
 	 *       + 1 desc for context descriptor,
 	 * otherwise try next time
 	 */
-	if (NETDEV_FRAG_PAGE_MAX_SIZE > IGB_MAX_DATA_PER_TXD) {
-		unsigned short f;
-
-		for (f = 0; f < skb_shinfo(skb)->nr_frags; f++)
-			count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);
-	} else {
-		count += skb_shinfo(skb)->nr_frags;
-	}
+	for (f = 0; f < skb_shinfo(skb)->nr_frags; f++)
+		count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);
 
 	if (igb_maybe_stop_tx(tx_ring, count + 3)) {
 		/* this is a hard error */
diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c
index ce5f7f9cff06..ecce8261ce3b 100644
--- a/drivers/net/ethernet/marvell/mvneta.c
+++ b/drivers/net/ethernet/marvell/mvneta.c
@@ -1359,7 +1359,7 @@ static void *mvneta_frag_alloc(const struct mvneta_port *pp)
 static void mvneta_frag_free(const struct mvneta_port *pp, void *data)
 {
 	if (likely(pp->frag_size <= PAGE_SIZE))
-		put_page(virt_to_head_page(data));
+		skb_free_frag(data);
 	else
 		kfree(data);
 }
diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c
index 43efc3a0cda5..0a28c07361cf 100644
--- a/drivers/net/ethernet/ti/netcp_core.c
+++ b/drivers/net/ethernet/ti/netcp_core.c
@@ -537,7 +537,7 @@ int netcp_unregister_rxhook(struct netcp_intf *netcp_priv, int order,
 static void netcp_frag_free(bool is_frag, void *ptr)
 {
 	if (is_frag)
-		put_page(virt_to_head_page(ptr));
+		skb_free_frag(ptr);
 	else
 		kfree(ptr);
 }