ibmveth: Copy tx skbs into a premapped buffer

Rather than DMA mapping and unmapping every outgoing skb, copy the skb
into a buffer that was mapped during the drivers open function. Copying
the skb and its frags have proven to be more time efficient than
mapping and unmapping. As an effect, performance increases by 3-5
Gbits/s.

Allocate and DMA map one continuous 64KB buffer at `ndo_open`. This
buffer is maintained until `ibmveth_close` is called. This buffer is
large enough to hold the largest possible linnear skb. During
`ndo_start_xmit`, copy the skb and all of it's frags into the continuous
buffer. By manually linnearizing all the socket buffers, time is saved
during memcpy as well as more efficient handling in FW.
As a result, we no longer need to worry about the firmware limitation
of handling a max of 6 frags. So, we only need to maintain 1 descriptor
instead of 6 and can hardcode 0 for the other 5 descriptors during
h_send_logical_lan.

Since, DMA allocation/mapping issues can no longer arise in xmit
functions, we can further reduce code size by removing the need for a
bounce buffer on DMA errors.

Signed-off-by: Nick Child <nnac123@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 12 insertions, 10 deletions

diff --git a/drivers/net/ethernet/ibm/ibmveth.h b/drivers/net/ethernet/ibm/ibmveth.h
index 27dfff200166..a46ead9b31de 100644
--- a/drivers/net/ethernet/ibm/ibmveth.h
+++ b/drivers/net/ethernet/ibm/ibmveth.h
@@ -46,23 +46,23 @@
 #define h_add_logical_lan_buffer(ua, buf) \
   plpar_hcall_norets(H_ADD_LOGICAL_LAN_BUFFER, ua, buf)
 
+/* FW allows us to send 6 descriptors but we only use one so mark
+ * the other 5 as unused (0)
+ */
 static inline long h_send_logical_lan(unsigned long unit_address,
-		unsigned long desc1, unsigned long desc2, unsigned long desc3,
-		unsigned long desc4, unsigned long desc5, unsigned long desc6,
-		unsigned long corellator_in, unsigned long *corellator_out,
-		unsigned long mss, unsigned long large_send_support)
+		unsigned long desc, unsigned long corellator_in,
+		unsigned long *corellator_out, unsigned long mss,
+		unsigned long large_send_support)
 {
 	long rc;
 	unsigned long retbuf[PLPAR_HCALL9_BUFSIZE];
 
 	if (large_send_support)
 		rc = plpar_hcall9(H_SEND_LOGICAL_LAN, retbuf, unit_address,
-				  desc1, desc2, desc3, desc4, desc5, desc6,
-				  corellator_in, mss);
+				  desc, 0, 0, 0, 0, 0, corellator_in, mss);
 	else
 		rc = plpar_hcall9(H_SEND_LOGICAL_LAN, retbuf, unit_address,
-				  desc1, desc2, desc3, desc4, desc5, desc6,
-				  corellator_in);
+				  desc, 0, 0, 0, 0, 0, corellator_in);
 
 	*corellator_out = retbuf[0];
 
@@ -98,6 +98,7 @@ static inline long h_illan_attributes(unsigned long unit_address,
 #define IBMVETH_BUFF_LIST_SIZE 4096
 #define IBMVETH_FILT_LIST_SIZE 4096
 #define IBMVETH_MAX_BUF_SIZE (1024 * 128)
+#define IBMVETH_MAX_TX_BUF_SIZE (1024 * 64)
 
 static int pool_size[] = { 512, 1024 * 2, 1024 * 16, 1024 * 32, 1024 * 64 };
 static int pool_count[] = { 256, 512, 256, 256, 256 };
@@ -137,6 +138,9 @@ struct ibmveth_adapter {
     unsigned int mcastFilterSize;
     void * buffer_list_addr;
     void * filter_list_addr;
+    void *tx_ltb_ptr;
+    unsigned int tx_ltb_size;
+    dma_addr_t tx_ltb_dma;
     dma_addr_t buffer_list_dma;
     dma_addr_t filter_list_dma;
     struct ibmveth_buff_pool rx_buff_pool[IBMVETH_NUM_BUFF_POOLS];
@@ -145,8 +149,6 @@ struct ibmveth_adapter {
     int rx_csum;
     int large_send;
     bool is_active_trunk;
-    void *bounce_buffer;
-    dma_addr_t bounce_buffer_dma;
 
     u64 fw_ipv6_csum_support;
     u64 fw_ipv4_csum_support;