devtools: add cocci script to rename mbuf offload flags
[dpdk.git] / doc / guides / nics / bnxt.rst
index e75f4fa..b3d10f3 100644 (file)
@@ -77,7 +77,7 @@ It can direct some traffic, for example data plane traffic, to DPDK.
 Rest of the traffic, for example control plane traffic, would be redirected to
 the traditional Linux networking stack.
 
-Refer to https://doc.dpdk.org/guides/howto/flow_bifurcation.html
+Refer to :doc:`../howto/flow_bifurcation`
 
 Benefits of the flow bifurcation include:
 
@@ -877,22 +877,21 @@ processing. This improved performance is derived from a number of optimizations:
     * TX: only the following reduced set of transmit offloads is supported in
       vector mode::
 
-       DEV_TX_OFFLOAD_MBUF_FAST_FREE
+       RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE
 
     * RX: only the following reduced set of receive offloads is supported in
       vector mode (note that jumbo MTU is allowed only when the MTU setting
-      does not require `DEV_RX_OFFLOAD_SCATTER` to be enabled)::
-
-       DEV_RX_OFFLOAD_VLAN_STRIP
-       DEV_RX_OFFLOAD_KEEP_CRC
-       DEV_RX_OFFLOAD_JUMBO_FRAME
-       DEV_RX_OFFLOAD_IPV4_CKSUM
-       DEV_RX_OFFLOAD_UDP_CKSUM
-       DEV_RX_OFFLOAD_TCP_CKSUM
-       DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM
-       DEV_RX_OFFLOAD_OUTER_UDP_CKSUM
-       DEV_RX_OFFLOAD_RSS_HASH
-       DEV_RX_OFFLOAD_VLAN_FILTER
+      does not require `RTE_ETH_RX_OFFLOAD_SCATTER` to be enabled)::
+
+       RTE_ETH_RX_OFFLOAD_VLAN_STRIP
+       RTE_ETH_RX_OFFLOAD_KEEP_CRC
+       RTE_ETH_RX_OFFLOAD_IPV4_CKSUM
+       RTE_ETH_RX_OFFLOAD_UDP_CKSUM
+       RTE_ETH_RX_OFFLOAD_TCP_CKSUM
+       RTE_ETH_RX_OFFLOAD_OUTER_IPV4_CKSUM
+       RTE_ETH_RX_OFFLOAD_OUTER_UDP_CKSUM
+       RTE_ETH_RX_OFFLOAD_RSS_HASH
+       RTE_ETH_RX_OFFLOAD_VLAN_FILTER
 
 The BNXT Vector PMD is enabled in DPDK builds by default. The decision to enable
 vector processing is made at run-time when the port is started; if no transmit