eal: refactor --huge-unlink storage

[dpdk.git] / lib / net / rte_net.h

diff --git a/lib/net/rte_net.h b/lib/net/rte_net.h
index 434435f..53a7f4d 100644 (file)
--- a/lib/net/rte_net.h
+++ b/lib/net/rte_net.h
@@ -34,9 +34,6 @@ struct rte_net_hdr_lens {
  * This function skips all IPv6 extensions, returning size of
  * complete header including options and final protocol value.
  *
- * @warning
- * @b EXPERIMENTAL: this API may change without prior notice
- *
  * @param proto
  *   Protocol field of IPv6 header.
  * @param m
@@ -51,7 +48,6 @@ struct rte_net_hdr_lens {
  *   Protocol that follows IPv6 header.
  *   -1 if an error occurs during mbuf parsing.
  */
-__rte_experimental
 int
 rte_net_skip_ip6_ext(uint16_t proto, const struct rte_mbuf *m, uint32_t *off,
        int *frag);
@@ -125,11 +121,22 @@ rte_net_intel_cksum_flags_prepare(struct rte_mbuf *m, uint64_t ol_flags)
         * Mainly it is required to avoid fragmented headers check if
         * no offloads are requested.
         */
-       if (!(ol_flags & (PKT_TX_IP_CKSUM | PKT_TX_L4_MASK | PKT_TX_TCP_SEG)))
+       if (!(ol_flags & (RTE_MBUF_F_TX_IP_CKSUM | RTE_MBUF_F_TX_L4_MASK | RTE_MBUF_F_TX_TCP_SEG |
+                         RTE_MBUF_F_TX_OUTER_IP_CKSUM)))
                return 0;
 
-       if (ol_flags & (PKT_TX_OUTER_IPV4 | PKT_TX_OUTER_IPV6))
+       if (ol_flags & (RTE_MBUF_F_TX_OUTER_IPV4 | RTE_MBUF_F_TX_OUTER_IPV6)) {
                inner_l3_offset += m->outer_l2_len + m->outer_l3_len;
+               /*
+                * prepare outer IPv4 header checksum by setting it to 0,
+                * in order to be computed by hardware NICs.
+                */
+               if (ol_flags & RTE_MBUF_F_TX_OUTER_IP_CKSUM) {
+                       ipv4_hdr = rte_pktmbuf_mtod_offset(m,
+                                       struct rte_ipv4_hdr *, m->outer_l2_len);
+                       ipv4_hdr->hdr_checksum = 0;
+               }
+       }
 
        /*
         * Check if headers are fragmented.
@@ -140,16 +147,16 @@ rte_net_intel_cksum_flags_prepare(struct rte_mbuf *m, uint64_t ol_flags)
                     inner_l3_offset + m->l3_len + m->l4_len))
                return -ENOTSUP;
 
-       if (ol_flags & PKT_TX_IPV4) {
+       if (ol_flags & RTE_MBUF_F_TX_IPV4) {
                ipv4_hdr = rte_pktmbuf_mtod_offset(m, struct rte_ipv4_hdr *,
                                inner_l3_offset);
 
-               if (ol_flags & PKT_TX_IP_CKSUM)
+               if (ol_flags & RTE_MBUF_F_TX_IP_CKSUM)
                        ipv4_hdr->hdr_checksum = 0;
        }
 
-       if ((ol_flags & PKT_TX_L4_MASK) == PKT_TX_UDP_CKSUM) {
-               if (ol_flags & PKT_TX_IPV4) {
+       if ((ol_flags & RTE_MBUF_F_TX_L4_MASK) == RTE_MBUF_F_TX_UDP_CKSUM) {
+               if (ol_flags & RTE_MBUF_F_TX_IPV4) {
                        udp_hdr = (struct rte_udp_hdr *)((char *)ipv4_hdr +
                                        m->l3_len);
                        udp_hdr->dgram_cksum = rte_ipv4_phdr_cksum(ipv4_hdr,
@@ -164,9 +171,9 @@ rte_net_intel_cksum_flags_prepare(struct rte_mbuf *m, uint64_t ol_flags)
                        udp_hdr->dgram_cksum = rte_ipv6_phdr_cksum(ipv6_hdr,
                                        ol_flags);
                }
-       } else if ((ol_flags & PKT_TX_L4_MASK) == PKT_TX_TCP_CKSUM ||
-                       (ol_flags & PKT_TX_TCP_SEG)) {
-               if (ol_flags & PKT_TX_IPV4) {
+       } else if ((ol_flags & RTE_MBUF_F_TX_L4_MASK) == RTE_MBUF_F_TX_TCP_CKSUM ||
+                       (ol_flags & RTE_MBUF_F_TX_TCP_SEG)) {
+               if (ol_flags & RTE_MBUF_F_TX_IPV4) {
                        /* non-TSO tcp or TSO */
                        tcp_hdr = (struct rte_tcp_hdr *)((char *)ipv4_hdr +
                                        m->l3_len);