From: Maciej Gajdzica <maciejx.t.gajdz...@intel.com> p->mtu field should be used instead.
Signed-off-by: Maciej Gajdzica <maciejx.t.gajdzica at intel.com> --- lib/librte_port/rte_port_frag.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/lib/librte_port/rte_port_frag.c b/lib/librte_port/rte_port_frag.c index ff0ab9b..dce33d5 100644 --- a/lib/librte_port/rte_port_frag.c +++ b/lib/librte_port/rte_port_frag.c @@ -38,10 +38,6 @@ #include "rte_port_frag.h" -/* Default byte size for the IPv4 Maximum Transfer Unit (MTU). - * This value includes the size of IPv4 header. */ -#define IPV4_MTU_DEFAULT ETHER_MTU - /* Max number of fragments per packet allowed */ #define IPV4_MAX_FRAGS_PER_PACKET 0x80 @@ -160,7 +156,7 @@ rte_port_ring_reader_ipv4_frag_rx(void *port, p->n_pkts--; /* If not jumbo, pass current packet to output */ - if (pkt->pkt_len <= IPV4_MTU_DEFAULT) { + if (pkt->pkt_len <= p->mtu) { pkts[n_pkts_out++] = pkt; n_pkts_to_provide = n_pkts - n_pkts_out; -- 1.7.9.5