The wrong offset was being used to specify the mask and value for the IPv6
traffic class field for fdir. Also, since the 1 byte field is not byte
aligned in the header, do the shifting in big endian and then convert the
32-bit vtc_flow to little endian as expected by the adapter.
Fixes: dfbd6a9cb504 ("net/enic: extend flow director support for 1300 series")
Cc: sta...@dpdk.org
Signed-off-by: John Daley <johnd...@cisco.com>
---
drivers/net/enic/enic_clsf.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/net/enic/enic_clsf.c b/drivers/net/enic/enic_clsf.c
index c2218fe..2665b6e 100644
--- a/drivers/net/enic/enic_clsf.c
+++ b/drivers/net/enic/enic_clsf.c
@@ -315,8 +315,10 @@ copy_fltr_v2(struct filter_v2 *fltr, struct
rte_eth_fdir_input *input,
input->flow.ipv6_flow.dst_ip[i];
}
if (input->flow.ipv6_flow.tc) {
- ipv6_mask.vtc_flow = masks->ipv6_mask.tc << 16);
- ipv6_val.vtc_flow = input->flow.ipv6_flow.tc << 16;
+ ipv6_mask.vtc_flow = rte_be_to_cpu_32((uint32_t)
+ masks->ipv6_mask.tc << 20);
+ ipv6_val.vtc_flow = rte_be_to_cpu_32((uint32_t)
+ input->flow.ipv6_flow.tc << 20);
}
if (input->flow.ipv6_flow.hop_limits) {
ipv6_mask.hop_limits = masks->ipv6_mask.hop_limits;
--
2.10.0
