[dpdk-stable] [PATCH] net/ena: TX L4 offloads should not be set in RX path

Luca Boccassi bluca at debian.org
Fri Jan 26 14:39:09 CET 2018


On Thu, 2018-01-25 at 16:12 +0100, Rafal Kozik wrote:
> Information about received packet type detected by NIC should be
> stored in packet_type field of rte_mbuf. TX L4 offload flags should
> not be set in RX path. Only fields that could be set in of_flags
> during packet receiving are information if L4 and L3 checksum is
> correct.
> 
> Fixes: 1173fca25af9 ("ena: add polling-mode driver")
> Cc: stable at dpdk.org
> 
> Reported-by: Matthew Smith <mgsmith at netgate.com>
> Signed-off-by: Rafal Kozik <rk at semihalf.com>
> ---
>  drivers/net/ena/ena_ethdev.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/ena/ena_ethdev.c
> b/drivers/net/ena/ena_ethdev.c
> index 83e0ae2..1e2af80 100644
> --- a/drivers/net/ena/ena_ethdev.c
> +++ b/drivers/net/ena/ena_ethdev.c
> @@ -275,16 +275,17 @@ static inline void ena_rx_mbuf_prepare(struct
> rte_mbuf *mbuf,
>  				       struct ena_com_rx_ctx
> *ena_rx_ctx)
>  {
>  	uint64_t ol_flags = 0;
> +	uint32_t packet_type = 0;
>  
>  	if (ena_rx_ctx->l4_proto == ENA_ETH_IO_L4_PROTO_TCP)
> -		ol_flags |= PKT_TX_TCP_CKSUM;
> +		packet_type |= RTE_PTYPE_L4_TCP;
>  	else if (ena_rx_ctx->l4_proto == ENA_ETH_IO_L4_PROTO_UDP)
> -		ol_flags |= PKT_TX_UDP_CKSUM;
> +		packet_type |= RTE_PTYPE_L4_UDP;
>  
>  	if (ena_rx_ctx->l3_proto == ENA_ETH_IO_L3_PROTO_IPV4)
> -		ol_flags |= PKT_TX_IPV4;
> +		packet_type |= RTE_PTYPE_L3_IPV4;
>  	else if (ena_rx_ctx->l3_proto == ENA_ETH_IO_L3_PROTO_IPV6)
> -		ol_flags |= PKT_TX_IPV6;
> +		packet_type |= RTE_PTYPE_L3_IPV6;
>  
>  	if (unlikely(ena_rx_ctx->l4_csum_err))
>  		ol_flags |= PKT_RX_L4_CKSUM_BAD;
> @@ -292,6 +293,7 @@ static inline void ena_rx_mbuf_prepare(struct
> rte_mbuf *mbuf,
>  		ol_flags |= PKT_RX_IP_CKSUM_BAD;
>  
>  	mbuf->ol_flags = ol_flags;
> +	mbuf->packet_type = packet_type;
>  }
>  
>  static inline void ena_tx_mbuf_prepare(struct rte_mbuf *mbuf,

Hi,

has a similar/related fix been applied to dpdk/master?

-- 
Kind regards,
Luca Boccassi


More information about the stable mailing list