Re: [dpdk-dev] [PATCH 1/6] lib/librte_ether: modify the structures for fdir new modes

[Thomas Monjalon <thomas.monjalon@6wind.com>]

2015-09-25 14:05, Wenzhuo Lu:
> +enum rte_fdir_mode {
> +	RTE_FDIR_MODE_NONE  = 0, /**< Disable FDIR support. */
> +	RTE_FDIR_MODE_SIGNATURE, /**< Enable FDIR signature filter mode. */
> +	RTE_FDIR_MODE_PERFECT,   /**< Enable FDIR perfect filter mode for IP. */
> +	RTE_FDIR_MODE_PERFECT_MAC_VLAN, /**< Enable FDIR filter mode - MAC VLAN. */
> +	RTE_FDIR_MODE_PERFECT_CLOUD,    /**< Enable FDIR filter mode - cloud. */
> +};

I know that some Intel NICs use the terminology "cloud" in their datasheet,
but it is meaningless.

[...]
> +/**
> + * A structure used to define the input for VxLAN NVGRE flow
> + */
> +struct rte_eth_cloud_flow {
> +	enum rte_eth_fdir_tunnel_type tunnel_type; /**< Tunnel type to match. */
> +	uint32_t tni_vni;                          /**< TNI or VNI to match. */

Isn't it Intel-specific?

> +	struct ether_addr mac_addr;                /**< Mac address to match. */
> +};

So it is a tunnel, currently only VXLAN or NVGRE.
And this kind of tunnel can be used in a cloud datacenter, yes. Or elsewhere.

Please use the "tunnel" word.