Hi Qi,
Will modify in next version.
-----Original Message-----
From: Zhang, Qi Z
Sent: Friday, September 20, 2019 8:55 AM
To: Pei, Andy <andy....@intel.com>; dev@dpdk.org
Cc: Xu, Rosen <rosen...@intel.com>; Zhang, Tianfei <tianfei.zh...@intel.com>;
Ye, Xiaolong <xiaolong...@intel.com>; Lomartire, David
<david.lomart...@intel.com>; Yigit, Ferruh <ferruh.yi...@intel.com>
Subject: RE: [PATCH v6 01/17] net/i40e: i40e support ipn3ke FPGA port bonding
> -----Original Message-----
> From: Pei, Andy
> Sent: Thursday, September 19, 2019 5:03 PM
> To: dev@dpdk.org
> Cc: Xu, Rosen <rosen...@intel.com>; Zhang, Tianfei
> <tianfei.zh...@intel.com>; Ye, Xiaolong <xiaolong...@intel.com>;
> Zhang, Qi Z <qi.z.zh...@intel.com>; Lomartire, David
> <david.lomart...@intel.com>; Yigit, Ferruh <ferruh.yi...@intel.com>
> Subject: [PATCH v6 01/17] net/i40e: i40e support ipn3ke FPGA port
> bonding
>
> In ipn3ke, each FPGA network side port bonding to an i40e pf, each
> i40e pf link status should get data from FPGA network, side port. This
> patch provide bonding relationship.
>
> Signed-off-by: Rosen Xu <rosen...@intel.com>
> Signed-off-by: Andy Pei <andy....@intel.com>
> ---
> drivers/net/i40e/base/i40e_type.h | 3 +++
> drivers/net/i40e/i40e_ethdev.c | 20 ++++++++++++++++++++
> drivers/net/i40e/rte_pmd_i40e.h | 4 ++++
> 3 files changed, 27 insertions(+)
>
> diff --git a/drivers/net/i40e/base/i40e_type.h
> b/drivers/net/i40e/base/i40e_type.h
> index 112866b..06863d7 100644
> --- a/drivers/net/i40e/base/i40e_type.h
> +++ b/drivers/net/i40e/base/i40e_type.h
> @@ -660,6 +660,9 @@ struct i40e_hw {
> struct i40e_nvm_info nvm;
> struct i40e_fc_info fc;
>
> + /* switch device is used to get link status when i40e is in ipn3ke */
> + struct rte_eth_dev *switch_dev;
> +
> /* pci info */
> u16 device_id;
> u16 vendor_id;
> diff --git a/drivers/net/i40e/i40e_ethdev.c
> b/drivers/net/i40e/i40e_ethdev.c index 4e40b7a..6c4b1e8 100644
> --- a/drivers/net/i40e/i40e_ethdev.c
> +++ b/drivers/net/i40e/i40e_ethdev.c
> @@ -1312,6 +1312,9 @@ static inline void i40e_config_automask(struct
> i40e_pf *pf)
> hw->adapter_stopped = 0;
> hw->adapter_closed = 0;
>
> + /* Init switch device pointer */
> + hw->switch_dev = NULL;
> +
> /*
> * Switch Tag value should not be identical to either the First Tag
> * or Second Tag values. So set something other than common
> Ethertype @@ -2782,6 +2785,20 @@ void
> i40e_flex_payload_reg_set_default(struct
> i40e_hw *hw)
> }
> }
>
> +void
> +i40e_set_switch_dev(struct rte_eth_dev *i40e_dev, struct rte_eth_dev
> +*switch_dev) {
> + struct i40e_hw *hw;
> +
> + if (!i40e_dev)
> + return;
> +
> + hw = I40E_DEV_PRIVATE_TO_HW(i40e_dev->data->dev_private);
> +
> + hw->switch_dev = switch_dev;
> +}
Better to move the function implementation into rte_pmd_i40e.c since it is
declared at rte_pmd_i40e.h
> +
> int
> i40e_dev_link_update(struct rte_eth_dev *dev,
> int wait_to_complete)
> @@ -2803,6 +2820,9 @@ void i40e_flex_payload_reg_set_default(struct
> i40e_hw *hw)
> else
> update_link_aq(hw, &link, enable_lse, wait_to_complete);
>
> + if (hw->switch_dev)
> + rte_eth_linkstatus_get(hw->switch_dev, &link);
> +
> ret = rte_eth_linkstatus_set(dev, &link);
> i40e_notify_all_vfs_link_status(dev);
>
> diff --git a/drivers/net/i40e/rte_pmd_i40e.h
> b/drivers/net/i40e/rte_pmd_i40e.h index faac9e2..9d77c85 100644
> --- a/drivers/net/i40e/rte_pmd_i40e.h
> +++ b/drivers/net/i40e/rte_pmd_i40e.h
> @@ -1061,4 +1061,8 @@ int rte_pmd_i40e_inset_set(uint16_t port,
> uint8_t pctype,
> return 0;
> }
>
> +void
> +i40e_set_switch_dev(struct rte_eth_dev *i40e_dev, struct rte_eth_dev
> +*switch_dev);
1. Missing doxygen header here for the new API.
2. Also as an external API, we should use port_id but not rte_eth_dev as
parameter.
3. you may also need to update the rte_pmd_i40e_version.map.
Regards
Qi
> +
> #endif /* _PMD_I40E_H_ */
> --
> 1.8.3.1
