Hi Ferruh, > -----Original Message----- > From: Yigit, Ferruh > Sent: Thursday, August 17, 2017 9:05 AM > To: Lu, Wenzhuo <wenzhuo...@intel.com>; dev@dpdk.org > Cc: Wu, Jingjing <jingjing...@intel.com>; Xing, Beilei > <beilei.x...@intel.com>; > Stephen Hurd <stephen.h...@broadcom.com>; Ajit Khaparde > <ajit.khapa...@broadcom.com> > Subject: Re: [dpdk-dev] [PATCH v2] net/i40e: new API to add VF MAC address > from PF > > On 7/25/2017 3:09 PM, Wenzhuo Lu wrote: > > Currently, on i40e the parameter 'pool' of API > > rte_eth_dev_mac_addr_add means the VMDq pool, not VF. > > The argument "pool" in rte_eth_dev_mac_addr_add() IS VMDq pool. > This is not just for i40e, this is by API definition. > > > So, it's wrong to use it to set the VF MAC address. > > Agreed, it seems testpmd function cmd_vf_mac_addr_parsed() implemented > wrong. > > > As this API is also used by the VMDq example, ideally we need a > > parameter to tell the pool is VMDq or VF. > > But it's hard to change it because of the ABI change concern. > > I think we shouldn't NOT update rte_eth_dev_mac_addr_add() API, it is not > wrong. Agree :)
> > But should fix testpmd ""mac_addr add port <port_id> vf <vf_id> > <mac_addr>" command. > > Can you please update this patch as two patches: > > 1- i40e rte_pmd_i40e_add_vf_mac_addr() API > > 2- Fix testpmd function, instead of inserting new i40e API replace it with > wrong rte_eth_dev_mac_addr_add() call. > And I guess bnxt driver also has API to add VF MAC, can you add that one too? Thanks for the suggestion. Will send a new version. > > > Now the solution is to provide a new API, users can call it to add VF > > MAC address from PF on i40e. > > > > Signed-off-by: Wenzhuo Lu <wenzhuo...@intel.com> > > <...>
