On 5/12/2017 4:48 PM, Thomas Monjalon wrote: > 12/05/2017 12:49, Ferruh Yigit: >> On 5/10/2017 2:51 PM, Thomas Monjalon wrote: >>> 10/05/2017 15:10, Ferruh Yigit: >>>> --- a/doc/guides/nics/features/bnx2x.ini >>>> +++ b/doc/guides/nics/features/bnx2x.ini >>>> @@ -4,6 +4,7 @@ >>>> ; Refer to default.ini for the full list of available PMD features. >>>> ; >>>> [Features] >>>> +Speed capabilities = Y >>> >>> We should validate this feature only if the driver advertise the >>> right speeds for the device. >> >> Hi Thomas, >> >> Can you please clarify more, what is expected implementation in PMD? > > It is expected to advertise only the speeds that the device is > capable to offer. > >> And perhaps a good and a bad sample can be helpful. > > Good example: > drivers/net/i40e/i40e_ethdev.c > if (I40E_PHY_TYPE_SUPPORT_40G(hw->phy.phy_types)) > /* For XL710 */ > dev_info->speed_capa = ETH_LINK_SPEED_40G; > else if (I40E_PHY_TYPE_SUPPORT_25G(hw->phy.phy_types)) > /* For XXV710 */ > dev_info->speed_capa = ETH_LINK_SPEED_25G; > else > /* For X710 */ > dev_info->speed_capa = ETH_LINK_SPEED_1G | ETH_LINK_SPEED_10G; > > Bad example: > drivers/net/bnx2x/bnx2x_ethdev.c > dev_info->speed_capa = ETH_LINK_SPEED_10G | ETH_LINK_SPEED_20G; > Looking at qlogic.com, only some 57840 adapters are capable of 20G.
OK, I will update the patch according. Does it make sense to mark PMDs that report speed_capabilities, but without checking actual hw, as partial "P", to differentiate them from the ones that doesn't report at all, also this may help to PMD maintainers about what to fix. Thanks, ferruh
