Hello Ferruh, On Fri, Mar 30, 2018 at 5:29 PM, David Marchand <david.march...@6wind.com> wrote: > On Fri, Mar 30, 2018 at 5:17 PM, Ferruh Yigit <ferruh.yi...@intel.com> wrote: >> diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c >> index 209796d46..68bdc3103 100644 >> --- a/lib/librte_ether/rte_ethdev.c >> +++ b/lib/librte_ether/rte_ethdev.c >> @@ -2421,6 +2421,7 @@ rte_eth_dev_info_get(uint16_t port_id, struct >> rte_eth_dev_info *dev_info) >> memset(dev_info, 0, sizeof(struct rte_eth_dev_info)); >> dev_info->rx_desc_lim = lim; >> dev_info->tx_desc_lim = lim; >> + dev_info->device = dev->device; >> >> RTE_FUNC_PTR_OR_RET(*dev->dev_ops->dev_infos_get); >> (*dev->dev_ops->dev_infos_get)(dev, dev_info); > > Just a little comment, do we want the pmd to be able to override this ?
I still don't see when the pmd would need to override the dev_info->device. The pmd is the one that populated dev->device in the first place. Why would it want to report something different in dev_infos ? Some comments from pmd maintainers, maybe ? -- David Marchand
