> -----Original Message-----
> From: dev <dev-boun...@dpdk.org> On Behalf Of murphy yang
> Sent: Friday, September 18, 2020 5:47 PM
> To: dev@dpdk.org
> Cc: Yang, MurphyX <murphyx.y...@intel.com>; sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH] net/iavf: fix link status
> 
> From: murphy <murphyx.y...@intel.com>
> 
> If the PF driver supports the new speed reporting capabilities then use
> link_event_adv instead of link_event to get the speed.
> 
> Fixes: 5330b042977c ("net/iavf: fix link speed")
> 
> Cc: sta...@dpdk.org
> 
> Signed-off-by: murphy <murphyx.y...@intel.com>
> ---
>  drivers/net/iavf/iavf_vchnl.c | 7 +++----
>  1 file changed, 3 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/iavf/iavf_vchnl.c b/drivers/net/iavf/iavf_vchnl.c 
> index
> 76f8e38d1..c4956dc58 100644
> --- a/drivers/net/iavf/iavf_vchnl.c
> +++ b/drivers/net/iavf/iavf_vchnl.c
> @@ -195,14 +195,13 @@ iavf_handle_pf_event_msg(struct rte_eth_dev *dev,
> uint8_t *msg,
>       case VIRTCHNL_EVENT_LINK_CHANGE:
>               PMD_DRV_LOG(DEBUG, "VIRTCHNL_EVENT_LINK_CHANGE event");
>               vf->link_up = pf_msg->event_data.link_event.link_status;
> +             enum virtchnl_link_speed speed;
>               if (vf->vf_res->vf_cap_flags & VIRTCHNL_VF_CAP_ADV_LINK_SPEED)
> {
> -                     vf->link_speed =
> -                             pf_msg->event_data.link_event_adv.link_speed;
> +                     speed = pf_msg->event_data.link_event_adv.link_speed;

According to vritual channel spec, link_event_adv.link_speed should store the 
link_speed provided in Mbps, it no need to be further converted.
If the value is wrong, it should be fixed in kernel driver.

>               } else {
> -                     enum virtchnl_link_speed speed;
>                       speed = pf_msg->event_data.link_event.link_speed;
> -                     vf->link_speed = iavf_convert_link_speed(speed);
>               }
> +             vf->link_speed = iavf_convert_link_speed(speed);
>               iavf_dev_link_update(dev, 0);
>               rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_LSC,
> NULL);
>               break;
> --
> 2.17.1

Reply via email to