Hi Selvin,
> From: linux-rdma-owner@xxxxxxxxxxxxxxx <linux-rdma-
> owner@xxxxxxxxxxxxxxx> On Behalf Of Selvin Xavier
>
[..]
> +int bnxt_re_ib_init(struct bnxt_re_dev *rdev) {
> + int rc = 0;
> +
> + /* Register ib dev */
> + rc = bnxt_re_register_ib(rdev);
> + if (rc) {
> + pr_err("Failed to register with IB: %#x\n", rc);
> + return rc;
> + }
> + set_bit(BNXT_RE_FLAG_IBDEV_REGISTERED, &rdev->flags);
> + dev_info(rdev_to_dev(rdev), "Device registered successfully");
> + ib_get_eth_speed(&rdev->ibdev, 1, &rdev->active_speed,
> + &rdev->active_width);
> + set_bit(BNXT_RE_FLAG_ISSUE_ROCE_STATS, &rdev->flags);
> + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1,
> IB_EVENT_PORT_ACTIVE);
What if the link is down at this point?
I see that it was done this way before, but since you are refactoring, may be you want to relook?
Do you still want to report it as active?
> + bnxt_re_dispatch_event(&rdev->ibdev, NULL, 1,
> IB_EVENT_GID_CHANGE);
> +
GID addition, deletion decisions for RoCE ports are taken care by the IB core.
So hw driver shouldn't report this event. Please remove this call.
