Dear Dan?
? 2016?11?17? 19:04, Dan Carpenter ??:
> Hello William Wu,
>
> The patch 0c42fe48fd23: "phy: rockchip-inno-usb2: support otg-port
> for rk3399" from Nov 7, 2016, leads to the following static checker
> warning:
>
> drivers/phy/phy-rockchip-inno-usb2.c:998 rockchip_usb2phy_otg_port_init()
> warn: 'rphy->edev' isn't an ERR_PTR
>
> drivers/phy/phy-rockchip-inno-usb2.c
> 989 ret = devm_request_threaded_irq(rphy->dev, rport->bvalid_irq, NULL,
> 990 rockchip_usb2phy_bvalid_irq,
> 991 IRQF_ONESHOT,
> 992 "rockchip_usb2phy_bvalid", rport);
> 993 if (ret) {
> 994 dev_err(rphy->dev, "failed to request otg-bvalid irq handle\n");
> 995 goto out;
> 996 }
> 997
> 998 if (!IS_ERR(rphy->edev)) {
>
> I think the static checker is correct that this check should be removed.
> The rest of the code assumes ->edev is a valid pointer.
Yes, you are right. This static checker can be removed safely.
Thanks very much for your help.
Best regards?
wulf
>
> 999 rport->event_nb.notifier_call = rockchip_otg_event;
> 1000
> 1001 ret = extcon_register_notifier(rphy->edev, EXTCON_USB_HOST,
> 1002 &rport->event_nb);
> 1003 if (ret)
> 1004 dev_err(rphy->dev, "register USB HOST notifier failed\n");
> 1005 }
> 1006
> 1007 out:
> 1008 return ret;
> 1009 }
>
> regards,
> dan carpenter
>
>
>
