> On Oct 28, 2019, at 11:58, Hayes Wang <hayeswang@xxxxxxxxxxx> wrote: > > Kai-Heng Feng [mailto:kai.heng.feng@xxxxxxxxxxxxx] >> Sent: Friday, October 25, 2019 6:59 PM > [...] >> @@ -6626,6 +6648,9 @@ static int rtl8152_probe(struct usb_interface *intf, >> netdev->hw_features &= ~NETIF_F_RXCSUM; >> } >> >> + if (id->driver_info & R8152_QUIRK_LENOVO_MACPASSTHRU) > > Do you really need this? > It seems the information of idVendor and idProduct is enough. Both idVendor and idProduct are just part of "struct usb_device_id". IMO it's clearer to add quirks there. > >> + set_bit(LENOVO_MACPASSTHRU, &tp->flags); >> + >> if (le16_to_cpu(udev->descriptor.bcdDevice) == 0x3011 && udev->serial >> && >> (!strcmp(udev->serial, "000001000000") || >> !strcmp(udev->serial, "000002000000"))) { >> @@ -6754,6 +6779,8 @@ static const struct usb_device_id rtl8152_table[] = { >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x304f, 0)}, >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x3062, 0)}, >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x3069, 0)}, >> + {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x3082, >> + R8152_QUIRK_LENOVO_MACPASSTHRU)}, > > This limits the usage of driver_info. For example, I couldn't > use it to store a pointer address anymore. But will the driver ever use .driver_info for pointers? There are many driver use it for quirks and quirks only. Kai-Heng > >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x7205, 0)}, >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x720c, 0)}, >> {REALTEK_USB_DEVICE(VENDOR_ID_LENOVO, 0x7214, 0)}, >> -- >> 2.17.1 >
