On Mon, Oct 19, 2020 at 2:22 PM Jakub Kicinski <k...@kernel.org> wrote:
>
> Looks correct to me. I spotted there is also IFF_WAN_HDLC added by
> 7cdc15f5f9db ("WAN: Generic HDLC now uses IFF_WAN_HDLC private flag.")
> would using that flag also be correct and cleaner potentially?
>
> Up to you, just wanted to make sure you considered it.Oh, Yes! I see IFF_WAN_HDLC is set for all HDLC devices. I also searched through the kernel code and see no other uses of IFF_WAN_HDLC. So I think we can use IFF_WAN_HDLC to reliably check if a device is an HDLC device. This should be cleaner than checking ndo_start_xmit. Thanks! I'll change this patch to use IFF_WAN_HDLC.
