On Thu, May 28, 2020 at 04:33:35PM +0200, Thomas Bogendoerfer wrote:
> below is the dts part for the two network interfaces. The switch to
> the outside has two ports, which correlate to the two internal ports.
> And the switch propagates the link state of the external ports to
> the internal ports.
Okay, so this DTS hasn't been reviewed...
> &cp0_eth1 {
> status = "okay";
> phy-mode = "2500base-x";
> mac-address = [00 00 00 00 00 01];
> interrupts = <41 IRQ_TYPE_LEVEL_HIGH>,
> <45 IRQ_TYPE_LEVEL_HIGH>,
> <49 IRQ_TYPE_LEVEL_HIGH>,
> <53 IRQ_TYPE_LEVEL_HIGH>,
> <57 IRQ_TYPE_LEVEL_HIGH>,
> <61 IRQ_TYPE_LEVEL_HIGH>,
> <65 IRQ_TYPE_LEVEL_HIGH>,
> <69 IRQ_TYPE_LEVEL_HIGH>,
> <73 IRQ_TYPE_LEVEL_HIGH>,
> <127 IRQ_TYPE_LEVEL_HIGH>;
> interrupt-names = "hif0", "hif1", "hif2",
> "hif3", "hif4", "hif5", "hif6", "hif7",
> "hif8", "link";
> port-id = <2>;
> gop-port-id = <3>;
This seems to correlate with the eth2 node in armada-cp11x.dtsi -
you do not need to specify the interrupts, interrupt-names, port-id
and gop-port-id unless you need to change them just because you want
to add a few properties to this node - you can just list the new or
altered properties. To delete a property, you need to prefix the
property name with /delete-property/.
> managed = "in-band-status";
This isn't correct - you are requesting that in-band status is used
(i.o.w. the in-band control word, see commit 4cba5c210365), but your
bug report wants to enable AN bypass because there is no in-band
control word. This seems to be rather contradictory.
May I suggest you use a fixed-link here, which will not have any
inband status, as there is no in-band control word being sent by
the switch? That is also the conventional way of handling switch
links.
Thanks.
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC for 0.8m (est. 1762m) line in suburbia: sync at 13.1Mbps down 424kbps up
