udev: net_id: Use devicetree aliases when available

[sean-anderson-seco]

Devicetree firmware contains an "aliases" node, containing various
aliases for devices described by the firmware. For ethernet devices,
these are named "ethernet0", "ethernet1", etc. They provide a convenient
means of numbering ethernet devices, especially on systems with no other
stable number other than the address. In particular, U-Boot already uses
these aliases to name its ethernet devices.

Previously, there have been attempts (such as [1]) to add support for
these aliases to Linux. However, these patches have been rejected
because it is the maintainers' view that naming policy be left to
userspace. Well, systemd is userspace, so here we are.

In terms of implementation, apparently there can be multiple device
trees at once. I have decided to dodge this problem for now, and just
use /proc/device-tree. If it is desired to support multiple device trees
later, then the scheme can be modified to something like enb2d3.

For the moment we only support "ethernetX" aliases. Future patches might
want to also handle "canX" and "wifiX".

It is common on boards with only one ethernet device to use an alias of
just "ethernet". In this case, the index is an implicit 0. In case the
author of the firmware made a mistake, we check to ensure that aliases
of "ethernet" and "ethernet0" do not both exist.

[1] https://patchwork.kernel.org/project/linux-arm-kernel/patch/1399390594-1409-1-git-send-email-boris.brezillon@free-electrons.com/

Signed-off-by: Sean Anderson sean.anderson@seco.com

Closes #17625.

[yuwata]

BTW, is this for RFE #17625? If so, please add Closes #17625. in the commit message.
Also, please drop signed-off-by tag in the commit message. We do not use it.

[yuwata]

In terms of implementation, apparently there can be multiple device
trees at once. I have decided to dodge this problem for now, and just
use /proc/device-tree. If it is desired to support multiple device trees
later, then the scheme can be modified to something like enb2d3.

What is enb2d3??

[sean-anderson-seco]

Thank you for the quick review.

BTW, is this for RFE #17625? If so, please add Closes #17625. in the commit message.

I did not write this specifically for that issue, but it looks like this might close that (depending on whether there is an appropriate alias). I will add that tag.

Also, please drop signed-off-by tag in the commit message. We do not use it.

OK (as it happened I quickly went through the history of this file and saw it used in d6eda67).

In terms of implementation, apparently there can be multiple device
trees at once. I have decided to dodge this problem for now, and just
use /proc/device-tree. If it is desired to support multiple device trees
later, then the scheme can be modified to something like enb2d3.

What is enb2d3??

According to the sysfs ABI documentation, there can be multiple devicetrees, but it does not document the naming scheme used. The actual attachment is done by __of_attach_node_sysfs, which uses safe_name to create the new top-level tree. I believe this results in something like base, base1, ..., base15. So the idea is if we needed to support this we could use b for "base".

[yuwata]

Looks much better now. One more round please.

[yuwata]

Mostly LGTM. Only several coding style comments.

[yuwata]

LGTM. Thank you.