This adds a simple boot blessing module, to be used, with the Zyxel NWA50AX.
There's a lot of elephant in the rooms: how do you upgrade kernel, etc.
Signed-off-by: Raito Bezarius <masterancpp@gmail.com>
This is a quite comprehensive example using maximally the hardware
available to reach nice performance.
In the future, I will even add RADIUS examples.
Signed-off-by: Raito Bezarius <masterancpp@gmail.com>
Zyxel "firmware" format is just… a FIT with some metadata on the models.
This FIT is like this:
--------------------------
uImage FIT header
--------------------------
Linux kernel
--------------------------
FDT DTB
--------------------------
Padding so that
this makes
8192kb [1]
--------------------------
UBI volume
as a root filesystem
--------------------------
We just reproduce this in a very brutal and naive way.
In the future, this seems worth to generalize and modularize this idea
so that zyxel-nwa-fit is just an instance of a more general output.
[1]: https://git.openwrt.org/?p=openwrt/openwrt.git;a=blob;f=target/linux/ramips/image/mt7621.mk;h=ab1b829ba0086cb9fc9ca8cbbf3cbc14735034d6;hb=refs/heads/main#l3097
Signed-off-by: Raito Bezarius <masterancpp@gmail.com>
I believe there should be another module exposing `rootubifs` as `rootfs`
or let any other module just subsume that component like `zyxel-nwa-fit` output.
Signed-off-by: Raito Bezarius <masterancpp@gmail.com>
Some devices have a U-boot variant that does not accept a third
parameter on the "bootm" command, meaning we can't override the dtb
in the bootloader so have to smush it back into the kernel image
This doesn't work in QEMU but I think the problem is with the
U-Boot configuration for QEMU. It does work on at least one
hardware device so I'm pushing it anyway
Based on
https://gti.telent.net/raboof/liminix/src/branch/tftp-old-uboot
Co-authored-by: Arnout Engelen <arnout@bzzt.net>
On this device we don't need the openwrt kernel or patches. The
newer kernel also fixes the weird one minute pause at boot when
it was doing something with either mmc or switch.
This changes the practice for building kernel modules: now we expect
that the appropriate Kconfig symbols are set to =m in
config.kernel.config, and then use pkgs.kmodloader to create
a service that loads and unloads all the modules depended on by
a particular requirement.
Note that modules won't be installed on the target device just by
virue of having been built: only the modules that are referenced by a
kmodloader package will be in the closure.
An example may make this clearer: see modules/firewall/default.nix
in this commit.
Why?
If you have a compiled Linux kernel source tree and you change some
symbol from "is not set" to m and then run make modules, you cannot in
general expect that newly compiled module to work. This is because
there are places in the build of the main kernel where it looks to see
which modules _may_ be defined and uses that information to
accommodate them.
For example in an in-kernel build of
https://github.com/torvalds/linux/blob/master/net/netfilter/core.c#L689
some symbols are defined only if CONFIG_NF_CONNTRACK is set, meaning
this code won't work if we have it unset initially then try later to
enable it and build modules only. Or see
https://github.com/torvalds/linux/blob/master/include/linux/netdevice.h#L160
this is step one towards getting rid of rotuer-secrets completely and
turning rotuer into a "profile" module that can be less hackily
customised for other people's networks
