On 2024-12-17, Vagrant Cascadian wrote:
> On 2024-12-17, Vagrant Cascadian wrote:
>> I just did some basic boot testing of u-boot-pinebook and
>> u-boot-rock64-rk3328, by manually installing and booting Debian with it,
>> though I have not tested with Guix System (and not sure I will be able
>> to).
>>
>> I might be able to test u-boot-sifive-unmatched next. Having troubles
>> unrelated to these patches with the rockpro64 board, but hope to get
>> those fixed.
>
> Boot tested u-boot-rockpro64-rk3399 booting Debian. So that is three
> targets not fundamentally broken, at least. :)

And also boot tested u-boot-sifive-unmatched, which had a glitch in how
it changed the boot device order and booted an old kernel from a device
that the previous defaults did not boot... so that is a bit
odd... worked around it by renaming the extlinux.conf file so that it
would not select it.

Other platforms that switched from u-boot's distro boot to u-boot's
bootstd might have similar issues, so it is something to note possibly
in a news entry.

This might be worth setting BOOTSTD_FULL=y in the configuration; without
it there is no way to select the boot device if the first one it picks
is not the one you actually want... and possibly trying to get that
upstream too.


live well,
  vagrant