Hi Tom, On Mon, 3 Apr 2023 at 02:42, Tom Rini <tr...@konsulko.com> wrote: > > On Sun, Apr 02, 2023 at 02:41:46PM +1200, Simon Glass wrote: > > Hi Mark, > > > > On Sun, 2 Apr 2023 at 10:31, Mark Kettenis <mark.kette...@xs4all.nl> wrote: > > > > > > > From: Simon Glass <s...@chromium.org> > > > > Date: Sun, 2 Apr 2023 09:45:33 +1200 > > > > > > > > Since CONFIG_BOOTSTD_DEFAULTS now selects ENV_VARS_UBOOT_CONFIG we > > cannot > > > > enable it for some Xilinx boards which have a very small environment. > > > > > > > > Disable it. > > > > > > Does that mean these configs lose distroboot support? That would be a > > > shame! > > > > We use 'select' to enable things, so we don't have the ability to enable > > just some of the defaults. So unfortunately this board needs to have > > BOOTSTD_DEFAULTS disabled. > > > > It certainly doesn't have anywhere near enough env space for distroboot, > > although I'm not sure why it is so limited. > > Distroboot seems to work historically on this platform, so you need to > ensure it still works afterwards. These platforms just don't enable the > "enable all the options" option because they're intentionally limited > (the "mini" in the name).
It looks like they set the bootcommand to 'run distro_bootcmd' rather than enabling DISTRO_DEFAULTS. But I just don't understand how it can work. The env size is 0x80 bytes and it doesn't include the config_distro_bootcmd.h file. Mark, maybe I just totally lost, but I just cannot see how this can work. Please take a look. Tom, do you have a board to test with? Anyway, as should be clear from my commit message, this is about BOOTSTD_DEFAULTS so it does not affect DISTRO_DEFAULTS. Regards, Simon