[U-Boot] Issues when saving environment in RK399 RockPI 4

[Ezequiel Garcia]

Hi Matthias, Fabio:

Thanks for the reply.

On Fri, 8 Nov 2019 at 13:45, Matthias Brugger <mbrugger at suse.com> wrote:
>
> Hi Ezequiel,
>
> On 07/11/2019 20:27, Fabio Estevam wrote:
> > Hi Ezequiel,
> >
> > On Thu, Nov 7, 2019 at 3:45 PM Ezequiel Garcia
> > <ezequiel at vanguardiasur.com.ar> wrote:
> >>
> >> I decided to test latest U-Boot, following instructions in
> >> doc/README.rockchip. The instructions seemed
> >> clear and I could build this easily.
> >>
> >> However, there seems to be an issue when I save the environment. Any ideas?
> >>
> >> => saveenv
> >> Saving Environment to MMC... Writing to MMC(0)... OK
> >> => reset
> >> resetting ...
> >> U-Boot TPL 2020.01-rc1-00213-g0f282c1876af-dirty (Nov 07 2019 - 15:21:44)
> >> Trying to boot from BOOTROM
> >> Returning to boot ROM...
> >>
> >> U-Boot SPL 2020.01-rc1-00213-g0f282c1876af-dirty (Nov 07 2019 - 15:21:44 -0300)
> >> Trying to boot from MMC2
> >>
> >>
> >> U-Boot 2020.01-rc1-00213-g0f282c1876af-dirty (Nov 07 2019 - 15:21:44 -0300)
> >>
> >> Model: Radxa ROCK Pi 4
> >> DRAM:  2 GiB
> >> Cannot find regulator pwm init_voltage
> >> MMC:   dwmmc at fe320000: 1, sdhci at fe330000: 0
> >> Loading Environment from MMC... OK
> >> In:    serial at ff1a0000
> >> Out:   serial at ff1a0000
> >> Err:   serial at ff1a0000
> >> Model: Radxa ROCK Pi 4
> >> ## Error: Can't overwrite "serial#"
> >> ## Error inserting "serial#" variable, errno=1
> >> initcall sequence 000000007ffc10b8 failed at call 0000000000202a20 (err=-1)
> >> ### ERROR ### Please RESET the board ###
> >
> > I have observed issues like this with i.MX when U-Boot size grew and
> > overlapped the environment variable region.
> >
> > Here is one commit that fixed the issue for mx53loco board:
> > https://gitlab.denx.de/u-boot/u-boot/commit/033f6ea5fa5fce63d52c8c2b63d8284144415b88
> >
> > Try to investigate if this could be cause of the issue you are seeing.
> >
>
> You could also try the first two patches from this series:
> https://patchwork.ozlabs.org/user/todo/uboot/?series=132338
>
> Maybe you are hit by the over-writing discontiguous files bug in the FAT code.
>

It's probably something along those lines. OTOH, it seems RK3399 is under going
some more work, judging by some patches recently sent. I decided to go back
to some vendor bootloader until upstream settles a bit on the stable side.

Will be getting back to this sooner or later.

Thanks again,
Ezequiel