[BUG] sandbox: './u-boot -l ' fails

[Simon Glass]

Hi Heinrich,

On Sat, 19 Sep 2020 at 13:48, Heinrich Schuchardt <xypron.glpk at gmx.de> wrote:
>
> Hello Simon,
>
> when I try to run ./u-boot -l the sandbox stalls. Shouldn't it run out
> of the box?
>
> $ ./u-boot -l -d arch/sandbox/dts/sandbox.dtb

For the record you should be able to use -D to get the same effect as
your -d above.

>
> U-Boot 2020.10-rc4-00018-g21a10244f9-dirty (Sep 19 2020 - 19:55:39 +0200)
>
> Model: sandbox
> DRAM:  128 MiB
>
> Warning: host_lo MAC addresses don't match:
> Address in ROM is               26:4b:ca:6c:98:f4
> Address in environment is       00:00:11:22:33:44
>
> Warning: host_virbr0 MAC addresses don't match:
> Address in ROM is               ee:3e:c9:ce:1f:9c
> Address in environment is       00:00:11:22:33:45
>
> Warning: host_docker0 MAC addresses don't match:
> Address in ROM is               c2:85:07:7b:9a:18
> Address in environment is       00:00:11:22:33:46
> WDT:   Not found!
> MMC:
>
> No output after this point.
>
> The problem also exists with U-Boot v2020.07, v2019.10, v2018.11.
>
> CONFIG_SANDBOX_SDL=y
>
> SDL_InitSubSystem() never returns. It is looping somewhere in U-Boot's
> __serial_getc(). I wonder how it gets there without returning from the
> function.
>
> I compiled SDL2.cpp from
> https://gist.github.com/miguelmartin75/6946310#file-sdl2-cpp-L18
> with
>
> g++ SDL2.cpp -D_REENTRANT -I/usr/include/SDL2 -lSDL2 -lGL -o test
>
> and it runs fine showing an X11 windows with red background.
>
> So there seems to be no general problem with the SDL2 library.

I hit this myself on another computer and it turned out to be that SDL
defined getc(), as does U-Boot, and things get confused. At least I
think it is getc.

I hacked around with changing the name of getc (I think it was getc)
in U-Boot and the problem went away.

Regards,
Simon