U-Boot SPL crashes after stack setup

[Simon Glass]

Hi Osman,

On Mon, 9 Sept 2024 at 21:20, Osman <aibaykaro at gmail.com> wrote:
>
> Hello,
>
> I am having a crash with u-boot-spl simulated in spike, to be more spesific
> the problem occurs after this line located in start.S:
>
> mv sp, t0

Which start.S and which architecture is this?

>
> and spike prints this:
>
> dmi_read(0x05 (data[1]) -> -1 because abstractcs.busy==true
>
> dmi_read(0x04 (data[0]) -> -1 because abstractcs.busy==true
>
> My first thought was that the memory was unaccesible but that isn't the
> case because i can see the contents of the memory through GDB (i will also
> attach memory.png that shows the address at the sp pointer) and i provided
> the memory for spl in spike like so:
> ./spike -H --rbb-port=9824
> -m0x80000000:0x20000000,0x90000000:0x100000,0x00000000:0x20000,0x70000000:0x10000
> $SPL_PATH
>
> My second thought was that maybe somehow the size wasn't enough and the sp
> just overlaps with some region but i checked the readelf (i will attach it
> as readelf_output.txt) but nothing seemed unusual.
>
> I am not really experienced in this topic so i would really appreciate it
> if someone can guide me on the right direction.

I'm not sure. Could the stack-pointer alignment be bad, perhaps?

OK I found that code in RISC-V.

+Heinrich Schuchardt might have some ideas?

>
> You can see the config files attached to end of this e-mail.
>
> Regards

Regards,
Simon