litex_json2renode: fix --bios-binary and add --opensbi-binary

[AndrewD]

--bios-binary is reverted to loading the bios into rom as per the documentation and litex defaults --opensbi-binary is added to load the openSBI binary into the opensbi memory region

The examples documented in the wiki now work again

Fixes #1785

The Generate Renode simulation wiki page should also have one more section:

Finally, you can run load the simulation by executing the command:

renode litex.resc

Then in the renode terminal window execute the start (or s) command to boot the bios

(litex_vexriscv) start

@MateuszKarlic I think this change makes sense but as you have done more significant changes in the past maybe you can confirm this approach.

[MateuszKarlic]

@AndrewD I confirm that both changes make sense. It was my mistake to change behavior of this switch.

In your commit, I'd suggest setting PC for all cores when loading Bios, not just cpu0, unless this somehow breaks the examples:

for cpu_id in range(0, number_of_cores):
    result += f"cpu{cpu_id} PC {hex(bios_base)}\n"

Since the options are not mutually exclusive, it might be helpful to print message to user when loading both, that the PC will be set to rom_base not opensbi_base.

[AndrewD]

@AndrewD I confirm that both changes make sense. It was my mistake to change behavior of this switch.

In your commit, I'd suggest setting PC for all cores when loading Bios, not just cpu0, unless this somehow breaks the examples:

for cpu_id in range(0, number_of_cores):
    result += f"cpu{cpu_id} PC {hex(bios_base)}\n"

I wasn't sure it made sense to boot the bios on all cores, but I'm only using single core ATM so either is fine with me.

Since the options are not mutually exclusive, it might be helpful to print message to user when loading both, that the PC will be set to rom_base not opensbi_base.

I wondered about making it mutually exclusive or adding a warning. As you've also raised this I'll make it mutually exclusive as I could not think of a reason to use both together.

[AndrewD]

@MateuszKarlic thanks for the review!

[MateuszKarlic]

Happy to help :)