Hi all,
Doing some labs (to see what claude code could do) and got a hand!

Aros does do the romtag stuff but doesnt show on the screen!

Thanks Bloodline!

As soon as i have more spare tokens i will ask to get uefi on emu68 to see how it goes!

https://github.co...berry-labs

Can you give some more context? What are we actually looking at?

Sure,
I started a raspberry labs (to try bare metal stuff), and found a bloodline bare metal emulator ( https://github.co...n1xp/Omega ) tried to play with it and somehow it worked. i can load roms and boots adfs. It is still labs. i am tring to add hdf support (using Zorro 2) atm.

But i would like once i gotta better understand, try to make with it more labs, like using emul68 as processor and add chipsets in (the real host) memory map, maybe in the future run munt ( https://crates.io... ) try to have aros in the in between, dunno how aros would be in the mmu and smp stuff, but, atm you are looking the bloodline emulator rom screen in bare metal. (far from be any optimized, but i got surprised that it even worked), there is a lot ahead to try, atm is really just labs. zero architect, it is almost the next, next, next, done from claude code.

Regards,
ps: i used rust cause it was just fun, but now i regreat the choice as the comunity is more C oriented, if things goes ok im thinking of stick with C, cause maybe others can be interested in too.

ps2: in qemu aros show the boot screen

Coolcat! o/

Ah ok so it's like running Amiga programs "nativelly" under RasperryPi. Cool!

A quick side question, does aros 68k 64bits exists?

http://www.apollo...ge.php?b=4¬e=11638

Ps: I wonder how aros arm be 64 bits would fit. But that is a far future question (have a mmu that could handle the 64 bits with 32 bits legacy support). Have such memory model and map, and the smp (multiple cores) are the two more intricated parts (but pistorm land with the emul68 idea of have ppc in another core make me think about ipc design in this kind of environment). As I said a far far away question, but it would be the ultimate scenario. AROS native with legacy support (and the source code level of compatibility maybe doesn’t give AROS native full freedom for how it would work and I’m not even sure how much is gained from it. All this kind of retro support is already emulated. I want to play with emulation and make it modular, but the ultimate scenario of a ideal environment is a full mistery for me)

I like a lot the raspberry labs, but have zero ideas where is the goal to reach. (Memory model and map, smp and the ABI that would make more sense)

Hi, Oli, how are you brother?

Hi Deadwood, it is in early stages, but yes, i jumped out of my labs to a project, dont know if it will works, but so far i can load the kickstart (only, no gfx yet)

https://github.co.../bellatrix

lets se how far it goes!

Best of luck Sounds like a great project

Hi argos, (I don’t know how to mention properly someone here).

Thanks, it is a new (old) project. 😃

Anyway, it seems that things are working (the rom is loaded, try to boot and the chipset and emu68 can talk to each other), but the kickstart “try” to initialize and stop.

I think it’s something timing related (vbl/tod thing).

It is somehow good news, Emu68 does all the heavy lifting (bare metal stuff) and the communication between “my code” and it is up.

I hope we can have the happy hand soon (it’s far from be usefull but it’s a good milestone, and my objective atm), after that, things can go in a cumulative pace. (But really, I’m not thinking about it, quite happy that the communication is ok, like intreq and ipl things, I didn’t knew this would work)

I hope to “stop” once I got the aros rom screen. 😎

Regards,

Hi all,
I will try to change my approach for emulate a full machine, the idea of have loosely coupled hardware doesn’t make the kickstart happy and the challenge becomes from the aros (kickstart) side, not from the emu68.

Winuae now has Draco emulation, have aros ported to draco maybe would be half path for decouple the hardware emulation, but it’s far more complexity for me atm.

So I will try to emulate a virtual machine good enough to meet kickstart criterias (timing hell).

After that, if, it would need to customize the roms (I don’t know even if I will sucess with the virtual machine, and customize the rom would be even harder for me)

So the objective is first emulate a full machine, using emu68, to boot regular roms, anything beside that is an after project (unfortunately).

But have an emulated machine with emu68 could be a good start, emu68 to the masses 😁

Regards,

Ps: I am using Claude code, and I don’t know how good it will handle the aros codebase, maybe it would need some external tool to index it and it would be even more token$.

A quick update, if someone is interested 😁

I had step back to musashi to test the machine, I still don’t get the happy hand (cat), but aros “boot” till trackdisk.device (aros rom serial debug is helping a lot)…

Let’s se how it goes. I think next I will add full support to adf to see if I can pass the trackdisk.device.

A quick question, if I want to boot from aros.iso I would need something like this or there is an easier way?

https://github.co...1-software

Best regards,

Hi again,
Better overall view about the iso booting, but not there yet. I passed the trackdisk.device problem, but still no happy hand.

Out of ideas.
I got white on white drawing=nothing.(gray in aros case)

Omega2 of bloodline does show the gfx but I don’t know what I am missing. I think I have implemented all that has in omega2 (for booting), but for sure I am missing something.

Is there a checklist of what is needed to reach the happy hand? Even reading aros source I couldn’t find and no hd space to build custom debug on aros rom.

Anyway. The good is that having a propper env I should overcome the problem, the bad, atm I am stalled.

It seems to be a missing bit in some register but atm I can’t pinpoint it, either with a custom rom or with 68k cross compiler tools. This will take more time than I first expected.

Regards,