MagiC/MagiCMacX source updated

[mikro]

Here is a ROM-version of the latest build from Thorsten. It is mostly translated, using some of Atarizoll's and some of mine. It is built from the MAGIC.RAM using r2 p12 a0. I like to see the versions and text on the boot screen, that's why 12 seconds. It's skippable anyway.

It's running as-is on my MonSTer STacy from IDE. None of that offscreen driver BS. I'm booting from the ROM and I have verified it to save 237KB of RAM.

Thank you for this image and your hardware description. Now I can safely say that MAGICROM definitely doesn't produce images valid for the MegaSTE. I feared that MonSTer somehow gets in the way but since it works for you, it must be the MegaSTE, it's even mentioned in MAGICROM's documentation.

So what makes MegaSTE so special that the converted ROM image doesn't work? Obviously not the 0x00E00000 address and since MAGIC.RAM normally works, obviously it's not an OS issue.

[joska]

It is quite possibly caused by the same issue as you've seen with 1.04 where you have to reset from 2.06/EmuTOS. I guess MagiC relies on TOS to initialise some MSTE-specific hardware.

[mikro]

It is quite possibly caused by the same issue as you've seen with 1.04 where you have to reset from 2.06/EmuTOS. I guess MagiC relies on TOS to initialise some MSTE-specific hardware.

That's an intriguing thought. Maybe there's a way to verify this in Hatari? TOS 1.04 on Hatari's MegaSTE should be easier but MAGIC.ROM in Hatari shouldn't cause much trouble either.

[mikro]

It is quite possibly caused by the same issue as you've seen with 1.04 where you have to reset from 2.06/EmuTOS. I guess MagiC relies on TOS to initialise some MSTE-specific hardware.

Actually, there must be more than meets the eye. While "jump starting" TOS 1.04 with TOS2.06/EmuTOS 256K/patched EmuTOS 192K works, MagiC stubbornly refuses to boot.

[ThorstenOtto]

Maybe its easier to analyze what magcrom6 does, that MagiC doesn't: https://github.com/th-otto/MagicMac/blo ... magcrom6.s

[mikro]

Maybe its easier to analyze what magcrom6 does, that MagiC doesn't: https://github.com/th-otto/MagicMac/blo ... magcrom6.s

Actually, Ppera has pulled this off and fixed the image for Mega STE. When I compared his older UK image and the one for MegaSTE, there is just a few bytes of code difference so perhaps it wouldn't be so hard to incorporate the changes into new builds?

That would allow people to create their own ROM images, with custom timeouts, maybe even Alt-RAM would be working out of the box, who knows.

[ThorstenOtto]

But he also writes "Adding Mega STE support is not so simple", so i guess his ROM image won't work in STE. And, as expected, he's violating the GPL by offering only the binaries.

[mikro]

I think he referred to the TOS 1.x image with MegaSTE support, not MagiC (but I may be wrong here). So there's still chance that the MagiC change is really only about changing few bytes. When you think about it, it's really not rocket science: disable interrupts, set scc interrupt vectors, reset scc, done (verified in EmuTOS).

[ThorstenOtto]

Yes, i think too that it can't be that much to do. The reason why i linked to the source of magrom6, is that i always wanted to get that integrated, but never got to it.
And maybe someone could verify whether his new version works on STE.

Apart from that, a separate step to create the ROM image from the RAM file is always needed, since Pure-C is not able to build just a binary file, without a gemdos header.

[mikro]

As long as MagiCROM works with 6.20 (which looks like it does), this shouldn't be a huge issue.

[mpattonm]

Pardon my ignorance, would the utility work on/for Falcon HW?

[mikro]

Pardon my ignorance, would the utility work on/for Falcon HW?

Perhaps not without some (slight?) modifications. MAGIC.RAM/ROM is meant for 256K space, so if nothing else, one would need to make the padding up to 512K. Also, perhaps other hardware would be needed to initialise, similar as SCC for MegaSTE.

[karnahai36]

Awesome work to get the sources to compile! I think the next year will be bring some good things for MagiC!

hellodear.in

teatv.ltd

[arf]

But he also writes "Adding Mega STE support is not so simple", so i guess his ROM image won't work in STE. And, as expected, he's violating the GPL by offering only the binaries.

… isn’t it that you don’t have to publish then alongside your binaries but that it’s sufficient to hand out the sources upon request?

[mikro]

I think Ppera is patching the binary directly, without using the sources, so I wouldn't be so harsh about it.

[ThorstenOtto]

A patched binary would have the same size, but his doesn't. Also, since most of MagiC is already in assembler, there is not much place to make room for new routines, and certainly not for a new harddisk driver.

[Gaiyan]

Hi,

I'm back to giving this a whack with the latest snapshot from Otto (magicmac-20220406-132308-bin).

Everything is fine except, again, with something in the XTENSION folder. If I copy everything else from the snapshot everything works fine. As soon as I copy the XTENSION-folder contents, I get the memory errors as before. This is on the STacy again, with or without TF536.

Unlike earlier though, the MAGIC.RAM is perfect. It's perfect from the root of C: and creates a perfect ROM with MagRom6.

I will try to narrow the offender in the XTENSION-folder and report back. I will also post an english ROM once I translate the bootscreen generated by the MAGROM6.

[ThorstenOtto]

Nice. Maybe you can figure out which of the slb in the extension folder causes problems. pdlg.slb is rarely needed, since only a few programs are aware of the functionality. editobjc.slb is needed eg. by mgnotice.

[Gaiyan]

Nice. Maybe you can figure out which of the slb in the extension folder causes problems. pdlg.slb is rarely needed, since only a few programs are aware of the functionality. editobjc.slb is needed eg. by mgnotice.

Alright, I did some more thorough test. I used the new MAGX.INF which bypassed my jinnee setup to a stock install. I made sure no AUTO-programs or ACCs, etc were contributing to the errors.

So basically having a clean install of only the components in the snapshot, I can't even boot to MAGXDESK as I get memory errors. If I replace XTENSION folder with my own install, everything is good.

Then, only replacing editobjc.slb from the snapshot into the XTENSION folder allowed me to boot but I was missing 62MB of RAM. So it's undeniable that at least editobjc.slb is causing issues and yes, MGNOTICE seems to load by default. I'll try others.

[Gaiyan]

and yes, MGNOTICE seems to load by default. I'll try others.

Correction. MGNOTICE does not load by default. Even without loading it 62MB are gone.

However, I forgot about the START-folder in GEMSYS\MAGIC. Once I disabled it and did a REAL clean boot, I did not get any terminating memory errors but am missing the roughly 62MB of RAM.

[ThorstenOtto]

By "missing 62MB" i assume you mean 62MB less memory available compared to a clean boot? Did you take a look a u:\proc, whether any process is using an extraordinary amount of ram? for example, this is what i get:

Screenshot_20220822_144506.png

[Gaiyan]

By "missing 62MB" i assume you mean 62MB less memory available compared to a clean boot? Did you take a look a u:\proc, whether any process is using an extraordinary amount of ram? for example, this is what i get:
Screenshot_20220822_144506.png

Ah, forgot about that trick. Sorry, don't have a screenshot but here's the output:

Code: Select all

AESSYS.002	34.7 KB
PDLG  .004	63.6 MB
JINNEE.005      609.6 KB	

So what does PDLG do again?

[Gaiyan]

Ok, when I restored all my files and settings and kept all files from the snapshot, this is the result:

SNAP_26.GIF

So it seems like those two are the offending ones because replacing them fixes everything it seems.

[ThorstenOtto]

So what does PDLG do again?

That are the printer dialog functions (pdlg_*). It is loaded automatically when any app uses them.

Edit: hm. just wonder whether the size in u:\proc is really correct. 42k for AESSYS seems a bit low. OTOH, 778 in my snapshot seems a bit much.

[Gaiyan]

So what does PDLG do again?

That are the printer dialog functions (pdlg_*). It is loaded automatically when any app uses them.

Edit: hm. just wonder whether the size in u:\proc is really correct. 42k for AESSYS seems a bit low. OTOH, 778 in my snapshot seems a bit much.

I'm sure it's legit because of the terminating memory errors. I'm running 640x400 in mono, maybe that has an effect?