EmuTOS for Amiga

[Amiga1992]

Originally Posted by elpiloto

Sorry about my ignorance, but, Whos him? some one related to emutos?

That guy is related to skick but as I posted right after I told you that, it seems he wouldn't be cooperating.

Fear not though! It seems like BlankVector can fix EmuTOS to be softkicked/create RTB and PAT files, so hang in there and wait to see what comes out of this. I'd love to test this out on my A600

[elpiloto]

i contacted jean, no answer yet, we ll have to wait for BlankVectors fix

[Jacques]

Hi, it's 2 years since last post, is there any update on EmuTOS mapping with SKICK?

[BlankVector]

Nothing new in EmuTOS specific to Amiga.
However, new EmuTOS releases still support Amiga, just as before:
http://sourceforge.net/projects/emut.../emutos/0.9.4/

[Jacques]

Thank you for news, BlankVector.
I intended to run it on unexpanded A600 using SKICK, but with that still not possible, I successfuly managed to run it on ACA500-equipped A500.

However the default Amiga binary does not support FDD, HDD, etc.
Could someone please put Emutos compiled for Amiga with AROS=1 in the zone?

EDITED:

Disregard, pls. I managed to compile EmuTOS with AROS=1 by myself using Cygwin and following tutorial:
http://vincent.riviere.free.fr/soft/m68k-atari-mint/

[Jacques]

Unfortunately while A: disk drive is now working for me on Amiga500 with ACA500, I can't make it see CF card in ACA500's A600-compatible IDE port.
I've prepared CF card with single FAT16 partition, but it's still a no go.

Any clue how to make EmuTOS see HDD under real Amiga?

[BlankVector]

Originally Posted by Jacques

Disregard, pls. I managed to compile EmuTOS with AROS=1 by myself using Cygwin

Excellent

Originally Posted by Jacques

Unfortunately while A: disk drive is now working for me on Amiga500 with ACA500, I can't make it see CF card in ACA500's A600-compatible IDE port.

Currently, the IDE controller detection is quite poor.
There is a flag named has_gayle, it should be accurately detected with AROS=1.
But there is a big approximation in bios/ide.c, at the top of detect_ide():
#ifdef MACHINE_AMIGA
has_ide = has_gayle ? 1 : 0;

I suspect you to have no Gayle on your A500.
Try to force has_ide = 1 there, that should not hurt.

IDE support worked fine on Gibs' Blizzard A1260.

[strim]

Originally Posted by BlankVector

Excellent

Currently, the IDE controller detection is quite poor.
There is a flag named has_gayle, it should be accurately detected with AROS=1.
But there is a big approximation in bios/ide.c, at the top of detect_ide():
#ifdef MACHINE_AMIGA
has_ide = has_gayle ? 1 : 0;

I suspect you to have no Gayle on your A500.
Try to force has_ide = 1 there, that should not hurt.

Indeed there is no Gayle in A500. The ACA500 has kinda-compatible CF controller, but there are minor differences compared to normal Gayle IDE. It's seamlessly working on AmigaOS just because there's a modified scsi.device in ACA500 ROM.

On Individual Computers wiki there's document "Inside ACA500" that explains it all. Based on this doc I wrote the NetBSD driver (IDE part here, ACA detection and interrupt stuff here).

[wXR]

strim, the ACA500 works with NetBSD? I thought it needed at least 68020?

[Toni Wilen]

Originally Posted by strim

Indeed there is no Gayle in A500. The ACA500 has kinda-compatible CF controller, but there are minor differences compared to normal Gayle IDE. It's seamlessly working on AmigaOS just because there's a modified scsi.device in ACA500 ROM.

It is (just enough) Gayle IDE compatible (first slot only), it does work without flash rom driver (there is option to disable it in boot menu).

EDIT:

the ACA500 works with NetBSD? I thought it needed at least 68020?

It needs A1200 accelerator (with MMU) connected.

[strim]

Originally Posted by Toni Wilen

It is (just enough) Gayle IDE compatible (first slot only)

I remember the NetBSD Gayle IDE driver didn't work with ACA500 as-is, though I don't remember what was the exact reason for it not working. It's possible it was just detection related problem.

That's why I wrote the ACA-specific driver (and to support the second slot too).

[Jacques]

After today's work, I'll recompile EmuTOS with forcing has_ide = 1 and will check.

As it is now it behaves following way with ACA500 and AROS=1:
- no CF card in either ACA500 slot: EmuTOS starts immediately up to the info screen and then onto the GEM
- CF card present in either ACA500's slot (or both): EmuTOS freezes for about 1 minute with white screen, then resumes boot up and enters GEM, but no HDD is detected.

I tested with 16MB (FAT12) and 2GB (FAT16) formatted CF cards.

Btw. ACA500 has this nice feature allowing to load A1000 Kickstart from diskette, so I could test mapping EmuTOS using both ACAMAPROM command together with freshly complied EmuTOS.rom and also generated A1000-like ADF :-)

Hopefully forcing IDE detection will help, it's definitely worth it and great fun. ACA500 also seems very well suited for this kind of play with its features

Originally Posted by strim

I remember the NetBSD Gayle IDE driver didn't work with ACA500 as-is, though I don't remember what was the exact reason for it not working. It's possible it was just detection related problem.

That's why I wrote the ACA-specific driver (and to support the second slot too).

Could you possibly recall the reason, please? Some time ago I failed to use ACA500 CF port with "A600 compatible" HDD driver under KCS Power PC emulator (PC/XT), although partition C was initially detected, I couldn't format it under emulated PC, though.
Maybe there would be possibility to nail this problem, too

[Jacques]

@BlankVector
I recompiled EmuTOS with has_ide = 1 but it hasn't helped, actually it behaves the same way as described in post #152.

[BlankVector]

Jacques, thanks for your tests.
The timeout probably indicates trouble in the IDE driver.
We will need further tests. I will come back when I have precise ideas.

[Jacques]

You're welcome

Apart from A500 + ACA500 reported IDE issue, I tested EmuTOS a bit more and there are following issues I faced:

1) A1200 + Blizzard 1230-IV + 64 MB FAST: no Fastram detected on welcome screen (this line is totally ommited, shows just ST-RAM). That was tested on 2 Commodore Amigas 1200 + 2 different Blizzards 1230 (another A1230 is my friend's who tested, too)
2) FDD LED keeps lightened even after floppy is not being read anymore (and keeps spinning, too), it stops only after (at least) warm reboot.
3) my friend has newer-type Commodore Amiga mouse (more round one) , which works 100% correctly under AmigaOS. Yet with EmuTOS it behaves weird, after few clicks of left button (in random situations), EmuTOS gets frozen like left button is still pushed. You can still move the pointer, but it acts like button is pushed as if it happens over a GEM window, it moves the window without button being actually pressed. There's no other help than reboot for it (till the next button "freeze"). It doesn't happen under AmigaOS (believe me), it's just EmuTOS mouse routine looks confused by this otherwise working mouse

The rest is looking really good and promising, I've had a lot of fun playing with it! If you need some more tests to conduct in the future, I'll be happy to help.

And yes, keeping fingers crossed for ACA500 IDE-CF working with EmuTOS one day
Thanks again for all the great effor you've put into EmuTOS.

[strim]

1) A1200 + Blizzard 1230-IV + 64 MB FAST: no Fastram detected on welcome screen (this line is totally ommited, shows just ST-RAM). That was tested on 2 Commodore Amigas 1200 + 2 different Blizzards 1230 (another A1230 is my friend's who tested, too)

That's because in the current design where EmuTOS replaces Kickstart, the Blizzard's ROM never had the chance to execute and add memory. The problem with this was already discussed earlier in this thread.

[AnnaWu]

@BlankVector

EmuTOS support on WinUAE 3.1.0 beta series

Please could you take a look at this problem?
http://eab.abime.net/showpost.php?p=...&postcount=236

[BlankVector]

Fixed in latest EmuTOS snapshot
http://sourceforge.net/projects/emut...hots/20150828/

Now EmuTOS works fine with WinUAE 3.1.0 (I may have a few additional things to fix, though).

[amiman99]

Does compiled version supports the floppy drive and HDD?
If my memory is correct, there was a problem with using floppy code from AROS??
I like to mess around with EmuTOS on my A1000, but w/o floppy support it's kinda hard and I can't compile it myself.

[BlankVector]

Originally Posted by amiman99

Does compiled version supports the floppy drive and HDD?

The compiled version:
- Supports IDE hard disks on A1200. But I dumbly limited IDE detection to Gayle machines, so it currently does not work on A600. This will have to be fixed.
- Does not support floppy, because of licence incompatibility with AROS sources. But the 2 actual authors of the relevant code were OK to give their work to EmuTOS under GPL, so there is hope to solve that issue. This problem is getting old, I must work on that until it's too late.

The AROS-enabled version has support for floppies, unfortunately the binary can't be redistributed due to license issues. One can legally build it for himself using "make amiga AROS=1", though.

Originally Posted by amiman99

I like to mess around with EmuTOS on my A1000, but w/o floppy support it's kinda hard and I can't compile it myself.

This is very nice from you. We will try to find a solution.
I would be specially interested by 2 questions:
- If it still boots on A1000, because we changed a few stuff in early CPU initialization
- If there is abnormal screen flickering. It is supposed to be the same interlaced mode as AmigaOS with high resolution on A500. I had some doubts when using WinUAE.