Minimig (Amiga) core discussion

[Glaurung]

OS must be installed in order to play games?
how play disk games?
edit:ok, just f12+load disk ... nice!!

The FPGA amiga is better than a pc-emulator amiga?

[mahen]

Nice guides ! I wonder to what extent it's more relevant to start from the newly released WB 3.1.4 and to what extent the tweaks can be applied...

[BBond007]

OS must be installed in order to play games?

To play ADF games, you do not need workbench installed.

For whdload games, you do. I typically prefer to run whdload games (and demos) for convenience but also like to use (cracked) ADFs from time to time because they more commonly contain trainers and cheats in the loader.

how play disk games?

You start ADF(Amiga Disk Format) games like you with any of the console cores with ROMs

Some of the cracked games fix certain incompatibilities with the backwards compatibility of older games, still It may be necessary to configure the core differently per game.

ADFs that are not "AGA" - KS/Workbench 1.3 with 68000/TURBO OFF/1MB CHIP/0MB FAST/OCS/PAL. Some older games might even need 1.2 ROM or 512K Chip. This config resembles an A1000/A500.

ADF games with "AGA" in the title - KS/Workbench 3.1 with 68020/TURBO OFF/2MB CHIP/0MB FAST/AGA/PAL. This config resembles an A1200.

The FPGA amiga is better than a pc-emulator amiga?

That matter of opinion. Minimig core is very nice and plays games very well, however pc-emulators are now very mature and can support higher-end Amiga systems with hardware expansions and levels of performance not currently available on the MiSTer Minimig core. Both solutions offer a wider range of compatibility than you would get with any single model of real Amiga.

[jemismyname]

You can use same kickstarts that FS-UAE is using:
https://fs-uae.net/docs/kickstarts

Before trying to configure/install Workbench 3.1 on Mister I highly suggest you trying to do in FS-UAE just to get hang of it, because some essential knowledge is required about using Amiga. Just follow their documentation:
https://fs-uae.net/docs

Also good guide about installing Workbench 3.1:
http://guide.abime.net/home3.htm

And superb customization guide after you've installed Workbench 3.1:
https://devtty.io/2018/07/computing/how ... kbench-os/

Ok, thanks for all this good information!

I keep seeing references to the "amiga forever" collection.

What is the general consensus of this? is it good/worth buying? who actually gets this money?

does any money go to any original amiga developers or is this just a cash grab by a holding company?

can anyone comment on which of the 3 packages is the best value?

(i'm assuming either the $10 or $30 package)

[BBond007]

who actually gets this money?
does any money go to any original amiga developers or is this just a cash grab by a holding company?

I don't believe anything goes to Toni Wilen the developer of WinUAE which their product is based on.
Cloanto is currently in a legal battle with Hyperion (the makers of the OS 3.14 update and 4.X PPC versions).
The money probably goes to lawyers.

can anyone comment on which of the 3 packages is the best value?
(i'm assuming either the $10 or $30 package)

It looks like the $10 version does not come with any ROMS besides Kickstart 1.3. That means no AGA games.

There is also Amiga Essentials for Android ($1.99) which claims to include Kickstart ROMS 1.2, 1.3, 2.04, 3.0 and 3.1.

https://www.amigaforever.com/android/

[Glaurung]

I can't pass the "crack screen" in a lot of disk games.
The trainer/loader works, to select "ininity energy" or whatever (the keyboard responds when press F1,f2,f2) , but thats all.I cannot start the game (or I dont know how to start it).With keyboard oy joystick.

doing something wrong?

[BBond007]

I cannot start the game (or I dont know how to start it).With keyboard oy joystick.

doing something wrong?

Did you try the mouse? Remember, all Amigas from day one were shipped with a mouse.

[Glaurung]

I cannot start the game (or I dont know how to start it).With keyboard oy joystick.

doing something wrong?

Did you try the mouse? Remember, all Amigas from day one were shipped with a mouse.

Finally working but It's a little confusing the way it works the keybaord/joystick selection in OSD (bloqnum+/selecting keyboard/joystick in OSD, etc etc

Barbarian2, tellsme to change the disk 5-6 times.

the game feel perfect

[R4MS]

I've been looking at the memory management of the Minimig core, mainly
to get hrtmon running with fastram enabled.

Somehow the core currently maps all non-fastram into the first 4M of the
sram, probably for historic reasons. This means in particular, that
the 1M ROM support limits slow RAM to 1M, respectively 0.5M when
combined with hrtmon. Moreover there is a bug in the fastram mapping
of the tg68k code. In KICK turbo mode it maps $e0 on the Amiga side to
the same sram slot as the first 512K slow ram. (That has a trivial fix)

The sram design allows the non-fastram to occupy the first 8M bank of
the sram module. Thus I did fix all these issues by changing the
mapping to

Amiga - sram
00-1f - 00-1f
a0-ff - 20-7f

This is a more natural an less bug prone mapping and leaves enough
space to enable all features simulaniously.

Currently working:

hrtmon in all configurations
up to 1.5M slow ram independent of other settings

Not working:

hard disks... have to look up the memory path

Not tested:

1MB Kickroms

Once I solved the problem with the hard disks I'll upload the modified core for testing.

[R4MS]

Here is the modified core for testing:

* does give you the amount of slow ram selected in OSD (normal core 1M max, 0.5M with hrtmon enabled)
* maps $e00000 to the correct region (normal core maps this to the same bank as $c00000, so 1M Kicks will not work)
* disables writing to kickram. (with normal core, kickram is not write protected)

I did not make progress on hrtmon. But I found out that it works, when the turbo mode is off (can be changed in running system). If it's been started once, it can be called again even when the turbo mode is on.

I already forked the sources and will upload my modifications within a couple of days. Need to do some cleaning first.

[Sorgelig]

Here is the modified core for testing

Thanks for improving the core! Make a pull request when you will be ready.

[R4MS]

Thanks for improving the core!

Thank you for this great project!

Would you prefer to stick to bug fixes only? I did also clean up some things... like for example in sdram_ctrl.v, chipAddr is [23:1], but bit 23 is not used, since it only accesses the first 8M block of the sdram. I therefore changed the wire to [22:1] in the module definition to make the interface more informative. Also I'd like to prune the cpuDMA line from tg68k to sdram_ctrl.v, since it is not used at all, but in the tg68k sources it looks like it would control fast ram access. These changes make it much easier to understand and modify the code, but I am not sure if they should stay to make porting to other platforms easier.

What is your opinion on this topic?

[mahen]

Hi ! In the first post of this page (http://www.atari-forum.com/viewtopic.ph ... &start=475) Mike mentions the M68K soft 020 core "which is a tidy up of" TG68K and supposedly fixes quite a few issues (like the Pinball Illusions crash I guess). Would anyone be willing to implement this ?

Personally I would definitely fund via Patreon anyone willing to maintain especially the Minimig core (in addition to Sorgelig & others of course !)
Maybe I should mail Rok about this

[Sorgelig]

Thanks for improving the core!

Thank you for this great project!

Would you prefer to stick to bug fixes only? I did also clean up some things... like for example in sdram_ctrl.v, chipAddr is [23:1], but bit 23 is not used, since it only accesses the first 8M block of the sdram. I therefore changed the wire to [22:1] in the module definition to make the interface more informative. Also I'd like to prune the cpuDMA line from tg68k to sdram_ctrl.v, since it is not used at all, but in the tg68k sources it looks like it would control fast ram access. These changes make it much easier to understand and modify the code, but I am not sure if they should stay to make porting to other platforms easier.

What is your opinion on this topic?

I'm open to any improvements being it fixes or new features. So, you are welcome to add all your changes.

Hi ! In the first post of this page (viewtopic.php?f=117&t=32761&start=475) Mike mentions the M68K soft 020 core "which is a tidy up of" TG68K and supposedly fixes quite a few issues (like the Pinball Illusions crash I guess). Would anyone be willing to implement this ?

M68K haven't been released. He gave test version of source to some people with condition to not release the sources till he finish his improvements. Minimig is open source core, so such condition is unacceptable. Thus i wait for release M68K first and only then it will be integrated (if it will be compatible).
So, ask Mike to release the sources.

[Sorgelig]

Personally I would definitely fund via Patreon anyone willing to maintain especially the Minimig core (in addition to Sorgelig & others of course !)
Maybe I should mail Rok about this

being non-stop user from 1993 you should be very familiar with Amiga HW internals So why don't try to improve the core by yourself?
No one born with HDL knowledge.

[mahen]

being non-stop user from 1993 you should be very familiar with Amiga HW internals So why don't try to improve the core by yourself? No one born with HDL knowledge.

Heh, it would have been great, in my former life. Impossible now sadly due to several reasons I won't state here as it's a public conversation

[slingshot]

Hi ! In the first post of this page (http://www.atari-forum.com/viewtopic.ph ... &start=475) Mike mentions the M68K soft 020 core "which is a tidy up of" TG68K and supposedly fixes quite a few issues (like the Pinball Illusions crash I guess). Would anyone be willing to implement this ?

Personally I would definitely fund via Patreon anyone willing to maintain especially the Minimig core (in addition to Sorgelig & others of course !)
Maybe I should mail Rok about this

That CPU core is still not released. The last test version I tried had a good 68020, but broken 68000 mode.

[R4MS]

Hrtmon now works also in turbo mode. The core sends the NMI_vector ($a1000c) to the CPU, when it accesses $7c (the location of the NMI_vector in the 68k exception table). A smart way to prevent programs to change the NMI_vector. The corresponding code in the fast path to memory (tg68k.vhd) was not functional. Which was good, because it would have crashed the system due to multiple bugs.

For one, it did not respect the boot overlay... so I found out, that the long word at $f8007c is not crucial for Kick 1.3 but when set to $a1000c immediately crashes 2.0 and 3.1

Once Hrtmon is started, it writes its NMI_vector to $7c. So if not overwritten, it does not need this special functionality for the second call. That explains why before it had to be started when not in turbo mode first.

Heh, it would have been great, in my former life. Impossible now sadly due to several reasons

You should really try HDL. Haven't had so much fun coding (ah... well mostly debugging) in a while.

[R4MS]

OK... here is the core with hrtmon working. I made sure that the processor only reads the NMI vector through the chipset (cart.v). This way, the only thing the CPU needs to know, is the address of the vector (VBR+$7c) which needs special treatment. Everything else is done in cart.v, which in particular avoids duplicate code.

I guess I'm out of obvious fixes around memory mapping. I'll clean the code and upload it within a couple of days.

Everbody interested in the Minimig core should test this version, before the changes get pulled into the main MiSTer repository. Thanks!

BTW. There is a reason why I was confused about when hrtmon works: The turbo modes depend on multiple settings:
You get turbo chip when selecting it in OSD and additionally AGA and 2M Chipram. Turbo kick is on when it is selected in OSD OR AGA is selected.

[Sorgelig]

It's better to submit your patches through PR, then i will make a formal release.
In this case it will get larger user base for testing.
Release is not something set in stone. So you can continue improvement and request for new release if there is something critical.

[R4MS]

I personally prefer to have a couple of voluntary people testing the core, before it is distributed (via the update-script) to a lot of (not necessarily voluntary) testers. This core just has too many options...

[BBond007]

Has anyone noticed issues with IDEFix?

I have noticed that my whdload games frequently hang on startup when I have IDEFix enabled in s:startup-sequence even if the HDF not mounted on the 3rd or 4th controller.

Another user was having issues with whdload games hanging on startup and I suggested disabling IDEFix and that apparently worked for them too.

I also notice a similar hang if I copy large files (ie. 100MB HAM videos) with IDEFix enabled. If I copy the same files without using IDEFix, it never hangs.

The mouse still works and workbench windows can be moved, but they don't redraw and the system is basically dead...

Thanks

[R4MS]

Do you mean http://aminet.net/package/driver/media/IDEfix97 ? What is the purpose of this driver in MiSTer? Does it increase transfer rates?

[remowilliams]

I've noticed the exact same issues with WHDLoad hanging and IDEFix was the first thing I disabled in that image to address that problem. It's not something I've ever used with my real A1200.

[Sorgelig]

Do you mean http://aminet.net/package/driver/media/IDEfix97 ? What is the purpose of this driver in MiSTer? Does it increase transfer rates?

It allows to use 4 IDE drives.
Would be good if you can check it. I think you are more familiar with Amiga HW than me.
But i didn't encountered problems with IDEFix. may be because i'm not extensive Amiga user.