Minimig (Amiga) core discussion

[DrOG]

For Amiga 500 create a config in the OSD that has "CPU = 68000, Turbo=NONE, Chipset=ECS, chipRAM=0.5MB, slowRAM=0.5MB, Kickstart 1.3".

AFAIK most A500 had OCS chipset, ECS arrived later with A500+ / A600.

[kolla]

Any possibility updates to MiSTer comes back to MiST ?

They typically do, yes.

[Threepwood]

For Amiga 500 create a config in the OSD that has "CPU = 68000, Turbo=NONE, Chipset=ECS, chipRAM=0.5MB, slowRAM=0.5MB, Kickstart 1.3".

AFAIK most A500 had OCS chipset, ECS arrived later with A500+ / A600.

You are right and I indeed use OCS. I missed that detail when I copied the line from the Wiki.

[slingshot]

Any possibility updates to MiSTer comes back to MiST ?

What updates do you think about? Fast RAM in DDR3, or what?

[apolkosnik]

I've looked at BFINS today, and it looks like there's a bug in the TG68K implementation. The source bits should be taken from the LSB side, and not from MSB side. Ugh. This is one of those weird corner cases where there's a MSB bit flipped in the second word of the opcode, where it shouldn't be.

[Sorgelig]

I've looked at BFINS today, and it looks like there's a bug in the TG68K implementation. The source bits should be taken from the LSB side, and not from MSB side. Ugh. This is one of those weird corner cases where there's a MSB bit flipped in the second word of the opcode, where it shouldn't be.

are you working on fix?

[apolkosnik]

Yes, though it will take me some time to iterate, fix and verify, as I'm pretty new to this vhd thing. I'm going through the output from cputest, but due to its brute-force approach, it's more like fuzzing.

I've looked at BFINS today, and it looks like there's a bug in the TG68K implementation. The source bits should be taken from the LSB side, and not from MSB side. Ugh. This is one of those weird corner cases where there's a MSB bit flipped in the second word of the opcode, where it shouldn't be.

are you working on fix?

[Sorgelig]

Anyway, i've decided to release the Minimig finally.

[raparici]

Probably the best bet is to go for whdload version. I think that last time I checked it didn't run on Vampire either.

It does work on current MiSTer Minimig for me, just one screen of them all does not render correctly on ECS (the toilet render). When in AGA mode in addition to that, the green cube does not render correctly either. I guess about Vampire you mean V4 right, as previous ones rely on original machine chipset.

[apolkosnik]

Anyway, i've decided to release the Minimig finally.

Good call. Let me summarize what I've got so far:

bfchg - eac0 31fe ... 2nd word should be 01fe since bits 15-12 of the second word should be 0. No dest in this opcode.
bfclr - ecc2 ec68 ... 2nd word should be 0c68 since bits 15-12 of the second word should be 0. No dest in this opcode.

bfexts - ebc0 f094 ... 2nd word should be 7094 since bit 15 of the second word should be 0 - makes it write into An.
bfextu - e9c0 116b ... bfextu d0{#5:d3},d1 ... if Dw(2nd word bit 5)=1 bits 4,3 should be 0. Also, if Do(2nd word bit 11)=1 bits 10,9 should be 0.
bfffo - edc0 d771 ... 2nd word should be 5771 since bit 15 of the second word should be 0 - makes it write into An.

bfins - efc0 7ee1 ... if Do(2nd word bit 11)=1 bits 10,9 should be 0. Also, if Dw(2nd word bit 5)=1 bits 4,3 should be 0.

bfset - eec0 8fc4 ... 2nd word should be 0fc4 since bits 15-12 of the second word should be 0. No dest in this opcode.
bftst - e8c3 3a9a ...2nd word should be 0a9a since bits 15-12 of the second word should be 0. No dest in this opcode.

[apolkosnik]

Probably the best bet is to go for whdload version. I think that last time I checked it didn't run on Vampire either.

It does work on current MiSTer Minimig for me, just one screen of them all does not render correctly on ECS (the toilet render). When in AGA mode in addition to that, the green cube does not render correctly either. I guess about Vampire you mean V4 right, as previous ones rely on original machine chipset.

There's an old v3 alpha core for V2 that has AGA emulation, but it seems based on the minimig base, as it was showing similar issues as with minimig.

[kolla]

Anyway, i've decided to release the Minimig finally.

Seems more stable so far, thank you

(I got the impression that Pinball Illusions now should work (as it now does on MiST), but it doesn’t)

[wins04]

Hi,

is there any routine in the Amiga Dos to show, what Amiga machine is running? A friend has created an multiboot workbench that runs on any Amiga machine and supports also all accelerator cards. Now we want to run this workbench on MiSTer (Minimig). Therefore it is important to get the information, that the current system is an MiSTer.

On base of this information the multiboot system switched certain configurations.

[Sorgelig]

Hi,

is there any routine in the Amiga Dos to show, what Amiga machine is running? A friend has created an multiboot workbench that runs on any Amiga machine and supports also all accelerator cards. Now we want to run this workbench on MiSTer (Minimig). Therefore it is important to get the information, that the current system is an MiSTer.

On base of this information the multiboot system switched certain configurations.

Use A1200 config. This is what MiSTer effectively emulates.

[wins04]

Sorry, maybe I didn't make myself perfectly clear:

I need to find out in a shell script if the script is executed on a mister.

Something like this (pseudocode):

If <fill in here> EQ TRUE

Echo "I am running on a Mister!!"

Else

Echo "I am not running on a Mister

EndIf


Is there any way to do this?

[Sorgelig]

Sorry, maybe I didn't make myself perfectly clear:

I need to find out in a shell script if the script is executed on a mister.

Something like this (pseudocode):

If <fill in here> EQ TRUE

Echo "I am running on a Mister!!"

Else

Echo "I am not running on a Mister

EndIf


Is there any way to do this?

No, and won't be such explicit flag. The goal is to simulate an original Amiga, so from inside point of view it must look like genuine Amiga. Probably in the future, when RTG will be available, it may have MiSTer in device name (if there is such).

Why WB for MiSTer should be different? I use HDD backup from my real A1200 on MiSTer and it works fine. FPGA is quite different from software emulators, you don't need to add special tweaks to make it work correctly.

[R4MS]

I can think of two (more or less) intended ways to detect a MiSTer:

1. When a hdd is autodetected (eg. with HDToolBox), the manufacturer will be "MiSTer".
2. The autoconfig ID of the memory expansion is 5020 / 1. I did not find this ID in the http://amiga.resource.cx database. Might be the same on the MIST tough.

Method 2 ist easy to check, but probably not very reliable.

[kolla]

http://aminet.net/package/util/libs/BoardsLib

Code: Select all

3.Minimig:> showboards allknown mid=5020

Manufacturer 5020: Rok Krajnc
  1: Minimig Z2 FastRAM                                  ID: 0
  2: Minimig Z3 FastRAM                                  ID: 16
  3: Minimig Z3 EthernetCard                             ID: 32
  4: Minimig Z3 GraphicsCard                             ID: 48
  5: Minimig Z3 SoundCard                                ID: 64

5 boards known of this manufacturer.

3.Minimig:>

I play around some more with it on the MiSTer tomorrow.
Of course it is still possible to test whether CPU is TG68 or not, hehe

[R4MS]

Serial ID on MiSTer is 1 or 2, which does not exist in that list.

[kolla]

Serial ID on MiSTer is 1 or 2, which does not exist in that list.

Well, then it should be easy to identify the MiSTer as long as fast ram is enabled

Boards.library has status as “semi official” since it is used, when available, by for example ShowConfig of OS 3.1.4 etc.

[kolla]

Serial ID on MiSTer is 1 or 2, which does not exist in that list.

Hm, on my MiSTer, they are "Minimig Z3 FastRAM Prod=5020/16", just like on the MiST.

[R4MS]

OK. You're right. The Product ID is 16. The 1 in 5020/1 displayed in Sysinfo is the optional serial number. The explanation in the HRM is

This is for the manufacturer’s use and can contain anything at all. The
main intent is to allow a manufacturer to uniquely identify individual
cards, but it can certainly be used for revision information or other data.

Don't know what that would be on MIST.

[kolla]

Yes, I figured this out too this morning - ShowBoard has a "serial" flag to display serial numbers, and I can find out later today what it says for the memory boards on the MiST, I suspect it says the same 1 and 2

Edit: yup, they are also 1 and 2.

[kolla]

Seems more stable so far, thank you

As it now has been running for days even as I have been using it quite actively, I can say that it is indeed stable again.

[apolkosnik]

Got some more fixes. This seems to fix the graphical corruption in Andromeda's Nexus7 demo. PR submitted.