Work on the Minimig core?

[vebxenon]

I've tested again. MiST makes noises and core doesn't boot. I've been working for hours today with 20190418 version testing some Midi modules, but new core doesn't work at all . I've tried even changing the SD card and no success.

[solskogen]

I use menu.rbf to load the other cores, could you try that as well?

[desin24]

20190924.rbf works here too

[vebxenon]

I use menu.rbf to load the other cores, could you try that as well?

I'm going to try It.

[vebxenon]

I use menu.rbf to load the other cores, could you try that as well?

I've tried it. No success. Kickstart doesn't boot.

[GoingDown]

I've tried it. No success. Kickstart doesn't boot.

Can you delete minimig settings file and create new configuration from scratch? Just in case something is messed up there?

[vebxenon]

I've tried it. No success. Kickstart doesn't boot.

Can you delete minimig settings file and create new configuration from scratch? Just in case something is messed up there?

I´ve done it twice

[retrofun]

Fixed the RTR instruction!

Does NOT fix any issue I currently know (testkey4, Pinball Illusions).

Latest cputest 20190924 from Toni Wilen:

Before:

Code: Select all

CPUlvl=2, Mask=ffffffff Code=00740000
RTR:
data/68020/rtr/0000.dat. 0...

0:      4e77    rtr

A7: expected 00777ffe but got 00778004
Registers before:
D0: 00000010 D1: 00000000 D2: ffffffff D3: ffffff00
D4: ffff0000 D5: 80008080 D6: 7fff7fff D7: aaaaaaaa
A0: 00000000 A1: 00000080 A2: 00008000 A3: 00007fff
A4: fffffffe A5: ffffff00 A6: 0073ff00 A7:*00777ff8
SR: 0400   PC: 00740000 ISP: 0077ff80 MSP: 00780000
T1=0 T0=0 S=0 M=0 X=0 N=0 Z=0 V=0 C*0
Registers after:
D0: 00000010 D1: 00000000 D2: ffffffff D3: ffffff00
D4: ffff0000 D5: 80008080 D6: 7fff7fff D7: aaaaaaaa
A0: 00000000 A1: 00000080 A2: 00008000 A3: 00007fff
A4: fffffffe A5: ffffff00 A6: 0073ff00 A7:*00778004
SR: 0401   PC: 00741c3a ISP: 0077ff80 MSP: 00000000
T1=0 T0=0 S=0 M=0 X=0 N=0 Z=0 V=0 C*1
OK: No exception generated


With https://github.com/retrofun/minimig-mis ... 429d2c4287

Code: Select all

CPUlvl=2, Mask=ffffffff Code=00740000
RTR:
data/68020/rtr/0000.dat. 0...
data/68020/rtr/0001.dat. 226...
All tests complete (total 226).


minimig_mist_20190926.rbf

[retrofun]

MiST 1.3+
Firmware 190820
minimig core 20190922
68000/OCS/Kick 1.3/0.5 Chip/0.5 Fast

0.5 Slow, not 0.5 Fast mem. Sorry.

Hi
The only difference is i use 1mb chip ram and no fast ram

Tested again. Lotus III does not work with 1MB Chip mem for me, too. It needs Slow or Fast mem (e.g. 0.5 Chip + 0.5 Slow mem or 0.5 Chip + 2.0 Fast mem works).

[slingshot]

Cool! Strange this CPU test doesn't even pass in WinUAE 68020 mode (or just used a wrong setup).
Maybe you can fix that pinball game, too?

cputest works fine for me on WinUAE.
I'm aware of the problem with Pinball Illusions but could not track down the problem yet... (also due to lack of time). Any hints, anyone?

Maybe the unimplemented CMP2/CHK2?

[vebxenon]

minimig_mist_20190926.rbf doesn't work to me also. MiST first light flashes, I get a blank screen and nothing happens. I only get the first information Minimig screen then hangs. Using minimig_mist-20190418.rbf everything works ok. I've done everything... I give up

[retrofun]

Cool! Strange this CPU test doesn't even pass in WinUAE 68020 mode (or just used a wrong setup).
Maybe you can fix that pinball game, too?

cputest works fine for me on WinUAE.

...with latest WinUAE 4.3.0 Public Beta, 68020/32Bit address space test. WinUAE 4.2.1 doesn't run the test suite successfully.

[rondc]

I just confirm that Fusion 3.2 Macintosh II emulator WORKS !!!
We're trying disk images and making some experiments in order to boot from floppy, and I'm testing with HD Images for ShapeShifter and Fusion.

Regards

[vebxenon]

minimig_mist_20190926.rbf doesn't work to me also. MiST first light flashes, I get a blank screen and nothing happens. I only get the first information Minimig screen then hangs. Using minimig_mist-20190418.rbf everything works ok. I've done everything... I give up

Anybody?

[retrofun]

Fixed the RTR instruction!

Does NOT fix any issue I currently know (testkey4, Pinball Illusions).

optiroc reported a lot of WHDLoad games that have issues with the TG68K core in the MiSTer subforum: viewtopic.php?f=117&t=32761&start=900#p382726

These WHDLoad games now work with the RTR instruction fix:

Code: Select all

Blasteroids
Deflektor
Epic
F-16 Combat Pilot
Gods
Road Wars

[slingshot]

These WHDLoad games now work with the RTR instruction fix:

Code: Select all

Blasteroids
Deflektor
Epic
F-16 Combat Pilot
Gods
Road Wars

Great! Hope you can find out some more broken instructions and fix them

I'm thinking about moving the repo to mist-devel, I don't know if Rok has any plans to work on it.

Upd.:
https://github.com/mist-devel/minimig-mist

[retrofun]

TG68K: fix ADDX.B, SUBX.B instructions
https://github.com/retrofun/minimig-mis ... 08bb065c06

TG68K: fix CMPM.B instruction
https://github.com/retrofun/minimig-mis ... d5ba0ad8ec

minimig_mist_20190928.rbf

[vebxenon]

TG68K: fix ADDX.B, SUBX.B instructions
https://github.com/retrofun/minimig-mis ... 08bb065c06

TG68K: fix CMPM.B instruction
https://github.com/retrofun/minimig-mis ... d5ba0ad8ec

minimig_mist_20190928.rbf

Wow! Hope it will boot on my MiST

[slingshot]

Does NOT fix any issue I currently know (testkey4, Pinball Illusions).

I did a little debugging with Pinball Illusions, and found out that no trap_illegal signal set inside the TG68K, so probably not an unimplemented instruction. trap_trace is set at $FFDC0004, and this address looks suspicious. Maybe jumping there is already an error.

[retrofun]

TG68K: fix PACK/UNPK pre-decrement for sp register
https://github.com/retrofun/minimig-mis ... 6b4e1453b0

Fixes UNPK, PACK needs additional fixing.

[slingshot]

I did a little debugging with Pinball Illusions, and found out that no trap_illegal signal set inside the TG68K, so probably not an unimplemented instruction. trap_trace is set at $FFDC0004, and this address looks suspicious. Maybe jumping there is already an error.

...seems a really weird thing happen: the last instruction opcode is 4e73, which should be a move.l, then no way the execution should jump to $ffdc0000. Seems somehow the next opcode is fetched from one of the move's operand.

Upd.: wrong, 4exx is a jmp or jsr.
Upd2: it's an RTE. Maybe something with the exception stack frame? This return address looks very wrong.

[vebxenon]

Still with the same problem: core starts, but no kickstart screen. Just a blank screen, so no game, hardisk, anything can be booted

Using the same settings I had before, I've tested four 3.1 roms, no one starts. My Mist.ini is:

[mist]
scandoubler_disable=0 ; set to 1 to run supported cores in 15khz
mouse_boot_mode=0 ; set to 1 if a mouse does not work well
joystick_disable_shortcuts=0 ; set to 1 to remove joystick -> keyboard commands
joystick_ignore_hat=0 ; set to 1 if having issues on gamepads with 'POV hat'
joystick_ignore_osd=0 ; set to 1 to prevent gamepad from controlling OSD
; joystick_remap=0583,2060,1,2,4,8,10,20,20,8,1400,2800,40,80
joystick_remap=0810,E501,1,2,4,8,200,20,10,100,400,40,0,0,0,80
joy_key_map=1000,E2
joy_key_map=2000,E6
joy_key_map=4,51
joy_key_map=8,52

I have a 1.2 MiST, bought in 2015, using last firmware, firmware_190820. Same card, same settings, with same kickstart and minimig_mist-20190418.rbf boots without any problem. What a mistery

[Higgy]

Vebxenon - so you get the initial Minimig info screen with text at the bottom showing memory config etc? Then it goes black so F12 does not work?

I have also a v1.2 and I am sure I have had same issues as you on other Cores so I think our hardware is same (memory chip same).
The only thing I have done is remove a SMD resistor so the audio channels have complete separation rather than blended as the MiST PCB design.
Not that I can see any difference.

I am only using USB devices at the moment. No dtype mouse/ joysticks

[vebxenon]

Vebxenon - so you get the initial Minimig info screen with text at the bottom showing memory config etc? Then it goes black so F12 does not work?

I get the initial Minimig info, F12 works, but no Workbench, no Kickstart screen appears, no hard disk file starts booting. If I use previous stable core, everything works well.

I have also a v1.2 and I am sure I have had same issues as you on other Cores so I think our hardware is same (memory chip same).
The only thing I have done is remove a SMD resistor so the audio channels have complete separation rather than blended as the MiST PCB design.
Not that I can see any difference.

I am only using USB devices at the moment. No dtype mouse/ joysticks

The only change I did to the board is that power supply port was desoldered, and now usb cable is soldered into the board instead of having a MicroUSB port. But every other core works , and as I've said, minimig_mist-20190418.rbf, here I'm playing Amiga core with a MT-32 module https://www.instagram.com/p/B219CT5IdTB/

[Higgy]

So can you boot an .adf and that works?