MiSTer FPGA Forum

Hi all
With a few games on NeoGeo I get a 'Not enough memory' message even though I have the full 128bm SDRAM. I have run a memory test on it using the memory test core and it's come back as all good so I'm a little confused. I get the message and corrupted graphics on Garou and KOF for example.

Anybody got any ideas as to what could be wrong?

Personally, I suspect a bad ROM set...can't say more because I don't know where you got yours, but the Darksoft pack is absolutely the way to go.

Thanks rhester72, I'll do some digging and see if I can work out if there's something up with the rom set. I thought I had got the Darksoft pack (from archive.org) but then it's quite probable I got the wrong one, I was still learning how to set up MiSTer at the time.

My Garou CRC32 is 25D2E39E, if it helps.

rhester72 wrote: ↑Fri Jul 17, 2020 6:23 pm My Garou CRC32 is 25D2E39E, if it helps.

Thanks, I'll check that out tomorrow

This happened to me too. The problem is using the .xml file with the .neo ROM set. (at least it was for me). Deleting the .xml file fixes the issue - BUT I think an update script replaces the .xml file after an update - so what I did was....

Create a blank romsets.xml file
Make it read only.

This means that the update script cannot overwrite it in the future.

I'm not sure that's the problem either. Mine works fine, and:

Are you using the Darksoft set @rhester72? The romsets.xml works fine with that set. I only had issues when using the .NEO ROM container.

I was pulling my hair out, emailing the shop where I purchased the RAM card from, until I started with a fresh card and worked backwards. In my case it was 100% reproducible. (Remove the .xml and works fine, add it and it corrupts the graphics and gives a warning. Was a while later I realised one of the update scripts kept pulling down a fresh copy. Hence making a dummy file set to read only finally stopped it once and for all.

rhester72 wrote: ↑Fri Jul 17, 2020 6:23 pm My Garou CRC32 is 25D2E39E, if it helps.

I've just had a look and realised that, of course, the NeoGeo roms are no zipped so I'm not sure what file to check the CRC32 of. I did try quickly zipping the directory and seeing what CRC32 it had and it didn't match, but that's not entirely surprising as it could be that it used a different compression amount etc. Which file are you checking the CRC of rhester72?

I tried removing romsets.xml but that completely broke it - there were no games to select, only the rom directories which appeared to be empty...

No, CRC32 isn't affected by compression (assuming you're looking at the calculated CRC32 inside the ZIP) - it's very likely yours is just different.

Going to PM you.

Just wanted to leave an update so that this doesn't become one of those unhelpful threads that I happen across online regularly.

in short - this is still not working for me. I have tried:

• I known working .neo file for Garou
• A complete "reinstall" of the core using the update script (I had previously set up the core manually)

I'm starting to run out of ideas, but I might try different bios files some time. Failing that, maybe I need to test my ram expansion again

Will come back with an update if I get any further

Using unibios4? Correctly named? Other bios files correctly named? There are 2 or 3 variations of filenames out there on the internet's, worth checking that the bios names are correct. You can find the correct names on the git page for the neogeo core. I had this issue, even though the bios files were fine on other emulators.

In addition, several bios files must be placed in the NEOGEO folder for the core to function properly:

000-lo.lo
sfix.sfix
sp-s2.sp1 (MVS)
neo-epo.sp1 (AES)
uni-bios.rom

Sometimes these files may be named slightly differently depending on where they are obtained, but they must be renamed to match the filenames above to work with MiSTer. You may choose between using original system BIOS (sp-s2.sp1/neo-epo.sp1) and uni-bios.rom. Using uni-bios is recommended, and can be obtained here.

Also, try a few versions back of the neogeo core. Ensure you're cold booting between ROM selections on large ROMs.

Make sure you've only got a single directory/location for ROMs on the sd - people have had issues with multiple locations and ROM files.

Thanks Gymcrash, I'll try all of those things soon

I had the same problem aToyRobot. It started when I changed the bootcore in the .ini file of my MiSTer to something else than the Menu core. My setup was booting the SNES core, then when I changed the core to NeoGeo there was this message on every game and there was corrupted graphics. I just commented the bootcore option and it works now.

Seems like the sdram_sz() is reporting the wrong sdram size (0) when not initialized via the Menu core.
https://github.com/MiSTer-devel/Main_Mi ... loader.cpp line 1195

artfakt wrote: ↑Mon Aug 03, 2020 2:20 pm I had the same problem aToyRobot. It started when I changed the bootcore in the .ini file of my MiSTer to something else than the Menu core. My setup was booting the SNES core, then when I changed the core to NeoGeo there was this message on every game and there was corrupted graphics. I just commented the bootcore option and it works now.

Seems like the sdram_sz() is reporting the wrong sdram size (0) when not initialized via the Menu core.
https://github.com/MiSTer-devel/Main_Mi ... loader.cpp line 1195

This is a massive revelation, because I was really running out of ideas and was beginning to wonder if my RAM was faulty in a weird way that the tester wasn't detecting! I have my bootcore set to Minimig, so immediately this looks like a probable cause to me. I'm going to try removing the bootcore option tonight after work and see what happens, I reckon you've cracked it... Fingers crossed

artfakt wrote: ↑Mon Aug 03, 2020 2:20 pm I had the same problem aToyRobot. It started when I changed the bootcore in the .ini file of my MiSTer to something else than the Menu core. My setup was booting the SNES core, then when I changed the core to NeoGeo there was this message on every game and there was corrupted graphics. I just commented the bootcore option and it works now.

Seems like the sdram_sz() is reporting the wrong sdram size (0) when not initialized via the Menu core.
https://github.com/MiSTer-devel/Main_Mi ... loader.cpp line 1195

Thanks artfakt, that was the exact issue. A reboot from the menu before loading the NeoGeo core fixes it and is an acceptable workaround for me.

I have raised an issue on github and referenced back to this post https://github.com/MiSTer-devel/NeoGeo_ ... issues/101

Well done, glad there was a resolution!

It's not an issue. It's a feature.
You have to start Menu core at least briefly.

Hi Sorgelig

I'm not sure I understand, because this sounds like an issue to me. On github you said "It's misusing the MiSTer" but all I have done is used a documented feature (bootcore) and it has broken another part of the system without warning.

At very least I feel like the documentation on MiSTer.ini should be updated to warn the user that using bootcore can break other cores

Before you can select any script, like the one where you could setup bootcore, it warns you that it could stop your system from functioning, to be fair.

You don't have to use a script to set the bootcore. I set my bootcore by editing the ini file on the SD card.

I'm just trying to help to prevent people from falling into this trap in the future that's all

I know this is an old thread but I just got my mister a few days ago and ran into the same error loading Garou.

I did not make any change to bootcore...

Rebooting didn't solve it. I had to unplug mister and plug it back to get rid of the error.

I am using a .neo set

Any idea what could cause the issue?

PS: I had ran the games menu script and added neo geo to it. Could that have been the problem?