News:

Printed Amstrad Addict magazine announced, check it out here!

Main Menu
avatar_freemac

FPGAmstrad on MiST-board

Started by freemac, 03:04, 21 March 15

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Gryzor

Whoa that's a lot of games tested... Not perfect at 74%, but getting there!

Bryce

Quote from: Gryzor on 09:50, 16 November 15
Whoa that's a lot of games tested... Not perfect at 74%, but getting there!

A success percentage doesn't give any indication of how many games were actually tested?? It could be 74% of 7 games that they tested.

Bryce.

Gryzor

Quote from: Bryce on 10:23, 16 November 15
A success percentage doesn't give any indication of how many games were actually tested?? It could be 74% of 7 games that they tested.

Bryce.


Check the link :)


What I'd like to know is how games were actually "tested". I met this issue with the ST games on MiST, where testing was a bitch and went quite slowly. It's not enough to say a 'yes' or 'no', since something can be playable with glitches on the menu, it may crash at some specific point, it may involve corrupted gfx at the end of a level etc etc...

Bryce

Ok, that is a lot, but I would also question how far they were tested. The same thing happens with Amiga and which version of Kickstart you use. Some games will start and seem to work until they get to a certain point.

Bryce.

Gryzor

That's what I said :) . Same thing with the ST, each time I would find a game that didn't work I would cycle through all configurations - ST/STe, TOS version, memory...

freemac

Quote from: Gryzor on 11:16, 16 November 15
That's what I said :) . Same thing with the ST, each time I would find a game that didn't work I would cycle through all configurations - ST/STe, TOS version, memory...


Tests were done using minimum configuration : original UK ROM only :)
I remark that Arkanoid 2, and Turbo Tortoise doesn't like ROMs added...


I test them quickly but I try to do 1 action in first level of each game.
Tests here are first about screen+sound+controllers and fun (and detect regressions in next tests)

freemac

Quote from: Bryce on 10:23, 16 November 15
A success percentage doesn't give any indication of how many games were actually tested?? It could be 74% of 7 games that they tested.

Bryce.
1016 games passed OK on a set of 1375 games  :)

freemac


Gryzor

Quote from: freemac on 13:38, 17 November 15

Here the full SDCARD testbench result for r004.7.2 (core + validated dsk)
http://github.com/renaudhelias/CoreAmstrad/blob/master/MarkusGameBaseTestbench/MarkusGB_test_r004.7.2.zip


Oh, that's really useful!! How are the files split among the dirs? Any special reasoning?

freemac


Quote from: Gryzor on 11:47, 18 November 15
Oh, that's really useful!! How are the files split among the dirs? Any special reasoning?


Game folder contains games running in 1011 VRAM (&8000-&FFFF)
Game 0011 folder contains games running in 0011 VRAM (&0000-&3FFF and &8000-&FFFF)
Game 0111 folder contains games running in 0111 VRAM (&4000-&7FFF and &C000-&FFFF)


1011 is selected by defaut in OSD menu (10 11)
0011 and 0111 can be manually selected in OSD menu (00 11, 01 11)

Gryzor

Ah thanks man, that's good to know...

tezet

Hello,

just got a question regarding the amstrad core: It refuses to work on my MiST! I flashed the 150910-firmware to the device (as recommended since today) and used the files contained in Markus testbench. All I get is an orange screen with random dots (and a very promising clear blue screen when switching VRAM). My MiST is a V1.3 with MIDI-Exts... Oh yes, the orange  LED in the middle is on, the on screen-menu works, my SDHC-Card is a 16GB Samsung Class 10  partitioned to 4GB and formatted with FAT32. I also tried other SD-Cards but some just produce a grey screen (which also does not seem quite right). Is the Amstrad core really this picky about SD-Cards or am I missing something obvious here? Trying it for some time now, the Amiga-Core is working fine.

Thanks for any help and best regards,

Tobi

freemac


VRAM default position in OSD menu are :
- LowerVRAM 10
- UpperVRAM 11

With this default position you can run all Games in "Games" folder (that folder shall be called "Games 1011" in fact)

When it turns blue, you can type letters, or quick reset using "page up" key.

Normaly SDCARD has just to be formated as FAT32, only with one partition, generaly using 4096bytes blocs (in PC, do create a simple text file containing just 1 letter, save it, and show the file size...)

I never tested using partitioned SDCARD...

You have to defrag your sdcard (or else dsk will have problems being inserted)

I'm testing currently last firmware (firmware_151118.upg), it seems really cool (faster to load dsk !)

tezet

Hi Freemac,

thanks for your reply! Sadly I was not able to get to the ready-prompt, even after doing a full optimization with UltraDefrag on the card (Cluster-Size is 4KiB with FAT32). I'm starting to believe the core is not reading the roms correctly. Am I right assuming rom loading is done there?

Just to be clear: All I see is a complete orange screen with grey borders top and bottom (ok, sometimes random bytes are set). After switching VRAM through the (working) OSD the screen turns to royal Amstrad-Blue (but clear, nothing else on it besides the grey borders, no ready-prompt and no nifty beep when pressing delete).

For me this indicates I have to get another SD-Card as I only have smaller or bigger ones than 4GB. Perhaps some kind of compatibility-list could help to chose one (they are becoming quite rare, bigger is better). Perhaps someone has some experience with this?

But even if I'm disappointed not being able to play Rick Dangerous in cpc colors I really appreciate the hard work this must have been! I will stay on it and try again with new cards - Amiga core cannot satisfy me ;)

tezet

Just realized there is the ability to do ram dumps... Perhaps someone could post a dump of a freshly initialized core so I could diff it with mine? There shouldn't be big differences or am I wrong about this? Thanks in advance!

freemac

#65

Quote from: tezet on 08:30, 05 December 15
Just realized there is the ability to do ram dumps... Perhaps someone could post a dump of a freshly initialized core so I could diff it with mine? There shouldn't be big differences or am I wrong about this? Thanks in advance!
Hi tezet,


I did test DUMP yesterday (both button pressed (not reset) after boot), but file result is too small (only 2MB)... last version of CoreAmstrad uses RAM for : 4MB ROM + 576KB RAM + 16kB LowerROM. Content of this actual 2MB DUMP.DMP is : .eZZ file content, garbage RAM (indeterminist not used), ghost RAM (indeterminist used in past)... and no more ^^'. So you cannot compare it with an emulator snapshoot.


Documentation is by here :
CoreDocAmstrad · mist-devel/mist-board Wiki · GitHub (user doc)
FPGAmstrad - CPCWiki (development strategy/TODO list/bugs)


I don't have time about creating/maintaining a more advanced end-user wiki page, if a fan want to make a FAQ like page, that shall be cool (explaining use of STarKos/use of ROMDOS/use of Protext/use of Maxam/use of SymbOS/use of LowerVRAM UpperVRAM jumper versus demos/ROM keyboard map I18n).
But I'll do my possible to still haunt this forum thread :)


To know :
- There is no detection about missing ROMs at root folder of SDCARD, if not ROMs are found at root folder (beside core), no ROMs are inserted and boot is inderteminist (as LowerROM area is not filled with 1 as UpperROM area)
- If no .dsk file is present at root of SDCARD, then no defrag is needed in order to show the "Ready" welcome message of Amstrad (as in MarkusGameBaseTestbench zip)
- The orange led is one when all ROM has finished loading at start and z80 is turned on. You can then do the sound test pressing the delete key.
- initial position of OSD is : LowerVRAM 10 UpperVRAM 11 Drive A, you can save OSD preferences in menu, that does create a file in sdcard, you can defrag after it creation if you want (a know bug in very old firmwares)

freemac


CoreAmstrad (MiST-board version of FPGAmstrad) doc revisited :
CoreDocAmstrad · mist-devel/mist-board Wiki · GitHub
ROMs/ folder does now contain only the minimum ROMs set for having an Amstrad Ready  :picard:


FPGAmstrad (dev board version) doc contain the TODO-list and exploits :
FPGAmstrad - CPCWiki

tezet

Hi Freemac,

I just ordered a bunch of different SD(HC)-Cards. I will try these and keep you informed! My idea about a RAM-Dump was to diff a dump of a freshly initialized running core (at ready-prompt) with mine (which is not running at a prompt). Perhaps I could spot some differences in the loaded roms. But thats just to insure my MiST is not damaged somehow... Would be nice to know more about the DUMP-Capabilities and the exact definition of this file. Perhaps I could do some documentation but I won't promise right now. It is really frustrating seeing these videos about a running core, following every step and not getting it to run (only this blue or orange screen) ;) And, to my utter surprise, no one else seems to have this problem, wtf is wrong with me?

Thanks & best regards!

freemac


Hi Tezet,


To install last Till firmware (HowToInstallTheFirmware · mist-devel/mist-board Wiki · GitHub)
* Download last firmware in .upg format : mist-binaries/firmware at master · mist-devel/mist-binaries · GitHub
* Rename firmware_151118.upg into firmware.upg, and then copy it into SDCARD
* have some core.rbf in your SDCARD (or else MiST-board shall not start (and then you don't access OSD menu using F12)
* Do use OSD menu (F12), go to second page, choose update firmware, check is both versions (sdcard one versus installed one) are differents, if differents, do update action.
* After update, it does auto-reboot, then you can use OSD menu and check firmware version installed.


To prepare minimum CoreAmstrad SDCARD (CoreDocAmstrad · mist-devel/mist-board Wiki · GitHub)
* Do format your SDCARD (as it, no more amstrad.ini file, so OSD menu come back to default values (not corrupted)
* Download last core in .rbf format : mist-binaries/cores/amstrad at master · mist-devel/mist-binaries · GitHub
* Rename amstrad_151013_r004.7.2.rbf into core.rbf, and then copy it into SDCARD
* Download the 3 Amstrad ROMs from here : mist-binaries/cores/amstrad/ROMs at master · mist-devel/mist-binaries · GitHub
* Copy them into SDCARD, resulting in a SDCARD with only 4 files, and no folders.
* At boot, you shall see the Amstrad "Ready" on screen.


Good luck !

freemac

* Btw, isn't it a Windows problem ? like not ejecting correctly SDCARD using right-click on drive>eject ?


Spoiler: ShowHide
ParaDOS 1.2 tested ok on CoreAmstrad, using ROMDOS-D2 dsk format in drive A, transfert between 720KB dsk and 128KB dsk OK, tested also with STarKos, lovely use of RAM ram expansion  :)

tezet

Would be nice if the problem boils down to windows... But unfortunately I (nearly) always unmount the cards correctly and follow the steps you have described above (I even used SAM-BA to update)... Up to now I have tested about ten different card-modells and none of it seems to work with the Amstrad-Core.

I contacted Till today and he assured me, my MiST should be in perfect working order as the Amiga-Core is running well. My guess, it could be something related to the MIDI-Input seems also pointless. Now I'm quite helpless as I do not have any other idea. Diffing the dumps seems to be the next logical step (I'm hoping to test the cards on a non-midi-MiST but I fear this won't happen untill next year).

Optimus

Just to say, I have the same problem with Tezet.


I just received the Mist with MIDI out.
Using a 4GB SD card that is not SDHC, format as Fat32.
Tried to do exactly the same process, update the firmware, format the card again, copy those exact latest files, nothing. Just the red screen which changes to blue when I change the roms.


The board works just fine with the AtariST core.

freemac

I invite you to create an issue in Issues · mist-devel/mist-binaries · GitHub


I don't have this MIDI version.


A cool test shall be coming back into some old firmware versions.

tezet

Thanks Optimus, in a way these are good news to me - I thought I'm the only one not getting this to work...

I contacted Till Harbaum last week just to make sure, it has nothing to do with MIDI. He tested the Amstrad-Core on his MIDI-Board and it worked fine with the latest firmware (he even used a 16GB SD-Card)... So now I only got one explanation: WE ARE CURSED! And I only made little fun out of the name MiST (well, in german this is so very tempting) - Now I will use the MiST to emulate Amiga and search for a proper AmigaOS-CPC-Emulator (I wanted it to do the other way round, just have seen a youtube-video) ;)

But ok, enough disappointment, my next steps will be:

- ask Lotharek if something has been changed in the production-process
- test some cards on a non-Midi-MiST a friend of mine bought (will be done in January due to holidays)
- Send a SD-Card to Till so he can prepare and test it with his working MiST (or better: send him my MiST to check on it, but I don't know, if he has the time to do this).




freemac

#74
[attachimg=2]
Perhaps you can test this alpha r005 version :)
scandoubler from Tobi, and original TV signal if you select TV in output mode (I can't test this last signal, I don't have any old TV at home ^^')

Powered by SMFPacks Menu Editor Mod