MMFS Development and Support

for bbc micro/electron hardware, peripherals & programming issues (NOT emulators!)
User avatar
fordp
Posts: 919
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England

Re: MMFS Development and Support

Postby fordp » Sat Jun 17, 2017 5:53 pm

Hi Dave,

1. Yes in the User Port
2. My interface is not a true Turbo MMC but generation one with the only chip on it being a 3.3V Regulator so the T version is not for me
3. I did not do a *FX 200 2

I can try an older version, hold on ;)
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
fordp
Posts: 919
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England

Re: MMFS Development and Support

Postby fordp » Sun Jun 18, 2017 12:05 pm

Sorry, the misses dragged me out to the "Medway in Flames" fireworks.

Version 1.29 does not work on my Master either. It is a shame the RAM on my IFEL cartridge is corrupt as what ever version was on that was working.
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Jun 18, 2017 2:57 pm

fordp wrote:Version 1.29 does not work on my Master either. It is a shame the RAM on my IFEL cartridge is corrupt as what ever version was on that was working.

Can you say a bit more about your setup?

Has MMFS worked in the past on this configuration of hardware?

The Card? error normally means a hardware problem of some description, or using the wrong build of MMFS for the hardware that you have.

Dave

User avatar
fordp
Posts: 919
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England

Re: MMFS Development and Support

Postby fordp » Sun Jun 18, 2017 4:26 pm

I only have one "BBC Micro MMC Interface" which I have been sharing between my BBC B and my Master 128 for a long time.

I have definitely been using MMFS in my BBC B (Stored in a Atmel AT28C256). It works great.

I have used two different ROM images provided by Steve Picton in the past. The second Steve Picton supplied ROM was stored in a BBC Master "IFEL Master ROM/RAM" board.

I thought I had stopped using the Steve Picton ROM in my master a while ago and switched to MMFS on the master too. I am not so sure now however. The BBC Master will no longer boot with the IFEL cartridge any more (corrupted image?).

In the last few days I got my new Ebay spare master and it came with two genuine Acorn "ROM CARTRIDGE (Accepts 2x27128 (32K) and 2x27513 (128K))". So I used some of my stash of EEPROMs to make them in to MMFS cartridges.

This is when I found neither version works.

I will post a picture of my set-up from my phone in a post in a moment.
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
fordp
Posts: 919
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England

Re: MMFS Development and Support

Postby fordp » Sun Jun 18, 2017 4:30 pm

14978033684401232350794.jpg


So my set-up is Pi3 Copro and a home made battery pack and a Steve Picton Turbo MMC ROM in the socket nearest the Acorn mask ROM.
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Jun 18, 2017 4:44 pm

If you give me a couple of minutes, I'll try to duplicate this exact configuration.

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Jun 18, 2017 4:56 pm

Edit: For solution see next post!

OK, so I don't have an internal Pi Tube Direct.

But I've programmed MMFS 1.36 (U/MAMMFS.rom) into the upper 16K of a AT28C256 and put this in an identical ROM cartridge, so it appears as ROM 0.

It's working fine.

The only software difference is that you also have the Steve Picton supplied DFS in ROM 7. Can you just *UNPLUG that and power cycle the machine?

If that doesn't help, this must be slightly obscure a hardware issue with this Master, right?

Am I right in thinking:
1. The error from MMFS is Card?
2. The MMC interface works fine on the model B?
3. Only MMFS is failing, i.e. using the older Steve Picton supplied DFS in ROM 7 on the Master works?

Dave
Last edited by hoglet on Sun Jun 18, 2017 5:07 pm, edited 1 time in total.

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Jun 18, 2017 5:01 pm

OK, I've spotted the problem from your HELP message.

You are using the wrong build.

You need to be using U/MAMMFS (which is for a normal user port connected MMC card).

I think you are actually using M/MAMMFS (which is for a memory mapped SPI based interface).

See this page for the release structure:
https://github.com/hoglet67/MMFS/wiki/Release-structure

Dave

User avatar
fordp
Posts: 919
Joined: Sun Feb 12, 2012 9:08 pm
Location: Kent, England

Re: MMFS Development and Support

Postby fordp » Sun Jun 18, 2017 6:34 pm

Thanks Dave.

I have just reprogrammed my EEPROM and now it is all working.

OK, In post mortem this is what happened:

I read the readme and established that a file called "MAMMFS.ROM" was needed. I downloaded the release and assumed the M directory was for the Master. I found a file with the exact name I was looking for ....

I am sorry for the wild goose chase.
FordP (Simon Ellwood)
Time is an illusion. Lunchtime, doubly so!

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Sun Aug 20, 2017 12:08 pm

Just updated my rom from 1.32 to 1.36,
now the *MMFS gives 'Bad command' :?

This is what is in the IMTOD / IMTOM progs for transfer, so hence found bug.

The SWMMFS+ rom is now fine in my B+64k, never checked the *MMFS though....
ImageImageImage

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 12:13 pm

I'll take a look at what changed....

Is this on the Model B?

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 12:30 pm

CMcDougall wrote:Just updated my rom from 1.32 to 1.36,
now the *MMFS gives 'Bad command' :?

I just tried this myself (using 1.36 in a Model B, i.e. the U/MMFS.rom build). I can't reproduce the issue, *MMFS seems to work fine.

Can you give a few more details?

Which build are you using? On which machine?

Dave

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Sun Aug 20, 2017 3:46 pm

^using U/mmfs on a B with 8271 , dfs1.21 & ADT2.

Fine when I put 1.32 back in :?
ImageImageImage

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 4:41 pm

CMcDougall wrote:^using U/mmfs on a B with 8271 , dfs1.21 & ADT2.

Fine when I put 1.32 back in :?

Are you loading into sideways RAM, or burning an EPROM?

I'm wondering if your copy of 1.36 has somehow got corrupted.

Does removing ADT2 and DFS 1.21 make any difference?

Does *DISC work?

Does *HELP work?

Dave

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Sun Aug 20, 2017 4:50 pm

Just figured it out, when I change DISC /DISK to DMMC /CARD (so can run both FSs)
I changed the disc & mmfs instead #-o

wondered why only *DISK got me back to floppy!

Oww well, another 15 mins for another blower :oops:

Luckily never burnt any more, cheers Dave.
ImageImageImage

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 5:02 pm

Ah, that would explain it.... :lol:

Dave

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Sun Aug 20, 2017 5:21 pm

Thought I was not seeing things, (github comments)
So does the *OPT5,1 do anything interesting (like *DTRAP on dcentre?)
Just tried the Elixir game again, with this opt51 & still says 'Disc not found , Bad prog'
Was using original MMFS with disc/k
ImageImageImage

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 5:40 pm

CMcDougall wrote:Thought I was not seeing things, (github comments)

Linky? So I know what your are referring to...
CMcDougall wrote:So does the *OPT5,1 do anything interesting (like *DTRAP on dcentre?)

As far as I know, all *OPT 5,1 does is to cause MMFS to ignore *DISK and *DISC (i.e. they just get passed on to the next ROM).

This isn't anything new, it's code there was in there since MMFS 1.00.
CMcDougall wrote:Just tried the Elixir game again, with this opt51 & still says 'Disc not found , Bad prog'
Was using original MMFS with disc/k

Are you saying Elixir worked with 1.32 and doesn't with 1.36??? I'm now confused :?

Dave

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Sun Aug 20, 2017 5:56 pm

https://github.com/hoglet67/MMFS/issues
linky /\
Elixir has never worked , but does with SuperMMC & the original 0.90. Men signed it in this thread page 2 maybe :?
It's the only one , but BITD mine would revert back to tape so had to load 4 blocks from tape, then it would continue the last 4B blocks from disc :lol:

it must dribble all over E00 space, so a *DISC , *DR. 0 80 to get it back like White Light Beta demo
ImageImageImage

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 8:56 pm

CMcDougall wrote:Elixir has never worked , but does with SuperMMC & the original 0.90. Men signed it in this thread page 2 maybe :?
It's the only one , but BITD mine would revert back to tape so had to load 4 blocks from tape, then it would continue the last 4B blocks from disc :lol:

it must dribble all over E00 space, so a *DISC , *DR. 0 80 to get it back like White Light Beta demo

The problem with Elixir is that ELIX2 loads over the top of MMFS's private workspace (normally page &1700), and then goes on to try to load ELIX3 and ELIX4. At this point all bets are off. The reason that it works on older DFSs is that they are more tolerant of workspace corruption (DFS 2.x added workspace consistency checking, giving the dreaded Bad Sum error).

The only way I have found to run Elixir from MMFS on the Model B is to load &E00 version of MMFS (SWMMFS) into sideways RAM and then load Elixir using this. I've just tried this and it works. Sideways RAM must be left write enabled for SWMMFS to work.

Elixir should work fine on a Master (not actually tried this myself...)

Dave

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Sun Aug 20, 2017 9:38 pm

I've been tracing Elixir a bit more with ICE T65.

This is what seems to happen:

1. MMFS's private workspace is normally at &1700. The ForceReset flag is at &17D3. Normal value for this is &FF. Any value with bit 7 = 0 will force a reset of the drive table on the next file system restart (e.g. when BREAK is pressed, or *DISC or *MMFS).

2. Elix2 loads to &1100, which sets ForceReset to &00.

3. Elix2 runs and executes the following via OSCLI:

Code: Select all

*DISC
*DIR :0.$
*L.Elix3
*/Elix4


4. DISC is handled by MMFS, but because ForceReset is now &00 it resets the drive table, causing the default images to be mounted. Hence the Not Found error.

5. At some point (not sure when exactly), ELIX2 also writes to &10E0-&10EF, which is MMFS's drive table where the currently mounted disk numbers are stored. This region is protected by a CRC7 checksum.

So if Elix2 is modified to remove *DISC (see here), then *L.Elix3 still fails, but this time with a Bad Sum error.

Thoughts:

1. ForceReset could be made more robust, needing a specific value (e.g. &DB) to be found before a reset is forced.

2. I wonder where DFS 0.9 based MMC file systems store the drive table?

Dave

User avatar
sweh
Posts: 1847
Joined: Sat Mar 10, 2012 12:05 pm
Location: New York, New York
Contact:

Re: MMFS Development and Support

Postby sweh » Mon Aug 21, 2017 12:32 am

hoglet wrote:2. I wonder where DFS 0.9 based MMC file systems store the drive table?

Probably page D, if page 6 of https://swhs.home.xs4all.nl/bbc/mmbeeb/ ... DUTILS.pdf is anything to go by.
Rgds
Stephen

duikkie
Posts: 2689
Joined: Fri Feb 07, 2014 3:28 pm

Re: MMFS Development and Support

Postby duikkie » Mon Aug 21, 2017 6:08 am

maybe 1 file solution is an option :) ?
if i have time where is the elix . ssd , (if i not fixed it already?)

User avatar
hoglet
Posts: 6605
Joined: Sat Oct 13, 2012 6:21 pm
Location: Bristol

Re: MMFS Development and Support

Postby hoglet » Mon Aug 21, 2017 6:19 am

duikkie wrote:maybe 1 file solution is an option :) ?
if i have time where is the elix . ssd , (if i not fixed it already?)

It's on disk 215.

I think it already works with SmartSPI.

Dave

duikkie
Posts: 2689
Joined: Fri Feb 07, 2014 3:28 pm

Re: MMFS Development and Support

Postby duikkie » Mon Aug 21, 2017 7:39 am

off course is works with SMART-SPI :D :evil:
but if maybe all in one file it will work with the other file system i will not name here :lol:


hoglet wrote:
duikkie wrote:maybe 1 file solution is an option :) ?
if i have time where is the elix . ssd , (if i not fixed it already?)

It's on disk 215.

I think it already works with SmartSPI.

Dave

duikkie
Posts: 2689
Joined: Fri Feb 07, 2014 3:28 pm

Re: MMFS Development and Support

Postby duikkie » Mon Aug 21, 2017 8:56 pm

schame schame on me. elix is not fixed also not for smart-spi !!
there is still *disc *dir 0:$ in file
a lot of stuff moving :)
but no discusion here it is a software problem
if fixed i will report it in software :)
maybe 2 files , because file 1 is instructions. 2 the game

duikkie wrote:off course is works with SMART-SPI :D :evil:
but if maybe all in one file it will work with the other file system i will not name here :lol:


hoglet wrote:
duikkie wrote:maybe 1 file solution is an option :) ?
if i have time where is the elix . ssd , (if i not fixed it already?)

It's on disk 215.

I think it already works with SmartSPI.

Dave

duikkie
Posts: 2689
Joined: Fri Feb 07, 2014 3:28 pm

Re: MMFS Development and Support

Postby duikkie » Tue Aug 22, 2017 8:48 am

sorry but elixir , is terrible on din 215 , a normal tape version ? , ssd version is all over swapping memory lost of time
so it will never work around &1000-&1100 , i cound so 4 swapings there

User avatar
CMcDougall
Posts: 5611
Joined: Wed Feb 02, 2005 3:13 pm
Location: Shadow in a Valley of Scotland

Re: MMFS Development and Support

Postby CMcDougall » Wed Aug 23, 2017 5:57 pm

http://www.ebay.co.uk/itm/BBC-B-Micro-S ... Ciid%253A1
^Duke, do you know about this? someone selling your work (priceless £~),
MMC interface 33p, EPROM £5, board ?£2 & £2 postage = £9.33 -£16 = £6.77 profit :shock: (seller = maisie899)
duikkie wrote:sorry but elixir , is terrible on din 215 , a normal tape version ?

all versions here : http://www.stairwaytohell.com/filesearc ... rch=Search
ImageImageImage

duikkie
Posts: 2689
Joined: Fri Feb 07, 2014 3:28 pm

Re: MMFS Development and Support

Postby duikkie » Wed Aug 23, 2017 6:44 pm

smart-spi rom is free software , i have no problem with it. someone did ask about it some time ago.
i am working on an other rom called user dos. and my ch375b china thing :) for tube.

mmc at user port sooooooooooooo 2016 :)


CMcDougall wrote:http://www.ebay.co.uk/itm/BBC-B-Micro-SD-Card-drive-SMART-SPI-ROM/202026549923?_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D20160908105057%26meid%3D08d51185d5b84dd48ad0134f50425c52%26pid%3D100675%26rk%3D2%26rkt%3D15%26sd%3D292221128855&_trksid=p2481888.c100675.m4236&_trkparms=pageci%253A672441b1-882b-11e7-8a59-74dbd1809b70%257Cparentrq%253A103691a215e0a861b3424749fffd01d4%257Ciid%253A1
^Duke, do you know about this? someone selling your work (priceless £~),
MMC interface 33p, EPROM £5, board ?£2 & £2 postage = £9.33 -£16 = £6.77 profit :shock: (seller = maisie899)
duikkie wrote:sorry but elixir , is terrible on din 215 , a normal tape version ?

all versions here : http://www.stairwaytohell.com/filesearc ... rch=Search

KenLowe
Posts: 157
Joined: Mon Oct 18, 2004 4:35 pm
Location: Scotland

Re: MMFS Development and Support

Postby KenLowe » Tue Nov 14, 2017 9:50 pm

I'm running SWMMFS in one of my SWRAM banks and generally it works very well. However, I've noticed that immediately after a BREAK, the *FREE command doesn't work. If I do a *CAT first, it then works. Is this a problem with my setup, or can others replicate this?


Return to “hardware”

Who is online

Users browsing this forum: No registered users and 7 guests