Welcome, Guest. Please login or register.
Amiga Kit Amiga Store Iridium Banner AMIStore App Store A1200/A600 4xIDE Interface

AuthorTopic: Weird CF card behavior with 3.1.4.1 - some apps not loading/failing silently.  (Read 531 times)

0 Members and 1 Guest are viewing this topic.

Offline hishamk

The gist of the issue is that some apps, like Calculator in the Tools directory, do not load. There is an "attempt" as per the message on Workbench, but then nothing happens.

My set up is as follows:

Amiga 3000@25MHz, 2mb+8mb, stock. CF-IDE adapter attached to Acard 7720UW bridge which in turn is attached to the SCSI port. Using a Transcend 133x 4GB card. AmigaOS partition is 1gb, the other 3gb is Work.

I've installed 3.1.4 without any issues as well as the .1 update. The system boots just fine. The only thing is that some apps refuse to run, including Calculator.

The interesting bit is that I've got another setup, an Amiga 1000 w/ a Classic 520 and a SanDisk 8gb card. 3.1.4.1 works perfectly fine on it, including Calculator :)

I plugged in the 3k's Transcend into the 520 and booted the A1k up. The same behavior happened. The Calculator refuses to show up.

I then copied the Calculator file from the 3k's Transcend onto a floppy disk and rebooted the A1k with it's usual SanDisk 8gb card. The calculator loaded up fine from the floppy disk!

I took that disk back to my Amiga 3000 and plugged back the Transcend as well to boot it up.

Clicking on the Calculator icon on the floppy resulted in nothing, just as if I clicked on it from the Transcend. So no joy there.

I then booted my 3k with the 1k's SanDisk card. Lo and behold, the Calculator on the floppy and on the SanDisk loads up just fine!

So what's going on here?

It seems to not be an issue with the SCSI/drive controllers, since the behavior for the Transcend card happens on both the A1k and A3k.

I've also installed MMULib to see if that had anything to do with it as I was writing this and the same behavior persists.

Also, when attempting to load the Calculator in CLI, nothing happens, no error message, nothing.
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA
 

Offline TribbleSmasher

Sounds like the devilish MaxTransfer problem...
 

Offline hishamk

FWIW I did change MaxTransfer to 0x1fe00 yet the issue persists.
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA
 

Offline TribbleSmasher

Can you run SnoopDos with all DOS functions activated while attempting to run several of those refusing apps?

Offline hishamk

Can you run SnoopDos with all DOS functions activated while attempting to run several of those refusing apps?

Interesting, when I run Calculator; SnoopDos reports:

ramlib   Load     mathieeedoubbas.library   FAIL

Moreover, when I tried using SGrab to get a screenshot, it complained that it failed to open mathieeedoubbas.library, version 37.

Hmm, could it have been something corrupted during the install? Would be strange since I didn't get any disk errors.

I do have the the library in LIBS. When I run version on it, it says cannot find version information... The size is 3596 bytes, the same as on the Workbench install disk.

I copied over the one from the Workbench install disk and did a cold reboot. Same thing,

ramlib   Load     mathieeedoubbas.library   FAIL

Could this be an issue related to the fact that the 3k has an FPU and perhaps some install configuration was incorrectly set by the installer script?


« Last Edit: April 12, 2020, 02:12:42 PM by hishamk »
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA
 

Offline TribbleSmasher

When i look here in the thread THOR has explained, they fixed something in those math libs so one could assume you should have libs with a higher version number present; so the v37 load should succeed.

Math stuff. Yes, there are new math libraries (again)....

mathieeedoubbas and singbas only changed in so far as they configure the FPU slightly different, namely "extended precision, round to zero". The results of the computation are still the same as the last operation in each call to the math code still implies a rounding to double or single precision, but the advantage of the different rounding mode is that a task can open both libraries simultaneously without a conflict, which was not possible before. ...

Is the lib read protected?

 

Offline hishamk

When i look here in the thread THOR has explained, they fixed something in those math libs so one could assume you should have libs with a higher version number present; so the v37 load should succeed.

Math stuff. Yes, there are new math libraries (again)....

mathieeedoubbas and singbas only changed in so far as they configure the FPU slightly different, namely "extended precision, round to zero". The results of the computation are still the same as the last operation in each call to the math code still implies a rounding to double or single precision, but the advantage of the different rounding mode is that a task can open both libraries simultaneously without a conflict, which was not possible before. ...

Is the lib read protected?

Yup, the bits are set exactly the same as on the working installs (and the 3.1.4 Workbench floppy): RWD.
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA
 

Offline hishamk

So the plot thickens.

I copied over the 68882 version of mathieeedoubbas.library from http://aminet.net/package/util/libs/FMath406.lha. It's version v40.6, the same as that supplied with 3.1.4, but the file size is much smaller (896 vs 3956 bytes - perhaps because it's just compiled for 68882 machines).

Calculator worked just fine.

So, I'm not really sure what's going on here.
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA
 

Offline duga

Try disabling/enable Direct SCSI Transfer in HDToolBox. Perhaps do a backup first.
 

Offline kolla

So, I'm not really sure what's going on here.

Bugs, bugs and more bugs?
I have some 020+888 systems, but none with OS 3.1.4.1, as the hardware in question is not supported.

I suggest you file a bug report with @Gulliver
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC
---
A3000/060CSPPC+CVPPC/128MB + 256MB BigRAM/Deneb USB
A4000/CS060/Mediator4000Di/Voodoo5/128MB
A1200/Blz1260/IndyAGA/192MB
A1200/Blz1260/64MB
A1200/Blz1230III/32MB
A1200/ACA1221
A600/V600v2/Subway USB
A600/Apollo630/32MB
CD32/SX32/32MB/Plipbox
CD32/TF328
A500/V500v2
A500/MTec520
CDTV
MiSTer, MiST, FleaFPGAs and original Minimig
Peg1, SAM460 and Mac minis with MorphOS
 

Offline hishamk

I ended up reinstalling on a 2gb SanDisk instead of 4gb Transcend and it's been fine. I'm also using a MaxTransfer of 0x1FE00.

Should you wonder, I've previously used the Transcend with Tom Thul's A1000's IDE adapter and it's always worked fine.

Now on to the 3000's fan. That thing is LOUD!

« Last Edit: April 15, 2020, 08:34:40 AM by hishamk »
2x A1000, 2x A2000, 1x A3000, 2x A1200, 2x A500, 1x CDTV, 1x CD32, 2x Pegasos II, 1x EFIKA