exxos's Atari, Amiga & retro forum

mikro wrote: 24 Sep 2025 13:48

stephen_usher wrote: 24 Sep 2025 13:43 The question is, how many Falcon owners have used SCSI for storage other than CDROMs?

Given that the main storage for the Falcon is the IDE drive I somewhat doubt there's a very large number of people trying to use SCSI storage as (a) the number of Falcons is small, (b) most were for music composition where the storage of MIDI data would be easily held on the internal IDE drive. So, probably a vanishingly small number.

I have to disagree. Just recently I have inherited a musician's Falcon and got a heap of SCSI ZIP disk media full of files with tens of MB each.

I was thinking maybe more than 80% of Falcon's use IDE. I inherited a musicians Falcon too, it had SCSI internally but most people I know just had IDE. Afterall there were a few years Atari only sold Falcon's with an IDE drive. SCSI was only external but the boot was IDE. Anyway there are a few options here and would be good to test it further.

These are the first tests with jookies program. I've been chatting to him about adding more tests which I'll test tonight..

mikro wrote: 24 Sep 2025 13:48 Just recently I have inherited a musician's Falcon and got a heap of SCSI ZIP disk media full of files with tens of MB each.

What driver?

exxos wrote: 24 Sep 2025 13:59 These are the first tests with jookies program. I've been chatting to him about adding more tests which I'll test tonight..

What would be useful would be locating the transfer buffer at specific addresses, starting below the 4MB barrier, straddling the barrier and having the buffer above the barrier and testing DMA data transfers directly to/from those (note: it may not be just write corruption). With each of these you can try different sizes.

I guess that the best way to rule out (as best as you can) any filesystem access is to preallocate a large file and only write/read within it. If using a filesystem the FAT will still be updated with the access time, however.

Does Jookie say what the SCSI status at the bottom actually means?

I'm guessing that he's doing raw access and this is coming from the SCSI controller, which would suggest that the SCSI device is getting a bad command not just that the data being transferred is corrupt.

stephen_usher wrote: 24 Sep 2025 14:11 What would be useful would be locating the transfer buffer at specific addresses

That is what he is coding now..

Does Jookie say what the SCSI status at the bottom actually means?

Nope.. I'll ask.

Command success" - 0/1 - just boolean, if my SCSI library successfully got to the end of SCSI command and there was no timeout and no error

"SCSI result" - if the "Command success" is 1, then this holds SCSI status byte (from last phase of SCSI command)

if the "Command success" is 0, e.g. due to timeout, then this is just 0xff... usually, with SCSI, value 0 means OK status (no error)

exxos wrote: 24 Sep 2025 14:03What driver?

An older HD Driver, definitely below 10 (I had multiple disks from him with different versions).

Depending what music applications the Falcon was being used for would have a big impact on whether a SCSI drive were used or not.
Both Cubase Audio & Logic Audio could only record/playback multiple audio tracks from a SCSI drive (and preferably an AV SCSI drive at that).
If just using MIDI then IDE would be fine (or for loading the app from disc).
The typical IDE drives in Falcons were too small to be any use for audio recording/playback too.

JezC wrote: 24 Sep 2025 14:39The typical IDE drives in Falcons were too small to be any use for audio recording/playback too.

Exactly my point. So I'm pretty sceptical that none of those musicians wouldn't have noticed.

mikro wrote: 24 Sep 2025 13:48

stephen_usher wrote: 24 Sep 2025 13:43 The question is, how many Falcon owners have used SCSI for storage other than CDROMs?
...
So, probably a vanishingly small number.

I have to disagree. Just recently I have inherited a musician's Falcon and got a heap of SCSI ZIP disk media full of files with tens of MB each.

Concur. And recall how we got here: Cubase Audio insists on SCSI and it was used for direct-to-disc recordings.

It does smack of a fault with that board. SDMA bridged address pin, perhaps?

BW

Badwolf wrote: 24 Sep 2025 14:53 [And recall how we got here: Cubase Audio insists on SCSI and it was used for direct-to-disc recordings.

depending on hdd fragmentation, partition cluster size (e.g. 64 kB), TOS might want to issue smaller transfers - to fit into empty parts of the drive, into individual clusters, etc.

It does smack of a fault with that board. SDMA bridged address pin, perhaps?

Same fault on 2 different falcons....