Converting Al Kossow's UniFLEX disk image
nosn...@gmail.com
Al,
I tried doing to your UniFLEX disk image .td file what I did to Adams UniFLEX .td file with no luck. I was able to get an extracted file generated on the BeagleBone, but when I transfer it to the PC with pscp, it makes no sense and FloppyMaintenance does not recognize it as a UniFLEX image. This is the command I used to convert from .td to .ext along with the output from mfm_util:
root@beaglebone:~/mfm# ./mfm_util --ana --emu ../ms3438_gimix_xe1420.emu --ext ms3438_gimix_xe1420.ext
Original decode arguments: --format Xebec_S1420 --sectors 17,0 --heads 4 --cylinders 615 --header_crc 0x0,0xa00805,32,2 --data_crc 0x0,0xa00805,32,2 --sector_length 512 --retries 50,4 --drive 1
Matches count 34 for controller Xebec_S1420
Header CRC: Polynomial 0xa00805 length 32 initial value 0x0
Sector length 512
Data CRC: Polynomial 0xa00805 length 32 initial value 0x0
Number of heads 4 number of sectors 17 first sector 0
Interleave (not checked): 0 13 9 5 1 14 10 6 2 15 11 7 3 16 12 8 4
Command line to read disk:
--format Xebec_S1420 --sectors 17,0 --heads 4 --cylinders 615 --header_crc 0x0,0xa00805,32,2 --data_crc 0x0,0xa00805,32,2 --sector_length 512 --retries 50,4 --drive 1
Bad sectors on cylinder 614 head 0: 0H 1H 2H 3H 4H 5H 6H 7H 8H 9H 10H 11H 12H 13H 14H 15H 16H
Bad sectors on cylinder 614 head 1: 0H 1H 2H 3H 4H 5H 6H 7H 8H 9H 10H 11H 12H 13H 14H 15H 16H
Bad sectors on cylinder 614 head 2: 0H 1H 2H 3H 4H 5H 6H 7H 8H 9H 10H 11H 12H 13H 14H 15H 16H
Bad sectors on cylinder 614 head 3: 0H 1H 2H 3H 4H 5H 6H 7H 8H 9H 10H 11H 12H 13H 14H 15H 16H
Found cyl 0 to 613, head 0 to 3, sector 0 to 16
Expected cyls 615 doesn't match cyls found 614
Expected 41820 sectors got 41752 good sectors, 68 bad header, 0 bad data
0 sectors marked bad or spare
0 sectors corrected with ECC. Max bits in burst corrected 0
One HUGE difference between the two images (other than actual content of course), is that your image seems to be interleaved and Adam’s does not. And yours is 17 sectors per track while Adam’s is 18 sectors per track. Yours came from an image that was connected to a Xebec controller while Adam’s drive was connected to an Adaptec controller. Also when I converted Adam’s I converted from an emulator file to an extract file. Yours I did a transitions file directly to an extract file.
I guess my question is really for David – when converting from a transitions file to an extract file, are the sectors un-interleaves so that they are in logical order instead of physical order in the extract file? If this is NOT the case, I would need to un-interleave them from the extract file to .dsk image used by FloppyMaintenance.
Michael Evenson
http://www.swtpcemu.com/swtpc/Downloads.htm
http://www.flexusergroup.com:8080 <- http link to the sftp site
David Gesswein
and emulation file. Some controllers use space at the beginning of the
disk for storing information they need so you may need to remove some
amount from the beginning of the image. The interleave should have been
removed.
> that they are in logical order instead of physical order in the extract
> file? If this is NOT the case, I would need to un-interleave them from the
> extract file to .dsk image used by FloppyMaintenance.
>
>
>
> Michael Evenson
>
> <http://www.swtpcemu.com/swtpc/Downloads.htm>
>
> <http://www.flexusergroup.com:8080/> http://www.flexusergroup.com:8080
>
>
>
>
>
>
>
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/004b01d9e6f4%24739cb9c0%245ad62d40%24%40gmail.com.
nosn...@gmail.com
Thanks for clarifying the interleave removal when converting. After reading the mfm_util document from your web site, I realize that you had already stated that. Sorry for not RTF(ine)M first. I kinda figured that the Al Kossow image needed some sectors removed at the beginning. I just haven't been able to figure out how many yet. I tried searching for a valid SIR sector, but so far it is eluding me. I am nothing if not persistent. I WILL find it. With UniFLEX, the SIR starts at sector 1 (offset 0x200) in a valid image. Not knowing how many cylinders, heads and sectors that UniFLEX was told it has when the drive was formatted makes it a bit of a challenge. Perhaps Al can shed some light on that. If I knew what to look for that would be helpful. If I could get a dump of the first two sectors from the actual drive, I would be able to easily trim the excess sector from the start of the image. I did manage to find the root directory start. It is at offset 0xA5E00 in the extract file.
Al - if you are reading this, any information about the geometry of the disk would be most helpful.
Michael Evenson
http://www.swtpcemu.com/swtpc/Downloads.htm
Al Kossow
> David,
>
> Thanks for clarifying the interleave removal when converting. After reading the mfm_util document from your web site, I realize that you had already stated that. Sorry for not RTF(ine)M first. I kinda figured that the Al Kossow image needed some sectors removed at the beginning. I just haven't been able to figure out how many yet. I tried searching for a valid SIR sector, but so far it is eluding me. I am nothing if not persistent. I WILL find it. With UniFLEX, the SIR starts at sector 1 (offset 0x200) in a valid image. Not knowing how many cylinders, heads and sectors that UniFLEX was told it has when the drive was formatted makes it a bit of a challenge. Perhaps Al can shed some light on that. If I knew what to look for that would be helpful. If I could get a dump of the first two sectors from the actual drive, I would be able to easily trim the excess sector from the start of the image. I did manage to find the root directory start. It is at offset 0xA5E00 in the extract file.
>
> Al - if you are reading this, any information about the geometry of the disk would be most helpful.
nosn...@gmail.com
Al,
Thanks for that. I was more interested in where the UniFLEX partition begins on the drive. Which cylinder would be a great help. I am going to try to work back from where I found the root directory. I would need to know how big the fdn space is though. If you could do a etc/diskinfo /dev/w0 an post the results that would tell me how big the fdn and swap spaces are as well as what UniFLEX sees as the disk geometry.
This is what I get when I run it on my UniFLEX 6809 emulation:
Disk Information for "/dev/w0":
------------------------------------------------
Creation date: 04:35:44 Tue Mar 07 1989
Last update: 07:17:17 Tue Mar 07 1989
Total disk size: 15.6 Mb (31212 blocks)
- Fdn space: 472.5 Kb (945 blocks, 7560 fdns)
- File space: 14.1 Mb (28225 blocks)
- Swap space: 1.0 Mb (2040 blocks)
Free space: 12.2 Mb (24318 blocks)
Free fdns: 6951
Disk type: Mini Winchester
306 cylinders
6 heads
Also a capture of etc/devcheck /dev/w0 would be nice.
Checking device '/dev/w0'.
Testing Boot - Block 0
Testing SIR - Block 1
Testing FDN space - Blocks 2 through 946 (0x3B2)
Testing Volume space - Blocks 947 (0x3B3) through 29171 (0x71F3)
Testing Swap space - Blocks 29172 (0x71F4) through 31211 (0x79EB)
Device check for '/dev/w0' complete - 0 bad blocks detected.
Michael Evenson
http://www.swtpcemu.com/swtpc/Downloads.htm
http://www.flexusergroup.com:8080 <- http link to the sftp site
-----Original Message-----
From: mfm-d...@googlegroups.com <mfm-d...@googlegroups.com> On Behalf Of Al Kossow
Sent: Thursday, September 14, 2023 9:06 AM
To: mfm-d...@googlegroups.com
Subject: Re: Converting Al Kossow's UniFLEX disk image
On 9/14/23 6:02 AM, nosn...@gmail.com wrote:
--
You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/356209c0-4c04-e460-1c6a-9ed56b942d41%40bitsavers.org.
Al Kossow
> Al,
>
> Thanks for that. I was more interested in where the UniFLEX partition begins on the drive. Which cylinder would be a great
> help. I am going to try to work back from where I found the root directory. I would need to know how big the fdn space is though. If you
> could do a etc/diskinfo /dev/w0 an post the results that would tell me how big the fdn and swap spaces are as well as what UniFLEX sees as
> the disk geometry.
nosn...@gmail.com
-----Original Message-----
From: mfm-d...@googlegroups.com <mfm-d...@googlegroups.com> On Behalf Of Al Kossow
Sent: Thursday, September 14, 2023 9:46 AM
To: mfm-d...@googlegroups.com
Subject: Re: Converting Al Kossow's UniFLEX disk image
You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
nosn...@gmail.com
Al,
If the reason your GIMIX Micro-20 machine does not run is not a hardware failure, then it could be your hard drive. While investigating the image you made available of your machine's hard drive, I discovered that all of the sectors on cylinder 0 (all 4 heads) are corrupted. Starting at cylinder 1, head 0, sector 0, there are fdn entries. But the fdn list does not start at fdn 0. The fdn list is supposed to start at off 0x400 (cylinder 0, head 0, sector 2. The layout of a valid UniFLEX hard drive is as follows:
// Boot - sector 0 0x00000000 - 0x000001FF 00/01 - 00/01
// SIR - sector 1 0x00000200 - 0x000003FF 00/02 - 00/02
// FDN space - sector 2 0x00000400 – (number of fdn sectors – 2) * blocksize
// Volume space - sectors after the FDN space
// Swap space - sectors after the volume space
At least I think that’s what it should be. When looking at the hard drive image for my 6809 emulation set up for UniFLEX, this is the case. This is also the case for Adam’s Tektronics 4404 hard disk. Yours has all 0x6C bytes except for the first few bytes of each sector that are:
02 66 04 10 02 02 66 00 80 0B FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF
You might want to take another shot at using the MFM reader to get a new image of your hard drive. Just a thought.
Michael Evenson
http://www.swtpcemu.com/swtpc/Downloads.htm
http://www.flexusergroup.com:8080 <- http link to the sftp site
-----Original Message-----
From: mfm-d...@googlegroups.com <mfm-d...@googlegroups.com> On Behalf Of Al Kossow
Sent: Thursday, September 14, 2023 9:46 AM
To: mfm-d...@googlegroups.com
Subject: Re: Converting Al Kossow's UniFLEX disk image
On 9/14/23 6:41 AM, nosn...@gmail.com wrote:
--
You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/5ac147cd-411a-2a7c-7fbe-6854127b40f8%40bitsavers.org.
David Gesswein
> 02 66 04 10 02 02 66 00 80 0B FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF
Frequently end of disk is spare. Somewhat suspicious the pattern is the same
for all sectors and track. I would expect at least the head to be encoded in it.
Possibly this is the controller storing the disk geometry.
614 cylinders, 4 heads. Would have expected the next byte to be 11 for 17
sectors per track that it looks like its using.
Looked at cyl 614 and it is all zero. Decoded the
transistions file again and cylinder 614 was not decodable. Poking at it I
don't see any valid sectors headers in the xebec_s1420 format the rest of
the disk is in. Does have one sector header that looks reasonable for Western
digital format with 0xa1 0xfc pattern. Seems to be folowed by data sector with
10240 bytes. Data is just a fill pattern. Assume this track wasn't formatted
by the GIMIX.
If its spared somewhere I don't know where. If you know what one of the missing
fdn's should look like you could search the file for it.
Checked the GIMIX image I have but its OS-9.
On Thu, Sep 14, 2023 at 07:47:58PM -0400, nosn...@gmail.com wrote:
> Al,
>
>
>
> If the reason your GIMIX Micro-20 machine does not run is not a hardware failure, then it could be your hard drive. While investigating the image you made available of your machine's hard drive, I discovered that all of the sectors on cylinder 0 (all 4 heads) are corrupted. Starting at cylinder 1, head 0, sector 0, there are fdn entries. But the fdn list does not start at fdn 0. The fdn list is supposed to start at off 0x400 (cylinder 0, head 0, sector 2. The layout of a valid UniFLEX hard drive is as follows:
>
>
>
> // Boot - sector 0 0x00000000 - 0x000001FF 00/01 - 00/01
>
> // SIR - sector 1 0x00000200 - 0x000003FF 00/02 - 00/02
>
> // FDN space - sector 2 0x00000400 – (number of fdn sectors – 2) * blocksize
>
> // Volume space - sectors after the FDN space
>
> // Swap space - sectors after the volume space
>
>
>
> At least I think that’s what it should be. When looking at the hard drive image for my 6809 emulation set up for UniFLEX, this is the case. This is also the case for Adam’s Tektronics 4404 hard disk. Yours has all 0x6C bytes except for the first few bytes of each sector that are:
>
>
>
>
>
>
>
>
>
> Michael Evenson
>
>
> <http://www.flexusergroup.com:8080/> http://www.flexusergroup.com:8080 <- http link to the sftp site
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: mfm-d...@googlegroups.com <mfm-d...@googlegroups.com> On Behalf Of Al Kossow
> Sent: Thursday, September 14, 2023 9:46 AM
> To: mfm-d...@googlegroups.com
> Subject: Re: Converting Al Kossow's UniFLEX disk image
>
>
>
> On 9/14/23 6:41 AM, <mailto:nosn...@gmail.com> nosn...@gmail.com wrote:
>
> > Al,
>
> >
>
> > Thanks for that. I was more interested in where the
>
> > UniFLEX partition begins on the drive. Which cylinder would be a great
>
> > help. I am going to try to work back from where I found the root
>
> > directory. I would need to know how big the fdn space is though. If you could do a etc/diskinfo /dev/w0 an post the results that would tell me how big the fdn and swap spaces are as well as what UniFLEX sees as the disk geometry.
>
>
>
> the machine doesn't work
>
>
>
> --
>
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
>
>
> To view this discussion on the web visit <https://groups.google.com/d/msgid/mfm-discuss/5ac147cd-411a-2a7c-7fbe-6854127b40f8%40bitsavers.org> https://groups.google.com/d/msgid/mfm-discuss/5ac147cd-411a-2a7c-7fbe-6854127b40f8%40bitsavers.org.
> --
> You received this message because you are subscribed to the Google Groups "MFM Discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to mfm-discuss...@googlegroups.com.
nosn...@gmail.com
David,
Since I know where the fdn is supposed to start, and I found the root directory which is always the first thing in the data space behind the fdn list, I should be able to figure out where all of the files are described in the portion of the fdn that is still intact. Unfortunately, there are 210 missing FDNs. That a lot of files whose whereabouts would be unknown. I am going to give it a shot though. The first files put on the disk would have been the files put there during installation, so I am hoping that they are all contiguous, since these file would be the ones pointed to by the missing fdns. Each directory entry has the first two bytes equal to the FDN number that holds the descriptor for the entry. Another helpful thing is that binary files from TSC always have the file info bytes at the end of each program and they are quite easy to relocate.
Maybe if I can reconstruct Al's disk image, he will be enticed t get his Micro-20 running again. 😊
I would really be disappointed if what appears to be the only GIMIX Micro-20 disk image is lost to histroy. Maybe it is time to switch to using the image posted by Adam Billyard. Time to start looking for technical specs for the Tektronics 4404. If I can reverse engineer the UniFLEX file from his image, perhaps I can adapt it to the CDS68020.
I am going to take a stab at getting what files I can from Al's disk image.
Does anyone else out there have an image of a 68K UniFLEX hard drive. I'd love to get my hands on one for the CDS68020.
Michael Evenson
http://www.swtpcemu.com/swtpc/Downloads.htm
http://www.flexusergroup.com:8080 <- http link to the sftp site
To view this discussion on the web visit https://groups.google.com/d/msgid/mfm-discuss/ZQO4VIj7uFDufTRM%40hugin3.