GIT broken?

[emu]

Just updated from GIT few days ago, but have problems with it.

(https://github.com/dgesswein/mfm/tree/master/mfm)

Complains about cables being reversed, and sometimes gets

"floating point exceptions"?

Am I the only one?

The much older version works ...

[William Stillwell]

Last PR was on 5/16/2026

Which commit did you try pulling to see if it was resolved? I assume it's not a compile error, can you provide your command-line options?

I have had no reason to update mine. Only reason would be a needed feature that has been added or building a new unit.

William Stillwell

--
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 visit https://groups.google.com/d/msgid/mfm-discuss/237ab062-a39f-4bce-b831-236d0123b796%40gmail.com.

[emu]

Just did a "git clone; make clean ; make"

Nothing special.

Had really no reason to update, just did it, as I didn't use my debug
station for a long time
(silly, I know )

cheers

On 2026-06-15 15:16, William Stillwell wrote:
> Last PR was on 5/16/2026
>
> Which commit did you try pulling to see if it was resolved? I assume
> it's not a compile error, can you provide your command-line options?
>
> I have had no reason to update mine. Only reason would be a needed
> feature that has been added or building a new unit.
>
> William Stillwell
>
>
>
>
> On Mon, Jun 15, 2026 at 3:10 PM emu <e2k...@gmail.com> wrote:
>
> Just updated from GIT few days ago, but have problems with it.
>
> (https://github.com/dgesswein/mfm/tree/master/mfm)
>
> Complains about cables being reversed, and sometimes gets
>
> "floating point exceptions"?
>
> Am I the only one?
>
> The much older version works ...
>
> --
> 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

> <mailto:mfm-discuss%2Bunsu...@googlegroups.com>.

> To view this discussion visit
> https://groups.google.com/d/msgid/mfm-discuss/237ab062-a39f-4bce-b831-236d0123b796%40gmail.com.
>
> --
> 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 visit

> https://groups.google.com/d/msgid/mfm-discuss/CADewWKW%2BaaC0c6MVqGToxhxqKL2SibjfZrg9iukvZc9t%2B0jMbA%40mail.gmail.com
> <https://groups.google.com/d/msgid/mfm-discuss/CADewWKW%2BaaC0c6MVqGToxhxqKL2SibjfZrg9iukvZc9t%2B0jMbA%40mail.gmail.com?utm_medium=email&utm_source=footer>.

[emu]

On 2026-06-15 15:30, emu wrote:
> Just did a "git clone; make clean ; make"
>
> Nothing special.
>
> Had really no reason to update, just did it, as I didn't use my debug
> station for a long time
> (silly, I know )
>

OK, tried again:

got the latest sd-image BBB-mfm-emu-v4.43-OS-11.8-r74.img

booted it, rebooted, so it adjusted the sd-card size

got the newest from git

git --clone https://github.com/dgesswein/mfm.git

make

and tried.

Get the "floating point exception" again.

The old mfm utilities, which are on the sd-card image work.

Am I really the only one, having this problem/issue?

Nobody else is using git?

Cheers & thanks in advance

P.S. HW revision "D", BeagleBoneBlack

[Joan Touzet]

I don't usually have availability on the weekend to look into these things,

but I found a bit of free time today.

Confirmed, the very latest commit from David (two months ago) causes

a floating point exception when attempting to read another drive.

Emulation is unaffected, which is probably why no one has reported it

so far.

A roll back to the previous commit (`git checkout 4.44`) and rebuild will

fix the problem.

I didn't have time to debug the issue; hopefully someone else can.

-Joan @ DECromancer

[d...@pdp8online.com]

Works for me with my standard test disk.

debian@BeagleBone:~$ mfm_read --version
Board revision C or D detected
Version 4.45
Drive must be between 1 and 4
debian@BeagleBone:~$ mfm_read --ana --ext /tmp/t
Board revision C or D detected
Found drive at select 1
Returning to track 0
Drive RPM 3600.4
Found matching format RQDX2: good count difference -1
Found matching format WD_3B1: good count difference 0
Found matching format CORVUS_OMNI: good count difference -1
Found matching format Souyz-Neon: good count difference -1
Retrying on cylinder 0 head 1
Found matching format RQDX2: good count difference -1
Found matching format WD_3B1: good count difference 0
Found matching format CORVUS_OMNI: good count difference -1
Found matching format Souyz-Neon: good count difference -1
Multiple matching formats found, using best
Selected head 8 found 0, last good head found 6
Read errors trying to determine sector numbering, results may be in error
Number of heads 7 number of sectors 17 first sector 0
Interleave (not checked): 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Drive supports buffered seeks (ST412)
Disk has recalibrated to track 0                                              
Stopping end of disk search due to recalibration
Number of cylinders 714, 43.5 MB

Command line to read disk:
--format WD_3B1 --sectors 17,0 --heads 7 --cylinders 714 --header_crc 0xffff,0x1021,16,0 --data_crc  0xffff,0x1021,16,0 --sector_length 512 --retries 50,4 --drive 1
All sectors recovered from multiple reads after 8 retries cyl 672 head 0      
Retries failed cyl 673 head 0
Bad sectors on cylinder 673 head 0: 2 3 14
All sectors recovered after 1 retries cyl 673 head 1
Retries failed cyl 674 head 0
Bad sectors on cylinder 674 head 0: 0 2 3 8 9 11 14
Retries failed cyl 674 head 1
Bad sectors on cylinder 674 head 1: 9
Retries failed cyl 675 head 0                                                  
...
Found cyl 0 to 713, head 0 to 6, sector 0 to 16
Expected 84966 sectors got 84758 good sectors, 14 bad header, 194 bad data
0 sectors marked bad or spare
0 sectors corrected with ECC. Max bits in burst corrected 0
Track read time in ms min 18.569833 max 71.599750 avg 35.106509

Will need more information such at the exact command run and the output from the command to investigate further.

[Joan Touzet]

Our standard QA image is Elektronika_85; do you have one of those you can try? If not I can get you a copy, it definitely seems to trigger the issue (when trying to read from a second device, doing the emulation)

[David Gesswein]

Probably. I'll check tomorrow.

I don't understand this

> (when trying to read from a second device, doing the emulation)

I thought you said emulation was fine. Not sure what read from a second device
means.

> --
> 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 visit https://groups.google.com/d/msgid/mfm-discuss/a774d013-ef03-4684-b744-5d0066b3db93n%40googlegroups.com.

[Joan Touzet]

On Saturday, July 11, 2026 at 10:34:28 p.m. UTC-4 David wrote:

I don't understand this
> (when trying to read from a second device, doing the emulation)

I thought you said emulation was fine. Not sure what read from a second device
means. 

Oh! As in, I use one device to emulate a disk, while reading from a second device. Our QA process does this both ways, to validate a new device's ability to do both (without having to spin up a vintage device for the process).

Emulation is indeed fine; reading (from an emulated Elektronica_85 disk running on a second machine) is not. The emulation continues even though the read crashes.

 

[emu]

On 2026-07-11 22:09, d...@pdp8online.com wrote:
> Works for me with my standard test disk.
>

> Track read time in ms min 18.569833 max 71.599750 avg 35.106509
>
> Will need more information such at the exact command run and the
> output from the command to investigate further.
>

I'm using it exactly as written in the examples.

./setup_mfm_read
./mfm_read --analyze --emulation_file xxx.ef --extracted_data_file xxx.ed

Probably it fails, if something is wrong/unexpected happening on the drive?
Most of my drives are flaky, but the old version doesn't crash ...

Cheers

[David Gesswein]

It was executing some code during analyze that it shouldn't have. Error
would occur based on what the data in the first sector was. New version
released.

> --
> 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 visit https://groups.google.com/d/msgid/mfm-discuss/d31fbf07-dfd0-44e2-b4f4-c34cf4c9e146%40gmail.com.

[emu]

On 2026-07-12 15:21, David Gesswein wrote:
> It was executing some code during analyze that it shouldn't have. Error
> would occur based on what the data in the first sector was. New version
> released.

Thank you Dave, for the quick fix.

Tried the current git, and it doesn't crash anymore.

Back to fooling with disks :)