Page 1 of 1
System Error: ARM CPU data abort catched
Posted: Mon Nov 23, 2009 6:15 am
by ak
Hi,
I got this MSG twice today for the first time while I was playing music both in Prog and Seq mode.
System Error: ARM CPU data abort catched
Anyone knows why?
My m3 has the latest OS 2.4, Radias, 256ram, and Firewire card.
Re: System Error: ARM CPU data abort catched
Posted: Mon Nov 23, 2009 3:02 pm
by cowtothesky
ak wrote:Hi,
I got this MSG twice today for the first time while I was playing music both in Prog and Seq mode.
System Error: ARM CPU data abort catched
Anyone knows why?
My m3 has the latest OS 2.4, Radias, 256ram, and Firewire card.
I had this problem a few months ago and it wasn't good. It happened to me a few times randomly before, but I was able to turn off my synth and turn it back on to recover. One day, it did it on start-up and I was unable to get the synth to come on. I had to take it to an authorized Korg repairman to fix it. It took about 3 weeks before I got it back.
What I think may be happening is a corruption in the internal hard drive. What I would do is update to the latest OS (again) and MAKE SURE you run the update 2 times. I am not 100% sure that this was what caused it for me, but I may have forgotten to run the update twice.
Good luck.
Posted: Mon Nov 23, 2009 5:18 pm
by BasariStudios
It means you should ARM your self Karim...lol...search the topic i opened on this,
it was a whole issue that run back and forth between me, Jerry, Daz, Sharp and
the engineers in Japan, it was one of the biggest issues. It can be caused by
certain type of samples and unproper USB stick and NOTHING else.
Posted: Mon Nov 23, 2009 6:45 pm
by ak
mmm..what kind of samples are you referring to?
Posted: Tue Nov 24, 2009 6:06 am
by ak
I found the problem and it is very strange.
Please read this...
I have created Category and named it "Orienta" and created a Sub_Category of Oriental and named it "Plucked/Guitar/Oud/Kanon" and as soon as I select the Sub_Category then the M3 gives this message:
System Error: ARM CPU data abort catched
It has nothing to do with samples or corrupted programs. I would like to see Korg Engineers looking into it.
The solution is to name the Sub_Category just "Plucked." Again, it is very strange for KM3 to act this way.

Posted: Sun Jan 03, 2010 11:10 pm
by kanthos
I'd bet that a) the keyboard uses the / character to denote a subfolder, just like Unix-based filesystems do b) since a program or combi can only be in one category and one subcategory, it uses a folder structure Category/SubCategory/Program, and c) by putting the extra / characters in the category name, the M3 is looking for a folder called Oriental/Plucked/Guitar/Oud/Kanon (i.e. a folder that's 5 levels deep) which never got created and doesn't exist.
Posted: Mon Jan 04, 2010 12:22 am
by ak
I agree with you. It is looking for a path and now it is all working. I changed the name to something else.
I wonder why Korg never paid attention to my post.
Posted: Mon Jan 04, 2010 1:52 pm
by kanthos
Probably because Korg Japan would be the ones making the fix, and they've never had any connection with these forums, to my knowledge.
Posted: Mon Jan 04, 2010 3:29 pm
by Rob Sherratt
The M3 software was developed in the USA. Jerry Kovarsky is USA marketing manager for Korg USA and he (jerrythek) writes on this forum quite often. But at the moment I think everyone in Korg and especially Jerry Kovarsky is focused on NAMM.
On my own M3 I confirmed that the bug in it with malformed names - it's not just "/" that causes it but there's other non alphanumeric characters that can do it as well. But given that everyone knows what causes the problem, and it's repeatable, it's easy to work around for now.
The "ARM CPU data abort catched" message is from an exception handler that is triggered when the ARM (Acorn Risk Machines) CPU attempts to execute an illegal instruction. This can be caused by a stack under or overflow if there is a coding error (e.g. a fault in the Korg software when parsing filenames containing non alphanumeric characters).
There are other reports about the problem over at Karma Lab Forums.
Regards,
Rob
Posted: Tue Jan 05, 2010 5:02 am
by cowtothesky
Rob Sherratt wrote:The M3 software was developed in the USA. Jerry Kovarsky is USA marketing manager for Korg USA and he (jerrythek) writes on this forum quite often. But at the moment I think everyone in Korg and especially Jerry Kovarsky is focused on NAMM.
On my own M3 I confirmed that the bug in it with malformed names - it's not just "/" that causes it but there's other non alphanumeric characters that can do it as well. But given that everyone knows what causes the problem, and it's repeatable, it's easy to work around for now.
The "ARM CPU data abort catched" message is from an exception handler that is triggered when the ARM (Acorn Risk Machines) CPU attempts to execute an illegal instruction. This can be caused by a stack under or overflow if there is a coding error (e.g. a fault in the Korg software when parsing filenames containing non alphanumeric characters).
There are other reports about the problem over at Karma Lab Forums.
Regards,
Rob
Thanks for the info, Rob. But, I think this is something that should be given a higher priority. All I was doing when it happened to me was recording and hitting the compare button several times. Standard fast recording stuff. I am terrified that this will happen again, and I have quite a few projects lined up. One thing I always counted on from Korg was reliability. This is a pretty common occurance and it seems the only "fix" without voiding the warranty is to take it in for repairs, which took me 3 weeks last time this happened. I'm hoping that Korg comes through on this issue.
Posted: Tue Jan 05, 2010 9:16 am
by Rob Sherratt
cowtothesky wrote:Thanks for the info, Rob. But, I think this is something that should be given a higher priority. All I was doing when it happened to me was recording and hitting the compare button several times. Standard fast recording stuff. I am terrified that this will happen again, and I have quite a few projects lined up. One thing I always counted on from Korg was reliability. This is a pretty common occurance and it seems the only "fix" without voiding the warranty is to take it in for repairs, which took me 3 weeks last time this happened. I'm hoping that Korg comes through on this issue.
Just out of interest, were you recording direct to a USB device? Just a theory .... the M3 USB interface does not work reliably with certain USB sticks ....
Have you taken the M3 in for warranty repair? If so, did it solve the problem? What was the fix?
Regards,
Rob
Posted: Tue Jan 05, 2010 9:28 am
by BasariStudios
This cannot be traced to a single reason for happening, it has been confirmed
on over 30 machines, 3 of my personal M3s, happens randomly on many things.
Before OS 2.0 was happening all over, even without Samples, after that it
continued, Me, Jerry and Sharp worked on it a little behind the scenes and it
was brought to the attention of the Engineers, then 2.04 came after 2.02.
Now its happening again, on each machine i touched or requested testing.
I can confirm that it does happen even without any KSC loaded even when the
machine is reinstalled and Factory Default loaded which has nothing to do with
Category name. Many people have problems which brings problems to me.
Many people are ignorant and many people dont have my knowledge of the
M3 so when they purchase something for it and it happens they come back to
me and blame me...i dont think anyone deserves that. It is frustrating, annoying,
it affects my daily life, career life and everything. I also have the affect on many
people to actually stop them and prevent them from purchasing an M3.
I am sorry. Sometimes i even demonstrate the problem right in front of
them, as i've done it many times in Sam Ash and GC and the Salesmen
are just stunned, they usually end up asking ME questions.
And for the ones having this problem, dont bother calling Korgs tech support,
its not worth the time nor the money for phone, they have no clue what is
going on...its more of Caveman watching TV in color.
Posted: Tue Jan 05, 2010 11:17 pm
by cowtothesky
Rob Sherratt wrote:cowtothesky wrote:Thanks for the info, Rob. But, I think this is something that should be given a higher priority. All I was doing when it happened to me was recording and hitting the compare button several times. Standard fast recording stuff. I am terrified that this will happen again, and I have quite a few projects lined up. One thing I always counted on from Korg was reliability. This is a pretty common occurance and it seems the only "fix" without voiding the warranty is to take it in for repairs, which took me 3 weeks last time this happened. I'm hoping that Korg comes through on this issue.
Just out of interest, were you recording direct to a USB device? Just a theory .... the M3 USB interface does not work reliably with certain USB sticks ....
Have you taken the M3 in for warranty repair? If so, did it solve the problem? What was the fix?
Regards,
Rob
Nope - I was using internal presets with the sequencer. It wasnt connected to anything. I did have a flash card usb, but only used it to save my sequences. I had problems from time to time with the crash, but was able to re-start the M3 (seemed to happen when I would change from the modes - combi to program or program to sequencer). One day, it crashed on load up and would not exit the load up screen, so I was forced to bring it in for repairs.
The tech who repaired it said that it had a corrupted internal hard drive and he had to completely re-load the memory. It was all covered under warranty, but had a long down time as he is the only authorized Korg repairman in my area and also works on Yamaha and Roland synths, so I had to take a number. At that time, I hadn't experimented with loading samples or soundfonts, so I know it wasn't that.
I haven't had the error since I got my synth back, but I am concerned that it might happen again, since so many have posted this issue.
One sidenote - I did have the latest version operating system when this happened. I am not 100% sure if I loaded it twice though. I think I did, but honestly can't remember.
Thanks
Posted: Tue Jan 05, 2010 11:59 pm
by Rob Sherratt
Hi cowtothesky,
The tech who repaired it said that it had a corrupted internal hard drive and he had to completely re-load the memory.
The M3 doesn't have an internal hard drive but it does have an internal SD card (a type of solid state disk), so I guess that must be what he said. Replacing the SD card and reloading it with the operating system takes about 1 hour max. to do. See this Wiki article:
http://karma-lab.wikidot.com/korg-m3:ex ... to-2-gbyte
I now keep a backup image of the SD card on my PC just in case I ever have to replace the SD card due to a similar fault.
Best regards,
Rob
Posted: Wed Jan 06, 2010 12:46 am
by cowtothesky
Rob Sherratt wrote:Hi cowtothesky,
The tech who repaired it said that it had a corrupted internal hard drive and he had to completely re-load the memory.
The M3 doesn't have an internal hard drive but it does have an internal SD card (a type of solid state disk), so I guess that must be what he said. Replacing the SD card and reloading it with the operating system takes about 1 hour max. to do. See this Wiki article:
http://karma-lab.wikidot.com/korg-m3:ex ... to-2-gbyte
I now keep a backup image of the SD card on my PC just in case I ever have to replace the SD card due to a similar fault.
Best regards,
Rob
Yep - that's what it was. That is a good idea about keeping a backup, but it still wouldn't have solved my problem because the keyboard wouldn't exit the loading screen in order to activate any menus or restore. Also, opening up the M3 would void my warranty, which I don't really want to do. So, I had no choice. It probably does take 1 hour to fix, but he had to do the research and diagnose the problem, which can be time consuming. He said he left it on for 48 hours straight after fixing it and couldn't get the error. Also, he wasn't all that familiar with the M3 either. I think mine was the first M3 he had worked on. Something is definitely going on that I don't believe is user-caused. But, I appreciate your feedback and hope this does get looked into.
Thanks for your input Rob.