Data abort, Undefined instruction, Prefetch abort (Update)

[Llorean]

You could always attempt to actually *fix* the problem, rather than trying to find a way to avoid the problem. That would not only allow you to listen to music again, but also solve the problem for anyone else who might also run into it, and contribute to the advancement of Rockbox.

[aarongoltz]

Do you have any ideas of things to try to go about fixing my problem? I have tried formatting my Ipod and removing the partitions, installing new and old Ipod firmware, using the newest bootloader and new and old Rockbox builds, but nothing seems to work. I've tried using the Apple OS to confirm that my Ipod is at least working. I've provided as much information as I can about my issue, as well as tried doing anything and everything that I know how to do, to no avail. I don't know at what point it seemed like I was trying to avoid this problem, as all I feel I have been doing is try to fix it with the help of other Rockbox users.

[Llorean]

Any step that isn't "figure out what actually causes the problem, then write new code for Rockbox so that the problem isn't happening any more" is avoiding it.

Using an old build, using an older Apple firmware, etc, are all ways to go around the problem, but leave whatever's causing it existing and just pretending like it's not there, aren't they?

To be fixed someone will have to do some programming, and it's always best done by people who experience the problem because they can actually know if they've succeeded in fixing it.

Any "fix" that doesn't involve changing Rockbox means the problem will still be there if Rockbox is run under the right conditions - so how can one actually say it's "fixed"?

[aarongoltz]

From all of my research on these forums, it seems like one way to identify a problem is by trying out different builds. I'm not looking to use an old build that may work so I can go ahead and listen to my music. I'm trying to find any build that works so I can look for changes to see at what point Rockbox stopped working for my Ipod.

[Llorean]

That's one way to identify an introduced bug. This seems to be a case of us having always done something wrong (or technically, not done something at all trusting to the Apple flash loader to set things up), and a change (the new cpu speed handling) having brought that problem to light because it now actually causes issues.

It may require someone investigating the Apple firmware and fixing our IDE timings, assuming the current diagnosis is correct.

[aarongoltz]

Do you think the next firmware issues would prevent me from using software that is very old? I tried loading a build from January and it wouldn't run either. Also, is there a chance that something in my Ipod is broken or corrupted that wouldn't effect Apple's OS, but may effect Rockbox?

[Llorean]

Well, here's some clear points:

1) Some people don't have this problem, in fact most regular users of Rockbox don't.
2) This same problem seems to happen on the Nano, and on that hardware it seems to be related to the Firmware Version.
3) Many people *on Nano* have been able to revert to Rockbox builds immediately before Amiconn's first CPU speed change and use it.

Because it affects multiple people, and seems to have been triggered by this commit, it's most likely the same issue. Though it could be possible that something is broken in just your iPod causing the same symptoms, it is somewhat unlikely, and the best route of investigation is the IDE timings.

[aarongoltz]

#Reistlehr-

Check what I wrote in the update I added to my first post. Let me know if that pertains to your situation.

Aaron