On 08/01/2019 18:45, Frank Scheiner wrote: > On 1/8/19 14:15, Mark Cave-Ayland wrote: >> Do you see the fsck error on all filesystems or just /dev/sda4? > > I only have one HFS (actually /dev/sda2) on the used disk. And the `fsck.hfs` > errors > happen with both clean and damaged HFSes. > >> I'd be interested to >> do a side-by-side comparison with the same filesystem image on my local G4 >> mini vs. >> your G5 to help work out what is happening, if it's something you can make >> available? > > I already tried that, `fsck.hfs` works on a G4 (Mac mini G4 here, too) and > can both > check and repair clean/damaged HFSes on the very same disk. See [1] for the > process, > where I first made sure that the HFS is ok on a Mac mini G4 and then tried to > check > it on a G5 with (1) a 64 bit version of fsck.hfs - which segfaulted - and (2) > a 32 > bit version of fsck.hfs - which worked the same way as it did on the G4. > > [1]: https://lists.debian.org/debian-powerpc/2019/01/msg00023.html > > Therefore think the problem is not the state of the HFS, as I got the > segfaults for > both damaged and undamaged HFSes. Quick test: Just create a HFS on a G4 (e.g. > using > `parted` and `mkfs.hfs`) and try to check it on a G5 with the 64 bit version > of > `fsck.hfs`.
Absolutely I understand that you've done the work to narrow down the issue, but the easiest way to do this will be to build full debug versions on both a G4 and a G5, step through them both via gdb, and then it should be really easy to spot where things start to go wrong. I don't think it will take long at all to track this down, but whilst I have some G4 hardware sadly I don't have any G5 hardware to test with. Does anyone out there have a G5 that they could temporarily let me have SSH access to for a short while? Non-root should be absolutely fine. ATB, Mark.
