On Mon, Dec 27, 2004 at 08:28:25PM +0100, Andreas Unterkircher wrote: > Hello List! > > I'm currently revive an old Digital AlphaStation 200 with an Alpha EV4 > processor (Avanti system). OS is a Debian SID (unstable) with a self > compiled 2.6.9 kernel (gcc 3.3.5). > On one of the two scsi disks I created one single partition and formated > it successfully with JFS. > > When I'm trying to mount that partition, the mount-binary segfaults and > the kernel throws following message into the logs... > <snip> > I've tested the disk with 'badblocks' - seems there is no problem. Also > a ext3-filesystem runs in a 'normal' way. > > Maybe it's the memory with some defects, but this only happens when I > try to mount an JFS filesystem. > > But could this be kernel or JFS related? Below is the actual .config. > I'm currently recompile the kernel with > JFS_DEBUG which hopefully give a hint where the problem is.... but this > will take sometime on this machine :) >
This is actually a known issue, which I think Shaggy has plans to work on in the coming year. Basically, JFS doesn't work properly on architectures without a 4k page size. This includes Alpha 8K, Sparc 8K, and more. I think maybe we should at least disable JFS support on architecutres where this is the case, so more people don't get confused by this. Sorry! Sonny _______________________________________________ Jfs-discussion mailing list [email protected] http://www-124.ibm.com/developerworks/oss/mailman/listinfo/jfs-discussion
