> Perhaps running tune2fs on the new ext2 filesystems might be the way
> to go.
Can the sparse_super flag be changed back after a filesystem is in place?
I'll have to test this.... It does seem to work! SO! we-have-a-solution!
?? seems-to work...
boot tomsrtbt
tune2fs -s 0 /dev/hda1
e2fsck /dev/hda1
mount /dev/hda1 /mnt
# work with filesystem from tomsrtbt
umount /mnt
tune2fs -s 1 /dev/hda1
e2fsck /dev/hda1
> good... if there was any blame to be thrown around then it is at the
> authors of the e2fs tools for not phasing in this fundamental change
They said:
-" Unfortunately, 2.0 kernels don't support the sparse superblock option,
-" and therein lies the problem. I had assumed that given that 2.4 is
-" almost ready to go out the door, it was time to change the default
-" options for mke2fs to create filesystems that scaled properly to
-" modern-sized disks. Unfortunately this creates a interoperability
-" problem with 2.0 kernels, but for many people, especially for fixed
-" disks, this doesn't matter. They're not likely to unplug their SCSI
-" disk and plug it into a system running a 2.0 kernel.
Oh well. I will look into back-porting sparse superblock support to 2.0.
-Tom
