On Fri, Feb 12, 2016 at 02:31:49PM +0100, Georg Altmann wrote:
> Previous message was hard to understand: When mounting a ext4 fs without a
> journal, but a mount option that controls journaling is used, clearly state
> that the fs cannot be mounted with this option because it has no journal.
Well, technically it's possible for the file system to have a journal,
but for journalling not to be enabled. For example:
root@kvm-xfstests:~# dmesg -n 7
root@kvm-xfstests:~# mke2fs -t ext4 -Fq /dev/vdc
/dev/vdc contains a ext4 file system
created on Fri Feb 12 11:59:10 2016
root@kvm-xfstests:~# dumpe2fs /dev/vdc | grep features
dumpe2fs 1.43-WIP (18-May-2015)
Filesystem features: has_journal ext_attr resize_inode dir_index filetype
extent 64bit flex_bg sparse_super large_file huge_file dir_nlink extra_isize
metadata_csum
Journal features: (none)
root@kvm-xfstests:~# mount -o noload,commit=1 /dev/vdc /vdc 2> /dev/null
[ 313.867505] EXT4-fs (vdc): can't mount with commit=1, fs mounted w/o journal
So saying "fs has no journal" isn't necessarily going to be correct.
Maybe "Can't mount with data=xxx, journalling not enabled" would be
less confusing to users?
Cheers,
- Ted
