On Fri, Mar 02, 2018 at 11:14:13AM -0500, Theodore Ts'o wrote:
> On Fri, Mar 02, 2018 at 03:20:22PM +0000, Joakim Tjernlund wrote:
> >
> > But I don't hav ext3, only ext4 in kernel:
> > # CONFIG_EXT2_FS is not set
> > # CONFIG_EXT3_FS is not set
> > CONFIG_EXT4_FS=y
>
> What version of the kernel are you using? The ext3 file system has
> been removed from the upstream kernel for quite some time, so this
> must be a kernel missing lots of security bug fixes. (E.g., an
> obsolete distribution kernel? :-)
Never mind, the "CONFIG_EXT3_FS is not set" confused me. We still
have this for backwards compatibility with old kernel configs.
So I'm not seeing this behavior at all. I created an ext4 file system
on /dev/cwcc/scratch, and then ran:
# mount -t auto /dev/cwcc/scratch /mnt
and the only thing in my dmesg from that mount is:
EXT4-fs (dm-4): mounted filesystem with ordered data mode. Opts: (null)
This assumes you have a mount command that uses blkid to identify the
file system type. If you have a mount -t auto which just blindly
probes, then sure, you can get this message if it does the equivalent of
mount -t ext2 /dev/cwcc/scratch /mnt
mount -t ext3 /dev/cwcc/scratch /mnt
mount -t ext4 /dev/cwcc/scratch /mnt
# mount -t ext3 /dev/cwcc/scratch /mnt
mount: /mnt: wrong fs type, bad option, bad superblock on /dev/mapper/cwcc-scratch, missing codepage or helper program, or other error.
In which case you will get:
EXT4-fs (dm-4): couldn't mount as ext3 due to feature incompatibilities
But that's because mount explicitly asked for ext3. So this may boil
down to whaht version of mount (and from what package; are you using
busybox?) and what system calls it is sending down into the kernel.
Also note that technically speaking this is not an *error* message.
If it had been an error message, it would have looked like this:
EXT4-fs error (dm-4): file system is on fire! Run away!! Run away!!
^^^^^
What was printed was just an informative message; sent to syslog with
a priority of LOG_ERR, true. But technically, still just a message
sent voa ext4_msg() as opposed to the ext4_error() function:
if (IS_EXT3_SB(sb)) {
if (ext3_feature_set_ok(sb))
ext4_msg(sb, KERN_INFO, "mounting ext3 file system "
"using the ext4 subsystem");
else {
ext4_msg(sb, KERN_ERR, "couldn't mount as ext3 due "
"to feature incompatibilities");
goto failed_mount;
}
}
In any case, this does not appear to be a kernel issue, but rather a
userspace program issue. I can't reproduce this on my Debian testing
laptop.
Cheers,
- Ted
