No special mount options for root file system:
/dev/sda1 / ext2 rw,relatime,errors=remount-ro 0 0
However, it is unusual to actually use ext2. Jerre mentioned that in an earlier comment.
I can create a new ext2 partition here with working ACLs. Even when I do
sudo tune2fs -E mount_opts="noacl" /dev/sdb
on my local test partition, dumpe2fs doesn't change the default mount options, but has a new fiield "Mount options", which isn't being considered. That smells like another bug, but is not quite related to this change.
Jerre is using the -lowlatency kernel, but as far as I can see, ACL support is enabled there for all relevant file systems (grep FS.*ACL /boot/config-3.5.0-18-lowlatency).
For more experiments I need to be able to load scsi_debug and/or do loop mounts for further testing, and thus need sudo privileges.
I'm logged into Jerre's box now, and this is what I can find out without root privileges:
- I confirm that ACLs are broken:
$ touch test.txt
$ getfacl test.txt
# file: test.txt
# owner: mpitt
# group: mpitt
user::rw-
group::rw-
other::r--
$ setfacl -m u:nv:r test.txt
setfacl: test.txt: Operation not supported
No special mount options for root file system: errors= remount- ro 0 0
/dev/sda1 / ext2 rw,relatime,
However, it is unusual to actually use ext2. Jerre mentioned that in an earlier comment.
I can create a new ext2 partition here with working ACLs. Even when I do
sudo tune2fs -E mount_opts="noacl" /dev/sdb
on my local test partition, dumpe2fs doesn't change the default mount options, but has a new fiield "Mount options", which isn't being considered. That smells like another bug, but is not quite related to this change.
Jerre is using the -lowlatency kernel, but as far as I can see, ACL support is enabled there for all relevant file systems (grep FS.*ACL /boot/config- 3.5.0-18- lowlatency) .
For more experiments I need to be able to load scsi_debug and/or do loop mounts for further testing, and thus need sudo privileges.