Jim Meyering wrote:
Curtis Gedak wrote:
The reason I have been following up on this, perhaps incorrectly, is
because there is legacy code in GParted that uses the
ped_disk_commit_to_os() to determine if a device can have it's
partition table re-read by the kernel.
But that is only useful if that partition table has been changed.
Is GParted changing partition tables that have mounted partitions?
If so, wouldn't it be better to refuse to do that, requesting that the
user first unmount those partitions?
The existing functionality in GParted does permit adding partitions and
manipulating unmounted partitions on a device that has a partition
mounted. Parted permits this functionality too including with BETA
parted-2.0. Whether this is planned functionality in parted or not is
another question. The only restriction is that mounted partitions can
not be modified and this makes sense to me.
This ability permits a person to run parted or gparted from a mounted
partition that contains the operating system. If the drive has
additional free space on it, then the user can create new partitions, or
manipulate unmounted partitions on the same device.
I know that a person can boot from a LiveCD and hence have no mounted
partitions on a device. This is what we often recommend for users
wishing to edit partitions.
I see that Karel Zak has contributed to this discussion as well. My
knowledge of the kernel internals is minimal. Perhaps some of Karel's
suggestions would permit the kernel's vision of the partition table to
be properly kept up-to-date while still permitting parted, and gparted
to add or remove partitions from a device that has at least one
partition mounted?
Again I defer to your decisions on what features parted will offer and
support.
Curtis
_______________________________________________
parted-devel mailing list
parted-devel@lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/parted-devel
