Bug#246841: qualified success on iBook G4
clone 246841 -1
reassign 246841 partman-newworld
retitle 246841 partman-newworld: doesn't set filesystem properly
On Sun, May 02, 2004 at 10:46:56PM +0100, Colin Watson wrote:
> On Sun, May 02, 2004 at 09:47:37PM +1000, Jonathan Matthew wrote:
> > On Sun, May 02, 2004 at 11:40:08AM +0100, Colin Watson wrote:
> > > It was perfectly fine here, so no, I don't think that's the case at all.
> > >
> > > I'd like to see the contents of the directory corresponding to that
> > > filesystem in /var/lib/partman/devices, particularly the 'method' file.
> > The 'method' file says 'newworld'. The 'filesystem' file says 'ext3',
> > which is what seems to be causing the problem. ext3 is the filesystem
> > set before I change the usage method to newworld. Maybe
> > choose_method/newworld/do_option should set 'filesystem' to 'hfs'?
> > Perhaps setting 'method' to 'newworld' should also set 'filesystem'
> > to 'hfs'? It looks like this would go in
> > partman-newworld/choose_method/newworld/do_option.
> Hm, I thought that the presence of /lib/partman/parted_names/newworld
> containing 'hfs' would take care of this ...
Aha, I see what's happening. Anton, your change to make partman-newworld
remove the new use_filesystem file and to drop its valid_filesystems
directory broke its ability to set the filesystem correctly.
The flow in partman-target/update.d/filesystems is something like this:
use_filesystems is missing, so property=newworld;
there's no valid_filesystems/*newworld, so filesystems is empty,
therefore available is empty;
acting_filesystem is removed.
What were you proposing to do with property="$method", if not pass it to
a valid_filesystems script? If it's supposed to be passed to
valid_filesystems, I'd like to reinstate
partman-newworld/valid_filesystems, and maybe partman-palo should be
checked as well. If not, I'm confused.
Colin Watson [email@example.com]