[systemd-devel] systemd-fsck change fsck arguments to "-a" -> "-y"

[Karel Zak]

On Wed, Feb 13, 2013 at 03:32:46AM +0100, Lennart Poettering wrote:
> On Mon, 11.02.13 09:46, Karel Zak (kzak at redhat.com) wrote:
> 
> > 
> > On Fri, Feb 08, 2013 at 12:38:55AM +0100, Lennart Poettering wrote:
> > > > diff --git a/src/fsck/fsck.c b/src/fsck/fsck.c
> > > > index 058f34d..b1938c7 100644
> > > > --- a/src/fsck/fsck.c
> > > > +++ b/src/fsck/fsck.c
> > > > @@ -321,9 +321,10 @@ int main(int argc, char *argv[]) {
> > > >                  }
> > > >  
> > > >          cmdline[i++] = "/sbin/fsck";
> > > > -        cmdline[i++] = "-a";
> > > > +//        cmdline[i++] = "-a";
> > > >          cmdline[i++] = "-T";
> > > >          cmdline[i++] = "-l";
> > > > +        cmdline[i++] = "-y";
> > > >  
> > > 
> > > Hmm, I wonder if -a or -y is the way to go. Karel, as util-linux/fsck
> > > maintainer, do you have an opinion whether we should use -a or -y for
> > > automatic, non-interactive fscking? Is -a obsolete and -y the future?
> > 
> >  This is gray zone... there is not explicit standard or conclusion
> >  that -a (or -p) or -y is supported by all fsck.<type> checkers.
> >  
> >  Anyway, it seems that -a is supported on more places.
> 
> Do you happen to know whether -a and -y are equivalent in the more
> common fsck implementations, modulo the RTC issue this thread was
> initially about?

If I know than -y is fsck.extN specific.


The ideal solution would be to standardize such options (or at least
-a) and inform FS developers that the option will be required. I'll
try to talk with the developers next week. I guess that extN, xfs,
btrfs and vfat is enough for now. (I maintain fsck.cramfs and fsck.minix.)

    Karel

-- 
 Karel Zak  <kzak at redhat.com>
 http://karelzak.blogspot.com