On Sat, 2009-10-10 at 10:54 -0500, Raphael Geissert wrote: > 2009/10/10 Soeren Sonnenburg <so...@debian.org>: > > On Fri, 2009-10-09 at 18:59 -0500, Raphael Geissert wrote: > >> severity 550322 important > >> tag 550322 moreinfo > >> thanks > >> > >> Downgrading the severity as, unless proved wrong, this is an exception > >> caused > >> by a bug somewhere else. > >> > >> On Friday 09 October 2009 16:50:45 Soeren Sonnenburg wrote: > >> > re-running with readlink -vf leads to the output attached in readlink_v. > >> > Seems like the migration of initserv (backup files) cause files to be > >> > not found... > >> > >> Do you mean insserv? > > > > Yes. > > Just to double check, did you manually backup those files?
no I think it was innserv doing this back in the days when it was announced and I was trying this out... > >> Could you also please provide the output of ls -lR /etc/rcS.d? > > > > Attached. > > > >> I will fix it in readahead-fedora by limiting the directory recursion, but > >> this case is strange. > > > > But why does readlink fail? What if there are stale symlinks even > > in /etc/rc?.d/ ? I would expect the same problem to occur. > > > > Because /etc/rcS.d/rcS.d/../init.d != /etc/init.d/, and since > /etc/rcS.d/init.d doesn't exist readlink fails. True, but still the postinst script shouldn't fail ... I guess your find -maxdepth approach is good enough to solve this but I am afraid reaslink will fail too when there is some stale symlink in etc/rc?.d/ ... Soeren Soeren -- For the one fact about the future of which we can be certain is that it will be utterly fantastic. -- Arthur C. Clarke, 1962-- For the one fact about the future of which we can be certain is that it will be utterly fantastic. -- Arthur C. Clarke, 1962
signature.asc
Description: This is a digitally signed message part
