Re: localdev and fixed devices

On Fri, Sep 05, 2008 at 01:57:46PM +0200, Helmut Lichtenberg wrote:
> Oliver Grawert schrieb am 27. Aug 2008 um 14:44:55 CEST:
> > it used to be without checks for removable devices in the original
> > implementation, due to massive complaints from skolelinux (schools with
> > dual booted client machines suddenly gave full acccess to the XP
> > partition to everyone logged in on a thin client) we decided to accept
> > their hardcoded patch to disable HDDs completely until someone else
> > comes up with a better solution ... so patches are indeed appreciated
> > any happily accepted ;)
 
> I sent a patch for this problem to the list, but Jim McQuillan redirected me
> to ltsp-developers. So here's my suggestion:

> 1. udev must provide all devices, not only removables. Otherwise we could not
>    select between them via lts.conf.

until very recently, we didn't have access to lts.conf values from the
udev scripts (mainly ltspfs_entry). now that we do, it's probably time
to reconsider this as a configurable value.

> 2. It seems easier for me not to allow certain devices but to deny them.

we'll probably want both allow and deny, i'm guessing.

> 3. Let's restrict it for now for the two traits 'removable' and 'fixed'.
>    Introduce a variable in lts.conf like this:
> 
>    LOCALDEV_DENY = fixed      # possible values: [fixed|removable]

> 4. Maybe later we can enhance it for certain named devices or partitions. This
>    needs a more complicated retrieval of udev information and decisions about
>    the syntax in lts.conf.

well, it's more complicated than one might think. USB external
hard-drive cages are not considered removable according to udev, because
the hard-drive itself is a fixed device, even though it's attached by
USB, which is clearly removable. i don't know for sure, but there may be
other cases like that (firewire devices?).

maybe this would be why we need a corresponding LOCALDEV_ALLOW.
 
> To implement this for the new ltspfs-trunk, Oliver pointed to, there are only
> two new functions for ltspfs_entry: is_removable() and is_denied().
> 
> I've put a first implementation into a shell script, where you can test it
> easily. 

the basic idea looks reasonable(if we can account for the
above-mentioned issues), but i'd prefer to see a patch against
ltspfs-trunk instead of an example script, if possible. this would make
it easier to test and see how it would be used.

live well,
  vagrant

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_____________________________________________________________________
Ltsp-developer mailing list.   To un-subscribe, or change prefs, goto:
      https://lists.sourceforge.net/lists/listinfo/ltsp-developer
For additional LTSP help,   try #ltsp channel on irc.freenode.net