On Thu, 2010-09-02 at 21:31 +0300, Martin Paljak wrote:
> Hello,
> 
> On Sep 2, 2010, at 9:16 PM, Andre Zepezauer wrote:
> >  But as an inspiration for the future, this problem can be solved throughout
> > exploiting logical channels.
> Which problem? How?

1. If only one application authenticates successfully, then the token
becomes unlocked and will be accessible without authentication for all
the other applications _and_ users too.

2. Application A authenticates successfully. Later application B fails
to authenticate. Now application A is unauthenticated too. Not sure if
mozilla will ask for a pin again. My assumption is, it won't.

3. Left as an exercise for the interested reader.

> However, logical channels may share application-dependent security status and 
> therefore may have security-related command interdependencies across logical 
> channels (e.g. password verification).

Source iso7816-4 Draft:

After a successful open function performed from

1) the basic logical channel (bits 6, 2 and 1 all set to zero in CLA,
coding number zero), the MF shall be implicitly selected as the current
DF and the security status for the new logical channel should be the
same as for the basic logical channel after the answer to reset. The
security status of the new logical channel should be separate from that
of any other logical channel.

2) a non-basic logical channel (bits 6, 2 and 1 not all set to zero in
CLA, coding a number from one to seven), the current DF of the logical
channel from which the command was issued shall be selected as the
current DF and the security status for the new logical channel should be
the same as for the logical channel from which the open function was
performed.


> Do you have a specific card in mind?

Every modern Java card and in particular GlobalPlatform is capable of
doing so, if the applet implements javacard.framework.MultiSelectable.

> Do you have a patch

No.

>  or a plan on how to apply the concept to OpenSC?

Not in detail. But if someone will begin a serious discussion on it,
then I will participate.


_______________________________________________
opensc-devel mailing list
opensc-devel@lists.opensc-project.org
http://www.opensc-project.org/mailman/listinfo/opensc-devel

Reply via email to