Re: [Opensc-devel] Connecting to a specific USB device

[Mat A. <arg...@gm...>]

Hy!

Sorry for bringing that old subject back up. I had to do some other work in 
between, but now really need a solution, which I haven't found yet.

On Wednesday 17. July 2013 13:46:01 Anthony Foiani wrote:
> Mat --
> 
> On Tue, Jul 16, 2013 at 8:30 AM, Mat Arge <arg...@gm...> wrote:
> > Hello!
> > 
> > I am tracking the connection of USB tokens via udev and want to do some
> > specific stuff with them (pass them through to certain virtual machines).
> > For that, I would like to get some token specifics (like the serial
> > number or the PKCS#11 label). So what I need is a connection (preferably
> > with PKCS#11) to the just inserted token. The problem is, that at the
> > udev level where I am, I only know the USB BUS Id, the vendor ID and such
> > stuff. Is there some way to get a PKCS#11 or pc/sc connection to the
> > correct token?
> 
> A quick look through the pcsc-lite stuff is discouraging; it seems
> that a string describing the specific port is indeed stored in the
> reader context structure, but there doesn't seem to be any existing
> way to get it out.
> 
> See line 213 of:
> http://anonscm.debian.org/viewvc/pcsclite/trunk/PCSC/src/readerfactory.c?rev
> ision=6668&view=markup
> 
> For lilbusb, that "device" value is ultimately generated around line 516 of:
> http://anonscm.debian.org/viewvc/pcsclite/trunk/PCSC/src/hotplug_libusb.c?r
> evision=6557&view=markup
> 
> But I don't see any place where that value can be accessed.
> 
> (But I'm certainly not an expert on this; I've just hacked a few
> things onto the source as I needed them -- see next point.)
> 
> The cleanest way might be a vendor-specified attribute, but as the
> comment at the top of p11_attr.c says: "The number of layers we stack
> on top of each other here is frightening."

I am considering writing a patch for pcscd, which writes information about the 
interface used when a connection is created into a structured file. Ludovic, 
would such a patch we acceptable for upstream?

> 
> > Or the other way round: Is there some way to find out for an existing
> > pcsc/pkcs11 connection which hardware address it is leading to?
> 
> It's not exactly what you're looking for, but I did propose a patch to
> pcscd that restricted it to a particular USB port:
> 
> http://opensc.1086184.n5.nabble.com/FYI-PATCH-restrict-pcscd-to-a-single-USB
> -port-path-td13800.html
> 
> Worst case, your udev script could:
> 1. if it looks like a crypto token ...
> 2. spawn a pcscd that looks only at that port...
> 3. then query that specific pcscd to get label etc...
> 4. kill the pcscd...
> 5. bind the port to the VM appropriately.
> 
> Not pretty, and not fast (pcscd takes a few seconds to come up even on
> my fast hardware, but maybe I just don't know how to strip out
> unnecessary bits).
> 
> But it should work.

That's interesting, too. As far as I can see, it was never actually merged 
upstream, was it?

cheers
Mat