[fprint] [PATCH] Support for password prompt while finger scanning.

Bastien Nocera hadess at hadess.net
Mon Feb 16 10:15:39 GMT 2009


On Sat, 2009-02-14 at 20:37 +0100, Timo Hoenig wrote:
> Hi!
> 
> On Sat, 2009-02-14 at 20:31 +0100, Timo Hoenig wrote:
> 
> > If you want to get this done properly please help enhancing fprintd.
> 
> I just saw that your patch is for fprintd.  Without having looked at it
> deeper, the uinput hack should go.  This ugly peace needs to be solved
> in a different way.

Exactly, otherwise we end with the same problems as in thinkfinger. I
know Ray Strode is currently working on a proof-of-concept patch for GDM
to allow having multiple PAM stacks, so you'd have one for
"devices" (say a fingerprint reader, or a smartcard reader), and one for
your usual password/kerberos/ldap/whatever. This could be extended in
the future to other applications.

The problem really lies with PAM, not with our use of it, and if there
was a non-hacky way to get password and fingerprint auth at the same
time, we'd probably be doing it already...

> Also, as this patch is just based on my code, please state this in the
> header of it.  It's completely misleading as is.
> 
> Everything else of my previous mail still applies.

The only thing in your other mail is a warning about the device heating
up because it would be left in an undefined state. That shouldn't happen
when using fprintd, as the device will be closed as soon as the D-Bus
connection to it is closed. So if your fprintd-using application
crashes, or exits without cleaning up after itself, then fprintd will
take care of mopping up.

Cheers



More information about the fprint mailing list