[PATCH] Remove incorrect expression leading to errors.

NIIBE Yutaka gniibe at fsij.org
Mon Jan 26 03:02:20 CET 2015


Thank you for your patch.

On 01/26/2015 03:30 AM, Joshua Rogers wrote:
> On 25/01/15 20:42, Werner Koch wrote:
>> Thanks.  Pushed to 2.0 and master.
> Any idea the implications of this? Doesn't look security related, but it
> looks like it wouldn't report that the card is not present, or that it
> is inactive, or it has too bit of a response(could this be sec related?
> probably not)

The function send_escape_cmd is to send raw command to card readers
(at initialization time, basically).  Since commands to be sent is for
card readers (not cards), we should ignore card status.

The function is exposed by ccid_transceive_escape, but
ccid_transceive_escape is not used at all, in fact.  Thus, we only
need to check internal use of send_escape_cmd in ccid-driver.c.

The function calls are only possible for specific card readers of:

  VEGA_ALPHA, VENDOR_VEGA, VENDOR_CHERRY, VENDOR_GEMPC, VENDOR_SCM

For all of those calls, no return messages are expected.  If a reader
sends back some message (say, after upgraded firmware in future), it
was just ignored (before the patch of yours).  The impact at large
would be such a reader won't work any more, or our intended behavior
of pinpad won't work well.
-- 



More information about the Gnupg-devel mailing list