Hello,
My CCID driver has a bug with the ATR "3B 80 81 1F 07 19". This ATR
defines a TA3 but that is not the IFSC but XI and UI information. My
online parser [1] is correct.
I had a look at OpenCT and the same bug is also present.
openct-0.6.20/src/ifd/ifd-ccid.c:
/* TA3 -> IFSC */
if (atr_info.TA[2] != -1)
parambuf[5] = atr_info.TA[2];
OpenCT is even less flexible since only TA3 is used as IFSC, not the
first TAi with i>2
The bug should not prevent the communication to happen. But the T=1
blocks will be limited to 7 bytes instead of the default 32 bytes. The
communication speed with the card will be lower.
I implemented a function to get the IFSC value. Feel free to reuse it
(under LGPL v2+)
static int get_IFSC(ATR_t *atr, int *idx)
{
int i, ifsc, protocol = -1;
/* default return values */
ifsc = -1;
*idx = -1;
for (i=0; i<ATR_MAX_PROTOCOLS; i++)
{
/* protocol T=? */
if (atr->ib[i][ATR_INTERFACE_BYTE_TD].present)
protocol = atr->ib[i][ATR_INTERFACE_BYTE_TD].value & 0x0F;
/* TAi (i>2) present and protocol=1 => IFSC */
if (i >= 2 && protocol == 1
&& atr->ib[i][ATR_INTERFACE_BYTE_TA].present)
{
ifsc = atr->ib[i][ATR_INTERFACE_BYTE_TA].value;
*idx = i+1;
/* only the first TAi (i>2) must be used */
break;
}
}
return ifsc;
} /* get_IFSC */
It is used as in:
ifsc = get_IFSC(&atr, &i);
if (ifsc > 0)
{
DEBUG_COMM3("IFSC (TA%d) present: %d", i, ifsc);
param[5] = ifsc;
}
[1] http://smartcard-atr.appspot.com/parse?ATR=3B80811F0719
--
Dr. Ludovic Rousseau
_______________________________________________
opensc-devel mailing list
[email protected]
http://www.opensc-project.org/mailman/listinfo/opensc-devel
