[Muscle] SCM SCR-301 USB driver on Linux

[Jamie Nicolson]

I'm having some trouble with the SCR-301 0.9.0 driver. Specifically, in IFDHCreateChannel(), Adm_Initialize() returns STATUS_SUCCESS, but Adm_PowerIFD() returns STATUS_UNSUCCESSFUL. Since the Adm code is shipped in object format, I can't tell why it's failing. Adm_PowerIFD() takes a few minutes to return.

I'm running Linux 2.4.20-rc4. I have mounted the usbdevfs filesystem on /proc/bus/usb. The output from pcscd looks like this:

# ./pcscd -d stdout
pcscdaemon.c:258 main: debug messages to stdout
readerfactory.c:1318 RFInitializeReader: Attempting startup of SCM301 0 0.
readerfactory.c:1060 RFBindFunctions: Loading IFD Handler 2.0
pcscdaemon.c:413 main: pcsc-lite daemon ready.
readerfactory.c:1318 RFInitializeReader: Attempting startup of SCM SCR 301 0 0.
readerfactory.c:1060 RFBindFunctions: Loading IFD Handler 2.0
readerfactory.c:1355 RFInitializeReader: Open Port 200000 Failed
readerfactory.c:1190 RFUnloadReader: Unloading reader driver.
readerfactory.c:219 RFAddReader: SCM SCR 301 init failed.
eventhandler.c:319 EHSpawnEventHandler: Error communicating to: SCM301 0 0

The reader shows up in /proc/bus/usb/devices with Vendor=0066 ProdID=1001.

I couldn't find an explanation of why the Adm code is only distributed in object form. Did someone sign an NDA with SCM that prevents him from giving out the source? I would like to try to debug it if I could. Also, I noticed the files usblinux.[ch] and usbserial.c are included with the driver package, but they don't appear to be used. Was the driver originally going to use raw USB, but then changed to use the Adm package instead?

Thanks for any pointers anyone can give me,
-jamie

smime.p7s
Description: S/MIME Cryptographic Signature