I ordered a SAM7-P64 board for running Simtrace (received board today, but still waiting for the REBELSIm connectors). When compiling the code (using Gnuarm3.4.3) I got two errors: 1) the --g($DEBUGF) option was not understood; omitting this, made the Makefile (for dfu and main) working 2) in /lib/vsprintf.c I had to add #include <limits.h> in order to prevent errors related to MAX_INT etc.
Are there any ideas for making a man-in-the-middle SIM device? One master interface to a real SIM, another slave interface providing a SIM interface towards a phone with the possibility to filter certain APDU's or to add files or commands on top of the real SIM. This would need an additional SIM master interface (to send and receive APDU's to the real SIM) on top of what already is in SIMtrace and the slave interface should be capable of sending messaged back to the real SIM master. I will look into it, first starting with implementing a simple SIM card master on the other UART.
Hi Ste7an,
good to see there are multiple pepole with interest in SIMtrace.
On Fri, Dec 17, 2010 at 12:03:35AM +0100, ste7an wrote:
I ordered a SAM7-P64 board for running Simtrace (received board today, but still waiting for the REBELSIm connectors). When compiling the code (using Gnuarm3.4.3) I got two errors:
- the --g($DEBUGF) option was not understood; omitting this, made the
Makefile (for dfu and main) working 2) in /lib/vsprintf.c I had to add #include <limits.h> in order to prevent errors related to MAX_INT etc.
those are probably all related to the fact that
Are there any ideas for making a man-in-the-middle SIM device? One master interface to a real SIM, another slave interface providing a SIM interface towards a phone with the possibility to filter certain APDU's or to add files or commands on top of the real SIM.
sure, there are ideas for this. The easiest way is to implement transmit support in the SIMtrace firmware + host software and use that implementation as the 'card emulation' part. You can then use any pcsc-lite supported card reader to actually talk to the card, and write some PC software that passes and filters or even manipulates the APDUs.
This would need an additional SIM master interface (to send and receive APDU's to the real SIM) on top of what already is in SIMtrace and the slave interface should be capable of sending messaged back to the real SIM master. I will look into it, first starting with implementing a simple SIM card master on the other UART.
Implementing the reader side inside the SAM7S (on the other UART) will work, too - but at the expense of having to write your own card reader firmware. Depending on where you want to go, simply using an existing USB card reader for the SIM will be the quicker road to success.
Regards, Harald
Sorry, I was missing the second half of one sentence:
On Fri, Dec 17, 2010 at 08:12:32AM +0100, Harald Welte wrote:
On Fri, Dec 17, 2010 at 12:03:35AM +0100, ste7an wrote:
I ordered a SAM7-P64 board for running Simtrace (received board today, but still waiting for the REBELSIm connectors). When compiling the code (using Gnuarm3.4.3) I got two errors:
- the --g($DEBUGF) option was not understood; omitting this, made the
Makefile (for dfu and main) working 2) in /lib/vsprintf.c I had to add #include <limits.h> in order to prevent errors related to MAX_INT etc.
those are probably all related to the fact that
... you are using a different toolchain than the gnuarm-4.0.2 that we use for OsmocomBB as well as OpenPCD and SIMtrace cross-compilation.
SimTrace compiles, I got the P64 board to load the main program. Connected the RebelSim flatcable+SIM. Without connecting the P64 board, the phione can use the SIm without a problem. When connecting the I/O and clk to the PA6+PA1 and PA2+PA4 pins the phone shows a SIM error. Is there a need for a resistor between the SIM signals and the P64 pins? Is there a need for a resistor between the input and output ports on the P64, or can they be directly connected? In some cases I got the ATR sniffed. The value was consistent with other readings. So, part of it i working.
I am using the toolchain provided by the summon-arm-toolchain script from esden. it uses binutils 2.20, newlib 1.18.0, and gcc 4.5.1
- the --g($DEBUGF) option was not understood; omitting this, made the
Makefile (for dfu and main) working
I do not get this error
- in /lib/vsprintf.c I had to add #include <limits.h> in order to
prevent errors related to MAX_INT etc.
I also get this error and solve it the same way
Additionally, I get : error : uses hardware FP, whereas main_simtrace.elf uses software FP which I could not get rid off yet.
Are there any ideas for making a man-in-the-middle SIM device?
I am designing some SIMtrace hardware. I think the schema is almost finished. I will publish it in the next days and begin with the board if no one finds an error. I already ordered most of the parts to build the PCB.
Kevin
baseband-devel@lists.osmocom.org
-
Harald Welte -
ml＠mail.tsaitgaist.info -
ste7an