tillitis-key/hw/application_fpga/fw

Firmware

Introduction

This text is an introduction to, a requirement specification of, and some implementation notes of the TKey firmware. It also gives a few hint on developing and debugging the firmware.

This text is specific for the firmware. For a more general description on how to implement device apps, see the TKey Developer Handbook.

Definitions

• Firmware - software in ROM responsible for loading applications. The firmware is included as part of the FPGA bitstream and not replacable on a usual consumer TKey.
• Device application or app - software supplied by the client which is received, loaded, measured, and started by the firmware.

CPU modes and firmware

The TKey has two modes of software operation: firmware mode and application mode. The TKey always starts in firmware mode and starts the firmware. When the firmware is about to start the application it switches to a more constrained environment, the application mode.

The TKey hardware cores are memory mapped. Firmware has complete access, except that the UDS is readable only once. The memory map is constrained when running in application mode, e.g. FW_RAM and UDS isn't readable, and several other hardware addresses are either not readable or not writable for the application.

See the table in the Developer Handbook for an overview about the memory access control.

Communication

The firmware communicates to the host via the UART_{RX,TX}_{STATUS,DATA} registers, using the framing protocol described in the Framing Protocol.

The firmware uses a protocol on top of this framing layer which is used to bootstrap the application. All commands are initiated by the client. All commands receive a reply. See Firmware protocol for specific details.

Memory constraints

• ROM: 6 kByte.
• FW_RAM: 2 kByte.
• RAM: 128 kByte.

Firmware behaviour

The purpose of the firmware is to load, measure, and start an application received from the client over the USB/UART.

The firmware binary is part of the FPGA bitstream as the initial values of the Block RAMs used to construct the FW_ROM. The FW_ROM start address is located at 0x0000_0000 in the CPU memory map, which is the CPU reset vector.

Firmware state machine

This is the state diagram of the firmware. There are only four states. Change of state occur when we receive specific I/O or a fatal error occurs.

stateDiagram-v2
     S1: initial
     S2: loading
     S3: running
     SE: failed

     [*] --> S1

     S1 --> S1: Commands
     S1 --> S2: LOAD_APP
     S1 --> SE: Error

     S2 --> S2: LOAD_APP_DATA
     S2 --> S3: Last block received
     S2 --> SE: Error

     S3 --> [*]

States:

• initial - At start. Allows the commands NAME_VERSION, GET_UDI, LOAD_APP.
• loading - Expect application data. Allows only the command LOAD_APP_DATA.
• run - Computes CDI and starts the application. Allows no commands.
• fail - Stops waiting for commands, flashes LED forever. Allows no commands.

Commands in state initial:

command	next state
FW_CMD_NAME_VERSION	unchanged
FW_CMD_GET_UDI	unchanged
FW_CMD_LOAD_APP	loading

Commands in state loading:

command	next state
FW_CMD_LOAD_APP_DATA	unchanged or run on last chunk

See Firmware protocol in the Dev Handbook for the definition of the specific commands and their responses.

State changes from "initial" to "loading" when receiving LOAD_APP, which also sets the size of the number of data blocks to expect. After that we expect several LOAD_APP_DATA commands until the last block is received, when state is changed to "running".

In "running", the loaded device app is measured, the Compound Device Identifier (CDI) is computed, we do some cleanup of firmware data structures, flip to application mode, and finally start the app, which ends the firmware state machine.

The device app is now running in application mode. There is no other means of getting back from application mode to firmware mode than resetting/power cycling the device. Note that ROM is still accessible in the memory map, so it's still possible to execute firmware code in application mode, but with no privileged access.

Firmware loads the application at the start of RAM (0x4000_0000). It uses the special FW_RAM for its own stack.

When the firmware starts it clears all FW_RAM, then sets up a stack there before jumping to main().

When reset is released, the CPU starts executing the firmware. It begins by clearing all CPU registers, and then sets up a stack for itself and then jumps to main().

Beginning at main() it sets up the "system calls", then fills the entire RAM with pseudo random data and setting up the RAM address and data hardware scrambling with values from the True Random Number Generator (TRNG). It then waits for data coming in through the UART.

Typical expected use scenario:

1. The client sends the FW_CMD_LOAD_APP command with the size of the device app and the optional 32 byte hash of the user-supplied secret as arguments and gets a FW_RSP_LOAD_APP back. After using this it's not possible to restart the loading of an application.

2. If the the client receive a sucessful response, it will send multiple FW_CMD_LOAD_APP_DATA commands, together containing the full application.

3. On receivingFW_CMD_LOAD_APP_DATA commands the firmware places the data into 0x4000_0000 and upwards. The firmware replies with a FW_RSP_LOAD_APP_DATA response to the client for each received block except the last data block.

4. When the final block of the application image is received with a FW_CMD_LOAD_APP_DATA, the firmware measure the application by computing a BLAKE2s digest over the entire application. Then firmware send back the FW_RSP_LOAD_APP_DATA_READY response containing the digest.

5. The Compound Device Identifier (CDI) is then computed by doing a new BLAKE2s using the Unique Device Secret (UDS), the application digest, and any User Supplied Secret (USS).

6. The start address of the device app, currently 0x4000_0000, is written to APP_ADDR and the size of the binary to APP_SIZE to let the device application know where it is loaded and how large it is, if it wants to relocate in RAM.

7. The firmware now clears the special FW_RAM where it keeps it stack. After this it performs no more function calls and uses no more automatic variables.

8. Firmware starts the application by first switching to from firmware mode to application mode by writing to the SWITCH_APP register. In this mode the MMIO region is restricted, e.g. some registers are removed (UDS), and some are switched from read/write to read-only (see the memory map).

   Then the firmware jumps to what is in APP_ADDR which starts the application.

If during this whole time any commands are received which are not allowed in the current state, we enter the "failed" state and execute an illegal instruction. An illegal instruction traps the CPU and hardware blinks the status LED red until a power cycle. No further instructions are executed.

User-supplied Secret (USS)

USS is a 32 bytes long secret provided by the user. Typically a client program gets a secret from the user and then does a key derivation function of some sort, for instance a BLAKE2s, to get 32 bytes which it sends to the firmware to be part of the CDI computation.

Compound Device Identifier computation

The CDI is computed by:

CDI = blake2s(UDS, blake2s(app), USS)

In an ideal world, software would never be able to read UDS at all and we would have a BLAKE2s function in hardware that would be the only thing able to read the UDS. Unfortunately, we couldn't fit a BLAKE2s implementation in the FPGA at this time.

The firmware instead does the CDI computation using the special firmware-only FW_RAM which is invisible after switching to app mode. We keep the entire firmware stack in FW_RAM and clear it just before switching to app mode just in case.

We sleep for a random number of cycles before reading out the UDS, call blake2s_update() with it and then immediately call blake2s_update() again with the program digest, destroying the UDS stored in the internal context buffer. UDS should now not be in FW_RAM anymore. We can read UDS only once per power cycle so UDS should now not be available to firmware at all.

Then we continue with the CDI computation by updating with an optional USS and then finalizing the hash, storing the resulting digest in CDI.

We also clear the entire FW_RAM where the stack lived, including the BLAKE2s context with the UDS, very soon after that, just before jumping to the application.

Firmware services

The firmware exposes a BLAKE2s function through a function pointer located in MMIO BLAKE2S (see memory map) with the with function signature:

int blake2s(void *out, unsigned long outlen, const void *key,
	    unsigned long keylen, const void *in, unsigned long inlen,
	    blake2s_ctx *ctx);

where blake2s_ctx is:

typedef struct {
	uint8_t b[64]; // input buffer
	uint32_t h[8]; // chained state
	uint32_t t[2]; // total number of bytes
	size_t c;      // pointer for b[]
	size_t outlen; // digest size
} blake2s_ctx;

The libcommon library in tkey-libs has a wrapper for using this function called blake2s() which needs to be maintained if you do any changes to the firmware call.

Developing firmware

Standing in hw/application_fpga/ you can run make firmware.elf to build just the firmware. You don't need all the FPGA development tools. See the Developer Handbook for the tools you need. The easiest is probably to use your OCI image, ghcr.io/tillitis/tkey-builder.

Our version of qemu is also useful for debugging the firmware. You can attach GDB, use breakpoints, et cetera.

If you want to use plain debug prints you can remove -DNOCONSOLE from the CFLAGS in the Makefile and using the helper functions in lib.c like htif_puts() htif_putinthex() htif_hexdump() and friends for printf-like debugging. Note that these functions are only usable in qemu.

Test firmware

The test firmware is in testfw. It's currently a bit of a hack and just runs through expected behavior of the hardware cores, giving special focus to access control in firmware mode and application mode.

It outputs results on the UART. This means that you have to attach a terminal program to the serial port device, even if it's running in qemu. It waits for you to type a character before starting the tests.

It needs to be compiled with -Os instead of -O2 in CFLAGS in the ordinary application_fpga/Makefile to be able to fit in the 6 kByte ROM.