Today I managed to get some senors working. I spent quite some time in order to get the TWI/I²C interface working properly, using interrupts etc for interfacing the magnetometer. Still reading some inconsistent values from it though, so I’ll have to look into that a bit further.
The SPI interface to the gyro and accelerometer is currently using old school bit banging, but should now be easily converted into native support. The value readings from the accelerometer are really good. It’s currently set to the ±2g scale, giving me a precision of 0.25mg/LSB. In the live scenario it could be better to use the ±4g scale instead with a 0.5mg/LSB precision.