We present a generic framework for runtime code polymorphism, applicable to a large class of computing platforms up to embedded systems with low computing resources (e.g. microcontrollers with few kilo-bytes of memory). Code polymorphism is dened as the ability to change the observable behaviour of a software component without changing its functional properties. In our framework, code polymorphism is achieved thanks to runtime code generation, which oers many levers for code transformations: we describe the use of random register allocation, random instruction selection, instruction shuing and insertion of noise instructions.We evaluate the eectiveness of our framework against dierential power analysis and its overhead impact. As compared to a reference implementation of AES where the cipher key could be recovered by DPA in less than 50 traces in average, in our implementation the key cipher could not be extracted after 10000 traces. Our experimental evaluation shows a moderate impact in terms of performance overhead.