Once the code has been located, we lift the program representation from disassembled instructions to a graph. Lifting enables representing multiple architectures. Graph nodes are code blocks and data blocks, and graph edges are control flow arcs from code blocks to other code blocks. The code blocks contain sequences of disassembled instructions, and any code-referent operands within these instructions (like a call instruction) are replaced by symbolic operands that refer to the associated code blocks. Control may only enter a code block at the start of the block, simplifying analyses like liveness.
Μήπως είναι ο εκθετικός τελεστής;
