"Panopticon is a disassembler that understands the semantics of opcodes. This way it's able to help the user by discovering and displaying invariants that would have to be discovered "by hand" in traditional disassemblers."
Doesn't every disassembler have to understand the opcode semantics in order to disassemble and make sense of them or am I misinterpreting that statement?
Also can anyone explain what the "invariants" are? When I hear the word I can only think of loop invariants and I'm guessing that is not what the author means here.
You say "in order to disassemble and make sense of them", but the "sense" that a disassembler makes can just be the direct conversion from machine code to equivalent assembley. Machine code is just another syntax, directly converting that into assembly can happen instruction-by-instruction.
Making more sensible assembly that a human would write is another thing, if you want to follow jumps to figure out where instructions are and regenerate labels, then that's one step towards semantics.
I don't know what kinds of invariants panopticon is looking at, but there seems to be code for static analysis using abstract interpretation[1]. This could do the data-flow analysis, determine types and recognize persistent variables.
"Panopticon is a disassembler that understands the semantics of opcodes. This way it's able to help the user by discovering and displaying invariants that would have to be discovered "by hand" in traditional disassemblers."
Doesn't every disassembler have to understand the opcode semantics in order to disassemble and make sense of them or am I misinterpreting that statement?
Also can anyone explain what the "invariants" are? When I hear the word I can only think of loop invariants and I'm guessing that is not what the author means here.