>Wouldn't it be enough to rearrange your replacement code to have >the nop in the beginning? The your jump would be at the same >address again. No it wouldn't. It might work just in the example I gave, but there could be situations where it won't work. For example a (really- dummy-I-hope-noone-will-ever-write) code like this: movsb movsb movsb movsb movsb (With no rep prefix!), would be transformed into: A4 A4 A4 A4 A4 byte sequence. The program would replace this sequence with a call (still not relocated): E8 FC FF FF FF no nops at all. Still, there could be a jump to every "original" A4, resulting in 4 potential SIGILL... Bye, Alessandro
