Jonathan Salwan

You should probably first post your snippet of code and a short description of the issue in this thread.

Thanks for this MR, > I could do that but its best to have all of the files formatted first Before updating all files, we should first agree on a...

So far for this `stosb` instruction, I see no problem. What I suggest is to make one PR per each instruction with a bunch of unit tests to be sure...

In a general point of view, removing all `evaluate()` from the instruction semantics could be a step forward to solve #473.

Yeah I think grouping them would be ok :)

STORE/LOAD nodes are already implemented (#1185) and we have modes for this now: * https://github.com/JonathanSalwan/Triton/blob/master/src/libtriton/includes/triton/modesEnums.hpp#L34 * https://github.com/JonathanSalwan/Triton/blob/master/src/libtriton/includes/triton/modesEnums.hpp#L39-L40 Some examples here: * https://github.com/JonathanSalwan/Triton/blob/master/src/testers/unittests/test_symbolic_array.py However, i'm pretty sure we can do better...

You have to update the memory only if there is a desynchronization between your real CPU state (via IntelPin/Dynamorio) and the Triton's state Take a look at the `mem_read` and...

> However, when I tried to solve branches How do you solve a branch? If you want to solve a branch you have to take into account previous constraints too...

> But comparison will be done via XMM registers. What should I do? I'm not sure to get this point. If you receive a model from the solver, you just...

Sorry for the delay, can you provide a unit test of this behaviour ?