Would you like to update microtrace code?

[xiaoyao377905]

Hi! Our team is interested in BCSD(Binary Code Similarity Detection) problem.
We tried to rebuild a pretrain model, but it performed bad on arm/arm64 arch.
Then we found that it seem missed something in microtrace code. (Or maybe I made some mistakes.)
Eg. In micro_trace/kadabra/arch/arm/arm_64.py, the conditional jump code is missed.
In micro_trace/kadabra/emulator/hooks.py the hook code seems missed hook for arm/arm64 arch.
And by the way, the microtrace seems always the same result on x86/x86-64 arch.(Traced about 100000 rounds) It seems ignored
some conditional jump then caused miss some code in the result. I think becasue of the conditional jump code, there should be
a different result after a enough trace round.
So would you like to update microtrace code or show more details about the mircotrace?
Thanks a lot!

[xiaoyao377905]

I found the problem about the x86 arch trace.
The hook code to handle conditional jump or unconditional jump inst only increase IP pointer by inst size, that means this choice dorpped some code of the jump inst.So the trace result of "static" maybe drop some code. May I have your mind of this problem? Whether this problem affect the result of similarity detect or not?