I am assuming this should place be appropriate for what I want to ask. I want to ask more of a philosophical and procedural question because I don't want to try things that have been tried and spend much time on what I might need to redo in the future.
I have experience in source code recovery and have recovered complete programs back to source, but I noticed that in many instances (in X86) when I didn't have a function completed, I was able to just use __asm {} for functions that weren't completed yet. Essentially, the process I used for recovering X86 programs was to convert them using IDA to pseudo C and then clean it up, restore structs and classes, throw it in Visual Studio and correct types and other things, but this process as you can tell is very painstaking.
Now, I am wanting to try restoring a X64 bit program and I see there is no inline assembly for X64 assembly in Visual Studio reference.
An approach that I am evaluating, is to have a system of inline assembly and then converting function by function from assembly to pseudo C and then correcting the pseudo C, structs, classes, et cetera. The functions which aren't completely recovered yet would be in asm {} and I am thinking that maybe since Visual Studio cannot use the x64 asm blocks, that it might be a safe idea to compile these uncompleted functions as assembly objects (.obj files) with another compiler then link them together with MSVC.
Another idea that I am currently evaluating is to use Intel compiler which supposedly can use inline X64 assembly and use Visual Studio as well.
The compiler is important, because often you want to use the same compiler as the authors used for recovery. I assume, though, that because the code will be assembly, the compiler here doesn't matter too strongly.
Is this a sound idea, would I not have to scrap it later?
