Cheat Engine

[Frouk]

So I was writing a simple code to get the ground level which returns float, the main thing that executeCodeEx returns the value that was in eax register, which may or may not be a return value for float or double types, usually functions that return float use fld instruction, fstp instruction gets the actual value that is stored in there

[Dark Byte]

Write a stub that converts the float into EAX/RAX and use executeCodeEx on that stub function

[Frouk]

what do I need to create a stub function?

[ParkourPenguin]

Allocate some memory and write some assembly that calls the function in that memory. After the function returns, use `movd eax,xmm0` to return the float in eax.

Follow correct calling conventions when calling the function. For 64-bit code, the stack must be aligned to 16 bytes and there must be at least 32 bytes of free space on the stack for the callee to use. executeCodeEx should deal with the parameters, as long as you don't mess with those registers.
For 32-bit code, figure out the calling convention yourself. Probably cdecl or stdcall, maybe something weirder. Depending on the calling convention, you might have to modify executeCodeEx too- that's now calling your stub and not the game's function.

[Frouk]

[ParkourPenguin]

[Frouk]

doesn't crash, but it makes target laggy

EDIT:

function result is returned as int, needs only conversion