Cheat Engine

[Dr.Disrespect]

The "Code Injection" function often picks more than one line of assembly code, for example:

[++METHOS]

Using the injection templates allow you to reroute your code to a code cave so that you do not have any size restrictions or overlapping of existing, functional code. That being the case, when you establish a jump to your code cave, it will require so many bytes for the jump. Depending on the size of the original instruction, your new jump code may require more bytes than what the original instruction needed...so, instead of overlapping existing code, which would most probably cause a crash, CE will NOP all of the bytes for both instructions, and just relocate them both so that there are no issues that might cause the target to crash.

Why do you not want to include them both inside your originalcode? If doing so causes problems (e.g. because of a jump to that location), you can fix this problem by injection at a different location. Remember, your access to certain values is not only by way of the same sub-routine or section of code, but also via other values that occur within that same data structure...so you have a lot of options.

[Dr.Disrespect]

[hhhuut]

That's because the jump CE uses to execute the injected code is exactly 5 bytes long.

The command "cmp edx,esi" only consists of 2 bytes, the "jne ..." command adds another 2 bytes so we've still one byte less than we need.

That's why CE also takes the third instruction "push 05" into its calculation, so that we've got 6 bytes in sum of which 5 bytes will contain the jump and one will be a "nop" (stands for no operation).

[Dr.Disrespect]

[++METHOS]

By the way, if all you're wanting to do is NOP the instruction in question, then you do not need to jump to a code cave at all...which would eliminate your problem, entirely. You can still write a script for this and assign hotkeys for it as well.

If you're just wanting to change the value, store it or whatever...then you do not need to inject at that instruction either...you can probably injection 10 instructions above/below and still be able to manipulate that value however you like (depending on the size of the sub-routine and where that address is being set etc.).

If there are no viable, alternative instructions that access the value that you're trying to manipulate, you can look at the data structure for the value in question, and inject your code by hooking one of the instructions that access any of the other values within that same data structure, because the offset of your target value will remain relative. Just be sure to use the correct offset for your target value, based on the offset that the new instruction happens to be using. It's not very common, but you may find that a different base address is used at one of the other instructions...but that shouldn't matter much since your target value will always be the same distance away from whatever value you decide to use for your injection point.

[Dr.Disrespect]

[++METHOS]

Yes. If you've completed the CE tutorial, the last step will show you how to use the dissect data structure feature in order to find an ID for filtering out unwanted addresses. The data structure feature is very useful for many reasons - if you haven't used it yet, you should definitely learn how.

[Dr.Disrespect]