Think about that you just reverse engineered a chunk of malware in painstaking element solely to seek out that the malware writer created a barely modified model of the malware the following day. You would not wish to redo all of your arduous work. One strategy to keep away from beginning over from scratch is to make use of code comparability methods to attempt to determine pairs of capabilities within the outdated and new model which might be “the identical.” (I’ve used quotes as a result of “identical” on this occasion is a little bit of a nebulous idea, as we’ll see).
A number of instruments will help in such conditions. A extremely popular industrial instrument is zynamics’ BinDiff, which is now owned by Google and is free. The SEI’s Pharos binary evaluation framework additionally features a code comparability utility referred to as fn2hash. On this SEI Weblog publish, the primary in a two-part collection on hashing, I first clarify the challenges of code comparability and current our resolution to the issue. I then introduce fuzzy hashing, a brand new sort of hashing algorithm that may carry out inexact matching. Lastly, I evaluate the efficiency of fn2hash and a fuzzy hashing algorithm utilizing quite a lot of experiments.
Background
Actual Hashing
fn2hash employs a number of varieties of hashing. Essentially the most generally used is named place impartial code (PIC) hashing. To see why PIC hashing is necessary, we’ll begin by a naive precursor to PIC hashing: merely hashing the instruction bytes of a operate. We’ll name this actual hashing.
For instance, I compiled this straightforward program oo.cpp with g++. Determine 1 exhibits the start of the meeting code for the operate myfunc (full code):
Determine 1: Meeting Code and Bytes from oo.gcc
Actual Bytes
Within the first highlighted line of Determine 1, you possibly can see that the primary instruction is a push r14, which is encoded by the hexadecimal instruction bytes 41 56. If we gather the encoded instruction bytes for each instruction within the operate, we get the next (Determine 2):
Determine 2: Actual Bytes in oo.gcc
We name this sequence the actual bytes of the operate. We are able to hash these bytes to get an actual hash, 62CE2E852A685A8971AF291244A1283A.
Shortcomings of Actual Hashing
The highlighted name at tackle 0x401210 is a relative name, which signifies that the goal is specified as an offset from the present instruction (technically, the following instruction). When you have a look at the instruction bytes for this instruction, it consists of the bytes bb fe ff ff, which is 0xfffffebb in little endian type;. Interpreted as a signed integer worth, that is -325. If we take the tackle of the following instruction (0x401210 + 5 == 0x401215) after which add -325 to it, we get 0x4010d0, which is the tackle of operator new, the goal of the decision. Now we all know that bb fe ff ff is an offset from the following instruction. Such offsets are referred to as relative offsets as a result of they’re relative to the tackle of the following instruction.
I created a barely modified program (oo2.gcc) by including an empty, unused operate earlier than myfunc (Determine 3). You’ll find the disassembly of myfunc for this executable right here.
If we take the precise hash of myfunc on this new executable, we get 05718F65D9AA5176682C6C2D5404CA8D. You’ll discover that is totally different from the hash for myfunc within the first executable, 62CE2E852A685A8971AF291244A1283A. What occurred? Let us take a look at the disassembly.
Determine 3: Meeting Code and Bytes from oo2.gcc
Discover that myfunc moved from 0x401200 to 0x401210, which additionally moved the tackle of the decision instruction from 0x401210 to 0x401220. The decision goal is specified as an offset from the (subsequent) instruction’s tackle, which modified by 0x10 == 16, so the offset bytes for the decision modified from bb fe ff ff (-325) to ab fe ff ff (-341 == -325 – 16). These modifications modify the precise bytes as proven in Determine 4.
Determine 4: Actual Bytes in oo2.gcc
Determine 5 presents a visible comparability. Purple represents bytes which might be solely in oo.gcc, and inexperienced represents bytes in oo2.gcc. The variations are small as a result of the offset is just altering by 0x10, however this is sufficient to break actual hashing.
Determine 5: Distinction Between Actual Bytes in oo.gcc and oo2.gcc
PIC Hashing
The purpose of PIC hashing is to compute a hash or signature of code in a manner that preserves the hash when relocating the code. This requirement is necessary as a result of, as we simply noticed, modifying a program usually leads to small modifications to addresses and offsets, and we do not need these modifications to change the hash. The instinct behind PIC hashing is very straight-forward: Establish offsets and addresses which might be more likely to change if this system is recompiled, comparable to bb fe ff ff, and easily set them to zero earlier than hashing the bytes. That manner, if they modify as a result of the operate is relocated, the operate’s PIC hash will not change.
The next visible diff (Determine 6) exhibits the variations between the precise bytes and the PIC bytes on myfunc in oo.gcc. Purple represents bytes which might be solely within the PIC bytes, and inexperienced represents the precise bytes. As anticipated, the primary change we will see is the byte sequence bb fe ff ff, which is modified to zeros.
Determine 6: Byte Distinction Between PIC Bytes (Purple) and Actual Bytes (Inexperienced)
If we hash the PIC bytes, we get the PIC hash EA4256ECB85EDCF3F1515EACFA734E17. And, as we’d hope, we get the identical PIC hash for myfunc within the barely modified oo2.gcc.
Evaluating the Accuracy of PIC Hashing
The first motivation behind PIC hashing is to detect equivalent code that’s moved to a special location. However what if two items of code are compiled with totally different compilers or totally different compiler flags? What if two capabilities are very related, however one has a line of code eliminated? These modifications would modify the non-offset bytes which might be used within the PIC hash, so it will change the PIC hash of the code. Since we all know that PIC hashing is not going to all the time work, on this part we talk about tips on how to measure the efficiency of PIC hashing and evaluate it to different code comparability methods.
Earlier than we will outline the accuracy of any code comparability approach, we’d like some floor fact that tells us which capabilities are equal. For this weblog publish, we use compiler debug symbols to map operate addresses to their names. Doing so gives us with a floor fact set of capabilities and their names. Additionally, for the needs of this weblog publish, we assume that if two capabilities have the identical identify, they’re “the identical.” (This, clearly, will not be true normally!)
Confusion Matrices
So, for example we now have two related executables, and we wish to consider how nicely PIC hashing can determine equal capabilities throughout each executables. We begin by contemplating all potential pairs of capabilities, the place every pair comprises a operate from every executable. This strategy is named the Cartesian product between the capabilities within the first executable and the capabilities within the second executable. For every operate pair, we use the bottom fact to find out whether or not the capabilities are the identical by seeing if they’ve the identical identify. We then use PIC hashing to foretell whether or not the capabilities are the identical by computing their hashes and seeing if they’re equivalent. There are two outcomes for every willpower, so there are 4 prospects in complete:
- True optimistic (TP): PIC hashing appropriately predicted the capabilities are equal.
- True adverse (TN): PIC hashing appropriately predicted the capabilities are totally different.
- False optimistic (FP): PIC hashing incorrectly predicted the capabilities are equal, however they don’t seem to be.
- False adverse (FN): PIC hashing incorrectly predicted the capabilities are totally different, however they’re equal.
To make it just a little simpler to interpret, we colour the great outcomes inexperienced and the dangerous outcomes purple. We are able to signify these in what is named a confusion matrix:
For instance, here’s a confusion matrix from an experiment by which I exploit PIC hashing to match openssl variations 1.1.1w and 1.1.1v when they’re each compiled in the identical manner. These two variations of openssl are fairly related, so we’d count on that PIC hashing would do nicely as a result of numerous capabilities can be equivalent however shifted to totally different addresses. And, certainly, it does:
Metrics: Accuracy, Precision, and Recall
So, when does PIC hashing work nicely and when does it not? To reply these questions, we will want a better strategy to consider the standard of a confusion matrix as a single quantity. At first look, accuracy looks like probably the most pure metric, which tells us: What number of pairs did hashing predict appropriately? This is the same as
For the above instance, PIC hashing achieved an accuracy of
You may assume 99.9 % is fairly good, however should you look carefully, there’s a refined downside: Most operate pairs are not equal. Based on the bottom fact, there are TP+FN = 344+1=345 equal operate pairs, and TN+FP=118,602+78=118,680 nonequivalent operate pairs. So, if we simply guessed that every one operate pairs had been nonequivalent, we’d nonetheless be proper 118680/(118680+345)=99.9 % of the time. Since accuracy weights all operate pairs equally, it isn’t probably the most acceptable metric.
As a substitute, we wish a metric that emphasizes optimistic outcomes, which on this case are equal operate pairs. This result’s in keeping with our purpose in reverse engineering, as a result of figuring out that two capabilities are equal permits a reverse engineer to switch data from one executable to a different and save time.
Three metrics that focus extra on optimistic instances (i.e., equal capabilities) are precision, recall, and F1 rating:
- Precision: Of the operate pairs hashing declared equal, what number of had been really equal? This is the same as TP/(TP+FP).
- Recall: Of the equal operate pairs, what number of did hashing appropriately declare as equal? This is the same as TP/(TP+FN).
- F1 rating: This can be a single metric that displays each the precision and recall. Particularly, it’s the harmonic imply of the precision and recall, or (2∗Recall∗Precision)/(Recall+Precision). In comparison with the arithmetic imply, the harmonic imply is extra delicate to low values. Because of this if both the precision or recall is low, the F1 rating may also be low.
So, wanting on the instance above, we will compute the precision, recall, and F1 rating. The precision is 344/(344+78) = 0.81, the recall is 344 / (344 + 1) = 0.997, and the F1 rating is 2∗0.81∗0.997/(0.81+0.997)=0.89. PIC hashing is ready to determine 81 % of equal operate pairs, and when it does declare a pair is equal it’s appropriate 99.7 % of the time. This corresponds to an F1 rating of 0.89 out of 1.0, which is fairly good.
Now, you may be questioning how nicely PIC hashing performs when there are extra substantial variations between executables. Let’s have a look at one other experiment. On this one, I evaluate an openssl executable compiled with GCC to 1 compiled with Clang. As a result of GCC and Clang generate meeting code otherwise, we’d count on there to be much more variations.
Here’s a confusion matrix from this experiment:
On this instance, PIC hashing achieved a recall of 23/(23+301)=0.07, and a precision of 23/(23+31) = 0.43. So, PIC hashing can solely determine 7 % of equal operate pairs, however when it does declare a pair is equal it’s appropriate 43 % of the time. This corresponds to an F1 rating of 0.12 out of 1.0, which is fairly dangerous. Think about that you just spent hours reverse engineering the 324 capabilities in one of many executables solely to seek out that PIC hashing was solely capable of determine 23 of them within the different executable. So, you’d be compelled to needlessly reverse engineer the opposite capabilities from scratch. Can we do higher?
The Nice Fuzzy Hashing Debate
Within the subsequent publish on this collection, we are going to introduce a really totally different sort of hashing referred to as fuzzy hashing, and discover whether or not it may well yield higher efficiency than PIC hashing alone. As with PIC hashing, fuzzy hashing reads a sequence of bytes and produces a hash. Not like PIC hashing, nonetheless, while you evaluate two fuzzy hashes, you might be given a similarity worth between 0.0 and 1.0, the place 0.0 means fully dissimilar, and 1.0 means fully related. We’ll current the outcomes of a number of experiments that pit PIC hashing and a well-liked fuzzy hashing algorithm in opposition to one another and look at their strengths and weaknesses within the context of malware reverse-engineering.
