2 title: "Undefined Behavior deserves a better reputation"
3 categories: rust research
4 reddit: /rust/comments/qx168t/undefined_behavior_deserves_a_better_reputation/
7 *This is a cross-post of an [article that I wrote for the SIGPLAN blog](https://blog.sigplan.org/2021/11/18/undefined-behavior-deserves-a-better-reputation/).*
9 "Undefined Behavior" often has a bad reputation. People see it as an excuse compiler writers use to break code, or an excuse by a lazy language designer to not complete the specification and properly define all this behavior.
10 But what, really, is Undefined Behavior, and is it as bad as its reputation?
11 In this blog post, I will look at this topic from a PL perspective, and argue that Undefined Behavior (or UB for short) is a valuable tool in a language designer's toolbox, and that it can be used responsibly to convey more of the programmer's insight about their code to the compiler with the goal of enabling more optimizations.
12 I will also explain why I spent a significant amount of time adding *more* UB to Rust.
18 In the best PL tradition, let us consider an artificial example to demonstrate the benefit of UB.
19 Imagine we want to implement a function that returns the element in the middle of an array.
20 If we are using Rust, we would probably write something like this:
22 fn mid(data: &[i32]) -> Option<i32> {
23 if data.is_empty() { return None; }
24 return Some(data[data.len()/2]);
27 The argument is of type `&[i32]`, which is called a "slice" and consists of a pointer to some array and information about how long the array is.
28 `mid` itself returns an integer wrapped in `Option` (corresponding to `Maybe` in Haskell) to properly signal the case where the array is empty.
29 In the non-empty case, it computes the index in the middle of `data`, and returns that element.
31 Now imagine this function is called in a tight loop in the benchmark for our next paper, so performance *really* matters.
32 Is there any performance improvement we can hope to achieve in this function?
33 It might seem like `mid` already does the absolute minimum amount of work required for the task, but there is some hidden cost in the array access `data[_]`:
34 the compiler has to insert a bounds-check here to ensure that we do not access data beyond the size of the array that `data` points to.
35 But as the programmer we know that bounds-check to be entirely unnecessary, since `data.len()/2` will always be smaller than `data.len()`!
36 Wouldn't it be great if there was a way to tell the compiler about this, such that we can be sure no bounds check happens?
38 Here is one way to accomplish that in Rust:
40 fn mid(data: &[i32]) -> Option<i32> {
41 if data.is_empty() { return None; }
42 match data.get(data.len()/2) {
43 Some(&x) => return Some(x),
44 None => unsafe { unreachable_unchecked() }
48 We are now using the `get` operation to access the array, which returns an `Option` that is `None` for out-of-bounds accesses.
49 And in case we get `None`, we call a special function `unreachable_unchecked` which makes a *binding promise to the compiler* that this piece of code is unreachable.
50 The keyword `unsafe` here indicates that what we are doing is not covered by the type safety guarantees of the language: the compiler will not actually check that the promise we made holds true, it will just trust us on that.
51 (The phrase "unchecked" is a Rust idiom; this is the "unchecked" version of `unreachable`, which inserts a run-time check that safely aborts the program should this code ever be reached -- or, to be more precise, it triggers a Rust panic.)
53 After some inlining, the relevant part of this code looks as follows:
55 let idx = data.len()/2;
56 if idx < data.len() { // Automatically inserted bounds-check.
57 ... // Access the array at `idx`.
59 unreachable_unchecked()
62 Since we told the compiler that the `else` branch is unreachable, it is easy to optimize away the conditional, so we end up with just a direct access to element `idx` in the array.
64 (In fact, Rust provides `get_unchecked` as an alternative to `get` where the caller has to promise that the index is in-bounds, so a Rust programmer would just write `data.get_unchecked(data.len()/2)` to implement `mid` efficiently.)
66 I expect some readers will not be happy with the way I achieved the desired optimization in the initial example, and argue that the compiler should be smart enough to do this automatically.
67 I will get back to this point later; for now, just note that the latest stable version of Rust at the time of writing does [not perform this optimization](https://rust.godbolt.org/z/G34Ezzb9c) (as indicated by the call to `panic_bounds_check`).
69 ## Where is the Undefined Behavior?
71 Hang on, you might say at this point, wasn't this blog post supposed to be about Undefined Behavior?
72 That term did not even appear in the discussion of the example!
73 Indeed, I was a bit sneaky and used different terminology that I think better captures a constructive way to think about Undefined Behavior.
74 In the typical terminology, I would have said that calling the special function `unreachable_unchecked` causes immediate Undefined Behavior.
75 Following the definition in the latest C standard (also shared by C++), the standard "imposes no requirements" on programs that exhibit Undefined Behavior.
76 The compiler can hence basically replace the `else` branch by whatever code it wants, famously including "to make [demons fly out of your nose](http://www.catb.org/jargon/html/N/nasal-demons.html)", but also including just executing the `then` branch instead.
78 This line of reasoning leads to the same result, but it paints an unnecessarily antagonistic picture of compiler writers. It makes it sound like compilers use complicated analyses to detect Undefined Behavior. Once they find UB, they have an excuse to emit broken code and hide behind the standard should anyone complain.
79 This is not what actually happens.
80 As we have seen in our example, the compiler really has no idea if this code has Undefined Behavior or not -- all it does is perform optimizations that are correct *under the extra assumption* that there is no Undefined Behavior.
82 ## UB is a double-edged sword
84 Another reaction you might have is that `unreachable_unchecked` is not a "typical" example of UB.
85 Most people probably associate that term with C or C++, which do not even have `unreachable_unchecked` (though many compilers provide an intrinsic with the same effect, e.g., `__builtin_unreachable` in GCC).
86 So it may seem like I picked a strange example.
87 Shouldn't I be talking about how, say, signed integer overflow is UB?
89 This is the right time to admit that I am *not* going to defend all UB in C/C++.
90 I think UB as a concept is a great idea, and `unreachable_unchecked` is the "most pure" form of UB that shows how it can be used by the programmer to convey extra information to the compiler -- but I also think that C and C++ are massively overusing UB.
91 Of course, it is easy to say this with the benefit of hindsight; the first C compilers were extremely simple and today's use of UB for optimizations only emerged over time.
92 It took a while for the implications of the modern interpretation of UB in the standard to become clear -- and C and C++, being very successful languages, have massive existing codebases which makes it super hard to revise any prior decision.
93 This post is about defending and promoting UB as a concept, not UB in C/C++.
95 Speaking of signed integer overflow, I think this is actually a good example for how to *not* use UB.
96 An innocent-looking `+` turns into a promise of the programmer that this addition will never overflow, but the programmer probably will not carefully do a mental no-overflow proof for every single addition in their program.
97 Instead, `+` could perform overflow checks or well-defined wrap-around, and the language could provide an `unchecked_add` function where overflows are UB.
98 This lets the programmer opt in to providing extra no-overflow promises, to be used in situations where it is really beneficial for performance that the compiler can make this assumption (such as [this example](https://youtu.be/yG1OZ69H_-o?t=2357)).
99 Basically, I am considering this a language (and library) design problem: UB is a sharp knife; when used well it gets the job done better, but it can also hurt a lot when used without enough care.
101 Language and library design is not everything that can be used to tame UB, however.
102 Good tooling can also make a big difference: if programmers can easily run their programs in "UB-checking mode", they can write tests to at least ensure the absence of UB for certain inputs.
103 (Shameless plug: I am working on [Miri](https://github.com/rust-lang/miri/), a tool that provides exactly this for Rust.)
104 Library authors can run their test suites with such a tool, and the tool can also be used exploratively to learn about what exactly is and is not UB in the first place.
105 I think this is absolutely crucial, and language designers should design UB in a way that makes UB-checking tools more feasible.
106 For the examples of UB we have seen so far (`unreachable_unchecked`, `get_unchecked`, and `unchecked_add`), this is obviously trivial.
108 ## How far can we push UB?
110 That said, not all UB is that simple to teach and test.
111 Even Rust, with its benefit of learning from several decades of experience with UB in C and C++, has UB that is a lot more subtle than this. The most glaring example of this is probably UB related to incorrect aliasing of mutable references.
112 (Other, less extreme examples would be UB due to using uninitialized memory, or UB due to data races.)
114 The Rust type system ensures that mutable references never alias any other reference that is currently being used in the program, i.e., they never point to the same memory as any other reference.
115 This is a juicy guarantee for compiler writers, because while reordering memory accesses is often beneficial, it can be very hard to figure out if the transformation is even allowed -- if two accesses alias, then their original order must be preserved.
117 However, `unsafe` code in Rust could easily create aliasing mutable references.
119 We make the programmer promise that they do not do this!
120 This is a lot like saying "the programmer promises that `unreachable_unchecked` is never called", so we can put on our UB lens and say that it is Undefined Behavior to have aliasing mutable references.
122 The devil is of course in the details of defining what exactly this means.
123 [Stacked Borrows](https://plv.mpi-sws.org/rustbelt/stacked-borrows/) (part of [my PhD thesis](https://www.ralfj.de/research/thesis.html) and also described in a series of blog posts: [v1.0](https://www.ralfj.de/blog/2018/11/16/stacked-borrows-implementation.html), [v2.0](https://www.ralfj.de/blog/2019/04/30/stacked-borrows-2.html), [v2.1](https://www.ralfj.de/blog/2019/05/21/stacked-borrows-2.1.html)) goes into all that detail by giving an operational semantics that exactly defines the promises programmers have to make.
124 And that semantics is non-trivial!
125 According to Stacked Borrows, the following code has UB:
127 let x = &mut 42; // Safely create a reference.
128 let xptr = x as *mut i32; // Turn that reference into a raw (unchecked) pointer.
129 let x1 = unsafe { &mut *xptr }; // Turn the pointer back into a reference...
130 let x2 = unsafe { &mut *xptr }; // ...twice, so uniqueness is violated.
131 *x1 = 0; // Undefined Behavior!
133 The reason this code has UB is that creating `x2` makes a promise that this is the unique reference created from `xptr`, so the previously created `x1` is invalidated when `x2` gets created.
134 This means future uses of `x1` are Undefined Behavior.
136 So this raises the question: can we really expect every author of `unsafe` Rust code to internalize Stacked Borrows to the extent that they can faithfully promise to the Rust compiler that their code will comply by this bespoke set of rules?
137 Is it a good idea to interpret `&mut expr` as a promise that all aliasing was carefully checked and this reference is definitely unique?
138 As with other UB, we can help programmers by providing tools; Miri contains an implementation of Stacked Borrows which both helps us to evaluate whether actual Rust code is compatible (or can reasonably be made compatible) with Stacked Borrows, and it helps Rust programmers by giving them a way to at least test for aliasing violations, and to interactively play with the semantics to gain a better understanding.
139 I think that puts us in a pretty good spot overall, but some people still argue that Stacked Borrows goes too far and Rust will end up in a situation similar to the one C and C++ find themselves in -- where too few programmers actually know how to write UB-free code, and a significant amount of the code people rely on exhibits UB.
141 Stacked Borrows is not part of the Rust spec, and is not the final word for aliasing-related UB in Rust.
142 So there is still the chance that future revisions of this model can be made to better align with programmer intuition.
143 The above code might get accepted because `x2` is not actually being used to access memory.
144 Or maybe `&mut expr` should only make such promises when used outside an `unsafe` block -- but then, should adding `unsafe` really change the semantics of the program?
145 As usual, language design is a game of trade-offs.
149 I have presented Undefined Behavior as a tool that enables the programmer to write code that the compiler cannot check for correctness, and argued that -- used responsibly -- it is a useful component in a language designer's toolbox.
151 As I alluded to earlier, the "obvious" alternative would be to make the compiler smarter.
152 However, real programs are typically a lot more complicated than my simple example (which already outsmarts Rust's LLVM backend), and the reasoning required to justify an optimization can become arbitrarily complicated.
153 Language designers should acknowledge that optimizers have their limitations and give programmers the tools they need to help the optimizer.
154 Indeed, I think the fact that Rust combines a clever type checker with the idea of using `unsafe` code for the cases where the type checker is not clever enough is crucial for its success: `unsafe` is not a bug; it is a feature without which Rust would not be able to make systems programming safer in practice.
155 It is also worth mentioning that many languages that we all know and love provide comparable "trusted" operations or annotations, e.g., `Obj.magic` in OCaml or the rewrite rules in GHC.
156 Rust only differs in how prevalent unsafe code is in the ecosystem (and in emphasizing the importance of [encapsulating such code within safe APIs](https://blog.sigplan.org/2019/10/17/what-type-soundness-theorem-do-you-really-want-to-prove/)).
158 In closing, I would like to propose that "Undefined Behavior" might need a rebranding.
159 The term focuses on the negative case, when really all we ever care about as programmers or compiler authors is that programs do *not* have Undefined Behavior.
160 Can we get rid of this double negation?
161 Maybe we should talk about "ensuring Well-Defined Behavior" instead of "avoiding Undefined Behavior".
163 To sum up: most of the time, ensuring Well-Defined Behavior is the responsibility of the type system, but as language designers we should not rule out the idea of sharing that responsibility with the programmer.
165 *Thanks to Anish Athalye and Adrian Sampson for feedback on earlier drafts of this post.*
projects / web.git / blob