Undefined Behavior != Unsafe Programming

Undefined behavior (UB) in C and C++ is a clear and present danger to developers, especially when they are writing code that will execute near a trust boundary. A less well-known kind of undefined behavior exists in the intermediate representation (IR) for most optimizing, ahead-of-time compilers. For example, LLVM IR has undef and poison in addition to true explodes-in-your-face C-style UB. When people become aware of this, a typical reaction is: “Ugh, why? LLVM IR is just as bad as C!” This piece explains why that is not the correct reaction.

Undefined behavior is the result of a design decision: the refusal to systematically trap program errors at one particular level of a system. The responsibility for avoiding these errors is delegated to a higher level of abstraction. For example, it is obvious that a safe programming language can be compiled to machine code, and it is also obvious that the unsafety of machine code in no way compromises the high-level guarantees made by the language implementation. Swift and Rust are compiled to LLVM IR; some of their safety guarantees are enforced by dynamic checks in the emitted code, other guarantees are made through type checking and have no representation at the LLVM level. Either way, UB at the LLVM level is not a problem for, and cannot be detected by, code in the safe subsets of Swift and Rust. Even C can be used safely if some tool in the development environment ensures that it will not execute UB. The L4.verified project does exactly this.

The essence of undefined behavior is the freedom to avoid a forced coupling between error checks and unsafe operations. The checks, once decoupled, can be optimized, for example by being hoisted out of loops or eliminated outright. The remaining unsafe operations can be — in a well-designed IR — mapped onto basic processor operations with little or no overhead. As a concrete example, consider this Swift code:

func add(a : Int, b : Int)->Int {
  return (a & 0xffff) + (b & 0xffff);
}

Although a Swift implementation must trap on integer overflow, the compiler observes that overflow is impossible and emits this LLVM IR:

define i64 @add(i64 %a, i64 %b) {
  %0 = and i64 %a, 65535
  %1 = and i64 %b, 65535
  %2 = add nuw nsw i64 %0, %1
  ret i64 %2
}

Not only has the checked addition operation been lowered to an unchecked one, but also the add instruction has been marked with LLVM’s nsw and nuw attributes, indicating that both signed and unsigned overflow are undefined. In isolation these attributes provide no benefit, but they may enable additional optimizations after this function is inlined. When the Swift benchmark suite is compiled to LLVM, about one in eight addition instructions has an attribute indicating that integer overflow is undefined.

In this particular example the nsw and nuw attributes are redundant since an optimization pass could re-derive the fact that the add cannot overflow. However, in general these attributes and others like them add real value by avoiding the need for potentially expensive static analyses to rediscover known program facts. Also, some facts cannot be rediscovered later, even in principle, since information is lost at some compilation steps.

In summary, undefined behavior in programmer-visible abstractions represents an aggressive and dangerous tradeoff: it sacrifices program correctness in favor of performance and compiler simplicity. On the other hand, UB at lower levels of the system, such as machine code or a compiler IR, is an internal design choice that needn’t have any effect on the programmer-facing parts of the system. This kind of UB simply requires us to accept that safety checks can be usefully factored out of their corresponding unsafe operations to afford efficient execution.

3 thoughts on “Undefined Behavior != Unsafe Programming”

  1. Are you arguing that undefined behaviour should not be a thing in, e.g., C? It sounds like you’re claiming that the fabled sufficiently smart compiler could take code in a C-like language with no UB, and emit something equivalent to what a C compiler would produce given equivalent input in C that correctly avoids UB. In other words, UB really is only there to make the compiler writer’s life easier. I’m not sure if I agree or not; just trying to clarify your position.

  2. Well, here I’m trying to argue that UB in an IR is fine and probably even a good thing.

    It’s at least conceivable that some UB is OK in a programming language if programmers are provided with sufficient tooling to get the UB under control, but in general UB in a PL is quite a bad thing since it mostly leads to silent defects.

  3. Thanks for clarifying this distinction. I expect that when we start adding more optimizations, I will have to walk back my claims of “no undefined behavior” in the Cretonne intermediate language.

    One example is the br_table instruction. Like LLVM’s switch instruction it has a default case, but real indirect jump instructions don’t have built-in bounds checks, so br_table needs to be expanded into an explicit bounds check, a load from a jump table, and an unsafe indirect jump. It’s clear that these low-level parts exhibit undefined behavior when examined individually.

    LLVM does have a few unfortunate cases of undefined behavior that are hard to work around. For example, a loop is assumed to terminate unless it can trivially be proven infinite. This property is probably not something that a higher level in the compiler stack has already proven to be true.

    I hope that we can add undefined behavior to Cretonne IL in a gradual, controlled manner. It would be good to have a subset of the language that can still encode everything you need without making any additional assumptions.

Comments are closed.