Rust Compiler Error Index

E0001

This error suggests that the expression arm corresponding to the noted pattern will never be reached as for all possible values of the expression being matched, one of the preceding patterns will match.

This means that perhaps some of the preceding patterns are too general, this one is too specific or the ordering is incorrect.

For example, the following match block has too many arms:

match foo {
    Some(bar) => {/* ... */}
    None => {/* ... */}
    _ => {/* ... */} // All possible cases have already been handled
}

match blocks have their patterns matched in order, so, for example, putting a wildcard arm above a more specific arm will make the latter arm irrelevant.

Ensure the ordering of the match arm is correct and remove any superfluous arms.

E0002

This error indicates that an empty match expression is invalid because the type it is matching on is non-empty (there exist values of this type). In safe code it is impossible to create an instance of an empty type, so empty match expressions are almost never desired. This error is typically fixed by adding one or more cases to the match expression.

An example of an empty type is enum Empty { }. So, the following will work:

enum Empty {}

fn foo(x: Empty) {
    match x {
        // empty
    }
}

However, this won't:

fn foo(x: Option<String>) {
    match x {
        // empty
    }
}

E0003

Not-a-Number (NaN) values cannot be compared for equality and hence can never match the input to a match expression. So, the following will not compile:

const NAN: f32 = 0.0 / 0.0;

let number = 0.1f32;

match number {
    NAN => { /* ... */ },
    _ => {}
}

To match against NaN values, you should instead use the is_nan() method in a guard, like so:

let number = 0.1f32;

match number {
    x if x.is_nan() => { /* ... */ }
    _ => {}
}

E0004

This error indicates that the compiler cannot guarantee a matching pattern for one or more possible inputs to a match expression. Guaranteed matches are required in order to assign values to match expressions, or alternatively, determine the flow of execution. Erroneous code example:

enum Terminator {
    HastaLaVistaBaby,
    TalkToMyHand,
}

let x = Terminator::HastaLaVistaBaby;

match x { // error: non-exhaustive patterns: `HastaLaVistaBaby` not covered
    Terminator::TalkToMyHand => {}
}

If you encounter this error you must alter your patterns so that every possible value of the input type is matched. For types with a small number of variants (like enums) you should probably cover all cases explicitly. Alternatively, the underscore _ wildcard pattern can be added after all other patterns to match "anything else". Example:

enum Terminator {
    HastaLaVistaBaby,
    TalkToMyHand,
}

let x = Terminator::HastaLaVistaBaby;

match x {
    Terminator::TalkToMyHand => {}
    Terminator::HastaLaVistaBaby => {}
}

// or:

match x {
    Terminator::TalkToMyHand => {}
    _ => {}
}

E0005

Patterns used to bind names must be irrefutable, that is, they must guarantee that a name will be extracted in all cases. Erroneous code example:

let x = Some(1);
let Some(y) = x;
// error: refutable pattern in local binding: `None` not covered

If you encounter this error you probably need to use a match or if let to deal with the possibility of failure. Example:

let x = Some(1);

match x {
    Some(y) => {
        // do something
    },
    None => {}
}

// or:

if let Some(y) = x {
    // do something
}

E0007

This error indicates that the bindings in a match arm would require a value to be moved into more than one location, thus violating unique ownership. Code like the following is invalid as it requires the entire Option<String> to be moved into a variable called op_string while simultaneously requiring the inner String to be moved into a variable called s.

let x = Some("s".to_string());

match x {
    op_string @ Some(s) => {}, // error: cannot bind by-move with sub-bindings
    None => {},
}

See also the error E0303.

E0008

Names bound in match arms retain their type in pattern guards. As such, if a name is bound by move in a pattern, it should also be moved to wherever it is referenced in the pattern guard code. Doing so however would prevent the name from being available in the body of the match arm. Consider the following:

match Some("hi".to_string()) {
    Some(s) if s.len() == 0 => {}, // use s.
    _ => {},
}

The variable s has type String, and its use in the guard is as a variable of type String. The guard code effectively executes in a separate scope to the body of the arm, so the value would be moved into this anonymous scope and therefore becomes unavailable in the body of the arm.

The problem above can be solved by using the ref keyword.

match Some("hi".to_string()) {
    Some(ref s) if s.len() == 0 => {},
    _ => {},
}

Though this example seems innocuous and easy to solve, the problem becomes clear when it encounters functions which consume the value:

struct A{}

impl A {
    fn consume(self) -> usize {
        0
    }
}

fn main() {
    let a = Some(A{});
    match a {
        Some(y) if y.consume() > 0 => {}
        _ => {}
    }
}

In this situation, even the ref keyword cannot solve it, since borrowed content cannot be moved. This problem cannot be solved generally. If the value can be cloned, here is a not-so-specific solution:

#[derive(Clone)]
struct A{}

impl A {
    fn consume(self) -> usize {
        0
    }
}

fn main() {
    let a = Some(A{});
    match a{
        Some(ref y) if y.clone().consume() > 0 => {}
        _ => {}
    }
}

If the value will be consumed in the pattern guard, using its clone will not move its ownership, so the code works.

E0009

In a pattern, all values that don't implement the Copy trait have to be bound the same way. The goal here is to avoid binding simultaneously by-move and by-ref.

This limitation may be removed in a future version of Rust.

Erroneous code example:

struct X { x: (), }

let x = Some((X { x: () }, X { x: () }));
match x {
    Some((y, ref z)) => {}, // error: cannot bind by-move and by-ref in the
                            //        same pattern
    None => panic!()
}

You have two solutions:

Solution #1: Bind the pattern's values the same way.

struct X { x: (), }

let x = Some((X { x: () }, X { x: () }));
match x {
    Some((ref y, ref z)) => {},
    // or Some((y, z)) => {}
    None => panic!()
}

Solution #2: Implement the Copy trait for the X structure.

However, please keep in mind that the first solution should be preferred.

#[derive(Clone, Copy)]
struct X { x: (), }

let x = Some((X { x: () }, X { x: () }));
match x {
    Some((y, ref z)) => {},
    None => panic!()
}

const X: i32 = 42 / 0;
// error: attempted to divide by zero in a constant expression

E0023

A pattern used to match against an enum variant must provide a sub-pattern for each field of the enum variant. This error indicates that a pattern attempted to extract an incorrect number of fields from a variant.

enum Fruit {
    Apple(String, String),
    Pear(u32),
}

Here the Apple variant has two fields, and should be matched against like so:

enum Fruit {
    Apple(String, String),
    Pear(u32),
}

let x = Fruit::Apple(String::new(), String::new());

// Correct.
match x {
    Fruit::Apple(a, b) => {},
    _ => {}
}

Matching with the wrong number of fields has no sensible interpretation:

enum Fruit {
    Apple(String, String),
    Pear(u32),
}

let x = Fruit::Apple(String::new(), String::new());

// Incorrect.
match x {
    Fruit::Apple(a) => {},
    Fruit::Apple(a, b, c) => {},
}

Check how many fields the enum was declared with and ensure that your pattern uses the same number.

E0024

This error indicates that a pattern attempted to extract the fields of an enum variant with no fields. Here's a tiny example of this error:

// This enum has two variants.
enum Number {
    // This variant has no fields.
    Zero,
    // This variant has one field.
    One(u32)
}

// Assuming x is a Number we can pattern match on its contents.
match x {
    Number::Zero(inside) => {},
    Number::One(inside) => {},
}

The pattern match Zero(inside) is incorrect because the Zero variant contains no fields, yet the inside name attempts to bind the first field of the enum.

E0025

Each field of a struct can only be bound once in a pattern. Erroneous code example:

struct Foo {
    a: u8,
    b: u8,
}

fn main(){
    let x = Foo { a:1, b:2 };

    let Foo { a: x, a: y } = x;
    // error: field `a` bound multiple times in the pattern
}

Each occurrence of a field name binds the value of that field, so to fix this error you will have to remove or alter the duplicate uses of the field name. Perhaps you misspelled another field name? Example:

struct Foo {
    a: u8,
    b: u8,
}

fn main(){
    let x = Foo { a:1, b:2 };

    let Foo { a: x, b: y } = x; // ok!
}

E0026

This error indicates that a struct pattern attempted to extract a non-existent field from a struct. Struct fields are identified by the name used before the colon : so struct patterns should resemble the declaration of the struct type being matched.

// Correct matching.
struct Thing {
    x: u32,
    y: u32
}

let thing = Thing { x: 1, y: 2 };

match thing {
    Thing { x: xfield, y: yfield } => {}
}

If you are using shorthand field patterns but want to refer to the struct field by a different name, you should rename it explicitly.

Change this:

struct Thing {
    x: u32,
    y: u32
}

let thing = Thing { x: 0, y: 0 };

match thing {
    Thing { x, z } => {}
}

To this:

struct Thing {
    x: u32,
    y: u32
}

let thing = Thing { x: 0, y: 0 };

match thing {
    Thing { x, y: z } => {}
}

E0027

This error indicates that a pattern for a struct fails to specify a sub-pattern for every one of the struct's fields. Ensure that each field from the struct's definition is mentioned in the pattern, or use .. to ignore unwanted fields.

For example:

struct Dog {
    name: String,
    age: u32,
}

let d = Dog { name: "Rusty".to_string(), age: 8 };

// This is incorrect.
match d {
    Dog { age: x } => {}
}

This is correct (explicit):

struct Dog {
    name: String,
    age: u32,
}

let d = Dog { name: "Rusty".to_string(), age: 8 };

match d {
    Dog { name: ref n, age: x } => {}
}

// This is also correct (ignore unused fields).
match d {
    Dog { age: x, .. } => {}
}

E0029

In a match expression, only numbers and characters can be matched against a range. This is because the compiler checks that the range is non-empty at compile-time, and is unable to evaluate arbitrary comparison functions. If you want to capture values of an orderable type between two end-points, you can use a guard.

// The ordering relation for strings can't be evaluated at compile time,
// so this doesn't work:
match string {
    "hello" ... "world" => {}
    _ => {}
}

// This is a more general version, using a guard:
match string {
    s if s >= "hello" && s <= "world" => {}
    _ => {}
}

E0033

This error indicates that a pointer to a trait type cannot be implicitly dereferenced by a pattern. Every trait defines a type, but because the size of trait implementors isn't fixed, this type has no compile-time size. Therefore, all accesses to trait types must be through pointers. If you encounter this error you should try to avoid dereferencing the pointer.

let trait_obj: &SomeTrait = ...;

// This tries to implicitly dereference to create an unsized local variable.
let &invalid = trait_obj;

// You can call methods without binding to the value being pointed at.
trait_obj.method_one();
trait_obj.method_two();

You can read more about trait objects in the Trait Object section of the Reference:

https://doc.rust-lang.org/reference.html#trait-objects

E0034

The compiler doesn't know what method to call because more than one method has the same prototype. Erroneous code example:

struct Test;

trait Trait1 {
    fn foo();
}

trait Trait2 {
    fn foo();
}

impl Trait1 for Test { fn foo() {} }
impl Trait2 for Test { fn foo() {} }

fn main() {
    Test::foo() // error, which foo() to call?
}

To avoid this error, you have to keep only one of them and remove the others. So let's take our example and fix it:

struct Test;

trait Trait1 {
    fn foo();
}

impl Trait1 for Test { fn foo() {} }

fn main() {
    Test::foo() // and now that's good!
}

However, a better solution would be using fully explicit naming of type and trait:

struct Test;

trait Trait1 {
    fn foo();
}

trait Trait2 {
    fn foo();
}

impl Trait1 for Test { fn foo() {} }
impl Trait2 for Test { fn foo() {} }

fn main() {
    <Test as Trait1>::foo()
}

One last example:

trait F {
    fn m(&self);
}

trait G {
    fn m(&self);
}

struct X;

impl F for X { fn m(&self) { println!("I am F"); } }
impl G for X { fn m(&self) { println!("I am G"); } }

fn main() {
    let f = X;

    F::m(&f); // it displays "I am F"
    G::m(&f); // it displays "I am G"
}

E0035

You tried to give a type parameter where it wasn't needed. Erroneous code example:

struct Test;

impl Test {
    fn method(&self) {}
}

fn main() {
    let x = Test;

    x.method::<i32>(); // Error: Test::method doesn't need type parameter!
}

To fix this error, just remove the type parameter:

struct Test;

impl Test {
    fn method(&self) {}
}

fn main() {
    let x = Test;

    x.method(); // OK, we're good!
}

E0036

This error occurrs when you pass too many or not enough type parameters to a method. Erroneous code example:

struct Test;

impl Test {
    fn method<T>(&self, v: &[T]) -> usize {
        v.len()
    }
}

fn main() {
    let x = Test;
    let v = &[0];

    x.method::<i32, i32>(v); // error: only one type parameter is expected!
}

To fix it, just specify a correct number of type parameters:

struct Test;

impl Test {
    fn method<T>(&self, v: &[T]) -> usize {
        v.len()
    }
}

fn main() {
    let x = Test;
    let v = &[0];

    x.method::<i32>(v); // OK, we're good!
}

Please note on the last example that we could have called method like this:

x.method(v);

E0038

Trait objects like Box<Trait> can only be constructed when certain requirements are satisfied by the trait in question.

Trait objects are a form of dynamic dispatch and use a dynamically sized type for the inner type. So, for a given trait Trait, when Trait is treated as a type, as in Box<Trait>, the inner type is 'unsized'. In such cases the boxed pointer is a 'fat pointer' that contains an extra pointer to a table of methods (among other things) for dynamic dispatch. This design mandates some restrictions on the types of traits that are allowed to be used in trait objects, which are collectively termed as 'object safety' rules.

Attempting to create a trait object for a non object-safe trait will trigger this error.

There are various rules:

The trait cannot require Self: Sized

When Trait is treated as a type, the type does not implement the special Sized trait, because the type does not have a known size at compile time and can only be accessed behind a pointer. Thus, if we have a trait like the following:

trait Foo where Self: Sized {

}

We cannot create an object of type Box<Foo> or &Foo since in this case Self would not be Sized.

Generally, Self : Sized is used to indicate that the trait should not be used as a trait object. If the trait comes from your own crate, consider removing this restriction.

Method references the Self type in its arguments or return type

This happens when a trait has a method like the following:

trait Trait {
    fn foo(&self) -> Self;
}

impl Trait for String {
    fn foo(&self) -> Self {
        "hi".to_owned()
    }
}

impl Trait for u8 {
    fn foo(&self) -> Self {
        1
    }
}

(Note that &self and &mut self are okay, it's additional Self types which cause this problem.)

In such a case, the compiler cannot predict the return type of foo() in a situation like the following:

trait Trait {
    fn foo(&self) -> Self;
}

fn call_foo(x: Box<Trait>) {
    let y = x.foo(); // What type is y?
    // ...
}

If only some methods aren't object-safe, you can add a where Self: Sized bound on them to mark them as explicitly unavailable to trait objects. The functionality will still be available to all other implementers, including Box<Trait> which is itself sized (assuming you impl Trait for Box<Trait>).

trait Trait {
    fn foo(&self) -> Self where Self: Sized;
    // more functions
}

Now, foo() can no longer be called on a trait object, but you will now be allowed to make a trait object, and that will be able to call any object-safe methods. With such a bound, one can still call foo() on types implementing that trait that aren't behind trait objects.

Method has generic type parameters

As mentioned before, trait objects contain pointers to method tables. So, if we have:

trait Trait {
    fn foo(&self);
}

impl Trait for String {
    fn foo(&self) {
        // implementation 1
    }
}

impl Trait for u8 {
    fn foo(&self) {
        // implementation 2
    }
}
// ...

At compile time each implementation of Trait will produce a table containing the various methods (and other items) related to the implementation.

This works fine, but when the method gains generic parameters, we can have a problem.

Usually, generic parameters get monomorphized. For example, if I have

fn foo<T>(x: T) {
    // ...
}

The machine code for foo::<u8>(), foo::<bool>(), foo::<String>(), or any other type substitution is different. Hence the compiler generates the implementation on-demand. If you call foo() with a bool parameter, the compiler will only generate code for foo::<bool>(). When we have additional type parameters, the number of monomorphized implementations the compiler generates does not grow drastically, since the compiler will only generate an implementation if the function is called with unparametrized substitutions (i.e., substitutions where none of the substituted types are themselves parametrized).

However, with trait objects we have to make a table containing every object that implements the trait. Now, if it has type parameters, we need to add implementations for every type that implements the trait, and there could theoretically be an infinite number of types.

For example, with:

trait Trait {
    fn foo<T>(&self, on: T);
    // more methods
}

impl Trait for String {
    fn foo<T>(&self, on: T) {
        // implementation 1
    }
}

impl Trait for u8 {
    fn foo<T>(&self, on: T) {
        // implementation 2
    }
}

// 8 more implementations

Now, if we have the following code:

fn call_foo(thing: Box<Trait>) {
    thing.foo(true); // this could be any one of the 8 types above
    thing.foo(1);
    thing.foo("hello");
}

We don't just need to create a table of all implementations of all methods of Trait, we need to create such a table, for each different type fed to foo(). In this case this turns out to be (10 types implementing Trait)*(3 types being fed to foo()) = 30 implementations!

With real world traits these numbers can grow drastically.

To fix this, it is suggested to use a where Self: Sized bound similar to the fix for the sub-error above if you do not intend to call the method with type parameters:

trait Trait {
    fn foo<T>(&self, on: T) where Self: Sized;
    // more methods
}

If this is not an option, consider replacing the type parameter with another trait object (e.g. if T: OtherTrait, use on: Box<OtherTrait>). If the number of types you intend to feed to this method is limited, consider manually listing out the methods of different types.

Method has no receiver

Methods that do not take a self parameter can't be called since there won't be a way to get a pointer to the method table for them.

trait Foo {
    fn foo() -> u8;
}

This could be called as <Foo as Foo>::foo(), which would not be able to pick an implementation.

Adding a Self: Sized bound to these methods will generally make this compile.

trait Foo {
    fn foo() -> u8 where Self: Sized;
}

The trait cannot use Self as a type parameter in the supertrait listing

This is similar to the second sub-error, but subtler. It happens in situations like the following:

trait Super<A> {}

trait Trait: Super<Self> {
}

struct Foo;

impl Super<Foo> for Foo{}

impl Trait for Foo {}

Here, the supertrait might have methods as follows:

trait Super<A> {
    fn get_a(&self) -> A; // note that this is object safe!
}

If the trait Foo was deriving from something like Super<String> or Super<T> (where Foo itself is Foo<T>), this is okay, because given a type get_a() will definitely return an object of that type.

However, if it derives from Super<Self>, even though Super is object safe, the method get_a() would return an object of unknown type when called on the function. Self type parameters let us make object safe traits no longer safe, so they are forbidden when specifying supertraits.

There's no easy fix for this, generally code will need to be refactored so that you no longer need to derive from Super<Self>.

E0040

It is not allowed to manually call destructors in Rust. It is also not necessary to do this since drop is called automatically whenever a value goes out of scope.

Here's an example of this error:

struct Foo {
    x: i32,
}

impl Drop for Foo {
    fn drop(&mut self) {
        println!("kaboom");
    }
}

fn main() {
    let mut x = Foo { x: -7 };
    x.drop(); // error: explicit use of destructor method
}

E0044

You can't use type parameters on foreign items. Example of erroneous code:

extern { fn some_func<T>(x: T); }

To fix this, replace the type parameter with the specializations that you need:

extern { fn some_func_i32(x: i32); }
extern { fn some_func_i64(x: i64); }

E0045

Rust only supports variadic parameters for interoperability with C code in its FFI. As such, variadic parameters can only be used with functions which are using the C ABI. Examples of erroneous code:

extern "rust-call" { fn foo(x: u8, ...); }

// or

fn foo(x: u8, ...) {}

To fix such code, put them in an extern "C" block:

extern "C" fn foo(x: u8, ...);

Or:

extern "C" {
    fn foo (x: u8, ...);
}

E0046

Items are missing in a trait implementation. Erroneous code example:

trait Foo {
    fn foo();
}

struct Bar;

impl Foo for Bar {}
// error: not all trait items implemented, missing: `foo`

When trying to make some type implement a trait Foo, you must, at minimum, provide implementations for all of Foo's required methods (meaning the methods that do not have default implementations), as well as any required trait items like associated types or constants. Example:

trait Foo {
    fn foo();
}

struct Bar;

impl Foo for Bar {
    fn foo() {} // ok!
}

E0049

This error indicates that an attempted implementation of a trait method has the wrong number of type parameters.

For example, the trait below has a method foo with a type parameter T, but the implementation of foo for the type Bar is missing this parameter:

trait Foo {
    fn foo<T: Default>(x: T) -> Self;
}

struct Bar;

// error: method `foo` has 0 type parameters but its trait declaration has 1
// type parameter
impl Foo for Bar {
    fn foo(x: bool) -> Self { Bar }
}

E0050

This error indicates that an attempted implementation of a trait method has the wrong number of function parameters.

For example, the trait below has a method foo with two function parameters (&self and u8), but the implementation of foo for the type Bar omits the u8 parameter:

trait Foo {
    fn foo(&self, x: u8) -> bool;
}

struct Bar;

// error: method `foo` has 1 parameter but the declaration in trait `Foo::foo`
// has 2
impl Foo for Bar {
    fn foo(&self) -> bool { true }
}

E0053

The parameters of any trait method must match between a trait implementation and the trait definition.

Here are a couple examples of this error:

trait Foo {
    fn foo(x: u16);
    fn bar(&self);
}

struct Bar;

impl Foo for Bar {
    // error, expected u16, found i16
    fn foo(x: i16) { }

    // error, values differ in mutability
    fn bar(&mut self) { }
}

E0054

It is not allowed to cast to a bool. If you are trying to cast a numeric type to a bool, you can compare it with zero instead:

let x = 5;

// Not allowed, won't compile
let x_is_nonzero = x as bool;

let x = 5;

// Ok
let x_is_nonzero = x != 0;

E0055

During a method call, a value is automatically dereferenced as many times as needed to make the value's type match the method's receiver. The catch is that the compiler will only attempt to dereference a number of times up to the recursion limit (which can be set via the recursion_limit attribute).

For a somewhat artificial example:

#![recursion_limit="2"]

struct Foo;

impl Foo {
    fn foo(&self) {}
}

fn main() {
    let foo = Foo;
    let ref_foo = &&Foo;

    // error, reached the recursion limit while auto-dereferencing &&Foo
    ref_foo.foo();
}

One fix may be to increase the recursion limit. Note that it is possible to create an infinite recursion of dereferencing, in which case the only fix is to somehow break the recursion.

E0057

When invoking closures or other implementations of the function traits Fn, FnMut or FnOnce using call notation, the number of parameters passed to the function must match its definition.

An example using a closure:

let f = |x| x * 3;
let a = f();        // invalid, too few parameters
let b = f(4);       // this works!
let c = f(2, 3);    // invalid, too many parameters

A generic function must be treated similarly:

fn foo<F: Fn()>(f: F) {
    f(); // this is valid, but f(3) would not work
}

E0059

The built-in function traits are generic over a tuple of the function arguments. If one uses angle-bracket notation (Fn<(T,), Output=U>) instead of parentheses (Fn(T) -> U) to denote the function trait, the type parameter should be a tuple. Otherwise function call notation cannot be used and the trait will not be implemented by closures.

The most likely source of this error is using angle-bracket notation without wrapping the function argument type into a tuple, for example:

fn foo<F: Fn<i32>>(f: F) -> F::Output { f(3) }

It can be fixed by adjusting the trait bound like this:

fn foo<F: Fn<(i32,)>>(f: F) -> F::Output { f(3) }

Note that (T,) always denotes the type of a 1-tuple containing an element of type T. The comma is necessary for syntactic disambiguation.

E0060

External C functions are allowed to be variadic. However, a variadic function takes a minimum number of arguments. For example, consider C's variadic printf function:

extern crate libc;
use libc::{ c_char, c_int };

extern "C" {
    fn printf(_: *const c_char, ...) -> c_int;
}

Using this declaration, it must be called with at least one argument, so simply calling printf() is invalid. But the following uses are allowed:

unsafe {
    use std::ffi::CString;

    printf(CString::new("test\n").unwrap().as_ptr());
    printf(CString::new("number = %d\n").unwrap().as_ptr(), 3);
    printf(CString::new("%d, %d\n").unwrap().as_ptr(), 10, 5);
}

E0061

The number of arguments passed to a function must match the number of arguments specified in the function signature.

For example, a function like:

fn f(a: u16, b: &str) {}

Must always be called with exactly two arguments, e.g. f(2, "test").

Note that Rust does not have a notion of optional function arguments or variadic functions (except for its C-FFI).

E0062

This error indicates that during an attempt to build a struct or struct-like enum variant, one of the fields was specified more than once. Erroneous code example:

struct Foo {
    x: i32
}

fn main() {
    let x = Foo {
                x: 0,
                x: 0, // error: field `x` specified more than once
            };
}

Each field should be specified exactly one time. Example:

struct Foo {
    x: i32
}

fn main() {
    let x = Foo { x: 0 }; // ok!
}

E0063

This error indicates that during an attempt to build a struct or struct-like enum variant, one of the fields was not provided. Erroneous code example:

struct Foo {
    x: i32,
    y: i32
}

fn main() {
    let x = Foo { x: 0 }; // error: missing field: `y`
}

Each field should be specified exactly once. Example:

struct Foo {
    x: i32,
    y: i32
}

fn main() {
    let x = Foo { x: 0, y: 0 }; // ok!
}

E0067

The left-hand side of a compound assignment expression must be an lvalue expression. An lvalue expression represents a memory location and includes item paths (ie, namespaced variables), dereferences, indexing expressions, and field references.

Let's start with some erroneous code examples:

use std::collections::LinkedList;

// Bad: assignment to non-lvalue expression
LinkedList::new() += 1;

// ...

fn some_func(i: &mut i32) {
    i += 12; // Error : '+=' operation cannot be applied on a reference !
}

And now some working examples:

let mut i : i32 = 0;

i += 12; // Good !

// ...

fn some_func(i: &mut i32) {
    *i += 12; // Good !
}

E0069

The compiler found a function whose body contains a return; statement but whose return type is not (). An example of this is:

// error
fn foo() -> u8 {
    return;
}

Since return; is just like return ();, there is a mismatch between the function's return type and the value being returned.

E0070

The left-hand side of an assignment operator must be an lvalue expression. An lvalue expression represents a memory location and can be a variable (with optional namespacing), a dereference, an indexing expression or a field reference.

More details can be found here: https://doc.rust-lang.org/reference.html#lvalues-rvalues-and-temporaries

Now, we can go further. Here are some erroneous code examples:

struct SomeStruct {
    x: i32,
    y: i32
}

const SOME_CONST : i32 = 12;

fn some_other_func() {}

fn some_function() {
    SOME_CONST = 14; // error : a constant value cannot be changed!
    1 = 3; // error : 1 isn't a valid lvalue!
    some_other_func() = 4; // error : we can't assign value to a function!
    SomeStruct.x = 12; // error : SomeStruct a structure name but it is used
                       // like a variable!
}

And now let's give working examples:

struct SomeStruct {
    x: i32,
    y: i32
}
let mut s = SomeStruct {x: 0, y: 0};

s.x = 3; // that's good !

// ...

fn some_func(x: &mut i32) {
    *x = 12; // that's good !
}

E0071

You tried to use structure-literal syntax to create an item that is not a struct-style structure or enum variant.

Example of erroneous code:

enum Foo { FirstValue(i32) };

let u = Foo::FirstValue { value: 0 }; // error: Foo::FirstValue
                                         // isn't a structure!
// or even simpler, if the name doesn't refer to a structure at all.
let t = u32 { value: 4 }; // error: `u32` does not name a structure.

To fix this, ensure that the name was correctly spelled, and that the correct form of initializer was used.

For example, the code above can be fixed to:

enum Foo {
    FirstValue(i32)
}

fn main() {
    let u = Foo::FirstValue(0i32);

    let t = 4;
}

E0072

When defining a recursive struct or enum, any use of the type being defined from inside the definition must occur behind a pointer (like Box or &). This is because structs and enums must have a well-defined size, and without the pointer, the size of the type would need to be unbounded.

Consider the following erroneous definition of a type for a list of bytes:

// error, invalid recursive struct type
struct ListNode {
    head: u8,
    tail: Option<ListNode>,
}

This type cannot have a well-defined size, because it needs to be arbitrarily large (since we would be able to nest ListNodes to any depth). Specifically,

size of `ListNode` = 1 byte for `head`
                   + 1 byte for the discriminant of the `Option`
                   + size of `ListNode`

One way to fix this is by wrapping ListNode in a Box, like so:

struct ListNode {
    head: u8,
    tail: Option<Box<ListNode>>,
}

This works because Box is a pointer, so its size is well-known.

struct Foo { x: Box<Foo> } // error

struct Foo { x: Option<Box<Foo>> }

#![feature(simd)]

#[simd]
struct Bad<T>(T, T, T);

#![feature(simd)]

#[simd]
struct Good(u32, u32, u32);

E0075

The #[simd] attribute can only be applied to non empty tuple structs, because it doesn't make sense to try to use SIMD operations when there are no values to operate on.

This will cause an error:

#![feature(simd)]

#[simd]
struct Bad;

This will not:

#![feature(simd)]

#[simd]
struct Good(u32);

E0076

When using the #[simd] attribute to automatically use SIMD operations in tuple struct, the types in the struct must all be of the same type, or the compiler will trigger this error.

This will cause an error:

#![feature(simd)]

#[simd]
struct Bad(u16, u32, u32);

This will not:

#![feature(simd)]

#[simd]
struct Good(u32, u32, u32);

E0077

When using the #[simd] attribute on a tuple struct, the elements in the tuple must be machine types so SIMD operations can be applied to them.

This will cause an error:

#![feature(simd)]

#[simd]
struct Bad(String);

This will not:

#![feature(simd)]

#[simd]
struct Good(u32, u32, u32);

E0079

Enum variants which contain no data can be given a custom integer representation. This error indicates that the value provided is not an integer literal and is therefore invalid.

For example, in the following code:

enum Foo {
    Q = "32"
}

We try to set the representation to a string.

There's no general fix for this; if you can work with an integer then just set it to one:

enum Foo {
    Q = 32
}

However if you actually wanted a mapping between variants and non-integer objects, it may be preferable to use a method with a match instead:

enum Foo { Q }
impl Foo {
    fn get_str(&self) -> &'static str {
        match *self {
            Foo::Q => "32",
        }
    }
}

E0080

This error indicates that the compiler was unable to sensibly evaluate an integer expression provided as an enum discriminant. Attempting to divide by 0 or causing integer overflow are two ways to induce this error. For example:

enum Enum {
    X = (1 << 500),
    Y = (1 / 0)
}

Ensure that the expressions given can be evaluated as the desired integer type. See the FFI section of the Reference for more information about using a custom integer type:

https://doc.rust-lang.org/reference.html#ffi-attributes

E0081

Enum discriminants are used to differentiate enum variants stored in memory. This error indicates that the same value was used for two or more variants, making them impossible to tell apart.

// Bad.
enum Enum {
    P = 3,
    X = 3,
    Y = 5
}

// Good.
enum Enum {
    P,
    X = 3,
    Y = 5
}

Note that variants without a manually specified discriminant are numbered from top to bottom starting from 0, so clashes can occur with seemingly unrelated variants.

enum Bad {
    X,
    Y = 0
}

Here X will have already been specified the discriminant 0 by the time Y is encountered, so a conflict occurs.

#[repr(u8)]
enum Thing {
    A = 1024,
    B = 5
}

enum DependsOnPointerSize {
    A = 1 << 32
}

E0084

It is impossible to define an integer type to be used to represent zero-variant enum values because there are no zero-variant enum values. There is no way to construct an instance of the following type using only safe code:

enum Empty {}

E0087

Too many type parameters were supplied for a function. For example:

fn foo<T>() {}

fn main() {
    foo::<f64, bool>(); // error, expected 1 parameter, found 2 parameters
}

The number of supplied parameters must exactly match the number of defined type parameters.

E0088

You gave too many lifetime parameters. Erroneous code example:

fn f() {}

fn main() {
    f::<'static>() // error: too many lifetime parameters provided
}

Please check you give the right number of lifetime parameters. Example:

fn f() {}

fn main() {
    f() // ok!
}

It's also important to note that the Rust compiler can generally determine the lifetime by itself. Example:

struct Foo {
    value: String
}

impl Foo {
    // it can be written like this
    fn get_value<'a>(&'a self) -> &'a str { &self.value }
    // but the compiler works fine with this too:
    fn without_lifetime(&self) -> &str { &self.value }
}

fn main() {
    let f = Foo { value: "hello".to_owned() };

    println!("{}", f.get_value());
    println!("{}", f.without_lifetime());
}

E0089

Not enough type parameters were supplied for a function. For example:

fn foo<T, U>() {}

fn main() {
    foo::<f64>(); // error, expected 2 parameters, found 1 parameter
}

Note that if a function takes multiple type parameters but you want the compiler to infer some of them, you can use type placeholders:

fn foo<T, U>(x: T) {}

fn main() {
    let x: bool = true;
    foo::<f64>(x);    // error, expected 2 parameters, found 1 parameter
    foo::<_, f64>(x); // same as `foo::<bool, f64>(x)`
}

E0091

You gave an unnecessary type parameter in a type alias. Erroneous code example:

type Foo<T> = u32; // error: type parameter `T` is unused
// or:
type Foo<A,B> = Box<A>; // error: type parameter `B` is unused

Please check you didn't write too many type parameters. Example:

type Foo = u32; // ok!
type Foo2<A> = Box<A>; // ok!

E0092

You tried to declare an undefined atomic operation function. Erroneous code example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn atomic_foo(); // error: unrecognized atomic operation
                     //        function
}

Please check you didn't make a mistake in the function's name. All intrinsic functions are defined in librustc_trans/trans/intrinsic.rs and in libcore/intrinsics.rs in the Rust source code. Example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn atomic_fence(); // ok!
}

E0093

You declared an unknown intrinsic function. Erroneous code example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn foo(); // error: unrecognized intrinsic function: `foo`
}

fn main() {
    unsafe {
        foo();
    }
}

Please check you didn't make a mistake in the function's name. All intrinsic functions are defined in librustc_trans/trans/intrinsic.rs and in libcore/intrinsics.rs in the Rust source code. Example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn atomic_fence(); // ok!
}

fn main() {
    unsafe {
        atomic_fence();
    }
}

E0094

You gave an invalid number of type parameters to an intrinsic function. Erroneous code example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn size_of<T, U>() -> usize; // error: intrinsic has wrong number
                                 //        of type parameters
}

Please check that you provided the right number of lifetime parameters and verify with the function declaration in the Rust source code. Example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn size_of<T>() -> usize; // ok!
}

E0101

You hit this error because the compiler lacks the information to determine a type for this expression. Erroneous code example:

fn main() {
    let x = |_| {}; // error: cannot determine a type for this expression
}

You have two possibilities to solve this situation: * Give an explicit definition of the expression * Infer the expression

Examples:

fn main() {
    let x = |_ : u32| {}; // ok!
    // or:
    let x = |_| {};
    x(0u32);
}

E0102

You hit this error because the compiler lacks the information to determine the type of this variable. Erroneous code example:

fn main() {
    // could be an array of anything
    let x = []; // error: cannot determine a type for this local variable
}

To solve this situation, constrain the type of the variable. Examples:

#![allow(unused_variables)]

fn main() {
    let x: [u8; 0] = [];
}

E0106

This error indicates that a lifetime is missing from a type. If it is an error inside a function signature, the problem may be with failing to adhere to the lifetime elision rules (see below).

Here are some simple examples of where you'll run into this error:

struct Foo { x: &bool }        // error
struct Foo<'a> { x: &'a bool } // correct

enum Bar { A(u8), B(&bool), }        // error
enum Bar<'a> { A(u8), B(&'a bool), } // correct

type MyStr = &str;        // error
type MyStr<'a> = &'a str; // correct

Lifetime elision is a special, limited kind of inference for lifetimes in function signatures which allows you to leave out lifetimes in certain cases. For more background on lifetime elision see the book.

The lifetime elision rules require that any function signature with an elided output lifetime must either have

• exactly one input lifetime
• or, multiple input lifetimes, but the function must also be a method with a &self or &mut self receiver

In the first case, the output lifetime is inferred to be the same as the unique input lifetime. In the second case, the lifetime is instead inferred to be the same as the lifetime on &self or &mut self.

Here are some examples of elision errors:

// error, no input lifetimes
fn foo() -> &str { }

// error, `x` and `y` have distinct lifetimes inferred
fn bar(x: &str, y: &str) -> &str { }

// error, `y`'s lifetime is inferred to be distinct from `x`'s
fn baz<'a>(x: &'a str, y: &str) -> &str { }

E0107

This error means that an incorrect number of lifetime parameters were provided for a type (like a struct or enum) or trait.

Some basic examples include:

struct Foo<'a>(&'a str);
enum Bar { A, B, C }

struct Baz<'a> {
    foo: Foo,     // error: expected 1, found 0
    bar: Bar<'a>, // error: expected 0, found 1
}

Here's an example that is currently an error, but may work in a future version of Rust:

struct Foo<'a>(&'a str);

trait Quux { }
impl Quux for Foo { } // error: expected 1, found 0

Lifetime elision in implementation headers was part of the lifetime elision RFC. It is, however, currently unimplemented.

E0109

You tried to give a type parameter to a type which doesn't need it. Erroneous code example:

type X = u32<i32>; // error: type parameters are not allowed on this type

Please check that you used the correct type and recheck its definition. Perhaps it doesn't need the type parameter.

Example:

type X = u32; // this compiles

Note that type parameters for enum-variant constructors go after the variant, not after the enum (Option::None::, not Option::::None).

E0110

You tried to give a lifetime parameter to a type which doesn't need it. Erroneous code example:

type X = u32<'static>; // error: lifetime parameters are not allowed on
                       //        this type

Please check that the correct type was used and recheck its definition; perhaps it doesn't need the lifetime parameter. Example:

type X = u32; // ok!

E0116

You can only define an inherent implementation for a type in the same crate where the type was defined. For example, an impl block as below is not allowed since Vec is defined in the standard library:

impl Vec<u8> { } // error

To fix this problem, you can do either of these things:

• define a trait that has the desired associated functions/types/constants and implement the trait for the type in question
• define a new type wrapping the type and define an implementation on the new type

Note that using the type keyword does not work here because type only introduces a type alias:

type Bytes = Vec<u8>;

impl Bytes { } // error, same as above

E0117

This error indicates a violation of one of Rust's orphan rules for trait implementations. The rule prohibits any implementation of a foreign trait (a trait defined in another crate) where

• the type that is implementing the trait is foreign
• all of the parameters being passed to the trait (if there are any) are also foreign.

Here's one example of this error:

impl Drop for u32 {}

To avoid this kind of error, ensure that at least one local type is referenced by the impl:

pub struct Foo; // you define your type in your crate

impl Drop for Foo { // and you can implement the trait on it!
    // code of trait implementation here
}

impl From<Foo> for i32 { // or you use a type from your crate as
                         // a type parameter
    fn from(i: Foo) -> i32 {
        0
    }
}

Alternatively, define a trait locally and implement that instead:

trait Bar {
    fn get(&self) -> usize;
}

impl Bar for u32 {
    fn get(&self) -> usize { 0 }
}

For information on the design of the orphan rules, see RFC 1023.

E0118

You're trying to write an inherent implementation for something which isn't a struct nor an enum. Erroneous code example:

impl (u8, u8) { // error: no base type found for inherent implementation
    fn get_state(&self) -> String {
        // ...
    }
}

To fix this error, please implement a trait on the type or wrap it in a struct. Example:

// we create a trait here
trait LiveLongAndProsper {
    fn get_state(&self) -> String;
}

// and now you can implement it on (u8, u8)
impl LiveLongAndProsper for (u8, u8) {
    fn get_state(&self) -> String {
        "He's dead, Jim!".to_owned()
    }
}

Alternatively, you can create a newtype. A newtype is a wrapping tuple-struct. For example, NewType is a newtype over Foo in struct NewType(Foo). Example:

struct TypeWrapper((u8, u8));

impl TypeWrapper {
    fn get_state(&self) -> String {
        "Fascinating!".to_owned()
    }
}

E0119

There are conflicting trait implementations for the same type. Example of erroneous code:

trait MyTrait {
    fn get(&self) -> usize;
}

impl<T> MyTrait for T {
    fn get(&self) -> usize { 0 }
}

struct Foo {
    value: usize
}

impl MyTrait for Foo { // error: conflicting implementations of trait
                       //        `MyTrait` for type `Foo`
    fn get(&self) -> usize { self.value }
}

When looking for the implementation for the trait, the compiler finds both the impl<T> MyTrait for T where T is all types and the impl MyTrait for Foo. Since a trait cannot be implemented multiple times, this is an error. So, when you write:

trait MyTrait {
    fn get(&self) -> usize;
}

impl<T> MyTrait for T {
    fn get(&self) -> usize { 0 }
}

This makes the trait implemented on all types in the scope. So if you try to implement it on another one after that, the implementations will conflict. Example:

trait MyTrait {
    fn get(&self) -> usize;
}

impl<T> MyTrait for T {
    fn get(&self) -> usize { 0 }
}

struct Foo;

fn main() {
    let f = Foo;

    f.get(); // the trait is implemented so we can use it
}

E0120

An attempt was made to implement Drop on a trait, which is not allowed: only structs and enums can implement Drop. An example causing this error:

trait MyTrait {}

impl Drop for MyTrait {
    fn drop(&mut self) {}
}

A workaround for this problem is to wrap the trait up in a struct, and implement Drop on that. An example is shown below:

trait MyTrait {}
struct MyWrapper<T: MyTrait> { foo: T }

impl <T: MyTrait> Drop for MyWrapper<T> {
    fn drop(&mut self) {}
}

Alternatively, wrapping trait objects requires something like the following:

trait MyTrait {}

//or Box<MyTrait>, if you wanted an owned trait object
struct MyWrapper<'a> { foo: &'a MyTrait }

impl <'a> Drop for MyWrapper<'a> {
    fn drop(&mut self) {}
}

E0121

In order to be consistent with Rust's lack of global type inference, type placeholders are disallowed by design in item signatures.

Examples of this error include:

fn foo() -> _ { 5 } // error, explicitly write out the return type instead

static BAR: _ = "test"; // error, explicitly write out the type instead

E0122

An attempt was made to add a generic constraint to a type alias. While Rust will allow this with a warning, it will not currently enforce the constraint. Consider the example below:

trait Foo{}

type MyType<R: Foo> = (R, ());

fn main() {
    let t: MyType<u32>;
}

We're able to declare a variable of type MyType<u32>, despite the fact that u32 does not implement Foo. As a result, one should avoid using generic constraints in concert with type aliases.

E0124

You declared two fields of a struct with the same name. Erroneous code example:

struct Foo {
    field1: i32,
    field1: i32, // error: field is already declared
}

Please verify that the field names have been correctly spelled. Example:

struct Foo {
    field1: i32,
    field2: i32, // ok!
}

E0128

Type parameter defaults can only use parameters that occur before them. Erroneous code example:

struct Foo<T=U, U=()> {
    field1: T,
    filed2: U,
}
// error: type parameters with a default cannot use forward declared
// identifiers

Since type parameters are evaluated in-order, you may be able to fix this issue by doing:

struct Foo<U=(), T=U> {
    field1: T,
    filed2: U,
}

Please also verify that this wasn't because of a name-clash and rename the type parameter if so.

E0130

You declared a pattern as an argument in a foreign function declaration. Erroneous code example:

extern {
    fn foo((a, b): (u32, u32)); // error: patterns aren't allowed in foreign
                                //        function declarations
}

Please replace the pattern argument with a regular one. Example:

struct SomeStruct {
    a: u32,
    b: u32,
}

extern {
    fn foo(s: SomeStruct); // ok!
}

Or:

extern {
    fn foo(a: (u32, u32)); // ok!
}

E0131

It is not possible to define main with type parameters, or even with function parameters. When main is present, it must take no arguments and return (). Erroneous code example:

fn main<T>() { // error: main function is not allowed to have type parameters
}

E0132

It is not possible to declare type parameters on a function that has the start attribute. Such a function must have the following type signature:

fn(isize, *const *const u8) -> isize;

E0133

Using unsafe functionality is potentially dangerous and disallowed by safety checks. Examples:

• Dereferencing raw pointers
• Calling functions via FFI
• Calling functions marked unsafe

These safety checks can be relaxed for a section of the code by wrapping the unsafe instructions with an unsafe block. For instance:

unsafe fn f() { return; }

fn main() {
    unsafe { f(); }
}

See also https://doc.rust-lang.org/book/unsafe.html

E0136

A binary can only have one entry point, and by default that entry point is the function main(). If there are multiple such functions, please rename one.

E0137

This error indicates that the compiler found multiple functions with the #[main] attribute. This is an error because there must be a unique entry point into a Rust program.

E0138

This error indicates that the compiler found multiple functions with the #[start] attribute. This is an error because there must be a unique entry point into a Rust program.

struct Foo<T>(Vec<T>);

fn foo<T>(x: Vec<T>) {
    // we are transmuting between Vec<T> and Foo<T> here
    let y: Foo<T> = unsafe { transmute(x) };
    // do something with y
}

struct Foo<T>(Vec<T>);

trait MyTransmutableType {
    fn transmute(Vec<Self>) -> Foo<Self>;
}

impl MyTransmutableType for u8 {
    fn transmute(x: Foo<u8>) -> Vec<u8> {
        transmute(x)
    }
}

impl MyTransmutableType for String {
    fn transmute(x: Foo<String>) -> Vec<String> {
        transmute(x)
    }
}

// ... more impls for the types you intend to transmute

fn foo<T: MyTransmutableType>(x: Vec<T>) {
    let y: Foo<T> = <T as MyTransmutableType>::transmute(x);
    // do something with y
}

ptr::read(&v as *const _ as *const SomeType) // `v` transmuted to `SomeType`

E0152

Lang items are already implemented in the standard library. Unless you are writing a free-standing application (e.g. a kernel), you do not need to provide them yourself.

You can build a free-standing crate by adding #![no_std] to the crate attributes:

#![no_std]

See also https://doc.rust-lang.org/book/no-stdlib.html

fn f() {
    // Variable declaration before import
    let x = 0;
    use std::io::Read;
    // ...
}

fn f() {
    use std::io::Read;
    let x = 0;
    // ...
}

E0158

const and static mean different things. A const is a compile-time constant, an alias for a literal value. This property means you can match it directly within a pattern.

The static keyword, on the other hand, guarantees a fixed location in memory. This does not always mean that the value is constant. For example, a global mutex can be declared static as well.

If you want to match against a static, consider using a guard instead:

static FORTY_TWO: i32 = 42;

match Some(42) {
    Some(x) if x == FORTY_TWO => {}
    _ => {}
}

E0162

An if-let pattern attempts to match the pattern, and enters the body if the match was successful. If the match is irrefutable (when it cannot fail to match), use a regular let-binding instead. For instance:

struct Irrefutable(i32);
let irr = Irrefutable(0);

// This fails to compile because the match is irrefutable.
if let Irrefutable(x) = irr {
    // This body will always be executed.
    foo(x);
}

Try this instead:

struct Irrefutable(i32);
let irr = Irrefutable(0);

let Irrefutable(x) = irr;
foo(x);

E0163

This error means that an attempt was made to match an enum variant as a struct type when the variant isn't a struct type:

enum Foo { B(u32) }

fn bar(foo: Foo) -> u32 {
    match foo {
        B{i} => i, // error E0163
    }
}

Try using () instead:

enum Foo { B(u32) }

fn bar(foo: Foo) -> u32 {
    match foo {
        Foo::B(i) => i,
    }
}

E0164

This error means that an attempt was made to match a struct type enum variant as a non-struct type:

enum Foo { B { i: u32 } }

fn bar(foo: Foo) -> u32 {
    match foo {
        Foo::B(i) => i, // error E0164
    }
}

Try using {} instead:

enum Foo { B { i: u32 } }

fn bar(foo: Foo) -> u32 {
    match foo {
        Foo::B{i} => i,
    }
}

E0165

A while-let pattern attempts to match the pattern, and enters the body if the match was successful. If the match is irrefutable (when it cannot fail to match), use a regular let-binding inside a loop instead. For instance:

struct Irrefutable(i32);
let irr = Irrefutable(0);

// This fails to compile because the match is irrefutable.
while let Irrefutable(x) = irr {
    ...
}

Try this instead:

struct Irrefutable(i32); let irr = Irrefutable(0);

loop { let Irrefutable(x) = irr; ... } ```

E0166

This error means that the compiler found a return expression in a function marked as diverging. A function diverges if it has ! in the place of the return type in its signature. For example:

fn foo() -> ! { return; } // error

For a function that diverges, every control path in the function must never return, for example with a loop that never breaks or a call to another diverging function (such as panic!()).

E0170

Enum variants are qualified by default. For example, given this type:

enum Method {
    GET,
    POST,
}

You would match it using:

enum Method {
    GET,
    POST,
}

let m = Method::GET;

match m {
    Method::GET => {},
    Method::POST => {},
}

If you don't qualify the names, the code will bind new variables named "GET" and "POST" instead. This behavior is likely not what you want, so rustc warns when that happens.

Qualified names are good practice, and most code works well with them. But if you prefer them unqualified, you can import the variants into scope:

use Method::*;
enum Method { GET, POST }

If you want others to be able to import variants from your module directly, use pub use:

pub use Method::*;
enum Method { GET, POST }

E0172

This error means that an attempt was made to specify the type of a variable with a combination of a concrete type and a trait. Consider the following example:

fn foo(bar: i32+std::fmt::Display) {}

The code is trying to specify that we want to receive a signed 32-bit integer which also implements Display. This doesn't make sense: when we pass i32, a concrete type, it implicitly includes all of the traits that it implements. This includes Display, Debug, Clone, and a host of others.

If i32 implements the trait we desire, there's no need to specify the trait separately. If it does not, then we need to impl the trait for i32 before passing it into foo. Either way, a fixed definition for foo will look like the following:

fn foo(bar: i32) {}

To learn more about traits, take a look at the Book:

https://doc.rust-lang.org/book/traits.html

E0178

In types, the + type operator has low precedence, so it is often necessary to use parentheses.

For example:

trait Foo {}

struct Bar<'a> {
    w: &'a Foo + Copy,   // error, use &'a (Foo + Copy)
    x: &'a Foo + 'a,     // error, use &'a (Foo + 'a)
    y: &'a mut Foo + 'a, // error, use &'a mut (Foo + 'a)
    z: fn() -> Foo + 'a, // error, use fn() -> (Foo + 'a)
}

More details can be found in RFC 438.

E0184

Explicitly implementing both Drop and Copy for a type is currently disallowed. This feature can make some sense in theory, but the current implementation is incorrect and can lead to memory unsafety (see issue #20126), so it has been disabled for now.

E0185

An associated function for a trait was defined to be static, but an implementation of the trait declared the same function to be a method (i.e. to take a self parameter).

Here's an example of this error:

trait Foo {
    fn foo();
}

struct Bar;

impl Foo for Bar {
    // error, method `foo` has a `&self` declaration in the impl, but not in
    // the trait
    fn foo(&self) {}
}

E0186

An associated function for a trait was defined to be a method (i.e. to take a self parameter), but an implementation of the trait declared the same function to be static.

Here's an example of this error:

trait Foo {
    fn foo(&self);
}

struct Bar;

impl Foo for Bar {
    // error, method `foo` has a `&self` declaration in the trait, but not in
    // the impl
    fn foo() {}
}

E0191

Trait objects need to have all associated types specified. Erroneous code example:

trait Trait {
    type Bar;
}

type Foo = Trait; // error: the value of the associated type `Bar` (from
                  //        the trait `Trait`) must be specified

Please verify you specified all associated types of the trait and that you used the right trait. Example:

trait Trait {
    type Bar;
}

type Foo = Trait<Bar=i32>; // ok!

E0192

Negative impls are only allowed for traits with default impls. For more information see the opt-in builtin traits RFC.

trait Foo {
    fn bar(&self);
}

#[derive(Copy,Clone)]
struct Wrapper<T> {
    Wrapped: T
}

impl Foo for Wrapper<u32> where Wrapper<u32>: Clone {
    fn bar(&self) { }
}

trait Foo {
    fn bar(&self);
}

#[derive(Copy,Clone)]
struct Wrapper<T> {
    Wrapped: T
}
impl <T> Foo for Wrapper<T> where Wrapper<T>: Clone {
    fn bar(&self) { }
}

E0194

A type parameter was declared which shadows an existing one. An example of this error:

trait Foo<T> {
    fn do_something(&self) -> T;
    fn do_something_else<T: Clone>(&self, bar: T);
}

In this example, the trait Foo and the trait method do_something_else both define a type parameter T. This is not allowed: if the method wishes to define a type parameter, it must use a different name for it.

E0195

Your method's lifetime parameters do not match the trait declaration. Erroneous code example:

trait Trait {
    fn bar<'a,'b:'a>(x: &'a str, y: &'b str);
}

struct Foo;

impl Trait for Foo {
    fn bar<'a,'b>(x: &'a str, y: &'b str) {
    // error: lifetime parameters or bounds on method `bar`
    // do not match the trait declaration
    }
}

The lifetime constraint 'b for bar() implementation does not match the trait declaration. Ensure lifetime declarations match exactly in both trait declaration and implementation. Example:

trait Trait {
    fn t<'a,'b:'a>(x: &'a str, y: &'b str);
}

struct Foo;

impl Trait for Foo {
    fn t<'a,'b:'a>(x: &'a str, y: &'b str) { // ok!
    }
}

E0197

Inherent implementations (one that do not implement a trait but provide methods associated with a type) are always safe because they are not implementing an unsafe trait. Removing the unsafe keyword from the inherent implementation will resolve this error.

struct Foo;

// this will cause this error
unsafe impl Foo { }
// converting it to this will fix it
impl Foo { }

E0198

A negative implementation is one that excludes a type from implementing a particular trait. Not being able to use a trait is always a safe operation, so negative implementations are always safe and never need to be marked as unsafe.

#![feature(optin_builtin_traits)]

struct Foo;

// unsafe is unnecessary
unsafe impl !Clone for Foo { }

This will compile:

#![feature(optin_builtin_traits)]

struct Foo;

trait Enterprise {}

impl Enterprise for .. { }

impl !Enterprise for Foo { }

Please note that negative impls are only allowed for traits with default impls.

E0199

Safe traits should not have unsafe implementations, therefore marking an implementation for a safe trait unsafe will cause a compiler error. Removing the unsafe marker on the trait noted in the error will resolve this problem.

struct Foo;

trait Bar { }

// this won't compile because Bar is safe
unsafe impl Bar for Foo { }
// this will compile
impl Bar for Foo { }

E0200

Unsafe traits must have unsafe implementations. This error occurs when an implementation for an unsafe trait isn't marked as unsafe. This may be resolved by marking the unsafe implementation as unsafe.

struct Foo;

unsafe trait Bar { }

// this won't compile because Bar is unsafe and impl isn't unsafe
impl Bar for Foo { }
// this will compile
unsafe impl Bar for Foo { }

E0201

It is an error to define two associated items (like methods, associated types, associated functions, etc.) with the same identifier.

For example:

struct Foo(u8);

impl Foo {
    fn bar(&self) -> bool { self.0 > 5 }
    fn bar() {} // error: duplicate associated function
}

trait Baz {
    type Quux;
    fn baz(&self) -> bool;
}

impl Baz for Foo {
    type Quux = u32;

    fn baz(&self) -> bool { true }

    // error: duplicate method
    fn baz(&self) -> bool { self.0 > 5 }

    // error: duplicate associated type
    type Quux = u32;
}

Note, however, that items with the same name are allowed for inherent impl blocks that don't overlap:

struct Foo<T>(T);

impl Foo<u8> {
    fn bar(&self) -> bool { self.0 > 5 }
}

impl Foo<bool> {
    fn bar(&self) -> bool { self.0 }
}

E0202

Inherent associated types were part of RFC 195 but are not yet implemented. See the tracking issue for the status of this implementation.

E0204

An attempt to implement the Copy trait for a struct failed because one of the fields does not implement Copy. To fix this, you must implement Copy for the mentioned field. Note that this may not be possible, as in the example of

struct Foo {
    foo : Vec<u32>,
}

impl Copy for Foo { }

This fails because Vec<T> does not implement Copy for any T.

Here's another example that will fail:

#[derive(Copy)]
struct Foo<'a> {
    ty: &'a mut bool,
}

This fails because &mut T is not Copy, even when T is Copy (this differs from the behavior for &T, which is always Copy).

E0205

An attempt to implement the Copy trait for an enum failed because one of the variants does not implement Copy. To fix this, you must implement Copy for the mentioned variant. Note that this may not be possible, as in the example of

enum Foo {
    Bar(Vec<u32>),
    Baz,
}

impl Copy for Foo { }

This fails because Vec<T> does not implement Copy for any T.

Here's another example that will fail:

#[derive(Copy)]
enum Foo<'a> {
    Bar(&'a mut bool),
    Baz
}

This fails because &mut T is not Copy, even when T is Copy (this differs from the behavior for &T, which is always Copy).

E0206

You can only implement Copy for a struct or enum. Both of the following examples will fail, because neither i32 (primitive type) nor &'static Bar (reference to Bar) is a struct or enum:

type Foo = i32;
impl Copy for Foo { } // error

#[derive(Copy, Clone)]
struct Bar;
impl Copy for &'static Bar { } // error

E0207

Any type parameter or lifetime parameter of an impl must meet at least one of the following criteria:

• it appears in the self type of the impl
• for a trait impl, it appears in the trait reference
• it is bound as an associated type

Error example 1

Suppose we have a struct Foo and we would like to define some methods for it. The following definition leads to a compiler error:

struct Foo;

impl<T: Default> Foo {
// error: the type parameter `T` is not constrained by the impl trait, self
// type, or predicates [E0207]
    fn get(&self) -> T {
        <T as Default>::default()
    }
}

The problem is that the parameter T does not appear in the self type (Foo) of the impl. In this case, we can fix the error by moving the type parameter from the impl to the method get:

struct Foo;

// Move the type parameter from the impl to the method
impl Foo {
    fn get<T: Default>(&self) -> T {
        <T as Default>::default()
    }
}

Error example 2

As another example, suppose we have a Maker trait and want to establish a type FooMaker that makes Foos:

trait Maker {
    type Item;
    fn make(&mut self) -> Self::Item;
}

struct Foo<T> {
    foo: T
}

struct FooMaker;

impl<T: Default> Maker for FooMaker {
// error: the type parameter `T` is not constrained by the impl trait, self
// type, or predicates [E0207]
    type Item = Foo<T>;

    fn make(&mut self) -> Foo<T> {
        Foo { foo: <T as Default>::default() }
    }
}

This fails to compile because T does not appear in the trait or in the implementing type.

One way to work around this is to introduce a phantom type parameter into FooMaker, like so:

use std::marker::PhantomData;

trait Maker {
    type Item;
    fn make(&mut self) -> Self::Item;
}

struct Foo<T> {
    foo: T
}

// Add a type parameter to `FooMaker`
struct FooMaker<T> {
    phantom: PhantomData<T>,
}

impl<T: Default> Maker for FooMaker<T> {
    type Item = Foo<T>;

    fn make(&mut self) -> Foo<T> {
        Foo {
            foo: <T as Default>::default(),
        }
    }
}

Another way is to do away with the associated type in Maker and use an input type parameter instead:

// Use a type parameter instead of an associated type here
trait Maker<Item> {
    fn make(&mut self) -> Item;
}

struct Foo<T> {
    foo: T
}

struct FooMaker;

impl<T: Default> Maker<Foo<T>> for FooMaker {
    fn make(&mut self) -> Foo<T> {
        Foo { foo: <T as Default>::default() }
    }
}

Additional information

For more information, please see RFC 447.

E0210

This error indicates a violation of one of Rust's orphan rules for trait implementations. The rule concerns the use of type parameters in an implementation of a foreign trait (a trait defined in another crate), and states that type parameters must be "covered" by a local type. To understand what this means, it is perhaps easiest to consider a few examples.

If ForeignTrait is a trait defined in some external crate foo, then the following trait impl is an error:

extern crate foo;
use foo::ForeignTrait;

impl<T> ForeignTrait for T { } // error

To work around this, it can be covered with a local type, MyType:

struct MyType<T>(T);
impl<T> ForeignTrait for MyType<T> { } // Ok

Please note that a type alias is not sufficient.

For another example of an error, suppose there's another trait defined in foo named ForeignTrait2 that takes two type parameters. Then this impl results in the same rule violation:

struct MyType2;
impl<T> ForeignTrait2<T, MyType<T>> for MyType2 { } // error

The reason for this is that there are two appearances of type parameter T in the impl header, both as parameters for ForeignTrait2. The first appearance is uncovered, and so runs afoul of the orphan rule.

Consider one more example:

impl<T> ForeignTrait2<MyType<T>, T> for MyType2 { } // Ok

This only differs from the previous impl in that the parameters T and MyType<T> for ForeignTrait2 have been swapped. This example does not violate the orphan rule; it is permitted.

To see why that last example was allowed, you need to understand the general rule. Unfortunately this rule is a bit tricky to state. Consider an impl:

impl<P1, ..., Pm> ForeignTrait<T1, ..., Tn> for T0 { ... }

where P1, ..., Pm are the type parameters of the impl and T0, ..., Tn are types. One of the types T0, ..., Tn must be a local type (this is another orphan rule, see the explanation for E0117). Let i be the smallest integer such that Ti is a local type. Then no type parameter can appear in any of the Tj for j < i.

For information on the design of the orphan rules, see RFC 1023.

E0211

You used a function or type which doesn't fit the requirements for where it was used. Erroneous code examples:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn size_of<T>(); // error: intrinsic has wrong type
}

// or:

fn main() -> i32 { 0 }
// error: main function expects type: `fn() {main}`: expected (), found i32

// or:

let x = 1u8;
match x {
    0u8...3i8 => (),
    // error: mismatched types in range: expected u8, found i8
    _ => ()
}

// or:

use std::rc::Rc;
struct Foo;

impl Foo {
    fn x(self: Rc<Foo>) {}
    // error: mismatched self type: expected `Foo`: expected struct
    //        `Foo`, found struct `alloc::rc::Rc`
}

For the first code example, please check the function definition. Example:

#![feature(intrinsics)]

extern "rust-intrinsic" {
    fn size_of<T>() -> usize; // ok!
}

The second case example is a bit particular : the main function must always have this definition:

fn main();

They never take parameters and never return types.

For the third example, when you match, all patterns must have the same type as the type you're matching on. Example:

let x = 1u8;

match x {
    0u8...3u8 => (), // ok!
    _ => ()
}

And finally, for the last example, only Box<Self>, &Self, Self, or &mut Self work as explicit self parameters. Example:

struct Foo;

impl Foo {
    fn x(self: Box<Foo>) {} // ok!
}

E0214

A generic type was described using parentheses rather than angle brackets. For example:

fn main() {
    let v: Vec(&str) = vec!["foo"];
}

This is not currently supported: v should be defined as Vec<&str>. Parentheses are currently only used with generic types when defining parameters for Fn-family traits.

E0220

You used an associated type which isn't defined in the trait. Erroneous code example:

trait Trait {
    type Bar;
}

type Foo = Trait<F=i32>; // error: associated type `F` not found for
                         //        `Trait`

Please verify you used the right trait or you didn't misspell the associated type name. Example:

trait Trait {
    type Bar;
}

type Foo = Trait<Bar=i32>; // ok!

E0221

An attempt was made to retrieve an associated type, but the type was ambiguous. For example:

trait T1 {}
trait T2 {}

trait Foo {
    type A: T1;
}

trait Bar : Foo {
    type A: T2;
    fn do_something() {
        let _: Self::A;
    }
}

In this example, Foo defines an associated type A. Bar inherits that type from Foo, and defines another associated type of the same name. As a result, when we attempt to use Self::A, it's ambiguous whether we mean the A defined by Foo or the one defined by Bar.

There are two options to work around this issue. The first is simply to rename one of the types. Alternatively, one can specify the intended type using the following syntax:

trait T1 {}
trait T2 {}

trait Foo {
    type A: T1;
}

trait Bar : Foo {
    type A: T2;
    fn do_something() {
        let _: <Self as Bar>::A;
    }
}

E0223

An attempt was made to retrieve an associated type, but the type was ambiguous. For example:

trait MyTrait {type X; }

fn main() {
    let foo: MyTrait::X;
}

The problem here is that we're attempting to take the type of X from MyTrait. Unfortunately, the type of X is not defined, because it's only made concrete in implementations of the trait. A working version of this code might look like:

trait MyTrait {type X; }
struct MyStruct;

impl MyTrait for MyStruct {
    type X = u32;
}

fn main() {
    let foo: <MyStruct as MyTrait>::X;
}

This syntax specifies that we want the X type from MyTrait, as made concrete in MyStruct. The reason that we cannot simply use MyStruct::X is that MyStruct might implement two different traits with identically-named associated types. This syntax allows disambiguation between the two.

E0225

You attempted to use multiple types as bounds for a closure or trait object. Rust does not currently support this. A simple example that causes this error:

fn main() {
    let _: Box<std::io::Read + std::io::Write>;
}

Builtin traits are an exception to this rule: it's possible to have bounds of one non-builtin type, plus any number of builtin types. For example, the following compiles correctly:

fn main() {
    let _: Box<std::io::Read + Send + Sync>;
}

E0229

An associated type binding was done outside of the type parameter declaration and where clause. Erroneous code example:

pub trait Foo {
    type A;
    fn boo(&self) -> <Self as Foo>::A;
}

struct Bar;

impl Foo for isize {
    type A = usize;
    fn boo(&self) -> usize { 42 }
}

fn baz<I>(x: &<I as Foo<A=Bar>>::A) {}
// error: associated type bindings are not allowed here

To solve this error, please move the type bindings in the type parameter declaration:

fn baz<I: Foo<A=Bar>>(x: &<I as Foo>::A) {} // ok!

Or in the where clause:

fn baz<I>(x: &<I as Foo>::A) where I: Foo<A=Bar> {}

E0232

The attribute must have a value. Erroneous code example:

#![feature(on_unimplemented)]

#[rustc_on_unimplemented] // error: this attribute must have a value
trait Bar {}

Please supply the missing value of the attribute. Example:

#![feature(on_unimplemented)]

#[rustc_on_unimplemented = "foo"] // ok!
trait Bar {}

E0243

This error indicates that not enough type parameters were found in a type or trait.

For example, the Foo struct below is defined to be generic in T, but the type parameter is missing in the definition of Bar:

struct Foo<T> { x: T }

struct Bar { x: Foo }

E0244

This error indicates that too many type parameters were found in a type or trait.

For example, the Foo struct below has no type parameters, but is supplied with two in the definition of Bar:

struct Foo { x: bool }

struct Bar<S, T> { x: Foo<S, T> }

E0248

This error indicates an attempt to use a value where a type is expected. For example:

enum Foo {
    Bar(u32)
}

fn do_something(x: Foo::Bar) { }

In this example, we're attempting to take a type of Foo::Bar in the do_something function. This is not legal: Foo::Bar is a value of type Foo, not a distinct static type. Likewise, it's not legal to attempt to impl Foo::Bar: instead, you must impl Foo and then pattern match to specify behavior for specific enum variants.

E0249

This error indicates a constant expression for the array length was found, but it was not an integer (signed or unsigned) expression.

Some examples of code that produces this error are:

const A: [u32; "hello"] = []; // error
const B: [u32; true] = []; // error
const C: [u32; 0.0] = []; // error

E0250

There was an error while evaluating the expression for the length of a fixed- size array type.

Some examples of this error are:

// divide by zero in the length expression
const A: [u32; 1/0] = [];

// Rust currently will not evaluate the function `foo` at compile time
fn foo() -> usize { 12 }
const B: [u32; foo()] = [];

// it is an error to try to add `u8` and `f64`
use std::{f64, u8};
const C: [u32; u8::MAX + f64::EPSILON] = [];

use foo::baz;
use bar::*; // error, do `use foo::baz as quux` instead on the previous line

fn main() {}

mod foo {
    pub struct baz;
}

mod bar {
    pub mod baz {}
}

E0252

Two items of the same name cannot be imported without rebinding one of the items under a new local name.

An example of this error:

use foo::baz;
use bar::baz; // error, do `use bar::baz as quux` instead

fn main() {}

mod foo {
    pub struct baz;
}

mod bar {
    pub mod baz {}
}

E0253

Attempt was made to import an unimportable value. This can happen when trying to import a method from a trait. An example of this error:

mod foo {
    pub trait MyTrait {
        fn do_something();
    }
}

use foo::MyTrait::do_something;

It's invalid to directly import methods belonging to a trait or concrete type.

E0255

You can't import a value whose name is the same as another value defined in the module.

An example of this error:

use bar::foo; // error, do `use bar::foo as baz` instead

fn foo() {}

mod bar {
     pub fn foo() {}
}

fn main() {}

use foo::Bar; // error

type Bar = u32;

mod foo {
    pub mod Bar { }
}

fn main() {}

E0259

The name chosen for an external crate conflicts with another external crate that has been imported into the current module.

Erroneous code example:

extern crate a;
extern crate crate_a as a;

The solution is to choose a different name that doesn't conflict with any external crate imported into the current module.

Correct example:

extern crate a;
extern crate crate_a as other_name;

E0260

The name for an item declaration conflicts with an external crate's name.

For instance:

extern crate abc;

struct abc;

There are two possible solutions:

Solution #1: Rename the item.

extern crate abc;

struct xyz;

Solution #2: Import the crate with a different name.

extern crate abc as xyz;

struct abc;

See the Declaration Statements section of the reference for more information about what constitutes an Item declaration and what does not:

https://doc.rust-lang.org/reference.html#statements

E0261

When using a lifetime like 'a in a type, it must be declared before being used.

These two examples illustrate the problem:

// error, use of undeclared lifetime name `'a`
fn foo(x: &'a str) { }

struct Foo {
    // error, use of undeclared lifetime name `'a`
    x: &'a str,
}

These can be fixed by declaring lifetime parameters:

fn foo<'a>(x: &'a str) {}

struct Foo<'a> {
    x: &'a str,
}

E0262

Declaring certain lifetime names in parameters is disallowed. For example, because the 'static lifetime is a special built-in lifetime name denoting the lifetime of the entire program, this is an error:

// error, invalid lifetime parameter name `'static`
fn foo<'static>(x: &'static str) { }

E0263

A lifetime name cannot be declared more than once in the same scope. For example:

// error, lifetime name `'a` declared twice in the same scope
fn foo<'a, 'b, 'a>(x: &'a str, y: &'b str) { }

E0264

An unknown external lang item was used. Erroneous code example:

#![feature(lang_items)]

extern "C" {
    #[lang = "cake"] // error: unknown external lang item: `cake`
    fn cake();
}

A list of available external lang items is available in src/librustc/middle/weak_lang_items.rs. Example:

#![feature(lang_items)]

extern "C" {
    #[lang = "panic_fmt"] // ok!
    fn cake();
}

E0269

Functions must eventually return a value of their return type. For example, in the following function:

fn foo(x: u8) -> u8 {
    if x > 0 {
        x // alternatively, `return x`
    }
    // nothing here
}

If the condition is true, the value x is returned, but if the condition is false, control exits the if block and reaches a place where nothing is being returned. All possible control paths must eventually return a u8, which is not happening here.

An easy fix for this in a complicated function is to specify a default return value, if possible:

fn foo(x: u8) -> u8 {
    if x > 0 {
        x // alternatively, `return x`
    }
    // lots of other if branches
    0 // return 0 if all else fails
}

It is advisable to find out what the unhandled cases are and check for them, returning an appropriate value or panicking if necessary. Check if you need to remove a semicolon from the last expression, like in this case:

fn foo(x: u8) -> u8 {
    inner(2*x + 1);
}

The semicolon discards the return value of inner, instead of returning it from foo.

E0270

Rust lets you define functions which are known to never return, i.e. are 'diverging', by marking its return type as !.

For example, the following functions never return:

fn foo() -> ! {
    loop {}
}

fn bar() -> ! {
    foo() // foo() is diverging, so this will diverge too
}

fn baz() -> ! {
    panic!(); // this macro internally expands to a call to a diverging function
}

Such functions can be used in a place where a value is expected without returning a value of that type, for instance:

fn foo() -> ! {
    loop {}
}

let x = 3;

let y = match x {
    1 => 1,
    2 => 4,
    _ => foo() // diverging function called here
};

println!("{}", y)

If the third arm of the match block is reached, since foo() doesn't ever return control to the match block, it is fine to use it in a place where an integer was expected. The match block will never finish executing, and any point where y (like the print statement) is needed will not be reached.

However, if we had a diverging function that actually does finish execution:

fn foo() -> ! {
    loop {break;}
}

Then we would have an unknown value for y in the following code:

fn foo() -> ! {
    loop {}
}

let x = 3;

let y = match x {
    1 => 1,
    2 => 4,
    _ => foo()
};

println!("{}", y);

In the previous example, the print statement was never reached when the wildcard match arm was hit, so we were okay with foo() not returning an integer that we could set to y. But in this example, foo() actually does return control, so the print statement will be executed with an uninitialized value.

Obviously we cannot have functions which are allowed to be used in such positions and yet can return control. So, if you are defining a function that returns !, make sure that there is no way for it to actually finish executing.

E0271

This is because of a type mismatch between the associated type of some trait (e.g. T::Bar, where T implements trait Quux { type Bar; }) and another type U that is required to be equal to T::Bar, but is not. Examples follow.

Here is a basic example:

trait Trait { type AssociatedType; }

fn foo<T>(t: T) where T: Trait<AssociatedType=u32> {
    println!("in foo");
}

impl Trait for i8 { type AssociatedType = &'static str; }

foo(3_i8);

Here is that same example again, with some explanatory comments:

trait Trait { type AssociatedType; }

fn foo<T>(t: T) where T: Trait<AssociatedType=u32> {
//                    ~~~~~~~~ ~~~~~~~~~~~~~~~~~~
//                        |            |
//         This says `foo` can         |
//           only be used with         |
//              some type that         |
//         implements `Trait`.         |
//                                     |
//                             This says not only must
//                             `T` be an impl of `Trait`
//                             but also that the impl
//                             must assign the type `u32`
//                             to the associated type.
    println!("in foo");
}

impl Trait for i8 { type AssociatedType = &'static str; }
~~~~~~~~~~~~~~~~~   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
//      |                             |
// `i8` does have                     |
// implementation                     |
// of `Trait`...                      |
//                     ... but it is an implementation
//                     that assigns `&'static str` to
//                     the associated type.

foo(3_i8);
// Here, we invoke `foo` with an `i8`, which does not satisfy
// the constraint `<i8 as Trait>::AssociatedType=u32`, and
// therefore the type-checker complains with this error code.

Here is a more subtle instance of the same problem, that can arise with for-loops in Rust:

let vs: Vec<i32> = vec![1, 2, 3, 4];
for v in &vs {
    match v {
        1 => {},
        _ => {},
    }
}

The above fails because of an analogous type mismatch, though may be harder to see. Again, here are some explanatory comments for the same example:

{
    let vs = vec![1, 2, 3, 4];

    // `for`-loops use a protocol based on the `Iterator`
    // trait. Each item yielded in a `for` loop has the
    // type `Iterator::Item` -- that is, `Item` is the
    // associated type of the concrete iterator impl.
    for v in &vs {
//      ~    ~~~
//      |     |
//      |    We borrow `vs`, iterating over a sequence of
//      |    *references* of type `&Elem` (where `Elem` is
//      |    vector's element type). Thus, the associated
//      |    type `Item` must be a reference `&`-type ...
//      |
//  ... and `v` has the type `Iterator::Item`, as dictated by
//  the `for`-loop protocol ...

        match v {
            1 => {}
//          ~
//          |
// ... but *here*, `v` is forced to have some integral type;
// only types like `u8`,`i8`,`u16`,`i16`, et cetera can
// match the pattern `1` ...

            _ => {}
        }

// ... therefore, the compiler complains, because it sees
// an attempt to solve the equations
// `some integral-type` = type-of-`v`
//                      = `Iterator::Item`
//                      = `&Elem` (i.e. `some reference type`)
//
// which cannot possibly all be true.

    }
}

To avoid those issues, you have to make the types match correctly. So we can fix the previous examples like this:

// Basic Example:
trait Trait { type AssociatedType; }

fn foo<T>(t: T) where T: Trait<AssociatedType = &'static str> {
    println!("in foo");
}

impl Trait for i8 { type AssociatedType = &'static str; }

foo(3_i8);

// For-Loop Example:
let vs = vec![1, 2, 3, 4];
for v in &vs {
    match v {
        &1 => {}
        _ => {}
    }
}

E0272

The #[rustc_on_unimplemented] attribute lets you specify a custom error message for when a particular trait isn't implemented on a type placed in a position that needs that trait. For example, when the following code is compiled:

fn foo<T: Index<u8>>(x: T){}

#[rustc_on_unimplemented = "the type `{Self}` cannot be indexed by `{Idx}`"]
trait Index<Idx> { /* ... */ }

foo(true); // `bool` does not implement `Index<u8>`

There will be an error about bool not implementing Index<u8>, followed by a note saying "the type bool cannot be indexed by u8".

As you can see, you can specify type parameters in curly braces for substitution with the actual types (using the regular format string syntax) in a given situation. Furthermore, {Self} will substitute to the type (in this case, bool) that we tried to use.

This error appears when the curly braces contain an identifier which doesn't match with any of the type parameters or the string Self. This might happen if you misspelled a type parameter, or if you intended to use literal curly braces. If it is the latter, escape the curly braces with a second curly brace of the same type; e.g. a literal { is {{.

E0273

The #[rustc_on_unimplemented] attribute lets you specify a custom error message for when a particular trait isn't implemented on a type placed in a position that needs that trait. For example, when the following code is compiled:

fn foo<T: Index<u8>>(x: T){}

#[rustc_on_unimplemented = "the type `{Self}` cannot be indexed by `{Idx}`"]
trait Index<Idx> { /* ... */ }

foo(true); // `bool` does not implement `Index<u8>`

there will be an error about bool not implementing Index<u8>, followed by a note saying "the type bool cannot be indexed by u8".

As you can see, you can specify type parameters in curly braces for substitution with the actual types (using the regular format string syntax) in a given situation. Furthermore, {Self} will substitute to the type (in this case, bool) that we tried to use.

This error appears when the curly braces do not contain an identifier. Please add one of the same name as a type parameter. If you intended to use literal braces, use {{ and }} to escape them.

E0274

The #[rustc_on_unimplemented] attribute lets you specify a custom error message for when a particular trait isn't implemented on a type placed in a position that needs that trait. For example, when the following code is compiled:

fn foo<T: Index<u8>>(x: T){}

#[rustc_on_unimplemented = "the type `{Self}` cannot be indexed by `{Idx}`"]
trait Index<Idx> { /* ... */ }

foo(true); // `bool` does not implement `Index<u8>`

there will be an error about bool not implementing Index<u8>, followed by a note saying "the type bool cannot be indexed by u8".

For this to work, some note must be specified. An empty attribute will not do anything, please remove the attribute or add some helpful note for users of the trait.

E0275

This error occurs when there was a recursive trait requirement that overflowed before it could be evaluated. Often this means that there is unbounded recursion in resolving some type bounds.

For example, in the following code:

trait Foo {}

struct Bar<T>(T);

impl<T> Foo for T where Bar<T>: Foo {}

To determine if a T is Foo, we need to check if Bar<T> is Foo. However, to do this check, we need to determine that Bar<Bar<T>> is Foo. To determine this, we check if Bar<Bar<Bar<T>>> is Foo, and so on. This is clearly a recursive requirement that can't be resolved directly.

Consider changing your trait bounds so that they're less self-referential.

E0276

This error occurs when a bound in an implementation of a trait does not match the bounds specified in the original trait. For example:

trait Foo {
    fn foo<T>(x: T);
}

impl Foo for bool {
    fn foo<T>(x: T) where T: Copy {}
}

Here, all types implementing Foo must have a method foo<T>(x: T) which can take any type T. However, in the impl for bool, we have added an extra bound that T is Copy, which isn't compatible with the original trait.

Consider removing the bound from the method or adding the bound to the original method definition in the trait.

E0277

You tried to use a type which doesn't implement some trait in a place which expected that trait. Erroneous code example:

// here we declare the Foo trait with a bar method
trait Foo {
    fn bar(&self);
}

// we now declare a function which takes an object implementing the Foo trait
fn some_func<T: Foo>(foo: T) {
    foo.bar();
}

fn main() {
    // we now call the method with the i32 type, which doesn't implement
    // the Foo trait
    some_func(5i32); // error: the trait bound `i32 : Foo` is not satisfied
}

In order to fix this error, verify that the type you're using does implement the trait. Example:

trait Foo {
    fn bar(&self);
}

fn some_func<T: Foo>(foo: T) {
    foo.bar(); // we can now use this method since i32 implements the
               // Foo trait
}

// we implement the trait on the i32 type
impl Foo for i32 {
    fn bar(&self) {}
}

fn main() {
    some_func(5i32); // ok!
}

Or in a generic context, an erroneous code example would look like: ``compile_fail fn some_func<T>(foo: T) { println!("{:?}", foo); // error: the traitcore::fmt::Debugis not // implemented for the typeT` }

fn main() { // We now call the method with the i32 type, // which does implement the Debug trait. some_func(5i32); } ```

Note that the error here is in the definition of the generic function: Although we only call it with a parameter that does implement Debug, the compiler still rejects the function: It must work with all possible input types. In order to make this example compile, we need to restrict the generic type we're accepting: ``` use std::fmt;

// Restrict the input type to types that implement Debug. fn some_func(foo: T) { println!("{:?}", foo); }

fn main() { // Calling the method is still fine, as i32 implements Debug. some_func(5i32);

// This would fail to compile now:
// struct WithoutDebug;
// some_func(WithoutDebug);

}

Rust only looks at the signature of the called function, as such it must already specify all requirements that will be used for every type parameter. ```

E0281

You tried to supply a type which doesn't implement some trait in a location which expected that trait. This error typically occurs when working with Fn-based types. Erroneous code example:

fn foo<F: Fn()>(x: F) { }

fn main() {
    // type mismatch: the type ... implements the trait `core::ops::Fn<(_,)>`,
    // but the trait `core::ops::Fn<()>` is required (expected (), found tuple
    // [E0281]
    foo(|y| { });
}

The issue in this case is that foo is defined as accepting a Fn with no arguments, but the closure we attempted to pass to it requires one argument.

E0282

This error indicates that type inference did not result in one unique possible type, and extra information is required. In most cases this can be provided by adding a type annotation. Sometimes you need to specify a generic type parameter manually.

A common example is the collect method on Iterator. It has a generic type parameter with a FromIterator bound, which for a char iterator is implemented by Vec and String among others. Consider the following snippet that reverses the characters of a string:

let x = "hello".chars().rev().collect();

In this case, the compiler cannot infer what the type of x should be: Vec<char> and String are both suitable candidates. To specify which type to use, you can use a type annotation on x:

let x: Vec<char> = "hello".chars().rev().collect();

It is not necessary to annotate the full type. Once the ambiguity is resolved, the compiler can infer the rest:

let x: Vec<_> = "hello".chars().rev().collect();

Another way to provide the compiler with enough information, is to specify the generic type parameter:

let x = "hello".chars().rev().collect::<Vec<char>>();

Again, you need not specify the full type if the compiler can infer it:

let x = "hello".chars().rev().collect::<Vec<_>>();

Apart from a method or function with a generic type parameter, this error can occur when a type parameter of a struct or trait cannot be inferred. In that case it is not always possible to use a type annotation, because all candidates have the same return type. For instance:

struct Foo<T> {
    num: T,
}

impl<T> Foo<T> {
    fn bar() -> i32 {
        0
    }

    fn baz() {
        let number = Foo::bar();
    }
}

This will fail because the compiler does not know which instance of Foo to call bar on. Change Foo::bar() to Foo::<T>::bar() to resolve the error.

E0283

This error occurs when the compiler doesn't have enough information to unambiguously choose an implementation.

For example:

trait Generator {
    fn create() -> u32;
}

struct Impl;

impl Generator for Impl {
    fn create() -> u32 { 1 }
}

struct AnotherImpl;

impl Generator for AnotherImpl {
    fn create() -> u32 { 2 }
}

fn main() {
    let cont: u32 = Generator::create();
    // error, impossible to choose one of Generator trait implementation
    // Impl or AnotherImpl? Maybe anything else?
}

To resolve this error use the concrete type:

trait Generator {
    fn create() -> u32;
}

struct AnotherImpl;

impl Generator for AnotherImpl {
    fn create() -> u32 { 2 }
}

fn main() {
    let gen1 = AnotherImpl::create();

    // if there are multiple methods with same name (different traits)
    let gen2 = <AnotherImpl as Generator>::create();
}

E0296

This error indicates that the given recursion limit could not be parsed. Ensure that the value provided is a positive integer between quotes, like so:

#![recursion_limit="1000"]

E0297

Patterns used to bind names must be irrefutable. That is, they must guarantee that a name will be extracted in all cases. Instead of pattern matching the loop variable, consider using a match or if let inside the loop body. For instance:

let xs : Vec<Option<i32>> = vec!(Some(1), None);

// This fails because `None` is not covered.
for Some(x) in xs {
    // ...
}

Match inside the loop instead:

let xs : Vec<Option<i32>> = vec!(Some(1), None);

for item in xs {
    match item {
        Some(x) => {},
        None => {},
    }
}

Or use if let:

let xs : Vec<Option<i32>> = vec!(Some(1), None);

for item in xs {
    if let Some(x) = item {
        // ...
    }
}

E0301

Mutable borrows are not allowed in pattern guards, because matching cannot have side effects. Side effects could alter the matched object or the environment on which the match depends in such a way, that the match would not be exhaustive. For instance, the following would not match any arm if mutable borrows were allowed:

match Some(()) {
    None => { },
    option if option.take().is_none() => {
        /* impossible, option is `Some` */
    },
    Some(_) => { } // When the previous match failed, the option became `None`.
}

E0302

Assignments are not allowed in pattern guards, because matching cannot have side effects. Side effects could alter the matched object or the environment on which the match depends in such a way, that the match would not be exhaustive. For instance, the following would not match any arm if assignments were allowed:

match Some(()) {
    None => { },
    option if { option = None; false } { },
    Some(_) => { } // When the previous match failed, the option became `None`.
}

E0303

In certain cases it is possible for sub-bindings to violate memory safety. Updates to the borrow checker in a future version of Rust may remove this restriction, but for now patterns must be rewritten without sub-bindings.

// Before.
match Some("hi".to_string()) {
    ref op_string_ref @ Some(s) => {},
    None => {},
}

// After.
match Some("hi".to_string()) {
    Some(ref s) => {
        let op_string_ref = &Some(s);
        // ...
    },
    None => {},
}

The op_string_ref binding has type &Option<&String> in both cases.

See also https://github.com/rust-lang/rust/issues/14587

E0306

In an array literal [x; N], N is the number of elements in the array. This must be an unsigned integer. Erroneous code example:

let x = [0i32; true]; // error: expected positive integer for repeat count,
                      //        found boolean

Working example:

let x = [0i32; 2];

E0307

The length of an array is part of its type. For this reason, this length must be a compile-time constant. Erroneous code example:

    let len = 10;
    let x = [0i32; len]; // error: expected constant integer for repeat count,
                         //        found variable

Working example:

let x = [0i32; 10];

E0308

This error occurs when the compiler was unable to infer the concrete type of a variable. It can occur for several cases, the most common of which is a mismatch in the expected type that the compiler inferred for a variable's initializing expression, and the actual type explicitly assigned to the variable.

For example:

let x: i32 = "I am not a number!";
//     ~~~   ~~~~~~~~~~~~~~~~~~~~
//      |             |
//      |    initializing expression;
//      |    compiler infers type `&str`
//      |
//    type `i32` assigned to variable `x`

Another situation in which this occurs is when you attempt to use the try! macro inside a function that does not return a Result<T, E>:

use std::fs::File;

fn main() {
    let mut f = try!(File::create("foo.txt"));
}

This code gives an error like this:

<std macros>:5:8: 6:42 error: mismatched types:
 expected `()`,
     found `core::result::Result<_, _>`
 (expected (),
     found enum `core::result::Result`) [E0308]

try! returns a Result<T, E>, and so the function must. But main() has () as its return type, hence the error.

E0309

Types in type definitions have lifetimes associated with them that represent how long the data stored within them is guaranteed to be live. This lifetime must be as long as the data needs to be alive, and missing the constraint that denotes this will cause this error.

// This won't compile because T is not constrained, meaning the data
// stored in it is not guaranteed to last as long as the reference
struct Foo<'a, T> {
    foo: &'a T
}

This will compile, because it has the constraint on the type parameter:

struct Foo<'a, T: 'a> {
    foo: &'a T
}

E0310

Types in type definitions have lifetimes associated with them that represent how long the data stored within them is guaranteed to be live. This lifetime must be as long as the data needs to be alive, and missing the constraint that denotes this will cause this error.

// This won't compile because T is not constrained to the static lifetime
// the reference needs
struct Foo<T> {
    foo: &'static T
}

This will compile, because it has the constraint on the type parameter:

struct Foo { foo: &'static T } ```

E0318

Default impls for a trait must be located in the same crate where the trait was defined. For more information see the opt-in builtin traits RFC.

E0321

A cross-crate opt-out trait was implemented on something which wasn't a struct or enum type. Erroneous code example:

#![feature(optin_builtin_traits)]

struct Foo;

impl !Sync for Foo {}

unsafe impl Send for &'static Foo {
// error: cross-crate traits with a default impl, like `core::marker::Send`,
//        can only be implemented for a struct/enum type, not
//        `&'static Foo`

Only structs and enums are permitted to impl Send, Sync, and other opt-out trait, and the struct or enum must be local to the current crate. So, for example, unsafe impl Send for Rc<Foo> is not allowed.

E0322

The Sized trait is a special trait built-in to the compiler for types with a constant size known at compile-time. This trait is automatically implemented for types as needed by the compiler, and it is currently disallowed to explicitly implement it for a type.

E0323

An associated const was implemented when another trait item was expected. Erroneous code example:

#![feature(associated_consts)]

trait Foo {
    type N;
}

struct Bar;

impl Foo for Bar {
    const N : u32 = 0;
    // error: item `N` is an associated const, which doesn't match its
    //        trait `<Bar as Foo>`
}

Please verify that the associated const wasn't misspelled and the correct trait was implemented. Example:

struct Bar;

trait Foo {
    type N;
}

impl Foo for Bar {
    type N = u32; // ok!
}

Or:

#![feature(associated_consts)]

struct Bar;

trait Foo {
    const N : u32;
}

impl Foo for Bar {
    const N : u32 = 0; // ok!
}

E0324

A method was implemented when another trait item was expected. Erroneous code example:

struct Bar;

trait Foo {
    const N : u32;

    fn M();
}

impl Foo for Bar {
    fn N() {}
    // error: item `N` is an associated method, which doesn't match its
    //        trait `<Bar as Foo>`
}

To fix this error, please verify that the method name wasn't misspelled and verify that you are indeed implementing the correct trait items. Example:

#![feature(associated_consts)]

struct Bar;

trait Foo {
    const N : u32;

    fn M();
}

impl Foo for Bar {
    const N : u32 = 0;

    fn M() {} // ok!
}

E0325

An associated type was implemented when another trait item was expected. Erroneous code example:

struct Bar;

trait Foo {
    const N : u32;
}

impl Foo for Bar {
    type N = u32;
    // error: item `N` is an associated type, which doesn't match its
    //        trait `<Bar as Foo>`
}

Please verify that the associated type name wasn't misspelled and your implementation corresponds to the trait definition. Example:

struct Bar;

trait Foo {
    type N;
}

impl Foo for Bar {
    type N = u32; // ok!
}

Or:

#![feature(associated_consts)]

struct Bar;

trait Foo {
    const N : u32;
}

impl Foo for Bar {
    const N : u32 = 0; // ok!
}

E0326

The types of any associated constants in a trait implementation must match the types in the trait definition. This error indicates that there was a mismatch.

Here's an example of this error:

trait Foo {
    const BAR: bool;
}

struct Bar;

impl Foo for Bar {
    const BAR: u32 = 5; // error, expected bool, found u32
}

E0327

You cannot use associated items other than constant items as patterns. This includes method items. Example of erroneous code:

enum B {}

impl B {
    fn bb() -> i32 { 0 }
}

fn main() {
    match 0 {
        B::bb => {} // error: associated items in match patterns must
                    // be constants
    }
}

Please check that you're not using a method as a pattern. Example:

enum B {
    ba,
    bb
}

fn main() {
    match B::ba {
        B::bb => {} // ok!
        _ => {}
    }
}

#![feature(associated_consts)]

trait Foo {
    const BAR: f64;
}

struct MyStruct;

impl Foo for MyStruct {
    const BAR: f64 = 0f64;
}

fn get_bar_bad<F: Foo>(t: F) -> f64 {
    F::BAR
}

#![feature(associated_consts)]

trait Foo {
    const BAR: f64;
}

struct MyStruct;

fn get_bar_good() -> f64 {
    <MyStruct as Foo>::BAR
}

E0364

Private items cannot be publicly re-exported. This error indicates that you attempted to pub use a type or value that was not itself public.

Here is an example that demonstrates the error:

mod foo {
    const X: u32 = 1;
}

pub use foo::X;

The solution to this problem is to ensure that the items that you are re-exporting are themselves marked with pub:

mod foo {
    pub const X: u32 = 1;
}

pub use foo::X;

See the 'Use Declarations' section of the reference for more information on this topic:

https://doc.rust-lang.org/reference.html#use-declarations

E0365

Private modules cannot be publicly re-exported. This error indicates that you attempted to pub use a module that was not itself public.

Here is an example that demonstrates the error:

mod foo {
    pub const X: u32 = 1;
}

pub use foo as foo2;

The solution to this problem is to ensure that the module that you are re-exporting is itself marked with pub:

pub mod foo {
    pub const X: u32 = 1;
}

pub use foo as foo2;

See the 'Use Declarations' section of the reference for more information on this topic:

https://doc.rust-lang.org/reference.html#use-declarations

E0366

An attempt was made to implement Drop on a concrete specialization of a generic type. An example is shown below:

struct Foo<T> {
    t: T
}

impl Drop for Foo<u32> {
    fn drop(&mut self) {}
}

This code is not legal: it is not possible to specialize Drop to a subset of implementations of a generic type. One workaround for this is to wrap the generic type, as shown below:

struct Foo<T> {
    t: T
}

struct Bar {
    t: Foo<u32>
}

impl Drop for Bar {
    fn drop(&mut self) {}
}

E0367

An attempt was made to implement Drop on a specialization of a generic type. An example is shown below:

trait Foo{}

struct MyStruct<T> {
    t: T
}

impl<T: Foo> Drop for MyStruct<T> {
    fn drop(&mut self) {}
}

This code is not legal: it is not possible to specialize Drop to a subset of implementations of a generic type. In order for this code to work, MyStruct must also require that T implements Foo. Alternatively, another option is to wrap the generic type in another that specializes appropriately:

trait Foo{}

struct MyStruct<T> {
    t: T
}

struct MyStructWrapper<T: Foo> {
    t: MyStruct<T>
}

impl <T: Foo> Drop for MyStructWrapper<T> {
    fn drop(&mut self) {}
}

E0368

This error indicates that a binary assignment operator like += or ^= was applied to a type that doesn't support it. For example:

let mut x = 12f32; // error: binary operation `<<` cannot be applied to
                   //        type `f32`

x <<= 2;

To fix this error, please check that this type implements this binary operation. Example:

let mut x = 12u32; // the `u32` type does implement the `ShlAssign` trait

x <<= 2; // ok!

It is also possible to overload most operators for your own type by implementing the [OP]Assign traits from std::ops.

Another problem you might be facing is this: suppose you've overloaded the + operator for some type Foo by implementing the std::ops::Add trait for Foo, but you find that using += does not work, as in this example:

use std::ops::Add;

struct Foo(u32);

impl Add for Foo {
    type Output = Foo;

    fn add(self, rhs: Foo) -> Foo {
        Foo(self.0 + rhs.0)
    }
}

fn main() {
    let mut x: Foo = Foo(5);
    x += Foo(7); // error, `+= cannot be applied to the type `Foo`
}

This is because AddAssign is not automatically implemented, so you need to manually implement it for your type.

E0369

A binary operation was attempted on a type which doesn't support it. Erroneous code example:

let x = 12f32; // error: binary operation `<<` cannot be applied to
               //        type `f32`

x << 2;

To fix this error, please check that this type implements this binary operation. Example:

let x = 12u32; // the `u32` type does implement it:
               // https://doc.rust-lang.org/stable/std/ops/trait.Shl.html

x << 2; // ok!

It is also possible to overload most operators for your own type by implementing traits from std::ops.

E0370

The maximum value of an enum was reached, so it cannot be automatically set in the next enum value. Erroneous code example:

enum Foo {
    X = 0x7fffffffffffffff,
    Y, // error: enum discriminant overflowed on value after
       //        9223372036854775807: i64; set explicitly via
       //        Y = -9223372036854775808 if that is desired outcome
}

To fix this, please set manually the next enum value or put the enum variant with the maximum value at the end of the enum. Examples:

enum Foo {
    X = 0x7fffffffffffffff,
    Y = 0, // ok!
}

Or:

enum Foo {
    Y = 0, // ok!
    X = 0x7fffffffffffffff,
}

E0371

When Trait2 is a subtrait of Trait1 (for example, when Trait2 has a definition like trait Trait2: Trait1 { ... }), it is not allowed to implement Trait1 for Trait2. This is because Trait2 already implements Trait1 by definition, so it is not useful to do this.

Example:

trait Foo { fn foo(&self) { } }
trait Bar: Foo { }
trait Baz: Bar { }

impl Bar for Baz { } // error, `Baz` implements `Bar` by definition
impl Foo for Baz { } // error, `Baz` implements `Bar` which implements `Foo`
impl Baz for Baz { } // error, `Baz` (trivially) implements `Baz`
impl Baz for Bar { } // Note: This is OK

E0373

This error occurs when an attempt is made to use data captured by a closure, when that data may no longer exist. It's most commonly seen when attempting to return a closure:

fn foo() -> Box<Fn(u32) -> u32> {
    let x = 0u32;
    Box::new(|y| x + y)
}

Notice that x is stack-allocated by foo(). By default, Rust captures closed-over data by reference. This means that once foo() returns, x no longer exists. An attempt to access x within the closure would thus be unsafe.

Another situation where this might be encountered is when spawning threads:

fn foo() {
    let x = 0u32;
    let y = 1u32;

    let thr = std::thread::spawn(|| {
        x + y
    });
}

Since our new thread runs in parallel, the stack frame containing x and y may well have disappeared by the time we try to use them. Even if we call thr.join() within foo (which blocks until thr has completed, ensuring the stack frame won't disappear), we will not succeed: the compiler cannot prove that this behaviour is safe, and so won't let us do it.

The solution to this problem is usually to switch to using a move closure. This approach moves (or copies, where possible) data into the closure, rather than taking references to it. For example:

fn foo() -> Box<Fn(u32) -> u32> {
    let x = 0u32;
    Box::new(move |y| x + y)
}

Now that the closure has its own copy of the data, there's no need to worry about safety.