3 [![Actions build status][actions-badge]][actions-url]
5 [actions-badge]: https://github.com/rust-lang/miri/workflows/CI/badge.svg?branch=master
6 [actions-url]: https://github.com/rust-lang/miri/actions
8 An experimental interpreter for [Rust][rust]'s
9 [mid-level intermediate representation][mir] (MIR). It can run binaries and
10 test suites of cargo projects and detect certain classes of
11 [undefined behavior](https://doc.rust-lang.org/reference/behavior-considered-undefined.html),
14 * Out-of-bounds memory accesses and use-after-free
15 * Invalid use of uninitialized data
16 * Violation of intrinsic preconditions (an [`unreachable_unchecked`] being
17 reached, calling [`copy_nonoverlapping`] with overlapping ranges, ...)
18 * Not sufficiently aligned memory accesses and references
19 * Violation of *some* basic type invariants (a `bool` that is not 0 or 1, for example,
20 or an invalid enum discriminant)
21 * **Experimental**: Violations of the [Stacked Borrows] rules governing aliasing
23 * **Experimental**: Data races (but no weak memory effects)
25 On top of that, Miri will also tell you about memory leaks: when there is memory
26 still allocated at the end of the execution, and that memory is not reachable
27 from a global `static`, Miri will raise an error.
29 You can use Miri to emulate programs on other targets, e.g. to ensure that
30 byte-level data manipulation works correctly both on little-endian and
31 big-endian systems. See
32 [cross-interpretation](#cross-interpretation-running-for-different-targets)
35 Miri has already discovered some [real-world bugs](#bugs-found-by-miri). If you
36 found a bug with Miri, we'd appreciate if you tell us and we'll add it to the
39 However, be aware that Miri will **not catch all cases of undefined behavior**
40 in your program, and cannot run all programs:
42 * There are still plenty of open questions around the basic invariants for some
43 types and when these invariants even have to hold. Miri tries to avoid false
44 positives here, so if you program runs fine in Miri right now that is by no
45 means a guarantee that it is UB-free when these questions get answered.
47 In particular, Miri does currently not check that integers/floats are
48 initialized or that references point to valid data.
49 * If the program relies on unspecified details of how data is laid out, it will
50 still run fine in Miri -- but might break (including causing UB) on different
51 compiler versions or different platforms.
52 * Program execution is non-deterministic when it depends, for example, on where
53 exactly in memory allocations end up, or on the exact interleaving of
54 concurrent threads. Miri tests one of many possible executions of your
55 program. You can alleviate this to some extent by running Miri with different
56 values for `-Zmiri-seed`, but that will still by far not explore all possible
58 * Miri runs the program as a platform-independent interpreter, so the program
59 has no access to most platform-specific APIs or FFI. A few APIs have been
60 implemented (such as printing to stdout) but most have not: for example, Miri
61 currently does not support SIMD or networking.
63 [rust]: https://www.rust-lang.org/
64 [mir]: https://github.com/rust-lang/rfcs/blob/master/text/1211-mir.md
65 [`unreachable_unchecked`]: https://doc.rust-lang.org/stable/std/hint/fn.unreachable_unchecked.html
66 [`copy_nonoverlapping`]: https://doc.rust-lang.org/stable/std/ptr/fn.copy_nonoverlapping.html
67 [Stacked Borrows]: https://github.com/rust-lang/unsafe-code-guidelines/blob/master/wip/stacked-borrows.md
72 Install Miri on Rust nightly via `rustup`:
75 rustup +nightly component add miri
78 If `rustup` says the `miri` component is unavailable, that's because not all
79 nightly releases come with all tools. Check out
80 [this website](https://rust-lang.github.io/rustup-components-history) to
81 determine a nightly version that comes with Miri and install that using
82 `rustup toolchain install nightly-YYYY-MM-DD`.
84 Now you can run your project in Miri:
86 1. Run `cargo clean` to eliminate any cached dependencies. Miri needs your
87 dependencies to be compiled the right way, that would not happen if they have
88 previously already been compiled.
89 2. To run all tests in your project through Miri, use `cargo miri test`.
90 3. If you have a binary project, you can run it through Miri using `cargo miri run`.
92 The first time you run Miri, it will perform some extra setup and install some
93 dependencies. It will ask you for confirmation before installing anything.
95 `cargo miri run/test` supports the exact same flags as `cargo run/test`. You
96 can pass arguments to Miri via `MIRIFLAGS`. For example,
97 `MIRIFLAGS="-Zmiri-disable-stacked-borrows" cargo miri run` runs the program
98 without checking the aliasing of references.
100 When compiling code via `cargo miri`, the `cfg(miri)` config flag is set. You
101 can use this to ignore test cases that fail under Miri because they do things
102 Miri does not support:
106 #[cfg_attr(miri, ignore)]
107 fn does_not_work_on_miri() {
108 std::thread::spawn(|| println!("Hello Thread!"))
114 There is no way to list all the infinite things Miri cannot do, but the
115 interpreter will explicitly tell you when it finds something unsupported:
118 error: unsupported operation: can't call foreign function: bind
120 = help: this is likely not a bug in the program; it indicates that the program \
121 performed an operation that the interpreter does not support
124 ### Cross-interpretation: running for different targets
126 Miri can not only run a binary or test suite for your host target, it can also
127 perform cross-interpretation for arbitrary foreign targets: `cargo miri run
128 --target x86_64-unknown-linux-gnu` will run your program as if it was a Linux
129 program, no matter your host OS. This is particularly useful if you are using
130 Windows, as the Linux target is much better supported than Windows targets.
132 You can also use this to test platforms with different properties than your host
133 platform. For example `cargo miri test --target mips64-unknown-linux-gnuabi64`
134 will run your test suite on a big-endian target, which is useful for testing
135 endian-sensitive code.
137 ### Running Miri on CI
139 To run Miri on CI, make sure that you handle the case where the latest nightly
140 does not ship the Miri component because it currently does not build. For
141 example, you can use the following snippet to always test with the latest
142 nightly that *does* come with Miri:
145 MIRI_NIGHTLY=nightly-$(curl -s https://rust-lang.github.io/rustup-components-history/x86_64-unknown-linux-gnu/miri)
146 echo "Installing latest nightly with Miri: $MIRI_NIGHTLY"
147 rustup set profile minimal
148 rustup default "$MIRI_NIGHTLY"
149 rustup component add miri
156 When using the above instructions, you may encounter a number of confusing compiler
159 ### "note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace"
161 You may see this when trying to get Miri to display a backtrace. By default, Miri
162 doesn't expose any environment to the program, so running
163 `RUST_BACKTRACE=1 cargo miri test` will not do what you expect.
165 To get a backtrace, you need to disable isolation
166 [using `-Zmiri-disable-isolation`](#miri-flags):
169 RUST_BACKTRACE=1 MIRIFLAGS="-Zmiri-disable-isolation" cargo miri test
172 #### "found possibly newer version of crate `std` which `<dependency>` depends on"
174 Your build directory may contain artifacts from an earlier build that have/have
175 not been built for Miri. Run `cargo clean` before switching from non-Miri to
176 Miri builds and vice-versa.
178 #### "found crate `std` compiled by an incompatible version of rustc"
180 You may be running `cargo miri` with a different compiler version than the one
181 used to build the custom libstd that Miri uses, and Miri failed to detect that.
182 Try deleting `~/.cache/miri`.
184 #### "no mir for `std::rt::lang_start_internal`"
186 This means the sysroot you are using was not compiled with Miri in mind. This
187 should never happen when you use `cargo miri` because that takes care of setting
188 up the sysroot. If you are using `miri` (the Miri driver) directly, see the
189 [contributors' guide](CONTRIBUTING.md) for how to use `./miri` to best do that.
192 ## Miri `-Z` flags and environment variables
193 [miri-flags]: #miri--z-flags-and-environment-variables
195 Miri adds its own set of `-Z` flags, which are usually set via the `MIRIFLAGS`
196 environment variable:
198 * `-Zmiri-compare-exchange-weak-failure-rate=<rate>` changes the failure rate of
199 `compare_exchange_weak` operations. The default is `0.8` (so 4 out of 5 weak ops will fail).
200 You can change it to any value between `0.0` and `1.0`, where `1.0` means it
201 will always fail and `0.0` means it will never fail.
202 * `-Zmiri-disable-alignment-check` disables checking pointer alignment, so you
203 can focus on other failures, but it means Miri can miss bugs in your program.
204 Using this flag is **unsound**.
205 * `-Zmiri-disable-data-race-detector` disables checking for data races. Using
206 this flag is **unsound**.
207 * `-Zmiri-disable-stacked-borrows` disables checking the experimental
208 [Stacked Borrows] aliasing rules. This can make Miri run faster, but it also
209 means no aliasing violations will be detected. Using this flag is **unsound**
210 (but the affected soundness rules are experimental).
211 * `-Zmiri-disable-validation` disables enforcing validity invariants, which are
212 enforced by default. This is mostly useful to focus on other failures (such
213 as out-of-bounds accesses) first. Setting this flag means Miri can miss bugs
214 in your program. However, this can also help to make Miri run faster. Using
215 this flag is **unsound**.
216 * `-Zmiri-disable-isolation` disables host isolation. As a consequence,
217 the program has access to host resources such as environment variables, file
218 systems, and randomness.
219 * `-Zmiri-env-exclude=<var>` keeps the `var` environment variable isolated from
220 the host so that it cannot be accessed by the program. Can be used multiple
221 times to exclude several variables. On Windows, the `TERM` environment
222 variable is excluded by default.
223 * `-Zmiri-ignore-leaks` disables the memory leak checker.
224 * `-Zmiri-seed=<hex>` configures the seed of the RNG that Miri uses to resolve
225 non-determinism. This RNG is used to pick base addresses for allocations.
226 When isolation is enabled (the default), this is also used to emulate system
227 entropy. The default seed is 0. **NOTE**: This entropy is not good enough
228 for cryptographic use! Do not generate secret keys in Miri or perform other
229 kinds of cryptographic operations that rely on proper random numbers.
230 * `-Zmiri-symbolic-alignment-check` makes the alignment check more strict. By
231 default, alignment is checked by casting the pointer to an integer, and making
232 sure that is a multiple of the alignment. This can lead to cases where a
233 program passes the alignment check by pure chance, because things "happened to
234 be" sufficiently aligned -- there is no UB in this execution but there would
235 be UB in others. To avoid such cases, the symbolic alignment check only takes
236 into account the requested alignment of the relevant allocation, and the
237 offset into that allocation. This avoids missing such bugs, but it also
238 incurs some false positives when the code does manual integer arithmetic to
239 ensure alignment. (The standard library `align_to` method works fine in both
240 modes; under symbolic alignment it only fills the middle slice when the
241 allocation guarantees sufficient alignment.)
242 * `-Zmiri-track-alloc-id=<id>` shows a backtrace when the given allocation is
243 being allocated or freed. This helps in debugging memory leaks and
245 * `-Zmiri-track-call-id=<id>` shows a backtrace when the given call id is
246 assigned to a stack frame. This helps in debugging UB related to Stacked
247 Borrows "protectors".
248 * `-Zmiri-track-pointer-tag=<tag>` shows a backtrace when the given pointer tag
249 is popped from a borrow stack (which is where the tag becomes invalid and any
250 future use of it will error). This helps you in finding out why UB is
251 happening and where in your code would be a good place to look for it.
252 * `-Zmiri-track-raw-pointers` makes Stacked Borrows track a pointer tag even for
253 raw pointers. This can make valid code fail to pass the checks, but also can
254 help identify latent aliasing issues in code that Miri accepts by default. You
255 can recognize false positives by "<untagged>" occurring in the message -- this
256 indicates a pointer that was cast from an integer, so Miri was unable to track
259 Some native rustc `-Z` flags are also very relevant for Miri:
261 * `-Zmir-opt-level` controls how many MIR optimizations are performed. Miri
262 overrides the default to be `0`; be advised that using any higher level can
263 make Miri miss bugs in your program because they got optimized away.
264 * `-Zalways-encode-mir` makes rustc dump MIR even for completely monomorphic
265 functions. This is needed so that Miri can execute such functions, so Miri
266 sets this flag per default.
267 * `-Zmir-emit-retag` controls whether `Retag` statements are emitted. Miri
268 enables this per default because it is needed for [Stacked Borrows].
270 Moreover, Miri recognizes some environment variables:
272 * `MIRI_LOG`, `MIRI_BACKTRACE` control logging and backtrace printing during
273 Miri executions, also [see "Testing the Miri driver" in `CONTRIBUTING.md`][testing-miri].
274 * `MIRIFLAGS` (recognized by `cargo miri` and the test suite) defines extra
275 flags to be passed to Miri.
276 * `MIRI_SYSROOT` (recognized by `cargo miri` and the test suite)
277 indicates the sysroot to use. To do the same thing with `miri`
278 directly, use the `--sysroot` flag.
279 * `MIRI_TEST_TARGET` (recognized by the test suite) indicates which target
280 architecture to test against. `miri` and `cargo miri` accept the `--target`
281 flag for the same purpose.
283 The following environment variables are internal, but used to communicate between
284 different Miri binaries, and as such worth documenting:
286 * `MIRI_BE_RUSTC` when set to any value tells the Miri driver to actually not
287 interpret the code but compile it like rustc would. This is useful to be sure
288 that the compiled `rlib`s are compatible with Miri.
289 When set while running `cargo-miri`, it indicates that we are part of a sysroot
290 build (for which some crates need special treatment).
291 * `MIRI_CWD` when set to any value tells the Miri driver to change to the given
292 directory after loading all the source files, but before commencing
293 interpretation. This is useful if the interpreted program wants a different
294 working directory at run-time than at build-time.
295 * `MIRI_VERBOSE` when set to any value tells the various `cargo-miri` phases to
296 perform verbose logging.
298 [testing-miri]: CONTRIBUTING.md#testing-the-miri-driver
300 ## Miri `extern` functions
302 Miri provides some `extern` functions that programs can import to access
303 Miri-specific functionality:
308 /// Miri-provided extern function to mark the block `ptr` points to as a "root"
309 /// for some static memory. This memory and everything reachable by it is not
310 /// considered leaking even if it still exists when the program terminates.
312 /// `ptr` has to point to the beginning of an allocated block.
313 fn miri_static_root(ptr: *const u8);
315 /// Miri-provided extern function to obtain a backtrace of the current call stack.
316 /// This returns a boxed slice of pointers - each pointer is an opaque value
317 /// that is only useful when passed to `miri_resolve_frame`
318 /// The `flags` argument must be `0`.
319 fn miri_get_backtrace(flags: u64) -> Box<[*mut ()]>;
321 /// Miri-provided extern function to resolve a frame pointer obtained
322 /// from `miri_get_backtrace`. The `flags` argument must be `0`,
323 /// and `MiriFrame` should be declared as follows:
327 /// struct MiriFrame {
328 /// // The name of the function being executed, encoded in UTF-8
330 /// // The filename of the function being executed, encoded in UTF-8
331 /// filename: Box<[u8]>,
332 /// // The line number currently being executed in `filename`, starting from '1'.
334 /// // The column number currently being executed in `filename`, starting from '1'.
336 /// // The function pointer to the function currently being executed.
337 /// // This can be compared against function pointers obtained by
338 /// // casting a function (e.g. `my_fn as *mut ()`)
343 /// The fields must be declared in exactly the same order as they appear in `MiriFrame` above.
344 /// This function can be called on any thread (not just the one which obtained `frame`).
345 fn miri_resolve_frame(frame: *mut (), flags: u64) -> MiriFrame;
347 /// Miri-provided extern function to begin unwinding with the given payload.
349 /// This is internal and unstable and should not be used; we give it here
350 /// just to be complete.
351 fn miri_start_panic(payload: *mut u8) -> !;
355 ## Contributing and getting help
357 If you want to contribute to Miri, great! Please check out our
358 [contribution guide](CONTRIBUTING.md).
360 For help with running Miri, you can open an issue here on
361 GitHub or use the [Miri stream on the Rust Zulip][zulip].
363 [zulip]: https://rust-lang.zulipchat.com/#narrow/stream/269128-miri
367 This project began as part of an undergraduate research course in 2015 by
368 @solson at the [University of Saskatchewan][usask]. There are [slides] and a
369 [report] available from that project. In 2016, @oli-obk joined to prepare miri
370 for eventually being used as const evaluator in the Rust compiler itself
371 (basically, for `const` and `static` stuff), replacing the old evaluator that
372 worked directly on the AST. In 2017, @RalfJung did an internship with Mozilla
373 and began developing miri towards a tool for detecting undefined behavior, and
374 also using miri as a way to explore the consequences of various possible
375 definitions for undefined behavior in Rust. @oli-obk's move of the miri engine
376 into the compiler finally came to completion in early 2018. Meanwhile, later
377 that year, @RalfJung did a second internship, developing miri further with
378 support for checking basic type invariants and verifying that references are
379 used according to their aliasing restrictions.
381 [usask]: https://www.usask.ca/
382 [slides]: https://solson.me/miri-slides.pdf
383 [report]: https://solson.me/miri-report.pdf
385 ## Bugs found by Miri
387 Miri has already found a number of bugs in the Rust standard library and beyond, which we collect here.
391 * [`Debug for vec_deque::Iter` accessing uninitialized memory](https://github.com/rust-lang/rust/issues/53566)
392 * [`Vec::into_iter` doing an unaligned ZST read](https://github.com/rust-lang/rust/pull/53804)
393 * [`From<&[T]> for Rc` creating a not sufficiently aligned reference](https://github.com/rust-lang/rust/issues/54908)
394 * [`BTreeMap` creating a shared reference pointing to a too small allocation](https://github.com/rust-lang/rust/issues/54957)
395 * [`Vec::append` creating a dangling reference](https://github.com/rust-lang/rust/pull/61082)
396 * [Futures turning a shared reference into a mutable one](https://github.com/rust-lang/rust/pull/56319)
397 * [`str` turning a shared reference into a mutable one](https://github.com/rust-lang/rust/pull/58200)
398 * [`rand` performing unaligned reads](https://github.com/rust-random/rand/issues/779)
399 * [The Unix allocator calling `posix_memalign` in an invalid way](https://github.com/rust-lang/rust/issues/62251)
400 * [`getrandom` calling the `getrandom` syscall in an invalid way](https://github.com/rust-random/getrandom/pull/73)
401 * [`Vec`](https://github.com/rust-lang/rust/issues/69770) and [`BTreeMap`](https://github.com/rust-lang/rust/issues/69769) leaking memory under some (panicky) conditions
402 * [`beef` leaking memory](https://github.com/maciejhirsz/beef/issues/12)
403 * [`EbrCell` using uninitialized memory incorrectly](https://github.com/Firstyear/concread/commit/b15be53b6ec076acb295a5c0483cdb4bf9be838f#diff-6282b2fc8e98bd089a1f0c86f648157cR229)
404 * [TiKV performing an unaligned pointer access](https://github.com/tikv/tikv/issues/7613)
405 * [`servo_arc` creating a dangling shared reference](https://github.com/servo/servo/issues/26357)
406 * [TiKV constructing out-of-bounds pointers (and overlapping mutable references)](https://github.com/tikv/tikv/pull/7751)
407 * [`encoding_rs` doing out-of-bounds pointer arithmetic](https://github.com/hsivonen/encoding_rs/pull/53)
408 * [TiKV using `Vec::from_raw_parts` incorrectly](https://github.com/tikv/agatedb/pull/24)
410 Violations of [Stacked Borrows] found that are likely bugs (but Stacked Borrows is currently just an experiment):
412 * [`VecDeque::drain` creating overlapping mutable references](https://github.com/rust-lang/rust/pull/56161)
413 * Various `BTreeMap` problems
414 * [`BTreeMap` iterators creating mutable references that overlap with shared references](https://github.com/rust-lang/rust/pull/58431)
415 * [`BTreeMap::iter_mut` creating overlapping mutable references](https://github.com/rust-lang/rust/issues/73915)
416 * [`BTreeMap` node insertion using raw pointers outside their valid memory area](https://github.com/rust-lang/rust/issues/78477)
417 * [`LinkedList` cursor insertion creating overlapping mutable references](https://github.com/rust-lang/rust/pull/60072)
418 * [`Vec::push` invalidating existing references into the vector](https://github.com/rust-lang/rust/issues/60847)
419 * [`align_to_mut` violating uniqueness of mutable references](https://github.com/rust-lang/rust/issues/68549)
420 * [`sized-chunks` creating aliasing mutable references](https://github.com/bodil/sized-chunks/issues/8)
421 * [`String::push_str` invalidating existing references into the string](https://github.com/rust-lang/rust/issues/70301)
422 * [`ryu` using raw pointers outside their valid memory area](https://github.com/dtolnay/ryu/issues/24)
423 * [ink! creating overlapping mutable references](https://github.com/rust-lang/miri/issues/1364)
424 * [TiKV creating overlapping mutable reference and raw pointer](https://github.com/tikv/tikv/pull/7709)
425 * [Windows `Env` iterator using a raw pointer outside its valid memory area](https://github.com/rust-lang/rust/pull/70479)
426 * [`VecDeque::iter_mut` creating overlapping mutable references](https://github.com/rust-lang/rust/issues/74029)
427 * [Various standard library aliasing issues involving raw pointers](https://github.com/rust-lang/rust/pull/78602)
431 Licensed under either of
433 * Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE) or
434 http://www.apache.org/licenses/LICENSE-2.0)
435 * MIT license ([LICENSE-MIT](LICENSE-MIT) or
436 http://opensource.org/licenses/MIT)
442 Unless you explicitly state otherwise, any contribution intentionally submitted
443 for inclusion in the work by you shall be dual licensed as above, without any
444 additional terms or conditions.