Ralf Jung f8f2255a91 readme: document permissive-provenance flag

2022-05-23 09:07:32 +02:00

33 KiB

Raw Blame History

Miri

An experimental interpreter for Rust's mid-level intermediate representation (MIR). It can run binaries and test suites of cargo projects and detect certain classes of undefined behavior, for example:

Out-of-bounds memory accesses and use-after-free
Invalid use of uninitialized data
Violation of intrinsic preconditions (an unreachable_unchecked being reached, calling copy_nonoverlapping with overlapping ranges, ...)
Not sufficiently aligned memory accesses and references
Violation of some basic type invariants (a bool that is not 0 or 1, for example, or an invalid enum discriminant)
Experimental: Violations of the Stacked Borrows rules governing aliasing for reference types
Experimental: Data races (but no weak memory effects)

On top of that, Miri will also tell you about memory leaks: when there is memory still allocated at the end of the execution, and that memory is not reachable from a global static, Miri will raise an error.

You can use Miri to emulate programs on other targets, e.g. to ensure that byte-level data manipulation works correctly both on little-endian and big-endian systems. See cross-interpretation below.

Miri has already discovered some real-world bugs. If you found a bug with Miri, we'd appreciate if you tell us and we'll add it to the list!

However, be aware that Miri will not catch all cases of undefined behavior in your program, and cannot run all programs:

There are still plenty of open questions around the basic invariants for some types and when these invariants even have to hold. Miri tries to avoid false positives here, so if your program runs fine in Miri right now that is by no means a guarantee that it is UB-free when these questions get answered.

In particular, Miri does currently not check that integers/floats are initialized or that references point to valid data.
If the program relies on unspecified details of how data is laid out, it will still run fine in Miri -- but might break (including causing UB) on different compiler versions or different platforms.
Program execution is non-deterministic when it depends, for example, on where exactly in memory allocations end up, or on the exact interleaving of concurrent threads. Miri tests one of many possible executions of your program. You can alleviate this to some extent by running Miri with different values for -Zmiri-seed, but that will still by far not explore all possible executions.
Miri runs the program as a platform-independent interpreter, so the program has no access to most platform-specific APIs or FFI. A few APIs have been implemented (such as printing to stdout, accessing environment variables, and basic file system access) but most have not: for example, Miri currently does not support networking. System API support varies between targets; if you run on Windows it is a good idea to use --target x86_64-unknown-linux-gnu to get better support.
Threading support is not finished yet. E.g., weak memory effects are not emulated and spin loops (without syscalls) just loop forever. There is no threading support on Windows.

Using Miri

Install Miri on Rust nightly via rustup:

rustup +nightly component add miri

If rustup says the miri component is unavailable, that's because not all nightly releases come with all tools. Check out this website to determine a nightly version that comes with Miri and install that using rustup toolchain install nightly-YYYY-MM-DD. Either way, all of the following commands assume the right toolchain is pinned via rustup override set nightly or rustup override set nightly-YYYY-MM-DD. (Alternatively, use cargo +nightly/cargo +nightly-YYYY-MM-DD for each of the following commands.)

Now you can run your project in Miri:

Run cargo clean to eliminate any cached dependencies. Miri needs your dependencies to be compiled the right way, that would not happen if they have previously already been compiled.
To run all tests in your project through Miri, use cargo miri test.
If you have a binary project, you can run it through Miri using cargo miri run.

The first time you run Miri, it will perform some extra setup and install some dependencies. It will ask you for confirmation before installing anything.

cargo miri run/test supports the exact same flags as cargo run/test. For example, cargo miri test filter only runs the tests containing filter in their name.

You can pass arguments to Miri via MIRIFLAGS. For example, MIRIFLAGS="-Zmiri-disable-stacked-borrows" cargo miri run runs the program without checking the aliasing of references.

When compiling code via cargo miri, the cfg(miri) config flag is set for code that will be interpret under Miri. You can use this to ignore test cases that fail under Miri because they do things Miri does not support:

#[test]
#[cfg_attr(miri, ignore)]
fn does_not_work_on_miri() {
    tokio::run(futures::future::ok::<_, ()>(()));
}

There is no way to list all the infinite things Miri cannot do, but the interpreter will explicitly tell you when it finds something unsupported:

error: unsupported operation: can't call foreign function: bind
    ...
    = help: this is likely not a bug in the program; it indicates that the program \
            performed an operation that the interpreter does not support

Cross-interpretation: running for different targets

Miri can not only run a binary or test suite for your host target, it can also perform cross-interpretation for arbitrary foreign targets: cargo miri run --target x86_64-unknown-linux-gnu will run your program as if it was a Linux program, no matter your host OS. This is particularly useful if you are using Windows, as the Linux target is much better supported than Windows targets.

You can also use this to test platforms with different properties than your host platform. For example cargo miri test --target mips64-unknown-linux-gnuabi64 will run your test suite on a big-endian target, which is useful for testing endian-sensitive code.

Running Miri on CI

To run Miri on CI, make sure that you handle the case where the latest nightly does not ship the Miri component because it currently does not build. rustup toolchain install --component knows how to handle this situation, so the following snippet should always work:

rustup toolchain install nightly --component miri
rustup override set nightly

cargo miri test

Here is an example job for GitHub Actions:

  miri:
    name: "Miri"
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Install Miri
        run: |
          rustup toolchain install nightly --component miri
          rustup override set nightly
          cargo miri setup          
      - name: Test with Miri
        run: cargo miri test

The explicit cargo miri setup helps to keep the output of the actual test step clean.

Testing for alignment issues

Miri can sometimes miss misaligned accesses since allocations can "happen to be" aligned just right. You can use -Zmiri-symbolic-alignment-check to definitely catch all such issues, but that flag will also cause false positives when code does manual pointer arithmetic to account for alignment. Another alternative is to call Miri with various values for -Zmiri-seed; that will alter the randomness that is used to determine allocation base addresses. The following snippet calls Miri in a loop with different values for the seed:

for seed in $({ echo obase=16; seq 255; } | bc); do
  MIRIFLAGS=-Zmiri-seed=$seed cargo miri test || { echo "Last seed: $seed"; break; };
done

Supported targets

Miri does not support all targets supported by Rust. The good news, however, is that no matter your host OS/platform, it is easy to run code for any target using --target!

The following targets are tested on CI and thus should always work (to the degree documented below):

The best-supported target is x86_64-unknown-linux-gnu. Miri releases are blocked on things working with this target. Most other Linux targets should also work well; we do run the test suite on i686-unknown-linux-gnu as a 32bit target and mips64-unknown-linux-gnuabi64 as a big-endian target.
x86_64-apple-darwin should work basically as well as Linux. We also test aarch64-apple-darwin. However, we might ship Miri with a nightly even when some features on these targets regress.
x86_64-pc-windows-msvc works, but supports fewer features than the Linux and Apple targets. For example, file system access and concurrency are not supported on Windows. We also test i686-pc-windows-msvc, with the same reduced feature set. We might ship Miri with a nightly even when some features on these targets regress.

Common Problems

When using the above instructions, you may encounter a number of confusing compiler errors.

"note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace"

You may see this when trying to get Miri to display a backtrace. By default, Miri doesn't expose any environment to the program, so running RUST_BACKTRACE=1 cargo miri test will not do what you expect.

To get a backtrace, you need to disable isolation using -Zmiri-disable-isolation:

RUST_BACKTRACE=1 MIRIFLAGS="-Zmiri-disable-isolation" cargo miri test

"found possibly newer version of crate `std` which `<dependency>` depends on"

Your build directory may contain artifacts from an earlier build that have/have not been built for Miri. Run cargo clean before switching from non-Miri to Miri builds and vice-versa.

"found crate `std` compiled by an incompatible version of rustc"

You may be running cargo miri with a different compiler version than the one used to build the custom libstd that Miri uses, and Miri failed to detect that. Try deleting ~/.cache/miri.

"no mir for `std::rt::lang_start_internal`"

This means the sysroot you are using was not compiled with Miri in mind. This should never happen when you use cargo miri because that takes care of setting up the sysroot. If you are using miri (the Miri driver) directly, see the contributors' guide for how to use ./miri to best do that.

Miri `-Z` flags and environment variables

Miri adds its own set of -Z flags, which are usually set via the MIRIFLAGS environment variable. Some of these are unsound, which means they can lead to Miri failing to detect cases of undefined behavior in a program.

-Zmiri-check-number-validity enables checking of integer and float validity (e.g., they must be initialized and not carry pointer provenance) as part of enforcing validity invariants. This has no effect when -Zmiri-disable-validation is present.
-Zmiri-compare-exchange-weak-failure-rate=<rate> changes the failure rate of compare_exchange_weak operations. The default is 0.8 (so 4 out of 5 weak ops will fail). You can change it to any value between 0.0 and 1.0, where 1.0 means it will always fail and 0.0 means it will never fail.
-Zmiri-disable-abi-check disables checking function ABI. Using this flag is unsound.
-Zmiri-disable-alignment-check disables checking pointer alignment, so you can focus on other failures, but it means Miri can miss bugs in your program. Using this flag is unsound.
-Zmiri-disable-data-race-detector disables checking for data races. Using this flag is unsound.
-Zmiri-disable-stacked-borrows disables checking the experimental Stacked Borrows aliasing rules. This can make Miri run faster, but it also means no aliasing violations will be detected. Using this flag is unsound (but the affected soundness rules are experimental).
-Zmiri-disable-validation disables enforcing validity invariants, which are enforced by default. This is mostly useful to focus on other failures (such as out-of-bounds accesses) first. Setting this flag means Miri can miss bugs in your program. However, this can also help to make Miri run faster. Using this flag is unsound.
-Zmiri-disable-isolation disables host isolation. As a consequence, the program has access to host resources such as environment variables, file systems, and randomness.
-Zmiri-isolation-error=<action> configures Miri's response to operations requiring host access while isolation is enabled. abort, hide, warn, and warn-nobacktrace are the supported actions. The default is to abort, which halts the machine. Some (but not all) operations also support continuing execution with a "permission denied" error being returned to the program. warn prints a full backtrace when that happen; warn-nobacktrace is less verbose. hide hides the warning entirely.
-Zmiri-env-exclude=<var> keeps the var environment variable isolated from the host so that it cannot be accessed by the program. Can be used multiple times to exclude several variables. The TERM environment variable is excluded by default to speed up the test harness. This has no effect unless -Zmiri-disable-validation is also set.
-Zmiri-env-forward=<var> forwards the var environment variable to the interpreted program. Can be used multiple times to forward several variables. This has no effect if -Zmiri-disable-validation is set.
-Zmiri-ignore-leaks disables the memory leak checker, and also allows some remaining threads to exist when the main thread exits.
-Zmiri-measureme=<name> enables measureme profiling for the interpreted program. This can be used to find which parts of your program are executing slowly under Miri. The profile is written out to a file with the prefix <name>, and can be processed using the tools in the repository https://github.com/rust-lang/measureme.
-Zmiri-mute-stdout-stderr silently ignores all writes to stdout and stderr, but reports to the program that it did actually write. This is useful when you are not interested in the actual program's output, but only want to see miri's errors and warnings.
-Zmiri-panic-on-unsupported will makes some forms of unsupported functionality, such as FFI and unsupported syscalls, panic within the context of the emulated application instead of raising an error within the context of Miri (and halting execution). Note that code might not expect these operations to ever panic, so this flag can lead to strange (mis)behavior.
-Zmiri-permissive-provenance is experimental. This will make Miri do a best-effort attempt to implement the semantics of expose_addr and ptr::from_exposed_addr for pointer-to-int and int-to-pointer casts, respectively. This will necessarily miss some bugs as those semantics are not efficiently implementable in a sanitizer, but it will only miss bugs that concerns memory/pointers which is subject to these operations. Also note that this flag is currently incompatible with Stacked Borrows, so you will have to also pass -Zmiri-disable-stacked-borrows to use this.
-Zmiri-seed=<hex> configures the seed of the RNG that Miri uses to resolve non-determinism. This RNG is used to pick base addresses for allocations. When isolation is enabled (the default), this is also used to emulate system entropy. The default seed is 0. NOTE: This entropy is not good enough for cryptographic use! Do not generate secret keys in Miri or perform other kinds of cryptographic operations that rely on proper random numbers.
-Zmiri-strict-provenance enables strict provenance checking in Miri. This means that casting an integer to a pointer yields a result with 'invalid' provenance, i.e., with provenance that cannot be used for any memory access. Also implies -Zmiri-tag-raw-pointers and -Zmiri-check-number-validity.
-Zmiri-symbolic-alignment-check makes the alignment check more strict. By default, alignment is checked by casting the pointer to an integer, and making sure that is a multiple of the alignment. This can lead to cases where a program passes the alignment check by pure chance, because things "happened to be" sufficiently aligned -- there is no UB in this execution but there would be UB in others. To avoid such cases, the symbolic alignment check only takes into account the requested alignment of the relevant allocation, and the offset into that allocation. This avoids missing such bugs, but it also incurs some false positives when the code does manual integer arithmetic to ensure alignment. (The standard library align_to method works fine in both modes; under symbolic alignment it only fills the middle slice when the allocation guarantees sufficient alignment.)
-Zmiri-track-alloc-id=<id1>,<id2>,... shows a backtrace when the given allocations are being allocated or freed. This helps in debugging memory leaks and use after free bugs. Specifying this argument multiple times does not overwrite the previous values, instead it appends its values to the list. Listing an id multiple times has no effect.
-Zmiri-track-call-id=<id1>,<id2>,... shows a backtrace when the given call ids are assigned to a stack frame. This helps in debugging UB related to Stacked Borrows "protectors". Specifying this argument multiple times does not overwrite the previous values, instead it appends its values to the list. Listing an id multiple times has no effect.
-Zmiri-track-pointer-tag=<tag1>,<tag2>,... shows a backtrace when a given pointer tag is popped from a borrow stack (which is where the tag becomes invalid and any future use of it will error). This helps you in finding out why UB is happening and where in your code would be a good place to look for it. Specifying this argument multiple times does not overwrite the previous values, instead it appends its values to the list. Listing a tag multiple times has no effect.
-Zmiri-tag-raw-pointers makes Stacked Borrows assign proper tags even for raw pointers. This can make valid code using int-to-ptr casts fail to pass the checks, but also can help identify latent aliasing issues in code that Miri accepts by default. You can recognize false positives by <untagged> occurring in the message -- this indicates a pointer that was cast from an integer, so Miri was unable to track this pointer. Note that it is not currently guaranteed that code that works with -Zmiri-tag-raw-pointers also works without -Zmiri-tag-raw-pointers, but for the vast majority of code, this will be the case.

Some native rustc -Z flags are also very relevant for Miri:

-Zmir-opt-level controls how many MIR optimizations are performed. Miri overrides the default to be 0; be advised that using any higher level can make Miri miss bugs in your program because they got optimized away.
-Zalways-encode-mir makes rustc dump MIR even for completely monomorphic functions. This is needed so that Miri can execute such functions, so Miri sets this flag per default.
-Zmir-emit-retag controls whether Retag statements are emitted. Miri enables this per default because it is needed for Stacked Borrows.

Moreover, Miri recognizes some environment variables:

MIRI_LOG, MIRI_BACKTRACE control logging and backtrace printing during Miri executions, also see "Testing the Miri driver" in CONTRIBUTING.md.
MIRIFLAGS (recognized by cargo miri and the test suite) defines extra flags to be passed to Miri.
MIRI_LIB_SRC defines the directory where Miri expects the sources of the standard library that it will build and use for interpretation. This directory must point to the library subdirectory of a rust-lang/rust repository checkout. Note that changing files in that directory does not automatically trigger a re-build of the standard library; you have to clear the Miri build cache manually (on Linux, rm -rf ~/.cache/miri).
MIRI_SYSROOT (recognized by cargo miri and the test suite) indicates the sysroot to use. Only set this if you do not want to use the automatically created sysroot. (The miri driver sysroot is controlled via the --sysroot flag instead.)
MIRI_TEST_TARGET (recognized by the test suite) indicates which target architecture to test against. miri and cargo miri accept the --target flag for the same purpose.

The following environment variables are internal and must not be used by anyone but Miri itself. They are used to communicate between different Miri binaries, and as such worth documenting:

MIRI_BE_RUSTC can be set to host or target. It tells the Miri driver to actually not interpret the code but compile it like rustc would. With target, Miri sets some compiler flags to prepare the code for interpretation; with host, this is not done. This environment variable is useful to be sure that the compiled rlibs are compatible with Miri.
MIRI_CALLED_FROM_XARGO is set during the Miri-induced xargo sysroot build, which will re-invoke cargo-miri as the rustc to use for this build.
MIRI_CALLED_FROM_RUSTDOC when set to any value tells cargo-miri that it is running as a child process of rustdoc, which invokes it twice for each doc-test and requires special treatment, most notably a check-only build before interpretation. This is set by cargo-miri itself when running as a rustdoc-wrapper.
MIRI_CWD when set to any value tells the Miri driver to change to the given directory after loading all the source files, but before commencing interpretation. This is useful if the interpreted program wants a different working directory at run-time than at build-time.
MIRI_LOCAL_CRATES is set by cargo-miri to tell the Miri driver which crates should be given special treatment in diagnostics, in addition to the crate currently being compiled.
MIRI_VERBOSE when set to any value tells the various cargo-miri phases to perform verbose logging.

Miri `extern` functions

Miri provides some extern functions that programs can import to access Miri-specific functionality:

#[cfg(miri)]
extern "Rust" {
    /// Miri-provided extern function to mark the block `ptr` points to as a "root"
    /// for some static memory. This memory and everything reachable by it is not
    /// considered leaking even if it still exists when the program terminates.
    ///
    /// `ptr` has to point to the beginning of an allocated block.
    fn miri_static_root(ptr: *const u8);

    // Miri-provided extern function to get the amount of frames in the current backtrace.
    // The `flags` argument must be `0`.
    fn miri_backtrace_size(flags: u64) -> usize;

    /// Miri-provided extern function to obtain a backtrace of the current call stack.
    /// This writes a slice of pointers into `buf` - each pointer is an opaque value
    /// that is only useful when passed to `miri_resolve_frame`.
    /// `buf` must have `miri_backtrace_size(0) * pointer_size` bytes of space.
    /// The `flags` argument must be `1`.
    fn miri_get_backtrace(flags: u64, buf: *mut *mut ());

    /// Miri-provided extern function to resolve a frame pointer obtained
    /// from `miri_get_backtrace`. The `flags` argument must be `1`,
    /// and `MiriFrame` should be declared as follows:
    ///
    /// ```rust
    /// #[repr(C)]
    /// struct MiriFrame {
    ///     // The size of the name of the function being executed, encoded in UTF-8
    ///     name_len: usize,
    ///     // The size of filename of the function being executed, encoded in UTF-8
    ///     filename_len: usize,
    ///     // The line number currently being executed in `filename`, starting from '1'.
    ///     lineno: u32,
    ///     // The column number currently being executed in `filename`, starting from '1'.
    ///     colno: u32,
    ///     // The function pointer to the function currently being executed.
    ///     // This can be compared against function pointers obtained by
    ///     // casting a function (e.g. `my_fn as *mut ()`)
    ///     fn_ptr: *mut ()
    /// }
    /// ```
    ///
    /// The fields must be declared in exactly the same order as they appear in `MiriFrame` above.
    /// This function can be called on any thread (not just the one which obtained `frame`).
    fn miri_resolve_frame(frame: *mut (), flags: u64) -> MiriFrame;

    /// Miri-provided extern function to get the name and filename of the frame provided by `miri_resolve_frame`.
    /// `name_buf` and `filename_buf` should be allocated with the `name_len` and `filename_len` fields of `MiriFrame`.
    /// The flags argument must be `0`.
    fn miri_resolve_frame_names(ptr: *mut (), flags: u64, name_buf: *mut u8, filename_buf: *mut u8);

    /// Miri-provided extern function to begin unwinding with the given payload.
    ///
    /// This is internal and unstable and should not be used; we give it here
    /// just to be complete.
    fn miri_start_panic(payload: *mut u8) -> !;
}

Contributing and getting help

If you want to contribute to Miri, great! Please check out our contribution guide.

For help with running Miri, you can open an issue here on GitHub or use the Miri stream on the Rust Zulip.

History

This project began as part of an undergraduate research course in 2015 by @solson at the University of Saskatchewan. There are slides and a report available from that project. In 2016, @oli-obk joined to prepare miri for eventually being used as const evaluator in the Rust compiler itself (basically, for const and static stuff), replacing the old evaluator that worked directly on the AST. In 2017, @RalfJung did an internship with Mozilla and began developing miri towards a tool for detecting undefined behavior, and also using miri as a way to explore the consequences of various possible definitions for undefined behavior in Rust. @oli-obk's move of the miri engine into the compiler finally came to completion in early 2018. Meanwhile, later that year, @RalfJung did a second internship, developing miri further with support for checking basic type invariants and verifying that references are used according to their aliasing restrictions.

Bugs found by Miri

Miri has already found a number of bugs in the Rust standard library and beyond, which we collect here.

Definite bugs found:

Violations of Stacked Borrows found that are likely bugs (but Stacked Borrows is currently just an experiment):

Scientific papers employing Miri

License

Licensed under either of

Apache License, Version 2.0 (LICENSE-APACHE or http://www.apache.org/licenses/LICENSE-2.0)
MIT license (LICENSE-MIT or http://opensource.org/licenses/MIT)

at your option.

Contribution

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you shall be dual licensed as above, without any additional terms or conditions.

33 KiB Raw Blame History