In user interface, added a new flag `-Zmiri-isolation-error` which
takes one of the four values -- hide, warn, warn-nobacktrace, and
abort. This option can be used to configure Miri to either abort or
return an error code upon executing isolated op. If not aborted, Miri
prints a warning, whose verbosity can be configured using this flag.
In implementation, added a new enum `IsolatedOp` to capture all the
settings related to ops requiring communication with the
host. Old `communicate` flag in both miri configs and machine
stats is replaced with a new helper function `communicate()` which
checks `isolated_op` internally.
Added a new helper function `reject_in_isolation` which can be called
by shims to reject ops according to the reject_with settings. Use miri
specific diagnostics function `report_msg` to print backtrace in the
warning. Update it to take an enum value instead of a bool, indicating
the level of diagnostics.
Updated shims related to current dir to use the new APIs. Added a new
test for current dir ops in isolation without halting machine.
This PR uses the `measureme` crate to profile the call stack of the
program being interpreted by Miri. This is accomplished by starting a
measureme 'event' when we enter a function call, and ending the event
when we exit the call. The `measureme` tooling can be used to produce a
call stack from the generated profile data.
Limitations:
* We currently record every single entry/exit. This might generate very
large profile outputs for programs with a large number of function
calls. In follow-up work, we might want to explore sampling (e.g. only
recording every N function calls).
* This does not integrate very well with Miri's concurrency support.
Each event we record starts when we push a frame, and ends when we pop
a frame. As a result, switching between virtual threads will cause
events from different threads to be interleaved. Additionally, the
recorded for a particular frame will include all of the work Miri does
before that frame completes, including executing another thread.
The `measureme` integration is off by default, and must be enabled via
`-Zmiri-measureme=<output_name>`
Throw UB if f*_fast intrinsic called with non-finite value
Calling these intrinsics with non-finite values is undefined behaviour, since they result in `f*` intrinsics in LLVM with the `fast` flag, and `fast` math on non-finite values results in `poison` values. (technically LLVM only considers it UB upon _using_ the value, but that shouldn't make much of a difference)
only check timeouts when a thread yields
Currently, we check for expired timeouts after each step of execution. That seems excessive. This changes the scheduler to only check for timeouts when the active thread cannot continue running any more.
`@vakaras` does this sound right? `pthread_cond_timedwait` anyway already yields, of course, since it blocks on getting the signal (or the timeout).
Add atomic min and max
Closes#1718
Previous attempt: #1653
TODO:
- [x] Merge `atomic_op` and `atomic_min_max` functions
- [x] Fix CI
**Note:** this PR also removes arbitrary trailing whitespace and generally formats the affected files
Add simple data-race detector
Partially fixes data-race detection, see #1372, based on Dynamic Race Detection for C++11
- This does not explore weak memory behaviour, only exploring one sequentially consistent ordering.
- Data-race detection is only enabled after the first thread is created, so should have minimal overhead for non-concurrent execution.
- ~~Does not attempt to re-use thread id's so creating and joining threads lots of time in an execution will result in the vector clocks growing in size and slowing down program execution~~ It does now
std now looks up `getrandom` and `statx` with `dlsym` before attempting
to use `syscall(SYS_.., ..)`.
It also now passes all arguments as a machine-sized word, instead of
their original types.
Rustc removed the target wrapper and exposed the target directly.
Result of running:
find . -type f -exec sed -i -e 's/target\.target\([)\.,;]\)/target\1/g' {} \;
Plus one manual edit of the rust-version file
Add an `fn_ptr` field to `MiriFrame`
The `backtrace-rs` crate can use this to implement
`Frame::symbol_address`, which is used to skip frames
above the call to `Backtrace::capture` on the stack.
The function pointer will not be useable for comparison purposes if the
function is generic, as CTFE creates a new function pointer for each
cast of a (monomorphized) generic function. However, this already
affects code running under Miri, and isn't a problem for `backtrace-rs`
(which only casts a non-generic function).
I've added logic to allow `MiriFrame` to have either 4 or 5 fields - if
a 5th field is present, we write the function pointer to it.
The `backtrace-rs` crate can use this to implement
`Frame::symbol_address`, which is used to skip frames
above the call to `Backtrace::capture` on the stack.
The function pointer will not be useable for comparison purposes if the
function is generic, as CTFE creates a new function pointer for each
cast of a (monomorphized) generic function. However, this already
affects code running under Miri, and isn't a problem for `backtrace-rs`
(which only casts a non-generic function).
I've added logic to allow `MiriFrame` to have either 4 or 5 fields - if
a 5th field is present, we write the function pointer to it.
This PR adds two new Miri-defined extern functions:
`miri_get_backtrace` and `miri_resolve_frame`, which are documented in
the README. Together, they allow obtaining a backtrace for the currently
executing program.
I've added a test showing how these APIs are used. I've also prepared a
companion PR `backtrace-rs`, which will allow
`backtrace::Backtrace::new()` to work automatically under Miri.
Once these two PRs are merged, we will be able to print backtraces from
the normal Rust panic hook (since libstd is now using backtrace-rs).
A few notes:
* Resolving the backtrace frames is *very* slow - you can actually see
each line being printed out one at a time. Some local testing showed
that this is not (primrary) caused by resolving a `Span` - it seems
to be just Miri being slow.
* For the first time, we now interact directly with a user-defined
struct (instead of just executing the user-provided MIR that
manipulates the struct). To allow for future changes, I've added
a 'version' parameter (currently required to be 0). This should allow
us to change the `MiriFrame` struct should the need ever arise.
* I used the approach suggested by @oli-obk - a returned backtrace
pointer consists of a base function allocation, with the 'offset'
used to encode the `Span.lo`. This allows losslessly reconstructing
the location information in `miri_resolve_frame`.
* There are a few quirks on the `backtrace-rs` side:
* `backtrace-rs` calls `getcwd()` by default to try to simplify
the filename. This results in an isolation error by default,
which could be annoying when printing a backtrace from libstd.
* `backtrace-rs` tries to remove 'internal' frames (everything between
the call to `Backtrace::new()` and the internal API call made by
backtrace-rs) by comparing the returned frame pointer value to
a Rust function pointer. This doesn't work due to the way we
construct the frame pointers passed to the caller. We could
attempt to support this kind of comparison, or just add a
`#[cfg(miri)]` and ignore the frames ourselves.
Redo cargo-miri logic
This rewrite the cargo-miri logic for running the requested crate(s) following what we outlined in https://github.com/rust-lang/miri/issues/739: `cargo miri run/test $FLAGS` (almost) literally invokes `cargo run/test $FLAGS` but with some environment variables set so that we can control what happens:
* `RUSTC_WRAPPER` is set so that we get invoked instead of `rustc`. We use that power to mess with the flags being used for the build (things to be interpreted by Miri use a different sysroot), and when we are detecting a binary crate meant to be run by Miri, we grab the info we care about and put it into a JSON file for later use.
* `CARGO_TARGET_$TARGET_RUNNER` is set so what we get invoked when cargo wants to run a binary. At that point we take that JSON info from before and use it to invoke Miri.
Overall this works great! We get all sorts of cargo magic for free, and do not even need `cargo-metadata` any more. There's one annoying point though, which I have not managed to entirely work around yet: this means we are doing a full build, not just a check-build. Given that our sysroot is MIR-only, I was surprised that this even worked, but still -- this means we are doing more work than we should. So I also added some patching of arguments to make sure `--emit` does not contain `link`, and then more patching was required of the `--extern` flags for the binary because those referenced the `.rlib` files but now only `.rmeta` exists, and that is still not fully working because cargo seems to expect those `.rmeta` files and now triggers a rebuild each time as those files are still missing. My current plan is to make our wrapper create some empty dummy files with the right names, but the amount of hacks we are stacking on top of each other here is getting worrysome.^^ `@ehuss` your input would be welcome on this issue.
Due to re-using cargo more literally, this also changes flag parsing to match `cargo`. So `-Zmiri` flags now have to be passed via an environment variable (Cc https://github.com/rust-lang/miri/issues/1416).
This PR is not ready yet, but the parts that are there I think can be reviewed already. TODO:
* [x] [Fix Windows](https://github.com/rust-lang/miri/pull/1540#issuecomment-688733741).
* [x] Figure out how we can do check-only builds without the rebuild problem above. ~~I am also worried about cases where `build.rs` script might detect check-only builds and then do less work; I think some crates in rustc are doing that and if this is a thing in the wider ecosystem we need to find a way to support this as well.~~ (postponed that until we have a concrete example)
* [x] Currently cargo runs doctests as well, and they are not run in Miri. We should at least figure out a way to not run them at all (resolving https://github.com/rust-lang/miri/issues/584 is left for a future PR).
* [x] For some time, detect the old way of passing `-Zmiri` flags and warn that this will need updating. For some simple situations we can probably make it still support the old way, but I plan to remove those hacks after a bit. This is just to give people and me time to go around and send PRs to all projects that use Miri on CI, and update their use of the flags.
* [x] Add a test for stdin handling (https://github.com/rust-lang/miri/issues/1505). This should work now but we should be sure.
* [x] Add a test for cargo env vars (https://github.com/rust-lang/miri/issues/1515).
* [x] Check if https://github.com/rust-lang/miri/issues/1516 is resolved.
* [x] Check if https://github.com/rust-lang/miri/issues/1001 and https://github.com/rust-lang/miri/issues/1514 are resolved.
* [x] Check if https://github.com/rust-lang/miri/issues/1312 is resolved (not sure if it is wort adding a test).
* [x] Check if https://github.com/rust-lang/miri/issues/1512 is resolved (not sure if it is wort adding a test).
Fixes https://github.com/rust-lang/miri/issues/700.
Fixes https://github.com/rust-lang/miri/issues/739.
Fixes https://github.com/rust-lang/miri/issues/1001.
Fixes https://github.com/rust-lang/miri/issues/1312 (without a test, as we run without cargo's stdin/stdout wrapping now, as the test for stdin confirms).
Fixes https://github.com/rust-lang/miri/issues/1416.
Fixes https://github.com/rust-lang/miri/issues/1505.
Fixes https://github.com/rust-lang/miri/issues/1512 (without a test, as cargo now does all that handling anyway, which various other tests confirm).
Fixes https://github.com/rust-lang/miri/issues/1514.
Fixes https://github.com/rust-lang/miri/issues/1516.
Cc `@alecmocatta` who reported many of the bugs above; would be great if you could help with the tests e.g. by providing some small examples I could try.
r? `@oli-obk`
Support F_DUPFD on stdin/stdout/stderr
Enable `close`-ing stdin/stdout/stderr
For `dup`, check if FD is `File` first
If not, clone the appropriate standard IO stream
Merge POSIX `close` and `dup` tests into same module
Also, add assertion that `write` on a closed FD returns an error.
Add `dup` as FileDescriptor trait fn
Also:
- Fix `close` so it drops `self` instead of reference to it
- Remove FD clamping in insert_fd_with_min_fd, since FDs 0-2 can be
closed
Fix fs_libc tests
Make error message when closing stdin/out/err more specific
Return io::Result from `FileDescriptor::dup`
Change error message when closing stdin/out/err
Refactor `FileDescriptor::dup` impl for `FileHandle`
Remove empty line
