rustc: Tweak funclet cleanups of ffi functions
This commit is targeted at addressing #48251 by specifically fixing a case where
a longjmp over Rust frames on MSVC runs cleanups, accidentally running the
"abort the program" cleanup as well. Added in #46833 `extern` ABI functions in
Rust will abort the process if Rust panics, and currently this is modeled as a
normal cleanup like all other destructors.
Unfortunately it turns out that `longjmp` on MSVC is implemented with SEH, the
same mechanism used to implement panics in Rust. This means that `longjmp` over
Rust frames will run Rust cleanups (even though we don't necessarily want it
to). Notably this means that if you `longjmp` over a Rust stack frame then that
probably means you'll abort the program because one of the cleanups will abort
the process.
After some discussion on IRC it turns out that `longjmp` doesn't run cleanups
for *caught* exceptions, it only runs cleanups for cleanup pads. Using this
information this commit tweaks the codegen for an `extern` function to
a catch-all clause for exceptions instead of a cleanup block. This catch-all is
equivalent to the C++ code:
try {
foo();
} catch (...) {
bar();
}
and in fact our codegen here is designed to match exactly what clang emits for
that C++ code!
With this tweak a longjmp over Rust code will no longer abort the process. A
longjmp will continue to "accidentally" run Rust cleanups (destructors) on MSVC.
Other non-MSVC platforms will not rust destructors with a longjmp, so we'll
probably still recommend "don't have destructors on the stack", but in any case
this is a more surgical fix than #48567 and should help us stick to standard
personality functions a bit longer.
Fix find_width_of_character_at_span bounds check
Commit 0bd96671f0 added bounds checking of our current target byte position to prevent infinite loops. Unfortunately it was comparing the file-relative `target` versus the global `file_start_pos` and `file_end_pos`.
The result is failing to detect multibyte characters unless their file-relative offset fit within their global offset. This causes other parts of the compiler to generate spans pointing to the middle of a
multibyte character which will ultimately panic in `bytepos_to_file_charpos`.
Fix by comparing the `target` to the total file size when moving forward and doing checked subtraction when moving backwards. This should preserve the intent of the bounds check while removing the offset confusion.
cc @davidtwco
Fixes#48508
Support parentheses in patterns under feature gate
This is a prerequisite for any other extensions to pattern syntax - `|` with multiple patterns, type ascription, `..PAT` in slice patterns.
Closes https://github.com/rust-lang/rfcs/issues/554
[Underspecified semantics] Type check defaults at declaration.
Fixes #46669. See the test for code that compiles on stable but will no longer compile. This falls under a "Underspecified language semantics" fix. **Needs crater**.
On type and trait declarations, we currently allow anything that name checks as a type parameter default. That allows the user to write a default that can never be applied, or even a default that may conditionally be applied depending on the type of another parameter. Mostly this just defers the error to use sites, but also allows clever hacks such as `Foo<T, U = <T as Iterator>::Item>` where `U` will be able to apply it's default only when `T: Iterator`. Maybe that means this bug is a feature, but it's a fiddly behaviour that seems undesirable.
This PR validates defaults at declaration sites by ensuring all predicates on the parameter are valid for the default. With the exception of `Self: Sized` which we don't want to check to allow things like `trait Add<RHS = Self>`.
This commit is targeted at addressing #48251 by specifically fixing a case where
a longjmp over Rust frames on MSVC runs cleanups, accidentally running the
"abort the program" cleanup as well. Added in #46833 `extern` ABI functions in
Rust will abort the process if Rust panics, and currently this is modeled as a
normal cleanup like all other destructors.
Unfortunately it turns out that `longjmp` on MSVC is implemented with SEH, the
same mechanism used to implement panics in Rust. This means that `longjmp` over
Rust frames will run Rust cleanups (even though we don't necessarily want it
to). Notably this means that if you `longjmp` over a Rust stack frame then that
probably means you'll abort the program because one of the cleanups will abort
the process.
After some discussion on IRC it turns out that `longjmp` doesn't run cleanups
for *caught* exceptions, it only runs cleanups for cleanup pads. Using this
information this commit tweaks the codegen for an `extern` function to
a catch-all clause for exceptions instead of a cleanup block. This catch-all is
equivalent to the C++ code:
try {
foo();
} catch (...) {
bar();
}
and in fact our codegen here is designed to match exactly what clang emits for
that C++ code!
With this tweak a longjmp over Rust code will no longer abort the process. A
longjmp will continue to "accidentally" run Rust cleanups (destructors) on MSVC.
Other non-MSVC platforms will not rust destructors with a longjmp, so we'll
probably still recommend "don't have destructors on the stack", but in any case
this is a more surgical fix than #48567 and should help us stick to standard
personality functions a bit longer.
Given a trait predicate, if all params appearing in the LHS have
defaults then it should be a backwards compatible predicate. We verify
that by checking the WF of predicate with all defaults substituted
simultaneously.
Fix FreeBSD struct returning ABI.
FreeBSD has had a patch similar to this for a while. See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=223047.
This reworks 6774e7a to be more specific about what `compute_abi_info` is checking for per target.
Fix error-format to properly send JSON to stdout
Since we take Cargo's JSON messages as well we need to specifically send
rustc's messages out so we don't hide them.
r? @Manishearth
Ensure main() always has external linkage
This ensures that the entry function is never elided due to inlining, even with `inline(always)`. Fixes#47783.
There were a couple of possible ways of addressing this issue; I simply picked the one that seemed most direct. A warning could be appropriate, but considering using inlining hints in other places it doesn't apply also throws no warnings, and it seems like an edge case anyway, I haven't added one for now.
Restrict the Termination impls to simplify stabilization
Make a minimal commitment in preparation for stabilization. More impls, or broader ones, are likely in future, but are not necessary at this time and are more controversial.
cc https://github.com/rust-lang/rust/issues/48453#issuecomment-368155082
r? @nikomatsakis
Handle gdb command failure gracefully in compiletest
Previously, if the gdb command was available, but threw an error, compiletest would panic. This is obviously not good. Now, gdb is treated as missing if calling `gdb --version` does not output anything on stdout.
Add support for powerpc-unknown-linux-gnuspe
This PR adds support for the embedded PowerPC variant "e500". On Linux, this architecture is usually called "powerpcspe", it is a 32-bit PowerPC architecture. The main difference between normal 32-bit PowerPC and PowerPCSPE is the lack of Altivec instructions and the additional SPE instruction set.
This architecture is supported in Debian through an unofficial port.
RefCell: document panics in Clone, PartialEq, PartialOrd, Ord.
This fixes#47400 by adding:
```rust
/// # Panics
///
/// Panics if the value is currently mutably borrowed.
```
to said impls. They may panic since they call `.borrow()`.
Add non-panicking variants of pow for integer types
Currently, calling pow may panic in case of overflow, and the function does not have non-panicking counterparts. Thus, it would be beneficial to add those in.
Closes#48291.
Relevant tracking issue: #48320
Report non-standard compile flags on ICE
Some ICEs (such as the recent #48248) only happens when a non-standard compiler flag is provided to rustc, but users don't always report the used flags. This can slow down reproducing the issue, so this PR shows all the non-standard compiler flags in the ICE error message.
For example, the output of #48248 with this PR is:
```
error: internal compiler error: [...]
thread 'rustc' panicked at [...]
note: Run with `RUST_BACKTRACE=1` for a backtrace.
note: the compiler unexpectedly panicked. this is a bug.
note: we would appreciate a bug report: [...]
note: rustc 1.25.0-dev running on x86_64-unknown-linux-gnu
note: compiler flags: -C link-dead-code
```
### Open questions
* At the moment, only `-C` and `-Z` flags are shown by default, and all the ones provided by cargo in a standard build are ignored: I did this to only show the flags that probably caused the ICE, and to remove some noise from the message. This removed flags like `opt-level` and `debuginfo` though, could those be useful for reproducing ICEs?
