Use a range to identify SIGSEGV in stack guards
Previously, the `guard::init()` and `guard::current()` functions were
returning a `usize` address representing the top of the stack guard,
respectively for the main thread and for spawned threads. The `SIGSEGV`
handler on `unix` targets checked if a fault was within one page below that
address, if so reporting it as a stack overflow.
Now `unix` targets report a `Range<usize>` representing the guard memory,
so it can cover arbitrary guard sizes. Non-`unix` targets which always
return `None` for guards now do so with `Option<!>`, so they don't pay any
overhead.
For `linux-gnu` in particular, the previous guard upper-bound was
`stackaddr + guardsize`, as the protected memory was *inside* the stack.
This was a glibc bug, and starting from 2.27 they are moving the guard
*past* the end of the stack. However, there's no simple way for us to know
where the guard page actually lies, so now we declare it as the whole range
of `stackaddr ± guardsize`, and any fault therein will be called a stack
overflow. This fixes#47863.
decline to lint technically-unnecessary parens in function or method arguments inside of nested macros
In #46980 ("in which the unused-parens lint..." (14982db2d6)), the
unused-parens lint was made to check function and method arguments,
which it previously did not (seemingly due to oversight rather than
willful design). However, in #47775 and discussion thereon,
user–developers of Geal/nom and graphql-rust/juniper reported that the
lint was seemingly erroneously triggering on certain complex macros in
those projects. While this doesn't seem like a bug in the lint in the
particular strict sense that the expanded code would, in fact, contain
unncecessary parentheses, it also doesn't seem like the sort of thing
macro authors should have to think about: the spirit of the
unused-parens lint is to prevent needless clutter in code, not to give
macro authors extra heartache in the handling of token trees.
We propose the expediency of declining to lint unused parentheses in
function or method args inside of nested expansions: we believe that
this should eliminate the petty, troublesome lint warnings reported
in the issue, without forgoing the benefits of the lint in simpler
macros.
It seemed like too much duplicated code for the `Call` and `MethodCall`
match arms to duplicate the nested-macro check in addition to each
having their own `for` loop, so this occasioned a slight refactor so
that the function and method cases could share code—hopefully the
overall intent is at least no less clear to the gentle reader.
This is concerning #47775.
rustc: prefer ParamEnvAnd and LayoutCx over tuples for LayoutOf.
This PR provides `tcx.layout_of(param_env.and(ty))` as the idiomatic replacement for the existing `(tcx, param_env).layout_of(ty)` and removes fragile (coherence-wise) layout-related tuple impls.
r? @nikomatsakis
copy_nonoverlapping example: Fixed typo
The comment referred to a variable using an incorrect name. (it has probably been renamed since the comment was written, or the comment was copied elsewhere - I noted the example in libcore has the `tmp` name for the temporary variable.)
Tweak presentation on lifetime trait mismatch
- On trait/impl method discrepancy, add label pointing at trait signature.
- Point only at method definition when referring to named lifetimes on lifetime mismatch.
- When the sub and sup expectations are the same, tweak the output to avoid repeated spans.
Fix#30790, CC #18759.
The comment referred to a variable using an incorrect name. (it has probably been renamed since the comment was written, or the comment was copied elsewhere - I noted the example in libcore has the `tmp` name for the temporary variable.)
Implement extensible syscall interface for wasm
Currently it's possible to run tests with the native wasm target, but it's not possible to tell whether they pass or to capture the output, because libstd throws away stdout, stderr and the exit code. While advanced libstd features should probably require more specific targets (eg. wasm-unknown-web) I think even the unknown target should at least support basic I/O.
Any solution is constrained by these factors:
- It must not be javascript specific
- There must not be too strong coupling between libstd and the host environment (because it's an "unknown" target)
- WebAssembly does not allow "optional" imports - all imports *must* be resolved.
- WebAssembly does not support calling the host environment through any channel *other* than imports.
The best solution I could find to these constraints was to give libstd a single required import, and implement a syscall-style interface through that import. Each syscall is designed such that a no-op implementation gives the most reasonable fallback behaviour. This means that the following import table would be perfectly valid:
```javascript
imports.env = { rust_wasm_syscall: function(index, data) {} }
```
Currently I have implemented these system calls:
- Read from stdin
- Write to stdout/stderr
- Set the exit code
- Get command line arguments
- Get environment variable
- Set environment variable
- Get time
It need not be extended beyond this set if being able to run tests for this target is the only goal.
edit:
As part of this PR I had to make a further change. Previously, the rust entry point would be automatically called when the webassembly module was instantiated. This was problematic because from the javascript side it was impossible to call exported functions, access program memory or get a reference to the instance.
To solve this, ~I changed the default behaviour to not automatically call the entry point, and added a crate-level attribute to regain the old behaviour. (`#![wasm_auto_run]`)~ I disabled this behaviour when building tests.
Sometimes the parser attempts to synthesize spans from within a macro
context with the span for the captured argument, leading to non-sensical
spans with very bad output. Given that an incorrect span is worse than
a partially incomplete span, when detecting this situation return only
one of the spans without mergin them.
The current `f32|f64.to_degrees` implementation uses a division to calculate 180/π, which causes a loss of precision. Using a constant is still not perfect (implementing a maximally-precise algorithm would come with a high performance cost), but improves precision with a minimal change.
remove intercrate ambiguity hints
The scheme was causing overflows during coherence checking (e.g. #47139). This is sort of a temporary fix; the proper fix I think involves reworking trait selection in deeper ways.
cc @sgrif -- this *should* fix diesel
cc @qnighy -- I'd like to discuss you with alternative techniques for achieving the same end. =) Actually, it might be good to put some energy into refactoring traits first.
r? @eddyb
Add approximate suggestions for rustfix
This adds `span_approximate_suggestion()` that lets you emit a
suggestion marked as "non-machine applicable" in the JSON output. UI
users see no difference. This is for when rustc and clippy wish to
emit suggestions which will make sense to the reader (e.g. they may
have placeholders like `<type>`) but are not source-applicable, so that
rustfix/etc can ignore these.
fixes#39254
Currently ', " and \ are escaped as \', \" and \\ respectively. This
leads to confusing messages such as `error: unknown start of token: \\`
when encountering a single backslash.
Fix by emitting printable ASCII characters directly. This will still
escape \r, \n, \t and Unicode characters.
Fixes#47902
Previously, the `guard::init()` and `guard::current()` functions were
returning a `usize` address representing the top of the stack guard,
respectively for the main thread and for spawned threads. The `SIGSEGV`
handler on `unix` targets checked if a fault was within one page below
that address, if so reporting it as a stack overflow.
Now `unix` targets report a `Range<usize>` representing the guard
memory, so it can cover arbitrary guard sizes. Non-`unix` targets which
always return `None` for guards now do so with `Option<!>`, so they
don't pay any overhead.
For `linux-gnu` in particular, the previous guard upper-bound was
`stackaddr + guardsize`, as the protected memory was *inside* the stack.
This was a glibc bug, and starting from 2.27 they are moving the guard
*past* the end of the stack. However, there's no simple way for us to
know where the guard page actually lies, so now we declare it as the
whole range of `stackaddr ± guardsize`, and any fault therein will be
called a stack overflow. This fixes#47863.
rustc: Move location of `codegen-backends` dir
Right now this directory is located under:
```
$sysroot/lib/rustlib/$target/lib/codegen-backends
```
but after seeing what we do in a few other places it seems that a more
appropriate location would be:
```
$sysroot/lib/rustlib/$target/codegen-backends
```
so this commit moves it!
Avoid underflow in render_source_line
While testing rust-lang/rust#47655 I was able to make the compiler panic when it's compiled with debug assertions:
```shell
> rustc /dev/null --crate-type proc-macro
error: internal compiler error: unexpected panic
note: the compiler unexpectedly panicked. this is a bug.
note: we would appreciate a bug report: https://github.com/rust-lang/rust/blob/master/CONTRIBUTING.md#bug-reports
note: rustc 1.25.0-dev running on x86_64-apple-darwin
note: run with `RUST_BACKTRACE=1` for a backtrace
thread 'rustc' panicked at 'attempt to subtract with overflow', librustc_errors/emitter.rs:287:49
```
Without debug assertions the following warning is emitted:
```shell
> rustc /dev/null --crate-type proc-macro
warning: unused variable: `registrar`
--> /dev/null:0:1
|
|
= note: #[warn(unused_variables)] on by default
= note: to avoid this warning, consider using `_registrar` instead
```
The panic is due to the unused variable warning being spanned to `/dev/null:0:1`. When `render_source_line` subtracts 1 from the line number to look up the source line it panics due to underflow. Without debug assertions this would wrap and cause us to return a blank string instead.
Fix by explicitly testing for 0 and exiting early. I'm unsure how to automatically test this now that rust-lang/rust#46655 has been approved.
Right now this directory is located under:
$sysroot/lib/rustlib/$target/lib/codegen-backends
but after seeing what we do in a few other places it seems that a more
appropriate location would be:
$sysroot/lib/rustlib/$target/codegen-backends
so this commit moves it!
rustc_trans: keep LLVM types for trait objects anonymous.
Fixes#47638 by reverting the addition of readable LLVM trait object type names.
r? @nikomatsakis
[MSP430] Don't enable trap_unreachable option by default on msp.
Since MSP430 doesn't meaningfully support faulting on illegal
instructions, LLVM generates a call to abort() function instead
of a trap instruction. Such calls are 4 bytes long, and that is
too much overhead for such small target.
r? @alexcrichton
rustc: Add an option to default hidden visibility
This commit adds a new option to target specifictions to specify that symbols
should be "hidden" visibility by default in LLVM. While there are no existing
targets that take advantage of this the `wasm32-unknown-unknown` target will
soon start to use this visibility. The LLD linker currently interprets `hidden`
as "don't export this from the wasm module" which is what we want for 90% of our
functions. While the LLD linker does have a "export this symbol" argument which
is what we use for other linkers, it was also somewhat easier to do this change
instead which'll involve less arguments flying around. Additionally there's no
need for non-`hidden` visibility for most of our symbols!
This change should not immediately impact the wasm targets as-is, but rather
this is laying the foundations for soon integrating LLD as a linker for wasm
code.
