In investigating [an issue][1] with `panic_implementation` defined in an
executable that's optimized I once again got to rethinking a bit about the
`rustc_std_internal_symbol` attribute as well as weak lang items. We've sort of
been non-stop tweaking these items ever since their inception, and this
continues to the trend.
The crux of the bug was that in the reachability we have a [different branch][2]
for non-library builds which meant that weak lang items (and std internal
symbols) weren't considered reachable, causing them to get eliminiated by
ThinLTO passes. The fix was to basically tweak that branch to consider these
symbols to ensure that they're propagated all the way to the linker.
Along the way I've attempted to erode the distinction between std internal
symbols and weak lang items by having weak lang items automatically configure
fields of `CodegenFnAttrs`. That way most code no longer even considers weak
lang items and they're simply considered normal functions with attributes about
the ABI.
In the end this fixes the final comment of #51342
[1]: https://github.com/rust-lang/rust/issues/51342#issuecomment-414368019
[2]: 35bf1ae257/src/librustc/middle/reachable.rs (L225-L238)
Add crate build test for `thumb*` targets. [IRR-2018-embedded]
## Summary
This PR adds `run-make` test that compiles `cortex-m` crate for all supported `thumb*-none-*` targets using `cargo` and stage2 `rustc`.
- Supported `thumb*-none-*` targets:
- thumbv6m-none-eabi (Bare Cortex-M0, M0+, M1)
- thumbv7em-none-eabi (Bare Cortex-M4, M7)
- thumbv7em-none-eabihf (Bare Cortex-M4F, M7F, FPU, hardfloat)
- thumbv7m-none-eabi (Bare Cortex-M3)
## How to run & Example output
I tested locally and all targets succeeded like below:
```
./x.py clean
./x.py test --target thumbv6m-none-eabi,thumbv7em-none-eabi,thumbv7em-none-eabihf,thumbv7m-none-eabi src/test/run-make
```
```
Check compiletest suite=run-make mode=run-make (x86_64-unknown-linux-gnu -> thumbv6m-none-eabi)
running 5 tests
.....
test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out
```
## How to re-run
Remove `stamp` file for the test run.
```
rm build/x86_64-unknown-linux-gnu/test/run-make/thumb-none-cortex-m/stamp
```
Then run `test`
```
./x.py test --target thumbv6m-none-eabi,thumbv7em-none-eabi,thumbv7em-none-eabihf,thumbv7m-none-eabi src/test/run-make
(snip)
running 5 tests
iiii.
test result: ok. 1 passed; 0 failed; 4 ignored; 0 measured; 0 filtered out
```
## Artifacts
You can examine the artifacts under the directory below:
```
sekineh@sekineh-VirtualBox:~/rustme10$ ls -l build/x86_64-unknown-linux-gnu/test/run-make/thumb-none-cortex-m/thumb-none-cortex-m/
total 4
drwxrwxr-x 7 sekineh sekineh 4096 8月 14 22:40 cortex-m
```
where `build/x86_64-unknown-linux-gnu/test/run-make/thumb-none-cortex-m/thumb-none-cortex-m/` is came from TMPDIR variable.
## Other notes
For `test.rs` modification, I used the same logic as:
- https://github.com/rust-lang/rust/blame/d8b3c830fbcdd14d085209a8dcc3399151f3286a/src/bootstrap/dist.rs#L652-L657
```
if builder.no_std(target) == Some(true) {
// the `test` doesn't compile for no-std targets
builder.ensure(compile::Std { compiler, target });
} else {
builder.ensure(compile::Test { compiler, target });
}
```
It is a useful snippet when adding `no_std` support to `src/bootstrap` code.
CC @kennytm @jamesmunns @nerdyvaishali
This commit tweaks the linker-level visibility of some lang items that rustc
uses and defines. Notably this means that `#[panic_implementation]` and
`#[alloc_error_handler]` functions are never marked as `internal`. It's up to
the linker to eliminate these, not rustc.
Additionally `#[global_allocator]` generated symbols are no longer forced to
`Default` visibility (fully exported), but rather they're relaxed to `Hidden`
visibility). This symbols are *not* needed across DLL boundaries, only as a
local implementation detail of the compiler-injected allocator symbols, so
`Hidden` should suffice.
Closes#51342Closes#52795
rustc: Work around an upstream wasm ThinLTO bug
This commit implements a workaround for an [upstream LLVM bug][1] where custom
sections were accidentally duplicated amongst codegen units when ThinLTO passes
were performed. This is due to the fact that custom sections for wasm are stored
as metadata nodes which are automatically imported into modules when ThinLTO
happens. The fix here is to forcibly delete the metadata node from imported
modules before LLVM has a chance to try to copy it over.
[1]: https://bugs.llvm.org/show_bug.cgi?id=38184
This commit implements a workaround for an [upstream LLVM bug][1] where custom
sections were accidentally duplicated amongst codegen units when ThinLTO passes
were performed. This is due to the fact that custom sections for wasm are stored
as metadata nodes which are automatically imported into modules when ThinLTO
happens. The fix here is to forcibly delete the metadata node from imported
modules before LLVM has a chance to try to copy it over.
[1]: https://bugs.llvm.org/show_bug.cgi?id=38184
This commit stabilizes the `#[wasm_import_module]` attribute as
`#[link(wasm_import_module = "...")]`. Tracked by #52090 this new directive in
the `#[link]` attribute is used to configured the module name that the imports
are listed with. The WebAssembly specification indicates two utf-8 names are
associated with all imported items, one for the module the item comes from and
one for the item itself. The item itself is configurable in Rust via its
identifier or `#[link_name = "..."]`, but the module name was previously not
configurable and defaulted to `"env"`. This commit ensures that this is also
configurable.
Closes#52090
This commit transitions definitions of custom sections on the wasm target from
the unstable `#[wasm_custom_section]` attribute to the
already-stable-for-other-targets `#[link_section]` attribute. Mostly the same
restrictions apply as before, except that this now applies only to statics.
Closes#51088
This commit removes allocation of the panic message in instances like
`panic!("foo: {}", "bar")` if we don't actually end up needing the message. We
don't need it in the case of wasm32 right now, and in general it's not needed
for panic=abort instances that use the default panic hook.
For now this commit only solves the wasm use case where with LTO the allocation
is entirely removed, but the panic=abort use case can be implemented at a later
date if needed.
This commit applies a few code size optimizations for the wasm target to
the standard library, namely around panics. We notably know that in most
configurations it's impossible for us to print anything in
wasm32-unknown-unknown so we can skip larger portions of panicking that
are otherwise simply informative. This allows us to get quite a nice
size reduction.
Finally we can also tweak where the allocation happens for the
`Box<Any>` that we panic with. By only allocating once unwinding starts
we can reduce the size of a panicking wasm module from 44k to 350 bytes.
This commit adds a new attribute to the Rust compiler specific to the wasm
target (and no other targets). The `#[wasm_import_module]` attribute is used to
specify the module that a name is imported from, and is used like so:
#[wasm_import_module = "./foo.js"]
extern {
fn some_js_function();
}
Here the import of the symbol `some_js_function` is tagged with the `./foo.js`
module in the wasm output file. Wasm-the-format includes two fields on all
imports, a module and a field. The field is the symbol name (`some_js_function`
above) and the module has historically unconditionally been `"env"`. I'm not
sure if this `"env"` convention has asm.js or LLVM roots, but regardless we'd
like the ability to configure it!
The proposed ES module integration with wasm (aka a wasm module is "just another
ES module") requires that the import module of wasm imports is interpreted as an
ES module import, meaning that you'll need to encode paths, NPM packages, etc.
As a result, we'll need this to be something other than `"env"`!
Unfortunately neither our version of LLVM nor LLD supports custom import modules
(aka anything not `"env"`). My hope is that by the time LLVM 7 is released both
will have support, but in the meantime this commit adds some primitive
encoding/decoding of wasm files to the compiler. This way rustc postprocesses
the wasm module that LLVM emits to ensure it's got all the imports we'd like to
have in it.
Eventually I'd ideally like to unconditionally require this attribute to be
placed on all `extern { ... }` blocks. For now though it seemed prudent to add
it as an unstable attribute, so for now it's not required (as that'd force usage
of a feature gate). Hopefully it doesn't take too long to "stabilize" this!
cc rust-lang-nursery/rust-wasm#29
This commit is an implementation of adding custom sections to wasm artifacts in
rustc. The intention here is to expose the ability of the wasm binary format to
contain custom sections with arbitrary user-defined data. Currently neither our
version of LLVM nor LLD supports this so the implementation is currently custom
to rustc itself.
The implementation here is to attach a `#[wasm_custom_section = "foo"]`
attribute to any `const` which has a type like `[u8; N]`. Other types of
constants aren't supported yet but may be added one day! This should hopefully
be enough to get off the ground with *some* custom section support.
The current semantics are that any constant tagged with `#[wasm_custom_section]`
section will be *appended* to the corresponding section in the final output wasm
artifact (and this affects dependencies linked in as well, not just the final
crate). This means that whatever is interpreting the contents must be able to
interpret binary-concatenated sections (or each constant needs to be in its own
custom section).
To test this change the existing `run-make` test suite was moved to a
`run-make-fulldeps` folder and a new `run-make` test suite was added which
applies to all targets by default. This test suite currently only has one test
which only runs for the wasm target (using a node.js script to use `WebAssembly`
in JS to parse the wasm output).
travis: Upgrade OSX builders
This upgrades the OSX builders to the `xcode9.3-moar` image which has 3 cores as
opposed to the 2 that our builders currently have. Should help make those OSX
builds a bit speedier!
Add relro-level tests
The `relro-level` debugging flag was added in #43170 which was merged in July 2017. This PR moves this flag to be a proper codegen flag.
This upgrades the OSX builders to the `xcode9.3-moar` image which has 3 cores as
opposed to the 2 that our builders currently have. Should help make those OSX
builds a bit speedier!
Currently rustc isn't always the best at producing deterministic builds of a
crate when the source directory of a crate is changed. This is happening due to
what appears two different sources:
* First the `-L` paths passed to rustc are hashed into the crate hash. These
paths through Cargo are typically absolute paths that can vary if the build
directory changes.
* Next the paths passed to `--extern` are also hashed which like `-L` can change
if the build directory changes.
This commit fixes these two sources of nondeterminism by ensuring that avoiding
tracking the hashes of these arguments on the command line. For `-L` paths
they're either related to loading crates (whose hashes are tracked elsewhere) or
native librarise used in the linking phase (which isn't incremental). The
`--extern` paths are similar in that they're related to crate resolution which
is already tracked independently of the command line arguments.
Closes#48019
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.
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.
