`codegen_allocator` and `write_metadata` mutate the underlying LLVM module. As
such, it makes sense for these two functions to receive a mutable reference to
the module (as opposed to an immutable one).
Stabilize slice_sort_by_cached_key
I was going to ask on the tracking issue (https://github.com/rust-lang/rust/issues/34447), but decided to just send this and hope for an FCP here. The method was added last March by https://github.com/rust-lang/rust/pull/48639.
Signature: https://doc.rust-lang.org/std/primitive.slice.html#method.sort_by_cached_key
```rust
impl [T] {
pub fn sort_by_cached_key<K, F>(&mut self, f: F)
where F: FnMut(&T) -> K, K: Ord;
}
```
That's an identical signature to the existing `sort_by_key`, so I think the questions are just naming, implementation, and the usual "do we want this?".
The implementation seems to have proven its use in rustc at least, which many uses: https://github.com/rust-lang/rust/search?l=Rust&q=sort_by_cached_key
(I'm asking because it's exactly what I just needed the other day:
```rust
all_positions.sort_by_cached_key(|&n|
data::CITIES.iter()
.map(|x| *metric_closure.get_edge(n, x.pos).unwrap())
.sum::<usize>()
);
```
since caching that key is a pretty obviously good idea.)
Closes#34447
rustc: Implement incremental "fat" LTO
Currently the compiler will produce an error if both incremental
compilation and full fat LTO is requested. With recent changes and the
advent of incremental ThinLTO, however, all the hard work is already
done for us and it's actually not too bad to remove this error!
This commit updates the codegen backend to allow incremental full fat
LTO. The semantics are that the input modules to LTO are all produce
incrementally, but the final LTO step is always done unconditionally
regardless of whether the inputs changed or not. The only real
incremental win we could have here is if zero of the input modules
changed, but that's so rare it's unlikely to be worthwhile to implement
such a code path.
cc #57968
cc rust-lang/cargo#6643
Cosmetic improvements to doc comments
This has been factored out from https://github.com/rust-lang/rust/pull/58036 to only include changes to documentation comments (throughout the rustc codebase).
r? @steveklabnik
Once you're happy with this, maybe we could get it through with r=1, so it doesn't constantly get invalidated? (I'm not sure this will be an issue, but just in case...) Anyway, thanks for your advice so far!
Currently the compiler will produce an error if both incremental
compilation and full fat LTO is requested. With recent changes and the
advent of incremental ThinLTO, however, all the hard work is already
done for us and it's actually not too bad to remove this error!
This commit updates the codegen backend to allow incremental full fat
LTO. The semantics are that the input modules to LTO are all produce
incrementally, but the final LTO step is always done unconditionally
regardless of whether the inputs changed or not. The only real
incremental win we could have here is if zero of the input modules
changed, but that's so rare it's unlikely to be worthwhile to implement
such a code path.
cc #57968
cc rust-lang/cargo#6643
The layout of a extern_type static is unsized, but may pass the
Well-Formed check in typeck. As a result, we cannot assume that
a static is sized when generating the `Place` for an r-value.
NVPTX target specification
This change adds a built-in `nvptx64-nvidia-cuda` GPGPU no-std target specification and a basic PTX assembly smoke tests.
The approach is taken here and the target spec is based on `ptx-linker`, a project started about 1.5 years ago. Key feature: bitcode object files being linked with LTO into the final module on the linker's side.
Prior to this change, the linker used a `ld` linker-flavor, but I think, having the special CLI convention is a more reliable way.
Questions about further progress on reliable CUDA workflow with Rust:
1. Is it possible to create a test suite `codegen-asm` to verify end-to-end integration with LLVM backend?
1. How would it be better to organise no-std `compile-fail` tests: add `#![no_std]` where possible and mark others as `ignore-nvptx` directive, or alternatively, introduce `compile-fail-no-std` test suite?
1. Can we have the `ptx-linker` eventually be integrated as `rls` or `clippy`? Hopefully, this should allow to statically link against LLVM used in Rust and get rid of the [current hacky solution](https://github.com/denzp/rustc-llvm-proxy).
1. Am I missing some methods from `rustc_codegen_ssa:🔙:linker::Linker` that can be useful for bitcode-only linking?
Currently, there are no major public CUDA projects written in Rust I'm aware of, but I'm expecting to have a built-in target will create a solid foundation for further experiments and awesome crates.
Related to #38789Fixes#38787Fixes#38786
Use pinning for generators to make trait safe
I'm unsure whether there needs to be any changes to the actual generator transform. Tests are passing so the fact that `Pin<&mut T>` is fundamentally the same as `&mut T` seems to allow it to still work, but maybe there's something subtle here that could go wrong.
This is specified in [RFC 2349 § Immovable generators](https://github.com/rust-lang/rfcs/blob/master/text/2349-pin.md#immovable-generators) (although, since that RFC it has become safe to create an immovable generator, and instead it's unsafe to resume any generator; with these changes both are now safe and instead the unsafety is moved to creating a `Pin<&mut [static generator]>` which there are safe APIs for).
CC #43122
Implement optimize(size) and optimize(speed) attributes
This PR implements both `optimize(size)` and `optimize(speed)` attributes.
While the functionality itself works fine now, this PR is not yet complete: the code might be messy in places and, most importantly, the compiletest must be improved with functionality to run tests with custom optimization levels. Otherwise the new attribute cannot be tested properly. Oh, and not all of the RFC is implemented – attribute propagation is not implemented for example.
# TODO
* [x] Improve compiletest so that tests can be written;
* [x] Assign a proper error number (E9999 currently, no idea how to allocate a number properly);
* [ ] Perhaps reduce the duplication in LLVM attribute assignment code…
Don't ICE when logging unusual types
MonoItemExt#to_string is used for both debug logging and LLVM symbol
name generation. When debugging, we want to print out any type we
encounter, even if it's something weird like GeneratorWitness. However,
during codegen, we still want to error if we encounter an unexpected
type when generating a name.
To resolve this issue, this commit introduces a new 'debug' parameter to
the relevant methods. When set to 'true', it allows any type to be
printed - when set to 'false', it 'bug!'s when encountering an
unexpected type.
This prevents an ICE when enabling debug logging (via RUST_LOG) while
running rustc on generator-related code.
MonoItemExt#to_string is used for both debug logging and LLVM symbol
name generation. When debugging, we want to print out any type we
encounter, even if it's something weird like GeneratorWitness. However,
during codegen, we still want to error if we encounter an unexpected
type when generating a name.
To resolve this issue, this commit introduces a new 'debug' parameter to
the relevant methods. When set to 'true', it allows any type to be
printed - when set to 'false', it 'bug!'s when encountering an
unexpected type.
This prevents an ICE when enabling debug logging (via RUST_LOG) while
running rustc on generator-related code.
Add a target option "merge-functions", and a corresponding -Z flag (works around #57356)
This commit adds a target option "merge-functions", which takes values in ("disabled", "trampolines", or "aliases" (default is "aliases")), to allow targets to opt out of the MergeFunctions LLVM pass. Additionally, the latest commit also adds an optional -Z flag, "merge-functions", which takes the same values and has precedence over the target option when both are specified.
This works around https://github.com/rust-lang/rust/issues/57356.
cc @eddyb @japaric @oli-obk @nox @nagisa
Also thanks to @denzp and @gnzlbg for discussing this on rust-cuda!
### Motivation
Basically, the problem is that the MergeFunctions pass, which rustc currently enables by default at -O2 and -O3 [1], and `extern "ptx-kernel"` functions (specific to the NVPTX target) are currently not compatible with each other. If the MergeFunctions pass is allowed to run, rustc can generate invalid PTX assembly (i.e. a PTX file that is not accepted by the native PTX assembler `ptxas`). Therefore we would like a way to opt out of the MergeFunctions pass, which is what our target option does.
### Related work
The current behavior of rustc is to enable MergeFunctions at -O2 and -O3 [1], and also to enable the use of function aliases within MergeFunctions [2] [3]. MergeFunctions seems to have some benefits, such as reducing code size and fixing a crash [4], which is why it is enabled. However, MergeFunctions both with and without function aliases is incompatible with the NVPTX target; a more detailed example for both cases is given below.
clang's "solution" is to have a "-fmerge-functions" flag that opts in to the MergeFunctions pass, but it is not enabled by default.
### Examples/more details
Consider an example Rust lib using `extern "ptx-kernel"` functions: https://github.com/peterhj/nvptx-mergefunc-bug/blob/master/nocore.rs. If we try to compile this with nightly rustc, we get the following compiler error:
LLVM ERROR: Module has aliases, which NVPTX does not support.
This error happens because: (1) functions `foo` and `bar` have the same body, so are candidates to be merged by MergeFunctions; and (2) rustc configures MergeFunctions to generate function aliases using the "mergefunc-use-aliases" LLVM option [2] [3], but the NVPTX backend does not support those aliases.
Okay, so we can try omitting "mergefunc-use-aliases", and then rustc will happily emit PTX assembly: https://github.com/peterhj/nvptx-mergefunc-bug/blob/master/nocore-mergefunc-nousealiases-bad.ptx. However, this PTX is invalid! When we try to assemble it with `ptxas` (I'm on the CUDA 9.2 toolchain), we get an assembler error:
ptxas nocore-mergefunc-nousealiases-bad.ptx, line 38; error : Illegal call target, device function expected
ptxas fatal : Ptx assembly aborted due to errors
What's happening is that MergeFunctions rewrites the `bar` function to call `foo`. However, directly calling an `extern "ptx-kernel"` function from another `extern "ptx-kernel"` is wrong.
If we disable the MergeFunctions pass from running at all, rustc generates correct PTX assembly: https://github.com/peterhj/nvptx-mergefunc-bug/blob/master/nocore-nomergefunc-ok.ptx
[1] a36b960df6/src/librustc_codegen_ssa/back/write.rs (L155)
[2] a36b960df6/src/librustc_codegen_llvm/llvm_util.rs (L64)
[3] https://github.com/rust-lang/rust/pull/56358
[4] https://github.com/rust-lang/rust/pull/49479
rustc: Place wasm linker args first instead of last
This ensures that arguments passed via `-C link-arg` can override the
first ones on the command line, for example allowing configuring of the
stack size.
"trampolines", or "aliases (the default)) to allow targets to opt out of
the MergeFunctions LLVM pass. Also add a corresponding -Z option with
the same name and values.
This works around: https://github.com/rust-lang/rust/issues/57356
Motivation:
Basically, the problem is that the MergeFunctions pass, which rustc
currently enables by default at -O2 and -O3, and `extern "ptx-kernel"`
functions (specific to the NVPTX target) are currently not compatible
with each other. If the MergeFunctions pass is allowed to run, rustc can
generate invalid PTX assembly (i.e. a PTX file that is not accepted by
the native PTX assembler ptxas). Therefore we would like a way to opt
out of the MergeFunctions pass, which is what our target option does.
Related work:
The current behavior of rustc is to enable MergeFunctions at -O2 and -O3,
and also to enable the use of function aliases within MergeFunctions.
MergeFunctions both with and without function aliases is incompatible with
the NVPTX target.
clang's "solution" is to have a "-fmerge-functions" flag that opts in to
the MergeFunctions pass, but it is not enabled by default.
panic when calling MaybeUninhabited::into_inner on uninhabited type
I do this by adding an internal-only intrinsic `panic_if_uninhabited`. I have no idea what I am doing here, just mindlessly copying code around, so please review carefully!
Add support for trait-objects without a principal
The hard-error version of #56481 - should be merged after we do something about the `traitobject` crate.
Fixes#33140.
Fixes#57057.
r? @nikomatsakis
This ensures that arguments passed via `-C link-arg` can override the
first ones on the command line, for example allowing configuring of the
stack size.
