| Age | Commit message (Collapse) | Author | Lines |
|---|
|
Cleanup src/test/ui/{simd,simd-intrinsic}
Initial motivation was to simplify a huge macro expansion using a tuple, since we can just use an array in `#[repr(simd)]` now for the same result. But also, several tests were going unnoticed during development of SIMD intrinsics because people kept looking in the wrong directory, and many are basically run-pass vs. build-fail versions of the same tests, so let's keep them close together and simplify their names, so they're easier to sift through.
|
|
|
|
Mostly test/ui/simd-intrinsic -> test/ui/simd/intrinsic
Intrinsics-per-se tests moved into that folder
Repetitive names were cut short.
Duplicate names given -pass annotations.
|
|
Simd([u8; 64]) is now valid for repr(simd), so simplify tests with
huge tuples instead. This also found some completely untested code,
so let's just ditch that.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This commit performs two changes to stabilize Rust support for
WebAssembly simd intrinsics:
* The stdarch submodule is updated to pull in rust-lang/stdarch#1179.
* The `wasm_target_feature` feature gate requirement for the `simd128`
feature has been removed, stabilizing the name `simd128`.
This should conclude the FCP started on #74372 and...
Closes #74372
|
|
Allow raw pointers in SIMD types
Closes #85915 by loosening the strictness in typechecking and adding a test to guarantee it passes.
This still might be too strict, as references currently do pass monomorphization, but my understanding is that they are not guaranteed to be "scalar" in the same way.
|
|
This commit updates wasm target specs to use `simd_types_indirect: true`
again. Long ago this was added since wasm simd types were always
translated to `v128` under-the-hood in LLVM, meaning that it didn't
matter whether that target feature was enabled or not. Now, however,
`v128` is conditionally used in codegen depending on target features
enabled, meaning that it's possible to get linker errors about different
signatures in code that correctly uses simd types. The fix is the same
as for all other platforms, which is to pass the type indirectly.
|
|
build-fail test for simd type with an element type that's a wide pointer
test for SIMD element type of optional nonnull ptr to extern type
test that wide ptr simd element type of *mut [u8] fails
|
|
|
|
|
|
Remove rustc_args_required_const attribute
Now that stdarch no longer needs it (thanks `@Amanieu!),` we can kill the `rustc_args_required_const` attribute. This means that lifetime extension of references to temporaries is the only remaining job that promotion is performing. :-)
r? `@oli-obk`
Fixes https://github.com/rust-lang/rust/issues/69493
|
|
|
|
When there are multiple macros in use, it can be difficult to tell
which one was responsible for producing an error.
|
|
LLVM supports many functions from math.h in its IR. Many of these have
single-instruction variants on various platforms. So, let's add them so
std::arch can use them.
Yes, exact comparison is intentional: rounding must always return a
valid integer-equal value, except for inf/NAN.
|
|
This is step 2 towards fixing #77548.
In the codegen and codegen-units test suites, the `//` comment markers
were kept in order not to affect any source locations. This is because
these tests cannot be automatically `--bless`ed.
|
|
|
|
|
|
|
|
|
|
Improve SIMD type element count validation
Resolves rust-lang/stdsimd#53.
These changes are motivated by `stdsimd` moving in the direction of const generic vectors, e.g.:
```rust
#[repr(simd)]
struct SimdF32<const N: usize>([f32; N]);
```
This makes a few changes:
* Establishes a maximum SIMD lane count of 2^16 (65536). This value is arbitrary, but attempts to validate lane count before hitting potential errors in the backend. It's not clear what LLVM's maximum lane count is, but cranelift's appears to be much less than `usize::MAX`, at least.
* Expands some SIMD intrinsics to support arbitrary lane counts. This resolves the ICE in the linked issue.
* Attempts to catch invalid-sized vectors during typeck when possible.
Unresolved questions:
* Generic-length vectors can't be validated in typeck and are only validated after monomorphization while computing layout. This "works", but the errors simply bail out with no context beyond the name of the type. Should these errors instead return `LayoutError` or otherwise provide context in some way? As it stands, users of `stdsimd` could trivially produce monomorphization errors by making zero-length vectors.
cc `@bjorn3`
|
|
|
|
|
|
vectors with non-power-of-two lengths.
|
|
|
|
Update tests to remove old numeric constants
Part of #68490.
Care has been taken to leave the old consts where appropriate, for testing backcompat regressions, module shadowing, etc. The intrinsics docs were accidentally referring to some methods on f64 as std::f64, which I changed due to being contrary with how we normally disambiguate the shadow module from the primitive. In one other place I changed std::u8 to std::ops since it was just testing path handling in macros.
For places which have legitimate uses of the old consts, deprecated attributes have been optimistically inserted. Although currently unnecessary, they exist to emphasize to any future deprecation effort the necessity of these specific symbols and prevent them from being accidentally removed.
|
|
Part of #68490.
Care has been taken to leave the old consts where appropriate, for testing backcompat regressions, module shadowing, etc. The intrinsics docs were accidentally referring to some methods on f64 as std::f64, which I changed due to being contrary with how we normally disambiguate the shadow module from the primitive. In one other place I changed std::u8 to std::ops since it was just testing path handling in macros.
For places which have legitimate uses of the old consts, deprecated attributes have been optimistically inserted. Although currently unnecessary, they exist to emphasize to any future deprecation effort the necessity of these specific symbols and prevent them from being accidentally removed.
|
|
|
|
|
|
|
|
This PR allows using `#[repr(simd)]` on ADTs containing a
single array field:
```rust
#[repr(simd)] struct S0([f32; 4]);
#[repr(simd)] struct S1<const N: usize>([f32; N]);
#[repr(simd)] struct S2<T, const N: usize>([T; N]);
```
This should allow experimenting with portable packed SIMD
abstractions on nightly that make use of const generics.
|
|
Support vectors with fewer than 8 elements for simd_select_bitmask
Resolves the issue raised here: https://github.com/rust-lang/stdarch/issues/310#issuecomment-693730094
|
|
|
|
|
|
|
|
|
|
|
|
|
|
LLVM 8 was released on March 20, 2019, over a year ago.
|
|
ecstatic-morse:promote-shuffle-no-special-case, r=petrochenkov"
This reverts commit 61d3b6dedb1ec1f3e3cbd3d66b1a3453225bc37c, reversing
changes made to c6ad1e2c2a0c7e48537617d36085f866fa6a65a3.
|
|
|
|
Reverts part of #67676
|
|
This extends the invalid cases we catch in const prop to include
overflowing integer casts using the same machinery as the overflowing
binary and unary operation logic.
|
|
LLVM 7 is over a year old, which should be plenty for compatibility. The
last LLVM 6 holdout was llvm-emscripten, which went away in #65501.
I've also included a fix for LLVM 8 lacking `MemorySanitizerOptions`,
which was broken by #66522.
|
|
|
