packed_simd

packed_simd copied to clipboard

I couldn't find the exact location so I think it's actually inside a macro expansion somewhere, possibly inside `packed_simd`.

workingjubilee

Found in https://github.com/rust-lang/packed_simd/pull/315#issuecomment-801564862. The failure is: ``` ---- v128::i8x16_ops_scalar_shifts::ops_scalar_shifts stdout ---- thread 'v128::i8x16_ops_scalar_shifts::ops_scalar_shifts' panicked at 'assertion failed: `(left == right)` left: `i8x16(1, 0, 1, 0, 1, 0, 1, 0, 1,...

JohnTitor

LLVM horizontal reduction intrinsics broken on AArch64

5

These include: * floating-point `sum` and `product` * integer `wrapping_sum` and `wrapping_product` * horizontal min/max reductions: `min_element`/`max_element` * bitwise horizontal reductions: `and`/`or`/`xor` * mask reductions `all`,`any`, `none` Reported to LLVM:...

gnzlbg

Add support for faster shuffles

2

Currently `u32x8` `shuffle1_dyn` are not optimized and fallback is used which results in a whole mess of extract intrinsics. It is not very fast. Can we please add support for...

velvia

Building for 32-bit ARM target fails in src/codegen/shuffle1_dyn.rs

2

This happens both when targeting Android and glibc: ``` (bionic)hsivonen@localhost:~/Projects/packed_simd$ RUSTFLAGS='-Z macro-backtrace' cargo build --target thumbv7neon-unknown-linux-gnueabihf --verbose Compiling cfg-if v0.1.10 Running `rustc --crate-name cfg_if --edition=2018 /home/hsivonen/.cargo/registry/src/github.com-1285ae84e5963aae/cfg-if-0.1.10/src/lib.rs --error-format=json --json=diagnostic-rendered-ansi,artifacts --crate-type lib...

hsivonen

Windows CI used to run on AppVeyor. AppVeyor isn't running right now, though. This probably should be fixed, and any issues it uncovers also resolved to the extent of convenience.

workingjubilee

gnzlbg

Differing shift behavior between u8x16 and u16x8

6

I was executing something similar to the following code: ```rust use packed_simd::{u16x8, u8x16}; fn main() { let x: u16x8 = u16x8::from([0xFFFF; 8]); let counted = x.count_ones(); let x = (u16x8::from([1;...

vadixidav

The `product_roundoff()` test in src/api/reductions/float_arithmetic.rs reports pretty lousy accuracy: there's a 0~3ulp discrepancy between the two multiplication methods on the f32x8 vector test, at least on some platforms. I unblocked...

workingjubilee

Compilation failure on 2019-12-20 nightly: unresolved imports `crate::codegen::pointer_sized_int::isize_`, `crate::codegen::pointer_sized_int::usize_`

1

I'm unable to compule packed_simd using `nightly-2019-12-20-x86_64-unknown-linux-gnu`. The error message is below: ``` Compiling packed_simd v0.3.3 error[E0432]: unresolved imports `crate::codegen::pointer_sized_int::isize_`, `crate::codegen::pointer_sized_int::usize_` --> /home/roger/.cargo/registry/src/github.com-1ecc6299db9ec823/packed_simd-0.3.3/src/codegen/vSize.rs:3:41 | 3 | use crate::codegen::pointer_sized_int::{isize_, usize_}; |...

rocurley