Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Use upstream intrinsics #257

Merged
merged 1 commit into from
Aug 7, 2019
Merged

Use upstream intrinsics #257

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 1 addition & 8 deletions build.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,13 +1,6 @@
#![cfg_attr(
all(feature = "simd_backend", target_feature = "avx512ifma"),
feature(simd_ffi)
)]
#![cfg_attr(
all(feature = "simd_backend", target_feature = "avx512ifma"),
feature(link_llvm_intrinsics)
)]
#![cfg_attr(all(feature = "alloc", not(feature = "std")), feature(alloc))]
#![cfg_attr(feature = "nightly", feature(doc_cfg))]
#![cfg_attr(feature = "simd_backend", feature(stdsimd))]
#![allow(unused_variables)]
#![allow(non_snake_case)]
#![allow(dead_code)]
Expand Down
24 changes: 13 additions & 11 deletions src/backend/vector/ifma/field.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,12 +14,18 @@ use packed_simd::{u64x4, IntoBits};

use backend::serial::u64::field::FieldElement51;

#[allow(improper_ctypes)]
extern "C" {
#[link_name = "llvm.x86.avx512.vpmadd52l.uq.256"]
fn madd52lo(z: u64x4, x: u64x4, y: u64x4) -> u64x4;
#[link_name = "llvm.x86.avx512.vpmadd52h.uq.256"]
fn madd52hi(z: u64x4, x: u64x4, y: u64x4) -> u64x4;
/// A wrapper around `vpmadd52luq` that works on `u64x4`.
#[inline(always)]
unsafe fn madd52lo(z: u64x4, x: u64x4, y: u64x4) -> u64x4 {
use core::arch::x86_64::_mm256_madd52lo_epu64;
_mm256_madd52lo_epu64(z.into_bits(), x.into_bits(), y.into_bits()).into_bits()
}

/// A wrapper around `vpmadd52huq` that works on `u64x4`.
#[inline(always)]
unsafe fn madd52hi(z: u64x4, x: u64x4, y: u64x4) -> u64x4 {
use core::arch::x86_64::_mm256_madd52hi_epu64;
_mm256_madd52hi_epu64(z.into_bits(), x.into_bits(), y.into_bits()).into_bits()
}

/// A vector of four field elements in radix 2^51, with unreduced coefficients.
Expand Down Expand Up @@ -203,11 +209,7 @@ use subtle::ConditionallySelectable;

impl ConditionallySelectable for F51x4Reduced {
#[inline]
fn conditional_select(
a: &F51x4Reduced,
b: &F51x4Reduced,
choice: Choice,
) -> F51x4Reduced {
fn conditional_select(a: &F51x4Reduced, b: &F51x4Reduced, choice: Choice) -> F51x4Reduced {
let mask = (-(choice.unwrap_u8() as i64)) as u64;
let mask_vec = u64x4::splat(mask);
F51x4Reduced([
Expand Down
8 changes: 1 addition & 7 deletions src/lib.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,17 +9,11 @@
// - Henry de Valence <hdevalence@hdevalence.ca>

#![no_std]
#![cfg_attr(
any(
all(feature = "simd_backend", target_feature = "avx512ifma"),
all(feature = "nightly", rustdoc)
),
feature(simd_ffi, link_llvm_intrinsics)
)]
#![cfg_attr(feature = "nightly", feature(test))]
#![cfg_attr(all(feature = "alloc", not(feature = "std")), feature(alloc))]
#![cfg_attr(feature = "nightly", feature(external_doc))]
#![cfg_attr(feature = "nightly", feature(doc_cfg))]
#![cfg_attr(feature = "simd_backend", feature(stdsimd))]
// Refuse to compile if documentation is missing, but only on nightly.
//
// This means that missing docs will still fail CI, but means we can use
Expand Down