Skip to content

Simplify runtime SIMD detection #215

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

Closed
CathalMullan wants to merge 1 commit into from
+12 −38
Closed

Simplify runtime SIMD detection #215

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
50 changes: 12 additions & 38 deletions src/simd/runtime.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,34 +1,4 @@
use std::sync::atomic::{AtomicU8, Ordering};
use crate::iter::Bytes;
use super::avx2;
use super::sse42;

const AVX2: u8 = 1;
const SSE42: u8 = 2;
const NOP: u8 = 3;

fn detect_runtime_feature() -> u8 {
if is_x86_feature_detected!("avx2") {
AVX2
} else if is_x86_feature_detected!("sse4.2") {
SSE42
} else {
NOP
}
}

static RUNTIME_FEATURE: AtomicU8 = AtomicU8::new(0);

#[inline]
fn get_runtime_feature() -> u8 {
let mut feature = RUNTIME_FEATURE.load(Ordering::Relaxed);
if feature == 0 {
feature = detect_runtime_feature();
RUNTIME_FEATURE.store(feature, Ordering::Relaxed);
}

feature
}

pub fn match_header_name_vectored(bytes: &mut Bytes) {
super::swar::match_header_name_vectored(bytes);
Expand All @@ -37,21 +7,25 @@ pub fn match_header_name_vectored(bytes: &mut Bytes) {
pub fn match_uri_vectored(bytes: &mut Bytes) {
// SAFETY: calls are guarded by a feature check
unsafe {
match get_runtime_feature() {
AVX2 => avx2::match_uri_vectored(bytes),
SSE42 => sse42::match_uri_vectored(bytes),
_ /* NOP */ => super::swar::match_uri_vectored(bytes),
if is_x86_feature_detected!("avx2") {
super::avx2::match_uri_vectored(bytes)
} else if is_x86_feature_detected!("sse4.2") {
super::sse42::match_uri_vectored(bytes)
} else {
super::swar::match_uri_vectored(bytes)
}
}
}

pub fn match_header_value_vectored(bytes: &mut Bytes) {
// SAFETY: calls are guarded by a feature check
unsafe {
match get_runtime_feature() {
AVX2 => avx2::match_header_value_vectored(bytes),
SSE42 => sse42::match_header_value_vectored(bytes),
_ /* NOP */ => super::swar::match_header_value_vectored(bytes),
if is_x86_feature_detected!("avx2") {
super::avx2::match_header_value_vectored(bytes)
} else if is_x86_feature_detected!("sse4.2") {
super::sse42::match_header_value_vectored(bytes)
} else {
super::swar::match_header_value_vectored(bytes)
}
}
}