Skip to content

specialize inflate_fast_help for avx2 #289

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

Merged
folkertdev merged 2 commits into from
Feb 6, 2025
Merged

specialize inflate_fast_help for avx2 #289

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
95b9be0
specialize `inflate_fast_help` for `avx2`
folkertdev Jan 30, 2025
81f2528
comment out the avx512 check (it did nothing anyway)
folkertdev Jan 30, 2025
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
23 changes: 22 additions & 1 deletion zlib-rs/src/inflate.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1856,7 +1856,28 @@ impl State<'_> {
}
}

fn inflate_fast_help(state: &mut State, _start: usize) {
fn inflate_fast_help(state: &mut State, start: usize) {
#[cfg(any(target_arch = "x86_64", target_arch = "x86"))]
if crate::cpu_features::is_enabled_avx2() {
// SAFETY: we've verified the target features
return unsafe { inflate_fast_help_avx2(state, start) };
}

inflate_fast_help_vanilla(state, start);
}

#[cfg(any(target_arch = "x86_64", target_arch = "x86"))]
#[target_feature(enable = "avx2")]
unsafe fn inflate_fast_help_avx2(state: &mut State, start: usize) {
inflate_fast_help_impl(state, start);
}

fn inflate_fast_help_vanilla(state: &mut State, start: usize) {
inflate_fast_help_impl(state, start);
}

#[inline(always)]
fn inflate_fast_help_impl(state: &mut State, _start: usize) {
let mut bit_reader = BitReader::new(&[]);
core::mem::swap(&mut bit_reader, &mut state.bit_reader);

Expand Down
20 changes: 12 additions & 8 deletions zlib-rs/src/inflate/writer.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -77,10 +77,12 @@ impl<'a> Writer<'a> {

#[inline(always)]
pub fn extend_from_window(&mut self, window: &super::window::Window, range: Range<usize>) {
#[cfg(target_arch = "x86_64")]
if crate::cpu_features::is_enabled_avx512() {
return self.extend_from_window_help::<core::arch::x86_64::__m512i>(window, range);
}
// NOTE: the dynamic check for avx512 makes avx2 slower. Measure this carefully before re-enabling
//
// #[cfg(target_arch = "x86_64")]
// if crate::cpu_features::is_enabled_avx512() {
// return self.extend_from_window_help::<core::arch::x86_64::__m512i>(window, range);
// }

#[cfg(target_arch = "x86_64")]
if crate::cpu_features::is_enabled_avx2() {
Expand Down Expand Up @@ -138,10 +140,12 @@ impl<'a> Writer<'a> {

#[inline(always)]
pub fn copy_match(&mut self, offset_from_end: usize, length: usize) {
#[cfg(target_arch = "x86_64")]
if crate::cpu_features::is_enabled_avx512() {
return self.copy_match_help::<core::arch::x86_64::__m512i>(offset_from_end, length);
}
// NOTE: the dynamic check for avx512 makes avx2 slower. Measure this carefully before re-enabling
//
// #[cfg(target_arch = "x86_64")]
// if crate::cpu_features::is_enabled_avx512() {
// return self.copy_match_help::<core::arch::x86_64::__m512i>(offset_from_end, length);
// }

#[cfg(target_arch = "x86_64")]
if crate::cpu_features::is_enabled_avx2() {
Expand Down