Skip to content

LoongArch64 LSX fast-path for str.contains(&str) #144393

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.

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
merged 1 commit into from
Jul 29, 2025
Merged
Changes from all commits
Commits
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
19 changes: 16 additions & 3 deletions library/core/src/str/pattern.rs
Original file line number Diff line number Diff line change
Expand Up @@ -996,7 +996,10 @@ impl<'b> Pattern for &'b str {
return haystack.as_bytes().contains(&self.as_bytes()[0]);
}

#[cfg(all(target_arch = "x86_64", target_feature = "sse2"))]
#[cfg(any(
all(target_arch = "x86_64", target_feature = "sse2"),
all(target_arch = "loongarch64", target_feature = "lsx")
))]
if self.len() <= 32 {
if let Some(result) = simd_contains(self, haystack) {
return result;
Expand Down Expand Up @@ -1770,11 +1773,18 @@ impl TwoWayStrategy for RejectAndMatch {
/// If we ever ship std with for x86-64-v3 or adapt this for other platforms then wider vectors
/// should be evaluated.
///
/// Similarly, on LoongArch the 128-bit LSX vector extension is the baseline,
/// so we also use `u8x16` there. Wider vector widths may be considered
/// for future LoongArch extensions (e.g., LASX).
///
/// For haystacks smaller than vector-size + needle length it falls back to
/// a naive O(n*m) search so this implementation should not be called on larger needles.
///
/// [0]: http://0x80.pl/articles/simd-strfind.html#sse-avx2
#[cfg(all(target_arch = "x86_64", target_feature = "sse2"))]
#[cfg(any(
all(target_arch = "x86_64", target_feature = "sse2"),
all(target_arch = "loongarch64", target_feature = "lsx")
))]
#[inline]
fn simd_contains(needle: &str, haystack: &str) -> Option<bool> {
let needle = needle.as_bytes();
Expand Down Expand Up @@ -1906,7 +1916,10 @@ fn simd_contains(needle: &str, haystack: &str) -> Option<bool> {
/// # Safety
///
/// Both slices must have the same length.
#[cfg(all(target_arch = "x86_64", target_feature = "sse2"))] // only called on x86
#[cfg(any(
all(target_arch = "x86_64", target_feature = "sse2"),
all(target_arch = "loongarch64", target_feature = "lsx")
))]
#[inline]
unsafe fn small_slice_eq(x: &[u8], y: &[u8]) -> bool {
debug_assert_eq!(x.len(), y.len());
Expand Down
Loading