Skip to content

Conversation

usamoi
Copy link
Contributor

@usamoi usamoi commented Feb 19, 2025

No description provided.

Signed-off-by: usamoi <usamoi@outlook.com>
@usamoi usamoi merged commit 14a95d9 into tensorchord:main Feb 19, 2025
15 checks passed
@usamoi usamoi requested a review from Copilot March 12, 2025 04:07
Copy link
Contributor

@Copilot Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull Request Overview

This pull request introduces a new implementation for u8::reduce_sum_of_x with multiple architecture-specific optimizations. Key changes include:

  • Renaming and updating test functions for the v3 implementation.
  • Adding a new optimized reduce_sum_of_x_v2 for x86_64 and reduce_sum_of_x_a2 for aarch64.
  • Adjusting attributes and inline annotations for improved performance and clarity.

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File Description
crates/simd/src/u8.rs Renamed test function and added new implementations for x86_64 (v2) and aarch64 (a2).
crates/simd/src/emulate.rs Removed a now-unnecessary #[expect(dead_code)] attribute in favor of inline attributes.
Comments suppressed due to low confidence (2)

crates/simd/src/u8.rs:310

  • [nitpick] The test function name changed from reduce_sum_of_x_as_u16_v3_test to reduce_sum_of_x_v3_test. Please confirm that the new name clearly conveys the intent of the test and aligns with the implementation under test.
fn reduce_sum_of_x_v3_test() {

crates/simd/src/u8.rs:367

  • [nitpick] Consider extracting the loop iteration count (256 and 1) into a named constant to improve test readability and maintainability.
for _ in 0..if cfg!(not(miri)) { 256 } else { 1 } {

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant