Skip to content

Prevent error from being fused with scalar in simd_op_check #8867

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
zvookin merged 3 commits into from
Nov 21, 2025
Merged

Prevent error from being fused with scalar in simd_op_check #8867

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
b8ed5c1
Prevent error from being fused with scalar in simd_op_check
stevesuzuki-arm Aug 26, 2025
d525073
Add comments
stevesuzuki-arm Nov 18, 2025
0246ae3
Update vector load*_splat checks in simd_op_check_wasm.cpp
alexreinking Nov 18, 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
3 changes: 3 additions & 0 deletions test/correctness/simd_op_check.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -327,6 +327,9 @@ class SimdOpCheckTest {
// Include a scalar version
Halide::Func f_scalar("scalar_" + name);
f_scalar(x, y) = e;
// Make sure scalar result is computed independently to prevent it
// from being fused into error() by optimization which complicates floating point errors.
f_scalar.compute_root();
Copy link
Member

@zvookin zvookin Nov 18, 2025

Choose a reason for hiding this comment

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

Might be worth a comment as to why this is necessary for correctness.

Copy link
Contributor Author

@stevesuzuki-arm stevesuzuki-arm Nov 18, 2025

Choose a reason for hiding this comment

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

Done


if (has_inline_reduction.result) {
// If there's an inline reduction, we want to vectorize it
Expand Down
10 changes: 5 additions & 5 deletions test/correctness/simd_op_check_wasm.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -385,11 +385,11 @@ class SimdOpCheckWASM : public SimdOpCheckTest {
// check("v128.load32_zero", 2 * w, in_u32(0));
// check("v128.load64_zero", 2 * w, in_u64(0));

// Load vector with identical lanes generates *.splat.
check("i8x16.splat", 16 * w, in_u8(0));
check("i16x8.splat", 8 * w, in_u16(0));
check("i32x4.splat", 4 * w, in_u32(0));
check("i64x2.splat", 2 * w, in_u64(0));
// Load vector with identical lanes generates load*_splat.
check("v128.load8_splat", 16 * w, in_u8(0));
check("v128.load16_splat", 8 * w, in_u16(0));
check("v128.load32_splat", 4 * w, in_u32(0));
check("v128.load64_splat", 2 * w, in_u64(0));

// Load Lane
// TODO: does Halide have any idiom that obviously generates these?
Expand Down
Loading