Skip to content

MSVC/AVX512F: Fix batch_const with 16b and 8b integers. #1150

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 24, 2025

Conversation

degasus
Copy link
Contributor

@degasus degasus commented Jul 24, 2025

There is no _setr_ intrinsic for 16b and 8b types, so we have to flip the order ourself.

Honestly, there is a CI for catching such bugs. How can we missed to test this case?

There is no _setr_ intrinsic for 16b and 8b types, so we have to flip the order ourself.

Honestly, why is there no CI for catching this bug? Can we have an MSVC+AVX512 CI runner please?
@degasus
Copy link
Contributor Author

degasus commented Jul 24, 2025

Oh dear: https://github.com/xtensor-stack/xsimd/blob/master/.github/workflows/windows.yml#L53 ...

FYI: With this patch, ninja xtest is working fine on my Windows+AVX512 machine.

@degasus degasus changed the title AVX512F: Fix batch_const with 16b and 8b integers. MSVC/AVX512F: Fix batch_const with 16b and 8b integers. Jul 24, 2025
@serge-sans-paille
Copy link
Contributor

Thanks! I'll have a look on this CI thing, it's frightening.

@serge-sans-paille serge-sans-paille merged commit ecbb0b4 into xtensor-stack:master Jul 24, 2025
63 checks passed
@serge-sans-paille
Copy link
Contributor

ok, we do test avx512 on Linux, (through emulation) and not on windows (see actions/runner#1069)

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.

2 participants