Skip to content

Conversation

@Vogtinator
Copy link
Contributor

If the build has no baseline features set, the test ended up setting e.g. NPY_ENABLE_CPU_FEATURES="ASIMDHP, None". This actually made the execution succeed, as the warning for decoding "None" overrode the error for the real feature. Fix the error handling there by removing the errorneous "return 0;", add a test for this, and avoid passing "None" by accident.

If the build has no baseline features set, the test ended up setting
e.g. NPY_ENABLE_CPU_FEATURES="ASIMDHP, None". This actually made the
execution succeed, as the warning for decoding "None" overrode the
error for the real feature. Fix the error handling there by removing
the errorneous "return 0;", add a test for this, and avoid passing
"None" by accident.
@Vogtinator
Copy link
Contributor Author

numpy/meson#12 is tangentially related. Without that PR, on arm it never built ASIMD.

@mattip mattip added triage review Issue/PR to be discussed at the next triage meeting and removed triage review Issue/PR to be discussed at the next triage meeting labels Apr 3, 2024
@mattip mattip merged commit 6ecfb44 into numpy:main Apr 17, 2024
@mattip
Copy link
Member

mattip commented Apr 17, 2024

Thanks @Vogtinator

@Vogtinator Vogtinator deleted the featuretest branch April 18, 2024 06:30
@rgommers rgommers added the component: SIMD Issues in SIMD (fast instruction sets) code or machinery label Oct 14, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

00 - Bug component: SIMD Issues in SIMD (fast instruction sets) code or machinery

Projects

Development

Successfully merging this pull request may close these issues.

3 participants