Function core::arch::x86::_mm256_movemask_ps

pub unsafe fn _mm256_movemask_ps(a: __m256) -> i32

This is supported on (non-crate feature miri-test-libstd or test or doctest) and x86 and target feature avx only.

Sets each bit of the returned mask based on the most significant bit of the corresponding packed single-precision (32-bit) floating-point element in a.

Intel’s documentation