Fix UB in _mm_movemask_ps, _mm_movemask_pd, _mm256_movemask_ps and _mm256_movemask_pd

[eduardosm]

The simd_bitmask intrinsic requires each element to be all-1 or all-0, while the x86 functions only check for the highest bit.

[rustbot]

r? @Amanieu

(rustbot has picked a reviewer for you, use r? to override)

[eduardosm]

Hmmm

LLVM ERROR: IO failure on output stream: No space left on device
error: could not compile `core_arch` (lib)
Error: Process completed with exit code 101.

[Amanieu]

Can you do a quick test on godbolt to ensure this still compiles down to 1 instruction?

[eduardosm]

Of course: https://godbolt.org/z/n6EMz7dbv