Wrong size for char* overwrite

[nick-dumas]

For the types specified in: xsimd_avx_double.hpp, xsimd_sse_float.hpp and xsimd_sse_int32.hpp

The SSE method store_aligned_int32(uint8_t* dst) (and similar) stores using the function _mm_storel_epi64. As this is a batch of 4 values, and each value will be saved to a single byte, this is 32 bits not 64 bits. I believe the correct function should be _mm_cvtsi128_si32 or _mm_storeu_si32.