char a[16]={1,0,0,1 ,0,0,1,0, 0,1,0,0, 0,0,0,1};
char sum1 = a[0]  + a[1]  + a[2]  + a[3];
char sum2 = a[4]  + a[5]  + a[6]  + a[7];
char sum3 = a[8]  + a[9]  + a[10] + a[11];
char sum4 = a[12] + a[13] + a[14] + a[15];

void myfunc( const char *vec1, const char *vec2, char *vec3, int *counts, int n){
  __m128i *r1   = (__m128i*)vec1;
  __m128i *r2   = (__m128i*)vec2;
  char *a = vec3;
  char temp[16] __attribute__ ((aligned (16)));

  for ( int i = 0; i < n; i+=16, r1++, r2++, a+=16 ) {
    _mm_store_si128((__m128i*)a, _mm_and_si128(*r1, *r2));

    _mm_store_si128((__m128i*)temp, _mm_or_si128(*r1, *r2));

    char size = a[0]+a[1]+a[2]+a[3];
    if( size == 0 ){
        memcpy(a, temp, 4*sizeof(char));
        counts[k]++;
    }
    k++;

    size = a[4]+a[5]+a[6]+a[7];
    if( size == 0 ){
        memcpy(a+4, temp+4, 4*sizeof(char));
        counts[k]++;
    }
    k++;

    size = a[8]+a[9]+a[10]+a[11];
    if( size == 0 ){
        memcpy(a+8, temp+8, 4*sizeof(char));
        counts[k]++;
    }
    k++;

    size = a[12]+a[13]+a[14]+a[15];
    if( size == 0 ){
        memcpy(a+12, temp+12, 4*sizeof(char));
        counts[k]++;
    }
    k++;
  }
}

for ( int i = 0; i < n; i+=16, r1++, r2++, a+=16, k+=4 ) {
    _mm_store_si128((__m128i*)a, _mm_and_si128(*r1, *r2));
    _mm_store_si128((__m128i*)temp, _mm_or_si128(*r1, *r2));

    __m128i a4 = _mm_load_si128((__m128i*)a);
    __m128i tmp4 = _mm_load_si128((__m128i*)tmp);
    __m128i counts4 = _mm_load_si128((__m128i*)&counts[k]);

    __m128i test = _mm_cmpeq_epi32(_mm_set1_epi32(0), a4);
    a4 = _mm_add_epi32(a4, _mm_and_si128(tmp4,test));
    counts4 = _mm_sub_epi32(counts4, test);

    _mm_store_si128((__m128i*)a, a4);
    _mm_store_si128((__m128i*)counts, counts4);
}