SSE内在 - _mm_and_ps古怪行为
下面一段代码:SSE内在 - _mm_and_ps古怪行为
__m128 a = _mm_setr_ps(1, 2, 3, 4); __m128 b = _mm_set1_ps(2);
__m128 res = _mm_and_ps(a, b);
cout << a[0] << " " << a[1] << " " << a[2] << " " << a[3] << endl;
cout << b[0] << " " << b[1] << " " << b[2] << " " << b[3] << endl;
cout << res[0] << " " << res[1] << " " << res[2] << " " << res[3] << endl;
cout<<endl;
cout << (1 & 2) << " " << (2 & 2) << " " << (3 & 2) << " " << (4 & 2) << endl;
结果:
1 2 3 4 2 2 2 2
0 2 2 2
0 2 2 0
不应在SSE操作的结果是因为0 2 2 0
2 = 010, 4 = 100 => 2&4 = 0
。
根据文档:
__m128 _mm_and_ps(__ M128一个,__m128 B)
计算的按位的a和b的四个SP FP值的AND。
R0 R1 R2 R3
A0 & B0 A1 & B1 A2 & B2 A3 & B3
回答:
的documentation我发现说:
计算四个的位与单精度,浮点数 a和b的值。
(我的强调)
2和4将具有相同的尾数(0,加上一个隐含的前导1个比特)分别为128和129,和指数。按位和这些是零尾数和128(= 2.0)的指数。
编辑
如果你想要做一个逐位非负整数,并且,你可以加一个偏移量。如果使用8388608的偏移量(== 1 < < 23),那么您可以像预期的那样对0..8388607执行按位运算。
const float offset=8388608; __m128 mm_offset = _mm_set1_ps();
__m128 a = _mm_setr_ps(1, 2, 3, 4);
a =_mm_add_ps(mm_offset,a);
__m128 b = _mm_set1_ps(2+offset);
__m128 res = _mm_and_ps(a, b);
res = _mm_sub_ps(res,mm_offset);
以上是 SSE内在 - _mm_and_ps古怪行为 的全部内容, 来源链接: utcz.com/qa/264730.html