blob: 21268607c281218fffab9858723a6b8e2766567d (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
|
#define __rose_lt(x,y) ((x)<(y)?(x):(y))
#define __rose_gt(x,y) ((x)>(y)?(x):(y))
void mm(float **A,float **B,float **C,int ambn,int an,int bm)
{
int t6;
int t4;
int t2;
int chill_t12;
int chill_t11;
int chill_t10;
int chill_t9;
int chill_t8;
int chill_t7;
int chill_t6;
int chill_t5;
int chill_t4;
int over1;
int chill_t3;
int chill_t2;
int chill_t1;
int i;
int j;
int n;
over1 = 0;
for (t2 = 0; t2 <= an - 1; t2 += 1) {
over1 = bm % 4;
for (t4 = 0; t4 <= -over1 + bm - 1; t4 += 4) {
C[t2][t4] = 0.0f;
C[t2][t4] += A[t2][0] * B[0][t4];
C[t2][t4 + 1] = 0.0f;
C[t2][t4 + 1] += A[t2][0] * B[0][t4 + 1];
C[t2][t4 + 2] = 0.0f;
C[t2][t4 + 2] += A[t2][0] * B[0][t4 + 2];
C[t2][t4 + 3] = 0.0f;
C[t2][t4 + 3] += A[t2][0] * B[0][t4 + 3];
for (t6 = 1; t6 <= ambn - 1; t6 += 1) {
C[t2][t4] += A[t2][t6] * B[t6][t4];
C[t2][t4 + 1] += A[t2][t6] * B[t6][t4 + 1];
C[t2][t4 + 2] += A[t2][t6] * B[t6][t4 + 2];
C[t2][t4 + 3] += A[t2][t6] * B[t6][t4 + 3];
}
}
for (t4 = __rose_gt(bm - over1,0); t4 <= bm - 1; t4 += 1) {
C[t2][t4] = 0.0f;
C[t2][t4] += A[t2][0] * B[0][t4];
for (t6 = 1; t6 <= ambn - 1; t6 += 1)
C[t2][t4] += A[t2][t6] * B[t6][t4];
}
}
}
|
