void dgem(int n, double *A, double *B, double *C)
{
for(int i = 0; i < n; ++i){
for(int j = 0; j < n; j++){
double cij = C[i+j*n]; /* cij = C[i][j] */
for(int k = 0; k < n; k++){
cij += A[i+k*n] * B[k+j*n]; /*cij = A[i][k]*b[k][j]*/
C[i+j*n] = cij; /* C[i][j] = cij */
}
}
}
}
This code is from Computer_Organization_and_Design_5th
Is it right? double cij = C[i+j*n];
As far as I know, it should be C[i*n + j]
int main(void){
double A[4][4] = {1,2,3,4,
5,6,7,8,
9,10,11,12,
13,14,15,16};
double * a = &A[0][0];
int n = 4;
printf("%f %f %f %f", *(*(A+1)+3), A[1][3], a[1*n + 3], a[1 + 3*n]); /*
when i == 1 and j == 3 */
return 0;
}
OUTPUT:
8.000000 8.000000 8.000000 14.000000
When I try with gcc, it doesn't make sense...
The below statement
double cij = C[i+j*n];
is correct. To understand this Lets assume double ptr[3] = {1.5,2.5,3.5]
where ptr
is array of three double variable. Now how will you access ptr[0]
or ptr[1]
etc.
----------------------------------------
| 1.5 | 2.5 | 3.5 | 4.5 |
-----------------------------------------
0x100 0x108 0x116 0x124 0x132 <-- lets say starting address
LSB of ptr is 0x100
|
ptr
for row = 1
ptr[row] == *(ptr + row * sizeof(ptr[row]))
2.5 == *(0x100 + 1*8)
2.5 == *(0x108)
2.5 == 2.5
From above you can't have *(ptr*8 + row)
it should be *(ptr + row*8)
.
Similarly if ptr
is 2D
array like double ptr[2][3]
then
ptr[row][col] == *( *(ptr + row) + col*8)
Similarly in your case valid one is C[i + j*n]
not C[i*n +j]
Edit :- you have an 2D array
like below
double A[4][4] = { {1,2,3,4} , {5,6,7,8} , {9,10,11,12} , {13,14,15,16} };
And
double *a = &A[0][0];
Now it looks like
A[0] | A[1] | A[2] | A[3] |
-------------------------------------------------------------------------
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 |
-------------------------------------------------------------------------
0x100 0x108...................0x156.......................0x156 (assume that 0x100 is starting address of A)
A
a
LSB -->
Now when you do a[1*n + 3])
How internally its expanded
a[1*n + 3]) == *(a + (1*n + 3) /*note that a is double pointer, it increments by 8 bytes
a[1*n + 3]) == *(0x100 + (1*4 + 3) *8)
a[1*n + 3]) == *(0x100 + 56)
== *(0x156)
== 8 /* it prints 8.000000 */
And when you do a[1+3*n])
How internally its expanded
a[1+3*n]) == *(a + (1+3*n)*8)
a[1+3*n]) == *(0x100 + (1+3*4)*8)
a[1+3*n]) == *(0x100 + 96)
== *(0x196)
== 14 /* it prints 14.000000 */
When you do *(*(A+1) +3))
it internally expanded as
*(*(A+1) +3)) == *( *(0x100 +1*32) + 3*8) /* A+1 means increment by size of A[0] i.e 32 */
*(*(A+1) +3)) == *( *(0x132) + 24)
*(*(A+1) +3)) == *( 0x132 + 24 ) == *(156)
*(*(A+1) +3)) == 8 /*it prints 8.000000 */
And When you do A[1][3]
which is same as above case.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.