Consider the following source code, where NR and NC are macro expressions declar
ID: 3786077 • Letter: C
Question
Consider the following source code, where NR and NC are macro expressions declared with #define that compute the dimensions of array A in terms of parameter n. This code computes the sum of the elements of column j of the array.
1 long sum_col (long n, long A[NR(n) ] [NC(n) ] , long j) {
2 long I;
3 long result =0;
4 for (i = 0; i < NR(n); i++)
5 result += A[i] [j];
6 return result;
7 }
In compiling this program, GCC generates the following assembly code:
long sum_col (long n, long A [NR(n) ] [NC(n) ], long j)
n in %rdi, A in %rsi, j in %rdx
1 sum_col:
2 leaq 1(,%rdi, 4), %r8
3 leaq (%rdi , %rdi , 2) , %rax
4 movq %rax, %rdi
5 testq %rax , %rax
6 jle .L4
7 salq $3, %r8
8 leaq (%rsi , %rdx, 8) , %rcx
9 movl $0 , %eax
10 movl $0 , %edx
11 .L3 :
12 addq (%rcx) , %rax
13 addq $1 , %rdx
14 addq %r8 , %rcx
15 compq %rdi , %rdx
16 jne .L3
17 rep ; ret
18 .L4:
19 mol $0 , &eax
20 ret
Use your reverse engineering skills to determine the definitions of NR and NC.
Explanation / Answer
NR(n) gives the number of rows in the array which also decides the range of values for array index .
NC(n) is the size of the offset value for increment which is size of type * number of columns which is size of long * n in the above case .
These are used for memory allocation by the compiler and return a pointer to the start of the memory called base address . For row major arrangement :