I have been writing small c programs and then disassembling them to try to understand what is actually happening under the hood.

I am using mingw-w64 on window 7 for my work, and I have run into a question about why certain amounts of space are being allocated on the stack.

program1: int main() { int i,j; return 0; }

The start of the disassembled code for the main function is:

00000000004015b0 <main>:
  4015b0:   55                      push   %rbp
  4015b1:   48 89 e5                mov    %rsp,%rbp
  4015b4:   48 83 ec 20             sub    $0x20,%rsp

The last line looks like it is allocating 32 bytes on teh stack.

Program 2: The same program except it assigns values to teh two variables and returns the sum of the variables. int main() { int i,j; i = 5; j = 6; return i+j; }

Start of disassembled code: 00000000004015b0 : 4015b0: 55 push %rbp 4015b1: 48 89 e5 mov %rsp,%rbp 4015b4: 48 83 ec 30 sub $0x30,%rsp

Why does this program allocate 48 bytes instead of the 32 bytes from the first program?

In both cases i and j are stored at %rbp-4 and %rbp-8 respectively, and the math done in the second program is all done in the registers.

Note I am using gcc 6.3.0 to compile the code and objdump 2.28 to disassemble it.

All code can be seen in my github repository http://github.com/draikes/rev-eng

I have been writing small C programs and then disassembling them to try to understand what is actually happening under the hood.

I am using mingw-w64 on Windows 7 for my work, and I have run into a question about why certain amounts of space are being allocated on the stack.

program1:

int main()
{
  int i,j;
  return 0;
}

The start of the disassembled code for the main function is:

00000000004015b0 <main>:
  4015b0:   55                      push   %rbp
  4015b1:   48 89 e5                mov    %rsp,%rbp
  4015b4:   48 83 ec 20             sub    $0x20,%rsp

The last line looks like it is allocating 32 bytes on teh stack.

Program 2: The same program except it assigns values to the two variables and returns the sum of the variables.

int main()
{
  int i,j;
  i = 5;
  j = 6;
  return i+j;
}

Start of disassembled code:

00000000004015b0 <main>:
  4015b0:   55                      push   %rbp
  4015b1:   48 89 e5                mov    %rsp,%rbp
  4015b4:   48 83 ec 30             sub    $0x30,%rsp

Why does this program allocate 48 bytes instead of the 32 bytes from the first program?

In both cases i and j are stored at %rbp-4 and %rbp-8 respectively, and the math done in the second program is all done in the registers.

Note I am using gcc 6.3.0 to compile the code and objdump 2.28 to disassemble it.

All code can be seen in my github repository http://github.com/draikes/rev-eng