Question

 size_t size = sizeof(int);
 printf("%d\n", size);

 int i;
 for (i = 0; i < size; i++) {
    printf("%d ", i);
 }

The above code (using gcc) outptus

4

0 1 2 3

 size_t size = sizeof(int);
 printf("%d\n", size);

 int i;
 for (i = -1; i < size; i++) {
    printf("%d ", i);
 }

This code (i is initialized to -1) outputs only 4 and nothing in the loop.

 size_t size = sizeof(int);
 printf("%d\n", size);

 int i;
 for (i = -1; i < (int) size; i++) {
    printf("%d ", i);
 }

Adding a cast makes the code run fine again. The output is

4

-1 0 1 2 3

What’s going wrong in the second code? Why doesn’t printf go wrong anywhere?

Answer 1

i < size

When i is signed and size is unsigned, then i is converted to unsigned before the comparison is performed. This is part of what are called the usual arithmetic conversions.

When -1 is converted to an unsigned type, the result is the largest possible value representable by the unsigned type, thus i < size is false when i is -1 for any value of size.

When you use i < (int)size instead, both operands of < are of type int, so no conversions need to be performed and since both operands are signed, you get the expected result.