...
For values of n where 0 < n <= INT_MAX, the loop executes n times, as expected.
For values of n > where INT_MAX < n <= (size_t)INT_MIN, the loop executes INT_MAX + 1 times. For each iteration, i is Once i becomes negative the loop stops, and i remains in the range 0 through INT_MAX. Once i becomes negative, the loop will stop, because a negative value for i converts into SIZE_MAX + 1 + i. Because sizeof(size_t) > sizeof(int), SIZE_MAX is well above INT_MAX * 2, as well as SIZE_MAX + 1 + INT_MIN, therefore the size_t values i converts to are always greater than values of n in the range INT_MAX + 1 to INT_MAX * 2
| Wiki Markup |
|---|
For values of {{n}} where {{(size_t)INT_MIN < n <= SIZE_MAX}}, {{i}} wraps and takes the values {{INT_MIN}} to {{INT_MIN + (n - (size_t)INT_MIN - 1)}}. Execution of the loop overwrites memory from {{p\[INT_MIN\]}} through {{p\[INT_MIN + (n - (size_t)INT_MIN - 1)\]}}. |
Compliant Solution (TR 24731-1)
...