Multiplication with double produces wrong result

[Christoph Rabel]

Hi!

The following program compiled with gcc 3.3.3

  #include <iostream>
  using namespace std;

  int main() {
    double x = 0.55;
    int v1 = (int) (x * UINT_MAX);
    int v2 = (int) (0.55 * UINT_MAX);

    cout << dec << v1 << "\t" << hex << v1 << endl;
    cout << dec << v2 << "\t" << hex << v2 << endl;
  }

produces the output:

  -2147483648     80000000
  -1932735284     8ccccccc

This is obviously a bug.
Changing "double x = 0.55;" to "const double x = 0.55;" produces 
expected results.

Christoph