[Issue 2697] New: Cast of float function return to ulong or uint gives bogus value

[d-bugmail]

http://d.puremagic.com/issues/show_bug.cgi?id=2697

           Summary: Cast of float function return to ulong or uint gives
                    bogus value
           Product: D
           Version: 1.040
          Platform: PC
        OS/Version: Windows
            Status: NEW
          Keywords: wrong-code
          Severity: major
          Priority: P2
         Component: DMD
        AssignedTo: bugzi...@digitalmars.com
        ReportedBy: s...@iname.com


import std.stdio;

float func() { return 5; }

void main() {
    writefln("float:  %f", func());
    writefln("ulong:  %d", cast(ulong)  func());
    writefln("long:   %d", cast(long)   func());
    writefln("uint:   %d", cast(uint)   func());
    writefln("int:    %d", cast(int)    func());
    writefln("ushort: %d", cast(ushort) func());
    writefln("short:  %d", cast(short)  func());
    writefln("ubyte:  %d", cast(ubyte)  func());
    writefln("byte:   %d", cast(byte)   func());

    float result = func();
    writefln("float:  %f", result);
    writefln("ulong:  %d", cast(ulong)  result);
    writefln("long:   %d", cast(long)   result);
    writefln("uint:   %d", cast(uint)   result);
    writefln("int:    %d", cast(int)    result);
    writefln("ushort: %d", cast(ushort) result);
    writefln("short:  %d", cast(short)  result);
    writefln("ubyte:  %d", cast(ubyte)  result);
    writefln("byte:   %d", cast(byte)   result);
}
----------
float:  5.000000
ulong:  2048
long:   5
uint:   0
int:    5
ushort: 5
short:  5
ubyte:  5
byte:   5
float:  5.000000
ulong:  5
long:   5
uint:   5
int:    5
ushort: 5
short:  5
ubyte:  5
byte:   5
----------

If the ulong and long cases are removed, the cast to uint gives 2048.
Seems to be only float that has the problem - it disappears if I change it to
double or real.


--