Article ID: 168012
Article Last Modified on 10/17/2003
APPLIES TO
- Microsoft Visual C++ 5.0 Enterprise Edition
- Microsoft Visual C++ 5.0 Professional Edition
This article was previously published under Q168012
SYMPTOMS
Incorrect results from floating point operations may be obtained when building with /Og and /Oi or a switch that includes /Og and /Oi (such as /Ox, /O1, or /O2.)
STATUS
Microsoft has confirmed this to be a bug in the Microsoft products listed at the beginning of this article. This bug has been corrected in Visual Studio 97, Service Pack 1.
For additional information about the Visual Studio 97 Service Pack 1, please see the following article in the Microsoft Knowledge Base:
170365 INFO: Visual Studio 97 Service Packs - What, Where, and Why
MORE INFORMATION
The following sample demonstrates one form of this bug. As you step through this code in a debugger, you will see that inconsistent results are obtained.
Sample Code
// compile options /Ogi
#include <math.h>
float e1(float x)
{
float value;
if (x < 0.0)
value = exp(x);
else
value = 1.0;
return value;
}
#pragma ("g", off )
void f()
{
float f1 = e1( -1.2f); // 1.00000
float f2 = exp(-1.2f); // 0.301194
}
Keywords: kbbug kbfix kbvs97sp1fix kbvs97sp2fix KB168012
