INFO: Handling Floating-Point Errors in C (43391)
The information in this article applies to:
- The C Run-Time (CRT), when used with:
- Microsoft Visual C++ for Windows, 16-bit edition 1.0
- Microsoft Visual C++ for Windows, 16-bit edition 1.5
- Microsoft Visual C++, 32-bit Editions 1.0
- Microsoft Visual C++, 32-bit Editions 2.0
- Microsoft Visual C++, 32-bit Editions 2.1
- Microsoft Visual C++, 32-bit Editions 2.2
- Microsoft Visual C++, 32-bit Editions 4.0
- Microsoft Visual C++, 32-bit Editions 4.1
- Microsoft Visual C++, 32-bit Editions 5.0
- Microsoft Visual C++, 32-bit Editions 6.0
This article was previously published under Q43391 SUMMARY
With Microsoft C, floating-point error-handling routines may be written
using the signal() function; however, "return" should not be used to go
back to the calling process. If "return" is used, the floating-point state
is left undefined.
Instead, use the setjmp() function to save the stack environment before
each section of the calling process code that may potentially generate a
floating-point error. In the error handler, use _fpreset() to reinitialize
the floating-point package and longjmp() to return to the calling process.
WARNING: Neither the setjmp nor longjmp function is compatible with the C++
language.
The sample program on page 296 of the "Microsoft C/C++ Run-time Library
Reference" that shipped with Microsoft C/C++ version 7.0 demonstrates the
use of setjmp(), _fpreset(), and longjmp() to handle a floating-point
error. The online help also provides samples for these functions.
The function setjmp() saves the environment (bp, si, di, sp, and ds
registers) and return address in a buffer. The buffer is used by longjmp()
to restore the environment.
| Modification Type: | Major | Last Reviewed: | 12/8/2003 |
|---|
| Keywords: | kbCRT kbinfo KB43391 kbAudDeveloper |
|---|
|