Standards
| Standards / Extensions |
C or C++ |
Dependencies |
|---|
| |
Both |
OS/390® V2R6 |
Syntax
#include <_Ieee754.h>
int __fp_swapmode(int mode);
General description
The __fp_swapmode()
function sets a flag to tell C/C++ runtime library functions whether
to interpret parameters as IEEE floating-point or hexadecimal floating-point
values based on the value of mode as follows:
- Value
- Description
- _FP_MODE_RESET
- Use the FLOAT compile option to determine the format of floating-point
parameters.
- _FP_HFP_MODE
- Interpret parameters as hexadecimal floating-point values.
- _FP_BFP_MODE
- Interpret parameters as IEEE floating-point values.
Usage Notes: - Language Environment Library
code and non-Language Environment Library
code which have the Language Environment library
bit set must use __fp_swapmode() to explicitly set floating point
behavior. Failure to do so could result in incorrect floating point
values.
- Users of the Language Environment library
bit which call c-rtl functions which could potentially use floating
point values, need to use the __fp_swapmode() function as in the following
example.
- Customer application compiled for IEEE floating point math, calls
the C++ class library.
- the C++ class library calls __fp_swapmode(), as follows:
int custmode;
/*save customer fp mode and set fp mode to HFP */
custmode = __fp_swapmode (_FP_HFP_MODE);
/* perform call to c-rtl */
sprintf();
/* restore customer fp mode */
__fp_setmode(custmode);
Return values
__fp_swapmode() returns a
flag (same values as above) with the changed from floating point mode
of operation. If __fp_swapmode() is passed a value for fpmode other
than the values shown above, the changed-to floating point mode will
remain unchanged. The return value will continue to be the changed-from
floating point mode.