-qfloat
In z/OS® XL C/C++, the -qfloat option selects different strategies for speeding up or improving the accuracy of floating-point calculations.
In Open XL C/C++ 1.1, some
of the -qfloat suboptions have functionally equivalent options as shown in the
following table.
https://ibmid.acrolinx.cloud
| Options supported by z/OS XL C/C++ | Similar/equivalent Clang options supported by Open XL C/C++ 1.1 |
|---|---|
| -qfloat=afp=volatile, -qfloat=afp=novolatile, -qfloat=noafp | None |
| -qfloat=[no]fold | None |
| -qfloat=hex | None |
| -qfloat=ieee | None; Open XL C/C++ 1.1 generates binary floating-point numbers and instructions by default. |
| -qfloat=maf | -ffp-contract=on; -ffp-contract=fast can be used for additional optimizations. |
| -qfloat=nomaf | -ffp-contract=off |
| -qfloat=rrm | -frounding-math |
| -qfloat=norrm | -fno-rounding-math |