| |||

Home > Floating-point support > fplib conversions between floats, doubles, and ints |

Table 16 describes routines to perform conversions between number formats, excluding long long types.

**Table 16. Number format conversion routines**

Function | Argument types | Return type |
---|---|---|

`__aeabi_f2d` | float | double |

`__aeabi_d2f` | double | float |

`_fflt` | int | float |

`_ffltu` | unsigned int | float |

`_dflt` | int | double |

`_dfltu` | unsigned int | double |

`_ffix` | float | int (see Notes on rounding) |

`_ffix_r` | float | int |

`_ffixu` | float | unsigned int (see Notes on rounding) |

`_ffixu_r` | float | unsigned int |

`_dfix` | double | int (see Notes on rounding) |

`_dfix_r` | double | int |

`_dfixu` | double | unsigned int (see Notes on rounding) |

`_dfixu_r` | double | unsigned int |

Rounded toward zero, independently of the current rounding
mode. This is because the C standard requires implicit conversions
to integers to round this way, so it is convenient not to have to
change the rounding mode to do so. Each function has a corresponding
function with `_r`

on the end of its name, that
performs the same operation but rounds according to the current mode.