QUERY · ISSUE

NRF52: build error, implicit conversion from 'float' to 'double'

openby permaBoxopened 2025-04-22updated 2025-06-09

bugport-nrf

Port, board and/or hardware

nrf port, nrf52 board

MicroPython version

mpy latest, nrf52 board

The build process has failed with a compiler configuration issue if support for double is configured.

Build error report:

CC ../../py/formatfloat.c
../../py/formatfloat.c: In function 'mp_format_float':
../../py/formatfloat.c:92:26: error: implicit conversion from 'float' to 'double' to match other operand of binary expression [-Werror=double-promotion]
92 | #define fp_isless1(x) (x < 1.0)
| ^
../../py/formatfloat.c:183:16: note: in expansion of macro 'fp_isless1'
183 | } else if (fp_isless1(f)) {
| ^~~~~~~~~~
../../py/formatfloat.c:336:15: error: implicit conversion from 'float' to 'double' to match other operand of binary expression [-Werror=double-promotion]
336 | f *= FPCONST(10.0);
| ^~
cc1: all warnings being treated as errors

Reproduction

mpconfigport.h

#if NRF51
#define MICROPY_FLOAT_IMPL (MICROPY_FLOAT_IMPL_NONE)
#else
#define MICROPY_FLOAT_IMPL (MICROPY_FLOAT_IMPL_DOUBLE)
#endif

Expected behaviour

No response

Observed behaviour

Related build error report.

Additional Information

No, I've provided everything above.

Code of Conduct

Yes, I agree

CANDIDATE · ISSUE

1.13+ doesn't compile on FreeBSD 13.0

closedby yggdropened 2021-07-11updated 2024-03-15

Starting with micropython 1.13 I cannot compile it on FreeBSD 13 anymore. 1.12 worked fine. Installation methods were installation via pyenv's python-build, as well as the ports/unix method described in the README.

[...]
CC ../../py/formatfloat.c
In file included from ../../py/formatfloat.c:33:
../../py/formatfloat.c: In function 'mp_format_float':
../../py/formatfloat.c:131:20: error: conversion from 'double' to 'float' may change value [-Werror=float-conversion]
  131 |     if (fp_signbit(f) && !fp_isnan(f)) {
      |                    ^
../../py/formatfloat.c:131:9: note: in expansion of macro 'fp_signbit'
  131 |     if (fp_signbit(f) && !fp_isnan(f)) {
      |         ^~~~~~~~~~
../../py/formatfloat.c:131:36: error: conversion from 'double' to 'float' may change value [-Werror=float-conversion]
  131 |     if (fp_signbit(f) && !fp_isnan(f)) {
      |                                    ^
../../py/formatfloat.c:131:27: note: in expansion of macro 'fp_isnan'
  131 |     if (fp_signbit(f) && !fp_isnan(f)) {
      |                           ^~~~~~~~
../../py/formatfloat.c:146:22: error: conversion from 'double' to 'float' may change value [-Werror=float-conversion]
  146 |         if (fp_isinf(f)) {
      |                      ^
../../py/formatfloat.c:146:13: note: in expansion of macro 'fp_isinf'
  146 |         if (fp_isinf(f)) {
      |             ^~~~~~~~
../../py/formatfloat.c:151:29: error: conversion from 'double' to 'float' may change value [-Werror=float-conversion]
  151 |         } else if (fp_isnan(f)) {
      |                             ^
../../py/formatfloat.c:151:20: note: in expansion of macro 'fp_isnan'
  151 |         } else if (fp_isnan(f)) {
      |                    ^~~~~~~~
cc1: all warnings being treated as errors
gmake: *** [../../py/mkrules.mk:77: build-standard/py/formatfloat.o] Error 1