The best way to disable floating-point support

[R. Diez]

Hi there:

I'm trying to port an embedded PowerPC application from a proprietary compiler with its own proprietary C run-time library to GCC 4.5.3 + newlib 1.19.0 .

What's the best way to completely disable floating-point support in newlib? With the proprietary compiler, if I inadvertently tried to use some floating-point routine like atof(), I got linker errors. With GCC/newlib, it just seems to use the software emulation without warning, making the resulting binary bigger.

Please copy me on the reply, as I'm not subscribed to this list.

Many thanks,
  R. Diez