I have a project that works with 32 bit integers. Recently I added a module (another project) to the initial project, but this one deals with 32 bit floating point numbers. So the initial project now behaves as the master project which has to deal with the int to float conversions and vice versa when the added module is to be used. I tried using the built in functions (__builtin_conv_FtoR) and (__builtin_conv_RtoF) per sample to deal with this, but there seems to be a problem particularly with the __builtin_conv_FtoR function. The output is always 0.
There is no problem while using these functions when I use floating point buffers for the input/output data from and to the ADC/DACs. The problem happens when I use integer buffers. I have to use integer buffers for compatibility purposes.
Is there anyway I could find out what these two functions expect in terms of input? I'd appreciate any general comments that might be helpful in this situation.