How to convert byte array to single point value in matlab function for compiling for TI C2000 hardware

I have a matlab function block I'm trying to compile for the C2000 hardware (F28069M chip in particular). The block brings in a 8-byte message from a canbus port read, which I'm trying to use the Typecast function to convert the 8 bytes into two "single" values. If I use something like:

byte1=DataIn(5);
byte2=DataIn(6);
byte3=DataIn(7);
byte4=DataIn(8);
X=uint8([byte1 byte2 byte3 byte4]);
DataOut = typecast(X,'single');

I get the error "For code generation, interger input or result classes for typecast must map directly to a C type in the target hardware"

if I try to specify the C2000 data type 'float' or 'float32', I get an error on compiling "Unsupported data type for conversion"

Is there a way to do this type of typecast (from 4-byte array to float) as part of a matlab function for the C2000 hardware?

Answer 1

[Mark McBroom]

byte1 = bitshift(uint32(DataIn(5)),24);

byte2 = bitshift(uint32(DataIn(6)),16);

byte3 = bitshift(uint32(DataIn(7)),8);

byte4 = uint32(DataIn(8));

DataOut = single(byte1+byte2+byte3+byte4);