-msd
This option directs the compiler to treat double as float and not to support 64-bit floating point arithmetic.
To treat double as float for a project:
It is not possible to set the Treat 'double' as 'float' property on a per-file basis