The MOS 6502 processor has only three 8-bit registers and only supports 8-bit integer arithmetic, so implementing one's own multiple precision floating point ar

The MOS 6502 processor has only three 8-bit registers and only supports 8-bit integer arithmetic, so implementing one's own multiple precision floating point arithmetic is a major pain in the ass. If the C compiler doesn't provide the float and double data types, which are basic C datatypes, this means that a lot of modern software written in C cannot be compiled to 6502 machine code, and that one must write one's own arithmetic routines, at which point one is compelled to ask oneself: what's the benefit of the aforementioned C compiler when one would not be able to compile existing C software with it?