Can I load a floating point # from a GP register

Question

Can I load a floating point from a general purpose register like this:

lwc1  $f12, 0($t4)

It give me this error:
"$f12": operand is of incorrect type

Answer 1

Usually processors don's support something like this. Floating point values can be stored/loaded only to/from memory.
If you want to perform an operation equivalent to C's union , then store the value from your GP register to memory and then read it back in a FP register. If you want to perform a conversion from int to float, then use appropriate load instruction.