Opcodes beginning with Hex 'A' (1010_xxxx_xxxx_xxxx) and Hex 'F' were traditionally used on the 68000 architecture to implement coprocessors.
For example, the 680x0 family originally had a separate Floating Point processor. All opcodes for floating point operations had an 'F' as the top nibble, and the 680x0 handed them off to the coprocessor for execution. If you tried to execute floating point code on a 680x0 processor without an FPU, you would get an "Unimplementd F-Line Opcode" exception.
The ColdFire processors use 'A-line' opcodes for MAC instructions, and still use 'F-line' opcodes for the FPU (where present). If a particular A-line opcode doesn't decode to a valid instruction, you get an A-Line exception.