1 floating-point arithmetic exceptions, Floating-point arithmetic exceptions -18 – Freescale Semiconductor MCF5480 User Manual
Page 208

MCF548x Reference Manual, Rev. 3
6-18
Freescale Semiconductor
In addition to these vectors, attempting to execute a FRESTORE instruction with a unsupported frame
value generates a format error exception (vector 14). See the FRESTORE instruction in the ColdFire
Programmer’s Reference Manual.
Attempting to execute an FPU instruction with an undefined or unsupported value in the 6-bit effective
address, the 3-bit source/destination specifier, or the 7-bit opmode generates a line-F emulator exception,
vector 11. See
6.6.1
Floating-Point Arithmetic Exceptions
This section describes floating-point arithmetic exceptions;
lists these exceptions in order of
priority:
Most floating-point exceptions are taken when the next floating-point arithmetic instruction is encountered
(this is called a pre-instruction exception). Exceptions set during a floating-point store to memory or to an
integer register are taken immediately (post-instruction exception).
Note that FMOVE is considered an arithmetic instruction because the result is rounded. Only FMOVE
with any destination other than a floating-point register (sometimes called FMOVE OUT) can generate
post-instruction exceptions. Post-instruction exceptions never write the destination. After a
post-instruction exception, processing continues with the next instruction.
Table 6-10. Floating-Point Exception Vectors
Vector Number
Vector Offset Program Counter
Assignment
48
0x0C0
Fault
Floating-point branch/set on unordered condition
49
0x0C4
NextFP or Fault
Floating-point inexact result
50
0x0C8
NextFP
Floating-point divide-by-zero
51
0x0CC
NextFP or Fault
Floating-point underflow
52
0x0D0
NextFP or Fault
Floating-point operand error
53
0x0D4
NextFP or Fault
Floating-point overflow
54
0x0D8
NextFP or Fault
Floating-point input NAN
55
0x0DC
NextFP or Fault
Floating-point input denormalized number
Table 6-11. Exception Priorities
Priority
Exception
1
Branch/set on unordered (BSUN)
2
Input Not-a-Number (INAN)
3
Input denormalized number (IDE)
4
Operand error (OPERR)
5
Overflow (OVFL)
6
Underflow (UNFL)
7
Divide-by-zero (DZ)
8
Inexact (INEX)