![](http://datasheet.mmic.net.cn/120000/MC68349FT16_datasheet_3559370/MC68349FT16_189.png)
5- 34
MC68349 USER’S MANUAL
MOTOROLA
After initial instruction prefetches, normal program execution begins at the address in the
PC. The reset exception does not save the value of either the PC or the SR.
If a bus error or address error occurs during reset exception processing, a double bus fault
occurs, the processor halts, and the HALT signal is asserted to indicate the halted
condition.
Execution of the RESET instruction does not cause a reset exception nor does it affect
any internal CPU register. The SIM49 registers and the module control register in each
internal peripheral module (DMA, quad data memory module (QDMM), and serial
modules) are not affected. All other internal peripheral module registers are reset the
same as for a hardware reset. The external devices connected to the RESET signal are
reset at the completion of the RESET instruction.
5.6.2.2 BUS ERROR. A bus error exception occurs when an assertion of the BERR signal
is acknowledged. The BERR signal can be asserted by one of three sources:
1. External logic by assertion of the BERR input pin
2. Direct assertion of the internal BERR signal by an internal module
3. Direct assertion of the internal BERR signal by the on-chip hardware watchdog
after detecting a no-response condition
Bus error exception processing begins when the processor attempts to use information
from an aborted bus cycle.
When the aborted bus cycle is an instruction prefetch, the processor will not initiate
exception processing unless the prefetched information is used. For example, if a branch
instruction flushes an aborted prefetch, that word is not accessed, and no exception
occurs.
When the aborted bus cycle is a data access, the processor initiates exception processing
immediately, except in the case of released operand writes. Released write bus errors are
delayed until the next instruction boundary or until another operand access is attempted.
Exception processing for bus error exceptions follows the regular sequence, but context
preservation is more involved than for other exceptions because a bus exception can be
initiated while an instruction is executing. Several bus error stack format organizations are
utilized to provide additional information regarding the nature of the fault.
First, any register altered by a faulted-instruction EA calculation is restored to its initial
value. Then an SSW is placed on the stack. The SSW contains specific information about
the aborted access—size, type of access (read or write), bus cycle type, and function
code. Finally, fault address, bus error exception vector number, PC value, and a copy of
the SR are saved.