![](http://datasheet.mmic.net.cn/370000/P9S12XEP100J1VVLR_datasheet_16728329/P9S12XEP100J1VVLR_218.png)
Chapter 3 Memory Mapping Control (S12XMMCV4)
MC9S12XE-Family Reference Manual , Rev. 1.07
218
Freescale Semiconductor
TheRTCinstructionterminatessubroutinesinvokedbyaCALLinstruction.TheRTCinstructionunstacks
the PPAGE value and the return address and refills the queue. Execution resumes with the next instruction
after the CALL instruction.
During the execution of an RTC instruction the CPU performs the following steps:
1. Pulls the previously stored PPAGE value from the stack
2. Pulls the 16-bit return address from the stack and loads it into the PC
3. Writes the PPAGE value into the PPAGE register
4. Refills the queue and resumes execution at the return address
This sequence is uninterruptable. The RTC can be executed from anywhere in the local CPU memory
space.
The CALL and RTC instructions behave like JSR and RTS instruction, they however require more
execution cycles. Usage of JSR/RTS instructions is therefore recommended when possible and
CALL/RTC instructions should only be used when needed. The JSR and RTS instructions can be used to
access subroutines that are already present in the local CPU memory map (i.e. in the same page in the
program memory page window for example). However calling a function located in a different page
requires usage of the CALL instruction. The function must be terminated by the RTC instruction. Because
the RTC instruction restores contents of the PPAGE register from the stack, functions terminated with the
RTC instruction must be called using the CALL instruction even when the correct page is already present
in the memory map. This is to make sure that the correct PPAGE value will be present on stack at the time
of the RTC instruction execution.
3.5.2
Port Replacement Registers (PRRs)
Registers used for emulation purposes must be rebuilt by the in-circuit emulator hardware to achieve full
emulation of single chip mode operation. These registers are called port replacement registers (PRRs) (see
Table 1-25
). PRRs are accessible from CPU, BDM and XGATE using different access types (word
aligned, word-misaligned and byte).
Each access to PRRs will be extended to 2 bus cycles for write or read accesses independent of the
operating mode. In emulation modes all write operations result in simultaneous writing to the internal
registers (peripheral access) and to the emulated registers (external access) located in the PRU in the
emulator. All read operations are performed from external registers (external access) in emulation modes.
In all other modes the read operations are performed from the internal registers (peripheral access).
Due to internal visibility of CPU accesses the CPU will be halted during XGATE or BDM access to any
PRR. This rule applies also in normal modes to ensure that operation of the device is the same as in
emulation modes.
A summary of PRR accesses:
An aligned word access to a PRR will take 2 bus cycles.
A misaligned word access to a PRRs will take 4 cycles. If one of the two bytes accessed by the
misaligned word access is not a PRR, the access will take only 3 cycles.
A byte access to a PRR will take 2 cycles.