PIC18F2420/2520/4420/4520
A summary of the instructions in the extended instruc-
tion set is provided in Table 24-3. Detailed descriptions
are provided in Section 24.2.2 “Extended Instruction
Set”. The opcode field descriptions in Table 24-1
(page 268) apply to both the standard and extended
PIC18 instruction sets.
24.2 Extended Instruction Set
In addition to the standard 75 instructions of the PIC18
instruction set, PIC18F2420/2520/4420/4520 devices
also provide an optional extension to the core CPU
functionality. The added features include eight addi-
tional instructions that augment indirect and indexed
addressing operations and the implementation of
Indexed Literal Offset Addressing mode for many of the
standard PIC18 instructions.
Note:
The instruction set extension and the
Indexed Literal Offset Addressing mode
were designed for optimizing applications
written in C; the user may likely never use
these instructions directly in assembler.
The syntax for these commands is pro-
vided as a reference for users who may be
reviewing code that has been generated
by a compiler.
The additional features of the extended instruction set
are disabled by default. To enable them, users must set
the XINST Configuration bit.
The instructions in the extended set can all be
classified as literal operations, which either manipulate
the File Select Registers, or use them for indexed
addressing. Two of the instructions, ADDFSR and
SUBFSR, each have an additional special instantiation
for using FSR2. These versions (ADDULNK and
SUBULNK) allow for automatic return after execution.
24.2.1
EXTENDED INSTRUCTION SYNTAX
Most of the extended instructions use indexed argu-
ments, using one of the File Select Registers and some
offset to specify a source or destination register. When
an argument for an instruction serves as part of
indexed addressing, it is enclosed in square brackets
(“[ ]”). This is done to indicate that the argument is used
as an index or offset. MPASM™ Assembler will flag an
error if it determines that an index or offset value is not
bracketed.
The extended instructions are specifically implemented
to optimize re-entrant program code (that is, code that
is recursive or that uses a software stack) written in
high-level languages, particularly C. Among other
things, they allow users working in high-level
languages to perform certain operations on data
structures more efficiently. These include:
When the extended instruction set is enabled, brackets
are also used to indicate index arguments in byte-
oriented and bit-oriented instructions. This is in addition
to other changes in their syntax. For more details, see
Section 24.2.3.1 “Extended Instruction Syntax with
Standard PIC18 Commands”.
• Dynamic allocation and deallocation of software
stack space when entering and leaving
subroutines
• Function Pointer invocation
• Software Stack Pointer manipulation
• Manipulation of variables located in a software
stack
Note:
In the past, square brackets have been
used to denote optional arguments in the
PIC18 and earlier instruction sets. In this
text and going forward, optional
arguments are denoted by braces (“{ }”).
TABLE 24-3: EXTENSIONS TO THE PIC18 INSTRUCTION SET
16-Bit Instruction Word
MSb LSb
Mnemonic,
Operands
Status
Affected
Description
Cycles
ADDFSR
ADDULNK
CALLW
f, k
k
Add Literal to FSR
Add Literal to FSR2 and Return
Call Subroutine using WREG
1
2
2
2
1110 1000 ffkk kkkk
1110 1000 11kk kkkk
0000 0000 0001 0100
1110 1011 0zzz zzzz
1111 ffff ffff ffff
1110 1011 1zzz zzzz
1111 xxxx xzzz zzzz
1110 1010 kkkk kkkk
None
None
None
None
MOVSF
zs, fd Move zs (source) to 1st word
fd (destination) 2nd word
zs, zd Move zs (source) to 1st word
MOVSS
PUSHL
2
1
None
None
zd (destination)
Store Literal at FSR2,
Decrement FSR2
2nd word
k
SUBFSR
SUBULNK
f, k
k
Subtract Literal from FSR
Subtract Literal from FSR2 and
Return
1
2
1110 1001 ffkk kkkk
1110 1001 11kk kkkk
None
None
© 2008 Microchip Technology Inc.
DS39631E-page 309
MICROCHIP [ MICROCHIP ]
