DSPIC33EP64MC204 - Electronic component Microchip - Free user manual and instructions

USER MANUAL DSPIC33EP64MC204 Microchip

16-Bit MCU and DSC Programmer's Reference Manual

High-Performance Microcontrollers (MCU) and Digital Signal Controllers (DSC)

Note the following details of the code protection feature on Microchip devices:

• Microchip products meet the specification contained in their particular Microchip Data Sheet.
- Microchip believes that its family of products is one of the most secure families of its kind on the market today, when used in the intended manner and under normal conditions.
- There are dishonest and possibly illegal methods used to breach the code protection feature. All of these methods, to our knowledge, require using the Microchip products in a manner outside the operating specifications contained in Microchip's Data Sheets. Most likely, the person doing so is engaged in theft of intellectual property.
• Microchip is willing to work with the customer who is concerned about the integrity of their code.
- Neither Microchip nor any other semiconductor manufacturer can guarantee the security of their code. Code protection does not mean that we are guaranteeing the product as "unbreakable."

Code protection is constantly evolving. We at Microchip are committed to continuously improving the code protection features of our products. Attempts to break Microchip's code protection feature may be a violation of the Digital Millennium Copyright Act. If such acts allow unauthorized access to your software or other copyrighted work, you may have a right to sue for relief under that Act.

Information contained in this publication regarding device applications and the like is provided only for your convenience and may be superseded by updates. It is your responsibility to ensure that your application meets with your specifications. MICROCHIP MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND WHETHER EXPRESS OR IMPLIED, WRITTEN OR ORAL, STATUTORY OR OTHERWISE, RELATED TO THE INFORMATION, INCLUDING BUT NOT LIMITED TO ITS CONDITION, QUALITY, PERFORMANCE, MERCHANTABILITY OR FITNESS FOR PURPOSE. Microchip disclaims all liability arising from this information and its use. Use of Microchip devices in life support and/or safety applications is entirely at the buyer's risk, and the buyer agrees to defend, indemnify and hold harmless Microchip from any and all damages, claims, suits, or expenses resulting from such use. No licenses are conveyed, implicitly or otherwise, under any Microchip intellectual property rights unless otherwise stated.

Microchip received ISO/TS-16949:2009 certification for its worldwide headquarters, design and wafer fabrication facilities in Chandler and Tempe, Arizona; Gresham, Oregon and design centers in California and India. The Company's quality system processes and procedures are for its PIC® MCUs and dsPIC® DSCs, KEELOQ® code hopping devices, Serial EEPROMs, microperipherals, nonvolatile memory and analog products. In addition, Microchip's quality system for the design and manufacture of development systems is ISO 9001:2000 certified.

QUALITY MANAGEMENT SYSTEM CERTIFIED BY DNV = ISO/TS 16949=

Trademarks

The Microchip name and logo, the Microchip logo, AnyRate, AVR, AVR logo, AVR Freaks, BeaconThings, BitCloud, CryptoMemory, CryptoRF, dsPIC, FlashFlex, flexPWR, Heldo, JukeBlox, KEELOQ, KEELOQ logo, Kleer, LANCheck, LINK MD, maXStylus, maXTouch, MediaLB, megaAVR, MOST, MOST logo, MPLAB, OptoLyzer, PIC, picoPower, PICSTART, PIC32 logo, Prochip Designer, QTouch, RightTouch, SAM-BA, SpyNIC, SST, SST Logo, SuperFlash, tinyAVR, UNI/O, and XMEGA are registered trademarks of Microchip Technology Incorporated in the U.S.A. and other countries.

ClockWorks, The Embedded Control Solutions Company, EtherSynch, Hyper Speed Control, HyperLight Load, IntelliMOS, mTouch, Precision Edge, and Quiet-Wire are registered trademarks of Microchip Technology Incorporated in the U.S.A. Adjacent Key Suppression, AKS, Analog-for-the-Digital Age, Any Capacitor, AnyIn, AnyOut, BodyCom, chipKIT, chipKIT logo, CodeGuard, CryptoAuthentication, CryptoCompanion, CryptoController, dsPICDEM, dsPICDEM.net, Dynamic Average Matching, DAM, ECAN, EtherGREEN, In-Circuit Serial Programming, ICSP, Inter-Chip Connectivity, JitterBlocker, KleerNet, KleerNet logo, Mindi, MiWi, motorBench, MPASM, MPF, MPLAB Certified logo, MPLIB, MPLINK, MultiTRAK, NetDetach, Omniscient Code Generation, PICDEM, PICDEM.net, PICkit, PICtail, PureSilicon, QMatrix, RightTouch logo, REAL ICE, Ripple Blocker, SAM-ICE, Serial Quad I/O, SMART-I.S., SQI, SuperSwitcher, SuperSwitcher II, Total Endurance, TSHARC, USBCheck, VariSense, ViewSpan, WiperLock, Wireless DNA, and ZENA are trademarks of Microchip Technology Incorporated in the U.S.A. and other countries.

SQTP is a service mark of Microchip Technology Incorporated in the U.S.A.

Silicon Storage Technology is a registered trademark of Microchip Technology Inc. in other countries.

GestIC is a registered trademark of Microchip Technology Germany II GmbH & Co. KG, a subsidiary of Microchip Technology Inc., in other countries.

All other trademarks mentioned herein are property of their respective companies.

© 2005-2018, Microchip Technology Incorporated, All Rights Reserved.

ISBN: 978-1-5224-2896-1

Table of Contents

PAGE

SECTION 1. INTRODUCTION 5

Introduction 6

Manual Objective 6

Development Support 6

Style and Symbol Conventions 7

Instruction Set Symbols 8

SECTION 2. PROGRAMMER'S MODEL 9

16-Bit MCU and DSC Core Architecture Overview 10

Programmer's Model 14

Working Register Array 19

Default Working Register (WREG) 20

Software Stack Frame Pointer 20

Software Stack Pointer 20

Stack Pointer Limit Register (SPLIM) 20

Accumulator A and Accumulator B (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 20

Program Counter 21

TBLPAG Register 21

PSVPAG Register (PIC24F, PIC24H, dsPIC30F and dsPIC33F) 21

RCOUNT Register 21

DCOUNT Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 2

DOSTART Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 22

DOEND Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 22

STATUS Register 22

Core Control Register 25

Shadow Registers 25

DO Stack (dsPIC33E and dsPIC33C Devices) 26

SECTION 3. INSTRUCTION SET OVERVIEW 39

Introduction 40

Instruction Set Overview 40

Instruction Set Summary Tables 42

SECTION 4. INSTRUCTION SET DETAILS 53

Data Addressing Modes 54

Program Addressing Modes 63

Instruction Stalls 64

Byte Operations 66

Word Move Operations 68

Using 10-Bit Literal Operands ....71

Bit Field Insert/Extract Instructions (dsPIC33C Devices Only) 71

Software Stack Pointer and Frame Pointer 72

Conditional Branch Instructions 78

Z Status Bit 79

Assigned Working Register Usage 80

DSP Data Formats (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 83

Accumulator Usage (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 85

Accumulator Access (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 86

DSP MAC Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 86

16-Bit MCU and DSC Programmer's Reference Manual

DSP Accumulator Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 90

Scaling Data with the FBCL Instruction (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 90

Data Range Limit Instructions (dsPIC33C Devices Only) 92

Normalizing the Accumulator with the FBCL Instruction (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) 93

Normalizing the Accumulator with the NORM Instruction (dsPIC33C Devices Only) 93

Extended Precision Arithmetic Using Mixed-Sign Multiplications (dsPIC33E and dsPIC33C Only) 94

SECTION 5. INSTRUCTION DESCRIPTIONS 95

Instruction Symbols 96

Instruction Encoding Field Descriptors Introduction 96

Instruction Description Example 101

Instruction Descriptions 102

SECTION 6. BUILT-IN FUNCTIONS 459

Introduction 460

Built-in Function List 461

SECTION 7. REFERENCE 497

Instruction Bit Map 498

Instruction Set Summary Table 501

Revision History 511

SECTION 8. INDEX 513

SECTION 9. WORLDWIDE SALES AND SERVICE 520

Section 1. Introduction

HIGHLIGHTS

This section of the manual contains the following major topics:

1.1 Introduction....6
1.2 Manual Objective 6
1.3 Development Support 6
1.4 Style and Symbol Conventions 7
1.5 Instruction Set Symbols 8

1.1 INTRODUCTION

Microchip Technology focuses on products for the embedded control market. Microchip is a leading supplier of the following devices and products:

• 8-Bit General Purpose Microcontrollers (PIC® MCUs)
  • 16-Bit Digital Signal Controllers (dsPIC® DSCs)
  • 16-Bit and 32-Bit Microcontrollers (MCUs)
  • Specialty and Standard Nonvolatile Memory Devices
  • Security Devices (K EELOQ® Security ICs)
  • Application-Specific Standard Products

Information about these devices and products, with corresponding technical documentation, is available on the Microchip web site (www.microchip.com).

1.2 MANUAL OBJECTIVE

This manual is a software developer's reference for the 16-bit MCU and DSC device families. It describes the Instruction Set in detail and also provides general information to assist the development of software for the 16-bit MCU and DSC device families.

This manual does not include detailed information about the core, peripherals, system integration or device-specific information. The user should refer to the specific device family reference manual for information about the core, peripherals and system integration. For device-specific information, the user should refer to the specific device data sheets. The information that can be found in the data sheets includes:

• Device memory map
  • Device pinout and packaging details
  • Device electrical specifications
• List of peripherals included on the device

Code examples are given throughout this manual. These examples are valid for any device in the 16-bit MCU and DSC families.

1.3 DEVELOPMENT SUPPORT

Microchip offers a wide range of development tools that allow users to efficiently develop and debug application code. Microchip's development tools can be broken down into four categories:

• Code Generation
• Hardware/Software Debug
• Device Programmer
  • Product Evaluation Boards

Information about the latest tools, product briefs and user guides can be obtained from the Microchip web site (www.microchip.com) or from your local Microchip Sales Office.

Microchip offers other reference tools to speed up the development cycle. These include:

• Application Notes
  • Reference Designs
• Microchip Web Site
• Local Sales Offices with Field Application Support
  • Corporate Support Line

The Microchip web site also lists other sites that may be useful references.

1.4 STYLE AND SYMBOL CONVENTIONS

Throughout this document, certain style and font format conventions are used. Table 1-1 provides a description of the conventions used in this document.

Table 1-1: Document Conventions

Symbol or Term	Description
set To force a bit/register to a value of logic ‘1’.
clear To force a bit/register to a value of logic ‘0’.
Reset 1. To force a register/bit to its default state.2. A condition in which the device places itself after a device Reset occurs. Some bits will be forced to ‘0’ (such as interrupt enable bits), while others will be forced to ‘1’ (such as the I/O data direction bits).
0xnnnn Designates the number ‘nnnn’ in the hexadecimal number system. These conventions are used in the code examples. For example, 0x013F or 0xA800.
: (colon) Used to specify a range or the concatenation of registers/bits/pins.One example is ACCAU:ACCAH:ACCAL, which is the concatenation of three registers to form the 40-bit Accumulator.Concatenation order (left-right) usually specifies a positional relationship (MSb to LSb, higher to lower).
< > Specifies bit locations in a particular register.One example is SR<7:5> (or IPL<2:0>), which specifies the register and associated bits or bit locations.
LSb, MSb Indicates the Least Significant or Most Significant bit in a field.
LSB, MSB Indicates the Least/Most Significant Byte in a field of bits.
Isw, msw Indicates the least/most significant word in a field of bits
Courier New Font Used for code examples, binary numbers and for Instruction mnemonics in the text.
Times New Roman Font, Italic Used for equations and variables.
Times New Roman Font, Bold Italic Used in explanatory text for items called out from a figure, equation or example.
Note:	A Note presents information that we want to re-emphasize, either to help you avoid a common pitfall or make you aware of operating differences between some device family members. A Note can be in a box, or when used in a table or figure, it is located at the bottom of the table or figure.

1.5 INSTRUCTION SET SYMBOLS

The summary tables in Section 3.2 “Instruction Set Overview” and Section 7.2 “Instruction Set Summary Table”, and the instruction descriptions in Section 5.4 “Instruction Descriptions” utilize the symbols shown in Table 1-2.

Table 1-2: Symbols Used in Instruction Summary Tables and Descriptions

Symbol(1)	Description
{ } Optional field or operation
[text] The location addressed by text
(text) The contents of text
#text The literal defined by text
a ∈ [b, c, d] “a” must be in the set of [b, c, d]
	Register bit field
{label:} Optional label name
Acc Accumulator	A or Accumulator B
AWB Accumulator	Write-Back
bit4 4-bit wide bit position (0:7 in Byte mode, 0:15 in Word mode)
Expr Absolute address, label or expression (resolved by the linker)
f File register address
lit1 1-bit literal (0:1)
lit4 4-bit literal (0:15)
lit5 5-bit literal (0:31)
lit8 8-bit literal (0:255)
lit10 10-bit literal (0:255 in Byte mode, 0:1023 in Word mode)
lit14 14-bit literal (0:16383)
lit16 16-bit literal (0:65535)
lit23 23-bit literal (0:8388607)
Slit4 Signed 4-bit literal (-8:7)
Slit6 Signed 6-bit literal (-32:31) (range is limited to -16:16)
Slit10 Signed 10-bit literal (-512:511)
Slit16 Signed 16-bit literal (-32768:32767)
TOS Top-of-Stack
Wb	Base Working register
Wd	Destination Working register (Direct and Indirect Addressing)
Wdo	Destination Working register (Direct and Indirect Addressing, including Indirect Addressing with Offset)
Wm, Wn Working register divide pair (dividend, divisor)
Wm * Wm	Working register multiplier pair (same source register)
Wm * Wn	Working register multiplier pair (different source registers)
Wn	Both source and destination Working register (Direct Addressing)
Wnd Destination Working register (Direct Addressing)
Wns Source Working register (Direct Addressing)
WREG Default Working register (assigned to W0)
Ws	Source Working register (Direct and Indirect Addressing)
Wso	Source Working register (Direct and Indirect Addressing, including Indirect Addressing with Offset)
Wx	Source Addressing mode and Working register for X data bus prefetch
Wxd Destination Working register for X data bus prefetch
Wy	Source Addressing mode and Working register for Y data bus prefetch
Wyd Destination Working register for Y data bus prefetch

Note 1: The range of each symbol is instruction-dependent. Refer to Section 5. "Instruction Descriptions" for the specific instruction range.

Section 2. Programmer's Model

HIGHLIGHTS

This section of the manual contains the following major topics:

2.1 16-Bit MCU and DSC Core Architecture Overview.... 10
2.2 Programmer's Model....14
2.3 Working Register Array 19
2.4 Default Working Register (WREG) 20
2.5 Software Stack Frame Pointer 20
2.6 Software Stack Pointer....20
2.7 Stack Pointer Limit Register (SPLIM)....20
2.8 Accumulator A and Accumulator B (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)....20
2.9 Program Counter 21
2.10 TBLPAG Register 21
2.11 PSVPAG Register (PIC24F, PIC24H, dsPIC30F and dsPIC33F) 21
2.12 RCOUNT Register 21
2.13 DCOUNT Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) ..... 21
2.14 DOSTART Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)..... 22
2.15 DOEND Register (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices).... 22
2.16 STATUS Register 22
2.17 Core Control Register 25
2.18 Shadow Registers.... 25
2.19 DO Stack (dsPIC33E and dsPIC33C Devices) 26

2.1 16-BIT MCU AND DSC CORE ARCHITECTURE OVERVIEW

This section provides an overview of the 16-bit architecture features and capabilities for the following families of devices:

• 16-Bit Microcontrollers (MCU):

• P I C 2 4 F
• P I C 2 4 H
• P I C 2 4 E

• 16-Bit Digital Signal Controllers (DSC):

• d s P I C 3 0 F
• d s P I C 3 3 F
• d s P I C 3 3 E
• d s P I C 3 3 C

2.1.1 Features Specific to 16-Bit MCU and DSC Core

The core of the 16-bit MCU and DSC devices is a 16-bit (data) modified Harvard architecture with an enhanced instruction set. The core has a 24-bit instruction word, with an 8-bit opcode field. The Program Counter (PC) is 23 bits wide and addresses up to 4M x 24 bits of user program memory space. An instruction prefetch mechanism is used to help maintain throughput and provides predictable execution. The majority of instructions execute in a single cycle.

2.1.1.1 REGISTERS

The 16-bit MCU and DSC devices have sixteen 16-bit Working registers. Each of the Working registers can act as a data, address or offset register. The 16th Working register (W15) operates as a Software Stack Pointer (SSP) for interrupts and calls.

2.1.1.2 INSTRUCTION SET

The instruction set is almost identical for the 16-bit MCU and DSC architectures. The instruction set includes many addressing modes and was designed for optimum C compiler efficiency.

2.1.1.3 DATA SPACE ADDRESSING

The data space can be addressed as 32K words or 64 Kbytes. The upper 32 Kbytes of the data space memory map can optionally be mapped into program space at any 16K program word boundary, which is a feature known as Program Space Visibility (PSV). The program to data space mapping feature lets any instruction access program space as if it were the data space, which is useful for storing data coefficients.

Note: Some devices families support Extended Data Space (EDS) Addressing. See the specific device data sheet and family reference manual for more details on this feature.

2.1.1.4 ADDRESSING MODES

The core supports Inherent (no operand), Relative, Literal, Memory Direct, Register Direct, Register Indirect and Register Offset Addressing modes. Each instruction is associated with a predefined addressing mode group, depending upon its functional requirements. As many as seven addressing modes are supported for each instruction.

For most instructions, the CPU is capable of executing a data (or program data) memory read, a Working register (data) read, a data memory write and a program (instruction) memory read per instruction cycle. As a result, 3-operand instructions can be supported, allowing A + B = C operations to be executed in a single cycle.

2.1.1.5 ARITHMETIC AND LOGIC UNIT

A high-speed, 17-bit by 17-bit multiplier is included to significantly enhance the core's arithmetic capability and throughput. The multiplier supports Signed, Unsigned, and Mixed modes, as well as 16-bit by 16-bit, or 8-bit by 8-bit integer multiplication. All multiply instructions execute in a single cycle.

The 16-bit Arithmetic Logic Unit (ALU) is enhanced with integer divide assist hardware that supports an iterative non-restoring divide algorithm. It operates in conjunction with the REPEAT instruction looping mechanism, and a selection of iterative divide instructions, to support 32-bit (or 16-bit) divided by 16-bit integer signed and unsigned division. All divide operations require 19 cycles to complete, but are interruptible at any cycle boundary.

2.1.1.6 EXCEPTION PROCESSING

The 16-bit MCU and DSC devices have a vectored exception scheme with support for up to eight sources of non-maskable traps and up to 246 interrupt sources. In both families, each interrupt source can be assigned to one of seven priority levels.

2.1.2 PIC24E, dsPIC33E and dsPIC33C Features

In addition to the information provided in Section 2.1.1 “Features Specific to 16-Bit MCU and DSC Core”, this section describes the enhancements that are available in the PIC24E, dsPIC33E and dsPIC33C families of devices.

2.1.2.1 DATA SPACE ADDRESSING

The Base Data Space address is used in conjunction with a Read or Write Page register (DSRPAG or DSWPAG) to form an Extended Data Space (EDS) address, which can also be used for PSV access. The EDS can be addressed as 8M words or 16 Mbytes. Refer to "Data Memory" (DS70595) in the "dsPIC33/PIC24 Family Reference Manual" for more details on EDS, PSV and table accesses.

Note: Some PIC24F devices also support Extended Data Space. Refer to “CPU with Extended Data Space (EDS)” (DS39732) and “Data Memory with Extended Data Space (EDS)” (DS39733) in the “dsPIC33/PIC24 Family Reference Manual” for details.

2.1.2.2 AUTOMATIC MIXED-SIGN MULTIPLICATION MODE (dsPIC33E AND dsPIC33C ONLY)

In addition to signed and unsigned DSP multiplications, dsPIC33E and dsPIC33C devices support mixed-sign (unsigned-signed and signed-unsigned) multiplications without the need to dynamically reconfigure the Multiplication mode and shift data to account for the difference in operand formats. This mode is particularly beneficial for dsPIC33C executing extended precision (32-bit and 64-bit) algorithms. Besides DSP instructions, MCU multiplication (MUL) instructions can also utilize either accumulator as a result destination, thereby enabling faster extended precision arithmetic. Refer to Section 4.11.1 "Implied DSP Operands (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)" and Section 4.21 "Extended Precision Arithmetic Using Mixed-Sign Multiplications (dsPIC33E and dsPIC33C Only)" for more details on mixed-sign DSP multiplications.

2.1.2.3 MCU MULTIPLICATIONS WITH 16-BIT RESULT

16x16-bit MUL instructions include an option to store the product in a single 16-bit Working register rather than a pair of registers. This feature helps free up a register for other purposes, in cases where the numbers being multiplied are small in magnitude, and therefore, expected to provide a 16-bit result. See the individual MUL instruction descriptions in Section 5.4 "Instruction Descriptions" for more details.

2.1.2.4 HARDWARE STACK FOR DO LOOPS (dsPIC33E AND dsPIC33C ONLY)

The single-level DO Loop Shadow register set has been replaced by a 4-level deep DO loop hardware stack. This provides automatic DO Loop register save/restore for up to 3 levels of DO loop nesting, resulting in more efficient implementation of nested loops. Refer to Section 2.19 "DO Stack (dsPIC33E and dsPIC33C Devices)" for more details on DO loop nesting in dsPIC33E and dsPIC33C devices.

2.1.2.5 DSP CONTEXT SWITCH SUPPORT (dsPIC33E AND dsPIC33C ONLY)

In dsPIC33E and dsPIC33C devices, the DSP Overflow and Saturation Status bits are writable. This allows the state of the DSP engine to be efficiently saved and restored while switching between DSP tasks. See Section 2.16.4 "DSP ALU Status Bits (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)" for more details on DSP Status bits. In addition, dsPIC33C devices have up to four additional sets of DSP Accumulators A and B for fast context switching. Please see the specific device data sheet for details.

2.1.2.6 EXTENDED CALL AND GOTO INSTRUCTIONS (PIC24E, dsPIC33E AND dsPIC33C ONLY)

The CALL.L Wn and GOTO.L Wn instructions extend the capabilities of the CALL Wn and GOTO Wn by enabling 32-bit addresses for computed branch/call destinations. In these enhanced instructions, the destination address is provided by a pair of Working registers, rather than a single 16-bit register. See the CALL.L and GOTO.L instruction descriptions in Section 5.4 "Instruction Descriptions" for more details.

2.1.2.7 COMPARE/BRANCH INSTRUCTIONS (PIC24E, dsPIC33E AND dsPIC33C ONLY)

PIC24E/dsPIC33E/dsPIC33C devices feature conditional Compare/Branch (CPBxx) instructions. These instructions extend the capabilities of the Compare/Skip (CPSxx) instructions by allowing branches, rather than only skipping over a single instruction. See the CPBEQ, CPBNE, CPBGT and CPBLT instruction descriptions in Section 5.4 "Instruction Descriptions" for more details on Compare/Branch instructions.

2.1.3 dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Features

In addition to the information provided in Section 2.1.1 “Features Specific to 16-Bit MCU and DSC Core”, this section describes the DSP enhancements that are available in the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C families of devices.

2.1.3.1 PROGRAMMING LOOP CONSTRUCTS

Overhead-free program loop constructs are supported using the DO instruction, which is interruptible.

2.1.3.2 DSP INSTRUCTION CLASS

The DSP class of instructions are seamlessly integrated into the architecture and execute from a single execution unit.

2.1.3.3 DATA SPACE ADDRESSING

The data space is split into two blocks, referred to as X and Y data memory. Each memory block has its own independent Address Generation Unit (AGU). The MCU class of instructions operates solely through the X memory AGU, which accesses the entire memory map as one linear data space. The DSP dual source class of instructions operates through the X and Y AGUs, which splits the data address space into two parts. The X and Y data space boundary is arbitrary and device-specific.

2.1.3.4 MODULO AND BIT-REVERSED ADDRESSING

Overhead-free circular buffers (Modulo Addressing) are supported in both X and Y address spaces. The Modulo Addressing removes the software boundary checking overhead for DSP algorithms. Furthermore, the X AGU Circular Addressing can be used with any of the MCU class of instructions. The X AGU also supports Bit-Reversed Addressing, to greatly simplify input or output data reordering for radix-2 FFT algorithms.

2.1.3.5 DSP ENGINE

The DSP engine features a high-speed, 17-bit by 17-bit multiplier, a 40-bit ALU, two 40-bit saturating accumulators and a 40-bit bidirectional barrel shifter. The barrel shifter is capable of shifting a 40-bit value, up to 16 bits right or up to 16 bits left, in a single cycle. The DSP instructions operate seamlessly with all other instructions and have been designed for optimal real-time performance. The MAC instruction and other associated instructions can concurrently fetch two data operands from memory while multiplying two Working registers. This requires that the data space be split for these instructions and linear for all others. This is achieved in a transparent and flexible manner through dedicating certain Working registers to each address space.

2.1.3.6 EXCEPTION PROCESSING

The dsPIC30F devices have a vectored exception scheme with support for up to eight sources of non-maskable traps and up to 54 interrupt sources. The dsPIC33F, dsPIC33E and dsPIC33C have a similar exception scheme, but support up to 118, and up to 246 interrupt sources, respectively. In all three families, each interrupt source can be assigned to one of seven priority levels.

Refer to “Interrupts” (DS70000600) of the “dsPIC33/PIC24 Family Reference Manual” for more details on exception processing.

2.2 PROGRAMMER'S MODEL

Figure 2-1 through Figure 2-5 show the programmer's model diagrams for the 16-bit MCU and DSC families of devices.
Figure 2-1: PIC24F and PIC24H Programmer's Model Diagram

DIV and KUL Result Registers 15 0 W0/WREG W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13 W14/Frame Pointer W15/Stack Pointer Working Registers SPLIM Stack Pointer Limit Register 22 0 Program Counter 7 0 TBLPAG Data Table Page Address 7 0 PSVPAG Program Space Visibility Page Address 13 0 RCOUNT REPEAT Loop Counter 15 0 CORCON CPU Core Control Register STATUS Register - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - SRH SRL PUSH.S Shadow Register Legend

Figure 2-2: PIC24E Programmer's Model Diagram

DIV and MUL Result Registers 15 0 W0/WREG W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13 W14/Frame Pointer W15/Stack Pointer Working Registers SPLIM Stack Pointer Limit Register 22 0 Program Counter 7 0 TBLPAG Data Table Page Address 9 0 DSRPAG Data Space Read Page Address 8 0 DSWPAG Data Space Write Page Address 15 0 REPEAT Loop Counter 15 0 CORCON CPU Core Control Register STATUS Register — — — — — — — D CPL2 IPR1 APL0 OWN C Z SRH SRL

Figure 2-3: dsPIC30F and dsPIC33F Programmer's Model Diagram

DIV and MUL Result Registers W0/WREG W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13 W14/Frame Pointer W15/Stack Pointer PUSH.S Shadow Register DO Shadow Register Legend Working Registers SPLIM Stack Pointer Limit Register 39 31 15 0 ACCA DSP ACC B Accumulators 22 0 Program Counter 7 0 Data Table Page Address 7 0 Program Space Visibility Page Address 13 RCOUNT REPEAT Loop Counter 13 DCOUNT DO Loop Counter 24 0 DO Start Address 24 0 DO End Address 15 CORCON CPU Core Control Register STATUS Register OA OB SA SB OAB SAB DA DC RA IPL2 IPL1 IPL0 OV Z SRH SRL

Figure 2-4: dsPIC33E Programmer's Model Diagram

DIV and MUL Result Registers MAC Operand Registers MAC Address Registers W0/WREG W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13 W14/Frame Pointer W15/Stack Pointer 015 Working Registers(1) SPLIM Stack Pointer Limit Register 39 031 15 ACCA ACC B DSP Accumulators 22 Program Counter TBLPAG Data Table Page Address 9 DSRPAG X Data Space Read Page Address 8 DSWPAG X Data Space Write Page Address 15 REPEAT Loop Counter RCOUNT DO Loop Counter 15 DCOUNT DO Loop Start Address 24 0 DOSTART 0 DO Loop End Address 24 0 DOEND 0 CPU Core Control Register 15 CORCON STATUS Register OA OB SA SB OAB SAB DA DC IPL2 IPL1 RA N CIPLO OV Z SRH SRL

Note 1: Some dsPIC33E devices have up to four additional sets of Working registers (W0-W14) for context switching. Please see the specific device data sheet for details.

Figure 2-5: dsPIC33C Programmer's Model

graph TD
    A["Working/Address Registers"] --> B["W0-W3"]
    A --> C["DSP Operand Registers"]
    A --> D["DSP Address Registers"]
    B --> E["W0 (WREG)"]
    C --> F["W1"]
    C --> G["W2"]
    C --> H["W3"]
    C --> I["W4"]
    C --> J["W5"]
    C --> K["W6"]
    C --> L["W7"]
    C --> M["W8"]
    C --> N["W9"]
    C --> O["W10"]
    C --> P["W11"]
    C --> Q["W12"]
    C --> R["W13"]
    C --> S["Frame Pointer/W14"]
    C --> T["Stack Pointer/W15"]
    D --> U["ACCA"]
    D --> V["ACCB"]
    U --> W["PC23"]
    V --> X["PC0"]
    W --> Y["Program Counter"]
    X --> Z["TBLPAG"]
    Y --> AA["X Data Space Read Page Address"]
    Z --> AB["DO Loop Counter and Stack"]
    W --> AC["RCOUNT"]
    X --> AD["DCOUNT"]
    AC --> AE["DO Loop Start Address and Stack"]
    AD --> AF["DO Loop End Address and Stack"]
    AE --> AG["CPU Core Control Register"]
    AF --> AH["CPU Core Control Register"]
    subgraph Stack Pointer Limit
        AI["SPLIM"] --> AJ["AD39"]
        AK["AD31"] --> AL["AD39"]
        AM["AD15"] --> AN["AD31"]
        AO["AD0"] --> AP["AD31"]
        AQ["AD15"] --> AR["AD31"]
        AS["AD0"] --> AT["AD31"]
        AU["AD0"] --> AV["AD39"]
        AW["AD0"] --> AX["AD39"]
        AY["AD0"] --> AZ["AD39"]
        BA["AD0"] --> BB["AD39"]
        BC["AD0"] --> BD["AD39"]
        BE["AD0"] --> BF["AD39"]
        BG["AD0"] --> BH["AD39"]
        BI["AD0"] --> BJ["AD39"]
        BK["AD0"] --> BL["AD39"]
        BM["AD0"] --> BN["AD39"]
        BO["AD0"] --> BP["AD39"]
        BQ["AD0"] --> BR["AD39"]
        BS["AD0"] --> BT["AD39"]
        BU["AD0"] --> BV["AD39"]
        BW["AD0"] --> BX["AD39"]
        BY["AD0"] --> BZ["AD39"]
        CA["AD0"] --> CB["AD39"]
        CC["AD0"] --> CD["AD39"]
        DE["AD0"] --> DF["AD39"]
        DG["AD0"] --> DH["AD39"]
        DI["AD0"] --> DJ["AD39"]
        DK["AD0"] --> DL["AD39"]
        DM["AD0"] --> DN["AD39"]
        DO["AD0"] --> DP["AD39"]
        QD["AD0"] --> DR["AD39"]
        EQ["AD0"] --> RQ["AD39"]
        SC["AD0"] --> SD["AD39"]
        TX["AD0"] --> TXA["AD39"]
        TXB["AD0"] --> TXB
        TXC["AD0"] --> TXC
        TXD["AD0"] --> TXD
        TXE["AD0"] --> TXE
        TXF["AD0"] --> TXF
        TXG["AD0"] --> TXG
        TXH["AD0"] --> TXH
        TXI["AD0"] --> TXI
        TXJ["AD0"] --> TXJ
        TXK["AD0"] --> TXK
        TXL["AD0"] --> TXL
        TXM["AD0"] --> TXM
        TXN["AD0"] --> TXN
        TXO["AD0"] --> TXO
        TXP["AD0"] --> TXP
        TXQ["AD0"] --> TXQ
        TXR["AD0"] --> TXR
        TXS["AD0"] --> TXS
        TXT["AD0"] --> TXT
        TXU["AD0"] --> TXU
        TXV["AD0"] --> TXV
        TXW["AD0"] --> TXW
        TXX["AD0"] --> TXX
        TXY["AD0"] --> TXY
        TXZ["AD0"] --> TXZ
        TXW' --> DC["SRL"]
    end

All registers in the programmer's model are memory-mapped and can be manipulated directly by the instruction set. A description of each register is provided in Table 2-1.

Note: Unless otherwise specified, the Programmer's Model register descriptions in Table 2-1 apply to all MCU and DSC device families.

Table 2-1: Programmer's Model Register Descriptions

Register Description
CORCON CPU Core Configuration register
PC 23-Bit Program Counter
PSVPAG^(1)	Program Space Visibility Page Address register
DSRPAG^(2)	Extended Data Space (EDS) Read Page register
DSWPAG^(2)	Extended Data Space (EDS) Write Page register
RCOUNT REPEAT Loop Counter register
SPLIM Stack Pointer Limit Value register
SR ALU and DSP Engine STATUS Register
TBLPAG	Table Memory Page Address register
W0-W15 ^(4)	Working register array
ACCA, ACCB^(3,5)	40-Bit DSP Accumulators
DCOUNT^(3)	DO Loop Counter register
DOSTART^(3)	DO Loop Start Address register
DOEND^(3)	DO Loop End Address register

Note 1: This register is only available on PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
2: This register is only available on PIC24E, dsPIC33E and dsPIC33C devices.
3: This register is only available on dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
4: dsPIC33C devices and some dsPIC33E devices have up to four additional sets of Working registers for context switching. Please see the device data sheet for details.
5: dsPIC33C devices have up to four additional sets of accumulators for context switching. Please see the device data sheet for details.

2.3 WORKING REGISTER ARRAY

The 16 Working (W) registers can function as data, address or offset registers. The function of a W register is determined by the instruction that accesses it.

Byte instructions, which target the Working register array, only affect the Least Significant Byte (LSB) of the target register. Since the Working registers are memory-mapped, the Least and Most Significant Bytes can be manipulated through byte-wide data memory space accesses.

Note: dsPIC33C devices and some dsPIC33E devices have up to four additional sets of Working registers for context switching. Please see the device data sheet to find out the exact number of register contexts available on a device. The context switching can be performed quickly using the CTXTSWP instruction.

2.4 DEFAULT WORKING REGISTER (WREG)

The instruction set can be divided into two instruction types: Working register instructions and file register instructions. The Working register instructions use the Working register array as data values or as addresses that point to a memory location. In contrast, file register instructions operate on a specific memory address contained in the instruction opcode.

File register instructions that also utilize a Working register do not specify the Working register that is to be used for the instruction. Instead, a default Working register (WREG) is used for these file register instructions. Working register, W0, is assigned to be the WREG. The WREG assignment is not programmable.

2.5 SOFTWARE STACK FRAME POINTER

A frame is a user-defined section of memory in the stack, used by a function to allocate memory for local variables. W14 has been assigned for use as a Stack Frame Pointer with the link (LNK) and unlink (ULNK) instructions. However, if a Stack Frame Pointer and the LNK and ULNK instructions are not used, W14 can be used by any instruction in the same manner as all other W registers. On dsPIC33E, dsPIC33C and PIC24E devices, a Stack Frame Active (SFA) Status bit is used to support nested stack frames. See Section 4.8.2 "Software Stack Frame Pointer" for detailed information about the Frame Pointer.

2.6 SOFTWARE STACK POINTER

W15 serves as a dedicated Software Stack Pointer, and will be automatically modified by function calls, exception processing and returns. However, W15 can be referenced by any instruction in the same manner as all other W registers. This simplifies reading, writing and manipulating the Stack Pointer. Refer to Section 4.8.1 "Software Stack Pointer" for detailed information about the Stack Pointer.

2.7 STACK POINTER LIMIT REGISTER (SPLIM)

The SPLIM is a 16-bit register associated with the Stack Pointer. It is used to prevent the Stack Pointer from overflowing and accessing memory beyond the user allocated region of stack memory. Refer to Section 4.8.3 "Stack Pointer Overflow" for detailed information about the SPLIM.

2.8 ACCUMULATOR A AND ACCUMULATOR B (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

Accumulator A (ACCA) and Accumulator B (ACCB) are 40-bit wide registers, utilized by DSP instructions to perform mathematical and shifting operations. Each accumulator is composed of three memory-mapped registers:

• AccxU (bits 39-32)
• AccxH (bits 31-16)
• AccxL (bits 15-0)

In dsPIC33E devices, Accumulator A and Accumulator B can also be used as destination registers in MCU MUL.xx instructions. This helps reduce the execution time of extended precision arithmetic operations.

Refer to Figure 4-13 for details on using ACCA and ACCB.

Note: dsPIC33C devices have up to four additional sets of accumulators for context switching. Please see the device data sheet to find out the exact number of register contexts available on a device. The context switching can be performed quickly using the CTXTSWP instruction.

2.9 PROGRAM COUNTER

The Program Counter (PC) is 23 bits wide. Instructions are addressed in the 4M x 24-bit user program memory space by PC<22:1>, where PC<0> is always set to '0' to maintain instruction word alignment and provide compatibility with Data Space Addressing. This means that during normal instruction execution, the PC increments by two.

Program memory, located at 0x800000 and above, is utilized for device configuration data, Unit ID and Device ID. This region is not available for user code execution and the PC cannot access this area. However, one may access this region of memory using table instructions. For details on accessing the configuration data, Unit ID and Device ID, refer to the specific device family reference manual.

2.10 TBLPAG REGISTER

The TBLPAG register is used to hold the upper eight bits of a program memory address during table read and write operations. Table instructions are used to transfer data between program memory space and data memory space. For details on accessing program memory with the table instructions, refer to the family reference manual of the specific device.

2.11 PSVPAG REGISTER (PIC24F, PIC24H, dsPIC30F AND dsPIC33F)

Program Space Visibility (PSV) allows the user to map a 32-Kbyte section of the program memory space into the upper 32 Kbytes of data address space. This feature allows transparent access of constant data through instructions that operate on data memory. The PSVPAG register selects the 32-Kbyte region of program memory space that is mapped to the data address space. For details on Program Space Visibility, refer to the specific device family reference manual.

2.12 RCOUNT REGISTER

The 14-bit RCOUNT register (16-bit for PIC24E, dsPIC33E and dsPIC33C devices) contains the loop counter for the REPEAT instruction. When a REPEAT instruction is executed, RCOUNT is loaded with the repeat count of the instruction, either "lit14" for the "REPEAT #lit14" instruction ("lit15" for the "REPEAT #lit15" instruction for PIC24E, dsPIC33E and dsPIC33C devices) or the 14 LSbs of the Wn register for the "REPEAT Wn" instruction (entire Wn for PIC24E, dsPIC33E and dsPIC33C devices). The REPEAT loop will be executed RCOUNT + 1 time.

Note 1: If a REPEAT loop is executing and gets interrupted, RCOUNT may be cleared by the Interrupt Service Routine (ISR) to break out of the REPEAT loop when the foreground code is re-entered.

2: Refer to the specific device family reference manual for complete details about REPEAT loops.

2.13 DCOUNT REGISTER (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The 14-bit DCOUNT register (16-bit for dsPIC33E and dsPIC33C devices) contains the loop counter for hardware DO loops. When a DO instruction is executed, DCOUNT is loaded with the loop count of the instruction, either "lit14" for the "DO #lit14, Expr" instruction ("lit15" for the "DO #lit15, Expr" instruction for dsPIC33E devices) or the 14 LSbs of the Ws register for the "DO Ws, Expr" instruction (entire Wn for dsPIC33E devices). The DO loop will be executed DCOUNT + 1 time.

Note 1: In dsPIC30F and dsPIC33F devices, the DCOUNT register contains a shadow register. See Section 2.18 "Shadow Registers" for information on shadow registers.

2: The dsPIC33E devices have a 4-level deep, nested DO stack instead of a shadow register.

3: Refer to the specific device family reference manual for complete details about DO loops.

2.14 DOSTART REGISTER (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The DOSTART register contains the starting address for a hardware DO loop. When a DO instruction is executed, DOSTART is loaded with the address of the instruction that follows the DO instruction. This location in memory is the start of the DO loop. When looping is activated, program execution continues with the instruction stored at the DOSTART address after the last instruction in the DO loop is executed. This mechanism allows for zero overhead looping.

Note 1: For dsPIC30F and dsPIC33F devices, DOSTART has a shadow register. See Section 2.18 "Shadow Registers" for information on shadowing.

2: The dsPIC33E and dsPIC33C devices have a 4-level deep, nested DO stack instead of a shadow register. The DOSTART register is read-only in dsPIC33E and dsPIC33C devices.
3: Refer to the specific device family reference manual for complete details about DO loops.

2.15 DOEND REGISTER (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The DOEND register contains the ending address for a hardware DO loop. When a DO instruction is executed, DOEND is loaded with the address specified by the expression in the DO instruction. This location in memory specifies the last instruction in the DO loop. When looping is activated and the instruction stored at the DOEND address is executed, program execution will continue from the DO loop start address (stored in the DOSTART register).

Note 1: For dsPIC30F and dsPIC33F devices, DOEND has a shadow register. See Section 2.18 "Shadow Registers" for information on shadow registers.

2: The dsPIC33E and dsPIC33C devices have a 4-level deep, nested DO stack instead of a shadow register.
3: Refer to the specific device family reference manual for complete details about DO loops.

2.16 STATUS REGISTER

The 16-bit STATUS Register maintains status information for the instructions which have been executed most recently. Operation Status bits exist for MCU operations, loop operations and DSP operations. Additionally, the STATUS Register contains the CPU Interrupt Priority Level bits, IPL<2:0>, which are used for interrupt processing.

Depending on the MCU and DSC family, one of the following STATUS Registers is used:

• Register 2-1 for PIC24F, PIC24H and PIC24E devices
• Register 2-2 for dsPIC30F and dsPIC33F devices
• Register 2-3 for dsPIC33E and dsPIC33C devices

2.16.1 MCU ALU Status Bits

The MCU operation Status bits are either affected or used by the majority of instructions in the instruction set. Most of the logic, math, rotate/shift and bit instructions modify the MCU Status bits after execution, and the conditional branch instructions use the state of individual Status bits to determine the flow of program execution. All conditional branch instructions are listed in Section 4.9 "Conditional Branch Instructions".

The Carry (C), Zero (Z), Overflow (OV), Negative (N) and Digit Carry (DC) bits show the immediate status of the MCU ALU by indicating whether an operation has resulted in a Carry, Zero, Overflow, Negative result or Digit Carry. When a subtract operation is performed, the C flag is used as a Borrow flag.

The Z Status bit is useful for extended precision arithmetic. The Z Status bit functions like a normal Z flag for all instructions except those that use a carry or borrow input (ADDC, CPB, SUBB and SUBBR). See Section 4.10 "Z Status Bit" for more detailed information.

Note 1: All MCU bits are shadowed during execution of the PUSH.S instruction and they are restored on execution of the POP.S instruction.

2: All MCU bits, except the DC flag (which is not in the SRL), are stacked during exception processing (see Section 4.8.1 "Software Stack Pointer").

2.16.2 REPEAT Loop Active (RA) Status Bit

The REPEAT Loop Active bit (RA) is used to indicate when looping is active. The RA flag indicates that a REPEAT instruction is being executed and it is only affected by the REPEAT instructions. The RA flag is set to '1' when the instruction being repeated begins execution and it is cleared when the instruction being repeated completes execution for the last time.

Since the RA flag is also read-only, it may not be directly cleared. However, if a REPEAT or its target instruction is interrupted, the Interrupt Service Routine may clear the RA flag of the SRL, which resides on the stack. This action will disable looping once program execution returns from the Interrupt Service Routine, because the restored RA will be '0'.

2.16.3 DO Loop Active (DA) Status Bit (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)

The DO Loop Active bit (DA) is used to indicate when looping is active. The DO instructions affect the DA flag, which indicates that a DO loop is active. The DA flag is set to '1' when the first instruction of the DO loop is executed and it is cleared when the last instruction of the loop completes final execution.

The DA flag is read-only. This means that looping is not initiated by writing a '1' to DA, nor is it terminated by writing a '0' to DA. If a DO loop must be terminated prematurely, the EDT bit (CORCON<11>) should be used.

2.16.4 DSP ALU Status Bits (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)

The high byte of the STATUS Register (SRH) is used by the DSP class of instructions and it is modified when data passes through one of the adders. The SRH provides status information about overflow and saturation for both accumulators. The Saturate A, Saturate B, Overflow A and Overflow B (SA, SB, OA, OB) bits provide individual accumulator status, while the Saturate AB and Overflow AB (SAB, OAB) bits provide combined accumulator status. The SAB and OAB bits provide an efficient method for the software developer to check the register for saturation or overflow.

The OA and OB bits are used to indicate when an operation has generated an overflow into the guard bits (bits 32 through 39) of the respective accumulator. This condition can only occur when the processor is in Super Saturation mode or if saturation is disabled. It indicates that the operation has generated a number which cannot be represented with the lower 31 bits of the accumulator. The OA and OB bits are writable in dsPIC33E and dsPIC33C devices.

The SA and SB bits are used to indicate when an operation has generated an overflow out of the MSb of the respective accumulator. The SA and SB bits are active, regardless of the Saturation mode (Disabled, Normal or Super) and may be considered "sticky". Namely, once the SA or SB bit is set to '1', it can only be cleared manually by software, regardless of subsequent DSP operations. When it is required, the BCLR instruction can be used to clear the SA or SB bit.

In addition, the SA and SB bits can be set by software in dsPIC33E and dsPIC33C devices, enabling efficient context state switching.

For convenience, the OA and OB bits are logically ORed together to form the OAB flag, and the SA and SB bits are logically ORed to form the SAB flag. These cumulative Status bits provide efficient overflow and saturation checking when an algorithm is implemented. Instead of interrogating the OA and OB bits independently for arithmetic overflows, a single check of OAB can be performed. Likewise, when checking for saturation, SAB may be examined instead of checking both the SA and SB bits. Note that clearing the SAB flag will clear both the SA and SB bits.

2.16.5 Interrupt Priority Level Status Bits

The three Interrupt Priority Level (IPL) bits of the SRL (SR<7:5>) and the IPL3 bit (CORCON<3>) set the CPU's IPL, which is used for exception processing. Exceptions consist of interrupts and hardware traps. Interrupts have a user-defined priority level between 0 and 7, while traps have a fixed priority level between 8 and 15. The fourth Interrupt Priority Level bit, IPL3, is a special IPL bit that may only be read or cleared by the user. This bit is only set when a hardware trap is activated and it is cleared after the trap is serviced.

The CPU's IPL identifies the lowest level exception which may interrupt the processor. The interrupt level of a pending exception must always be greater than the CPU's IPL for the CPU to process the exception. This means that if the IPL is 0, all exceptions at Priority Level 1 and above may interrupt the processor. If the IPL is 7, only hardware traps may interrupt the processor.

When an exception is serviced, the IPL is automatically set to the priority level of the exception being serviced, which will disable all exceptions of equal and lower priority. However, since the IPL field is read/write, one may modify the lower three bits of the IPL in an Interrupt Service Routine to control which exceptions may preempt the exception processing. Since the SRL is stacked during exception processing, the original IPL is always restored after the exception is serviced. If required, one may also prevent exceptions from nesting by setting the NSTDIS bit (INTCON1<15>).

Note: For more detailed information on exception processing, refer to the family reference manual of the specific device.

2.17 CORE CONTROL REGISTER

For all MCU and DSC devices, the 16-bit CPU Core Control register (CORCON) is used to set the configuration of the CPU. This register provides the ability to map program space into data space.

In addition to setting CPU modes, the CORCON register contains status information about the IPL<3> Status bit, which indicates if a trap exception is being processed.

Depending on the MCU and DSC family, one of the following CORCON registers is used:

• Register 2-4 for PIC24F and PIC24H devices
• Register 2-5 for PIC24E devices
• Register 2-6 for dsPIC30F and dsPIC33F devices
• Register 2-7 for dsPIC33E and dsPIC33C devices

2.17.1 dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Specific Bits

In addition to setting CPU modes, the following features are available through the CORCON register:

• Sets the ACCA and ACCB saturation enable
• Sets the Data Space Write Saturation mode
• Sets the Accumulator Saturation and Rounding modes
• Sets the Multiplier mode for DSP operations
• Terminates DO loops prematurely
• Provides status information about the DO loop nesting level (DL<2:0>)
• Selects fixed or variable interrupt latency (dsPIC33E and dsPIC33C only)

2.17.1.1 PIC24E, dsPIC33E AND dsPIC33C SPECIFIC BITS

A Status bit (SFA) is available that indicates whether the stack frame is active.

Note: PIC24E, dsPIC33E and dsPIC33C devices do not have a PSV control bit; it has been replaced by the SFA bit.

2.18 SHADOW REGISTERS

A shadow register is used as a temporary holding register and can transfer its contents to or from the associated host register when instructed. Some of the registers in the programmer's model have a shadow register, which is utilized during the execution of a DO, POP.S or PUSH.S instruction. Shadow register usage is shown in Table 2-2.

Note: The DO instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

Table 2-2: Automatic Shadow Register Usage

Location DO	(1)	POP.S/PUSH.S
DCOUNT^(1)	Yes	—
DOSTART^(1)	Yes	—
DOEND^(1)	Yes	—
STATUS Register – DC, N, OV, Z and C bits	—	Yes
W0-W3	—	Yes

Note 1: The DO Shadow registers are only available in dsPIC30F and dsPIC33F devices.

For dsPIC30F and dsPIC33F devices, since the DCOUNT, DOSTART and DOEND registers are shadowed, the ability to nest DO loops without additional overhead is provided. Since all shadow registers are one register deep, up to one level of DO loop nesting is possible. Further nesting of DO loops is possible in software, with support provided by the DO Loop Nesting Level Status bits (DL<2:0>) in the CORCON register (CORCON<10:8>).

Note: All shadow registers are one register deep and not directly accessible. Additional shadowing may be performed in software using the software stack.

2.19 DO STACK (dsPIC33E AND dsPIC33C DEVICES)

The DO stack is used to preserve the following elements associated with a DO loop underway when another DO loop is encountered (i.e., a nested DO loop).

• DOSTART register value
• DOEND register value
- DCOUNT register value

Note that the DO Level Status field (DL<2:0>) also acts as a pointer to address the DO stack. After the DO instruction is executed, the DO Level Status field (DL<2:0>) points to the next free entry.

The DOSTART, DOEND, and DCOUNT registers each have an associated hardware stack that allows the DO loop hardware to support up to three levels of nesting. A conceptual representation of the DO stack is shown in Figure 2-6.

Figure 2-6: do Stack Conceptual Diagram

DCOUNTDOENDDOSTARTDL<2:0> Empty Level 1 Registers Level 2 Registers Level 3 Registers

Note 1: For DO register entries, DL<2:0> bits represent the value before the DO stack is executed.

2: For DO instruction buffer entries, DL<2:0> bits represent the value after the DO stack is executed.

3: If DL<2:0>=000, no DO loops are active (DA=0).

Register 2-1: SR: CPU STATUS Register (PIC24H, PIC24F and PIC24E Devices)

U-0 U-0 U-0 U-0 U-0 U-0 R/W-0
——	——	D	C
bit 15 bit 8

R/W-0	R/W-0	R/W-0	R-0	R/W-0	R/W-0	R/W-0	R/W-0
IPL2^(1,2)	IPL1^(1,2)	IPL0^(1,2)	RA NOVZC
bit 7 bit 0

Legend:

R = Readable bit	W = Writable bit	U = Unimplemented bit, read as '0'
-n = Value at POR	'1' = Bit is set	'0' = Bit is cleared x = Bit is unknown

bit 15-9 Unimplemented: Read as '0'

bit 8 DC: MCU ALU Half Carry/Borrow bit

1 = A carry-out from the 4th low-order bit (for byte-sized data) or 8th low-order bit (for word-sized data) of the result occurred 0 = No carry-out from the 4th low-order bit (for byte-sized data) or 8th low-order bit (for word-sized data) of the result occurred

bit 7-5 IPL<2:0>: CPU Interrupt Priority Level Status bits ^(1,2)

111 = CPU Interrupt Priority Level is 7 (15); user interrupts are disabled 110 = CPU Interrupt Priority Level is 6 (14) 101 = CPU Interrupt Priority Level is 5 (13) 100 = CPU Interrupt Priority Level is 4 (12) 011 = CPU Interrupt Priority Level is 3 (11) 010 = CPU Interrupt Priority Level is 2 (10) 001 = CPU Interrupt Priority Level is 1 (9) 000 = CPU Interrupt Priority Level is 0 (8)

bit 4 RA: REPEAT Loop Active bit

1 = REPEAT loop is in progress 0 = REPEAT loop is not in progress

bit 3 N: MCU ALU Negative bit

1 = Result was negative 0 = Result was non-negative (zero or positive)

bit 2 OV: MCU ALU Overflow bit

This bit is used for signed arithmetic (2's complement). It indicates an overflow of the magnitude that causes the sign bit to change state.

1 = Overflow occurred for signed arithmetic (in this arithmetic operation) 0 = No overflow occurred

bit 1 Z: MCU ALU Zero bit

1 = An operation that affects the Z bit has set it at some time in the past 0 = The most recent operation that affects the Z bit has cleared it (i.e., a non-zero result)

bit 0 C: MCU ALU Carry/Borrow bit

1 = A carry-out from the MSb occurred 0 = No carry-out from the MSb occurred

Note 1: The IPL<2:0> bits are concatenated with the IPL3 bit (CORCON<3>) to form the CPU Interrupt Priority Level. The value in parentheses indicates the IPL, if IPL3 = 1. User interrupts are disabled when IPL<3> = 1.

2: The IPL<2:0> Status bits are read-only when the NSTDIS bit (INTCON1<15>) = 1. Refer to the family reference manual of the specific device family to see the associated interrupt register.

Register 2-2: SR: CPU STATUS Register (dsPIC30F and dsPIC33F Devices)

R-0 R-0 R/C-0 R/C-0 R-0 R/C-0 R-0 R/W-0
OA OB SA		(1,2)	SB(1,2)	OAB SAB	(1,2,3)	DA(4)	DC
bit 15 bit 8

R/W-0	R/W-0	R/W-0	R-0	R/W-0	R/W-0	R/W-0	R/W-0
IPL2^(5)	IPL1^(5)	IPL0^(5)	RA	N	OV	Z	C
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15 OA: Accumulator A Overflow bit

1 = Accumulator A overflowed
0 = Accumulator A has not overflowed

bit 14 OB: Accumulator B Overflow bit

1 = Accumulator B overflowed
0 = Accumulator B has not overflowed

bit 13 SA: Accumulator A Saturation bit (1,2)

1 = Accumulator A is saturated or has been saturated since this bit was last cleared
0 = Accumulator A is not saturated

bit 12 SB: Accumulator B Saturation bit (1,2)

1 = Accumulator B is saturated or has been saturated at since this bit was last cleared
0 = Accumulator B is not saturated

bit 11 OAB: OA || OB Combined Accumulator Overflow bit

1 = Accumulator A or B has overflowed
0 = Neither Accumulator A nor B has overflowed

bit 10 SAB: SA || SB Combined Accumulator bit (1,2,3)

1 = Accumulator A or B is saturated or has been saturated since this bit was last cleared
0 = Neither Accumulator A nor B is saturated

bit 9 DA: DO Loop Active bit ^(4)

1 = DO loop is in progress
0 = DO loop is not in progress

bit 8 DC: MCU ALU Half Carry bit

1 = A carry-out from the MSb of the lower nibble occurred
0 = No carry-out from the MSb of the lower nibble occurred

Note 1: This bit may be read or cleared, but not set.

2: Once this bit is set, it must be cleared manually by software.
3: Clearing this bit will clear SA and SB.
4: This bit is read-only.
5: The IPL<2:0> bits are concatenated with the IPL3 bit (CORCON<3>) to form the CPU Interrupt Priority Level. The value in parentheses indicates the IPL, if IPL3 = 1.

Register 2-2: SR: CPU STATUS Register (dsPIC30F and dsPIC33F Devices) (Continued)
bit 7-5 IPL<2:0>: Interrupt Priority Level bits (5) 111 = CPU Interrupt Priority Level is 7 (15); user interrupts are disabled 110 = CPU Interrupt Priority Level is 6 (14) 101 = CPU Interrupt Priority Level is 5 (13) 100 = CPU Interrupt Priority Level is 4 (12) 011 = CPU Interrupt Priority Level is 3 (11) 010 = CPU Interrupt Priority Level is 2 (10) 001 = CPU Interrupt Priority Level is 1 (9) 000 = CPU Interrupt Priority Level is 0 (8)

bit 4 RA: REPEAT Loop Active bit 1 = REPEAT loop is in progress 0 = REPEAT loop is not in progress

bit 3 N: MCU ALU Negative bit 1 = The result of the operation was negative 0 = The result of the operation was not negative

bit 2 OV: MCU ALU Overflow bit 1 = Overflow occurred 0 = No overflow occurred

bit 1 Z: MCU ALU Zero bit 1 = The result of the operation was zero 0 = The result of the operation was not zero bit 0 C: MCU ALU Carry/Borrow bit 1 = A carry-out from the MSb occurred 0 = No carry-out from the MSb occurred

Note 1: This bit may be read or cleared, but not set. 2: Once this bit is set, it must be cleared manually by software. 3: Clearing this bit will clear SA and SB. 4: This bit is read-only. 5: The IPL<2:0> bits are concatenated with the IPL3 bit (CORCON<3>) to form the CPU Interrupt Priority Level. The value in parentheses indicates the IPL, if IPL3 = 1.

Register 2-3: SR: CPU STATUS Register (dsPIC33E and dsPIC33C Devices)

R/W-0 R/W-0 R/W-0 R/W-0 R/C-0 R/C-0 R-0 R/W-0
OA OB SA		(3)	SB(3)	OAB SAB	DA DC
bit 15 bit 8

R/W-0	R/W-0	R/W-0	R-0	R/W-0	R/W-0	R/W-0	R/W-0
IPL2^(1,2)	IPL1^(1,2)	IPL0^(1,2)	RA	N	OV	Z	C
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15 OA: Accumulator A Overflow Status bit

1 = Accumulator A has overflowed
0 = Accumulator A has not overflowed

bit 14 OB: Accumulator B Overflow Status bit

1 = Accumulator B has overflowed
0 = Accumulator B has not overflowed

bit 13 SA: Accumulator A Saturation Status bit ^(3)

1 = Accumulator A is saturated or has been saturated since this bit was last cleared
0 = Accumulator A is not saturated

bit 12 SB: Accumulator B Saturation Status bit ^(3)

1 = Accumulator B is saturated or has been saturated since this bit was last cleared
0 = Accumulator B is not saturated

bit 11 OAB: OA || OB Combined Accumulator Overflow Status bit

1 = Accumulator A or B has overflowed
0 = Neither Accumulator A nor B has overflowed

bit 10 SAB: SA || SB Combined Accumulator Status bit

1 = Accumulator A or B is saturated or has been saturated since this bit was last cleared
0 = Neither Accumulator A nor B is saturated

bit 9 DA: DO Loop Active bit

1 = DO loop is in progress
0 = DO loop is not in progress

bit 8 DC: MCU ALU Half Carry/Borrow bit

1 = A carry-out from the 4th low-order bit (for byte-sized data) or 8th low-order bit (for word-sized data) of the result occurred
0 = No carry-out from the 4th low-order bit (for byte-sized data) or 8th low-order bit (for word-sized data) of the result occurred

Note 1: The IPL<2:0> bits are concatenated with the IPL3 bit (CORCON<3>) to form the CPU Interrupt Priority Level. The value in parentheses indicates the IPL, if IPL3 = 1. User interrupts are disabled when IPL3 = 1.

2: The IPL<2:0> Status bits are read-only when the NSTDIS bit (INTCON1<15>) = 1. Refer to the family reference manual of the specific device family to see the associated interrupt register.
3: A data write to SR can modify the SA or SB bits by either a data write to SA and SB or by clearing the SAB bit. To avoid a possible SA/SB bit write race condition, the SA and SB bits should not be modified using bit operations.

Register 2-3: SR: CPU STATUS Register (dsPIC33E and dsPIC33C Devices) (Continued)
bit 7-5 IPL<2:0>: CPU Interrupt Priority Level Status bits ^(1,2) 111 = CPU Interrupt Priority Level is 7 (15); user interrupts are disabled

110 = CPU Interrupt Priority Level is 6 (14)

101 = CPU Interrupt Priority Level is 5 (13)

100 = CPU Interrupt Priority Level is 4 (12)

011 = CPU Interrupt Priority Level is 3 (11)

010 = CPU Interrupt Priority Level is 2 (10)

001 = CPU Interrupt Priority Level is 1 (9)

000 = CPU Interrupt Priority Level is 0 (8)

bit 4 RA: REPEAT Loop Active bit 1 = REPEAT loop is in progress 0 = REPEAT loop is not in progress

bit 3 N: MCU ALU Negative bit 1 = Result was negative 0 = Result was non-negative (zero or positive)

bit 2 OV: MCU ALU Overflow bit This bit is used for signed arithmetic (2's complement). It indicates an overflow of the magnitude that causes the sign bit to change state. 1 = Overflow occurred for signed arithmetic (in this arithmetic operation) 0 = No overflow occurred

bit 1 Z: MCU ALU Zero bit 1 = The result of the operation was zero 0 = The result of the operation was not zero

bit 0 C: MCU ALU Carry/Borrow bit 1 = A carry-out from the MSb of the result occurred 0 = No carry-out from the MSb of the result occurred

Note 1: The IPL<2:0> bits are concatenated with the IPL3 bit (CORCON<3>) to form the CPU Interrupt Priority Level. The value in parentheses indicates the IPL, if IPL3 = 1. User interrupts are disabled when IPL3 = 1. 2: The IPL<2:0> Status bits are read-only when the NSTDIS bit (INTCON1<15>) = 1. Refer to the family reference manual of the specific device family to see the associated interrupt register. 3: A data write to SR can modify the SA or SB bits by either a data write to SA and SB or by clearing the SAB bit. To avoid a possible SA/SB bit write race condition, the SA and SB bits should not be modified using bit operations.

Register 2-4: CORCON: Core Control Register (PIC24F and PIC24H Devices)

U-0 U-0 U-0 U-0 U-0 U-0 U-0 U-0
———	————
bit 15							bit 8

U-0 U-0 U-0 U-0		R/C-0	R/W-0	U-0 U-0
———	IPL3			(1,2)	PSV — —
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15-4 Unimplemented: Read as '0'

bit 3 IPL3: Interrupt Priority Level 3 Status bit ^(1,2)

1 = CPU Interrupt Priority Level is 8 or greater (trap exception activated)
0 = CPU Interrupt Priority Level is 7 or less (no trap exception activated)

bit 2 PSV: Program Space Visibility in Data Space Enable bit

1 = Program space is visible in data space
0 = Program space is not visible in data space

bit 1-0 Unimplemented: Read as '0'

Note 1: This bit may be read or cleared, but not set.
2: This bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU Interrupt Priority Level.

Register 2-5: CORCON: Core Control Register (PIC24E Devices)

R/W-0 U-0 U-0 U-0 U-0 U-0 U-0
VAR	———	———
bit 15 bit 8

U-0 U-0 U-0 U-0		R/C-0	R-0 U-0 U-0
— — —	IPL3		(1,2)	SFA	— —
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15 VAR: Variable Exception Processing Latency Control bit

1 = Variable (bounded deterministic) exception processing latency

0 = Fixed (fully deterministic) exception processing latency

bit 14-4 Unimplemented: Read as '0'

bit 3 IPL3: CPU Interrupt Priority Level Status bit 3 ^(1,2)

1 = CPU Interrupt Priority Level is greater than 7

0 = CPU Interrupt Priority Level is 7 or less

bit 2 SFA: Stack Frame Active Status bit

1 = Stack frame is active; W14 and W15 address 0x0000 to 0xFFFF, regardless of DSRPAG and DSWPAG values

0 = Stack frame is not active; W14 and W15 address of EDS or Base Data Space

bit 1-0 Unimplemented: Read as '0'

Note 1: This bit may be read or cleared, but not set.

2: The IPL3 bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU Interrupt Priority Level.

Register 2-6: CORCON: Core Control Register (dsPIC30F and dsPIC33F Devices)

U-0 U-0 U-0 R/W-0 R(0)/W-0 R-0 R-0 R-0
———	U	S	E	D^(1)	DL^(2,3)	DL1^(2,3)	DL0^(3)
bit 15 bit 8

R/W-0	R/W-0	R/W-1	R/W-0	R/C-0	R/W-0	R/W-0	R/W-0
SATA	SATB	SATDW	ACCSAT	IPL3^(4,5)	PSV RND	IF
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15-13 Unimplemented: Read as '0'

bit 12 US: Unsigned or Signed Multiplier Mode Select bit

1 = Unsigned mode enabled for DSP multiply operations

0 = Signed mode enabled for DSP multiply operations

bit 11 EDT: Early DO Loop Termination Control bit ^(1)

1 = Terminates executing DO loop at the end of current iteration

0 = No effect

bit 10-8 DL<2:0>: DO Loop Nesting Level Status bits ^(2,3)

111 = DO looping is nested at 7 levels

110 = DO looping is nested at 6 levels

110 = DO looping is nested at 5 levels

110 = DO looping is nested at 4 levels

011 = DO looping is nested at 3 levels

010 = DO looping is nested at 2 levels

001 = DO looping is active, but not nested (just 1 level)

000 = DO looping is not active

bit 7 SATA: ACCA Saturation Enable bit

1 = Accumulator A saturation is enabled

0 = Accumulator A saturation is disabled

bit 6 SATB: ACCB Saturation Enable bit

1 = Accumulator B saturation is enabled

0 = Accumulator B saturation is disabled

bit 5 SATDW: Data Space Write from DSP Engine Saturation Enable bit

1 = Data space write saturation is enabled

0 = Data space write saturation is disabled

bit 4 ACCSAT: Accumulator Saturation Mode Select bit

1 = 9.31 saturation (super saturation)

0 = 1.31 saturation (normal saturation)

bit 3 IPL3: Interrupt Priority Level 3 Status bit ^(4,5)

1 = CPU Interrupt Priority Level is 8 or greater (trap exception is activated)

0 = CPU Interrupt Priority Level is 7 or less (no trap exception is activated)

Note 1: This bit will always read as '0'.

2: DL<2:1> bits are read-only.

3: The first two levels of DO loop nesting are handled by hardware.

4: This bit may be read or cleared, but not set.

5: This bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU Interrupt Priority Level.

Register 2-6: CORCON: Core Control Register (dsPIC30F and dsPIC33F Devices) (Continued)

bit 2 PSV: Program Space Visibility in Data Space Enable bit

1 = Program space is visible in data space
0 = Program space is not visible in data space

bit 1 RND: Rounding Mode Select bit

1 = Biased (conventional) rounding is enabled
0 = Unbiased (convergent) rounding is enabled

bit 0 IF: Integer or Fractional Multiplier Mode Select bit

1 = Integer mode enabled for DSP multiply operations
0 = Fractional mode enabled for DSP multiply operations

Note 1: This bit will always read as '0'.

2: DL<2:1> bits are read-only.
3: The first two levels of DO loop nesting are handled by hardware.
4: This bit may be read or cleared, but not set.
5: This bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU Interrupt Priority Level.

Register 2-7: CORCON: Core Control Register (dsPIC33E and dsPIC33C Devices)

R/W-0 U-0 R/W-0 R/W-0 R/W-0 R-0 R-0 R-0
VAR	— US1	US0 EDT		(1)	DL2 DL1	DL0
bit 15 bit 8

R/W-0	R/W-0	R/W-1	R/W-0	R/C-0	R-0	R/W-0	R/W-0
SATA	SATB	SATDW	ACCSAT	IPL3^(2,3)	SFA	RND IF
bit 7 bit 0

Legend:	C = Clearable bit
R = Readable bit	W = Writable bit	U = Unimplemented bit, read as ‘0’
-n = Value at POR	‘1’ = Bit is set	‘0’ = Bit is cleared	x = Bit is unknown

bit 15 VAR: Variable Exception Processing Latency Control bit 1 = Variable (bounded deterministic) exception processing latency 0 = Fixed (fully deterministic) exception processing latency bit 14 Unimplemented: Read as '0'

bit 13-12 US<1:0>: DSP Multiply Unsigned/Signed Control bits 11 = Reserved 10 = DSP engine multiplies are mixed-sign 01 = DSP engine multiplies are unsigned 00 = DSP engine multiplies are signed

bit 11 EDT: Early DO Loop Termination Control bit ^(1) 1 = Terminates executing DO loop at end of current loop iteration 0 = No effect

bit 10-8 DL<2:0>: DO Loop Nesting Level Status bits 111 = 7 DO loops are active . . . 001 = 1 DO loop is active 000 = 0 DO loops are active

bit 7 SATA: ACCA Saturation Enable bit 1 = Accumulator A saturation is enabled 0 = Accumulator A saturation is disabled

bit 6 SATB: ACCB Saturation Enable bit 1 = Accumulator B saturation is enabled 0 = Accumulator B saturation is disabled

bit 5 SATDW: Data Space Write from DSP Engine Saturation Enable bit 1 = Data space write saturation is enabled 0 = Data space write saturation is disabled

bit 4 ACCSAT: Accumulator Saturation Mode Select bit 1 = 9.31 saturation (super saturation) 0 = 1.31 saturation (normal saturation)

bit 3 IPL3: CPU Interrupt Priority Level Status bit 3 ^(2,3) 1 = CPU Interrupt Priority Level is greater than 7 0 = CPU Interrupt Priority Level is 7 or less

Note 1: This bit always reads as '0'. 2: This bit may be read or cleared, but not set. 3: The IPL3 bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU interrupt priority level.

Register 2-7: CORCON: Core Control Register (dsPIC33E and dsPIC33C Devices) (Continued)

bit 2 SFA: Stack Frame Active Status bit

1 = Stack frame is active; W14 and W15 address 0x0000 to 0xFFFF, regardless of DSRPAG and DSWPAG values 0 = Stack frame is not active; W14 and W15 address of EDS or Base Data Space

bit 1 RND: Rounding Mode Select bit

1 = Biased (conventional) rounding is enabled

0 = Unbiased (convergent) rounding is enabled

bit 0 IF: Integer or Fractional Multiplier Mode Select bit

1 = Integer mode is enabled for DSP multiply

0 = Fractional mode is enabled for DSP multiply

Note 1: This bit always reads as '0'.

2: This bit may be read or cleared, but not set.
3: The IPL3 bit is concatenated with the IPL<2:0> bits (SR<7:5>) to form the CPU interrupt priority level.

NOTES:

Section 3. Instruction Set Overview

HIGHLIGHTS

This section of the manual contains the following major topics:

3.1 Introduction 40
3.2 Instruction Set Overview 40
3.3 Instruction Set Summary Tables 42

3.1 INTRODUCTION

The 16-bit MCU and DSC instruction set provides a broad suite of instructions that support traditional microcontroller applications and a class of instructions that support math-intensive applications. Since almost all of the functionality of the 8-bit PIC ^® MCU instruction set has been maintained, this hybrid instruction set allows an easy 16-bit migration path for users already familiar with the PIC microcontroller.

3.2 INSTRUCTION SET OVERVIEW

Depending on the device family, the 16-bit MCU and DSC instruction set contains up to 105 instructions, which can be grouped into the functional categories shown in Table 3-1. Table 1-2 defines the symbols used in the instruction summary tables. Table 3-2 through Table 3-11 define the syntax, description, storage and execution requirements for each instruction. Storage requirements are represented in 24-bit instruction words and execution requirements are represented in instruction cycles.

Table 3-1: Instruction Groups

Functional Group Summary Table Page Number
Move Instructions Table 3-2 42
Math Instructions Table 3-3 43
Logic Instructions Table 3-4 45
Rotate/Shift Instructions Table 3-5 46
Bit Instructions	Table 3-6 47
Compare/Skip and Compare/Branch Instructions	Table 3-7	48
Program Flow Instructions	Table 3-8	49
Shadow/Stack Instructions	Table 3-9	51
Control Instructions	Table 3-10	51
DSP Instructions ^(1)	Table 3-11	52

Note 1: DSP instructions are only available in the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C device families.

Most instructions have several different addressing modes and execution flows, which require different instruction variants. For instance, depending on the device family, there are up to six unique ADD instructions and each instruction variant has its own instruction encoding. Instruction format descriptions and specific instruction operation are provided in Section 5. "Instruction Descriptions". Additionally, a composite alphabetized instruction set table is provided in Section 7. "Reference".

3.2.1 Multicycle Instructions

As shown in the instruction summary tables, most instructions execute in a single cycle with the following exceptions:

Note: The DO and DIVF instructions are only available in the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C device families.

• Instructions, DO, MOV.D, POP.D, PUSH.D, TBLRDH, TBLRDL, TBLWTH and TBLWTL, require two cycles to execute.
• Instructions, DIV.S, DIV.U and DIVE, are single-cycle instructions, which should be executed 18 consecutive times as the target of a REPEAT instruction.
• Instructions that change the Program Counter also require two cycles to execute, with the extra cycle executed as a NOP. Compare/Skip instructions, which skip over a two-word instruction, require three instruction cycles to execute, with two cycles executed as a NOP. Compare/Branch instructions (dsPIC33E/dsPIC33C/PIC24E devices only) require five instruction cycles to execute when the branch is taken.
• The RETFIE, RETLW and RETURN are a special case of an instruction that changes the Program Counter. These execute in three cycles, unless an exception is pending, and then they execute in two cycles.

Note 1: Instructions which access program memory as data, using Program Space Visibility (PSV), will incur a one or two-cycle delay for PIC24F, PIC24H, dsPIC30F and dsPIC33F devices, whereas using PSV in dsPIC33E and PIC24E devices incurs a four-cycle delay based on Flash memory access time. However, regardless of which device is being used, when the target instruction of a REPEAT loop accesses program memory as data, only the first execution of the target instruction is subject to the delay. See the specific device family reference manual for details.

2: All instructions may incur an additional delay on some device families depending on Flash memory access time. For example, PIC24E, dsPIC33E and dsPIC33C devices have a three-cycle Flash memory access time. However, instruction pipelining increases the effective instruction execution throughput. Refer to "CPU" in the "dsPIC33/PIC24 Family Reference Manual" for details on instruction timing.

3: All read and Read-Modify-Write (RMW) operations (including bit operations) on non-CPU Special Function Registers (e.g., I/O Port, Peripheral Control or STATUS Registers; interrupt flags, etc.) in PIC24E, dsPIC33E and dsPIC33C devices require two instruction cycles to execute. However, all write operations on both CPU and non-CPU Special Function Registers, and all read and Read-Modify-Write operations on CPU Special Function Registers, require one instruction cycle.

3.2.2 Multiword Instructions

As defined by Table 3-2, almost all instructions consume one instruction word (24 bits), with the exception of the CALL, DO and GOTO instructions, which are program flow Instructions listed in Table 3-8. These instructions require two words of memory because their opcodes embed large literal operands.

3.3 INSTRUCTION SET SUMMARY TABLES

Table 3-2: Move Instructions

Assembly Syntax Description Words Cycles					Page Number
EXCH Wns, Wnd Swap Wns and Wnd 1 1 254
LDSLV Wns, Wnd, #lit2 (5)		Move a single instruction word from Master to Slave PRAM	1	1	279
MOV f {,WREG} (1)		Move f to destination	1 1 299
MOV WREG, f		Move WREG to f	1	1	300
MOV f, Wnd		Move f to Wnd	1	1(4)	301
MOV Wns, f		Move Wns to f	1	1	302
MOV.B #lit8, Wnd		Move 8-bit literal to Wnd	1	1	303
MOV #lit16, Wnd		Move 16-bit literal to Wnd	1	1	304
MOV [Ws+Slit10], Wnd		Move [Ws with offset] to Wnd	1	1(4)	305
MOV Wns, [Wd+Slit10]		Move Wns to [Wd with offset]	1	1	306
MOV Ws, Wd		Move Ws to Wd	1	1(4)	307
MOV.D Wns, Wnd		Move double Wns to Wnd: Wnd + 1	1	2(4)	309
MOV.D Ws, Wnd		Move double Ws: Ws + 1 to Wnd	1	2(4)	309
MOVPAG #lit10, DSRPAG (2)		Move 10-bit literal to DSRPAG	1 1	311
MOVPAG #lit9, DSWPAG (2)		Move 9-bit literal to DSWPAG	1	1	311
MOVPAG #lit8, TBLPAG (2)		Move 8-bit literal to TBLPAG	1	1	311
MOVPAG Wn, DSRPAG (2)		Move Wn to DSRPAG	1 1 312
SWAP Wn		Wn = byte or nibble swap Wn	1	1	439
TBLRDH [Ws], Wd		Read high program word to Wd	1	2(3)	440
TBLRDL [Ws], Wd		Read low program word to Wd	1	2(3)	442
TBLWTH Ws, [Wd]		Write Ws to high program word	1	2(4)	444
TBLWTL Ws, [Wd]		Write Ws to low program word	1	2(4)	446
VFSLV Wns, Wnd, #lit2 (5)		Verify Slave processor program RAM	1 1 450

Note 1: When the optional {, WREG} operand is specified, the destination of the instruction is WREG. When {, WREG} is not specified, the destination of the instruction is the file register f.
2: The MOVPAG instruction is only available in PIC24E, dsPIC33E and dsPIC33C devices.
3: In dsPIC33E and PIC24E devices, and in dsPIC33C Master cores, these instructions require three additional cycles – compared to dsPIC30F, dsPIC33F, PIC24F and PIC24H devices and in dsPIC33C Slave cores.
4: In dsPIC33E, dsPIC33C and PIC24E devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to dsPIC30F, dsPIC33F, PIC24F and PIC24H devices.
5: These instructions are only available in dsPIC33C devices.

Table 3-3: Math Instructions

Assembly Syntax Description Words Cycles				Page Number
ADD f {,WREG} (1)	Destination = f + WREG 1 1		(5)	102
ADD #lit10,Wn	Wn = lit10 + Wn 1 1 103
ADD Wb,#lit5,Wd	Wd = Wb + lit5 1 1 104
ADD Wb,Ws,Wd	Wd = Wb + Ws 1 1		(5)	105
ADDC f {,WREG} (1)	Destination = f + WREG + (C)	1 1	(5)	110
ADDC #lit10,Wn	Wn = lit10 + Wn + (C)	1 1 11
ADDC Wb,#lit5,Wd	Wd = Wb + lit5 + (C)	1 1 112
ADDC Wb,Ws,Wd	Wd = Wb + Ws + (C)	1 1	(5)	114
DAW.B Wn	Wn = decimal adjust Wn	1 1 225
DEC f {,WREG} (1)	Destination = f - 1	1 1	(5)	226
DEC Ws,Wd	Wd = Ws - 1	1 1	(5)	227
DEC2 f {,WREG} (1)	Destination = f - 2	1 1	(5)	229
DEC2 Ws,Wd	Wd = Ws - 2	1 1	(5)	230
DIV.S Wm,Wn	Signed 16/16-bit integer divide, Q → W0, R → W1 1	18/6	(2)	233
DIV.U Wm,Wn	Unsigned 16/16-bit integer divide, Q - W0, R → W1	1	18/6(2)	235
DIVF Wm,Wn	Signed 16/16-bit fractional divide, Q - W0, R → W1	1	18/6(2)	236
DIVF2 Wm,Wn(6)	Signed 16/16-bit fractional divide (W1:W0 preserved)	1 6 238
DIV2.S Wm,Wn(6)	Signed 16/16-bit fractional divide (W1:W0 preserved)	1 6 240
DIV2.U Wm,Wn(6)	Unsigned 16/16-bit integer divide (W1:W0 preserved)	1 6 241
FLIM Wb,Ws(6)	Force data (upper and lower) range limit without limit excess result	1	1	261
FLIM.V Wb,Ws,Wnd(6)	Force data (upper and lower) range limit with limit excess result	1	1	262
INC f {,WREG} (1)	Destination = f + 1	1 1	(5)	267
INC Ws,Wd	Wd = Ws + 1	1 1	(5)	268
INC2 f {,WREG} (1)	Destination = f + 2	1 1	(5)	269
INC2 Ws,Wd	Wd = Ws + 2	1 1	(5)	270
MUL f	W3:W2 = f * WREG	1 1	(5)	323
MUL.SS Wb,Ws,Wnd	{Wnd + 1,Wnd} = signed(Wb) * signed(Ws)	1 1	(5)	325
MUL.SS Wb,Ws,Acc(4)	Accumulator = signed(Wb) * signed(Ws)	1 1	(5)	327
MUL.SU Wb,#lit5,Wnd	{Wnd + 1, Wind} = signed(Wb) * unsigned(lit5)	1 1 328

Note 1: When the optional {, WREG} operand is specified, the destination of the instruction is WREG. When {, WREG} is not specified, the destination of the instruction is the file register f.
2: In PIC24F, PIC24H, PIC24E, dsPIC30F, dsPIC33F and dsPIC33E devices, the divide instructions must be preceded with a "REPEAT #17" instruction, such that they are executed 18 consecutive times, thus taking 18 instruction cycles. In dsPIC33C devices, the divide instructions must be preceded with a "REPEAT #5" instruction, such that they are executed six consecutive times, thus taking six instruction cycles.
3: These instructions are only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: These instructions are only available in dsPIC33E and dsPIC33C devices.
5: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: These instructions are only available in dsPIC33C devices.

Table 3-3: Math Instructions (Continued)

Assembly Syntax	Description	Words	Cycles	Page Number
MUL.SU Wb,Ws,Wnd	{Wnd + 1, Wind} = signed(Wb) * unsigned(Ws) 1 1		(5)	329
MUL.SU Wb,Ws,Acc (4)	Accumulator = signed(Wb) * unsigned(Ws) 1 1		(5)	331
MUL.SU Wb,#lit5,Acc (4)	Accumulator = signed(Wb) * unsigned(lit5) 1 1 332
MUL.US Wb,Ws,Wnd	{Wnd + 1, Wind} = unsigned(Wb) * signed(Ws) 1 1		(5)	333
MUL.US Wb,Ws,Acc (4)	Accumulator = unsigned(Wb) * signed(Ws) 1 1		(5)	335
MUL.UU Wb,#lit5,Wnd	{Wnd + 1, Wind} = unsigned(Wb) * unsigned(lit5) 1 1 336
MUL.UU Wb,Ws,Wnd	{Wnd + 1, Wind} = unsigned(Wb) * unsigned(Ws) 1 1		(5)	337
MUL.UU Wb,Ws,Acc (4)	Accumulator = unsigned(Wb) * unsigned(Ws) 1 1		(5)	339
MUL.UU Wb,#lit5,Acc (4)	Accumulator = unsigned(Wb) * unsigned(lit5) 1 1 340
MULW.SS Wb,Ws,Wnd (3)	Wnd = signed(Wb) * signed(Ws) 1 1		(5)	341
MULW.SU Wb,Ws,Wnd (3)	Wnd = signed(Wb) * unsigned(Ws)	1 1	(5)	343
MULW.SU Wb,#lit5,Wnd (3)	Wnd = signed(Wb) * unsigned(lit5)	1 1 345
MULW.US Wb,Ws,Wnd (3)	Wnd = unsigned(Wb) * signed(Ws)	1 1	(5)	346
MULW.UU Wb,Ws,Wnd (3)	Wnd = unsigned(Wb) * unsigned(Ws)	1 1	(5)	348
MULW.UU Wb,#lit5,Wnd (3)	Wnd = unsigned(Wb) * unsigned(lit5)	1 1 349
SE Ws,Wnd	Wnd = sign-extended Ws	1 1	(5)	406
SUB f {,WREG} (1)	Destination = f - WREG	1 1	(5)	418
SUB #lit10,Wn	Wn = Wn - lit10	1 1 419
SUB Wb,#lit5,Wd	Wd = Wb - lit5	1 1 420
SUB Wb,Ws,Wd	Wd = Wb - Ws	1 1	(5)	421
SUBB f {,WREG} (1)	Destination = f - WREG - (C)	1	1(5)	424
SUBB #lit10,Wn	Wn = Wn - lit10 - (C)	1	1	425
SUBB Wb,#lit5,Wd	Wd = Wb - lit5 - (C)	1	1	426
SUBB Wb,Ws,Wd	Wd = Wb - Ws - (C)	1	1(5)	428
SUBBR f {,WREG} (1)	Destination = WREG - f - (C)	1	1(5)	430
SUBBR Wb,#lit5,Wd	Wd = lit5 - Wb - (C)	1	1	431
SUBBR Wb,Ws,Wd	Wd = Ws - Wb - (C)	1	1(5)	433
SUBR f {,WREG} (1)	Destination = WREG - f	1 1	(5)	435
SUBR Wb,#lit5,Wd	Wd = lit5 - Wb	1 1 436
SUBR Wb,Ws,Wd	Wd = Ws - Wb	1 1	(5)	437
ZE Ws,Wnd	Wnd = zero-extended Ws	1 1	(5)	456

Note 1: When the optional {, WREG} operand is specified, the destination of the instruction is WREG. When {, WREG} is not specified, the destination of the instruction is the file register f.
2: In PIC24F, PIC24H, PIC24E, dsPIC30F, dsPIC33F and dsPIC33E devices, the divide instructions must be preceded with a "REPEAT #17" instruction, such that they are executed 18 consecutive times, thus taking 18 instruction cycles. In dsPIC33C devices, the divide instructions must be preceded with a "REPEAT #5" instruction, such that they are executed six consecutive times, thus taking six instruction cycles.
3: These instructions are only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: These instructions are only available in dsPIC33E and dsPIC33C devices.
5: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: These instructions are only available in dsPIC33C devices.

Table 3-4: Logic Instructions

Assembly Syntax Description Words Cycles				Page Number
AND f {,WREG} (1)	Destination = f.AND.WREG 1 1		(2)	116
AND #lit10,Wn	Wn = lit10.AND.Wn	1	1	117
AND Wb,#lit5,Wd	Wd = Wb.AND.lit5	1	1	118
AND Wb,Ws,Wd	Wd = Wb.AND.Ws	1	1^(2)	119
CLR f	f = 0x0000	1	1	192
CLR WREG	WREG = 0x0000	1	1	192
CLR Wd	Wd = 0x0000	1	1	193
COM f {,WREG} (1)	Destination =	1	1^(2)	197
COM Ws,Wd Wd = Ws ————	1	1^(2)	198
IOR f {,WREG} (1)	Destination = f.IOR.WREG	1 1	(2)	271
IOR #lit10,Wn	Wn = lit10.IOR.Wn	1	1	272
IOR Wb,#lit5,Wd	Wd = Wb.IOR.lit5	1	1	273
IOR Wb,Ws,Wd	Wd = Wb.IOR.Ws	1	1^(2)	274
NEG f {,WREG} (1)	Destination = + 1	1 1	(2)	350
NEG Ws,Wd Wd = Ws —+ 1	1 1	(2)	351
SETM f	f = 0xFFFF	1	1	408
SETM WREG	WREG = 0xFFFF	1	1	409
SETM Wd	Wd = 0xFFFF	1	1	409
XOR f {,WREG} (1)	Destination = f.XOR.WREG	1 1	(2)	451
XOR #lit10,Wn	Wn = lit10.XOR.Wn	1	1	452
XOR Wb,#lit5,Wd	Wd = Wb.XOR.lit5	1	1	453
XOR Wb,Ws,Wd Wd = Wb .XOR. Ws	1 1	(2)	454

Note 1: When the optional {, WREG} operand is specified, the destination of the instruction is WREG. When {, WREG} is not specified, the destination of the instruction is the file register f.
2: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.

Table 3-5: Rotate/Shift Instructions

Assembly Syntax Description Words Cycles					Page Number
ASR f {,WREG} (1)		Destination = arithmetic right shift f, LSb → C	1	(2) 1	121
ASR	Ws,Wd	Wd = arithmetic right shift Ws, LSb → C	1	1^(2)	123
ASR	Wb,#lit4,Wnd	Wnd = arithmetic right shift Wb by lit4	1	1	125
ASR	Wb,Wns,Wnd	Wnd = arithmetic right shift Wb by Wns	1	1	126
LSR f {,WREG} (1)		Destination = logical right shift f, LSb → C	1	1^(2)	282
LSR	Ws,Wd	Wd = logical right shift Ws, LSb → C	1	1^(2)	284
LSR	Wb,#lit4,Wnd	Wnd = logical right shift Wb by lit4	1	1	286
LSR	Wb,Wns,Wnd	Wnd = logical right shift Wb by Wns	1	1	287
RLC f {,WREG} (1)		Destination = rotate left through Carry f	1 1	(2)	388
RLC	Ws,Wd	Wd = rotate left through Carry Ws	1	1^(2)	389
RLNC	f {,WREG} (1)	Destination = rotate left (no Carry) f	1	1^(2)	391
RLNC	Ws,Wd	Wd = rotate left (no Carry) Ws	1	1^(2)	392
RRC f {,WREG} (1)		Destination = rotate right through Carry f	1	1^(2)	394
RRC	Ws,Wd	Wd = rotate right through Carry Ws	1	1^(2)	396
RRNC	f {,WREG} (1)	Destination = rotate right (no Carry) f	1	1^(2)	398
RRNC	Ws,Wd	Wd = rotate right (no Carry) Ws	1	1^(2)	399
SL	f {,WREG} (1)	Destination = left shift f, MSb → C	1	1^(2)	412
SL	Ws,Wd	Wd = left shift Ws, MSb → C	1	1^(2)	414
SL	Wb,#lit4,Wnd	Wnd = left shift Wb by lit4	1	1	416
SL	Wb,Wns,Wnd	Wnd = left shift Wb by Wns	1	1	417

Note 1: When the optional {, WREG} operand is specified, the destination of the instruction is WREG. When {, WREG} is not specified, the destination of the instruction is the file register f.
2: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.

Table 3-6: Bit Instructions

Assembly Syntax Description		Words Cycles		(1)	Page Number
BCLR f,#bit4 Bit clear in f 1 1 127
BCLR Ws,#bit4 Bit clear in Ws 1 1 128
BFEXT #bit4,#wid5,Ws,Wb ^(2)		Bit field extract from Ws to Wb 2 2 130
BFEXT #bit4,#wid5,f,Wb ^(2)		Bit field extract from f to Wb 2 2 131
BFINS #bit4,#wid5,Wb,Ws ^(2)		Bit field insert from Wb into Ws	2 2 132
BFINS #bit4,#wid5,Wb,f ^(2)		Bit field insert from Wb into f	2 2 133
BFINS #bit4,#wid5,#lit8,Ws ^(2)		Bit field insert from #lit8 into Ws	2 2 134
BSET f,#bit4		Bit set in f	1	1	160
BSET Ws,#bit4		Bit set in Ws	1	1	161
BSW Ws,Wb		Write C bit to Ws11163	1	1	163
BTG f,#bit4		Bit toggle in f	1	1	165
BTG Ws,#bit4		Bit toggle in Ws	1	1	166
BTST f,#bit4		Bit test in f	1	1	175
BTST Ws,#bit4		Bit test in Ws	1	1	176
BTST Ws,Wb		Bit test in Ws	1	1	178
BTSTS f,#bit4		Bit test f to Z, then set f	1	1	180
BTSTS Ws,#bit4		Bit test Ws to C, then set Ws	1	1	181
FBCL Ws,Wnd		Find bit change from left (MSb) side	1	1	255
FF1L Ws,Wnd		Find first one from left (MSb) side	1	1	257
FF1R Ws,Wnd		Find first one from right (LSb) side	1	1	259

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to dsPIC30F, dsPIC33F, PIC24F and PIC24H devices.
2: These instructions are only available in dsPIC33C devices.

Table 3-7: Compare/Skip and Compare/Branch Instructions

Assembly Syntax Description Words Cycles				(1)	Page Number
BTSC f,#bit4 Bit test f,skip if clear 1 1 (2 or 3)				(5)	168
BTSC Ws,#bit4 Bit test Ws,skip if clear 1 1 (2 or 3)				(5)	170
BTSS f,#bit4 Bit test f,skip if set 1 1 (2 or 3)				(5)	172
BTSS Ws,#bit4 Bit test Ws,skip if set 1 1 (2 or 3)				(5)	173
CP f		Compare (f-WREG)	1	1^(5)	200
CP Wb,#lit5(2)		Compare (Wb-lit5)	1	1	201
CP Wb,#lit8(3)		Compare (Wb-lit8)	1	1	202
CP Wb,Ws		Compare (Wb-Ws)	1	1^(5)	203
CP0 f		Compare (f-0x0000)	1	1^(5)	204
CP0 Ws		Compare (Ws-0x0000)	1	1^(5)	205
CPB f		Compare with Borrow (f-WREG-C)	1	1^(5)	206
CPB Wb,#lit5(2)		Compare with Borrow (Wb-lit5-C)	1	1	207
CPB Wb,#lit8(3)		Compare with Borrow (Wb-lit8-C)	1	1	208
CPB Wb,Ws		Compare with Borrow (Wb-Ws-C)	1	1^(5)	209
CPBEQ Wb,Wn,Expr(3)		Compare Wb with Wn, branch if =	1	1(5)^(4)	211
CPBGT Wb,Wn,Expr(3)		Signed compare Wb with Wn, branch if >	1	1(5)^(4)	212
CPBLT Wb,Wn,Expr(3)		Signed compare Wb with Wn, branch if <	1	1(5)^(4)	213
CPBNE Wb,Wn,Expr(3)		Compare Wb with Wn, branch if ≠	1	1(5)^(4)	212
CPSEQ Wb,Wn(2)		Compare (Wb-Wn), skip if =	1	1 (2 or 3)	215
CPSEQ Wb,Wn(3)		Compare (Wb-Wn), skip if =	1	1 (2 or 3)	216
CPSGT Wb,Wn(2)		Signed compare (Wb-Wn), skip if >	1	1 (2 or 3)	217
CPSGT Wb,Wn(3)		Signed compare (Wb-Wn), skip if >	1	1 (2 or 3)	218
CPSLT Wb,Wn(2)		Signed compare (Wb-Wn), skip if <	1	1 (2 or 3)	219
CPSLT Wb,Wn(3)		Signed compare (Wb-Wn), skip if <	1	1 (2 or 3)	220
CPSNE Wb,Wn(2)		Signed compare (Wb-Wn), skip if ≠	1	1 (2 or 3)	221
CPSNE Wb,Wn(3)		Signed compare (Wb-Wn), skip if ≠	1	1 (2 or 3)	222

Note 1: Conditional skip instructions execute in one cycle if the skip is not taken, two cycles if the skip is taken over a one-word instruction and three cycles if the skip is taken over a two-word instruction.
2: This instruction is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
3: This instruction is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: Compare/Branch instructions in PIC24E/dsPIC33E devices and in dsPIC33C Master cores execute in one cycle if the branch is not taken, and five cycles if the branch is taken. Compare/Branch instructions in dsPIC33C Slave cores execute in one cycle if the branch is not taken and two cycles if the branch is taken.
5: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.

Table 3-8: Program Flow Instructions

Assembly Syntax Description Words Cycles				Page Number
BRA Expr Branch unconditionally 1 2			(8)	136
BRA Wn (5)	Computed branch 1 2		(8)	137
BRA Wn (4)	Computed branch 1 2		(8)	138
BRA C Expr Branch if Carry (no Borrow) 1 1 (2)			(1,8)	139
BRA GE Expr Branch if signed greater than or equal 1 1 (2)			(1,8)	141
BRA GEU Expr	Branch if unsigned greater than or equal	1	1(2)^(1,8)	142
BRA GT Expr	Branch if signed greater than	1	1(2)^(1,8)	143
BRA GTU Expr	Branch if unsigned greater than	1	1(2)^(1,8)	144
BRA LE Expr	Branch if signed less than or equal	1	1(2)^(1,8)	145
BRA LEU Expr	Branch if unsigned less than or equal	1	1(2)^(1,8)	146
BRA LT Expr	Branch if signed less than	1	1(2)^(1,8)	147
BRA LTU Expr	Branch if unsigned less than	1	1(2)^(1,8)	148
BRA N Expr	Branch if Negative	1	1(2)^(1,8)	149
BRA NC Expr	Branch if not Carry (Borrow)	1	1(2)^(1,8)	150
BRA NN Expr	Branch if not Negative	1	1(2)^(1,8)	151
BRA NOV Expr	Branch if not Overflow	1	1(2)^(1,8)	152
BRA NZ Expr Branch if not Zero 1 1 (2)			(1,8)	153
BRA OA Expr (3)	Branch if Accumulator A Overflow	1 1 (2)	(1,8)	154
BRA OB Expr (3)	Branch if Accumulator B Overflow	1 1 (2)	(1,8)	155
BRA OV Expr Branch if Overflow 1 1 (2)			(1,8)	156
BRA SA Expr (3)	Branch if Accumulator A Saturate	1 1 (2)	(1,8)	157
BRA SB Expr (3)	Branch if Accumulator B Saturate	1 1 (2)	(1,8)	158
BRA Z Expr	Branch if Zero	1	1(2)^(1,8)	159
CALL Expr(5)	Call subroutine	2 2	(8)	183
CALL Expr(4)	Call subroutine	2 2	(8)	185
CALL Wn(5)	Call indirect subroutine	1 2	(8)	187
CALL Wn(4)	Call indirect subroutine	1 2	(8)	189
CALL.L Wn (4)	Call indirect subroutine long (long address)	1	4	191
DO #lit14,Expr(6)	Do code through PC + Expr, (lit14 + 1) times	2	2	242
DO #lit15,Expr(7)	Do code through PC + Expr, (lit15 + 1) times	2	2	244
DO Wn,Expr(6)	Do code through PC + Expr, (Wn + 1) times	2	2	246
DO Wn,Expr(7)	Do code through PC + Expr, (Wn + 1) times	2	2	248

Note 1: Conditional branch instructions execute in one cycle if the branch is not taken or two cycles if the branch is taken.
2: RETURN instructions execute in three cycles, but if an exception is pending, they execute in two cycles.
3: This instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
4: This instruction is only available in PIC24E, dsPIC33E and dsPIC33C devices.
5: This instruction is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction is only available in dsPIC30F and dsPIC33F devices.
7: This instruction is only available in dsPIC33E and dsPIC33C devices.
8: In PIC24E and dsPIC33E devices, and in dsPIC33C Master cores, these instructions require two additional cycles (four cycles overall) when the branch is taken when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices, and dsPIC33C Slave cores.
9: In dsPIC33E and PIC24E devices, and in dsPIC33C Master cores, these instructions require three additional cycles when compared to dsPIC30F, dsPIC33F, PIC24F and PIC24H devices, and dsPIC33C Slave cores.

Table 3-8: Program Flow Instructions (Continued)

Assembly Syntax	Description	Words Cycles	Page Number
GOTO Expr Go to address 2	2		(8)
GOTO Wn (5)	Go to address indirectly 1 2		(8)
GOTO Wn (4)	Go to address indirectly 1 2		(8)
GOTO.L Wn (4)	Go to indirect (long address) 1 4 266
RCALL Expr (5)	Relative call 1 2		(8)
RCALL Expr (4)	Relative call 1 2		(8)
RCALL Wn (5)	Computed call 1 2		(8)
RCALL Wn (4)	Computed call 1 2		(8)
REPEAT #lit14 (5)	Repeat next instruction (lit14 + 1) times 1 1 375
REPEAT #lit15 (4)	Repeat next instruction (lit15 + 1) times 1 1 376
REPEAT Wn (5)	Repeat next instruction (Wn + 1) times	1 1 377
REPEAT Wn (4)	Repeat next instruction (Wn + 1) times	1 1 378
RETFIE^(5)	Return from interrupt enable	1	3 (2) ^(2,9)
RETFIE^(4)	Return from interrupt enable	1	3 (2) ^(2,9)
RETLW #lit10,Wn (5)	Return with lit10 in Wn	1	3 (2) ^(2,9)
RETLW #lit10,Wn (4)	Return with lit10 in Wn	1	3 (2) ^(2,9)
RETURN^(5)	Return from subroutine	1	3 (2) ^(2,9)
RETURN^(4)	Return from subroutine	1	3 (2) ^(2,9)

Note 1: Conditional branch instructions execute in one cycle if the branch is not taken or two cycles if the branch is taken.
2: RETURN instructions execute in three cycles, but if an exception is pending, they execute in two cycles.
3: This instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
4: This instruction is only available in PIC24E, dsPIC33E and dsPIC33C devices.
5: This instruction is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction is only available in dsPIC30F and dsPIC33F devices.
7: This instruction is only available in dsPIC33E and dsPIC33C devices.
8: In PIC24E and dsPIC33E devices, and in dsPIC33C Master cores, these instructions require two additional cycles (four cycles overall) when the branch is taken when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices, and dsPIC33C Slave cores.
9: In dsPIC33E and PIC24E devices, and in dsPIC33C Master cores, these instructions require three additional cycles when compared to dsPIC30F, dsPIC33F, PIC24F and PIC24H devices, and dsPIC33C Slave cores.

Table 3-9: Shadow/Stack/Context Instructions

Assembly Syntax Description Words Cycles				Page Number
BOOTSWP^(4)	Swap the active and inactive program Flash spaces 1 2 135
CTXTSWP #lit3^(2,3)	Switch CPU register context to context defined by #lit3 1 2 223
CTXTSWP Wn^(2,3)	Switch CPU register context to context defined by Wn 1 2 224
LNK #lit14^(5)	Link Frame Pointer 1 1 280
LNK #lit14^(6)	Link Frame Pointer 1 1 281
POP f	Pop TOS to f	1	1	357
POP Wd	Pop TOS to Wd	1	1	358
POP.D Wnd	Double pop from TOS to Wnd:Wnd + 1	1	2	359
POP.S	POP shadow registers	1	1	360
PUSH f	Push f to TOS	1	1^(1)	361
PUSH Ws	Push Ws to TOS	1	1^(1)	362
PUSH.D Wns	Push double Wns:Wns + 1 to TOS	1	2	364
PUSH.S	Push shadow registers	1	1	365
ULNK^(5)	Unlink Frame Pointer	1 1 448
ULNK^(6)	Unlink Frame Pointer	1 1 449

Note 1: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
2: These instructions are only available in dsPIC33C and some dsPIC33E devices. Please see the specific device data sheet for details.
3: In dsPIC33C devices, these instructions also switch the accumulator context in addition to the CPU register context.
4: These instructions are only available in some PIC24F, dsPIC33E and dsPIC33C devices. Please see the specific device data sheet for details.
5: These instructions are only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices. Please see the specific device data sheet for details.
6: These instructions are only available in PIC24E, dsPIC33E and dsPIC33C devices. Please see the specific device data sheet for details.

Table 3-10: Control Instructions

Assembly Syntax	Description	Words	Cycles	Page Number
CLRWDT	Clear Watchdog Timer	1	1	196
DISI #lit14	Disable interrupts for (lit14 + 1) instruction cycles	1	1	232
NOP	No operation	1	1	354
NOPR No operation	1 1 355
PWRSAV #lit1	Enter Power-Saving mode lit1	1	1	366
RESET	Software Device Reset	1	1	379

Table 3-11: DSP Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)

Assembly Syntax Description Words Cycles				Page Number
ADD Acc	Add accumulators 1 1 107
ADD Ws, #Slit4,Acc	16-bit signed add to accumulator 1 1		(1)	108
CLR Acc, [Wx], Wxd, [Wy], Wyd, AWB	Clear accumulator, prefetch operands 1	1	194
ED Wm*Wm, Acc, [Wx], [Wy], Wxd	Euclidean distance (no accumulate)	1 1 2	50
EDAC Wm*Wm, Acc, [Wx], [Wy], Wxd	Euclidean distance	1 1 2	52
LAC Ws, #Slit4,Acc	Load accumulator	1 1	(1)	276
LAC.D Wso, #Slit4,Acc(2)	Load accumulator double word	1 1 2	78
MAC Wm*Wn, Acc, [Wx], Wxd, [Wy], Wyd, AWB	Multiply and accumulate	1 1 2	88
MAC Wm*Wm, Acc, [Wx], Wxd, [Wy], Wyd	Square and accumulate	1 1 2	90
MAX Acc (2)	Force accumulator maximum data limit	1	1	292
MAX.V Acc, Wd(2)	Force accumulator maximum data limit and store limit excess result	1	1	293
MIN Acc (2)	Force accumulator minimum data limit 1	1	294
MIN.V Acc, Wd(2)	Force accumulator minimum data limit and store limit excess result	1	1	295
MINZ Acc (2)	Conditionally force accumulator minimum data limit if Z flag is set	1	1	296
MINZ.V Acc, Wd(2)	Conditionally force accumulator minimum data limit and store limit excess result if Z flag is set	1	1	297
MOVSAC Acc, [Wx], Wxd, [Wy], Wyd, AWB	Move Wx to Wxd and Wy to Wyd	1 1 3	13
MPY Wm*Wn, Acc, [Wx], Wxd, [Wy], Wyd	Multiply Wm by Wn to accumulator	1 1 3	15
MPY Wm*Wn, Acc, [Wx], Wxd, [Wy], Wyd	Square to accumulator	1 1 3	17
MPY.N Wm*Wn, Acc, [Wx], Wxd, [Wy], Wyd	(Multiply -Wm by Wn) to accumulator	1 1 3	19
MSC Wm*Wn, Acc, [Wx], Wxd, [Wy], Wyd, AWB	Multiply and subtract from accumulator	1	1	321
NEG Acc	Negate accumulator	1 1 3	50
NORM Acc, Wd (2)	Normalize accumulator	1 1 3	56
SAC Acc, #Slit4, Wd	Store accumulator	1 1 4	01
SAC.D Acc, #Slit4, Wnd(2)	Store accumulator double word	1 1 4	03
SAC.R Acc, #Slit4, Wd	Store rounded accumulator	1 1 4	04
SFTAC Acc, #Slit6	Arithmetic shift accumulator by Slit6	1 1 4	10
SFTAC Acc, Wb	Arithmetic shift accumulator by (Wb)	1	1	411
SUB Acc	Subtract accumulators	1 1 4	18

Note 1: In PIC24E, dsPIC33E and dsPIC33C devices, read and Read-Modify-Write operations on non-CPU Special Function Registers require an additional cycle when compared to PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
2: These instructions are only available in dsPIC33C devices.

Section 4. Instruction Set Details

HIGHLIGHTS

This section of the manual contains the following major topics:

4.1 Data Addressing Modes....54
4.2 Program Addressing Modes 63
4.3 Instruction Stalls....64
4.4 Byte Operations 66
4.5 Word Move Operations 68
4.6 Using 10-Bit Literal Operands....71
4.8 Software Stack Pointer and Frame Pointer....72
4.9 Conditional Branch Instructions 78
4.10 Z Status Bit....79
4.11 Assigned Working Register Usage 80
4.12 DSP Data Formats (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)....83
4.13 Accumulator Usage (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)....85
4.14 Accumulator Access (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices) ..... 86
4.15 DSP MAC Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)..... 86
4.16 DSP Accumulator Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices).....90
4.17 Scaling Data with the FBCL Instruction (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)....90
4.19 Normalizing the Accumulator with the FBCL Instruction (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)....93
4.21 Extended Precision Arithmetic Using Mixed-Sign Multiplications (dsPIC33E and dsPIC33C Only) 94

4.1 DATA ADDRESSING MODES

The 16-bit MCU and DSC devices support three native addressing modes for accessing data memory, along with several forms of Immediate Addressing. Data accesses may be performed using File Register Addressing, Register Direct or Indirect Addressing, and Immediate Addressing, allowing a fixed value to be used by the instruction.

File Register Addressing provides the ability to operate on data stored in the lower 8K of data memory (Near RAM), and also move data between the Working registers and the entire 64K data space. Register Direct Addressing is used to access the 16 memory-mapped Working registers, W0:W15. Register Indirect Addressing is used to efficiently operate on data stored in the entire 64K data space (and also Extended Data Space in the case of dsPIC33E/dsPIC33C/PIC24E and some PIC24F devices), using the contents of the Working registers as an Effective Address (EA). Immediate Addressing does not access data memory, but provides the ability to use a constant value as an instruction operand. The address range of each mode is summarized in Table 4-1.

Table 4-1: 16-Bit MCU and DSC Addressing Modes

Addressing Mode Address Range
File Register 0x0000-0x1FFF	(1)
Register Direct 0x0000-0x001F (Working register array, W0:W15)
Register Indirect 0x0000-0xFFFF
Immediate N/A (constant value)

Note 1: The address range for the File Register MOV is 0x0000-0xFFFE.

4.1.1 File Register Addressing

File Register Addressing is used by instructions which use a predetermined data address as an operand for the instruction. The majority of instructions that support File Register Addressing provide access to the lower 8 Kbytes of data memory, which is called the Near RAM. However, the MOV instruction provides access to all 64 Kbytes of memory using File Register Addressing. This allows the loading of the data from any location in data memory to any Working register and storing the contents of any Working register to any location in data memory. It should be noted that File Register Addressing supports both byte and word accesses of data memory, with the exception of the MOV instruction, which accesses all 64K of memory as words. Examples of File Register Addressing are shown in Example 4-1.

Most instructions which support File Register Addressing perform an operation on the specified file register and the default Working register, WREG (see Section 2.4 "Default Working Register (WREG)"). If only one operand is supplied in the instruction, WREG is an implied operand and the operation results are stored back to the file register. In these cases, the instruction is effectively a Read-Modify-Write instruction. However, when both the file register and the WREG register are specified in the instruction, the operation results are stored in the WREG register and the contents of the file register are unchanged. Sample instructions that show the interaction between the file register and the WREG register are shown in Example 4-2.

Note: Instructions which support File Register Addressing use 'f' as an operand in the instruction summary tables of Section 3. "Instruction Set Overview".

Example 4-1: File Register Addressing

DEC 0x1000 ; decrement data stored at 0x1000
Before Instruction:
Data Memory 0x1000 = 0x5555
After Instruction:
Data Memory 0x1000 = 0x5554
MOV 0x27FE, W0 ; move data stored at 0x27FE to W0
Before Instruction:
W0 = 0x5555
Data Memory 0x27FE = 0x1234
After Instruction:
W0 = 0x1234
Data Memory 0x27FE = 0x1234 

Example 4-2: File Register Addressing and WREG

AND 0x1000 ; AND 0x1000 with WREG, store to 0x1000
Before Instruction:
    WO (WREG) = 0x332C
    Data Memory 0x1000 = 0x5555
After Instruction:
    WO (WREG) = 0x332C
    Data Memory 0x1000 = 0x1104
    AND 0x1000, WREG ; AND 0x1000 with WREG, store to WREG
Before Instruction:
    WO (WREG) = 0x332C
    Data Memory 0x1000 = 0x5555
After Instruction:
    WO (WREG) = 0x1104
    Data Memory 0x1000 = 0x5555 

4.1.2 Register Direct Addressing

Register Direct Addressing is used to access the contents of the 16 Working registers (W0:W15). The Register Direct Addressing mode is fully orthogonal, which allows any Working register to be specified for any instruction that uses Register Direct Addressing, and it supports both byte and word accesses. Instructions which employ Register Direct Addressing use the contents of the specified Working register as data to execute the instruction; therefore, this addressing mode is useful only when data already resides in the Working register core. Sample instructions which utilize Register Direct Addressing are shown in Example 4-3.

Another feature of Register Direct Addressing is that it provides the ability for dynamic flow control. Since variants of the DO and REPEAT instruction support Register Direct Addressing, flexible looping constructs may be generated using these instructions.

Note: Instructions which must use Register Direct Addressing, use the symbols Wb, Wn, Wns and Wnd in the summary tables of Section 3. "Instruction Set Overview". Commonly, Register Direct Addressing may also be used when Register Indirect Addressing may be used. Instructions which use Register Indirect Addressing, use the symbols Wd and Ws in the summary tables of Section 3. "Instruction Set Overview".

Example 4-3: Register Direct Addressing
EXCH W2, W3 ; Exchange W2 and W3 Before Instruction: W2 = 0x3499 W3 = 0x003D After Instruction: W2 = 0x003D W3 = 0x3499 IOR #0x44, W0 ; Inclusive-OR 0x44 and W0 Before Instruction: W0 = 0x9C2E After Instruction: W0 = 0x9C6E SL W6, W7, W8 ; Shift left W6 by W7, and store to W8 Before Instruction: W6 = 0x000C W7 = 0x0008 W8 = 0x1234 After Instruction: W6 = 0x000C W7 = 0x0008 W8 = 0x0C00

4.1.3 Register Indirect Addressing

Register Indirect Addressing is used to access any location in data memory by treating the contents of a Working register as an Effective Address (EA) to data memory. Essentially, the contents of the Working register become a pointer to the location in data memory which is to be accessed by the instruction.

This addressing mode is powerful, because it also allows one to modify the contents of the Working register, either before or after the data access is made, by incrementing or decrementing the EA. By modifying the EA in the same cycle that an operation is being performed, Register Indirect Addressing allows for the efficient processing of data that is stored sequentially in memory. The modes of Indirect Addressing supported by the 16-bit MCU and DSC devices are shown in Table 4-2.

Table 4-2: Indirect Addressing Modes

Indirect Mode Syntax		Function (Byte Instruction)	Function (Word Instruction)	Description
No Modification	[Wn]	EA = [Wn]	EA = [Wn]	The contents of Wn form the EA.
Pre-Increment	[++Wn]	EA = [Wn + = 1]	EA = [Wn + = 2]	Wn is pre-incremented to form the EA.
Pre-Decrement	[--Wn]	EA = [Wn - = 1]	EA = [Wn - = 2]	Wn is pre-decremented to form the EA.
Post-Increment	[Wn++]	EA = [Wn] + = 1	EA = [Wn] + = 2	The contents of Wn form the EA, then Wn is post-incremented.
Post-Decrement	[Wn--]	EA = [Wn] - = 1	EA = [Wn] - = 2	The contents of Wn form the EA, then Wn is post-decremented.
Register Offset	[Wn+Wb]	EA = [Wn + Wb]	EA = [Wn + Wb]	The sum of Wn and Wb forms the EA. Wn and Wb are not modified.

Table 4-2 shows that four addressing modes modify the EA used in the instruction, and this allows the following updates to be made to the Working register: post-increment, post-decrement, pre-increment and pre-decrement. Since all EAs must be given as byte addresses, support is provided for Word mode instructions by scaling the EA update by two. Namely, in Word mode, pre/post-decrements subtract two from the EA stored in the Working register and pre/post-increments add two to the EA. This feature ensures that after an EA modification is made, the EA will point to the next adjacent word in memory. Example 4-4 shows how Indirect Addressing may be used to update the EA.

Table 4-2 also shows that the Register Offset mode addresses data which is offset from a base EA stored in a Working register. This mode uses the contents of a second Working register to form the EA by adding the two specified Working registers. This mode does not scale for Word mode instructions, but offers the complete offset range of 64 Kbytes. Note that neither of the Working registers used to form the EA is modified. Example 4-5 shows how Register Offset Indirect Addressing may be used to access data memory.

Note: The MOV with offset instructions (see pages 299 and 300) provides a literal addressing offset ability to be used with Indirect Addressing. In these instructions, the EA is formed by adding the contents of a Working register to a signed 10-bit literal. Example 4-6 shows how these instructions may be used to move data to and from the Working register array.

Example 4-4: Indirect Addressing with Effective Address Update
MOV.B [W0++], [W13--] ; byte move [W0] to [W13] ; post-inc W0, post-dec W13

Before Instruction: W0 = 0x2300 W13 = 0x2708 Data Memory 0x2300 = 0x7783 Data Memory 0x2708 = 0x904E

After Instruction: W0 = 0x2301 W13 = 0x2707 Data Memory 0x2300 = 0x7783 Data Memory 0x2708 = 0x9083

ADD W1, [--W5], (++W8] ; pre-dec W5, pre-inc W8 ; add W1 to [W5], store in [W8]

Before Instruction: W1 = 0x0800 W5 = 0x2200 W8 = 0x2400 Data Memory 0x21FE = 0x7783 Data Memory 0x2402 = 0xAACC

After Instruction: W1 = 0x0800 W5 = 0x21FE W8 = 0x2402 Data Memory 0x21FE = 0x7783 Data Memory 0x2402 = 0x7F83

Example 4-5: Indirect Addressing with Register Offset

MOV.B [W0+W1], [W7++] ; byte move [W0+W1] to W7, post-inc W7
Before Instruction:
W0 = 0x2300
W1 = 0x01FE
W7 = 0x1000
Data Memory 0x24FE = 0x7783
Data Memory 0x1000 = 0x11DC
After Instruction:
W0 = 0x2300
W1 = 0x01FE
W7 = 0x1001
Data Memory 0x24FE = 0x7783
Data Memory 0x1000 = 0x1183
LAC [W0+W8], A ; load ACCA with [W0+W8]
; (sign-extend and zero-backfill)
Before Instruction:
W0 = 0x2344
W8 = 0x0008
ACCA = 0x00 7877 9321
Data Memory 0x234C = 0xE290
After Instruction:
W0 = 0x2344
W8 = 0x0008
ACCA = 0xFF E290 0000
Data Memory 0x234C = 0xE290 

Example 4-6: Move with Literal Offset Instructions

MOV [W0+0x20], W1 ; move [W0+0x20] to W1
Before Instruction:
W0 = 0x1200
W1 = 0x01FE
Data Memory 0x1220 = 0xFD27
After Instruction:
W0 = 0x1200
W1 = 0xFD27
Data Memory 0x1220 = 0xFD27
MOV W4, [W8-0x300] ; move W4 to [W8-0x300]
Before Instruction:
W4 = 0x3411
W8 = 0x2944
Data Memory 0x2644 = 0xCB98
After Instruction:
W4 = 0x3411
W8 = 0x2944
Data Memory 0x2644 = 0x3411 

4.1.3.1 REGISTER INDIRECT ADDRESSING AND THE INSTRUCTION SET

The addressing modes presented in Table 4-2 demonstrate the Indirect Addressing mode capability of the 16-bit MCU and DSC devices. Due to operation encoding and functional considerations, not every instruction which supports Indirect Addressing supports all modes shown in Table 4-2. The majority of instructions which use Indirect Addressing support the No Modify, Pre-Increment, Pre-Decrement, Post-Increment and Post-Decrement Addressing modes. The MOV instructions, and several accumulator-based DSP instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices only), are also capable of using the Register Offset Addressing mode.

Note: Instructions which use Register Indirect Addressing use the operand symbols, Wd and Ws, in the summary tables of Section 3. "Instruction Set Overview".

4.1.3.2 DSP MAC INDIRECT ADDRESSING MODES (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

A special class of Indirect Addressing modes is utilized by the DSP MAC instructions. As is described later in Section 4.15 "DSP MAC Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)", the DSP MAC class of instructions is capable of performing two fetches from memory using Effective Addressing. Since DSP algorithms frequently demand a broader range of address updates, the addressing modes offered by the DSP MAC instructions provide greater range in the size of the Effective Address update which may be made. Table 4-3 shows that both X and Y prefetches support Post-Increment and Post-Decrement Addressing modes, with updates of two, four and six bytes. Since DSP instructions only execute in Word mode, no provisions are made for odd-sized EA updates.

Table 4-3: DSP MAC Indirect Addressing Modes

Addressing Mode X Memory	Y Memory
Indirect with No Modification EA = [Wx] EA = [Wy]
Indirect with Post-Increment by two EA = [Wx] + = 2 EA = [Wy] + = 2
Indirect with Post-Increment by four	EA = [Wx] + = 4	EA = [Wy] + = 4
Indirect with Post-Increment by six	EA = [Wx] + = 6	EA = [Wy] + = 6
Indirect with Post-Decrement by two	EA = [Wx] - = 2	EA = [Wy] - = 2
Indirect with Post-Decrement by four	EA = [Wx] - = 4	EA = [Wy] - = 4
Indirect with Post-Decrement by six	EA = [Wx] - = 6	EA = [Wy] - = 6
Indirect with Register Offset	EA = [W9 + W12]	EA = [W11 + W12]

Note: As described in Section 4.15 "DSP MAC Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)", only W8 and W9 may be used to access X memory, and only W10 and W11 may be used to access Y memory.

4.1.3.3 MODULO AND BIT-REVERSED ADDRESSING MODES (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The 16-bit DSC architecture provides support for two special Register Indirect Addressing modes, which are commonly used to implement DSP algorithms. Modulo (or circular) Addressing provides an automated means to support circular data buffers in X and/or Y memory. Modulo buffers remove the need for software to perform address boundary checks, which can improve the performance of certain algorithms. Similarly, Bit-Reversed Addressing allows one to access the elements of a buffer in a nonlinear fashion. This addressing mode simplifies data re-ordering for radix-2 FFT algorithms and provides a significant reduction in FFT processing time.

Both of these addressing modes are powerful features of the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C architectures, which can be exploited by any instruction that uses Indirect Addressing. Refer to the specific device family reference manual for details on using Modulo and Bit-Reversed Addressing.

4.1.4 Immediate Addressing

In Immediate Addressing, the instruction encoding contains a predefined constant operand, which is used by the instruction. This addressing mode may be used independently, but it is more frequently combined with the File Register, Direct and Indirect Addressing modes. The size of the immediate operand which may be used varies with the instruction type. Constants of size 1-bit (#lit1), 4-bit (#bit4, #lit4 and #Slit4), 5-bit (#lit5), 6-bit (#Slit6), 8-bit (#lit8), 10-bit (#lit10 and #Slit10), 14-bit (#lit14) and 16-bit (#lit16) may be used. Constants may be signed or unsigned and the symbols, #Slit4, #Slit6 and #Slit10, designate a signed constant. All other immediate constants are unsigned. Table 4-4 shows the usage of each immediate operand in the instruction set.

Note:

The 6-bit (#Slit6) operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

Table 4-4: Immediate Operands in the Instruction Set

Operand Instruction Usage

#lit1 PWRSAV

#lit3 CTXTSWP (3)

#bit4 BCLR, BSET, BTG, BTSC, BTSS, BTST, BTST.C, BTST.Z, BTSTS, BTSTS.C, BTSTS.Z

#lit4 ASR, LSR, SL

#Slit4 ADD, LAC, SAC, SAC.R

#wid4 BFEXT, BFINS (6)

#lit5 ADD, ADDC, AND, CP (5), CPB(5), IOR, MUL.SU, MUL.UU, SUB, SUBB, SUBBR, SUBR, XOR

#Slit6(1) SFTAC

#lit8 MOV.B, CP (4), CPB(4)

#lit10 ADD, ADDC, AND, CP, CPB, IOR, RETLW, SUB, SUBB, XOR

#Slit10 MOV

#lit14 DISI, DO (2), LNK, REPEAT(5)

#lit15 DO (3), REPEAT(4)

#lit16 MOV

Note 1: This operand or instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
2: This operand or instruction is only available in dsPIC30F and dsPIC33F devices.
3: This operand or instruction is only available in dsPIC33E and dsPIC33C devices.
4: This operand or instruction is only available in dsPIC33E, dsPIC33C and PIC24E devices.
5: This operand or instruction is only available in dsPIC30F, dsPIC33F, PIC24F and PIC24H devices.
6: This operand or instruction is only available in dsPIC33C devices.

The syntax for Immediate Addressing requires that the number sign (#) must immediately precede the constant operand value. The “#” symbol indicates to the assembler that the quantity is a constant. If an out-of-range constant is used with an instruction, the assembler will generate an error. Several examples of Immediate Addressing are shown in Example 4-7.

Example 4-7: Immediate Addressing

PWRSAV #1 ; Enter IDLE mode
ADD.B #0x10, W0 ; Add 0x10 to W0 (byte mode)
Before Instruction:
W0 = 0x12A9
After Instruction:
W0 = 0x12B9
    XOR W0, #1, [W1++] ; Exclusive-OR W0 and 0x1
    ; Store the result to [W1]
    ; Post-increment W1
Before Instruction:
W0 = 0xFFFF
W1 = 0x0890
Data Memory 0x0890 = 0x0032
After Instruction:
W0 = 0xFFFF
W1 = 0x0892
Data Memory 0x0890 = 0xFFFE 

4.1.5 Data Addressing Mode Tree

The Data Addressing modes of the PIC24F, PIC24H and PIC24E families are summarized in Figure 4-1.

Figure 4-1: Data Addressing Mode Tree (PIC24F, PIC24H, PIC24E)

graph TD
    A["Data Addressing Modes"] --> B["Immediate"]
    A --> C["File Register"]
    A --> D["Direct"]
    A --> E["Indirect"]
    A --> F["No Modification"]
    F --> G["Pre-Increment"]
    F --> H["Pre-Decrement"]
    F --> I["Post-Increment"]
    F --> J["Post-Decrement"]
    F --> K["Literal Offset"]
    F --> L["Register Offset"]

The Data Addressing modes of the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C are summarized in Figure 4-2.

Figure 4-2: Data Addressing Mode Tree (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

graph TD
    A["Data Addressing Modes"] --> B["Basic"]
    A --> C["DSP MAC"]
    B --> D["Immediate"]
    B --> E["File Register"]
    B --> F["Direct"]
    B --> G["Indirect"]
    C --> H["Direct"]
    C --> I["Indirect"]
    H --> J["No Modification"]
    H --> K["Pre-Increment"]
    H --> L["Pre-Decrement"]
    H --> M["Post-Increment"]
    H --> N["Post-Decrement"]
    H --> O["Literal Offset"]
    H --> P["Register Offset"]
    I --> Q["No Modification"]
    I --> R["Post-Increment (2, 4 and 6)"]
    I --> S["Post-Decrement (2, 4 and 6)"]
    I --> T["Register Offset"]

4.2 PROGRAM ADDRESSING MODES

The 16-bit MCU and DSC devices have a 24-bit Program Counter (PC). The PC addresses the 24-bit wide program memory to fetch instructions for execution and it may be loaded in several ways. For byte compatibility with the table read and table write instructions, each instruction word consumes two locations in program memory. This means that during serial execution, the PC is loaded with PC + 2.

Several methods may be used to modify the PC in a non-sequential manner, and both absolute and relative changes may be made to the PC. The change to the PC may be from an immediate value encoded in the instruction or a dynamic value contained in a Working register. In dsPIC30F, dsPIC33F and dsPIC33E devices, when DO looping is active, the PC is loaded with the address stored in the DOSTART register after the instruction at the DOEND address is executed. For exception handling, the PC is loaded with the address of the exception handler, which is stored in the Interrupt Vector Table (IVT). When required, the software stack is used to return scope to the foreground process from where the change in program flow occurred.

Table 4-5 summarizes the instructions which modify the PC. When performing function calls, it is recommended that RCALL be used instead of CALL, since RCALL only consumes one word of program memory.

Table 4-5: Methods of Modifying Program Flow

Condition/Instruction PC Modification Software Stack Usage
Sequential Execution PC = PC + 2 None
BRA Expr^(1) (Branch Unconditionally)	PC = PC + 2 * Slit16 None
BRA Condition, Expr^(1) (Branch Conditionally)	PC = PC + 2 (condition false)PC = PC + 2 * Slit16 (condition true)	None
CALL Expr^(1) (Call Subroutine)	PC = lit23 PC + 4 is PUSHed on the stack	(2)
CALL Wn(Call Subroutine Indirect)	PC = Wn PC + 2 is PUSHed on the stack	(2)
CALL.L Wn^(5) (Call Indirect Subroutine Long)	PC = {Wn+1:Wn} PC + 2 is PUSHed on the stack	(2)
GOTO Expr^(1) (Unconditional Jump)	PC = lit23 None
GOTO Wn(Unconditional Indirect Jump)	PC = Wn None
GOTO.L Wn^(5) (Unconditional Indirect Long Jump)	PC = {Wn+1:Wn} None
RCALL Expr^(1) (Relative Call)	PC = PC + 2 * Slit16 PC + 2 is PUSHed on the stack	(2)
RCALL Wn(Computed Relative Call)	PC = PC + 2 * Wn	PC + 2 is PUSHed on the stack^(2)
Exception Handling	PC = Address of the exception handler (read from vector table)	PC + 2 is PUSHed on the stack^(3)
PC = Target REPEAT instruction(REPEAT Looping)	PC not modified (if REPEAT active)	None
PC = DOEND address^(4) (DO Looping)	PC = DOSTART (if DO active)	None

Note 1: For BRA, CALL and GOTO, the Expr may be a label, absolute address or expression, which is resolved by the linker to a 16-bit or 23-bit value (Slit16 or lit23). When representing an address offset value, Expr can also be indicated by using a “.” and a sign, “+” or “-”. For example, the expression, “.+2”, means an address offset of +2 (i.e., the next instruction address relative to the current position of the Program Counter). See Section 5. “Instruction Descriptions” for details.

2: After CALL or RCALL is executed, RETURN or RETLW will POP the Top-of-Stack (TOS) back into the PC.

3: After an exception is processed, RETFIE will POP the Top-of-Stack (TOS) back into the PC.

4: This condition/instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

5: This condition instruction is only available in dsPIC33E, dsPIC33C and PIC24E devices.

4.3 INSTRUCTION STALLS

In order to maximize the data space EA calculation and operand fetch time, the X data space read and write accesses are partially pipelined. A consequence of this pipelining is that address register data dependencies may arise between successive read and write operations using common registers.

'Read-After-Write' (RAW) dependencies occur across instruction boundaries and are detected by the hardware. An example of a RAW dependency would be a write operation that modifies W5, followed by a read operation that uses W5 as an Address Pointer. The contents of W5 will not be valid for the read operation until the earlier write completes. This problem is resolved by stalling the instruction execution for one instruction cycle, which allows the write to complete before the next read is started.

4.3.1 RAW Dependency Detection

During the instruction predecode, the core determines if any address register dependency is imminent across an instruction boundary. The Stall detection logic compares the W register (if any) used for the destination EA of the instruction currently being executed with the W register to be used by the source EA (if any) of the prefetched instruction. When a match between the destination and source registers is identified, a set of rules is applied to decide whether or not to stall the instruction by one cycle. Table 4-6 lists various RAW conditions which cause an instruction execution Stall.

Table 4-6: Raw Dependency Rules (Detection By Hardware)

Destination Addressing Mode Using Wn	Source Addressing Mode Using Wn	Stall Required?	Examples(2)(Wn = W2)
Direct Direct No Stall ADD.W W0, W1, W2			MOV.W W2, W3
Indirect Direct No Stall ADD.W W0, W1, [W2]			MOV.W W2, W3
Indirect	Indirect	No	StallMOV.W [W2], W3
Indirect Indirect with Pre/Post-Modification		No Stall ADD.W W0, W1, [W2]MOV.W [W2++, W3
Indirect with Pre/Post-Modification	Direct No Stall ADD.W	W0, W1, [W2++]MOV.W W2, W3
Direct	Indirect	Stall^(1)	ADD.W W0, W1, W2MOV.W [W2], W3
Direct Indirect with Pre/Post-Modification		Stall^(1)	ADD.W W0, W1, W2MOV.W [W2++, W3
Indirect	Indirect	Stall	ADD.W W0, W1, [W2] (2)MOV.W [W2], W3 (2)
Indirect Indirect with Pre/Post-Modification		Stall^(1)	ADD.W W0, W1, [W2] (2)MOV.W [W2++, W3 (2)
Indirect with Pre/Post-Modification	Indirect	Stall^(1)	ADD.W W0, W1, [W2++]MOV.W [W2], W3
Indirect with Pre/Post-Modification	Indirect with Pre/Post-Modification	Stall^(1)	ADD.W W0, W1, [W2++]MOV.W [W2++, W3

Note 1: When Stalls are detected, one cycle is added to the instruction execution time.
2: For these examples, the contents of W2 = the mapped address of W2 (0x0004).

Note: When Register Indirect with Offset Addressing is used to specify the destination for an instruction, and Ws is the same register as Wd, the old value of Ws is used for Wd (i.e., the address offset is ignored).

ADD.W W0, W1, [

4.3.2 Instruction Stalls and Exceptions

In order to maintain deterministic operation, instruction Stalls are allowed to happen, even if they occur immediately prior to exception processing.

4.3.3 Instruction Stalls and Instructions that Change Program Flow

CALL and RCALL write to the stack using W15 and may, therefore, be subject to an instruction Stall if the source read of the subsequent instruction uses W15.

GOTO, RETFIE and RETURN instructions are never subject to an instruction Stall because they do not perform write operations to the Working registers.

4.3.4 Instruction Stalls and DO/REPEAT Loops

Instructions operating in a DO or REPEAT loop are subject to instruction Stalls, just like any other instruction. Stalls may occur on loop entry, loop exit and also during loop processing.

Note: DO loops are only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

4.3.5 Instruction Stalls and PSV

Instructions operating in PSV address space are subject to instruction Stalls, just like any other instruction. Should a data dependency be detected in the instruction immediately following the PSV data access, the second cycle of the instruction will initiate a Stall. Should a data dependency be detected in the instruction immediately before the PSV data access, the last cycle of the previous instruction will initiate a Stall.

Note: Refer to the specific device family reference manual for more detailed information about RAW instruction Stalls.

4.4 BYTE OPERATIONS

Since the data memory is byte-addressable, most of the base instructions may operate in either Byte mode or Word mode. When these instructions operate in Byte mode, the following rules apply:

• All direct Working register references use the Least Significant Byte of the 16-bit Working register and leave the Most Significant Byte (MSB) unchanged
• All indirect Working register references use the data byte specified by the 16-bit address stored in the Working register
• All file register references use the data byte specified by the byte address
• The STATUS Register is updated to reflect the result of the byte operation

It should be noted that data addresses are always represented as byte addresses. Additionally, the native data format is little-endian, which means that words are stored with the Least Significant Byte at the lower address and the Most Significant Byte at the adjacent, higher address (as shown in Figure 4-3). Example 4-8 shows sample byte move operations and Example 4-9 shows sample byte math operations.

Note: Instructions that operate in Byte mode must use the “.b” or “.B” instruction extension to specify a byte instruction. For example, the following two instructions are valid forms of a byte clear operation:

- CLR.b W0
- CLR.B W0 

Example 4-8: Sample Byte Move Operations

MOV.B #0x30, W0 ; move the literal byte 0x30 to W0
Before Instruction:
    W0 = 0x5555
After Instruction:
    W0 = 0x5530
    MOV.B 0x1000, W0 ; move the byte at 0x1000 to W0
Before Instruction:
    W0 = 0x5555
    Data Memory 0x1000 = 0x1234
After Instruction:
    W0 = 0x5534
    Data Memory 0x1000 = 0x1234
    MOV.B W0, 0x1001 ; byte move W0 to address 0x1001
Before Instruction:
    W0 = 0x1234
    Data Memory 0x1000 = 0x5555
After Instruction:
    W0 = 0x1234
    Data Memory 0x1000 = 0x3455
    MOV.B W0, [W1++] ; byte move W0 to [W1], then post-inc W1
Before Instruction:
    W0 = 0x1234
    W1 = 0x1001
    Data Memory 0x1000 = 0x5555
After Instruction:
    W0 = 0x1234
    W1 = 0x1002
    Data Memory 0x1000 = 0x3455 

Example 4-9: Sample Byte Math Operations
CLR.B [W6--] ; byte clear [W6], then post-dec W6 Before Instruction: W6 = 0x1001 Data Memory 0x1000 = 0x5555 After Instruction: W6 = 0x1000 Data Memory 0x1000 = 0x0055 SUB.B W0, #0x10, W1 ; byte subtract literal 0x10 from W0 ; and store to W1 Before Instruction: W0 = 0x1234 W1 = 0xFFFF After Instruction: W0 = 0x1234 W1 = 0xFF24 ADD.B W0, W1, [W2++] ; byte add W0 and W1, store to [W2] ; and post-inc W2 Before Instruction: W0 = 0x1234 W1 = 0x5678 W2 = 0x1000 Data Memory 0x1000 = 0x5555 After Instruction: W0 = 0x1234 W1 = 0x5678 W2 = 0x1001 Data Memory 0x1000 = 0x55AC

4.5 WORD MOVE OPERATIONS

Even though the data space is byte-addressable, all move operations made in Word mode must be word-aligned. This means that for all source and destination operands, the Least Significant address bit must be '0'. If a word move is made to or from an odd address, an address error exception is generated. Likewise, all double words must be word-aligned. Figure 4-3 shows how bytes and words may be aligned in data memory. Example 4-10 contains several legal word move operations.

When an exception is generated due to a misaligned access, the exception is taken after the instruction executes. If the illegal access occurs from a data read, the operation will be allowed to complete, but the Least Significant bit of the source address will be cleared to force word alignment. If the illegal access occurs during a data write, the write will be inhibited. Example 4-11 contains several illegal word move operations.

Figure 4-3: Data Alignment in Memory

Legend:
b0 - byte stored at 0x1000
b1 - byte stored at 0x1003
b3:b2 – word stored at 0x1005:1004 (b2 is LSB)
b7:b4 - double word stored at 0x1009:0x1006 (b4 is LSB)
b8 - byte stored at 0x100A

Note: Instructions that operate in Word mode are not required to use an instruction extension. However, they may be specified with an optional “.w” or “.w” extension, if desired. For example, the following instructions are valid forms of a word clear operation:

• CLR W0
• CLR.w W0
• CLR.W W0

Example 4-10: Legal Word Move Operations

MOV #0x30, W0 ; move the literal word 0x30 to W0
Before Instruction:
W0 = 0x5555
After Instruction:
W0 = 0x0030
MOV 0x1000, W0 ; move the word at 0x1000 to W0
Before Instruction:
W0 = 0x5555
Data Memory 0x1000 = 0x1234
After Instruction:
W0 = 0x1234
Data Memory 0x1000 = 0x1234
MOV [W0], [W1++] ; word move [W0] to [W1], ; then post-inc W1
Before Instruction:
W0 = 0x1234
W1 = 0x1000
Data Memory 0x1000 = 0x5555
Data Memory 0x1234 = 0xAAAA
After Instruction:
W0 = 0x1234
W1 = 0x1002
Data Memory 0x1000 = 0xAAAA
Data Memory 0x1234 = 0xAAAA 

Example 4-11: Illegal Word Move Operations
MOV 0x1001, W0 ; move the word at 0x1001 to W0 Before Instruction: W0 = 0x5555 Data Memory 0x1000 = 0x1234 Data Memory 0x1002 = 0x5678 After Instruction: W0 = 0x1234 Data Memory 0x1000 = 0x1234 Data Memory 0x1002 = 0x5678 ADDRESS ERROR TRAP GENERATED (source address is misaligned, so MOV is performed) MOV W0, 0x1001 ; move W0 to the word at 0x1001 Before Instruction: W0 = 0x1234 Data Memory 0x1000 = 0x5555 Data Memory 0x1002 = 0x6666 After Instruction: W0 = 0x1234 Data Memory 0x1000 = 0x5555 Data Memory 0x1002 = 0x6666 ADDRESS ERROR TRAP GENERATED (destination address is misaligned, so MOV is not performed) MOV [W0], [W1++] ; word move [W0] to [W1], ; then post-inc W1 Before Instruction: W0 = 0x1235 W1 = 0x1000 Data Memory 0x1000 = 0x1234 Data Memory 0x1234 = 0xAAAA Data Memory 0x1236 = 0xBBBB After Instruction: W0 = 0x1235 W1 = 0x1002 Data Memory 0x1000 = 0xAAAA Data Memory 0x1234 = 0xAAAA Data Memory 0x1236 = 0xBBBB ADDRESS ERROR TRAP GENERATED (source address is misaligned, so MOV is performed)

4.6 USING 10-BIT LITERAL OPERANDS

Several instructions that support Byte and Word mode have 10-bit operands. For byte instructions, a 10-bit literal is too large to use. So when 10-bit literals are used in Byte mode, the range of the operand must be reduced to eight bits or the assembler will generate an error. Table 4-7 shows that the range of a 10-bit literal is 0:1023 in Word mode and 0:255 in Byte mode.

Instructions which employ 10-bit literals in Byte and Word mode are: ADD, ADDC, AND, IOR, RETLW, SUB, SUBB and XOR. Example 4-12 shows how positive and negative literals are used in Byte mode for the ADD instruction.

Table 4-7: 10-Bit Literal Coding

Literal Value	Word Modekk kkkk kkkk	Byte Modekkkk kkkk
0 00 000 0 0000 0000 0000
1 00 000 0 0001 0000 0001
2 00 000 0 0010 0000 0010
127 00 01 11 1111 0111 1111
128 00 10 00 0000 1000 0000
255 00 11 11 1111 1111 1111
256 01 00 00 0000 N/A
512 10 00 00 0000 N/A
1023 11 11 11 1111 N/A

Example 4-12: Using 10-Bit Literals for Byte Operands

ADD.B	#0x80, WO	; add 128 (or -128) to WO
ADD.B	#0x380, WO	; ERROR... Illegal syntax for byte mode
ADD.B	#0xFF, WO	; add 255 (or -1) to WO
ADD.B	#0x3FF, WO	; ERROR... Illegal syntax for byte mode
ADD.B	#0xF, WO	; add 15 to WO
ADD.B	#0x7F, WO	; add 127 to WO
ADD.B	#0x100, WO	; ERROR... Illegal syntax for byte mode

Note: Using a literal value greater than 127 in Byte mode is functionally identical to using the equivalent negative two's complement value, since the Most Significant bit of the byte is set. When operating in Byte mode, the assembler will accept either a positive or negative literal value (i.e., #-10).

4.7 BIT FIELD INSERT/EXTRACT INSTRUCTIONS (dsPIC33C DEVICES ONLY)

The dsPIC33C family provides a set of instructions that operate on bit fields within a target word.

4.7.1 BFEXT

This instruction can extract multiple bits from a W register or data memory location into a destination W register.

4.7.2 BFINs

This instruction can insert multiple bits from a source W register, or 8-bit literal value into a W register or data memory location.

In both instructions, the location and width of the bit field within the target word are defined as literal values within the instruction.

4.8 SOFTWARE STACK POINTER AND FRAME POINTER

4.8.1 Software Stack Pointer

The 16-bit MCU and DSC devices feature a software stack which facilitates function calls and exception handling. W15 is the default Stack Pointer (SP) and after any Reset, it is initialized to 0x0800 (0x1000 for PIC24E, dsPIC33E and dsPIC33C devices). This ensures that the SP will point to valid RAM and permits stack availability for exceptions, which may occur before the SP is set by the user software. The user may reprogram the SP during initialization to any location within data space.

The SP always points to the first available free word (Top-of-Stack) and fills the software stack, working from lower addresses towards higher addresses. It pre-decrements for a stack POP (read) and post-increments for a stack PUSH (write).

The software stack is manipulated using the PUSH and POP instructions. The PUSH and POP instructions are the equivalent of a MOV instruction with W15 used as the destination pointer. For example, the contents of W0 can be PUSHed onto the Top-of-Stack (TOS) by:

PUSH W0

This syntax is equivalent to:

MOV W0, [W15++]

The contents of the TOS can be returned to W0 by:

POP WO

This syntax is equivalent to:

MOV [--W15], W0

During any CALL instruction, the PC is PUSHed onto the stack, such that when the subroutine completes execution, program flow may resume from the correct location. When the PC is PUSHed onto the stack, PC<15:0> are PUSHed onto the first available stack word, then PC<22:16> are PUSHed. When PC<22:16> are PUSHed, the Most Significant seven bits of the PC are zero-extended before the PUSH is made, as shown in Figure 4-4. During exception processing, the Most Significant seven bits of the PC are concatenated with the lower byte of the STATUS Register (SRL) and IPL<3> (CORCON<3>). This allows the primary STATUS Register contents and CPU Interrupt Priority Level to be automatically preserved during interrupts.

Note: In order to protect against misaligned stack accesses, W15<0> is always clear.

Figure 4-4: Stack Operation for CALL Instruction

0x0000 15 0 Stack Grows Towards Higher Address PC<15:0> 0x0 PC<22:16> Top-of-Stack W15 (before CALL) W15 (after CALL) 0xFFFE

Note: For exceptions, the upper nine bits of the second PUSHed word contains the SRL and IPL<3>.

4.8.1.1 STACK POINTER EXAMPLE

Figure 4-5 through Figure 4-8 show how the software stack is modified for the code snippet shown in Example 4-13. Figure 4-5 shows the software stack before the first PUSH has executed. Note that the SP has the initialized value of 0x0800. Furthermore, the example loads 0x5A5A and 0x3636 to W0 and W1, respectively. The stack is PUSHed for the first time in Figure 4-6 and the value contained in W0 is copied to TOS. W15 is automatically updated to point to the next available stack location and the new TOS is 0x0802. In Figure 4-7, the contents of W1 are PUSHed onto the stack and the new TOS becomes 0x0804. In Figure 4-8, the stack is POPped, which copies the last PUSHed value (W1) to W3. The SP is decremented during the POP operation and at the end of the example, the final TOS is 0x0802.

Example 4-13: Stack Pointer Usage

MOV #0x5A5A, W0 ; Load W0 with 0x5A5A
MOV #0x3636, W1 ; Load W1 with 0x3636
PUSH W0 ; Push W0 to TOS (see Figure 4-5)
PUSH W1 ; Push W1 to TOS (see Figure 4-7)
POP W3 ; Pop TOS to W3 (see Figure 4-8) 

Figure 4-5: Stack Pointer Before the First PUSH

0x0000 0x0800 0xFFFE W0 = 0x5A5A W1 = 0x3636 W15 = 0x0800 ← W15 (SP)

Figure 4-6: Stack Pointer After " PUSH w0" Instruction

0x0000 0x0800 5A5A 0x0802← W15 (SP) 0xFFFE W0 = 0x5A5A W1 = 0x3636 W15 = 0x0802

Figure 4-7: Stack Pointer After " PUSH w1" Instruction

0x0000 0x0800 5A5A 0x0802 3636 0x0804 W15 (SP) 0xFFFE W0 = 0x5A5A W1 = 0x3636 W15 = 0x0804

Figure 4-8: Stack Pointer After "POP w3" Instruction

0x0000 0x0800 5A5A 0x0802 0x0804 0xFFFE W15 (SP) W0 = 0x5A5A W1 = 0x3636 W3 = 0x3636 W15 = 0x0802

Note: The contents of 0x802, the new TOS, remain unchanged (0x3636).

4.8.2 Software Stack Frame Pointer

A stack frame is a user-defined section of memory residing in the software stack. It is used to allocate memory for temporary variables, which a function uses, and one stack frame may be created for each function. W14 is the default Stack Frame Pointer (FP) and it is initialized to 0x0000 on any Reset. If the Stack Frame Pointer is not used, W14 may be used like any other Working register.

The Link (LNK) and Unlink (ULNK) instructions provide stack frame functionality. The LNK instruction is used to create a stack frame. It is used during a call sequence to adjust the SP, such that the stack may be used to store temporary variables utilized by the called function. After the function completes execution, the ULNK instruction is used to remove the stack frame created by the LNK instruction. The LNK and ULNK instructions must always be used together to avoid stack overflow.

4.8.2.1 STACK FRAME POINTER EXAMPLE

Figure 4-9 through Figure 4-11 show how a stack frame is created and removed for the code snippet shown in Example 4-14. This example demonstrates how a stack frame operates and is not indicative of the code generated by the compiler. Figure 4-9 shows the stack condition at the beginning of the example, before any registers are pushed to the stack. Here, W15 points to the first free stack location (TOS) and W14 points to a portion of stack memory allocated for the routine that is currently executing.

Before calling the function, "COMPUTE", the parameters of the function (W0, W1 and W2) are PUSHed on the stack. After the "CALL COMPUTE" instruction is executed, the PC changes to the address of "COMPUTE" and the return address of the function, "TASKA", is placed on the stack (Figure 4-10). Function "COMPUTE" then uses the "LNK #4" instruction to PUSH the calling routine's Frame Pointer value onto the stack and the new Frame Pointer will be set to point to the current Stack Pointer. Then, the literal 4 is added to the Stack Pointer address in W15, which reserves memory for two words of temporary data (Figure 4-11).

Inside the function, "COMPUTE", the FP is used to access the function parameters and temporary (local) variables. [W14 + n] will access the temporary variables used by the routine and [W14 - n] is used to access the parameters. At the end of the function, the ULNK instruction is used to copy the Frame Pointer address to the Stack Pointer and then POP the calling subroutine's Frame Pointer back to the W14 register. The ULNK instruction returns the stack back to the state shown in Figure 4-10.

A RETURN instruction will return to the code that called the subroutine. The calling code is responsible for removing the parameters from the stack. The RETURN and POP instructions restore the stack to the state shown in Figure 4-9.

Example 4-14: Frame Pointer Usage

TASKA:
...
PUSH W0 ; Push parameter 1
PUSH W1 ; Push parameter 2
PUSH W2 ; Push parameter 3
CALL COMPUTE ; Call COMPUTE function
POP W2 ; Pop parameter 3
POP W1 ; Pop parameter 2
POP W0 ; Pop parameter 1
...
COMPUTE:
LNK #4 ; Stack FP, allocate 4 bytes for local variables
...
ULNK ; Free allocated memory, restore original FP
RETURN ; Return to TASKA 

Figure 4-9: Stack at the Beginning of Example 4-14

0x0000 0x0800 Frame of TASKA W14 (FP) W15 (SP) 0xFFFE

Figure 4-10: Stack After "CALL COMPUTE" Executes

0x0000 0x0800 Frame of TASKA Parameter 1 Parameter 2 Parameter 3 PC<15:0>(1) 0:PC<22:16> W14 (FP) W15 (SP) 0xFFFE

Note 1: In dsPIC33E/PIC24E devices, the SFA bit is stacked instead of PC<0>.

Figure 4-11: Stack After “LNK #4” Executes

0x0000 0x0800 Frame of TASKA Parameter 1 Parameter 2 Parameter 3 PC<15:0>(1) 0:PC<22:16> FP of TASKA Temp Word 1 Temp Word 2 0xFFFE W14 (FP) W15 (SP)

Note 1: In dsPIC33E/PIC24E devices, the SFA bit is stacked instead of PC<0>.

4.8.3 Stack Pointer Overflow

There is a Stack Limit register (SPLIM) associated with the Stack Pointer that is reset to 0x0000. SPLIM is a 16-bit register, but SPLIM<0> is fixed to '0', because all stack operations must be word-aligned.

The stack overflow check will not be enabled until a word write to SPLIM occurs; after which time, it can only be disabled by a device Reset. All Effective Addresses generated using W15 as a source or destination are compared against the value in SPLIM. Should the Effective Address be greater than the contents of SPLIM, then a stack error trap is generated.

If stack overflow checking has been enabled, a stack error trap will also occur if the W15 Effective Address calculation wraps over the end of data space (0xFFFF).

Refer to the specific device family reference manual for more information on the stack error trap.

4.8.4 Stack Pointer Underflow

The stack is initialized to 0x0800 during Reset (0x1000 for PIC24E, dsPIC33E and dsPIC33C devices). A stack error trap will be initiated should the Stack Pointer address ever be less than 0x0800 (0x1000 for PIC24E, dsPIC33E and dsPIC33C devices).

Note: Locations in data space, between 0x0000 and 0x07FF (0x0FFF for PIC24E, dsPIC33E and dsPIC33C devices), are in general, reserved for core and peripheral Special Function Registers (SFRs).

4.8.5 Stack Frame Active (SFA) Control (dsPIC33E, dsPIC33C and PIC24E Devices)

W15 is never subject to paging and is therefore restricted to address range, 0x000000 to 0x00FFFF. However, the Stack Frame Pointer (W14) for any user software function is only dedicated to that function when a stack frame addressed by W14 is active (i.e., after a LNK instruction). Therefore, it is desirable to have the ability to dynamically switch W14 between use as a general purpose W register and use as a Stack Frame Pointer. The SFA Status bit (CORCON<2>) achieves this function without additional software overhead.

When the SFA bit is clear, W14 may be used with any page register. When SFA is set, W14 is not subject to paging and is locked into the same address range as W15 (0x000000 to 0x00FFFF). Operation of the SFA register lock is as follows:

• The LNK instruction sets SFA (and creates a stack frame).
• The ULNK instruction clears SFA (and deletes the stack frame).
• The CALL, CALL.L and RCALL instructions also stack the SFA bit (placing it in the LSb of the stacked PC), and clear the SFA bit after the stacking operation is complete. The called procedure is now free to either use W14 as a general purpose register or create another stack frame using the LNK instruction.
• The RETURN, RETLW and RETFIE instructions all restore the SFA bit from its previously stacked value.

The SFA bit is a read-only bit. It can only be set by execution of the LNK instruction, and cleared by the ULNK, CALL, CALL.L and RCALL instructions.

Note: In dsPIC33E, dsPIC33C and PIC24E devices, the SFA bit is stacked instead of PC<0>.

4.9 CONDITIONAL BRANCH INSTRUCTIONS

Conditional branch instructions are used to direct program flow based on the contents of the STATUS Register. These instructions are generally used in conjunction with a compare class instruction, but they may be employed effectively after any operation that modifies the STATUS Register.

The compare instructions, CP, CP0 and CPB, perform a subtract operation (minuend – subtrahend), but do not actually store the result of the subtraction. Instead, compare instructions just update the flags in the STATUS Register, such that an ensuing conditional branch instruction may change program flow by testing the contents of the updated STATUS Register. If the result of the STATUS Register test is true, the branch is taken. If the result of the STATUS Register test is false, the branch is not taken.

The conditional branch instructions supported by the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices are shown in Table 4-8. This table identifies the condition in the STATUS Register which must be true for the branch to be taken. In some cases, just a single bit is tested (as in BRA C), while in other cases, a complex logic operation is performed (as in BRA GT). For dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices, it is worth noting that both signed and unsigned conditional tests are supported, and that support is provided for DSP algorithms with the OA, OB, SA and SB condition mnemonics.

Table 4-8: Conditional Branch Instructions

Condition Mnemonic ^(1)	Description Status Test
C Carry (not Borrow) C
GE Signed Greater Than or Equal (N		&& || (N&&OV)
GEU ^(2)	Unsigned Greater Than or Equal	C
GT	Signed Greater Than	(&&&&) || (&&&&)
GTU	Unsigned Greater Than	C&&
LE	Signed Less Than or Equal	Z || (&&) || (N&& )
LEU	Unsigned Less Than or Equal	|| Z
LT	Signed Less Than	(&&) || (N&& )
LTU ^(3)	Unsigned Less Than	C
N Negative	N
NC Not Carry (Borrow)	C
NN Not Negative	N
NOV	Not Overflow	OV
NZ	Not Zero	Z
OA ^(4)	Accumulator A Overflow OA
OB ^(4)	Accumulator B Overflow OB
OV Overflow	OV
SA ^(4)	Accumulator A Saturate	SA
SB ^(4)	Accumulator B Saturate	SB
Z	Zero	Z

Note 1: Instructions are of the form: BRA mnemonic, Expr.
2: GEU is identical to C and will reverse assemble to BRA C, Expr.
3: LTU is identical to NC and will reverse assemble to BRA NC, Expr.
4: This condition is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

Note: The "Compare and Skip" instructions (CPBEQ, CPBGT, CPBLT, CPBNE, CPSEQ, CPSGT, CPSLT and CPSNE) do not modify the STATUS Register.

4.10 Z STATUS BIT

The Z Status bit is a special Zero Status bit that is useful for extended precision arithmetic. The Z bit functions like a normal Z flag for all instructions, except those that use the Carry/Borrow input (ADDC, CPB, SUBB and SUBBR). For the ADDC, CPB, SUBB and SUBBR instructions, the Z bit can only be cleared and never set. If the result of one of these instructions is non-zero, the Z bit will be cleared and will remain cleared, regardless of the result of subsequent ADDC, CPB, SUBB or SUBBR operations. This allows the Z bit to be used for performing a simple zero check on the result of a series of extended precision operations.

A sequence of instructions working on multiprecision data (starting with an instruction with no Carry/Borrow input) will automatically logically AND the successive results of the zero test. All results must be zero for the Z flag to remain set at the end of the sequence of operations. If the result of the ADDC, CPB, SUBB or SUBBR instruction is non-zero, the Z bit will be cleared and remain cleared for all subsequent ADDC, CPB, SUBB or SUBBR instructions. Example 4-15 shows how the Z bit operates for a 32-bit addition. It shows how the Z bit is affected for a 32-bit addition implemented with an ADD/ADDC instruction sequence. The first example generates a zero result for only the most significant word, and the second example generates a zero result for both the least significant word and most significant word.

Example 4-15: 'Z' Status Bit Operation for 32-Bit Addition

; Add two doubles (W0:W1 and W2:W3)
; Store the result in W5:W4
ADD W0, W2, W4 ; Add LSWord and store to W4
ADDC W1, W3, W5 ; Add MSWord and store to W5 

Before 32-Bit Addition (zero result for the most significant word):

w0 = 0x2342
w1 = 0xFFFF0
w2 = 0x39AA
w3 = 0x0010
w4 = 0x0000
w5 = 0x0000
SR = 0x0000 

After 32-Bit Addition:

W0 = 0x2342
W1 = 0xFFFF0
W2 = 0x39AA
W3 = 0x0010
W4 = 0x5CEC
W5 = 0x0000
SR = 0x0201 (DC, C=1) 

Before 32-Bit Addition (zero result for the least significant word and most significant word):

W0 = 0xB76E
W1 = 0xFB7B
W2 = 0x4892
W3 = 0x0484
W4 = 0x0000
W5 = 0x0000
SR = 0x0000 

After 32-Bit Addition:

W0 = 0xB76E
W1 = 0xFB7B
W2 = 0x4892
W3 = 0x0485
W4 = 0x0000
W5 = 0x0000
SR = 0x0103 (DC, Z, C=1) 

4.11 ASSIGNED WORKING REGISTER USAGE

The 16 Working registers of the 16-bit MCU and DSC devices provide a large register set for efficient code generation and algorithm implementation. In an effort to maintain an instruction set that provides advanced capability, a stable run-time environment and backwards compatibility with earlier Microchip processor cores, some Working registers have a preassigned usage. Table 4-9 summarizes these Working register assignments. For the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C, additional details are provided in subsections, Section 4.11.1 "Implied DSP Operands (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)" through Section 4.11.3 "PIC® Microcontroller Compatibility".

Table 4-9: Special Working Register Assignments

Register Special Assignment
W0 Default WREG, Divide Quotient for DIV instructions
W1 Divide Remainder for DIV instructions
W2 “MUL f” Product least significant word
W3 “MUL f” Product most significant word
W4 MAC Operand	(1)
W5 MAC Operand	(1)
W6 MAC Operand	(1)
W7 MAC Operand	(1)
W8 MAC Prefetch Address (X Memory)	(1)
W9 MAC Prefetch Address (X Memory)	(1)
W10 MAC Prefetch Address (Y Memory)	(1)
W11 MAC Prefetch Address (Y Memory)	(1)
W12 MAC Prefetch Offset	(1)
W13 MAC Write-Back Destination	(1)
W14 Frame Pointer
W15 Stack Pointer

Note 1: This assignment is only applicable in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

4.11.1 Implied DSP Operands (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)

To assist instruction encoding and maintain uniformity among the DSP class of instructions, some Working registers have preassigned functionality. For all DSP instructions which have prefetch ability, the following ten register assignments must be adhered to:

• W4-W7 are used for arithmetic operands
- W8-W11 are used for prefetch addresses (pointers)
- W12 is used for the prefetch register offset index
- W13 is used for the accumulator write-back destination

These restrictions only apply to the DSP MAC class of instructions, which utilize Working registers and have prefetch ability (described in Section 4.16 "DSP Accumulator Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)). These instructions are CLR, ED, EDAC, MAC, MOVSAC, MPY, MPY.N and MSC.

In dsPIC33E devices, mixed-sign DSP multiplication operations are supported without the need to dynamically modify the US<1:0> bits. In this mode (US<1:0> = 10), each input operand is treated as unsigned or signed, based on which register is being used for that operand. W4 and W6 are always unsigned operands, whereas W5 and W7 are always signed operands. This feature can be used to efficiently execute extended precision DSP multiplications.

The DSP accumulator class of instructions (described in Section 4.16 "DSP Accumulator Instructions (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)") is not required to follow the Working register assignments in Table 4-9 and may freely use any Working register when required.

4.11.2 Implied Frame and Stack Pointer

To accommodate software stack usage, W14 is the implied Frame Pointer (used by the LNK and ULNK instructions) and W15 is the implied Stack Pointer (used by the CALL, LNK, POP, PUSH, RCALL, RETFIE, RETLW, RETURN, TRAP and ULNK instructions). Even though W14 and W15 have this implied usage, they may still be used as generic operands in any instruction with the exceptions outlined in Section 4.11.1 "Implied DSP Operands (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C Devices)". If W14 and W15 must be used for other purposes (it is strongly advised that they remain reserved for the Frame and Stack Pointers), extreme care must be taken such that the run-time environment is not corrupted.

4.11.3 PIC ^® Microcontroller Compatibility

4.11.3.1 DEFAULT WORKING REGISTER (WREG)

To ease the migration path for users of the Microchip 8-bit PIC MCU families, the 16-bit MCU and DSC devices have matched the functionality of the PIC MCU instruction sets as closely as possible. One major difference between the 16-bit MCU and DSC, and the 8-bit PIC MCU processors is the number of Working registers provided. The 8-bit PIC MCU families only provide one 8-bit Working register, while the 16-bit MCU and DSC families provide sixteen, 16-bit Working registers. To accommodate for the one Working register of the 8-bit PIC MCU, the 16-bit MCU and DSC device instruction set has designated one Working register to be the default Working register for all legacy file register instructions. The default Working register is set to W0 and it is used by all instructions which use File Register Addressing.

Additionally, the syntax used by the 16-bit MCU and DSC device assembler to specify the default Working register is similar to that used by the 8-bit PIC MCU assembler. As shown in the detailed instruction descriptions in Section 5. "Instruction Descriptions", "WREG" must be used to specify the default Working register. Example 4-16 shows several instructions that use WREG.

Example 4-16: Using the Default Working Register, WREG

ADD RAM100 ; add RAM100 and WREG, store in RAM100
ASR RAM100, WREG ; shift RAM100 right, store in WREG
CLR.B WREG ; clear the WREG LS Byte
DEC RAM100, WREG ; decrement RAM100, store in WREG
MOV WREG, RAM100 ; move WREG to RAM100
SETM WREG ; set all bits in the WREG
XOR RAM100 ; XOR RAM100 and WREG, store in RAM100 

4.11.3.2 PRODH:PRODL REGISTER PAIR

Another significant difference between the Microchip 8-bit PIC MCU and 16-bit MCU and DSC architectures is the multiplier. Some PIC MCU families support an 8-bit x 8-bit multiplier, which places the multiply product in the PRODH:PRODL register pair. The 16-bit MCU and DSC devices have a 17-bit x 17-bit multiplier, which may place the result into any two successive Working registers (starting with an even register) or an accumulator.

Despite this architectural difference, the 16-bit MCU and DSC devices still support the legacy file register multiply instruction (MULWF) with the "MUL{.B} f" instruction (described on page 323). Supporting the legacy MULWF instruction has been accomplished by mapping the PRODH:PRODL registers to the Working register pair W3:W2. This means that when "MUL{.B} f" is executed in Word mode, the multiply generates a 32-bit product which is stored in W3:W2, where W3 has the most significant word of the product and W2 has the least significant word of the product. When "MUL{.B} f" is executed in Byte mode, the 16-bit product is stored in W2 and W3 is unaffected. Examples of this instruction are shown in Example 4-17.

Example 4-17: Unsigned f and WREG Multiply (Legacy MULWF Instruction)
MUL.B 0x100 ; (0x100)WREG (byte mode), store to W2 Before Instruction: W0 (WREG) = 0x7705 W2 = 0x1235 W3 = 0x1000 Data Memory 0x0100 = 0x1255 After Instruction: W0 (WREG) = 0x7705 W2 = 0x01A9 W3 = 0x1000 Data Memory 0x0100 = 0x1255 MUL 0x100 ; (0x100)WREG (word mode), store to W3:W2 Before Instruction: W0 (WREG) = 0x7705 W2 = 0x1235 W3 = 0x1000 Data Memory 0x0100 = 0x1255 After Instruction: W0 (WREG) = 0x7705 W2 = 0xDEA9 W3 = 0x0885 Data Memory 0x0100 = 0x1255

4.11.3.3 MOVING DATA WITH WREG

The "MOV{.B} f {,WREG}" instruction (described on page 299) and "MOV{.B} WREG, f" instruction (described on page 300) allow for byte or word data to be moved between file register memory and the WREG (Working register, W0). These instructions provide equivalent functionality to the legacy Microchip PIC MCU MOVF and MOVWF instructions.

The "MOV{.B} f {,WREG}" and "MOV{.B} WREG, f" instructions are the only MOV instructions which support moves of byte data to and from file register memory. Example 4-18 shows several MOV instruction examples using the WREG.

Note: When moving word data between file register memory and the Working register array, the "MOV Wns, f" and "MOV f, Wnd" instructions allow any Working register (W0:W15) to be used as the source or destination register, not just WREG.

Example 4-18: Moving Data with WREG
MOV.B 0x1001, WREG ; move the byte stored at location 0x1001 to W0 MOV 0x1000, WREG ; move the word stored at location 0x1000 to W0 MOV.B WREG, TBLPAG ; move the byte stored at W0 to the TBLPAG register MOV WREG, 0x804 ; move the word stored at W0 to location 0x804

4.12 DSP DATA FORMATS (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

4.12.1 Integer and Fractional Data

The dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices support both integer and fractional data types. Integer data is inherently represented as a signed two's complement value, where the Most Significant bit is defined as a sign bit. Generally speaking, the range of an N-bit two's complement integer is -2^N-1 to 2^N-1-1 . For a 16-bit integer, the data range is -32768 (0x8000) to 32767 (0x7FFF), including '0'. For a 32-bit integer, the data range is -2,147,483,648 (0x8000 0000) to 2,147,483,647 (0x7FFF FFFF).

Fractional data is represented as a two's complement number, where the Most Significant bit is defined as a sign bit and the radix point is implied to lie just after the sign bit. This format is commonly referred to as 1.15 (or Q15) format, where 1 is the number of bits used to represent the integer portion of the number and 15 is the number of bits used to represent the fractional portion. The range of an N-bit two's complement fraction with this implied radix point is -1.0 to (1 - 2^1-N) . For a 16-bit fraction, the 1.15 data range is -1.0 (0x8000) to 0.999969482 (0x7FFF), including 0.0 and it has a precision of 3.05176 × 10^5 . In Normal Saturation mode, the 32-bit accumulators use a 1.31 format, which enhances the precision to 4.6566 × 10^10 .

The dynamic range of the accumulators can be expanded by using the eight bits of the Upper Accumulator register (ACCxU) as guard bits. Guard bits are used if the value stored in the accumulator overflows beyond the 32^nd bit and they are useful for implementing DSP algorithms. This mode is enabled when the ACCSAT bit (CORCON<4>) is set to '1' and it expands the accumulators to 40 bits. The guard bits are also used when the accumulator saturation is disabled. The accumulators then support an integer range of -5.498 × 10^11 (0x80 0000 0000) to 5.498 × 10^11 (0x7F FFFF FFFF). In Fractional mode, the guard bits of the accumulator do not modify the location of the radix point and the 40-bit accumulators use a 9.31 fractional format. Note that all fractional operation results are stored in the 40-bit accumulator, justified with a 1.31 radix point. As in Integer mode, the guard bits merely increase the dynamic range of the accumulator. 9.31 fractions have a range of -256.0 (0x80 0000 0000) to (256.0 - 4.65661x10(0x7F FFFF FFFF). Table 4-10 identifies the range and precision of integers and fractions on the dsPIC30F/33F/33E/33C devices for 16-bit, 32-bit and 40-bit registers.

It should be noted that, with the exception of DSP multiplies, the ALU operates identically on integer and fractional data. Namely, an addition of two integers will yield the same result (binary number) as the addition of two fractional numbers. The only difference is how the result is interpreted by the user. However, multiplies performed by DSP operations are different. In these instructions, data format selection is made by the IF bit (CORCON<0>) and it must be set accordingly ('0' for Fractional mode, '1' for Integer mode). This is required because of the implied radix point used by dsPIC30F/33F/33E/33C fractional numbers. In Integer mode, multiplying two 16-bit integers produces a 32-bit integer result. However, multiplying two 1.15 values generates a 2.30 result. Since the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices use a 1.31 format for the accumulators, a DSP multiply in Fractional mode also includes a left shift of one bit to keep the radix point properly aligned. This feature reduces the resolution of the DSP multiplier to 2^-30 , but has no other effect on the computation (e.g., 0.5 × 0.5 = 0.25 ).

Table 4-10: dsPIC30F/33F/33E/33CData Ranges

Register Size	Integer Range Fraction Range	Range Fraction Resolution
16-bit -32768 to 32	2767-1.0 to (1.0	-2-15)	3.052 × 10^-5
32-bit -2,147,483,648 to 2,147,483,647		-1.0 to (1.0 - 2^-31)	4.657 × 10^-10
40-bit	-549,755,813,888 to 549,755,813,887	-256.0 to (256.0 - 2^-31)	4.657 × 10^-10

4.12.2 Integer and Fractional Data Representation

Having a working knowledge of how integer and fractional data is represented on the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C is fundamental to working with the devices. Both integer and fractional data treat the Most Significant bit as a sign bit, and the binary exponent decreases by one as the bit position advances toward the Least Significant bit. The binary exponent for an N-bit integer starts at (N-1) for the Most Significant bit and ends at '0' for the Least Significant bit. For an N-bit fraction, the binary exponent starts at '0' for the Most Significant bit and ends at (1-N) for the Least Significant bit (as shown in Figure 4-12 for a positive value and in Figure 4-13 for a negative value).

Conversion between integer and fractional representations can be performed using simple division and multiplication. To go from an N-bit integer to a fraction, divide the integer value by 2^N-1 . Similarly, to convert an N-bit fraction to an integer, multiply the fractional value by 2^N-1 .

Figure 4-12: Different Representations of 0x4001

Integer: 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 -2^15 2^14 2^13 2^12...... 0x4001 = 2^14 + 2^0 = 16384 + 1 = 16385 1.15 Fractional: 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 -2^0 . 2^-1 2^-2 2^-3...... Implied Radix Point 0x4001 = 2^-1 + 2^-15 = 0.5 + .000030518 = 0.500030518

Figure 4-13: Different Representations of 0xC002

Integer: 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 -2^15 2^14 2^13 2^12...... 0xC002 = -2^15 + 2^14 + 2^1 = -32768 + 16384 + 2 = -16382 1.15 Fractional: 1 1 0 0 0 0 0 0 0 0 0 0 0 1 0 -2^0 . 2^-1 2^-2 2^-3...... Implied Radix Point 0xC002 = -2^0 + 2^-1 + 2^-14 = -1.0 + 0.5 + 0.000061035 = -0.499938965

4.13 ACCUMULATOR USAGE (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

Accumulators A and B are utilized by DSP instructions to perform mathematical and shifting operations. Since the accumulators are 40 bits wide, and the X and Y data paths are only 16 bits, the method to load and store the accumulators must be understood.

Item A in Figure 4-14 shows that each 40-bit accumulator (ACCA and ACCB) consists of an 8-bit upper register (ACCxU), a 16-bit high register (ACCxH) and a 16-bit low register (ACCxL). To address the bus alignment requirement and provide the ability for 1.31 math, ACCxH is used as a destination register for loading the accumulator (with the LAC instruction), and also as a source register for storing the accumulator (with the SAC.R instruction). This is represented by Item B in Figure 4-14, where the upper and lower portions of the accumulator are shaded. In reality, during accumulator loads, ACCxL is zero backfilled and ACCxU is sign-extended to represent the sign of the value loaded in ACCxH.

Note: dsPIC33C devices provide double-word LAC.D and SAC.D instructions, which allow both ACCxH and ACCxL to be loaded or stored in a single instruction.

When normal (31-bit) saturation is enabled, DSP operations (such as ADD, MAC, MSC, etc.) solely utilize ACCxH:ACCxL (Item C in Figure 4-14) and ACCxU is only used to maintain the sign of the value stored in ACCxH:ACCxL. For instance, when an MPY instruction is executed, the result is stored in ACCxH:ACCxL and the sign of the result is extended through ACCxU.

When super saturation is enabled, or when saturation is disabled, all registers of the accumulator may be used (Item D in Figure 4-14) and the results of DSP operations are stored in ACCxU:ACCxH:ACCxL. The benefit of ACCxU is that it increases the dynamic range of the accumulator, as described in Section 4.12.1 "Integer and Fractional Data". Refer to Table 4-10 to see the range of values which may be stored in the accumulator when in Normal and Super Saturation modes.

Figure 4-14: Accumulator Alignment and Usage

A) ACCxU ACCxH ACCxL 31.30 015163239 Implied Radix Point (between bits 31 and 30) B) C) D)

A) 40-bit accumulator consists of ACCxU:ACCxH:ACCxL

B) Load and store operations

C) Operations in Normal Saturation mode

D) Operations in Super Saturation mode or with saturation disabled

4.14 ACCUMULATOR ACCESS (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The six registers of Accumulator A and Accumulator B are memory-mapped like any other Special Function Register. This feature allows them to be accessed with File Register or Indirect Addressing, using any instruction which supports such addressing. However, it is recommended that the DSP instructions, LAC, SAC and SAC.R, be used to load and store the accumulators, since they provide sign-extension, shifting and rounding capabilities. LAC, SAC and SAC.R instruction details are provided in Section 5. "Instruction Descriptions".

Note 1: For convenience, ACCAU and ACCBU are sign-extended to 16 bits. This provides the flexibility to access these registers using either Byte or Word mode (when File Register or Indirect Addressing is used).

2: The OA, OB, SA or SB bit cannot be set by writing overflowed values to the memory-mapped accumulators using MOV instructions, as these Status bits are only affected by DSP operations.
3: dsPIC33C devices provide double-word LAC.D and SAC.D instructions, which allow both ACCxH and ACCxL to be loaded or stored in a single instruction.

4.15 DSP MAC INSTRUCTIONS (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The DSP Multiply and Accumulate (MAC) operations are a special suite of instructions which provide the most efficient use of the dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C architectures. The DSP MAC instructions, shown in Table 4-11, utilize both the X and Y data paths of the CPU core, which enables these instructions to perform the following operations all in one cycle:

• Two reads from data memory using prefetch Working registers (MAC Prefetches)
• Two updates to prefetch Working registers (MAC Prefetch Register Updates)
  • One mathematical operation with an accumulator (MAC Operations)

In addition, four of the ten DSP MAC instructions are also capable of performing an operation with one accumulator, while storing out the rounded contents of the alternate accumulator. This feature is called accumulator Write-Back (WB) and it provides flexibility for the software developer. For instance, the Accumulator WB may be used to run two algorithms concurrently, or efficiently process complex numbers, among other things.

Table 4-11: DSP MAC Instructions

Instruction Description Accumulator WB?
CLR Clear Accumulator Yes
ED Euclidean Distance (no accumulate)	No
EDAC	Euclidean Distance	No
MAC Multiply and Accumulate	Yes
MAC Square and Accumulate	No
MOVSAC	Move from X and Y Bus	Yes
MPY Multiply to Accumulator	No
MPY Square to Accumulator	No
MPY.N	Negative Multiply to Accumulator	No
MSC Multiply and Subtract	Yes

4.15.1 MAC Prefetches

Prefetches (or data reads) are made using the Effective Address stored in the Working register. The two prefetches from data memory must be specified using the Working register assignments shown in Table 4-9. One read must occur from the X data bus using W8 or W9, and one read must occur from the Y data bus using W10 or W11. The allowed destination registers for both prefetches are W4-W7.

As shown in Table 4-3, one special addressing mode exists for the MAC class of instructions. This mode is the Register Offset Addressing mode and utilizes W12. In this mode, the prefetch is made using the Effective Address of the specified Working register, plus the 16-bit signed value stored in W12. Register Offset Addressing may only be used in the X space with W9 and in the Y space with W11.

4.15.2 MAC Prefetch Register Updates

After the MAC prefetches are made, the Effective Address stored in each prefetch Working register may be modified. This feature enables efficient single-cycle processing for data stored sequentially in X and Y memory. Since all DSP instructions execute in Word mode, only even numbered updates may be made to the Effective Address stored in the Working register. Allowable address modifications to each prefetch register are -6, -4, -2, 0 (no update), +2, +4 and +6. This means that Effective Address updates may be made up to three words in either direction.

When the Register Offset Addressing mode is used, no update is made to the base prefetch register (W9 or W11) or the offset register (W12).

4.15.3 MAC Operations

The mathematical operations performed by the MAC class of DSP instructions center around multiplying the contents of two Working registers and either adding or storing the result to either Accumulator A or Accumulator B. This is the operation of the MAC, MPY, MPY.N and MSC instructions. Table 4-9 shows that W4-W7 must be used for data source operands in the MAC class of instructions. W4-W7 may be combined in any fashion, and when the same Working register is specified for both operands, a square or square and accumulate operation is performed.

For the ED and EDAC instructions, the same multiplicand operand must be specified by the instruction, because this is the definition of the euclidean distance operation. Another unique feature about this instruction is that the values prefetched from X and Y memory are not actually stored in W4-W7. Instead, only the difference of the prefetched data words is stored in W4-W7.

The two remaining MAC class instructions, CLR and MOVSAC, are useful for initiating or completing a series of MAC or EDAC instructions and do not use the multiplier. CLR has the ability to clear Accumulator A or B, prefetch two values from data memory and store the contents of the other accumulator. Similarly, MOVSAC has the ability to prefetch two values from data memory and store the contents of either accumulator.

4.15.4 MAC Write-Back

The Write-Back ability of the MAC class of DSP instructions facilitates efficient processing of algorithms. This feature allows one mathematical operation to be performed with one accumulator and the rounded contents of the other accumulator to be stored in the same cycle. As indicated in Table 4-9, register W13 is assigned for performing the Write-Back and two addressing modes are supported: Direct and Indirect with Post-Increment.

The CLR, MOVSAC and MSC instructions support accumulator Write-Back, while the ED, EDAC, MPY and MPY.N instructions do not support accumulator Write-Back. The MAC instruction, which multiplies two Working registers which are not the same, also supports accumulator Write-Back. However, the square and accumulate MAC instruction does not support accumulator Write-Back (see Table 4-11).

4.15.5 MAC Syntax

The syntax of the MAC class of instructions can have several formats, which depend on the instruction type and the operation it is performing, with respect to prefetches and accumulator Write-Back. With the exception of the CLR and MOVSAC instructions, all MAC class instructions must specify a target accumulator along with two multiplicands, as shown in Example 4-19.

Example 4-19: Base MAC Syntax

; MAC with no prefetch MAC W4*W5, A ; MAC with no prefetch MAC W7*W7, B Multiply W7*W7, Accumulate to ACCB

If a prefetch is used in the instruction, the assembler is capable of discriminating between the X or Y data prefetch based on the register used for the Effective Address. [W8] or [W9] specifies the X prefetch and [W10] or [W11] specifies the Y prefetch. Brackets around the Working register are required in the syntax and they designate that Indirect Addressing is used to perform the prefetch. When address modification is used, it must be specified using a minus-equals or plus-equals "C"-like syntax (i.e., "[W8] -= 2" or "[W8] += 6"). When Register Offset Addressing is used for the prefetch, W12 is placed inside the brackets ([W9 + W12] for X prefetches and [W11 + W12] for Y prefetches). Each prefetch operation must also specify a prefetch destination register (W4-W7). In the instruction syntax, the destination register appears before the prefetch register. Legal forms of prefetch are shown in Example 4-20.

Example 4-20: MAC Prefetch Syntax

graph TD
    A["; MAC with X only prefetch<br>MAC W5*W6, A, [W8"]+=2, W5] --> B["ACCA=ACCA+W5*W6"]
    B --> C["X ([W8"]+=2) → W5]
    D["; MAC with Y only prefetch<br>MAC W5*W5, B, [W11+W12"], W5] --> E["ACCB=ACCB+W5*W5"]
    E --> F["Y ([W11+W12"]) → W5]
    G["; MAC with X/Y prefetch<br>MAC W6*W7, B, [W9"], W6, [W10]+=4, W7] --> H["ACCB=ACCB+W6*W7"]
    H --> I["X ([W9"]) → W6]
    I --> J["Y ([W10"]+=4) → W7]

If an accumulator Write-Back is used in the instruction, it is specified last. The Write-Back must use the W13 register, and allowable forms for the Write-Back are "W13" for Direct Addressing and "[W13] += 2" for Indirect Addressing with Post-Increment. By definition, the accumulator not used in the mathematical operation is stored, so the Write-Back accumulator is not specified in the instruction. Legal forms of accumulator Write-Back (WB) are shown in Example 4-21.

Example 4-21: MAC Accumulator WB Syntax

graph TD
    A["CLR A, W13"] --> B["0 → ACCA"]
    B --> C["ACCB → W13"]
    D["MAC with indirect WB of ACCB"] --> E["MAC W4*W5, A [W13"] += 2]
    E --> F["ACCA=ACCA+W4*W5"]
    F --> G["ACCB → [W13"] += 2]
    H["MAC with Y prefetch, direct WB of ACCA"] --> I["MAC W4*W5, B, [W10"] += 2, W4, W13]
    I --> J["ACCB=ACCB+W4*W5"]
    J --> K["Y ([W10"] += 2) → W4]
    K --> L["ACCA → W13"]

Putting it all together, an MSC instruction which performs two prefetches and a Write-Back is shown in Example 4-22.

Example 4-22: MSC Instruction with Two Prefetches and Accumulator Write-Back

graph TD
    A["MSC w6*W7, B, [W8"] += 2, W6, [W10] -= 6, W7["W13"] += 2] --> B["ACCB=ACCB-W6*W7"]
    A --> C["X ([W8"] += 2) → W6]
    A --> D["Y ([W10"] -= 6) → W7]
    A --> E["ACCA→[W13"] += 2]

4.16 DSP ACCUMULATOR INSTRUCTIONS (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The DSP accumulator instructions do not have prefetch or accumulator WB ability, but they do provide the ability to add, negate, shift, load and store the contents of either 40-bit accumulator. In addition, the ADD and SUB instructions allow the two accumulators to be added or subtracted from each other. DSP accumulator instructions are shown in Table 4-12 and instruction details are provided in Section 5. "Instruction Descriptions".

Table 4-12: DSP Accumulator Instructions

Instruction Description Accumulator WB?
ADD Add Accumulators No
ADD 16-Bit Signed Accumulator Add No
LAC Load Accumulator	No
LAC.D	Load Accumulator Double Word	No
NEG Negate Accumulator	No
SAC	Store Accumulator	No
SAC.D	Store Accumulator Double Word	No
SAC.R	Store Rounded Accumulator	No
SFTAC	Arithmetic Shift Accumulator by Literal	No
SFTAC	Arithmetic Shift Accumulator by (Wn)	No
SUB Subtract Accumulators	No

4.17 SCALING DATA WITH THE FBCL INSTRUCTION (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

To minimize quantization errors that are associated with data processing using DSP instructions, it is important to utilize the complete numerical result of the operations. This may require scaling data up to avoid underflow (i.e., when processing data from a 12-bit ADC) or scaling data down to avoid overflow (i.e., when sending data to a 10-bit DAC). The scaling, which must be performed to minimize quantization error, depends on the dynamic range of the input data which is operated on and the required dynamic range of the output data. At times, these conditions may be known beforehand and fixed scaling may be employed. In other cases, scaling conditions may not be fixed or known and then dynamic scaling must be used to process data.

The FBCL instruction (Find First Bit Change Left) can efficiently be used to perform dynamic scaling, because it determines the exponent of a value. A fixed point or integer value's exponent represents the amount which the value may be shifted before overflowing. This information is valuable, because it may be used to bring the data value to "full scale", meaning that its numeric representation utilizes all the bits of the register it is stored in.

The FBCL instruction determines the exponent of a word by detecting the first bit change, starting from the value's sign bit and working towards the LSB. Since the dsPIC DSC device's barrel shifter uses negative values to specify a left shift, the FBCL instruction returns the negated exponent of a value. If the value is being scaled up, this allows the ensuing shift to be performed immediately with the value returned by FBCL. Additionally, since the FBCL instruction only operates on signed quantities, FBCL produces results in the range of -15:0. When the FBCL instruction returns 0, it indicates that the value is already at full scale. When the instruction returns -15, it indicates that the value cannot be scaled (as is the case with 0x0 and 0xFFFF). Table 4-13 shows word data with various dynamic ranges, their exponents and the value after scaling each data to maximize the dynamic range. Example 4-23 shows how the FBCL instruction may be used for block processing.

Table 4-13: Scaling Examples

Word Value Exponent	Full-Scale Value(Word Value << Exponent)
0x0001 14 0x4000
0x0002 13 0x4000
0x0004 12 0x4000
0x0100 6 0x4000
0x01FF 6 0x7FC0
0x0806 3 0x4030
0x2007 1 0x400E
0x4800 0 0x4800
0x7000 0 0x7000
0x8000 0 0x8000
0x900A 0 0x900A
0xE001 2 0x8004
0xFF07 7 0x8380

Note: For the word values, 0x0000 and 0xFFFF, the FBCL instruction returns -15.

As a practical example, assume that block processing is performed on a sequence of data with very low dynamic range stored in 1.15 fractional format. To minimize quantization errors, the data may be scaled up to prevent any quantization loss which may occur as it is processed. The FBCL instruction can be executed on the sample with the largest magnitude to determine the optimal scaling value for processing the data. Note that scaling the data up is performed by left shifting the data. This is demonstrated with the code snippet below.

Example 4-23: Scaling with FBCL

; assume W0 contains the largest absolute value of the data block
; assume W4 points to the beginning of the data block
; assume the block of data contains BLOCK_SIZE words

; determine the exponent to use for scaling
FBCL W0, W2    ; store exponent in W2

; scale the entire data block before processing
DO    #(BLOCK_SIZE-1), SCALE
LAC    [W4], A    ; move the next data sample to ACCA
SFTAC    A, W2    ; shift ACCA by W2 bits
SCALE:
SAC    A, [W4++]    ; store scaled input (overwrite original)

; now process the data
; (processing block goes here) 

4.18 DATA RANGE LIMIT INSTRUCTIONS (dsPIC33C DEVICES ONLY)

The dsPIC33C family provides special instructions that automatically limit the data in a W register or an accumulator to lie within a user-specified numerical range. These include the FLIM, MAX, MIN and MINZ instructions.

4.18.1 FLIM/FLIM.V

The FLIM instruction simultaneously compares a maximum and a minimum data limit value with the specified W register (or data pointed to by the W register), and clamps the target data to the user-specified limit if the limit is exceeded. SR Status bits are set accordingly for subsequent signed branching. In the FLIM.V instruction, an additional W register is specified, in which the limit test result (known as "limit excess") value is loaded.

4.18.2 MAX/MAX.V

The MAX instruction compares a maximum data limit value (stored in a W register or the other accumulator) with the target accumulator and clamps the target accumulator to the user-specified limit if this upper limit is exceeded. SR Status bits are set accordingly for subsequent signed branching. In the MAX .V instruction, an additional W register (or W register in Indirect Addressing mode) is specified, in which the limit excess value is loaded.

4.18.3 MIN/MIN.V/MINZ

The MIN instruction compares a minimum data limit value (stored in a W register or the other accumulator) with the target accumulator and clamps the target accumulator to the user-specified limit if the data is smaller than this minimum limit. SR Status bits are set accordingly for subsequent signed branching. In the MIN.V instruction, an additional W register (or W register in Indirect Addressing mode) is specified, in which the limit excess value is loaded. The MINZ instruction is simply a conditional version of the MIN instruction, which is executed only when Z = 1.

4.19 NORMALIZING THE ACCUMULATOR WITH THE FBCL INSTRUCTION (dsPIC30F, dsPIC33F, dsPIC33E AND dsPIC33C DEVICES)

The process of scaling a quantized value for its maximum dynamic range is known as normalization (the data in the third column in Table 4-13 contains normalized data). Accumulator normalization is a technique used to ensure that the accumulator is properly aligned before storing data from the accumulator and the FBCL instruction facilitates this function.

The two 40-bit accumulators each have eight guard bits from the ACCxU register, which expands the dynamic range of the accumulators from 1.31 to 9.31 when operating in Super Saturation mode (see Section 4.12.1 "Integer and Fractional Data"). However, even in Super Saturation mode, the Store Rounded Accumulator (SAC.R) instruction only stores 16-bit data (in 1.15 format) from ACCxH, as described in Section Figure 4-13: "Different Representations of 0xC002". Under certain conditions, this may pose a problem.

Proper data alignment for storing the contents of the accumulator may be achieved by scaling the accumulator down if ACCxU is in use or scaling the accumulator up if all of the ACCxH bits are not being used. To perform such scaling, the FBCL instruction must operate on the ACCxU byte and it must operate on the ACCxH word. If a shift is required, the ALU's 40-bit shifter is employed using the SFTAC instruction to perform the scaling. Example 4-24 contains a code snippet for accumulator normalization.

Note: dsPIC33C devices provide a special NORM (normalize accumulator) instruction, which allows an accumulator to be normalized in a single instruction, eliminating the need to use the FBCL and SFTAC instructions for this purpose.

Example 4-24: Normalizing with FBCL

; assume an operation in ACCA has just completed (SR intact)
; assume the processor is in super saturation mode
; assume ACCAH is defined to be the address of ACCAH (0x24)

MOV #ACCAH, W5 ; W5 points to ACCAH
BRA OA, FBCL_GUARD ; if overflow we right shift
FBCL_HI:
FBCL [W5], W0 ; extract exponent for left shift
BRA SHIFT_ACC ; branch to the shift
FBCL_GUARD:
FBCL[++W5], W0 ; extract exponent for right shift
ADD.B W0, #15, W0 ; adjust the sign for right shift
SHIFT_ACC:
SFTAC A, W0 ; shift ACCA to normalize 

4.20 NORMALIZING THE ACCUMULATOR WITH THE NORM INSTRUCTION (dsPIC33C DEVICES ONLY)

The NORM instruction automatically normalizes the target accumulator to obtain the largest fractional value possible without overflow. The target accumulator must contain signed fractional data for the instruction result to be valid. It will shift the target accumulator right or left by as many bits as required to normalize the data, keeping the sign consistent. The shift value is stored in a destination address. The N Status bit reflects the direction of the accumulator shift.

If the accumulator cannot be normalized, the accumulator contents will not be affected.

4.21 EXTENDED PRECISION ARITHMETIC USING MIXED-SIGN MULTIPLICATIONS (dsPIC33E AND dsPIC33C ONLY)

Many DSP algorithms utilize extended precision arithmetic operations (operations with 32-bit or 64-bit operands and results) to enhance the resolution and accuracy of computations. These can be implemented using 16-bit signed or unsigned multiplications; however, this would require some additional processing and shifting of the data to obtain the correct results. To enable such extended precision algorithms to be computed faster, dsPIC33E devices support an optional implicit Mixed-Sign Multiplication mode, which is selected by setting US<1:0>(CORCON<13:12>) = 10.

In this mode, mixed-sign (unsigned x signed and signed x unsigned) multiplications can be performed without the need to dynamically reconfigure the US<1:0> bits and shift data to account for the difference in operand formats. Moreover, signed x signed and unsigned x unsigned multiplications can also be performed without changing the multiplication mode. Each input operand is implicitly treated as an unsigned number if the Working register being used to specify the operand is either W4 or W6. Similarly, an operand is treated as a signed number if the register used is either W5 or W7. The DSP engine selects the type of multiplication to be performed based on the operand registers used, thereby eliminating the need for the user software to modify the US<1:0> bits.

The execution time reductions provided by the implicit mixed-sign multiplication feature is illustrated in the following code example, where the instruction cycle count for performing a 32-bit multiplication is reduced from seven cycles to four cycles when the Mixed-Sign Multiplication mode is enabled.

Example 4-25: 32-Bit Signed Multiplication Using Implicit Mixed-Sign Mode
Case A: Mixed-Sign Multiplication Mode Not Enabled

MUL.SU W5, W6, W0 ; Word1 (signed) x Word2 (unsigned)
MUL.US W4, W7, W2 ; Word0 (unsigned) x Word3 (signed)
CLR B ; Clear Accumulator B
ADD W1, B
ADD W3, B
SFTAC B, #15 ; Shift right by 15 bits to align for Q31 format
MAC W5*W7, B ; Word1 (signed) x Word 3 (signed) 

Case B: Mixed-Sign Multiplication Mode Enabled

MPY W5*W6, B ; Word1 (signed) x Word2 (unsigned)
MAC W4*W7, B ; Word0 (unsigned) x Word3 (signed)
SFTAC B, #15 ; Shift right by 15 bits to align for Q31 format
MAC W5*W7, B ; Word1 (signed) x Word 3 (signed) 

Besides DSP instructions, MCU Multiplication (MUL) instructions can also utilize Accumulator A or Accumulator B as a result destination, which enables faster extended precision arithmetic, even when not using DSP multiplication instructions such as MPY or MAC.

Section 5. Instruction Descriptions

HIGHLIGHTS

This section of the manual contains the following major topics:

5.1 Instruction Symbols....96
5.2 Instruction Encoding Field Descriptors Introduction.... 96
5.3 Instruction Description Example 101
5.4 Instruction Descriptions....102

5.1 INSTRUCTION SYMBOLS

All the symbols used in Section 5.4 "Instruction Descriptions" are listed in Table 1-2.

5.2 INSTRUCTION ENCODING FIELD DESCRIPTORS INTRODUCTION

All instruction encoding field descriptors used in Section 5.4 "Instruction Descriptions" are shown in Table 5-2 through Table 5-15.

Table 5-1: Instruction Encoding Field Descriptors

Field Description
A^(1)	Accumulator selection bit: 0 = ACCA; 1 = ACCB
aa^(1)	Accumulator Write-Back mode (see Table5-13)
B	Byte mode selection bit: 0 = word operation; 1 = byte operation
bbbb	4-bit bit position select: 0000 = LSB; 1111 = MSB
D	Destination address bit: 0 = result stored in WREG; 1 = result stored in file register
dddd	Wd Destination register select: 0000 = W0; 1111 = W15
f ffff ffff ffff 13-bit register file address (0x0000 to 0x1FFF)
fff ffff ffff ffff 15-bit register file word address – implied 0 LSB (0x0000 to 0xFFFE)
ffff ffff ffff ffff 16-bit register file byte address (0x0000 to 0xFFFF)
ggg	Register Offset Addressing mode for Ws Source register (see Table 5-5)
hhh	Register Offset Addressing mode for Wd Destination register (see Table 5-6)
iiii^(1)	Prefetch X operation (see Table5-7)
jjjj^(1)	Prefetch Y operation (see Table5-9)
k 1-bit literal field, constant data or expression
kkkk 4-bit literal field, constant data or expression
kk kkkk 6-bit literal field, constant data or expression
kkkk kkkk 8-bit literal field, constant data or expression
kk kkkk kkkk 10-bit literal field, constant data or expression
kk kkkk kkkk kkkk 14-bit literal field, constant data or expression
kkkk kkkk kkkk kkkk 16-bit literal field, constant data or expression
mm	Multiplier source select with same Working registers (see Table 5-11)
mmm	Multiplier source select with different Working registers (see Table 5-12)
nnnn nnnn nnnn nnn0	23-bit program address for CALL and GOTO instructions
nnnn nnnn nnnn nnnn 16-bit program offset field for relative branch/call instructions
ppp	Addressing mode for Ws Source register (see Table 5-2)
qqq	Addressing mode for Wd Destination register (see Table 5-3)
rrrrr Barrel shift count
ssss	Ws Source register select: 0000 = W0; 1111 = W15
ttttt Dividend select, most significant word
vvvv Dividend select, least significant word
W Double-Word mode selection bit:0 = word operation;1 = double-word operation
www	Wb Base register select: 0000 = W0; 1111 = W15
xx^(1)	Prefetch X destination (see Table5-8)
xxxx xxxx xxxx xxxx 16-bit unused field (don't care)
yy^(1)	Prefetch Y destination (see Table5-10)
z	Bit test destination: 0 = C flag bit; 1 = Z flag bit

Note 1: This field is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.

Table 5-2: Addressing Modes for Ws Source Register

ppp	Addressing Mode Source Operand
000 Register Direct Ws
001 Indirect [Ws]
010 Indirect with Post-Decrement [Ws--]
011 Indirect with Post-Increment [Ws++]
100 Indirect with Pre-Decrement [--Ws]
101 Indirect with Pre-Increment (++Ws]
11x Unused

Table 5-3: Addressing Modes for Wd Destination Register

qqq	Addressing Mode Destination Operand
000 Register Direct Wd
001 Indirect [Wd]
010 Indirect with Post-Decrement [Wd--]
011 Indirect with Post-Increment [Wd++]
100 Indirect with Pre-Decrement [--Wd]
101 Indirect with Pre-Increment (++Wd]
11x Unused (an attempt to use this addressing mode will force a RESET instruction)

Table 5-4: Destination Addressing Modes for MCU Multiplications

dddd	Destination
0000	W1:W0
0001	W0
0010	W3:W2
0011	W2
0100	W5:W4
0101	W4
0110	W7:W6
0111	W6
1000	W9:W8
1001	W8
1010	W11:W10
1011	W10
1100	W13:W12
1101	W12
1110	ACCA<39:0>
1111	ACCB<39:0>

Table 5-5: Offset Addressing Modes for Ws Source Register (with Register Offset)

ggg	Addressing Mode Source Operand
000 Register Direct Ws
001 Indirect [Ws]
010 Indirect with Post-Decrement [Ws--]
011 Indirect with Post-Increment [Ws++]
100 Indirect with Pre-Decrement [--Ws]
101 Indirect with Pre-Increment (++Ws]
11x Indirect with Register Offset [Ws+Wb]

Table 5-6: Offset Addressing Modes for Wd Destination Register (with Register Offset)

hhh	Addressing Mode	Source Operand
000 Register Direct	Wd
001 Indirect	[Wd]
010 Indirect with Post-Decrement	[Wd--]
011 Indirect with Post-Increment	[Wd++]
100 Indirect with Pre-Decrement	[--Wd]
101 Indirect with Pre-Increment	[++Wd]
11x Indirect with Register Offset	[Wd+Wb]

Table 5-7: X Data Space Prefetch Operation (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

iii	Operation
0000	Wxd = [W8]
0001	Wxd = [W8], W8 = W8 + 2
0010	Wxd = [W8], W8 = W8 + 4
0011	Wxd = [W8], W8 = W8 + 6
0100	No Prefetch for X Data Space
0101	Wxd = [W8], W8 = W8 - 6
0110	Wxd = [W8], W8 = W8 - 4
0111	Wxd = [W8], W8 = W8 - 2
1000	Wxd = [W9]
1001	Wxd = [W9], W9 = W9 + 2
1010	Wxd = [W9], W9 = W9 + 4
1011	Wxd = [W9], W9 = W9 + 6
1100	Wxd = [W9 + W12]
1101	Wxd = [W9], W9 = W9 - 6
1110	Wxd = [W9], W9 = W9 - 4
1111	Wxd = [W9], W9 = W9 - 2

Table 5-8: X Data Space Prefetch Destination (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

xx	Wxd
00	W4
01	W5
10	W6
11	W7

Table 5-9: Y Data Space Prefetch Operation (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

jjjj	Operation
0000 Wyd = [W10]
0001 Wyd = [W10], W10 = W10 + 2
0010 Wyd = [W10], W10 = W10 + 4
0011 Wyd = [W10], W10 = W10 + 6
0100 No Prefetch for Y Data Space
0101 Wyd = [W10], W10 = W10 - 6
0110 Wyd = [W10], W10 = W10 - 4
0111 Wyd = [W10], W10 = W10 - 2
1000 Wyd = [W11]
1001 Wyd = [W11], W11 = W11 + 2
1010 Wyd = [W11], W11 = W11 + 4
1011 Wyd = [W11], W11 = W11 + 6
1100 Wyd = [W11 + W12]

Table 5-10: Y Data Space Prefetch Destination (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

yy	Wyd
00 W4
01 W5
10 W6
11 W7

Table 5-11: MAC or MPY Source Operands – Same Working Register (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

mm	Multiplicands
00 W4 * W4
01 W5 * W5
10 W6 * W6
11 W7 * W7

Table 5-12: MAC or MPY Source Operands – Different Working Register (dsPIC30F, dsPIC33F, dsPIC33E, dsPIC33C)

mmm	Multiplicands
000 W4 * W5
001 W4 * W6
010 W4 * W7
011 Invalid
100 W5 * W6
101 W5 * W7
110 W6 * W7
111 Invalid

Table 5-13: MAC Accumulator Write-Back Selection (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C)

aa	Write-Back Selection
00 W13 = Other Accumulator (Direct Addressing)
01 [W13] += 2 = Other Accumulator (Indirect Addressing with Post-Increment)
10 No Write-Back
11 Invalid

Table 5-14: MOVPA G Destination Selection (dsPIC33E, dsPIC33C and PIC24E)

PP	Target Page Register
00 DSRPAG
01 DSWPAG
10 TBLPAG
11 Invalid (results in Illegal Opcode Reset) – do not use

Table 5-15: Accumulator Selection (dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C)

A	Target Accumulator
0 Accumulator A
1 Accumulator B

5.3 INSTRUCTION DESCRIPTION EXAMPLE

The example description below is for the fictitious instruction, FOO. The following example instruction was created to demonstrate how the table fields (syntax, operands, operation, etc.) are used to describe the instructions presented in Section 5.4 "Instruction Descriptions".

FOO

The Header field summarizes what the instruction does

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Cells marked with an 'X' indicate the instruction is implemented for that device family.

Syntax: The Syntax field consists of an optional label, the instruction mnemonic, any optional extensions which exist for the instruction and the operands for the instruction. Most instructions support more than one operand variant to support the various addressing modes. In these circumstances, all possible instruction operands are listed beneath each other and are enclosed in braces.

Operands: The Operands field describes the set of values which each of the operands may take. Operands may be accumulator registers, file registers, literal constants (signed or unsigned) or Working registers.

Operation: The Operation field summarizes the operation performed by the instruction.

Status Affected: The Status Affected field describes which bits of the STATUS Register are affected by the instruction. Status bits are listed by bit position in descending order.

Encoding: The Encoding field shows how the instruction is bit encoded. Individual bit fields are explained in the Description field and complete encoding details are provided in Table 5.2.

Description: The Description field describes in detail the operation performed by the instruction. A key for the encoding bits is also provided.

Words: The Words field contains the number of program words that are used to store the instruction in memory.

Cycles: The Cycles field contains the number of instruction cycles that are required to execute the instruction.

Examples: The Examples field contains examples that demonstrate how the instruction operates. "Before" and "After" register snapshots are provided, which allow the user to clearly understand what operation the instruction performs.

5.4 INSTRUCTION DESCRIPTIONS

ADD

Add f to WREG

Implemented in: PIC24F PIC24	H PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} ADD{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) + (WREG) → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding:

1011

0100

0BDf

ffff

ffff

ffff

Description: Add the contents of the default Working register WREG to the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register, W0.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: ADD.B RAM100 ; Add WREG to RAM100 (Byte mode)

Before Instruction WREG CC80 RAM100 FFC0 SR 0000 After Instruction WREG CC80 RAM100 FF40 SR 0005 (OV, C = 1)

Example 2: ADD RAM200, WREG ; Add RAM200 to WREG (Word mode)

Before Instruction WREG CC80 RAM200 FFC0 SR 0000 After Instruction WREG CC40 RAM200 FFC0 SR 0001 (C = 1)

ADD

Add Literal to Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} ADD{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation lit10 ∈ [0 ... 1023] for word operation Wn ∈ [W0 ... W15]

Operation: lit10 + (Wn) → Wn

Status Affected: DC, N, OV, Z, C

Encoding:

1011

0000

0Bkk

kkkk

kkkk

dddd

Description: Add the 10-bit unsigned literal operand to the contents of the Working register Wn and place the result back into the Working register Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

Words: 1

Cycles: 1

Example 1: ADD.B #0xFF, W7 ; Add -1 to W7 (Byte mode)

Before
Instruction

W7	12C0
SR	0000

After
Instruction

W7	12BF
SR	0009 (N, C = 1)

Example 2: ADD #0xFF, W1 ; Add 255 to W1 (Word mode)

Before
Instruction

W1	12C0
SR	0000

After
Instruction

W1	13BF
SR	0000

ADD

Add Wb to Short Literal

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} ADD{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb) + lit5 → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0100

0www

wBqq

qddd

d11k

kkkk

Description: Add the contents of the base register Wb to the 5-bit unsigned sh

place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: ADD.B W0, #0x1F, W7 ; Add W0 and 31 (Byte mode)

; Store the result in W7

Before

Instruction
W0	2290
W7	12C0
SR	0000

After

Instruction
W0	2290
W7	12AF
SR	0008 (N = 1)

Example 2: ADD W3, #0x6, [--W4] ; Add W3 and 6 (Word mode)

; Store the result in [--W4]

Before

	Instruction
W3	6006
W4	1000
Data 0FFE	DDEE
Data 1000	DDEE
SR	0000

After

Instruction
W3	6006
W4	0FFE
Data 0FFE	600C
Data 1000	DDEE
SR	0000

ADD
Add Wb to Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} ADD{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb) + (Ws) Wd

Status Affected: DC, N, OV, Z, C

Encoding: 0100 0www wBqq qddd dppp ssss

Description: Add the contents of the source register Ws and the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1:

ADD.B

W5, W6, W7

; Add W5 to W6, store result in W7; (Byte mode)

Before

Instruction

W5	AB00
W6	0030
W7	FFFF
SR	0000

After
Instruction

W5	AB00
W6	0030
W7	FF30
SR	0000

Example 2: ADD W5, W6, W7 ; Add W5 to W6, store result in W7 ; (Word mode)

	Before Instruction		After Instruction
W5	AB00 W5	AB00
W6	0030 W6	0030
W7	FFFF W7	AB30
SR	0000 SR	0008 (N = 1)

ADD

Add Accumulators

Implemented in: PIC24F PIC24	H PIC24E ds	sPIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} ADD Acc

Operands: Acc [A, B]

Operation: If (Acc = A):

$$ \begin{array}{l} \overline {{(\mathrm{ACCA}) + (\mathrm{ACCB}) \rightarrow \mathrm{ACCA}}} \ \text { Else: } \ (\mathrm{ACCA}) + (\mathrm{ACCB}) \rightarrow \mathrm{ACCB} \ \end{array} $$

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1011

A000

0000

0000

0000

Description:

Add the contents of Accumulator A to the contents of Accumulator B and place the result in the selected accumulator. This instruction performs a 40-bit addition.

The 'A' bit specifies the destination accumulator.

Words: 1

Cycles: 1

Example 1:

ADD

A

; Add ACCB to ACCA

Before

	Instruction
ACCA	00 0022 3300
ACCB	00 1833 4558
SR	0000

After

	Instruction
ACCA 00	1855 7858
ACCB 00	1833 4558
SR	0000

Example 2:

ADD

B

; Add ACCA to ACCB

; Assume Super Saturation mode enabled

; (ACCSAT = 1, SATA = 1, SATB = 1)

Before
Instruction

ACCA	00 E111 2222
ACCB	00 7654 3210
SR	0000

After
Instruction

ACCA	00 E111 2222
ACCB	01 5765 5432
SR	4800 (OB, OAB = 1)

ADD

16-Bit Signed Add to Accumulator

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Syntax: {label:} ADD Ws, {#Slit4,} Acc
[Ws],
[Ws++]
[Ws--],
[--Ws],
[++Ws],

[Ws+Wb],
Operands: Ws ∈ [W0 ... W15]
Wb ∈ [W0 ... W15]
Slit4 ∈ [-8 ... +7]
Acc ∈ [A,B]
Operation: Shift _Slit4 (Extend(Ws)) + (Acc) → Acc
Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1001

Awww

wrrr

rggg

ssss

Description: Add a 16-bit value specified by the source Working register to the most significant word of the selected accumulator. The source operand may specify the direct contents of a Working register or an Effective Address. The value specified is added to the most significant word of the accumulator by sign-extending and zero backfilling the source operand prior to the operation. The value added to the accumulator may also be shifted by a 4-bit signed literal before the addition is made.
The 'A' bit specifies the destination accumulator.
The 'w' bits specify the offset register Wb.
The 'r' bits encode the optional shift.
The 'g' bits select the source addressing mode.
The 's' bits specify the source register Ws.
Note: Positive values of operand Slit4 represent an arithmetic shift right and negative values of operand Slit4 represent an arithmetic shift left. The contents of the source register are not affected by Slit4.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: ADD W0, #2, A ; Add W0 right-shifted by 2 to ACCA
Before
Instruction

W0	8000
ACCA	00 7000 0000
SR	0000

After
Instruction

W0	8000
ACCA	00 5000 0000
SR	0000

Example 2: ADD [W5++, A ; Add the effective value of W5 to ACCA ; Post-increment W5

Before Instruction W5 2000 W5 2002 ACCA 00 0067 2345 AC CA 00 5067 2345 Data 2000 5000 Data 2000 5000 SR 0000 SR 0000 After Instruction

ADDC

Add f to WREG with Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	C33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} ADDC{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) + (WREG) + (C) → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding:

1011

0100

1BDf

ffff

ffff

ffff

Description:

Add the contents of the default Working register WREG, the contents of the file register and the Carry bit, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

1

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

ADDC.B RAM100

; Add WREG and C bit to RAM100

; (Byte mode)

Before
Instruction

WREG	CC60	WREG	CC60
RAM100	8006	RAM100	8067
SR	0001	(C = 1)	SR

Example 2:

ADDC RAM200, WREG

; Add RAM200 and C bit to the WREG

; (Word mode)

Before
Instruction

WREG	5600	WREG	8A01
RAM200	3400	RAM200	3400
SR	0001	(C = 1)	SR

ADDC

Add Literal to Wn with Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X	X	X	X	X	X

Syntax: {label:} ADDC{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation lit10 ∈ [0 ... 1023] for word operation Wn ∈ [W0 ... W15]

Operation: lit10 + (Wn) + (C) → Wn

Status Affected: DC, N, OV, Z, C

Encoding:

1011

0000

1Bkk

kkkk

kkkk

dddd

Description: Add the 10-bit unsigned literal operand, the contents of the Working register Wn and the Carry bit, and place the result back into the Working register Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: ADDC.B #0xFF, W7 ; Add -1 and C bit to W7 (Byte mode)

Before Instruction		After Instruction
W7	12C0	W7	12BF
SR	0000 (C = 0)	SR	0009 (N, C = 1)

Example 2: ADDC #0xFF, W1 ; Add 255 and C bit to W1 (Word mode)

Before Instruction		After Instruction
W1	12C0	W1	13C0
SR	0001	(C = 1)	SR

ADDC

Add Wb to Short Literal with Carry

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X	X	X	X	X

Syntax: {label:} ADDC{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb) + lit5 + (C) → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0100

lwww

wBqq

qddd

d11k

kkkk

Description: Add the contents of the base register Wb, the 5-bit unsigned short literal operand and the Carry bit, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1:	ADDC.B	W0, #0x1F, [W7]	; Add W0, 31 and C bit (Byte mode)
			; Store the result in [W7]

Before Instruction		After Instruction
W0	CC80	W0	CC80
W7	12C0	W7	12C0
Data 12C0	B000	Data 12C0	B09F
SR	0000 (C = 0)	SR	0008 (N = 1)

Example 2: ADDC W3, #0x6, [--W4]; Add W3, 6 and C bit (Word mode); Store the result in [--W4]

	Before Instruction		After Instruction
	W3	6006 W3 6006
	W4	1000 W4 0FFE
	Data 0FFE	DDEE Data 0FFE 600D
	Data 1000	DDEE Data 1000	DDEE
	SR	0001 (C = 1) 0000	SR

ADDC

Add Wb to Ws with Carry

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} ADDC{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb) + (Ws) + (C) → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0100

lwww

wBqq

qddd

dppp

ssss

Description: Add the contents of the source register Ws, the contents of the base register Wb and the Carry bit, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: ADDC.B W0, [W1++], [W2++] ; Add W0, [W1] and C bit (Byte mode)

; Store the result in [W2]

; Post-increment W1, W2

Before Instruction W0 CC20 W0 CC20 W1 0800 W1 0801 W2 1000 W2 1001 Data 0800 AB25 Data 0800 AB25 Data 1000 FFFF Data 1000 FF46 SR 0001 (C = 1) S After Instruction R 0 0 0 0

Example 2: ADDC W3, [W2++, [W1++] ; Add W3, [W2] and C bit (Word mode)

; Store the result in [W1]

; Post-increment W1, W2

AND

AND f and WREG

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} AND{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f).AND.(WREG) → destination designated by D

Status Affected:

N, Z

Encoding:

1011

0110

0BDF

ffff

ffff

ffff

Description:

Compute the logical AND operation of the contents of the default Working register WREG and the contents of the file register, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'D' bit selects the destination register ('0' for WREG, '1' for file register). The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: AND.B RAM100 ; AND WREG to RAM100 (Byte mode)

	Before Instruction
WREG	CC80
RAM100	FFC0
SR	0000

	After Instruction
WREG	CC80
RAM100	FF80
SR	0008 (N = 1)

Example 2: AND RAM200, WREG ; AND RAM200 to WREG (Word mode)

	Before Instruction
WREG	CC80
RAM200	12C0
SR	0000

	After Instruction
WREG	0080
RAM200	12C0
SR	0000

AND

AND Literal and Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} AND{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation

lit10 ∈ [0 ... 1023] for word operation

Wn ∈ [W0 ... W15]

Operation: lit10.AND.(Wn) → Wn

Status Affected: N, Z

Encoding:

1011

0010

OBkk

kkkk

kkkk

dddd

Description: Compute the logical AND operation of the 10-bit literal operand and the contents of the Working register Wn, and place the result back into the Working register Wn. Register Direct Addressing must be used for Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

Words: 1

Cycles: 1

Example 1: AND.B #0x83, W7 ; AND 0x83 to W7 (Byte mode)

Before
Instruction

W7	12C0
SR	0000

After
Instruction

W7	1280
SR	0008(N=1)

Example 2: AND #0x333, W1 ; AND 0x333 to W1 (Word mode)

Before
Instruction

W1	12D0
SR	0000

After
Instruction

W1	0210
SR	0000

AND

AND Wb and Short Literal

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} AND{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb).AND.lit5 → Wd

Status Affected: N, Z

Encoding: 0110 0www wBqq qddd d11k kkkk

Description: Compute the logical AND operation of the contents of the base register Wb and the 5-bit literal, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: AND.B W0,#0x3,[W1++] ; AND W0 and 0x3 (Byte mode)

; Store to [W1]

; Post-increment W1

Before
Instruction

W0	23A5
W1	2211
Data 2210	9999
SR	0000

After
Instruction

W0	23A5
W1	2212
Data 2210	0199
SR	0000

Example 2: AND W0, #0x1F, W1 ; AND W0 and 0x1F (Word mode)

; Store to W1

Before
Instruction

W0	6723
W1	7878
SR	0000

After
Instruction

W0	6723
W1	0003
SR	0000

AND

AND Wb and Ws

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} AND{.B} Wb, Ws, Wd
[Ws], [Wd]
[Ws++]， [Wd++]
[Ws--], [Wd--]
[++Ws], [++Wd]

[--Ws], [--Wd]
Operands: Wb ∈ [W0 ... W15]
Ws ∈ [W0 ... W15]
Wd ∈ [W0 ... W15]
Operation: (Wb).AND.(Ws) → Wd
Status Affected: N, Z

Encoding:

0110

0www

wBqq

qddd

dppp

ssss

Description:

Compute the logical AND operation of the contents of the source register Ws and the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.
The 'w' bits select the address of the base register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The ‘q’ bits select the destination addressing mode.
The 'd' bits select the destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1:

AND.B

W0, W1 [W2++] ; AND W0 and W1, and; store to [W2] (Byte mode); Post-increment W2

Before
Instruction

W0	AA55
W1	2211
W2	1001
Data 1000	FFFF
SR	0000

After
Instruction

W0	AA55
W1 22	11
W2	1002
Data 1000	11FF
SR	0000

Example 2:

AND W0, [W1++], W2; AND W0 and [W1], and

; store to W2 (Word mode)

; Post-increment W1

Before

Instruction

W0 AA55 W0 AA55

W1 1000 W1 1002

W2 55AA W2 2214

Data 1000 2634 Data 1000 2634

SR 0000 SR 0000

After

Instruction

ASR

Arithmetic Shift Right f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X

Syntax: {label:} ASR{.B} f {,WREG}
Operands: f [0 8191]
Operation: For Byte Operation:

$$ (f < 7 >) \rightarrow \text {Dest} < 7 > $$

$$ (f < 7 >) \rightarrow \text {Dest} < 6 > $$

$$ (f < 6: 1 >) \rightarrow \text {Dest} < 5: 0 > $$

$$ (f < 0 >) \rightarrow C $$

For Word Operation:

$$ (f < 1 5 >) \rightarrow \text {Dest} < 1 5 > $$

$$ (f < 1 5 >) \rightarrow \text {Dest} < 1 4 > $$

$$ (f < 1 4: 1 >) \rightarrow \text {Dest} < 1 3: 0 > $$

$$ (f < 0 >) \rightarrow C $$

Status Affected: N, Z, C

Encoding:

Description:

1101

0101

1BDf

ffff

ffff

ffff

Shift the contents of the file register one bit to the right and place the result in the destination register. The Least Significant bit of the file register is shifted into the Carry bit of the STATUS Register. After the shift is performed, the result is sign-extended. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	ASR.B RAM400, WREG	; ASR RAM400 and store to WREG
		; (Byte mode)

Before
Instruction

WREG	0600
RAM400	0823
SR	0000

After
Instruction

WREG	0611
RAM400	0823
SR	0001 (C = 1)

Example 2: ASR RAM200 ; ASR RAM200 (Word mode)

	Before Instruction	After Instruction
RAM200	8009 RAM M200 C004
SR	0000 SR 0009 (N, C = 1)

ASR

Arithmetic Shift Right Ws

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} ASR{.B} Ws, Wd

$$ [ \mathrm{Ws} ], \quad [ \mathrm{Wd} ] $$

$$ [ \mathrm{Ws} + + ], [ \mathrm{Wd} + + ] $$

$$ [ \mathrm{Ws} - - ], \quad [ \mathrm{Wd} - - ] $$

$$ [ + + \mathrm{Ws} ], \quad [ + + \mathrm{Wd} ] $$

$$ [ -- W s ], \qquad [ -- W d ] $$

Operands: Ws ∈ [W0 ... W15]

$$ \mathrm{Wd} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: For Byte Operation:

$$ (W s < 7 >) \rightarrow W d < 7 > $$

$$ (W s < 7 >) \rightarrow W d < 6 > $$

$$ (W s < 6: 1 >) \rightarrow W d < 5: 0 > $$

$$ (W s < 0 >) \rightarrow C $$

For Word Operation:

$$ (W s < 1 5 >) \rightarrow W d < 1 5 > $$

$$ (W s < 1 5 >) \rightarrow W d < 1 4 > $$

$$ (W s < 1 4: 1 >) \rightarrow W d < 1 3: 0 > $$

$$ (W s < 0 >) \rightarrow C $$

Status Affected: N, Z, C

Encoding:

1101

0001

1Bqq

qddd

dppp

ssss

Description:

Shift the contents of the source register Ws one bit to the right and place the result in the destination register Wd. The Least Significant bit of Ws is shifted into the Carry bit of the STATUS Register. After the shift is performed, the result is sign-extended. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: _1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: ASR.B [W0++, [W1++] ; ASR [W0] and store to [W1] (Byte mode) ; Post-increment W0 and W1

Before Instruction
W0	0600 W0 0601
W1	0801 W1 0802
Data 600	2366 Data 600 2366
Data 800	FFC0 Data 800 33C0
SR	0000 SR 0000

After Instruction

Example 2: ASR W12, W13; ASR W12 and store to W13 (Word mode)

Before Instruction
W12	AB01
W13	0322
SR	0000

	After Instruction
W12	AB01
W13	D580
SR 0009 (N, C = 1)

ASR

Arithmetic Shift Right by Short Literal

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC38C
	X	X		X	X	X

Syntax: {label:} ASR Wb, #lit4, Wnd

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 4 \in [ 0 \dots 1 5 ] $$

$$ \text { Wnd } \in [ \text { W0 } \dots \text { W15 } ] $$

Operation: lit4<3:0> → Shift_Val

$$ \mathrm{Wb} < 1 5 > \rightarrow \text { Wnd } < 1 5: 1 5 - \text { Shift_Val } + 1 > $$

$$ \mathrm {Wb < 15:Shift_ {V} al > \rightarrow Wnd < 15 - Shift_ {V} al:0 > } $$

Status Affected: N, Z

Encoding:

1101

1110

lwww

wddd

d100

kkkk

Description: Arithmetic shift right the contents of the source register Wb by the 4-bit unsigned literal and store the result in the destination register Wnd. After the shift is performed, the result is sign-extended. Direct Addressing must be used for Wb and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1

Example 1: ASR W0, #0x4, W1 ; ASR W0 by 4 and store to W1

Before
Instruction

W0	060F
W1	1234
SR	0000

After
Instruction

W0	060F
W1	0060
SR	0000

Example 2: ASR W0, #0x6, W1 ; ASR W0 by 6 and store to W1

Before
Instruction

W0	80FF
W1	0060
SR	0000

After
Instruction

W0	80FF
W1	FE03
SR	0008 (N = 1)

Example 3: ASR W0, #0xF, W1 ; ASR W0 by 15 and store to W1

Before
Instruction

W0	70FF
W1	CC26
SR	0000

After
Instruction

W0	70FF
W1	0000
SR	0002 (Z = 1)

ASR

Arithmetic Shift Right by Wns

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F ds	PIC33E ds	PIC33C
X	X		X	X		X

Syntax: {label:} ASR Wb, Wns, Wnd

Operands: Wb ∈ [W0 ... W15]

Wns ∈ [W0 ...W15]

Wnd ∈ [W0 ... W15]

Operation: Wns<3:0> → Shift_Val

Wb<15> → Wnd<15:15-Shift_Val + 1>

Wb<15:Shift_Val> → Wnd<15-Shift_Val:0>

Status Affected: N, Z

Encoding:

1101

1110

1www

wddd

d000

ssss

Description: Arithmetic shift right the contents of the source register Wb by the 4 Least Significant bits of Wns (up to 15 positions) and store the result in the destination register Wnd. After the shift is performed, the result is sign-extended. Direct Addressing must be used for Wb, Wns and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the destination register.

The 's' bits select the source register.

Note 1: This instruction operates in Word mode only.

2: If Wns is greater than 15, Wnd = 0x0 if Wb is positive and Wnd = 0xFFFF if Wb is negative.

Words: 1

Cycles: 1

Example 1: ASR W0, W5, W6 ; ASR W0 by W5 and store to W6

Before

Instruction
W0	80FF
W5	0004
W6	2633
SR	0000

After

Instruction
W0	80FF
W5	0004
W6	F80F
SR	0000

Example 2: ASR W0, W5, W6 ; ASR W0 by W5 and store to W6

Before

Instruction
W0	6688
W5	000A
W6	FF00
SR	0000

After

Instruction
W0	6688
W5	000A
W6	0019
SR	0000

Example 3: ASR W11, W12, W13 ; ASR W11 by W12 and store to W13

Before

Instruction
W11	8765
W12	88E4
W13	A5A5
SR	0000

After

Instruction
W11	8765
W12	88E4
W13	F876
SR	0008 (N = 1)

BCLR

Bit Clear in f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} BCLR{.B} f, #bit4
Operands: f ∈ [0 ... 8191] for byte operation f ∈ [0 ... 8190] (even only) for word operation bit4 ∈ [0 ... 7] for byte operation bit4 ∈ [0 ... 15] for byte operation

Operation: 0 → f
Status Affected: None

Encoding:

1010

1001

bbb f

ffff

ffff

ffffb

Description: Clear the bit in the file register f specified by 'bit4'. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations).
The 'b' bits select value bit 4 of the bit position to be cleared.
The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: When this instruction operates in Word mode, the file register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words:	1
Cycles:	1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

BCLR.B 0x800, #0x7

; Clear bit 7 in 0x800

Before Instruction Data 0800 66EF SR 0000 After Instruction Data 0800 666F SR 0000

Example 2:

BCLR

0x400, #0x9

; Clear bit 9 in 0x400

Before Instruction Data 0400 AA55 SR 0000 After Instruction Data 0400 A855 SR 0000

X

BCLR

Bit Clear in Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X	X	X	X	X	X

Syntax: {label:} BCLR{.B} Ws, #bit4

[Ws],

[Ws++],

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation: 0 → Ws

Status Affected: None

Encoding:

1010

0001

bbbb

0B00

0ppp

ssss

Description:

Clear the bit in register Ws specified by 'bit4'. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations). Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select value bit4 of the bit position to be cleared.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 's' bits select the source/destination register.

The 'p' bits select the source addressing mode.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the source register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.
4: In dsPIC33E, dsPIC33C and PIC24E devices, this instruction uses the DSRPAG register for Indirect Addressing generation in Extended Data Space (EDS).

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: BCLR.B W2, #0x2 ; Clear bit 3 in W2

Example 2: BCLR [W0++, #0x0 ; Clear bit 0 in [W0] ; Post-increment W0

Before Instruction W0 2300 W0 2302 Data 2300 5607 Data 2300 5606 SR 0000 After Instruction SR 0000

BFEXT

Bit Field Extract from Ws into Wnd

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
						X

Syntax: {label:} BFEXT #bit4, #wid5, Ws, Wnd

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: bit4 ∈ [0 ... 15]; wid5 ∈ [1 ... 16]; Ws ∈ [W0 ... W15]; Wnd ∈ [W0 ... W15]

Operation: See text

Status Affected: None

Encoding:	1st word	0000	1010	1000	www	MMMM	LLLL
	2nd word	0000	0000	0000	0000	0ppp	ssss

Description: A bit field is extracted (copied) from (Ws) and written into Wnd. The bit field data loaded into Wnd starts at Wnd<0>, and all MSbs within Wnd that are beyond the defined bit field width, will be cleared.

The bit location within Ws of the LSb of the bit field to be extracted is defined by operand bit4. The width of the bit field may be up to 16 bits and is defined by operand wid5.

The 'w' bits select the address of the bit field destination register.

The 's' bits select the address of the data source register.

The 'p' bits select the source addressing mode.

The 'LLLL' bits define the bit field LSb position within the target word.

The ‘MMMM’ bits define the bit field MSb position within the target word.

Words: 2

Cycles: 2

BFEXT

Bit Field Extract from f into Wnd

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
							X

Syntax: {label:} BFEXT #bit4, #wid5, f, Wnd

Operands: bit4 ∈ [0 ... 15]; wid5 ∈ [1 ... 16]; Wnd ∈ [W0 ... W15]; f ∈ [0 ... 65534]

Operation: See text

Status Affected: None

Encoding:	1st word	0000	1010	1010	www	MMMM	LLLL
	2nd word	0000	0000	ffff	ffff	ffff	fff0

Description: A bit field is extracted (copied) from the file register address and written into Wnd. The bit field data loaded into Wnd starts at Wnd<0> and all MSbs within Wnd, that are beyond the defined bit field width, will be cleared.

The bit location within Ws of the LSb of the bit field to be extracted is defined by operand bit4. The width of the bit field may be up to 16 bits and is defined by operand wid5.

The 'w' bits select the address of the bit field destination register. The 'f' bits select the (word) address of the source file register. The 'LLLL' bits define the bit field LSb position within the target word. The 'MMMM' bits define the bit field MSb position within the target word.

Words: 2

Cycles: 2

BFINS

Bit Field Insert from Wb into Wd

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
						X

Syntax: {label:} BFINs #bit4 #wid5, Wns, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: bit4 ∈ [0 ... 15]; wid5 ∈ [1 ... 16]; Wns ∈ [W0 ... W15]; Wd ∈ [W0 ... W15]

Operation: See text

Status Affected: None

Encoding:	1st word	0000	1010	0000	www	MMMM	LLLL
	2nd word	0000	0000	0000	0000	0ppp	dddd

Description: A bit field is read from (Wb) and inserted (copied) into Wd. The bit field data sourced from Wns starts at Wns<0>. All MSbs within Wns, that are beyond the defined bit field width, are ignored and may be set to any value.

The bit location within Wd of the LSb of the bit field to be inserted is defined by operand bit4. The width of the bit field may be up to 16 bits and is defined by operand wid5. The insert operation overwrites the existing bits within the insert range (i.e., it does not shift the existing bits to accommodate the inserted bits).

The 'w' bits select the address of the bit field source register. The 'd' bits select the address of the data destination register. The 'p' bits select Source Addressing Mode 1. The 'LLLL' bits define the bit field LSb position within the target word. The 'MMMM' bits define the bit field MSb position within the target word.

Words: 2

Cycles: 2

BFINS

Bit Field Insert from Wns into f

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPI	C33E dsPIC3	3C
							X

Syntax: {label:} BFINs #bit4, #wid5, Wns, f

Operands:

bit4 ∈ [0 ... 15]; wid5 ∈ [1 ... 16];

Wns ∈ [W0 ... W15]; f ∈ [0 ... 65534]

Operation: See text

Status Affected:

None

Encoding:	1st word	0000	1010	0010	www	MMMM	LLLL
	2nd word	0000	0000	ffff	ffff	ffff	fff0

Description:

A bit field is read from (Wns) and inserted (copied) into the file register address. The bit field data sourced from Wns starts at Wns<0>. All MSbs within Wns, that are beyond the defined bit field width, are ignored and may be set to any value.

The bit location within f of the LSb of the bit field to be inserted is defined by operand bit4. The width of the bit field may be up to 16 bits and is defined by operand wid5. The insert operation overwrites the existing bits within the insert range (i.e., it does not shift the existing bits to accommodate the inserted bits).

The 'f' bits select the (word) address of the destination file register. The 'LLLL' bits define the bit field LSb position within the target word. The 'MMMM' bits define the bit field MSb position within the target word.

Words: 2

Cycles: 2

BFINS

Bit Field Insert Literal into Ws

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
						X

Syntax: {label:} BFINS #bit4, #wid5, #lit8, Ws

[Ws]

[Ws++]

[Ws--]

[++Ws]

[--Ws]

Operands: bit4 ∈ [0 ... 15]; wid5 ∈ [1 ... 16]; lit8 ∈ [0 ... 255]; Ws ∈ [W0 ... W15]

Operation: See text

Status Affected: None

Encoding:	1st word	0000	1010	0100	0000	MMMM	LLLL
	2nd word	0000	0000	kkkk	kkkk	0ppp	ssss

Description: A bit field literal value is inserted (copied) into Ws. The bit field data sourced from the literal starts at the LSb of the literal. All MSbs within the literal value, that are beyond the defined bit field width, are ignored and may be set to any value.

The bit location within Ws of the LSb of the bit field to be inserted is defined by operand bit4. The width of the bit field may be up to 16 bits and is defined by operand wid5.

The 'k' bits contain the bit field source value.

The 's' bits select the address of the source/destination register.

The 'p' bits select Source Addressing Mode 1.

The 'LLLL' bits define the bit field LSb position within the target word.

The 'MMMM' bits define the bit field MSb position within the target word.

Words: 2

Cycles: 2

BOOTSWP ^(1)

Swap Active and Inactive Flash Address Panel

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F dsPIC33E dsPIC33C
	X					X	X

Syntax: {label:} BOOTSWP

Operands: None

Operation: If

(Dual Boot Operating mode and the BOOTSWP instruction are enabled (via device-specific Configuration bits))

Then

(P2ACTIV (NVMCON<10>) → P2ACTIV

1 → SFTSWP (NVMCON<11>))

Else

Execute as NOP

Status Affected:

None

Encoding:

1111

1110

0010

0000

0000

0000

Description:

If the BOOTSWP instruction is enabled (via device-specific Configuration bit) and the device is operating in a Dual Boot mode, and the NVMKEY software interlock sequence has been satisfied, the BOOTSWP instruction will:

1. Toggle the state of the P2ACTIV (NVMCON<10>) status bit, which will swap the Active and Inactive Flash address space within the program space address map.
2. Set SFTSWP (NVMCON<11>), indicating a successful panel swap.

Words: 1

Cycles: 2

Note 1: This instruction is present only in some devices of the device families listed above. Please see the specific device data sheet to ensure that this instruction is supported on a specific device.

BRA

Branch Unconditionally

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
X	X		X	X	X	X

X

Syntax: {label:} BRA Expr
Operands: Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].
Operation: (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register
Status Affected: None

Encoding:

0011

0111

nnnn

nnnn

nnnn

nnnn

Description: The program will branch unconditionally, relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches of up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression. After the branch is taken, the new address will be (PC + 2) + 2* Slit16, since the PC will have incremented to fetch the next instruction.

The 'n' bits are a signed literal that specifies the number of program words offset from (PC + 2). Words: 1 Cycles: 2 (PIC24F, PIC24H, dsPIC30F, dsPIC33F) 4 (PIC24E, dsPIC33E, dsPIC33C)

Example 1:	002000 HERE:	BRA THERE	; Branch to THERE
	002002	. . .
	002004	. . .
	002006	. . .
	002008	. . .
	00200A THERE:	. . .
	00200C	. . .

	Before Instruction
PC	00 2000
SR	0000

	After Instruction
PC	00 200A
SR	0000

Example 2:	002000 HERE:	BRA THERE+0x2	; Branch to THERE+0x2
	002002	. . .
	002004	. . .
	002006	. . .
	002008	. . .
	00200A THERE:	. . .
	00200C	. . .

	Before Instruction
PC	00 2000
SR	0000

	After Instruction
PC	00 200C
SR	0000

Example 3:	002000 HERE:	BRA 0x1366	; Branch to 0x1366
	002002	. . .
	002004	. . .

	Before Instruction
PC	00 2000
SR	0000

	After Instruction
PC	00 1366
SR	0000

BRA

Computed Branch

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsP	C33F dsPIC	33E dsPIC33C
X	X	X	X

Syntax: {label:} BRA Wn
Operands: Wn ∈ [W0 ... W15]
Operation: (PC + 2) + (2 * Wn) → PC
NOP → Instruction Register
Status Affected: None

Encoding:

The program branches unconditionally, relative to the next PC. The offset of the branch is the sign-extended 17-bit value (2 * Wn), which supports branches up to 32K instructions, forward or backward. After this instruction executes, the new PC will be (PC + 2) + 2 * Wn, since the PC will have incremented to fetch the next instruction.
The 's' bits select the source register.
Words: 1
Cycles: 2

Example 1:	002000 HERE:	BRA W7	; Branch forward (2 + 2 * W7)
	002002	. . .
	...	. . .
	...	. . .
	002108	. . .
	00210A TABLE7:	. . .
	00210C	. . .

Before
Instruction

PC	00 2000
W7	0084
SR	0000

After
Instruction

PC	00 210A
W7	0084
SR	0000

BRA

Computed Branch

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC33C
			X			X

Syntax: {label:} BRA Wn
Operands: Wn ∈ [W0 ... W15]
Operation: (PC + 2) + (2 * Wn) → PC
NOP → Instruction Register
Status Affected: None

Encoding:

The program branches unconditionally, relative to the next PC. The offset of the branch is the sign-extended 17-bit value (2 * Wn), which supports branches up to 32K instructions, forward or backward. After this instruction executes, the new PC will be (PC + 2) + 2 * Wn, since the PC will have incremented to fetch the next instruction.
The 's' bits select the source register.
Words: 1
Cycles: 4

Example 1:	002000 HERE:	BRA W7	; Branch forward (2 + 2 * W7)
	002002	. . .
	...	. . .
	...	. . .
	002108	. . .
	00210A TABLE7:	. . .
	00210C	. . .

Before
Instruction

PC	00 2000
W7	0084
SR	0000

After
Instruction

PC	00 210A
W7	0084
SR	0000

BRA C
Branch if Carry

Implemented in: PIC24F PIC	24H PIC24E	dsPIC30F ds	PIC33F dsPIC	33E dsPIC3	3C
	X	X		X	X	X	X

Syntax: {label:} BRA C, Expr
Operands: Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: Condition = C If (Condition) (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register
Status Affected: None

Encoding:

0011

0001

nnnn

nnnn

nnnn

nnnn

Description: If the Carry flag bit is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a 16-bit signed literal that specify the offset from (PC + 2) in instruction words.
Words: 1 Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33C 1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA C, CARRY	; If C is set, branch to CARRY
	002002 NO_C:	...	; Otherwise... continue
	002004	...
	002006	GOTO THERE
	002008 CARRY:	...
	00200A	...
	00200C THERE:	...
	00200E	...

Before Instruction
PC	00 2000
SR	0001 (C = 1)

After Instruction
PC	00 2008
SR	0001 (C = 1)

Example 2:	002000 HERE:	BRA C, CARRY	; If C is set, branch to CARRY
	002002 NO_C:	...	; Otherwise... continue
	002004	...
	002006	GOTO THERE
	002008 CARRY:	...
	00200A	...
	00200C THERE:	...
	00200E	...

	Before Instruction
PC	00 2000
SR	0000

	After Instruction
PC	00 2002
SR	0000

Example 3:	006230 HERE: BRA C, CARRY ; If C is set, branch to CARRY
	006232 NO_C: . . . ; Otherwise... continue
	006234 . . .
	006236 GOTO THERE
	006238 CARRY: . . .
	00623A . . .
	00623C THERE: . . .
	00623E . . .

BeforeInstruction
PC	00 6230 PC 00 6238
SR	0001 (C = 1)

After Instruction
SR	0001 (C = 1)

Example 4:	006230 START: . . .
	006232 . . .
	006234 CARRY: . . .
	006236 . . .
	006238 . . .
	00623A . . .
	00623C HERE: BRA C, CARRY
	00623E . . .

; If C is set, branch to CARRY

; Otherwise... continue

BeforeInstruction
PC	00 623C PC
SR	0001 (C = 1)

After Instruction
SR	0001 (C = 1)

BRA GE

Branch if Signed Greater Than or Equal

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} BRA GE, Expr
Operands: Expr may be a label, absolute address or expression.
Expr is resolved by the linker to a Slit16, where
Slit16 ∈ [-32768 ... +32767].
Operation: Condition = (N&&OV)||(!N&&!OV)
If (Condition)
(PC + 2) + 2 * Slit16 → PC
NOP → Instruction Register
Status Affected: None

Encoding:

0011

1101

nnnn

nnnn

nnnn

nnnn

Description: If the logical expression, (N&&OV)||(!N&&!OV), is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a 16-bit signed literal that specify the offset from (PC + 2) in instruction words.
Note: The assembler will convert the specified label into the offset to be used.
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F
1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	007600 LOOP:	. . .
	007602	. . .
	007604	. . .
	007606	. . .
	007608 HERE:	BRA GE, LOOP	; If GE, branch to LOOP
	00760A NO_GE:	. . .	; Otherwise... continue

Before
Instruction

PC	00 7608
SR	0000

After

	Instruction
PC 00	7600
SR	0000

Example 2: 007600 LOOP: . . .

007602 . . .

007604 . . .

007606 . . .

007608 HERE: BRA GE, LOOP ; If GE, branch to LOOP

00760A NO_GE: . . . ; Otherwise... continue

Before
Instruction

PC	00 7608
SR	0008(N=1)

After
Instruction

PC	00 760A
SR	0008(N=1)

BRA GEU

Branch if Unsigned Greater Than or Equal

Implemented in: PIC24F PIC24H	PIC24E dsP	IC30F dsPIC	C33F dsPIC	C33E dsPIC	C33C
	X	X		X	X		X

Syntax: {label:} BRA GEU, Expr
Operands: Expr may be a label, absolute address or expression.
Expr is resolved by the linker to a Slit16 offset that supports an offset range of [-32768 ... +32767] program words.
Operation: Condition = C
If (Condition)
(PC + 2) + 2^*Slit16 PC
NOP → Instruction Register
Status Affected: None

Encoding:

Description:

0011

0001

nnnn

nnnn

nnnn

nnnn

If the Carry flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.
The 'n' bits are a 16-bit signed literal that specify the offset from (PC + 2) in instruction words.

Note: This instruction is identical to the BRA C, Expr (Branch if Carry) instruction and has the same encoding. It will reverse assemble as BRA C, Slit16.
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F
1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA GEU, BYPASS	; If C is set, branch
	002002 NO_GEU:	...	; to BYPASS
	002004	...	; Otherwise... continue
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction
PC 00 2000
SR 0001 (C = 1)

BRA GT
Branch if Signed Greater Than

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:}

BRA GT, Expr

Operands:

Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = (!Z&&N&&OV)||(!Z&&!N&&!OV)

If (Condition)

(PC + 2) + 2^*Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

1100

nnnn

nnnn

nnnn

nnnn

Description:

If the logical expression, (!Z&&N&&OV)||(!Z&&!N&&!OV), is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a 16-bit signed literal that specify the offset from (PC + 2) in instruction words.

Words:

1

Cycles:

1 (2 if branch taken) - PIC24F, PIC24H, dsPIC30F, dsPIC33F
1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA GT, BYPASS	; If GT, branch to BYPASS
	002002 NO_GT:	. . .	; Otherwise... continue
	002004	. . .
	002006	. . .
	002008	. . .
	00200A	GOTO THERE
	00200C BYPASS:	. . .
	00200E	. . .

Before Instruction
PC	00 2000
SR	0001 (C = 1)

AfterInstruction
PC	00 200C
SR	0001(C = 1)

BRA GTU

Branch if Unsigned Greater Than

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} BRA GTU, Expr

Operands: Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation: Condition = (C&&!Z)

If (Condition)

(PC + 2) + 2 * Slit16 → PC

NOP → Instruction Register

Status Affected: None

Encoding:

None

Description:

0011

1110

nnnn

nnnn

nnnn

nnnn

If the logical expression, (C&&!Z), is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words: 1

Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA GTU, BYPASS	; If GTU, branch to BYPASS
	002002 NO_GTU:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before Instruction		After Instruction
PC	00 2000	PC	00 200C
SR	0001 (C = 1)	SR	0001 (C = 1)

BRA LE

Branch if Signed Less Than or Equal

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
	X	X		X	X

X

Syntax: {label:} BRA LE,

Expr

Operands:

Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = Z||(N&&!OV)||(!N&&OV)

If (Condition)

(PC + 2) + 2^* Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

0100

nnnn

nnnn

nnnn

nnnn

Description:

If the logical expression, Z||(N&&!OV)||(!N&&OV), is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words: 1

Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA LE, BYPASS	; If LE, branch to BYPASS
	002002 NO_LE:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

After Instruction PC 00 2002 SR 0001 (C = 1)

BRA LEU

Branch if Unsigned Less Than or Equal

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X X

Syntax: {label:} BRA LEU, Expr
Operands: Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].
Operation: Condition = !C||Z If (Condition) (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register
Status Affected: None

Encoding:

0011

0110

nnnn

nnnn

nnnn

nnnn

Description: If the logical expression, !C||Z, is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F 1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE: BRA LEU, BYPASS ; If LEU, branch to BYPASS
	002002 NO_LEU: . . . ; Otherwise... continue
	002004 . . .
	002006 . . .
	002008 . . .
	00200A GOTO THERE
	00200C BYPASS: . . .
	00200E . . .

Before Instruction		After Instruction
PC	00 2000	PC	00 200C
SR	0001	(C = 1)	SR

BRA LT
Branch if Signed Less Than

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} BRA LT, Expr
Operands: Expr may be a label, absolute address or expression.
Expr is resolved by the linker to a Slit16, where
Slit16 ∈ [-32768 ... +32767].
Operation: Condition = (N&&!OV)||(!N&&OV)
If (Condition)
(PC + 2) + 2 * Slit16 → PC
NOP → Instruction Register
Status Affected: None
Encoding: 0011 0101 nnnn nnnn nnnn nnnn
Description: If the logical expression, (N&&!OV)||(!N&&OV), is true, then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.
The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F
1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA LT, BYPASS	; If LT, branch to BYPASS
	002002 NO_LT:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before Instruction		After Instruction
PC	00 2000	PC	00 2002
SR	0001 (C = 1)	SR	0001 (C = 1)

BRA LTU

Branch if Unsigned Less Than

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} BRA LTU, Expr

Operands: Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation: Condition = !C

If (Condition)

(PC + 2) + 2 * Slit16 → PC

NOP → Instruction Register

Status Affected: None

Encoding:

None

Description:

0011

1001

nnnn

nnnn

nnnn

nnnn

If the Carry flag is '0', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Note : This instruction is identical to the BRA NC, Expr (Branch if Not Carry) instruction and has the same encoding. It will reverse assemble as BRA NC, Slit16.

Words: 1

Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE: BRA LTU, BYPASS ; If LTU, branch to BYPASS
	002002 NO_LTU: . . . ; Otherwise... continue
	002004 . . .
	002006 . . .
	002008 . . .
	00200A GOTO THERE
	00200C BYPASS: . . .
	00200E . . .

Before Instruction		After Instruction
PC	00 2000	PC	00 2002
SR	0001	(C = 1)	SR	0001

BRA N

Branch if Negative

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} BRA N,

Expr

Operands:

Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = N If (Condition) (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register.

Status Affected:

None

Encoding:

0011

0011

nnnn

nnnn

nnnn

nnnn

Description:

If the Negative flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F 1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA N, BYPASS	; If N, branch to BYPASS
	002002 NO_N:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction

PC	00 2000
SR	0008 (N = 1)

After

Instruction

PC	00 200C
SR	0008 (N = 1)

BRA NC

Branch if Not Carry

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} BRA NC, Expr

Operands: Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation: Condition = !C

If (Condition)

(PC + 2) + 2^*Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

1001

nnnn

nnnn

nnnn

nnnn

Description:

If the Carry flag is '0', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.
The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F
1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA NC, BYPASS	; If NC, branch to BYPASS
	002002 NO_NC:	. . .	; Otherwise... continue
	002004	. . .
	002006	. . .
	002008	. . .
	00200A	GOTO THERE
	00200C BYPASS:	. . .
	00200E	. . .

Before
Instruction

PC	00 2000		PC	00 2002
SR	0001	(C = 1)	SR	0001	(C = 1)

BRA NN

Branch if Not Negative

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} BRA NN,

Expr

Operands:

Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = !N

If (Condition)

(PC + 2) + 2 * Slit16 → PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

1011

nnnn

nnnn

nnnn

nnnn

Description:

If the Negative flag is '0', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) - PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA NN, BYPASS	; If NN, branch to BYPASS
	002002 NO_NN:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction

PC	00 2000
SR	0000

After

PC	00 200C
	0000

BRA NOV

Branch if Not Overflow

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
	X	X		X	X		X

Syntax: {label:} BRA NOV, Expr
Operands: Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: Condition = !OV If (Condition) (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register
Status Affected: None

Encoding:

0011

1000

nnnn

nnnn

nnnn

nnnn

Description: If the Overflow flag is '0', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F 1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA NOV, BYPASS	; If NOV, branch to BYPASS
	002002 NO_NOV:	. . .	; Otherwise... continue
	002004	. . .
	002006	. . .
	002008	. . .
	00200A	GOTO THERE
	00200C BYPASS:	. . .
	00200E	. . .

	Before Instruction
PC	00 2000
SR	0008 (N = 1)

	After Instruction
PC	00 200C
SR	0008 (N = 1)

BRA NZ

Branch if Not Zero

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} BRA NZ, Expr
Operands: Expr may be a label, absolute address or expression. Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].
Operation: Condition = !Z If (Condition) (PC + 2) + 2 * Slit16 → PC NOP → Instruction Register
Status Affected: None

Encoding:

0011

1010

nnnn

nnnn

nnnn

nnnn

Description: If the Z flag is '0', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.
If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .
Words: 1
Cycles: 1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F 1 (4 if branch taken) – PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA NZ, BYPASS	; If NZ, branch to BYPASS
	002002 NO_NZ:	. . .	; Otherwise... continue
	002004	. . .
	002006	. . .
	002008	. . .
	00200A	GOTO THERE
	00200C BYPASS:	. . .
	00200E	. . .

Before
Instruction

PC	00 2000
SR	0002(Z = 1)

After
Instruction

PC	00 2002
SR	0002(Z=1)

BRA OA

Branch if Overflow Accumulator A

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} BRA OA, Expr
Operands: Expr may be a label, absolute address or expression.
Expr is resolved by the linker to a Slit16, where
Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = OA

If (Condition)

(PC + 2) + 2^* Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0000

1100

nnnn

nnnn

nnnn

nnnn

Description:

If the Overflow Accumulator A flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Note: The assembler will convert the specified label into the offset to be used.

Words:

1

Cycles:

1 (2 if branch taken) - dsPIC30F, dsPIC33F
1 (4 if branch taken) - dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA OA, BYPASS	; If OA, branch to BYPASS
	002002 NO_OA:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

	Before Instruction		After Instruction
PC	00 2000	PC	00 200C
SR	8800	(OA, OAB = 1)	SR

BRA OB

Branch if Overflow Accumulator B

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} BRA OB, Expr

Operands: Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation: Condition = OB

If (Condition)

(PC + 2) + 2^*Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0000

1101

nnnn

nnnn

nnnn

nnnn

Description:

If the Overflow Accumulator B flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) - dsPIC30F, dsPIC33F
1 (4 if branch taken) - dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA OB, BYPASS	; If OB, branch to BYPASS
	002002 NO_OB:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before
Instruction

PC	00 2000	PC	00 2002
SR	8800	(OA, OAB = 1)	SR

BRA OV
Branch if Overflow

Implemented in: PIC24F PIC24	H PIC24E ds	sPIC30F dsP	IC33F dsPIC	c33E dsPIC3	3C
	X	X	X	X	X	X	X

Syntax: {label:} BRA OV,

Expr

Operands:

Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = OV

If (Condition)

(PC + 2) + 2^* Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

0000

nnnn

nnnn

nnnn

nnnn

Description:

If the Overflow flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) - PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA OV, BYPASS	; If OV, branch to BYPASS
	002002 NO_OV	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction

PC	00 2000
SR	0002 (Z = 1)

After

Instruction

PC	00 2002
SR	0002 (Z=1)

BRA SA

Branch if Saturation Accumulator A

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Syntax: {label:} BRA SA, Expr
Operands: Expr may be a label, absolute address or expression.
Expr is resolved by the linker to a Slit16, where
Slit16 ∈ [-32768 ... +32767].
Operation: Condition = SA
If (Condition)
(PC + 2) + 2 * Slit16 → PC
NOP → Instruction Register

Status Affected:

None

Encoding:

0000

1110

nnnn

nnnn

nnnn

nnnn

Description:

If the Saturation Accumulator A flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) - dsPIC30F, dsPIC33F
1 (4 if branch taken) - dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA SA, BYPASS	; If SA, branch to BYPASS
	002002 NO_SA:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction

PC	00 2000	PC	00 200C
SR	2400	(SA, SAB = 1)	SR

After

Instruction

BRA SB

Branch if Saturation Accumulator B

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} BRA SB, Expr

Operands: Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation: Condition = SB

if (Condition)

(PC + 2) + 2 * Slit16 → PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0000

1111

nnnn

nnnn

nnnn

nnnn

Description:

If the Saturation Accumulator B flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2) .

Words:

1

Cycles:

1 (2 if branch taken) - dsPIC30F, dsPIC33F
1 (4 if branch taken) - dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA SB, BYPASS	; If SB, branch to BYPASS
	002002 NO_SB:	. . .	; Otherwise... continue
	002004	. . .
	002006	. . .
	002008	. . .
	00200A	GOTO THERE
	00200C BYPASS:	. . .
	00200E	. . .

Before
Instruction

PC	00 2000
SR	0000

After
Instruction

PC	00 2002
SR 0000

BRA Z
Branch if Zero

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

X

Syntax: {label:} BRA Z,

Expr

Operands:

Expr may be a label, absolute address or expression.

Expr is resolved by the linker to a Slit16, where

Slit16 ∈ [-32768 ... +32767].

Operation:

Condition = Z

if (Condition)

(PC + 2) + 2^*Slit16 PC

NOP → Instruction Register

Status Affected:

None

Encoding:

0011

0010

nnnn

nnnn

nnnn

nnnn

Description:

If the Zero flag is '1', then the program will branch relative to the next PC. The offset of the branch is the two's complement number, '2 * Slit16', which supports branches up to 32K instructions, forward or backward. The Slit16 value is resolved by the linker from the supplied label, absolute address or expression.

If the branch is taken, the new address will be (PC + 2) + 2 × Slit16 , since the PC will have incremented to fetch the next instruction. The instruction then becomes a two-cycle instruction, with a NOP executed in the second cycle.

The 'n' bits are a signed literal that specifies the number of instructions offset from (PC + 2).

Words:

1

Cycles:

1 (2 if branch taken) – PIC24F, PIC24H, dsPIC30F, dsPIC33F

1 (4 if branch taken) - PIC24E, dsPIC33E, dsPIC33C

Example 1:	002000 HERE:	BRA Z, BYPASS	; If Z, branch to BYPASS
	002002 NO_Z:	...	; Otherwise... continue
	002004	...
	002006	...
	002008	...
	00200A	GOTO THERE
	00200C BYPASS:	...
	00200E	...

Before

Instruction

PC	00 2000
SR	0002(Z = 1)

After

Instruction

PC	00 200C
SR	0002 (Z = 1)

BSET

Bit Set in f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X X

Syntax: {label:} BSET{.B} f, #bit4

Operands:

f [0 8191] for byte operation

f [0 8190] (even only) for word operation

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation:

1 → f

Status Affected:

None

Encoding:

1010

1000

bbbf

ffff

ffff

ffffb

Description:

Set the bit in the file register 'f' specified by 'bit4'. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations).

The 'b' bits select value bit4 of the bit position to be set.

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the file register address must be word-aligned.

3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words:

1

Cycles:

1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: BSET.B 0x601, #0x3 ; Set bit 3 in 0x601

Before Instruction Data 0600 F234 SR 0000 After Instruction Data 0600 FA34 SR 0000

Example 2:

BSET 0x444, #0xF

; Set bit 15 in 0x444

Before Instruction Data 0444 5604 SR 0000 After Instruction Data 0444 D604 SR 0000

BSET

Bit Set in Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X	X	X	X	X

Syntax: {label:} BSET{.B} Ws, #bit4

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation: 1 → Ws

Status Affected: None

Encoding:

1010

0000

bbbb

0B00

0ppp

ssss

Description:

Set the bit in register Ws specified by 'bit4'. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations). Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select value bit4 of the bit position to be cleared.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'p' bits select the source addressing mode.

The 's' bits select the source/destination register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the source register address must be word-aligned.

3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

4: In dsPIC33E, dsPIC33C and PIC24E devices, this instruction uses the DSRPAG register for indirect address generation in Extended Data Space.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: BSET.B W3, #0x7 ; Set bit 7 in W3

Before

Instruction

W3 0026 W3 00A6

SR 0000 SR 0000

After

Instruction

Example 2: BSET [W4++, #0x0 ; Set bit 0 in [W4]

; Post-increment W4

Before

Instruction

W4 6700 W4 6702

Data 6700 1734 Data 6700 1735

SR 0000

After

Instruction

BSW

Bit Write in Ws

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X

Syntax: {label:} BSW.C Ws, Wb

BSW.Z [Ws],
[Ws++],
[Ws--],
[++Ws],
[--Ws], 

Operands: Ws ∈ [W0 ... W15]

Wb ∈ [W0 ... W15] 

Operation: For ".c" Operation:

C → Ws<(Wb)> 

For “.z” Operation (default):
Ws<(Wb)>
Status Affected: None

Encoding:

1010

1101

Zwww

w000

0ppp

ssss

Description:

The (Wb) bit in register Ws is written with the value of the C or Zflag from the STATUS Register. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 15) of the Working register. Only the four Least Significant bits of Wb are used to determine the destination bit number. Register Direct Addressing must be used for Wb, and either Register Direct or Indirect Addressing may be used for Ws.
The 'Z' bit selects the C or Z flag as source.
The 'w' bits select the address of the bit select register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note: This instruction only operates in Word mode. If no extension is provided, the “.z” operation is assumed.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1:	BSW.C W2, W3	; Set bit W3 in W2 to the value
		; of the C bit

Before
Instruction

W2	F234		W2	7234
W3	111F		W3	111F
SR	0002	(Z = 1, C = 0)	SR	0002	(Z = 1, C = 0)

Example 2: BSW.Z W2, W3 ; Set bit W3 in W2 to the complement ; of the Z bit

Example 3: BSW.C [++W0], W6; Set bit W6 in [W0++] to the value; of the C bit

Example 4: BSW.Z [W1--], W5 ; Set bit W5 in [W1] to the ; complement of the Z bit ; Post-decrement W1

BTG

Bit Toggle in f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} BTG{.B} f, #bit4
Operands: f ∈ [0 ... 8191] for byte operation f ∈ [0 ... 8190] (even only) for word operation bit4 ∈ [0 ... 7] for byte operation bit4 ∈ [0 ... 15] for word operation
Operation: (f)(f)
Status Affected: None

Encoding:

1010

1010

bbb

ffff

ffff

ffffb

Description: Bit 'bit4' in file register 'f' is toggled (complemented). For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operation, bit 15 for word operation) of the byte.

The 'b' bits select value bit4, the bit position to toggle.

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: When this instruction operates in Word mode, the file register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: BTG.B 0x1001, #0x4 ; Toggle bit 4 in 0x1001

Before Instruction Data 1000 F234 SR 0000 After Instruction Data 1000 E234 SR 0000

Example 2: BTG 0x1660, #0x8 ; Toggle bit 8 in RAM660

Before Instruction After Instruction Data 1660 5606 SR 0000 Data 1660 5706 SR 0000

BTG

Bit Toggle in Ws

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F ds	C33E ds	PIC33C
X	X		X	X		X

Syntax: {label:} BTG{.B} Ws, #bit4

[Ws],

[Ws++],

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation: (Ws) Ws

Status Affected: None

Encoding:

1010

0010

bbbb

0B00

0ppp

ssss

Description:

Bit 'bit4' in register Ws is toggled (complemented). For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations). Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select value bit4, the bit position to test.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 's' bits select the source/destination register.

The 'p' bits select the source addressing mode.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the source register address must be word-aligned.

3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

4: In dsPIC33E, dsPIC33C and PIC24E devices, this instruction uses the DSRPAG register for indirect address generation in Extended Data Space.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: BTG W2, #0x0 ; Toggle bit 0 in W2

Before

Instruction

W2 F234 W2 F235

SR 0000 SR 0000

After

Instruction

Example 2: BTG [W0++, #0x0 ; Toggle bit 0 in [W0]

; Post-increment W0

Before

Instruction

W0 2300 W0 2302

Data 2300 5606 Data 2300 5607

SR 0000 SR 0000

After

Instruction

BTSC

Bit Test f, Skip if Clear

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	IC33E dsPIC3	3C
	X	X		X	X		X

X X

Syntax: {label:} BTSC{.B} f, #bit4

Operands:

f [0 8191] for byte operation

f ∈ [0 ... 8190] (even only) for word operation

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation:

Test (f), skip if clear

Status Affected:

None

Encoding:

1010

1111

bbbf

ffff

ffff

fffb

Description:

Bit 'bit4' in the file register is tested. If the tested bit is '0', the next instruction (fetched during the current instruction execution) is discarded and on the next cycle, a NOP is executed instead. If the tested bit is '1', the next instruction is executed as normal. In either case, the contents of the file register are not changed. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations).

The 'b' bits select value bit4, the bit position to test.

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the file register address must be word-aligned.

3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words:

1

Cycles:

1 (2 or 3) ^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

002000 HERE: BTSC.B 0x1201, #2 ; If bit 2 of 0x1201 is 0,002002 GOTO BYPASS ; skip the GOTO002004 . . .002006 . . .002008 BYPASS: . . .00200A . . .

	BeforeInstruction
PC	00	2000
Data 1200	264F
SR	0000

	After Instruction
PC	00 2002
264F
SR	0000

Example 2:

002000 HERE: BTSC 0x804, #14 ; If bit 14 of 0x804 is 0,002002 GOTO BYPASS ; skip the GOTO002004 . . .002006 . . .002008 BYPASS: . . .00200A . . .

Before
Instruction

PC	00 2000 PC	00 2006
Data 0804	2647 Data	0804 2647
SR	0000 SR 0000

After
Instruction

BTSC

Bit Test Ws, Skip if Clear

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} BTSC Ws, #bit4

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 15]

Operation: Test (Ws), skip if clear

Status Affected: None

Encoding:

1010

0111

bbbb

0000

0ppp

ssss

Description:

Bit 'bit4' in Ws is tested. If the tested bit is '0', the next instruction (fetched during the current instruction execution) is discarded and on the next cycle, a NOP is executed instead. If the tested bit is '1', the next instruction is executed as normal. In either case, the contents of Ws are not changed. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 15) of the word. Either Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select value bit4, the bit position to test.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1 (2 or 3 if the next instruction is skipped) ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	002000 HERE:	BTSC	WO, #0x0	; If bit 0 of WO is 0,
	002002	GOTO	BYPASS	; skip the GOTO
	002004	...
	002006	...
	002008 BYPASS:	...
	00200A	...

	Before Instruction
PC	00 2000
W0	264F
SR	0000

	After Instruction
PC	00 2002
W0	264F
SR	0000

Example 2:

002000 HERE: BTSC W6, #0xF ; If bit 15 of W6 is 0,002002 GOTO BYPASS ; skip the GOTO002004 . . .002006 . . .002008 BYPASS: . . .00200A . . .

Before Instruction
PC	00 2000 PC 00 2006
W6	264F W6 264F
SR	0000 SR 0000

After Instruction

Example 3:	003400 HERE: BTSC [W6++, #0xC	; If bit 12 of [W6] is 0,
	003402 GOTO BYPASS	; skip the GOTO
	003404 . . .	; Post-increment W6
	003406 . . .
	003408 BYPASS: . . .
	00340A . . .

	Before Instruction
PC	00 3400 PC 00 3402
W6	1800
Data	1800 1000
SR	0000

	After Instruction
W6	1802
Data 1800	1000
SR	0000

BTSS

Bit Test f, Skip if Set

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} BTSS{.B} f, #bit4

Operands: f [0 8191] for byte operation

f [0 8190] (even only) for word operation

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation: Test (f), skip if set

Status Affected: None

Encoding:

1010

1110

bbbF

ffff

ffff

ffffb

Description: Bit 'bit4' in file register 'f' is tested. If the tested bit is '1', the next instruction (fetched during the current instruction execution) is discarded and on the next cycle, a NOP is executed instead. If the tested bit is '0', the next instruction is executed as normal. In either case, the contents of the file register are not changed. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operation, bit 15 for word operation).

The 'b' bits select value bit4, the bit position to test.

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the file register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words: 1

Cycles: 1 (2 or 3 if the next instruction is skipped) ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	007100 HERE:	BTSS.B	0x1401, #0x1	; If bit 1 of 0x1401 is 1,
	007102	CLR	WREG	; don't clear WREG
	007104	...

	Before Instruction		After Instruction
PC	00 7100	PC	00 7104
Data 1400	0280	Data 1400	0280
SR	0000	SR	0000

Example 2:	007100 HERE:	BTSS	0x890, #0x9	; If bit 9 of 0x890 is 1,
	007102	GOTO	BYPASS	; skip the GOTO
	007104	...
	007106 BYPASS:	...

	Before Instruction		After Instruction
PC	00 7100	PC	00 7102
Data 0890	00FE	Data 0890	00FE
SR	0000	SR	0000

BTSS

Bit Test Ws, Skip if Set

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} BTSS Ws, #bit4

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 15]

Operation: Test (Ws), skip if set.

Status Affected: None

Encoding:

1010

0110

bbbb

0000

0ppp

ssss

Description:

Bit 'bit4' in Ws is tested. If the tested bit is '1', the next instruction (fetched during the current instruction execution) is discarded and on the next cycle, a NOP is executed instead. If the tested bit is '0', the next instruction is executed as normal. In either case, the contents of Ws are not changed. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 15) of the word. Either Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select the value bit4, the bit position to test.

The 's' bits select the source register.

The 'p' bits select the source addressing mode.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1 (2 or 3 if the next instruction is skipped) ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	002000 HERE:	BTSS	WO, #0x0	; If bit 0 of WO is 1,
	002002	GOTO	BYPASS	; skip the GOTO
	002004	...
	002006	...
	002008 BYPASS:	...
	00200A	...

Before
Instruction

PC	00 2000
W0	264F
SR	0000

After
Instruction

PC	00 2006
W0	264F
SR	0000

Example 2:

002000 HERE: BTSS W6, #0xF ; If bit 15 of W6 is 1,002002 GOTO BYPASS ; skip the GOTO002004 . . .002006 . . .002008 BYPASS: . . .00200A . . .

Before Instruction
PC	00 2000 PC 00 2002
W6	264F W6 264F
SR	0000 SR 0000

After Instruction

Example 3:

003400 HERE: BTSS [W6++, 0xC ; If bit 12 of [W6] is 1,003402 GOTO BYPASS ; skip the GOTO003404 . . . ; Post-increment W6003406 . . .003408 BYPASS: . . .00340A . . .

	Before Instruction				After Instruction
PC	00	3400	PC	00	3406
W6	1800	W6	1802
Data 1800	1000			Data 1800	1000
SR	0000			SR	0000

BTST

Bit Test in f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} BTST{.B} f, #bit4
Operands: f [0 8191] for byte operation f [0 8190] (even only) for word operation
bit4 [0 7] for byte operation
bit4 [0 15] for word operation
Operation: (f) Z
Status Affected: Z

Encoding:

1010

1011

bbb f

ffff

ffff

ffffb

Description: Bit 'bit4' in file register 'f' is tested and the complement of the tested bit is stored to the Z flag in the STATUS Register. The contents of the file register are not changed. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operation, bit 15 for word operation).

The 'b' bits select value bit4, the bit position to be tested. The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: When this instruction operates in Word mode, the file register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: BTST.B 0x1201, #0x3 ; Set Z = complement of ; bit 3 in 0x1201

Before Instruction Data 1200 F7FF SR 0000 After Instruction Data 1200 F7FF SR 0002 (Z = 1)

Example 2: BTST 0x1302, #0x7 ; Set Z = complement of ; bit 7 in 0x1302

Before Instruction Data 1302 F7FF Data 1302 F7FF SR 0002 (Z = 1) SR 0000

BTST
Bit Test in Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} BTST.C Ws, #bit4

BTST.Z [Ws],

[Ws++],

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

bit4 ∈ [0 ... 15]

Operation: For ".c" Operation:

(WS) → C

For “.z” Operation (default):

(Ws) → Z

Status Affected: Z or C

Encoding:

1010

0011

bbbb

z000

0ppp

ssss

Description: Bit 'bit4' in register Ws is tested. If the ".Z" option of the instruction is specified, the complement of the tested bit is stored to the Zero flag in the STATUS Register. If the ".C" option of the instruction is specified, the value of the tested bit is stored to the Carry flag in the STATUS Register. In either case, the contents of Ws are not changed.

For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 15) of the word. Either Register Direct or Indirect Addressing may be used for Ws.

The 'b' bits select value bit4, the bit position to test.

The 'Z' bit selects the C or Z flag as destination.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction only operates in Word mode. If no extension is provided, the “.2” operation is assumed.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: BTST.C [W0++, #0x3 ; Set C = bit 3 in [W0] ; Post-increment W0

Before Instruction W0 1200 W0 1202 Data 1200 FFF7 Data 1200 FFF7 SR 0001 (C = 1) S After Instruction R 0 0 0 0

Example 2: BTST.Z W0, #0x7 ; Set Z = complement of bit 7 in W0

Before Instruction W0 F234 W0 F234 SR 0000 SR 0002 (Z = 1) After Instruction

BTST
Bit Test in Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} BTST.C Ws, Wb

BTST.Z [Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wb ∈ [W0 ... W15]

Operation: For “.c” Operation:

(Ws) < (Wb) > C

For “.z” Operation (default):

(Ws) < (Wb) > Z

Status Affected: Z or C

Encoding:

1010

0101

Zwww

w000

0ppp

ssss

Description:

The (Wb) bit in register Ws is tested. If the “.c” option of the instruction is specified, the value of the tested bit is stored to the Carry flag in the STATUS Register. If the “.z” option of the instruction is specified, the complement of the tested bit is stored to the Zero flag in the STATUS Register. In either case, the contents of Ws are not changed.

Only the four Least Significant bits of Wb are used to determine the bit number. Bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 15) of the Working register. Register Direct or Indirect Addressing may be used for Ws.

The 'Z' bit selects the C or Z flag as destination.

The 'w' bits select the address of the bit select register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction only operates in Word mode. If no extension is provided, the “.z” operation is assumed.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: BTST.C W2, W3 ; Set C = bit W3 of W2

Before Instruction W2 F234 W2 F234 W3 2368 W3 2368 SR 0001 (C = 1) SR 0000 After Instruction

Example 2: BTST.Z [W0++, W1 ; Set Z = complement of ; bit W1 in [W0], ; Post-increment W0

Before Instruction W0 1200 W0 1202 W1 CCC0 W1 CCC0 Data 1200 6243 Data 1200 6243 SR 0002 (Z = 1) SR 0000 After Instruction

BTSTS

Bit Test/Set in f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} BTSTS{.B} f, #bit4

Operands: f [0 8191] for byte operation

f [0 8190] (even only) for word operation

bit4 ∈ [0 ... 7] for byte operation

bit4 ∈ [0 ... 15] for word operation

Operation: (f)< bit4> → Z

1 → (f)

Status Affected: Z

Encoding:

1010

1100

bbb

ffff

ffff

ffffb

Description: Bit 'bit4' in file register 'f' is tested and the complement of the tested bit is stored to the Zero flag in the STATUS Register. The tested bit is then set to '1' in the file register. For the bit4 operand, bit numbering begins with the Least Significant bit (bit 0) and advances to the Most Significant bit (bit 7 for byte operations, bit 15 for word operations).

The 'b' bits select value bit4, the bit position to test/set.

The 'f' bits select the address of the file register.

Words: 1

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: When this instruction operates in Word mode, the file register address must be word-aligned.
3: When this instruction operates in Byte mode, 'bit4' must be between 0 and 7.
4: The file register 'f' must not be the CPU STATUS Register (SR).

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: BTSTS.B 0x1201, #0x3 ; Set Z = complement of bit 3 in 0x1201, ; then set bit 3 of 0x1201 = 1

Before Instruction Data 1200 F7FF SR 0000 After Instruction Data 1200 FFFF SR 0002 (Z = 1)

Example 2: BTSTS 0x808, #15 ; Set Z = complement of bit 15 in 0x808, then set bit 15 of 0x808 = 1

Before Instruction RAM300 8050 SR 0002 (Z = 1) After Instruction RAM300 8050 SR 0000

BTSTS
Bit Test/Set in Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C

X

Syntax:	{label:}	BTSTS.C	Ws,	#bit4
				BTSTS.Z	[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands:	Ws ∈ [W0 ... W15]
	bit4 ∈ [0 ... 15]

Operation:

For “.C” Operation:(Ws)<bit4> → C1 → Ws<bit4>

For “.z” Operation (default):

(Ws) < bit4> Z

1 Ws < bit4>

Status Affected:

Z or C

Encoding:

1010

0100

bbbb

Z000

0ppp

ssss

Description:

Bit 'bit4' in register Ws is tested. If the “.z” option of the instruction is specified, the complement of the tested bit is stored to the Zero flag in the STATUS Register. If the “.c” option of the instruction is specified, the value of the tested bit is stored to the Carry flag in the STATUS Register. In both cases, the tested bit in Ws is set to '1'.

The 'b' bits select the value bit4, the bit position to test/set.
The 'Z' bit selects the C or Z flag as destination.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.

Note 1: This instruction only operates in Word mode. If no extension is provided, the “.z” operation is assumed.

2: If Ws is used as a pointer, it must not contain the address of the CPU STATUS Register (SR).

3: In dsPIC33E, dsPIC33C and PIC24E devices, this instruction uses the DSRPAG register for indirect address generation in Extended Data Space.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: BTSTS.C [W0++, #0x3 ; Set C = bit 3 in [W0] ; Set bit 3 in [W0] = 1 ; Post-increment W0

Before Instruction W0 1200 W0 1202 Data 1200 FFFF7 Data 1200 FFFF SR 0001 (C = 1) SR 0000 After Instruction

Example 2: BTSTS.Z W0, #0x7 ; Set Z = complement of bit 7 ; in W0, and set bit 7 in W0 = 1

Before Instruction W0 F234 W0 F2BC SR 0000 SR 0002 (Z = 1) After Instruction

CALL
Call Subroutine

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
	X	X		X	X

Syntax: {label:} CALL Expr

Operands: Expr may be a label or expression (but not a literal).

Expr is resolved by the linker to a lit23, where lit23 ∈ [0 ... 8388606].

Operation: (PC) + 4 → PC

$$ (P C < 1 5: 0 >) \rightarrow (T O S) $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ (P C < 2 3: 1 6 >) \rightarrow (T O S) $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ \text { lit23 } \rightarrow \text { PC } $$

$$ \mathrm{NOP} \rightarrow \text { Instruction Register } $$

Status Affected: None

Encoding:	1st word	0000	0010	nnnn	nnnn	nnnn	nnn0
	2nd word	0000	0000	0000	0000	0nnn	nnnn

Description: Direct subroutine call over the entire 4-Mbyte instruction program memory range. Before the CALL is made, the 24-bit return address (PC + 4) is PUSHed onto the stack. After the return address is stacked, the 23-bit value, 'lit23', is loaded into the PC.

The 'n' bits form the target address.

Note: The linker will resolve the specified expression into the lit23 to be used.

Words: 2

Cycles: 2

Example 1:	026000	CALL	_FIR	; Call _FIR subroutine
	026004	MOV	W0, W1
	.	...
	.	...
	026844 _FIR:	MOV	#0x400, W2	; _FIR subroutine start
	026846	...

	Before Instruction		After Instruction
PC	02 6000	PC	02 6844
W15	A268 W15	A26C
Data A268	FFFF	Data A268	6004
Data A26A	FFFF	Data A26A	0002
SR	0000	SR	0000

Example 2:	072000	CALL	_G66	; call routine _G66
	072004	MOV	W0, W1
	...
	077A28 _G66:	INC	W6, [W7++]	; routine start
	077A2A ...
	077A2C

Before
Instruction

PC	07 2000 PC	07 7A28
W15	9004 W15	9008
Data 9004	FFFF Data	9004 2004
Data 9006	FFFF Data	9006 0007
SR	0000 SR	0000

After
Instruction

CALL
Call Subroutine

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F ds	PIC33F dsPI	C33E dsPIC	33C
			X			X	X

Syntax: {label:} CALL Expr
Operands: Expr may be a label or expression (but not a literal). Expr is resolved by the linker to a lit23, where lit23 ∈ [0 ... 8388606].

Operation: (PC) + 4 → PC (PC<15:1>) → TOS<15:1>, SFA Status bit → TOS<0> (W15) + 2 → W15 (PC<23:16>) → TOS (W15) + 2 → W15 0 → SFA Status bit lit23 → PC NOP → Instruction Register
Status Affected: SFA

Encoding:	1st word	0000	0010	nnnn	nnnn	nnnn	nnnn0
	2nd word	0000	0000	0000	0000	0nnn	nnnn

Description: Direct subroutine call over the entire 4-Mbyte instruction program memory range. Before the CALL is made, the 24-bit return address (PC + 4) is PUSHed onto the stack. After the return address is stacked, the 23-bit value, 'lit23', is loaded into the PC.

The 'n' bits form the target address.

Note: The linker will resolve the specified expression into the lit23 to be used.

Words: 2

Cycles: 4

Example 1:	026000	CALL	_FIR	; Call _FIR subroutine
	026004	MOV	W0, W1
	.	...
	.	...
	026844 _FIR:	MOV	#0x400, W2	; _FIR subroutine start
	026846	...

	Before Instruction		After Instruction
PC	02 6000	PC	02 6844
W15	A268	W15	A26C
Data A268	FFFF	Data A268	6004
Data A26A	FFFF	Data A26A	0002
SR	0000	SR	0000

Example 2:	072000	CALL	_G66	; call routine _G66
	072004	MOV	W0, W1
	...
	077A28 _G66:	INC	W6, [W7++]	; routine start
	077A2A ...
	077A2C

Before
Instruction

PC	07 2000 PC	07 7A28
W15	9004 W15	9008
Data 9004	FFFF Data	9004 2004
Data 9006	FFFF Data	9006 0007
SR	0000 SR	0000

After
Instruction

CALL
Call Indirect Subroutine

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	IC33F dsPIC	33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} CALL Wn

Operands: Wn ∈ [W0 ... W15]

Operation: (PC) + 2 → PC

$$ (P C < 1 5: 0 >) \rightarrow T O S $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ (P C < 2 3: 1 6 >) \rightarrow T O S $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ 0 \rightarrow \mathrm{PC} < 2 2: 1 6 > $$

$$ (W n < 1 5: 1 >) \rightarrow P C < 1 5: 1 > $$

$$ \mathrm{NOP} \rightarrow \text { Instruction Register } $$

Status Affected: None

Encoding:

0000

0001

0000

0000

0000

ssss

Description: Indirect subroutine call over the first 32K instructions of program memory. Before the CALL is made, the 24-bit return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, Wn<15:1> is loaded into PC<15:1> and PC<22:16> is cleared. Since PC<0> is always '0', Wn<0> is ignored.

The 's' bits select the source register.

Words: 1

Cycles: 2

Example 1:	001002	CALL W0	; Call BOOT subroutine indirectly
	001004	...	; using W0
	.	...
	001600 _BOOT:	MOV #0x400, W2	; _BOOT starts here
	001602	MOV #0x300, W6
	.	...

Before
Instruction

PC	00 1002
W0	1600
W15	6F00
Data 6F00	FFFF
Data 6F02	FFFF
SR	0000

After
Instruction

PC	00 1600
W0	1600
W15	6F04
Data 6F00	1004
Data 6F02	0000
SR	0000

Example 2:	004200	CALL W7	; Call TEST subroutine indirectly
	004202 ...		; using W7
	...
	005500 _TEST:	INC W1, W2	; _TEST starts here
	005502	DEC W1, W3	;
	...

	Before Instruction			After Instruction
PC	00 4200 PC	00 5500
W7	5500 W7	5500
W15	6F00 W15	6F04
Data 6F00	FFFF Data	6F00	4202
Data 6F02	FFFF Data	6F02	0000
SR	0000	SR	0000

CALL
Call Indirect Subroutine

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} CALL Wn

Operands: Wn ∈ [W0 ... W15]

Operation: (PC) + 2 → PC

(PC<15:1>) → TOS, SFA Status bit → TOS<0>

(W15) + 2 → W15

(PC<23:16>) → TOS

(W15) + 2 → W15

0 → SFA Status bit

0 → PC<22:16>

(Wn<15:1>) → PC<15:1>

NOP → Instruction Register

Status Affected: SFA

Encoding: 0000 0001 0000 0000 0000 ssss

Description: Indirect subroutine call over the first 32K instructions of program memory. Before the CALL is made, the 24-bit return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, Wn<15:1> is loaded into PC<15:1> and PC<22:16> is cleared. Since PC<0> is always '0', Wn<0> is ignored.

The 's' bits select the source register.

Words: 1

Cycles: 4

Example 1:	001002	CALL W0	; Call BOOT subroutine indirectly
	001004	...	; using W0
	.	...
	001600 _BOOT:	MOV #0x400, W2	; _BOOT starts here
	001602	MOV #0x300, W6
	.	...

	Before Instruction		After Instruction
PC	00 1002	PC	00 1600
W0	1600	W0	1600
W15	6F00	W15	6F04
Data 6F00	FFFF	Data 6F00	1004
Data 6F02	FFFF	Data 6F02	0000
SR	0000	SR	0000

16-Bit MCU and DSC Programmer's Reference Manual

Example 2:	004200	CALL W7	; Call TEST subroutine indirectly
	004202 ...		; using W7
	...
	005500 _TEST:	INC W1, W2	; _TEST starts here
	005502	DEC W1, W3	;
	...

	Before Instruction
PC	00 4200 PC	00 5500
W7	5500 W7 5500
W15	6F00 W15 6F04
Data 6F00	FFFF Data 6F00	4202
Data 6F02	FFFF Data 6F02	0000
SR	0000 SR	0000

After Instruction

CALL.L
Call Indirect Subroutine Long

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F ds	PIC33F dsP	C33E dsPIC	33C
			X			X	X

Syntax: {label:} CALL.L Wn

Operands: Wn ∈ [W0, W2, W4, W6, W8, W10, W12]

Operation: (PC) + 2 → PC

(PC<15:1>) → TOS<15:1>, SFA Status bit → TOS<0>

(W15) + 2 → W15

(PC<23:16>) → TOS

(W15) + 2 → W15

0 → SFA Status bit

PC<23> → PC<23> (see text); (Wn+1)<6:0> → PC<22:16>; (Wn) →

PC<15:0>

NOP → Instruction Register

Status Affected: SFA

Encoding: 0000 0001 1www w000 0000 ssss

Description: Indirect subroutine call to any user program memory address. First, the return address (PC+2) and the state of the Stack Frame Active bit (SFA) are pushed onto the system stack, after which, the SFA bit is cleared.

Then, the Least Significant 7 bits of (Wn+1) are loaded in PC<22:16> and the 16-bit value (Wn) is loaded into PC<15:0>.

PC<23> is not modified by this instruction.

The contents of (Wn+1)<15:7> are ignored.

The value of Wn<0> is also ignored and PC<0> is always set to '0'.

The 's' bits specify the address of the Wn source register.

The 'w' bits specify the address of the Wn+1 source register.

Words: 1

Cycles: 4

Example 1:	026000	CALL.L W4	; Call _FIR subroutine
	026004	MOV W0, W1
	.	...
	.	...
	026844 _FIR:	MOV #0x400, W2	; _FIR subroutine start
	026846	...

	Before Instruction		After Instruction
PC	02 6000	PC	02 6844
W4	6844	W4	6844
W5	0002	W5	0002
W15	A268	A26C
Data A268	FFFF	Data A268	6004
Data A26A	FFFF	Data A26A	0002
SR	0000	SR	0000

CLR

Clear f or WREG

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} CLR{.B} f

WREG

Operands: f [0 8191]

Operation: 0 → destination designated by D

Status Affected: None

Encoding:

1110

1111

0BDF

ffff

ffff

ffff

Description:

Clear the contents of a file register or the default Working register WREG. If WREG is specified, the WREG is cleared. Otherwise, the specified file register 'f' is cleared.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1

Example 1: CLR.B RAM200 ; Clear RAM200 (Byte mode)

Before Instruction		After Instruction
RAM200	8009	RAM200	8000
SR	0000	SR	0000

Example 2: CLR WREG ; Clear WREG (Word mode)

	Before Instruction	After Instruction
WREG	0600	WREG
SR	0000	SR

CLR

Clear Wd

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

X

Syntax: {label:} CLR{.B} Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wd ∈ [W0 ... W15]

Operation: 0 → Wd

Status Affected:

None

Encoding:

1110

1011

0Bqq

qddd

d000

0000

Description:

Clear the contents of register Wd. Either Register Direct or Indirect Addressing may be used for Wd.

The 'B' bit select byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: CLR.B W2 ; Clear W2 (Byte mode)

Before

Instruction
W2	3333
SR	0000

After

Instruction
W2	3300
SR	0000

Example 2: CLR [W0++] ; Clear [W0]

; Post-increment W0

Before
Instruction

W0	2300
Data 2300	5607
SR	0000

After
Instruction

W0	2302
Data 2300	0000
SR	0000

CLR
Clear Accumulator, Prefetch Operands

Implemented in: PIC24F	PIC24H PIC24E dsPIC30F		dsPIC33F dsPIC33E dsPIC33C
				X	X	X	X
Syntax:	{label:}	CLR Acc	{, [Wx], Wxd}	{, [Wy], Wyd}	{, AWB}
			{, [Wx] + = kx, Wxd}	{, [Wy] + = ky, Wyd}
			{, [Wx] - = kx, Wxd}	{, [Wy] - = ky, Wyd}
			{, [W9 + W12], Wxd}	{, [W11 + W12], Wyd}
Operands:	Acc ∈ [A,B]Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]; Wxd ∈ [W4 ... W7]Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]; Wyd ∈ [W4 ... W7]AWB ∈ [W13, [W13] + = 2]
Operation:	0 → Acc(A or B)([Wx]) → Wxd; (Wx) +/- kx → Wx([Wy]) → Wyd; (Wy) +/- ky → Wy(Acc(B or A)) rounded → AWB
Status Affected:	OA, OB, SA, SB
Encoding:	1100	0011	A0xx	yyii	iijj	jjaa
Description:	Clear all 40 bits of the specified accumulator. Optionally prefetch operands in preparation for a MAC type instruction and optionally store the non-specified accumulator results. This instruction clears the respective overflow and saturate flags (either OA, SA or OB, SB).Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing, as described in Section 4.15.1 “MAC Prefetches”.Operand AWB specifies the optional register direct or indirect store of the convergently rounded contents of the “other” accumulator, as described in Section 4.15.4 “MAC Write-Back”.The ‘A’ bit selects the other accumulator used for Write-Back.The ‘x’ bits select the prefetch Wxd destination.The ‘y’ bits select the prefetch Wyd destination.The ‘i’ bits select the Wx prefetch operation.The ‘j’ bits select the Wy prefetch operation.The ‘a’ bits select the accumulator Write-Back destination.
Words:	1
Cycles:	1

Example 1: CLR A, [W8] += 2, W4, W13 ; Clear ACCA ; Load W4 with [W8], post-inc W8 ; Store ACCB to W13

	Before Instruction		After Instruction
W4	F001 W4	1221
W8	2000 W8	2002
W13	C623 W13	5420
ACCA	00 0067 2345	ACCA 00 0000 0000
ACCB	00 5420 3BDD	ACCB 00 5420 3BDD
Data 2000	1221 Data 2000	1221
SR	0000 SR 0000

Example 2: CLR B, [W8] += 2, W6, [W10] += 2, W7, [W13] += 2 ; Clear ACCB ; Load W6 with [W8] ; Load W7 with [W10] ; Save ACCA to [W13] ; Post-inc W8, W10, W13

Before Instruction		After Instruction
W6	F001	W6	1221
W7	C783	W7	FF80
W8	2000	W8	2002
W10	3000	W10	3002
W13	4000	W13	4002
ACCA	00 0067 2345	ACCA	00 0067 2345
ACCB	00 5420 ABDD	ACCB	00 0000 0000
Data 2000	1221	Data 2000	1221
Data 3000	FF80	Data 3000	FF80
Data 4000	FFC3	Data 4000	0067
SR	0000	SR	0000

CLRWDT

Clear Watchdog Timer

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} CLRWDT

Operands: None

Operation:	0 → WDT Count register
	0 → WDT Prescaler A count
	0 → WDT Prescaler B count

Status Affected: None

Encoding:

1111

1110

0110

0000

0000

0000

Description: Clear the contents of the Watchdog Timer Count register and the Prescaler Count registers. The Watchdog Prescaler A and Prescaler B settings, set by Configuration fuses in the FWDT, are not changed.

Words: 1

Cycles: 1

Example 1: CLRWDT ; Clear Watchdog Timer

Before Instruction
SR	0000

After Instruction
SR	0000

COM

Complement f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} COM{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) destination designated by D

Status Affected: N, Z

Encoding:

1110

1110

1BDF

ffff

ffff

ffff

Description:

Compute the 1's complement of the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: COM.b RAM200 ; COM RAM200 (Byte mode)

Before Instruction RAM200 80FF SR 0000 After Instruction RAM200 8000 SR 0002 (Z)

Example 2: COM RAM400, WREG ; COM RAM400 and store to WREG ; (Word mode)

COM

Complement Ws

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} COM{.B} Ws, Wd

$$ [ \mathrm{Ws} ], \quad [ \mathrm{Wd} ] $$

$$ [ W s + + ], \quad [ W d + + ] $$

$$ [ \mathrm{Ws} - ], \quad [ \mathrm{Wd} - ] $$

$$ [ + + \mathrm{Ws} ], [ + + \mathrm{Wd} ] $$

$$ [ - - W s ], \quad [ - - W d ] $$

Operands: Ws ∈ [W0 ... W15]

$$ \mathrm{Wd} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Ws) Wd

Status Affected: N, Z

Encoding: 1110 1010 1Bqq qddd dppp ssss

Description: Compute the 1's complement of the contents of the source register Ws and place the result in the destination register Wd. Either Register Direct or Indirect Addressing may be used for both Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	COM.B [W0++, [W1++]	; COM [W0] and store to [W1] (Byte mode)
		; Post-increment W0, W1

Before Instruction		After Instruction
W0	2301	W0	2302
W1	2400	W1	2401
Data 2300	5607	Data 2300	5607
Data 2400	ABCD	Data 2400	ABA9
SR	0000	SR	0008 (N = 1)

Example 2: COM W0, [W1++] ; COM W0 and store to [W1] (Word mode) ; Post-increment W1

Before

Instruction

W0 D004 W0 D004

W1 1000 W1 1002

Data 1000 ABA9 Data 1000 2FFB

SR 0000 SR 0000

After

Instruction

CP

Compare f with WREG, Set Status Flags

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} CP{.B} f

Operands: f [0 8191]

Operation: (f) - (WREG)

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0011

0B0f

ffff

ffff

ffff

Description: Compute (f) – (WREG) and update the STATUS Register. This instruction is equivalent to the SUBWF instruction, but the result of the subtraction is not stored.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

CP.B

RAM400

; Compare RAM400 with WREG (Byte mode)

	Before Instruction
WREG	8823
RAM400	0823
SR	0000

After Instruction
WREG	8823
RAM400	0823
SR	0003 (C = 1)

Example 2:

CP

0x1200

; Compare (0x1200) with WREG (Word mode)

	Before Instruction
WREG	2377
Data 1200	2277
SR	0000

	After Instruction
WREG	2377
Data 1200	2277
SR	0008 (N = 1)

CP

Compare Wb with lit5, Set Status Flags

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X		X

Syntax: {label:} CP{.B} Wb, #lit5

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Operation: (Wb) - lit5

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0001

0www

wB00

011k

kkkk

Description: Compute (Wb) – lit5 and update the STATUS Register. This instruction is equivalent to the SUB instruction, but the result of the subtraction is not stored. Register Direct Addressing must be used for Wb.

The 'w' bits select the address of the Wb Base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: CP.B W4, #0x12 ; Compare W4 with 0x12 (Byte mode)

Example 2: CP W4, #0x12 ; Compare W4 with 0x12 (Word mode)

CP

Compare Wb with lit8, Set Status Flags

Implemented in: PIC24F PIC24	H PIC24E ds	sPIC30F dsP	IC33F dsPIC	c33E dsPIC3	3C
			X			X	X

Syntax: {label:} CP{.B} Wb, #lit8

Operands: Wb ∈ [W0 ... W15]

lit8 ∈ [0 ... 255]

Operation: (Wb) - lit8

Status Affected:

DC, N, OV, Z, C

Encoding:

1110

0001

0www

wBkk

k11k

kkkk

Description:

Compute (Wb) – lit8 and update the STATUS Register. This instruction is equivalent to the SUB instruction, but the result of the subtraction is not stored. Register Direct Addressing must be used for Wb.

The 'w' bits select the address of the Wb Base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: CP.B W4, #0x12 ; Compare W4 with 0x12 (Byte mode)

Before
Instruction

W4	7711
SR	0000

After
Instruction

W4	7711
SR	0009 (N, C = 1)

Example 2: CP W4, #0x12 ; Compare W4 with 0x12 (Word mode)

Before
Instruction

W4	7713
SR	0000

After
Instruction

W4	7713
SR	0001 (C = 1)

CP

Compare Wb with Ws, Set Status Flags

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsP	C33E dsPIC	33C
X	X		X	X		X

X

Syntax: {label:} CP{.B} Wb, Ws
[Ws]
[Ws++]
[Ws--]
[++Ws]
[--Ws]
Operands: Wb ∈ [W0 ... W15]
Ws ∈ [W0 ... W15]
Operation: (Wb) - (Ws)
Status Affected: DC, N, OV, Z, C

Encoding:

1110

0001

0www

wB00

0ppp

ssss

Description: Compute (Wb) – (Ws) and update the STATUS Register. This instruction is equivalent to the SUB instruction, but the result of the subtraction is not stored. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing may be used for Ws.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'p' bits select the source addressing mode.
The 's' bits select the address of the Ws source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1
Cycles: 1 ^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	CP.B W0, [W1++]	; Compare [W1] with W0 (Byte mode)
		; Post-increment W1

	Before Instruction
W0	ABA9
W1	2000
Data 2000	D004
SR	0000

After Instruction
W0	ABA9
W1	2001
Data 2000	D004
SR	0009 (N, C = 1)

Example 2:

CP

W5, W6

; Compare W6 with W5 (Word mode)

Before Instruction
W5	2334
W6	8001
SR	0000

After Instruction
W5	2334
W6	8001
SR	000C (N, OV = 1)

CP0

Compare f with 0x0, Set Status Flags

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} CP0{.B} f

Operands: f [0 8191]

Operation: (f) - 0x0

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0010

0B0f

ffff

ffff

ffff

Description:

Compute (f) - 0x0 and update the STATUS Register. The result of the subtraction is not stored.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'f' bits select the address of the file register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: CP0.B RAM100 ; Compare RAM100 with 0x0 (Byte mode)

After Instruction RAM100 44C3 SR 0009 (N, C = 1)

Example 2: CP0 0x1FFE ; Compare (0x1FFE) with 0x0 (Word mode)

After Instruction Data 1FFE 0001 SR 0001 (C = 1)

CP0

Compare Ws with 0x0, Set Status Flags

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
X	X		X	X		X

X

Syntax: {label:} CP0{.B} Ws

[Ws]

[Ws++]

[Ws--]

[++Ws]

[--Ws]

Operands: Ws ∈ [W0 ... W15]

Operation: (Ws) - 0x0000

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0000

0000

0B00

0ppp

ssss

Description:

Compute (Ws) - 0x0000 and update the STATUS Register. The result of the subtraction is not stored. Register Direct or Indirect Addressing may be used for Ws.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'p' bits select the source addressing mode.

The 's' bits select the address of the Ws source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	CP0.B [W4--]	; Compare [W4] with 0 (Byte mode)
		; Post-decrement W4

	Before Instruction
W4	1001
Data 1000	0034
SR	0000

AfterInstruction
W4	1000
Data 1000	0034
SR	0001 (C = 1)

Example 2:

CPO

[--W5]

; Compare [--W5] with 0 (Word mode)

	Before Instruction
W5	2400
Data 23FE	9000
SR	0000

After Instruction
W5	23FE
Data 23FE	9000
SR	0009 (N, C = 1)

CPB

Compare f with WREG Using Borrow, Set Status Flags

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} CPB{.B} f

Operands: f [0 8191]

Operation: (f) - (WREG) - (C)

Status Affected:

Encoding:

DC, N, OV, Z, C

1110

0011

1B0f

ffff

ffff

ffff

Description:

Compute (f) - (WREG) - (C) and update the STATUS Register. This instruction is equivalent to the SUBB instruction, but the result of the subtraction is not stored.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: CPB.B RAM400 ; Compare RAM400 with WREG using C (Byte mode)

	Before Instruction
WREG	8823
RAM400	0823
SR	0000

After Instruction
WREG	8823
RAM400	0823
SR	0008 (N = 1)

Example 2: CPB 0x1200 ; Compare (0x1200) with WREG using C (Word mode)

	Before Instruction		After Instruction
WREG	2377		WREG	2377
Data 1200	2377		Data 1200	2377
SR	0001	(C = 1)	SR	0001

CPB

Compare Wb with lit5 Using Borrow, Set Status Flags

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
	X	X		X		X

Syntax: {label:} CPB{.B} Wb, #lit5

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 5 \in [ 0 \dots 3 1 ] $$

Operation: (Wb) - lit5 - (C

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0001

1www

wB00

011k

kkkk

Description:

Compute (Wb) - lit5 - (C) and update the STATUS Register. This instruction is equivalent to the SUBB instruction, but the result of the subtraction is not stored.

Register Direct Addressing must be used for Wb.

The 'w' bits select the address of the Wb source register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits provide the literal operand, a five-bit integer number.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: CPB.B W4, #0x12 ; Compare W4 with 0x12 using C (Byte mode)

Before
Instruction

After
Instruction

Example 2: CPB.B W4, #0x12 ; Compare W4 with 0x12 using C (Byte mode)

Before
Instruction

After
Instruction

Example 3: CPB W12, #0x1F ; Compare W12 with 0x1F using C (Word mode)

Before
Instruction

After
Instruction

Example 4: CPB W12, #0x1F ; Compare W12 with 0x1F using C (Word mode)

Before
Instruction

After
Instruction

CPB

Compare Wb with lit8 Using Borrow, Set Status Flags

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F dsP	IC33F dsPIC	C33E dsPIC3	3C
			X			X	X

Syntax: {label:} CPB{.B} Wb, #lit8

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 8 \in [ 0 \dots 2 5 5 ] $$

Operation: (Wb) - lit8 - (C

Status Affected: DC, N, OV, Z, C

Encoding:

1110

0001

1www

wBkk

kllk

kkkk

Description:

Compute (Wb) - lit8 - (C) and update the STATUS Register. This instruction is equivalent to the SUBB instruction, but the result of the subtraction is not stored.

Register Direct Addressing must be used for Wb.

The 'w' bits select the address of the Wb register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits provide the literal operand, a five-bit integer number.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: CPB.B W4, #0x12 ; Compare W4 with 0x12 using C (Byte mode)

Before
Instruction

W4	7711
SR	0001(C = 1)

After
Instruction

W4	7711
SR	0008 (N = 1)

Example 2: CPB.B W4, #0x12 ; Compare W4 with 0x12 using C (Byte mode)

Before
Instruction

W4	7711
SR	0000

After
Instruction

W4	7711
SR	0008 (N = 1)

Example 3: CPB W12, #0x1F ; Compare W12 with 0x1F using C (Word mode)

Before
Instruction

W12	0020
SR	0002(Z = 1)

After
Instruction

W12	0020
SR	0003 (Z, C = 1)

Example 4: CPB W12, #0x1F ; Compare W12 with 0x1F using C (Word mode)

Before
Instruction

W12	0020
SR	0003(Z, C = 1)

After
Instruction

W12	0020
SR	0001 (C = 1)

CPB

Compare Ws with Wb Using Borrow, Set Status Flags

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
	X	X		X	X

Syntax: {label:} CPB{.B} Wb, Ws
[Ws]
[Ws++]
[Ws--]

[++Ws]
[--Ws]
Operands: Wb ∈ [W0 ... W15]
Ws ∈ [W0 ... W15]
Operation: (Wb) - (Ws) - (C)
Status Affected: DC, N, OV, Z, C

Encoding:

1110

0001

lwww

wB00

0ppp

ssss

Description: Compute (Wb) - (Ws) - (C) and update the STATUS Register. This instruction is equivalent to the SUBB instruction, but the result of the subtraction is not stored.
Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing may be used for Ws.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'p' bits select the source addressing mode.
The 's' bits select the address of the Ws source register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1
2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: CPB.B W0, [Wl++] ; Compare [Wl] with W0 using (Byte mode); Post-increment Wl

Before Instruction		After Instruction
W0	ABA9	W0	ABA9
W1	1000	W1	1001
Data 1000	D0A9	Data 1000	D0A9
SR	0002	(Z = 1)	SR

Example 2: CPB.B W0, [W1++]; Compare [W1] with W0 using C ^- (Byte mode); Post-increment W1

Before Instruction W0 ABA9 W0 ABA9 W1 1000 W1 1001 Data 1000 D0A9 Data 1000 D0A9 SR 0001 (C = 1) After Instruction SR 0001 (C = 1)

Example 3: CPB W4, W5 ; Compare W5 with W4 using C (Word mode)

Before Instruction W4 4000 W4 4000 W5 3000 W5 3000 SR 0001 (C = 1) After Instruction SR 0001 (C = 1)

CPBEQ

Compare Wb with Wn, Branch if Equal (Wb = Wn)

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F dsP	PIC33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} CPBEQ{.B} Wb, Wn, Expr

Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

If (Wb) = (Wn), [(PC+2) + 2 * Expr] → PC and NOP → Instruction Register

Status Affected:

None

Encoding:

1110

0111

1www

wBnn

nnnn

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, the PC is recalculated based on the 6-bit signed offset specified by Expr, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal (branch is not taken).

The 'w' bits select the address of the Wb source register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 's' bits select the address of the Wn source register.

The 'n' bits select the offset of the branch destination.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 (5 if branch taken)

Example 1:	002000 HERE: CPBEQ.B W0, W1, BYPASS ; If W0 = W1 (Byte mode)
	002002 ADD W2, W3, W4 ; Perform branch to BYPASS
	002004 ...
	002006 ...
	002008 BYPASS: ...
	00200A ...

Before
Instruction

PC	00 2000
W0	1000
W1	1000
SR	0000

After
Instruction

PC	00 2008
W0	1000
W1	1000
SR	0002 (Z = 1)

CPBGT

Signed Compare Wb with Wn, Branch if Greater Than (Wb > Wn)

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	PIC33F ds	PIC33C	3E ds	PIC33C
			X				X	X

Syntax: {label:} CPBGT{.B} Wb, Wn, Expr
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { If } (W b) = (W n), [ (P C + 2) + 2 * \text { Expr } ] \rightarrow P C \text { and } N O P \rightarrow I n s t r u c t i o n R e g i s t e r $$

Status Affected:

None

Encoding:

1110

0110

0www

wBnn

nnnn

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, the PC is recalculated based on the 6-bit signed offset specified by Expr, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal (branch is not taken).
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
The 'n' bits select the offset of the branch destination.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (5 if branch taken)

Example 1:	002000 HERE:	CPBGT.B W0, W1, BYPASS ; If W0 > W1 (Byte mode),
	002002	ADD W2, W3, W4 ; Perform branch to BYPASS
	002004	...
	002006	...
	002008 BYPASS	...
	00200A	...

Before
Instruction

PC	00 2000
W0	30FF
W1	26FE
SR	0000

After
Instruction

PC	00 2008
W0	00FF
W1	26FE
SR	0000 (N, C = 0)

CPBLT

Signed Compare Wb with Wn, Branch if Less Than (Wb < Wn)

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
			X			X	X

Syntax: {label:} CPBLT{.B} Wb, Wn, Expr
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)
If (Wb) = (Wn), [(PC+2) + 2 * Expr] PC and NOP → Instruction Register
Status Affected: None

Encoding:

1110

0110

1www

wBnn

nnnn

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, the PC is recalculated based on the 6-bit signed offset specified by Expr, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal (branch is not taken).
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
The 'n' bits select the offset of the branch destination.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (5 if branch taken)

Example 1:	002000 HERE:	CPBLT.B	W8, W9, BYPASS	; If W8 < W9 (Byte mode),
	002002	ADD	W2, W3, W4	; Perform branch to BYPASS
	002004	...
	002006	...
	002008 BYPASS:	...
	00200A	...

Before
Instruction

PC	00 2000
W8	00FF
W9	26FE
SR	0000

After
Instruction

PC	00 2008
W8	00FF
W9	26FE
SR	0008 (N = 1)

CPBNE

Compare Wb with Wn, Branch if Not Equal (Wb ≠ Wn)

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsPIC33E	dsPIC33C
		X			X

Syntax: {label:} CPBNE{.B} Wb, Wn, Expr

Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

If (Wb) = (Wn), [(PC+2) + 2 * Expr] → PC and NOP → Instruction Register

Status Affected:

None

Encoding:

1110

0111

0www

wBnn

nnnn

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, the PC is recalculated based on the 6-bit signed offset specified by Expr, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal (branch is not taken).

The 'w' bits select the address of the Wb source register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 's' bits select the address of the Wn source register.

The 'n' bits select the offset of the branch destination.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 (5 if branch taken)

Example 1:	002000 HERE:	CPBNE.B	W2, W3, BYPASS	; If W2 != W3 (Byte mode),
	002002	ADD	W2, W3, W4	; Perform branch to BYPASS
	002004	...
	002006	...
	002008 BYPASS:	...
	00200A	...

	Before Instruction
PC	00 2000
W2	00FF
W3	26FE
SR	0000

After Instruction
PC	00 200A
W2	00FF
W3	26FE
SR	0001 (C = 1)

CPSEQ

Compare Wb with Wn, Skip if Equal (Wb = Wn)

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F ds	PIC33F dsPI	C33E dsPIC	33C
	X	X		X	X

Syntax: {label:} CPSEQ{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { Skip if } (W b) = (W n) $$

Status Affected: None

Encoding:

Description:

1110

0111

1www

wB00

0000

ssss

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:

002000 HERE: CPSEQ.B W0, W1 ; If W0 = W1 (Byte mode),002002 GOTO BYPASS ; skip the GOTO002004 ...002006 ...002008 BYPASS: ...00200A ...

	Before Instruction
PC	00 2000
W0	1001
W1	1000
SR	0000

	After Instruction
PC	00 2002
W0	1001
W1	1000
SR	0000

Example 2:	018000 HERE:	CPSEQ	W4, W8 ; If W4 = W8 (Word mode),
	018002	CALL	_FIR ; skip the subroutine call
	018006	...
	018008	...

Before Instruction
PC	01 8000
W4	3344
W8	3344
SR	0002 (Z = 1)

After Instruction
PC	01 8006
W4	3344
W8	3344
SR	0002 (Z = 1)

CPSEQ

Compare Wb with Wn, Skip if Equal (Wb = Wn)

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
			X			X

Syntax: {label:} CPSEQ{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15] Wn ∈ [W0 ... W15]
Operation: (Wb) - (Wn) Skip if (Wb) = (Wn)
Status Affected: None
Encoding: 1110 0111 1www wB00 0001 ssss
Description: Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) = (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. If (Wb) ≠ (Wn), the next instruction is executed as normal.

The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1 Cycles: 1 (2 or 3 if skip taken)

Example 1:

002000 HERE: CPSEQ.B WO, W1 ; If WO = W1 (Byte mode),002002 GOTO BYPASS ; skip the GOTO002004 ...002006 ...002008 BYPASS: ...00200A ...

	Before Instruction
PC	00 2000
W0	1001
W1	1000
SR	0000

	After Instruction
PC	00 2002
W0	1001
W1	1000
SR	0000

Example 2:	018000 HERE:	CPSEQ	W4, W8	; If W4 = W8 (Word mode),
	018002	CALL	_FIR ; skip the subroutine call
	018006	...
	018008	...

Before Instruction
PC	01 8000
W4	3344
W8	3344
SR	0002 (Z = 1)

After Instruction
PC	01 8006
W4	3344
W8	3344
SR	0002 (Z = 1)

CPSGT

Signed Compare Wb with Wn, Skip if Greater Than (Wb > Wn)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} CPSGT{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { Skip if } (W b) > (W n) $$

Status Affected:

None

Encoding:

1110

0110

0www

wB00

0000

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) > (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000	HERE:	CPSGT.B	W0, W1	; If W0 > W1 (Byte mode),
	002002		GOTO	BYPASS	; skip the GOTO
	002006		...
	002008		...
	00200A	BYPASS	...
	00200C		...

Before

Instruction
PC	00 2000
W0	00FF
W1	26FE
SR	0009 (N, C = 1)

After

Instruction
PC	00 2006
W0	00FF
W1	26FE
SR	0009(N, C = 1)

Example 2:	018000 HERE:	CPSGT	W4, W5	; If W4 > W5 (Word mode),
	018002	CALL	_FIR	; skip the subroutine call
	018006	...
	018008	...

Before

Instruction
PC	01 8000
W4	2600
W5	2600
SR	0004 (OV = 1)

After

Instruction
PC 01 8002
W4	2600
W5	2600
SR	0004 (OV = 1)

CPSGT

Signed Compare Wb with Wn, Skip if Greater Than (Wb > Wn)

Implemented in: PIC24F PIC24	H PIC24E ds	sPIC30F dsPIC	C33F dsPIC	33E dsPIC33C
			X			X	X

Syntax: {label:} CPSGT{.B} Wb, Wn

Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

Skip if (Wb) > (Wn)

Status Affected:

None

Encoding:

1110

0110

0www

wB00

0001

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) – (Wn), but do not store the result. If (Wb) > (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000 HERE:	CPSGT.B	W0, W1	; If W0 > W1 (Byte mode),
	002002	GOTO	BYPASS	; skip the GOTO
	002006	...
	002008	...
	00200A BYPASS	...
	00200C	...

Before

	Instruction		Instruction
PC	00 2000		PC 00 2006
W0	00FF		W0 00FF
W1	26FE		26FE
SR	0009	(N, C = 1)	SR 0009 (N, C = 1)

Example 2:	018000 HERE:	CPSGT	W4, W5	; If W4 > W5 (Word mode),
	018002	CALL	_FIR	; skip the subroutine call
	018006	...
	018008	...

Before

	Instruction		Instruction
PC	01 8000		PC	01 8002
W4	2600		W4 2600
W5	2600		W5 2600
SR	0004	(OV = 1)	SR	0004

CPSLT

Signed Compare Wb with Wn, Skip if Less Than (Wb < Wn)

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F dsP	IC33F dsPIC	C33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} CPSLT{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15] Wn ∈ [W0 ... W15]

Operation: (Wb) - (Wn) Skip if (Wb) < (Wn)
Status Affected: None

Encoding:

1110

0110

1www

wB00

0000

ssss

Description: Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) < (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000 HERE: CPSLT.B W8, W9 ; If W8 < W9 (Byte mode),
	002002 GOTO BYPASS ; skip the GOTO
	002006 ...
	002008 ...
	00200A BYPASS: ...
	00200C ...

Before Instruction		After Instruction
PC	00 2000	PC	00 2002
W8	00FF	W8	00FF
W9	26FE	W9	26FE
SR	0008 (N = 1)	SR	0008 (N = 1)

Example 2:	018000 HERE:	CPSLT W3, W6 ; If W3 < W6 (Word mode),
	018002	CALL _FIR ; skip the subroutine call
	018006	...
	018008	...

Before Instruction
PC	01 8000
W3	2600
W6	3000
SR	0000

	After Instruction
PC	01 8006
W3	2600
W6	3000
SR	0000

CPSLT

Signed Compare Wb with Wn, Skip if Less Than (Wb < Wn)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax: {label:} CPSLT{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { Skip if } (W b) < (W n) $$

Status Affected:

None

Encoding:

1110

0110

1www

wB00

0001

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) < (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000 HERE:	CPSLT.B W8, W9 ; If W8 < W9 (Byte mode),
	002002	GOTO BYPASS ; skip the GOTO
	002006	...
	002008	...
	00200A BYPASS:	...
	00200C	...

Before Instruction
PC	00 2000
W8	00FF
W9	26FE
SR	0008 (N = 1)

After Instruction
PC	00 2002
W8	00FF
W9	26FE
SR	0008 (N = 1)

Example 2:	018000	HERE:	CPSLT	W3, W6	; If W3 < W6 (Word mode),
	018002		CALL _FIR	; skip the subroutine call
	018006		...
	018008		...

	Before Instruction
PC	01 8000
W3	2600
W6	3000
SR	0000

	After Instruction
PC	01 8006
W3	2600
W6	3000
SR	0000

CPSNE

Signed Compare Wb with Wn, Skip if Not Equal (Wb ≠ Wn)

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} CPSNE{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { Skip if } (W b) \neq (W n) $$

Status Affected: None

Encoding:

Description:

1110

0111

0www

wB00

0000

ssss

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) ≠ (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000 HERE: PSNE.B W2, W3 ; If W2 != W3 (Byte mode),
	002002 GOTO BYPASS ; skip the GOTO
	002006 ...
	002008 ...
	00200A BYPASS: ...
	00200C ...

Before Instruction
PC	00 2000
W2	00FF
W3	26FE
SR	0001 (C = 1)

After Instruction
PC	00 2006
W2	00FF
W3	26FE
SR	0001 (C = 1)

Example 2:	018000 HERE:	CPSNE WO, W8	; If WO != W8 (Word mode),
	018002	CALL _FIR	; skip the subroutine call
	018006	...
	018008	...

	Before Instruction
PC	01 8000
W0	3000
W8	3000
SR	0000

	After Instruction
PC	01 8002
W0	3000
W8	3000
SR	0000

CPSNE

Signed Compare Wb with Wn, Skip if Not Equal (Wb ≠ Wn)

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax: {label:} CPSNE{.B} Wb, Wn
Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: (Wb) - (Wn)

$$ \text { Skip if } (W b) \neq (W n) $$

Status Affected:

None

Encoding:

1110

0111

0www

wB00

0001

ssss

Description:

Compare the contents of Wb with the contents of Wn by performing the subtraction, (Wb) - (Wn), but do not store the result. If (Wb) ≠ (Wn), the next instruction (fetched during the current instruction execution) is discarded, and on the next cycle, a NOP is executed instead. Otherwise, the next instruction is executed as normal.
The 'w' bits select the address of the Wb source register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 's' bits select the address of the Wn source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1 (2 or 3 if skip taken)

Example 1:	002000 HERE: CPSNE.B W2, W3 ; If W2 != W3 (Byte mode),
	002002	GOTO	BYPASS ; skip the GOTO
	002006	...
	002008	...
	00200A BYPASS: ...
	00200C	...

Before Instruction
PC	00 2000
W2	00FF
W3	26FE
SR	0001 (C = 1)

After Instruction
PC	00 2006
W2	00FF
W3	26FE
SR	0001 (C = 1)

Example 2:	018000 HERE:	CPSNE	W0, W8	; If W0 != W8 (Word mode),
	018002	CALL	_FIR	; skip the subroutine call
	018006	...
	018008	...

	Before Instruction
PC	01 8000
W0	3000
W8	3000
SR	0000

	After Instruction
PC	01 8002
W0	3000
W8	3000
SR	0000

CTXTSWP ^(1)

CPU Register Context Swap Literal

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPI	C33E dsPIC3	3C
						X	X

Syntax: {label:} CTXTSWP #lit3

Operands: lit3 ∈ [0 ... 4]

Operation: If context defined by lit3 is valid,

Then

Switch CPU register context to context defined by lit3

Else

Execute as 2-cycle NOP

Status Affected: None

Encoding:

1111

1110

1110

0000

0000

0kkk

Description:

This instruction will force a CPU register context switch (W0 through W14, and Accumulators A and B) from the current context to the target context defined by the value defined by #lit3. If the specified context is not implemented on the device, this instruction will execute as a 2-cycle NOP.

A successful context switch will update the current context identifier and the manual context identifier (held in CCTXI<2:0> (CTXTSTAT<10:8>) and MCTXI<2:0> (CTXTSTAT<2:0>), respectively) to reflect the new active CPU register context.

Words: 1

Cycles: 2

Note 1: This instruction is present only in some devices of the device families. Please see the specific device data sheet to ensure that this instruction is supported on a specific device.

CTXTSWP ^(1)

CPU Register Context Swap Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
					X	X

Syntax: {label:} CTXTSWP Wn

Operands: Wn ∈ [W0 ... W15]

Operation: If context defined by the contents of Wn<2:0> is valid,

Then

Switch CPU register context to context defined by the contents of Wn<2:0>se

Execute as 2-cycle NOP

Status Affected: None

Encoding:

1111

1110

1111

0000

0000

ssss

Description:

This instruction will force a CPU register context switch (W0 through W14, and Accumulators A and B) from the current context to the target context defined by the value in the three Least Significant bits of Wn. If the specified context is not implemented on the device, this instruction will execute as a 2-cycle NOP.

A successful context switch will update the current context identifier and the manual context identifier (held in CCTXI<2:0> (CTXTSTAT<10:8>) and MCTXI<2:0> (CTXTSTAT<2:0>), respectively) to reflect the new active CPU register context.

Words: 1

Cycles: 2

Note 1: This instruction is present only in some devices of the device families. Please see the specific device data sheet to ensure that this instruction is supported on a specific device.

DAW.B

Decimal Adjust Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
X	X		X	X		X

X

Syntax: {label:} DAW.B Wn

Operands: Wn ∈ [W0 ... W15]

Operation: If (Wn<3:0>>9) or (DC=1)

$$ (W n < 3: 0 >) + 6 \rightarrow W n < 3: 0 > $$

Else

$$ (W n < 3: 0 >) \rightarrow W n < 3: 0 > $$

$$ \text { If } (W n < 7: 4 > > 9) \text { or } (C = 1) $$

$$ (W n < 7: 4 >) + 6 \rightarrow W n < 7: 4 > $$

Else

$$ (W n < 7: 4 >) \rightarrow W n < 7: 4 > $$

Status Affected: C

Encoding:

Description:

1111

1101

0100

0000

0000

ssss

Adjust the Least Significant Byte in Wn to produce a Binary Coded Decimal (BCD) result. The Most Significant Byte of Wn is not changed and the Carry flag is used to indicate any decimal rollover. Register Direct Addressing must be used for Wn.

The 's' bits select the source/destination register.

Note 1: This instruction is used to correct the data format after two packed BCD bytes have been added.

2: This instruction operates in Byte mode only and the .B extension must be included with the opcode.

Words: 1

Cycles: 1

Example 1: DAW.B W0 ; Decimal adjust W0

Before

Instruction

W0	771A
SR	0002(DC=1)

After

Instruction

W0	7720
SR	0002(DC = 1)

Example 2: DAW.B W3 ; Decimal adjust W3

Before

Instruction

W3	77AA
SR	0000

After

Instruction

W3	7710
SR	0001(C = 1)

DEC

Decrement f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} DEC{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) - 1 → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding: 1110 1101 0BDF ffff ffff ffff

Description: Subtract one from the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: DEC.B 0x200 ; Decrement (0x200) (Byte mode)

After Instruction Data 200 □ 80FE SR □ 0009 (N, C = 1)

Example 2: DEC RAM400, WREG ; Decrement RAM400 and store to WREG ; (Word mode)

DEC

Decrement Ws

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} DEC{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Ws) - 1 → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

1110

1001

0Bqq

qddd

dppp

ssss

Description: Subtract one from the contents of the source register Ws and place the result in the destination register Wd. Either Register Direct or Indirect Addressing may be used by Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

DEC.B [W7++, [W8++] ; DEC [W7] and store to [W8] (Byte mode) ; Post-increment W7, W8

	Before Instruction		After Instruction
W7	2301	W7	2302
W8	2400	W8	2401
Data 2300	5607	Data 2300	5607
Data 2400	ABCD	Data 2400	AB55
SR	0000	SR	0000

Example 2: DEC W5, [W6++] ; Decrement W5 and store to [W6] (Word mode) ; Post-increment W6

	Before Instruction				After Instruction
	W5	D004	W5	D004
	W6	2000	W6	2002
	Data 2000	ABA9	Data 2000	D003
	SR	0000	SR	0009 (N, C = 1)

DEC2

Decrement f by 2

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} DEC2{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) - 2 → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding: 1110 1101 lBDf ffff ffff ffff

Description: Subtract two from the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: DEC2.B 0x200 ; Decrement (0x200) by 2 (Byte mode)

After Instruction Data 200 80FD SR 0009 (N, C = 1)

Example 2: DEC2 RAM400, WREG ; Decrement RAM400 by 2 and ; store to WREG (Word mode)

DEC2

Decrement Ws by 2

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} DEC2{.B} Ws, Wd
[Ws], [Wd]
[Ws++], [Wd++]
[Ws--], [Wd--]

[++Ws], [++Wd]
[--Ws], [--Wd]
Operands: Ws ∈ [W0 ... W15]
Wd ∈ [W0 ... W15]
Operation: (Ws) - 2 → Wd
Status Affected: DC, N, OV, Z, C

Encoding:

1110

1001

1Bqq

qddd

dppp

ssss

Description: Subtract two from the contents of the source register Ws and place the result in the destination register Wd. Either Register Direct or Indirect Addressing may be used by Ws and Wd.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'q' bits select the destination addressing mode.
The 'd' bits select the destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: DEC2.B [W7--], [W8--] ; DEC [W7] by 2, store to [W8] (Byte mode) ; Post-decrement W7, W8

Before Instruction		After Instruction
W7	2301	W7	2300
W8	2400	W8	23FF
Data 2300	0107	Data 2300	0107
Data 2400	ABCD	Data 2400	ABFF
SR	0000	SR	0008 (N = 1)

Example 2: DEC2 W5, [W6++] ; DEC W5 by 2, store to [W6] (Word mode) ; Post-increment W6

	Before Instruction				After Instruction
	W5	D004	W5	D004
	W6	1000	W6	1002
	Data 1000	ABA9		Data 1000 D002
	SR	0000	SR	0009 (N, C = 1)

DISI

Disable Interrupts Temporarily

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} DISI #lit14
Operands: lit14 ∈ [0 ... 16383]
Operation: lit14 → DISICNT
1 → DISI
Disable interrupts for (lit14 + 1) cycles

Status Affected:

None

Encoding:

1111

1100

00kk

kkkk

kkkk

kkkk

Description:

Disable interrupts of Priority 0 through Priority 6 for (lit14 + 1) instruction cycles. Priority 0 through Priority 6 interrupts are disabled, starting in the cycle that DISI executes, and remain disabled for the next (lit 14) cycles. The lit14 value is written to the DISICNT register and the DISI flag (INTCON2<14>) is set to '1'. This instruction can be used before executing time-critical code to limit the effects of interrupts.
Note 1: This instruction does not prevent Priority 7 interrupts and traps from running. See the specific device family reference manual for details.
2: This instruction does not prevent any interrupts when the device is in Sleep mode.
Words: 1
Cycles: 1

Example 1:	002000	HERE:	DISI	#100	; Disable interrupts for 101 cycles
	002002				; next 100 cycles protected by DISI
	002004		...

	Before Instruction
PC	00 2000
DISICNT	0000
INTCON2	0000
SR	0000

	After Instruction
PC	00 2002
DISICNT	0100
INTCON2	4000
SR	0000

DIV.S
Signed Integer Divide

Implemented in: PIC24F	PIC24H PIC2	4E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X

Syntax:	{label:}	DIV.S{W}	Wm, Wn
			DIV.SD	Wm, Wn

Operands:	Wm ∈ [W0 ... W15] for word operation
	Wm ∈ [W0, W2, W4 ... W14] for double operation
	Wn ∈ [W2 ... W15]

Operation:	For Word Operation (default):
	Wm → W0
	If (Wm<15> = 1):
	0xFFFF → W1

Else:

0x0 → W1

W1:W0/Wn → W0

Remainder → W1

For Double Operation (DIV.SD):

Wm + 1:Wm → W1:W0

W1:W0/Wn → W0

Remainder → W1

Status Affected:

N, OV, Z, C

Encoding:

1101

1000

0ttt

tvvv

vW00

ssss

Description: Iterative, signed integer divide, where the dividend is stored in Wm (for a 16-bit by 16-bit divide) or Wm + 1:Wm (for a 32-bit by 16-bit divide) and the divisor is stored in Wn. In the default word operation, Wm is first copied to W0 and sign-extended through W1 to perform the operation. In the double operation, Wm + 1:Wm is first copied to W1:W0. The 16-bit quotient of the divide operation is stored in W0 and the 16-bit remainder is stored in W1.

This instruction must be executed 18 times using the REPEAT instruction (with an iteration count of 17) to generate the correct quotient and remainder. The N flag will be set if the remainder is negative and cleared otherwise. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.

The 't' bits select the most significant word of the dividend for the double operation. These bits are clear for the word operation. The 'v' bits select the least significant word of the dividend. The 'W' bit selects the dividend size ('0' for 16-bit, '1' for 32-bit).

Note 1: The 's' bits select the Divisor register. The extension .D in the instruction denotes a double-word (32-bit) dividend rather than a word dividend. You may use a .W extension to denote a word operation, but it is not required.
2: Unexpected results will occur if the quotient can not be represented in 16 bits. When this occurs for the double operation (DIV.SD), the OV Status bit will be set and the quotient and remainder should not be used. For the word operation (DIV.S), only one type of overflow may occur (0x8000/0xFFFF = +32768 or 0x00008000), which allows the OV Status bit to interpret the result.
3: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution. 4: This instruction is interruptible on each instruction cycle boundary.

Words: 1

Cycles: 18 (plus 1 for REPEAT execution) for PIC24F, PIC24H, PIC24E, dsPIC30F, dsPIC33F, dsPIC33E

6 (plus 1 for REPEAT execution) for dsPIC33C

Example 1:

REPEAT #17	; Execute DIV.S 18 times
DIV.S W3, W4	; Divide W3 by W4
	; Store quotient to W0, remainder to W1

Before
Instruction

W0	5555 W0	013B
W1	1234 W1	0003
W3	3000 W3	3000
W4	0027 W4	0027
SR	0000 SR	0000

After
Instruction

Example 2:

REPEAT #17	; Execute DIV.SD 18 times
DIV.SD W0, W12	; Divide W1:W0 by W12
	; Store quotient to W0, remainder to W1

Before
Instruction

W0	2500 W0	FA6B
W1	-F42 W1	EF00
W12	2200 W12	2200
SR	0000 SR	0008 (N = 1)

After
Instruction

DIV.U

Unsigned Integer Divide

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} DIV.U{W} Wm, Wn

DIV.UD Wm, Wn

Operands: Wm ∈ [W0 ... W15] for word operation

Wm ∈ [W0, W2, W4 ... W14] for double operation

Wn ∈ [W2 ... W15]

Operation: For Word Operation (default):

Wm → W0

0x0 → W1

W1:W0/Wn → W0

Remainder → W1

For Double Operation (DIV.UD):

Wm + 1:Wm → W1:W0

W1:W0/Wns → W0

Remainder → W1

Status Affected: N, OV, Z, C

Encoding:

1101

1000

1ttt

tvvv

vW00

ssss

Description:

Iterative, unsigned integer divide, where the dividend is stored in Wm (for a 16-bit by 16-bit divide) or Wm + 1:Wm (for a 32-bit by 16-bit divide) and the divisor is stored in Wn. In the word operation, Wm is first copied to W0 and W1 is cleared to perform the divide. In the double operation, Wm + 1:Wm is first copied to W1:W0. The 16-bit quotient of the divide operation is stored in W0 and the 16-bit remainder is stored in W1.

This instruction must be executed 18 times using the REPEAT instruction (with an iteration count of 17) to generate the correct quotient and remainder. The N flag will always be cleared. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.

The 't' bits select the most significant word of the dividend for the double operation. These bits are clear for the word operation.

The 'v' bits select the least significant word of the dividend.

The 'W' bit selects the dividend size ('0' for 16-bit, '1' for 32-bit).

The 's' bits select the Divisor register.

Note 1: The extension .D in the instruction denotes a double-word (32-bit) dividend rather than a word dividend. You may use a .W extension to denote a word operation, but it is not required.

2: Unexpected results will occur if the quotient can not be represented in 16 bits. This may only occur for the double operation (DIV.UD). When an overflow occurs, the OV Status bit will be set, and the quotient and remainder should not be used.
3: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution.
4: This instruction is interruptible on each instruction cycle boundary.

Words: 1

Cycles: 18 (plus 1 for REPEAT execution) for PIC24F, PIC24H, PIC24E, dsPIC30F, dsPIC33F, dsPIC33E

6 (plus 1 for REPEAT execution) for dsPIC33C

DIVF

Fractional Divide

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} DIVF Wm, Wn

Operands: Wm ∈ [W0 ... W15]

$$ \mathrm{Wn} \in [ \mathrm{W2} \dots \mathrm{W15} ] $$

Operation: 0x0 → W0

Iterative, signed fractional 16-bit by 16-bit divide, where the dividend is stored in Wm and the divisor is stored in Wn. To perform the operation, W0 is first cleared and Wm is copied to W1. The 16-bit quotient of the divide operation is stored in W0 and the 16-bit remainder is stored in W1. The sign of the remainder will be the same as the sign of the dividend.

This instruction must be executed 18 times using the REPEAT instruction (with an iteration count of 17) to generate the correct quotient and remainder. The N flag will be set if the remainder is negative and cleared otherwise. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.

The 't' bits select the Dividend register.

The 's' bits select the Divisor register.

Note 1: For the fractional divide to be effective, Wm must be less than Wn. If Wm is greater than or equal to Wn, unexpected results will occur because the fractional result will be greater than or equal to 1.0. When this occurs, the OV Status bit will be set, and the quotient and remainder should not be used.
2: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution.
3: This instruction is interruptible on each instruction cycle boundary.

Words: 1

Cycles: 18 (plus 1 for REPEAT execution) for dsPIC30F, dsPIC33F, dsPIC33E

6 (plus 1 for REPEAT execution) for dsPIC33C

Example 1:	REPEAT #17 ; Execute DIVF 18 timesDIVF W8, W9 ; Divide W8 by W9; Store quotient to W0, remainder to W1
	Before Instruction	After Instruction
	W0 8000 W0 2000
	W1 1234 W1 0000
	W8 1000 W8 1000
	W9 4000 W9 4000
	SR 0000 SR 0002 (Z = 1)
Example 2:	REPEAT #17 ; Execute DIVF 18 timesDIVF W8, W9 ; Divide W8 by W9; Store quotient to W0, remainder to W1
	Before Instruction	After Instruction
	W0 8000 W0 F000
	W1 1234 W1 0000
	W8 1000 W8 1000
	W9 8000 W9 8000
	SR 0000 SR 0002 (Z = 1)
Example 3:	REPEAT #17 ; Execute DIVF 18 timesDIVF W0, W1 ; Divide W0 by W1; Store quotient to W0, remainder to W1
	Before Instruction	After Instruction
	W0 8002 W0 7FFE
	W1 8001 W1 8002
	SR 0000 SR 0008 (N = 1)

DIVF2

Signed Fractional Divide, 16/16 (W1:W0 Preserved)

Implemented in: PIC24F P	IC24H PIC2	4E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
						X

Syntax: {label:} DIVF2 Wm, Wn
Operands: Wn ∈ [W2 ... W15]; Wm ∈ [W2 ... W15]

Operation: Wm = Dividend, Wn = Divisor: 0x0000 → W(m-1) Wm:W(m-1)/Wn → W(m-1); Remainder → Wm
Status Affected: C, N, OV, Z

Encoding:

1101

1001

Ottt

t000

0010

ssss

Description: Iterative, signed fractional 16-bit by 16-bit divide, producing a 16-bit quotient and a 16-bit remainder. The sign of the remainder will be the same as that of the dividend.
This instruction must be executed 6 times to generate the correct quotient and remainder. This may only be achieved by executing a REPEAT with an iteration count of 5 (i.e., 5+1 iterations in all) and the DIVF instruction as its target.
The N flag will be set if the remainder is negative and cleared otherwise. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.
The 's' bits select the address of the source (divisor) register. The 't' bits select the address of the source (dividend) register.

Note 1: For the fractional divide to be effective, Wm must be less than Wn. If Wm is greater than or equal to Wn, unexpected results will occur because the fractional result will be greater than or equal to 1.0. When this occurs, the OV Status bit will be set, and the quotient and remainder should not be used.
2: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution.
3: This instruction is interruptible on each instruction cycle boundary.
Words: 1
Cycles: 6 (plus 1 for REPEAT instruction execution)

Example 1:	REPEAT #17	; Execute DIV.U 18 times
	DIV.U W2, W4	; Divide W2 by W4
		; Store quotient to W0, remainder to W1

	Before Instruction
W0	5555
W1	1234
W2	8000
W4	0200
SR	0000

After Instruction
W0	0040
W1	0000
W2	8000
W4	0200
SR	0002(Z = 1)

Example 2:

REPEAT #17	; Execute DIV.UD 18 times
DIV.UD W10, W12	; Divide W11:W10 by W12
	; Store quotient to W0, remainder to W1

Before

Instruction

W0	5555 W0	01F2
W1	1234 W1	0100
W10	2500 W1	0 2500
W11	0042 W1	1 0042
W12	2200 W1	2 2200
SR	0000 SR	0000

After
Instruction

DIV2.S

Signed Integer Divide (W1:W0 Preserved)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
						X

Syntax: {label:} DIV2.S{W} Wm, Wn

DIV2.SD Wm, Wn

Operands: Wm ∈ [W0 ... W15] for word operation

Wm ∈ [W0, W2, W4 ... W14] for double operation

Wn ∈ [W2 ... W15]

Operation: For Word Operation (default):

W(m+1):Wm/Wn → Wm; Remainder → W(m+1)

For Double Operation (DIV2.SD):

W(m+1):Wm/Wn → Wm; Remainder → W(m+1)

Status Affected: C, N, OV, Z

Encoding:

1101

1000

0ttt

tvvv

v110

ssss

Description:

Iterative, signed integer 32-bit by 16-bit divide to a 16-bit quotient and a 16-bit remainder. The sign of the remainder will be the same as that of the dividend. Wm must be an even number and holds the least significant word of the dividend. The most significant word of the dividend is held in W(m+1).

This instruction must be executed 6 times to generate the correct quotient and remainder. This may only be achieved by executing a REPEAT with an iteration count of 5 (i.e., 5+1 iterations in all) and the DIV2.S instruction as its target.

The N flag will be set if the remainder is negative and cleared otherwise. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.

The 's' bits select the address of the source (divisor) register.

The 't' bits select the address of the source (dividend, most significant word) register.

The 'v' bits select the address of the source (dividend, least significant word) register.

Note 1: The extension .D in the instruction denotes a double-word (32-bit) dividend rather than a word dividend. You may use a .W extension to denote a word operation, but it is not required.

2: Unexpected results will occur if the quotient can not be represented in 16 bits. When this occurs for the double operation (DIV2.SD), the OV Status bit will be set, and the quotient and remainder should not be used. For the word operation (DIV2.S), only one type of overflow may occur (0x8000/0xFFFF = +32768 or 0x00008000), which allows the OV Status bit to interpret the result.
3: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution.
4: This instruction is interruptible on each instruction cycle boundary.

Words: 1

Cycles: 6 (plus 1 for REPEAT instruction execution)

DIV2.U

Unsigned Integer Divide (W1:W0 Preserved)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
						X

Syntax: {label:} DIV2.U{W} Wm, Wn

DIV2.UD Wm, Wn

Operands: Wm ∈ [W0 ... W15] for word operation

Wm ∈ [W0, W2, W4 ... W14] for double operation

Wn ∈ [W2 ... W15]

Operation: W(m+1):Wm = Dividend, Wn = Divisor:

W(m+1):Wm/Wn → Wm; Remainder → W(m+1)

For Word Operation (default):

0 W(m + 1)

W(m+1):Wm/Wn → Wm; Remainder → W(m+1)

For Double Operation (DIV2.SD):

W(m+1):Wm/Wn → Wm; Remainder → W(m+1)

Status Affected: C, N, OV, Z

Encoding:

1101

1000

1ttt

tvvv

v110

ssss

Description: Iterative, unsigned integer 16-bit by 16-bit or 32-bit by 16-bit divide, producing a 16-bit quotient and a 16-bit remainder. Wm must be an even number and holds the least significant word of the dividend. The most significant word of the dividend is held in W(m+1).

This instruction must be executed 6 times to generate the correct quotient and remainder. This may only be achieved by executing a REPEAT with an iteration count of 5 (i.e., 5+1 iterations in all) and the DIV.UD instruction as its target.

The N flag is always cleared. The OV flag will be set if the divide operation resulted in an overflow and cleared otherwise. The Z flag will be set if the remainder is '0' and cleared otherwise. The C flag is used to implement the divide algorithm and its final value should not be used.

The 's' bits select the address of the source (divisor) register.

The 't' bits select the address of the source (dividend, most significant word) register.

Note 1: The extension .D in the instruction denotes a double-word (32-bit) dividend rather than a word dividend. You may use a .W extension to denote a word operation, but it is not required.

2: Unexpected results will occur if the quotient can not be represented in 16 bits. This may only occur for the double operation (DIV2.UD). When an overflow occurs, the OV Status bit will be set, and the quotient and remainder should not be used.

3: Dividing by zero will initiate an arithmetic error trap during the first cycle of execution.

4: This instruction is interruptible on each instruction cycle boundary.

Words: 1

Cycles: 6 (plus 1 for REPEAT instruction execution)

DO

Initialize Hardware Loop Literal

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
				X	X

Syntax: {label:} DO #lit14, Expr

Operands: lit14 ∈ [0 ... 16383]

Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: PUSH DO shadows (DCOUNT, DOEND, DOSTART)

(lit14) → DCOUNT

(PC) + 4 → PC

(PC) → DOSTART

(PC) + (2 * Slit16) → DOEND

Increment DL<2:0> (CORCON<10:8>)

Status Affected: DA

Encoding:

DA

0000	1000	00kk	kkkk	kkkk	kkkk
0000	0000	nnnn	nnnn	nnnn	nnnn

Description: Initiate a no overhead hardware DO loop, which is executed (lit14 + 1) times. The DO loop begins at the address following the DO instruction and ends at the address 2 * Slit16 instruction words away. The 14-bit count value (lit14) supports a maximum loop count value of 16384 and the 16-bit offset value (Slit16) supports offsets of 32K instruction words in both directions.

When this instruction executes, DCOUNT, DOSTART and DOEND are first PUSHed into their respective shadow registers, and then updated with the new DO loop parameters specified by the instruction. The DO level count, DL<2:0>

(CORCON<8:10>), is then incremented. After the DO loop completes execution, the PUSHed DCOUNT, DOSTART and DOEND registers are restored, and DL<2:0> are decremented.

The 'k' bits specify the loop count.

The 'n' bits are a signed literal that specifies the number of instructions that are offset from the PC to the last instruction executed in the loop.

Special Features, Restrictions:

The following features and restrictions apply to the DO instruction.

1. Using a loop count of 0 will result in the loop being executed one time.
2. Using a loop size of -2, -1 or 0 is invalid. Unexpected results may occur if these offsets are used.
3. The very last two instructions of the DO loop cannot be:

4. an instruction which changes program control flow
   • a DO or REPEAT instruction
   Unexpected results may occur if any of these instructions are used.

5. If a hard trap occurs in the second to last instruction or third to last instruction of a DO loop, the loop will not function properly. The hard trap includes exceptions of Priority Level 13 through Level 15, inclusive.

Note 1: The DO instruction is interruptible and supports 1 level of hardware nesting. Nesting up to an additional 5 levels may be provided in software by the user. See the specific device family reference manual for details.

2: The linker will convert the specified expression into the offset to be used.

Words: 2

Cycles: 2

Example 1: 002000 LOOP6: DO #5, END6 ; Initiate DO loop (6 reps)

002004 ADD W1, W2, W3 ; First instruction in loop

002006 ...

002008 ...

00200A END6: SUB W2, W3, W4 ; Last instruction in loop

00200C ...

Before

Instruction PC 00 2000 PC 00 2004 DCOUNT 0000 DCOUNT 0005 DOSTART FF FFFF DOEND FF FFFF CORCON 0000 SR 0001 (C = 1)

After

Instruction DOSTART 00 2004 DOEND 00 200A CORCON 0100 (DL = 1) SR 0201 (DA, C = 1)

Example 2: 01C000 LOOP12: DO #0x160, END12 ; Init DO loop (353 reps)

01C004 DEC W1, W2 ; First instruction in loop

01C006 ...

01C008 ...

01C00A ...

01C00C ...

01C00E CALL _FIR88 ; Call the FIR88 subroutine

01C012 NOP

01C014 END12: NOP ; Last instruction in loop

; (Required NOP filler)

Before
Instruction

PC 01 C000 PC 01 DCOUNT 0000 DCOUNT 0160 DOSTART FF FFFF DOEND FF FFFF CORCON 0000 SR 0008 (N = 1)

After
Instruction

C004 DOSTART 01 C004 DOEND 01 C014 CORCON 0100 (DL = 1) SR 0208 (DA, N = 1)

DO

Initialize Hardware Loop Literal

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
						X	X

Syntax: {label:} DO #lit15, Expr

Operands: lit15 ∈ [0 ... 32767]

Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: PUSH DO shadows (DCOUNT, DOEND, DOSTART)

(lit15) → DCOUNT

(PC) + 4 → PC

(PC) → DOSTART

(PC) + (2 * Slit16) → DOEND

Increment DL<2:0> (CORCON<10:8>)

Status Affected: DA

Encoding:

0000	1000	0kkk	kkkk	kkkk	kkkk
0000	0000	nnnn	nnnn	nnnn	nnnn

Description: Initiate a no overhead hardware DO loop, which is executed (lit15 + 1) times. The DO loop begins at the address following the DO instruction and ends at the address 2 * Slit16 instruction words away. The 15-bit count value (lit15) supports a maximum loop count value of 32768 and the 16-bit offset value (Slit16) supports offsets of 32K instruction words in both directions.

When this instruction executes, DCOUNT, DOSTART and DOEND are first PUSHed into their respective shadow registers, and then updated with the new DO loop parameters specified by the instruction. The DO level count, DL<2:0> bits (CORCON<8:10>), is then incremented. After the DO loop completes execution, the PUSHed DCOUNT, DOSTART and DOEND registers are restored and DL<2:0> are decremented.

The 'k' bits specify the loop count.

The 'n' bits are a signed literal that specifies the number of instructions that are offset from the PC to the last instruction executed in the loop.

Special Features, Restrictions:

The following features and restrictions apply to the DO instruction.

1. Using a loop count of 0 will result in the loop being executed one time.
2. Using a loop size of -2, -1 or 0 is invalid. Unexpected results may occur if these offsets are used.
3. The very last two instructions of the DO loop cannot be:

4. an instruction which changes program control flow
   • a DO or REPEAT instruction
   Unexpected results may occur if any of these instructions are used.

5. If a hard trap occurs in the second to last instruction or third to last instruction of a DO loop, the loop will not function properly. The hard trap includes exceptions of Priority Level 13 through Level 15, inclusive.

6. The first and last instructions of the DO loop should not be a PSV read, table read or table write.

Note 1: The DO instruction is interruptible and supports 1 level of hardware nesting. Nesting up to an additional 5 levels may be provided in software by the user. See the specific device family reference manual for details.

2: The linker will convert the specified expression into the offset to be used.

Words: 2

Cycles: 2

Example 1: 002000 LOOP6: DO #5, END6 ; Initiate DO loop (6 reps) 002004 ADD W1, W2, W3 ; First instruction in loop 002006 ... 002008 ... 00200A END6: SUB W2, W3, W4 ; Last instruction in loop 00200C ...

Before Instruction PC 00 2000 PC 00 2004 DCOUNT 0000 DCOUNT 0005 DOSTART FF FFFF DOEND FF FFFF CORCON 0000 SR 0001 (C = 1) After Instruction DOSTART 00 2004 DOEND 00 200A CORCON 0100 SR 0201 (DL = 1) (DA, C = 1)

Example 2: 01C000 LOOP12: DO #0x160, END12 ; Init DO loop (353 reps) 01C004 DEC W1, W2 ; First instruction in loop 01C006 ... 01C008 ... 01C00A ... 01C00C ... 01C00E CALL _FIR88 ; Call the FIR88 subroutine 01C012 NOP 01C014 END12: NOP ; Last instruction in loop ; (Required NOP filler)

DO

Initialize Hardware Loop Wn

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsPIC33E	dsPIC33C
			X	X

Syntax: {label:} DO Wn, Expr

Operands: Wn ∈ [W0 ... W15]

Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: PUSH Shadows (DCOUNT, DOEND, DOSTART)

(Wn<13:0>) → DCOUNT

(PC) + 4 → PC

(PC) → DOSTART

(PC) + (2 * Slit16) → DOEND

Increment DL<2:0> (CORCON<10:8>)

Status Affected: DA

0000	1000	1000	0000	0000	ssss
0000	0000	nnnn	nnnn	nnnn	nnnn

Encoding:

Description:

Initiate a no overhead hardware DO loop, which is executed (Wn + 1) times. The DO loop begins at the address following the DO instruction and ends at the address 2 * Slit16 instruction words away. The lower 14 bits of Wn support a maximum count value of 16384 and the 16-bit offset value (Slit16) supports offsets of 32K instruction words in both directions.

When this instruction executes, DCOUNT, DOSTART and DOEND are first PUSHed into their respective shadow registers, and then updated with the new DO loop parameters specified by the instruction. The DO level count, DL<2:0> (CORCON<8:10>), is then incremented. After the DO loop completes execution, the PUSHed DCOUNT, DOSTART and DOEND registers are restored, and DL<2:0> are decremented.

The 's' bits specify the register Wn that contains the loop count.

The 'n' bits are a signed literal that specifies the number of instructions that are offset from (PC + 4) , which is the last instruction executed in the loop.

Special Features, Restrictions:

The following features and restrictions apply to the DO instruction.

1. Using a loop count of 0 will result in the loop being executed one time.
2. Using an offset of -2, -1 or 0 is invalid. Unexpected results may occur if these offsets are used.
3. The very last two instructions of the DO loop cannot be:
4. an instruction which changes program control flow
   • a DO or REPEAT instruction
   Unexpected results may occur if these last instructions are used.

Note 1: The DO instruction is interruptible and supports 1 level of nesting. Nesting up to an additional 5 levels may be provided in software by the user. See the specific device family reference manual for details.

2: The linker will convert the specified expression into the offset to be used.

Words: 2

Cycles: 2

Example 1:	002000 LOOP6:	DO	W0, END6	; Initiate DO loop (W0 reps)
	002004	ADD	W1, W2, W3	; First instruction in loop
	002006	...
	002008	...
	00200A	...
	00200C	REPEAT #6
	00200E	SUB	W2, W3, W4
	002010 END6:	NOP		; Last instruction in loop
				; (Required NOP filler)

	Before Instruction	After Instruction
PC	00 2000 PC 00 2004
W0	0012 W0 0012
DCOUNT	0000 DCOUNT 0012
DOSTART	FF FFFF DOSTART 00 2004
DOEND	FF FFFF DOEND 00 2010
CORCON	0000	CORCON 0100 (DL = 1)
SR	0000	SR 0080 (DA = 1)

Example 2:	002000 LOOPA:	DO	W7, ENDA	; Initiate DO loop (W7 reps)
	002004	SWAP	WO	; First instruction in loop
	002006	...
	002008	...
	00200A	...
	002010 ENDA:	MOV	W1, [W2++]	; Last instruction in loop

	Before Instruction			After Instruction
PC	00 2000 PC	00 2004
W7	E00F	W7	E00F
DCOUNT	0000 DCOUNT		200F
DOSTART	FF FFFF DOSTART 00 2004
DOEND	FF FFFF DOEND 00 2010
CORCON	0000	CORCON 0100 (DL = 1)
SR	0000 SR 0080 (DA = 1)

DO

Initialize Hardware Loop Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} DO Wn, Expr

Operands: Wn ∈ [W0 ... W15]

Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... +32767].

Operation: PUSH Shadows (DCOUNT, DOEND, DOSTART)

(Wn) → DCOUNT

(PC) + 4 → PC

(PC) → DOSTART

(PC) + (2 * Slit16) → DOEND

Increment DL<2:0> (CORCON<10:8>)

Status Affected: DA

0000	1000	1000	0000	0000	ssss
0000	0000	nnnn	nnnn	nnnn	nnnn

Encoding:

Description:

Initiate a no overhead hardware DO loop, which is executed (Wn + 1) times. The DO loop begins at the address following the DO instruction and ends at the address 2 * Slit16 instruction words away. The 16 bits of Wn support a maximum count value of 65536 and the 16-bit offset value (Slit16) supports offsets of 32K instruction words in both directions.

When this instruction executes, DCOUNT, DOSTART and DOEND are first PUSHed into their respective shadow registers, and then updated with the new DO loop parameters specified by the instruction. The DO level count, DL<2:0> (CORCON<8:10>), is then incremented. After the DO loop completes execution, the PUSHed DCOUNT, DOSTART and DOEND registers are restored, and DL<2:0> are decremented.

The 's' bits specify the register Wn that contains the loop count.

The 'n' bits are a signed literal that specifies the number of instructions that are offset from (PC + 4) , which is the last instruction executed in the loop.

Special Features, Restrictions:

The following features and restrictions apply to the DO instruction.

1. Using a loop count of 0 will result in the loop being executed one time.
2. Using an offset of -2, -1 or 0 is invalid. Unexpected results may occur if these offsets are used.
3. The very last two instructions of the DO loop cannot be:

4. an instruction which changes program control flow
   • a DO or REPEAT instruction
   Unexpected results may occur if these last instructions are used.

5. The first and last instructions of the DO loop should not be a PSV read, table read or table write.

Note 1: The DO instruction is interruptible and supports 1 level of nesting. Nesting up to an additional 5 levels may be provided in software by the user. See the specific device family reference manual for details.

2: The linker will convert the specified expression into the offset to be used.

Words: 2

Cycles: 2

Example 1:	002000 LOOP6:	DO	W0, END6	; Initiate DO loop (W0 reps)
	002004	ADD	W1, W2, W3	; First instruction in loop
	002006	...
	002008	...
	00200A	...
	00200C	REPEAT #6
	00200E	SUB	W2, W3, W4
	002010 END6:	NOP		; Last instruction in loop
				; (Required NOP filler)

	Before Instruction	After Instruction
PC	00 2000 PC 00 2004
W0	0012 W0 0012
DCOUNT	0000 DCOUNT 0012
DOSTART	FF FFFF DOSTART 00 2004
DOEND	FF FFFF DOEND 00 2010
CORCON	0000	CORCON 0100 (DL = 1)
SR	0000	SR 0080 (DA = 1)

Example 2:	002000 LOOPA:	DO	W7, ENDA	; Initiate DO loop (W7 reps)
	002004	SWAP	WO	; First instruction in loop
	002006	...
	002008	...
	00200A	...
	002010 ENDA:	MOV	W1, [W2++]	; Last instruction in loop

	Before Instruction			After Instruction
PC	00 2000 PC	00 2004
W7	E00F	W7	E00F
DCOUNT	0000 DCOUNT		200F
DOSTART	FF FFFF DOSTART 00 2004
DOEND	FF FFFF DOEND 00 2010
CORCON	0000	CORCON 0100 (DL = 1)
SR	0000 SR 0080 (DA = 1)

ED
Euclidean Distance (No Accumulate)

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	PIC33F ds	PIC33E dsP	IC33C
				X	X		X

Syntax:	{label:}	ED	Wm * Wm,	Acc,	[Wx],	[Wy],	Wxd
					[Wx] += kx, [Wy] += ky,
					[Wx] -= kx, [Wy] -= ky,
					[W9 + W12],	[W11 + W12],

Operands:	Acc ∈ [A,B]
	Wm * Wm ∈ [W4 * W4, W5 * W5, W6 * W6, W7 * W7]
	Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]
	Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]
	Wxd ∈ [W4 ... W7]

Operation:

(Wm) ^* (Wm) Acc(A or B) ([Wx] - [Wy]) Wxd (Wx) + kx Wx (Wy) + ky Wy

Status Affected:

OA, OB, OAB, SA, SB, SAB

Encoding:

1111

00mm

A1xx

00ii

iijj

jj11

Description:	Compute the square of Wm, and compute the difference of the prefetch values specified by [Wx] and [Wy]. The results of Wm * Wm are sign-extended to 40 bits and stored in the specified accumulator. The results of [Wx] – [Wy] are stored in Wxd, which may be the same as Wm.
	Operands, Wx, Wxd and Wyd, specify the prefetch operations which support Indirect and Register Offset Addressing, as described in Section 4.15.1 “MAC Prefetches”.
	The ‘m’ bits select the operand register Wm for the square.
	The ‘A’ bit selects the accumulator for the result.
	The ‘x’ bits select the prefetch difference Wxd destination.
	The ‘i’ bits select the Wx prefetch operation.
	The ‘j’ bits select the Wy prefetch operation.

Words:	1
Cycles:	1

Example 1: ED W4*W4, A, [W8]+=2, [W10]==2, W4 ; Square W4 to ACCA; [W8]-[W10] to W4; Post-increment W8; Post-decrement W10

	Before Instruction
W4	009A
W8	1100
W10	2300
ACCA 00 3D0A 0000
Data 1100	007F
Data 2300	0028
SR	0000

	After Instruction
W4	0057
W8	1102
W10	22FE
ACCA 00	0000 5CA4
Data 1100	007F
Data 2300	0028
SR	0000

X

Example 2: ED W5*W5, B, [W9]+=2, [W11+W12], W5; Square W5 to ACCB ; [W9]-[W11+W12] to W5 ; Post-increment W9

	Before Instruction			After Instruction
W5	43C2 W5	3F3F
W9	1200 W9	1202
W11	2500 W11	2500
W12	0008 W12	0008
ACCB	00 28E3 F14C ACCB 00 11EF 1F04
Data 1200		6A7C	Data 1200	6A7C
Data 2508		2B3D	Data 2508	2B3D
SR	0000		SR 0000

EDAC

Euclidean Distance

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F ds	PIC33F dsP	C33E dsP	C33C
			X	X		X

Syntax: {label:} EDAC Wm * Wm, Acc, [Wx], [Wy], Wxd

$$ [ W x ] + = k x, \quad [ W y ] + = k y, $$

$$ [ W x ] - = k x, \quad [ W y ] - = k y, $$

$$ [ W 9 + W 1 2 ], [ W 1 1 + W 1 2 ], $$

Operands:	Acc ∈ [A,B]Wm * Wm ∈ [W4 * W4, W5 * W5, W6 * W6, W7 * W7]Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]Wxd ∈ [W4 ... W7]
Operation:	(Acc(A or B)) + (Wm) * (Wm) → Acc(A or B)[[Wx] - [Wy]) → Wxd(Wx) + kx → Wx(Wy) + ky → Wy

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1111

00mm

A1xx

00ii

iijj

jj10

Description:

Compute the square of Wm, and also the difference of the prefetch values specified by [Wx] and [Wy]. The results of Wm * Wm are sign-extended to 40 bits and added to the specified accumulator. The results of [Wx] – [Wy] are stored in Wxd, which may be the same as Wm.Operands, Wx, Wxd and Wyd, specify the prefetch operations which support Indirect and Register Offset Addressing, as described in Section 4.15.1 “MAC Prefetches”.The ‘m’ bits select the operand register Wm for the square.The ‘A’ bit selects the accumulator for the result.The ‘x’ bits select the prefetch difference Wxd destination.The ‘i’ bits select the Wx prefetch operation.The ‘j’ bits select the Wy prefetch operation.

Words: 1

Cycles: 1

Example 1: EDAC W4*W4, A, [W8] += 2, [w10] -= 2, W4; Square W4 and

; add to ACCA
; [W8] - [W10] to W4
; Post-increment W8
; Post-decrement W10

Before
Instruction

W4	009A W4 0057
W8	1100 W8 1102
W10	2300 W10 22FE
ACCA	00 3D0A 3D0A ACCA 00 3D0A 99AE
Data 1100	007F Data 1100	007F
Data 2300	0028 Data 2300	0028
SR	0000 SR	0000

After
Instruction

Example 2: EDAC W5*W5, B, [w9] += 2, [W11+W12], W5; Square W5 and

; add to ACCB
; [W9] - [W11+W12] to W5
; Post-increment W9

Before
Instruction

W5	43C2	W5	3F3F
W9	1200	W9	1202
W11	2500	W11 2500
W12	0008	W12	0008
ACCB	00 28E3 F14C	ACCB 00 3AD8	1050
Data 1200	6A7C	Data 1200	6A7C
Data 2508	2B3D	Data 2508	2B3D
SR	0000	SR	0000

After
Instruction

EXCH

Exchange Wns and Wnd

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} EXCH Wns, Wnd

Operands: Wns ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: (Wns) ↔ (Wnd)

Status Affected:

None

Encoding:	1111	1101 0000	0ddd d000	ssss
Description:	Exchange the word contents of two Working registers. Register Direct Addressing must be used for Wns and Wnd.

The 'd' bits select the address of the first register.

The 's' bits select the address of the second register.

Note: This instruction only executes in Word mode.

Words: 1

Cycles: 1

Example 1: EXCH W1, W9 ; Exchange the contents of W1 and W9

Before
Instruction

W1	55FF
W9	A3A3
SR	0000

After
Instruction

W1	A3A3
W9	55FF
SR	0000

Example 2: EXCH W4, W5 ; Exchange the contents of W4 and W5

Before
Instruction

W4	ABCD
W5	4321
SR	0000

After
Instruction

W4	4321
W5	ABCD
SR	0000

FBCL
Find First Bit Change from Left

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
	X	X	X	X	X	X

Syntax: {label:} FBCL Ws, Wnd

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: Max_Shift = 15

Sign = (Ws) & 0x8000

Temp = (Ws) << 1

Shift = 0

While ((Shift < Max_Shift) && (Temp & 0x8000) == Sign))

Temp = Temp << 1

Shift = Shift + 1

-Shift → (Wnd)

Status Affected:

C

Encoding:

1101

1111

0000

0ddd

dppp

ssss

Description:

Find the first occurrence of a one (for a positive value) or zero (for a negative value), starting from the Most Significant bit after the sign bit of Ws and working towards the Least Significant bit of the word operand. The bit number result is sign-extended to 16 bits and placed in Wnd.

The next Most Significant bit after the sign bit is allocated bit number 0 and the Least Significant bit is allocated bit number -14. This bit ordering allows for the immediate use of Wd with the SFTAC instruction for scaling values up. If a bit change is not found, a result of -15 is returned and the C flag is set. When a bit change is found, the C flag is cleared.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: FBCL W1, W9; Find 1st bit change from left in W1; and store result to W9

Before
Instruction

After
Instruction

Example 2: FBCL W1, W9 ; Find 1st bit change from left in W1 ; and store result to W9

Before
Instruction

After
Instruction

Example 3: FBCL [W1++, W9 ; Find 1st bit change from left in [W1]

; and store result to W9

; Post-increment W1

Before
Instruction

After
Instruction

FF1L
Find First One from Left

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} FF1L Ws, Wnd

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: Max_Shift = 17

Temp = (Ws)

Shift = 1

While ((Shift < Max_Shift) && !(Temp & 0x8000))

Temp = Temp << 1

Shift = Shift + 1

If (Shift == Max_Shift)

0 → (Wnd)

Else

Shift → (Wnd)

Status Affected: C

Encoding:

1100

1111

1000

0ddd

dppp

ssss

Description: Finds the first occurrence of a one starting from the Most Significant bit of Ws and working towards the Least Significant bit of the word operand. The bit number result is zero-extended to 16 bits and placed in Wnd.

Bit numbering begins with the Most Significant bit (allocated number 1) and advances to the Least Significant bit (allocated number 16). A result of zero indicates a '1' was not found and the C flag will be set. If a '1' is found, the C flag is cleared.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: FF1L W2, W5 ; Find the 1st one from the left in W2; and store result to W5

Before

Instruction

W2 000A W2 000A

W5 BBBBB W5 000D

SR 0000 SR 0000

After

Instruction

Example 2: FF1L [W2++, W5 ; Find the 1st one from the left in [W2]

; and store the result to W5

; Post-increment W2

Before

Instruction

W2 2000 W2 2002

W5 BBBB W5 0000

Data 2000 0000 Data 2000 0000

SR 0000 SR 0001 (C = 1)

After

Instruction

FF1R
Find First One from Right

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C

X

Syntax: {label:} FF1R Ws, Wnd

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: Max_Shift = 17

Temp = (Ws)

Shift = 1

While ((Shift < Max_Shift) && !(Temp & 0x1))

Temp = Temp >> 1

Shift = Shift + 1

If (Shift == Max_Shift)

0 → (Wnd)

Else

Shift → (Wnd)

Status Affected: C

Encoding:

1100

1111 0000

0ddd dppp

ssss

Description: Finds the first occurrence of a one starting from the Least Significant bit of Ws and working towards the Most Significant bit of the word operand. The bit number result is zero-extended to 16 bits and placed in Wnd.

Bit numbering begins with the Least Significant bit (allocated number 1) and advances to the Most Significant bit (allocated number 16). A result of zero indicates a '1' was not found and the C flag will be set. If a '1' is found, the C flag is cleared.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: FF1R W1, W9 ; Find the 1st one from the right in W1 ; and store the result to W9

Before

Instruction

W1 000A W1 000A

W9 BBBBB W9 0002

SR 0000 SR 0000

After

Instruction

Example 2: FF1R [W1++], W9 ; Find the 1st one from the right in [W1]

; and store the result to W9

; Post-increment W1

Before

Instruction

W1 2000 W1 2002

W9 BBBBB W9 0010

Data 2000 8000 Data 2000 8000

SR 0000 SR 0000

After

Instruction

FLIM
Force (Signed) Data Range Limit

Implemented in: PIC24F PIC	24H PIC24E	dsPIC30F	dsPIC33F	dsPIC33E	dsPIC33C
							X

Syntax: {label:} FLIM Wb, Ws,

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15];

Wb ∈ [W0, W2, W4, W6, W8, W10, W12, W14];

Operation: If (Ws) > (Wb)

Then

(Wb) → (Ws);

0 Z;0 N;0 OV;

If (Ws) < (Wb+1)

Then

(Wb+1) → Ws;

0 Z;1 N;0 OV;

Else

1 Z; 0 N; 0 OV;

N, Z, OV

Encoding: 1110 0100 0www w000 0ppp ssss

Description: Simultaneously compare a 16-bit signed data value in Ws to a maximum signed limit value held in Wb and a minimum signed limit value held in W(b+1).

If Ws is greater than Wb, set Ws to the limit value held in Wb. The Z, N and OV Status bits are set such that a subsequent BRA GT instruction will take a branch.

If Ws is less than W(b + 1) , set Ws to the limit value held in W(b + 1) . The Z, N and OV Status bits are set such that a subsequent BRA LT instruction will take a branch.

If Ws is less than or equal to the maximum limit in Wb, and greater than or equal to the minimum limit in W(b+1), Ws is not modified (i.e., data is within range and limits are not applied). The Z Status bit is set such that a subsequent BRA Z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 's' bits select the address of the source (data value) register.

The 'w' bits select the address of the base (data limit) register.

The 'p' bits select the source addressing mode.

Note 1: Although the instruction assumes signed values for all operands, both upper and lower limit values may be of the same sign.

2: The Status bits are set based upon the value loaded into Wnd.
3: If the operand is greater than the maximum limit value in Wb, the CPU will write back the Wb value, regardless of whether the operand is less than the minimum value held in W(b+1) or not.

Words: 1

Cycles: 1

FLIM.V
Force (Signed) Data Range Limit with Limit Excess Result

Implemented in: P	IC24F PIC24H	PIC24E dsPIC	C30F dsPIC33	F dsPIC33E ds	sPIC33C
							X

Syntax:	{label:}	FLIM.V	Wb,	Ws,	Wnd
					[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands: Ws ∈ [W0 ... W15]; Wb ∈ [W0, W2, W4, W6, W8, W10, W12, W14]; Wnd ∈ [W0 ... W15]

Operation: If (Ws) > (Wb) Then (0x0001 → Wnd OR (Ws-Wb) → Wnd; (Wb) → (Ws); 0 → Z; 0 → N; 0 → OV;) If (Ws) < (Wb+1) Then (0xFFFF → Wnd OR (Ws-W(b+1)) → Wnd; W(b+1) → Ws; 0 → Z; 1 → N; 0 → OV;) Else (0 → Wnd; 1 → Z; 0 → N; 0 → OV;)
N, Z, OV

Encoding:

1110

0101

xwww

wddd

dppp

ssss

Description: Simultaneously compare a 16-bit signed data value in Ws to a maximum signed limit value held in Wb and a minimum signed limit value held in W(b+1). Write the limit excess value into Wnd.

If Ws is greater than Wb, either write the (signed) value by which the limit is exceeded to Wnd (FLIM.V, where instruction bit x = 1) or set Wnd to +1 (FLIM, where instruction bit x = 0). In both cases, set Ws to the limit value held in Wb. Whenever Ws is greater than Wb, Wnd will always be a positive value. The Z, N and OV Status bits are set such that a subsequent BRA GT instruction will take a branch.

If Ws is less than W(b+1), either write the (signed) value by which the limit is exceeded to Wnd (FLIM.V, where instruction bit x = 1) or set Wnd to -1 (FLIM, where instruction bit x = 0). In both cases, set Ws to the limit value held in W(b+1). Whenever Ws is less than W(b+1), Wnd will always be a negative value. The Z, N and OV Status bits are set such that a subsequent BRA LT instruction will take a branch.

If Ws is less than or equal to the maximum limit in Wb, and greater than or equal to the minimum limit in W(b+1), Ws is not modified (i.e., data is within range and limits are not applied). Wnd is cleared and the Z Status bit is set such that a subsequent BRA z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 's' bits select the address of the source (data value) register.

The 'w' bits select the address of the base (data limit) register.

The 'd' bits select the address of the destination (limit test result) register.

The 'p' bits select the source addressing mode.

The 'x' bit defines the presence and result format for Wnd.

Note 1: Although the instruction assumes signed values for all operands, both upper and lower limit values may be of the same sign.

2: The Status bits are set based upon the value loaded into Wnd.

3: If the operand is greater than the maximum limit value in Wb, the CPU will write back the Wb value, regardless of whether the operand is less than the minimum value held in W(b+1) or not.

Words:	1
Cycles:	1

GOTO

Unconditional Jump

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} GOTO Expr

Operands: Expr may be label or expression (but not a literal). Expr is resolved by the linker to a lit23, where lit23 ∈ [0 ... 8388606].

Operation: lit23 → PC NOP → Instruction Register

Status Affected: None

Encoding:	1st word	0000	0100	nnnn	nnnn	nnnn	nnn0
	2nd word	0000	0000	0000	0000	0nnn	nnnn

Description: Unconditional jump to anywhere within the 4M instruction word program memory range. The PC is loaded with the 23-bit literal specified in the instruction. Since the PC must always reside on an even address boundary, lit23<0> is ignored.

The 'n' bits form the target address.

Note: The linker will resolve the specified expression into the lit23 to be used.

Words: 2

Cycles: 2 (PIC24F, PIC24H, dsPIC30F, dsPIC33F) 4 (PIC24E, dsPIC33E, dsPIC33C)

Example 1:	026000	GOTO	_THERE	; Jump to _THERE
	026004	MOV	W0, W1
	.	...
	.	...
	027844 _THERE:	MOV	#0x400, W2	; Code execution
	027846	...		; resumes here

	Before Instruction		After Instruction
PC	02 6000		PC 02 7844
SR	0000		SR 0000

Example 2:	000100 _code: ...			; start of code
	.	...
	026000	GOTO _code+2		; Jump to _code+2
	026004	...

	Before Instruction
PC	02 6000
SR	0000

	After Instruction
PC	00 0102
SR	0000

GOTO

Unconditional Indirect Jump

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F dsP	IC33F dsPIC	C33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} GOTO Wn

Operands: Wn ∈ [W0 ... W15]

Operation: 0 → PC<22:16>

$$ \begin{array}{l} (W n < 1 5: 1 >) \rightarrow P C < 1 5: 1 > \ 0 \rightarrow \mathrm{PC} < 0 > \ \mathrm{NOP} \rightarrow \text { Instruction Register } \ \end{array} $$

Status Affected: None

Encoding:

Unconditional indirect jump within the first 32K words of program memory. Zero is loaded into PC<22:16> and the value specified in (Wn) is loaded into PC<15:1>. Since the PC must always reside on an even address boundary, Wn<0> is ignored.

The 's' bits select the source register.

Words: 1

Cycles: 2

Example 1:	006000	GOTO	W4	; Jump unconditionally
	006002	MOV	W0, W1	; to 16-bit value in W4
	.	...
	.	...
	007844 _THERE:	MOV	#0x400, W2	; Code execution
	007846	...		; resumes here

	Before Instruction
W4	7844
PC	00 6000
SR	0000

	After Instruction
W4	7844
PC	00 7844
SR	0000

GOTO

Unconditional Indirect Jump

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E dsPIC33C
		X			X	X

Syntax: {label:} GOTO Wn

Operands: Wn ∈ [W0 ... W15]

Operation: 0 → PC<22:16>

$$ \begin{array}{l} (W n < 1 5: 1 >) \rightarrow P C < 1 5: 1 > \ 0 \rightarrow \mathrm{PC} < 0 > \ \mathrm{NOP} \rightarrow \text { Instruction Register } \ \end{array} $$

Status Affected: None

Encoding:

Unconditional indirect jump within the first 32K words of program memory. Zero is loaded into PC<22:16> and the value specified in (Wn) is loaded into PC<15:1>. Since the PC must always reside on an even address boundary, Wn<0> is ignored.

The 's' bits select the source register.

Words: 1

Cycles: 4

Example 1:	006000	GOTO	W4	; Jump unconditionally
	006002	MOV	W0, W1	; to 16-bit value in W4
	.	...
	.	...
	007844 _THERE:	MOV	#0x400, W2	; Code execution
	007846	...		; resumes here

Before
Instruction

W4	7844
PC	00 6000
SR	0000

After
Instruction

W4	7844
PC	00 7844
SR	0000

GOTO.L

Unconditional Indirect Jump Long

Implemented in: PIC24F PIC24	H PIC24E dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
		X			X	X

Syntax: {label:} GOTO.L Wn

Operands: Wn ∈ [W0, W2, W4, W6, W8, W10, W12]

Operation: PC<23> → PC<23> (see text); (Wn+1)<6:0> → PC<22:16>; (Wn) → PC<15:0>

Status Affected: None

Encoding:

Unconditional indirect jump to any user program memory address.

The Least Significant 7 bits of (Wn+1) are loaded in PC<22:16> and the 16-bit value (Wn) is loaded into PC<15:0>.

PC<23> is not modified by this instruction.

The contents of (Wn+1)<15:7> are ignored.

The value of Wn<0> is also ignored and PC<0> is always set to '0'.

GOTO is a two-cycle instruction.

The 's' bits select the address of the Wn source register.

The 'w' bits specify the address of the Wn+1 source register.

Words: 1

Cycles: 4

Example 1:	026000	GOTO.L	W4	; Call _FIR subroutine
	026004	MOV	W0, W1
	.	...
	.	...
	026844 _FIR:	MOV	#0x400, W2	; _FIR subroutine start
	026846	...

	Before Instruction		After Instruction
PC	02 6000	PC	02 6844
W4	6844	W4	6844
W5	0002	W5	0002
W15	A268	W15	A26C
Data A268	FFFF	Data A268	6004
Data A26A	FFFF	Data A26A	0002
SR	0000	SR	0000

INC

Increment f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

X

Syntax: {label:} INC{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) + 1 → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding: 1110 1100 0BDF ffff ffff ffff

Description: Add one to the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: INC.B 0x1000 ; Increment 0x1000 (Byte mode)

After Instruction Data 1000 □ 8F00 SR □ 0101 (DC, C = 1)

Example 2: INC 0x1000, WREG ; Increment 0x1000 and store to WREG ; (Word mode)

After Instruction WREG 9000 Data 1000 8FFF SR 0108 (DC, N = 1)

INC

Increment Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} INC{.B} Ws, Wd
[Ws], [Wd]
[Ws++]， [Wd++]
[Ws--], [Wd--]
[++Ws], [++Wd]
[--Ws], [--Wd]
Operands: Ws ∈ [W0 ... W15]
Wd ∈ [W0 ... W15]
Operation: (Ws) + 1 → Wd
Status Affected: DC, N, OV, Z, C
Encoding: 1110 1000 0Bqq qddd dppp ssss
Description: Add one to the contents of the source register Ws and place the result in the destination register Wd. Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'q' bits select the destination addressing mode.
The 'd' bits select the destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	INC.B W1, [++W2]	; Pre-increment W2
		; Increment W1 and store to W2
		; (Byte mode)

Before Instruction		After Instruction
W1	FF7F	W1	FF7F
W2	2000	W2	2001
Data 2000	ABCD	Data 2000	80CD
SR	0000	SR	010C (DC, N, OV = 1)

Example 2:

INC

W1, W2

; Increment W1 and store to W2; (Word mode)

Before Instruction		After Instruction
W1	FF7F	W1	FF7F
W2	2000	W2	FF80
SR	0000	SR	0108 (DC, N = 1)

INC2

Increment f by 2

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} INC2{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) + 2 → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding:

1110

1100

1BDf

ffff

ffff

ffff

Description: Add two to the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: INC2.B 0x1000 ; Increment 0x1000 by 2 ; (Byte mode)

After Instruction Data 1000 □ 8F01 SR □ 0101 (DC, C = 1)

Example 2: INC2 0x1000, WREG ; Increment 0x1000 by 2 and store to WREG ; (Word mode)

After Instruction WREG 9001 Data 1000 8FFF SR 0108 (DC, N = 1)

INC2

Increment Ws by 2

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
	X	X		X	X

Syntax: {label:} INC2{.B} Ws, Wd
[Ws], [Wd]
[Ws++, [Wd++]
[Ws--], [Wd--]

[++Ws], [++Wd]
[--Ws], [--Wd]
Operands: Ws ∈ [W0 ... W15]
Wd ∈ [W0 ... W15]
Operation: (Ws) + 2 → Wd
Status Affected: DC, N, OV, Z, C

Encoding:

1110

1000

1Bqq

qddd

dppp

ssss

Description: Add two to the contents of the source register Ws and place the result in the destination register Wd. Register Direct or Indirect Addressing may be used for Ws and Wd.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'q' bits select the destination addressing mode.
The 'd' bits select the destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1
Cycles: 1(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	INC2.B W1, [++W2]	; Pre-increment W2
		; Increment by 2 and store to W1
		; (Byte mode)

Before Instruction		After Instruction
W1	FF7F	W1	FF7F
W2	2000	W2	2001
Data 2000	ABCD	Data 2000	81CD
SR	0000	SR	010C (DC, N, OV = 1)

Example 2:

INC2

W1, W2

; Increment W1 by 2 and store to W2; (word mode)

Before Instruction		After Instruction
W1	FF7F	W1	FF7F
W2	2000	W2	FF81
SR	0000	SR	0108 (DC, N = 1)

IOR

Inclusive OR f and WREG

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C

X

Syntax: {label:} IOR{.B} f {,WREG}
Operands: f [0 8191]
Operation: (f).IOR.(WREG) → destination designated by D
Status Affected: N, Z

Encoding:

1011

0111

0BDF

ffff

ffff

ffff

Description:

Compute the logical inclusive OR operation of the contents of the Working register WREG and the contents of the file register, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:

IOR.B 0x1000

;IOR WREG to (0x1000) (Byte mode); (Byte mode)

	Before Instruction		After Instruction
WREG	1234	WREG 12	34
Data 1000	FF00	Data 1000	FF34
SR	0000	SR 0000

Example 2:	IOR	0x1000, WREG	; IOR (0x1000) to WREG
			; (Word mode)

	Before Instruction		After Instruction
WREG	1234	WREG	1FBF
Data 1000	0FAB	Data 1000	0FAB
SR	0008 (N = 1)	SR 0000

IOR

Inclusive OR Literal and Wn

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F	dsPIC33F	dsPIC33E	dsPIC33C
	X	X		X	X		X

Syntax: {label:} IOR{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation

lit10 ∈ [0 ... 1023] for word operation

Wn ∈ [W0 ... W15]

Operation: lit10.IOR.(Wn) → Wn

Status Affected: N, Z

Encoding:

1011

0011

0Bkk

kkkk

kkkk

dddd

Description: Compute the logical inclusive OR operation of the 10-bit literal operand and the contents of the Working register Wn, and place the result back into the Working register Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

Words: 1

Cycles: 1

Example 1: IOR.B #0xAA, W9 ; IOR 0xAA to W9

; (Byte mode)

Before
Instruction

W9	1234
SR	0000

After
Instruction

W9	12BE
SR	0008 (N = 1)

Example 2: IOR #0x2AA, W4 ; IOR 0x2AA to W4

; (Word mode)

Before
Instruction

W4	A34D
SR	0000

After
Instruction

W4	A3EF
SR	0008 (N = 1)

IOR

Inclusive OR Wb and Short Literal

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F dsP	IC33F dsPIC	C33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} IOR{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb).IOR.lit5 → Wd

Status Affected: N, Z

Encoding:

0111

0www

wBqq

qddd

d11k

kkkk

Description: Compute the logical inclusive OR operation of the contents of the base register Wb and the 5-bit literal operand, and place the result in the destination register Wd.

Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: IOR.B W1, #0x5, [W9++] ; IOR W1 and 0x5 (Byte mode)

; Store to [W9]

; Post-increment W9

	Before Instruction
W1	AAAA
W9	2000
Data 2000	0000
SR	0000

After Instruction
W1	AAAA
W9	2001
Data 2000	00AF
SR	0008 (N = 1)

Example 2: IOR W1, #0x0, W9 ; IOR W1 with 0x0 (Word mode)

; Store to W9

Before
Instruction

W1	0000
W9	A34D
SR	0000

After
Instruction

W1	0000
W9	0000
SR	0002 (Z = 1)

IOR
Inclusive OR Wb and Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X	X	X	X	X	X

Syntax: {label:} IOR{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb).IOR.(Ws) → Wd

Status Affected: N, Z

Encoding:

0111

0www

wBqq

qddd

dppp

ssss

Description: Compute the logical inclusive OR operation of the contents of the source register Ws and the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: IOR.B W1, [W5++], [W9++] ; IOR W1 and [W5] (Byte mode)

; Store result to [W9]

; Post-increment W5 and W9

Before
Instruction

W1	AAAA W1	AAAA
W5	2000 W5	2001
W9	2400 W9	2401
Data 2000	1155 Data	2000 1155
Data 2400	0000 Data	2400 00FF
SR	0000 SR	0008 (N = 1)

After

Instruction

Example 2: IOR W1, W5, W9 ; IOR W1 and W5 (Word mode)

; Store the result to W9

Before
Instruction

W1	AAAA
W5	5555
W9	A34D
SR	0000

After

	Instruction
W1	AAAA
W5	5555
W9	FFFF
SR	0008 (N = 1)

LAC
Load Accumulator

Implemented in: PIC24F P	IC24H PIC24E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
			X	X	X	X
Syntax:{label:}	LAC	Ws,	{#Slit4,}Acc[Ws],[Ws++],[Ws--],[--Ws],[++Ws],[Ws+Wb],
Operands:	Ws ∈ [W0 ... W15]Wb ∈ [W0 ... W15]Slit4 ∈ [-8 ... +7]Acc ∈ [A,B]
Operation:	ShiftSlit4(Extend(Ws)) → Acc(A or B)
Status Affected:	OA, OB, OAB, SA, SB, SAB
Encoding:	1100	1010	Awww	wrrr	rggg	ssss
Description:	Read the contents of the source register. Optionally perform a signed 4-bit shift and store the result in the specified accumulator. The shift range is -8:7, where a negative operand indicates an arithmetic left shift and a positive operand indicates an arithmetic right shift. The data stored in the source register is assumed to be 1.15 fractional data, and is automatically sign-extended (through bit 39) and zero-backfilled (bits<15:0>) prior to shifting.The ‘A’ bit specifies the destination accumulator.The ‘w’ bits specify the offset register Wb.The ‘r’ bits encode the accumulator preshift.The ‘g’ bits select the source addressing mode.The ‘s’ bits specify the source register Ws.Note: If the operation moves more than sign-extension data into the Accumulator Upper register (ACCxU), or causes a saturation, the appropriate overflow and saturation bits will be set.
Words:	1
Cycles:	_1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1:	LAC [W4++], #-3, B	; Load ACCB with [W4] << 3
		; Contents of [W4] do not change
		; Post increment W4
		; Assume saturation disabled
		; (SATB = 0)

Before Instruction W4 2000 W4 2002 ACCB 00 5125 ABCD AC CB FF 9108 0000 Data 2000 1221 Data 2000 1221 SR 0000 SR 4800 (OB, OAB = 1)

Example 2:	LAC [--W2], #7, A	; Pre-decrement W2
		; Load ACCA with [W2] >> 7
		; Contents of [W2] do not change
		; Assume saturation disabled
		; (SATA = 0)

Before Instruction W2 4002 W2 4000 ACCA 00 5125 ABCD ACCA FF FF22 1000 Data 4000 9108 Data 4000 9108 Data 4002 1221 Data 4002 1221 SR 0000 SR 0000

After Instruction

LAC.D
Load Accumulator Double

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
						X

Syntax:	{label:}	LAC.D	Ws,	[,	#Slit4],	Acc
					[Ws],
					[Ws++]
					[Ws--]
					[--Ws],
					[++Ws]

Operands: Register Direct: Wns ∈ [W0, W2, W4, W6, W8, W10, W12, W14];

Register Indirect: Wns ∈ [W0 ... W15];

Slit4 ∈ [-8 ... +7] Acc ∈ [A,B]

Operation: ShiftSlit4(Extend(Ws)) → ACC (A,B)

Status Affected: OA, SA or OB, SB

Encoding:

1101

1011

A000

0rrr

rppp

ssss

Description: Read the contents of the source register. Optionally perform a signed 4-bit shift and store the result in the specified accumulator. The shift range is -8:7, where a negative operand indicates an arithmetic left shift and a positive operand indicates an arithmetic right shift. The data stored in the source register is assumed to be 1.31 fractional data, and is automatically sign-extended (through bit 39) and zero-backfilled (bits<15:0>) prior to shifting.

The 'A' bit specifies the destination accumulator.

The 's' bits specify the source register Wns.

The 'p' bits select the source addressing mode.

The 'r' bits encode the optional operand Slit4, which determines the amount of the accumulator preshift; if the operand Slit4 is absent, a '0' is encoded.

See Table 5-7 for modifier addressing information.

Note 1:	Unlike the LAC instruction, the LAC.D instruction does not support Indirect with Register Offset Addressing mode.
	2: Positive values of operand Slit4 represent arithmetic shift right. Negative values of operand Slit4 represent shift left.
	3: The LAC.D instruction cannot be executed within a REPEAT loop.

Words: 1

Cycles: 2^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

LDSLV

Load Slave Processor Program RAM

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
							X

Syntax:	{label:}	LDSLV	[Wns]	[Wnd++]	#lit2
					[Wns++]

Operands: Wns ∈ [W0 ... W15];

$$ \text { Wnd } \in [ \text { W0 } \dots \text { W15 } ]; $$

$$ \operatorname{lit} 2 \in [ 0 \dots 3 ] $$

Operation: Master (EAs) → Slave EAd

Status Affected: None

Encoding:

0000

0011

00kk

0ddd

d0p1

ssss

Description:

This instruction moves a single instruction word from the target Slave PRAM image (held in the Master program space Flash) into the Slave PRAM. The source address must be located within PSV address space (i.e., DSRPAG ≥ 0x200). The destination address uses DSWPAG and the destination EA to create a 24-bit Slave PS PRAM write address.

Starting with an aligned double instruction word destination address (see note), move the contents of the source Effective Address (in Master program space) to the destination Effective Address (in the Slave PRAM address space).

If the (single instruction word) destination address is even, capture the data in the Slave PRAM wrapper. If the (single instruction word) destination address is odd, the ECC parity bits are calculated from the current and captured source data (48-bits), then stored together with the data into the PRAM double instruction word destination Effective Address.

The target Slave processor is selected by the value defined by lit2.

The instruction may be regarded as a PSV operation, and hence, may be executed within a REPEAT loop to accelerate data processing.

The 's' bits select the address of the source register.

The 'd' bits select the address of the destination register.

The 'k' bits select the target Slave processor.

The 'p' bit selects the destination addressing mode (see note).

Note 1: This instruction supports a subset of addressing modes. The Source Addressing mode bit field is constrained to 2 options and the Destination Addressing mode bit field is not required.

2: An aligned double instruction word destination address is an even address that addresses the least significant word of a double instruction word.

3: This instruction only supports Word mode.

Words: 1

Cycles: 1

LNK

Allocate Stack Frame

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
	X	X		X		X

Syntax: {label:} LNK #lit14
Operands: lit14 ∈ [0 ... 16382]

Operation: (W14) → (TOS)

(W15) + 2 → W15

(W15) → W14

(W15) + lit14 → W15

Status Affected: None

Encoding:

1111

1010

00kk

kkkk

kkkk

kkk0

Description:

This instruction allocates a stack frame of size lit14 bytes for a subroutine calling sequence. The stack frame is allocated by PUSHing the contents of the Frame Pointer (W14) onto the stack, storing the updated Stack Pointer (W15) to the Frame Pointer and then incrementing the Stack Pointer by the unsigned 14-bit literal operand. This instruction supports a maximum stack frame of 16382 bytes.
The 'k' bits specify the size of the stack frame.
Note: Since the Stack Pointer can only reside on a word boundary, lit14 must be even.
Words: 1
Cycles: 1
Example 1: LNK #0xA0 ; Allocate a stack frame of 160 bytes

	Before Instruction
W14	2000
W15	2000
Data 2000	0000
SR	0000

	After Instruction
W14	2002
W15	20A2
Data 2000	2000
SR	0000

LNK

Allocate Stack Frame

Implemented in: PIC24F PIC24	H PIC24E	dsPIC30F dsP	IC33F dsPIC	C33E dsPIC3	3C
			X			X	X

Syntax: {label:} LNK #lit14
Operands: lit14 ∈ [0 ... 16382]

Operation: (W14) → (TOS)

(W15) + 2 → W15

(W15) → W14

1 → SFA Status bit

(W15) + lit14 → W15

Status Affected: SFA

Encoding:

1111

1010

00kk

kkkk

kkkk

kkk0

Description:

This instruction allocates a stack frame of size lit14 bytes for a subroutine calling sequence. The stack frame is allocated by PUSHing the contents of the Frame Pointer (W14) onto the stack, storing the updated Stack Pointer (W15) to the Frame Pointer and then incrementing the Stack Pointer by the unsigned 14-bit literal operand. This instruction supports a maximum stack frame of 16382 bytes.
The 'k' bits specify the size of the stack frame.
Note: Since the Stack Pointer can only reside on a word boundary, lit14 must be even.
Words: 1
Cycles: 1
Example 1: LNK #0xA0 ; Allocate a stack frame of 160 bytes

	Before Instruction
W14	2000
W15	2000
Data 2000	0000
SR	0000
CORCON	0000

	After Instruction
W14	2002
W15	20A2
Data 2000	2000
SR	0000
CORCON	0004

LSR

Logical Shift Right f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F ds PIC33F ds PIC33E ds PIC33C
	X	X		X	X		X

Syntax: {label:} LSR{.B} f {,WREG}

Operands: f [0 8191]

Operation: For Byte Operation:

0 → Dest<7>

(f<7:1>) → Dest<6:0>

(f<0>) → C

For Word Operation:

0 → Dest<15>

(f<15:1>) → Dest<14:0>

(f<0>) → C

Status Affected:

N, Z, C

Encoding:

1101

0101

OBDf

ffff

ffff

ffff

Description:

Shift the contents of the file register one bit to the right and place the result in the destination register. The Least Significant bit of the file register is shifted into the Carry bit of the STATUS Register. Zero is shifted into the Most Significant bit of the destination register.

The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words:

1

Cycles:

1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: LSR.B 0x600 ; Logically shift right (0x600) by one ; (Byte mode)

Before

Instruction

Data 600 55FF Data 600 557F

SR 0000 SR 0001 (C = 1)

After

Instruction

Example 2: LSR 0x600, WREG ; Logically shift right (0x600) by one

; Store to WREG

; (Word mode)

Before

Instruction

Data 600 55FF Data 600 55FF

WREG 0000 WREG 2AFF

SR 0000 SR 0001 (C = 1)

After

Instruction

LSR

Logical Shift Right Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} LSR{.B} Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

0 → Wd<7>

(Ws<7:1>) Wd<6:0>

(Ws<0>) C

For Word Operation:

0 → Wd<15>

(Ws<15:1>) Wd<14:0>

(Ws<0>) C

Status Affected: N, Z, C

Encoding:

1101

0001

0Bqq

qddd

dppp

ssss

Description: Shift the contents of the source register Ws one bit to the right and place the result in the destination register Wd. The Least Significant bit of Ws is shifted into the Carry bit of the STATUS Register. Zero is shifted into the Most Significant bit of Wd. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: LSR.B W0, W1 ; LSR W0 (Byte mode) ; Store result to W1

Before
Instruction
W0	FF03 W0	FF03
W1	2378 W1	2301
SR	0000 SR	0001 (C = 1)

After Instruction

Example 2: LSR W0, W1 ; LSR W0 (Word mode) ; Store the result to W1

Before Instruction
W0	8000 W0 8000
W1	2378 W1 4000
SR	0000 SR 0000

After Instruction

LSR

Logical Shift Right by Short Literal

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} LSR Wb, #lit4, Wnd

Operands: Wb ∈ [W0 ... W15]

lit4 ∈ [0 ... 15]

Wnd ∈ [W0 ... W15]

Operation: lit4<3:0> → Shift_Val 0 → Wnd<15:15-Shift_Val + 1> Wb<15:Shift_Val> → Wnd<15-Shift_Val:0>

Status Affected: N, Z

Encoding: 1101 1110 0www wddd d100 kkkk

Description: Logical shift right the contents of the source register Wb by the 4-bit unsigned literal and store the result in the destination register Wnd. Direct Addressing must be used for Wb and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1

Example 1: LSR W4, #14, W5 ; LSR W4 by 14 ; Store result to W5

	Before Instruction
W4	C800
W5	1200
SR	0000

	After Instruction
W4	C800
W5	0003
SR	0000

Example 2: LSR W4, #1, W5 ; LSR W4 by 1 ; Store result to W5

	Before Instruction		After Instruction
W4	0505	W4	0505
W5	F000	W5	0282
SR	0000	SR	0000

LSR

Logical Shift Right by Wns

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} LSR Wb, Wns, Wnd

Operands: Wb ∈ [W0 ... W15] Wns ∈ [W0 ...W15] Wnd ∈ [W0 ... W15]

Operation: Wns<4:0> → Shift_Val 0 → Wnd<15:15-Shift_Val + 1> Wb<15:Shift_Val> → Wnd<15 - Shift_Val:0>

Status Affected: N, Z

Encoding: 1101 1110 0www wddd d000 ssss

Description: Logical shift right the contents of the source register Wb by the 5 Least Significant bits of Wns (only up to 15 positions) and store the result in the destination register Wnd. Direct Addressing must be used for Wb and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the destination register.

The 's' bits select the source register.

Note 1: This instruction operates in Word mode only.

2: If Wns is greater than 15, Wnd will be loaded with 0x0.

Words: 1

Cycles: 1

Example 1: LSR W0, W1, W2 ; LSR W0 by W1 ; Store result to W2

	Before Instruction
W0	C00C
W1	0001
W2	2390
SR	0000

After Instruction
W0 C00C
W1	0001
W2	6006
SR	0000

Example 2: LSR W5, W4, W3 ; LSR W5 by W4 ; Store result to W3

	Before Instruction
W3	DD43
W4	000C
W5	0800
SR	0000

After Instruction
W3	0000
W4	000C
W5	0800
SR	0002 (Z = 1)

Before

MAC

Multiply and Accumulate

Implemented in: PIC24F	PIC24H	PIC2	4E dsPIC30F	dsPIC33F dsP	IC33E dsP	C33C
					X	X	X	X

Syntax: {label:} MAC Wm*Wn, Acc {,[Wx], Wxd} {,[Wy], Wyd} {,AWB}

$$ {, [ W x ] + = k x, W x d } {, [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } {, [ W 1 1 + W 1 2 ], W y d } $$

Operands: Wm * Wn ∈ [W4 * W5, W4 * W6, W4 * W7, W5 * W6, W5 * W7, W6 * W7]

Acc ∈ [A,B]

Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]; Wxd ∈ [W4 ... W7]

Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]; Wyd ∈ [W4 ... W7]

AWB ∈ [W13, [W13] += 2]

Operation: (Acc(A or B)) + (Wm) * (Wn) → Acc(A or B)

(Acc(B or A)) rounded → AWB

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

0mmm

A0xx

yyii

iijj

jjaa

Description: Multiply the contents of two Working registers. Optionally prefetch operands in preparation for another MAC type instruction and optionally store the unspecified accumulator results. The 32-bit result of the signed multiply is sign-extended to 40 bits and added to the specified accumulator.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing, as described in Section 4.14.1 "MAC Prefetches". Operand AWB specifies the optional store of the "other" accumulator, as described in Section 4.15.4 "MAC Write-Back".

The 'm' bits select the operand registers Wm and Wn for the multiply.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

The 'a' bits select the accumulator Write-Back destination.

Note 1: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

2: The US<1:0> bits (CORCON<13:12> in dsPIC33E/dsPIC33C, CORCON<12> in dsPIC30F/dsPIC33F) determine if the multiply is unsigned, signed or mixed-sign. Only dsPIC33E/dsPIC33C devices support mixed-sign multiplication.

Words: 1

Cycles: 1

Example 1: MAC W4*W5, A, [W8] += 6, W4, [W10] += 2, W5

; Multiply W4*W5 and add to ACCA
; Fetch [W8] to W4, Post-increment W8 by 6
; Fetch [W10] to W5, Post-increment W10 by 2
; CORCON = 0x00C0 (fractional multiply, normal saturation)

Before Instruction
W4	A022 W4 2567
W5	B900 W5 909C
W8	0A00 W8 0A06
W10	1800 W10 1802
ACCA	00 1200 0000 ACCA 00 472D 2400
Data 0A00	2567 Data 0A00 2567
Data 1800	909C Data 1800 909C
CORCON	00C0 CORCON 00C0
SR	0000 SR 0000

After Instruction

Example 2: MAC W4*W5, A, [W8]-=2, W4, [W10]+=2, W5, W13

; Multiply W4*W5 and add to ACCA
; Fetch [W8] to W4, Post-decrement W8 by 2
; Fetch [W10] to W5, Post-increment W10 by 2
; Write Back ACCB to W13
; CORCON = 0x00D0 (fractional multiply, super saturation)

Before Instruction
W4	1000
W5	3000
W8	0A00
W10	1800
W13	2000
ACCA	23 5000 2000
ACCB	00 0000 8F4C
Data 0A00	5BBE
Data 1800	C967
CORCON	00D0
SR	0000

After Instruction
W4	5BBE
W5	C967
09FE
W10	1802
W13	0001
ACCA	23 5600 2000
ACCB	00 0000 1F4C
Data 0A00	5BBE
Data 1800	C967
CORCON	00D0
SR	8800 (OA, OAB = 1)

MAC

Square and Accumulate

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Syntax: {label:} MAC

Wm * Wm, Acc {,[Wx], Wxd} {,[Wy], Wyd}

$$ {, [ W x ] + = k x, W x d } {, [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } \quad {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } \quad {, [ W 1 1 + W 1 2 ], W y d } $$

Operands:

Wm * Wm ∈ [W4 * W4, W5 * W5, W6 * W6, W7 * W7]

Acc ∈ [A,B]

Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]; Wxd ∈ [W4 ... W7]

Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]; Wyd ∈ [W4 ... W7]

Operation:

(Acc(A or B)) + (Wm) * (Wm) → Acc(A or B)

([Wx]) → Wxd; (Wx) + kx → Wx

([Wy]) → Wyd; (Wy) + ky → Wy

Status Affected:

OA, OB, OAB, SA, SB, SAB

Encoding:

1111

00mm

A0xx

yyii

iijj

jj00

Description:

Square the contents of a Working register. Optionally prefetch operands in preparation for another MAC type instruction. The 32-bit result of the signed multiply is sign-extended to 40 bits and added to the specified accumulator.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing, as described in Section 4.14.1 "MAC Prefetches".

The 'm' bits select the operand register Wm for the square.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

Note 1: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

2: The US<1:0> bits (CORCON<13:12> in dsPIC33E/dsPIC33C, CORCON<12> in dsPIC30F/dsPIC33F) determine if the multiply is unsigned, signed or mixed-sign. Only dsPIC33E/dsPIC33C devices support mixed-sign multiplication.

Words: 1

Cycles: 1

Example 1: MAC W4*W4, B, [W9+W12], W4, [W10]-=2, W5

; Square W4 and add to ACCB
; Fetch [W9+W12] to W4
; Fetch [W10] to W5, Post-decrement W10 by 2
; CORCON = 0x00C0 (fractional multiply, normal saturation)

Before
Instruction

W4	A022 W4 A230
W5	B200 W5 650B
W9	0C00 W9 0C00
W10	1900 W10 18FE
W12	0020 W12 0020
ACCB	00 2000 0000 ACCB 00 67CD 0908
Data 0C20	A230 Data 0C20 A230
Data 1900	650B Data 1900 650B
CORCON	00C0 CORCON 00C0
SR	0000 SR 0000

After
Instruction

Example 2: MAC W7*W7, A, [W11] == 2, W7

; Square W7 and add to ACCA
; Fetch [W11] to W7, Post-decrement W11 by 2
; CORCON = 0x00D0 (fractional multiply, super saturation)

Before
Instruction

W7	76AE		W7 23FF
W11		2000
ACCA	FE 9834 4500		A
Data 2000	23FF	Data	2000 23FF
CORCON		00D0	CORCON
SR		0000

After
Instruction

F
W11	1FFE
ACCA	FF 063E 0188
F
	00D0
SR	8800 (OA, OAB = 1)

MAX

Accumulator Force Maximum Data Range Limit

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsP C33E dsPIC 33C
						X

Syntax: {label:} MAX

Acc

Operands: Acc ∈ [A,B]

Operation: If (MAX A) Then

If ACCA - ACCB > 0 Then

(ACCB → ACCA;

0 Z;0 N;0 OV;)

Else

(1 → Z; 0 → N; 0 → OV;)

If (MAX B) Then

If ACCB - ACCA > 0 Then

(ACCA → ACCB;

0 Z;0 N;0 OV;)

Else

(1 → Z; 0 → N; 0 → OV;)

Status Affected: N, OV, Z

Encoding:

1100 1110

A00x

x000

0000 0000

Description:

The target accumulator (defined in the instruction) is clamped to the maximum limit value previously loaded into the other accumulator (sign-extended 32-bit value). The comparison examines the full 40-bit value of the target accumulator, and will therefore, clamp an overflowed accumulator.

If the target accumulator is greater than the limit accumulator, load the target accumulator with the contents of the limit accumulator. The Z and N Status bits are set such that a subsequent BRA GT instruction will take a branch. In addition, Z is set such that a subsequent MIN instruction will execute as a NOP if the limit is exceeded. If the limit is not exceeded (Z = 1), the MIN instruction will execute as normal.

If the target accumulator is not greater than the limit accumulator, the target accumulator is unaffected. The Z Status bit is set such that a subsequent BRA z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 'A' bit specifies the destination accumulator.

The 'x' bits define the presence and result format for Wd.

Note: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC , #0 after MAX operation to update DSP status to reflect contents of AccX.

Words: 1

Cycles: 1

MAX.V
Accumulator Force Maximum Data Range Limit with Limit Excess Result

Implemented in: PIC24F P	C24H PIC24	dsPIC30F	dsPIC33F dsP	C33E dsPIC33C
							X
Syntax: {label:} MAX.V	Acc			Wd[Wd][Wd++]Wd--][++Wd] [--Wd]
Operands: Wd ∈ [W0 ... W15]A ∈ [A,B]
Operation: If (MAX A) ThenIf ACCA – ACCB > 0 Then(0x0001 → Wd or ACCA – ACCB → Wd (see text);ACCB → ACCA;0 → Z; 0 → N; 0 → OV;)Else(0 → Wd;1 → Z; 0 → N; 0 → OV;)If (MAX B) ThenIf ACCB – ACCA > 0 Then(0x0001 → Wd or ACCB – ACCA → Wd (see text);ACCA → ACCB;1 → Z; 0 → N; 0 → OV;)Else(0 → Wd;1 → Z; 0 → N; 0 → OV;)
Status Affected: N, OV, Z
Encoding:	1100 1110		A00x	x000	0qqq dddd
Description:	The target accumulator (defined in the instruction) is clamped to the maximum limit value previously loaded into the other accumulator. The comparison examines the full 40-bit value of the target accumulator, and will therefore, clamp an overflowed accumulator.If the target accumulator is greater than the limit accumulator, load the target accumulator with the contents of the limit accumulator. For MAX (instruction bit field xx = 2'b10), set Wd to +1. For MAX.V (instruction bit field xx = 2'b11), write the (signed) value by which the limit is exceeded to Wd. This is sourced from the Least Significant 16 bits of the 40-bit result. If the limit is exceeded by a value greater than that which can be represented by a signed 16-bit number, saturate the Wd write to the maximum positive value (i.e., set Wd to 0x7FFF).The Z and N Status bits are set such that a subsequent BRA GT instruction will take a branch if the limit is exceeded. In addition, Z is set such that a subsequent MIN{ .V} instruction will execute as a NOP if the limit is exceeded. If the limit is not exceeded (Z = 1), the MIN{ .V} instruction will execute as normal.If the target accumulator is not greater than the limit accumulator, the target accumulator is unaffected and Wd is cleared. The Z Status bit is set such that a subsequent BRA Z instruction will take a branch.The OV Status bit is always cleared by this instruction.The ‘A’ bit specifies the destination accumulator.The ‘d’ bits select the address of the destination register.The ‘q’ bits select the destination addressing mode.The ‘x’ bits define the presence and result format for Wd.Note: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC <AccX>, #0 after MAX.V operation to update DSP status to reflect contents of AccX.
Words:	1
Cycles:	1

MIN

Accumulator Force Minimum Data Range Limit (Unconditional Execution)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsP C33E dsPIC 33C
						X

Syntax: {label:} MIN

Acc

Operands: A ∈ [A,B]

Operation: If (MIN A) Then

If ACCA - ACCB < 0 Then

(ACCB → ACCA;

0 Z;1 N;0 OV;)

Else

(1 → Z; 0 → N; 0 → OV;)

If (MIN B) Then

If ACCB - ACCA < 0 Then

(ACCA → ACCB;

0 Z;1 N;0 OV;)

Else

(1 → Z; 0 → N; 0 → OV;)

Status Affected: N, OV, Z

Encoding:

1100

1110

A01x

x000

0000

0000

Description:

The target accumulator (defined in the instruction) is clamped to the minimum limit value previously loaded into the other accumulator. The comparison examines the full 40-bit value of the target accumulator, and will therefore, clamp an overflowed accumulator.

If the target accumulator is greater than the limit accumulator, load the target accumulator with the contents of the limit accumulator. The Z and N Status bits are set such that a subsequent BRA LT instruction will take a branch.

If the target accumulator is not less than the limit accumulator, the target accumulator is unaffected. The Z Status bit is set (Z = 1) such that a subsequent BRA z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 'A' bit specifies the destination accumulator.

The 'x' bits define the presence and result format for Wd.

Note: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC , #0 after MIN execution to update DSP status to reflect contents of AccX.

Words: 1

Cycles: 1

MIN.V

Accumulator Force Minimum Data Range Limit with Limit Excess Result (Unconditional Execution)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsP C33E dsPIC 33C
						X

Syntax: {label:} MIN.V

Acc

Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wd ∈ [W0 ... W15]

A ∈ [A,B]

Operation: If (MIN A) Then

If ACCA - ACCB < 0 Then

(0xFFFF → Wd or ACCA - ACCB → Wd (see text);

ACCB → ACCA;

0 Z;1 N;0 OV;)

Else

(0 → Wd;

1 Z;0 N;0 OV;)

If (MIN B) Then

If ACCB - ACCA < 0 Then

(0xFFFF → Wd or ACCB - ACCA → Wd (see text);

ACCA → ACCB;

0 Z;1 N;0 OV;)

Else

(0 → Wd;

1 Z;0 N;0 OV;)

Status Affected: N, OV, Z

Encoding:

Description:

1100 1110

A01x

x000

0qqq dddd

The target accumulator (defined in the instruction) is clamped to the minimum limit value previously loaded into the other accumulator. The comparison examines the full 40-bit value of the target accumulator.

If the target accumulator is greater than the limit accumulator, load the target accumulator with the contents of the limit accumulator. For MIN (instruction bit field xx = 2'b10), set Wd to -1. For MIN.V (instruction bit field xx = 2'b11), write the (signed) value by which the limit is exceeded to Wd. This is sourced from the Least Significant 16 bits of the 40-bit result. If the limit is exceeded by a value greater than that which can be represented by a signed 16-bit number, saturate the Wd write to the maximum negative value (i.e., set Wd to 0x8000).

The Z and N Status bits are set such that a subsequent BRA LT instruction will take a branch if the limit is exceeded.

If the target accumulator is not less than the limit accumulator, the target accumulator is unaffected and Wd is cleared. The Z Status bit is set such that a subsequent BRA 2 instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 'A' bit specifies the destination accumulator.

The 'd' bits select the address of the destination register.

The 'q' bits select the destination addressing mode.

The 'x' bits define the presence and result format for Wd.

Note: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC , #0 after MIN.V execution to update DSP status to reflect contents of AccX.

Words: 1

Cycles: 1

MINZ

Accumulator Force Minimum Data Range Limit (Conditional Execution)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
						X

Syntax: {label:} MINZ

Acc

Operands: Acc ∈ [A,B]

Operation:

If (Z = 0) execute as NOP

Else

If (MINZ A) Then

If ACCA - ACCB < 0 Then

ACCB → ACCA;

0 Z;1 N;0 OV;)

Else

1 → Z; 0 → N; 0 → OV;)

If (MINZ B) Then

If ACCB - ACCA < 0 Then

ACCA → ACCB;

0 Z;1 N;0 OV;)

Else

(0 Wd;

1 Z;0 N;0 OV;)

Status Affected: N, OV, Z

Encoding:

1100 1110

A01x

x100

0000 0000

Description:

If MINZ is executed when Z = 1 (see note), the target accumulator (defined in the instruction) is clamped to the minimum limit value previously loaded into the other accumulator. If MINZ is executed when Z = 0, the instruction is skipped (executed as a NOP).

The comparison examines the full 40-bit value of the target accumulator, and will therefore, clamp an overflowed accumulator.

If the target accumulator is less than the limit accumulator, load the target accumulator with the contents of the limit accumulator. The Z and N Status bits are set such that a subsequent BRA LT instruction will take a branch.

If the target accumulator is not less than the limit accumulator, the target accumulator is unaffected. The Z Status bit is set (Z = 1) such that a subsequent BRA Z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 'A' bit specifies the destination accumulator.

The 'x' bits define the presence and result format for Wd.

Note 1: Execution of the accumulator maximum clamp instruction (MAX) is expected to be immediately followed by execution of the conditionally executed accumulator minimum clamp instruction (MINZ). If MAX resulted in a clamp condition (Z = 0), MINZ will be skipped. Use the unconditionally executed MIN instruction if it is required to be executed in isolation.

2: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC , #0 after MINZ execution to update DSP status to reflect the contents of AccX.

Words: 1

Cycles: 1

MINZ.V
Accumulator Force Minimum Data Range Limit with Limit Excess Result (Conditional Execution)

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsP C33E dsPIC 33C
						X

Syntax: {label:} MINZ.V

Acc

Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wd ∈ [W0 ... W15]

A ∈ [A,B]

Operation:

If (Z = 0) execute as a NOP

Else

If (MINZ A) Then

If (ACCA - ACCB < 0 Then

(0xFFFF → Wd or ACCA - ACCB → Wd (see text);

ACCB → ACCA;

0 Z;1 N;0 OV;)

Else

(0 → Wd;

1 Z;0 N;0 OV;)

If (MINZ B) Then

If ACCB - ACCA < 0 Then

(0xFFFF → Wd or ACCB - ACCA → Wd (see text);

ACCA → ACCB;

0 Z;1 N;0 OV;)

Else

(0 → Wd;

1 Z;0 N;0 OV;)

Status Affected: N, OV, Z

Encoding:

1100 11

10

A01x

x100

0qqq dddd

MINZ.V

Accumulator Force Minimum Data Range Limit with Limit Excess Result (Conditional Execution)

Description:

If MINZ is executed when Z = 1 (see note), the target accumulator (defined in the instruction) is clamped to the minimum limit value previously loaded into the other accumulator. If MINZ is executed when Z = 0, the instruction is skipped (executed as a NOP).

The comparison examines the full 40-bit value of the target accumulator, and will therefore, clamp an overflowed accumulator.

If the target accumulator is less than the limit accumulator, load the target accumulator with the contents of the limit accumulator. For MINZ (instruction bit field xx = 2'b10), set Wd to -1. For MINZ.V (instruction bit field xx = 2'b11), write the (signed) value by which the limit is exceeded to Wd. This is sourced from the Least Significant 16 bits of the 40-bit result. If the limit is exceeded by a value greater than that which can be represented by a signed 16-bit number, saturate the Wd write to the maximum negative value (i.e., set Wd to 0x8000).

The Z and N Status bits are set such that a subsequent BRA LT instruction will take a branch if the limit is exceeded.

If the target accumulator is not less than the limit accumulator, the target accumulator is unaffected and Wd is cleared. The Z Status bit is set such that a subsequent BRA Z instruction will take a branch.

The OV Status bit is always cleared by this instruction.

The 'A' bit specifies the destination accumulator.

The 'd' bits select the address of the destination register.

The 'q' bits select the destination addressing mode.

The 'x' bits define the presence and result format for Wd.

Note 1: Execution of the MINZ.V instruction is intended to immediately follow execution of a MAX instruction. If MAX resulted in a clamp condition (Z = 0), the MINZ.V instruction will be skipped.

2: OA and SA or OB and SB Status bits are not modified by this instruction. Execute SFTAC , #0 after MINZ.V execution to update DSP status to reflect the contents of AccX.

Words: 1

Cycles: 1

MOV

Move f to Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F		dsPIC33F dsPIC33E dsPIC33C
X	X	X	X	X	X	X

Syntax: {label:} MOV{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) → destination designated by D

Status Affected: N, Z

Encoding:

Description:

1011 11

11 1BDf ffff ffff ffff

Move the contents of the specified file register to the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored back to the file register and the only effect is to modify the STATUS Register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.
3: When moving word data from file register memory, the "MOV f to Wnd" (page 301) instruction allows any Working register (W0:W15) to be the destination register.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: MOV.B TMRO, WREG ; move (TMRO) to WREG (Byte mode)

Before Instruction WREG (W0) 9080 WREG (W0) 9055 TMR0 2355 TMR0 2355 SR 0000 SR 0000

Example 2: MOV 0x800 ; update SR based on (0x800) (Word mode)

Before Instruction Data 0800 B29F SR 0000 After Instruction Data 0800 B29F SR 0008 (N = 1)

MOV

Move WREG to f

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} MOV{.B} WREG, f

Operands: f [0 8191]

Operation: (WREG) → f

Status Affected: None

Encoding:

1011

0111

1B1f

ffff

ffff

ffff

Description: Move the contents of the default Working register WREG into the specified file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .w extension to denote a word move, but it is not required.

2: The WREG is set to Working register W0.
3: When moving word data from the Working register array to file register memory, the "MOV Wns to f" (page 302) instruction allows any Working register (W0:W15) to be the source register.

Words: 1

Cycles: 1

Example 1: MOV.B WREG, 0x801 ; move WREG to 0x801 (Byte mode)

Example 2: MOV WREG, DISICNT ; move WREG to DISICNT

Before Instruction WREG (W0) 00A0 DISICNT 0000 SR 0000 After Instruction WREG (W0) 00A0 DISICNT 00A0 SR 0000

MOV

Move f to Wnd

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

X

Syntax: {label:} MOV f,

Wnd

Operands:

f ∈ [0 ... 65534]

Wnd ∈ [W0 ... W15]

Operation:

(f) → Wnd

Status Affected:

None

Encoding:

1000

0fff

ffff

ffff

ffff

dddd

Description:

Move the word contents of the specified file register to Wnd. The file register may reside anywhere in the 32K words of data memory, but must be word-aligned. Register Direct Addressing must be used for Wnd.

The 'f' bits select the address of the file register.

The 'd' bits select the destination register.

Note 1: This instruction operates on word operands only.

2: Since the file register address must be word-aligned, only the upper 15 bits of the file register address are encoded (bit 0 is assumed to be '0').
3: To move a byte of data from file register memory, the "MOV f to Destination" instruction (page 299) may be used.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MOV CORCON, W12 ; move CORCON to W12

Before Instruction W12 78FA CORCON 00F0 SR 0000 After Instruction W12 00F0 CORCON 00F0 SR 0000

Example 2: MOV 0x27FE, W3 ; move (0x27FE) to W3

Before Instruction W3 0035 Data 27FE ABCD SR 0000 After Instruction W3 ABCD Data 27FE ABCD SR 0000

MOV

Move Wns to f

Implemented in: PIC24F	PIC24H PIC2	4E dsPIC30F	dsPIC33F ds	PIC33E ds	C33C
	X	X		X	X		X

Syntax: {label:} MOV Wns, f

Operands: f ∈ [0 ... 65534] Wns ∈ [W0 ... W15]

Operation: (Wns) → f

Status Affected: None

Encoding:

1000

1fff

ffff

ffff

ffff

ssss

Description: Move the word contents of the Working register Wns to the specified file register. The file register may reside anywhere in the 32K words of data memory, but must be word-aligned. Register Direct Addressing must be used for Wn.

The 'f' bits select the address of the file register. The 's' bits select the source register.

Note 1: This instruction operates on word operands only.

2: Since the file register address must be word-aligned, only the upper 15 bits of the file register address are encoded (bit 0 is assumed to be '0').
3: To move a byte of data to file register memory, the "MOV WREG to f" instruction (page 300) may be used.

Words: 1

Cycles: 1

Example 1: MOV W4, XMDOSRT ; move W4 to XMODSRT

	Before Instruction		After Instruction
W4	1200		1200
XMODSRT	1340		1200
SR	0000		0000

Example 2: MOV W8, 0x1222 ; move W8 to data address 0x1222

	Before Instruction		After Instruction
W8	F200	W8	F200
Data 1222	FD88	Data 1222	F200
SR	0000	SR	0000

MOV.B

Move 8-Bit Literal to Wnd

Implemented in: PIC24F	PIC24H PIC	24E dsPIC30F	dsPIC33F	dsPIC33E dsP	IC33C
	X	X		X	X		X

X

Syntax: {label:} MOV.B #lit8, Wnd

Operands: lit8 ∈ [0 ... 255] Wnd ∈ [W0 ... W15]

Operation: lit8 → Wnd

Status Affected: None

Encoding:

1011

0011

1100

kkkk

kkkk

dddd

Description: The unsigned 8-bit literal 'k' is loaded into the lower byte of Wnd. The upper byte of Wnd is not changed. Register Direct Addressing must be used for Wnd.

The 'k' bits specify the value of the literal. The 'd' bits select the address of the Working register.

Note: This instruction operates in Byte mode and the .B extension must be provided.

Words: 1

Cycles: 1

Example 1: MOV.B #0x17, w5 ; load w5 with #0x17 (Byte mode)

Before
Instruction

W5	7899
SR	0000

After
Instruction

W5	7817
SR	0000

Example 2: MOV.B #0xFE, W9 ; load W9 with #0xFE (Byte mode)

Before
Instruction

W9	AB23
SR	0000

After
Instruction

W9	ABFE
SR	0000

MOV

Move 16-Bit Literal to Wnd

Implemented in: PIC24F	PIC24H PIC2	4E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} MOV #lit16, Wnd

Operands: lit16 ∈ [-32768 ... 65535]

$$ \text { Wnd } \in [ \text { W0 } \dots \text { W15 } ] $$

Operation: lit16 → Wnd

Status Affected: None

Encoding:

0010

kkkk

kkkk

kkkk

kkkk

dddd

Description: The 16-bit literal 'k' is loaded into Wnd. Register Direct Addressing must be used for Wnd.

The 'k' bits specify the value of the literal.

The 'd' bits select the address of the Working register.

Note 1: This instruction operates only in Word mode.

2: The literal may be specified as a signed value [-32768:32767] or unsigned value [0:65535].

Words: 1

Cycles: 1

Example 1: MOV #0x4231, W13 ; load W13 with #0x4231

Example 2: MOV #0x4, W2 ; load W2 with #0x4

Example 3: MOV #-1000, W8 ; load W8 with #-1000

MOV

Move [Ws with Offset] to Wnd

Implemented in: PIC24F	PIC24H	PIC24	4E dsPIC30F	dsPIC33F dsP	IC33E dsPIC	33C
	X		X	X	X	X	X	X

Syntax: {label:} MOV{.B} [Ws + Slit10], Wnd
Operands: Ws ∈ [W0 ... W15]
Slit10 ∈ [-512 ... 511] for byte operation

Slit10 ∈ [-1024 ... 1022] (even only) for word operation
Wnd ∈ [W0 ... W15]
Operation: [Ws + Slit10] → Wnd
Status Affected: None

Encoding:

1001

0kkk

kBkk

kddd

dkkk

ssss

Description: The contents of [Ws + Slit10] are loaded into Wnd. In Word mode, the range of Slit10 is increased to [-1024 ... 1022] and Slit10 must be even to maintain word address alignment. Register Indirect Addressing must be used for the source and Direct Addressing must be used for Wnd.
The 'k' bits specify the value of the literal.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'd' bits select the destination register.
The 's' bits select the source register.
Note 1: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .W extension to denote a word move, but it is not required.

2: In Byte mode, the range of Slit10 is not reduced as specified in Section 4.6 "Using 10-bit Literal Operands", since the literal represents an address offset from Ws.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1:

MOV.B

[W8+0x13], W10 ; load W10 with [W8+0x13] ; (Byte mode)

	Before Instruction		After Instruction
W8	1008	W8	1008
W10	4009	W10	4033
Data 101A	3312	Data 101A	3312
SR	0000	SR	0000

Example 2:

MOV

[W4+0x3E8], W2 ; load W2 with [W4+0x3E8] ; (Word mode)

	Before Instruction		After Instruction
W2	9088	W2	5634
W4	0800	W4	0800
Data 0BE8	5634	Data 0BE8	5634
SR	0000	SR	0000

MOV

Move Wns to [Wd with Offset]

Implemented in: PIC24F	PIC24H PIC2	4E dsPIC30F	dsPIC33F	dsPIC33E	dsPIC33C
	X	X		X		X

X X

Syntax: {label:} MOV{.B} Wns, [Wd + Slit10]
Operands: Wns ∈ [W0 ... W15]
Slit10 ∈ [-512 ... 511] in Byte mode
Slit10 ∈ [-1024 ... 1022] (even only) in Word mode
Wd ∈ [W0 ... W15]
Operation: (Wns) → [Wd + Slit10]
Status Affected: None

Encoding:

1001

1kkk

kBkk

kddd

dkkk

ssss

Description: The contents of Wns are stored to [Wd + Slit10] . In Word mode, the range of Slit10 is increased to [-1024 1022] and Slit10 must be even to maintain word address alignment. Register Direct Addressing must be used for Wns and Indirect Addressing must be used for the destination.
The 'k' bits specify the value of the literal.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'd' bits select the destination register.
The 's' bits select the source register.

Note 1: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .w extension to denote a word move, but it is not required.
2: In Byte mode, the range of Slit10 is not reduced as specified in Section 4.6 "Using 10-bit Literal Operands", since the literal represents an address offset from Wd.
Words: 1
Cycles: 1

Example 1:	MOV.B W0, [W1+0x7]	; store W0 to [W1+0x7]
		; (Byte mode)

Before Instruction		After Instruction
W0	9015	W0	9015
W1	1800	W1	1800
Data 1806	2345	Data 1806	1545
SR	0000	SR	0000

Example 2:	MOV	W11, [W1-0x400]	; store W11 to [W1-0x400]
			; (Word mode)

	Before Instruction		After Instruction
W1	1000	W1	1000
W11	8813	W11	8813
Data 0C00	FFEA	Data 0C00	8813
SR	0000	SR	0000

MOV
Move Ws to Wd

Implemented in: PIC24F P	IC24H PIC24E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} MOV{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[--Ws], [--Wd]

[++Ws], [++Wd]

[Ws + Wb], [Wd + Wb]

Operands: Ws ∈ [W0 ... W15]

Wb ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Ws) → Wd

Status Affected: None

Encoding:

0111

1www

wBhh

hddd

dggg

ssss

Description: Move the contents of the source register into the destination register. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits define the offset register Wb.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'h' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'g' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .w extension to denote a word move, but it is not required.

2: When Register Offset Addressing mode is used for both the source and destination, the offset must be the same because the 'w' encoding bits are shared by Ws and Wd.

3: The instruction, "PUSH Ws", translates to "MOV Ws, [W15++]".

4: The instruction, "POP Wd", translates to "MOV [--W15], Wd".

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MOV.B [W0--], W4 ; Move [W0] to W4 (Byte mode) ; Post-decrement W0

Before Instruction

After Instruction

W0

0A01

W0

0A00

W4

2976

W4

2989

Data 0A00

8988 Data 0A00 8988

SR

0000

SR

0000

Example 2: MOV [W6++, [W2+W3] ; Move [W6] to [W2+W3] (Word mode) ; Post-increment W6

	Before Instruction
	W2	0800	W2	0800
	W3	0040	W3	0040

MOV.D
Double-Word Move from Source to Wnd

Implemented in: PIC24F P	IC24H PIC2	4E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X

Syntax:	{label:}	MOV.D	Wns,	Wnd
				[Ws],
				[Ws++],
				[Ws--],
				[++Ws],
				[--Ws],

Operands:	Wns ∈ [W0, W2, W4 ... W14]
	Ws ∈ [W0 ... W15]
	Wnd ∈ [W0, W2, W4 ... W14]

Operation:	For Direct Addressing of Source:
	Wns → Wnd
	Wns + 1 → Wnd + 1
	For Indirect Addressing of Source
	See Description

Status Affected:

None

Encoding:

1011

1110

0000

0ddd

0ppp

ssss

Description: Move the double word specified by the source to a destination Working register pair (Wnd:Wnd + 1). If Register Direct Addressing is used for the source, the contents of two successive Working registers (Wns:Wns + 1) are moved to Wnd:Wnd + 1. If Indirect Addressing is used for the source, Ws specifies the Effective Address for the least significant word of the double word. Any pre/post-increment or pre/post-decrement will adjust Ws by 4 bytes to accommodate for the double word.

The 'd' bits select the destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the address of the first source register.

Note 1: This instruction only operates on double words. See Figure 4-3 for information on how double words are aligned in memory. 2: Wnd must be an even numbered Working register. 3: The instruction, "POP.D Wind", translates to "MOV.D [--W15], Wnd".

Words: 1

Cycles: 2^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MOV.D W2, W6 ; Move W2 to W6 (Double mode)

Before

Instruction

W2	12FB W2	12FB
W3	9877 W3	9877
W6	9833 W6	12FB
W7	FCC6 W7	9877
SR	0000 SR	0000

After

Instruction

Example 2: MOV.D [W7--], W4; Move [W7] to W4 (Double mode)

; Post-decrement W7

Before

Instruction

W4 B012 W4 A319

W5 FD89 W5 9927

W7 0900 W7 08FC

Data 0900 A319 Data 0900 A319

Data 0902 9927 Data 0902 9927

SR 0000

After

Instruction

00

MOVPAG
Move Literal to Page Register

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax: {label:}

MOVPAG

lit10,

DSRPAG

lit9,

DSWPAG

lit8,

TBLPAG

Operands:

lit10 ∈ [0 ... 1023], lit9 ∈ [0 ... 511], lit8 ∈ [0 ... 255]

Operation:

lit10 → DSRPAG or lit9 → DSWPAG or lit8 → TBLPAG

Status Affected:

None

Encoding:

1111

1110

1100

PPkk

kkkk

kkkk

Description:

The appropriate number of bits from the unsigned literal 'k' is loaded into the DSRPAG, DSWPAG or TBLPAG register. The assembler restricts the literal to a 9-bit unsigned value when the destination is DSWPAG and an 8-bit unsigned value when the destination is TBLPAG.

The 'P' bits select the destination register.

The 'k' bits specify the value of the literal.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1

Example 1: MOVPAG #0x02, DSRPAG

Before Instruction	After Instruction
DSRPAG 0000	DSRPAG 0002

MOVPAG

Move Wn to Page Register

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax: {label:} MOVPAG Wn, DSRPAG

DSWPAG

TBLPAG

Operands: Wn ∈ [W0 ... W15]

Operation: Wn<9:0> → DSRPAG or Wn<8:0> → DSWPAG or Wn<7:0> → TBLPAG

Status Affected: None

Encoding:

1111

1110

1101

PP00

0000

ssss

Description: The appropriate number of bits from the register Ws is loaded into the DSRPAG, DSWPAG or TBLPAG register. The assembler restricts the literal to a 9-bit unsigned value when the destination is DSWPAG and an 8-bit unsigned value when the destination is TBLPAG.

The 'P' bits select the destination register.

The 's' bits specify the source register.

Note: This instruction operates in word mode only.

Words: 1

Cycles: 1

Example 1: MOVPAG W2, DSRPAG

	Before Instruction		After Instruction
DSRPAG	0000	DSRPAG	0002
W2	0002	W2	0002

MOVSAC

Prefetch Operands and Store Accumulator

Implemented in: PIC24F P	IC24H PIC24E dsPIC30F	dsPIC33F ds	PIC33E dsP	IC33C
			X	X	X	X

Syntax: {label:} MOVSAC Acc {,[Wx], Wxd} {,[Wy], Wyd} {,AWB}

$$ {, [ W x ] + = k x, W x d } {, [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } \quad {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } \quad {, [ W 1 1 + W 1 2 ], W y d } $$

Operands: Acc ∈ [A,B]

$$ \mathrm{Wx} \in [ \mathrm{W8}, \mathrm{W9} ]; \mathrm{kx} \in [ - 6, - 4, - 2, 2, 4, 6 ]; \mathrm{Wxd} \in [ \mathrm{W4} \dots \mathrm{W7} ] $$

$$ \mathrm{Wy} \in [ \mathrm{W10}, \mathrm{W11} ]; \mathrm{ky} \in [ - 6, - 4, - 2, 2, 4, 6 ]; \mathrm{Wyd} \in [ \mathrm{W4} \dots \mathrm{W7} ] $$

$$ \mathrm{AWB} \in [ \mathrm{W13}, [ \mathrm{W13} ] + = 2 ] $$

Operation: ([Wx]) → Wxd; (Wx) + kx → Wx

$$ ([ W y ]) \rightarrow W y d; (W y) + k y \rightarrow W y $$

$$ (\text { Acc } (B \text { or } A)) \text { rounded } \rightarrow \text { AWB } $$

Status Affected: None

Encoding:

1100

0111

A0xx

yyii

iijj

jjaa

Description:

Optionally prefetch operands in preparation for another MAC type instruction and optionally store the unspecified accumulator results. Even though an accumulator operation is not performed in this instruction, an accumulator must be specified to designate which accumulator to Write-Back.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing, as described in Section 4.15.1 "MAC Prefetches". Operand AWB specifies the optional store of the "other" accumulator, as described in Section 4.15.4 "MAC Write-Back".

The 'A' bit selects the other accumulator used for Write-Back.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

The 'a' bits select the accumulator Write-Back destination.

Words: 1

Cycles: 1

Example 1: MOVSAC B, [W9], W6, [W11] += 4, W7, W13

; Fetch [W9] to W6 ; Fetch [W11] to W7, Post-increment W11 by 4 ; Store ACCA to W13

Before
Instruction

W6	A022 W6 7811
W7	B200 W7 B2AF
W9	0800 W9 0800
W11	1900 W11 1904
W13	0020 W13 3290
ACCA	00 3290 5968 ACCA 00 3290 5968
Data 0800	7811 Data 0800 7811
Data 1900	B2AF Data 1900 B2AF
SR	0000 SR 0000

After
Instruction

Example 2: MOVSAC A, [W9] == 2, W4, [W11+W12], W6, [W13] += 2

; Fetch [W9] to W4, Post-decrement W9 by 2

; Fetch [W11+W12] to W6

; Store ACCB to [W13], Post-increment W13 by 2

Before
Instruction

W4	76AE	W4 BB00
W6	2000	W6	52CE
W9	1200	W9	11FE
W11	2000	W11	2000
W12	0024	W12	0024
W13	2300	W13	2302
ACCB	00 9834 4500	ACCB	00 9834 4500
Data 1200	BB00 Data	1200 BB00
Data 2024	52CE	Data 2024	52CE
Data 2300	23FF	Data 2300	9834
SR	0000	SR	0000

After
Instruction

MPY

Multiply Wm by Wn to Accumulator

Implemented in: PIC24F	PIC24H PIC2	4E dsPIC30F	dsPIC33F	dsPIC33E	dsPIC33C
				X	X	X	X

Syntax: {label:} MPY Wm * Wn, Acc {,[Wx], Wxd} {,[Wy], Wyd}

$$ {., [ W x ] + = k x, W x d } {., [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } \quad {, [ W 1 1 + W 1 2 ], W y d } $$

Operands: Wm * Wn ∈ [W4 * W5, W4 * W6, W4 * W7, W5 * W6, W5 * W7, W6 * W7]

$$ \mathrm{Acc} \in [ \mathrm{A}, \mathrm{B} ] $$

$$ \mathrm{Wx} \in [ \mathrm{W8}, \mathrm{W9} ]; \mathrm{kx} \in [ - 6, - 4, - 2, 2, 4, 6 ]; \mathrm{Wxd} \in [ \mathrm{W4} \dots \mathrm{W7} ] $$

$$ \mathrm{Wy} \in [ \mathrm{W10}, \mathrm{W11} ]; \mathrm{ky} \in [ - 6, - 4, - 2, 2, 4, 6 ]; \mathrm{Wyd} \in [ \mathrm{W4} \dots \mathrm{W7} ] $$

$$ \mathrm{AWB} \in [ \mathrm{W13} ], [ \mathrm{W13} ] + = 2 $$

Operation: (Wm) * (Wn) → Acc(A or B)

$$ ([ W x ]) \rightarrow W x d; (W x) + k x \rightarrow W x $$

$$ ([ W y ]) \rightarrow W y d; (W y) + k y \rightarrow W y $$

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

0mmm

A0xx

yyii

iijj

jj11

Description: Multiply the contents of two Working registers and optionally prefetch operands in preparation for another MAC type instruction. The 32-bit result of the signed multiply is sign-extended to 40 bits and stored to the specified accumulator.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations which support Indirect and Register Offset Addressing, as described in Section 4.15.1 "MAC Prefetches".

The 'm' bits select the operand registers, Wm and Wn, for the multiply.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

Note 1: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

2: The US<1:0> bits (CORCON<13:12> in dsPIC33E/dsPIC33C, CORCON<12> in dsPIC30F/dsPIC33F) determine if the multiply is unsigned, signed or mixed-sign. Only dsPIC33E/dsPIC33C devices support mixed-sign multiplication.

Words: 1

Cycles: 1

Example 1:

MPY W4*W5, A, [W8]+=2, W6, [W10]==2, W7; Multiply W4*W5 and store to ACCA; Fetch [W8] to W6, Post-increment W8 by 2; Fetch [W10] to W7, Post-decrement W10 by 2; CORCON = 0x0000 (fractional multiply, no saturation)

	Before Instruction			After Instruction
W4	C000 W4	C000
W5	9000 W5	9000
W6	0800 W6	671F
W7	B200 W7	E3DC
W8	1780 W8	1782
W10	2400 W10	23FE
ACCA	FF F780	2087 ACCA	00 3800 0000
Data 1780		671F	Data 1780	671F
Data 2400	E3DC	Data 2400	E3DC
CORCON	0000	CORCON	0000
SR	0000		SR 0000

Example 2:

MPY W6*W7, B, [W8]+=2, W4, [W10]==2, W5; Multiply W6*W7 and store to ACCB; Fetch [W8] to W4, Post-increment W8 by 2; Fetch [W10] to W5, Post-decrement W10 by 2; CORCON = 0x0000 (fractional multiply, no saturation)

	Before Instruction		After Instruction
W4	C000	W6	8FDC
W5	9000		W5	0078
W6	671F		671F
W7	E3DC		W7	E3DC
W8	1782		W8	1784
W10	23FE		W10	23FC
ACCB	00 9834 4500	ACCB	FF E954 3748
Data 1782	8FDC Data 1782 8FDC
Data 23FE	0078 Data 23FE 0078
CORCON	0000	CORCON	0000
SR	0000	SR	0000

MPY

Square to Accumulator

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Syntax: {label:} MPY Wm * Wm, Acc {,[Wx], Wxd} {,[Wy], Wyd}

$$ {, [ W x ] + = k x, W x d } \quad {, [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } \quad {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } \quad {, [ W 1 1 + W 1 2 ], W y d } $$

Operands: Wm * Wm ∈ [W4 * W4, W5 * W5, W6 * W6, W7 * W7]

Operation: (Wm) * (Wm) → Acc(A or B)

$$ ([ W x ]) \rightarrow W x d; (W x) + k x \rightarrow W x $$

$$ ([ W y ]) \rightarrow W y d; (W y) + k y \rightarrow W y $$

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1111

00mm

A0xx

yyii

iijj

jj01

Description: Square the contents of a Working register and optionally prefetch operands in preparation for another MAC type instruction. The 32-bit result of the signed multiply is sign-extended to 40 bits and stored in the specified accumulator.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing, as described in Section 4.15.1 "MAC Prefetches".

The 'm' bits select the operand register Wm for the square.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

Note 1: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

2: The US<1:0> bits (CORCON<13:12> in dsPIC33E/dsPIC33C, CORCON<12> in dsPIC30F/dsPIC33F) determine if the multiply is unsigned, signed or mixed-sign. Only dsPIC33E/dsPIC33C devices support mixed-sign multiplication.

Words: 1

Cycles: 1

Example 1: MPY W6*W6, A, [W9]+=2, W6 ; Square W6 and store to ACCA ; Fetch [W9] to W6, Post-increment W9 by 2 ; CORCON = 0x0000 (fractional multiply, no saturation)

Before Instruction W6 6500 W6 B865 W9 0900 W9 0902 ACCA 00 7C80 0908 ACCA 00 4FB2 0000 Data 0900 B865 Data 0900 B865 CORCON 0000 CORCON 0000 SR 0000 SR 0000 After Instruction

Example 2: MPY W4*W4, B, [W9+W12], W4, [W10]+=2, W5 ; Square W4 and store to ACCB ; Fetch [W9+W12] to W4 ; Fetch [W10] to W5, Post-increment W10 by 2 ; CORCON = 0x0000 (fractional multiply, no saturation)

MPY.N

Multiply -Wm by Wn to Accumulator

Implemented in: PIC24F	PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
				X	X	X

Syntax: {label:} MPY.N Wm * Wn, Acc {,[Wx], Wxd}	{,[Wy], Wyd}
	{,[Wx] += kx, Wxd} {,[Wy] += ky, Wyd}
	{,[Wx] -= kx, Wxd} {,[Wy] -= ky, Wyd}
	{,[W9 + W12], Wxd} {,[W11 + W12], Wyd}

Operands: Wm * Wn ∈ [W4 * W5; W4 * W6; W4 * W7; W5 * W6; W5 * W7; W6 * W7] Acc ∈ [A,B]

Wx ∈ [W8, W9]; kx ∈ [-6, -4, -2, 2, 4, 6]; Wxd ∈ [W4 ... W7]

Wy ∈ [W10, W11]; ky ∈ [-6, -4, -2, 2, 4, 6]; Wyd ∈ [W4 ... W7]

Operation: -(Wm)^*(Wn) Acc(A or B) ([Wx]) Wxd;(Wx)+kx Wx ([Wy]) Wyd;(Wy)+ky Wy

Status Affected: OA, OB, OAB

Encoding: 1100 0mmm A1xx yyii iijj jj11

Description: Multiply the contents of a Working register by the negative of the contents of another Working register. Optionally prefetch operands in preparation for another MAC type instruction and optionally store the unspecified accumulator results. The 32-bit result of the signed multiply is sign-extended to 40 bits and stored to the specified accumulator.

The 'm' bits select the operand registers, Wm and Wn, for the multiply.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

Note 1: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

2: The US<1:0> bits (CORCON<13:12> in dsPIC33E/dsPIC33C, CORCON<12> in dsPIC30F/dsPIC33F) determine if the multiply is unsigned, signed or mixed-sign. Only dsPIC33E/dsPIC33C devices support mixed-sign multiplication.

Words: 1

Cycles: 1

Example 1:

MPY.N W4*W5, A, [W8]+=2, W4, [W10]+=2, W5; Multiply W4*W5, negate the result and store to ACCA; Fetch [W8] to W4, Post-increment W8 by 2; Fetch [W10] to W5, Post-increment W10 by 2; CORCON = 0x0001 (integer multiply, no saturation)

	Before Instruction		After Instruction
W4	3023 W4	0054
W5	1290 W5	660A
W8	0B00 W8	0B02
W10	2000 W10	2002
ACCA	00 0000	2387 ACCA	FF FC82 7650
Data 0B00	0054 Data	0B00	0054
Data 2000	660A Data	2000	660A
CORCON	0001 CORCON		0001
SR	0000		SR 0000

Example 2:

MPY.N W4*W5, A, [W8]+=2, W4, [W10]+=2, W5; Multiply W4*W5, negate the result and store to ACCA; Fetch [W8] to W4, Post-increment W8 by 2; Fetch [W10] to W5, Post-increment W10 by 2; CORCON = 0x0000 (fractional multiply, no saturation)

	Before Instruction					After Instruction
W4	3023 W4	0054
W5	1290 W5	660A
W8	0B00 W8	0B02
W10	2000 W10	2002
ACCA	00 0000	2387 ACCA	FF F904 ECA0
Data 0B00	0054 Data	0B00	0054
Data 2000	660A Data	2000	660A
CORCON	0000 CORCON		0000
SR	0000		SR 0000

MSC

Multiply and Subtract from Accumulator

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X	X	X	X

Syntax: {label:} MSC Wm * Wn, Acc {,[Wx], Wxd} {,[Wy], Wyd} {,AWB}

$$ {, [ W x ] + = k x, W x d } {, [ W y ] + = k y, W y d } $$

$$ {, [ W x ] - = k x, W x d } {, [ W y ] - = k y, W y d } $$

$$ {, [ W 9 + W 1 2 ], W x d } {, [ W 1 1 + W 1 2 ], W y d } $$

Operands: Wm * Wn ∈ [W4 * W5, W4 * W6, W4 * W7, W5 * W6, W5 * W7, W6 * W7]

AWB ∈ [W13, [W13] += 2]

Operation: (Acc(A or B)) - (Wm) * (Wn) → Acc(A or B)

([Wx]) Wxd; (Wx) + kx Wx

([Wy]) Wyd;(Wy)+ky Wy

(Acc(B or A)) rounded → AWB

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

0mmm

A1xx

yyii

iijj

jjaa

Description: Multiply the contents of two Working registers. Optionally prefetch operands in preparation for another MAC type instruction and optionally store the unspecified accumulator results. The 32-bit result of the signed multiply is sign-extended to 40 bits and subtracted from the specified accumulator.

Operands, Wx, Wxd, Wy and Wyd, specify optional prefetch operations, which support Indirect and Register Offset Addressing as described in Section 4.15.1 "MAC Prefetches". Operand AWB specifies the optional store of the "other" accumulator as described in Section 4.15.4 "MAC Write-Back".

The 'm' bits select the operand registers, Wm and Wn, for the multiply.

The 'A' bit selects the accumulator for the result.

The 'x' bits select the prefetch Wxd destination.

The 'y' bits select the prefetch Wyd destination.

The 'i' bits select the Wx prefetch operation.

The 'j' bits select the Wy prefetch operation.

The 'a' bits select the accumulator Write-Back destination.

Note: The IF bit (CORCON<0>) determines if the multiply is fractional or an integer.

Words: 1

Cycles: 1

Example 1: MSC W6*W7, A, [W8]-=4, W6, [W10]-=4, W7

; Multiply W6*W7 and subtract the result from ACCA

; Fetch [W8] to W6, Post-decrement W8 by 4

; Fetch [W10] to W7, Post-decrement W10 by 4

; CORCON = 0x0001 (integer multiply, no saturation)

	Before Instruction			After Instruction
W6	9051 W6	D309
W7	7230 W7	100B
W8	0C00 W8	0BFC
W10	1C00 W10	1BFC
ACCA	00 0567	8000	ACCA 00 3738	5ED0
Data 0C00	D309 Data	0C00	D309
Data 1C00		100B	Data 1C00	100B
CORCON	0001	CORCON	0001
SR	0000		SR 0000

Example 2: MSC W4*W5, B, [W11+W12], W5, W13

; Multiply W4*W5 and subtract the result from ACCB

; Fetch [W11+W12] to W5

; Write Back ACCA to W13

; CORCON = 0x0000 (fractional multiply, no saturation)

	Before Instruction			After Instruction
W4	0500 W4	0500
W5	2000 W5	3579
W11	1800		W11	1800
W12	0800 W12	0800
W13	6233 W13	3738
ACCA	00 3738 5ED0	ACCA	00 3738 5ED0
ACCB	00 1000 0000	ACCB	00 0EC0 0000
Data 2000	3579	Data	2000 3579
CORCON	0000	CORCON	0000
SR	0000		SR 0000

MUL

Integer Unsigned Multiply f and WREG

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} MUL{.B} f

Operands: f [0 8191]

Operation: For Byte Operation:

$$ (W R E G) < 7: 0 > ^ {*} (f) < 7: 0 > \rightarrow W 2 $$

For Word Operation:

$$ (\text { WREG }) ^ {*} (f) \rightarrow \text { W2:W3 } $$

Status Affected: None

Encoding:

1011

1100

0B0f

ffff

ffff

ffff

Description:

Multiply the default Working register WREG with the specified file register and place the result in the W2:W3 register pair. Both operands and the result are interpreted as unsigned integers. If this instruction is executed in Byte mode, the 16-bit result is stored in W2. In Word mode, the most significant word of the 32-bit result is stored in W3 and the least significant word of the 32-bit result is stored in W2.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.
3: The IF bit (CORCON<0>) has no effect on this operation.
4: This is the only instruction which provides for an 8-bit multiply.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.B 0x800 ; Multiply (0x800)*WREG (Byte mode)

	Before Instruction		After Instruction
WREG (W0)	9823	WREG (W0)	9823
W2	FFFF	W2	13B0
W3	FFFF	W3	FFFF
Data 0800	2690	Data 0800	2690
SR	0000	SR	0000

Example 2: MUL TMR1 ; Multiply (TMR1)*WREG (Word mode)

	Before Instruction				After Instruction
WREG (W0)	F001	WREG (W0)	F001
W2	0000	W2	C287
W3	0000	W3	2F5E
TMR1	3287	TMR1	3287
SR	0000	SR	0000

MUL.SS

Integer 16x16-Bit Signed Multiply

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} MUL.SS Wb, Ws, Wnd
[Ws],
[Ws++]
[Ws--],
[++Ws],

[--Ws],
Operands: Wb ∈ [W0 ... W15]
Ws ∈ [W0 ... W15]
Wnd ∈ [W0, W2, W4 ... W12]
Operation: signed (Wb) * signed (Ws) → Wnd:Wnd + 1
Status Affected: None

Encoding:

1011

1001

lwww

wddd

dppp

ssss

Description: Multiply the contents of Wb with the contents of Ws and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. Both source operands and the result Wnd are interpreted as two's complement signed integers. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.
The 'w' bits select the address of the base register.
The 'd' bits select the address of the lower destination register.
The 'p' bits select the source addressing mode.
The 's' bits select the source register.
Note 1: This instruction operates in Word mode only.
2: Since the product of the multiplication is 32 bits, Wnd must be an even Working register. See Figure 4-2 for information on how double words are aligned in memory.
3: Wnd may not be W14, since W15<0> is fixed to zero.
4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multi-Cycle Instructions”.

Example 1:	MUL.SS W0, W1, W12	; Multiply W0*W1
		; Store the result to W12:W13

Before
Instruction

W0	9823
W1	67DC
W12	FFFF
W13	FFFF
SR	0000

After
Instruction

W0	9823
W1	67DC
W12	D314
W13	D5DC
SR	0000

Example 2: MUL.SS W2, [--W4], W0 ; Pre-decrement W4 ; Multiply W2*[W4] ; Store the result to W0:W1

Before

Instruction

W0 FFFF W0 28F8

W1 FFFF W1 0000

W2 0045 W2 0045

W4 27FE W4 27FC

Data 27FC 0098 Data 27FC 0098

SR 0000 SR 0000

After

Instruction

MUL.SS

Integer 16x16-Bit Signed Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.SS Wb, Ws, Acc

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Acc ∈ [A, B]

Operation: signed (Wb) * signed (Ws) → Acc(A or B)

Status Affected: None

Encoding:

1011

1001

lwww

w111

Appp

ssss

Description:

Performs a 16-bit x 16-bit signed multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is sign-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). Both source operands are treated as signed values.

The 'w' bits select the address of the base register.

The 'p' bits select Source Addressing Mode 2.

The 'A' bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.SS W0, W1, A

Before

Instruction

W0	9823
W1	67DC
ACCA	00 0000 0000
SR	0000

After

Instruction

W0	9823
W1	67DC
ACCA	FF D5DC D314
SR	0000

MUL.SU

Integer 16x16-Bit Signed-Unsigned Short Literal Multiply

Implemented in: PIC24F P	C24H PIC24E dsPIC30F dsPIC33F ds			PIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} MUL.SU Wb, #lit5, Wnd

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wnd ∈ [W0, W2, W4 ... W12]

Operation: signed (Wb) * unsigned lit5 → Wnd:Wnd + 1

Status Affected: None

Encoding:

1011

1001

0www

wddd

d11k

kkkk

Description: Multiply the contents of Wb with the 5-bit literal and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. The Wb operand and the result Wnd are interpreted as a two's complement signed integer. The literal is interpreted as an unsigned integer. Register Direct Addressing must be used for Wb and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the address of the lower destination register.

The 'k' bits define a 5-bit unsigned integer literal.

Note 1: This instruction operates in Word mode only.

2: Since the product of the multiplication is 32 bits, Wnd must be an even numbered Working register. See Figure 4-3 for information on how double words are aligned in memory.
3: Wnd may not be W14, since W15<0> is fixed to zero.
4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words: 1

Cycles: 1

Example 1: MUL.SU W0, #0x1F, W2 ; Multiply W0 by literal 0x1F

; Store the result to W2:W3

Before
Instruction

W0	C000
W2	1234
W3	C9BA
SR	0000

After
Instruction

W0 C000
W2	4000
W3	FFF8
SR	0000

Example 2: MUL.SU W2, #0x10, W0 ; Multiply W2 by literal 0x10

; Store the result to W0:W1

Before
Instruction

W0	ABCD
W1	89B3
W2	F240
SR	0000

After
Instruction

W0	2400
W1	000F
W2	F240
SR	0000

MUL.SU

Integer 16x16-Bit Signed-Unsigned Multiply

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

X

Syntax:	{label:}	MUL.SU	Wb,	Ws,	Wnd
					[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands:	Wb ∈ [W0 ... W15]
	Ws ∈ [W0 ... W15]
	Wnd ∈ [W0, W2, W4 ... W12]

Operation:	signed (Wb) * unsigned (Ws) → Wnd:Wnd + 1
Status Affected:	None

Encoding:

1011

1001

0www

wddd

dppp

ssss

Description:

Multiply the contents of Wb with the contents of Ws and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. The Wb operand and the result Wnd are interpreted as a two's complement signed integer. The Ws operand is interpreted as an unsigned integer. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.

The ‘w’ bits select the address of the base register.

The ‘d’ bits select the address of the lower destination register.

The ‘p’ bits select the source addressing mode.

The ‘s’ bits select the source register.

Note 1: This instruction operates in Word mode only.

2: Since the product of the multiplication is 32 bits, Wnd must be an even Working register. See Figure 4-3 for information on how double words are aligned in memory.

3: Wnd may not be W14, since W15<0> is fixed to zero.

4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words:	1
Cycles:	1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.SU W8, [W9], W0 ; Multiply W8*[W9] ; Store the result to W0:W1

Before Instruction
W0	68DC W0	0000
W1	AA40 W1	F100
W8	F000 W8	F000
W9	178C W9	178C
Data 178C	F000 Data 178C F000
SR	0000 SR	0000

After Instruction

Example 2: MUL.SU W2, [++W3], W4 ; Pre-Increment W3 ; Multiply W2*[W3] ; Store the result to W4:W5

	Before Instruction		After Instruction
W2	0040 W2	0040
W3	0280 W3	0282
W4	1819 W4	1A00
W5	2021 W5	0000
Data 0282	0068	Data 0282	0068
SR	0000 SR	0000

MUL.SU

Integer 16x16-Bit Signed-Unsigned Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.SU Wb, Ws, Acc

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Acc ∈ [A, B]

Operation: signed (Wb) * unsigned (Ws) → Acc(A or B)

Status Affected: None

Encoding:

1011

1001

0www

w111

Appp

ssss

Description:

Performs a 16-bit x 16-bit signed multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is sign-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). The first source operand is interpreted as a two's complement signed value and the second source operand is interpreted as an unsigned value.

The 'w' bits select the address of the base register.

The 'p' bits select Source Addressing Mode 2.

The 'A' bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.SU W8, W9, A

Before

Instruction

W8	F000
W9	F000
ACCA	00 0000 0000
SR	0000

After

Instruction

W8	F000
W9	F000
ACCA	FF F100 0000
SR	0000

MUL.SU

Integer 16x16-Bit Signed-Unsigned Short Literal Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.SU Wb, #lit5, Acc

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 5 \in [ 0 \dots 3 1 ] $$

$$ \mathrm{Acc} \in [ \mathrm{A}, \mathrm{B} ] $$

Operation: signed (Wb) * unsigned (lit5) → Acc(A or B)

Status Affected: None

Encoding:

1011

1001

0www

w111

A11k

kkkk

Description:

Performs a 16-bit x 16-bit signed multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is sign-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). The first source operand is interpreted as a two's complement signed value and the second source operand is interpreted as an unsigned value.

The 'w' bits select the address of the base register.

The 'k' bits select the 5-bit literal value.

The 'A' bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1

Example 1: MUL.SU W8, #0x02, A

Before
Instruction

W8	0042
ACCA	00 0000 0000
SR	0000

After
Instruction

W8	0042
ACCA	00 0000 0084
SR	0000

MUL.US
Integer 16x16-Bit Unsigned-Signed Multiply

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax:	{label:}	MUL.US	Wb,	Ws,	Wnd
					[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands:	Wb ∈ [W0 ... W15]
	Ws ∈ [W0 ... W15]
	Wnd ∈ [W0, W2, W4 ... W12]

Operation:	unsigned (Wb) * signed (Ws) → Wnd:Wnd + 1
Status Affected:	None

Encoding:

1011

1000

1www

wddd

dppp

ssss

Description:

Multiply the contents of Wb with the contents of Ws and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. The Wb operand is interpreted as an unsigned integer. The Ws operand and the result Wnd are interpreted as a two's complement signed integer. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.

The ‘w’ bits select the address of the base register.

The ‘d’ bits select the address of the lower destination register.

The ‘p’ bits select the source addressing mode.

The ‘s’ bits select the source register.

Note 1: This instruction operates in Word mode only.

2: Since the product of the multiplication is 32 bits, Wnd must be an even numbered Working register. See Figure 4-3 for information on how double words are aligned in memory.

3: Wnd may not be W14, since W15<0> is fixed to zero.

4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words:	1
Cycles:	_1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.US W0, [W1], W2 ; Multiply W0*[W1] (unsigned-signed) ; Store the result to W2:W3

Before Instruction
W0	C000 W0	C000
W1	2300 W1	2300
W2	00DA W2	0000
W3	CC25 W3	F400
Data 2300	F000 Data	2300 F000
SR	0000 SR	0000

After Instruction

Example 2: MUL.US W6, [W5++], W10 ; Mult. W6*[W5] (unsigned-signed) ; Store the result to W10:W11 ; Post-Increment W5

	Before Instruction		After Instruction
	W5	0C00 W5 0C02
	W6	FFFF W6 FFFF
	W10	0908 W10 8001
	W11	6EEB	W11 7FFE
	Data 0C00	7FFF Data 0C00 7FFF
	SR	0000 SR 0000

MUL.US

Integer 16x16-Bit Unsigned-Signed Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.US Wb, Ws, Acc

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Acc ∈ [A, B]

Operation: unsigned (Wb) * signed (Ws) → Acc(A or B)

Status Affected: None

Encoding:

1011

1000

0www

w111

Appp

ssss

Description:

Performs a 16-bit x 16-bit signed multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is sign-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). The first source operand is interpreted as an unsigned value and the second source operand is interpreted as a two's complement signed value.

The 'w' bits select the address of the base register.

The 'p' bits select Source Addressing Mode 2.

The ‘A’ bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.US W0, W1, B

Before

Instruction

W0	C000
W1	F000
ACCB	00 0000 0000
SR	0000

After

Instruction

W0	0000
W1	F000
ACCB	FF F400 0000
SR	0000

MUL.UU

Integer 16x16-Bit Unsigned Short Literal Multiply

Implemented in: PIC24F PIC	24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} MUL.UU Wb, #lit5, Wnd
Operands: Wb ∈ [W0 ... W15]
lit5 ∈ [0 ... 31]

Wnd ∈ [W0, W2, W4 ... W12]
Operation: unsigned (Wb) * unsigned lit5 → Wnd:Wnd + 1
Status Affected: None

Encoding:

1011

1000

0www

wddd

dllk

kkkk

Description: Multiply the contents of Wb with the 5-bit literal and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. Both operands and the result are interpreted as unsigned integers. Register Direct Addressing must be used for Wb and Wnd.
The 'w' bits select the address of the base register.
The 'd' bits select the address of the lower destination register.
The 'k' bits define a 5-bit unsigned integer literal.
Note 1: This instruction operates in Word mode only.
2: Since the product of the multiplication is 32 bits, Wnd must be an even Working register. See Figure 4-3 for information on how double words are aligned in memory.
3: Wnd may not be W14, since W15<0> is fixed to zero.
4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words: 1
Cycles: 1

Example 1:

MUL.UU W0, #0xF, W12 ; Multiply W0 by literal 0xF; Store the result to W12:W13

Before
Instruction

W0	2323
W12	4512
W13	7821
SR	0000

After
Instruction

W0	2323
W12	0F0D
W13	0002
SR	0000

Example 2:

MUL.UU W7, #0x1F, W0 ; Multiply W7 by literal 0x1F

; Store the result to W0:W1

Before
Instruction

W0	780B
W1	3805
W7	F240
SR	0000

After
Instruction

W0	55C0
W1	001D
W7	F240
SR	0000

MUL.UU
Integer 16x16-Bit Unsigned Multiply

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax:	{label:}	MUL.UU	Wb,	Ws,	Wnd
					[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands:	Wb ∈ [W0 ... W15]
	Ws ∈ [W0 ... W15]
	Wnd ∈ [W0, W2, W4 ... W12]

Operation:	unsigned (Wb) * unsigned (Ws) → Wnd:Wnd + 1
Status Affected:	None

Encoding:

1011

1000

0www

wddd

dppp

ssss

Description:

Multiply the contents of Wb with the contents of Ws and store the 32-bit result in two successive Working registers. The least significant word of the result is stored in Wnd (which must be an even numbered Working register) and the most significant word of the result is stored in Wnd + 1. Both source operands and the result are interpreted as unsigned integers. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Indirect Addressing may be used for Ws.The ‘w’ bits select the address of the base register.The ‘d’ bits select the address of the lower destination register.The ‘p’ bits select the source addressing mode.The ‘s’ bits select the source register.

Note 1: This instruction operates in Word mode only.

2: Since the product of the multiplication is 32 bits, Wnd must be an even numbered Working register. See Figure 4-3 for information on how double words are aligned in memory.

3: Wnd may not be W14, since W15<0> is fixed to zero.

4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words:	1
Cycles:	_1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.UU W4, W0, W2 ; Multiply W4*W0 (unsigned-unsigned) ; Store the result to W2:W3

Before

Instruction
W0	FFFF W0	FFFF
W2	2300 W2	0001
W3	00DA W3	FFFE
W4	FFFF W4	FFFF
SR	0000 SR	0000

After

Instruction

Example 2: MUL.UU W0, [W1++, W4 ; Mult. W0*[W1] (unsigned-unsigned)

; Store the result to W4:W5 ; Post-Increment W1

Before

Instruction
W0	1024 W0	1024
W1	2300 W1	2302
W4	9654 W4	6D34
W5	BDBC W5	0D80
Data 2300	D625 Data	2300 D625
SR	0000 SR	0000

After

Instruction

MUL.UU

Integer 16x16-Bit Unsigned Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.UU Wb, Ws, Acc

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Acc ∈ [A, B]

Operation: unsigned (Wb) * unsigned (Ws) → Acc(A or B)

Status Affected: None

Encoding:

1011

1000

0www

w111

Appp

ssss

Description:

Performs a 16-bit x 16-bit unsigned multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is zero-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). Both source operands are treated as unsigned values.

The 'w' bits select the address of the base register.

The 'p' bits select Source Addressing Mode 2.

The 'A' bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MUL.UU W4, W0, B

Before

Instruction

W0	FFFFFF
W4	FFFFFF
ACCB	00 0000 0000
SR	0000

After

Instruction

W0	FFFFFF
W4	FFFFFF
ACCB	FF FFFE 0001
SR	0000

MUL.UU

Integer 16x16-Bit Unsigned Short Literal Multiply with Accumulator Destination

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
					X	X

Syntax: {label:} MUL.UU Wb, #lit5, Acc

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 5 \in [ 0 \dots 3 1 ] $$

$$ \mathrm{Acc} \in [ \mathrm{A}, \mathrm{B} ] $$

Operation: unsigned (Wb) * unsigned (lit5) → Acc(A or B)

Status Affected: None

Encoding:

1011

1000

0www

w111

A11k

kkkk

Description: Performs a 16-bit x 16-bit signed multiply with a 32-bit result, which is stored in one of the DSP engine accumulators: ACCA or ACCB. The 32-bit result is zero-extended to bit 39 prior to being loaded into the target accumulator.

The source operands are treated as integer or fractional values depending upon the operating mode of the DSP engine (as defined by the IF bit in CORCON<0>). Both source operands are treated as unsigned values.

The 'w' bits select the address of the base register.

The 'k' bits select the 5-bit literal.

The 'A' bit selects the destination accumulator for the product.

Note 1: This instruction operates in Word mode only.

2: The state of the Multiplier mode bits (US<1:0> in CORCON) has no effect upon the operation of this instruction.

Words: 1

Cycles: 1

Example 1: MUL.UU W8, #0x02, A

Before
Instruction

W8	0042
ACCA	00 0000 0000
SR	0000

After
Instruction

W8	0042
ACCA	00 0000 0084
SR	0000

MULW.SS

Integer 16x16-Bit Signed Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax:	{label:}	MULW.SS	Wb,	Ws,	Wnd
					[Ws],
					[Ws++],
					[Ws--],
					[++Ws],
					[--Ws],

Operands:	Wb ∈ [W0 ... W15]
	Ws ∈ [W0 ... W15]
	Wnd ∈ [W0, W2, W4 ... W12]

Operation:

signed (Wb) * signed (Ws) → Wnd

Status Affected:

None

Encoding:

1011

1001

lwww

wddd

dppp

ssss

Description:

Multiply the contents of Wb with the contents of Ws and store the result in a Working register, which must be an even numbered Working register. Both source operands and the result Wnd are interpreted as two's complement signed integers. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.

The ‘w’ bits select the address of the base register.

The ‘d’ bits select the address of the lower destination register.

The ‘p’ bits select the source addressing mode.

The ‘s’ bits select the source register.

Note 1: This instruction operates in Word mode only.

2: Wnd must be an even numbered Working register.

3: Wnd may not be W14, since W15<0> is fixed to zero.

4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words:

1

Cycles:

1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: MULW.SS W0, W1, W12 ; Multiply W0*W1 ; Store the result to W12

Before Instruction
W0	9823 W0	9823
W1	67DC W1	67DC
W12	FFFF W1	D314
SR	0000 SR	0000

After Instruction

Example 2: MULW.SS W2, [--W4], W0 ; Pre-decrement W4 ; Multiply W2*[W4] ; Store the result to W0

	Before Instruction		After Instruction
	W0	FFFF W0	28F8
	W2	0045	W2 0045
	W4	27FE	W4 27FC
	Data 27FC	0098	Data 27FC
	SR	0000 SR	0000

MULW.SU

Integer 16x16-Bit Signed-Unsigned Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
			X			X	X
Syntax: {label:} MULW.SU Wb, Ws, Wnd [Ws], [Ws++], [Ws--], [++Ws], [--Ws],
Operands: Wb ∈ [W0 ... W15] Ws ∈ [W0 ... W15] Wnd ∈ [W0, W2, W4 ... W12]
Operation: signed (Wb) * unsigned (Ws) → Wnd
Status Affected: None
Encoding: 1011	1001	0www	wddd	dppp	ssss
Description: Multiply the contents of Wb with the contents of Ws and store the result in a Working register, which must be an even numbered Working register. The Wb operand and the result Wnd are interpreted as a two's complement signed integer. The Ws operand is interpreted as an unsigned integer. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.The 'w' bits select the address of the base register.The 'd' bits select the address of the lower destination register.The 'p' bits select the source addressing mode.The 's' bits select the source register.Note 1: This instruction operates in Word mode only.2: Wnd must be an even numbered Working register.3: Wnd may not be W14, since W15<0> is fixed to zero.4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.
Words: 1
Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multi-Cycle Instructions”.

Example 1: MULW.SU W8, [W9], W0 ; Multiply W8*[W9] ; Store the result to W0

	Before Instruction						After Instruction
	W0	68DC W0	0000
	W8	F000 W8	F000
	W9	178C W9	178C
	Data 178C	F000 Data 178C F000
	SR	0000 SR	0000

Example 2: MULW.SU W2, [++W3], W4 ; Pre-Increment W3 ; Multiply W2*[W3] ; Store the result to W4

	Before Instruction		After Instruction
W2	0040 W2	0040
W3	0280 W3	0282
W4	1819 W4	1A00
Data 0282	0068	Data 0282	0068
SR	0000 SR	0000

MULW.SU

Integer 16x16-Bit Signed-Unsigned Short Literal Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X

Syntax: {label:} MULW.SU Wb, #lit5, Wnd
Operands: Wb ∈ [W0 ... W15]
lit5 ∈ [0 ... 31]

Wnd ∈ [W0, W2, W4 ... W12]
Operation: signed (Wb) * unsigned (lit5) → Wnd
Status Affected: None

Encoding:

1011

1001

0www

wddd

d11k

kkkk

Description: Multiply the contents of Wb with a 5-bit literal value and store the result in a Working register, which must be an even numbered Working register. The Wb operand and the result Wnd are interpreted as a two's complement signed integer. Register Direct Addressing must be used for Wb and Wnd.
The 'w' bits select the address of the base register.
The 'd' bits select the address of the lower destination register.
The 'k' bits select the 5-bit literal value.
Note 1: This instruction operates in Word mode only.
2: Wind must be an even numbered Working register.
3: Wnd may not be W14, since W15<0> is fixed to zero.
4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words: 1
Cycles: 1

Example 1:	MULW.SU W8, #0x04, W0	; Multiply W8 * #0x04
		; Store the result to W0

Before Instruction		After Instruction
W0	68DC	W0	4000
W8	1000	W8	1000
SR	0000	SR	0000

MULW.US

Integer 16x16-Bit Unsigned-Signed Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X
Syntax:{label:}	MULW.US	Wb,	Ws,	Wnd[Ws],[Ws++],[Ws--],[++Ws],[-Ws],
Operands:	Wb ∈ [W0 ... W15]Ws ∈ [W0 ... W15]Wnd ∈ [W0, W2, W4 ... W12]
Operation:	unsigned (Wb) * signed (Ws) → Wnd
Status Affected:	None
Encoding:	1011	1000	1www	wddd	dppp	ssss
Description:	Multiply the contents of Wb with the contents of Ws and store the result in a Working register, which must be an even numbered Working register. The Wb operand is interpreted as an unsigned integer. The Ws operand and the result Wnd are interpreted as a two's complement signed integer. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Register Indirect Addressing may be used for Ws.The 'w' bits select the address of the base register.The 'd' bits select the address of the lower destination register.The 'p' bits select the source addressing mode.The 's' bits select the source register.Note 1: This instruction operates in Word mode only.2: Wnd must be an even numbered Working register.3: Wnd may not be W14, since W15<0> is fixed to zero.4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.
Words:	1
Cycles:	_1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multi-Cycle Instructions”.

Example 1: MULW.US W0, [W1], W2 ; Multiply W0*[W1] (unsigned-signed) ; Store the result to W2

Before Instruction
W0	C000 W0	C000
W1	2300 W1	2300
W2	00DA W2	0000
Data 2300	F000 Data	2300 F000
SR	0000 SR	0000

After Instruction

Example 2: MULW.US W6, [W5++, W10 ; Mult. W6*[W5] (unsigned-signed)

; Store the result to W10

; Post-Increment W5

	Before Instruction					After Instruction
	W5	0C00	W5	0C02
	W6	FFFF	W6	FFFF
	W10	0908			W10	8001
	Data 0C00	7FFF		Data 0C00	7FFF
	SR	0000	SR	0000

MULW.UU

Integer 16x16-Bit Unsigned Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
		X			X	X
Syntax: {label:} MULW.UU Wb, Ws, Wnd [Ws], [Ws++], [Ws--], [++Ws], [--Ws],
Operands: Wb ∈ [W0 ... W15] Ws ∈ [W0 ... W15] Wnd ∈ [W0, W2, W4 ... W12]
Operation: unsigned (Wb) * unsigned (Ws) → Wnd
Status Affected: None
Encoding: 1011	1000	0www	wddd	dppp	ssss
Description: Multiply the contents of Wb with the contents of Ws and store the result in a Working register, which must be an even numbered Working register. Both source operands and the result are interpreted as unsigned integers. Register Direct Addressing must be used for Wb and Wnd. Register Direct or Indirect Addressing may be used for Ws.The ‘w’ bits select the address of the base register.The ‘d’ bits select the address of the lower destination register.The ‘p’ bits select the source addressing mode.The ‘s’ bits select the source register.Note 1: This instruction operates in Word mode only.2: Wnd must be an even numbered Working register.3: Wnd may not be W14, since W15<0> is fixed to zero.4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.
Words: 1
Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multi-Cycle Instructions”.

Example 1:

MULW.UU W4, W0, W2 ; Multiply W4*W0 (unsigned-unsigned) ; Store the result to W2

	Before Instruction			After Instruction
W0	FFFF		W0	FFFF
W2	2300		W2	0001
W4	FFFF		W4	FFFF
SR	0000		SR	0000

MULW.UU

Integer 16x16-Bit Unsigned Short Literal Multiply with 16-Bit Result

Implemented in: PIC24F PIC24H PIC24	E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
		X		X	X

Syntax: {label:} MULW.UU Wb, #lit5, Wnd

Operands: Wb ∈ [W0 ... W15] lit5 ∈ [0 ... 31] Wnd ∈ [W0, W2, W4 ... W12]
Operation: unsigned (Wb) * unsigned → Wnd
Status Affected: None

Encoding:

1011

1000

0www

wddd

d11k

kkkk

Description: Multiply the contents of Wb with a 5-bit literal value and store the result in a Working register, which must be an even numbered Working register. Both source operands and the result are interpreted as unsigned integers. Register Direct Addressing must be used for Wb and Wnd.
The 'w' bits select the address of the base register.
The 'd' bits select the address of the lower destination register.
The 'k' bits select the 5-bit literal value.
Note 1: This instruction operates in Word mode only.
2: Wind must be an even numbered Working register.
3: Wnd may not be W14, since W15<0> is fixed to zero.
4: The IF bit and the US<1:0> bits in the CORCON register have no effect on this operation.

Words: 1
Cycles: 1

Example 1:	MULW.UU W4, #0x04, W2	; Multiply W4*W0 (unsigned-unsigned)
		; Store the result to W2

Before

	Instruction
W2	2300
W4	1000
SR	0000

After

	Instruction
W2	4000
W4	1000
SR	0000

NEG

Negate f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} NEG{.B} f {,WREG}
Operands: f [0 8191]

Operation: (f) + 1 destination designated by D
Status Affected: DC, N, OV, Z, C

Encoding:

1110

1110

OBDF

ffff

ffff

ffff

Description:

Compute the two's complement of the contents of the file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1:	NEG.B 0x880, WREG	; Negate (0x880) (Byte mode)
		; Store result to WREG

	Before Instruction	After Instruction
WREG (W0)	9080	90AB
Data 0880	2355	2355
SR	0000	0008(N = 1)

Example 2:

NEG

0x1200

; Negate (0x1200) (Word mode)

	Before Instruction		After Instruction
Data 1200	8923	Data 1200	76DD
SR	0000	SR	0000

NEG

Negate Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} NEG{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Ws)+1 Wd

Status Affected: DC, N, OV, Z, C

Encoding: 1110 1010 0Bqq qddd dppp ssss

Description: Compute the two's complement of the contents of the source register Ws and place the result in the destination register Wd. Either Register Direct or Indirect Addressing may be used for both Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: NEG.B W3, [W4++] ; Negate W3 and store to [W4] (Byte mode) ; Post-increment W4

Before Instruction W3 7839 W3 7839 W4 1005 W4 1006 Data 1004 2355 Data 1004 C755 SR 0000 SR 0008 (N = 1) After Instruction

Example 2: NEG [W2++], [--W4] ; Pre-decrement W4 (Word mode) ; Negate [W2] and store to [W4] ; Post-increment W2

Before Instruction W2 0900 W2 0902 W4 1002 W4 1000 Data 0900 870F Data 0900 870F Data 1000 5105 Data 1000 78F1 SR 0000 SR 0000 After Instruction

NEG

Negate Accumulator

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} NEG Acc

Operands: Acc ∈ [A,B]

Operation: If (Acc = A):

$$ \overline {{- \mathrm{ACCA} \rightarrow \mathrm{ACCA}}} $$

$$ \begin{array}{c}\underline {{\text { Else:}}}\- \text { ACCB } \rightarrow \text { ACCB }\end{array} $$

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1011

A001

0000

0000

0000

Description: Compute the two's complement of the contents of the specified accumulator. Regardless of the Saturation mode, this instruction operates on all 40 bits of the accumulator.

The 'A' bit specifies the selected accumulator.

Words: 1

Cycles: 1

Example 1:	NEG A	; Negate ACCA
		; Store result to ACCA
		; CORCON = 0x0000 (no saturation)

	Before Instruction
ACCA	00 3290 59C8
CORCON	0000
SR	0000

	After Instruction
ACCA FF	CD6F A638
CORCON	0000
SR	0000

Example 2:	NEG B	; Negate ACCB
		; Store result to ACCB
		; CORCON = 0x00C0 (normal saturation)

	Before Instruction
ACCB	FF F230 10DC
CORCON	00C0
SR	0000

	After Instruction
ACCB	00 0DCF EF24
CORCON	00C0
SR	0000

NOP

No Operation

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	C33F ds	PIC33E ds	PIC33C
	X	X		X	X		X

X X

Syntax: {label:} NOP

Operands: None

Operation: No Operation

Status Affected: None

Encoding:

No Operation is performed.

The 'x' bits can take any value.

Words: 1

Cycles: 1

Example 1: NOP ; execute no operation

	Before Instruction		After Instruction
PC	00 1092	PC	00 1094
SR	0000	SR	0000

Example 2: NOP ; execute no operation

	Before Instruction		After Instruction
PC	00 08AE	PC	00 08B0
SR	0000	SR	0000

NOPR

No Operation

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

×

Syntax: {label:} NOPR

Operands: None

Operation: No Operation

Status Affected: None

Encoding:

No Operation is performed.

The 'x' bits can take any value.

Words: 1

Cycles: 1

Example 1: NOPR ; execute no operation

	Before Instruction
PC	00 2430
SR	0000

	After Instruction
PC	00 2432
SR	0000

Example 2: NOPR ; execute no operation

	Before Instruction
PC	00 1466
SR	0000

	After Instruction
PC	00 1468
SR	0000

NORM

Normalize Accumulator

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC33C
						X

Syntax: {label:} NORM Acc, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wnd ∈ [W0 ... W15]

Acc ∈ [A,B]

Operation: Refer to text.

Status Affected: OA, SA or OB, SB, N, Z

Encoding:

1100

1110

A110

0000

0qqq

dddd

Description:

Normalize the contents of the target accumulator. If the accumulator contains an overflowed value, the contents of the accumulator are shifted right by the minimum number of bits required to remove the overflow. If the accumulator does not contain an overflowed value, the contents of the accumulator are shifted left by the minimum number of bits required to produce the largest fractional data value without an overflow.

If it is not possible to normalize the target accumulator (i.e., it is already normalized, or it is all '0's or all '1's), Wd is cleared, the Z bit is set and the N bit is cleared. The target accumulator is unaffected.

If it is possible to normalize the target accumulator, the exponent (shift value) required to normalize the target accumulator is written into Wd. A positive result indicates that a right shift of the accumulator was required for normalization. A negative result indicates that a left shift of the accumulator was required for normalization. The N bit is set to reflect the sign of the result and the Z bit is cleared.

The 'A' bit specifies the destination accumulator.

The 'd' bits select the address of the destination register.

The 'q' bits select Destination Address Mode 2.

Note 1: OA and SA or OB and SB Status bits are set based on the content of the target accumulator. Consequently, as the NORM instruction removes any overflow, OA or OB will always be cleared.

2: The SA/SB bits will remain set if they were already set prior to execution of the NORM instruction, but these bits can never be affected by this instruction.

Words: 1

Cycles: 1

POP

Pop TOS to f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} POP f

Operands: f [0 65534]

Operation: (W15) - 2 → W15

(TOS) → f

Status Affected: None

Encoding:

1111

1001

ffff

ffff

ffff

fff0

Description: The Stack Pointer (W15) is pre-decremented by 2 and the Top-of-Stack (TOS) word is written to the specified file register, which may reside anywhere in the lower 32K words of data memory.

The 'f' bits select the address of the file register.

Note 1: This instruction operates in Word mode only.

2: The file register address must be word-aligned.

Words: 1

Cycles: 1

Example 1: POP 0x1230 ; Pop TOS to 0x1230

Before
Instruction

W15	1006
Data 1004	A401
Data 1230	2355
SR	0000

After
Instruction

W15	1004
Data 1004	A401
Data 1230	A401
SR	0000

Example 2: POP 0x880 ; Pop TOS to 0x880

Before
Instruction

W15	2000
Data 0880	E3E1
Data 1FFE	A090
SR	0000

After
Instruction

W15	1FFE
Data 0880	A090
Data 1FFE	A090
SR	0000

POP

Pop TOS to Wd

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} POP Wd

[Wd]

[Wd++]

[Wd--]

[--Wd]

[++Wd]

[Wd+Wb]

Operands: Wd ∈ [W0 ... W15]

Wb ∈ [W0 ... W15]

Operation: (W15) - 2 → W15

(TOS) → Wd

Status Affected:

None

Encoding:

0111

1www

w0hh

hddd

d100

1111

Description:

The Stack Pointer (W15) is pre-decremented by 2 and the Top-of-Stack (TOS) word is written to Wd. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits define the offset register Wb.

The 'h' bits select the destination addressing mode.

The 'd' bits select the destination register.

Note 1: This instruction operates in Word mode only.

2: This instruction is a specific version of the "MOV Ws, Wd" instruction, (MOV [--W15], Wd); it reverse assembles as MOV.

Words: 1

Cycles: 1

Example 1: POP W4 ; Pop TOS to W4

Before
Instruction

W4	EDA8
W15	1008
Data 1006	C45A
SR	0000

After
Instruction

W4	C45A
W15	1006
Data 1006	C45A
SR	0000

Example 2: POP [++W10] ; Pre-increment W10

; Pop TOS to [W10]

Before

Instruction
W10	0E02
W15	1766
Data 0E04	E3E1
Data 1764	C7B5
SR	0000

After
Instruction

W10	0E04
W15	1764
Data 0E04	C7B5
Data 1764	C7B5
SR	0000

POP.D

Double Pop TOS to Wnd:Wnd+1

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} POP.D Wnd

Operands: Wnd ∈ [W0, W2, W4, ... W14]

Operation: (W15) - 2 → W15

$$ (T O S) \rightarrow W n d + 1 $$

$$ (W 1 5) - 2 \rightarrow W 1 5 $$

$$ (T O S) \rightarrow W n d $$

Status Affected: None

Encoding:

Description:

1011

1110

0000

0ddd

0100

1111

A double word is POPped from the Top-of-Stack (TOS) and stored to Wnd:Wnd + 1. The most significant word is stored to Wnd + 1 and the least significant word is stored to Wnd. Since a double word is POPped, the Stack Pointer (W15) gets decremented by 4.

The 'd' bits select the address of the destination register pair.

Note 1: This instruction operates on double words. See Figure 4-3 for information on how double words are aligned in memory.

2: Wind must be an even numbered Working register.

3: This instruction is a specific version of the "MOV.D Ws, Wnd" instruction, (MOV.D [--W15], Wnd); it reverse assembles as MOV.D.

Words: 1

Cycles: 2

Example 1: POP.D W6 ; Double pop TOS to W6

Before Instruction		After Instruction
W6	07BB	W6	3210
W7	89AE	W7	7654
W15	0850	W15	084C
Data 084C	3210	Data 084C	3210
Data 084E	7654	Data 084E	7654
SR	0000	SR	0000

Example 2: POP.D W0 ; Double pop TOS to W0

Before Instruction		After Instruction
W0	673E	W0	791C
W1	DD23	W1	D400
W15	0BBC	W15	0BB8
Data 0BB8	791C	Data 0BB8	791C
Data 0BBA	D400	Data 0BBA	D400
SR	0000	SR	0000

POP.S

Pop Shadow Registers

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X

X X

Syntax: {label:} POP.S

Operands: None

Operation: POP shadow registers.

Status Affected:

Encoding:

Description:

DC, N, OV, Z, C

1111

1110

1000

0000

0000

0000

The values in the shadow registers are copied into their respective primary registers. The following registers are affected: W0-W3, and the C, Z, OV, N and DC STATUS Register flags.

Note 1: The shadow registers are not directly accessible. They may only be accessed with PUSH.S and POP.S.

2: The shadow registers are only one-level deep.

Words: 1

Cycles: 1

Example 1: POP.S ; Pop the shadow registers ; (See PUSH.S Example 1 for contents of shadows)

Before Instruction
W0	07BB
W1	03FD
W2	9610
W3	7249
SR	00E0 (IPL = 7)

	After Instruction
W0	0000
W1	1000
W2	2000
W3	3000
SR	00E1 (IPL = 7, C = 1)

Note: After instruction execution, the contents of shadow registers are NOT modified.

PUSH

Push f to TOS

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X	X		X

X

Syntax: {label:} PUSH f

Operands: f [0 65534]

Operation: (f) → (TOS)

(W15) + 2 → W15

Status Affected: None

Encoding:

1111

1000

ffff

ffff

ffff

fff0

Description: The contents of the specified file register are written to the Top-of-Stack (TOS) location and then the Stack Pointer (W15) is incremented by 2.

The file register may reside anywhere in the lower 32K words of data memory.

The 'f' bits select the address of the file register.

Note 1: This instruction operates in Word mode only.

2: The file register address must be word-aligned.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: PUSH 0x2004 ; Push (0x2004) to TOS

	Before Instruction
W15	0B00
Data 0B00	791C
Data 2004	D400
SR	0000

	After Instruction
W15	0B02
Data 0B00	D400
Data 2004	D400
SR	0000

Example 2: PUSH 0xC0E ; Push (0xC0E) to TOS

	Before Instruction
W15	0920
Data 0920	0000
Data 0C0E	67AA
SR	0000

	After Instruction
W15	0922
Data 0920	67AA
Data 2004	67AA
SR	0000

PUSH
Push Ws to TOS

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	C33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} PUSH Ws

[Ws]

[Ws++]

[Ws--]

[--Ws]

[++Ws]

[Ws+Wb]

Operands: Ws ∈ [W0 ... W15] Wb ∈ [W0 ... W15]

Operation: (Ws) → (TOS) (W15) + 2 → W15

Status Affected: None

Encoding:

0111

1www

w001

1111

1ggg

ssss

Description:

The contents of Ws are written to the Top-of-Stack (TOS) location and then the Stack Pointer (W15) is incremented by 2.

The 'w' bits define the offset register Wb.

The 'g' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: This instruction operates in Word mode only.

2: This instruction is a specific version of the "MOV Ws, Wd" instruction, (MOV Ws, [W15++]); it reverse assembles as MOV.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: PUSH W2 ; Push W2 to TOS

Before

Instruction

W2 6889 W2 6889

W15 1566 W15 1568

Data 1566 0000 Data 1566 6889

SR 0000 SR 0000

After

Instruction

Example 2: PUSH [W5+W10] ; Push [W5+W10] to TOS

Before

Instruction

W5 1200 W5 1200

W10 0044 W10 0044

W15 0806 W15 0808

Data 0806 216F Data 0806 B20A

Data 1244 B20A Data 1244 B20A

SR 0000 SR 0000

After

Instruction

PUSH.D

Double Push Wns:Wns+1 to TOS

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X

Syntax: {label:} PUSH.D Wns

Operands: Wns ∈ [W0, W2, W4 ... W14]

Operation: (Wns) → (TOS)

(W15) + 2 → W15

(Wns + 1) → (TOS)

(W15) + 2 → W15

Status Affected: None

Encoding:

A double word (Wns:Wns + 1) is PUSHed to the Top-of-Stack (TOS). The least significant word (Wns) is PUSHed to the TOS first and the most significant word (Wns + 1) is PUSHed to the TOS last. Since a double word is PUSHed, the Stack Pointer (W15) gets incremented by 4.

Description:

The 's' bits select the address of the source register pair.

Note 1: This instruction operates on double words. See Figure 4-3 for information on how double words are aligned in memory.
2: Wns must be an even numbered Working register.
3: This instruction is a specific version of the "MOV.D Wns, Wd" instruction, (MOV.D Wns, [W15++]); it reverse assembles as MOV.D.

Words: 1

Cycles: 2

Example 1: PUSH.D W6 ; Push W6:W7 to TOS

Before Instruction		After Instruction
W6	C451	W6	C451
W7	3380	W7	3380
W15	1240	W15	1244
Data 1240	B004	Data 1240	C451
Data 1242	0891	Data 1242	3380
SR	0000	SR	0000

Example 2: PUSH.D W10 ; Push W10:W11 to TOS

	Before Instruction		After Instruction
W10	80D3	W10	80D3
W11	4550	W11	4550
W15	0C08	W15	0C0C
Data 0C08	79B5	Data 0C08	80D3
Data 0C0A	008E	Data 0C0A	4550
SR	0000	SR	0000

PUSH.S

Push Shadow Registers

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	C33F ds	PIC33E ds	PIC33C
	X	X		X	X		X

×

Syntax: {label:} PUSH.S

Operands: None

Operation: Push shadow registers.

Status Affected: None

Encoding:

1111

1110

1010

0000

0000

0000

Description:

The contents of the primary registers are copied into their respective shadow registers. The following registers are shadowed: W0-W3, and the C, Z, OV, N and DC STATUS Register flags.

Note 1: The shadow registers are not directly accessible. They may only be accessed with PUSH.S and POP.S.

2: The shadow registers are only one-level deep.

Words: 1

Cycles: 1

Example 1: PUSH.S ; Push primary registers into shadow registers

Before Instruction		After Instruction
W0	0000
W1	1000	W0	0000
W2	2000	W1	1000
W3	3000	W2	2000
SR	0001	W3	3000
		SR	0001

Note: After an instruction execution, the contents of the shadow registers are updated.

PWRSAV

Enter Power-Saving Mode

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} PWRSAV #lit1

Operands: lit1 ∈ [0,1]

Operation:	0 → WDT Count register
	0 → WDT Prescaler A count
	0 → WDT Prescaler B count
	0 → WDTO (RCON<4>)
	0 → SLEEP (RCON<3>)
	0 → IDLE (RCON<2>)

If (lit1 = 0):

Enter Sleep mode

Else:

Enter Idle mode

Status Affected:

Encoding:

Description:

None

1111

1110

0100

0000

0000

000k

Place the processor into the specified power-saving mode. If lit1 = 0, Sleep mode is entered. In Sleep mode, the clock to the CPU and peripherals is shut down. If an on-chip oscillator is being used, it is also shut down. If lit1 = 1, Idle mode is entered. In Idle mode, the clock to the CPU shuts down, but the clock source remains active and the peripherals continue to operate.

This instruction resets the Watchdog Timer Count register and the Prescaler Count registers. In addition, the WDTO, SLEEP and IDLE flags of the Reset System and Control register (RCON) are reset.

Note 1: The processor will exit from Idle or Sleep through an interrupt, processor Reset or Watchdog Timer time-out. See the specific device data sheet for details.
2: If awakened from Idle mode, the IDLE bit (RCON<2>) is set to '1' and the clock source is applied to the CPU.
3: If awakened from Sleep mode, the SLEEP bit (RCON<3>) is set to '1' and the clock source is started.
4: If awakened from a Watchdog Timer time-out, the WDTO bit (RCON<4>) is set to '1'.

Words: 1

Cycles: 1

Example 1: PWRSAV #0 ; Enter SLEEP mode

Before

Instruction

$$ \mathrm{SR} \boxed {0 0 4 0} (\mathrm{IPL} = 2) $$

After

Instruction

$$ \mathrm{SR} \boxed {0 0 4 0} (\mathrm{IPL} = 2) $$

Example 2: PWRSAV #1 ; Enter IDLE mode

Before

Instruction

$$ \mathrm{SR} \boxed {0 0 2 0} (\mathrm{IPL} = 1) $$

After

Instruction

$$ \mathrm{SR} \boxed {0 0 2 0} (\mathrm{IPL} = 1) $$

RCALL

Relative Call

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPI	C33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} RCALL Expr

Operands: Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... 32767].

Operation: (PC) + 2 → PC

$$ (\mathrm{PC} < 1 5: 0 >) \rightarrow (\mathrm{TOS}) $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ (P C < 2 2: 1 6 >) \rightarrow (T O S) $$

$$ (W 1 5) + 2 \rightarrow W 1 5 $$

$$ (P C) + (2 * S l i t 1 6) \rightarrow P C $$

$$ \mathrm{NOP} \rightarrow \text { Instruction Register } $$

Status Affected: None

Encoding:

0000

0111

nnnn

nnnn

nnnn

nnnn

Description:

Relative subroutine call with a range of 32K program words forward or backward from the current PC. Before the call is made, the return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, the sign-extended 17-bit value (2 * Slit16) is added to the contents of the PC and the result is stored in the PC.

The 'n' bits are a signed literal that specifies the size of the relative call (in program words) from (PC + 2) .

Note: When possible, this instruction should be used instead of CALL, since it only consumes one word of program memory.

Words: 1

Cycles: 2

Example 1:	012004 RCALL _Task1	; Call _Task1
	012006 ADD W0, W1, W2
	. . . .
	. . . .
	012458 _Task1: SUB W0, W2, W3	; _Task1 subroutine
	01245A ...

Before Instruction PC 01 2004 PC 01 2458 W15 0810 W15 0814 Data 0810 FFFF Data 0810 2006 Data 0812 FFFF Data 0812 0001 SR 0000 SR 0000

Example 2:	00620E RCALL _Init	; Call _Init
	006210 MOV W0, [W4++]
	. . . .
	. . . .
	007000 _Init: CLR W2	; _Init subroutine
	007002 ...

Before Instruction PC 00 620E PC 00 7000 W15 0C50 W15 0C54 Data 0C50 FFFF Data 0C50 6210 Data 0C52 FFFF Data 0C52 0000 SR 0000 SR 0000

RCALL

Relative Call

Implemented in: PIC24F PIC24	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
			X			X	X

Syntax: {label:} RCALL Expr

Operands: Expr may be an absolute address, label or expression.

Expr is resolved by the linker to a Slit16, where Slit16 ∈ [-32768 ... 32767].

Operation: (PC) + 2 → PC

(PC<15:1>) → TOS<15:1>, SFA Status bit → TOS<0>

(W15) + 2 → W15

(PC<22:16>) → (TOS)

(W15) + 2 → W15

0 → SFA Status bit

(PC) + (2 * Slit16) → PC

NOP → Instruction Register

Status Affected: SFA

Encoding:

0000

0111

nnnn

nnnn

nnnn

nnnn

Description:

Relative subroutine call with a range of 32K program words forward or backward from the current PC. Before the call is made, the return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, the sign-extended 17-bit value (2 * Slit16) is added to the contents of the PC and the result is stored in the PC.

The 'n' bits are a signed literal that specifies the size of the relative call (in program words) from (PC + 2) .

Note: When possible, this instruction should be used instead of CALL, since it only consumes one word of program memory.

Words: 1

Cycles: 4

Example 1:	012004 RCALL _Task1	; Call _Task1
	012006 ADD W0, W1, W2
	. . . .
	. . . .
	012458 _Task1: SUB W0, W2, W3	; _Task1 subroutine
	01245A ...

Before Instruction PC 01 2004 PC 01 2458 W15 0810 W15 0814 Data 0810 FFFF Data 0810 2006 Data 0812 FFFF Data 0812 0001 SR 0000 SR 0000

Example 2:	00620E CALL _Init ; Call _Init
	006210 MOV W0, [W4++]
	. . . .
	. . . .
	007000 _Init: CLR W2 ; _Init subroutine
	007002 ...

Before Instruction PC 00 620E PC 00 7000 W15 0C50 W15 0C54 Data 0C50 FFFF Data 0C50 6210 Data 0C52 FFFF Data 0C52 0000 SR 0000 SR 0000

RCALL

Computed Relative Call

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} RCALL Wn

Operands: Wn ∈ [W0 ... W15]

Operation: (PC) + 2 → PC

(PC<15:0>) → (TOS)

(W15) + 2 → W15

(PC<22:16>) → (TOS)

(W15) + 2 → W15

(PC) + (2 * (Wn)) → PC

NOP → Instruction Register

Status Affected: None

Encoding:

Computed, relative subroutine call specified by the Working register Wn. The range of the call is 32K program words forward or backward from the current PC. Before the call is made, the return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, the sign-extended 17-bit value (2 * (Wn)) is added to the contents of the PC and the result is stored in the PC. Register Direct Addressing must be used for Wn.

The 's' bits select the source register.

Words: 1

Cycles: 2

Example 1:	00FF8C EX1:	INC	W2, W3	; Destination of RCALL
	00FF8E	...
	.	...
	.	...
	010008
	01000A	RCALL	W6	; RCALL with W6
	01000C	MOVE	W4, [W10]

Before Instruction		After Instruction
PC	01 000A	PC	00 FF8C
W6	FFC0	W6	FFC0
W15	1004	W15	1008
Data 1004	98FF	Data 1004	000C
Data 1006	2310	Data 1006	0001
SR	0000	SR	0000

Example 2: 000302 RCALL W2

000304 FF1L W0, W1

. . . .

. . . .

000450 EX2: CLR W2

000452 ...

; RCALL with W2

; Destination of RCALL

Before
Instruction

PC 00 0302 PC 00 0450 W2 00A6 W2 00A6 W15 1004 Data 1004 32BB Data 1006 901A SR 0000 SR

After
Instruction

W15

Data 1004

Data 1006

0000

1008

0304

0000

RCALL

Computed Relative Call

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} RCALL Wn

Operands: Wn ∈ [W0 ... W15]

Operation: (PC) + 2 → PC

(PC<15:1>) → TOS<15:1>, SFA Status bit → TOS<0>

(W15) + 2 → W15

(PC<22:16>)→(TOS)

(W15) + 2 → W15

0 → SFA Status bit

(PC) + (2 * (Wn)) → PC

NOP → Instruction Register

Status Affected: SFA

Encoding:

Computed, relative subroutine call specified by the Working register Wn. The range of the call is 32K program words forward or backward from the current PC. Before the call is made, the return address (PC + 2) is PUSHed onto the stack. After the return address is stacked, the sign-extended 17-bit value (2 * (Wn)) is added to the contents of the PC and the result is stored in the PC. Register Direct Addressing must be used for Wn.

The 's' bits select the source register.

Words: 1

Cycles: 4

Example 1:	00FF8C	EX1:	INC	W2, W3	; Destination of RCALL
	00FF8E		...
	.		...
	.		...
	010008
	01000A		RCALL	W6	; RCALL with W6
	01000C		MOVE	W4, [W10]

Before Instruction		After Instruction
PC	01 000A	PC	00 FF8C
W6	FFC0	W6	FFC0
W15	1004	W15	1008
Data 1004	98FF	Data 1004	000C
Data 1006	2310	Data 1006	0001
SR	0000	SR	0000

Example 2:	000302 RCALL W2	; RCALL with W2
	000304 FF1L W0, W1
	. . . .
	. . . .
	000450 EX2: CLR	W2 ; Destination of RCALL
	000452 ...

REPEAT

Repeat Next Instruction 'lit14+1' Times

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X		X

Syntax: {label:} REPEAT #lit14
Operands: lit14 ∈ [0 ... 16383]
Operation: (lit14) → RCOUNT

$$ (\mathrm{PC}) + 2 \rightarrow \mathrm{PC} $$

Enable code looping.
Status Affected: RA

Encoding:

0000

1001

00kk

kkkk

kkkk

kkkk

Description:

Repeat the instruction immediately following the REPEAT instruction (lit14 + 1) times. The repeated instruction (or target instruction) is held in the Instruction Register (IR) for all iterations and is only fetched once.

When this instruction executes, the RCOUNT register is loaded with the repeat count value specified in the instruction. RCOUNT is decremented with each execution of the target instruction. When RCOUNT equals zero, the target instruction is executed one more time and then normal instruction execution continues with the instruction following the target instruction.

The 'k' bits are an unsigned literal that specifies the loop count.

Special Features, Restrictions:

1. When the repeat literal is '0', REPEAT has the effect of a NOP and the RA bit is not set.

2. The target REPEAT instruction cannot be:

- An instruction that changes program flow

• A DO, DISI, LNK, MOV.D, PWRSAV, REPEAT or UNLK instruction

• A 2-word instruction

Unexpected results may occur if these target instructions are used.
Note: The REPEAT and target instruction are interruptible.
Words: 1
Cycles: 1

Example 1:	000452 REPEAT	#9	; Execute ADD 10 times
	000454 ADD	[W0++], W1, [W2++]	; Vector update

Before Instruction PC 00 0452 RCOUNT 0000 SR 0000

After Instruction PC 00 0454 RCOUNT 0009 SR 0010 (RA = 1)

Example 2:	00089E REPEAT	#0x3FF	; Execute CLR 1024 times
	0008A0 CLR	[W6++]	; Clear the scratch space

Before Instruction PC 00 089E RCOUNT 0000 SR 0000

After Instruction PC 00 08A0 RCOUNT 03FF SR 0010 (RA = 1)

REPEAT

Repeat Next Instruction 'lit15+1' Times

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
			X			X

Syntax: {label:} REPEAT #lit15
Operands: lit15 ∈ [0 ... 32767]
Operation: (lit15) → RCOUNT
(PC) + 2 → PC
Enable code looping.

Status Affected:

RA

Encoding:

0000

1001

0kkk

kkkk

kkkk

kkkk

Description:

Repeat the instruction immediately following the REPEAT instruction (lit15 + 1) times. The repeated instruction (or target instruction) is held in the Instruction Register for all iterations and is only fetched once.

When this instruction executes, the RCOUNT register is loaded with the repeat count value specified in the instruction. RCOUNT is decremented with each execution of the target instruction. When RCOUNT equals zero, the target instruction is executed one more time and then normal instruction execution continues with the instruction following the target instruction.

The 'k' bits are an unsigned literal that specifies the loop count.

Special Features, Restrictions:

1. When the repeat literal is '0', REPEAT has the effect of a NOP and the RA bit is not set.

2. The target REPEAT instruction cannot be:

- An instruction that changes program flow

• A DISI, LNK, MOV.D, PWRSAV, REPEAT or UNLK instruction

• A 2-word instruction

Unexpected results may occur if these target instructions are used.
Note: The REPEAT and target instruction are interruptible.
Words: 1
Cycles: 1

Example 1:	000452	REPEAT	#9	; Execute ADD 10 times
	000454	ADD	[W0++, W1, [W2++]	; Vector update

	Before Instruction		After Instruction
PC	00 0452		PC 00 0454
RCOUNT	0000		RCOUNT 0009
SR	0000		SR 0010 (RA = 1)

Example 2:	00089E	REPEAT	#0x3FF	; Execute CLR 1024 times
	0008A0	CLR	[W6++]	; Clear the scratch space

	BeforeInstruction		AfterInstruction
PC	00 089E		PC 00 08A0
RCOUNT	0000		RCOUNT 03FF
SR	0000		SR 0010 (RA = 1)

REPEAT

Repeat Next Instruction Wn+1 Times

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} REPEAT Wn
Operands: Wn ∈ [W0 ... W15]

Operation: (Wn<13:0>) → RCOUNT
(PC) + 2 → PC
Enable code looping.
Status Affected: RA

Encoding:

0000

1001

1000

0000

0000

ssss

Description:

Repeat the instruction immediately following the REPEAT instruction (Wn<13:0>) times. The instruction to be repeated (or target instruction) is held in the Instruction Register for all iterations and is only fetched once.

When this instruction executes, the RCOUNT register is loaded with the lower 14 bits of Wn. RCOUNT is decremented with each execution of the target instruction. When RCOUNT equals zero, the target instruction is executed one more time and then normal instruction execution continues with the instruction following the target instruction.

The 's' bits specify the Wn register that contains the repeat count.

Special Features, Restrictions:

1. When (Wn) = 0, REPEAT has the effect of a NOP and the RA bit is not set.

2. The target REPEAT instruction cannot be:

- An instruction that changes program flow

- A DO, DISI, LNK, MOV.D, PWRSAV, REPEAT or ULNK instruction

• A 2-word instruction

Unexpected results may occur if these target instructions are used.
Note: The REPEAT and target instruction are interruptible.
Words: 1
Cycles: 1

Example 1:	000A26 REPEAT W4 ; Execute COM (W4+1) times
	000A28 COM [W0++, [W2++] ; Vector complement

	Before Instruction
PC	00 0A26
W4	0023
RCOUNT	0000
SR	0000

	After Instruction
PC	00 0A28
W4	0023
RCOUNT	0023
SR	0010 (RA = 1)

Example 2:	00089E REPEAT W10 ; Execute TBLRD (W10+1) times
	0008A0 TBLRDL [W2++], [W3++] ; Decrement (0x840)

	Before Instruction
PC	00 089E
W10	00FF
RCOUNT	0000
SR	0000

	After Instruction
PC	00 08A0
W10	00FF
RCOUNT	00FF
SR	0010 (RA = 1)

REPEAT

Repeat Next Instruction Wn+1 Times

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} REPEAT Wn

Operands: Wn ∈ [W0 ... W15]

Operation: (Wn) → RCOUNT

(PC) + 2 → PC

Enable code looping.

Status Affected:

RA

Encoding:

0000

1001

1000

0000

0000

ssss

Description:

Repeat the instruction immediately following the REPEAT instruction (Wn) times. The instruction to be repeated (or target instruction) is held in the Instruction Register for all iterations and is only fetched once.

When this instruction executes, the RCOUNT register is loaded with Wn. RCOUNT is decremented with each execution of the target instruction. When RCOUNT equals zero, the target instruction is executed one more time and then normal instruction execution continues with the instruction following the target instruction.

The 's' bits specify the Wn register that contains the repeat count.

Special Features, Restrictions:

1. When (Wn) = 0, REPEAT has the effect of a NOP and the RA bit is not set.

2. The target REPEAT instruction cannot be:

3. An instruction that changes program flow

4. A DO, DISI, LNK, MOV.D, PWRSAV, REPEAT or ULNK instruction
   • A 2-word instruction

Unexpected results may occur if these target instructions are used.

Note: The REPEAT and target instruction are interruptible.

Words: 1

Cycles: 1

Example 1:	000A26 REPEAT W4 ; Execute COM (W4+1) times
	000A28 COM [W0++, [W2++] ; Vector complement

	Before Instruction		After Instruction
PC	00 0A26		PC 00 0A28
W4	0023		W4 0023
RCOUNT	0000		RCOUNT 0023
SR	0000		SR 0010 (RA = 1)

Example 2:	00089E	REPEAT W10	; Execute TBLRD (W10+1) times
	0008A0	TBLRDL [W2++], [W3++]	; Decrement (0x840)

	Before Instruction		After Instruction
PC	00 089E		PC 00 08A0
W10	00FF		W10 00FF
RCOUNT	0000		RCOUNT 00FF
SR	0000		SR 0010 (RA = 1)

RESET

Reset

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

X

Syntax: {label:} RESET

Operands: None

Operation: Force all registers that are affected by a MCLR

$$ 1 \rightarrow \text { SWR } (\text { RCON } < 6 >) $$

$$ 0 \rightarrow \mathrm{PC} $$

Reset to their Reset condition.

Status Affected: OA, OB, OAB, SA, SB, SAB, DA, DC, IPL<2:0>, RA, N, OV, Z, C, SFA

Encoding:

1111

1110

0000

0000

0000

0000

Description:

This instruction provides a way to execute a software Reset. All core and peripheral registers will take their power-on value. The PC will be set to '0', the location of the RESET GOTO instruction. The SWR bit (RCON<6>) will be set to '1' to indicate that the RESET instruction was executed.

Note: Refer to the specific device family reference manual for the power-on value of all registers.

Words: 1

Cycles: 1

Before Instruction		After Instruction
PC	00 202A	PC	00 0000
W0	8901	W0	0000
W1	08BB	W1	0000
W2	B87A	W2	0000
W3	872F	W3	0000
W4	C98A	W4	0000
W5	AAD4	W5	0000
W6	981E	W6	0000
W7	1809	W7	0000
W8	C341	W8	0000
W9	90F4	W9	0000
W10	F409	W10	0000
W11	1700	W11	0000
W12	1008	W12	0000
W13	6556	W13	0000
W14	231D	W14	0000
W15	1704	W15	0800
SPLIM	1800	SPLIM	0000
TBLPAG	007F	TBLPAG	0000
PSVPAG	0001	PSVPAG	0000
CORCON	00F0	CORCON	0020 (SATDW = 1)
RCON	0000	RCON	0040 (SWR = 1)
SR	0021 (IPL, C = 1)	SR	0000

Before Instruction		After Instruction
PC	00 202A	PC	00 0000
W0	8901	W0	0000
W1	08BB	W1	0000
W2	B87A	W2	0000
W3	872F	W3	0000
W4	C98A	W4	0000
W5	AAD4	W5	0000
W6	981E	W6	0000
W7	1809	W7	0000
W8	C341	W8	0000
W9	90F4	W9	0000
W10	F409	W10	0000
W11	1700	W11	0000
W12	1008	W12	0000
W13	6556	W13	0000
W14	231D	W14	0000
W15	1704	W15	0800
SPLIM	1800	SPLIM	0000
TBLPAG	007F	TBLPAG	0000
PSVPAG	0001	PSVPAG	0000
CORCON	00F0	CORCON	0020 (SATDW = 1)
RCON	0000	RCON	0040 (SWR = 1)
SR	0021 (IPL, C = 1)	SR	0000

RETFIE

Return from Interrupt

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} RETFIE

Operands: None

Operation: (W15) - 2 → W15

$$ (T O S < 1 5: 8 >) \rightarrow (S R < 7: 0 >) $$

$$ (T O S < 7 >) \rightarrow (I P L 3, C O R C O N < 3 >) $$

$$ (T O S < 6: 0 >) \rightarrow (P C < 2 2: 1 6 >) $$

$$ (W 1 5) - 2 \rightarrow W 1 5 $$

$$ (T O S < 1 5: 0 >) \rightarrow (P C < 1 5: 0 >) $$

$$ \mathrm{NOP} \rightarrow \text { Instruction Register } $$

Status Affected: IPL<3:0>, RA, N, OV, Z, C

Encoding:

0000

0110

0100

0000

0000

0000

Description:

Return from Interrupt Service Routine. The stack is POPped, which loads the low byte of the STATUS Register, IPL<3> (CORCON<3>) and the Most Significant Byte of the PC. The stack is POPped again, which loads the lower 16 bits of the PC.

Note 1: Restoring IPL<3> and the low byte of the STATUS Register restores the Interrupt Priority Level to the level before the execution was processed.

2: Before RETFIE is executed, the appropriate interrupt flag must be cleared in software to avoid recursive interrupts.

Words: 1

Cycles: 3 (2 if exception pending)

Example 1: 000A26 RETFIE ; Return from ISR

	Before Instruction
PC	00 0A26
W15	0834
Data 0830	0230
Data 0832	8101
CORCON	0001
SR	0000

After Instruction
PC	01 0230
W15	0830
Data 0830	0230
Data 0832	8101
CORCON	0001
SR	0081 (IPL = 4, C = 1)

Example 2: 008050 RETFIE ; Return from ISR

	Before Instruction
PC	00 8050
W15	0926
Data 0922	7008
Data 0924	0300
CORCON	0000
SR	0000

After Instruction
PC	00 7008
W15	0922
Data 0922	7008
Data 0924	0300
CORCON	0000
SR	0003 (Z, C = 1)

RETFIE

Return from Interrupt

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} RETFIE

Operands: None

Operation: (W15) - 2 → W15

$$ (T O S < 1 5: 8 >) \rightarrow (S R < 7: 0 >) $$

$$ (T O S < 7 >) \rightarrow (I P L 3, C O R C O N < 3 >) $$

$$ (T O S < 6: 0 >) \rightarrow (P C < 2 2: 1 6 >) $$

$$ (W 1 5) - 2 \rightarrow W 1 5 $$

$$ (T O S < 1 5: 1 >) \rightarrow (P C < 1 5: 1 >) $$

$$ \mathrm{TOS} < 0 > \rightarrow \text { SFA Status bit } $$

$$ \mathrm{NOP} \rightarrow \text { Instruction Register } $$

Status Affected: IPL<3:0>, RA, N, OV, Z, C, SFA

Encoding:

0000

0110

0100

0000

0000

0000

Description:

Return from Interrupt Service Routine. The stack is POPped, which loads the low byte of the STATUS Register, IPL<3> (CORCON<3>) and the Most Significant Byte of the PC. The stack is POPped again, which loads the lower 16 bits of the PC.

Note 1: Restoring IPL<3> and the low byte of the STATUS Register restores the Interrupt Priority Level to the level before the execution was processed.

2: Before RETFIE is executed, the appropriate interrupt flag must be cleared in software to avoid recursive interrupts.

Words: 1

Cycles: 6 (5 if exception pending)

Example 1: 000A26 RETFIE ; Return from ISR

Before Instruction		After Instruction
PC	00 0A26	PC	01 0230
W15	0834	W15	0830
Data 0830	0230	Data 0830	0230
Data 0832	8101	Data 0832	8101
CORCON	0001	CORCON	0001
SR	0000	SR	0081 (IPL = 4, C = 1)

Example 2: 008050 RETFIE ; Return from ISR

Before Instruction		After Instruction
PC	00 8050	PC	00 7008
W15	0926	W15	0922
Data 0922	7008	Data 0922	7008
Data 0924	0300	Data 0924	0300
CORCON	0000	CORCON	0000
SR	0000	SR	0003 (Z, C = 1)

RETLW

Return with Literal in Wn

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F dsPIC	33E dsPIC3	3C
	X	X		X		X

Syntax: {label:} RETLW{.B} #lit10, Wn
Operands: lit10 ∈ [0 ... 255] for byte operation lit10 ∈ [0 ... 1023] for word operation Wn ∈ [W0 ... W15]
Operation: (W15) - 2 → W15 TOS<15:8> → SR<7:0> TOS<7:0> → IPL<3>: PC<22:16> (W15) - 2 → W15 (TOS) → (PC<15:0>) lit10 → Wn NOP → Instruction Register
Status Affected: None
Encoding: 0000 0101 0Bkk kkkk kkkk dddd

Description: Return from subroutine with the specified, unsigned 10-bit literal stored in Wn. The software stack is POPped twice to restore the PC and the signed literal is stored in Wn. Since two POPs are made, the Stack Pointer (W15) is decremented by 4.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'k' bits specify the value of the literal.
The 'd' bits select the destination register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.
Words: 1
Cycles: 3 (2 if exception pending)
Example 1: 000440 RETLW.B #0xA, WO ; Return with 0xA in WO

	Before Instruction		After Instruction
PC	00 0440	PC	00 7006
W0	9846	W0	980A
W15	1988	W15	1984
Data 1984	7006	Data 1984	7006
Data 1986	0000	Data 1986	0000
SR	0000	SR	0000

Example 2: 00050A RETLW #0x230, W2 ; Return with 0x230 in W2

	Before Instruction					After Instruction
	PC	00	050A	PC	01	7008
	W2	0993	W2	02	30
	W15	1200	W15	11	FC
	Data 11FC	7008	Data 11FC	7008
	Data 11FE	0001			Data 11FE	0001
	SR	0000			SR	0000

RETLW

Return with Literal in Wn

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} RETLW{.B} #lit10, Wn
Operands: lit10 ∈ [0 ... 255] for byte operation lit10 ∈ [0 ... 1023] for word operation Wn ∈ [W0 ... W15]
Operation: (W15) - 2 → W15
TOS<15:8> → SR<7:0> TOS<7:0> → IPL<3>: PC<22:16> (W15) - 2 → W15 (TOS<15:1>) → (PC<15:1>) TOS<0> → SFA Status bit lit10 → Wn NOP → Instruction Register

Status Affected:

SFA

Encoding:

0000

0101

0Bkk

kkkk

kkkk

dddd

Description:

Return from subroutine with the specified, unsigned 10-bit literal stored in Wn. The software stack is POPped twice to restore the PC and the signed literal is stored in Wn. Since two POPs are made, the Stack Pointer (W15) is decremented by 4.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'k' bits specify the value of the literal.
The 'd' bits select the destination register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.
Words: 1
Cycles: 6 (5 if exception pending)
Example 1: 000440 RETLW.B #0xA, WO ; Return with 0xA in WO

	Before Instruction		After Instruction
PC	00 0440	PC	00 7006
W0	9846	W0	980A
W15	1988	W15	1984
Data 1984	7006	Data 1984	7006
Data 1986	0000	Data 1986	0000
SR	0000	SR	0000

Example 2: 00050A RETLW #0x230, W2 ; Return with 0x230 in W2

	Before Instruction					After Instruction
	PC	00	050A	PC	01	7008
	W2	0993	W2	02	30
	W15	1200	W15	11	FC
	Data 11FC	7008	Data 11	FC	7008
	Data 11FE	0001			Data 11FE	0001
	SR	0000			SR	0000

RETURN

Return

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X		X

Syntax: {label:} RETURN

Operands: None

Operation: (W15) - 2 → W15

(TOS) → (PC<22:16>)

(W15) - 2 → W15

(TOS) → (PC<15:0>)

NOP → Instruction Register

Status Affected: None

Encoding:

0000

0110

0000

0000

0000

0000

Description:

Return from subroutine. The software stack is POPped twice to restore the PC. Since two POPs are made, the Stack Pointer (W15) is decremented by 4.

Words:

1

Cycles:

3 (2 if exception pending)

Example 1: 001A06 RETURN ; Return from subroutine

Before Instruction PC 00 1A06 W15 1248 Data 1244 0004 Data 1246 0001 SR 0000

After Instruction PC 01 0004 W15 1244 Data 1244 0004 Data 1246 0001 SR 0000

Example 2: 005404 RETURN ; Return from subroutine

Before Instruction PC 00 5404 W15 090A Data 0906 0966 Data 0908 0000 SR 0000

RETURN

Return

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
			X			X	X

Syntax: {label:} RETURN

Operands: None

Operation: (W15) - 2 → W15

$$ \begin{array}{l} (T O S) \rightarrow (P C < 2 2: 1 6 >) \ (W 1 5) - 2 \rightarrow W 1 5 \ (T O S < 1 5: 1) \rightarrow (P C < 1 5: 1 >) \ \mathrm{TOS} < 0 > \rightarrow \text { SFA Status bit } \ \mathrm{NOP} \rightarrow \text { Instruction Register } \ \end{array} $$

Status Affected: SFA

Encoding:

0000

0110

0000

0000

0000

0000

Description:

Return from subroutine. The software stack is POPped twice to restore the PC. Since two POPs are made, the Stack Pointer (W15) is decremented by 4.

Words: 1

Cycles: 6 (5 if exception pending)

Example 1: 001A06 RETURN ; Return from subroutine

Example 2: 005404 RETURN ; Return from subroutine

RLC

Rotate Left f through Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} RLC{.B} f {,WREG}
Operands: f [0 8191]
Operation: For Byte Operation:

$$ (C) \rightarrow \text { Dest } < 0 > $$

$$ (f < 6: 0 >) \rightarrow \text {Dest} < 7: 1 > $$

$$ (f < 7 >) \rightarrow C $$

For Word Operation:

$$ (C) \rightarrow D e s t < 0 > $$

$$ (f < 1 4: 0 >) \rightarrow \text { Dest } < 1 5: 1 > $$

$$ (f < 1 5 >) \rightarrow C $$

Status Affected:

Encoding:

Description:

N, Z, C

1101

0110

1BDF

ffff

ffff

ffff

Rotate the contents of the file register f, one bit to the left through the Carry flag, and place the result in the destination register. The Carry flag of the STATUS Register is shifted into the Least Significant bit of the destination and it is then overwritten with the Most Significant bit of Ws.
The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for f, '1' for WREG).
The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1
Example 1: RLC.B 0x1233 ; Rotate Left w/ C (0x1233) (Byte mode)

Before Instruction		After Instruction
Data 1232	E807	Data 1232	D007
SR	0000	SR	0009

Example 2: RLC 0x820, WREG ; Rotate Left w/ C (0x820) (Word mode) ; Store result in WREG

	Before Instruction		After Instruction
WREG (W0)	5601	WREG (W0)	42DD
Data 0820	216E	Data 0820	216E
SR	0001	(C = 1)	SR

RLC

Rotate Left Ws through Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X	X		X

X

Syntax: {label:} RLC{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

(C) → Wd<0>

(Ws<6:0>) Wd<7:1>

(Ws<7>) C

For Word Operation:

(C) → Wd<0>

(Ws<14:0>) Wd<15:1>

(Ws<15>) C

Status Affected: N, Z, C

Encoding:

1101

0010

1Bqq

qddd

dppp

ssss

Description:

Rotate the contents of the source register Ws, one bit to the left through the Carry flag, and place the result in the destination register Wd. The Carry flag of the STATUS Register is shifted into the Least Significant bit of Wd and it is then overwritten with the Most Significant bit of Ws. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: RLC.B W0, W3 ; Rotate Left w/C (W0) (Byte mode) ; Store the result in W3

Example 2: RLC [W2++, [W8] ; Rotate Left w/C [W2] (Word mode) ; Post-increment W2 ; Store result in [W8]

Before Instruction W2 2008 W2 200A W8 094E W8 094E Data 094E 3689 Data 094E 8082 Data 2008 C041 Data 2008 C041 SR 0001 (C = 1) SR 0009 (N, C = 1)

RLNC

Rotate Left f without Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X	X		X

Syntax: {label:} RLNC{.B} f {,WREG}

Operands: f [0 8191]

Operation: For Byte Operation: (f<6:0>) → Dest<7:1> (f<7>) → Dest<0>

For Word Operation: (f<14:0>) → Dest<15:1> (f<15>) → Dest<0>

Status Affected: N, Z

Encoding: 1101 0110 0BDF ffff ffff ffff

Description: Rotate the contents of the file register f, one bit to the left, and place the result in the destination register. The Most Significant bit of f is stored in the Least Significant bit of the destination and the Carry flag is not affected.

The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'D' bit selects the destination register ('0' for WREG, '1' for file register). The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: RLNC.B 0x1233 ; Rotate Left (0x1233) (Byte mode)

Before Instruction Data 1232 E807 SR 0000 After Instruction Data 1233 D107 SR 0008 (N = 1)

Example 2: RLNC 0x820, WREG ; Rotate Left (0x820) (Word mode) ; Store result in WREG

RLNC

Rotate Left Ws without Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	C33F ds	PIC33E ds	PIC33C
	X	X		X	X		X

Syntax: {label:} RLNC{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

(Ws<6:0>) Wd<7:1>

(Ws<7>) Wd<0>

For Word Operation:

(Ws<14:0>) Wd<15:1>

(Ws<15>) Wd<0>

Status Affected: N, Z

Encoding:

1101

0010

0Bqq

qddd

dppp

ssss

Description:

Rotate the contents of the source register Ws, one bit to the left, and place the result in the destination register Wd. The Most Significant bit of Ws is stored in the Least Significant bit of Wd and the Carry flag is not affected. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for byte, '1' for word).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: RLNC.B W0, W3 ; Rotate Left (W0) (Byte mode) ; Store the result in W3

Before Instruction
W0	9976 W0	9976
W3	5879 W3	58EC
SR	0001	(C = 1)

After Instruction
SR	0009 (N, C = 1)

Example 2: RLNC [W2++, [W8] ; Rotate Left [W2] (Word mode) ; Post-increment W2 ; Store result in [W8]

	Before Instruction					After Instruction
	W2	2008	W2	200A
	W8	094E	W8	094E
	Data 094E	3689	Data 094E	8083
	Data 2008	C041			Data 2008	C041
	SR	0001			(C = 1)	SR	0009			(N, C = 1)

RRC

Rotate Right f through Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X

Syntax: {label:} RRC{.B} f {,WREG}

Operands: f [0 8191]

Operation: For Byte Operation:

$$ (C) \rightarrow D e s t < 7 > $$

$$ (f < 7: 1 >) \rightarrow \text { Dest } < 6: 0 > $$

$$ (f < 0 >) \rightarrow C $$

For Word Operation:

$$ (C) \rightarrow \text { Dest } < 1 5 > $$

$$ (f < 1 5: 1 >) \rightarrow \text { Dest } < 1 4: 0 > $$

$$ (f < 0 >) \rightarrow C $$

Status Affected:

N, Z, C

Encoding:

1101

0111

1BDf

ffff

ffff

ffff

Description:

Rotate the contents of the file register f, one bit to the right through the Carry flag, and place the result in the destination register. The Carry flag of the STATUS Register is shifted into the Most Significant bit of the destination and it is then overwritten with the Least Significant bit of Ws.

The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for byte, '1' for word).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: RRC.B 0x1233 ; Rotate Right w/C (0x1233) (Byte mode)

Before

Instruction

Data 1232 E807 Data 1232 7407

SR 0000 SR 0000

After

Instruction

Example 2: RRC 0x820, WREG ; Rotate Right w/C (0x820) (Word mode)

; Store result in WREG

Before

Instruction

WREG (W0) 5601 WREG (W0) 90B7

Data 0820 216E Data 0820 216E

SR 0001 (C = 1) SR 0008 (N = 1)

After

Instruction

RRC

Rotate Right Ws through Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	IC33E dsPIC	33C
	X	X		X	X		X

Syntax: {label:} RRC{.B} Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

(C) → Wd<7>

(Ws<7:1>) Wd<6:0>

(Ws<0>) C

For Word Operation:

(C) → Wd<15>

(Ws<15:1>) Wd<14:0>

(Ws<0>) C

Status Affected: N, Z, C

Encoding:

1101

0011

1Bqq

qddd

dppp

ssss

Description:

Rotate the contents of the source register Ws, one bit to the right through the Carry flag, and place the result in the destination register Wd. The Carry flag of the STATUS Register is shifted into the Most Significant bit of Wd and it is then overwritten with the Least Significant bit of Ws. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: RRC.B W0, W3 ; Rotate Right w/ C (W0) (Byte mode) ; Store the result in W3

Before Instruction
W0	9976 W0	9976
W3	5879 W3	58BB
SR	0001	(C = 1)

After Instruction
SR	0008 (N = 1)

Example 2: RRC [W2++, [W8] ; Rotate Right w/ C [W2] (Word mode) ; Post-increment W2 ; Store result in [W8]

	Before Instruction					After Instruction
	W2	2008	W2	200A
	W8	094E	W8	094E
	Data 094E	3689	Data 094E	E020
	Data 2008	C041			Data 2008	C041
	SR	0001			(C = 1)	SR	0009			(N, C = 1)

RRNC

Rotate Right f without Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} RRNC{.B} f {,WREG}

Operands: f [0 8191]

Operation:	For Byte Operation:(f<7:1>) → Dest<6:0>(f<0>) → Dest<7>
	For Word Operation:(f<15:1>) → Dest<14:0>(f<0>) → Dest<15>

Status Affected: N, Z

Encoding:

1101

0111

0BDf

ffff

ffff

ffff

Description: Rotate the contents of the file register f, one bit to the right, and place the result in the destination register. The Least Significant bit of f is stored in the Most Significant bit of the destination and the Carry flag is not affected.

The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'D' bit selects the destination register ('0' for WREG, '1' for file register). The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required. 2: The WREG is set to Working register W0.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: RRNC.B 0x1233 ; Rotate Right (0x1233) (Byte mode)

	Before Instruction	After Instruction
Data 1232	E807	7407
SR	0000	0000

Example 2: RRNC 0x820, WREG ; Rotate Right (0x820) (Word mode) ; Store result in WREG

	Before Instruction		After Instruction
WREG (W0)	5601	WREG (W0)	10B7
Data 0820	216E	Data 0820	216E
SR	0001	(C = 1)	SR	0001

RRNC
Rotate Right Ws without Carry

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} RRNC{.B} Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

(Ws<7:1>) Wd<6:0>

(Ws<0>) Wd<7>

For Word Operation:

(Ws<15:1>) Wd<14:0>

(Ws<0>) Wd<15>

Status Affected: N, Z

Encoding:

1101

0011

0Bqq

qddd

dppp

ssss

Description: Rotate the contents of the source register Ws, one bit to the right, and place the result in the destination register Wd. The Least Significant bit of Ws is stored in the Most Significant bit of Wd and the Carry flag is not affected. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: RRNC.B W0, W3 ; Rotate Right (W0) (Byte mode) ; Store the result in W3

Before
Instruction

After
Instruction

Example 2: RRNC [W2++], [W8]; Rotate Right [W2] (Word mode)

; Post-increment W2 ; Store result in [W8]

Before

Instruction

W2 2008 W2 200A

W8 094E W8 094E

Data 094E 3689 Data 094E E020

Data 2008 C041 Data 2008 C041

SR 0000 SR 0008 (N = 1)

After
Instruction

SAC
Store Accumulator

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} SAC Acc, {#Slit4,} Wd

[Wd]

[Wd++]

[Wd--]

[--Wd]

[++Wd]

[Wd + Wb]

Operands: Acc ∈ [A,B]

Slit4 ∈ [-8 ... +7]

Wb, Wd ∈ [W0 ... W15]

Operation: Shift _Slit4 (Acc) (optional)

(Acc[31:16]) → Wd

Status Affected: None

Encoding:

1100

1100

Awww

wrrr

rhhh

dddd

Description:

Perform an optional, signed 4-bit shift of the specified accumulator, then store the shifted contents of ACCxH (Acc[31:16]) to Wd. The shift range is -8:7, where a negative operand indicates an arithmetic left shift and a positive operand indicates an arithmetic right shift. Either Register Direct or Indirect Addressing may be used for Wd.

The 'A' bit specifies the source accumulator.

The 'w' bits specify the offset register Wb.

The 'r' bits encode the optional accumulator preshift.

The 'h' bits select the destination addressing mode.

The 'd' bits specify the destination register Wd.

Note 1: This instruction does not modify the contents of Acc.

2: This instruction stores the truncated contents of Acc. The instruction, SAC.R, may be used to store the rounded accumulator contents.

3: If data write saturation is enabled (SATDW (CORCON<5>) = 1), the value stored to Wd is subject to saturation after the optional shift is performed.

Words: 1

Cycles: 1

Example 1: SAC A, #4, W5 ; Right shift ACCA by 4 ; Store result to W5 ; CORCON = 0x0010 (SATDW = 1)

Before Instruction W5 B900 W5 0120 ACCA 00 120F FF00 ACCA 00 120F FF00 CORCON 0010 CORCON 0010 SR 0000 SR 0000 After Instruction

Example 2: SAC B, #-4, [W5++] ; Left shift ACCB by 4 ; Store result to [W5], Post-increment W5 ; CORCON = 0x0010 (SATDW = 1)

Before Instruction W5 2000 W5 2002 ACCB FF C891 8F4C AC CB FF C891 1F4C Data 2000 5BBE Data 2000 8000 CORCON 0010 CORCON 0010 SR 0000 SR 0000

SAC.D
Store Accumulator Double

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsPIC	C33E dsPIC3	3C
							X

Syntax: {label:} SAC.D Acc, [, #Slit4], Wnd

[Wnd]

[Wnd++]

[Wnd--]

[--Wnd]

[++Wnd]

Operands: Register Direct: Wnd ∈ [W0, W2, W4, W6, W8, W10, W12, W14];

Register Indirect: Wnd ∈ [W0 ... W15];

Operation: ShiftSlit4(Acc) (optional); (Acc[31:0]) → Wnd

Status Affected: None

Encoding:

1101

1100

A0qq

qrrr

r000

dddd

Description: Optionally shift accumulator, then store accumulator, Acc<31:0>, to the destination Effective Address.

The 'A' bit specifies the source accumulator.

The 'd' bits specify the destination register Wnd.

The 'q' bits select the destination addressing mode.

The 'r' bits encode the optional operand Slit4, which determines the amount of the accumulator preshift; if the operand Slit4 is absent, a '0' is encoded.

Note 1: Unlike SAC and SAC.R instructions, the SAC.D instruction does not support Indirect with Register Offset Addressing mode.

2: Positive values of operand Slit4 represent arithmetic shift right. Negative values of operand Slit4 represent shift left.

3: The SAC.D instruction cannot be executed within a REPEAT loop.

Words: 1

Cycles: 2

SAC.R
Store Rounded Accumulator

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} SAC.R Acc, {#Slit4,} Wd

[Wd]

[Wd++]

[Wd--]

[--Wd]

[++Wd]

[Wd + Wb]

Operands: Acc ∈ [A,B]

Slit4 ∈ [-8 ... +7]

Wb ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: Shift _Slit4 (Acc) (optional)

Round(Acc)

(Acc[31:16]) → Wd

Status Affected: None

Encoding:

1100

1101

Awww

wrrr

rhhh

dddd

Description: Perform an optional, signed 4-bit shift of the specified accumulator, then store the rounded contents of ACCxH (Acc[31:16]) to Wd. The shift range is -8:7, where a negative operand indicates an arithmetic left shift and a positive operand indicates an arithmetic right shift. The Rounding mode (Conventional or Convergent) is set by the RND bit (CORCON<1>). Either Register Direct or Indirect Addressing may be used for Wd.

The 'A' bit specifies the source accumulator.

The 'w' bits specify the offset register Wb.

The 'r' bits encode the optional accumulator preshift.

The 'h' bits select the destination addressing mode.

The 'd' bits specify the destination register Wd.

Note 1: This instruction does not modify the contents of the Acc.

2: This instruction stores the rounded contents of Acc. The instruction, SAC, may be used to store the truncated accumulator contents.

3: If data write saturation is enabled (SATDW (CORCON<5>) = 1), the value stored to Wd is subject to saturation after the optional shift is performed.

Words: 1

Cycles: 1

Example 1: SAC.R A, #4, W5 ; Right shift ACCA by 4 ; Store rounded result to W5 ; CORCON = 0x0010 (SATDW = 1)

Before Instruction W5 B900 W5 0121 ACCA 00 120F FF00 ACCA 00 120F FF00 CORCON 0010 CORCON 0010 SR 0000 SR 0000 After Instruction

Example 2: SAC.R B, #-4, [W5++] ; Left shift ACCB by 4 ; Store rounded result to [W5], Post-increment W5 ; CORCON = 0x0010 (SATDW = 1)

Before Instruction W5 2000 W5 2002 ACCB FF F891 8F4C AC CB FF F891 8F4C Data 2000 5BBE Data 2000 8919 CORCON 0010 COR CON 0010 SR 0000 SR 0000

SE
Sign-Extend Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SE Ws, Wnd

[Ws],

[Ws++],

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: Ws<7:0> → Wnd<7:0>

If (Ws<7>=1):

0xFF → Wnd<15:8>

Else:

0 → Wnd<15:8>

Status Affected: N, Z, C

Encoding:

1111

1011

0000

0ddd

dppp

ssss

Description: Sign-extend the byte in Ws and store the 16-bit result in Wnd. Either Register Direct or Indirect Addressing may be used for Ws and Register Direct Addressing must be used for Wnd. The C flag is set to the complement of the N flag.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: This operation converts a byte to a word and it uses no .B or .W extension.

2: The source Ws is addressed as a byte operand, so any address modification is by '1'.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: SE W3, W4 ; Sign-extend W3 and store to W4

Before Instruction W3 7839 W3 7839 W4 1005 W4 0039 SR 0000 SR 0001 (C = 1) After Instruction

Example 2: SE [W2++], W12 ; Sign-extend [W2] and store to W12 ; Post-increment W2

Before Instruction W2 0900 W2 0901 W12 1002 W12 FF8F Data 0900 008F Data 0900 008F SR 0000 SR 0008 (N = 1) After Instruction

SETM

Set f or WREG

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SETM{.B} f

WREG

Operands: f [0 8191]

Operation: For Byte Operation ____ :

0xFF → destination designated by D

For Word Operation:

0xFFFF → destination designated by D

Status Affected: None

Encoding:

1110

1111

1BDf

ffff

ffff

ffff

Description:

All the bits of the specified register are set to '1'. If WREG is specified, the bits of WREG are set. Otherwise, the bits of the specified file register are set.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1

Example 1: SETM.B 0x891 ; Set 0x891 (Byte mode)

	Before Instruction		After Instruction
Data 0890	2739	Data 0890	FF39
SR	0000	SR	0000

Example 2: SETM WREG ; Set WREG (Word mode)

	Before Instruction		After Instruction
WREG (W0)	0900	WREG (W0)	FFFF
SR	0000	SR	0000

SETM
Set Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	38C
	X	X		X	X		X

X

Syntax: {label:} SETM{.B} Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

0xFF → Wd for byte operation

For Word Operation:

0xFFFF → Wd for word operation

Status Affected: None

Encoding:

1110

1011

1Bqq

qddd

d000

0000

Description: All the bits of the specified register are set to '1'. Either Register Direct or Indirect Addressing may be used for Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: SETM.B W13 ; Set W13 (Byte mode)

Before

Instruction

W13	2739
SR	0000

After

Instruction

W13	27FF
SR	0000

Example 2: SETM [--W6] ; Pre-decrement W6 (Word mode)

; Set [W6]

Before

Instruction

W6	1250
Data 124E	3CD9
SR	0000

After

Instruction

W6	124E
Data 124E	FFFF
SR	0000

SFTAC

Arithmetic Shift Accumulator by Slit6

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} SFTAC Acc, #Slit6
Operands: Acc ∈ [A,B]

$$ \text { S l i t } 6 \in [ - 1 6 \dots 1 6 ] $$

Operation: Shift _k (Acc) → Acc
Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1000

A000

0000

01kk

kkkk

Description: Arithmetic shift the 40-bit contents of the specified accumulation, literal and store the result back into the accumulator. The negative operand indicates a left shift and a positive output. Any bits which are shifted out of the accumulator are left-shifted.
The 'A' bit selects the accumulator for the result.
The 'k' bits determine the number of bits to be shifted.

Note 1: If saturation is enabled for the target accumulator (SATA, CORCON<7> or SATB, CORCON<6>), the value stored to the accumulator is subject to saturation.
2: If the shift amount is greater than 16 or less than -16, no modification will be made to the accumulator and an arithmetic trap will occur.
Words: 1
Cycles: 1

Example 1:	SFTAC A, #12
	; Arithmetic right shift ACCA by 12
	; Store result to ACCA
	; CORCON = 0x0080 (SATA = 1)

	Before Instruction
ACCA	00 120F FF00
CORCON	0080
SR	0000

	After Instruction
ACCA	00 0001 20FF
CORCON	0080
SR	0000

Example 2:	SFTAC B, #-10
	; Arithmetic left shift ACCB by 10
	; Store result to ACCB
	; CORCON = 0x0040 (SATB = 1)

	Before Instruction
ACCB	FF FFF1 8F4C
CORCON	0040
SR	0000

	After Instruction
ACCB	FF C63D 3000
CORCON	0040
SR	0000

SFTAC

Arithmetic Shift Accumulator by Wb

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
				X	X	X	X

Syntax: {label:} SFTAC Acc, Wb
Operands: Acc ∈ [A,B]

$$ \mathrm{Wb} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: Shift _(Wb) (Acc) → Acc
Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1000

A000

0000

0000

ssss

Description:

Arithmetic shift the 40-bit contents of the specified accumulator and store the result back into the accumulator. The Least Significant 6 bits of Wb are used to specify the shift amount. The shift range is -16:16, where a negative value indicates a left shift and a positive value indicates a right shift. Any bits which are shifted out of the accumulator are lost.
The 'A' bit selects the accumulator for the source/destination.
The 's' bits select the address of the Shift Count register.
Note 1: If saturation is enabled for the target accumulator (SATA, CORCON<7> or SATB, CORCON<6>), the value stored to the accumulator is subject to saturation.
2: If the shift amount is greater than 16 or less than -16, no modification will be made to the accumulator and an arithmetic trap will occur.
Words: 1
Cycles: 1

Example 1:

SFTAC A, W0; Arithmetic shift ACCA by (W0); Store result to ACCA; CORCON = 0x0000 (saturation disabled)

	Before Instruction
W0	FFFC
ACCA	00 320F AB09
CORCON	0000
SR	0000

After Instruction
W0	FFFC
ACCA	03 20FA B090
CORCON	0000
SR	8800 (OA, OAB = 1)

Example 2:	SFTAC B, W12
	; Arithmetic shift ACCB by (W12)
	; Store result to ACCB
	; CORCON = 0x0040 (SATB = 1)

	Before Instruction
W12	000F
ACCB	FF FFF1 8F4C
CORCON	0040
SR	0000

	After Instruction
W12	000F
ACCB	FF FFFF FFE3
CORCON	0040
SR	0000

SL

Shift Left f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} SL{.B} f {,WREG}

Operands: f [0 8191]

Operation: For Byte Operation:

$$ \begin{array}{l} (f < 7 >) \rightarrow (C) \ (f < 6: 0 >) \rightarrow \text { Dest } < 7: 1 > \ 0 \rightarrow \text { Dest } < 0 > \ \end{array} $$

For Word Operation:

$$ (f < 1 5 >) \rightarrow (C) $$

$$ (f < 1 4: 0 >) \rightarrow \text { Dest } < 1 5: 1 > $$

$$ 0 \rightarrow \text { Dest } < 0 > $$

Status Affected:

N, Z, C

Encoding:

1101

0100

0BDf

ffff

ffff

ffff

Description:

Shift the contents of the file register, one bit to the left, and place the result in the destination register. The Most Significant bit of the file register is shifted into the Carry bit of the STATUS Register and '0' is shifted into the Least Significant bit of the destination register.

The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'D' bit selects the destination register ('0' for WREG, '1' for file register).

The 'f' bits select the address of the file register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The WREG is set to Working register W0.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: SL.B 0x909 ; Shift left (0x909) (Byte mode)

Before

Instruction

Data 0908	9439 Data 0908 0839
SR	0000 SR 0001 (C = 1)

After

Instruction

Example 2: SL 0x1650, WREG ; Shift left (0x1650) (Word mode)

; Store result in WREG

Before

Instruction

WREG (W0)	0900	WREG (W0) 80CA
Data 1650	4065	Data 1650 4065
SR	0000	SR 0008 (N = 1)

After

Instruction

SL
Shift Left Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} SL{.B} Ws, Wd

[Ws], [Wd]

[Ws++, [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

(Ws<7>) C

(Ws<6:0>) Wd<7:1>

0 → Wd<0>

For Word Operation:

(Ws<15>) C

(Ws<14:0>) Wd<15:1>

0 → Wd<0>

Status Affected: N, Z, C

Encoding:

1101

0000

0Bqq

qddd

dppp

ssss

Description:

Shift the contents of the source register Ws, one bit to the left, and place the result in the destination register Wd. The Most Significant bit of Ws is shifted into the Carry bit of the STATUS Register and '0' is shifted into the Least Significant bit of Wd. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1 ^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: SL.B W3, W4 ; Shift left W3 (Byte mode) ; Store result to W4

Before Instruction
W3	78A9 W3	78A9
W4	1005 W4	1052
SR	0000 SR	0001 (C = 1)

After Instruction

Example 2: SL [W2++, [W12] ; Shift left [W2] (Word mode) ; Store result to [W12] ; Post-increment W2

	Before Instruction					After Instruction
	W2	0900	W2	0902
	W12	1002	W12	1002
	Data 0900	800F	Data 0900	800F
	Data 1002	6722	Data 1002	001E
	SR	0000	SR	0001 (C = 1)

SL

Shift Left by Short Literal

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SL Wb, #lit4, Wnd

Operands: Wb ∈ [W0 ... W15]

$$ \operatorname{lit} 4 \in [ 0 \dots 1 5 ] $$

$$ \text { Wnd } \in [ \text { W0 } \dots \text { W15 } ] $$

Operation: lit4<3:0> → Shift_Val

$$ \text { Wnd } < 1 5: \text { Shift_Val } > = \text { Wb } < 1 5 - \text { Shift_Val }: 0 > $$

$$ \text { Wd } < \text { Shift_Val } - 1: 0 > = 0 $$

Status Affected: N, Z

Encoding: 1101 1101 0www wddd d100 kkkk

Description: Shift left the contents of the source register Wb by the 4-bit unsigned literal and store the result in the destination register Wnd. Any bits shifted out of the source register are lost. Direct Addressing must be used for Wb and Wnd.

The 'w' bits select the address of the base register.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note: This instruction operates in Word mode only.

Words: 1

Cycles: 1

Example 1: SL W2, #4, W2 ; Shift left W2 by 4 ; Store result to W2

Before
Instruction

W2	78A9
SR	0000

After
Instruction

W2	8A90
SR	0008(N=1)

Example 2: SL W3, #12, W8 ; Shift left W3 by 12

; Store result to W8

Before
Instruction

W3	0912
W8	1002
SR	0000

After
Instruction

W3	0912
W8	2000
SR	0000

SL

Shift Left by Wns

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	C33F ds	PIC33E ds	PIC33C
	X	X		X	X		X

X

Syntax: {label:} SL Wb, Wns, Wnd

Operands: Wb ∈ [W0 ... W15]

$$ \mathrm{Wns} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

$$ \text { Wnd } \in [ \text { W0 } \dots \text { W15 } ] $$

Operation: Wns<4:0> → Shift_Val

$$ \text { Wnd } < 1 5: \text { Shift_Val } > = \text { Wb } < 1 5 - \text { Shift_Val }: 0 > $$

$$ W d < S h i f t _ V a l - 1: 0 > = 0 $$

Status Affected: N, Z

Encoding:

1101

1101

0www

wddd

d000

ssss

Description: Shift left the contents of the source register Wb by the 5 Least Significant bits of Wns (only up to 15 positions) and store the result in the destination register Wnd. Any bits shifted out of the source register are lost. Register Direct Addressing must be used for Wb, Wns and Wnd.
The 'w' bits select the address of the base register.
The 'd' bits select the destination register.
The 's' bits select the source register.

Note 1: This instruction operates in Word mode only.
2: If Wns is greater than 15, Wnd will be loaded with 0x0.
Words: 1
Cycles: 1

Example 1:	SL	W0, W1, W2	; Shift left W0 by W1<0:4>
			; Store result to W2

Before
Instruction

W0	09A4
W1	8903
W2	78A9
SR	0000

After
Instruction

W0	09A4
W1	8903
W2	4D20
SR	0000

Example 2:	SL	W4, W5, W6	; Shift left W4 by W5<0:4>
			; Store result to W6

Before
Instruction

W4	A409
W5	FF01
W6	0883
SR	0000

After
Instruction

W4	A409
W5	FF01
W6	4812
SR	0000

SUB

Subtract WREG from f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	C33F ds	PIC33E ds	PIC33C
	X	X	X	X	X	X	X

Syntax: {label:} SUB{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) - (WREG) → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding: 1011 0101 0BDf ffff ffff ffff

Description:

Subtract the contents of the default Working register WREG from the contents of the specified file register and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.
Example 1: SUB.B 0x1FFF ; Sub. WREG from (0x1FFF) (Byte mode) ; Store result to 0x1FFF

	Before Instruction		After Instruction
WREG (W0)	7804		7804
Data 1FFE	9439		9039
SR	0000		0001 (C = 1)

Example 2: SUB 0xA04, WREG ; Sub. WREG from (0xA04) (Word mode); Store result to WREG

	Before Instruction	After Instruction
WREG (W0)	6234	E2EF
Data 0A04	4523	4523
SR	0000	0008 (N = 1)

SUB

Subtract Literal from Wn

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X	X		X

X

Syntax: {label:} SUB{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation

lit10 ∈ [0 ... 1023] for word operation

Wn ∈ [W0 ... W15]

Operation: (Wn) - lit10 → Wn

Status Affected: DC, N, OV, Z, C

Encoding: 1011 0001 0Bkk kkkk kkkk dddd

Description: Subtract the 10-bit unsigned literal operand from the contents of the Working register Wn and store the result back in the Working register Wn. Register Direct Addressing must be used for Wn.

The 'B' bit selects byte or word operation.

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

Words: 1

Cycles: 1

Example 1: SUB.B #0x23, W0 ; Sub. 0x23 from W0 (Byte mode)

; Store result to W0

Before
Instruction

W0	7804
SR	0000

After
Instruction

W0	78E1
SR	0008(N=1)

Example 2: SUB #0x108, W4 ; Sub. 0x108 from W4 (Word mode)

; Store result to W4

Before
Instruction

W4	6234
SR	0000

After
Instruction

W4	612C
SR	0001(C=1)

SUB

Subtract Short Literal from Wb

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SUB{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[-Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb) - lit5 → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0101

0www

wBqq

qddd

d11k

kkkk

Description: Subtract the 5-bit unsigned literal operand from the contents of the base register Wb and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing must be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: SUB.B W4, #0x10, W5 ; Sub. 0x10 from W4 (Byte mode)

; Store result to W5

Before
Instruction

W4	1782
W5	7804
SR	0000

After
Instruction

W4	1782
W5	7872
SR	0005 (OV, C = 1)

Example 2: SUB W0, #0x8, [W2++] ; Sub. 0x8 from W0 (Word mode)

; Store result to [W2]

; Post-increment W2

Before
Instruction

W0	F230
W2	2004
Data 2004	A557
SR	0000

After
Instruction

W0	F230
W2	2006
Data 2004	F228
SR	0009(N, C = 1)

SUB
Subtract Ws from Wb

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SUB{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb) - (Ws) → Wd

Status Affected: DC, N, OV, Z, C

Encoding: 0101 0www wBqq qddd dppp ssss

Description:

Subtract the contents of the source register Ws from the contents of the base register Wb and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: SUB.B W0, W1, W0 ; Sub. W1 from W0 (Byte mode)

; Store result to W0

Before

Instruction W0 1732 W0 17EE W1 7844 W1 7844 SR 0000 SR 0108 (DC, N = 1)

Example 2: SUB W7, [W8++, [W9++] ; Sub. [W8] from W7 (Word mode)

; Store result to [W9]

; Post-increment W8

; Post-increment W9

Before

Instruction W7 2450 W7 2450 W8 1808 W8 180A W9 2020 W9 2022 808 92E4 Data 1808 92E4 020 A557 Data 2020 916C SR 0000 SR 010C (DC, N, OV = 1)

SUB

Subtract Accumulators

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
				X	X	X	X

Syntax: {label:} SUB Acc

Operands: Acc [A, B]

Operation: If (Acc = A):

$$ \overline {{\mathrm{ACCA} - \mathrm{ACCB}}} \rightarrow \mathrm{ACCA} $$

Else:

$$ \mathrm{ACCB} - \mathrm{ACCA} \rightarrow \mathrm{ACCB} $$

Status Affected: OA, OB, OAB, SA, SB, SAB

Encoding:

1100

1011

A011

0000

0000

0000

Description: Subtract the contents of the unspecified accumulator from the contents of Acc and store the result back into Acc. This instruction performs a 40-bit subtraction.

The 'A' bit specifies the destination accumulator.

Words: 1

Cycles: 1

Example 1:	SUB A	; Subtract ACCB from ACCA
		; Store the result to ACCA
		; CORCON = 0x0000 (no saturation)

	Before Instruction		After Instruction
ACCA	76 120F 098A	ACCA	52 1EFC 4D73
ACCB	23 F312 BC17	ACCB	23 F312 BC17
CORCON	0000	CORCON	0000
SR	0000	SR	1100 (OA, OB = 1)

Example 2:	SUB B	; Subtract ACCA from ACCB
		; Store the result to ACCB
		; CORCON = 0x0040 (SATB = 1)

	Before Instruction		After Instruction
ACCA	FF 9022 2EE1	ACCA	FF 9022 2EE1
ACCB	00 2456 8F4C	ACCB	00 7FFF FFFF
CORCON	0040	CORCON	0040
SR	0000	SR	1400 (SB, SAB = 1)

SUBB

Subtract WREG and Carry Bit from f

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	IC33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} SUBB{.B} f {,WREG}

Operands: f [0 8191]

Operation: (f) - (WREG) - (C) → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding:

Description:

1011

0101

1BDf

ffff

ffff

ffff

Subtract the contents of the default Working register WREG and the Borrow flag (Carry flag inverse, ) from the contents of the specified file register, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: SUBB.B 0x1FFF ; Sub. WREG and C from (0x1FFF) (Byte mode) ; Store result to 0x1FFF

	Before Instruction	After Instruction
WREG (W0)	7804	7804
Data 1FFE	9439	8F39
SR	0000	0011 (DC, C = 1)

Example 2: SUBB 0xA04, WREG ; Sub. WREG and C from (0xA04) (Word mode); Store result to WREG

	Before Instruction	After Instruction
WREG (W0)	6234	WREG (W0)
Data 0A04	6235	Data 0A04
SR	0000	SR

SUBB

Subtract Wn from Literal with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

X

Syntax: {label:} SUBB{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation

lit10 ∈ [0 ... 1023] for word operation

Wn ∈ [W0 ... W15]

Operation: (Wn) - lit10 - (C) → Wn

Status Affected: DC, N, OV, Z, C

Encoding:

1011

0001

1Bkk

kkkk

kkkk

dddd

Description: Subtract the unsigned 10-bit literal operand and the Borrow flag (Carry flag inverse, ) from the contents of the Working register Wn, and store the result back in the Working register Wn. Register Direct Addressing must be used for Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-Bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: SUBB.B #0x23, W0 ; Sub. 0x23 and from W0 (Byte mode); Store result to W0

Example 2: SUBB #0x108, W4 ; Sub. 0x108 and from W4 (Word mode) ; Store result to W4

SUBB

Subtract Short Literal from Wb with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPI	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SUBB{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb) - lit5 - (C) → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0101

lwww

wBqq

qddd

d11k

kkkk

Description: Subtract the 5-bit unsigned literal operand and the Borrow flag (Carry flag inverse, ) from the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: SUBB.B W4, #0x10, W5 ; Sub. 0x10 and C from W4 (Byte mode) ; Store result to W5

Before Instruction W4 1782 W4 1782 W5 7804 W5 7871 SR 0000 SR 0005 (OV, C = 1) After Instruction

Example 2: SUBB W0, #0x8, [W2++] ; Sub. 0x8 and C from W0 (Word mode) ; Store result to [W2] ; Post-increment W2

Before Instruction W0 0009 W0 0009 W2 2004 W2 2006 Data 2004 A557 Data 2004 0000 SR 0002 (Z = 1) SR 0103 (DC, Z, C = 1) After Instruction

SUBB

Subtract Ws from Wb with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F ds	C33F ds	PIC33E ds	PIC33C
	X	X		X	X		X

Syntax: {label:} SUBB{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb) - (Ws) - (C) → Wd

Status Affected: DC, N, OV, Z, C

Encoding: 0101 lwww wBqq qddd dppp ssss

Description: Subtract the contents of the source register Ws and the Borrow flag (Carry flag inverse, ) from the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: SUBB.B W0, W1, W0 ; Sub. W1 and C from W0 (Byte mode) ; Store result to W0

	Before Instruction		After Instruction
W0	1732 W0	17ED
W1	7844 W1	7844
SR	0000 SR	0108 (DC, N = 1)

Example 2: SUBB W7, [W8++, [W9++] ; Sub. [W8] and C from W7 (Word mode) ; Store result to [W9] ; Post-increment W8 ; Post-increment W9

	Before Instruction		After Instruction
W7	2450 W7	2450
W8	1808 W8	180A
W9	2022 W9	2024
Data 1808	92E4 Data	1808 92E4
Data 2022	A557 Data	2022 916B
SR	0000 SR	010C (DC, N, OV = 1)

SUBBR

Subtract f from WREG with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	IC33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} SUBBR{.B} f {,WREG}

Operands: f [0 8191]

Operation: (WREG) - (f) - (C) → destination designated by D

Status Affected: DC, N, OV, Z, C

Encoding:

Description:

1011

1101

1BDf

ffff

ffff

ffff

Subtract the contents of the specified file register f and the Borrow flag (Carry flag inverse, ) from the contents of WREG, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
3: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".
Example 1: SUBBR.B 0x803 ; Sub. (0x803) and from WREG (Byte mode); Store result to 0x803

	Before Instruction		After Instruction
WREG (W0)	7804		WREG (W0)	7804
Data 0802	9439		Data 0802	6F39
SR	0002	(Z = 1)	SR	0000

Example 2: SUBBR 0xA04, WREG ; Sub. (0xA04) and from WREG (Word mode); Store result to WREG

	Before Instruction		After Instruction
WREG (W0)	6234	WREG (W0)	FFFE
Data 0A04	6235	Data 0A04	6235
SR	0000	SR	0008(N = 1)

SUBBR
Subtract Wb from Short Literal with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC3	3C
	X	X		X	X		X

Syntax: {label:} SUBBR{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: lit5 - (Wb) - (C) → Wd

Status Affected: DC, N, OV, Z, C

Encoding:

0001

lwww

wBqq

qddd

d11k

kkkk

Description:

Subtract the contents of the base register Wb and the Borrow flag (Carry flag inverse, ) from the 5-bit unsigned literal, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing must be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a five-bit integer number.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1

Example 1: SUBBR.B W0, #0x10, W1 ; Sub. W0 and C from 0x10 (Byte mode) ; Store result to W1

Before Instruction W0 F310 W0 F310 W1 786A W1 7800 SR 0003 (Z, C = 1)

After Instruction SR 0103 (DC, Z, C = 1)

Example 2: SUBBR W0, #0x8, [W2++] ; Sub. W0 and C from 0x8 (Word mode) ; Store result to [W2] ; Post-increment W2

Before Instruction W0 0009 W0 0009 W2 2004 W2 2006 Data 2004 A557 Data 2004 FFFE SR 0020 (Z = 1) SR 0108 (DC, N = 1) After Instruction

SUBBR
Subtract Wb from Ws with Borrow

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} SUBBR{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Ws) - (Wb) - ()

Status Affected: DC, N, OV, Z, C

Encoding: 0001 lwww wBqq qddd dppp ssss

Description:

Subtract the contents of the base register Wb and the Borrow flag (Carry flag inverse, ) from the contents of the source register Ws, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

2: The Z flag is "sticky" for ADDC, CPB, SUBB and SUBBR. These instructions can only clear Z.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: SUBBR.B W0, W1, W0; Sub. W0 and C from W1 (Byte mode)

; Store result to W0

Before

Instruction
W0	1732 W0	1711
W1	7844 W1	7844
SR	0000 SR	0001 (C = 1)

Example 2: SUBBR W7, [W8++, [W9++] ; Sub. W7 and C from [W8] (Word mode)

; Store result to [W9]

; Post-increment W8

; Post-increment W9

Before

Instruction			Instruction
W7	2450 W7	2450
W8	1808 W8	180A
W9	2022 W9	2024
808	92E4 Data	1808 92E4
022	A557 Data	2022 6E93
SR	0000 SR	0005 (OV, C = 1)

SUBR

Subtract f from WREG

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	IC33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} SUBR{.B} f {,WREG}
Operands: f [0 8191]
Operation: (WREG) - (f) → destination designated by D
Status Affected: DC, N, OV, Z, C

Encoding:

1011

1101

0BDF

ffff

ffff

ffff

Description:

Subtract the contents of the specified file register from the contents of the default Working register WREG and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to Working register W0.
Words: 1
Cycles: 1^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1:	SUBR.B 0x1FFF	; Sub. (0x1FFF) from WREG (Byte mode)
		; Store result to 0x1FFF

	Before Instruction
WREG (W0)	7804
Data 1FFE	9439
SR	0000

	After Instruction
WREG (W0)	7804
Data 1FFE	7039
SR	0000

Example 2:

SUBR 0xA04, WREG ; Sub. (0xA04) from WREG (Word mode) ; Store result to WREG

	Before Instruction
WREG (W0)	6234
Data 0A04	6235
SR	0000

	After Instruction
WREG (W0)	FFFF
Data 0A04	6235
SR	0008 (N = 1)

SUBR

Subtract Wb from Short Literal

Implemented in: PIC24F PIC24H	PIC24E dsP	PIC30F dsPIC	C33F dsPIC	33E dsPIC33C
	X	X		X	X

Syntax: {label:} SUBR{.B} Wb, #lit5, Wd
[Wd]
[Wd++]
[Wd--]
[++Wd]
[--Wd]
Operands: Wb ∈ [W0 ... W15]
lit5 ∈ [0 ... 31]
Wd ∈ [W0 ... W15]
Operation: lit5 - (Wb) → Wd
Status Affected: DC, N, OV, Z, C
Encoding: 0001 0www wBqq qddd d11k kkkk

Description: Subtract the contents of the base register Wb from the unsigned 5-bit literal operand and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd. The 'w' bits select the address of the base register. The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 'q' bits select the destination addressing mode. The 'd' bits select the destination register. The 'k' bits provide the literal operand, a five-bit integer number.
Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.
Words: 1
Cycles: 1

Example 1:	SUBR.B W0, #0x10, W1	; Sub. W0 from 0x10 (Byte mode)
		; Store result to W1

Before

Instruction
W0	F310
W1	786A
SR	0000

After

Instruction
W0	F310
W1	7800
SR	0103(DC, Z, C = 1)

Example 2: SUBR W0, #0x8, [W2++] ; Sub. W0 from 0x8 (Word mode)

; Store result to [W2]

; Post-increment W2

Before Instruction		After Instruction
W0	0009	W0	0009
W2	2004	W2	2006
Data 2004	A557	Data 2004	FFFF
SR	0000	SR	0108(DC, N = 1)

SUBR
Subtract Wb from Ws

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsPIC	C33F dsPIC	33E dsPIC38C
	X	X		X	X		X

Syntax: {label:} SUBR{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Ws) - (Wb) → Wd

Status Affected: DC, N, OV, Z, C

Encoding: 0001 0www wBqq qddd dppp ssss

Description:

Subtract the contents of the base register Wb from the contents of the source register Ws and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: SUBR.B W0, W1, W0 ; Sub. W0 from W1 (Byte mode) ; Store result to W0

Before
Instruction

W0	1732 W0	1712
W1	7844 W1	7844
SR	0000 SR	0001 (C = 1)

After
Instruction

Example 2: SUBR W7, [W8++], [W9++] ; Sub. W7 from [W8] (Word mode)

; Store result to [W9]
; Post-increment W8
; Post-increment W9

Before
Instruction

W7	2450 W7	2450
W8	1808 W8	180A
W9	2022 W9	2024
808	92E4 Data	1808 92E4
022	A557 Data	2022 6E94
SR	0000 SR	0005 (OV, C)

After
Instruction

1)

SWAP

Byte or Nibble Swap Wn

Implemented in: PIC24F PIC24H	PIC24E ds	PIC30F dsP	IC33F dsPIC	C33E dsPIC	33C
	X	X	X	X	X	X	X

Syntax: {label:} SWAP{.B} Wn

Operands: Wn ∈ [W0 ... W15]

Operation: For Byte Operation:

$$ (W n) < 7: 4 > \leftrightarrow (W n) < 3: 0 > $$

For Word Operation:

$$ (W n) < 1 5: 8 > \leftrightarrow (W n) < 7: 0 > $$

Status Affected: None

Encoding: 1111 1101 1B00 0000 0000 ssss

Description:

Swap the contents of the Working register Wn. In Word mode, the two bytes of Wn are swapped. In Byte mode, the two nibbles of the Least Significant Byte of Wn are swapped and the Most Significant Byte of Wn is unchanged. Register Direct Addressing must be used for Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte). The 's' bits select the address of the Working register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: SWAP.B W0 ; Nibble swap (W0)

Before
Instruction

W0	AB87
SR	0000

After
Instruction

W0	AB78
SR	0000

Example 2: SWAP W0 ; Byte swap (W0)

Before
Instruction

W0	8095
SR	0000

After
Instruction

W0	9580
SR	0000

TBLRDH
Table Read High

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X	X	X	X	X	X

Syntax: {label:} TBLRDH{.B} [Ws], Wd

[Ws++, [Wd]

[Ws--], [Wd++]

[++Ws], [Wd--]

[--Ws], [++Wd]

[--Wd]

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

If (LSB(Ws) = 1):

0 → Wd

Else:

Program Mem [(TBLPAG),(Ws)] <23:16> → Wd

For Word Operation:

Program Mem [(TBLPAG),(Ws)] <23:16> → Wd <7:0>

0 → Wd <15:8>

Status Affected: None

Encoding:

1011

1010

1Bqq

qddd

dppp

ssss

Description:

Read the contents of the most significant word of program memory and store it to the destination register Wd. The target word address of program memory is formed by concatenating the 8-bit Table Pointer register, TBLPAG<7:0>, with the Effective Address specified by Ws. Indirect Addressing must be used for Ws and either Register Direct or Indirect Addressing may be used for Wd.

In Word mode, zero is stored to the Most Significant Byte of the destination register (due to non-existent program memory), and the third program memory byte (PM<23:16>) at the specified program memory address, is stored to the Least Significant Byte of the destination register.

In Byte mode, the source address depends on the contents of Ws. If Ws is not word-aligned, zero is stored to the destination register (due to non-existent program memory). If Ws is word-aligned, the third program memory byte (PM<23:16>), at the specified program memory address, is stored to the destination register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .W extension to denote a word move, but it is not required.

Words: 1

Cycles: 2 (PIC24F, PIC24H, dsPIC30F, dsPIC33F)

5 (PIC24E, dsPIC33E, dsPIC33C)

Example 1: TBLRDH.B [W0], [W1++] ; Read PM (TBLPAG:[W0]) (Byte mode)

; Store to [W1]

; Post-increment W1

Before

Instruction
W0	0812 W0	0812
W1	0F71 W1	0F72
F70	0944 Data	0F70 B

Program 01 0812 EF 2042 Program 01 0812 EF 2042

TBLPAG	0001 TBLPAG
SR	0000

After
Instruction

2

0000

Example 2: TBLRDH [W6++, W8 ; Read PM (TBLPAG:[W6]) (Word mode)

; Store to W8

; Post-increment W6

Before
Instruction

W6

3406

W6

3408

W8

65B1

Program 00 3406 29 2E40 Program 00 3406

TBLPAG 0000 TBLPAG

SR 0000 SR

After
Instruction

W8

0029

29 2E40

0000

TBLRDL
Table Read Low

Implemented in: PIC24F PIC	24H PIC24E dsPIC30F ds	PIC33F ds	PIC33E dsP	IC33C
	X	X	X	X	X	X

Syntax: {label:} TBLRDL{.B} [Ws], Wd

$$ [ \mathrm{Ws} + + ], [ \mathrm{Wd} ] $$

$$ [ \mathrm{Ws} - - ], \quad [ \mathrm{Wd} + + ] $$

$$ [ + + \mathrm{Ws} ], [ \mathrm{Wd} - - ] $$

$$ [ - - W s ], \quad [ + + W d ] $$

$$ [ - - W d ] $$

Operands: Ws ∈ [W0 ... W15]

$$ \mathrm{Wd} \in [ \mathrm{W0} \dots \mathrm{W15} ] $$

Operation: For Byte Operation:

$$ \underline {{\text { If } (\text { LSB } (W s) = 1)}}: $$

$$ \text { Program Mem } [ (\text { TBLPAG }), (\text { Ws }) ] < 1 5: 8 > \rightarrow \text { Wd } $$

$$ \underline {{\text { Else: }}} $$

$$ \text { Program Mem } [ (\text { TBLPAG }), (\text { Ws }) ] < 7: 0 > \rightarrow \text { Wd } $$

For Word Operation:

$$ \text { Program Mem } [ (\text { TBLPAG }), (\text { Ws }) ] < 1 5: 0 > \rightarrow \text { Wd } $$

Status Affected: None

Encoding:

1011

1010

0Bqq

qddd

dppp

ssss

Description: Read the contents of the least significant word of program memory and store it to the destination register Wd. The target word address of program memory is formed by concatenating the 8-bit Table Pointer register, TBLPAG<7:0>, with the Effective Address specified by Ws. Indirect Addressing must be used for Ws and either Register Direct or Indirect Addressing may be used for Wd.

In Word mode, the lower 2 bytes of program memory are stored to the destination register. In Byte mode, the source address depends on the contents of Ws. If Ws is not word-aligned, the second byte of the program memory word (PM<15:7>) is stored to the destination register. If Ws is word-aligned, the first byte of the program memory word (PM<7:0>) is stored to the destination register.

The 'B' bit selects byte or word operation ('0' for word mode, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .W extension to denote a word move, but it is not required.

Words: 1

Cycles: 2 (PIC24F, PIC24H, dsPIC30F, dsPIC33F)

5 (PIC24E, dsPIC33E, dsPIC33C)

Example 1: TBLRDL.B [W0++, W1 ; Read PM (TBLPAG:[W0]) (Byte mode)

; Store to W1

; Post-increment W0

Before

	Instruction
W0	0813 W0	0814
W1	0F71 W1	0F20
F70	0944 Data 0F	70 B
812	EF 2042 Program
PAG	0001 TBLPAG
SR	0000

After

Instruction

2

0000

Example 2: TBLRDL [W6], [W8++] ; Read PM (TBLPAG:[W6]) (Word mode)

; Store to W8

; Post-increment W8

Before

	Instruction
W6	3406 W6	3406
W8	1202 W8	1204
202	658B	Data 1202 2E40
406	29 2E40	Program 00 3406
PAG	0000 TBLPAG	0000
SR	0000	SR

After

Instruction

29 2E40

0000

TBLWTH
Table Write High

Implemented in: PIC24F	PIC24H PIC24E dsPIC30F		dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X	X	X
Syntax:	{label:}	TBLWTH{.B}	Ws, [Wd][Ws], [Wd++] [Ws++], [Wd--] [Ws--], [++Wd] [++Ws], [--Wd] [--Ws],
Operands:	Ws ∈ [W0 ... W15]Wd ∈ [W0 ... W15]
Operation:	For Byte Operation:If (LSB(Wd) = 1):NOPElse:(Ws) → Program Mem [(TBLPAG),(Wd)]<23:16>For Word Operation:(Ws)<7:0> → Program Mem [(TBLPAG),(Wd)] <23:16>
Status Affected:	None
Encoding:	1011	1011	1Bqq	qddd	dppp	ssss
Description:	Store the contents of the working source register Ws to the most significant word of program memory. The destination word address of program memory is formed by concatenating the 8-bit Table Pointer register, TBLPAG<7:0>, with the Effective Address specified by Wd. Either Direct or Indirect Addressing may be used for Ws and Indirect Addressing must be used for Wd.Since program memory is 24 bits wide, this instruction can only write to the upper byte of program memory (PM<23:16>). This may be performed using a Wd that is word-aligned in Byte mode or Word mode. If Byte mode is used with a Wd that is not word-aligned, no operation is performed.The ‘B’ bit selects byte or word operation (‘0’ for word, ‘1’ for byte).The ‘q’ bits select the destination addressing mode.The ‘d’ bits select the destination register.The ‘p’ bits select the source addressing mode.The ‘s’ bits select the source register.Note: The extension .B in the instruction denotes a byte move rather than a word move.You may use a .W extension to denote a word move, but it is not required.
Words:	1
Cycles:	2(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multicycle Instructions".

Example 1: TBLWTH.B [W0++, [W1] ; Write [W0]... (Byte mode) ; to PM Latch High (TBLPAG:[W1]) ; Post-increment W0

Note: Only the program latch is written to. The contents of program memory are not updated until the Flash memory is programmed using the procedure described in the specific device family reference manual.

Example 2: TBLWTH W6, [W8++] ; Write W6... (Word mode) ; to PM Latch High (TBLPAG:[W8]) ; Post-increment W8

Note: Only the program latch is written to. The contents of program memory are not updated until the Flash memory is programmed using the procedure described in the specific device family reference manual.

TBLWTL
Table Write Low

Implemented in: PIC24F	PIC24H	PIC24	4E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X	X	X

Syntax:	{label:}	TBLWTL{.B}	Ws,	[Wd]
				[Ws],	[Wd++]
				[Ws++],	[Wd--]
				[Ws--],	[++Wd]
				[++Ws],	[--Wd]
				[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: For Byte Operation:

If (LSB(Wd) = 1): (Ws) → Program Mem [(TBLPAG),(Wd)] <15:8>

Else: (Ws) → Program Mem [(TBLPAG),(Wd)] <7:0>

For Word Operation: (Ws) → Program Mem [(TBLPAG),(Wd)] <15:0>

Status Affected: None

Encoding: 1011 1011 0Bqq qddd dppp ssss

Description: Store the contents of the working source register Ws to the least significant word of program memory. The destination word address of program memory is formed by concatenating the 8-bit Table Pointer register, TBLPAG<7:0>, with the Effective Address specified by Wd. Either Direct or Indirect Addressing may be used for Ws and Indirect Addressing must be used for Wd.

In Word mode, Ws is stored to the lower 2 bytes of program memory. In Byte mode, the Least Significant bit of Wd determines the destination byte. If Wd is not word-aligned, Ws is stored to the second byte of program memory (PM<15:8>). If Wd is word-aligned, Ws is stored to the first byte of program memory (PM<7:0>).

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte move rather than a word move. You may use a .W extension to denote a word move, but it is not required.

Words: 1

Cycles: 2^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 “Multicycle Instructions”.

Example 1: TBLWTL.B W0, [W1++] ; Write W0... (Byte mode) ; to PM Latch Low (TBLPAG:[W1]) ; Post-increment W1

Note: Only the program latch is written to. The contents of program memory are not updated until the Flash memory is programmed using the procedure described in the specific device family reference manual.

Example 2: TBLWTL [W6], [W8] ; Write [W6]... (Word mode) ; to PM Latch Low (TBLPAG:[W8]) ; Post-increment W8

Note: Only the program latch is written to. The contents of program memory are not updated until the Flash memory is programmed using the procedure described in the specific device family reference manual.

ULNK

Deallocate Stack Frame

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X		X

Syntax: {label:} ULNK

Operands: None

Operation: W14 → W15

$$ (W 1 5) - 2 \rightarrow W 1 5 $$

$$ (T O S) \rightarrow W 1 4 $$

Status Affected: None

Encoding:

1111

1010

1000

0000

0000

0000

Description:

This instruction deallocates a stack frame for a subroutine calling sequence. The stack frame is deallocated by setting the Stack Pointer (W15) equal to the Frame Pointer (W14) and then POPping the stack to reset the Frame Pointer (W14).

Words: 1

Cycles: 1

Example 1: ULNK ; Unlink the stack frame

	Before Instruction		After Instruction
W14	2002	W14	2000
W15	20A2	W15	2000
Data 2000	2000	Data 2000	2000
SR	0000	SR	0000

Example 2: ULNK ; Unlink the stack frame

	Before Instruction		After Instruction
W14	0802	W14	0800
W15	0812	W15	0800
Data 0800	0800	Data 0800	0800
SR	0000	SR	0000

ULNK

Deallocate Stack Frame

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	C33E dsPIC	33C
			X			X	X

Syntax: {label:} ULNK

Operands: None

Operation: W14 → W15

$$ (W 1 5) - 2 \rightarrow W 1 5 $$

$$ (T O S) \rightarrow W 1 4 $$

$$ 0 \rightarrow \text { S F A b i t } $$

Status Affected: SFA

Encoding:

1111 1010 1000

0000 0000

0000

Description:

This instruction deallocates a stack frame for a subroutine calling sequence. The stack frame is deallocated by setting the Stack Pointer (W15) equal to the Frame Pointer (W14) and then POPping the stack to reset the Frame Pointer (W14).

Words: 1

Cycles: 1

Example 1: ULNK ; Unlink the stack frame

	Before Instruction		After Instruction
W14	2002	W14	2000
W15	20A2	W15	2000
Data 2000	2000	Data 2000	2000
SR	0000	SR	0000

Example 2: ULNK ; Unlink the stack frame

	Before Instruction		After Instruction
W14	0802	W14	0800
W15	0812	W15	0800
Data 0800	0800	Data 0800	0800
SR	0000	SR	0000

VFSLV

Verify Slave Processor Program RAM

Implemented in: PIC24F PIC2	4H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
							X

Syntax: {label:} VFSLV [Wns], [Wnd++,] #lit2

[Wns++]

Operands: Wns ∈ [W0 ... W15];

Wnd ∈ [W0 ... W15];

lit2 ∈ [0 ... 3];

Operation: If Master (EAs) != Slave EAd

Master VERFERR (MSIxSTAT<11>) = 1

Status Affected:

None

Encoding:

0000

0011

10kk

0ddd

d0p1

ssss

Description:

This instruction reads a single instruction word from the target Slave PRAM image (held in the Master program space Flash) and compares it to the value in the Slave PRAM at the destination address. The source address must be located within PSV address space (i.e., DSRPAG ≥ 0x200). The destination address uses DSWPAG and the destination EA to create a 24-bit Slave program space PRAM write address.

Starting with an aligned double instruction word (destination address, see Note 1), the contents of the source Effective Address (in Master program space) are compared with the destination Effective Address (in the Slave PRAM address space) in order to verify the PRAM contents.

If the (single instruction word) destination address is even, the data is captured in the Slave PRAM wrapper. If the (single instruction word) destination address is odd, the ECC parity bits are calculated from the current and captured source data (48 bits), and compared. If the data and ECC parity are not the same, the VERFERR (MSIxSTAT<11>) status bit is set.

The target Slave processor is selected by the value defined by lit2.

The instruction may be regarded as a PSV operation, and therefore, may be executed within a REPEAT loop to accelerate data processing.

The 's' bits select the address of the source register.

The 'd' bits select the address of the destination register.

The 'k' bits select the target Slave processor.

The 'p' bit selects the destination addressing mode (see Note 1).

Note 1: This instruction supports a subset of addressing modes. The source addressing mode bit field is constrained to 2 options and the destination addressing mode bit field is not required.

2: An aligned double instruction word destination address is an even address that addresses the least significant word of a double instruction word.
3: This instruction only supports Word mode.

Words: 1

Cycles: 1

XOR

Exclusive OR f and WREG

Implemented in: PIC24F PIC24H PIC24E	dsPIC30F ds	PIC33F dsP	IC33E dsPIC	33C
	X	X		X	X

X

Syntax: {label:} XOR{.B} f {,WREG}
Operands: f [0 8191]
Operation: (f).XOR.(WREG) → destination designated by D
Status Affected: N, Z

Encoding:

1011

0110

1BDf

ffff

ffff

ffff

Description:

Compute the logical exclusive OR operation of the contents of the default Working register WREG and the contents of the specified file register, and place the result in the destination register. The optional WREG operand determines the destination register. If WREG is specified, the result is stored in WREG. If WREG is not specified, the result is stored in the file register.
The 'B' bit selects byte or word operation ('0' for word, '1' for byte).
The 'D' bit selects the destination register ('0' for WREG, '1' for file register).
The 'f' bits select the address of the file register.
Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.
2: The WREG is set to working register W0.
Words: 1
Cycles: 1 ^(1)
Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: XOR.B 0x1FFF ; XOR (0x1FFF) and WREG (Byte mode) ; Store result to 0x1FFF

	Before Instruction		After Instruction
WREG (W0)	7804	WREG (W0)	7804
Data 1FFE	9439	Data 1FFE	9039
SR	0000	SR	0008	(N = 1)

Example 2:

XOR

0xA04, WREG ; XOR (0xA04) and WREG (Word mode) ; Store result to WREG

	Before Instruction	After Instruction
WREG (W0)	6234	C267
Data 0A04	A053	A053
SR	0000	0008(N=1)

XOR

Exclusive OR Literal and Wn

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} XOR{.B} #lit10, Wn

Operands: lit10 ∈ [0 ... 255] for byte operation lit10 ∈ [0 ... 1023] for word operation Wn ∈ [W0 ... W15]

Operation: lit10.XOR.(Wn) → Wn Status Affected: N, Z

Encoding: 1011 0010 1Bkk kkkk kkkk dddd

Description: Compute the logical exclusive OR operation of the unsigned 10-bit literal operand and the contents of the Working register Wn, and store the result back in the Working register Wn. Register Direct Addressing must be used for Wn.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'k' bits specify the literal operand.

The 'd' bits select the address of the Working register.

Note 1: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .w extension to denote a word operation, but it is not required.

2: For byte operations, the literal must be specified as an unsigned value [0:255]. See Section 4.6 "Using 10-bit Literal Operands" for information on using 10-bit literal operands in Byte mode.

Words: 1

Cycles: 1

Example 1: XOR.B #0x23, W0 ; XOR 0x23 and W0 (Byte mode) ; Store result to W0

Before Instruction W0 7804 W0 7827 SR 0000 After Instruction SR 0000

Example 2: XOR #0x108, W4 ; XOR 0x108 and W4 (Word mode) ; Store result to W4

Before Instruction W4 6134 W4 603C SR 0000 After Instruction SR 0000

XOR
Exclusive OR Wb and Short Literal

Implemented in: PIC24F PIC24H PIC24	E dsPIC30F	dsPIC33F ds	PIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} XOR{.B} Wb, #lit5, Wd

[Wd]

[Wd++]

[Wd--]

[++Wd]

[--Wd]

Operands: Wb ∈ [W0 ... W15]

lit5 ∈ [0 ... 31]

Wd ∈ [W0 ... W15]

Operation: (Wb).XOR.lit5 → Wd

Status Affected: N, Z

Encoding: 0110 1www wBqq qddd d11k kkkk

Description: Compute the logical exclusive OR operation of the contents of the base register Wb and the unsigned 5-bit literal operand, and place the result in the destination register Wd.

Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'k' bits provide the literal operand, a 5-bit integer number.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1

Example 1: XOR.B W4, #0x14, W5 ; XOR W4 and 0x14 (Byte mode)

; Store result to W5

Before
Instruction

W4	C822
W5	1200
SR	0000

After
Instruction

W4 C822
W5	1234
SR	0000

Example 2: XOR W2, #0x1F, [W8++] ; XOR W2 by 0x1F (Word mode)

; Store result to [W8]

; Post-increment W8

Before
Instruction

W2	8505
W8	1004
Data 1004	6628
SR	0000

After
Instruction

W2	8505
W8	1006
Data 1004	851A
SR	0008 (N = 1)

XOR

Exclusive OR Wb and Ws

Implemented in: PIC24F P	C24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X

Syntax: {label:} XOR{.B} Wb, Ws, Wd

[Ws], [Wd]

[Ws++], [Wd++]

[Ws--], [Wd--]

[++Ws], [++Wd]

[--Ws], [--Wd]

Operands: Wb ∈ [W0 ... W15]

Ws ∈ [W0 ... W15]

Wd ∈ [W0 ... W15]

Operation: (Wb).XOR.(Ws) → Wd

Status Affected: N, Z

Encoding:

0110

lwww

wBqq

qddd

dppp

ssss

Description: Compute the logical exclusive OR operation of the contents of the source register Ws and the contents of the base register Wb, and place the result in the destination register Wd. Register Direct Addressing must be used for Wb. Either Register Direct or Indirect Addressing may be used for Ws and Wd.

The 'w' bits select the address of the base register.

The 'B' bit selects byte or word operation ('0' for word, '1' for byte).

The 'q' bits select the destination addressing mode.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note: The extension .B in the instruction denotes a byte operation rather than a word operation. You may use a .W extension to denote a word operation, but it is not required.

Words: 1

Cycles: 1(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: XOR.B W1, [W5++], [W9++]; XOR W1 and [W5] (Byte mode)

; Store result to [W9]

; Post-increment W5 and W9

Before

Instruction
W1	AAAA W1	AAAA
W5	2000 W5	2001
W9	2600 W9	2601
Data 2000	115A Data	2000 115A
Data 2600	0000 Data	2600 00F0
SR	0000 SR	0008 (N = 1)

After

Instruction

Example 2: XOR W1, W5, W9; XOR W1 and W5 (Word mode)

; Store the result to W9

Before

Instruction
W1	FEDC W1	FEDC
W5	1234 W5	1234
W9	A34D W9	ECE8
SR	0000 SR	0008 (N = 1)

After

Instruction

ZE
Zero-Extend Ws

Implemented in: PIC24F PIC24H PIC24E dsPIC30F dsPIC33F dsPIC33E dsPIC33C
	X	X		X	X		X

Syntax: {label:} ZE Ws, Wnd

[Ws],

[Ws++]

[Ws--],

[++Ws],

[--Ws],

Operands: Ws ∈ [W0 ... W15]

Wnd ∈ [W0 ... W15]

Operation: Ws<7:0> → Wnd<7:0>

0 → Wnd<15:8>

Status Affected: N, Z, C

Encoding:

1111

1011

1000

0ddd

dppp

ssss

Description: Zero-extend the Least Significant Byte in source Working register Ws to a 16-bit value and store the result in the destination Working register Wnd. Either Register Direct or Indirect Addressing may be used for Ws and Register Direct Addressing must be used for Wnd. The N flag is cleared and the C flag is set because the zero-extended word is always positive.

The 'd' bits select the destination register.

The 'p' bits select the source addressing mode.

The 's' bits select the source register.

Note 1: This operation converts a byte to a word and it uses no .B or .W extension.

2: The source Ws is addressed as a byte operand, so any address modification is by one.

Words: 1

Cycles: 1^(1)

Note 1: In dsPIC33E, dsPIC33C and PIC24E devices, the listed cycle count does not apply to read and Read-Modify-Write operations on non-CPU Special Function Registers. For more details, see Note 3 in Section 3.2.1 "Multi-Cycle Instructions".

Example 1: ZE W3, W4 ; zero-extend W3 ; Store result to W4

	Before Instruction		After Instruction
W3	7839 W3	7839
W4	1005 W4	0039
SR	0000 SR	0001 (C = 1)

Example 2: ZE [W2++], W12 ; Zero-extend [W2] ; Store to W12 ; Post-increment W2

	Before Instruction					After Instruction
	W2	0900	W2	0901
	W12	1002	W12	008F
	Data 0900	268F		Data 0900	268F
	SR	0000	SR	0001 (C = 1)

NOTES:

Section 6. Built-in Functions

HIGHLIGHTS

This section of the manual contains the following major topics:

6.1 Introduction 460
6.2 Built-in Function List....461

6.1 INTRODUCTION

This section describes the built-in functions that are specific to the MPLAB ^® C Compiler for PIC24 MCUs and dsPIC ^® DSCs (formerly MPLAB C30).

Built-in functions give the C programmer access to assembler operators or machine instructions that are currently only accessible using in-line assembly, but are sufficiently useful that they are applicable to a broad range of applications. Built-in functions are coded in C source files syntactically like function calls, but they are compiled to assembly code that directly implements the function and does not involve function calls or library routines.

There are a number of reasons why providing built-in functions is preferable to requiring programmers to use in-line assembly. They include the following:

1. Providing built-in functions for specific purposes simplifies coding.
2. Certain optimizations are disabled when in-line assembly is used. This is not the case for built-in functions.
3. For machine instructions that use dedicated registers, coding in-line assembly while avoiding register allocation errors can require considerable care. The built-in functions make this process simpler as you do not need to be concerned with the particular register requirements for each individual machine instruction.

The built-in functions are listed below followed by their individual detailed descriptions.

• __builtin_addab
• _builtin_add
• __builtin_btg
• __builtin_clr
• _builtin_clr_prefetch
  • _builtin_divf
  • _builtin_divmodsd
• _builtin_divmodud
  • builtin_divsd
  • builtin_divud
• _builtin_dmaoffset
• _builtin_ed
• _builtin_edac
• builtin_edsoffset
• __builtin_edspage
  • _builtin_fbcl
  • builtin lac
• _builtin_mac
• _builtin_modsd
• _builtin_modud
• _builtin_movsac
• __builtin_mpy
• _builtin_mpyn
• _builtin_msc
  • builtin mulss
• __builtin_mulsu
  • _builtin_mulus
  • _builtin_muluu
  • _builtin_nop
• __builtin_psvpage
  • builtin_psvoffset
  • builtin_readsfr
  • __builtin_return_address
• _builtin_sac
• _builtin_sacr
  • builtin_sftac
• _builtin_subab
• _builtin_tbladdress
• builtin tblpage
• _builtin_tbloffset
  • _builtin_tblrdh
  • _builtin_tblrdl
  • _builtin_tblwth
  • _builtin_tblwtl

This section describes only the built-in functions related to the CPU operations. The compiler provides additional built-in functions for operations, such as writing to Flash program memory and changing the oscillator settings. Refer to the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284) for a complete list of compiler built-in functions.

6.2 BUILT-IN FUNCTION LIST

This section describes the programmer interface to the compiler built-in functions. Since the functions are "built-in", there are no header files associated with them. Similarly, there are no command-line switches associated with the built-in functions – they are always available. The built-in function names are chosen such that they belong to the compiler's namespace (they all have the prefix: __builtin_), so they will not conflict with function or variable names in the programmer's namespace.

\_builtin\_addab

Description:

Adds Accumulators A and B with the result written back to the specified accumulator. For example:

register int result asm("A");
register int B asm("A");

result = __builtin_addab(result, B);
will generate:
add A 

Prototype:

int _builtin_addab(int Accum_a, int Accum_b);

Argument:

Accum_aFirst accumulator to add.
Accum_bSecond accumulator to add. 

Return Value:

Returns the addition result to an accumulator.

Assembler Operator/Machine Instruction:

add

Error Messages:

An error message appears if the result is not an Accumulator register.

\_builtin\_add

Description:

Adds value to the accumulator specified by result with a shift specified by literal shift. For example:

register int result asm("A");
int value;
result = __builtin_add(result, value, 0); 

If value is held in w0, the following will be generated:

add w0, #0, A 

Prototype:

int __builtin_add(int Accum, int value, const int shift); 

Argument:

AccumAccumulator to add.
valueInteger number to add to accumulator value.
shiftAmount to shift resultant accumulator value. 

Return Value:

Returns the shifted addition result to an accumulator.

Assembler Operator/Machine Instruction:

add 

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- Argument 0 is not an accumulator
• The shift value is not a literal within range

\_builtin\_btg

Description:

This function will generate a btg machine instruction. Some examples include:

int i;    /* near by default */
int l    __attribute__(far);

struct foo {
    int bit1:1;
} barbits;

int    bar;

void some_bittoggles() {
    register int j asm("w9");
    int k;

    k = i;

    __builtin_btg(&i,1);
    __builtin_btg(&j,3);
    __builtin_btg(&k,4);
    __builtin_btg(&l,11);

    return j+k;
} 

Note that taking the address of a variable in a register will produce a warning by the compiler and cause the register to be saved onto the stack (so that its address may be taken); this form is not recommended. This caution only applies to variables explicitly placed in registers by the programmer.

Prototype:

void __builtin_btg(unsigned int *, unsigned int 0xn);

Argument:

*A pointer to the data item for which a bit should be toggled.

0x_nA literal value in the range of 0 to 15.

Return Value:

Returns a btg machine instruction.

Assembler Operator/Machine Instruction:

btg

Error Messages:

An error message appears if the parameter values are not within range.

\_builtin\_clr

Description:

Clears the specified accumulator. For example:

register int result asm("A");
result = __builtin_clr(); 

will generate:

clr A 

Prototype:

int __builtin_clr(void); 

Argument:

None.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

clr 

Error Messages:

An error message appears if the result is not an Accumulator register.

builtin\_clr\_prefetch

Description:

Clears an accumulator and prefetch data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

If AWB is non-null, the other accumulator will be written back into the referenced variable.

For example:

register int result asm("A");
register int B asm("B");
int x_memory_buffer[256]
__attribute__((space(xmemory)));
int y_memory_buffer[256]
__attribute__((space(ymemory)));
int *xmemory;
int *ymemory;
int awb;
int xVal, yVal;

xmemory = x_memory_buffer;
ymemory = y_memory_buffer;
result = __builtin_clr(&xmemory, &xVal, 2, &ymemory, &yVal, 2, &awb, B); 

May generate:

clr A, [w8] += 2, w4, [w10] += 2, w5, w13 

The compiler may need to spill w13 to ensure that it is available for the Write-Back. It may be recommended to users that the register be claimed for this purpose.

After this instruction:

• Result will be cleared
- xVal will contain x_memory_buffer[0]
- yVal will contain y_memory_buffer[0]
- xmemory and ymemory will be incremented by 2, ready for the next MAC operation

Prototype:

int __builtin_clr_prefetch(
int **xptr, int *xval, int xincr,
int **yptr, int *yval, int yincr, int *AWB,
int AWB_accum); 

\_builtin\_clr\_prefetch (Continued)

Argument:

xptrInteger Pointer to X prefetch. xvalInteger value of X prefetch. xincrInteger increment value of X prefetch. yptrInteger Pointer to Y prefetch. yvalInteger value of Y prefetch. yincrInteger increment value of Y prefetch. AWBAccumulator Write-Back location. AWB_accumAccumulator to Write-Back.

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

clr

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- xval is a null value but xptr is not null
- yval is a null value but yptr is not null
- AWB_accum is not an accumulator and AWB is not null

\_builtin\_divf

Description:

Computes the quotient: num / den. A math error exception occurs if den is zero. Function arguments are unsigned, as is the function result.

Prototype:

unsigned int __builtin_divf(unsigned int num, unsigned int den);

Argument:

numNumerator. denDenominator.

Return Value:

Returns the unsigned integer value of the quotient: num / den.

Assembler Operator/Machine Instruction:

div.f

\_builtin\_divmodsd

Description:

Issues the 16-bit architecture's native signed divide support. Notably, if the quotient does not fit into a 16-bit result, the results (including remainder) are unexpected. This form of the built-in function will capture both the quotient and remainder.

Prototype:

signed int __builtin_divmodsd( signed long dividend, signed int divisor, signed int *remainder);

Argument:

dividendNumber to be divided. divisorNumber to divide by. remainderPointer to remainder.

Return Value:

Quotient and remainder.

Assembler Operator/Machine Instruction:

divmodsd

Error Messages:

None.

\_builtin\_divmodud

Description:

Issues the 16-bit architecture's native unsigned divide support. Notably, if the quotient does not fit into a 16-bit result, the results (including remainder) are unexpected. This form of the built-in function will capture both the quotient and remainder.

Prototype:

unsigned int __builtin_divmodud(
unsigned long dividend, unsigned int divisor,
unsigned int *remainder); 

Argument:

dividendNumber to be divided.
divisorNumber to divide by.
remainderPointer to remainder. 

Return Value:

Quotient and remainder.

Assembler Operator/Machine Instruction:

divmodud

Error Messages:

None.

\_builtin\_divsd

Description:

Computes the quotient: num / den. A math error exception occurs if den is zero. Function arguments are signed, as is the function result. The command-line option, -Wconversions, can be used to detect unexpected sign conversions.

Prototype:

int __builtin_divsd(const long num, const int den);

Argument:

numNumerator.
denDenominator. 

Return Value:

Returns the signed integer value of the quotient: num / den.

Assembler Operator/Machine Instruction:

div.sd

\_builtin\_divud

Description:

Computes the quotient: num / den. A math error exception occurs if den is zero. Function arguments are unsigned, as is the function result. The command-line option, -Wconversions, can be used to detect unexpected sign conversions.

Prototype:

unsigned int __builtin_divud(const unsigned long num, const unsigned int den);

Argument:

numNumerator. denDenominator.

Return Value:

Returns the unsigned integer value of the quotient: num / den.

Assembler Operator/Machine Instruction:

div.ud

\_builtin\_dmaoffset

Description:

Obtains the offset of a symbol within DMA memory.

For example:

unsigned int result; char buffer[256] __attribute____((space(dma)));

result = __builtin_dmaoffset(&buffer);

May generate:

mov #dmaoffset(buffer), w0

Prototype:

unsigned int __builtin_dmaoffset(const void *p);

Argument:

^* pPointer to DMA address value.

Return Value:

Returns the offset to a variable located in DMA memory.

Assembler Operator/Machine Instruction:

dmaoffset

Error Messages:

An error message appears if the parameter is not the address of a global symbol.

\_builtin\_ed

Description:

Squares sqr, returning it as the result. Also prefetches data for future square operation by computing **xptr - **yptr and storing the result in *distance.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

For example:

register int result asm("A");
int *xmemory, *ymemory;
int distance; 

result = __builtin_ed(distance,
    &xmemory, 2,
    &ymemory, 2,
    &distance); 

May generate:

ed w4*w4, A, [w8] += 2, [w10] += 2, w4 

Prototype:

int __builtin_ed(int sqr, int **xptr, int xincr, int **yptr, int yincr, int *distance); 

Argument:

sqlInteger squared value.

xptrInteger Pointer to pointer to X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yincrInteger increment value of Y prefetch.
distanceInteger Pointer to distance. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the squared result to an accumulator.

Assembler Operator/Machine Instruction:

ed

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• xptr is null
• yptr is null
• distance is null

\_builtin\_edac

Description:

Squares sqr and sums with the nominated Accumulator register, returning it as the result. Also prefetches data for future square operation by computing **xptr - **yptr and storing the result in *distance.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

For example:

register int result asm("A");
int *xmemory, *ymemory;
int distance;

result = __builtin_ed(result, distance,
    &xmemory, 2,
    &ymemory, 2,
    &distance); 

May generate:

edac w4*w4, A, [w8] += 2, [W10] += 2, w4 

Prototype:

int __builtin_edac(int Accum, int sqr,
int **xptr, int xincr, int **yptr, int yincr,
int *distance); 

Argument:

AccumAccumulator to sum.
sqlInteger squared value.
xptrInteger Pointer to pointer to X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yincrInteger increment value of Y prefetch.
distanceInteger Pointer to distance. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the squared result to specified accumulator.

Assembler Operator/Machine Instruction:

edac

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• Accum is not an Accumulator register
- xptr is null
- yptr is null
- distance is null

\_builtin\_edsoffset

Description:

Returns the EDS page offset of the object whose address is given as a parameter. The argument p must be the address of an object in Extended Data Space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 "Specifying Attributes of Variables" of the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284).

Prototype:

unsigned int __builtin_edsoffset(int *p);

Argument:

pObject address.

Return Value:

Returns the EDS page number of the object whose address is given as a parameter

Assembler Operator/Machine Instruction:

edsoffset

\_builtin\_edspage

Description:

Returns the EDS page number of the object whose address is given as a parameter. The argument p must be the address of an object in Extended Data Space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 "Specifying Attributes of Variables" of the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284).

Prototype:

unsigned int __builtin_edspage(int *p);

Argument:

pObject address.

Return Value:

Returns the EDS page number of the object whose address is given as a parameter.

Assembler Operator/Machine Instruction:

edspage

\_builtin\_fbcl

Description:

Finds the first bit change from left in value. This is useful for dynamic scaling of fixed-point data.

For example:

int result, value;
result = __builtin_fbcl(value); 

May generate:

fbcl w4, w5 

Prototype:

int __builtin_fbcl(int value); 

Argument:

valuelnteger number of first bit change.

Return Value:

Returns the shifted addition result to an accumulator.

Assembler Operator/Machine Instruction:

fbcl

Error Messages:

An error message appears if the result is not an Accumulator register.

\_builtin\_lac

Description:

Shifts value by shift (a literal between -8 and 7) and returns the value to be stored into the Accumulator register. For example:

register int result asm("A");
int value;
result = __builtin_lac(value, 3); 

May generate:

lac w4, #3, A 

Prototype:

int __builtin_lac(int value, int shift); 

Argument:

valueInteger number to be shifted. shiftLiteral amount to shift.

Return Value:

Returns the shifted addition result to an accumulator.

Assembler Operator/Machine Instruction:

lac

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• The shift value is not a literal within range

\_builtin\_mac

Description:

Computes a x b and sums with accumulator; also, prefetches data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

If AWB is non-null, the other accumulator will be written back into the referenced variable.

For example:

register int result asm("A");
register int B asm("B");
int *xmemory;
int *ymemory;
int xVal, yVal;

result = __builtin_mac(result, xVal, yVal,
    &xmemory, &xVal, 2,
    &ymemory, &yVal, 2, 0, B); 

May generate:

mac w4*w5, A, [w8] += 2, w4, [w10] += 2, w5 

Prototype:

int __builtin_mac(int Accum, int a, int b, int **xptr, int *xval, int xincr, int **yptr, int *yval, int yincr, int *AWB, int AWB_accum); 

Argument:

AccumAccumulator to sum.
aInteger multiplicand.
bInteger multiplier.
xptrInteger Pointer to pointer to X prefetch.
xvalInteger Pointer to value of X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yvalInteger Pointer to value of Y prefetch.
yincrInteger increment value of Y prefetch.
AWBAccumulator Write-Back location.
AWB_accumAccumulator to Write-Back. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

mac

\_builtin\_mac (Continued)

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• Accum is not an Accumulator register
- xval is a null value but xptr is not null
- yval is a null value but yptr is not null

- AWB_accum is not an Accumulator register and AWB is not null

\_builtin\_modsd

Description:

Issues the 16-bit architecture's native signed divide support. Notably, if the quotient does not fit into a 16-bit result, the results (including remainder) are unexpected. This form of the built-in function will capture only the remainder.

Prototype:

signed int __builtin_modsd(signed long dividend, signed int divisor);

Argument:

dividendNumber to be divided. divisorNumber to divide by.

Return Value:

Remainder.

Assembler Operator/Machine Instruction:

modsd

Error Messages:

None.

\_builtin\_modud

Description:

Issues the 16-bit architecture's native unsigned divide support. Notably, if the quotient does not fit into a 16-bit result, the results (including remainder) are unexpected. This form of the built-in function will capture only the remainder.

Prototype:

unsigned int __builtin_modud(unsigned long dividend, unsigned int divisor);

Argument:

dividendNumber to be divided. divisorNumber to divide by.

Return Value:

Remainder.

Assembler Operator/Machine Instruction:

modud

Error Messages:

None.

\_builtin\_movac

Description:

Computes nothing, but prefetches data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

If AWB is not null, the other accumulator will be written back into the referenced variable.

For example:

register int result asm("A");
int *xmemory;
int *ymemory;
int xVal, yVal;
result = __builtin_movsac(&xmemory, &xVal, 2, &ymemory, &yVal, 2, 0, 0); 

May generate:

movsac A, [w8] += 2, w4, [w10] += 2, w5 

Prototype:

int __builtin_movsac(
int **xptr, int *xval, int xincr,
int **yptr, int *yval, int yincr, int *AWB
int AWB_accum); 

Argument:

xptrInteger Pointer to pointer to X prefetch.
xvalInteger Pointer to value of X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yvalInteger Pointer to value of Y prefetch.
yincrInteger increment value of Y prefetch.
AWBAccumulator Write-Back location.
AWB_accumAccumulator to Write-Back. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns prefetch data.

Assembler Operator/Machine Instruction:

movsac

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• xval is a null value but xptr is not null
• yval is a null value but yptr is not null
• AWB_accum is not an Accumulator register and AWB is not null

\_builtin\_mpy

Description:

Computes a x b; also, prefetches data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

For example:

register int result asm("A");
int *xmemory;
int *ymemory;
int xVal, yVal;
result = __builtin_mpy(xVal, yVal,
    &xmemory, &xVal, 2,
    &ymemory, &yVal, 2); 

May generate:

mac w4*w5, A, [w8] += 2, w4, [w10] += 2, w5 

Prototype:

int __builtin_mpy(int a, int b,
int **xptr, int *xval, int xincr,
int **yptr, int *yval, int yincr); 

Argument:

alnteger multiplicand.
bInteger multiplier.
xptrInteger Pointer to pointer to X prefetch.
xvalInteger Pointer to value of X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yvalInteger Pointer to value of Y prefetch.
yincrInteger increment value of Y prefetch.
AWBInteger Pointer to accumulator selection. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

mpy

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- xval is a null value but xptr is not null
- yval is a null value but yptr is not null

\_builtin\_mpyn

Description:

Computes -a x b; also, prefetches data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

For example:

register int result asm("A");
int *xmemory;
int *ymemory;
int xVal, yVal;
result = __builtin_mpy(xVal, yVal,
    &xmemory, &xVal, 2,
    &ymemory, &yVal, 2); 

May generate:

mac w4*w5, A, [w8] += 2, w4, [w10] += 2, w5 

Prototype:

int __builtin_mpyn(int a, int b,
int **xptr, int *xval, int xincr,
int **yptr, int *yval, int yincr); 

Argument:

alnteger multiplicand.
bInteger multiplier.
xptrInteger Pointer to pointer to X prefetch.
xvalInteger Pointer to value of X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yvalInteger Pointer to value of Y prefetch.
yincrInteger increment value of Y prefetch.
AWBInteger Pointer to accumulator selection. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

mpyn

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- xval is a null value but xptr is not null
- yval is a null value but yptr is not null

\_builtin\_msc

Description:

Computes a x b and subtracts from accumulator; also, prefetches data ready for a future MAC operation.

xptr may be null to signify no X prefetch to be performed; in which case, the values of xincr and xval are ignored, but required.

yptr may be null to signify no Y prefetch to be performed; in which case, the values of yincr and yval are ignored, but required.

xval and yval nominate the address of a C variable where the prefetched value will be stored.

xincr and yincr may be the literal values: -6, -4, -2, 0, 2, 4, 6 or an integer value.

If AWB is non-null, the other accumulator will be written back into the referenced variable.

For example:

register int result asm("A");
int *xmemory;
int *ymemory;
int xVal, yVal;
result = __builtin_msc(result, xVal, yVal,
    &xmemory, &xVal, 2,
    &ymemory, &yVal, 2, 0, 0); 

May generate:

msc w4*w5, A, [w8]+=2, w4, [w10]+=2, w5 

Prototype:

int __builtin_msc(int Accum, int a, int b,
int **xptr, int *xval, int xincr,
int **yptr, int *yval, int yincr, int *AWB,
int AWB_accum); 

Argument:

AccumAccumulator to sum.
alnteger multiplicand.
blnteger multiplier.
xptrInteger Pointer to pointer to X prefetch.
xvalInteger Pointer to value of X prefetch.
xincrInteger increment value of X prefetch.
yptrInteger Pointer to pointer to Y prefetch.
yvalInteger Pointer to value of Y prefetch.
yincrInteger increment value of Y prefetch.
AWBAccumulator Write-Back location.
AWB_accumAccumulator to Write-Back. 

Note: The arguments, xptr and yptr, must point to the arrays located in the X data memory and Y data memory, respectively.

Return Value:

Returns the cleared value result to an accumulator.

Assembler Operator/Machine Instruction:

msc

\_builtin\_msc (Continued)

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- Accum is not an Accumulator register
- xval is a null value but xptr is not null
- yval is a null value but yptr is not null

- AWB_accum is not an Accumulator register and AWB is not null

\_builtin\_mulss

Description:

Computes the product p0 × p1 . Function arguments are signed integers and the function result is a signed long integer. The command-line option, -Wconversions, can be used to detect unexpected sign conversions.

Prototype:

signed long __builtin_mulss(const signed int p0, const signed int p1);

Argument:

p0Multiplicand.

p1Multiplier.

Return Value:

Returns the signed long integer value of the product p0 × p1 .

Assembler Operator/Machine Instruction:

mul.ss

\_builtin\_mulsu

Description:

Computes the product p0 × p1 . Function arguments are integers with mixed signs and the function result is a signed long integer. The command-line option, -Wconversions, can be used to detect unexpected sign conversions. This function supports the full range of addressing modes of the instruction, including Immediate mode for operand p1 .

Prototype:

signed long __builtin_mulsu(const signed int p0, const unsigned int p1);

Argument:

p0Multiplicand.

p1Multiplier.

Return Value:

Returns the signed long integer value of the product p0 × p1 .

Assembler Operator/Machine Instruction:

mul.su

\_builtin\_mulus

Description:

Computes the product p0 × p1 . Function arguments are integers with mixed signs and the function result is a signed long integer. The command-line option, -Wconversions, can be used to detect unexpected sign conversions. This function supports the full range of addressing modes of the instruction.

Prototype:

signed long __builtin_mulus(const unsigned int p0, const signed int p1);

Argument:

p0Multiplicand.

p1Multiplier.

Return Value:

Returns the signed long integer value of the product p0 × p1 .

Assembler Operator/Machine Instruction:

mul.us

\_builtin\_muluu

Description:

Computes the product p0 × p1 . Function arguments are unsigned integers and the function result is an unsigned long integer. The command-line option, -Wconversions, can be used to detect unexpected sign conversions. This function supports the full range of addressing modes of the instruction, including Immediate mode for operand p1 .

Prototype:

unsigned long __builtin_muluu(const unsigned int p0, const unsigned int p1);

Argument:

p0Multiplicand.

p1Multiplier.

Return Value:

Returns the signed long integer value of the product p0 × p1 .

Assembler Operator/Machine Instruction:

mul.uu

\_builtin\_nop

Description:

Generates a NOP instruction.

Prototype:

void __builtin_nop(void);

Argument:

None.

Return Value:

Returns a no operation (NOP).

Assembler Operator/Machine Instruction:

NOP

\_builtin\_psvoffset

Description:

Returns the PSV page offset of the object whose address is given as a parameter. The argument p must be the address of an object in an EE data, PSV or executable memory space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 "Specifying Attributes of Variables" of the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284).

Prototype:

unsigned int __builtin_psvoffset(const void *p);

Argument:

pObject address.

Return Value:

Returns the PSV page number offset of the object whose address is given as a parameter.

Assembler Operator/Machine Instruction:

psvoffset

Error Messages:

The following error message is produced when this function is used incorrectly:

"Argument to __builtin_psvoffset() is not the address of an object in code, PSV or EE data section".

The argument must be an explicit object address.

For example, if obj is an object in an executable or read-only section, the following syntax is valid:

unsigned page = __builtin_psvoffset(&obj);

\_builtin\_psvpage

Description:

Returns the PSV page number of the object whose address is given as a parameter. The argument p must be the address of an object in an EE data, PSV or executable memory space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 “Specifying Attributes of Variables” of the “MPLAB® C Compiler for PIC24 MCUs and dsPIO® DSCs User's Guide” (DS51284).

Prototype:

unsigned int __builtin_psvpage(const void *p);

Argument:

pObject address.

Return Value:

Returns the PSV page number of the object whose address is given as a parameter.

Assembler Operator/Machine Instruction:

psvpage

Error Messages:

The following error message is produced when this function is used incorrectly:

"Argument to __builtin_psvpage() is not the address of an object in code, PSV or EE data section".

The argument must be an explicit object address.

For example, if obj is an object in an executable or read-only section, the following syntax is valid:

unsigned page = __builtin_psvpage(&obj);

\_builtin\_readsfr

Description:

Reads the SFR.

Prototype:

unsigned int __builtin_readsfr(const void *p);

Argument:

pObject address.

Return Value:

Returns the SFR.

Assembler Operator/Machine Instruction:

readsfr

\_builtin\_return\_address

Description:

Returns the return address of the current function or of one of its callers. For the level argument, a value of 0 yields the return address of the current function, a value of 1 yields the return address of the caller of the current function, and so forth. When level exceeds the current stack depth, 0 will be returned. This function should only be used with a non-zero argument for debugging purposes.

Prototype:

int __builtin_return_address (const int level);

Argument:

levelNumber of frames to scan up the call stack.

Return Value:

Returns the return address of the current function or of one of its callers.

Assembler Operator/Machine Instruction:

return_address

\_builtin\_sac

Description:

Shifts value by shift (a literal between -8 and 7) and returns the value.

For example:

register int value asm("A"); int result;

result = __builtin_sac(value, 3);

May generate:

sac A, #3, w0

Prototype:

int __builtin_sac(int value, int shift);

Argument:

valueInteger number to be shifted. shiftLiteral amount to shift.

Return Value:

Returns the shifted result to an accumulator.

Assembler Operator/Machine Instruction:

sac

Error Messages:

An error message appears if:

• The result is not an Accumulator register
• The shift value is not a literal within range

\_builtin\_sacr

Description:

Shifts value by shift (a literal between -8 and 7) and returns the value, which is rounded using the Rounding mode determined by the RND (CORCON<1>) control bit.

For example:

register int value asm("A");
int result; 

result = __builtin_sac(value, 3); 

May generate:

sac.r A, #3, w0 

Prototype:

int __builtin_sacr(int value, int shift); 

Argument:

valueInteger number to be shifted. shiftLiteral amount to shift.

Return Value:

Returns the shifted result to the CORCON register.

Assembler Operator/Machine Instruction:

sacr

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- The shift value is not a literal within range

\_builtin\_sftac

Description:

Shifts accumulator by shift. The valid shift range is -16 to 16.

For example:

register int result asm("A");
int i; 

result = __builtin_sftac(result, i); 

May generate:

sftac A, w0 

Prototype:

int __builtin_sftac(int Accum, int shift); 

Argument:

AccumAccumulator to shift.

shiftAmount to shift.

Return Value:

Returns the shifted result to an accumulator.

Assembler Operator/Machine Instruction:

sftac 

Error Messages:

An error message appears if:

• The result is not an Accumulator register
- Accum is not an Accumulator register
• The shift value is not a literal within range

\_builtin\_subab

Description:

Subtracts Accumulators A and B with the result written back to the specified accumulator. For example:

register int result asm("A");
register int B asm("B");
result = __builtin_subab(result, B); 

Will generate:

sub A 

Prototype:

int __builtin_subab(int Accum_a, int Accum_b); 

Argument:

Accum_a Accumulator from which to subtract. Accum_b Accumulator to subtract.

Return Value:

Returns the subtraction result to an accumulator.

Assembler Operator/Machine Instruction:

sub 

Error Messages:

An error message appears if the result is not an Accumulator register.

\_builtin\_tbladdress

Description:

Returns a value that represents the address of an object in program memory. The argument p must be the address of an object in an EE data, PSV or executable memory space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 "Specifying Attributes of Variables" of the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284).

Prototype:

unsigned long __builtin_tblpage(const void *p);

Argument:

pObject address.

Return Value:

Returns an unsigned long value that represents the address of an object in program memory.

Assembler Operator/Machine Instruction:

tbladdress 

Error Messages:

The following error message is produced when this function is used incorrectly:

"Argument to __builtin_tbladdress() is not the address of an object in code, PSV or EE data section".

The argument must be an explicit object address.

For example, if obj is an object in an executable or read-only section, the following syntax is valid: unsigned long page = __builtin_tbladdress(&obj);

\_builtin\_tbloffset

Description:

Returns the table page offset of the object whose address is given as a parameter. The argument p must be the address of an object in an EE data, PSV or executable memory space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 "Specifying Attributes of Variables" of the "MPLAB® C Compiler for PIC24 MCUs and dsPIC® DSCs User's Guide" (DS51284).

Prototype:

unsigned int __builtin_tboffset(const void *p);

Argument:

pObject address.

Return Value:

Returns the table page number offset of the object whose address is given as a parameter.

Assembler Operator/Machine Instruction:

tbloffset

Error Messages:

The following error message is produced when this function is used incorrectly:

"Argument to __builtin_tbloffset() is not the address of an object in code, PSV or EE data section".

The argument must be an explicit object address.

For example, if obj is an object in an executable or read-only section, the following syntax is valid:

unsigned page = __builtin_tboffset(&obj);

\_builtin\_tblpage

Description:

Returns the table page number of the object whose address is given as a parameter. The argument p must be the address of an object in an EE data, PSV or executable memory space; otherwise, an error message is produced and the compilation fails. See the space attribute in Section 2.3.1 “Specifying Attributes of Variables” of the “MPLAB® C Compiler for PIC24 MCUs and dsPIO® DSCs User's Guide” (DS51284).

Prototype:

unsigned int __builtin_tblpage(const void *p);

Argument:

pObject address.

Return Value:

Returns the table page number of the object whose address is given as a parameter.

Assembler Operator/Machine Instruction:

tblpage

Error Messages:

The following error message is produced when this function is used incorrectly:

"Argument to __builtin_tblpage() is not the address of an object in code, PSV or EE data section".

The argument must be an explicit object address.

For example, if obj is an object in an executable or read-only section, the following syntax is valid:

unsigned page = __builtin_tblpage(&obj);

\_builtin\_tblrdh

Description:

Issues the TBLRDH.W instruction to read a word from Flash or EE data memory. You must set up the TBLPAG to point to the appropriate page. To do this, you may make use of: __builtin_tbloffset() and __builtin_tblpage().

Please refer to the specific device data sheet or the appropriate family reference manual for complete details regarding reading and writing program Flash.

Prototype:

unsigned int __builtin_tblrdh(unsigned int offset);

Argument:

offsetDesired memory offset.

Return Value:

None.

Assembler Operator/Machine Instruction:

tblrdh

Error Messages:

None.

\_builtin\_tblrdl

Description:

Issues the TBLRDL.W instruction to read a word from Flash or EE data memory. You must set up the TBLPAG to point to the appropriate page. To do this, you may make use of:

__builtin_tbloffset() and__builtin_tblpage().

Please refer to the specific device data sheet or the appropriate family reference manual for complete details regarding reading and writing program Flash.

Prototype:

unsigned int __builtin_tblrdl(unsigned int offset);

Argument:

offsetDesired memory offset.

Return Value:

None.

Assembler Operator/Machine Instruction:

tblrdl

Error Messages:

None.

\_builtin\_tblwth

Description:

Issues the TBLWTH.W instruction to write a word to Flash or EE data memory. You must set up the TBLPAG to point to the appropriate page. To do this, you may make use of:

_builtin_tboffset() and __builtin_tblpage().

Please refer to the specific device data sheet or the appropriate family reference manual for complete details regarding reading and writing program Flash.

Prototype:

void __builtin_tblwth(unsigned int offset

unsigned int data);

Argument:

offsetDesired memory offset.

dataData to be written.

Return Value:

None.

Assembler Operator/Machine Instruction:

tblwth

Error Messages:

None.

\_builtin\_tblwt1

Description:

Issues the TBLRDL.W instruction to write a word to Flash or EE data memory. You must set up the TBLPAG to point to the appropriate page. To do this, you may make use of: __builtin_tbloffset() and __builtin_tblpage().

Please refer to the specific device data sheet or the appropriate family reference manual for complete details regarding reading and writing program Flash.

Prototype:

void __builtin_tblwtl(unsigned int offset unsigned int data);

Argument:

offsetDesired memory offset. dataData to be written.

Return Value:

None.

Assembler Operator/Machine Instruction:

tblwt1

Error Messages:

None.

Example 6-1: Additional In-Line Functions

#include "p33fxxxx.h"

volatile long Result_mpy1616;
volatile long Result_addab;
volatile long Result_subab;
volatile long Result_mpy3216;
volatile long Result_div3216;

register int Accu_A asm("A");
register int Accu_B asm("B");

inline static long mpy_32_16 (long, int);

inline static long mpy_32_16 (long x, int y)
{
    long result;
    int templ, temp2;
    temp1 = (x>>1)&0x7FFF;
    temp2 = x>>16;
    Accu_A = __builtin_mpy (temp1, y, 0,0,0,0,0,0);
    Accu_A = __builtin_sftac (15);
    Accu_A = __builtin_mac (temp2, y, 0,0,0,0,0,0,0);
    asm("mov _ACCAL,%0\n\t"
    "mov _ACCAH,%d0" : "=r"(result) : "w"(Accu_A));
    return result;
}

int main (void)
{
    // Variable declarations
    int Input1;
    int Input2;
    int Input3;
    int Input4;
    long Input5;
    int Input6;
    long Input7;
    int Input8;

    // Enable 32-bit saturation, signed and fractional modes for both ACCA and ACCB
    CORCON = 0x00C0;

    // Example of 16*16-bit fractional multiplication using ACCA
    Input1 = 32767;
    Input2 = 32767;
    Accu_A = __builtin_mpy (Input1, Input2, 0,0,0,0,0,0);
    asm("mov _ACCAL,%0\n\t"
    "mov _ACCAH,%d0" : "=r"(Result_mpy1616) : "w"(Accu_A));

    // Example of 16*16-bit fractional multiplication using ACCB
    Input3 = 16384;
    Input4 = 16384;
    Accu_B = __builtin_mpy (Input3, Input4, 0,0,0,0,0,0);
    asm("mov _ACCBL,%0\n\t"
    "mov _ACCBH,%d0" : "=r"(Result_mpy1616) : "w"(Accu_B));

    // Example of 32-bit addition using ACCA (ACCA = ACCA + ACCB)
    Accu_A = __builtin_addab();
    asm("mov _ACCAL,%0\n\t"
    "mov _ACCAH,%d0" : "=r"(Result_addab) : "w"(Accu_A));

    // Example of 32-bit subtraction using ACCB (ACCB = ACCB - ACCA)
    Accu_B = __builtin_subab();
    asm("mov _ACCBL,%0\n\t"
    "mov _ACCBH,%d0" : "=r"(Result_subab) : "w"(Accu_B));

    // Example of 32*16-bit fractional multiplication using ACCA
    Input5 = 0x7FFFFFFF;
    Input6 = 32767;
    Result_mpy3216 = mpy_32_16 (Input5, Input6);

    while(1);
} 

Example 6-2: Divide_32_by_16

#include <p33Fxxxx.h>
#include "divide.h"

_FOSCSEL(FNOSC_FRC);
_FOSC(FCKSM_CSDCMD & OSCIOFNC_OFF & POSCMD_NONE);
_FWDT(FWDTEN_OFF);

unsigned int divide_(long a, int b) {
    union convert {
    unsigned long 1;
    unsigned int i[2];
    } c;

    int sign;
    unsigned int result;

    c.l = a;
    sign = c.i[1] ^ b;

    if (a < 0) a = (-a);
    if (b < 0) b = -b;
    result = __builtin_divud(a,b);
    result >>= 1;
    if (sign < 0) result = -result;
    return result;
}

int main(void)
{
    unsigned long dividend;
    unsigned int divisor;
    unsigned int quotient;

    dividend = 0x3FFFFFFF;
    divisor = 0x7FFF;

    quotient = divide_(long)dividend, (int)divisor);
    while(1);
} 

NOTES:

Section 7. Reference

HIGHLIGHTS

This section of the manual contains the following major topics:

7.1 Instruction Bit Map 498
7.2 Instruction Set Summary Table 501
7.3 Revision History 511

7.1 INSTRUCTION BIT MAP

Instruction encoding for the 16-bit MCU and DSC family devices is summarized in Table 7-1. This table contains the encoding for the MSB of each instruction. The first column in the table represents bits<23:20> of the opcode and the first row of the table represents bits 19:16 of the opcode. The first byte of the opcode is formed by taking the first column bit value and appending the first row bit value. For instance, the MSB of the PUSH instruction (last row, ninth column) is encoded with '11111000b' (0xF8).

Note: The complete opcode for each instruction may be determined by the instruction descriptions in Section 5. "Instruction Descriptions", using Table5-1 through Table5-15.

Table 7-1: Instruction Encoding

Opcode<19:16>

0000

0001

0010

0011

01

00

0101

0110

0111

1000

10

01

1010

1011

1100

1101

1110

11

11

Opcode<23:20>

0000

NOP

BRA

CALL

LDSLV(4)VFSLV(4)

GOTO

RETLW

RETFIERETURN

RCALL

DO(1)

REPEAT

BFEXT(4)BFINS(4)

—

BRA(1)(OA)

BRA(1)(OB)

BRA(1)(SA)

BRA(1)(SB)

0001

SUBR

SUBBR

0010

MOV

0011

BRA(OV)

BRA(C)

BRA(Z)

BRA(N)

BRA(LE)

BRA(LE)

BRA(LT)

BRA(LEU)

BRA

BRA(NOV)

BRA(NC)

BRA(NZ)

BRA(NN)

BRA(GT)

BRA(GE)

BRA(GTU)

—

0100

ADD

ADDC

0101

SUB

SUBB

0110

AND

XOR

0111

IOR

MOV

1000

MOV

1001

MOV

1010

BSET

BCLR

BTG

BTST

BTSTS

BTST

BTSS

BTSC

BSET

BCLR

BTG

BTST

BTSTS

BSW

BTSS

BTSC

1011

ADDADDC

SUBSUBB

ANDXOR

IORMOV

ADDADDC

SUBSUBB

ANDXOR

IORMOV

MUL.USMUL.UU

MUL.SSMUL.SU

TBLRDHTBLRDL

TBLWTHTBLWTL

MUL

SUBSUBB

MOV.D

MOV

1100

MAC(1)MPY(1)MPYN(1)MSC(1)

CLRAC(1)

MAC(1)MPY(1)MPYN(1)MSC(1)

MOVSAC(1)

SFTAC(1)

ADD(1)

LAC(1)LAC.D(4)

ADD(1)NEG(1)SUB(1)

SAC(1)SAC.D(4)

SAC.R(1)

MAX(4)MAX.V(4)MIN(4)MIN.V(4)MINZ(4)MINZ.V(4)NORM(4)

FF1LFF1R

1101

SL

ASRLSR

RLCRLNC

RRCRRNC

SL

ASRLSR

RLCRLNC

RRCRRNC

DIV.SDIV.UDIV2.S(4)DIV2.U(4)

DIVF(1)DIVF2(4)

—

—

—

SL

ASRLSR

FBCL

Note 1: This instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C family devices.
2: This instruction is only available in PIC24E and dsPIC33E family devices.
3: This instruction is only available in dsPIC33C and some dsPIC33E family devices.
4: This instruction is only available in dsPIC33C family devices.
5: This instruction is only available in some dsPIC33C, dsPIC33E and PIC24F family devices.

Table 7-1: Instruction Encoding (Continued)

Opcode<19:16>

0000

0001

0010

0011

0100

0101

0110

0111

1000

1001

1010

1011

1100

1101

1110

1111

Opcode<23:20>

1110

CPU CP

CPB

CPU CP

CPB

FLIM^(4) FLIM.V^(4)

CPBGT^(2) CPBLT^(2) CPSGT CPSLT

CPBEQ^(2) CPBNE^(2) CPSEQ CPSNE

INC INC2

DEC DEC2

COM NEG

CLR SETM

INC INC2

DEC DEC2

COM NEG

CLR SETM

1111

ED

(1) EDAC^(1) MAC^(1) MPY^(1)

———

P

U

S

H

P

OULNK

SE PZE

DISI DAW

EXCH SWAP

BOOTSWP^(5)K CLRWDT CTXTSWP(3)MOVPAG(2)PWRSAVPOP.S PUSH.S RESET

NOPR

Note 1: This instruction is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C family devices.
2: This instruction is only available in PIC24E and dsPIC33E family devices.
3: This instruction is only available in dsPIC33C and some dsPIC33E family devices.
4: This instruction is only available in dsPIC33C family devices.
5: This instruction is only available in some dsPIC33C, dsPIC33E and PIC24F family devices.

7.2 INSTRUCTION SET SUMMARY TABLE

The complete 16-bit MCU and DSC device instruction set is summarized in Table 7-2. This table contains an alphabetized listing of the instruction set. It includes instruction assembly syntax, description, size (in 24-bit words), execution time (in instruction cycles), affected Status bits and the page number in which the detailed description can be found. Table 1-2 identifies the symbols that are used in the Instruction Set Summary Table.

Note: The instruction cycle counts listed here are for PIC24F, PIC24H, dsPIC30F and dsPIC33F devices. Some instructions require additional cycles in PIC24E and dsPIC33E devices. Refer to Section 3.3 "Instruction Set Summary Tables" and Section 5.4 "Instruction Descriptions" for details.

Table 7-2: Instruction Set Summary Table

Assembly Syntax Mnemonic, Operands		Description Words Cycles OA			(2)	OB(2)	SA(1,2)	SB(1,2)	OAB(2)	SAB(1,2)	DC N	OV Z	C			Page Number
ADD	f {,XREG}	Destination = f + WREG	1	1	—	—	—	—	—	—						102
ADD	#Lit10,Wn	Wn = lit10 + Wn	1	1	—	—	—	—	—	—						103
ADD	Wb,#Lit5,Wd	Wd = Wb + lit5	1	1	—	—	—	—	—	—						105
ADD	Wb,Ws,Wd	Wd = Wb + Ws	1	1	—	—	—	—	—	—						105
ADD	Acc(2)	Add Accumulators	1	1							—	—	—	—	—	107
ADD	Ws,#Lit4,Acc	16-Bit Signed Add to Accumulator	1	1							—	—	—	—	—	108
ADDC	f {,XREG}	Destination = f + WREG + (C)	1	1	—	—	—	—	—	—						110
ADDC	#Lit10,Wn	Wn = lit10 + Wn + (C)	1	1	—	—	—	—	—	—						111
ADDC	Wb,#Lit5,Wd	Wd = Wb + lit5 + (C)	1	1	—	—	—	—	—	—						112
ADDC	Wb,Ws,Wd	Wd = Wb + Ws + (C)	1	1	—	—	—	—	—	—						114
AND	f {,XREG}	Destination = f.AND.WREG	1	1	—	—	—	—	—	—	—		—		—	116
AND	#Lit10,Wn	Wn = lit10.AND.Wn	1	1	—	—	—	—	—	—	—		—		—	117
AND	Wb,#Lit5,Wd	Wd = Wb.AND.It5	1	1	—	—	—	—	—	—	—		—		—	118
AND	Wb,Ws,Wd	Wd = Wb.AND.Ws	1	1	—	—	—	—	—	—	—		—		—	119
ASR	f {,XREG}	Destination = Arithmetic Right Shift f,LSb→C	1	1	—	—	—	—	—	—	—		—			121
ASR	Ws,Wd	Wd = Arithmetic Right Shift Ws,LSb→C	1	1	—	—	—	—	—	—	—		—			123
ASR	Wb,#Lit4,And	Wnd = Arithmetic Right Shift Wb by lit4,LSb→C	1	1	—	—	—	—	—	—	—		—		—	125
ASR	Wb,Ws,Wnd	Wnd = Arithmetic Right Shift Wb by Wns,LSb→C	1	1	—	—	—	—	—	—	—		—		—	126
BCLR	f,#Lit4	Bit Clear f	1	1	—	—	—	—	—	—	—	—	—	—	—	127

Legend: ♂ set or cleared; 0 may be cleared, but never set; ↑ may be set, but never cleared; 1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Reference

↓

Table 7-2: Instruction Set Summary Table (Continued)

Legend: ♂ set or cleared; ♦ may be cleared, but never set; ♦ may be set, but never cleared; 1' always set; 0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description								Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z	C	Page Number
DSET Ws,#bit4	B	i	t	S	e	t	i	n	W	s	1	1	—	—	—	—	—	—	—	—	—	16
DSN Ws,Nb	B	i	t	W	r	i	t	e	i	n	W	s	<	W	b	>	1	1	—	—	—	—
RTG f,#bit4	Bit Toggle in f								1	1	—	—	—	—	—	—	—	—	—	—	—	165
RTG Ws,#bit4	Bit Toggle in Ws								1	1	—	—	—	—	—	—	—	—	—	—	—	166
ETSC f,#bit4	Bit Test f, Skip if Clear								1	1(2 or 3)	—		—	—		—	—		—		—	—
ETSC Ws,#bit4	Bit Test Ws, Skip if Clear								1	1(2 or 3)	—		—	—		—	—		—		—	—
MTSS f,#bit4	Bit Test f, Skip if Set								1	1(2 or 3)	—		—	—		—	—		—		—	—
RTSS Ws,#bit4	Bit Test Ws, Skip if Set								1	1(2 or 3)	—		—	—		—	—		—		—	—
BTST f,#bit4	B	i	t	Te	s	t	i	n	I						1	1	—	—	—	—	-8	—
BTST Ws,#bit4	B	i	t	Te	s	t	i	n	W	s						1	1	—	—	—	—	176
DTST Ws,Nb	B	i	t	Te	s	t	i	n	W	s						1	1	—	—	—	3	—
DTSTS t,#bit4	Bit Test/Set in f								1	1	—	—	—	—	—	—	—	—	—	3	—	180
BTSTS Ws,#bit4	B	i	t	Te	s	t	/	S	e	t	i n	W	s					1	1	—	—	6
CALL Expr (5)	C	a	l	l	S	u	b	r	o	u	t i n	e						2	2	—	—	—
CALL Expr (3)	C	a	l	l	S	u	b	r	o	u	t i n	e						2	2	—	—	—
CALL Kn (5)	C	a	l	l	l	n	d	i	r	e	p t	S	u	b r o	u	t i n	e					1
CALL Kn (3)	C	a	l	l	l	n	d	i	r	e	p t	S	u	b r o	u	t i n	e					1
CALL.1 Nb(5)	Call Indirect Subroutine Long (long address)								1	4	—	—	—	—	—	—	—	—	—	—	—	191
CLR f,NRSG	C	l	e	a	r	f	o	r	W	R	E	G					1	1	—	—	—	—
CLR Ws	C	l	e	a	r	W	d							1	1	—	—	—	—	—	—	—
CLR Acc,[Wx],Nxd,[Ny],Kyd,AWE (2)	Clear Accumulator								1	1	0	0	0			0	0	0—	—	—	—	19
CLRMDT	C	l	e	a	r	W	a	t	c	h	d o g		T	i m e r						1	1	—
CCM f{,NRAKS}	Destination =f								1	1	—	—	—	—	—	—	—	0	—	3	—	197
CCM Ws,Nd	Wd = Ws								1	1	—	—	—	—	—	—	—	0	—	3	—	198
CP f	Compare (f-WREG)								1	1	—	—	—	—	—	—	0	0	0	0	3	200
CP Nb,¥:1£(5)	Compare (Wb - lit5)								1	1	—	—	—	—	—	—	0	0	0	0	3	201
CP Nb,¥:1£(5)	Compare (Wb - lit8)								1	1	—	—	—	—	—	—	0	0	0	0	3	202

Legend: ♂ set or cleared; ♀ may be cleared, but never set; ↑ may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24F, dsPIC33F and dsPIC33C devices.
4: This instructionOperand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description	Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z C		Page Number
CP Wb, Ws	Compare (Wb - Ws) 1 1 — — — — —									3	6	0	3	6	203
CP0 f	Compare (f - 0x0) 1 1 — — — — —									1	6	0	3	1	204
CP0 Ws	Compare (Ws - 0x0) 1 1 — — — — —									1	6	0	3	1	205
CPB f	Compare with Borrow (f - WREG - )	1	1	—	—	—	—	—	—	3	6	0	3	6	206
CPB Wb, #llt5 (5)	Compare with Borrow (Wb - lit5 - )	1	1	—	—	—	—	—	—	3	6	0	3	6	207
CPB Wb, #llt8 (3)	Compare with Borrow (Wb - lit8 - )	1	1	—	—	—	—	—	—	3	6	0	3	6	208
CPB Wb, Ns	Compare with Borrow (Wb - Ws - )	1	1	—	—	—	—	—	—	3	6	0	3	6	209
CPSEQ Wb, Wn, Expr(3)	Compare Wb with Wn, Branch if =	1	1 (5)	—	—	—	—	—	—	—	—	—	—	—	211
CPSGT Wb, Wn, Expr(3)	Signed Compare Wb with Wn, Branch if >	1	1 (5)	—	—	—	—	—	—	—	—	—	—	—	212
CPSLT Wb, Wn, Expr(3)	Signed Compare Wb with Wn, Branch if <	1	1 (5)	—	—	—	—	—	—	—	—	—	—	—	213
CPSNE Wb, Wn, Expr(3)	Compare Wb with Wn, Branch if≠	1	1 (5)	—	—	—	—	—	—	—	—	—	—	—	214
CPSEQ Wb, Wn(5)	Compare (Wb with Wn), Skip if =	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	215
CPSEQ Wb, Wn(3)	Compare (Wb with Wn), Skip if =	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	216
CPSGT Wb, Wn(5)	Signed Compare (Wb with Wn), Skip if >	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	217
CPSGT Wb, Wn(3)	Signed Compare (Wb with Wn), Skip if >	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	218
CPSLT Wb, Wn(5)	Signed Compare (Wb with Wn), Skip if <	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	219
CPSLT Wb, Wn(3)	Signed Compare (Wb with Wn), Skip if <	1 1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	—	220
CPSNE Wb, Wn(5)	Signed Compare (Wb with Wn), Skip if ≠	1	1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	221
CPSNE Wb, Wn(3)	Signed Compare (Wb with Wn), Skip if =	1	1(2 or 3)	—	—	—	—	—	—	—	—	—	—	—	222
CTXTGKP #llt5(8)	CPU Register Context Swap Literal	1	2	—	—	—	—	—	—	—	—	—	—	—	223
CTXTEWP Wc(6)	CPU Register Context Swap Wn	1	2	—	—	—	—	—	—	—	—	—	—	—	224
DAN.B Wn	Wn = Decimal Adjust Wn	1	1	—	—	—	—	—	—	—	—	—	—	6	225
DEC f ( ,WREG )	Destination = f - 1 1 1 — — — — —									3	6	0	3	6	226
DEC Ws, Nd	Wd = Ws - 1	1 1— — — — — —								3	6	0	3	6	227

Legend: set or cleared; may be cleared, but never set; may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3. This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description	Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z	C	Page Number
DEC2 f {,WREG}	Destination = f - 2 1 1 — — — — —									0	0	0	0	0	229
DEC2 Ws,Wd	Wd = Ws - 2 1 1 — — — — —									0	0	0	0	0	230
DIS1 #lit14	Disable Interrupts for lit14 Instruction Cycles	1	1	—	—	—	—	—	—	—	—	—	—	—	232
DIV.S Wm,Wn	Signed 16/16-Bit Integer Divide	1	18	—	—	—	—	—	—	—	0	0	0	0	233
DIV.U Wm,Wn	Unsigned 16/16-Bit Integer Divide	1	18	—	—	—	—	—	—	—	0	0	0	0	235
DIVF Wm,Wn (2)	Signed 16/16-Bit Fractional Divide	1	18	—	—	—	—	—	—	—	0	0	0	0	236
DIVF2 Wm,Wn(7)	Signed 16/16-Bit Fractional Divide (W1:W0 preserved)	1	6	—	—	—	—	—	—	—	0	0	0	0	238
DIVZ.G Wm,Wn(7)	Signed 16/16-Bit Integer Divide (W1:W0 preserved)	1	6	—	—	—	—	—	—	—	0	0	0	0	240
DIVZ.U Wm,Wn(7)	Unsigned 16/16-Bit Integer Divide (W1:W0 preserved)	1 6 — — — — —			—						0	0	0	0	241
DO #lit14,Expr(6)	Do Code to PC + Expr, (lit14 + 1) Times	2	2	—	—	—	—	—	—	—	—	—	—	—	242
DO #lit15,Expr(4)	Do Code to PC + Expr, (lit15 + 1) Times	2	2	—	—	—	—	—	—	—	—	—	—	—	244
DO Wn,Expr(6)	Do Code to PC + Expr, (Wn + 1) Times	2	2	—	—	—	—	—	—	—	—	—	—	—	246
DO Wn,Expr(4)	Do Code to PC + Expr, (Wn + 1) Times	2	2	—	—	—	—	—	—	—	—	—	—	—	248
ED Wm*Wn,Acc, [Wx], [Ky], Kxd(2)	Euclidean Distance (no accumulate)	1 1		0	0	0	0	0	0	—	—	—	—	—	250
EDAC Wm*Wm,Acc, [Wx], [Wy], Wxd (2)	Euclidean Distance	1 1		0	0	0	0	0	0	—	—	—	—	—	252
EXCE Wns,Wnd	Swap Wns and Wnd	1	1	—	—	—	—	—	—	—	—	—	—	—	254
FDCL Ws,Wnd	Find First Bit Change from Left (MSb) Side	1	1	—	—	—	—	—	—	—	—	—	—	0	255
FTIL Ws,Wnd	Find First One from Left (MSb) Side	1	1	—	—	—	—	—	—	—	—	—	—	0	257
FF1R Ws,Wnd	Find First One from Right (LSb) Side	1	1	—	—	—	—	—	—	—	—	—	—	0	259
FLIM Wb,Ws (7)	Force (signed) Data Range Limit	1	1	—	—	—	—	—	—	—	0	0	0	—	251
FLIM.V Nb,Ws,Wnd(7)	Force (signed) Data Range Limit with Limit Excess Result	1	1	—	—	—	—	—	—	—	0	0	0	—	252
GCTO Expr	Unconditional Jump	2	2	—	—	—	—	—	—	—	—	—	—	—	263
GCTO Kn (5)	Unconditional Indirect Jump	1	2	—	—	—	—	—	—	—	—	—	—	—	264
GCTO Wn (3)	Unconditional Indirect Jump	1	2	—	—	—	—	—	—	—	—	—	—	—	265
GCTO.L Wn(3)	Unconditional Indirect Jump Long	1	4	—	—	—	—	—	—	—	—	—	—	—	266
TNC f {,WREG}	Destination = f + 1	1 1 — — — — —								0	0	0	0	0	267
TNC Ws,Wnd	Wd = Ws + 1 1 1 — — — — —									0	0	0	0	0	268
INC2 f {,WREG}	Destination = f + 2	1 1 — — — — —								0	0	0	0	0	269
INC2 Ws,Wnd	Wd = Ws + 2 1 1 — — — — —									0	0	0	0	0	270
ICR f {,WREG}	Destination = f.IOR.WREG	1	1	—	—	—	—	—	—	—	0	—	0	—	271

Legend: ♂ set or cleared; ↓ may be cleared, but never set; ↑ may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Reference

↓

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description	Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z C		Page Number
IOR flit10, Wn	Wn = lit10 .IOR. Wn 1 1 — — — — — —										∅	—	∅	—	272
IOR Wb, flit5, Wnd	Wd = Wb .IOR. flit5 1 1 — — — — — —										∅	—	∅	—	273
IOR Wb, Ws, Wd	Wd = Wb .IOR. Ws	1 1 — — — — —									∅	—	∅	—	274
IAC Ws, (#slit4,), Acc (2)	Load Accumulator	1 1		∅	∅	∅	∅	∅	∅	—	—	—	—	—	276
IAC.D Ws, [, #slit4], Acc(7)	Load Accumulator Double Word	1 2		∅	∅	∅	∅	∅	∅	—	—	—	—	—	278
LOSLV [Wns], [Wnd++, flit2(7)	Move Single Instruction Word from Master to Slave PRAM	1	2	—	—	—	—	—	—	—	—	—	—	—	279
LNX flit14 (5)	Link Frame Pointer	1	1	—	—	—	—	—	—	—	—	—	—	—	280
LNX flit14 (3)	Link Frame Pointer	1	1	—	—	—	—	—	—	—	—	—	—	—	281
LSR f ( ,WREG)	Destination = Logical Right Shift f, MSb → C	1	1	—	—	—	—	—	—	—	∅	—	∅	∅	282
LSR Ws, Wd	Wd = Logical Right Shift Ws, MSb → C	1	1	—	—	—	—	—	—	—	∅	—	∅	∅	284
LSR Wb, flit4,Wnd	Wnd = Logical Right Shift Wb by liM, MSb → C	1	1	—	—	—	—	—	—	—	∅	—	∅	—	286
LSR Wb, Wns, Wnd	Wnd = Logical Right Shift Wb by Wns, MSb → C	1	1	—	—	—	—	—	—	—	∅	—	∅	—	287
MAC Wm*Nw, Acc, [Wx], Nxd, [Wy], Wyd, Acc(2)	Multiply and Accumulate	1 1		∅	∅	∅	∅	∅	∅	—	—	—	—	—	288
MAC Wm*Nw, Acc, [Wx], Nxd, [Wy], Wyd (2)	Square and Accumulate	1 1		∅	∅	∅	∅	∅	∅	—	—	—	—	—	290
MAX Acc (7)	Force Accumulator Maximum Data Range Limit	1	1	—	—	—	—	—	—	∅	∅	∅	∅	∅	292
MAX.V Acc, Wd(7)	Force Accumulator Maximum Data Range Limit and Store Limit Excess Result	1	1	—	—	—	—	—	—	∅	∅	∅	∅	∅	293
MIN Acc (7)	Force Accumulator Minimum Data Range Limit	1 1 — — — —								∅	∅	∅	∅	∅	294
MIN.V Acc, Wd(7)	Force Accumulator Minimum Data Range Limit and Store Limit Excess Result	1	1	—	—	—	—	—	—	∅	∅	∅	∅	∅	295
MTNZ Acc (7)	Conditionally Force Accumulator Minimum Data Range Limit if Z Flag is Set	1	1	—	—	—	—	—	—	∅	∅	∅	∅	∅	296
MTNZ.V Acc, Wd(7)	Conditionally Force Accumulator Minimum Data Range Limit and Store Limit Excess Result if Z Flag is Set	1	1	—	—	—	—	—	—	∅	∅	∅	∅	∅	297
MCV f ( ,WREG)	Move f to Destination	1 1 — — — —									∅	—	∅	—	299
MCV WRREG, f	Move WREG to f	1	1	—	—	—	—	—	—	—	—	—	—	—	300
MCV f, And	Move f to Wnd	1	1	—	—	—	—	—	—	—	—	—	—	—	301
MCV Wns, f	Move Wns to f	1	1	—	—	—	—	—	—	—	—	—	—	—	302
MCV.B flit8,Wnd	Move 8-Bit Unsigned Literal to Wnd	1	1	—	—	—	—	—	—	—	—	—	—	—	303
MCV #lt15, Wnd	Move 16-Bit Literal to Wnd	1	1	—	—	—	—	—	—	—	—	—	—	—	304

Legend: ♂ set or cleared; ♂ may be cleared, but never set; ♂ may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description	Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z	C	Page Number
MCV [Ws-Slit10], Wnd	M o v e [ W s +	S I	i	t	1	0	]	t	o	W	n	d	1	1	—	—	—
MCV Wns, [Ws-Slit10]	M o v e W n s t	o [	W	d	+	S	i	t	1	0	]	1	1	—	—	—	—
MCV Ws, Nd	M o v e W s t o	W d	1	1	—	—	—	—	—	—	—	—	—	—	—	—	—
MCV.D Wns, Wnd	M o v e D o u b	I e	W	n	s	;	W	n	s	+ 1	t	o	W	n	d
MCVDAC #Lit10, DSRPAG(3)	Move 10-Bit Literal to DSRPAG	1	1	—	—	—	—	—	—	—	—	—	—	—	—	311
MCVDAC Ws, DSRPAG(3)	Move Wn to DSRPAG	1	1	—	—	—	—	—	—	—	—	—	—	—	—	312
MCVSAC Acc, [Nx], Kxd, [Wy], Wyd, AMB(2)	Move [Wx] to Wxd and [Wy] to Wyd	1	1	—	—	—	—	—	—	—	—	—	—	—	—	313
MPY Wm*Nm, Acc, [Wx], Kxd, [Wy], Wyd (2)	Multiply Wm by Wn to Accumulator	1	1	0	0	0	↑		↑	0	0	0	—	—	—	—	315
MPY Wm*Nm, Acc, [Wx], Kxd, [Wy], Wyd (2)	Square to Accumulator	1	1	0	0	0	↑		↑	0	0	0	—	—	—	—	317
MPY.N Wm*Nm, Acc, [Kx], Kxd, [Wy], Wyd (2)	-(Multiply Wn by Wm) to Accumulator	1	1	0	0	—		—	0	—		—	—	—	—	—	—
MSC Wm*Nm, Acc, [Wx], Kxd, [Wy], Wyd, AMB(2)	Multiply and Subtract from Accumulator	1	1	0	0	0		0	0	0	0	—	—	—	—	—	321
MJL f	W 3 : W 2 = f *	W R E G							1	1	—	—	—	—	—	—
MJL.SS Wb, Ws, Wnd	{Wnd + 1,Wnd} = Signed(Wb) * Signed(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	325
MJL.SS Wb, Ws, Acc(4)	Accumulator = Signed(Wb) * Signed(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	327
MJL.SU Wb, #Lit5, Wnd	{Wnd + 1,Wnd} = Signed(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	328
MJL.SU Wb, Ws, Wnd	{Wnd + 1,Wnd} = Signed(Wb) * Unsigned(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	329
MJL.SU Wb, Ws, Acc(4)	Accumulator = Signed(Wb) * Unsigned(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	331
MJL.SU Wb, #Lit5, Acc(4)	Accumulator = Signed(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	332
MJL.US Wb, Ws, Wnd	{Wnd + 1,Wnd} = Unsigned(Wb) * Signed(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	333
MJL.US Wb, Ws, Acc(4)	Accumulator = Unsigned(Wb) * Signed(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	335
MJL.UU Wb, #Lit5, Knd	{Wnd + 1,Wnd} = Unsigned(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	336
MJL.UU Wb, Ws, Wnd	{Wnd + 1,Wnd} = Unsigned(Wb) * Unsigned(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	337
MJL.UU Wb, Ws, Acc(4)	Accumulator = Unsigned(Wb) * Unsigned(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	339
MJL.UU Wb, #Lit5, Acc(4)	Accumulator = Unsigned(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	340
MJL.W.SS Wb, Ws, Wnd(2)	W n d = S i g n e d (W b ) * S i g n e d (W s )
MJL.W.SU Wb, Ws, Wnd(2)	Wnd = Signed(Wb) * Unsigned(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	343
MJL.W.SU Wb, #Lit5, Wnd(3)	Wnd = Signed(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	345
MJL.W.S U Bw, Ws, Wnd(2)	Wnd = Unsigned(Wb) * Signed(Ws)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	346
MJL.W.UU Wb, Ws, Wnd(2)	W n d = U n s i g n e d (W b ) * U n s i g n e d (W s )
MJL.W.UU Wb, #Lit5, Wnd(3)	Wnd = Unsigned(Wb) * Unsigned(lit5)	1	1	—	—	—	—	—	—	—	—	—	—	—	—	349

Legend: ♂ set or cleared; ↓ may be cleared, but never set; ↑ may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details)

Table 7-2: Instruction Set Summary Table (Continued)

Legend: ♂ set or cleared; ♀ may be cleared, but never set; ↑ may be set, but never cleared; 1' always set; ∇' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Legend: ♂ set or cleared; ↓ may be cleared, but never set; ↑ may be set, but never cleared; '1' always set; '0' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

Table 7-2: Instruction Set Summary Table (Continued)

Assembly Syntax Mnemonic, Operands	Description	Words	Cycles	OA^(2)	OB^(2)	SA^(1,2)	SB^(1,2)	OAB^(2)	SAB^(1,2)	DC	N	OV	Z	C	Page Number
SUBR Nb,flitb,Ws	Wd = lit5 - Wb - (C)	1					1			0	0-	0	3	0	431
SUBR Nb,Ws,Wd	Wd = Ws - Wb - (C)	1					1			3	6-	0	3	0	433
SUBR f [ ,WRSG]	Destination = WREG -f	1	1	—	—	—	—	—	—	0	0	0	3	0	435
SUBR Nb,flitb,Wd	Wd = lit5 - Wb	1	1	—	—	—	—	—	—	0	0	0	3	0	436
SUBR Nb,Ws,Wd	Wd = Ws - Wb	1	1	—	—	—	—	—	—	0	0	0	3	0	437
SNAP Wn	Wn = Byte or Nibble Swap Wn	1	1	—	—	—	—	—	—	—	—	—	—	—	439
TBLRDH [Ws],Wd	Read High Program Word to Wd	1	2	—	—	—	—	—	—	—	—	—	—	—	440
TBLRDU [Ws],Wd	Read Low Program Word to Wd	1	2	—	—	—	—	—	—	—	—	—	—	—	442
TBLWH Ws,[Xd]	Write Ws to High Program Word	1	2	—	—	—	—	—	—	—	—	—	—	—	444
TBLWTL Ws,[Wd]	Write Ws to Low Program Word	1	2	—	—	—	—	—	—	—	—	—	—	—	446
ULNK(5)	Deallocate Stack Frame	1	1	—	—	—	—	—	—	—	—	—	—	—	448
UTNK(3)	Deallocate Stack Frame	1	1	—	—	—	—	—	—	—	—	—	—	—	449
VPSLV Kns,Knd,flit2(7)	Verify Slave Processor Program RAM	1	1	—	—	—	—	—	—	—	—	—	—	—	450
XCR f [ ,WREG]	Destination = f .XOR. WREG	1	1	—	—	—	—	—	—	—	0	—	3	—	451
XCR #lit10,Wn	Wn = lit10 .XOR. Wn	1	1	—	—	—	—	—	—	—	0	—	3	—	452
XCR Nb,flit5,Wd	Wd = Wb .XOR. lit5	1	1	—	—	—	—	—	—	—	0	—	3	—	453
XCR Nb,Ws,Wd	Wd = Wb .XOR. Ws	1	1	—	—	—	—	—	—	—	0	—	3	—	454
ZK Ws,Knd	Wnd = Zero-Extended Ws	1	1	—	—	—	—	—	—	—	0	—	3	1	456

Legend: ♂ set or cleared; ♀ may be cleared, but never set; ♀ may be set, but never cleared; 1' always set; 2' always cleared; — unchanged
Note 1: SA, SB and SAB are only modified if the corresponding saturation is enabled, otherwise unchanged.
2: This instruction/operand is only available in dsPIC30F, dsPIC33F, dsPIC33E and dsPIC33C devices.
3: This instruction/operand is only available in PIC24E, dsPIC33E and dsPIC33C devices.
4: This instruction/operand is only available in dsPIC33E and dsPIC33C devices.
5: This instruction/operand is only available in PIC24F, PIC24H, dsPIC30F and dsPIC33F devices.
6: This instruction/operand is only available in dsPIC30F and dsPIC33F devices.
7: These instructions are only available in dsPIC33C devices.
8: These instructions are only available in all dsPIC33C devices and some dsPIC33E devices (see device data sheet for details).
9: These instructions are only available in all dsPIC33C devices, and some PIC24F and dsPIC33E devices (see device data sheet for details).

7.3 REVISION HISTORY

Revision A (May 2005)

This is the initial release of this document.

Revision B (September 2005)

This revision incorporates all known errata at the time of this document update.

Revision C (February 2008)

This revision includes the following corrections and updates:

- Instruction Updates:

• Updated BRA Instruction (see "BRA")
• Updated DIVF Instruction (see "DIVF")
• Updated DO Instruction (see "DO")
• Updated SUB instruction (see "SUB")

Revision D (November 2009)

This revision includes the following corrections and updates:

• Document renamed from dsPIC30F/33F Programmer's Reference Manual to 16-bit MCU and DSC Programmer's Reference Manual
• Document has been completely redesigned to accommodate all current 16-bit families: dsPIC30F, dsPIC33F, PIC24F and PIC24H

Revision E (June 2010)

This revision includes the following corrections and updates:

• Information specific to dsPIC33E and PIC24E devices has been added throughout the document

Revision F (July 2011)

This revision includes the following corrections and updates:

• Added a new section "Built-in Functions"
• Added and updated the cross-references throughout the document
• Updated the bit characteristics from U to U-0 in Register 2-4 and Register 2-6
• Added a note throughout the document specifying the requirement of an additional cycle for read and read-modify-write operations on non-CPU special function registers in dsPIC33E and PIC24E devices
• Updates to formatting and minor text changes were incorporated throughout the document

Revision G (April 2018)

This revision includes the following corrections and updates:

• Information specific to dsPIC33C devices has been added throughout the document
- Updates to formatting and minor text changes were incorporated throughout the document

NOTES:

INDEX

Symbols

__builtin_add.... 462

builtin addab.... 461

builtin btg.... 463

__builtin_clr 464

__builtin_clr_prefetch 465

__builtin_divf 467

__builtin_divmodsd 467

__builtin_divmodud 468

builtin divsd 468

_builtin_divud 469

_builtin dmaoffset 469

_builtin_ed....470

__builtin_edac 471

__builtin_edsoffset 472

__builtin_edspage 472

_builtin_fbcl 473

builtin lac 473

_builtin_mac 474

_builtin_modsd 476

__builtin_modud.... 476

__builtin_movsac 477

__builtin_mpy 478

__builtin_mpyn 479

__builtin_msc 480

_builtin_mulss 482

_builtin_mulsu 482

_builtin_mulus 483

_builtin_muluu 483

__builtin_nop.... 484

__builtin_psvoffset 484

__builtin_psvpage 485

builtin readsfr.... 485

_builtin_return_address.... 486

_builtin_sac 486

_builtin_sacr 487

builtin sftac 488

__builtin_subab 489

__builtin_tbladdress 489

_builtin_tbloffset 490

_builtin tblpage 491

_builtin_tblrdh.... 491

__builtin_tblrdl.... 492

__builtin_tblwth 492

__builtin_tblwtl.... 493

A

Accumulator A, Accumulator B 20

Accumulator Access 86

Accumulator Selection 100

Accumulator Usage 85

Addressing Modes for Wd Destination Register 97

Addressing Modes for Ws Source Register 97

Architecture Overview....10

Assigned Working Register Usage 80

B

Bit Field Insert/Extract Instructions 71

Block Diagrams

DO Stack Conceptual 26

dsPIC30F/dsPIC33F Programmer's Model 16

dsPIC33C Programmer's Model 18

dsPIC33E Programmer's Model 17

PIC24E Programmer's Model 15

PIC24F/PIC24H Programmer's Model.... 14

Built-In Functions

builtin add 462

_builtin_addab 461

_builtin_btg 463

__builtin_clr 464

__builtin_clr_prefetch.... 465

__builtin_divf.... 467

__builtin_divmodsd 467

__builtin_divmodud.... 468

__builtin_divsd.... 468

__builtin_divud.... 469

__builtin_dmaoffset 469

_builtin_ed 470

__builtin_edac 471

__builtin_edsoffset.... 472

__builtin_edspage 472

__builtin_fbcl.... 473

builtin lac.... 473

__builtin_mac 474

__builtin_modsd 476

__builtin_modud 476

__builtin_movsac 477

__builtin_mpy 478

__builtin_mpyn 479

_builtin_msc 480

__builtin_mulss.... 482

__builtin_mulsu.... 482

__builtin_mulus.... 483

__builtin_muluu 483

__builtin_nop 484

__builtin_psvoffset.... 484

_builtin_psvpage.... 485

_builtin_readsfr 485

__builtin_return_address 486

__builtin_sac.... 486

__builtin_sacr 487

__builtin_sftac.... 488

__builtin_subab 489

__builtin_tbladdress.... 489

__builtin_tbloffset.... 490

__builtin_tblpage 491

__builtin_tblrdh 491

__builtin_tblrdl 492

__builtin_tblwth.... 492

__builtin_tblwtl 493

Byte Operations 66

C

Code Examples

'Z' Status Bit Operation for 32-Bit Addition ...... 79

32-Bit Signed Multiplication Using Implicit Mixed-Sign Mode....94

Additional In-Line Functions.... 494

Base MAC Syntax.... 88

Divide 32 by 16 495

File Register Addressing....55

File Register Addressing and WREG.... 55

Frame Pointer Usage....75

Illegal Word Move Operations.... 70

Immediate Addressing 61

Indirect Addressing with Effective Address Update 57

Indirect Addressing with Register Offset....58

Legal Word Move Operations 69

MAC Accumulator WB Syntax 89

MAC Prefetch Syntax.... 88

Move with Literal Offset Instructions ....58

Moving Data with WREG 82

MSC Instruction with Two Prefetches and Accumulator Write-Back 89

Normalizing with FBCL 93

Register Direct Addressing ....56

Sample Byte Math Operations 67

Sample Byte Move Operations 66

Scaling with FBCL....91

Stack Pointer Usage 73

Unsigned f and WREG Multiply (Legacy MULWF Instruction) 82

Using 10-Bit Literals for Byte Operands....71

Using the Default Working Register, WREG.... 81

Conditional Branch Instructions 78

Core Control Register 25

D

Data Addressing Mode Tree 61

Data Addressing Modes.... 54

Data Range Limit Instructions 92

FLIM/FLIM.V 92

MAX/MAX.V 92

MIN/MIN.V/MINZ 92

DCOUNT Register 21

Default Working Register (WREG).... 20, 81

Destination Addressing Modes for MCU Multiplications 97

Development Support 6

DO Stack....26

DOEND Register....22

DOSTART Register.... 22

DSP Accumulator Instructions 90

DSP Data Formats....83

DSP MAC Indirect Addressing Modes 59

DSP MAC Instructions 86

E

Extended Precision Arithmetic using Mixed-Sign Multiplications 94

F

File Register Addressing ....54

|

Immediate Addressing....60

Operands in the Instruction Set 60

Implied DSP Operands 80

Implied Frame and Stack Pointer 81

Instruction Bit Map 498

Instruction Description Example 101

Instruction Descriptions 102

ADD (16-Bit Signed Add to Accumulator).... 108

ADD (Add Accumulators).... 107

ADD (Add f to WREG) 102

ADD (Add Literal to Wn) 103

ADD (Add Wb to Short Literal).... 104

ADD (Add Wb to Ws).... 105

ADDC (Add f to WREG with Carry) 110

ADDC (Add Literal to Wn with Carry) 111

ADDC (Add Wb to Short Literal with Carry).... 112

ADDC (Add Wb to Ws with Carry).... 114

AND (AND f and WREG).... 116

AND (AND Literal and Wn).... 117

AND (AND Wb and Short Literal) 118

AND (AND Wb and Ws).... 119

ASR (Arithmetic Shift Right by Short Literal) ...... 125

ASR (Arithmetic Shift Right by Wns) 126

ASR (Arithmetic Shift Right f) 121

ASR (Arithmetic Shift Right Ws) 123

BCLR (Bit Clear in f) 127

BCLR (Bit Clear in Ws).... 128

BFEXT (Bit Field Extract from f into Wnd) 131

BFEXT (Bit Field Extract from Ws into Wnd).... 130

BFINS (Bit Field Insert from Wb into Wd) 132

BFINS (Bit Field Insert from Wns into f).... 133

BFINS (Bit Field Insert Literal into Ws).... 134

BOOTSWP (Swap Active and Inactive Flash Address Panel) 135

BRA (Branch Unconditionally) 136

BRA (Computed Branch).... 137, 138

BRA C (Branch if Carry) 139

BRA GE (Branch if Signed Greater Than or Equal) 141

BRA GEU (Branch if Unsigned Greater Than or Equal) 142

BRA GT (Branch if Signed Greater Than) 143

BRA GTU (Branch if Unsigned Greater Than)...... 144

BRA LE (Branch if Signed Less Than or Equal)..... 145

BRA LEU (Branch if Unsigned Less Than or Equal) 146

BRA LT (Branch if Signed Less Than).... 147

BRA LTU (Branch if Unsigned Less Than).... 148

BRA N (Branch if Negative) 149

BRA NC (Branch if Not Carry) 150

BRA NN (Branch if Not Negative).... 151

BRA NOV (Branch if Not Overflow) 152

BRA NZ (Branch if Not Zero).... 153

BRA OA (Branch if Overflow Accumulator A) 154

BRA OB (Branch if Overflow Accumulator B) ..... 155

BRA OV (Branch if Overflow) 156

BRA SA (Branch if Saturation Accumulator A) ..... 157

BRA SB (Branch if Saturation Accumulator B) ..... 158

BRA Z (Branch if Zero) 159

BSET (Bit Set in f).... 160

BSET (Bit Set in Ws) 161

BSW (Bit Write in Ws).... 163

BTG (Bit Toggle in f) 165

BTG (Bit Toggle in Ws) 166

BTSC (Bit Test f, Skip if Clear) 168

BTSC (Bit Test Ws, Skip if Clear) 170

BTSS (Bit Test f, Skip if Set).... 172

BTSS (Bit Test Ws, Skip if Set).... 173

BTST (Bit Test in f) 175

BTST (Bit Test in Ws) 176, 178

BTSTS (Bit Test/Set in f).... 180

BTSTS (Bit Test/Set in Ws) 181

CALL (Call Indirect Subroutine) 187, 189

CALL (Call Subroutine) 183, 185

CALL.L (Call Indirect Subroutine Long) 191

CLR (Clear Accumulator, Prefetch Operands)...... 194

CLR (Clear f or WREG) 192

CLR (Clear Wd) 193

CLRWDT (Clear Watchdog Timer) 196

COM (Complement f).... 197

COM (Complement Ws).... 198

CP (Compare f with WREG, Set Status Flags)..... 200

CP (Compare Wb with lit5, Set Status Flags) ...... 201

CP (Compare Wb with lit8, Set Status Flags) ...... 202

CP (Compare Wb with Ws, Set Status Flags) ...... 203

CP0 (Compare f with 0x0, Set Status Flags) ...... 204

CP0 (Compare Ws with 0x0, Set Status Flags) ..... 205

CPB (Compare f with WREG Using Borrow, Set Status Flags) 206

CPB (Compare Wb with lit5 Using Borrow, Set Status Flags) 207

CPB (Compare Wb with lit8 Using Borrow, Set Status Flags) 208

CPB (Compare Ws with Wb Using Borrow, Set Status Flags) 209

CPBEQ (Compare Wb with Wn, Branch if Equal) 211

CPBGT (Signed Compare Wb with Wn, Branch if Greater Than) 212

CPBLT (Signed Compare Wb with Wn, Branch if Less Than) 213

CPBNE (Compare Wb with Wn, Branch if Not Equal).... 214

CPSEQ (Compare Wb with Wn, Skip if Equal).... 215, 216

CPSGT (Signed Compare Wb with Wn, Skip if Greater Than) 217

CPSGT (Signed Compare Wb with Wn, Skip if Greater Than) 218

CPSLT (Signed Compare Wb with Wn, Skip if Less Than) 219, 220

CPSNE (Signed Compare Wb with Wn, Skip if Not Equal) 221, 222

CTXTSWP (CPU Register Context Swap Literal).... 223

CTXTSWP (CPU Register Context Swap Wn) ...... 224

DAW.B (Decimal Adjust Wn) 225

DEC (Decrement f) 226

DEC (Decrement Ws) 227

DEC2 (Decrement f by 2).... 229

DEC2 (Decrement Ws by 2) 230

DISI (Disable Interrupts Temporarily) 232

DIV.S (Signed Integer Divide)...... 233

DIV.U (Unsigned Integer Divide).... 235

DIV2.S (Signed Integer Divide).... 240

DIV2.U (Unsigned Integer Divide).... 241

DIVF (Fractional Divide).... 236

DIVF2 (Signed Fractional Divide, 16/16) ...... 238

DO (Initialize Hardware Loop Literal).... 242, 244

DO (Initialize Hardware Loop Wn).... 246, 248

ED (Euclidean Distance.... 250

ED (Euclidean Distance, No Accumulate) 250

EDAC (Euclidean Distance) 252

EXCH (Exchange Wns and Wnd) 254

FBCL (Find First Bit Change from Left) 255

FF1L (Find First One from Left).... 257

FF1R (Find First One from Right).... 259

FLIM (Force (Signed) Data Range Limit) 261

FLIM.V (Force (Signed) Data Range Limit with Limit Excess Result).... 262

GOTO (Unconditional Indirect Jump) 264, 265

GOTO (Unconditional Jump) 263

GOTO.L (Unconditional Indirect Jump Long) ...... 266

INC (Increment f) 267

INC (Increment Ws) 268

INC2 (Increment f by 2) 269

INC2 (Increment Ws by 2) 270

IOR (Inclusive OR f and WREG) 271

IOR (Inclusive OR Literal and Wn) 272

IOR (Inclusive OR Wb and Short Literal).... 273

IOR (Inclusive OR Wb and Ws).... 274

LAC (Load Accumulator) 276

LAC.D (Load Accumulator Double) 278

LDSLV (Load Slave Processor Program RAM)..... 279

LNK (Allocate Stack Frame) 280, 281

LSR (Logical Shift Right by Short Literal) 286

LSR (Logical Shift Right by Wns) 287

LSR (Logical Shift Right f) 282

LSR (Logical Shift Right Ws) 284

MAC (Multiply and Accumulate) 288

MAC (Square and Accumulate).... 290

MAX (Accumulator Force Maximum Data Range Limit).... 292

MAX.V (Accumulator Force Maximum Data Range Limit with Limit Excess Result) 293

MIN (Accumulator Force Minimum Data Range Limit).... 294

MIN.V (Accumulator Force Minimum Data Range Limit with Limit Excess Result) 295

MINZ (Accumulator Force Minimum Data Range Limit).... 296

MINZ.V (Accumulator Force Minimum Data Range Limit with Limit Excess Result) 297

MOV (Move 16-Bit Literal to Wnd).... 304

MOV (Move f to Destination) 299

MOV (Move f to Wnd).... 301

MOV (Move Wns to f) 302

MOV (Move WREG to f) 300

MOV (Move Ws to Wd).... 307

MOV.B (Move 8-Bit Literal to Wnd) 303

MOV.D (Double-Word Move from Source to Wnd).... 309

MOVPAG (Move Literal to Page Register) 311

MOVPAG (Move Wn to Page Register) 312

MOVSAC (Prefetch Operands and Store Accumulator).... 313

MPY (Multiply Wm by Wn to Accumulator).... 315

MPY (Square to Accumulator).... 317

MPY.N (Multiply -Wm by Wn to Accumulator)...... 319

MSC (Multiply and Subtract from Accumulator) ..... 321

MUL (Integer Unsigned Multiply f and WREG)...... 323

MUL.SS (Integer 16x16-Bit Signed Multiply with Accumulator Destination) 327

MUL.SS (Integer 16x16-Bit Signed Multiply) ...... 325

MUL.SU (Integer 16x16-Bit Signed-Unsigned Multiply with Accumulator Destination).... 331

MUL.SU (Integer 16x16-Bit Signed-Unsigned Multiply) 329

MUL.SU (Integer 16x16-Bit Signed-Unsigned Short Literal Multiply with Accumulator Destination)..... 332

MUL.SU (Integer 16x16-Bit Signed-Unsigned Short Literal Multiply).... 328

MUL.US (Integer 16x16-Bit Unsigned-Signed Multiply with Accumulator Destination) 335

MUL.US (Integer 16x16-Bit Unsigned-Signed Multiply) .... 333

MUL.UU (Integer 16x16-Bit Unsigned Multiply with Accumulator Destination) 339

MUL.UU (Integer 16x16-Bit Unsigned Multiply) ...... 337

MUL.UU (Integer 16x16-Bit Unsigned Short Literal Multiply with Accumulator Destination) ...... 340

MUL.UU (Integer 16x16-Bit Unsigned Short Literal Multiply) 336

MULW.SS (Integer 16x16-Bit Signed Multiply with 16-Bit Result).... 341

MULW.SU (Integer 16x16-Bit Signed-Unsigned Multiply with 16-Bit Result)....343

MULW.SU (Integer 16x16-Bit Signed-Unsigned Short Literal Multiply with 16-Bit Result) 345

MULW.US (Integer 16x16-Bit Unsigned-Signed Multiply with 16-Bit Result)....346

MULW.UU (Integer 16x16-Bit Unsigned Multiply with 16-Bit Result)....348

MULW.UU (Integer 16x16-Bit Unsigned Short Literal Multiply with 16-Bit Result)....349

NEG (Negate Accumulator) 353

NEG (Negate f) 350

NEG (Negate Ws) 351

NOP (No Operation) 354

NOPR (No Operation)......355

NORM (Normalize Accumulator) 356

POP (Pop TOS to f) 357

POP (Pop TOS to Wd).... 358

POP.D (Double Pop TOS to Wnd/Wnd+1) 359

POP.S (Pop Shadow Registers) 360

PUSH (Push f to TOS) 361

PUSH (Push Ws to TOS) 362

PUSH.D (Double Push Wns/Wns+1 to TOS)...... 364

PUSH.S (Push Shadow Registers)....365

PWRSAV (Enter Power-Saving Mode) 366

RCALL (Computed Relative Call) 371, 373

RCALL (Relative Call).... 367, 369

REPEAT (Repeat Next Instruction 'lit14+1' Times)....375

REPEAT (Repeat Next Instruction 'lit15+1' Times)....376

REPEAT (Repeat Next Instruction Wn+1 Times) 377, 378

RESET (Reset) 379

RETFIE (Return from Interrupt) 380, 381

RETLW (Return with Literal in Wn) 382, 384

RETURN (Return).... 386, 387

RLC (Rotate Left f through Carry).... 388

RLC (Rotate Left Ws through Carry).... 389

RLNC (Rotate Left f without Carry) 391

RLNC (Rotate Left Ws without Carry).... 392

RRC (Rotate Right f through Carry).... 394

RRC (Rotate Right Ws through Carry).... 396

RRNC (Rotate Right f without Carry) 398

RRNC (Rotate Right Ws without Carry).... 399

SAC (Store Accumulator).... 40

SAC.D (Store Accumulator Double).... 403

SAC.R (Store Rounded Accumulator) 404

SE (Sign-Extend Ws) 406

SETM (Set f or WREG).... 408

SETM (Set Ws)....409

SFTAC (Arithmetic Shift Accumulator by Slit6)...... 410

SFTAC (Arithmetic Shift Accumulator by Wb) ...... 411

SL (Shift Left by Short Literal)...... 416

SL (Shift Left by Wns).... 417

SL (Shift Left f)...... 412

SL (Shift Left Ws)...... 414

SUB (Subtract Accumulators).... 423

SUB (Subtract Literal from Wn) 419

SUB (Subtract Short Literal from Wb)...... 420

SUB (Subtract WREG from f) 418

SUB (Subtract Ws from Wb).... 421

SUBB (Subtract Short Literal from Wb with Borrow) 426

SUBB (Subtract Wn from Literal with Borrow) ...... 425

SUBB (Subtract WREG and Carry Bit from f)...... 424

SUBB (Subtract Ws from Wb with Borrow)...... 428

SUBBR (Subtract f from WREG with Borrow) ..... 430

SUBBR (Subtract Wb from Short Literal with Borrow) 431

SUBBR (Subtract Wb from Ws with Borrow) 433

SUBR (Subtract f from WREG).... 435

SUBR (Subtract Wb from Short Literal).... 436

SUBR (Subtract Wb from Ws) 437

SWAP (Byte or Nibble Swap Wn).... 439

TBLRDH (Table Read High) 440

TBLRDL (Table Read Low) 442

TBLWTH (Table Write High)...... 444

TBLWTL (Table Write Low) 446

ULNK (Deallocate Stack Frame) 448, 449

VFSLV (Verify Slave Processor Program RAM)...... 450

XOR (Exclusive OR f and WREG)...... 451

XOR (Exclusive OR Literal and Wn)...... 452

XOR (Exclusive OR Wb and Short Literal) 453

XOR (Exclusive OR Wb and Ws) 454

ZE (Zero-Extend Ws) 456

Instruction Encoding 499

Instruction Encoding Field Descriptors Introduction ..... 96

Instruction Set Overview.... 40

Bit Instructions 47

Compare/Skip and Compare/Branch Instructions ..... 48

Control Instructions.... 51

DSP Instructions.... 52

Instruction Groups 40

Logic Instructions.... 45

Math Instructions 43

Move Instructions.... 42

Program Flow Instructions.... 49

Rotate/Shift Instructions.... 46

Shadow/Stack/Context Instructions.... 51

Instruction Set Summary Table 501

Instruction Set Symbols....8

(text) 8

[text]....8

{ }....8

{label:} 8

text....8

8

Acc....8

AWB 8

bit4 8

Expr 8

f....8

lit1

lit10 8

lit14 8

lit16 8

lit23 8

lit4 8

lit5 8

lit8 8

Slit10 8

Slit16 8

Slit4....8

Slit6....8

TOS....8

Wb....8

Wd....8

Wdb....8

Wm * Wm....8

Wm * Wn 8

Wm, Wn 8

Wn....8

Wnd....8

Wns....8

WREG 8

Ws....8

Wsb 8

Wx....8

Wxd....8

Wy....8

Wyd....8

Instruction Stalls.... 64

DO/REPEAT Loops 65

Exceptions 65

Instructions that Change Program Flow.... 65

PSV....65

RAW Dependency Detection 64

Instruction Symbols.... 96

Integer and Fractional Data 83

Representation...... 84

M

MAC

Operations 87

Prefetch Register Updates.... 87

Prefetches....87

Syntax 88

Write-Back 87

MAC Accumulator Write-Back Selection.... 100

MAC or MPY Source Operands

(Different Working Register) 99

MAC or MPY Source Operands

(Same Working Register).... 99

Manual Objective 6

Modulo and Bit-Reversed Addressing Modes.... 59

MOVPAG Destination Selection 100

Multicycle Instructions.... 41

Multiword Instructions 41

N

Normalizing the Accumulator with the

FBCL Instruction 93

Normalizing the Accumulator with the

NORM Instruction 93

0

Offset Addressing Modes for Wd Destination Register

(with Register Offset) 98

Offset Addressing Modes for Ws Source Register

(with Register Offset) 98

P

PIC Microcontroller Compatibility....81

PRODH

PRODL Register Pair 81

Program Addressing Modes 63

Methods of Modifying Flow.... 63

Program Counter (PC).... 21

Programmer's Model 14

Register Descriptions 19

PSVPAG Register....21

R

RCOUNT Register 21

Register Direct Addressing 55

Register Indirect Addressing.... 56

and the Instruction Set.... 59

Modes....56

Registers

CORCON (Core Control - dsPIC30F, dsPIC33F)..... 34

CORCON (Core Control - dsPIC33E, dsPIC33C) ..... 36

CORCON (Core Control - PIC24E) 33

CORCON (Core Control - PIC24F, PIC24H).... 32

SR (CPU STATUS - dsPIC30F, dsPIC33F) 28

SR (CPU STATUS - dsPIC33E, dsPIC33C).... 30

SR (CPU STATUS - PIC24H, PIC24F, PIC24E) ..... 27

Revision History.... 511

S

Scaling Data with the FBCL Instruction 90

Scaling Examples 91

Shadow Registers.... 25

Automatic Usage 25

Software Stack Frame Pointer.... 20,74

Example....75

Overflow 76

Underflow 77

Software Stack Pointer 72

Example....73

Software Stack Pointer (SSP).... 20

Stack Frame Active (SFA) Control.... 77

Stack Pointer Limit Register (SPLIM) 20

STATUS Register 22

DO Loop Active (DA) Status Bit.... 23

DSP ALU Status Bits.... 23

Interrupt Priority Level Bits 24

MCU ALU Status Bits 22

REPEAT Loop Active (RA) Status Bit.... 23

Style and Symbol Conventions....7

Document Conventions 7

T

TBLPAG Register 21

to Wnd) 305

U

Using 10-Bit Literal Operands....71

10-Bit Literal Coding 71

16-Bit MCU and DSC Programmer's Reference Manual

W

Instruction Descriptions

MOV (Move Wns to....306

Word Move Operations 68

Data Alignment in Memory....68

Working Register Array.... 19

Instruction Descriptions

MOV (Move.... 305

x

X Data Space Prefetch Operation 98

Y

Y Data Space Prefetch Destination 99

Y Data Space Prefetch Operation 99

Z

Z Status Bit 79

NOTES:

Worldwide Sales and Service

AMERICAS

Corporate Office

2355 West Chandler Blvd.

Chandler, AZ 85224-6199

Tel: [phone]

Fax: [phone]

Technical Support:

http://www.microchip.com/

support

Web Address:

www.microchip.com

Atlanta

Duluth, GA

Tel: [phone]

Fax: [phone]

Austin, TX

Tel: [phone]

Boston

Westborough, MA

Tel: [phone]

Fax: [phone]

Chicago

Itasca, IL

Tel: [phone]

Fax: [phone]

Dallas

Addison, TX

Tel: [phone]

Fax: [phone]

Detroit

Novi, MI

Tel: [phone]

Houston, TX

Tel: [phone]

Indianapolis

Noblesville, IN

Tel: [phone]

Fax: [phone]

Tel: [phone]

Los Angeles

Mission Viejo, CA

Tel: [phone]

Fax: [phone]

Tel: [phone]

Raleigh, NC

Tel: [phone]

New York, NY

Tel: [phone]

San Jose, CA

Tel: [phone]

Tel: [phone]

Canada - Toronto

Tel: [phone]

Fax: [phone]

ASIA/PACIFIC

Australia - Sydney

Tel: [phone]

China - Beijing

Tel: [phone]

China - Chengdu

Tel: [phone]

China - Chongqing

Tel: [phone]

China - Dongguan

Tel: [phone]

China - Guangzhou

Tel: [phone]

China - Hangzhou

Tel: [phone]

China - Hong Kong SAR

Tel: [phone]

China - Nanjing

Tel: [phone]

China - Qingdao

Tel: [phone]

China - Shanghai

Tel: [phone]

China - Shenyang

Tel: [phone]

China - Shenzhen

Tel: [phone]

China - Suzhou

Tel: [phone]

China - Wuhan

Tel: [phone]

China - Xian

Tel: [phone]

China - Xiamen

Tel: [phone]

China - Zhuhai

Tel: [phone]

ASIA/PACIFIC

India - Bangalore

Tel: [phone]

India - New Delhi

Tel: [phone]

India - Pune

Tel: [phone]

Japan - Osaka

Tel: [phone]

Japan - Tokyo

Tel: [phone]

Korea - Daegu

Tel: [phone]

Korea - Seoul

Tel: [phone]

Malaysia - Kuala Lumpur

Tel: [phone]

Malaysia - Penang

Tel: [phone]

Philippines - Manila

Tel: [phone]

Singapore

Tel: [phone]

Taiwan - Hsin Chu

Tel: [phone]

Taiwan - Kaohsiung

Tel: [phone]

Taiwan - Taipei

Tel: [phone]

Thailand - Bangkok

Tel: [phone]

Tel: [phone]

Fax: [phone]

Denmark - Copenhagen

Tel: [phone]

Fax: [phone]

Finland - Espoo

Tel: [phone]

France - Paris

Tel: [phone]

Fax: [phone]

Germany - Garching

Tel: [phone]

Germany - Haan

Tel: [phone]

Germany - Heilbronn

Tel: [phone]

Germany - Karlsruhe

Tel: [phone]

Germany - Munich

Tel: [phone]

Fax: [phone]

Germany - Rosenheim

Tel: [phone]

Israel - Ra'anana

Tel: [phone]

Italy - Milan

Tel: [phone]

Fax: [phone]

Italy - Padova

Tel: [phone]

Netherlands - Drunen

Tel: [phone]

Fax: [phone]

Norway - Trondheim

Tel: [phone]

Poland - Warsaw

Tel: [phone]

Romania - Bucharest

Tel: [phone]

Spain - Madrid

Tel: [phone]

Fax: [phone]

Sweden - Gothenberg

Tel: [phone]

Sweden - Stockholm

Tel: [phone]

UK - Wokingham

Tel: [phone]

Fax: [phone]