MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture
Document Number: MD00082 Revision 3.02
March 21, 2011
MIPS Technologies, Inc.
955 East Arques Avenue Sunnyvale, CA 94085-4521
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies, Inc. All rights reserved.
Unpublished rights (if any) reserved under the copyright laws of the United States of America and other countries.
This document contains information that is proprietary to MIPS Technologies, Inc. (“MIPS Technologies”). Any copying, reproducing, modifying or use of this information (in whole or in part) that is not expressly permitted in writing by MIPS Technologies or an authorized third party is strictly prohibited. At a minimum, this information is protected under unfair competition and copyright laws. Violations thereof may result in criminal penalties and fines.
Any document provided in source format (i.e., in a modifiable form such as in FrameMaker or Microsoft Word format) is subject to use and distribution restrictions that are independent of and supplemental to any and all confidentiality restrictions. UNDER NO CIRCUMSTANCES MAY A DOCUMENT PROVIDED IN SOURCE FORMAT BE DISTRIBUTED TO A THIRD PARTY IN SOURCE FORMAT WITHOUT THE EXPRESS WRITTEN PERMISSION OF MIPS TECHNOLOGIES, INC.
MIPS Technologies reserves the right to change the information contained in this document to improve function, design or otherwise. MIPS Technologies does not assume any liability arising out of the application or use of this information, or of any error or omission in such information. Any warranties, whether express, statutory, implied or otherwise, including but not limited to the implied warranties of merchantability or fitness for a particular purpose, are excluded. Except as expressly provided in any written license agreement from MIPS Technologies or an authorized third party, the furnishing of this document does not give recipient any license to any intellectual property rights, including any patent rights, that cover the information in this document.
The information contained in this document shall not be exported, reexported, transferred, or released, directly or indirectly, in violation of the law of any country or international law, regulation, treaty, Executive Order, statute, amendments or supplements thereto. Should a conflict arise regarding the export, reexport, transfer, or release of the information contained in this document, the laws of the United States of America shall be the governing law.
The information contained in this document constitutes one or more of the following: commercial computer software, commercial computer software documentation or other commercial items. If the user of this information, or any related documentation of any kind, including related technical data or manuals, is an agency, department, or other entity of the United States government (“Government”), the use, duplication, reproduction, release, modification, disclosure, or transfer of this information, or any related documentation of any kind, is restricted in accordance with Federal Acquisition Regulation 12.212 for civilian agencies and Defense Federal Acquisition Regulation Supplement 227.7202 for military agencies. The use of this information by the Government is further restricted in accordance with the terms of the license agreement(s) and/or applicable contract terms and conditions covering this information from MIPS Technologies or an authorized third party.
MIPS, MIPS I, MIPS II, MIPS III, MIPS IV, MIPS V, MIPS-3D, MIPS16, MIPS16e, MIPS32, MIPS64, MIPS-Based, MIPSsim, MIPSpro, MIPS Technologies logo, MIPS-VERIFIED, MIPS-VERIFIED logo, 4K, 4Kc, 4Km, 4Kp, 4KE, 4KEc, 4KEm, 4KEp, 4KS, 4KSc, 4KSd, M4K, M14K, 5K, 5Kc, 5Kf, 24K, 24Kc, 24Kf, 24KE, 24KEc, 24KEf, 34K, 34Kc, 34Kf, 74K, 74Kc, 74Kf, 1004K, 1004Kc, 1004Kf, R3000, R4000, R5000, ASMACRO, Atlas, “At the core of the user experience.”, BusBridge, Bus Navigator, CLAM, CorExtend, CoreFPGA, CoreLV, EC, FPGA View, FS2, FS2 FIRST SILICON SOLUTIONS logo, FS2 NAVIGATOR, HyperDebug, HyperJTAG, JALGO, Logic Navigator, Malta, MDMX, MED, MGB, microMIPS, OCI, PDtrace, the Pipeline, Pro Series, SEAD, SEAD-2, SmartMIPS, SOC-it, System Navigator, and YAMON are trademarks or registered trademarks of MIPS Technologies, Inc. in the United States and other countries.
All other trademarks referred to herein are the property of their respective owners.
Template: nB1.03, Built with tags: 2B ARCH MIPS32
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Contents
Chapter 1: About This Book …………………………………………………………………………………………………. 11
1.1: Typographical Conventions ……………………………………………………………………………………………………….. 11 1.1.1: Italic Text………………………………………………………………………………………………………………………… 11 1.1.2: Bold Text………………………………………………………………………………………………………………………… 12 1.1.3: Courier Text ……………………………………………………………………………………………………………………. 12
1.2: UNPREDICTABLE and UNDEFINED …………………………………………………………………………………………. 12 1.2.1: UNPREDICTABLE…………………………………………………………………………………………………………… 12 1.2.2: UNDEFINED …………………………………………………………………………………………………………………… 13 1.2.3: UNSTABLE …………………………………………………………………………………………………………………….. 13
1.3: Special Symbols in Pseudocode Notation ……………………………………………………………………………………. 13 1.4: For More Information ………………………………………………………………………………………………………………… 16
Chapter 2: The MIPS Architecture: An Introduction………………………………………………………………..17
2.1: MIPS Instruction Set Overview …………………………………………………………………………………………………… 17 2.1.1: Historical Perspective……………………………………………………………………………………………………….. 17 2.1.2: Architectural Evolution ……………………………………………………………………………………………………… 18 2.1.3: Architectural Changes Relative to the MIPS I through MIPS V Architectures……………………………. 21
2.2: Compliance and Subsetting……………………………………………………………………………………………………….. 21 2.3: Components of the MIPS Architecture ………………………………………………………………………………………… 23 2.3.1: MIPS Instruction Set Architecture (ISA) ………………………………………………………………………………. 23 2.3.2: MIPS Privileged Resource Architecture (PRA) …………………………………………………………………….. 23 2.3.3: MIPS Application Specific Extensions (ASEs) ……………………………………………………………………… 24 2.3.4: MIPS User Defined Instructions (UDIs)……………………………………………………………………………….. 24 2.4: Architecture Versus Implementation……………………………………………………………………………………………. 24 2.5: Relationship between the MIPSr3 Architectures …………………………………………………………………………… 24 2.6: Pipeline Architecture…………………………………………………………………………………………………………………. 26 2.6.1: Pipeline Stages and Execution Rates …………………………………………………………………………………. 26 2.6.2: Parallel Pipeline ………………………………………………………………………………………………………………. 27 2.6.3: Superpipeline ………………………………………………………………………………………………………………….. 27 2.6.4: Superscalar Pipeline ………………………………………………………………………………………………………… 28 2.7: Load/Store Architecture…………………………………………………………………………………………………………….. 28 2.8: Programming Model …………………………………………………………………………………………………………………. 29 2.8.1: CPU Data Formats…………………………………………………………………………………………………………… 29 2.8.2: FPU Data Formats …………………………………………………………………………………………………………… 29 2.8.3: Coprocessors (CP0-CP3) …………………………………………………………………………………………………. 30 2.8.4: CPU Registers ………………………………………………………………………………………………………………… 30 2.8.5: FPU Registers…………………………………………………………………………………………………………………. 32 2.8.6: Byte Ordering and Endianness ………………………………………………………………………………………….. 37 2.8.7: Memory Access Types……………………………………………………………………………………………………… 39 2.8.8: Implementation-Specific Access Types ………………………………………………………………………………. 40 2.8.9: Cacheability and Coherency Attributes and Access Types…………………………………………………….. 40 2.8.10: Mixing Access Types ……………………………………………………………………………………………………… 40 2.8.11: Instruction Fetches…………………………………………………………………………………………………………. 41
Chapter 3: Application Specific Extensions ………………………………………………………………………….. 47
3.1: Description of ASEs………………………………………………………………………………………………………………….. 47 3.2: List of Application Specific Instructions ……………………………………………………………………………………….. 48
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 3
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
3.2.1: The MIPS16eTM Application Specific Extension to the MIPS32 and MIPS64 Architecture………….. 48 3.2.2: The MDMXTM Application Specific Extension to the MIPS64 Architectures………………………………. 48 3.2.3: The MIPS-3D® Application Specific Extension to the MIPS Architecture…………………………………. 48 3.2.4: The SmartMIPS® Application Specific Extension to the MIPS32 Architecture ………………………….. 48 3.2.5: The MIPS® DSP Application Specific Extension to the MIPS Architecture ………………………………. 49 3.2.6: The MIPS® MT Application Specific Extension to the MIPS Architecture ………………………………… 49 3.2.7: The MIPS® MCU Application Specific Extension to the MIPS Architecture ……………………………… 49
Chapter 4: Overview of the CPU Instruction Set ……………………………………………………………………. 51
4.1: CPU Instructions, Grouped By Function………………………………………………………………………………………. 51 4.1.1: CPU Load and Store Instructions……………………………………………………………………………………….. 51 4.1.2: Computational Instructions………………………………………………………………………………………………… 54 4.1.3: Jump and Branch Instructions……………………………………………………………………………………………. 58 4.1.4: Miscellaneous Instructions ………………………………………………………………………………………………… 60 4.1.5: Coprocessor Instructions ………………………………………………………………………………………………….. 63
4.2: CPU Instruction Formats …………………………………………………………………………………………………………… 64
Chapter 5: Overview of the FPU Instruction Set……………………………………………………………………..67
5.1: Binary Compatibility ………………………………………………………………………………………………………………….. 67 5.2: Enabling the Floating Point Coprocessor …………………………………………………………………………………….. 68 5.3: IEEE Standard 754…………………………………………………………………………………………………………………… 68 5.4: FPU Data Types ………………………………………………………………………………………………………………………. 68
5.4.1: Floating Point Formats ……………………………………………………………………………………………………… 68
5.4.2: Fixed Point Formats …………………………………………………………………………………………………………. 72 5.5: Floating Point Register Types ……………………………………………………………………………………………………. 73 5.5.1: FPU Register Models ……………………………………………………………………………………………………….. 73 5.5.2: Binary Data Transfers (32-Bit and 64-Bit) ……………………………………………………………………………. 73 5.5.3: FPRs and Formatted Operand Layout ………………………………………………………………………………… 74 5.6: Floating Point Control Registers (FCRs) ……………………………………………………………………………………… 75 5.6.1: Floating Point Implementation Register (FIR, CP1 Control Register 0) ……………………………………. 75 5.6.2: Floating Point Control and Status Register (FCSR, CP1 Control Register 31)………………………….. 77 5.6.3: Floating Point Condition Codes Register (FCCR, CP1 Control Register 25)…………………………….. 80 5.6.4: Floating Point Exceptions Register (FEXR, CP1 Control Register 26) …………………………………….. 80 5.6.5: Floating Point Enables Register (FENR, CP1 Control Register 28)…………………………………………. 81 5.7: Formats of Values Used in FP Registers …………………………………………………………………………………….. 82 5.8: FPU Exceptions……………………………………………………………………………………………………………………….. 82 5.8.1: Exception Conditions ……………………………………………………………………………………………………….. 83 5.9: FPU Instructions ………………………………………………………………………………………………………………………. 85 5.9.1: Data Transfer Instructions…………………………………………………………………………………………………. 86 5.9.2: Arithmetic Instructions………………………………………………………………………………………………………. 87 5.9.3: Conversion Instructions…………………………………………………………………………………………………….. 89 5.9.4: Formatted Operand-Value Move Instructions ………………………………………………………………………. 90 5.9.5: Conditional Branch Instructions …………………………………………………………………………………………. 91 5.9.6: Miscellaneous Instructions ………………………………………………………………………………………………… 92 5.10: Valid Operands for FPU Instructions …………………………………………………………………………………………. 92 5.11: FPU Instruction Formats………………………………………………………………………………………………………….. 94 5.11.1: Implementation Note ………………………………………………………………………………………………………. 95
Appendix A: Instruction Bit Encodings………………………………………………………………………………….99
A.12: Instruction Encodings and Instruction Classes …………………………………………………………………………… 99 A.13: Instruction Bit Encoding Tables………………………………………………………………………………………………… 99 A.14: Floating Point Unit Instruction Format Encodings ……………………………………………………………………… 107
4 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Appendix B: Revision History …………………………………………………………………………………………….. 109
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 5
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Figures
Figure 2-1: MIPS Architectures ………………………………………………………………………………………………………………. 18 Figure 2-2: Relationship of the Binary Representations of MIPSr3 Architectures…………………………………………… 25 Figure 2-3: Relationships of the Assembler Source Code Representations of the MIPSr3 Architectures ………….. 26 Figure 2-4: One-Deep Single-Completion Instruction Pipeline ……………………………………………………………………. 27 Figure 2-5: Four-Deep Single-Completion Pipeline …………………………………………………………………………………… 27 Figure 2-6: Four-Deep Superpipeline………………………………………………………………………………………………………. 28 Figure 2-7: Four-Way Superscalar Pipeline ……………………………………………………………………………………………… 28 Figure 2-8: CPU Registers …………………………………………………………………………………………………………………….. 32 Figure 2-9: FPU Registers for a 32-bit FPU ……………………………………………………………………………………………… 34 Figure 2-10: FPU Registers for a 64-bit FPU if StatusFR is 1 …………………………………………………………………….. 35 Figure 2-11: FPU Registers for a 64-bit FPU if StatusFR is 0 ……………………………………………………………………… 36 Figure 2-12: Big-Endian Byte Ordering ……………………………………………………………………………………………………. 37 Figure 2-13: Little-Endian Byte Ordering………………………………………………………………………………………………….. 37 Figure 2-14: Big-Endian Data in Doubleword Format ………………………………………………………………………………… 38 Figure 2-15: Little-Endian Data in Doubleword Format………………………………………………………………………………. 38 Figure 2-16: Big-Endian Misaligned Word Addressing ………………………………………………………………………………. 39 Figure 2-17: Little-Endian Misaligned Word Addressing …………………………………………………………………………….. 39 Figure 2-18: Two instructions placed in a 64-bit wide, little-endian memory………………………………………………….. 41 Figure 2-19: Two instructions placed in a 64-bit wide, big-endian memory …………………………………………………… 42 Figure 3-1: MIPS ISAs and ASEs …………………………………………………………………………………………………………… 47 Figure 4-1: Immediate (I-Type) CPU Instruction Format…………………………………………………………………………….. 65 Figure 4-2: Jump (J-Type) CPU Instruction Format …………………………………………………………………………………… 65 Figure 4-3: Register (R-Type) CPU Instruction Format………………………………………………………………………………. 65 Figure 5-1: Single-Precisions Floating Point Format (S)…………………………………………………………………………….. 69 Figure 5-2: Double-Precisions Floating Point Format (D) …………………………………………………………………………… 70 Figure 5-3: Paired Single Floating Point Format (PS)………………………………………………………………………………… 70 Figure 5-4: Word Fixed Point Format (W) ………………………………………………………………………………………………… 72 Figure 5-5: Longword Fixed Point Format (L) …………………………………………………………………………………………… 72 Figure 5-6: FPU Word Load and Move-to Operations ………………………………………………………………………………. 74 Figure 5-7: FPU Doubleword Load and Move-to Operations………………………………………………………………………. 74 Figure 5-8: Single Floating Point or Word Fixed Point Operand in an FPR ………………………………………………….. 74 Figure 5-9: Double Floating Point or Longword Fixed Point Operand in an FPR …………………………………………… 75 Figure 5-10: Paired-Single Floating Point Operand in an FPR ……………………………………………………………………. 75 Figure 5-11: FIR Register Format ………………………………………………………………………………………………………….. 75 Figure 5-12: FCSR Register Format ………………………………………………………………………………………………………. 78 Figure 5-13: FCCR Register Format ………………………………………………………………………………………………………. 80 Figure 5-14: FEXR Register Format ………………………………………………………………………………………………………. 81 Figure 5-15: FENR Register Format ………………………………………………………………………………………………………. 81 Figure 5-16: I-Type (Immediate) FPU Instruction Format …………………………………………………………………………… 95 Figure 5-17: R-Type (Register) FPU Instruction Format …………………………………………………………………………….. 95 Figure 5-18: Register-Immediate FPU Instruction Format ………………………………………………………………………….. 95 Figure 5-19: Condition Code, Immediate FPU Instruction Format ……………………………………………………………….. 95 Figure 5-20: Formatted FPU Compare Instruction Format …………………………………………………………………………. 95 Figure 5-21: FP RegisterMove, Conditional Instruction Format …………………………………………………………………… 95 Figure 5-22: Four-Register Formatted Arithmetic FPU Instruction Format ……………………………………………………. 95 Figure 5-23: Register Index FPU Instruction Format …………………………………………………………………………………. 96 Figure 5-24: Register Index Hint FPU Instruction Format …………………………………………………………………………… 96
6 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Figure 5-25: Condition Code, Register Integer FPU Instruction Format ……………………………………………………….. 96 Figure A.26: Sample Bit Encoding Table ……………………………………………………………………………………………….. 100
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 7
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Tables
Table 1.1: Symbols Used in Instruction Operation Statements……………………………………………………………………. 13 Table 2.1: Unaligned Load and Store Instructions…………………………………………………………………………………….. 38 Table 2.2: Speculative instruction fetches ……………………………………………………………………………………………….. 42 Table 4.1: Load and Store Operations Using Register + Offset Addressing Mode…………………………………………. 52 Table 4.2: Aligned CPU Load/Store Instructions……………………………………………………………………………………….. 52 Table 4.3: Unaligned CPU Load and Store Instructions …………………………………………………………………………….. 53 Table 4.4: Atomic Update CPU Load and Store Instructions ………………………………………………………………………. 53 Table 4.5: Coprocessor Load and Store Instructions…………………………………………………………………………………. 54 Table 4.6: FPU Load and Store Instructions Using Register + Register Addressing………………………………………. 54 Table 4.7: ALU Instructions With a 16-bit Immediate Operand……………………………………………………………………. 55 Table 4.8: Three-Operand ALU Instructions …………………………………………………………………………………………….. 55 Table 4.9: Two-Operand ALU Instructions……………………………………………………………………………………………….. 56 Table 4.10: Shift Instructions …………………………………………………………………………………………………………………. 56 Table 4.11: Multiply/Divide Instructions……………………………………………………………………………………………………. 57 Table 4.12: Unconditional Jump Within a 256 Megabyte Region ………………………………………………………………… 59 Table 4.13: Unconditional Jump using Absolute Address…………………………………………………………………………… 59 Table 4.14: PC-Relative Conditional Branch Instructions Comparing Two Registers……………………………………… 59 Table 4.15: PC-Relative Conditional Branch Instructions Comparing With Zero…………………………………………….59 Table 4.16: Deprecated Branch Likely Instructions……………………………………………………………………………………. 60 Table 4.17: Serialization Instruction………………………………………………………………………………………………………… 61 Table 4.18: System Call and Breakpoint Instructions ………………………………………………………………………………… 61 Table 4.19: Trap-on-Condition Instructions Comparing Two Registers ………………………………………………………… 61 Table 4.20: Trap-on-Condition Instructions Comparing an Immediate Value ………………………………………………… 61 Table 4.21: CPU Conditional Move Instructions ……………………………………………………………………………………….. 62 Table 4.22: Prefetch Instructions ……………………………………………………………………………………………………………. 62 Table 4.23: NOP Instructions…………………………………………………………………………………………………………………. 63 Table 4.24: Coprocessor Definition and Use in the MIPS Architecture…………………………………………………………. 63 Table 4.25: CPU Instruction Format Fields………………………………………………………………………………………………. 64 Table 5.1: Parameters of Floating Point Data Types …………………………………………………………………………………. 69 Table 5.2: Value of Single or Double Floating Point DataType Encoding……………………………………………………… 70 Table 5.3: Value Supplied When a New Quiet NaN Is Created…………………………………………………………………… 72 Table 5.4: FIR Register Field Descriptions ………………………………………………………………………………………………. 75 Table 5.5: FCSR Register Field Descriptions …………………………………………………………………………………………… 78 Table 5.6: Cause, Enable, and Flag Bit Definitions……………………………………………………………………………………. 79 Table 5.8: FCCR Register Field Descriptions …………………………………………………………………………………………… 80 Table 5.7: Rounding Mode Definitions…………………………………………………………………………………………………….. 80 Table 5.9: FEXR Register Field Descriptions……………………………………………………………………………………………. 81 Table 5.10: FENR Register Field Descriptions …………………………………………………………………………………………. 81 Table 5.11: Default Result for IEEE Exceptions Not Trapped Precisely ……………………………………………………….. 83 Table 5.12: FPU Data Transfer Instructions……………………………………………………………………………………………… 86 Table 5.13: FPU Loads and Stores Using Register+Offset Address Mode …………………………………………………… 86 Table 5.16: FPU IEEE Arithmetic Operations …………………………………………………………………………………………… 87 Table 5.14: FPU Loads and Using Register+Register Address Mode………………………………………………………….. 87 Table 5.15: FPU Move To and From Instructions ……………………………………………………………………………………… 87 Table 5.17: FPU-Approximate Arithmetic Operations ………………………………………………………………………………… 88 Table 5.18: FPU Multiply-Accumulate Arithmetic Operations ……………………………………………………………………… 89 Table 5.19: FPU Conversion Operations Using the FCSR Rounding Mode………………………………………………….. 89
8 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.21: FPU Formatted Operand Move Instructions ……………………………………………………………………………. 90 Table 5.22: FPU Conditional Move on True/False Instructions……………………………………………………………………. 90 Table 5.20: FPU Conversion Operations Using a Directed Rounding Mode …………………………………………………. 90 Table 5.23: FPU Conditional Move on Zero/Nonzero Instructions……………………………………………………………….. 91 Table 5.24: FPU Conditional Branch Instructions ……………………………………………………………………………………… 92 Table 5.25: Deprecated FPU Conditional Branch Likely Instructions …………………………………………………………… 92 Table 5.26: CPU Conditional Move on FPU True/False Instructions ……………………………………………………………. 92 Table 5.27: FPU Operand Format Field (fmt, fmt3) Encoding …………………………………………………………………….. 93 Table 5.28: Valid Formats for FPU Operations…………………………………………………………………………………………. 93 Table 5.29: FPU Instruction Format Fields ………………………………………………………………………………………………. 96 Table A.30: Symbols Used in the Instruction Encoding Tables …………………………………………………………………. 100 Table A.31: MIPS32 Encoding of the Opcode Field ………………………………………………………………………………… 101 Table A.32: MIPS32 SPECIAL Opcode Encoding of Function Field…………………………………………………………… 102 Table A.33: MIPS32 REGIMM Encoding of rt Field …………………………………………………………………………………. 102 Table A.34: MIPS32 SPECIAL2 Encoding of Function Field …………………………………………………………………….. 102 Table A.35: MIPS32 SPECIAL3 Encoding of Function Field for Release 2 of the Architecture………………………. 103 Table A.36: MIPS32 MOVCI Encoding of tf Bit ……………………………………………………………………………………….. 103 Table A.37: MIPS32 SRL Encoding of Shift/Rotate …………………………………………………………………………………. 103 Table A.38: MIPS32 SRLV Encoding of Shift/Rotate……………………………………………………………………………….. 103 Table A.39: MIPS32 BSHFL Encoding of sa Field…………………………………………………………………………………… 104 Table A.40: MIPS32 COP0 Encoding of rs Field …………………………………………………………………………………….. 104 Table A.41: MIPS32 COP0 Encoding of Function Field When rs=CO………………………………………………………… 104 Table A.42: MIPS32 COP1 Encoding of rs Field …………………………………………………………………………………….. 105 Table A.43: MIPS32 COP1 Encoding of Function Field When rs=S…………………………………………………………… 105 Table A.44: MIPS32 COP1 Encoding of Function Field When rs=D ………………………………………………………….. 105 Table A.45: MIPS32 COP1 Encoding of Function Field When rs=W or L …………………………………………………… 106 Table A.46: MIPS64 COP1 Encoding of Function Field When rs=PS ………………………………………………………… 106 Table A.47: MIPS32 COP1 Encoding of tf Bit When rs=S, D, or PS, Function=MOVCF……………………………….. 106 Table A.48: MIPS32 COP2 Encoding of rs Field …………………………………………………………………………………….. 107 Table A.49: MIPS64 COP1X Encoding of Function Field …………………………………………………………………………. 107 Table A.50: Floating Point Unit Instruction Format Encodings…………………………………………………………………… 107
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 9
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
10 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Chapter 1
About This Book
The MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture comes as part of a multi-volume set.
• Volume I-A describes conventions used throughout the document set, and provides an introduction to the MIPS32® Architecture
• Volume I-B describes conventions used throughout the document set, and provides an introduction to the microMIPS32TM Architecture
• Volume II-A provides detailed descriptions of each instruction in the MIPS32® instruction set
• Volume II-B provides detailed descriptions of each instruction in the microMIPS32TM instruction set
• Volume III describes the MIPS32® and microMIPS32TM Privileged Resource Architecture which defines and governs the behavior of the privileged resources included in a MIPS® processor implementation
• Volume IV-a describes the MIPS16eTM Application-Specific Extension to the MIPS32® Architecture. Beginning with Release 3 of the Architecture, microMIPS is the preferred solution for smaller code size.
• Volume IV-b describes the MDMXTM Application-Specific Extension to the MIPS64® Architecture and microMIPS64TM. It is not applicable to the MIPS32® document set nor the microMIPS32TM document set
• Volume IV-c describes the MIPS-3D® Application-Specific Extension to the MIPS® Architecture
• Volume IV-d describes the SmartMIPS®Application-Specific Extension to the MIPS32® Architecture and the microMIPS32TM Architecture
• Volume IV-e describes the MIPS® DSP Application-Specific Extension to the MIPS® Architecture
• Volume IV-f describes the MIPS® MT Application-Specific Extension to the MIPS® Architecture
• Volume IV-h describes the MIPS® MCU Application-Specific Extension to the MIPS® Architecture
1.1 Typographical Conventions
This section describes the use of italic, bold and courier fonts in this book. 1.1.1 Italic Text
• is used for emphasis
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 11
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
About This Book
• is used for bits, fields, registers, that are important from a software perspective (for instance, address bits used by software, and programmable fields and registers), and various floating point instruction formats, such as S, D, and PS
• is used for the memory access types, such as cached and uncached 1.1.2 Bold Text
• represents a term that is being defined
• is used for bits and fields that are important from a hardware perspective (for instance, register bits, which are
not programmable but accessible only to hardware)
• is used for ranges of numbers; the range is indicated by an ellipsis. For instance, 5..1 indicates numbers 5 through 1
• is used to emphasize UNPREDICTABLE and UNDEFINED behavior, as defined below. 1.1.3 Courier Text
Courier fixed-width font is used for text that is displayed on the screen, and for examples of code and instruction pseudocode.
1.2 UNPREDICTABLE and UNDEFINED
The terms UNPREDICTABLE and UNDEFINED are used throughout this book to describe the behavior of the processor in certain cases. UNDEFINED behavior or operations can occur only as the result of executing instructions in a privileged mode (i.e., in Kernel Mode or Debug Mode, or with the CP0 usable bit set in the Status register). Unprivileged software can never cause UNDEFINED behavior or operations. Conversely, both privileged and unprivileged software can cause UNPREDICTABLE results or operations.
1.2.1 UNPREDICTABLE
UNPREDICTABLE results may vary from processor implementation to implementation, instruction to instruction, or as a function of time on the same implementation or instruction. Software can never depend on results that are UNPREDICTABLE. UNPREDICTABLE operations may cause a result to be generated or not. If a result is gener- ated, it is UNPREDICTABLE. UNPREDICTABLE operations may cause arbitrary exceptions.
UNPREDICTABLE results or operations have several implementation restrictions:
12
• •
•
Implementations of operations generating UNPREDICTABLE results must not depend on any data source (memory or internal state) which is inaccessible in the current processor mode
UNPREDICTABLE operations must not read, write, or modify the contents of memory or internal state which is inaccessible in the current processor mode. For example, UNPREDICTABLE operations executed in user mode must not access memory or internal state that is only accessible in Kernel Mode or Debug Mode or in another process
UNPREDICTABLE operations must not halt or hang the processor
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
1.2.2 UNDEFINED
UNDEFINED operations or behavior may vary from processor implementation to implementation, instruction to instruction, or as a function of time on the same implementation or instruction. UNDEFINED operations or behavior may vary from nothing to creating an environment in which execution can no longer continue. UNDEFINED opera- tions or behavior may cause data loss.
UNDEFINED operations or behavior has one implementation restriction:
• UNDEFINED operations or behavior must not cause the processor to hang (that is, enter a state from which there is no exit other than powering down the processor). The assertion of any of the reset signals must restore the processor to an operational state
1.2.3 UNSTABLE
UNSTABLE results or values may vary as a function of time on the same implementation or instruction. Unlike UNPREDICTABLE values, software may depend on the fact that a sampling of an UNSTABLE value results in a legal transient value that was correct at some point in time prior to the sampling.
UNSTABLE values have one implementation restriction:
• Implementations of operations generating UNSTABLE results must not depend on any data source (memory or
internal state) which is inaccessible in the current processor mode
1.3 Special Symbols in Pseudocode Notation
In this book, algorithmic descriptions of an operation are described as pseudocode in a high-level language notation resembling Pascal. Special symbols used in the pseudocode notation are listed in Table 1.1.
Table 1.1 Symbols Used in Instruction Operation Statements
1.3 Special Symbols in Pseudocode Notation
Symbol
Meaning
←
Assignment
=, ≠
Tests for equality and inequality
||
Bit string concatenation
xy
A y-bit string formed by y copies of the single-bit value x
b#n
A constant value n in base b. For instance 10#100 represents the decimal value 100, 2#100 represents the binary value 100 (decimal 4), and 16#100 represents the hexadecimal value 100 (decimal 256). If the “b#” prefix is omitted, the default base is 10.
0bn
A constant value n in base 2. For instance 0b100 represents the binary value 100 (decimal 4).
0xn
A constant value n in base 16. For instance 0x100 represents the hexadecimal value 100 (decimal 256).
xy..z
Selection of bits y through z of bit string x. Little-endian bit notation (rightmost bit is 0) is used. If y is less than z, this expression is an empty (zero length) bit string.
+, −
2’s complement or floating point arithmetic: addition, subtraction
*, ×
2’s complement or floating point multiplication (both used for either)
div
2’s complement integer division
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 13
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
About This Book
Table 1.1 Symbols Used in Instruction Operation Statements (Continued)
Symbol
Meaning
mod
2’s complement modulo
/
Floating point division
<
2’s complement less-than comparison
>
2’s complement greater-than comparison
≤
2’s complement less-than or equal comparison
≥
2’s complement greater-than or equal comparison
nor
Bitwise logical NOR
xor
Bitwise logical XOR
and
Bitwise logical AND
or
Bitwise logical OR
GPRLEN
The length in bits (32 or 64) of the CPU general-purpose registers
GPR[x]
CPU general-purpose register x. The content of GPR[0] is always zero. In Release 2 of the Architecture, GPR[x] is a short-hand notation for SGPR[ SRSCtlCSS, x].
SGPR[s,x]
In Release 2 of the Architecture and subsequent releases, multiple copies of the CPU general-purpose regis- ters may be implemented. SGPR[s,x] refers to GPR set s, register x.
FPR[x]
Floating Point operand register x
FCC[CC]
Floating Point condition code CC. FCC[0] has the same value as COC[1].
FPR[x]
Floating Point (Coprocessor unit 1), general register x
CPR[z,x,s]
Coprocessor unit z, general register x, select s
CP2CPR[x]
Coprocessor unit 2, general register x
CCR[z,x]
Coprocessor unit z, control register x
CP2CCR[x]
Coprocessor unit 2, control register x
COC[z]
Coprocessor unit z condition signal
Xlat[x]
Translation of the MIPS16e GPR number x into the corresponding 32-bit GPR number
BigEndianMem
Endian mode as configured at chip reset (0 →Little-Endian, 1 → Big-Endian). Specifies the endianness of the memory interface (see LoadMemory and StoreMemory pseudocode function descriptions), and the endi- anness of Kernel and Supervisor mode execution.
BigEndianCPU
The endianness for load and store instructions (0 → Little-Endian, 1 → Big-Endian). In User mode, this endianness may be switched by setting the RE bit in the Status register. Thus, BigEndianCPU may be com- puted as (BigEndianMem XOR ReverseEndian).
ReverseEndian
Signal to reverse the endianness of load and store instructions. This feature is available in User mode only, and is implemented by setting the RE bit of the Status register. Thus, ReverseEndian may be computed as (SRRE and User mode).
LLbit
Bit of virtual state used to specify operation for instructions that provide atomic read-modify-write. LLbit is set when a linked load occurs and is tested by the conditional store. It is cleared, during other CPU operation, when a store to the location would no longer be atomic. In particular, it is cleared by exception return instruc- tions.
14 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 1.1 Symbols Used in Instruction Operation Statements (Continued)
1.3 Special Symbols in Pseudocode Notation
Symbol
Meaning
I:, I+n:, I-n:
This occurs as a prefix to Operation description lines and functions as a label. It indicates the instruction time during which the pseudocode appears to “execute.” Unless otherwise indicated, all effects of the current instruction appear to occur during the instruction time of the current instruction. No label is equivalent to a time label of I. Sometimes effects of an instruction appear to occur either earlier or later — that is, during the instruction time of another instruction. When this happens, the instruction operation is written in sections labeled with the instruction time, relative to the current instruction I, in which the effect of that pseudocode appears to occur. For example, an instruction may have a result that is not available until after the next instruction. Such an instruction has the portion of the instruction operation description that writes the result register in a section labeled I+1.
The effect of pseudocode statements for the current instruction labelled I+1 appears to occur “at the same time” as the effect of pseudocode statements labeled I for the following instruction. Within one pseudocode sequence, the effects of the statements take place in order. However, between sequences of statements for different instructions that occur “at the same time,” there is no defined order. Programs must not depend on a particular order of evaluation between such sections.
PC
The Program Counter value. During the instruction time of an instruction, this is the address of the instruc- tion word. The address of the instruction that occurs during the next instruction time is determined by assign- ing a value to PC during an instruction time. If no value is assigned to PC during an instruction time by any pseudocode statement, it is automatically incremented by either 2 (in the case of a 16-bit MIPS16e instruc- tion) or 4 before the next instruction time. A taken branch assigns the target address to the PC during the instruction time of the instruction in the branch delay slot.
In the MIPS Architecture, the PC value is only visible indirectly, such as when the processor stores the restart address into a GPR on a jump-and-link or branch-and-link instruction, or into a Coprocessor 0 register on an exception. The PC value contains a full 32-bit address all of which are significant during a memory ref- erence.
ISA Mode
In processors that implement the MIPS16e Application Specific Extension or the microMIPS base architec- tures, the ISA Mode is a single-bit register that determines in which mode the processor is executing, as fol- lows:
In the MIPS Architecture, the ISA Mode value is only visible indirectly, such as when the processor stores a combined value of the upper bits of PC and the ISA Mode into a GPR on a jump-and-link or branch-and-link instruction, or into a Coprocessor 0 register on an exception.
Encoding
Meaning
0
The processor is executing 32-bit MIPS instructions
1
The processor is executing MIIPS16e instructions
PABITS
The number of physical address bits implemented is represented by the symbol PABITS. As such, if 36 physical address bits were implemented, the size of the physical address space would be 2PABITS = 236 bytes.
FP32RegistersMode
Indicates whether the FPU has 32-bit or 64-bit floating point registers (FPRs). In MIPS32 Release 1, the FPU has 32 32-bit FPRs in which 64-bit data types are stored in even-odd pairs of FPRs. In MIPS64, (and option- ally in MIPS32 Release2 and MIPSr3) the FPU has 32 64-bit FPRs in which 64-bit data types are stored in any FPR.
In MIPS32 Release 1 implementations, FP32RegistersMode is always a 0. MIPS64 implementations have a compatibility mode in which the processor references the FPRs as if it were a MIPS32 implementation. In such a case FP32RegisterMode is computed from the FR bit in the Status register. If this bit is a 0, the pro- cessor operates as if it had 32 32-bit FPRs. If this bit is a 1, the processor operates with 32 64-bit FPRs. The value of FP32RegistersMode is computed from the FR bit in the Status register.
InstructionInBranchDe- laySlot
Indicates whether the instruction at the Program Counter address was executed in the delay slot of a branch or jump. This condition reflects the dynamic state of the instruction, not the static state. That is, the value is false if a branch or jump occurs to an instruction whose PC immediately follows a branch or jump, but which is not executed in the delay slot of a branch or jump.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 15
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
About This Book
Table 1.1 Symbols Used in Instruction Operation Statements (Continued)
Symbol
Meaning
SignalException(excep- tion, argument)
Causes an exception to be signaled, using the exception parameter as the type of exception and the argument parameter as an exception-specific argument). Control does not return from this pseudocode function—the exception is signaled at the point of the call.
1.4 For More Information
Various MIPS RISC processor manuals and additional information about MIPS products can be found at the MIPS URL: http://www.mips.com
For comments or questions on the MIPS32® Architecture or this document, send Email to support@mips.com.
16 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Chapter 2
The MIPS Architecture: An Introduction
2.1 MIPS Instruction Set Overview 2.1.1 Historical Perspective
The MIPS® Instruction Set Architecture (ISA) has evolved over time from the original MIPS ITM ISA, through the MIPS VTM ISA, to the current MIPS32®, MIPS64® and microMIPSTM Architectures. As the ISA evolved, all exten- sions have been backward compatible with previous versions of the ISA. In the MIPS IIITM level of the ISA, 64-bit integers and addresses were added to the instruction set. The MIPS IVTM and MIPS VTM levels of the ISA added improved floating point operations, as well as a set of instructions intended to improve the efficiency of generated code and of data movement. Because of the strict backward-compatible requirement of the ISA, such changes were unavailable to 32-bit implementations of the ISA which were, by definition, MIPS ITM or MIPS IITM implementations.
While the user-mode ISA was always backward compatible, the privileged environment was allowed to change on a per-implementation basis. As a result, the R3000® privileged environment was different from the R4000® privileged environment, and subsequent implementations, while similar to the R4000 privileged environment, included subtle differences. Because the privileged environment was never part of the MIPS ISA, an implementation had the flexibil- ity to make changes to suit that particular implementation. Unfortunately, this required kernel software changes to every operating system or kernel environment on which that implementation was intended to run.
Many of the original MIPS implementations were targeted at computer-like applications such as workstations and servers. In recent years MIPS implementations have had significant success in embedded applications. Today, most of the MIPS parts that are shipped go into some sort of embedded application. Such applications tend to have differ- ent trade-offs than computer-like applications including a focus on cost of implementation, and performance as a function of cost and power.
The MIPS32 and MIPS64 Architectures are intended to address the need for a high-performance but cost-sensitive MIPS instruction set. The MIPS32 Architecture is based on the MIPS II ISA, adding selected instructions from MIPS III, MIPS IV, and MIPS V to improve the efficiency of generated code and of data movement. The MIPS64 Architec- ture is based on the MIPS V ISA and is backward compatible with the MIPS32 Architecture. Both the MIPS32 and MIPS64 Architectures bring the privileged environment into the Architecture definition to address the needs of oper- ating systems and other kernel software. Both also include provision for adding MIPS Application Specific Exten- sions (ASEs), User Defined Instructions (UDIs), and custom coprocessors to address the specific needs of particular markets.
MIPS32 and MIPS64 Architectures provides a substantial cost/performance advantage over microprocessor imple- mentations based on traditional architectures. This advantage is a result of improvements made in several contiguous disciplines: VLSI process technology, CPU organization, system-level architecture, and operating system and com- piler design.
The microMIPS32 and microMIPS64 Architectures deliver the same functionality of MIPS32 and MIPS64 with the additional benefit of smaller codesizes. The microMIPS architectures are supersets of MIPS32/MIPS64 architectures, with almost the same sets of 32-bit sized instructions and additional 16-bit instructions to help with codesize. micro- MIPS is especially compelling for systems in which the cost of memory dominate the entire bill of materials cost.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 17
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Unlike the earlier versions of the architectures, microMIPS supplies assembler-source code compatibility with its predecessors instead of binary compatibility.
Figure 2-1 MIPS Architectures
32-bit Address & Data Handling MIPS I
MIPS II
64-bit Address & Data Handling
MIPS III
MIPS IV
MIPS V
MIPS64 Release 1
MIPS64 Release 2
MIPS32 Release 1
MIPS32 Release 2
MIPS32 Release 3 microMIPS32
2.1.2 Architectural Evolution
Release 1
Release 2
18
MIPSr3TM
MIPS64 Release 3 microMIPS64
The evolution of an architecture is a dynamic process that takes into account both the need to provide a stable plat- form for implementations, as well as new market and application areas that demand new capabilities. Enhancements to an architecture are appropriate when they:
• • • • •
are applicable to a wide market
provide long-term benefit
maintain architectural scalability
are standardized to prevent fragmentation are a superset of the existing architecture
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture community constantly evaluates suggestions for architectural changes and enhancements against these criteria. New releases of the architecture, while infrequent, are made at appropriate points, following these criteria. At present, there are three releases of the MIPS Architecture: Release 1 (the original version of the MIPS32 Architecture) ; Release 2 which was added in 2002 and Release 3 (called MIPSr3TM) which was added in 2010.
2.1.2.1 Release 2 of the MIPS32 Architecture
Enhancements included in Release 2 of the MIPS32 Architecture are:
• Vectored interrupts: This enhancement provides the ability to vector interrupts directly to a handler for that inter- rupt. Vectored interrupts are an option in Release 2 implementations and the presence of that option is denoted by the Config3VInt bit.
• Support for an external interrupt controller: This enhancement reconfigures the on-core interrupt logic to take full advantage of an external interrupt controller. This support is an option in Release 2 implementations and the presence of that option is denoted by the Config3EIC bit.
• Programmable exception vector base: This enhancement allows the base address of the exception vectors to be moved for exceptions that occur when StatusBEV is 0. Doing so allows multi-processor systems to have separate exception vectors for each processor, and allows any system to place the exception vectors in memory that is appropriate to the system environment. This enhancement is required in a Release 2 implementation.
• Atomic interrupt enable/disable: Two instructions have been added to atomically enable or disable interrupts, and return the previous value of the Status register. These instructions are required in a Release 2 implementation.
• The ability to disable the Count register for highly power-sensitive applications. This enhancement is required in a Release 2 implementation.
• GPR shadow registers: This addition provides the addition of GPR shadow registers and the ability to bind these registers to a vectored interrupt or exception. Shadow registers are an option in Release 2 implementations and the presence of that option is denoted by a non-zero value in SRSCtlHSS. While shadow registers are most useful when either vectored interrupts or support for an external interrupt controller is also implemented, neither is required.
• Field, Rotate and Shuffle instructions: These instructions add additional capability in processing bit fields in reg- isters. These instructions are required in a Release 2 implementation.
• Explicit hazard management: This enhancement provides a set of instructions to explicitly manage hazards, in place of the cycle-based SSNOP method of dealing with hazards. These instructions are required in a Release 2 implementation.
• Access to a new class of hardware registers and state from an unprivileged mode. This enhancement is required in a Release 2 implementation.
• Coprocessor 0 Register changes: These changes add or modify CP0 registers to indicate the existence of new and optional state, provide L2 and L3 cache identification, add trigger bits to the Watch registers, and add support for 64-bit performance counter count registers. This enhancement is required in a Release 2 implementation.
• Support for 64-bit coprocessors with 32-bit CPUs: These changes allow a 64-bit coprocessor (including an FPU) to be attached to a 32-bit CPU. This enhancement is optional in a Release 2 implementation.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 19
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.1 MIPS Instruction Set Overview
The MIPS Architecture: An Introduction
20
• New Support for Virtual Memory: These changes provide support for a 1KByte page size. This change is optional in Release 2 implementations, and support is denoted by Config3SP.
2.1.2.2 Releases 2.5+ of the MIPS32 Architecture
Some optional features were added after Revision 2.5:
• Support for a MMU with more than 64 TLB entries. This feature aids in reducing the frequency of TLB misses.
• Scratch registers within Coprocessor0 for kernel mode software. This feature aids in quicker exception handling by not requiring the saving of usermode registers onto the stack before kernelmode software uses those registers.
• A MMU configuration which supports both larger set-associative TLBs and variable page-sizes. This feature aids in reducing the frequency of TLB misses.
• The CDMM memory scheme for the placement of small I/O devices into the physical address space. This scheme allows for efficient placement of such I/O devices into a small memory region.
• An EIC interrupt mode where the EIC controller supplies a 16-bit interrupt vector. This allows different inter- rupts to share code.
• The PAUSE instruction to deallocate a (virtual) processor when arbitration for a lock doesn’t succeed. This allows for lower power consumption as well as lower snoop traffic when multiple (virtual) processors are arbi- trating for a lock.
• More flavors of memory barriers that are available through stype field of the SYNC instruction. The newer mem- ory barriers attempt to minimize the amount of pipeline stalls while doing memory synchronization operations.
2.1.2.3 MIPSr3TM Architecture
MIPSr3TM is a family of architectures which includes Release 3.0 of the MIPS32 Architecture as well as the first release of the microMIPS32 architecture.
Enhancements included in MIPSr3TM Architecture are:
•
The •
•
•
• • •
•
microMIPSTM instruction set.
This instruction set contains both 16-bit and 32-bit sized instructions.
This mixed size ISA has all of the functionality of MIPS32 while also delivering smaller code sizes. microMIPS is assembler source code compatible with MIPS32.
microMIPS replaces the MIPS16eTM ASE.
microMIPS is an additional base instruction set architecture that is supported along with MIPS32.
A device can implement either base ISA or both. The ISA field of Config3 denotes which ISA is imple- mented.
A device can implement any other ASE with either base architecture.1
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• microMIPS shares the same privileged resource architecture with MIPS32.
• Branch Likely instructions are not supported in the microMIPS hardware architecture. Instead the micro- MIPS toolchain replaces these instructions with equivalent code sequences.
• A more flexible version of the Context Register that can point to any power-of-two sized data structure. This optional feature is denoted by CTXTC field of Config3.
• Additional protection bits in the TLB entries that allow for non-executable and write-only virtual pages. This optional feature is denoted by RXI field of Config3.
2.1.3 Architectural Changes Relative to the MIPS I through MIPS V Architectures
In addition to the MIPS Architecture described in this document set, the following changes were made to the architec- ture relative to the earlier MIPS RISC Architecture Specification, which describes the MIPS I through MIPS V Architectures.
• The MIPS IV ISA added a restriction to the load and store instructions which have natural alignment require- ments (all but load and store byte and load and store left and right) in which the base register used by the instruc- tion must also be naturally aligned (the restriction expressed in the MIPS RISC Architecture Specification is that the offset be aligned, but the implication is that the base register is also aligned, and this is more consistent with the indexed load/store instructions which have no offset field). The restriction that the base register be naturally- aligned is eliminated by the MIPS32 Architecture, leaving the restriction that the effective address be naturally- aligned.
• Early MIPS implementations required two instructions separating a MFLO or MFHI from the next integer multi- ply or divide operation. This hazard was eliminated in the MIPS IV ISA, although the MIPS RISC Architecture Specification does not clearly explain this fact. The MIPS32 Architecture explicitly eliminates this hazard and requires that the hi and lo registers be fully interlocked in hardware for all integer multiply and divide instruc- tions (including, but not limited to, the MADD, MADDU, MSUB, MSUBU, and MUL instructions introduced in this specification).
• The Implementation and Programming Notes included in the instruction descriptions for the madd, maddu, msub, msubu, and mul instructions should also be applied to all integer multiply and divide instructions in the MIPS RISC Architecture Specification.
2.2 Compliance and Subsetting
To be compliant with the MIPS32 Architecture, designs must implement a set of required features, as described in this document set. To allow flexibility in implementations, the MIPS32 Architecture does provide subsetting rules. An implementation that follows these rules is compliant with the MIPS32 Architecture as long as it adheres strictly to the rules, and fully implements the remaining instructions.Supersetting of the MIPS32 Architecture is only allowed by adding functions to the SPECIAL2 major opcode, by adding control for co-processors via the COP2, LWC2, SWC2, LDC2, and/or SDC2, or via the addition of approved Application Specific Extensions.
Note: The use of COP3 as a customizable coprocessor has been removed in the Release 2 of the MIPS32 architecture. The use of the COP3 is now reserved for the future extension of the architecture. Implementations using Release1 of the MIPS32 architecture are strongly discouraged from using the COP3 opcode for a user-available coprocessor as doing so will limit the potential for an upgrade path to a 64-bit floating point unit.
1. Except for MIPS16e.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 21
2.2 Compliance and Subsetting
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
22
The •
•
instruction set subsetting rules are as follows:
All CPU instructions must be implemented – no subsetting is allowed.
The FPU and related support instructions, including the MOVF and MOVT CPU instructions, may be omitted. Software may determine if an FPU is implemented by checking the state of the FP bit in the Config1 CP0 regis- ter. If the FPU is implemented, it must include S, D, and W formats, operate instructions, and all supporting instructions. Software may determine which FPU data types are implemented by checking the appropriate bit in the FIR CP1 register. The following allowable FPU subsets are compliant with the MIPS32 architecture:
• No FPU
• FPU with S, D, and W formats and all supporting instructions
• FPU with S, D, PS, and W formats and all supporting instructions
Coprocessor 2 is optional and may be omitted. Software may determine if Coprocessor 2 is implemented by checking the state of the C2 bit in the Config1 CP0 register. If Coprocessor 2 is implemented, the Coprocessor 2 interface instructions (BC2, CFC2, COP2, CTC2, LDC2, LWC2, MFC2, MTC2, SDC2, and SWC2) may be omitted on an instruction-by-instruction basis.
Supervisor Mode is optional. If Supervisor Mode is not implemented, bit 3 of the Status register must be ignored on write and read as zero.
The standard TLB-based memory management unit may be replaced with:
• a simpler MMU (e.g., a Fixed Mapping MMU or a Block Address Translation MMU or a Base-Bounds MMU).
• The Dual TLB MMU – (e.g. FTLB and VTLB MMU described in the Alternative MMU Organizations Appendix of Volume III)
If this is done, the rest of the interface to the Privileged Resource Architecture must be preserved. Software may determine the type of the MMU by checking the MT field in the Config CP0 register.
The Privileged Resource Architecture includes several implementation options and may be subsetted in accor- dance with those options. An incomplete list of these options include:
•
• •
•
• • • • • • • •
Interrupt Modes
Shadow Register Sets Common Device Memory Map Parity/ECC support
UserLocal register ContextConfig register PageGrain register
Config1-4 registers
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• Performance Counter, WatchPoint and Trace Registers
• Cache control/diagnostic registers
• Kernelmode scratch registers
• Instruction, CP0 Register, and CP1 Control Register fields that are marked “Reserved” or shown as “0” in the description of that field are reserved for future use by the architecture and are not available to implementations. Implementations may only use those fields that are explicitly reserved for implementation dependent use.
• Supported ASEs are optional and may be subsetted out. If most cases, software may determine if a supported ASE is implemented by checking the appropriate bit in the Config1 or Config3 CP0 register. If they are imple- mented, they must implement the entire ISA applicable to the component, or implement subsets that are approved by the ASE specifications.
• EJTAG is optional and may be subsetted out. If it is implemented, it must implement only those subsets that are approved by the EJTAG specification.
• If any instruction is subsetted out based on the rules above, an attempt to execute that instruction must cause the appropriate exception (typically Reserved Instruction or Coprocessor Unusable).
• In MIPSr3 (also called Release 3), there are two architecture branches (MIPS32/64 and microMIPS32/64). A single device is allowed to implement both architecture branches. The Privileged Resource Architecture (COP0) registers do not mode-switch in width (32-bit vs. 64-bit). For this reason, if a device implements both architec- ture branches, the address/data widths must be consistent. If a device implements MIPS64 and also implements microMIPS, it must implement microMIPS64 not just microMIPS32. Simiarly, If a device implements microMIPS64 and also implements MIPS32/64, it must implement MIPS64 not just MIPS32.
• If both of the architecture branches are implemented (MIPS32/64 and microMIPS32/64) or if MIPS16e is imple- mented then the JALX instructions are required. If only one branch of the architecture family and MIPS16e is not implemented then the JALX instruction is not implemented. That is, the JALX instruction is required if and only if when ISA mode-switching is possible.
2.3 Components of the MIPS Architecture 2.3.1 MIPS Instruction Set Architecture (ISA)
The MIPS32 and MIPS64 Instruction Set Architectures define a compatible family of instructions dealing with 32-bit data and 64-bit data (respectively) within the framework of the overall MIPS Architectures. Included in the ISA are all instructions, both privileged and unprivileged, by which the programmer interfaces with the processor. The ISA guarantees object code compatibility for unprivileged and, often, privileged programs executing on any MIPS32 or MIPS64 processor; all instructions in the MIPS64 ISA are backward compatible with those instructions in the MIPS32 ISA. Using conditional compilation or assembly language macros, it is often possible to write privileged programs that run on both MIPS32 and MIPS64 implementations.
2.3.2 MIPS Privileged Resource Architecture (PRA)
The MIPS32 and MIPS64 Privileged Resource Architecture defines a set of environments and capabilities on which the ISA operates. The effects of some components of the PRA are visible to unprivileged programs; for instance, the virtual memory layout. Many other components are visible only to privileged programs and the operating system. The
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 23
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.3 Components of the MIPS Architecture
The MIPS Architecture: An Introduction
PRA provides the mechanisms necessary to manage the resources of the processor: virtual memory, caches, excep- tions, user contexts, etc.
2.3.3 MIPS Application Specific Extensions (ASEs)
The MIPS32 and MIPS64 Architectures provide support for optional application specific extensions. As optional extensions to the base architecture, the ASEs do not burden every implementation of the architecture with instructions or capability that are not needed in a particular market. An ASE can be used with the appropriate ISA and PRA to meet the needs of a specific application or an entire class of applications.
2.3.4 MIPS User Defined Instructions (UDIs)
In addition to support for ASEs as described above, the MIPS32 and MIPS64 Architectures define specific instruc- tions for the use of each implementation. The Special2 instruction function fields and Coprocessor 2 are reserved for capability defined by each implementation.
2.4 Architecture Versus Implementation
When describing the characteristics of MIPS processors, architecture must be distinguished from the hardware implementation of that architecture.
• Architecture refers to the instruction set, registers and other state, the exception model, memory management, virtual and physical address layout, and other features that all hardware executes.
• Implementation refers to the way in which specific processors apply the architecture.
Here are two examples:
1. A floating point unit (FPU) is an optional part of the MIPS32 Architecture. A compatible implementation of the FPU may have different pipeline lengths, different hardware algorithms for performing multiplication or divi- sion, etc.
2. Most MIPS processors have caches; however, these caches are not implemented in the same manner in all MIPS processors. Some processors implement physically-indexed, physically tagged caches. Other implement virtu- ally-indexed, physically-tagged caches. Still other processor implement more than one level of cache.
The MIPS32 architecture is decoupled from specific hardware implementations, leaving microprocessor designers free to create their own hardware designs within the framework of the architectural definition.
2.5 Relationship between the MIPSr3 Architectures
The MIPS Architectures evolved as a compromise between software and hardware resources. The MIPS has a family of related architectures. Within each “branch of the family”, the architecture guarantees object-code compatibility for User-Mode programs executed on any MIPS processor.
MIPS32 and MIPS64 form one branch of the architecture family. In User Mode MIPS64 processors are backward- compatible with their MIPS32 predecessors. As such, the MIPS32 Architecture is a strict subset of the MIPS64 Architecture.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
24
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Similarly, microMIPS32 and microMIPS64 form another branch of the architecture family. In User Mode microMIPS64 processors are backward-compatible with their microMIPS predecessors. As such, the microMIPS Architecture is a strict subset of the MIPS64 Architecture.
The relationship between the binary representations of the architectures is shown in Figure 2-2. Figure 2-2 Relationship of the Binary Representations of MIPSr3 Architectures
microMIPS32/64 is not binary compatible with MIPS32/64
2.5 Relationship between the MIPSr3 Architectures
microMIPS64 is binary compatible with microMIPS32
MIPS64 is binary compatible with MIPS32
MIPS64
MIPS32
MIPS32 is proper subset of MIPS64
microMIPS64
microMIPS32
instructions dealing with
64-bit
data
microMIPS32 is proper subset of microMIPS64
As of 2010, there are two branches of the architecture family – the MIPS32/64 branch and the microMIPS32/64 branch. For these two branches, some levels of compatibility are available:
1. The microMIPS32/64 branch supplies a superset of the functionality that is available from the MIPS32/64 branch. The additional functionality that the microMIPS branch delivers is smaller code size.
2. It is allowed for implementations to implement both branches of the architecture family for compatibility rea- sons. For such implementations, the architectures define methods of switching from one instruction set to the other. This allows one binary program to use both instruction sets or call a library that is using the other instruc- tion set.
3. At the assembler source code level, the two architecture branches are fully compatible. That is, all of the MIPS32/64 assembler instruction mnemonics and directives are fully usable and understood by the microMIPS32/64 toolchains.
The relationships between the assembler source-code representations of the architectures is shown in Figure 2-3.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 25
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Figure 2-3 Relationships of the Assembler Source Code Representations of the MIPSr3 Architectures
MIPS64
MIPS32
microMIPS64
microMIPS32
Note 1
instructions dealing with
64-bit
data
Note 1 – microMIPS toolchain emulates branch-likely instrs
2.6 Pipeline Architecture
26
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
This section describes the basic pipeline architecture, along with two types of improvements: superpipelines and superscalar pipelines. (Pipelining and multiple issuing are not defined by the ISA, but are implementation dependent.)
2.6.1 Pipeline Stages and Execution Rates
MIPS processors all use some variation of a pipeline in their architecture. A pipeline is divided into the following dis- crete parts, or stages, shown in Figure 2-4:
• • • •
Fetch
Arithmetic operation Memory access Write back
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
16-bit & 32-bit instructions
for smaller code size
Figure 2-4 One-Deep Single-Completion Instruction Pipeline
2.6 Pipeline Architecture
Fetch
ALU
Memory
Write
Instruction 1
Cycle 1
Stage 1
Cycle 2 Cycle 3
Stage 2 Stage 3
Execution Rate
Cycle 4
Stage 4 Instruction 2
Cycle 5 Cycle 6 Instruction completion
Cycle 7
Cycle 3
Stage 3
Cycle 8
Stage 4
In the example shown in Figure 2-4, each stage takes one processor clock cycle to complete. Thus it takes four clock cycles (ignoring delays or stalls) for the instruction to complete. In this example, the execution rate of the pipeline is one instruction every four clock cycles. Conversely, because only a single execution can be fetched before comple- tion, only one stage is active at any time.
2.6.2 Parallel Pipeline
Figure 2-5 illustrates a remedy for the latency (the time it takes to execute an instruction) inherent in the pipeline shown in Figure 2-4.
Instead of waiting for an instruction to be completed before the next instruction can be fetched (four clock cycles), a new instruction is fetched each clock cycle. There are four stages to the pipeline so the four instructions can be exe- cuted simultaneously, one at each stage of the pipeline. It still takes four clock cycles for the first instruction to be completed; however, in this theoretical example, a new instruction is completed every clock cycle thereafter. Instruc- tions in Figure 2-5 are executed at a rate four times that of the pipeline shown in Figure 2-4.
Cycle 1
Instruction 1
Cycle 2
Instruction 2
Cycle 3
Cycle 4
Cycle 5
Cycle 6
Cycle 7
Figure 2-5 Four-Deep Single-Completion Pipeline
Fetch
Stage 1
Stage 2
ALU
Memory
Write
Fetch
ALU
Memory
Write
Fetch
ALU
Memory
Write
Fetch
ALU
Memory
Write
2.6.3 Superpipeline
Instruction 3
Fetch
ALU
Memory
Write
Instruction 4
Figure 2-6 shows a superpipelined architecture. Each stage is designed to take only a fraction of an external clock cycle—in this case, half a clock. Effectively, each stage is divided into more than one substage. Therefore more than one instruction can be completed each cycle.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 27
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Figure 2-6 Four-Deep Superpipeline
Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5 Cycle 6 Cycle 7 Cycle 8 Phase 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
Clock
Fetch
ALU
Mem
Write
Fetch
ALU
Mem
Fetch
ALU
Mem
Write
Fetch A
Write
LU M
em
Write
Fetch A
LU
Mem
Write
Fetch
ALU
Mem
2.6.4 Superscalar Pipeline
A superscalar architecture also allows more than one instruction to be completed each clock cycle. Figure 2-7 shows a four-way, five-stage superscalar pipeline.
Figure 2-7 Four-Way Superscalar Pipeline
Write
Fetch
ALU
Mem
Write
Fetch A
LU M
em
Write
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
Instruction 1 Instruction 2 Instruction 3 Instruction 4
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
IF
ID
IS
EX
WB
Instruction 5 Instruction 6 Instruction 7 Instruction 8
Four-way
IF = instruction fetch
ID = instruction decode and dependency IS = instruction issue
EX = execution
WB = write back
2.7 Load/Store Architecture
28
Generally, it takes longer to perform operations in memory than it does to perform them in on-chip registers. This is because of the difference in time it takes to access a register (fast) and main memory (slower).
To eliminate the longer access time, or latency, of in-memory operations, MIPS processors use a load/store design. The processor has many registers on chip, and all operations are performed on operands held in these processor regis- ters. Main memory is accessed only through load and store instructions. This has several benefits:
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Five-stage
• Reducing the number of memory accesses, easing memory bandwidth requirements
• Simplifying the instruction set
• Making it easier for compilers to optimize register allocation
2.8 Programming Model
This section describes the following aspects of the programming model:
• CPU Data Formats
• Coprocessors (CP0-CP3)
• CPU Registers
• FPU Data Formats
• Byte Ordering and Endianness
• Memory Access Types
2.8.1 CPU Data Formats
The CPU defines the following data formats:
• Bit (b)
• Byte (8 bits, B)
• Halfword (16 bits, H)
• Word (32 bits, W)
• Doubleword (64 bits, D)2 2.8.2 FPU Data Formats
The FPU defines the following data formats:
• 32-bit single-precision floating point (.fmt type S)
• 32-bit single-precision floating point paired-single (.fmt type PS)2
• 64-bit double-precision floating point (.fmt type D)
• 32-bit Word fixed point (.fmt type W)
2. The CPU Doubleword and FPU floating point paired-single and Long fixed point data formats are available in a Release 1 implementation of the MIPS64 Architecture, or in a Release 2 (or subsequent releases) implementation that includes a 64-bit floating point unit
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 29
2.8 Programming Model
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
30
• 64-bit Long fixed point (.fmt type L)2 2.8.3 Coprocessors (CP0-CP3)
The MIPS Architecture defines four coprocessors (designated CP0, CP1, CP2, and CP3):
• Coprocessor 0 (CP0) is incorporated on the CPU chip and supports the virtual memory system and exception
handling. CP0 is also referred to as the System Control Coprocessor.
• Coprocessor 1 (CP1) is reserved for the floating point coprocessor, the FPU.
• Coprocessor 2 (CP2) is available for specific implementations.
• Coprocessor 3 (CP3) is reserved for the floating point unit in a Release 1 implementation of the MIPS64 Archi- tecture, and on all Release 2 (and subsequent releases) implementations of the Architecture.
CP0 translates virtual addresses into physical addresses, manages exceptions, and handles switches between kernel, supervisor, and user states. CP0 also controls the cache subsystem, as well as providing diagnostic control and error recovery facilities. The architectural features of CP0 are defined in Volume III.
2.8.4 CPU Registers
The MIPS32 Architecture defines the following CPU registers:
• 32 32-bit general purpose registers (GPRs)
• a pair of special-purpose registers to hold the results of integer multiply, divide, and multiply-accumulate opera- tions (HI and LO)
• a special-purpose program counter (PC), which is affected only indirectly by certain instructions – it is not an architecturally-visible register.
2.8.4.1 CPU General-Purpose Registers
Two of the CPU general-purpose registers have assigned functions:
• r0 is hard-wired to a value of zero, and can be used as the target register for any instruction whose result is to be
discarded. r0 can also be used as a source when a zero value is needed.
• r31 is the destination register used by JAL, BLTZAL, BLTZALL, BGEZAL, and BGEZALL without being
explicitly specified in the instruction word. Otherwise r31 is used as a normal register.
The remaining registers are available for general-purpose use.
2.8.4.2 CPU Special-Purpose Registers
The CPU contains three special-purpose registers:
• •
PC—Program Counter register
HI—Multiply and Divide register higher result
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• LO—Multiply and Divide register lower result
• During a multiply operation, the HI and LO registers store the product of integer multiply.
• During a multiply-add or multiply-subtract operation, the HI and LO registers store the result of the integer multiply-add or multiply-subtract.
• During a division, the HI and LO registers store the quotient (in LO) and remainder (in HI) of integer divide.
• During a multiply-accumulate, the HI and LO registers store the accumulated result of the operation. Figure 2-8 shows the layout of the CPU registers.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 31
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.8 Programming Model
The MIPS Architecture: An Introduction
Figure 2-8 CPU Registers
31 031 0
r0 (hardwired to zero)
r1
r2
r3
r4
r5
r6
r7
r8
r9
r10
r11
r12
r13
r14
r15
r16
r17
r18
r19
r20
r21
r22
r23
r24
r25
r26
r27
r28
r29
r30
r31
HI
LO
31 0
General Purpose Registers Special Purpose Registers
PC
32
2.8.5 FPU Registers
The MIPS32 Architecture defines the following FPU registers:
• 32 floating point registers (FPRs). These registers are 32 bits wide in a 32-bit FPU and 64 bits wide on a 64-bit FPU.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• Five FPU control registers are used to identify and control the FPU.
• Eight floating point condition codes that are part of the FCSR register
In Release 1 of the Architecture, 64-bit floating point units were supported only by implementations of the MIPS64 Architecture. Similarly, implementations of MIPS32 of the Architecture only supported 32-bit floating point units. In Release 2 of the Architecture and subsequent releases, a 64-bit floating point unit is optional on implementations of both the MIPS32 and MIPS64 Architectures.
A 32-bit floating point unit contains 32 32-bit FPRs, each of which is capable of storing a 32-bit data type. Double- precision (type D) data types are stored in even-odd pairs of FPRs, and the long-integer (type L) and paired single (type PS) data types are not supported. Figure 2-9 shows the layout of these registers.
A 64-bit floating point unit contains 32 64-bit FPRs, each of which is capable of storing any data type. For compati- bility with 32-bit FPUs, the FR bit in the CP0 Status register is used by a MIPS64 Release 1, or any Release 2 (or subsequent releases) processor that supports a 64-bit FPU to configure the FPU in a mode in which the FPRs are treated as 32 32-bit registers, each of which is capable of storing only 32-bit data types. In this mode, the double-pre- cision floating point (type D) data type is stored in even-odd pairs of FPRs, and the long-integer (type L) and paired single (type PS) data types are not supported.
Figure 2-10 shows the layout of the FPU Registers when the FR bit in the CP0 Status register is 1; Figure 2-11 shows the layout of the FPU Registers when the FR bit in the CP0 Status register is 0.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 33
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.8 Programming Model
The MIPS Architecture: An Introduction
Figure 2-9 FPU Registers for a 32-bit FPU
31 0
f0
f1
f2
f3
f4
f5
f6
f7
f8
f9
f10
f11
f12
f13
f14
f15
f16
f17
f18
f19
f20
f21
f22
f23
f24
f25
f26
f27
f28
f29
f30
f31
31 0
FIR
FCCR
FEXR
FENR
FCSR
General Purpose Registers Special Purpose Registers
34 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Figure 2-10 FPU Registers for a 64-bit FPU if StatusFR is 1
2.8 Programming Model
63 32 31 0
f0
f1
f2
f3
f4
f5
f6
f7
f8
f9
f10
f11
f12
f13
f14
f15
f16
f17
f18
f19
f20
f21
f22
f23
f24
f25
f26
f27
f28
f29
f30
f31
FIR
FCCR
FEXR
FENR
FCSR
General Purpose Registers
31 0
Special Purpose Registers
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 35
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Figure 2-11 FPU Registers for a 64-bit FPU if StatusFR is 0
63 32 31 0
f0
f1
f2
f3
f4
f5
f6
f7
f8
f9
f10
f11
f12
f13
f14
f15
f16
f17
f18
f19
f20
f21
f22
f23
f24
f25
f26
f27
f28
f29
f30
f31
FCR0
FCR25
FCR26
FCR28
FCSR
General Purpose Registers
31 0
Special Purpose Registers
36 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
UNPREDICTABLE
2.8.6 Byte Ordering and Endianness
Bytes within larger CPU data formats—halfword, word, and doubleword—can be configured in either big-endian or little-endian order, as described in the following subsections:
• Big-Endian Order
• Little-Endian Order
• MIPS Bit Endianness
Endianness defines the location of byte 0 within a larger data structure (in this book, bits are always numbered with 0 on the right). Figures 2-12 and 2-13 show the ordering of bytes within words and the ordering of words within mul- tiple-word structures for both big-endian and little-endian configurations.
2.8.6.1 Big-Endian Order
When configured in big-endian order, byte 0 is the most-significant (left-hand) byte. Figure 2-12 shows this config- uration.
Higher Word Address Address
12 8 4
Lower 0 Address
2.8.6.2 Little-Endian Order
Figure 2-12 Big-Endian Byte Ordering
Bit #
2.8 Programming Model
31 2423 1615 87 0
12
13
14
15
8
9
10
11
4
5
6
7
0
1
2
3
1 word = 4 bytes
When configured in little-endian order, byte 0 is always the least-significant (right-hand) byte. Figure 2-13 shows this configuration.
Figure 2-13 Little-Endian Byte Ordering
Higher Word Address Address
12 8 4 0
Bit #
31 2423 1615 87 0
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
Lower Address
2.8.6.3 MIPS Bit Endianness
In this book, bit 0 is always the least-significant (right-hand) bit. Although no instructions explicitly designate bit positions within words, MIPS bit designations are always little-endian.
2-14 shows big-endian and 2-15 shows little-endian byte ordering in doublewords.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 37
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Most-significant byte
Least-significant byte 24 23 16 15 8 7 0
Byte
Figure 2-14 Big-Endian Data in Doubleword Format
Word
Bit # 63 Byte #
56 55
48 47 40 39
Halfword
32 31
0
1
2
3
4
5
7
Most-significant byte
Least-significant byte 24 23 16 15 8 7 0
Byte
Bit # 7 6 5 4 3 2 1 0 Bits in a byte
Bit # 7 6 5 4 3 2 1 0 Bits in a byte
Figure 2-15 Little-Endian Data in Doubleword Format
Word
Bit # 63 Byte #
56 55
48 47 40 39
Halfword
32 31
7
6
5
4
3
2
1
0
Alignment
Instructions
Instruction Set
Word
LWL, LWR, SWL, SWR
MIPS32 ISA
Doubleword
LDL, LDR, SDL, SDR
MIPS64 ISA
38
The CPU uses byte addressing for halfword, word, and doubleword accesses with the following alignment con- straints:
• Halfword accesses must be aligned on an even byte boundary (0, 2, 4…).
• Word accesses must be aligned on a byte boundary divisible by four (0, 4, 8…).
• Doubleword accesses must be aligned on a byte boundary divisible by eight (0, 8, 16…).
2.8.6.5 Unaligned Loads and Stores
The following instructions load and store words that are not aligned on word (W) or doubleword (D) boundaries:
Table 2.1 Unaligned Load and Store Instructions
2-16 show a big-endian access of a misaligned word that has byte address 3, and 2-17 shows a little-endian access of a misaligned word that has byte address 1.3
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
2.8.6.4 Addressing Alignment Constraints
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
6
Figure 2-16 Big-Endian Misaligned Word Addressing
Higher Address
Lower Address
31 24 23
Bit # 16 15
8 7 0
2.8 Programming Model
4
5
6
3
Higher Address
Lower Address
2.8.7 Memory Access Types
Bit #
Figure 2-17 Little-Endian Misaligned Word Addressing
31 24 23 16 15 8 7 0
4
3
2
1
3.
MIPS systems provide several memory access types. These are characteristic ways to use physical memory and caches to perform a memory access.
The memory access type is identified by the Cacheability and Coherency Attribute (CCA) bits in the TLB entry for each mapped virtual page. The access type used for a location is associated with the virtual address, not the physical address or the instruction making the reference. Memory access types are available for both uniprocessor and multi- processor (MP) implementations.
All implementations must provide the following memory access types:
• Uncached
• Cached
These memory access types are described in the following sections:
• Uncached Memory Access
• Cached Memory Access
2.8.7.1 Uncached Memory Access
In an uncached access, physical memory resolves the access. Each reference causes a read or write to physical mem- ory. Caches are neither examined nor modified.
These two figures show left-side misalignment.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 39
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
40
2.8.7.2 Cached Memory Access
In a cached access, physical memory and all caches in the system containing a copy of the physical location are used to resolve the access. A copy of a location is coherent if the copy was placed in the cache by a cached coherent access; a copy of a location is noncoherent if the copy was placed in the cache by a cached noncoherent access. (Coherency is dictated by the system architecture, not the processor implementation.)
Caches containing a coherent copy of the location are examined and/or modified to keep the contents of the location coherent. It is not possible to predict whether caches holding a noncoherent copy of the location will be examined and/or modified during a cached coherent access.
Prefetches for data and instructions are allowed. Speculative prefetching of data that may never be used or instruc- tions which may never be executed are allowed.
2.8.8 Implementation-Specific Access Types
An implementation may provide memory access types other than uncached or cached. Implementation-specific doc- umentation accompanies each processor, and defines the properties of the new access types and their effect on all memory-related operations.
2.8.9 Cacheability and Coherency Attributes and Access Types
Memory access types are specified by architecturally-defined and implementation-specific Cacheability and Coher- ency Attribute bits (CCAs) kept in TLB entries.
Slightly different cacheability and coherency attributes such as “cached coherent, update on write” and “cached coherent, exclusive on write” can map to the same memory access type; in this case they both map to cached coher- ent. In order to map to the same access type, the fundamental mechanisms of both CCAs must be the same.
When the operation of the instruction is affected, the instructions are described in terms of memory access types. The load and store operations in a processor proceed according to the specific CCA of the reference, however, and the pseudocode for load and store common functions uses the CCA value rather than the corresponding memory access type.
2.8.10 Mixing Access Types
It is possible to have more than one virtual location mapped to the same physical location (known as aliasing). The memory access type used for the virtual mappings may be different, but it is not generally possible to use mappings with different access types at the same time.
For all accesses to virtual locations with the same memory access type, a processor executing load and store instruc- tions on a physical location must ensure that the instructions occur in proper program order.
A processor can execute a load or store to a physical location using one access type, but any subsequent load or store to the same location using a different memory access type is UNPREDICTABLE, unless a privileged instruction sequence to change the access type is executed between the two accesses. Each implementation has a privileged implementation-specific mechanism to change access types.
The memory access type of a location affects the behavior of I-fetch, load, store, and prefetch operations to that loca- tion. In addition, memory access types affect some instruction descriptions. Load Linked (LL, LLD) and Store Con- ditional (SC, SCD) have defined operation only for locations with cached memory access type.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.8.11 Instruction Fetches 2.8.11.1 Instruction fields layout
For MIPS instructions, the layout of the bit fields within the instructions stays the same regardless of the endianness mode in which the processor is executing. The MIPS architecture only uses Little-Endian bit orderings. Bit 0 of an instruction is always the right-most bit within the instruction while bit 31 is always the left-most bit within a 32-bit instruction. The major opcode is always the left-most 6 bits within the instruction.
2.8.11.2 MIPS32 and MIPS64 Instruction placement and endianness
For the MIPS32 and MIPS64 base architectures, instructions are always 32 bits. All instructions are naturally aligned in memory (address bits 1:0 are 0b00).
Instruction words are always placed in memory according to the endianness.
Figure 2-18 shows an example where the width of external memory is 64-bits (two words) and the processor is exe- cuting in little-endian mode and the instructions are placed in memory for little-endian execution. In this case, the less significant address is the the right-most word of the dword while the more significant address is the left-most word within the dword.
Figure 2-18 Two instructions placed in a 64-bit wide, little-endian memory
Most-significant byte Double Word Least-significant byte
Bit # within dword 63 56 55 48 47 40 39 32 31 24 23 16 15 8 7 0 Address Bits[2:0] 7 6 5 4 3 2 1 0
2.8 Programming Model
Bit # within word 31 24 23 Byte # within word 3
Word 16 15
2 1
8 7
0 31
24 23
Word 16 15
8 7
0
0
3
2
1
0
Younger Instruction
Older Instruction
Program order
Major opcode here
Major opcode here
Figure 2-19 shows the equivalent Big-Endian example where the less significant address refers to the left-most word within the dword and the more significant address refers to the right-most word within the dword. In both BE and LE examples, the bit locations within the instruction words has not changed. The location of the major opcode is always at the left-most bits within the word.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 41
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The MIPS Architecture: An Introduction
Figure 2-19 Two instructions placed in a 64-bit wide, big-endian memory
Least-significant byte Most-significant byte Double Word
Bit # within dword 63 56 55 48 47 40 39 32 31 24 23 16 15 8 7 0 Address Bits[2:0] 0 1 2 3 4 5 6 7
Bit # within word 31 24 23 Byte # within word 0
Word 16 15
1 2
8 7
0 31
24 23
Word 16 15
8 7
0
Program order
on. The major opcode is always the left-most 6 bits within the instruction.
3
0
1
2
3
Older Instruction
Younger Instruction
Major opcode here
Major opcode here
2.8.11.3 Instruction fetches using uncached access to memory without side-effects
Memory regions having no access side-effects can be read an infinite amount of times without changing the value received. For such regions accessed with uncached instruction fetches, the following behaviors are allowed:
It is allowed for the fetch transfer size for uncached memory access to be larger than one instruction word. In this case, it is implementation specific whether multiple instruction fetches are done to the same memory location. It is not required for the processor to have a register to buffer the un-used instructions of the transfer for subsequent execution.
Speculative instruction fetches are allowed. Table 2.2 list some types of speculative instruction fetches. Table 2.2 Speculative instruction fetches
Sequential instructions located after branch/jump fetched before the branch/jump taken/not-taken decision has been determined.
Predicted branch/jump target addresses fetched before branch/jump taken/not-taken decision has been determined or before when target address has been calculated.
Predicted jump target register values before target register has been read.
Predicted return addresses before return register has been read.
Any other type of prefetching ahead of execution.
42
2.8.11.4 Instruction fetches using uncached access to memory with side-effects
Access side-effects for a memory region might include FIFO behavior, stack behavior or have location-specific behavior (one memory location defining the behavior of another memory location). For such regions accessed with uncached instruction fetches, these are the architectural requirements:
The transfer size can only be one instruction word per instruction fetch.
Speculative instruction fetches are not allowed. The types of instruction fetches listed in Table 2.2 are not allowed.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
The architecture defines this memory segment with access side-effects:
• EJTAG Debug Memory space (dmseg). Please refer to MIPS document – MD00047 EJTAG Specification.
Beyond this defined segment, the system programmer/designer is reminded that it is possible to memory map an IO device with access side-effects to any uncached memory location, even within segments which the architecture does not define to have access side-effects. For that reason, any implementation which allows behaviors listed in
2.8.11.3 “Instruction fetches using uncached access to memory without side-effects” should restrict software from executing code within any memory region with side-effects.
2.8.11.5 Instruction fetches using cacheable access to memory
The minimum transfer size for cacheable access is one cacheline. The transfer size may be multiple whole cachelines. Speculative accesses to cacheable memory spaces are allowed as cacheable memory spaces are defined to have no
access side-effects. Table 2.2 list some types of speculative instruction fetches. 2.8.11.6 Instruction fetchs and exceptions
Precise exception model for instruction fetches
The MIPS architecture uses the precise exception model for instruction fetches. A precise exception means that for an instruction-sourced exception, the cause of an exception is reported on the exact instruction which the processor has attempted to execute and has caused the exception.
It is not allowed to report an exception for an instruction which could not be executed due to program control flow. For example, if a branch/jump is taken and the instruction after the branch is not to be executed, the address checks (alignment, MMU match/validity, access priviledge) for that not-to-be-executed instruction may not generate any exception.
Instruction fetch exceptions on branch delay-slots
For instructions occupying a branch delay-slot, any exceptions, including those generated by the fetch of that instruc- tion, should report the exception results so that the branch can be correctly replayed upon return from the exception handler.
2.8.11.7 Self-Modified Code
When the processor writes memory with new instructions at run-time, there are some software steps that must be taken to ensure that the new instructions are fetched properly.
1. The path of instruction fetches to external memory may not be the same as the path of data loads/stores to exter- nal memory (this feature is known as a Harvard architecture). The new instructions must be flushed out to the first level of the memory hierarchy which is shared by both the instruction fetchs and the data load/stores.
2. The processor must wait until all of the new instructions have actually been written to this shared level of the memory hierarchy.
3. If there are caches which hold instructions between that first shared level of memory hierarchy and the processor pipeline, any stale instructions within those caches must be first invalidated before executing the new instruc- tions.
4. Some processors might implement some type of instruction prefetching. Precautions must be used to ensure that the prefetching does not interfere with the previous steps.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 43
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.8 Programming Model
The MIPS Architecture: An Introduction
44 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 45
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
2.8 Programming Model
The MIPS Architecture: An Introduction
46 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Chapter 3
Application Specific Extensions
This section gives an overview of the Architecture Specific Extensions that are supported by the MIPS Architecture Family.
3.1 Description of ASEs
As the MIPS architecture is adopted into a wider variety of markets, the need to extend this architecture in different directions becomes more and more apparent. Therefore various optional application-specific extensions are provided for use with the base ISAs (MIPS32/MIPS64 and microMIPS32/microMIPS64). The ASEs are optional, so the archi- tecture is not permanently bound to support them and the ASEs are used only as needed.
Extensions to the ISA are driven by the requirements of the computer segment, or by customers whose focus is prima- rily on performance. An ASE can be used with the appropriate ISA to meet the needs of a specific application or an entire class of applications.
Figure 3-1shows how ASEs interrelate with the MIPS32 and MIPS64 ISAs. Figure 3-1 MIPS ISAs and ASEs
Enhanced Geometry Processing
Code Compaction (also could use microMIPS)
MIPS-3D ASE
MIPS32 Architecture
MIPS MT ASE
Multi-Threading
MIPS16e ASE
MDMX ASE
MIPS64 Architecture
SmartMIPS ASE
MCU ASE
microControllers
The MIPS32 Architecture is a strict subset of the MIPS64 Architecture. ASEs are applicable to one or both of the base architectures as dictated by market need and the requirements placed on the base architecture by the ASE defini- tion.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 47
Smart Cards
Media Processing
MIPS DSP ASE
Signal Processing
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Application Specific Extensions
3.2 List of Application Specific Instructions
As of the publishing date of this document, the following Application Specific Extensions were supported by the architecture.
ASE
MIPS16eTM MDMXTM MIPS-3D® SmartMIPS® MIPS® DSP MIPS® MT MCU
Supported Base Architectures
MIPS32 or MIPS64 MIPS64 MIPS32 or MIPS64 MIPS32 MIPS32 or MIPS64 MIPS32 or MIPS64 MIPS32 or MIPS64
Use
Code Compaction
Digital Media
Geometry Processing
Smart Cards and Smart Objects Signal Processing Multi-Threading
Fast Interrupt Response & I/O register
programming
48
3.2.1 The MIPS16eTM Application Specific Extension to the MIPS32 and MIPS64 Archi- tecture
The MIPS16e ASE is composed of 16-bit compressed code instructions, designed for the embedded processor market and situations with tight memory constraints. The core can execute both 16- and 32-bit instructions intermixed in the same program, and is compatible with both the MIPS32 and MIPS64 Architectures. Volume IV-a of this document set describes the MIPS16e ASE.
The microMIPS Architectures supercedes the MIPS16e ASE as the small code-size solution. microMIPS provides for small code sizes for kernelmode code, floating-point code. These were not available through MIPS16e.
3.2.2 The MDMXTM Application Specific Extension to the MIPS64 Architectures
The MIPS Digital Media Extension (MDMX) provides video, audio, and graphics pixel processing through vectors of small integers. Although not a part of the MIPS ISA, this extension is included for informational purposes. Because the MDMX ASE requires one of the 64-bit Architectures, it is not discussed in this document set.
3.2.3 The MIPS-3D® Application Specific Extension to the MIPS Architecture
The MIPS-3D ASE provides enhanced performance of geometry processing calculations by building on the paired single floating point data type, and adding specific instructions to accelerate computations on these data types.Volume IV-c of this document set describes the MIPS-3D ASE. Because the MIPS-3D ASE requires a 64-bit floating point unit, it is only available with a Release 1 MIPS64 processor, or a Release 2 (or subsequent releases) processor that includes a 64-bit FPU.
3.2.4 The SmartMIPS® Application Specific Extension to the MIPS32 Architecture
The SmartMIPS ASE extends the MIPS32 Architectures with a set of new and modified instruction designed to improve the performance and reduce the memory consumption of MIPS-based smart card or smart object systems. Volume IV-d of this document set describes the SmartMIPS ASE.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
3.2.5 The MIPS® DSP Application Specific Extension to the MIPS Architecture
The MIPS DSP ASE provides enhanced performance of signal-processing applications by providing computational support for fractional data types, SIMD, saturation, and other elements that are commonly used in such applications. Volume IV-e of this document set describes the MIPS DSP ASE.
3.2.6 The MIPS® MT Application Specific Extension to the MIPS Architecture
The MIPS MT ASE provides the architecture to support multi-threaded implementations of the Architecture. This includes support for both virtual processors and lightweight thread contexts. Volume IV-f of this document set describes the MIPS MT ASE.
3.2.7 The MIPS® MCU Application Specific Extension to the MIPS Architecture
The MIPS MCU ASE provides enhanced handling of memory-mapped I/O registers and lower interrupt latencies. Volume IV-g of this document set describes the MIPS MCU ASE.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 49
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
3.2 List of Application Specific Instructions
Application Specific Extensions
50 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Chapter 4
Overview of the CPU Instruction Set
This chapter gives an overview of the CPU instructions, including a description of CPU instruction formats. An over- view of the FPU instructions is given in Chapter 5.
4.1 CPU Instructions, Grouped By Function
CPU instructions are organized into the following functional groups:
• Load and store
• Computational
• Jump and branch
• Miscellaneous
• Coprocessor
Each instruction is 32 bits long.
4.1.1 CPU Load and Store Instructions
MIPS processors use a load/store architecture; all operations are performed on operands held in processor registers and main memory is accessed only through load and store instructions.
4.1.1.1 Types of Loads and Stores
There are several different types of load and store instructions, each designed for a different purpose:
• Transferring variously-sized fields (for example, LB, SW)
• Trading transferred data as signed or unsigned integers (for example, LHU)
• Accessing unaligned fields (for example, LWR, SWL)
• Selecting the addressing mode (for example, SDXC1, in the FPU)
• Atomic memory update (read-modify-write: for instance, LL/SC)
Regardless of the byte ordering (big- or little-endian), the address of a halfword, word, or doubleword is the lowest byte address among the bytes forming the object:
• For big-endian ordering, this is the most-significant byte.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 51
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
• For a little-endian ordering, this is the least-significant byte.
Refer to “Byte Ordering and Endianness” on page 37 for more information on big-endian and little-endian data order- ing.
4.1.1.2 Load and Store Access Types
Table 4.1 lists the data sizes that can be accessed through CPU load and store operations. These tables also indicate the particular ISA within which each operation is defined.
Table 4.1 Load and Store Operations Using Register + Offset Addressing Mode
Data Size
CPU
Coprocessors 1 and 2
Load Signed
Load Unsigned
Store
Load
Store
Byte
MIPS32
MIPS32
MIPS32
Halfword
MIPS32
MIPS32
MIPS32
Word
MIPS32
MIPS64
MIPS32
MIPS32
MIPS32
Doubleword (FPU)
MIPS32
MIPS32
Unaligned word
MIPS32
MIPS32
Linked word (atomic modify)
MIPS32
MIPS32
4.1.1.3 List of CPU Load and Store Instructions
The following data sizes (as defined in the AccessLength field) are transferred by CPU load and store instructions:
• Byte
• Halfword
• Word
Signed and unsigned integers of different sizes are supported by loads that either sign-extend or zero-extend the data loaded into the register.
Table 4.2 lists aligned CPU load and store instructions, while unaligned loads and stores are listed in Table 4.3. Each table also lists the MIPS ISA within which an instruction is defined.
Table 4.2 Aligned CPU Load/Store Instructions
Mnemonic
Instruction
Defined in MIPS ISA
LB
Load Byte
MIPS32
LBU
Load Byte Unsigned
MIPS32
LH
Load Halfword
MIPS32
LHU
Load Halfword Unsigned
MIPS32
LW
Load Word
MIPS32
52
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 4.2 Aligned CPU Load/Store Instructions (Continued)
Unaligned words and doublewords can be loaded or stored in just two instructions by using a pair of the special instructions listed in Table 4.3. The load instructions read the left-side or right-side bytes (left or right side of register) from an aligned word and merge them into the correct bytes of the destination register.
Unaligned CPU load and store instructions are listed in Table 4.3, along with the MIPS ISA within which an instruc- tion is defined.
Table 4.3 Unaligned CPU Load and Store Instructions
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
SB
Store Byte
MIPS32
SH
Store Halfword
MIPS32
SW
Store Word
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
LWL
Load Word Left
MIPS32
LWR
Load Word Right
MIPS32
SWL
Store Word Left
MIPS32
SWR
Store Word Right
MIPS32
4.1.1.4 Loads and Stores Used for Atomic Updates
The paired instructions, Load Linked and Store Conditional, can be used to perform an atomic read-modify-write of word or doubleword cached memory locations. These instructions are used in carefully coded sequences to provide one of several synchronization primitives, including test-and-set, bit-level locks, semaphores, and sequencers and event counts. Table 4.4 lists the LL and SC instructions, along with the MIPS ISA within which an instruction is defined.
Table 4.4 Atomic Update CPU Load and Store Instructions
4.1.1.5 Coprocessor Loads and Stores
If a particular coprocessor is not enabled, loads and stores to that processor cannot execute and the attempted load or store causes a Coprocessor Unusable exception. Enabling a coprocessor is a privileged operation provided by the System Control Coprocessor, CP0.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 53
Mnemonic
Instruction
Defined in MIPS ISA
LL
Load Linked Word
MIPS32
SC
Store Conditional Word
MIPS32
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
Table 4.5 lists the coprocessor load and store instructions.
Table 4.5 Coprocessor Load and Store Instructions
Mnemonic
Instruction
Defined in MIPS ISA
LDCz
Load Doubleword to Coprocessor-z, z = 1 or 2
MIPS32
LWCz
Load Word to Coprocessor-z, z = 1 or 2
MIPS32
SDCz
Store Doubleword from Coprocessor-z, z = 1 or 2
MIPS32
SWCz
Store Word from Coprocessor-z, z = 1 or 2
MIPS32
Table 4.6 lists the specific FPU load and store instructions;1 it also lists the MIPS ISA within which an instruction was first defined.
Table 4.6 FPU Load and Store Instructions Using Register + Register Addressing
Mnemonic
Instruction
Defined in MIPS ISA
LWXC1
Load Word Indexed to Floating Point
MIPS64 MIPS32 Release 2
SWXC1
Store Word Indexed from Floating Point
MIPS64 MIPS32 Release 2
LDXC1
Load Doubleword Indexed to Floating Point
MIPS64 MIPS32 Release 2
SDXC1
Store Doubleword Indexed from Floating Point
MIPS64 MIPS32 Release 2
LUXC1
Load Doubleword Indexed Unaligned to Floating Point
MIPS64 MIPS32 Release 2
SUXC1
Store Doubleword Indexed Unaligned from Floating Point
MIPS64 MIPS32 Release 2
4.1.2 Computational Instructions
54
1.
This section describes the following:
• ALU Immediate and Three-Operand Instructions
• ALU Two-Operand Instructions
• Shift Instructions
• Multiply and Divide Instructions
2’s complement arithmetic is performed on integers represented in 2’s complement notation. These are signed ver- sions of the following operations:
• Add
FPU loads and stores are listed here with the other coprocessor loads and stores for convenience.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• Subtract
• Multiply
• Divide
The add and subtract operations labelled “unsigned” are actually modulo arithmetic without overflow detection.
There are also unsigned versions of multiply and divide, as well as a full complement of shift and logical operations. Logical operations are not sensitive to the width of the register.
MIPS32 provided 32-bit integers and 32-bit arithmetic.
4.1.2.1 ALU Immediate and Three-Operand Instructions
Table 4.7 lists those arithmetic and logical instructions that operate on one operand from a register and the other from a 16-bit immediate value supplied by the instruction word. This table also lists the MIPS ISA within which an instruc- tion is defined.
The immediate operand is treated as a signed value for the arithmetic and compare instructions, and treated as a logi- cal value (zero-extended to register length) for the logical instructions.
Table 4.7 ALU Instructions With a 16-bit Immediate Operand
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
ADDI
Add Immediate Word
MIPS32
ADDIU1
Add Immediate Unsigned Word
MIPS32
ANDI
And Immediate
MIPS32
LUI
Load Upper Immediate
MIPS32
ORI
Or Immediate
MIPS32
SLTI
Set on Less Than Immediate
MIPS32
SLTIU
Set on Less Than Immediate Unsigned
MIPS32
XORI
Exclusive Or Immediate
MIPS32
1. The term “unsigned” in the instruction name is a misnomer; this operation is 32-bit modulo arithmetic that does not trap on overflow.
Table 4.8 describes ALU instructions that use three operands, along with the MIPS ISA within which an instruction is defined.
Table 4.8 Three-Operand ALU Instructions
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 55
Mnemonic
Instruction
Defined in MIPS ISA
ADD
Add Word
MIPS32
ADDU1
Add Unsigned Word
MIPS32
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
Table 4.8 Three-Operand ALU Instructions (Continued)
Mnemonic
Instruction
Defined in MIPS ISA
AND
And
MIPS32
NOR
Nor
MIPS32
OR
Or
MIPS32
SLT
Set on Less Than
MIPS32
SLTU
Set on Less Than Unsigned
MIPS32
SUB
Subtract Word
MIPS32
SUBU1
Subtract Unsigned Word
MIPS32
XOR
Exclusive Or
MIPS32
1. The term “unsigned” in the instruction name is a misnomer; this operation is 32-bit modulo arithmetic that does not trap on overflow.
4.1.2.2 ALU Two-Operand Instructions
Table 4.8 describes ALU instructions that use two operands, along with the MIPS ISA within which an instruction is defined.
Table 4.9 Two-Operand ALU Instructions
4.1.2.3 Shift Instructions
The ISA defines two types of shift instructions:
• Those that take a fixed shift amount from a 5-bit field in the instruction word (for instance, SLL, SRL)
• Those that take a shift amount from the low-order bits of a general register (for instance, SRAV, SRLV) Shift instructions are listed in Table 4.10, along with the MIPS ISA within which an instruction is defined.
Table 4.10 Shift Instructions
Mnemonic
Instruction
Defined in MIPS ISA
CLO
Count Leading Ones in Word
MIPS32
CLZ
Count Leading Zeros in Word
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
ROTR
Rotate Word Right
MIPS32 Release 2
ROTRV
Rotate Word Right Variable
MIPS32 Release 2
SLL
Shift Word Left Logical
MIPS32
SLLV
Shift Word Left Logical Variable
MIPS32
SRA
Shift Word Right Arithmetic
MIPS32
56
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 4.10 Shift Instructions (Continued)
4.1.2.4 Multiply and Divide Instructions
The multiply and divide instructions produce twice as many result bits as is typical with other processors. With one exception, they deliver their results into the HI and LO special registers. The MUL instruction delivers the lower half of the result directly to a GPR.
• Multiply produces a full-width product twice the width of the input operands; the low half is loaded into LO and the high half is loaded into HI.
• Multiply-Add and Multiply-Subtract produce a full-width product twice the width of the input operations and adds or subtracts the product from the concatenated value of HI and LO. The low half of the addition is loaded into LO and the high half is loaded into HI.
• Divide produces a quotient that is loaded into LO and a remainder that is loaded into HI.
The results are accessed by instructions that transfer data between HI/LO and the general registers.
Table 4.11 lists the multiply, divide, and HI/LO move instructions, along with the MIPS ISA within which an instruc- tion is defined.
Table 4.11 Multiply/Divide Instructions
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
SRAV
Shift Word Right Arithmetic Variable
MIPS32
SRL
Shift Word Right Logical
MIPS32
SRLV
Shift Word Right Logical Variable
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
DIV
Divide Word
MIPS32
DIVU
Divide Unsigned Word
MIPS32
MADD
Multiply and Add Word
MIPS32
MADDU
Multiply and Add Word Unsigned
MIPS32
MFHI
Move From HI
MIPS32
MFLO
Move From LO
MIPS32
MSUB
Multiply and Subtract Word
MIPS32
MSUBU
Multiply and Subtract Word Unsigned
MIPS32
MTHI
Move To HI
MIPS32
MTLO
Move To LO
MIPS32
MUL
Multiply Word to Register
MIPS32
MULT
Multiply Word
MIPS32
MULTU
Multiply Unsigned Word
MIPS32
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 57
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
58
4.1.3 Jump and Branch Instructions
This section describes the following:
• Types of Jump and Branch Instructions Defined by the ISA
• Branch Delays and the Branch Delay Slot
• Delay Slot Behavior
• List of Jump and Branch Instructions
4.1.3.1 Types of Jump and Branch Instructions Defined by the ISA
The architecture defines the following jump and branch instructions:
• PC-relative conditional branch
• PC-region unconditional jump
• Absolute (register) unconditional jump
• A set of procedure calls that record a return link address in a general register.
4.1.3.2 Branch Delays and the Branch Delay Slot
All branches have an architectural delay of one instruction. The instruction immediately following a branchis said to be in the branch delay slot. If a branch or jump instruction is placed in the branch delay slot, the operation of both instructions is UNPREDICTABLE.
By convention, if an exception or interrupt prevents the completion of an instruction in the branch delay slot, the instruction stream is continued by re-executing the branch instruction. To permit this, branches must be restartable; procedure calls may not use the register in which the return link is stored (usually GPR 31) to determine the branch target address.
4.1.3.3 Delay Slot Behavior
There are two versions of branches and jumps; they differ in the manner in which they handle the instruction in the delay slot when the branch is not taken and execution falls through.
• •
Branch and Jump instructions execute the instruction in the delay slot.
Branch likely instructions do not execute the instruction in the delay slot if the branch is not taken (they are said
to nullify the instruction in the delay slot).
Although the Branch Likely instructions are included in this specification, software is strongly encouraged to avoid the use of the Branch Likely instructions, as they will be removed from a future revision of the MIPS Architecture.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
4.1.3.4 List of Jump and Branch Instructions
Table 4.12 lists instructions that jump to a procedure call within the current 256 MB-aligned region.
Table 4.13 lists instructions that jump to an absolute address held in a register.
Table 4.12 lists the unconditional jump instructions within a given 256 MByte region. Table 4.14 lists branch instruc- tions that compare two registers before conditionally executing a PC-relative branch. Table 4.15 lists branch instruc- tions that test a register—compare with zero—before conditionally executing a PC-relative branch. Table 4.16 lists the deprecated Branch Likely Instructions.
Each table also lists the MIPS ISA within which an instruction is defined.
Table 4.12 Unconditional Jump Within a 256 Megabyte Region
Table 4.13 Unconditional Jump using Absolute Address
Table 4.14 PC-Relative Conditional Branch Instructions Comparing Two Registers
Table 4.15 PC-Relative Conditional Branch Instructions Comparing With Zero
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
J
Jump
MIPS32
JAL
Jump and Link
MIPS32
JALX
Jump and Link Exchange
MIPS16e MIPS32 Release 3
Mnemonic
Instruction
Defined in MIPS ISA
JALR
Jump and Link Register
MIPS32
JALR.HB
Jump and Link Register with Hazard Barrier
MIPS32 Release 2
JR
Jump Register
MIPS32
JR.HB
Jump Register with Hazard Barrier
MIPS32 Release 2
Mnemonic
Instruction
Defined in MIPS ISA
BEQ
Branch on Equal
MIPS32
BNE
Branch on Not Equal
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
BGEZ
Branch on Greater Than or Equal to Zero
MIPS32
BGEZAL
Branch on Greater Than or Equal to Zero and Link
MIPS32
BGTZ
Branch on Greater Than Zero
MIPS32
BLEZ
Branch on Less Than or Equal to Zero
MIPS32
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 59
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
Table 4.15 PC-Relative Conditional Branch Instructions Comparing With Zero (Continued)
Table 4.16 Deprecated Branch Likely Instructions
Mnemonic
Instruction
Defined in MIPS ISA
BLTZ
Branch on Less Than Zero
MIPS32
BLTZAL
Branch on Less Than Zero and Link
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
BEQL
Branch on Equal Likely
MIPS32
BGEZALL
Branch on Greater Than or Equal to Zero and Link Likely
MIPS32
BGEZL
Branch on Greater Than or Equal to Zero Likely
MIPS32
BGTZL
Branch on Greater Than Zero Likely
MIPS32
BLEZL
Branch on Less Than or Equal to Zero Likely
MIPS32
BLTZALL
Branch on Less Than Zero and Link Likely
MIPS32
BLTZL
Branch on Less Than Zero Likely
MIPS32
BNEL
Branch on Not Equal Likely
MIPS32
60
4.1.4 Miscellaneous Instructions
Miscellaneous instructions include:
• Instruction Serialization (SYNC and SYNCI)
• Exception Instructions
• Conditional Move Instructions
• Prefetch Instructions
• NOP Instructions
4.1.4.1 Instruction Serialization (SYNC and SYNCI)
In normal operation, the order in which load and store memory accesses appear to a viewer outside the executing pro- cessor (for instance, in a multiprocessor system) is not specified by the architecture.
The SYNC instruction can be used to create a point in the executing instruction stream at which the relative order of some loads and stores can be determined: loads and stores executed before the SYNC are completed before loads and stores after the SYNC can start.
The SYNCI instruction synchronizes the processor caches with previous writes or other modifications to the instruc- tion stream.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 4.17 lists the synchronization instructions, along with the MIPS ISA within which it is defined. Table 4.17 Serialization Instruction
4.1.4.2 Exception Instructions
Exception instructions transfer control to a software exception handler in the kernel. There are two types of excep- tions, conditional and unconditional. These are caused by the following instructions:
Trap instructions, which cause conditional exceptions based upon the result of a comparison System call and breakpoint instructions, which cause unconditional exceptions
Table 4.18 lists the system call and breakpoint instructions. Table 4.19 lists the trap instructions that compare two registers. Table 4.20 lists trap instructions, which compare a register value with an immediate value.
Each table also lists the MIPS ISA within which an instruction is defined.
Table 4.18 System Call and Breakpoint Instructions
Table 4.19 Trap-on-Condition Instructions Comparing Two Registers
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
SYNC
Synchronize Shared Memory
MIPS32
SYNCI
Synchronize Caches to Make Instruction Writes Effective
MIPS32 Release 2
Mnemonic
Instruction
Defined in MIPS ISA
BREAK
Breakpoint
MIPS32
SYSCALL
System Call
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
TEQ
Trap if Equal
MIPS32
TGE
Trap if Greater Than or Equal
MIPS32
TGEU
Trap if Greater Than or Equal Unsigned
MIPS32
TLT
Trap if Less Than
MIPS32
TLTU
Trap if Less Than Unsigned
MIPS32
TNE
Trap if Not Equal
MIPS32
Table 4.20 Trap-on-Condition Instructions Comparing an Immediate Value
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 61
Mnemonic
Instruction
Defined in MIPS ISA
TEQI
Trap if Equal Immediate
MIPS32
TGEI
Trap if Greater Than or Equal Immediate
MIPS32
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
Table 4.20 Trap-on-Condition Instructions Comparing an Immediate Value
4.1.4.3 Conditional Move Instructions
MIPS32 includes instructions to conditionally move one CPU general register to another, based on the value in a third general register. For floating point conditional moves, refer to Chapter 4.
Table 4.21 lists conditional move instructions, along with the MIPS ISA within which an instruction is defined. Table 4.21 CPU Conditional Move Instructions
Mnemonic
Instruction
Defined in MIPS ISA
TGEIU
Trap if Greater Than or Equal Immediate Unsigned
MIPS32
TLTI
Trap if Less Than Immediate
MIPS32
TLTIU
Trap if Less Than Immediate Unsigned
MIPS32
TNEI
Trap if Not Equal Immediate
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
MOVF
Move Conditional on Floating Point False
MIPS32
MOVN
Move Conditional on Not Zero
MIPS32
MOVT
Move Conditional on Floating Point True
MIPS32
MOVZ
Move Conditional on Zero
MIPS32
Mnemonic
Instruction
Addressing Mode
Defined in MIPS ISA
PREF
Prefetch
Register+Offset
MIPS32
PREFX
Prefetch Indexed
Register+Register
MIPS64 MIPS32 Release 2
62
4.1.4.4 Prefetch Instructions
There are two prefetch advisory instructions:
• One with register+offset addressing (PREF)
• One with register+register addressing (PREFX)
These instructions advise that memory is likely to be used in a particular way in the near future and should be prefetched into the cache. The PREFX instruction is encoded in the FPU opcode space, along with the other opera- tions using register+register addressing
Table 4.22 Prefetch Instructions
4.1.4.5 NOP Instructions
The NOP instruction is actually encoded as an all-zero instruction. MIPS processors special-case this encoding as performing no operation, and optimize execution of the instruction. In addition, SSNOP instruction, takes up one issue cycle on any processor, including super-scalar implementations of the architecture.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 4.23 lists conditional move instructions, along with the MIPS ISA within which an instruction is defined. Table 4.23 NOP Instructions
4.1.5 Coprocessor Instructions
This section contains information about the following:
• What Coprocessors Do
• System Control Coprocessor 0 (CP0)
• Floating Point Coprocessor 1 (CP1)
• Coprocessor Load and Store Instructions 4.1.5.1 What Coprocessors Do
Coprocessors are alternate execution units, with register files separate from the CPU. In abstraction, the MIPS archi- tecture provides for up to four coprocessor units, numbered 0 to 3. Each level of the ISA defines a number of these coprocessors, as listed in Table 4.24.
Table 4.24 Coprocessor Definition and Use in the MIPS Architecture
4.1 CPU Instructions, Grouped By Function
Mnemonic
Instruction
Defined in MIPS ISA
NOP
No Operation
MIPS32
SSNOP
Superscalar Inhibit NOP
MIPS32
Coprocessor
MIPS32
MIPS64
CP0
Sys Control
Sys Control
CP1
FPU
FPU
CP2
implementation specific
CP3
FPU (COP1X)
Coprocessor 0 is always used for system control and coprocessor 1 and 3 are used for the floating point unit. Copro- cessor 2 is reserved for implementation-specific use.
A coprocessor may have two different register sets:
• Coprocessor general registers
• Coprocessor control registers
Each set contains up to 32 registers. Coprocessor computational instructions may use the registers in either set.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 63
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
4.1.5.2 System Control Coprocessor 0 (CP0)
The system controller for all MIPS processors is implemented as coprocessor 0 (CP02), the System Control Copro- cessor. It provides the processor control, memory management, and exception handling functions.
4.1.5.3 Floating Point Coprocessor 1 (CP1)
If a system includes a Floating Point Unit, it is implemented as coprocessor 1 (CP13). In Release 1 of the MIPS64 Architecture, and in Release 2 of the MIPS32 and MIPS64 Architectures, the FPU also uses the computation opcode space assigned to coprocessor unit 3, renamed COP1X. Details of the FPU instructions are documented in “Overview of the FPU Instruction Set” on page 67.
Coprocessor instructions are divided into two main groups:
• Load and store instructions (move to and from coprocessor), which are reserved in the main opcode space
• Coprocessor-specific operations, which are defined entirely by the coprocessor
4.1.5.4 Coprocessor Load and Store Instructions
Explicit load and store instructions are not defined for CP0; for CP0 only, the move to and from coprocessor instruc- tions must be used to write and read the CP0 registers. The loads and stores for the remaining coprocessors are sum- marized in “Coprocessor Loads and Stores” on page 53.
4.2 CPU Instruction Formats
A CPU instruction is a single 32-bit aligned word. The CPU instruction formats are shown below:
• Immediate (see Figure 4-1)
• Jump (see Figure 4-2)
• Register (see Figure 4-3)
Table 4.25 describes the fields used in these instructions.
Table 4.25 CPU Instruction Format Fields
Field
Description
opcode
6-bit primary operation code
rd
5-bit specifier for the destination register
rs
5-bit specifier for the source register
rt
5-bit specifier for the target (source/destination) register or used to specify functions within the primary opcode REGIMM
64
2. 3.
CP0 instructions use the COP0 opcode, and as such are differentiated from the CP0 designation in this book.
FPU instructions (such as LWC1, SDC1, etc.) that use the COP1 opcode are differentiated from the CP1 designation in this book. See “Overview of the FPU Instruction Set” on page 67 for more information about the FPU instructions.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 4.25 CPU Instruction Format Fields
4.2 CPU Instruction Formats
Field
Description
immediate
16-bit signed immediate used for logical operands, arithmetic signed operands, load/store address byte offsets, and PC-relative branch signed instruction displacement
instr_index
26-bit index shifted left two bits to supply the low-order 28 bits of the jump target address
sa
5-bit shift amount
function
6-bit function field used to specify functions within the primary opcode SPECIAL
31
31
31
Figure 4-1 Immediate (I-Type) CPU Instruction Format
26 25 21 20 16 15
655 16
Figure 4-2 Jump (J-Type) CPU Instruction Format
26 25 21 20 16 15 11 10
6 26
Figure 4-3 Register (R-Type) CPU Instruction Format
26 25 21 20 16 15 11 10
6 5
6 5
0
0
0
opcode
rs
rt
immediate
opcode
instr_index
opcode
rs
rt
rd
sa
function
655556
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
65
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the CPU Instruction Set
66 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Chapter 5
Overview of the FPU Instruction Set
This chapter describes the instruction set architecture (ISA) for the floating point unit (FPU) in the MIPS32 architec- ture. In the MIPS architecture, the FPU is implemented via Coprocessor 1 and Coprocessor 3, an optional processor
implementing IEEE Standard 7541 floating point operations. The FPU also provides a few additional operations not defined by the IEEE standard.
This chapter provides an overview of the following FPU architectural details:
• “Binary Compatibility” on page 67
• “Enabling the Floating Point Coprocessor” on page 68
• “IEEE Standard 754” on page 68
• “FPU Data Types” on page 68
• “Floating Point Register Types” on page 73
• “Floating Point Control Registers (FCRs)” on page 75
• “Formats of Values Used in FP Registers” on page 82
• “FPU Exceptions” on page 82
• “FPU Instructions” on page 85
• “Valid Operands for FPU Instructions” on page 92
• “FPU Instruction Formats” on page 94
The FPU instruction set is summarized by functional group. Each instruction is also described individually in alpha- betical order in Volume II.
5.1 Binary Compatibility
In addition to an Instruction Set Architecture, the MIPS architecture definition includes processing resources such as the set of coprocessor general registers. In Release 1 of the Architecture, the 32-bit registers in MIPS32 were enlarged to 64-bits in MIPS64; however, these 64-bit FPU registers are not backwards compatible. Instead, processors imple- menting the MIPS64 Architecture provide a mode bit to select either the 32-bit or 64-bit register model. In Release 2
1. In this chapter, references to “IEEE standard” and “IEEE Standard 754” refer to IEEE Standard 754-1985, “IEEE Standard for Binary Floating Point Arithmetic.” For more information about this standard, see the IEEE web page at http:// grouper.ieee.org/groups/754/.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 67
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
of the Architecture and subsequent releases, a 32-bit CPU may include a full 64-bit coprocessor, including a floating point unit which implements the same mode bit to select 32-bit or 64-bit FPU register model.
Any processor implementing MIPS64 can also run MIPS32 binary programs, built for the same, or a lower release of the Architecture, without change.
5.2 Enabling the Floating Point Coprocessor
Enabling the Floating Point Coprocessor is done by enabling Coprocessor 1, and is a privileged operation provided by the System Control Coprocessor. If Coprocessor 1 is not enabled, an attempt to execute a floating point instruction causes a Coprocessor Unusable exception. Every system environment either enables the FPU automatically or pro- vides a means for an application to request that it is enabled.
5.3 IEEE Standard 754
IEEE Standard 754 defines the following:
• Floating point data types
• The basic arithmetic, comparison, and conversion operations
• A computational model
The IEEE standard does not define specific processing resources nor does it define an instruction set.
The MIPS architecture includes non-IEEE FPU control and arithmetic operations (multiply-add, reciprocal, and reciprocal square root) which may not supply results that match the IEEE precision rules.
5.4 FPU Data Types
The FPU provides both floating point and fixed point data types, which are described in the next two sections.
• The single and double precision floating point data types are those specified by the IEEE standard.
• The fixed point types are signed integers provided by the CPU architecture.
5.4.1 Floating Point Formats
68
The •
•
• The •
following two floating point formats are provided by the FPU:
32-bit single precision floating point (type S, shown in Figure 5-1)
64-bit double precision floating point (type D, shown in Figure 5-2)
64-bit paired single floating point, combining two single precision data types (Type PS, shown in Figure 5-3) floating point data types represent numeric values as well as other special entities, such as the following:
Two infinities, +∞ and -∞
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• Signaling non-numbers (SNaNs)
• Quiet non-numbers (QNaNs)s
• Numbers of the form: (-1)s 2E b0.b1 b2..bp-1, where:
• s=0 or 1
• E=any integer between E_min and E_max, inclusive
• bi=0 or 1 (the high bit, b0, is to the left of the binary point)
• p is the signed-magnitude precision
Table 5.1 Parameters of Floating Point Data Types
5.4 FPU Data Types
Parameter
Single (or each half of Paired Single)
Double
Bits of mantissa precision, p
24
53
Maximum exponent, E_max
+127
+1023
Minimum exponent, E_min
-126
-1022
Exponent bias
+127
+1023
Bits in exponent field, e
8
11
Representation of b0 integer bit
hidden
hidden
Bits in fraction field, f
23
52
Total format width in bits
32
64
The single and double floating point data types are composed of three fields—sign, exponent, fraction—whose sizes are listed in Table 5.1.
Layouts of these fields are shown in Figures 5-1, 5-2, and 5-3 below. The fields are
• 1-bit sign, s
• Biased exponent, e=E + bias
• Binary fraction, f=.b1 b2..bp-1 (the b0 bit is not recorded)
Figure 5-1 Single-Precisions Floating Point Format (S)
33 22 10320
18 23
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 69
S
Exponent
Fraction
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Figure 5-2 Double-Precisions Floating Point Format (D)
66 55
3221 0
1 11 52
Figure 5-3 Paired Single Floating Point Format (PS)
66 55 333 22
3254 21032 0
18 23 18 23
Values are encoded in the specified format by using unbiased exponent, fraction, and sign values listed in Table 5.2. The high-order bit of the Fraction field, identified as b1, is also important for NaNs.
Table 5.2 Value of Single or Double Floating Point DataType Encoding
S
Exponent
Fraction
S
Exponent
fraction
S
Exponent
Fraction
Unbiased E
f
s
b1
Value V
Type of Value
Typical Single Bit Pattern1
Typical Double Bit Pattern1
E_max + 1
≠0
1
SNaN
Signaling NaN
0x7fffffff
0x7fffffff ffffffff
0
QNaN
Quiet NaN
0x7fbfffff
0x7ff7ffff ffffffff
E_max +1
0
1
-∞
minus infinity
0xff800000
0xfff00000 00000000
0
+∞
plus infinity
0x7f800000
0x7ff00000 00000000
E_max
to
E_min
1
– (2E)(1.f)
negative normalized num- ber
0x80800000
through
0xff7fffff
0x80100000 00000000
through
0xffefffff ffffffff
0
+ (2E)(1.f)
positive normalized number
0x00800000
through
0x7f7fffff
0x00100000 00000000
through
0x7fefffff ffffffff
E_min -1
≠0
1
– (2E_min)(0.f)
negative denormalized number
0x807fffff
0x800fffff ffffffff
0
+ (2E_min)(0.f)
positive denormalized num- ber
0x007fffff
0x000fffff ffffffff
E_min -1
0
1
-0
negative zero
0x80000000
0x80000000 00000000
0
+0
positive zero
0x00000000
0x00000000 00000000
1. The “Typical” nature of the bit patterns for the NaN and denormalized values reflects the fact that the sign may have either value (NaN) and the fact that the fraction field may have any non-zero value (both). As such, the bit patterns shown are one value in a class of potential values that represent these special values.
5.4.1.1 Normalized and Denormalized Numbers
For single and double data types, each representable nonzero numerical value has just one encoding; numbers are kept in normalized form. The high-order bit of the p-bit mantissa, which lies to the left of the binary point, is “hid- den,” and not recorded in the Fraction field. The encoding rules permit the value of this bit to be determined by look- ing at the value of the exponent. When the unbiased exponent is in the range E_min to E_max, inclusive, the number is normalized and the hidden bit must be 1. If the numeric value cannot be normalized because the exponent would be
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
70
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
less than E_min, then the representation is denormalized and the encoded number has an exponent of E_min-1 and the hidden bit has the value 0. Plus and minus zero are special cases that are not regarded as denormalized values.
5.4.1.2 Reserved Operand Values—Infinity and NaN
A floating point operation can signal IEEE exception conditions, such as those caused by uninitialized variables, vio- lations of mathematical rules, or results that cannot be represented. If a program does not choose to trap IEEE excep- tion conditions, a computation that encounters these conditions proceeds without trapping but generates a result indicating that an exceptional condition arose during the computation. To permit this, each floating point format defines representations, listed in Table 5.2, for plus infinity (+∞), minus infinity (-∞), quiet non-numbers (QNaN), and signaling non-numbers (SNaN).
5.4.1.3 Infinity and Beyond
Infinity represents a number with magnitude too large to be represented in the format; in essence it exists to represent a magnitude overflow during a computation. A correctly signed ∞ is generated as the default result in division by zero and some cases of overflow; details are given in the IEEE exception condition described in 5.8.1 “Exception Conditions” on page 83.
Once created as a default result, ∞ can become an operand in a subsequent operation. The infinities are interpreted such that -∞ < (every finite number) < +∞. Arithmetic with ∞ is the limiting case of real arithmetic with operands of arbitrarily large magnitude, when such limits exist. In these cases, arithmetic on ∞ is regarded as exact and exception conditions do not arise. The out-of-range indication represented by ∞ is propagated through subsequent computa- tions. For some cases there is no meaningful limiting case in real arithmetic for operands of ∞, and these cases raise the Invalid Operation exception condition (see “Invalid Operation Exception” on page 83).
5.4.1.4 Signalling Non-Number (SNaN)
SNaN operands cause the Invalid Operation exception for arithmetic operations. SNaNs are useful values to put in uninitialized variables. An SNaN is never produced as a result value.
IEEE Standard 754 states that “Whether copying a signaling NaN without a change of format signals the Invalid Operation exception is the implementor’s option.” The MIPS architecture has chosen to make the formatted operand move instructions (MOV.fmt MOVT.fmt MOVF.fmt MOVN.fmt MOVZ.fmt) non-arithmetic and they do not signal IEEE 754 exceptions.
5.4.1.5 Quiet Non-Number (QNaN)
QNaNs are intended to afford retrospective diagnostic information inherited from invalid or unavailable data and results. Propagation of the diagnostic information requires information contained in a QNaN to be preserved through arithmetic operations and floating point format conversions.
QNaN operands do not cause arithmetic operations to signal an exception. When a floating point result is to be deliv- ered, a QNaN operand causes an arithmetic operation to supply a QNaN result. When possible, this QNaN result is one of the operand QNaN values. QNaNs do have effects similar to SNaNs on operations that do not deliver a floating point result—specifically, comparisons. (For more information, see the detailed description of the floating point com- pare instruction, C.cond.fmt.)
When certain invalid operations not involving QNaN operands are performed but do not trap (because the trap is not enabled), a new QNaN value is created. Table 5.3 shows the QNaN value generated when no input operand QNaN value can be copied. The values listed for the fixed point formats are the values supplied to satisfy the IEEE standard
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 71
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
5.4 FPU Data Types
Overview of the FPU Instruction Set
when a QNaN or infinite floating point value is converted to fixed point. There is no other feature of the architecture that detects or makes use of these “integer QNaN” values.
Table 5.3 Value Supplied When a New Quiet NaN Is Created
Format
New QNaN value
Single floating point
0x7fbf ffff
Double floating point
0x7ff7 ffff ffff ffff
Word fixed point
0x7fff ffff
Longword fixed point
0x7fff ffff ffff ffff
S
Integer
S
Integer
72
5.4.1.6 Paired Single Exceptions
Exception conditions that arise while executing the two halves of a floating point vector operation are ORed together, and the instruction is treated as having caused all the exceptional conditions arising from both operations. The hard- ware makes no effort to determine which of the two operations encountered the exceptional condition.
5.4.1.7 Paired Single Condition Codes
The c.cond.PS instruction compares the upper and lower halves of FPR fs and FPR ft independently and writes the results into condition codes CC +1 and CC respectively. The CC number must be even. If the number is not even the operation of the instruction is UNPREDICTABLE.
5.4.2 Fixed Point Formats
The FPU provides two fixed point data types:
• 32-bit Word fixed point (type W), shown in Figure 5-4
• 64-bit Longword fixed point (type L), shown in Figure 5-5
The fixed point values are held in the 2’s complement format used for signed integers in the CPU. Unsigned fixed point data types are not provided by the architecture; application software may synthesize computations for unsigned integers from the existing instructions and data types.
Figure 5-4 Word Fixed Point Format (W)
33 100
1 31
Figure 5-5 Longword Fixed Point Format (L)
66 320
1 63
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
5.5 Floating Point Register Types
This section describes the organization and use of the two types of FPU register sets:
In Release 1 of the Architecture, 64-bit floating point units were supported only by implementations of the MIPS64 Architecture. Similarly, implementations of MIPS32 of the Architecture only supported 32-bit floating point units. In Release 2 of the Architecture and MIPSr3, a 64-bit floating point unit is supported on implementations of both the MIPS32 and MIPS64 Architectures.
Floating Point registers (FPRs) are 32 or 64 bits wide. A 32-bit floating point unit contains 32 32-bit FPRs, each of which is capable of storing a 32-bit data type. Double-precision (type D) data types are stored in even-odd pairs of FPRs, and the long-integer (type L) and paired single (type PS) data types are not supported. A 64-bit floating point unit contains 32 64-bit FPRs, each of which is capable of storing any data type. For compatibility with 32-bit FPUs, the FR bit in the CP0 Status register is used by a MIPS64 Release 1, or any Release 2 or subsequent releases proces- sor that supports a 64-bit FPU to configure the FPU in a mode in which the FPRs are treated as 32 32-bit registers, each of which is capable of storing only 32-bit data types. In this mode, the double-precision floating point (type D) data type is stored in even-odd pairs of FPRs, and the long-integer (type L) and paired single (type PS) data types are not supported.
• These registers transfer binary data between the FPU and the system, and are also used to hold formatted FPU operand values. Refer to Volume III, The MIPS Privileged Architecture Manual, for more information on the CP0 Registers.
• Floating Point Control registers (FCRs), which are 32 bits wide. There are five FPU control registers, used to identify and control the FPU. These registers are indicated by the fs field of the instruction word. Three of these registers, FCCR, FEXR, and FENR, select subsets of the floating point Control/Status register, the FCSR.
5.5.1 FPU Register Models
There are separate FPU register models in Release 1 of the Architecture:
• MIPS32 defines 32 32-bit registers, with D-format values stored in even-odd pairs of registers.
• MIPS64 defines 32 64-bit registers, with all formats supported in a register.
To support MIPS32 programs, MIPS64 processors also provide the MIPS32 register model, which is available as a mode selection through the FR Bit of the CP0 Status Register.
If the value of FR bit is changed, the contents of the FPRs becomes UNPREDICTABLE. For some implementations, it might be necessary for software to re-initialize the FPRs.
In Release 2 of the Architecture and subsequent releases, both FPU register models are available for implementations, and the FR bit of the CP0 Status Register.
5.5.2 Binary Data Transfers (32-Bit and 64-Bit)
The data transfer instructions move words and doublewords between the FPU FPRs and the remainder of the system. The operations of the word and doubleword load and move-to instructions are shown in Figure 5-6 and Figure 5-7.
The store and move-from instructions operate in reverse, reading data from the location which the corresponding load or move-to instruction wrote.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 73
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
5.5 Floating Point Register Types
Overview of the FPU Instruction Set
Initial value 1
Initial value 2
Initial value 1
Initial value 2
Reg 0 Reg 1
Reg 0 Reg 1
Reg 0 Reg 1
Reg 0 Reg 1
Reg 0 Reg 1
Figure 5-6 FPU Word Load and Move-to Operations
FR BIT = 1 FR BIT = 0
63 0 63 0
Reg 0 Reg 2
LWC1 f0, 0(r0) / MTC1 f0,r0
63 0 63 0 Reg 0
Reg 2
LWC1 f1, 4(r0) / MTC1 f1,r4
63 0 63 0
Reg 0 Reg 2
Figure 5-7 FPU Doubleword Load and Move-to Operations
FR BIT = 1 FR BIT = 0
63 063 0
Undefined/Unused
Data word (0)
Initial value 2
Undefined/Unused
Data word (0)
Initial value 2
Undefined/Unused
Data word (0)
Undefined/Unused
Data word (4)
Data word (4)
Data word (0)
Initial value 2
Initial value 1
Initial value 2
Initial value 1
Initial value 2
Reg 0 Reg 2
LDC1 f0, 0(r0) / DMTC1 f0,r0
63 063 0 Reg 0
Reg 2
LDC1 f1, 8(r0) / DMTC1 f1,r8
Data doubleword (0)
Initial value 2
Data doubleword (0)
Initial value 2
63 0 Reg 0
Reg 1
5.5.3 FPRs and Formatted Operand Layout
(Illegal when FP32RegistersMode = 0)
Data doubleword (0)
Data doubleword (8)
Undefined/Unused
Data word
74
FPU instructions that operate on formatted operand values specify the floating point register (FPR) that holds the value. Operands that are only 32 bits wide (W and S formats), use only half the space in a 64-bit FPR.
The FPR organization and the way that operand data is stored in them is shown in Figures 5-8, 5-9 and 5-10. Figure 5-8 Single Floating Point or Word Fixed Point Operand in an FPR
63 32 31 0 Reg 0
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Figure 5-9 Double Floating Point or Longword Fixed Point Operand in an FPR
63 0 Reg 0 Data doubleword/Longword
Figure 5-10 Paired-Single Floating Point Operand in an FPR
63 32 31 0 Reg 0
5.6 Floating Point Control Registers (FCRs)
The MIPS32 Architecture supports the following five floating point Control registers (FCRs):
• FIR, FP Implementation and Revision register
• FCCR, FP Condition Codes register
• FEXR, FP Exceptions register
• FENR, FP Enables register
• FCSR, FP Control/Status register (used to be known as FCR31).
FCCR, FEXR, and FENR access portions of the FCSR through CTC1 and CFC1 instructions.
Access to the Floating Point Control Registers is not privileged; they can be accessed by any program that can exe- cute floating point instructions. The FCRs can be accessed via the CTC1 and CFC1 instructions.
5.6.1 Floating Point Implementation Register (FIR, CP1 Control Register 0)
Compliance Level: Required if floating point is implemented
The Floating Point Implementation Register (FIR) is a 32-bit read-only register that contains information identifying the capabilities of the floating point unit, the floating point processor identification, and the revision level of the float- ing point unit. Figure 5-11 shows the format of the FIR register; Table 5.4 describes the FIR register fields.
Figure 5-11 FIR Register Format
31 2827 2423 22 2120 19 18171615 8 7
Table 5.4 FIR Register Field Descriptions
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
0
5.6 Floating Point Control Registers (FCRs)
Paired-Single
Paired-Single
0 0000
Impl
0
F64
L
W
3D
PS
D
S
ProcessorID
Revision
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
0
31:28
Reserved for future use; reads as zero
0
0
Reserved
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
75
Overview of the FPU Instruction Set
Table 5.4 FIR Register Field Descriptions (Continued)
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
Impl
27..24
These bits are implementation dependent and are not defined by the architecture, other than the fact that they are read-only. This bits are explicitly not intended to be used for mode control functions.
R
Preset
Optional
0
23
Reserved for future use; reads as zero
0
0
Reserved
F64
22
Indicates that the floating point unit has registers and data paths that are 64-bits wide. This bit was added in Release 2 of the Architecture, and is a one on either any processors with a 64-bit floating point unit, and a zero on any processors with a 32-bit floating point unit. A value of one in this bit indicates that StatusFR is implemented.
Encoding
0 1
Meaning
FPU is 32 bits FPU is 64 bits
R
Preset
Required (Release 2)
L
21
Indicates that the longword fixed point (L) data type and instructions are implemented:
Encoding
Meaning
0
L fixed point not implemented
1
L fixed point implemented
R
Preset
Required (Release 2)
W
20
Indicates that the word fixed point (W) data type and instructions are implemented:
Encoding
Meaning
0
W fixed point not implemented
1
W fixed point implemented
R
Preset or Externally Set
Required (Release 2)
3D
19
In Release 1 of the Architecture, this bit is used by MIPS64 processors to indicate that the MIPS-3D ASE is implemented. It is not used by MIPS32 processors and reads as zero.
In Release 2 of the Architecture and subsequent releases, the MIPS-3D ASE is supported on any processors with a 64-bit floating point unit, and this bit indicates that the MIPS-3D ASE is implemented:
Encoding
Meaning
0
MIPS-3D ASE not implemented
1
MIPS-3D ASE implemented
R
Preset
Required
76 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.4 FIR Register Field Descriptions (Continued)
5.6 Floating Point Control Registers (FCRs)
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
PS
18
In Release 1 of the Architecture, this bit is used by MIPS64 processors to indicate that the paired single floating point data type is implemented. It is not used by MIPS32 Release 1 processors and reads as zero.
In Release 2 of the Architecture and subsequent releases, this bit indicates that the paired single floating point data type is implemented:
Encoding
Meaning
0
PS floating point not implemented
1
PS floating point implemented
R
Preset
Required
D
17
Indicates that the double-precision (D) floating point data type and instructions are implemented:
Encoding
Meaning
0
D floating point not implemented
1
D floating point implemented
R
Preset
Required
S
16
Indicates that the single-precision (S) floating point data type and instructions are implemented:
Encoding
Meaning
0
S floating point not implemented
1
S floating point implemented
R
Preset
Required
Proces- sorID
15:8
Identifies the floating point processor.
R
Preset
Required
Revision
7:0
Specifies the revision number of the floating point unit. This field allows software to distinguish between one revision and another of the same floating point processor type. If this field is not implemented, it must read as zero.
R
Preset
Optional
5.6.2 Floating Point Control and Status Register (FCSR, CP1 Control Register 31)
Compliance Level: Required if floating point is implemented.
The Floating Point Control and Status Register (FCSR) is a 32-bit register that controls the operation of the floating
point unit, and shows the following status information:
• selects the default rounding mode for FPU arithmetic operations
• selectively enables traps of FPU exception conditions
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 77
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
• controls some denormalized number handling options
• reports any IEEE exceptions that arose during the most recently executed instruction
• reports IEEE exceptions that arose, cumulatively, in completed instructions
• indicates the condition code result of FP compare instructions
Access to FCSR is not privileged; it can be read or written by any program that has access to the floating point unit (via the coprocessor enables in the Status register). Figure 5-12 shows the format of the FCSR register; Table 5.5 describes the FCSR register fields.
Figure 5-12 FCSR Register Format
31 30 29 28 27 26 25 24 23 22 21 20 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
Table 5.5 FCSR Register Field Descriptions
FCC
FS
FCC
Impl
0 000
Cause
Enables
Flags
RM
7
6
5
4
3
2
1
0
E
V
Z
O
U
I
V
Z
O
U
I
V
Z
O
U
I
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
FCC
31:25, 23
Floating point condition codes. These bits record the result of floating point compares and are tested for float- ing point conditional branches and conditional moves. The FCC bit to use is specified in the compare, branch, or conditional move instruction. For backward compati- bility with previous MIPS ISAs, the FCC bits are sepa- rated into two, non-contiguous fields.
R/W
Undefined
Required
FS
24
Flush to Zero. When FS is one, denormalized results are flushed to zero instead of causing an Unimplemented Operation exception. It is implementation dependent whether denormalized operand values are flushed to zero before the operation is carried out.
R/W
Undefined
Required
Impl
22:21
Available to control implementation dependent features of the floating point unit. If these bits are not imple- mented, they must be ignored on write and read as zero.
R/W
Undefined
Optional
0
20:18
Reserved for future use; Must be written as zero; returns zero on read.
0
0
Reserved
Cause
17:12
Cause bits. These bits indicate the exception conditions that arise during execution of an FPU arithmetic instruc- tion. A bit is set to 1 if the corresponding exception con- dition arises during the execution of an instruction and is set to 0 otherwise. By reading the registers, the exception condition caused by the preceding FPU arithmetic instruction can be determined.
Refer to Table 5.6 for the meaning of each bit.
R/W
Undefined
Required
78
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.5 FCSR Register Field Descriptions (Continued)
5.6 Floating Point Control Registers (FCRs)
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
Enables
11:7
Enable bits. These bits control whether or not a excep- tion is taken when an IEEE exception condition occurs for any of the five conditions. The exception occurs when both an Enable bit and the corresponding Cause bit are set either during an FPU arithmetic operation or by moving a value to FCSR or one of its alternative repre- sentations. Note that Cause bit E has no corresponding Enable bit; the non-IEEE Unimplemented Operation exception is defined by MIPS as always enabled.
Refer to Table 5.6 for the meaning of each bit.
R/W
Undefined
Required
Flags
6:2
Flag bits. This field shows any exception conditions that have occurred for completed instructions since the flag was last reset by software.
When a FPU arithmetic operation raises an IEEE excep- tion condition that does not result in a Floating Point Exception (i.e., the Enable bit was off), the correspond- ing bit(s) in the Flag field are set, while the others remain unchanged. Arithmetic operations that result in a Floating Point Exception (i.e., the Enable bit was on) do not update the Flag bits.
This field is never reset by hardware and must be explic- itly reset by software.
Refer to Table 5.6 for the meaning of each bit.
R/W
Undefined
Required
RM
1:0
Rounding mode. This field indicates the rounding mode used for most floating point operations (some operations use a specific rounding mode).
Refer to Table 5.7 for the meaning of the encodings of this field.
R/W
Undefined
Required.
The FCC, FS, Cause, Enables, Flags and RM fields in the FCSR, FCCR, FEXR, and FENR registers always display the correct state. That is, if a field is written via FCCR, the new value may be read via one of the alternate registers. Similarly, if a value is written via one of the alternate registers, the new value may be read via FCSR.
Table 5.6 Cause, Enable, and Flag Bit Definitions
Bit Name
Bit Meaning
E
Unimplemented Operation (this bit exists only in the Cause field)
V
Invalid Operation
Z
Divide by Zero
O
Overflow
U
Underflow
I
Inexact
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 79
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Table 5.7 Rounding Mode Definitions
RM Field Encoding
Meaning
0
RN - Round to Nearest
Rounds the result to the nearest representable value. When two representable values are equally near, the result is rounded to the value whose least significant bit is zero (that is, even)
1
RZ - Round Toward Zero
Rounds the result to the value closest to but not greater than in magnitude than the result.
2
RP - Round Towards Plus Infinity
Rounds the result to the value closest to but not less than the result.
3
RM - Round Towards Minus Infinity
Rounds the result to the value closest to but not greater than the result.
0
0000 0000 0000 0000 0000 0000
FCC
7
6
5
4
3
2
1
0
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
0
31:8
Must be written as zero; returns zero on read
0
0
Reserved
FCC
7:0
Floating point condition code. Refer to the description of this field in the FCSR register.
R/W
Undefined
Required
80
5.6.3 Floating Point Condition Codes Register (FCCR, CP1 Control Register 25)
Compliance Level: Required if floating point is implemented.
The Floating Point Condition Codes Register (FCCR) is an alternative way to read and write the floating point con- dition code values that also appear in FCSR. Unlike FCSR, all eight FCC bits are contiguous in FCCR. Figure 5-13 shows the format of the FCCR register; Table 5.8 describes the FCCR register fields.
Figure 5-13 FCCR Register Format
31 870
Table 5.8 FCCR Register Field Descriptions
5.6.4 Floating Point Exceptions Register (FEXR, CP1 Control Register 26)
Compliance Level: Required if floating point is implemented.
The Floating Point Exceptions Register (FEXR) is an alternative way to read and write the Cause and Flags fields that also appear in FCSR. Figure 5-14 shows the format of the FEXR register; Table 5.9 describes the FEXR register fields.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Figure 5-14 FEXR Register Format
31 18 17 16 15 14 13 12 11 7 6 5 4 3 2 1 0
Table 5.9 FEXR Register Field Descriptions
5.6 Floating Point Control Registers (FCRs)
0
0000 0000 0000 00
V
Cause
O
Flags
U
0 00
E
Z
U
I
V
Z
O
I
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
0
31:18, 11:7, 1:0
Must be written as zero; returns zero on read
0
0
Reserved
Cause
17:12
Cause bits. Refer to the description of this field in the FCSR register.
R/W
Undefined
Required
Flags
6:2
Flags bits. Refer to the description of this field in the FCSR register.
R/W
Undefined
Optional
5.6.5 Floating Point Enables Register (FENR, CP1 Control Register 28)
31
Figure 5-15 FENR Register Format
12 11 10 9 8 7 6
Table 5.10 FENR Register Field Descriptions
3 2 1 0
Compliance Level: Required if floating point is implemented.
The Floating Point Enables Register (FENR) is an alternative way to read and write the Enables, FS, and RM fields that also appear in FCSR. Figure 5-15 shows the format of the FENR register; Table 5.10 describes the FENR regis- ter fields.
0
0000 0000 0000 0000 0000
Enables
0 000 0
FS
RM
V
Z
O
U
I
Fields
Description
Read/ Write
Reset State
Compliance
Name
Bits
0
31:12, 6:3
Must be written as zero; returns zero on read
0
0
Reserved
Enables
11:7
Enable bits. Refer to the description of this field in the FCSR register.
R/W
Undefined
Required
FS
2
Flush to Zero bit. Refer to the description of this field in the FCSR register.
R/W
Undefined
Required
RM
1:0
Rounding mode. Refer to the description of this field in the FCSR register.
R/W
Undefined
Required
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
81
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
0 00 000
Overview of the FPU Instruction Set
5.7 Formats of Values Used in FP Registers
Unlike the CPU, the FPU does not interpret the binary encoding of source operands nor produce a binary encoding of results for every operation. The contents of the floating point operand register is interpreted as the format defined by the instruction which is being executed. That is, there is no persistent interpretation of the register values.
5.8 FPU Exceptions
This section provides the following information FPU exceptions:
• Precise exception mode
• Descriptions of the exceptions
• Non-Arithmetic Instructions
FPU exceptions are implemented in the MIPS FPU architecture with the Cause, Enable, and Flag fields of the Con- trol/Status register. The Flag bits implement IEEE exception status flags, and the Cause and Enable bits control exception trapping. Each field has a bit for each of the five IEEE exception conditions and the Cause field has an additional exception bit, Unimplemented Operation, used to trap for software emulation assistance.
5.8.0.1 Precise Exception Mode
In precise exception mode, a trap occurs before the instruction that causes the trap, or any following instruction, can complete and write its results. If desired, the software trap handler can resume execution of the interrupted instruction stream after handling the exception.
The Cause field reports per-bit instruction exception conditions. The Cause bits are written during each floating point arithmetic operation to show any exception conditions that arise during the operation. The bit is set to 1 if the corre- sponding exception condition arises; otherwise it is set to 0.
A floating point trap is generated any time both a Cause bit and its corresponding Enable bit are set. This occurs either during the execution of a floating point operation or by moving a value into the FCSR. There is no Enable for Unimplemented Operation; this exception always generates a trap.
In a trap handler, exception conditions that arise during any trapped floating point operations are reported in the Cause field. Before returning from a floating point interrupt or exception, or before setting Cause bits with a move to the FCSR, software must first clear the enabled Cause bits by executing a move to FCSR to prevent the trap from being erroneously retaken.
User-mode programs cannot observe enabled Cause bits being set. If this information is required in a User-mode han- dler, it must be available someplace other than through the Status register.
If a floating point operation sets only non-enabled Cause bits, no trap occurs and the default result defined by the IEEE standard is stored (see Table 5.11). When a floating point operation does not trap, the program can monitor the exception conditions by reading the Cause field.
The Flag field is a cumulative report of IEEE exception conditions that arise as instructions complete; instructions that trap do not update the Flag bits. The Flag bits are set to 1 if the corresponding IEEE exception is raised, other- wise the bits are unchanged. There is no Flag bit for the MIPS Unimplemented Operation exception. The Flag bits are never cleared as a side effect of floating point operations, but may be set or cleared by moving a new value into the FCSR.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
82
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Addressing exceptions are precise.
5.8.1 Exception Conditions
The following five exception conditions defined by the IEEE standard are described in this section:
• Invalid Operation Exception
• Division By Zero Exception
• Underflow Exception
• Overflow Exception
• Inexact Exception
This section also describes a MIPS-specific exception condition, Unimplemented Operation, that is used to signal a need for software emulation of an instruction. Normally an IEEE arithmetic operation can cause only one exception condition; the only case in which two exceptions can occur at the same time are Inexact With Overflow and Inexact With Underflow.
At the program’s direction, an IEEE exception condition can either cause a trap or not cause a trap. The IEEE stan- dard specifies the result to be delivered in case the exception is not enabled and no trap is taken. The MIPS architec- ture supplies these results whenever the exception condition does not result in a precise trap (that is, no trap or an imprecise trap). The default action taken depends on the type of exception condition, and in the case of the Overflow, the current rounding mode. The default results are summarized in Table 5.11.
Table 5.11 Default Result for IEEE Exceptions Not Trapped Precisely
5.8 FPU Exceptions
Bit
Description
Default Action
V
Invalid Operation
Supplies a quiet NaN.
Z
Divide by zero
Supplies a properly signed infinity.
U
Underflow
Supplies a rounded result.
I
Inexact
Supplies a rounded result. If caused by an overflow without the overflow trap enabled, sup- plies the overflowed result.
O
Overflow
Depends on the rounding mode, as shown below.
0 (RN)
Supplies an infinity with the sign of the intermediate result.
1 (RZ)
Supplies the format’s largest finite number with the sign of the intermediate result.
2 (RP)
For positive overflow values, supplies positive infinity. For negative overflow values, sup- plies the format’s most negative finite number.
3 (RM)
For positive overflow values, supplies the format’s largest finite number. For negative over- flow values, supplies minus infinity.
5.8.1.1 Invalid Operation Exception
The Invalid Operation exception is signaled if one or both of the operands are invalid for the operation to be per- formed. The result, when the exception condition occurs without a precise trap, is a quiet NaN.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 83
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
84
These are invalid operations:
• One or both operands are a signaling NaN (except for the non-arithmetic MOV.fmt, MOVT.fmt, MOVF.fmt, MOVN.fmt, and MOVZ.fmt instructions).
• Addition or subtraction: magnitude subtraction of infinities, such as (+∞) + (-∞) or (-∞) - (-∞).
• Multiplication: 0 × ∞, with any signs.
• Division: 0/0 or ∞/∞, with any signs.
• Square root: An operand of less than 0 (-0 is a valid operand value).
• Conversion of a floating point number to a fixed point format when either an overflow or an operand value of infinity or NaN precludes a faithful representation in that format.
• Some comparison operations in which one or both of the operands is a QNaN value. (The detailed definition of the compare instruction, C.cond.fmt, in Volume II has tables showing the comparisons that do and do not signal the exception.)
5.8.1.2 Division By Zero Exception
An implemented divide operation signals a Division By Zero exception if the divisor is zero and the dividend is a finite nonzero number. The result, when no precise trap occurs, is a correctly signed infinity. Divisions (0/0) and (∞/0) do not cause the Division By Zero exception. The result of (0/0) is an Invalid Operation exception. The result of (∞/0) is a correctly signed infinity.
5.8.1.3 Underflow Exception
Two related events contribute to underflow:
• Tininess: the creation of a tiny nonzero result between ±2E_min which, because it is tiny, may cause some other exception later such as overflow on division
• Loss of accuracy: the extraordinary loss of accuracy during the approximation of such tiny numbers by denor- malized numbers
Tininess: The IEEE standard allows choices in detecting these events, but requires that they be detected in the same manner for all operations. The IEEE standard specifies that “tininess” may be detected at either of these times:
• After rounding, when a nonzero result computed as though the exponent range were unbounded would lie strictly between ±2E_min
• Before rounding, when a nonzero result computed as though both the exponent range and the precision were unbounded would lie strictly between ±2E_min
The MIPS architecture specifies that tininess be detected after rounding.
Loss of Accuracy: The IEEE standard specifies that loss of accuracy may be detected as a result of either of these conditions:
• Denormalization loss, when the delivered result differs from what would have been computed if the exponent range were unbounded
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
• Inexact result, when the delivered result differs from what would have been computed if both the exponent range and precision were unbounded
The MIPS architecture specifies that loss of accuracy is detected as inexact result.
Signalling an Underflow: When an underflow trap is not enabled, underflow is signaled only when both tininess and
loss of accuracy have been detected. The delivered result might be zero, denormalized, or ±2E_min.
When an underflow trap is enabled (through the FCSR Enable field bit), underflow is signaled when tininess is
detected regardless of loss of accuracy.
5.8.1.4 Overflow Exception
An Overflow exception is signaled when the magnitude of a rounded floating point result, were the exponent range unbounded, is larger than the destination format’s largest finite number.
When no precise trap occurs, the result is determined by the rounding mode and the sign of the intermediate result.
5.8.1.5 Inexact Exception
An Inexact exception is signaled if one of the following occurs:
• The rounded result of an operation is not exact
• The rounded result of an operation overflows without an overflow trap
5.8.1.6 Unimplemented Operation Exception
The Unimplemented Operation exception is a MIPS defined exception that provides software emulation support. This exception is not IEEE-compliant.
The MIPS architecture is designed so that a combination of hardware and software may be used to implement the architecture. Operations that are not fully supported in hardware cause an Unimplemented Operation exception so that software may perform the operation.
There is no Enable bit for this condition; it always causes a trap. After the appropriate emulation or other operation is done in a software exception handler, the original instruction stream can be continued.
5.8.1.7 Non-Arithmetic Instructions
Some FPU conversion and FPU Formatted Operand-Value Move instructions (see next section) do not perform float- ing-point arithmetic operations on their input operands. For that reason, such instructions do not generate IEEE arith- metic exceptions. These instructions include MOV.fmt, MOVF.fmt, MOVT.fmt, MOVZ.fmt, MOVN.fmt.
5.9 FPU Instructions
The FPU instructions comprise the following functional groups:
• Data Transfer Instructions
• Arithmetic Instructions
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 85
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
5.9 FPU Instructions
Overview of the FPU Instruction Set
• Conversion Instructions
• Formatted Operand-Value Move Instructions
• Conditional Branch Instructions
• Miscellaneous Instructions
5.9.1 Data Transfer Instructions
The FPU has two separate register sets: coprocessor general registers and coprocessor control registers. The FPU has a load/store architecture; all computations are done on data held in coprocessor general registers. The control regis- ters are used to control FPU operation. Data is transferred between registers and the rest of the system with dedicated load, store, and move instructions. The transferred data is treated as unformatted binary data; no format conversions are performed, and therefore no IEEE floating point exceptions can occur.
The supported transfer operations are listed in Table 5.12.
Table 5.12 FPU Data Transfer Instructions
5.9.1.1 Data Alignment in Loads, Stores, and Moves
All coprocessor loads and stores operate on naturally-aligned data items. An attempt to load or store to an address that is not naturally aligned for the data item causes an Address Error exception. Regardless of byte-ordering (the endian- ness), the address of a word or doubleword is the smallest byte address in the object. For a big-endian machine, this is the most-significant byte; for a little-endian machine, this is the least-significant byte (endianness is described in “Byte Ordering and Endianness” on page 37).
5.9.1.2 Addressing Used in Data Transfer Instructions
The FPU has loads and stores using the same register+offset addressing as that used by the CPU. Moreover, for the FPU only, there are load and store instructions using register+register addressing.
Tables 5.13 through 5.15 list the FPU data transfer instructions.
Table 5.13 FPU Loads and Stores Using Register+Offset Address Mode
Transfer Direction
Data Transferred
FPU general reg
↔
Memory
Word/doubleword load/store
FPU general reg
↔
CPU general reg
Word move
FPU control reg
↔
CPU general reg
Word move
Mnemonic
Instruction
Defined in MIPS ISA
LDC1
Load Doubleword to Floating Point
MIPS32
LWC1
Load Word to Floating Point
MIPS32
SDC1
Store Doubleword to Floating Point
MIPS32
SWC1
Store Word to Floating Point
MIPS32
86
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.14 FPU Loads and Using Register+Register Address Mode
5.9 FPU Instructions
Mnemonic
Instruction
Defined in MIPS ISA
LDXC1
Load Doubleword Indexed to Floating Point
MIPS64 MIPS32 Release 2
LUXC1
Load Doubleword Indexed Unaligned to Floating Point
MIPS64 MIPS32 Release 2
LWXC1
Load Word Indexed to Floating Point
MIPS64 MIPS32 Release 2
SDXC1
Store Doubleword Indexed to Floating Point
MIPS64 MIPS32 Release 2
SUXC1
Store Doubleword Indexed Unaligned to Floating Point
MIPS64 MIPS32 Release 2
SWXC1
Store Word Indexed to Floating Point
MIPS64 MIPS32 Release 2
Table 5.15 FPU Move To and From Instructions
Mnemonic
Instruction
Defined in MIPS ISA
CFC1
Move Control Word From Floating Point
MIPS32
CTC1
Move Control Word To Floating Point
MIPS32
MFC1
Move Word From Floating Point
MIPS32
MFHC1
Move Word from High Half of Floating Point Register
MIPS32 Release 2
MTC1
Move Word To Floating Point
MIPS32
MTHC1
Move Word to High Half of Floating Point Register
MIPS32 Release 2
5.9.2 Arithmetic Instructions
Arithmetic instructions operate on formatted data values. The results of most floating point arithmetic operations meet the IEEE standard specification for accuracy—a result is identical to an infinite-precision result that has been rounded to the specified format, using the current rounding mode. The rounded result differs from the exact result by less than one unit in the least-significant place (ULP).
FPU IEEE-approximate arithmetic operations are listed in Table 5.16.
Table 5.16 FPU IEEE Arithmetic Operations
Mnemonic
Instruction
Defined in MIPS ISA
ABS.fmt
Floating Point Absolute Value
MIPS32
ABS.fmt (PS)
Floating Point Absolute Value (Paired Single)
MIPS64 MIPS32 Release 2
ADD.fmt
Floating Point Add
MIPS32
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 87
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Table 5.16 FPU IEEE Arithmetic Operations (Continued)
Mnemonic
Instruction
Defined in MIPS ISA
ADD.fmt (PS)
Floating Point Add (Paired Single)
MIPS64 MIPS32 Release 2
C.cond.fmt
Floating Point Compare
MIPS32
C.cond.fmt (PS)
Floating Point Compare (Paired Single)
MIPS64 MIPS32 Release 2
DIV.fmt
Floating Point Divide
MIPS32
MUL.fmt
Floating Point Multiply
MIPS32
MUL.fmt (PS)
Floating Point Multiply (Paired Single)
MIPS64 MIPS32 Release 2
NEG.fmt
Floating Point Negate
MIPS32
NEG.fmt (PS)
Floating Point Negate (Paired Single)
MIPS64 MIPS32 Release 2
SQRT.fmt
Floating Point Square Root
MIPS32
SUB.fmt
Floating Point Subtract
MIPS32
SUB.fmt (PS)
Floating Point Subtract (Paired Single)
MIPS64 MIPS32 Release 2
Mnemonic
Instruction
Defined in MIPS ISA
RECIP.fmt
Floating Point Reciprocal Approximation
MIPS64 MIPS32 Release 2
RSQRT.fmt
Floating Point Reciprocal Square Root Approximation
MIPS64 MIPS32 Release 2
88
Two operations, Reciprocal Approximation (RECIP) and Reciprocal Square Root Approximation (RSQRT), may be less accurate than the IEEE specification:
• The result of RECIP differs from the exact reciprocal by no more than one ULP.
• The result of RSQRT differs from the exact reciprocal square root by no more than two ULPs. Within these error limits, the results of these instructions are implementation specific.
A list of FPU-approximate arithmetic operations is given in Table 5.17..
Table 5.17 FPU-Approximate Arithmetic Operations
Four compound-operation instructions perform variations of multiply-accumulate—that is, multiply two operands, accumulate the result to a third operand, and produce a result. These instructions are listed in Table 5.18. The product is rounded according to the current rounding mode prior to the accumulation. The accumulated result is also rounded. This model meets the IEEE-754-1985 accuracy specification; the result is numerically identical to an equivalent com- putation using a sequence of multiply, add/subtract, or negate instructions. Similarly, exceptions and flags behave as if the operation was implemented with a sequence of multiply, add/subtract and negate instructions. This behavior is often known as “Non-Fused”.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.18 lists the FPU Multiply-Accumulate arithmetic operations.
Table 5.18 FPU Multiply-Accumulate Arithmetic Operations
5.9 FPU Instructions
Mnemonic
Instruction
Defined in MIPS ISA
MADD.fmt
Floating Point Multiply Add
MIPS64 MIPS32 Release 2
MADD.fmt (PS)
Floating Point Multiply Add (Paired Single)
MIPS64 MIPS32 Release 2
MSUB.fmt
Floating Point Multiply Subtract
MIPS64 MIPS32 Release 2
MSUB.fmt (PS)
Floating Point Multiply Subtract (Paired Single)
MIPS64 MIPS32 Release 2
NMADD.fmt
Floating Point Negative Multiply Add
MIPS64 MIPS32 Release 2
NMADD.fmt (PS)
Floating Point Negative Multiply Add (Paired Single)
MIPS64 MIPS32 Release 2
NMSUB.fmt
Floating Point Negative Multiply Subtract
MIPS64 MIPS32 Release 2
NMSUB.fmt (PS)
Floating Point Negative Multiply Subtract (Paired Single)
MIPS64 MIPS32 Release 2
5.9.3 Conversion Instructions
These instructions perform conversions between floating point and fixed point data types. Each instruction converts values from a number of operand formats to a particular result format. Some conversion instructions use the rounding mode specified in the Floating Control/Status register (FCSR), while others specify the rounding mode directly. Tables 5.19 and 5.20 list the FPU conversion instructions according to their rounding mode.
Table 5.19 FPU Conversion Operations Using the FCSR Rounding Mode
Mnemonic
Instruction
Defined in MIPS ISA
CVT.D.fmt
Floating Point Convert to Double Floating Point
MIPS32
CVT.L.fmt
Floating Point Convert to Long Fixed Point
MIPS64 MIPS32 Release 2
CVT.PS.S
Floating Point Convert Pair to Paired Single
MIPS64 MIPS32 Release 2
CVT.S.fmt
Floating Point Convert to Single Floating Point
MIPS32
CVT.S.fmt (PL, PU)
Floating Point Convert to Single Floating Point (Paired Lower, Paired Upper)
MIPS64 MIPS32 Release 2
CVT.W.fmt
Floating Point Convert to Word Fixed Point
MIPS32
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 89
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Table 5.20 FPU Conversion Operations Using a Directed Rounding Mode
Mnemonic
Instruction
Defined in MIPS ISA
CEIL.L.fmt
Floating Point Ceiling to Long Fixed Point
MIPS64 MIPS32 Release 2
CEIL.W.fmt
Floating Point Ceiling to Word Fixed Point
MIPS32
FLOOR.L.fmt
Floating Point Floor to Long Fixed Point
MIPS64 MIPS32 Release 2
FLOOR.W.fmt
Floating Point Floor to Word Fixed Point
MIPS32
ROUND.L.fmt
Floating Point Round to Long Fixed Point
MIPS64 MIPS32 Release 2
ROUND.W.fmt
Floating Point Round to Word Fixed Point
MIPS32
TRUNC.L.fmt
Floating Point Truncate to Long Fixed Point
MIPS64 MIPS32 Release 2
TRUNC.W.fmt
Floating Point Truncate to Word Fixed Point
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
MOV.fmt
Floating Point Move
MIPS32
MOV.fmt (PS)
Floating Point Move (Paired Single)
MIPS64 MIPS32 Release 2
Mnemonic
Instruction
Defined in MIPS ISA
MOVF.fmt
Floating Point Move Conditional on FP False
MIPS32
90
5.9.4 Formatted Operand-Value Move Instructions
These instructions all move formatted operand values among FPU general registers. A particular operand type must be moved by the instruction that handles that type. There are three kinds of move instructions:
• Unconditional move
• Conditional move that tests an FPU true/false condition code
• Conditional move that tests a CPU general-purpose register against zero
Conditional move instructions operate in a way that may be unexpected. They always force the value in the destina- tion register to become a value of the format specified in the instruction. If the destination register does not contain an operand of the specified format before the conditional move is executed, the contents become UNPREDICTABLE. (For more information, see the individual descriptions of the conditional move instructions in Volume II.)
These instructions are listed in Tables 5.21 through 5.23.
Table 5.21 FPU Formatted Operand Move Instructions
Table 5.22 FPU Conditional Move on True/False Instructions
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.22 FPU Conditional Move on True/False Instructions
5.9 FPU Instructions
Mnemonic
Instruction
Defined in MIPS ISA
MOVF.fmt (PS)
Floating Point Move Conditional on FP False (Paired Single)
MIPS64 MIPS32 Release 2
MOVT.fmt
Floating Point Move Conditional on FP True
MIPS32
MOVT.fmt (PS)
Floating Point Move Conditional on FP True (Paired Single)
MIPS64 MIPS32 Release 2
Table 5.23 FPU Conditional Move on Zero/Nonzero Instructions
Mnemonic
Instruction
Defined in MIPS ISA
MOVN.fmt
Floating Point Move Conditional on Nonzero
MIPS32
MOVN.fmt (PS)
Floating Point Move Conditional on Nonzero (Paired Single)
MIPS64 MIPS32 Release 2
MOVZ.fmt
Floating Point Move Conditional on Zero
MIPS32
MOVZ.fmt (PS)
Floating Point Move Conditional on Zero (Paired Single)
MIPS64 MIPS32 Release 2
5.9.5 Conditional Branch Instructions
The FPU has PC-relative conditional branch instructions that test condition codes set by FPU compare instructions (C.cond.fmt).
All branches have an architectural delay of one instruction. When a branch is taken, the instruction immediately fol- lowing the branch instruction is said to be in the branch delay slot, and it is executed before the branch to the target instruction takes place. Conditional branches come in two versions, depending upon how they handle an instruction in the delay slot when the branch is not taken and execution falls through:
• Branch instructions execute the instruction in the delay slot.
Branch likely instructions do not execute the instruction in the delay slot if the branch is not taken (they are said
to nullify the instruction in the delay slot).
Although the Branch Likely instructions are included in this specification, software is strongly encouraged to avoid the use of the Branch Likely instructions, as they will be removed from a future revision of the MIPS Architecture.
The MIPS32 Architecture defines eight condition codes for use in compare and branch instructions. For backward compatibility with previous revision of the ISA, condition code bit 0 and condition code bits 1 thru 7 are in discontig- uous fields in FCSR.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 91
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Table 5.24 lists the conditional branch FPU instructions; Table 5.25 lists the deprecated conditional branch likely instructions.
Table 5.24 FPU Conditional Branch Instructions
Table 5.25 Deprecated FPU Conditional Branch Likely Instructions
5.9.6 Miscellaneous Instructions
The MIPS ISA defines various miscellaneous instructions that conditionally move one CPU general register to another, based on an FPU condition code. It also defines an instruction to align a misaligned pair of paired-single val- ues (ALNV.PS) and a quartet of instructions that merge a pair of paired-single values (PLL.PS, PLU.PS, PUL.PS, PUU.PS).
Table 5.26 lists these conditional move instructions.
Table 5.26 CPU Conditional Move on FPU True/False Instructions
Mnemonic
Instruction
Defined in MIPS ISA
BC1F
Branch on FP False
MIPS32
BC1T
Branch on FP True
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
BC1FL
Branch on FP False Likely
MIPS32
BC1TL
Branch on FP True Likely
MIPS32
Mnemonic
Instruction
Defined in MIPS ISA
ALNV.PS
FP Align Variable
MIPS64 MIPS32 Release 2
MOVN
Move Conditional on FP False
MIPS32
MOVZ
Move Conditional on FP True
MIPS32
PLL.PS
Pair Lower Lower
MIPS64 MIPS32 Release 2
PLU.PS
Pair Lower Upper
MIPS64 MIPS32 Release 2
PUL.PS
Pair Upper Lower
MIPS64 MIPS32 Release 2
PUU.PS
Pair Upper Upper
MIPS64 MIPS32 Release 2
5.10 Valid Operands for FPU Instructions
The floating point unit arithmetic, conversion, and operand move instructions operate on formatted values with dif- ferent precision and range limits and produce formatted values for results. Each representable value in each format has a binary encoding that is read from or stored to memory. The fmt or fmt3 field of the instruction encodes the oper-
92 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
and format required for the instruction. A conversion instruction specifies the result type in the function field; the result of other operations is given in the same format as the operands. The encodings of the fmt and fmt3 field are shown in Table 5.27.
Table 5.27 FPU Operand Format Field (fmt, fmt3) Encoding
5.10 Valid Operands for FPU Instructions
fmt
fmt3
Instruction Mnemonic
Size
Data Type
Name
Bits
0-15
-
Reserved
16
0
S
single
32
Floating point
17
1
D
double
64
Floating point
18-19
2-3
Reserved
20
4
W
word
32
Fixed point
21
5
L
long
64
Fixed point
22
6
PS
paired single
64 (2x32)
Floating point
23–31
7
Reserved
The result of an instruction using operand formats marked U in Table 5.28 is not currently specified by this architec- ture and causes a Reserved Instruction exception.
Table 5.28 Valid Formats for FPU Operations
Mnemonic
Operation
Operand Fmt
COP1 Function Value
COP1X op4 Value
Float
Fixed
S
D
PS
W
L
ABS
Absolute value
•
•
•
U
U
5
ADD
Add
•
•
•
U
U
0
C.cond
Floating Point compare
•
•
•
U
U
48–63
CEIL.L, (CEIL.W)
Convert to longword (word) fixed point, round toward +∞
•
•
U
U
U
10 (14)
CVT.D
Convert to double floating point
•
U
U
•
•
33
CVT.L
Convert to longword fixed point
•
•
U
U
U
37
CVT.S
Convert to single floating point
U
•
U
•
•
32
CVT. PU, PL
Convert to single floating point (paired upper, paired lower)
U
U
•
U
U
32, 40
CVT.W
Convert to 32-bit fixed point
•
•
U
U
U
36
DIV
Divide
•
•
U
U
U
3
FLOOR.L, (FLOOR.W)
Convert to longword (word) fixed point, round toward −∞
•
•
U
U
U
11 (15)
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 93
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
Table 5.28 Valid Formats for FPU Operations (Continued)
Mnemonic
Operation
Operand Fmt
COP1 Function Value
COP1X op4 Value
Float
Fixed
S
D
PS
W
L
MADD
Multiply-Add
•
•
•
U
U
4
MOV
Move Register
•
•
•
U
U
6
MOVC
FP Move conditional on condition
•
•
•
U
U
17
MOVN
FP Move conditional on GPR≠zero
•
•
•
U
U
19
MOVZ
FP Move conditional on GPR=zero
•
•
•
U
U
18
MSUB
Multiply-Subtract
•
•
•
U
U
5
MUL
Multiply
•
•
•
U
U
2
NEG
Negate
•
•
•
U
U
7
NMADD
Negative Multiply-Add
•
•
•
U
U
6
NMSUB
Negative Multiply-Subtract
•
•
•
U
U
7
PLL, PLU, PUL, PUU
Pair (Lower Lower, Lower Upper, Upper Lower, Upper Upper)
U
U
•
U
U
44-47
RECIP
Reciprocal Approximation
•
•
U
U
U
21
ROUND.L, (ROUND.W)
Convert to longword (word) fixed point, round to nearest/even
•
•
U
U
U
8 (12)
RSQRT
Reciprocal square root approximation
•
•
U
U
U
22
SQRT
Square Root
•
•
U
U
U
4
SUB
Subtract
•
•
•
U
U
1
TRUNC.L, (TRUNC.W)
Convert to longword (word) fixed point, round toward zero
•
•
U
U
U
9 (13)
Key: • − Valid. U − Unimplemented and causes Reserved Instruction Exception.
5.11 FPU Instruction Formats
An FPU instruction is a single 32-bit aligned word. FP instruction formats are shown in Figures 5-16 through 5-25.
In these figures, variables are labelled in lowercase, such as offset. Constants are labelled in uppercase, as are numer- als. Following these figures, Table 5.29 explains the fields used in the instruction layouts. Note that the same field may have different names in different instruction layouts.
The field name is mnemonic to the function of that field in the instruction layout. The opcode tables and the instruc- tion encode discussion use the canonical field names: opcode, fmt, nd, tf, and function. The remaining fields are not used for instruction encode.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
94
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
5.11.1 Implementation Note
31
31
31
31
31
31
31
When present, the destination FPR specifier may be in the fs, ft, or fd field.
Figure 5-16 I-Type (Immediate) FPU Instruction Format
26 25 21 20 16 15
655 16
Immediate: Load/Store using register + offset addressing
Figure 5-17 R-Type (Register) FPU Instruction Format
26 25 21 20 16 15 11 10
6 5
0
0
0
0
0
0
0
5.11 FPU Instruction Formats
opcode
base
ft
offset
COP1
fmt
ft
fs
fd
function
655556
Register: Two-register and Three-register formatted arithmetic operations
Figure 5-18 Register-Immediate FPU Instruction Format
26 25 21 20 16 15 11
6 5 5 5 11
Register Immediate: Data transfer, CPU ↔ FPU register
Figure 5-19 Condition Code, Immediate FPU Instruction Format
26 25 2120 18171615
65311 16
Condition Code, Immediate: Conditional branches on FPU cc using PC + offset
Figure 5-20 Formatted FPU Compare Instruction Format
26 25 21 20 16 15 11 10 8 7 6 5
6555326
Register to Condition Code: Formatted FP compare
Figure 5-21 FP RegisterMove, Conditional Instruction Format
26 25 21 20 18 17 16 15 11 10 6 5
65311556
Condition Code, Register FP: FPU register move-conditional on FP, cc
Figure 5-22 Four-Register Formatted Arithmetic FPU Instruction Format
26 25 21 20 16 15 11 10 6 5
6555533
Register-4: Four-register formatted arithmetic operations
COP1
sub
rt
fs
0
COP1
BCC1
cc
nd
tf
offset
COP1
fmt
ft
fs
cc
0
function
COP1
fmt
cc
0
tf
fs
fd
MOVCF
COP1X
fr
ft
fs
fd
op4
fmt3
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
95
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
31
31
31
Figure 5-23 Register Index FPU Instruction Format
26 25 21 20 16 15 11 10 6 5 0
655556
Register Index: Load and Store using register + register addressing
Figure 5-24 Register Index Hint FPU Instruction Format
26 25 21 20 16 15 11 10 6 5 0
655556
Register Index Hint: Prefetch using register + register addressing
Figure 5-25 Condition Code, Register Integer FPU Instruction Format
26 25 21 20 18 17 16 15 11 10 6 5 0
65311556
Condition Code, Register Integer: CPU register move-conditional on FP, cc
Table 5.29 FPU Instruction Format Fields
COP1X
base
index
0
fd
function
COP1X
base
index
hint
0
PREFX
SPECIAL
rs
cc
0
tf
rd
0
MOVCI
Field
Description
BC1
Branch Conditional instruction subcode (op=COP1).
base
CPU register: base address for address calculations.
COP1
Coprocessor 1 primary opcode value in op field.
COP1X
Coprocessor 1 eXtended primary opcode value in op field.
cc
Condition Code specifier; for architectural levels prior to MIPS IV, this must be set to zero.
fd
FPU register: destination (arithmetic, loads, move-to) or source (stores, move-from).
fmt
Destination and/or operand type (format) specifier.
fr
FPU register: source.
fs
FPU register: source.
ft
FPU register: source (for stores, arithmetic) or destination (for loads).
function
Field specifying a function within a particular op operation code.
function: op4 + fmt3
op4 is a 3-bit function field specifying a 4-register arithmetic operation for COP1X. fmt3 is a 3- bit field specifying the format of the operands and destination. The combinations are shown as distinct instructions in the opcode tables.
hint
Hint field made available to cache controller for prefetch operation.
index
CPU register that holds the index address component for address calculations.
MOVC
Value in function field for a conditional move. There is one value for the instruction when op=COP1, another value for the instruction when op=SPECIAL.
96
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table 5.29 FPU Instruction Format Fields (Continued)
5.11 FPU Instruction Formats
Field
Description
nd
Nullify delay. If set, the branch is Likely, and the delay slot instruction is not executed.
offset
Signed offset field used in address calculations.
op
Primary operation code (see COP1, COP1X, LWC1, SWC1, LDC1, SDC1, SPECIAL).
PREFX
Value in function field for prefetch instruction when op=COP1X.
rd
CPU register: destination.
rs
CPU register: source.
rt
CPU register: can be either source or destination.
SPECIAL
SPECIAL primary opcode value in op field.
sub
Operation subcode field for COP1 register immediate-mode instructions.
tf
True/False. The condition from an FP compare that is tested for equality with the tf bit.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 97
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Overview of the FPU Instruction Set
98 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Appendix A
Instruction Bit Encodings
A.12 Instruction Encodings and Instruction Classes
Instruction encodings are presented in this section; field names are printed here and throughout the book in italics. When encoding an instruction, the primary opcode field is encoded first. Most opcode values completely specify an
instruction that has an immediate value or offset.
Opcode values that do not specify an instruction instead specify an instruction class. Instructions within a class are further specified by values in other fields. For instance, opcode REGIMM specifies the immediate instruction class, which includes conditional branch and trap immediate instructions.
A.13 Instruction Bit Encoding Tables
This section provides various bit encoding tables for the instructions of the MIPS32® ISA.
Figure A.26 shows a sample encoding table and the instruction opcode field this table encodes. Bits 31..29 of the opcode field are listed in the leftmost columns of the table. Bits 28..26 of the opcode field are listed along the topmost rows of the table. Both decimal and binary values are given, with the first three bits designating the row, and the last three bits designating the column.
An instruction’s encoding is found at the intersection of a row (bits 31..29) and column (bits 28..26) value. For instance, the opcode value for the instruction labelled EX1 is 33 (decimal, row and column), or 011011 (binary). Sim- ilarly, the opcode value for EX2 is 64 (decimal), or 110100 (binary).
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 99
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Instruction Bit Encodings
31
26 25 21 20 16 15 0
Figure A.26 Sample Bit Encoding Table
opcode
rs
rt
immediate
655 16
Binary encoding of opcode (28..26)
Decimal encoding of opcode (28..26)
opcode
bits 28..26
bits 31..29
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
1
001
2
010
3
011
EX1
4
100
EX2
5
101
6
110
7
111
Decimal encoding of opcode (31..29)
Binary encoding of opcode (31..29)
Tables A.31 through A.49 describe the encoding used for the MIPS32 ISA. Table A.30 describes the meaning of the symbols used in the tables.
Table A.30 Symbols Used in the Instruction Encoding Tables
Symbol
Meaning
∗
Operation or field codes marked with this symbol are reserved for future use. Executing such an instruction must cause a Reserved Instruction Exception.
δ
(Also italic field name.) Operation or field codes marked with this symbol denotes a field class. The instruction word must be further decoded by examining additional tables that show values for another instruction field.
β
Operation or field codes marked with this symbol represent a valid encoding for a higher-order MIPS ISA level or a new revision of the Architecture. Executing such an instruction must cause a Reserved Instruction Exception.
∇
Operation or field codes marked with this symbol represent instructions which were only legal if 64-bit operations were enabled on implementations of Release 1 of the Architecture. In Release 2 of the architecture, operation or field codes marked with this symbol represent instructions which are legal if 64-bit floating point operations are enabled. In other cases, executing such an instruc- tion must cause a Reserved Instruction Exception (non-coprocessor encodings or coprocessor instruction encodings for a coprocessor to which access is allowed) or a Coprocessor Unusable Exception (coprocessor instruction encodings for a coprocessor to which access is not allowed).
100
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table A.30 Symbols Used in the Instruction Encoding Tables (Continued)
A.13 Instruction Bit Encoding Tables
Symbol
Meaning
θ
Operation or field codes marked with this symbol are available to licensed MIPS partners. To avoid multiple conflicting instruction definitions, MIPS Technologies will assist the partner in selecting appropriate encodings if requested by the partner. The partner is not required to consult with MIPS Technologies when one of these encodings is used. If no instruction is encoded with this value, executing such an instruction must cause a Reserved Instruction Exception (SPECIAL2 encodings or coprocessor instruction encodings for a coprocessor to which access is allowed) or a Coprocessor Unusable Exception (coprocessor instruction encodings for a copro- cessor to which access is not allowed).
σ
Field codes marked with this symbol represent an EJTAG support instruction and implementa- tion of this encoding is optional for each implementation. If the encoding is not implemented, executing such an instruction must cause a Reserved Instruction Exception. If the encoding is implemented, it must match the instruction encoding as shown in the table.
ε
Operation or field codes marked with this symbol are reserved for MIPS Application Specific Extensions. If the ASE is not implemented, executing such an instruction must cause a Reserved Instruction Exception.
φ
Operation or field codes marked with this symbol are obsolete and will be removed from a future revision of the MIPS32 ISA. Software should avoid using these operation or field codes.
⊕
Operation or field codes marked with this symbol are valid for Release 2 implementations of the architecture. Executing such an instruction in a Release 1 implementation must cause a Reserved Instruction Exception.
opcode
bits 28..26
Table A.31 MIPS32 Encoding of the Opcode Field
bits 31..29
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
SPECIAL δ
ADDI
COP0 δ
β
LB
SB
LL
SC
1
001
REGIMM δ
ADDIU
COP1 δ
β
LH
SH
LWC1
SWC1
2
010
J
SLTI
COP2 θδ
β
LWL
SWL
LWC2 θ
SWC2 θ
3
011
JAL
SLTIU
COP1X1 δ
β
LW
SW
PREF
∗
4
100
BEQ
ANDI
BEQL φ
SPECIAL2 δ
LBU
β
β
β
5
101
BNE
ORI
BNEL φ
JALX ε
LHU
β
LDC1
SDC1
6
110
BLEZ
XORI
BLEZL φ
ε
LWR
SWR
LDC2 θ
SDC2 θ
7
111
BGTZ
LUI
BGTZL φ
SPECIAL32 δ⊕
β
CACHE
β
β
1. In Release 1 of the Architecture, the COP1X opcode was called COP3, and was available as another user-available coprocessor. In Release 2 of the Architecture, a full 64-bit floating point unit is available with 32-bit CPUs, and the COP1X opcode is reserved for that purpose on all Release 2 CPUs. 32-bit implementations of Release 1 of the architecture are strongly discouraged from using this opcode for a user-available coprocessor as doing so will limit the potential for an upgrade path to a 64-bit floating point unit.
2. Release 2 of the Architecture added the SPECIAL3 opcode. Implementations of Release 1 of the Architecture sig- naled a Reserved Instruction Exception for this opcode.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 101
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Instruction Bit Encodings
Table A.32 MIPS32 SPECIAL Opcode Encoding of Function Field bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
SLL1
JR2
MFHI
MULT
ADD
∗
TGE
β
1
001
MOVCI δ
JALR2
MTHI
MULTU
ADDU
∗
TGEU
∗
2
010
SRL δ
MOVZ
MFLO
DIV
SUB
SLT
TLT
β
3
011
SRA
MOVN
MTLO
DIVU
SUBU
SLTU
TLTU
β
4
100
SLLV
SYSCALL
β
β
AND
β
TEQ
β
5
101
∗
BREAK
∗
β
OR
β
∗
∗
6
110
SRLV δ
∗
β
β
XOR
β
TNE
β
7
111
SRAV
SYNC
β
β
NOR
β
∗
β
1. Specific encodings of the rt, rd, and sa fields are used to distinguish among the SLL, NOP, SSNOP, EHB and PAUSE functions.
2. Specific encodings of the hint field are used to distinguish JR from JR.HB and JALR from JALR.HB
rt
bits 18..16
Table A.33 MIPS32 REGIMM Encoding of rt Field
bits 20..19
0
1
2
3
00
01
10
11
0
000
BLTZ
TGEI
BLTZAL
∗
1
001
BGEZ
TGEIU
BGEZAL
∗
2
010
BLTZL φ
TLTI
BLTZALL φ
∗
3
011
BGEZL φ
TLTIU
BGEZALL φ
∗
4
100
∗
TEQI
∗
∗
5
101
∗
∗
∗
*
6
110
∗
TNEI
∗
∗
7
111
∗
∗
∗
SYNCI ⊕
Table A.34 MIPS32 SPECIAL2 Encoding of Function Field bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
MADD
θ
θ
θ
CLZ
θ
θ
θ
1
001
MADDU
θ
θ
θ
CLO
θ
θ
θ
2
010
MUL
θ
θ
θ
θ
θ
θ
θ
3
011
θ
θ
θ
θ
θ
θ
θ
θ
4
100
MSUB
θ
θ
θ
β
θ
θ
θ
5
101
MSUBU
θ
θ
θ
β
θ
θ
θ
6
110
θ
θ
θ
θ
θ
θ
θ
θ
7
111
θ
θ
θ
θ
θ
θ
θ
SDBBP σ
102
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Table A.35 MIPS32 SPECIAL31 Encoding of Function Field for Release 2 of the Architecture bits 2..0
bits 5..3
1. Release 2 of the Architecture added the SPECIAL3 opcode. Implementations of Release 1 of the Architecture sig- naled a Reserved Instruction Exception for this opcode and all function field values shown above.
Table A.36 MIPS32 MOVCI Encoding of tf Bit bit 16
Table A.37 MIPS321 SRL Encoding of Shift/Rotate bit 21
1. Release 2 of the Architecture added the ROTR instruction. Implementations of Release 1 of the Architecture ignored bit 21 and treated the instruction as an SRL
Table A.38 MIPS321 SRLV Encoding of Shift/Rotate bit 6
1. Release 2 of the Architecture added the ROTRV instruction. Implementations of Release 1 of the Architecture ignored bit 6 and treated the instruction as an SRLV
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 103
A.13 Instruction Bit Encoding Tables
function
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
EXT ⊕
∗
∗
∗
BSHFL ⊕δ
∗
∗
∗
1
001
β
∗
∗
∗
∗
∗
∗
∗
2
010
β
∗
*
∗
∗
∗
∗
∗
3
011
β
*
∗
∗
∗
∗
∗
RDHWR ⊕
4
100
INS ⊕
∗
∗
∗
β
∗
∗
∗
5
101
β
∗
*
∗
∗
∗
∗
∗
6
110
β
∗
∗
∗
∗
∗
*
∗
7
111
β
∗
∗
∗
∗
∗
∗
∗
tf
0
1
MOVF
MOVT
R
0
1
SRL
ROTR
R
0
1
SRLV
ROTRV
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Instruction Bit Encodings
sa
bits 8..6
Table A.39 MIPS32 BSHFL Encoding of sa Field1
bits 10..9
0
1
2
3
00
01
10
11
0
000
SEB
SEH
1
001
2
010
WSBH
3
011
4
100
5
101
6
110
7
111
1. The sa field is sparsely decoded to identify the final instructions. Entries in this table with no mnemonic are reserved for future use by MIPS Technologies and may or may not cause a Reserved Instruction exception.
rs
bits 23..21
Table A.40 MIPS32 COP0 Encoding of rs Field
bits 25..24
0
1
2
3
00
01
10
11
0
000
MFC0
∗
1
001
β
∗
2
010
∗
RDPGPR ⊕
3
011
∗
MFMC01 δ⊕
C0 δ
4
100
MTC0
∗
5
101
β
∗
6
110
∗
WRPGPR ⊕
7
111
∗
∗
1. Release 2 of the Architecture added the MFMC0 function, which is further decoded as the DI (bit 5 = 0) and EI (bit 5 = 1) instructions.
Table A.41 MIPS32 COP0 Encoding of Function Field When rs=CO bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
∗
TLBP
∗
ERET
WAIT
∗
∗
∗
1
001
TLBR
∗
∗
∗
∗
∗
∗
∗
2
010
TLBWI
*
∗
∗
∗
∗
∗
∗
3
011
∗
∗
∗
∗
∗
∗
∗
∗
4
100
∗
∗
∗
∗
∗
∗
∗
∗
5
101
∗
∗
∗
∗
∗
∗
∗
∗
6
110
TLBWR
∗
∗
∗
∗
∗
∗
∗
7
111
∗
∗
∗
DERET σ
∗
∗
∗
∗
104
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
bits 23..21 bits 25..24
rs
Table A.42 MIPS32 COP1 Encoding of rs Field
A.13 Instruction Bit Encoding Tables
0
1
2
3
00
01
10
11
0
000
MFC1
BC1 δ
Sδ
∗
1
001
β
BC1ANY2
δε∇
Dδ
∗
2
010
CFC1
BC1ANY4
δε∇
∗
*
3
011
MFHC1 ⊕
∗
∗
∗
4
100
MTC1
∗
Wδ
*
5
101
β
∗
Lδ
∗
6
110
CTC1
∗
PS δ
*
7
111
MTHC1 ⊕
∗
∗
∗
Table A.43 MIPS32 COP1 Encoding of Function Field When rs=S bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
ADD
ROUND.L ∇
∗
∗
∗
∗
C.F CABS.F ε∇
C.SF CABS.SF ε∇
1
001
SUB
TRUNC.L ∇
MOVCF δ
∗
CVT.D
∗
C.UN CABS.UN ε∇
C.NGLE CABS.NGLE ε∇
2
010
MUL
CEIL.L ∇
MOVZ
∗
∗
∗
C.EQ CABS.EQ ε∇
C.SEQ CABS.SEQ ε∇
3
011
DIV
FLOOR.L ∇
MOVN
∗
∗
∗
C.UEQ CABS.UEQ ε∇
C.NGL CABS.NGL ε∇
4
100
SQRT
ROUND.W
∗
RECIP2 ε∇
CVT.W
∗
C.OLT CABS.OLT ε∇
C.LT CABS.LT ε∇
5
101
ABS
TRUNC.W
RECIP ∇
RECIP1 ε∇
CVT.L ∇
∗
C.ULT CABS.ULT ε∇
C.NGE CABS.NGE ε∇
6
110
MOV
CEIL.W
RSQRT ∇
RSQRT1 ε∇
CVT.PS ∇
∗
C.OLE CABS.OLE ε∇
C.LE CABS.LE ε∇
7
111
NEG
FLOOR.W
∗
RSQRT2 ε∇
∗
∗
C.ULE CABS.ULE ε∇
C.NGT CABS.NGT ε∇
Table A.44 MIPS32 COP1 Encoding of Function Field When rs=D bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
ADD
ROUND.L ∇
∗
∗
CVT.S
∗
C.F CABS.F ε∇
C.SF CABS.SF ε∇
1
001
SUB
TRUNC.L ∇
MOVCF δ
*
∗
∗
C.UN CABS.UN ε∇
C.NGLE CABS.NGLE ε∇
2
010
MUL
CEIL.L ∇
MOVZ
∗
∗
∗
C.EQ CABS.EQ ε∇
C.SEQ CABS.SEQ ε∇
3
011
DIV
FLOOR.L ∇
MOVN
∗
∗
∗
C.UEQ CABS.UEQ ε∇
C.NGL CABS.NGL ε∇
4
100
SQRT
ROUND.W
∗
RECIP2 ε∇
CVT.W
∗
C.OLT CABS.OLT ε∇
C.LT CABS.LT ε∇
5
101
ABS
TRUNC.W
RECIP ∇
RECIP1 ε∇
CVT.L ∇
*
C.ULT CABS.ULT ε∇
C.NGE CABS.NGE ε∇
6
110
MOV
CEIL.W
RSQRT ∇
RSQRT1 ε∇
∗
∗
C.OLE CABS.OLE ε∇
C.LE CABS.LE ε∇
7
111
NEG
FLOOR.W
∗
RSQRT2 ε∇
∗
∗
C.ULE CABS.ULE ε∇
C.NGT CABS.NGT ε∇
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 105
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Instruction Bit Encodings
Table A.45 MIPS32 COP1 Encoding of Function Field When rs=W or L1 bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
∗
∗
∗
∗
CVT.S
∗
∗
∗
1
001
∗
∗
∗
∗
CVT.D
∗
∗
∗
2
010
∗
∗
∗
∗
∗
∗
∗
∗
3
011
∗
∗
∗
∗
∗
∗
∗
∗
4
100
∗
∗
∗
∗
∗
∗
∗
∗
5
101
∗
∗
∗
∗
∗
∗
∗
∗
6
110
∗
∗
∗
∗
CVT.PS.PW ε∇
∗
∗
∗
7
111
∗
∗
∗
∗
∗
∗
∗
∗
1. Format type L is legal only if 64-bit floating point operations are enabled.
Table A.46 MIPS64 COP1 Encoding of Function Field When rs=PS1 bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
ADD ∇
∗
∗
ADDR ε∇
CVT.S.PU ∇
CVT.S.PL ∇
C.F ∇ CABS.F ε∇
C.SF ∇ CABS.SF ε∇
1
001
SUB ∇
∗
MOVCF δ∇
∗
∗
∗
C.UN ∇ CABS.UN ε∇
C.NGLE ∇ CABS.NGLEε∇
2
010
MUL ∇
∗
MOVZ ∇
MULR ε∇
∗
∗
C.EQ ∇ CABS.EQ ε∇
C.SEQ ∇ CABS.SEQ ε∇
3
011
∗
∗
MOVN ∇
∗
∗
∗
C.UEQ ∇ CABS.UEQ ε∇
C.NGL ∇ CABS.NGL ε∇
4
100
∗
∗
∗
RECIP2 ε∇
CVT.PW.PS ε∇
PLL.PS ∇
C.OLT ∇ CABS.OLT ε∇
C.LT ∇ CABS.LT ε∇
5
101
ABS ∇
∗
∗
RECIP1 ε∇
∗
PLU.PS ∇
C.ULT ∇ CABS.ULT ε∇
C.NGE ∇ CABS.NGE ε∇
6
110
MOV ∇
∗
∗
RSQRT1 ε∇
∗
PUL.PS ∇
C.OLE ∇ CABS.OLE ε∇
C.LE ∇ CABS.LE ε∇
7
111
NEG ∇
∗
∗
RSQRT2 ε∇
∗
PUU.PS ∇
C.ULE ∇ CABS.ULE ε∇
C.NGT ∇ CABS.NGT ε∇
1. Format type PS is legal only if 64-bit floating point operations are enabled.
Table A.47 MIPS32 COP1 Encoding of tf Bit When rs=S, D, or PS, Function=MOVCF bit 16
tf
0
1
MOVF.fmt
MOVT.fmt
106
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
rs
bits 23..21
Table A.48 MIPS32 COP2 Encoding of rs Field
A.14 Floating Point Unit Instruction Format Encodings
bits 25..24
0
1
2
3
00
01
10
11
0
000
MFC2 θ
BC2 θ
1
001
β
∗
2
010
CFC2 θ
∗
3
011
MFHC2 θ⊕
∗
C2 θδ
4
100
MTC2 θ
∗
5
101
β
∗
6
110
CTC2 θ
∗
7
111
MTHC2 θ⊕
∗
Table A.49 MIPS64 COP1X Encoding of Function Field1 bits 2..0
function
bits 5..3
0
1
2
3
4
5
6
7
000
001
010
011
100
101
110
111
0
000
LWXC1 ∇
SWXC1 ∇
∗
∗
MADD.S ∇
MSUB.S ∇
NMADD.S ∇
NMSUB.S ∇
1
001
LDXC1 ∇
SDXC1 ∇
∗
∗
MADD.D ∇
MSUB.D ∇
NMADD.D ∇
NMSUB.D ∇
2
010
∗
∗
∗
∗
∗
∗
∗
∗
3
011
∗
∗
∗
∗
∗
∗
∗
∗
4
100
∗
∗
∗
∗
∗
∗
∗
∗
5
101
LUXC1 ∇
SUXC1 ∇
∗
∗
∗
∗
∗
∗
6
110
∗
∗
∗
ALNV.PS ∇
MADD.PS ∇
MSUB.PS ∇
NMADD.PS ∇
NMSUB.PS ∇
7
111
∗
PREFX ∇
∗
∗
∗
∗
∗
∗
1. COP1X instructions are legal only if 64-bit floating point operations are enabled.
A.14 Floating Point Unit Instruction Format Encodings
Instruction format encodings for the floating point unit are presented in this section. This information is a tabular pre- sentation of the encodings described in tables Table A.42 and Table A.49 above.
Table A.50 Floating Point Unit Instruction Format Encodings
fmt field (bits 25..21 of COP1 opcode)
fmt3 field (bits 2..0 of COP1X opcode)
Mnemonic
Name
Bit Width
Data Type
Decimal
Hex
Decimal
Hex
0..15
00..0F
—
—
Used to encode Coprocessor 1 interface instructions (MFC1, CTC1, etc.). Not used for format encoding.
16
10
0
0
S
Single
32
Floating Point
17
11
1
1
D
Double
64
Floating Point
18..19
12..13
2..3
2..3
Reserved for future use by the architecture.
20
14
4
4
W
Word
32
Fixed Point
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 107
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Instruction Bit Encodings
Table A.50 Floating Point Unit Instruction Format Encodings
fmt field (bits 25..21 of COP1 opcode)
fmt3 field (bits 2..0 of COP1X opcode)
Mnemonic
Name
Bit Width
Data Type
Decimal
Hex
Decimal
Hex
21
15
5
5
L
Long
64
Fixed Point
22
16
6
6
PS
Paired Sin- gle
2 × 32
Floating Point
23
17
7
7
Reserved for future use by the architecture.
24..31
18..1F
—
—
Reserved for future use by the architecture. Not available for fmt3 encoding.
108 MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Appendix B
Revision History
In the left hand page margins of this document you may find vertical change bars to note the location of significant changes to this document since its last release. Significant changes are defined as those which you should take note of as you use the MIPS IP. Changes to correct grammar, spelling errors or similar may or may not be noted with change bars. Change bars will be removed for changes which are more than one revision old.
Please note: Limitations on the authoring tools make it difficult to place change bars on changes to figures. Change bars on figure titles are used to denote a potential change in the figure itself.
Revision
0.95
1.00
1.90
2.00
2.50
2.60
2.61
Date
March 12, 2001
August 29, 2002
September 1, 2002
June 8, 2003
July 1, 2005
June 25, 2008
December 5, 2009
Description
External review copy of reorganized and updated architecture documentation.
Update based on all feedback received:
• Fix bit numbering in FEXR diagram
• Clarify the description of the width of FPRs in 32-bit implementations
• Correct tag on FIR diagram.
• Update the compatibility and subsetting rules to capture the current require-
ments.
• Remove the requirement that a licensee must consult with MIPS Technolo-
gies when assigning SPECIAL2 function fields.
Update the specification with the changes due to Release 2 of the Architecture. Changes included in this revision are:
• The Coprocessor 1 FIR register was updated with new fields and interpreta-
tions.
• Update architecture and ASE summaries with the new instructions and
information introduced by Release 2 of the Architecture.
Continue the update of the specification for Release 2 of the Architecture. Changes included in this revision are:
• Correct the revision history year for Revision 1.00 (above). It should be
2002, not 2001.
• Remove NOR, OR, and XOR from the 2-operand ALU instruction table.
Changes in this revision:
• Correct the wording of the hidden modes section (see Section 2.2,
"Compliance and Subsetting").
• Update all files to FrameMaker 7.1.
• Allow shadow sets to be implemented without vectored interrupts or sup-
port for an external interrupt controller. In such an implementation, they are software-managed.
• COP3 no longer extendable by customer.
• Section on Instruction fetches added - 1. fetches & endian-ness 2. fetches &
CCA 3. self-modified code
• Fixed paragraph numbering between chapters.
• FPU chapter didn’t make it clear that MADD/MSUB were non-fused.
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02 109
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.
Revision History
Revision
Date
March 25, 2010 December 10, 2010
March 06, 2010
Description
Changes for microMIPS.
List changes in Release 2.5+ and non-microMIPS changes in Release 3. List PRA implementation options.
Change Security Classification for microMIPS AFP versions.
There is no persietent interpretation of FPR values between instructions. The interpretation comes from the instruction being executed. Clarification that the PS format availability is solely defined by the FIR.PS bit.
3.00 3.01
3.02
• • •
• •
•
110
MIPS® Architecture For Programmers Volume I-A: Introduction to the MIPS32® Architecture, Revision 3.02
Copyright © 2001-2003,2005,2008-2011 MIPS Technologies Inc. All rights reserved.