Manuals / Freescale Semiconductor / Computer Equipment / Computer Hardware

Freescale Semiconductor SC140 specifications Instruction Reordering Within an Execution Set

Models: SC140

1 760

Download 760 pages 48.94 Kb

Page 192

Image 192

Instruction Grouping

5.2.5 Instruction Reordering Within an Execution Set

The SC140 can execute up to four DALU instructions and up to two AGU instructions concurrently. These instructions are grouped together in an execution set and dispatched in parallel to the execution units by the PDU. Since the execution units of each type are identical (in principle), any ALU can receive any DALU instruction. As well, any AAU can receive any AGU instruction. The hardware takes advantage of this fact to reduce internal routing from the PDU to the execution units. As a result of this reduction, some reordering may be necessary concerning instruction positions within an execution set.

In general, execution set reordering is transparent to application developers. The assembler appropriately reorders the instruction encoding in an execution set. However, the assembler’s behavior may become apparent upon disassembly of the binary code when the order of instructions in the set may be different from the source code. In some rare cases, the assembler may add a NOP instruction in order to accomplish the reordering.

An execution set can include up to eight instruction words, occupying positions 0 through 7 within the set. The position of a multi-word instruction is defined as the position of its first word. Example 5-3 shows the positions occupied by 3 one-word (1w) instructions and 2 two-word (2w) instructions grouped with a one-word prefix:

5-12	SC140 DSP Core Reference Manual

Page 192

Image 192

Freescale Semiconductor SC140 specifications Instruction Reordering Within an Execution Set

Contents

SC140 DSP Core Reference ManualSC140 DSP Core Reference Manual Table of Contents SC140 DSP Core Reference Manual Control Registers SC140 DSP Core Reference Manual Program Control Instruction Set Accelerator Plug-In Programming Rules SC140 DSP Core Instruction Set Appendix aStarCore Registry Appendix BXii List of Figures Xiv List of Tables Xvi SC140 DSP Core Reference Manual Xvii Xviii List of Examples SC140 DSP Core Reference Manual SC140 DSP Core Reference Manual Xxi Xxii About This Book Abbreviations used in this manual are listed below AbbreviationsAbbreviation Description ISR Revision History Revision Date DescriptionChapter Introduction Target MarketsArchitectural Differentiation Core Architecture Features Typical System-On-Chip Configuration Variable Length Execution Set Vles Software Model SoC DSP expansion areaSystem expansion area SC140 platformCore Architecture Features Chapter Core Architecture Architecture OverviewBlock Diagram of the SC140 Core Data Arithmetic Logic Unit DaluAddress Generation Unit AGU Data Register FileMultiply-Accumulate MAC Unit Bit-Field Unit BFUStack Pointer Registers Bit Mask Unit BMUProgram Sequencer Unit Pseq Enhanced On-Chip Emulator EOnCEInstruction Set Accelerator Plug-in Isap Interface Memory InterfaceDalu Dalu ArchitectureDalu Programming Model Limit EXTData Registers D0-D15 Write to Data Registers Read from Data RegistersOperand Type Dn.e Dn.h Dn.l Data Registers Access Width Dalu Arithmetic Instructions MACOperand Type Data Width Bits Instruction DescriptionDIV NEG Data Shifter/Limiter Dalu Logical Instructions BFUScaling LimitingCalculating the Ln Bit Scaling ExampleScaling Mode Bits Defining the Ln bit Calculation Limiting with the Moves InstructionsLn Bit Calculation Scaling and Arithmetic Saturation Mode Interactions Selected Special Six Other Dalu Instructions ModeLimiting Example 10. Scaling and Limiting InteractionsDalu Arithmetic and Rounding Data Representation11. Saturation and Rounding Interactions Data Formats Signed FractionalSigned Integer 12. Two’s Complement Word RepresentationsSigned Fractional Signed Integer Unsigned Integer Multiplication DivisionUnsigned Arithmetic Unsigned Multiplication13. Rounding Position in Relation to Scaling Mode Scaling Mode High Portion Low PortionRounding Modes Unsigned ComparisonConvergent Rounding No Scaling 2.6.2 Two’s Complement Rounding Two’s Complement Rounding No Scaling Arithmetic Saturation Mode 14. Arithmetic Saturation ExampleMulti-Precision Arithmetic Support Fractional Multi-Precision ArithmeticFractional Double-Precision Multiplication Integer Multi-Precision Arithmetic Fractional Mixed-Precision Multiplication10. Signed Integer Double-Precision Multiplication Viterbi Decoding Support 11. Unsigned Integer Double-Precision MultiplicationAddress Generation Unit AGU ArchitectureAddress ArithmeticUnit AAU Address Generation Unit AGU Programming Model 13. AGU Programming ModelAddress Registers R0-R15 Stack Pointer Registers NSP, ESPOffset Registers N0-N3 Base Address Registers B0-B7Modifier Registers M0-M3 Shadow Stack Pointer RegistersModifier Control Register Mctl 17. Address Modifier AM BitsAddress Modifier Modes Addressing Modes Register Direct ModesAddress Register Indirect Modes Address Generation Unit PC Relative Mode Special Addressing Modes Memory Access Width Memory Access MisalignmentAccess Type Aligned Address Addressing Modes Summary19. Memory Address Alignment 20. Addressing Modes SummaryAddress Register Indirect PC RelativeSpecial Linear Addressing Mode Reverse-carry Addressing ModeModulo Addressing Mode Address Modifier Modes15. Modulo Addressing Example Multiple Wrap-Around Modulo Addressing Mode 21. Modulo Register Values for Modulo Addressing ModeModifier Mj Address Calculation Arithmetic Arithmetic Instructions on Address Registers 23. AGU Arithmetic InstructionsBit Mask Instructions Bit Mask Test and Set Semaphore Support Instruction 24. AGU Bit Mask Instructions BMUMove Instructions Semaphore Hardware ImplementationExample of Normal Usage of the Semaphoring Mechanism Label BMTSET.W #mask,R0 JT label25. AGU Move Instructions MOVE.W16. Integer Move Instructions 17. Fractional Move Instructions Memory Interface 18. Bit Allocation in MOVE.L D0.eD1.e1 SC140 Endian Support 1.1 SC140 Bus StructureMemory Organization 20. Basic Connection between SC140 Core and MemoryData Moves 26. Data Representation in MemoryRepresentation Type Value 22. Data Transfer in Big and Little Endian Modes Multi-Register Moves Address DataMulti-Register Transfer in Big and Little Endian Modes Instruction Word Transfers 25. Instruction Moves in Big and Little Endian Modes Memory Access Behavior in Big/Little Endian Modes Example MOVE.L D0.ED1.E, A0 Example MOVE.2L D0D1, R0Example MOVE.F D0, R0 Example MOVE.2F D0D1, R0Example VSL.4W D2D6D1D3, R0 + N0 D6 =Example VSL.4F D2D6D1D3, R0 + N0 Example VSL.2W D1D3, R0 + N0Example Push D0 Example Push D0 Push D1Example BMSET.W #$1234, A0 Data =31. Control Instructions in Big and Little Endian Modes Instruction Register Operands Big Little EndianCore Control Registers Status Register SRName Description Settings Describes the various SR bitsStatus Register Description I2-I0 Interrupt Mask Bits Reflect ExceptionsOverflow Exception Enable Bit Disable Interrupts Bit When this bitException Mode Bit Selects ReservedEquation Mode S1-S0 Scaling Mode Bits SpecifyScaling Mode Bit Rounding Mode Bit Selects the typeName Description Settings Arithmetic Saturation Mode Selects Exception and Mode Register EMR Exception and Mode Register EMRDescribes the EMR fields EMR DescriptionIlst Illegal Execution Set Indicates whether anPLL and Clock Registers Clearing EMR BitsBmclr #$fffb,EMR.L Example 3-1. Clearing an EMR BitEmulation and Debug EOnCE Debugging SystemJtag Interface Signal Descriptions Signal Name Signal DescriptionOverview of the Combined Jtag and EOnCE Interface Cascading Multiple SC140 EOnCE Modules in a SoCJtag Scan Paths Jtag InstructionsLoadgpr TAP Controller State Machine Jtag Scan Paths Select-DR Scan Path Select-IR Scan PathActivating the EOnCE Through the Jtag Port Enabling the EOnCE ModuleDebugrequest and Enableeonce Commands Reading/Writing EOnCE Registers Through JtagReading and Writing EOnCE Registers Via Jtag EOnCE register write operation through Jtag EOnCE register read capture operation through JtagEOnCE Signals EOnCE Signals Jtag SignalsMain Capabilities of the EOnCE Module Core InterfaceEOnCE Dedicated Instructions Debug StateDebug Exception Executing an Instruction while in Debug StateSoftware Downloading Software Downloading EOnCE Events EOnCE Event TypesEvent type Occurs when EOnCE Actions Event and Action SummaryEOnCE Event and Action Summary Event typeEOnCE Enabling and Power Considerations EOnCE Module Internal ArchitectureEOnCE Controller Command Register Transmit Register Update Signal from the TAP ControllerAddress Monitor and Control Register Address Control Decoder Logic Receive RegisterEvent Counter Shows a block diagram of the event counter Event Counter Register SetEvent Detection Unit EDU EE5..0 Address BusesXDBxx Data Buses EventD Event0 Event1 Event2 Event5 Count eventAddress Event Detection Channel Edca EEiEdca Register Set Data Event Detection Channel Edcd Event External Event 6,7 Count EventEventD Edcd register set is shown belowEvent Selector ES Optional External Event Detection Address ChannelsEE40 ES block diagram is shown in Figure Trace UnitEvent0..Event5 External Event6, Event7 EventD Count event 10. Event Selector Register SetEOnCE Module Internal Architecture Change of Flow and Interrupt Tracing Change of FlowTrace Buffer TB Off-Core Writing to the Trace Buffer Reading the Trace Buffer TbbuffTrace Unit Programming Model EOnCE Register Addressing 11. Trace Buffer Register Set12 displays the EOnCE register addressing offsets 12. EOnCE Register Addressing OffsetsOffset EDCA4REFA Reading or Writing EOnCE Registers Using Core Software Real-Time Jtag AccessGeneral EOnCE Register Issues Real-Time Data TransferEOnCE Register Addressing EOnCE Command Register ECR Read/Write Command SpecifiesEOnCE Controller Registers 13 describes the ECR fields16 displays the bit configuration of the ESR EOnCE Status Register ESR14 describes the ESR fields 14. ESR DescriptionName Description Section DREE3 EOnCE Monitor and Control Register Emcr 15 describes the Emcr fields15. Emcr Description Name Description ReservedDebugerst EOnCE Receive Register Ercv EOnCE Transmit Register EtrsmtEE Signals EE Signals as OutputsDetection by the Event Detection Channels Detecting Entry into Debug StateEE Signals as Inputs EE Signals Control Register Eectrl16. Eectrl Description EeddefEE2DEF Core Command Register Corecmd Length control bits are described in -17, below17. Length Control Bits Length Control Bits DescriptionPC of the Exception Execution Set Pcexcp PC of the Next Execution Set PcnextPC of Last Execution Set Pclast PC Breakpoint Detection Register PcdetectEvent Counter Registers Event Counter Control Register EcntctrlExtended Mode of Operation Bit 18 describes the Ecntctrl fields18. Ecntctrl Description Reserved for TestEvent Counter Enable Used to Event Counter Value Register EcntvalEvents to be Counted Determines EC Signals Extension Counter Value Register EcntextEvent Detection Unit EDU Channels and Registers Address Event Detection Channel EdcaEdca Control Registers EDCAiCTRL 19 describes the EDCAiCTRL fieldsEvent Detection Channel EDCAi Comparators Selection Used toAccess Type Selection These bits Comparator B Condition SelectionComparator a Condition Selection Edca Reference Value Registers a and B EDCAiREFA, EDCAiREFB Edca Mask Register EDCAiMASKData Event Detection Channel Edcd Edcd Control Register Edcdctrl20 describes the Edcdctrl fields 20. Edcdctrl DescriptionAccess Width Selection AWSAccess Type Selection The ATS bit Comparator Condition SelectionEvent Selector ES Registers Event Selector Control Register EselctrlEdcd Reference Value Register Edcdref Edcd Mask Register EdcdmaskEselctrl fields are described in Table 21. Eselctrl Description24 displays the bit configuration of Eseldm Event Selector Mask Debug State Register EseldmEvent Selector Mask Enable Trace Register Eseletb Event Selector Mask Debug Exception Register EseldiEvent Selector Mask Disable Trace Register Eseldtb Trace Unit RegistersTrace Buffer Control Register Tbctrl This mode is usefull only with the Tcount mode 22. Allowed tracing mode combinationsTrace mode Upon a trace event, trace the counter value EcntvalTrace Buffer Counter Mode Tbctrl fields are described in the following table23. Tbctrl Description Trace Buffer Extension CounterTrace Loops Mode Enables tracing Trace Buffer Enable Mode EnablesTrace Mark Instruction Mode Trace Issue of Execution Sets EnableTrace Buffer Read Pointer Register Tbrd Trace Buffer Write Pointer Register TbwrTrace Buffer Register Tbbuff Trace Unit Registers Chapter Program Control PipelineInstruction Pipeline Stages Illustrates the five instruction pipeline stagesPipeline Example Pipeline Stages OverviewInstruction Cycle Operation Pipeline Stage DescriptionInstruction Pre-Fetch and Fetch Instruction DispatchAddress Generation Instruction Grouping ExecutionExample 5-1. Four SC140 Instructions in an Execution Set Grouping Types Instruction Grouping MethodsSerial Grouping Prefix GroupingPrefix Types Two-Word PrefixConditional Execution One-Word Low Register PrefixFor example Prefix InstructionsAssembly Syntax Meaning Prefix Selection AlgorithmLow Register Prefix Selection Algorithm Instruction Reordering Within an Execution Set Example 5-4. Conditional Vles Having Two Subgroups Example 5-5. Set of 2 Two-word Instructions Requiring a NOPInstruction Timing Sequential Instruction Timing Instruction Categories Timing SummaryDalu Instruction Timing Move Instruction TimingBit Mask Instruction Timing Compare Shift TestExample 5-6. Delayed Change-of-Flow and Its Delay Slot Change-Of-Flow Instruction TimingNon-Loop Change-of-Flow Instructions Direct, PC-Relative, and Conditional COF Loop Change-Of-Flow InstructionsDelayed COF COF Execution CyclesExample 5-7 shows a case when a stall cycle is added Highest cycle count of instructions grouped with CallNumber of Cycles Needed by Change-of-Flow Instructions Example 5-7. Subroutine Call TimingMemory Access Timing Memory Access Examples MOVE.L Example 5-8. Parallel Execution of Two Move InstructionsImplicit Push/Pop Memory Timing Memory Stall ConditionsHardware Loops Loop Programming ModelLoop Start Address Registers SAn Loop Notation and Encoding Loop Counter Registers LCnStatus Register SR Loop Flag Bits Loop Initiation and Execution Lpmarka and Lpmarkb Bits in Short and Long LoopsLoop Type Location FunctionalityLoop Nesting Loop Iteration and TerminationLoop Control Instructions 10 lists the loop instructions10. Loop Control Instructions Instruction OperationExample 5-13. Long Loop Disassembly Example 5-12. Long LoopExample 5-14. Short Loop, Two Execution Sets Following is an example of a nested loop Example 5-15. Short Loop, One Execution SetExample 5-16. Nested Loop Stack Support Loop Timing1 SC140 Single Stack Memory Use 2 SC140 Dual Stack Memory Use Shows the stack structureStack Support Instructions 11. Stack Push/Pop Instructions12. Even and Odd Registers Even Register De File Odd Register Do FileAddressing Mode Description Shadow Stack Pointer Registers13. Stack Memory Map 14. Stack Move InstructionsFast Return from Subroutines Normal Working Mode Exception Working ModeWorking Mode EXP bit Active SP Working ModesTypical Working Mode Usage Scenarios Dual-stack RtosWorking Mode Transitions From Exception to Normal modeFrom Normal to Exception mode Single-stack RtosWorking Modes Processing States Processing State Change Instructions16. Processing State Change Instructions 10. Core State Diagram Processing State TransitionsReset Processing State Execution State17. Processing State Transitions Processing State Transitions DescriptionWait Processing State Stop Processing State 18. Exit Wait Processing State due to an Interrupt or NMIException Processing SC140 11. Core-PIC InterfaceInterrupt Vector Address Vector Base Address RegisterProgramming Exception Routine Addresses Return From Exception Instructions 19. Exception Vector Address TableException Address Priority Type Description Offset Highest Maskable Interrupts Non-Maskable Interrupts NMIInternal Exceptions Interrupt Priority LevelIllegal Exception Illegal InstructionIllegal Execution Set Exception Interface to the Pipeline Dalu OverflowTrap Exception Debug ExceptionException Mode Execution Exception Timing20. Exception Pipeline Example 5-17. Basic Exception TimingException Processing 12 provides a flow chart for Example 21. Pipeline Example Instruction Set Accelerator Plug-In IntroductionIsap SC140 Schematic Connection Single IsapData Memory SC140CoreCore to Multiple Isap Connection Schematic Multiple IsapIsap Memory Access Isap instructions and instruction encodingIsap Encoding Fields Binary Encoding Words BitsExample 6-1. Isap memory access ISAP-core register transfersTo understand this, look at the following lines of code Immediate Data Transfer to Isap registers Following line of codeExample 6-2. ISAP-Core register transfers Example 6-3. ISAP-Core register transfersCore Assembly Syntax with an Isap Identification of Isap instructionsWorking with One Isap Vles that uses an implicit Isap ID stringWorking with Multiple ISAPs One Isap in a Single-Line VlesOne Isap in a Multi-Line Vles An Example of the Definition Flexibility of an Isap Multiple ISAPs in a Multi-Line VlesExample 6-5. Multiple Isap coding Example 6-6. Conditional Execution Example Example 6-7. Conditional Execution ExampleProgramming Rules Isap Functions that Interact With the CoreGrouping rules for explicit Isap instructions Rules for implicit AGU instructionsSequencing rules for T bit update D.2, D.3Programming Rules Vles Sequencing Semantics Vles Grouping SemanticsVles Grouping Semantics SC140 Pipeline Exposure Programming Rule NotationGrouping Rules Sequencing Rules Register Read/WriteStatus Bit Updates Instruction WordsMOVE-like Instructions Register AliasingDelayed COF Instructions Delay SlotAGU Arithmetic Instructions Change-Of-Flow DestinationsEnabled Loop Static Programming RulesHardware Loops Hardware Loop DetectionGeneral Grouping Rules Rule G.G.1Rule G.G.2 Rule G.G.3Rule G.G.4 Example 7-5 Duplicate PC DestinationsExample 7-6 Duplicate Address Pointer Register Destinations Example 7-7 Duplicate Stack Pointer DestinationsRule G.G.4 Exceptions Example 7-8 Duplicate Register DestinationsExample 7-9 Duplicate SR/EMR Register Destinations Example 7-10 Duplicate Status Bit DestinationsPrefix Grouping Rules Rule G.G.5Following rules only apply to prefix-grouped Vles Example 7-15. Dalu Register Use Exceeds Four TimesRule G.P.1 Example 7-16 Vles Extension Words Exceed TwoExample 7-17 Two-Word Instructions Exceed Two Rule G.P.3 Rule G.P.4Rule G.P.5 Example 7-18. Vles Has Mutually Exclusive InstructionsRule G.P.6 Rule G.P.7Example 7-20. Data Source Use of Nn and Mn Registers Example 7-21. IFc Having Two SubgroupsRule G.P.8 Rule G.P.9Example 7-24. Isap instructions in same IFc group AGU Rules Rule A.1Example 7-25. Mctl Write to R0-R7 Use Rule A.2 Rule A.3Example 7-26. Rn, Nn, Mn Write to AGU Use Rule A.4 Example 7-27. Rn or Nn Write to MOVE-like UseExample 7-28. LCn Write to MOVE-like Use Delayed COF Rules Example 7-29. Nmid Update to EMR ReadExample 7-30. Instructions in a Delay Slot Rule A.7Example 7-31. Instructions in a Rted Delay Slot Example RTE/D with SR UpdatesRule D.2 Rule D.3Rule D.4 Rule D.5Rule D.5a Rule D.6Status Bit Rules Rule D.8Rule D.9 Rule T.1Rule T.2.a Rule T.2.bRule T.2.c Rule SR.2Static Programming Rules Example 7-43. SR Write to SR Status Bit Use Example 7-44. SR Write to SR Status Bit Update Rule SR.3Rule SR.4 Example 7-45. Dovf Update to SR Read or Write Example 7-46. Dovf Update grouped with Move-like SR updatesRule SR.4a Loop Nesting Rules Rule SR.7Rule L.N.1 DI and EI DOENn and DOENSHnRule L.N.2 Rule L.N.3Example 7-49. Nested Loops with Ordered Index Example 7-50. Nested DOENn/DOENSHn InstructionsExample 7-52. Loopend between Doen and Loopend Example 7-53. Changing a loop typeLoop LA Rules Rule L.L.1Rule L.L.2 Example 7-54. Instructions at the End of Long LoopsRule L.L.3 Rule L.L.4LA of a short loop cannot be at LA-1 of a long loop Example 7-56. Instructions in Short LoopsLoop Sequencing Rules Rule L.L.5Rule L.L.6 Rule L.D.1Rule L.D.3 Rule L.D.5Rule L.D.6 Example 7-60. LCn Write at the Start of Short Loop nRule L.D.7 Rule L.D.8Rule L.D.9 Loop COF Rules Rule L.C.1Rule L.C.2 Rule L.C.3Rule L.C.5 Bc or Jc instruction is not allowed at LA-3 of a long loopExample 7-68. Bc/Jc at LA-3 of a Long Loop Rule L.C.7 Example 7-69. Loop COF Destination in the Same LoopRule L.C.9 Rule L.C.10Example 7-70. Loop COF at End of Nested Long Loops Example 7-71. Subroutine Call to End of LoopsGeneral Looping Rules Rule L.C.11Rule L.C.12 Rule L.G.3Dynamic Programming Rules AGU Dynamic RulesRule L.G.5 Rule A.2aMemory Access Rules Rule A.5Rule A.6 Rule D.7RAS Rules Loop RulesRule J.4 Rule L.N.6Example 7-83. A.2 from a Delay Slot to a COF Destination Rule Detection Across COF BoundariesCycle-Based COF Rules Example 7-82. SR.2 Across a COF BoundaryVLES-Based COF Rules Example 7-85. EMR access at the start of an exception Rule Detection Across Exception BoundariesRule SR.2a Rule SR.4bRule A.1a Example 7-86. Mctl Write to R0-R7 UseCOF destination cannot be a delay slot Programming GuidelinesRule J.1 Rule J.2Rules Not Detected Across COF Boundaries Good Programming PracticesSource Code Practices Rule J.5Binary Code Practices Lpmark Rules Lpmark Instruction TypeSoftware Development Practices Static Programming Rules Dynamic Programming RulesGeneral Grouping Rules Prefix Grouping RulesLoop Nesting Rules Loop LA RulesLpmark Rule L.L.2 Lpmark Rule L.L.3Lpmark Rule L.L.5 Example 7-93. Active LCn Write at the End of Long LoopsLoop Sequencing Rules Lpmark Rule L.L.6Lpmark Rule L.D.2 + L.D.3 Lpmark Rule L.D.6Loop COF Rules Lpmark Rule L.C.2COF instructions are not allowed at LPB of a long loop Example 7-97. Active LCn Read at the Start of a LoopLpmark Rule L.C.3 + L.C.5 Example 7-98. COF Instructions at LPB of a Long LoopExample 7-99. Bc/Jc at the Start of a Loop Lpmark Rule L.C.9 Lpmark Rule L.C.10Example 7-100. Loop COF at End of Nested Long Loops Example 7-101. Subroutine Call to End of LoopsLpmark Programming Guidelines General Looping RulesRule Detection Across Exception Boundaries Example 7-104. COF Destination to Loop Delay Slots NOP DefinitionLpmark Rule L.C.1 Grouping Examples Is encoded asIs assembler mapped to the IFT prefix and encoded as Is assembler mapped to the IFF prefix and encoded asIs encoded ignoring the NOP subgroup as NOP Definition Appendix a SC140 DSP Core Instruction Set Conventions Table A-1. Instruction ConventionsConvention Definition Table A-2. Operations Syntax Table A-3. Register AbbreviationsOperator Description Abbreviation Register NameBrackets as Isap indicators Brackets as address indicatorsTable A-4. Assembler Syntax Addressing Mode Notation Table A-5. Addressing Mode Notation for the EA OperandTable A-6. Addressing Mode Notation for the ea Operand Addressing Mode Definition Notation in Instruction FieldData Representation in Memory for the Examples Encoding NotationDefinition for the field is Prefix Word Encoding Instruction Formats and Opcodes Instruction FieldsAaa CccExample, 2-w prefix + 2 grouped instruction words, aaa = If true D0, D2, A0, if false D1, D3, A1If true, all the set Prefix Words Cycles TypeLast, or to last-1 Example First execution set of the loopHigh data register is used for the op3 field E3 is set High data register is used for the op1 field E2 is setDSP Core Instruction Set Instruction Types Instruction Sub-typesTable A-7. Dalu Arithmetic Instructions MAC Table A-8. Dalu Logical Instructions BFU Table A-9. AGU Arithmetic Instructions Table A-10. AGU Move Instructions Table A-11. AGU Stack Support InstructionsTable A-12. AGU Bit-Mask Instructions BMU Table A-13. AGU Non-Loop Change-of-Flow InstructionsTable A-15. AGU Program Control Instructions Table A-16. Prefix InstructionsInst InstructionsInstruction Definition Layout ABS Single Source/Destination Data Register InstructionADC Add Long With Carry DaluDc + Dd + C → Dd ADC Dc,DdDc,Dd Data Register Pairs Register/Memory Address BeforeADD Add DaluOperation Assembler Syntax Add d0,d1,d2Add d1,d0,d2 Da,Db Da,Da Data Register Pairs#u5 ADD2 Add Two 16-Bit Values DaluAdd2 d0,d1 JJJ Single Source Data RegisterAdda Add AGUAdda #u5,Rx Adda #s16,rx,RnAdda r0,r1 Address RegisterRrrr AGU Source Register AGU Source/Destination Register#s16 ADDL1A Add With One-Bit Arithmetic Shift Left ADDL1A Source Operand AGUAddl1a r0,r1 Rx1 + Rx → RxADDL1A rx,Rx ADDL2A Add With Two-Bit Arithmetic Shift Left ADDL2A Addl2a r0,r1Rx2 + Rx → Rx ADDL2A rx,RxADDL2A rx,Rx ADDNC.W Add Without Changing ADDNC.WCarry Bit Dalu Addnc.w #$ca3e,d1,d2Instruction Words Cycles Type ADR Add and Round DaluAdr d3,d4 RndDa + Dn → DnADR Da,Dn Bitwise and Dalu #0u16,Da,Dn#u16$0000,Da,Dn Da,DnD2,d1 #$0ff2e,d2,d1#$ff2e0000,d2,d1 #0u16 #u16$0000#u16 #$a70e,d1.h #u16 DR.L → DR.L#u16 DR.H → DR.H #u16,DR.LData/Address Register Cycles Type OpcodeAND.W #u16,Rn AND.W #u16,SP-u5AND.W #u16,a16 AND.W #u16,SP+s16And.w #$54a1,r7 A16 S16ASL Asl d0,d1Da 1→ Dn ASL Da,DnASL Da,Dn ASL2A Asl2a r0Rx2 → Rx ASL2A RxAsla Asla r0Rx1 → Rx Asla RxAsll Multiple-Bit Arithmetic Shift Left DaluAsll #u5,Dn Asll Da,DnAsll d0,d1 Asll Aslw Word Arithmetic Shift Left 16 Bits Dalu AslwAslw d0,d1 Da16 → DnAslw Da,Dn ASR Asr d5,d3Da1 → Dn ASR Da,DnRegister/Memory Address Before After Asra Asra r2Asra Rx Asrr Multiple-Bit Arithmetic Shift Right DaluAsrr #$3,d5 Asrr d3,d5#u5 Asrw d5,d0 Asrw Da,DnAsrw Da,Dn If T==0, then PC + displacement → PC Branch If False AGUBF lbl BF labelDisplacement label Instruction Words Cycles1 Type OpcodeBFD Branch If False Using a Delay Slot AGU Operation BFDBFD lbl BFD labelLabel Displacement Bmchg ~C1.Hi → C1.Hi i denotes bits=1 in #u16~C1.Li → C1.Li ~DR.Hi → DR.HiBmchg #$f0f0,d1.h Clears the Ln bit in the destination data registerControl Registers Iiiiiiiiiiiiiiii 16-bit unsigned immediate data BMCHG.W Bit-Masked Change aBmchg.w #$661f,$800c BMCHG.W #u16,SP-u5Bit signed SP address offset Bmclr Bit-Masked Clear a 16-Bit Operand BMU Bmclr Operation Bmclr #u16,C1.HBmclr #u16,C1.L Bmclr #u16,DR.HBmclr #$b646,d7.l #u16 BMCLR.W Bit-Masked Clear aBit Operand in Memory BMU Operation Assembler Syntax BMCLR.W #u16,SP-u5 Bmset #u16,C1.H Bmset #u16,C1.LBmset #u16,DR.H Bmset #u16,DR.LBmset #$2436,d1.l BMSET.W Bit Operand in Memory BMUBmset.w #$f111,$800c Register/Memory Address Before Immediate$800C Bmtset Bmtset #$111f,d1.lBmtset #u16,DR.H Bmtset #u16,DR.LBmtset #$4238,d4.l BMTSET.W Bit-Masked Test and Set a BMTSET.WBmtset.w #$4328,$c BMTSET.W #u16,SP-u5 Bmtstc Bmtstc #$8a59,d7.h$0$0024A60000 $00E40000 BMTSTC.W BMTSTC.W #u16,SP-u5BMTSTC.W #u16,SP+s16 BMTSTC.W #u16,RnBmtstc.w #$8A59,r0 BMTSTC.W #u16,SP-u5 Bmtsts Bmtsts #u16,C1.LBmtsts #u16,DR.H Bmtsts #u16,DR.LBmtsts #$24a6,d7.h BMTSTS.W BMTSTS.W #u16,SP-u5BMTSTS.W #u16,SP+s16 BMTSTS.W #u16,RnBmtsts.w #$0428,r0 BMTSTS.W #u16,SP-u5 PC + displacement → PC BRABranch AGU BRA labelAAAAAAAAAA0 Brad Brad labelSource Code Comments $0000 000A $0000 000E PC + displacement → PC BreakBreak label → LFnEncoding is the displacement with bit BSR Branch to Subroutine AGUStatus and Conditions Changed by Instruction None Example Bsr labelBSR PC + displacement → PC, next* PC→RAS Next* PC → SP SR → SP + 4 SP + 8 → SPBsrd label Bsrd label If T==1, then PC + displacement → PC Branch If True AGUBT lbl BT labelRegister/Memory Address Before BT After BTD Branch If True Using a Delay Slot AGU Operation BTDBTD lbl BTD label$0035 $0000 $0006 $002A $001A $0016 CLB Count Leading Bits DaluClb d3,d7 CLB Da,DnCLB Da,Dn CLR Clear a Data Register DaluClr d1 → DnDestination Data Register Source Data RegisterCmpeq Compare for Equal DaluCmpeq d2,d3 If Da == Dn, then 1→ T, else 0 → T118 CMPEQ.W CMPEQ.W Compare for Equal DaluCmpeq.w #$5,d3 CMPEQ.W #u5,DnCMPEQ.W Cmpeqa Compare for Equal AGU Cmpeqa Operation Cmpeqa r1,r2If rx == Rx, then 1 → T, else 0 → T Cmpeqa rx,RxCmpeqa rx,Rx Cmpgt Compare for Greater Than DaluCmpgt d2,d3 Dn Da → T124 CMPGT.W Cmpgt.w #$8002,d2CMPGT.W #u5,Dn CMPGT.W #s16,DnCMPGT.W #u5,Dn Cmpgta Compare for Greater Than AGU CmpgtaCmpgta r2,r3 Rx rx → TCmpgta rx,Rx Cmphi Unsigned Compare for Higher Dalu CmphiCmphi d1,d0 Cmphi Da,Dn130 Cmphia Unsigned Compare for Higher AGU CmphiaCmphia r0,r1 Cmphia rx,RxCmphia rx,Rx Cont Continue to the Next Loop Iteration AGULabel LabelCycles1 Type Opcode Contd Contd labelCycles1 Type Enter Debug Mode AGU DebugDebug Signal a Debug Event AGU Debugev DebugevDeca Decrement a Register AGUDeca r0 Rx 1 → RxBit unsigned immediate data = 1, set by the assembler Deceq d7 Dn 1 → Dn if Dn==0, then 1→ T, else 0 → TDeceq Dn 142 Deceqa Decrement and Set T If Equal Zero Deceqa Deceqa r0Rx 1 → Rx if Rx==0, then 1 → T, else 0 → T Deceqa Rx Deceqa RxDecge Example decgeDn 1 → Dn Dn≥0 → T Decge DnSC140 DSP Core Reference Manual 145 Decgea Decgea r4Rx 1 → Rx Rx ≥ 0 → T Decgea RxDecgea Rx Determines execution working mode SR19 Set disable interrupt bit→ DI SR18Page DIV Divide Iteration DaluIf Dn39 ⊕ Da39 = Then 2 * Dn + C + Da & $FF Ffff 0000 → DnDiv d2,d1 DIV Da,Dn Dmacss Dmacss d2,d3,d5Dn16 + Dc.H * Dd.H → Dn Dc signed, Dd signedDmacss Dc,Dd,Dn 1 1 Dmacsu Multiply Signed By Unsigned and Dmacsu Accumulate With Right Shifted Data Register DaluDmacsu d2,d3,d5 156 Do Enable Long Loop AGU Doen2 d0DOENn #u6 DOENn #u16Loop Identifier #u6Do Enable Short Loop AGU DOENSHn Doensh2 d0DOENSHn #u6 DOENSHn #u16$00E4 $A0E4 Setup Long Loop DOSETUPn Dosetup1 labelPC + displacement → SAn DOSETUPn labelEncoding is the displacement with SR19 Clears disable interrupt bit → DI164 EOR Bitwise Exclusive or DaluEor d4,d5 Da ⊕ Dn → DnEOR Da,Dn Eor #$5,d5.l EOR #u16,DR.LEOR #u16,DR.H EOR #u16,DR.L EOR #u16,DR.H EOR.W Bitwise Exclusive or onEor.w #$aaaa,r0 Extract Extract Extract Signed Bit Field DaluExtract #$c,#$e,d2,d4 Extract #U6,#u6,Db,DnJjj Single Source/Destination Data Register Extractu Extractu Extract Unsigned Bit FieldExtractu #$c,#$e,d2,d4 Extractu #U6,#u6,Db,DnExtractu #U6,#u6,Da,Dn Iaddnc.w #$a002,d2 IADDNC.W #s16,DnConditionally Execute a Group or Subgroup Prefix IFc If T == Then execute group/subgroupElse treat as NOP If T == Then execute group/subgroup Else treat as NOP Execute group/subgroup unconditionallyIft move.w #$ffff,d0 Ccc Conditional execution of the entire execution setIllegal Illegal Imac Imac d4,d5,d6Dn ± Da.L * Db.L → Dn Imac ±Da,Db,DnImac -d4,d5,d6 Accumulation Notation182 Dn + Da.L * Db.H → Dn Imaclhuu Da,Db,DnImaclhuu Da,Db,Dn Imacus Integer Multiply AccumulateUnsigned By Signed Dalu Operation Assembler Syntax Imacus d3,d4,d0$0002 x -64 $FFC0 -128 $FF80 +0 $0000 -128 $FF80 Impy Integer Multiply DaluDa.L * Db.L → Dn Impy Da,Db,DnD1,D1 D3,D3 D5,D5 D7,D7 Impy.w #$fffe,d3 #s16 * Dn.L → DnIMPY.W #s16,Dn +16 Impyhluu Integer Multiply Upper ImpyhluuImpyhluu d4,d3,d0 Da.H * Db.L → DnImpyhluu Da,Db,Dn Impysu Impysu d3,d5,d1Impysu Da,Db,Dn Register Bit Name Description AddressImpysu Da,Db,Dn Impyuu Impyuu d5,d3,d1Impyuu Da,Db,Dn 196 INC INC Increment a Data Register By One Dalu OperationInc d0 INC DnInc d15 Inc.f d15 Dn + $0000010000 → DnINC.F Dn INC.F Dn Inca Increment Register AGUInca r0 Rx + 1 → RxInca Rx Insert Insert Bit Field DaluInsert #12,#22,d6,d7 Insert #U6,#u6,Db,DnInsert #U6,#u6,Db,Dn Jump If False AGU JF lblJF label JF RnBit absolute long address JFD JFD labelJFD Rn $00E0 $00 0000 $0000 $00 0000 002A $00 0000 001A JMP Jump AGUJmp label JMP labelJMP label Jump Using a Delay Slot AGU JmpdExample jmpd lbl Jmpd labelAaaaaaaaaaaaaaaaAAAAAAAAAAAAAAAAA JSR Jump to Subroutine AGUJsr r6 JSR labelAbsolute long address Example jsrd r6 Next* PC → RAS Rn → PCJsrd label Jsrd RnJsrd label Jump If True AGU Jt r0JT label JT RnJT label Jump If True Using Delay Slot AGU JTDExample jtd r0 JTD labelJTD label LPMARKx End-of-Loop Mark Prefix LPMARKx Operation If LCn Then SAn → PCLCn 1 → LCn Else next PC → PC → LFn If LCn Then SAn → PC LCn 1 → LCn Else next PC → PC → LFn → SLFStatus and Conditions that Affect Lpmark Execution Table A-17. Combinations of LPMARKx UseLFn LCn DescriptionStatus and Conditions Changed by Lpmark Execution Prefix Formats and OpcodesInsertion of lpmarkb by assembler Instruction Disassembled Instruction CommentsLsll Multiple-Bit Bitwise Shift Left DaluLsll d4,d2 Lsll Da,Dn$00E4 $0$FF 8765 LSR Bitwise Shift Right One Bit DaluLsr d4 Dn1 → Dn 0 → Dn39Lsra Bitwise Shift Right By One Bit AGULsra r2 Rx1 → Rx 0 → Rx31Lsrr Multiple-Bit Bitwise Shift Right DaluLsrr Da,Dn Lsrr #u5,DnLsrr d4,d2 Before AfterBit unsigned immediate data Lsrw Word Bitwise Shift Right DaluLsrw d4,d2 Lsrw Da,DnLsrw Da,Dn MAC Signed Fractional Multiply-Accumulate DaluMac d4,d5,d6 MAC #s16,Da,DnMac #$1000,d5,d6 SC140 DSP Core Reference Manual 235 Macr Macr d4,d5,d6RndDn ± Da.H * Db.H → Dn Macr ±Da,Db,Dn000 0000 0000 1000 $0008 Instruction Formats and Opcodes 238 Macsu Macsu d0,d1,d4Dn + Dc.H * Dd.L → Dn Macsu Dc,Dd,Dn111 1111 1111 1111 $FFFF Instruction Formats and Opcodes Macus Dn + Dc.L * Dd.H → DnMacus Dc,Dd,Dn Macus Dc,Dd,Dn Macuu Fractional Multiply-AccumulateUnsigned By Unsigned Dalu Operation Assembler Syntax Macuu d2,d3,d1Macuu Dc,Dd,Dn Mark Push the PC into the Trace Buffer AGU MarkPC → trace buffer MAX Transfer Maximum Signed Value DaluMax d0,d4 If Dg Dh, then Dg → DhMAX2 Max2 d0,d4If Dg.H Dh.H, then Dg.H → Dh.H If Dg.L Dh.L, then Dg.L → Dh.L00 D0,D4 01 D1,D5 10 D2,D6 11 D3,D7 248 MAX2VIT If Da.L Db.L, then 0 → VFn, Da.L → Db.LElse 1 → VFn MAX2VIT Da,DbMax2vit d4,d2 Maxm Transfer Maximum Absolute Value Dalu Maxm Operation Maxm d2,d6Maxm Dg,Dh 00 D0,D4 01 D1,D5 10 D2,D6 11 D3,D7 252 MIN Transfer Minimum Signed Value DaluMin d1,d5 MIN Dg,DhMOVE.2F Move Two Fractional Words from MOVE.2FMemory to a Register Pair AGU DescriptionDaDb Data Register Pairs MOVE.2F EA,DaDbMOVE.2L Move.2l d0d1,r0Da,Db ↔ EA MOVE.2L DaDb,EA MOVE.2L EA,DaDbRead/Write Notation MOVE.2W Move.2w d0d1,r0EA ↔ DaDb MOVE.2W EA,DaDb MOVE.2W DaDb,EA$FF Ffff AF44 MOVE.4F Move Four Fractional Words from MOVE.4FMemory to a Register Quad AGU EA → DaDbDcDdMove.4f r0,d0d1d2d3 DaDbDcDd Data Register QuadMOVE.4W EA ↔ DaDbDcDdMOVE.4W EA,DaDbDcDd MOVE.4W DaDbDcDd,EA Move.4w d0d1d2d3,r0 MOVE.B Byte Move AGUMove.b d3,r7+$3 MOVE.B DR,eaMOVE.B SP+s15,DR MOVE.B DR,SP+s15MOVE.B a16,DR A32 S15MOVE.F Move Fractional WordTo/from Memory AGU Operation Assembler Syntax Move.f $54,d10 MOVE.F SP+s15,DbMOVE.F Db,ea MOVE.F #s16,Db MOVE.F a16,Db SC140 DSP Core Reference Manual 271 MOVE.L Move Long Word AGUCccc General Registers#s32 #u32MOVE.L SP+s15,De.E MOVE.L Da.EDb.E,SP+s15Move.l d0.ed1.e,$1224 MOVE.L SP+s15,Do.EMOVE.L a32,De.E MOVE.L Da.EDb.E,a32Data Register Da.EDb.E ff Data Register Extension Pair278 Move Long AGU MOVE.L a32,DR MOVE.L DR,a32 MOVE.L a16,C4 MOVE.L C4,a16MOVE.L Rn+u3,DR MOVE.L DR,Rn+u3 MOVE.L Rn+s15,DR MOVE.L DR,Rn+s15Move.l d0,r0 MOVE.L SP+s15,C4 MOVE.L C4,SP+s15MOVE.L EA,DR MOVE.L DR,EA Rrr Address Register Read/Write NotationUnsigned 3-bit offset MOVE.W #s7,DR MOVE.W #s16,C4MOVE.W #s16,a16 MOVE.W #s16,SP-u5Move.w #$0050,r7 MOVE.W #s7,DR MOVE.W #s16,C4 #s7 Sa16MOVE.W Move Integer Word AGUMOVE.W a32,DR MOVE.W DR,a32 MOVE.W a16,C4 MOVE.W C4,a16MOVE.W Rn+u3,DR MOVE.W DR,Rn+u3 MOVE.W Rn+s15,DR MOVE.W DR,Rn+s15MOVE.W Rn+Rr,DR MOVE.W DR,Rn+Rr MOVE.W Rn,C3 MOVE.W C3,RnMove.w d1,r7+4 MOVE.W a32,DR MOVE.W DR,a32 Write Sss0 MOVEc Conditional Address Register Move AGU MOVEc Operation Movet r0,r1Movet Rq,Rn Movef Rq,RnQqq Address Register MOVES.2F Move Two Fractional Words to MOVES.2F DaDb → EAMOVES.2F DaDb,EA $7FFF $7EAC Moves.4f d0d1d2d3,r0 DaDbDcDd → EAMOVES.4F DaDbDcDd,EA $7FFF MOVES.F Move Fractional Word toMoves.f d0,r0 MOVES.F Db,a16 304 MOVES.L Move Long toMemory With Scaling and Saturation AGU Operation Moves.l d0,r0MOVES.L Db,EA MOVEU.B Move Unsigned Byte fromMemory AGU Operation Assembler Syntax Moveu.b $0053,d10 MOVEU.B a16,DR 310 MOVEU.L Moveu.l #$fffffff8,d3#u32 → Db MOVEU.L #u32,Db31IIIIIIIIIIIIIIII16 Moveu.w #$2345,d10.l #u16 → Db3116#u16 → Db150 MOVEU.W #u16,Db.H#u16 Iiiiiiiiiiiiiiii Bit unsigned immediate dataMOVEU.W Move Unsigned Word fromMemory to a Register AGU Operation Moveu.w r7+2,d10 MOVEU.W a16,C4 318 MPY Mpy d4,d5,d6Da.H * Db.H → Dn MPY Da,Db,DnMpy d6,d6,d7 SC140 DSP Core Reference Manual 321 Mpyr Mpyr d4,d5,d6RndDa.H * Db.H → Dn Mpyr Da,Db,Dn$0000 Register/Memory Address Before After L6D6324 Mpysu Mpysu d4,d5,d6Dc.H * Dd.L → Dn Mpysu Dc,Dd,Dn326 Mpyus Dc.L * Dd.H → DnMpyus Dc,Dd,Dn 328 Mpyuu Mpyuu d4,d5,d6Dc.L * Dd.L → Dn Mpyuu Dc,Dd,Dn330 NEG Negate DaluNeg d3 NEG DnNEG Dn NOP No Operation PrefixNo operation NopNot Bitwise Complement DaluNot d4,d5 ~Da → DnSC140 DSP Core Reference Manual 335 Binary Inversion of a 16-Bit Operand BMU Not D0.L~DR.L → DR.L Not DR.L NOT.W Binary Inversion of a 16-Bit OperandMemory BMU Operation Assembler Syntax Not.w r1 Bitwise Inclusive or Dalu Example or d3,d0Da Dn → Dn Or Da,DnSC140 DSP Core Reference Manual 341 Or #$0f0a,d0.l #u16 DR.L → DR.L#u16 DR.H → DR.H Or #u16,DR.LOr #u16,DR.L Or #u16,DR.H OR.W #u16,Rn OR.W #u16,SP-u5OR.W #u16,SP+s16 OR.W #u16,a16Or.w #$f01a,r1 OR.W #u16,Rn346 SP 8 → De SP 8 → SP SP 4 → Do SP 8 → SPPOP De POP DoPop d3 Eeeee Extension Pairs, Even Registers, and Loop Start RegistersNSP 8 → De NSP 8 → ΝSP NSP 4 → Do NSP 8 → ΝSPPopn d6.ed7.e Popn DePopn Do 352 De → SP SP + 8 → SP Do → SP + 4 SP + 8 → SPPush De Push DoPush d0.ed1.e SC140 DSP Core Reference Manual 355 De → NSP NSP + 8 → ΝSP Do → NSP + 4 NSP + 8 → ΝSPPushn De Pushn DoPushn d0.ed1.e Pushn De Pushn Do RND Round DaluRndDa → Dn RND Da,DnRnd d1,d5 Rnd d2,d1RND Da,Dn Rol d5 Dn3801 → Dn391Dn39 → C → Dn0 ROL DnROL Dn Ror d15 Dn39-11 → Dn38-0→ Dn39 Dn0 → C ROR DnROR Dn RTE SP 8 → PCSP 4 → SR SP 8 → SP → Nmid Rte Instruction Words Cycles1 TypeRted Example rted TrapRTS Return From Subroutine AGURts If RAS valid, then RAS → PCRTS Rtsd RtsdRtsd Restore PC from Stack AGU RtstkCleared Register Address Bit Name Description EMR3Example rtstk Rtstkd SP 8 → PCSP 8 → SP Example rtstkd SAT.F Sat.f d2,d3If Da $007FFFFFFF then $007FFF0000 → Dn SAT.F Da,DnSAT.F Da,Dn SAT.L Sat.l d6SAT.L Dn SC140 DSP Core Reference Manual 381 SBC Subtract With Borrow DaluDb Dc C → Dd SBC Dc,DdSBC Dc,Dd SBR Subtract And Round DaluSbr d3,d0 RndDn Da → Dn0010 1010 1110 0111 0000 0000 1000$2AE7 If LCn ≤ Then PC + displacement → PC Skipls label → LFn SkiplsSkipls label Skipls labelSkipls label Stop Stop Stop Instruction Processing AGU OperationEnter the stop processing state SUB Subtract DaluSub d1,d0,d2 SUB #u5,DnSub d0,d1,d2 SC140 DSP Core Reference Manual 391 SUB2 Subtract Two 16-Bit Values DaluSub2 d0,d1 SUB2 Da,Dn Suba Subtract AGUSuba r1,r0 Suba #u5,RxSuba Subl Shift Left and Subtract DaluSubl d0,d1 Dn Da → Dn$0$FF Ffff Fffe SUBNC.W Subnc.w #$15,d0Dn #s16 → Dn SUBNC.W #s16,DnSUBNC.W #s16,Dn Sign-Extension Dalu Sxt.b d3,d0Sxt.w d3,d2 Sxt.l d3 Sign-Extension AGU Sxta.b r3,r1Sxta.w r3 SXTA.B TFR Transfer Data Register to Data Register DaluTfr d15,d14 Tfr d7,d6TFR Da,Dn Tfra Tfra r0,r1Rx → Rx Tfra rx,RxSC140 DSP Core Reference Manual 407 To/from a Register AGU If Srexp = Then NSP → RnElse ESP → Rn If Srexp = Then Rn → NSP Else Rn → ESPTfra r0,osp Tfrt d14,d15 If T=1, then Da → DnIf T=0, then Da → Dn Tfrt Da, DnTfrt Trap Execute a Software Exception AGU Trap Operation TRAPnTrap Tsteq Test for Equal to Zero DaluTsteq d1 If Dn == 0, then 1 → T, else 0 → TTSTEQA.x Test for Equal to Zero AGU TSTEQA.x Operation Tsteqa.w r4Tsteqa.l r1 TSTEQA.W RxTSTEQA.W TSTEQA.L Tstge Test for Greater Than Or Equal to Zero DaluTstge d4 If Dn = 0, then 1 → T, else 0 → TTstgea.l r7 If Rx ≥ 0, then 1 → T, else 0 → TTESTGEA.L Rx TSTGEA.L Rx Tstgt Test for Greater Than Zero Dalu Tstgt Operation Tstgt d6If Dn 0, then 1 → T, else 0 → Τ Tstgt DnTstgta Test for Greater Than Zero AGU Tstgta Operation Tstgta r2If Rx 0, then 1 → T, else 0 → Τ Tstgta RxWord Big Endian Mode Little Endian ModeVSL Viterbi Shift Left Move AGUVSL.4W D2D6D1D3,Rn+N0 VSL.4F D2D6D1D3,Rn+N0VSL.2W D1D3,Rn+N0 VSL.2F D1D3,Rn+N0Vsl.2w d1d3,r0+n0 After Little EndianAfter Big Endian VSL.4W Enters the low-power standby Wait processing WaitState ResponseWait Zero Extension Dalu Zxt.b d2,d5Zxt.w d3,d6 Zxt.l d0 Zero Extension AGU Zxta.b r3,n2Zxta.w r4 ZXTA.B ZXTA.x 432 Using the StarCore Registry Appendix B StarCore RegistryTable B-1. Scid Assignments Hex Bits Instruction Cores ExampleSet Version SoC / platform Index Ecnten Eeddef Eselctrl 4-26ESELDI 4-26ESELDM 4-26ESELDTB 4-26 Eseletb MOVES.F A-299 MOVES.L A-301 MOVEU.B A-307 Macsu A-239 Macus A-241 Macuu A-243 Mpysu A-325 Mpyus A-327 Mpyuu A-329 Rtstk A-374 Index Index SC140 DSP Core Reference Manual