U s e r ’ s M a n u a l , V 1 .0 , J u n e 2004 XC167-32 16-Bit Single-Chip Microcontroller with C166SV2 Core Volume 1 (of 2): System Units Microcontrollers N e v e r s t o p t h i n k i n g . Edition 2004-06 Published by Infineon Technologies AG, St.-Martin-Strasse 53, 81669 München, Germany © Infineon Technologies AG 2004. All Rights Reserved. Attention please! The information herein is given to describe certain components and shall not be considered as a guarantee of characteristics. Terms of delivery and rights to technical change reserved. We hereby disclaim any and all warranties, including but not limited to warranties of non-infringement, regarding circuits, descriptions and charts stated herein. Information For further information on technology, delivery terms and conditions and prices please contact your nearest Infineon Technologies Office (www.infineon.com). Warnings Due to technical requirements components may contain dangerous substances. For information on the types in question please contact your nearest Infineon Technologies Office. Infineon Technologies Components may only be used in life-support devices or systems with the express written approval of Infineon Technologies, if a failure of such components can reasonably be expected to cause the failure of that life-support device or system, or to affect the safety or effectiveness of that device or system. Life support devices or systems are intended to be implanted in the human body, or to support and/or maintain and sustain and/or protect human life. If they fail, it is reasonable to assume that the health of the user or other persons may be endangered. U s e r ’ s M a n u a l , V 1 .0 , J u n e 2004 XC167-32 16-Bit Single-Chip Microcontroller with C166SV2 Core Volume 1 (of 2): System Units Microcontrollers N e v e r s t o p t h i n k i n g . XC167-32 Volume 1 (of 2): System Units Revision History: V1.0, 2004-06 Previous Version: None Page Subjects (major changes since last version) Controller Area Network (CAN): License of Robert Bosch GmbH We Listen to Your Comments Any information within this document that you feel is wrong, unclear or missing at all? Your feedback will help us to continuously improve the quality of this document. Please send your proposal (including a reference to this document) to: [email protected] Template: mc_tmplt_a5.fm / 3 / 2003-09-01 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page This User’s Manual consists of two Volumes, “System Units” and “Peripheral Units”. For your convenience this table of contents (and also the keyword index) lists both volumes, so you can immediately find the reference to the desired section in the corresponding document ([1] or [2]). 1 1.1 1.2 1.3 1.4 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 [1] Members of the 16-bit Microcontroller Family . . . . . . . . . . . . . . . . . . . 1-3 [1] Summary of Basic Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5 [1] Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-9 [1] Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-10 [1] 2 2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5 2.1.6 2.2 2.3 2.4 2.5 2.6 2.7 Architectural Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 [1] Basic CPU Concepts and Optimizations . . . . . . . . . . . . . . . . . . . . . . . 2-2 [1] High Instruction Bandwidth/Fast Execution . . . . . . . . . . . . . . . . . . . 2-4 [1] Powerful Execution Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5 [1] High Performance Branch-, Call-, and Loop-Processing . . . . . . . . . 2-6 [1] Consistent and Optimized Instruction Formats . . . . . . . . . . . . . . . . 2-7 [1] Programmable Multiple Priority Interrupt System . . . . . . . . . . . . . . 2-8 [1] Interfaces to System Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9 [1] On-Chip System Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-10 [1] On-Chip Peripheral Blocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-14 [1] Clock Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-29 [1] Power Management Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-29 [1] On-Chip Debug Support (OCDS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-31 [1] Protected Bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-32 [1] 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.9.1 3.9.2 3.9.3 3.9.4 3.9.5 3.9.6 Memory Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 [1] Address Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 [1] Special Function Register Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-4 [1] Data Memory Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-9 [1] Program Memory Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-11 [1] System Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13 [1] IO Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-14 [1] External Memory Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-15 [1] Crossing Memory Boundaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16 [1] The On-Chip Program Flash Module . . . . . . . . . . . . . . . . . . . . . . . . . 3-17 [1] Flash Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19 [1] Command Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20 [1] Error Correction and Data Integrity . . . . . . . . . . . . . . . . . . . . . . . . 3-26 [1] Protection and Security Features . . . . . . . . . . . . . . . . . . . . . . . . . 3-28 [1] Flash Status Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-33 [1] Operation Control and Error Handling . . . . . . . . . . . . . . . . . . . . . . 3-36 [1] User’s Manual I-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 3.10 3.10.1 3.10.2 3.10.3 Program Memory Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Flash Memory Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Program SRAM Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IMB Control Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 4.1 4.2 4.2.1 4.2.2 4.2.3 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.5 4.5.1 4.5.2 4.6 4.7 4.7.1 4.7.2 4.7.3 4.7.4 4.7.5 4.8 4.8.1 4.8.2 4.8.3 4.9 4.9.1 4.9.2 4.9.3 4.9.4 4.9.5 4.9.6 4.9.7 4.9.8 4.9.9 4.9.10 4.10 Central Processing Unit (CPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 [1] Components of the CPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4 [1] Instruction Fetch and Program Flow Control . . . . . . . . . . . . . . . . . . . . 4-5 [1] Branch Detection and Branch Prediction Rules . . . . . . . . . . . . . . . . 4-7 [1] Correctly Predicted Instruction Flow . . . . . . . . . . . . . . . . . . . . . . . . 4-7 [1] Incorrectly Predicted Instruction Flow . . . . . . . . . . . . . . . . . . . . . . . 4-9 [1] Instruction Processing Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-11 [1] Pipeline Conflicts Using General Purpose Registers . . . . . . . . . . . 4-13 [1] Pipeline Conflicts Using Indirect Addressing Modes . . . . . . . . . . . 4-15 [1] Pipeline Conflicts Due to Memory Bandwidth . . . . . . . . . . . . . . . . 4-17 [1] Pipeline Conflicts Caused by CPU-SFR Updates . . . . . . . . . . . . . 4-20 [1] CPU Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-26 [1] Use of General Purpose Registers . . . . . . . . . . . . . . . . . . . . . . . . . . 4-29 [1] GPR Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-31 [1] Context Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-33 [1] Code Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-37 [1] Data Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-39 [1] Short Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-39 [1] Long Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-41 [1] Indirect Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-45 [1] DSP Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-47 [1] The System Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-53 [1] Standard Data Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-57 [1] 16-bit Adder/Subtracter, Barrel Shifter, and 16-bit Logic Unit . . . . 4-61 [1] Bit Manipulation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-61 [1] Multiply and Divide Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-63 [1] DSP Data Processing (MAC Unit) . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-65 [1] Representation of Numbers and Rounding . . . . . . . . . . . . . . . . . . 4-66 [1] The 16-bit by 16-bit Signed/Unsigned Multiplier and Scaler . . . . . 4-67 [1] Concatenation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-67 [1] One-bit Scaler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-67 [1] The 40-bit Adder/Subtracter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-67 [1] The Data Limiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-68 [1] The Accumulator Shifter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-68 [1] The 40-bit Signed Accumulator Register . . . . . . . . . . . . . . . . . . . . 4-69 [1] The MAC Unit Status Word MSW . . . . . . . . . . . . . . . . . . . . . . . . . 4-70 [1] The Repeat Counter MRW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-72 [1] Constant Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-74 [1] User’s Manual I-2 3-38 [1] 3-39 [1] 3-41 [1] 3-42 [1] V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 5 5.1 5.2 5.3 5.4 5.4.1 5.4.2 5.4.3 5.4.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 Interrupt and Trap Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 [1] Interrupt System Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-2 [1] Interrupt Arbitration and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4 [1] Interrupt Vector Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10 [1] Operation of the Peripheral Event Controller Channels . . . . . . . . . . 5-18 [1] The PEC Source and Destination Pointers . . . . . . . . . . . . . . . . . . 5-22 [1] PEC Transfer Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-24 [1] Channel Link Mode for Data Chaining . . . . . . . . . . . . . . . . . . . . . . 5-26 [1] PEC Interrupt Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-27 [1] Prioritization of Interrupt and PEC Service Requests . . . . . . . . . . . . 5-29 [1] Context Switching and Saving Status . . . . . . . . . . . . . . . . . . . . . . . . 5-31 [1] Interrupt Node Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-34 [1] External Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-35 [1] OCDS Requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-40 [1] Service Request Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-41 [1] Trap Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-43 [1] 6 6.1 6.1.1 6.1.2 6.1.3 6.1.4 6.1.5 6.1.6 6.1.7 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.4 6.4.1 6.4.2 6.4.3 General System Control Functions . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 [1] System Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2 [1] Reset Sources and Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3 [1] Status After Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-6 [1] Application-Specific Initialization Routine . . . . . . . . . . . . . . . . . . . 6-11 [1] System Startup Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-14 [1] Hardware Configuration in External Start Mode . . . . . . . . . . . . . . 6-18 [1] Default Configuration in Single-Chip Mode . . . . . . . . . . . . . . . . . . 6-23 [1] Reset Behavior Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-24 [1] Clock Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-26 [1] Oscillators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-27 [1] Clock Generation and Frequency Control . . . . . . . . . . . . . . . . . . . 6-30 [1] Clock Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-37 [1] Oscillator Watchdog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-38 [1] Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-38 [1] Generation of an External Clock Signal . . . . . . . . . . . . . . . . . . . . 6-39 [1] Central System Control Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-43 [1] Status Indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-45 [1] Reset Source Indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-46 [1] Peripheral Shutdown Handshake . . . . . . . . . . . . . . . . . . . . . . . . . 6-47 [1] Debug System Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-48 [1] Register Security Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-50 [1] Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-54 [1] Power Reduction Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-54 [1] Reduction of Clock Frequencies . . . . . . . . . . . . . . . . . . . . . . . . . . 6-57 [1] Flexible Peripheral Management . . . . . . . . . . . . . . . . . . . . . . . . . . 6-57 [1] User’s Manual I-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 6.5 6.6 Watchdog Timer (WDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-59 [1] Identification Control Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-64 [1] 7 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13 Parallel Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 [1] Input Threshold Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2 [1] Output Driver Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3 [1] Alternate Port Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-8 [1] PORT0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9 [1] PORT1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13 [1] Port 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-24 [1] Port 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-29 [1] Port 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41 [1] Port 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-50 [1] Port 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-54 [1] Port 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-65 [1] Port 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-72 [1] Port 20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-82 [1] 8 Dedicated Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 [1] 9 9.1 9.2 9.2.1 9.2.1.1 9.2.1.2 9.2.2 9.2.2.1 9.2.2.2 9.2.2.3 9.2.2.4 9.2.2.5 9.2.2.6 9.2.3 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.3.5 9.3.6 9.3.6.1 9.3.6.2 9.3.7 The External Bus Controller EBC . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1 [1] External Bus Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-3 [1] Timing Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 [1] Basic Bus Cycle Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 [1] Demultiplexed Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-5 [1] Multiplexed Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-6 [1] Bus Cycle Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7 [1] A Phase - CS Change Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7 [1] B Phase - Address Setup/ALE Phase . . . . . . . . . . . . . . . . . . . . . 9-7 [1] C Phase - Delay Phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7 [1] D Phase - Write Data Setup/MUX Tristate Phase . . . . . . . . . . . . 9-7 [1] E Phase - RD/WR Command Phase . . . . . . . . . . . . . . . . . . . . . . 9-7 [1] F Phase - Address/Write Data Hold Phase . . . . . . . . . . . . . . . . . 9-8 [1] Bus Cycle Examples: Fastest Access Cycles . . . . . . . . . . . . . . . . . 9-8 [1] Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-10 [1] Configuration Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . 9-10 [1] The EBC Mode Register 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-12 [1] The EBC Mode Register 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-14 [1] The Timing Configuration Registers TCONCSx . . . . . . . . . . . . . . 9-15 [1] The Function Configuration Registers FCONCSx . . . . . . . . . . . . . 9-16 [1] The Address Window Selection Registers ADDRSELx . . . . . . . . . 9-18 [1] Definition of Address Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18 [1] Address Window Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-20 [1] Ready Controlled Bus Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-21 [1] User’s Manual I-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 9.3.7.1 9.3.7.2 9.3.7.3 9.3.8 9.3.9 9.3.9.1 9.3.9.2 9.3.9.3 9.3.9.4 9.3.9.5 9.3.10 9.4 9.5 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Synchronous/Asynchronous READY . . . . . . . . . . . . . . . . . Combining the READY Function with Predefined Wait States . Access Control to TwinCAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Bus Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Initialization of Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arbitration Master Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arbitration Slave Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bus Lock Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Master Slave Connection . . . . . . . . . . . . . . . . . . . . . . . . Shutdown Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LXBus Access Control and Signal Generation . . . . . . . . . . . . . . . . . EBC Register Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-21 [1] 9-22 [1] 9-22 [1] 9-23 [1] 9-24 [1] 9-24 [1] 9-24 [1] 9-26 [1] 9-27 [1] 9-27 [1] 9-28 [1] 9-29 [1] 9-29 [1] 10 10.1 10.2 10.3 10.4 The Bootstrap Loader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Entering the Bootstrap Loader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Loading the Startup Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exiting Bootstrap Loader Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Choosing the Baudrate for the BSL . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 [1] 10-2 [1] 10-4 [1] 10-4 [1] 10-5 [1] 11 11.1 11.2 11.3 11.3.1 11.3.2 11.4 11.4.1 Debug System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OCDS Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cerberus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 [1] 11-1 [1] 11-2 [1] 11-3 [1] 11-5 [1] 11-6 [1] 11-7 [1] 11-7 [1] 12 Instruction Set Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1 [1] 13 Device Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1 [1] 14 14.1 14.1.1 14.1.2 14.1.3 14.1.4 14.1.5 14.1.6 14.1.7 14.2 14.2.1 The General Purpose Timer Units . . . . . . . . . . . . . . . . . . . . . . . . . 14-1 [2] Timer Block GPT1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-2 [2] GPT1 Core Timer T3 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-4 [2] GPT1 Core Timer T3 Operating Modes . . . . . . . . . . . . . . . . . . . . . 14-8 [2] GPT1 Auxiliary Timers T2/T4 Control . . . . . . . . . . . . . . . . . . . . . 14-15 [2] GPT1 Auxiliary Timers T2/T4 Operating Modes . . . . . . . . . . . . . 14-18 [2] GPT1 Clock Signal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-27 [2] GPT1 Timer Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-29 [2] Interrupt Control for GPT1 Timers . . . . . . . . . . . . . . . . . . . . . . . . 14-30 [2] Timer Block GPT2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-31 [2] GPT2 Core Timer T6 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-33 [2] User’s Manual I-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 14.2.2 14.2.3 14.2.4 14.2.5 14.2.6 14.2.7 14.2.8 14.3 GPT2 Core Timer T6 Operating Modes . . . . . . . . . . . . . . . . . . . . GPT2 Auxiliary Timer T5 Control . . . . . . . . . . . . . . . . . . . . . . . . GPT2 Auxiliary Timer T5 Operating Modes . . . . . . . . . . . . . . . . . GPT2 Register CAPREL Operating Modes . . . . . . . . . . . . . . . . . GPT2 Clock Signal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . GPT2 Timer Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Control for GPT2 Timers and CAPREL . . . . . . . . . . . . . Interfaces of the GPT Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 15.1 15.2 15.3 15.3.1 15.3.2 15.3.3 15.4 Real Time Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1 [2] Defining the RTC Time Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-2 [2] RTC Run Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-5 [2] RTC Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-7 [2] 48-bit Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-10 [2] System Clock Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-10 [2] Cyclic Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-11 [2] RTC Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-12 [2] 16 16.1 16.1.1 16.1.2 16.2 16.2.1 16.2.2 16.2.3 16.2.4 16.3 16.4 16.5 16.6 The Analog/Digital Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1 [2] Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-3 [2] Compatibility Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-3 [2] Enhanced Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-5 [2] ADC Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-8 [2] Fixed Channel Conversion Modes . . . . . . . . . . . . . . . . . . . . . . . . 16-11 [2] Auto Scan Conversion Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-12 [2] Wait for Read Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-13 [2] Channel Injection Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-14 [2] Automatic Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-17 [2] Conversion Timing Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-18 [2] A/D Converter Interrupt Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-21 [2] Interfaces of the ADC Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-22 [2] 17 17.1 17.2 17.3 17.4 17.5 17.5.1 17.5.2 17.5.3 17.5.4 17.5.5 17.6 Capture/Compare Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1 [2] The CAPCOM Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-4 [2] CAPCOM Timer Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-9 [2] Capture/Compare Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-10 [2] Capture Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-13 [2] Compare Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-14 [2] Compare Mode 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-15 [2] Compare Mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-15 [2] Compare Mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-18 [2] Compare Mode 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-18 [2] Double-Register Compare Mode . . . . . . . . . . . . . . . . . . . . . . . . 17-22 [2] Compare Output Signal Generation . . . . . . . . . . . . . . . . . . . . . . . . 17-25 [2] User’s Manual I-6 14-37 [2] 14-40 [2] 14-42 [2] 14-46 [2] 14-51 [2] 14-54 [2] 14-55 [2] 14-56 [2] V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 17.7 17.8 17.9 17.10 17.11 Single Event Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Staggered and Non-Staggered Operation . . . . . . . . . . . . . . . . . . . . CAPCOM Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Input Signal Requirements . . . . . . . . . . . . . . . . . . . . . . . . Interfaces of the CAPCOM Units . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 18.1 18.1.1 18.1.2 18.1.3 18.1.4 18.1.5 18.2 18.2.1 18.2.2 18.3 18.4 18.5 18.5.1 18.5.2 18.5.3 18.5.4 18.6 18.7 18.8 18.9 18.10 18.11 Capture/Compare Unit 6 (CAPCOM6) . . . . . . . . . . . . . . . . . . . . . . 18-1 [2] Timer T12 Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-4 [2] Timer T12 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-7 [2] T12 Compare Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-12 [2] Dead-Time Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-22 [2] T12 Capture Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-25 [2] Hysteresis-Like Control Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-29 [2] Timer T13 Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-30 [2] T13 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-33 [2] T13 Compare Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-38 [2] Timer Block Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-42 [2] Multi-Channel Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-48 [2] Hall Sensor Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-51 [2] Hall Pattern Compare Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-52 [2] Sampling of the Hall Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-53 [2] Brushless DC-Motor Control with Timer T12 Block . . . . . . . . . . . 18-55 [2] Hall Mode Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-57 [2] Trap Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-63 [2] Output Modulation Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-67 [2] Shadow Register Transfer Control . . . . . . . . . . . . . . . . . . . . . . . . . 18-71 [2] Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-73 [2] Suspend Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-82 [2] Interfaces of the CAPCOM6 Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-83 [2] 19 19.1 19.2 19.2.1 19.2.2 19.2.3 19.2.4 19.2.5 19.2.6 19.2.7 19.2.8 19.3 19.3.1 Asynchronous/Synchronous Serial Interface (ASC) . . . . . . . . . . 19-1 [2] Operational Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-3 [2] Asynchronous Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-5 [2] Asynchronous Data Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-6 [2] Asynchronous Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-9 [2] Transmit FIFO Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-9 [2] Asynchronous Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-12 [2] Receive FIFO Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-12 [2] FIFO Transparent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-15 [2] IrDA Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-16 [2] RxD/TxD Data Path Selection in Asynchronous Modes . . . . . . . 19-17 [2] Synchronous Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-19 [2] Synchronous Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-20 [2] User’s Manual I-7 17-27 [2] 17-29 [2] 17-34 [2] 17-36 [2] 17-37 [2] V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 19.3.2 19.3.3 19.4 19.4.1 19.4.2 19.5 19.5.1 19.5.2 19.5.3 19.5.4 19.6 19.7 19.8 19.9 Synchronous Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synchronous Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Baudrate Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Baudrate in Asynchronous Mode . . . . . . . . . . . . . . . . . . . . . . . . . Baudrate in Synchronous Mode . . . . . . . . . . . . . . . . . . . . . . . . . Autobaud Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Serial Frames for Autobaud Detection . . . . . . . . . . . . . . . . . . . . . Baudrate Selection and Calculation . . . . . . . . . . . . . . . . . . . . . . . Overwriting Registers on Successful Autobaud Detection . . . . . Hardware Error Detection Capabilities . . . . . . . . . . . . . . . . . . . . . . Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interfaces of the ASC Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 20.1 20.2 20.2.1 20.2.2 20.2.3 20.2.4 20.2.5 20.2.6 20.2.7 20.2.8 20.3 High-Speed Synchronous Serial Interface (SSC) . . . . . . . . . . . . 20-1 [2] Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 [2] Operational Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 [2] Operating Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-3 [2] Full-Duplex Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-8 [2] Half-Duplex Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-11 [2] Continuous Transfers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-12 [2] Baudrate Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-12 [2] Error Detection Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-14 [2] SSC Register Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-16 [2] Port Configuration Requirements . . . . . . . . . . . . . . . . . . . . . . . . 20-17 [2] Interfaces of the SSC Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-18 [2] 21 21.1 21.2 21.3 21.3.1 21.3.2 21.3.3 21.3.4 21.3.5 21.3.6 21.4 21.5 21.6 21.7 IIC-Bus Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1 [2] Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-2 [2] Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-5 [2] IIC-Bus Module Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-12 [2] Operation in Single-Master Mode . . . . . . . . . . . . . . . . . . . . . . . . 21-12 [2] Operation in Multimaster Mode . . . . . . . . . . . . . . . . . . . . . . . . . . 21-12 [2] Operation in Slave Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-13 [2] Transmit/Receive Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-14 [2] Baud Rate Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-15 [2] Notes for Programming the IIC-Bus Module . . . . . . . . . . . . . . . . 21-16 [2] Interrupt Request Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-17 [2] Port Connection and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . 21-19 [2] Interfaces of the IIC-Bus Module . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-21 [2] IIC-Bus Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-22 [2] 22 TwinCAN Module User’s Manual 19-20 [2] 19-20 [2] 19-22 [2] 19-22 [2] 19-26 [2] 19-27 [2] 19-27 [2] 19-28 [2] 19-29 [2] 19-33 [2] 19-34 [2] 19-35 [2] 19-39 [2] 19-56 [2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 [2] I-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents Page 22.1 Kernel Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 [2] 22.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 [2] 22.1.2 TwinCAN Control Shell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-4 [2] 22.1.2.1 Initialization Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-4 [2] 22.1.2.2 Interrupt Request Compressor . . . . . . . . . . . . . . . . . . . . . . . . . 22-5 [2] 22.1.2.3 Global Control and Status Logic . . . . . . . . . . . . . . . . . . . . . . . . 22-6 [2] 22.1.3 CAN Node Control Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-7 [2] 22.1.3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-7 [2] 22.1.3.2 Timing Control Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-9 [2] 22.1.3.3 Bitstream Processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-11 [2] 22.1.3.4 Error Handling Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-11 [2] 22.1.3.5 Node Interrupt Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-12 [2] 22.1.3.6 Message Interrupt Processing . . . . . . . . . . . . . . . . . . . . . . . . 22-13 [2] 22.1.3.7 Interrupt Indication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-13 [2] 22.1.4 Message Handling Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-15 [2] 22.1.4.1 Arbitration and Acceptance Mask Register . . . . . . . . . . . . . . . 22-16 [2] 22.1.4.2 Handling of Remote and Data Frames . . . . . . . . . . . . . . . . . . 22-17 [2] 22.1.4.3 Handling of Transmit Message Objects . . . . . . . . . . . . . . . . . . 22-18 [2] 22.1.4.4 Handling of Receive Message Objects . . . . . . . . . . . . . . . . . . 22-21 [2] 22.1.4.5 Single Data Transfer Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-23 [2] 22.1.5 CAN Message Object Buffer (FIFO) . . . . . . . . . . . . . . . . . . . . . . 22-24 [2] 22.1.5.1 Buffer Access by the CAN Controller . . . . . . . . . . . . . . . . . . . 22-26 [2] 22.1.5.2 Buffer Access by the CPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-27 [2] 22.1.6 Gateway Message Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-28 [2] 22.1.6.1 Normal Gateway Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-29 [2] 22.1.6.2 Normal Gateway with FIFO Buffering . . . . . . . . . . . . . . . . . . . 22-33 [2] 22.1.6.3 Shared Gateway Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-36 [2] 22.1.7 Programming the TwinCAN Module . . . . . . . . . . . . . . . . . . . . . . 22-40 [2] 22.1.7.1 Configuration of CAN Node A/B . . . . . . . . . . . . . . . . . . . . . . . 22-40 [2] 22.1.7.2 Initialization of Message Objects . . . . . . . . . . . . . . . . . . . . . . . 22-40 [2] 22.1.7.3 Controlling a Message Transfer . . . . . . . . . . . . . . . . . . . . . . . 22-41 [2] 22.1.8 Loop-Back Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-44 [2] 22.1.9 Single Transmission Try Functionality . . . . . . . . . . . . . . . . . . . . 22-45 [2] 22.1.10 Module Clock Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-46 [2] 22.2 TwinCAN Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-47 [2] 22.2.1 Register Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-47 [2] 22.2.2 CAN Node A/B Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-49 [2] 22.2.3 CAN Message Object Registers . . . . . . . . . . . . . . . . . . . . . . . . . 22-64 [2] 22.2.4 Global CAN Control/Status Registers . . . . . . . . . . . . . . . . . . . . . 22-80 [2] 22.3 XC167 Module Implementation Details . . . . . . . . . . . . . . . . . . . . . . 22-82 [2] 22.3.1 Interfaces of the TwinCAN Module . . . . . . . . . . . . . . . . . . . . . . . 22-82 [2] 22.3.2 TwinCAN Module Related External Registers . . . . . . . . . . . . . . . 22-83 [2] User’s Manual I-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Table of Contents 22.3.2.1 22.3.2.2 22.3.2.3 22.3.3 23 23.1 23.2 Page System Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Port Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-84 [2] 22-85 [2] 22-90 [2] 22-91 [2] Register Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-1 [2] PD+BUS Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-1 [2] LXBUS Peripherals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23-16 [2] Keyword Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i-1 [1+2] User’s Manual I-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction 1 Introduction The rapidly growing area of embedded control applications is representing one of the most time-critical operating environments for today’s microcontrollers. Complex control algorithms have to be processed based on a large number of digital as well as analog input signals, and the appropriate output signals must be generated within a defined maximum response time. Embedded control applications also are often sensitive to board space, power consumption, and overall system cost. Embedded control applications therefore require microcontrollers, which: • • • • offer a high level of system integration eliminate the need for additional peripheral devices and the associated software overhead provide system security and fail-safe mechanisms provide effective means to control (and reduce) the device’s power consumption The increasing complexity of embedded control applications requires microcontrollers for new high-end embedded control systems to possess a significant increase in CPU performance and peripheral functionality over conventional 8-bit controllers. To achieve this high performance goal Infineon has decided to develop its families of 16-bit CMOS microcontrollers without the constraints of backward compatibility. Nonetheless the architectures of the 16-bit microcontroller families pursue successful hardware and software concepts, which have been established in Infineon’s popular 8-bit controller families. User’s Manual Introduction_X73, V2.1 1-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction About this Manual This manual describes the functionality of a number of 16-bit microcontrollers of the Infineon XC166 Family. These microcontrollers provide identical functionality to a large extent, but each device type has specific unique features as indicated here. The descriptions in this manual cover a superset of the provided features and refer to the following derivatives: • XC167CS-32F – 256 Kbytes Program Flash, 12 Kbytes on-chip RAM, – 16 analog input channels, – 7 serial interfaces (2 × ASC, 2 × SSC, 2 × CAN, IIC) This manual is valid for these derivatives and describes all variations of the different available temperature ranges and packages. For simplicity, these various device types are referred to by the collective term XC167 throughout this manual. The complete pro-electron conforming designations are listed in the respective data sheets. Some sections of this manual do not refer to all of the XC167 derivatives which are currently available or planned (such as devices with different types of on-chip memory or peripherals). These sections contain respective notes wherever possible. User’s Manual Introduction_X73, V2.1 1-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction 1.1 Members of the 16-bit Microcontroller Family The microcontrollers in the Infineon 16-bit family have been designed to meet the high performance requirements of real-time embedded control applications. The architecture of this family has been optimized for high instruction throughput and minimized response time to external stimuli (interrupts). Intelligent peripheral subsystems have been integrated to reduce the need for CPU intervention to a minimum extent. This also minimizes the need for communication via the external bus interface. The high flexibility of this architecture allows to serve the diverse and varying needs of different application areas such as automotive, industrial control, or data communications. The core of the 16-bit family has been developed with a modular family concept in mind. All family members execute an efficient control-optimized instruction set (additional instructions for members of the second generation). This allows easy and quick implementation of new family members with different internal memory sizes and technologies, different sets of on-chip peripherals, and/or different numbers of IO pins. The XBUS concept (internal representation of the external bus interface) provides a straightforward path for building application-specific derivatives by integrating application-specific peripheral modules with the standard on-chip peripherals. As programs for embedded control applications become larger, high level languages are favored by programmers, because high level language programs are easier to write, to debug and to maintain. The C166 Family supports this starting with its 2nd generation. The 80C166-type microcontrollers were the first generation of the 16-bit controller family. These devices established the C166 architecture. The C165-type and C167-type devices are members of the second generation of this family. This second generation is even more powerful due to additional instructions for HLL support, an increased address space, increased internal RAM, and highly efficient management of various resources on the external bus. Enhanced derivatives of this second generation provide more features such as additional internal high-speed RAM, an integrated CAN-Module, an on-chip PLL, etc. The design of more efficient systems may require the integration of application-specific peripherals to boost system performance while minimizing the part count. These efforts are supported by the XBUS, defined for the Infineon 16-bit microcontrollers (second generation). The XBUS is an internal representation of the external bus interface which opens and simplifies the integration of peripherals by standardizing the required interface. One representative taking advantage of this technology is the integrated CAN module. The C165-type devices are reduced functionality versions of the C167 because they do not have the A/D converter, the CAPCOM units, and the PWM module. This results in a smaller package, reduced power consumption, and design savings. User’s Manual Introduction_X73, V2.1 1-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction The C164-type devices, the C167CS derivatives, and some of the C161-type devices are further enhanced by a flexible power management and form the third generation of the 16-bit controller family. This power management mechanism provides an effective means to control the power that is consumed in a certain state of the controller and thus minimizes the overall power consumption for a given application. The XC16x derivatives represent the fourth generation of the 16-bit controller family. The XC166 Family dramatically increases the performance of 16-bit microcontrollers by several major improvements and additions. The MAC-unit adds DSP-functionality to handle digital filter algorithms and greatly reduces the execution time of multiplications and divisions. The 5-stage pipeline, single-cycle execution of most instructions, and PEC-transfers within the complete addressing range increase system performance. Debugging the target system is supported by integrated functions for On-Chip Debug Support (OCDS). A variety of different versions is provided which offer various kinds of on-chip program memory1): • • • • Mask-programmable ROM Flash memory OTP memory ROMless without non-volatile memory Also there are devices with specific functional units. The devices may be offered in different packages, temperature ranges and speed classes. Additional standard and application-specific derivatives are planned and are in development. Note: Not all derivatives will be offered in all temperature ranges, speed classes, packages, or program memory variations. Information about specific versions and derivatives will be made available with the devices themselves. Contact your Infineon representative for up-to-date material or refer to http://www.infineon.com/microcontrollers. Note: As the architecture and the basic features, such as the CPU core and built-in peripherals, are identical for most of the currently offered versions of the XC167, descriptions within this manual that refer to the “XC167” also apply to the other variations, unless otherwise noted. 1) Not all derivatives are offered with all kinds of on-chip memory. User’s Manual Introduction_X73, V2.1 1-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction 1.2 Summary of Basic Features The XC167 devices are enhanced members of the Infineon family of full featured 16-bit single-chip CMOS microcontrollers. The XC167 combines the extended functionality and performance of the C166SV2 Core with powerful on-chip peripheral subsystems and on-chip memory units and provides a means for power reduction. Several key features contribute to the high performance of the XC167: High Performance 16-bit CPU with Five-Stage Pipeline and MAC Unit • • • • • • • • • • • • • • Single clock cycle instruction execution 1 cycle minimum instruction cycle time (most instructions) 1 cycle multiplication (16-bit × 16-bit) 4 + 17 cycles division (32-bit/16-bit), 4 cycles delay, 17 cycles background execution 1 cycle multiply and accumulate instruction (MAC) execution Automatic saturation or rounding included Multiple high bandwidth internal data buses Register-based design with multiple, variable register banks Two additional fast register banks Fast context switching support 16 Mbytes of linear address space for code and data (Von Neumann architecture) System stack cache support with automatic stack overflow/underflow detection High performance branch, call, and loop processing Zero-cycle jump execution Control Oriented Instruction Set with High Efficiency • • • • • Bit, byte, and word data types Flexible and efficient addressing modes for high code density Enhanced boolean bit manipulation with direct addressability of 6 kbits for peripheral control and user-defined flags Hardware traps to identify exception conditions during runtime HLL support for semaphore operations and efficient data access Power Management Features • • • • • Gated clock concept for improved power consumption and EMC Programmable system slowdown via clock generation unit Flexible management of peripherals, can be individually disabled Sleep-mode supports wake-up via fast external interrupts or on-chip RTC Programmable frequency output User’s Manual Introduction_X73, V2.1 1-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction Integrated On-Chip Memory • • • • Up to 2 Kbytes Dual-Port RAM (DPRAM) for variables, register banks, and stacks Up to 4 Kbytes on-chip high-speed Data SRAM (DSRAM) for variables and stacks Up to 6 Kbytes on-chip high-speed Program/Data SRAM (PSRAM) for code and data 256 Kbytes on-chip Program Memory for instruction code or constant data (Flash or Mask ROM, not for ROMless devices) Note: The system stack can be located in any memory area within the complete addressing range. External Bus Interface • • • • • • Up to 12 Mbytes external address space for code and data Multiplexed or demultiplexed bus configurations Segmentation capability and chip select signal generation 8-bit or 16-bit data bus Bus cycle characteristics selectable for five programmable address areas Hold- and Hold-Acknowledge bus arbitration support for external multimaster bus 16-Priority-Level Interrupt System • • • • • 80 interrupt nodes with separate interrupt vectors on 15 priority levels (8 group levels) 13 cycles minimum interrupt latency in case of internal program execution Fast external interrupts Programmable external interrupt source selection Programmable vector table (start location and step-width) 8-Channel Peripheral Event Controller (PEC) • • • • • • Interrupt driven single cycle data transfer Programmable PEC interrupt request level, (15 down to 8) Transfer count option (standard CPU interrupt after programmable number of PEC transfers) Separate interrupt level for PEC termination interrupts selectable Overhead from saving and restoring system state for interrupt requests eliminated Full 24-bit addresses for source and destination pointers, supporting transfers within the total address space Intelligent On-Chip Peripheral Subsystems • • 16-channel A/D Converter with programmable resolution (10-bit or 8-bit) and conversion time (down to 2.55 µs or 2.15 µs), auto scan modes, channel injection Two Capture/Compare Units with 2 independent time bases each, very flexible PWM unit/event recording unit with different operating modes, includes four 16-bit timers/counters, maximum resolution fSYS User’s Manual Introduction_X73, V2.1 1-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction • • • • • • • • • • Capture/Compare Unit for flexible PWM Signal Generation (CAPCOM6) (3/6 Capture/Compare Channels and 1 Compare Channel) Two Multifunctional General Purpose Timer Units: – GPT1: three 16-bit timers/counters, maximum resolution fSYS/4 – GPT2: two 16-bit timers/counters, maximum resolution fSYS/2 Two Asynchronous/Synchronous Serial Channels (USARTs) with baud rate generator, parity, framing, and overrun error detection, with auto baud rate detection, receive/transmit FIFOs, and IrDA support Two High Speed Synchronous Serial Channels (SPI-compatible) with programmable data length and shift direction Controller Area Network (TwinCAN) Module, Rev. 2.0B active, two nodes operating independently or exchanging data via a gateway function, Full-CAN/Basic-CAN IIC Bus Interface (10-bit addressing, 400 kbit/s) with 3 channels (multiplexed) Real Time Clock with alarm interrupt Watchdog Timer with programmable time intervals Bootstrap Loader for flexible system initialization Protection management for system configuration and control registers On-Chip Debug Support • • • • • • • • On-chip debug controller and related interface to JTAG controller JTAG interface and break interface on separate pins Hardware, software and external pin breakpoints Up to 4 instruction pointer breakpoints Debug event control, e.g. with monitor call or CPU halt or trigger of data transfer Dedicated DEBUG instructions with control via JTAG interface Access to any internal register or memory location via JTAG interface Single step support and watchpoints with MOV-injection Up to 103 IO Lines with Individual Bit Addressability • • • • • Tri-stated in input mode Selectable input thresholds (not on all pins) Push/pull or open drain output mode Programmable port driver control I/O voltage is 5 V (core-logic and oscillator input voltage is 2.5 V) Various Temperature Ranges1) • • • 0 to +70 °C -40 to +85 °C -40 to +125 °C 1) Not all derivatives are offered in all temperature ranges. User’s Manual Introduction_X73, V2.1 1-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction Infineon CMOS Process • Low power CMOS technology enables power saving Idle, Sleep, and Power Down modes with flexible power management. 144-Pin Plastic Thin Quad Flat Pack (TQFP) Package • P-TQFP, 20 × 20 mm body, 0.5 mm (19.7 mil) lead spacing, surface mount technology Complete Development Support For the development tool support of its microcontrollers, Infineon follows a clear third party concept. Currently around 120 tool suppliers world-wide, ranging from local niche manufacturers to multinational companies with broad product portfolios, offer powerful development tools for the Infineon C500, C166, and XC166 microcontroller families, guaranteeing a remarkable variety of price-performance classes as well as early availability of high quality key tools such as compilers, assemblers, simulators, debuggers or in-circuit emulators. Infineon incorporates its strategic tool partners very early into the product development process, making sure embedded system developers get reliable, well-tuned tool solutions, which help them unleash the power of Infineon microcontrollers in the most effective way and with the shortest possible learning curve. The tool environment for the Infineon 16-bit microcontrollers includes the following tools: • • • • • • • • • • • • • • Compilers (C, MODULA2, FORTH) Macro-assemblers, linkers, locators, library managers, format-converters Architectural simulators HLL debuggers Real-time operating systems VHDL chip models In-circuit emulators (based on bondout or standard chips) Plug-in emulators Emulation and clip-over adapters, production sockets Logic analyzer disassemblers Starter kits Evaluation boards with monitor programs Industrial boards (also for CAN, FUZZY, PROFIBUS, FORTH applications) Network driver software (CAN, PROFIBUS) User’s Manual Introduction_X73, V2.1 1-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction 1.3 Abbreviations The following acronyms and terms are used within this document: JTAG Joint Test Access Group ADC Analog Digital Converter ALE Address Latch Enable ALU Arithmetic and Logic Unit ASC Asynchronous/synchronous Serial Channel CAN Controller Area Network (License Bosch) CAPCOM CAPture and COMpare unit CISC Complex Instruction Set Computing CMOS Complementary Metal Oxide Silicon CPU Central Processing Unit DMU Data Management Unit EBC External Bus Controller ESFR Extended Special Function Register Flash Non-volatile memory that may be electrically erased GPR General Purpose Register GPT General Purpose Timer unit HLL High Level Language IIC Inter Integrated Circuit (Bus) IO Input/Output LXBus Internal representation of the external bus OCDS On-Chip Debug Support OTP One-Time Programmable memory PEC Peripheral Event Controller PLA Programmable Logic Array PLL Phase Locked Loop PMU Program Management Unit PWM Pulse Width Modulation RAM Random Access Memory RISC Reduced Instruction Set Computing User’s Manual Introduction_X73, V2.1 1-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Introduction ROM Read Only Memory RTC Real Time Clock SFR Special Function Register SSC Synchronous Serial Channel 1.4 Naming Conventions The manifold bitfields used for control functions and status indication and the registers housing them are equipped with unique names wherever applicable. Thereby these control structured can be referred to by their names rather than by their location. This makes the descriptions by far more comprehensible. To describe regular structures (such as ports) indices are used instead of a plethora of similar bit names, so bit 3 of port 5 is referred to as P5.3. Where it helps to clarify the relation between several named structures, the next higher level is added to the respective name to make it unambiguous. The term ADC_CTR0 clearly identifies register CTR0 as part of module ADC, the term SYSCON1.CPSYS clearly identifies bitfield CPSYS as part of register SYSCON1. User’s Manual Introduction_X73, V2.1 1-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2 Architectural Overview The architecture of the XC167 core combines the advantages of both RISC and CISC processors in a very well-balanced way. This computing and controlling power is completed by the DSP-functionality of the MAC-unit. The XC167 integrates this powerful CPU core with a set of powerful peripheral units into one chip and connects them very efficiently. On-chip memory blocks with dedicated buses and control units store code and data. This combination of features results in a high performance microcontroller, which is the right choice not only for today’s applications, but also for future engineering challenges. One of the buses used concurrently on the XC167 is the LXBus, an internal representation of the external bus interface. This bus provides a standardized method for integrating additional application-specific peripherals into derivatives of the standard XC167. PSRAM DPRAM DSRAM Flash 256 KBytes DMU EBC PMU ProgMem CPU XBUS Control External Bus Control C166SV2-Core OCDS Debug Support XTAL Osc / PLL RTC Clock Generation WDT Interrupt & PEC Interrupt Bus Peripheral Data Bus ADC GPT 8/10-Bit 16 Channels ASC0 ASC1 (USART) T2 (USART) SSC0 SSC1 (SPI) (SPI) T3 CC1 CC2 T0 T7 IIC CC6 T12 T1 T8 T13 Twin CAN T4 A B T5 T6 BRGen P 20 Port 9 P 7 6 6 Port 6 4 8 BRGen Port 5 16 BRGen BRGen Port 4 BRGen Port 3 8 15 Port 2 8 PORT1 PORT0 16 16 MCB04323_x73.vsd Figure 2-1 XC167 Functional Block Diagram User’s Manual Architecture_X73, V2.1 2-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1 Basic CPU Concepts and Optimizations The main core of the CPU consists of a set of optimized functional units including the instruction fetch/processing pipelines, a 16-bit Arithmetic and Logic Unit (ALU), a 40-bit Multiply and Accumulate Unit (MAC), an Address and Data Unit (ADU), an Instruction Fetch Unit (IFU), a Register File (RF), and dedicated Special Function Registers (SFRs). Single clock cycle execution of instructions results in superior CPU performance, while maintaining C166 code compatibility. Impressive DSP performance, concurrent access to different kinds of memories and peripherals boost the overall system performance. PSRAM Flash/RO M PM U CPU P refetch U nit B ranch U nit FIFO CSP VECSEG CPUCON1 CPUCON2 R etu rn S tack ID X 0 ID X 1 QX0 QX1 QR0 QR1 + /- + /- M u ltip ly U nit IP MRW + /- MCW MSW MAH MAL 2-S tage P refetch P ipeline TF R Injection/ Exception Handler 5-S tage P ipeline IFU DPRAM IPIP DPP0 DPP1 DPP2 DPP3 SPSEG SP STKO V STKUN ADU D ivisio n U n it B it-M a sk-G e n . M u ltip ly U n it B a rre l-S h ifte r MDC CP RR1515 RR1414 R 15 R 14 R 15 R 14 GPRs GPRs GPRs GPRs RR1 1 RR0 0R 1 R0 R1 R0 RF PSW + /- MDH MDL ZE R O S ONES M AC B uffer ALU WB DSRAM EBC Peripherals DM U m ca04917_x.vsd Figure 2-2 CPU Block Diagram User’s Manual Architecture_X73, V2.1 2-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Summary of CPU Features • • • • • • • • • • • • • • • • • Opcode fully upward compatible with C166 Family 2-stage instruction fetch pipeline with FIFO for instruction pre-fetching 5-stage instruction execution pipeline Pipeline forwarding controls data dependencies in hardware Multiple high bandwidth buses for data and instructions Linear address space for code and data (von Neumann architecture) Nearly all instructions executed in one CPU clock cycle Fast multiplication (16-bit × 16-bit) in one CPU clock cycle Fast background execution of division (32-bit/16-bit) in 21 CPU clock cycles Built-in advanced MAC (Multiply Accumulate) Unit: – Single cycle MAC instruction with zero cycle latency including a 16 × 16 multiplier – 40-bit barrel shifter and 40-bit accumulator to handle overflows – Automatic saturation to 32 bits or rounding included with the MAC instruction – Fractional numbers supported directly – One Finite Impulse Response Filter (FIR) tap per cycle with no circular buffer management Enhanced boolean bit manipulation facilities High performance branch-, call-, and loop-processing Zero cycle jump execution Register-based design with multiple variable register banks (byte or word operands) Two additional fast register banks Variable stack with automatic stack overflow/underflow detection “Fast interrupt” and “Fast context switch” features The high performance and flexibility of the CPU is achieved by a number of optimized functional blocks (see Figure 2-2). Optimizations of the functional blocks are described in detail in the following sections. User’s Manual Architecture_X73, V2.1 2-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.1 High Instruction Bandwidth/Fast Execution Based on the hardware provisions, most of the XC167’s instructions can be executed in just one clock cycle (1/fCPU). This includes arithmetic instructions, logic instructions, and move instructions with most addressing modes. Special instructions such as SRST or PWRDN take more than one machine cycle. Divide instructions are mainly executed in the background, so other instructions can be executed in parallel. Due to the prediction mechanism (see Section 4.2), correctly predicted branch instructions require only one cycle or can even be overlaid with another instruction (zero-cycle jump). The instruction cycle time is dramatically reduced through the use of instruction pipelining. This technique allows the core CPU to process portions of multiple sequential instruction stages in parallel. Up to seven stages can operate in parallel: The two-stage instruction fetch pipeline fetches and preprocesses instructions from the respective program memory: PREFETCH: Instructions are prefetched from the PMU in the predicted order. The instructions are preprocessed in the branch detection unit to detect branches. The prediction logic determines if branches are assumed to be taken or not. FETCH: The instruction pointer for the next instruction to be fetched is calculated according to the branch prediction rules. The branch folding unit preprocesses detected branches and combines them with the preceding instructions to enable zero-cycle branch execution. Prefetched instructions are stored in the instruction FIFO, while stored instructions are moved from the instruction FIFO to the instruction processing pipeline. The five-stage instruction processing pipeline executes the respective instructions: DECODE: The previously fetched instruction is decoded and the GPR used for indirect addressing is read from the register file, if required. ADDRESS: All operand addresses are calculated. For instructions implicitly accessing the stack the stack pointer (SP) is decremented or incremented. MEMORY: All required operands are fetched. EXECUTE: The specified operation (ALU or MAC) is performed on the previously fetched operands. The condition flags are updated. Explicit write operations to CPUSFRs are executed. GPRs used for indirect addressing are incremented or decremented, if required. WRITE BACK: The result operands are written to the specified locations. Operands located in the DPRAM are stored via the write-back buffer. User’s Manual Architecture_X73, V2.1 2-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.2 Powerful Execution Units The 16-bit Arithmetic and Logic Unit (ALU) performs all standard (word) arithmetic and logical operations. Additionally, for byte operations, signals are provided from bits 6 and 7 of the ALU result to set the condition flags correctly. Multiple precision arithmetic is provided through a ‘CARRY-IN’ signal to the ALU from previously calculated portions of the desired operation. Most internal execution blocks have been optimized to perform operations on either 8-bit or 16-bit quantities. Instructions have been provided as well to allow byte packing in memory while providing sign extension of bytes for word wide arithmetic operations. The internal bus structure also allows transfers of bytes or words to or from peripherals based on the peripheral requirements. A set of consistent flags is updated automatically in the PSW after each arithmetic, logical, shift, or movement operation. These flags allow branching on specific conditions. Support for both signed and unsigned arithmetic is provided through user-specifiable branch tests. These flags are also preserved automatically by the CPU upon entry into an interrupt or trap routine. A 16-bit barrel shifter provides multiple bit shifts in a single cycle. Rotates and arithmetic shifts are also supported. The Multiply and Accumulate Unit (MAC) performs extended arithmetic operations such as 32-bit addition, 32-bit subtraction, and single-cycle 16-bit × 16-bit multiplication. The combined MAC operations (multiplication with cumulative addition/subtraction) represent the major part of the DSP performance of the CPU. The Address Data Unit (ADU) contains two independent arithmetic units to generate, calculate, and update addresses for data accesses. The ADU performs the following major tasks: • • The Standard Address Unit supports linear arithmetic for the short, long, and indirect addressing modes. It also supports data paging and stack handling. The DSP Address Generation Unit contains an additional set of address pointers and offset registers which are used in conjunction with the CoXXX instructions only. The CPU provides a lot of powerful addressing modes for word, byte, and bit data accesses (short, long, indirect). The different addressing modes use different formats and have different scopes. Dedicated bit processing instructions provide efficient control and testing of peripherals while enhancing data manipulation. These instructions provide direct access to two operands in the bit-addressable space without requiring them to be moved into temporary flags. Logical instructions allow the user to compare and modify a control bit for a peripheral in one instruction. Multiple bit shift instructions (single cycle execution) avoid long instruction streams of single bit shift operations. Bitfield instructions allow the modification of multiple bits from one operand in a single instruction. User’s Manual Architecture_X73, V2.1 2-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.3 High Performance Branch-, Call-, and Loop-Processing Pipelined execution delivers maximum performance with a stream of subsequent instructions. Any disruption requires the pipeline to be refilled and the new instruction to step through the pipeline stages. Due to the high percentage of branching in controller applications, branch instructions have been optimized to require pipeline refilling only in special cases. This is realized by detecting and preprocessing branch instructions in the prefetch stage and by predicting the respective branch target address. Prefetching then continues from the predicted target address. If the prediction was correct subsequent instructions can be fed to the execution pipeline without a gap, even if a branch is executed, i.e. the code execution is not linear. Branch target prediction (see also Section 4.2.1) uses the following rules: • • • • Unconditional branches: Branch prediction is trivial in this case, as the branches will always be taken and the target address is defined. This applies to implicitly unconditional branches such as JMPS, CALLR, or RET as well as to branches with condition code “unconditional” such as JMPI cc_UC. Fixed prediction: Branch instructions which are often used to realize loops are assumed to be taken if they branch backward to a previous location (the begin of the loop). This applies to conditional branches such as JMPR cc_XX or JNB. Variable prediction: In this case the respective prediction (taken or not taken) is coded into the instruction and can, therefore, be selected for each individual branch instruction. Thus, the software designer can optimize the instruction flow to the specific code to be executed1). This applies to the branch instructions JMPA and CALLA. Conditional indirect branches: These branches are always assumed to be not taken. This applies to branch instructions JMPI cc_XX, [Rw] and CALLI cc_XX, [Rw]. The system state information is saved automatically on the internal system stack, thus avoiding the use of instructions to preserve state upon entry and exit of interrupt or trap routines. Call instructions push the value of the IP on the system stack, and require the same execution time as branch instructions. Additionally, instructions have been provided to support indirect branch and call instructions. This feature supports implementation of multiple CASE statement branching in assembler macros and high level languages. 1) The programming tools accept either dedicated mnemonics for each prediction leaving the choice up to programmer, or they accept generic mnemonics and apply their own prediction rules. User’s Manual Architecture_X73, V2.1 2-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.4 Consistent and Optimized Instruction Formats To obtain optimum performance in a pipelined design, an instruction set has been designed which incorporates concepts from Reduced Instruction Set Computing (RISC). These concepts primarily allow fast decoding of the instructions and operands while reducing pipeline holds. These concepts, however, do not preclude the use of complex instructions required by microcontroller users. The instruction set was designed to meet the following goals: • • • Provide powerful instructions for frequently-performed operations which traditionally have required sequences of instructions. Avoid transfer into and out of temporary registers such as accumulators and carry bits. Perform tasks in parallel such as saving state upon entry into interrupt routines or subroutines. Avoid complex encoding schemes by placing operands in consistent fields for each instruction and avoid complex addressing modes which are not frequently used. Consequently, the instruction decode time decreases and the development of compilers and assemblers is simplified. Provide most frequently used instructions with one-word instruction formats. All other instructions use two-word formats. This allows all instructions to be placed on word boundaries: this alleviates the need for complex alignment hardware. It also has the benefit of increasing the range for relative branching instructions. The high performance of the CPU-hardware can be utilized efficiently by a programmer by means of the highly functional XC167 instruction set which includes the following instruction classes: • • • • • • • • • • • • • Arithmetic Instructions DSP Instructions Logical Instructions Boolean Bit Manipulation Instructions Compare and Loop Control Instructions Shift and Rotate Instructions Prioritize Instruction Data Movement Instructions System Stack Instructions Jump and Call Instructions Return Instructions System Control Instructions Miscellaneous Instructions Possible operand types are bits, bytes, words, and doublewords. Specific instructions support the conversion (extension) of bytes to words. Various direct, indirect, and immediate addressing modes are provided to specify the required operands. User’s Manual Architecture_X73, V2.1 2-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.5 Programmable Multiple Priority Interrupt System The XC167 provides 80 separate interrupt nodes that may be assigned to 16 priority levels with 8 group priorities on each level. Most interrupt sources are connected to a dedicated interrupt node. In some cases, multi-source interrupt nodes are incorporated for efficient use of system resources. These nodes can be activated by several source requests and are controlled via interrupt subnode control registers. The following enhancements within the XC167 allow processing of a large number of interrupt sources: • • • Peripheral Event Controller (PEC): This processor is used to off-load many interrupt requests from the CPU. It avoids the overhead of entering and exiting interrupt or trap routines by performing single-cycle interrupt-driven byte or word data transfers between any two locations with an optional increment of the PEC source pointer, the destination pointer, or both. Only one cycle is ‘stolen’ from the current CPU activity to perform a PEC service. Multiple Priority Interrupt Controller: This controller allows all interrupts to be assigned any specified priority. Interrupts may also be grouped, which enables the user to prevent similar priority tasks from interrupting each other. For each of the interrupt nodes, there is a separate control register which contains an interrupt request flag, an interrupt enable flag, and an interrupt priority bitfield. After being accepted by the CPU, an interrupt service can be interrupted only by a higher prioritized service request. For standard interrupt processing, each of the interrupt nodes has a dedicated vector location. Multiple Register Banks: Two local register banks for immediate context switching add to a relocatable global register bank. The user can specify several register banks located anywhere in the internal DPRAM and made of up to sixteen general purpose registers. A single instruction switches from one register bank to another (switching banks flushes the pipeline, changing the global bank requires a validation sequence). The XC167 is capable of reacting very quickly to non-deterministic events because its interrupt response time is within a very narrow range of typically 13 clock cycles (in the case of internal program execution). Its fast external interrupt inputs are sampled every clock cycle and allow even very short external signals to be recognized. The XC167 also provides an excellent mechanism to identify and process exceptions or error conditions that arise during run-time, so called ‘Hardware Traps’. A hardware trap causes an immediate non-maskable system reaction which is similar to a standard interrupt service (branching to a dedicated vector table location). The occurrence of a hardware trap is additionally signified by an individual bit in the trap flag register (TFR). Unless another, higher prioritized, trap service is in progress, a hardware trap will interrupt any current program execution. In turn, a hardware trap service can normally not be interrupted by a standard or PEC interrupt. Software interrupts are supported by means of the ‘TRAP’ instruction in combination with an individual trap (interrupt) number. User’s Manual Architecture_X73, V2.1 2-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.6 Interfaces to System Resources The CPU of the XC167 interfaces to the system resources via several bus systems which contribute to the overall performance by transferring data concurrently. This avoids stalling the CPU because instructions or operands need to be transferred. The Dual Port RAM (DPRAM) is directly coupled to the CPU because it houses the global register banks. Transfers from/to these locations affect the performance and are, therefore, carefully optimized. The Program Management Unit (PMU) controls accesses to the on-chip program memory blocks such as the ROM/Flash module and the Program/Data RAM (PSRAM) and also fetches instructions from external memory. The 64-bit interface between the PMU and the CPU delivers the instruction words, which are requested by the CPU. The PMU decides whether the requested instruction word has to be fetched from on-chip memory or from external memory. The Data Management Unit (DMU) controls accesses to the on-chip Data RAM (DSRAM), to the on-chip peripherals connected to the peripheral bus, and to resources on the external bus. External accesses (including accesses to peripherals connected to the on-chip LXBus) are executed by the External Bus Controller (EBC). The 16-bit interface between the DMU and the CPU handles all data transfers (operands). Data accesses by the CPU are distributed to the appropriate buses according to the defined address map. PMU and DMU are directly coupled to perform cross-over transfers with high speed. Crossover transfers are executed in both directions: • • PMU via DMU: Code fetches from external locations are redirected via the DMU to EBC. Thus, the XC167 can execute code from external resources. No code can be fetched from the Data RAM (DSRAM). DMU via PMU: Data accesses can also be executed to on-chip resources controlled by the PMU. This includes the following types of transfers: – Read a constant from the on-chip program ROM/Flash – Read data from the on-chip PSRAM – Write data to the on-chip PSRAM (required prior to executing out of it) – Program/Erase the on-chip Flash memory User’s Manual Architecture_X73, V2.1 2-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.2 On-Chip System Resources The XC167 controllers provide a number of powerful system resources designed around the CPU. The combination of CPU and these resources results in the high performance of the members of this controller family. Peripheral Event Controller (PEC) and Interrupt Control The Peripheral Event Controller enables response to an interrupt request with a single data transfer (word or byte) which consumes only one instruction cycle and does not require saving and restoring the machine status. Each interrupt source is prioritized for every machine cycle in the interrupt control block. If PEC service is selected, a PEC transfer is started. If CPU interrupt service is requested, the current CPU priority level stored in the PSW register is tested to determine whether a higher priority interrupt is currently being serviced. When an interrupt is acknowledged, the current state of the machine is saved on the internal system stack and the CPU branches to the system specific vector for the peripheral. The PEC contains a set of SFRs which store the count value and control bits for eight data transfer channels. In addition, the PEC uses a dedicated area of RAM which contains the source and destination addresses. The PEC is controlled in a manner similar to any other peripheral: through SFRs containing the desired configuration of each channel. An individual PEC transfer counter is implicitly decremented for each PEC service except in the continuous transfer mode. When this counter reaches zero, a standard interrupt is performed to the vector location related to the corresponding source. PEC services are very well suited, for example, to moving register contents to/from a memory table. The XC167 has eight PEC channels, each of which offers such fast interruptdriven data transfer capabilities. Memory Areas The memory space of the XC167 is configured in a Von Neumann architecture. This means that code memory, data memory, registers, and IO ports are organized within the same linear address space which covers up to 16 Mbytes. The entire memory space can be accessed bytewise or wordwise. Particular portions of the on-chip memory have been made directly bit addressable as well. 256 Kbytes of on-chip Flash memory store code or constant data. The on-chip Flash memory is organized as four 8-Kbyte sectors, one 32-Kbyte sector, and three 64-Kbyte sectors. Each sector can be separately write protected1), erased and programmed (in blocks of 128 bytes). The complete Flash area can be read-protected. A password sequence temporarily unlocks protected areas. The Flash module combines very fast 1) Each two 8-Kbyte sectors are combined for write-protection purposes. User’s Manual Architecture_X73, V2.1 2-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 64-bit one-cycle1) read accesses with protected and efficient writing algorithms for programming and erasing. Dynamic error correction provides extremely high read data security for all read accesses. Programming typically takes 2 ms per 128-byte block (5 ms max.), erasing a sector typically takes 200 ms (500 ms max.). Note: Program execution from on-chip program memory is the fastest of all possible alternatives and results in maximum performance. The type of the on-chip program memory depends on the chosen derivative. On-chip program memory also includes the PSRAM. 6 Kbytes of on-chip Program SRAM (PSRAM) are provided to store user code or data. The PSRAM is accessed via the PMU and is therefore optimized for code fetches. 4 Kbytes of on-chip Data SRAM (DSRAM) are provided as a storage for general user data.The DSRAM is accessed via the DMU and is therefore optimized for data accesses. 2 Kbytes of on-chip Dual-Port RAM (DPRAM) are provided as a storage for user defined variables, for the system stack, and in particular for general purpose register banks. A register bank can consist of up to 16 wordwide (R0 to R15) and/or bytewide (RL0, RH0, …, RL7, RH7) so-called General Purpose Registers (GPRs). The upper 256 bytes of the DPRAM are directly bitaddressable. When used by a GPR, any location in the DPRAM is bitaddressable. The CPU has an actual register context of up to 16 wordwide and/or bytewide global GPRs at its disposal, which are physically located within the on-chip RAM area. A Context Pointer (CP) register determines the base address of the active global register bank to be accessed by the CPU at a time. The number of register banks is restricted only by the available internal RAM space. For easy parameter passing, a register bank may overlap other register banks. A system stack of up to 32 Kwords is provided as storage for temporary data. The system stack can be located anywhere within the complete addressing range and it is accessed by the CPU via the Stack Pointer (SP) register and the Stack Pointer Segment (SPSEG) register. Two separate SFRs, STKOV and STKUN, are implicitly compared against the stack pointer value upon each stack access for the detection of a stack overflow or underflow. This mechanism also supports the control of a bigger virtual stack. Maximum performance for stack operations is achieved by allocating the system stack to internal data RAM areas (DPRAM, DSRAM). Hardware detection of the selected memory space is placed at the internal memory decoders and allows the user to specify any address directly or indirectly and obtain the desired data without using temporary registers or special instructions. 1) Flash accesses may require waitstates, depending on the actual operating frequency. For the exact Flash memory access timing and the required waitstates please refer to Section 3.10.1. User’s Manual Architecture_X73, V2.1 2-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview For Special Function Registers three areas of the address space are reserved: The standard Special Function Register area (SFR) uses 512 bytes, while the Extended Special Function Register area (ESFR) uses the other 512 bytes. A range of 4 Kbytes is provided for the internal IO area (XSFR). SFRs are wordwide registers which are used for controlling and monitoring functions of the different on-chip units. Unused SFR addresses are reserved for future members of the XC166 Family with enhanced functionality. Therefore, they should either not be accessed, or written with zeros, to ensure upward compatibility. In order to meet the needs of designs where more memory is required than is provided on chip, up to 12 Mbytes (approximately, see Table 2-1) of external RAM and/or ROM can be connected to the microcontroller. The External Bus Interface also provides access to external peripherals. Table 2-1 XC167 Memory Map1) Address Area Start Loc. End Loc. Area Size2) Notes Flash register space FF’F000H FF’FFFFH 4 Kbytes 3) Reserved (Acc. trap) FE’0000H FF’EFFFH 60 Kbytes Minus Flash regs Reserved for EPSRAM F8’1800H FD’FFFFH 378 Kbytes – Emul. Program SRAM4) F8’0000H F8’17FFH 6 Kbytes 2nd way to PSRAM Reserved for PSRAM E0’1800H F7’FFFFH < 1.5 Mbytes Minus PSRAM Program SRAM E0’0000H E0’17FFH 6 Kbytes Maximum Reserved for pr. mem. C4’0000H DF’FFFFH < 2 Mbytes Minus Flash Program Flash C0’0000H C3’FFFFH 256 Kbytes – Reserved BF’0000H BF’FFFFH 64 Kbytes – External memory area 40’0000H BE’FFFFH < 8 Mbytes Minus res. seg. External IO area5) 20’0800H 3F’FFFFH < 2 Mbytes Minus TwinCAN TwinCAN registers 20’0000H 20’07FFH 2 Kbytes – External memory area 01’0000H 1F’FFFFH < 2 Mbytes Minus segment 0 Data RAMs and SFRs 00’8000H 00’FFFFH 32 Kbytes Partly used External memory area 00’0000H 00’7FFFH 32 Kbytes – 1) Accesses to the shaded areas generate external bus accesses. 2) The areas marked with “<” are slightly smaller than indicated, see column “Notes”. 3) Not defined register locations return a trap code. 4) The Emulation PSRAM (EPSRAM) realizes a 2nd access path to the PSRAM with a different timing. 5) Several pipeline optimizations are not active within the external IO area. This is necessary to control external peripherals properly. User’s Manual Architecture_X73, V2.1 2-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview External Bus Interface To meet the needs of designs where more memory is required than is provided on chip, up to 12 Mbytes of external RAM and/or ROM can be connected to the XC167 microcontroller via its external bus interface. All of the external memory accesses are performed by a particular on-chip External Bus Controller (EBC). It can be programmed either to Single Chip Mode when no external memory is required, or to one of four different external memory access modes1), which are as follows: • • • • 16 … 24-bit Addresses, 16-bit Data, Demultiplexed 16 … 24-bit Addresses, 16-bit Data, Multiplexed 16 … 24-bit Addresses, 8-bit Data, Multiplexed 16 … 24-bit Addresses, 8-bit Data, Demultiplexed In the demultiplexed bus modes, addresses are output on PORT1 and data is input/output on PORT0 or P0L, respectively. In the multiplexed bus modes both addresses and data use PORT0 for input/output. The high order address (segment) lines use Port 4. The number of active segment address lines is selectable, restricting the external address space to 8 Mbytes … 64 Kbytes. This is required when interface lines are assigned to Port 4. For up to five address areas the bus mode (multiplexed/demultiplexed), the data bus width (8-bit/16-bit) and even the length of a bus cycle (waitstates, signal delays) can be selected independently. This allows access to a variety of memory and peripheral components directly and with maximum efficiency. Access to very slow memories or modules with varying access times is supported via a particular ‘Ready’ function. The active level of the control input signal is selectable. A HOLD/HLDA protocol is available for bus arbitration and allows the sharing of external resources with other bus masters. The external bus timing is related to the rising edge of the reference clock output CLKOUT. The external bus protocol is compatible with that of the standard C166 Family. For applications which require less than 64 Kbytes of address space, a non-segmented memory model can be selected, where all locations can be addressed by 16 bits. Thus, Port 4 is not needed as an output for the upper address bits (Axx … A16), as is the case when using the segmented memory model. The EBC also controls accesses to resources connected to the on-chip LXBus. The LXBus is an internal representation of the external bus and allows accessing integrated peripherals and modules in the same way as external components. The TwinCAN module is connected and accessed via the LXBus. 1) Bus modes are switched dynamically if several address windows with different mode settings are used. User’s Manual Architecture_X73, V2.1 2-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.3 On-Chip Peripheral Blocks The XC166 Family clearly separates peripherals from the core. This structure permits the maximum number of operations to be performed in parallel and allows peripherals to be added or deleted from family members without modifications to the core. Each functional block processes data independently and communicates information over common buses. Peripherals are controlled by data written to the respective Special Function Registers (SFRs). These SFRs are located within either the standard SFR area (00’FE00H … 00’FFFFH), the extended ESFR area (00’F000H … 00’F1FFH), or within the internal IO area (00’E000H … 00’EFFFH). These built-in peripherals either allow the CPU to interface with the external world or provide functions on-chip that otherwise would need to be added externally in the respective system. The XC167 generic peripherals are: • • • • • • • • • • Two General Purpose Timer Blocks (GPT1 and GPT2) Two Asynchronous/Synchronous Serial Interfaces (ASC0 and ASC1) Two High Speed Serial Interfaces (SSC0 and SSC1) IIC Bus Module A Watchdog Timer Two Capture/Compare units (CAPCOM1 and CAPCOM2) Enhanced Capture/Compare unit (CAPCOM6) A 10-bit Analog/Digital Converter (ADC) A Real Time Clock (RTC) Ten I/O ports with a total of 103 I/O lines Because the LXBus is the internal representation of the external bus, it does not support bit-addressing. Accesses are executed by the EBC as if it were external accesses. The LXBus connects on-chip peripherals to the CPU: • TwinCAN module with 2 CAN nodes and gateway functionality Each peripheral also contains a set of Special Function Registers (SFRs) which control the functionality of the peripheral and temporarily store intermediate data results. Each peripheral has an associated set of status flags. Individually selected clock signals are generated for each peripheral from binary multiples of the master clock. User’s Manual Architecture_X73, V2.1 2-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Peripheral Interfaces The on-chip peripherals generally have two different types of interfaces: an interface to the CPU and an interface to external hardware. Communication between the CPU and peripherals is performed through Special Function Registers (SFRs) and interrupts. The SFRs serve as control/status and data registers for the peripherals. Interrupt requests are generated by the peripherals based on specific events which occur during their operation, such as operation complete, error, etc. To interface with external hardware, specific pins of the parallel ports are used, when an input or output function has been selected for a peripheral. During this time, the port pins are controlled either by the peripheral (when used as outputs) or by the external hardware which controls the peripheral (when used as inputs). This is called the ‘alternate (input or output) function’ of a port pin, in contrast to its function as a general purpose I/O pin. Peripheral Timing Internal operation of the CPU and peripherals is based on the master clock (fMC). The clock generation unit uses the on-chip oscillator to derive the master clock from the crystal or from the external clock signal. The clock signal gated to the peripherals is independent from the clock signal that feeds the CPU. During Idle mode, the CPU’s clock is stopped while the peripherals continue their operation. Peripheral SFRs may be accessed by the CPU once per state. When an SFR is written to by software in the same state where it is also to be modified by the peripheral, the software write operation has priority. Further details on peripheral timing are included in the specific sections describing each peripheral. Programming Hints • Access to SFRs: All SFRs reside in data page 3 of the memory space. The following addressing mechanisms allow access to the SFRs: – Indirect or direct addressing with 16-bit (mem) addresses must guarantee that the used data page pointer (DPP0 … DPP3) selects data page 3. – Accesses via the Peripheral Event Controller (PEC) use the SRCPx and DSTPx pointers instead of the data page pointers. – Short 8-bit (reg) addresses to the standard SFR area do not use the data page pointers but directly access the registers within this 512-byte area. – Short 8-bit (reg) addresses to the extended ESFR area require switching to the 512-byte Extended SFR area. This is done via the EXTension instructions EXTR, EXTP(R), EXTS(R). User’s Manual Architecture_X73, V2.1 2-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview • • Byte Write Operations to wordwide SFRs via indirect or direct 16-bit (mem) addressing or byte transfers via the PEC force zeros in the non-addressed byte. Byte write operations via short 8-bit (reg) addressing can access only the low byte of an SFR and force zeros in the high byte. It is therefore recommended, to use the bitfield instructions (BFLDL and BFLDH) to write to any number of bits in either byte of an SFR without disturbing the non-addressed byte and the unselected bits. Reserved Bits: Some of the bits which are contained in the XC167’s SFRs are marked as ‘Reserved’. User software should never write ‘1’s to reserved bits. These bits are currently not implemented and may be used in future products to invoke new functions. In that case, the active state for those new functions will be ‘1’, and the inactive state will be ‘0’. Therefore writing only ‘0’s to reserved locations allows portability of the current software to future devices. After read accesses, reserved bits should be ignored or masked out. Capture/Compare Units (CAPCOM1/2) The CAPCOM units support generation and control of timing sequences on up to 32 channels with a maximum resolution of 1 system clock cycle (8 cycles in staggered mode). The CAPCOM units are typically used to handle high speed I/O tasks such as pulse and waveform generation, pulse width modulation (PMW), Digital to Analog (D/A) conversion, software timing, or time recording relative to external events. Four 16-bit timers (T0/T1, T7/T8) with reload registers provide two independent time bases for each capture/compare register array. The input clock for the timers is programmable to several prescaled values of the internal system clock, or may be derived from an overflow/underflow of timer T6 in module GPT2. This provides a wide range of variation for the timer period and resolution and allows precise adjustments to the application specific requirements. In addition, external count inputs for CAPCOM timers T0 and T7 allow event scheduling for the capture/compare registers relative to external events. Both of the two capture/compare register arrays contain 16 dual purpose capture/compare registers, each of which may be individually allocated to either CAPCOM timer T0 or T1 (T7 or T8, respectively), and programmed for capture or compare function. All registers of each module have each one port pin associated with it which serves as an input pin for triggering the capture function, or as an output pin to indicate the occurrence of a compare event. User’s Manual Architecture_X73, V2.1 2-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-2 Compare Modes (CAPCOM1/2) Compare Modes Function Mode 0 Interrupt-only compare mode; several compare interrupts per timer period are possible Mode 1 Pin toggles on each compare match; several compare events per timer period are possible Mode 2 Interrupt-only compare mode; only one compare interrupt per timer period is generated Mode 3 Pin set ‘1’ on match; pin reset ‘0’ on compare timer overflow; only one compare event per timer period is generated Double Register Mode Two registers operate on one pin; pin toggles on each compare match; several compare events per timer period are possible Single Event Mode Generates single edges or pulses; can be used with any compare mode When a capture/compare register has been selected for capture mode, the current contents of the allocated timer will be latched (‘captured’) into the capture/compare register in response to an external event at the port pin which is associated with this register. In addition, a specific interrupt request for this capture/compare register is generated. Either a positive, a negative, or both a positive and a negative external signal transition at the pin can be selected as the triggering event. The contents of all registers which have been selected for one of the five compare modes are continuously compared with the contents of the allocated timers. When a match occurs between the timer value and the value in a capture/compare register, specific actions will be taken based on the selected compare mode. User’s Manual Architecture_X73, V2.1 2-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Capture/Compare Unit CAPCOM6 The CAPCOM6 unit supports generation and control of timing sequences on up to three 16-bit capture/compare channels plus one independent 10-bit compare channel. In compare mode the CAPCOM6 unit provides two output signals per channel which have inverted polarity and non-overlapping pulse transitions (deadtime control). The compare channel can generate a single PWM output signal and is further used to modulate the capture/compare output signals. In capture mode the contents of compare timer T12 is stored in the capture registers upon a signal transition at pins CCx. The output signals can be generated in edge-aligned or center-aligned PWM mode. They are generated continuously or in single-shot mode. Compare timers T12 (16-bit) and T13 (10-bit) are free running timers which are clocked by the prescaled system clock. For motor control applications (brushless DC-drives) both subunits may generate versatile multichannel PWM signals which are basically either controlled by compare timer T12 or by a typical hall sensor pattern at the interrupt inputs (block commutation). The latter mode provides noise filtering for the hall inputs and supports automatic rotational speed measurement. The trap function offers a fast emergency stop without CPU activity. Triggered by an external signal (CTRAP) the outputs are switched to selectable logic levels which can be adapted to the connected power stages. User’s Manual Architecture_X73, V2.1 2-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview General Purpose Timer (GPT12E) Unit The GPT12E unit represents a very flexible multifunctional timer/counter structure which may be used for many different time related tasks such as event timing and counting, pulse width and duty cycle measurements, pulse generation, or pulse multiplication. The GPT12E unit incorporates five 16-bit timers which are organized in two separate blocks, GPT1 and GPT2. Each timer in each block may operate independently in a number of different modes, or may be concatenated with another timer of the same block. Each of the three timers T2, T3, T4 of block GPT1 can be configured individually for one of four basic modes of operation, which are Timer, Gated Timer, Counter, and Incremental Interface Mode. In Timer Mode, the input clock for a timer is derived from the system clock, divided by a programmable prescaler, while Counter Mode allows a timer to be clocked in reference to external events. Pulse width or duty cycle measurement is supported in Gated Timer Mode, where the operation of a timer is controlled by the ‘gate’ level on an external input pin. For these purposes, each timer has one associated port pin (TxIN) which serves as gate or clock input. The maximum resolution of the timers in block GPT1 is 4 system clock cycles. The count direction (up/down) for each timer is programmable by software or may additionally be altered dynamically by an external signal on a port pin (TxEUD) to facilitate e.g. position tracking. In Incremental Interface Mode the GPT1 timers (T2, T3, T4) can be directly connected to the incremental position sensor signals A and B via their respective inputs TxIN and TxEUD. Direction and count signals are internally derived from these two input signals, so the contents of the respective timer Tx corresponds to the sensor position. The third position sensor signal TOP0 can be connected to an interrupt input. Timer T3 has an output toggle latch (T3OTL) which changes its state on each timer overflow/underflow. The state of this latch may be output on pin T3OUT e.g. for time out monitoring of external hardware components. It may also be used internally to clock timers T2 and T4 for measuring long time periods with high resolution. In addition to their basic operating modes, timers T2 and T4 may be configured as reload or capture registers for timer T3. When used as capture or reload registers, timers T2 and T4 are stopped. The contents of timer T3 is captured into T2 or T4 in response to a signal at their associated input pins (TxIN). Timer T3 is reloaded with the contents of T2 or T4 triggered either by an external signal or by a selectable state transition of its toggle latch T3OTL. When both T2 and T4 are configured to alternately reload T3 on opposite state transitions of T3OTL with the low and high times of a PWM signal, this signal can be constantly generated without software intervention. With its maximum resolution of 2 system clock cycles, the GPT2 block provides precise event control and time measurement. It includes two timers (T5, T6) and a capture/reload register (CAPREL). Both timers can be clocked with an input clock which User’s Manual Architecture_X73, V2.1 2-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview is derived from the CPU clock via a programmable prescaler or with external signals. The count direction (up/down) for each timer is programmable by software or may additionally be altered dynamically by an external signal on a port pin (TxEUD). Concatenation of the timers is supported via the output toggle latch (T6OTL) of timer T6, which changes its state on each timer overflow/underflow. The state of this latch may be used to clock timer T5, and/or it may be output on pin T6OUT. The overflows/underflows of timer T6 can additionally be used to clock the CAPCOM1/2 timers, and to cause a reload from the CAPREL register. The CAPREL register may capture the contents of timer T5 based on an external signal transition on the corresponding port pin (CAPIN), and timer T5 may optionally be cleared after the capture procedure. This allows the XC167 to measure absolute time differences or to perform pulse multiplication without software overhead. The capture trigger (timer T5 to CAPREL) may also be generated upon transitions of GPT1 timer T3’s inputs T3IN and/or T3EUD. This is especially advantageous when T3 operates in Incremental Interface Mode. User’s Manual Architecture_X73, V2.1 2-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Real Time Clock The Real Time Clock (RTC) module of the XC167 is directly clocked via a separate clock driver either with the on-chip auxiliary oscillator frequency (fRTC = fOSCa) or with the prescaled on-chip main oscillator frequency (fRTC = fOSCm/32). It is therefore independent from the selected clock generation mode of the XC167. The RTC basically consists of a chain of divider blocks: • • • a selectable 8:1 divider (on - off) the reloadable 16-bit timer T14 the 32-bit RTC timer block (accessible via registers RTCH and RTCL), made of: – a reloadable 10-bit timer – a reloadable 6-bit timer – a reloadable 6-bit timer – a reloadable 10-bit timer All timers count up. Each timer can generate an interrupt request. All requests are combined to a common node request. Additionally, T14 can generate a separate node request. Note: The registers associated with the RTC are not affected by a reset in order to maintain the correct system time even when intermediate resets are executed. The RTC module can be used for different purposes: • • • • System clock to determine the current time and date, optionally during idle mode, sleep mode, and power down mode Cyclic time based interrupt, to provide a system time tick independent of CPU frequency and other resources, e.g. to wake-up regularly from idle mode 48-bit timer for long term measurements (maximum timespan is > 100 years) Alarm interrupt for wake-up on a defined time User’s Manual Architecture_X73, V2.1 2-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview A/D Converter For analog signal measurement, a 10-bit A/D converter with 16 multiplexed input channels and a sample and hold circuit has been integrated on-chip. It uses the method of successive approximation. The sample time (for loading the capacitors) and the conversion time is programmable (in two modes) and can thus be adjusted to the external circuitry. The A/D converter can also operate in 8-bit conversion mode, where the conversion time is further reduced. Overrun error detection/protection is provided for the conversion result register (ADDAT): either an interrupt request will be generated when the result of a previous conversion has not been read from the result register at the time the next conversion is complete, or the next conversion is suspended in such a case until the previous result has been read. For applications which require less analog input channels, the remaining channel inputs can be used as digital input port pins. The A/D converter of the XC167 supports four different conversion modes. In the standard Single Channel conversion mode, the analog level on a specified channel is sampled once and converted to a digital result. In the Single Channel Continuous mode, the analog level on a specified channel is repeatedly sampled and converted without software intervention. In the Auto Scan mode, the analog levels on a prespecified number of channels are sequentially sampled and converted. In the Auto Scan Continuous mode, the prespecified channels are repeatedly sampled and converted. In addition, the conversion of a specific channel can be inserted (injected) into a running sequence without disturbing this sequence. This is called Channel Injection Mode. The Peripheral Event Controller (PEC) may be used to automatically store the conversion results into a table in memory for later evaluation, without requiring the overhead of entering and exiting interrupt routines for each data transfer. After each reset and also during normal operation the ADC automatically performs calibration cycles. This automatic self-calibration constantly adjusts the converter to changing operating conditions (e.g. temperature) and compensates process variations. These calibration cycles are part of the conversion cycle, so they do not affect the normal operation of the A/D converter. The calibration cycles after a conversion can be disabled, so the overall conversion time is reduced again. In order to decouple analog inputs from digital noise and to avoid input trigger noise those pins used for analog input can be disconnected from the digital IO or input stages under software control. This can be selected for each pin separately via register P5DIDIS (Port 5 Digital Input Disable). The Auto-Power-Down feature of the A/D converter minimizes the power consumption when no conversion is in progress. User’s Manual Architecture_X73, V2.1 2-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Asynchronous/Synchronous Serial Interfaces (ASC0/ASC1) The Asynchronous/Synchronous Serial Interfaces ASC0/ASC1 (USARTs) provide serial communication with other microcontrollers, processors, terminals or external peripheral components. They are upward compatible with the serial ports of the Infineon 8-bit microcontroller families and support full-duplex asynchronous communication and halfduplex synchronous communication. A dedicated baud rate generator with a fractional divider precisely generates all standard baud rates without oscillator tuning. In asynchronous mode, 8- or 9-bit data frames (with optional parity bit) are transmitted or received, preceded by a start bit and terminated by one or two stop bits. For multiprocessor communication, a mechanism to distinguish address from data bytes has been included (8-bit data plus wake-up bit mode). IrDA data transmissions up to 115.2 kbit/s with fixed or programmable IrDA pulse width are supported. An autobaud detection unit allows to detect asynchronous data frames with its baudrate and mode with automatic initialization of the baudrate generator and the mode control bits. In synchronous mode, bytes (8 bits) are transmitted or received synchronously to a shift clock which is generated by the ASC0/1. The LSB is always shifted first. In both modes, transmission and reception of data is FIFO-buffered (8 entries per direction). A loop-back option is available for testing purposes. Five separate interrupt vectors are provided for transmit buffer, transmission, reception, autobaud detection, and error handling. A number of optional hardware error detection capabilities has been included to increase the reliability of data transfers. A parity bit can automatically be generated on transmission or be checked on reception. Framing error detection allows to recognize data frames with missing stop bits. An overrun error will be generated, if the last character received has not been read out of the receive buffer register at the time the reception of a new character is complete. Summary of Features • • • • • Full-duplex asynchronous operating modes – 8- or 9-bit data frames, LSB first, one or two stop bits, parity generation/checking – Baudrate from 2.5 Mbit/s to 0.6 bit/s (@ 40 MHz) – Multiprocessor mode for automatic address/data byte detection – Support for IrDA data transmission/reception up to max. 115.2 kbit/s (@ 40 MHz) – Loop-back capability – Auto baudrate detection Half-duplex 8-bit synchronous operating mode at 5 Mbit/s to 406.9 bit/s (@ 40 MHz) Buffered transmitter/receiver with FIFO support (8 entries per direction) Loop-back option available for testing purposes Interrupt generation on transmitter buffer empty condition, last bit transmitted condition, receive buffer full condition, error condition (frame, parity, overrun error), start and end of an autobaud detection User’s Manual Architecture_X73, V2.1 2-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview High Speed Synchronous Serial Channels (SSC0/SSC1) The High Speed Synchronous Serial Channels SSC0/SSC1 support full-duplex and halfduplex synchronous communication. They may be configured so they interface with serially linked peripheral components, full SPI functionality is supported. A dedicated baud rate generator allows to set up all standard baud rates without oscillator tuning. The SSC transmits or receives characters of 2 … 16 bits length synchronously to a shift clock which can be generated by the SSC (master mode) or by an external master (slave mode). The SSC can start shifting with the LSB or with the MSB and allows the selection of shifting and latching clock edges as well as the clock polarity. A loop-back option is available for testing purposes. Three separate interrupt vectors are provided for transmission, reception, and error handling. A number of optional hardware error detection capabilities has been included to increase the reliability of data transfers. Transmit error and receive error supervise the correct handling of the data buffer. Phase error and baudrate error detect incorrect serial data. Summary of Features • • • • • • • Master or Slave mode operation Full-duplex or Half-duplex transfers Baudrate generation from 20 Mbit/s to 305.18 bit/s (@ 40 MHz) Flexible data format – Programmable number of data bits: 2 to 16 bits – Programmable shift direction: LSB-first or MSB-first – Programmable clock polarity: idle low or idle high – Programmable clock/data phase: data shift with leading or trailing clock edge Loop back option available for testing purposes Interrupt generation on transmitter buffer empty condition, receive buffer full condition, error condition (receive, phase, baudrate, transmit error) Three pin interface with flexible SSC pin configuration User’s Manual Architecture_X73, V2.1 2-24 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview On-Chip TwinCAN Module The integrated TwinCAN module handles the completely autonomous transmission and reception of CAN frames in accordance with the CAN specification V2.0 part B (active), i.e. the on-chip TwinCAN module can receive and transmit standard frames with 11-bit identifiers as well as extended frames with 29-bit identifiers. Two Full-CAN nodes share the TwinCAN module’s resources to optimize the CAN bus traffic handling and to minimize the CPU load. The module provides up to 32 message objects, which can be assigned to one of the CAN nodes and can be combined to FIFOstructures. Each object provides separate masks for acceptance filtering. The flexible combination of Full-CAN functionality and FIFO architecture reduces the efforts to fulfill the real-time requirements of complex embedded control applications. Improved CAN bus monitoring functionality as well as the number of message objects permit precise and comfortable CAN bus traffic handling. Gateway functionality allows automatic data exchange between two separate CAN bus systems, which reduces CPU load and improves the real time behavior of the entire system. The bit timing for both CAN nodes is derived from the master clock and is programmable up to a data rate of 1 Mbit/s. Each CAN node uses two pins (configurable) to interface to an external bus transceiver. The interface pins are assigned via software. Summary of Features • • • • • • • CAN functionality according to CAN specification V2.0 B active Data transfer rate up to 1 Mbit/s Flexible and powerful message transfer control and error handling capabilities Full-CAN functionality and Basic CAN functionality for each message object 32 flexible message objects – Assignment to one of the two CAN nodes – Configuration as transmit object or receive object – Concatenation to a 2-, 4-, 8-, 16-, or 32-message buffer with FIFO algorithm – Handling of frames with 11-bit or 29-bit identifiers – Individual programmable acceptance mask register for filtering for each object – Monitoring via a frame counter – Configuration for Remote Monitoring Mode Up to eight individually programmable interrupt nodes can be used CAN Analyzer Mode for bus monitoring is implemented User’s Manual Architecture_X73, V2.1 2-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview On-Chip IIC Bus Module The integrated IIC Bus Module handles the transmission and reception of frames over the two-line IIC bus in accordance with the IIC Bus specification. The IIC Module can operate in slave mode, in master mode or in multi-master mode. It can receive and transmit data using 7-bit or 10-bit addressing. Up to 4 send/receive data bytes can be stored in the extended buffers. Up to three physical interfaces (port pin pairs) can be established dynamically under software control. The respective pins must be configured for open drain mode to enable IIC bus communication. Data can be transferred at the standard speed of 100 kbit/s or up to 400 kbit/s. Two interrupt nodes dedicated to the IIC module allow efficient interrupt service and also support operation via PEC transfers. Watchdog Timer The Watchdog Timer represents one of the fail-safe mechanisms which have been implemented to prevent the controller from malfunctioning for longer periods of time. The Watchdog Timer is always enabled after a reset of the chip, and can be disabled until the EINIT instruction has been executed (compatible mode), or it can be disabled and enabled at any time by executing instructions DISWDT and ENWDT (enhanced mode). Thus, the chip’s start-up procedure is always monitored. The software has to be designed to restart the Watchdog Timer before it overflows. If, due to hardware or software related failures, the software fails to do so, the Watchdog Timer overflows and generates an internal hardware reset and pulls the RSTOUT pin low in order to allow external hardware components to be reset. The Watchdog Timer is a 16-bit timer, clocked with the system clock divided by 2/4/128/256. The high byte of the Watchdog Timer register can be set to a prespecified reload value (stored in WDTREL) to allow further variation of the monitored time interval. Each time it is serviced by the application software, the high byte of the Watchdog Timer is reloaded and the low byte is cleared. Thus, time intervals between 13 µs and 419 ms can be monitored (@ 40 MHz). The default Watchdog Timer interval after reset is 3.28 ms (@ 40 MHz). User’s Manual Architecture_X73, V2.1 2-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Parallel Ports The XC167 provides up to 103 I/O lines which are organized into nine input/output ports and one input port. All port lines are bit-addressable, and all input/output lines are individually (bit-wise) programmable as inputs or outputs via direction registers. The I/O ports are true bidirectional ports which are switched to high impedance state when configured as inputs. The output drivers of some I/O ports can be configured (pin by pin) for push/pull operation or open-drain operation via control registers. During the internal reset, all port pins are configured as inputs (except for pin RSTOUT). The edge characteristics (shape) and driver characteristics (output current) of the port drivers can be selected via registers POCONx. The input threshold of some ports is selectable (TTL or CMOS like), where the special CMOS like input threshold reduces noise sensitivity due to the input hysteresis. The input threshold may be selected individually for each byte of the respective ports. All port lines have programmable alternate input or output functions associated with them. All port lines that are not used for these alternate functions may be used as general purpose IO lines. Table 2-3 Summary of the XC167’s Parallel Ports Port Control Alternate Functions PORT0 Pad drivers Address/Data lines or data lines1) PORT1 Pad drivers Address lines2) Capture inputs or compare outputs, Serial interface lines Port 2 Pad drivers, Open drain, Input threshold Capture inputs or compare outputs, Timer control signal, Fast external interrupt inputs Port 3 Pad drivers, Open drain, Input threshold Timer control signals, serial interface lines, Optional bus control signal BHE/WRH, System clock output CLKOUT (or FOUT) Port 4 Pad drivers, Open drain, Input threshold Segment address lines3) CAN interface lines4) Port 5 Input stage disable Analog input channels to the A/D converter, Timer control signals Port 6 Pad drivers, Open drain, Input threshold Capture inputs or compare outputs, Bus arbitration signals BREQ, HLDA, HOLD, Optional chip select signals User’s Manual Architecture_X73, V2.1 2-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-3 Summary of the XC167’s Parallel Ports (cont’d) Port Control Alternate Functions Port 7 Pad drivers, Open drain, Input threshold Capture inputs or compare outputs, CAN interface lines4) Port 9 Pad drivers, Open drain, Input threshold Capture inputs or compare outputs Port 20 Pad drivers, Input threshold CAN interface lines4), IIC bus interface lines4) Bus control signals RD, WR/WRL, READY, ALE, External access enable pin EA, Reset indication output RSTOUT 1) For multiplexed bus cycles. 2) For demultiplexed bus cycles. 3) For more than 64 Kbytes of external resources. 4) Can be assigned by software. User’s Manual Architecture_X73, V2.1 2-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.4 Clock Generation The Clock Generation Unit uses a programmable on-chip PLL with multiple prescalers to generate the clock signals for the XC167 with high flexibility. The master clock fMC is the reference clock signal, and is used for TwinCAN and is output to the external system. The CPU clock fCPU and the system clock fSYS are derived from the master clock either directly (1:1) or via a 2:1 prescaler (fSYS = fCPU = fMC/2). The on-chip oscillator can drive an external crystal or accepts an external clock signal. The oscillator clock frequency can be multiplied by the on-chip PLL (by a programmable factor) or can be divided by a programmable prescaler factor. If the bypass mode is used (direct drive or prescaler) the PLL can deliver an independent clock to monitor the clock signal generated by the on-chip oscillator. This PLL clock is independent from the XTAL1 clock. When the expected oscillator clock transitions are missing the Oscillator Watchdog (OWD) activates the PLL Unlock/OWD interrupt node and supplies the CPU with an emergency clock, the PLL clock signal. Under these circumstances the PLL will oscillate with its basic frequency. The oscillator watchdog can be disabled by switching the PLL off. This reduces power consumption, but also no interrupt request will be generated in case of a missing oscillator clock. 2.5 Power Management Features The basic power reduction modes (Idle and Power Down) are enhanced by additional power management features (see below). These features can be combined to reduce the controller’s power consumption to correspond to the application’s possible minimum. • • • • Basic power saving modes Flexible clock generation Flexible peripheral management (peripherals can be disabled and enabled) Periodic wake-up from Idle mode via RTC timer The listed features provide effective means to realize standby conditions for the system with an optimum balance between power reduction (standby time) and peripheral operation (system functionality). Basic Power Saving Modes The XC167 can be switched into special operating modes (control via instructions) where its power consumption (and functionality) is reduced. Idle Mode stops the CPU while the peripherals can continue to operate. Sleep Mode and Power Down Mode stop all clock signals and all operation (RTC may optionally continue running). Sleep Mode can be terminated by external interrupt signals. User’s Manual Architecture_X73, V2.1 2-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Flexible Clock Generation The flexible clock generation system combines a variety of improved mechanisms (partly user controllable) to provide the XC167 modules with clock signals. This is especially important in power sensitive modes such as standby operation. The power optimized oscillator generally reduces the amount of power which is consumed in order to generate the clock signal within the XC167. The clock system controls the distribution and the frequency of internal and external clock signals. The user can reduce the XC167’s CPU clock frequency which drastically reduces the consumed power. External circuitry can be controlled via the programmable frequency output FOUT. Flexible Peripheral Management Flexible peripheral management provides a mechanism to enable and disable each peripheral module separately. In each situation (such as several system operating modes, standby, etc.) only those peripherals may be kept running which are required for the specified functionality, for example, to maintain communication channels. All others may be switched off. The registers of disabled peripherals can still be accessed. Periodic Wake-up from Idle or Sleep Mode Periodic wake-up from Idle mode or from Sleep mode combines the drastically reduced power consumption in Idle/Sleep mode (in conjunction with the additional power management features) with a high level of system availability. External signals and events can be scanned (at a lower rate) by periodically activating the CPU and selected peripherals which then return to powersave mode after a short time. This greatly reduces the system’s average power consumption. Idle/Sleep mode can also be terminated by external interrupt signals. User’s Manual Architecture_X73, V2.1 2-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.6 On-Chip Debug Support (OCDS) The On-Chip Debug Support system provides a broad range of debug and emulation features built into the XC167. The user software running on the XC167 can thus be debugged within the target system environment. The OCDS is controlled by an external debugging device via the debug interface, consisting of the IEEE-1149-conforming JTAG port and a break interface. The debugger controls the OCDS via a set of dedicated registers accessible via the JTAG interface. Additionally, the OCDS system can be controlled by the CPU, e.g. by a monitor program. An injection interface allows the execution of OCDS-generated instructions by the CPU. Multiple breakpoints can be triggered by on-chip hardware, by software, or by an external trigger input. Single stepping is supported as well as the injection of arbitrary instructions and read/write access to the complete internal address space. A breakpoint trigger can be answered with a CPU-halt, a monitor call, a data transfer, or/and the activation of an external signal. The data transferred at a watchpoint (see above) can be obtained via the JTAG interface or via the external bus interface for increased performance. The debug interface uses a set of 6 interface signals (4 JTAG lines, 2 break lines) to communicate with external circuitry. These interface signals use dedicated pins. Complete system emulation is supported by an emulation device. Via this full-featured emulation interface (including internal buses, control, status, and pad signals) the functions of the XC167 chip can be emulated in an emulation system. User’s Manual Architecture_X73, V2.1 2-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.7 Protected Bits The XC167 provides a special mechanism to protect bits which can be modified by the on-chip hardware from being changed unintentionally by software accesses to related bits (see also Section 4.8.2). The following bits are protected: Table 2-4 XC167 Protected Bits Register Bit Name Notes GPT12E_T3CON T3OTL GPT1 timer output toggle latches GPT12E_T6CON T6OTL GPT2 timer output toggle latches ASC0_CON REN ASC0 receiver enable flag ASC1_CON REN ASC1 receiver enable flag SSC0_CON BSY SSC0 busy flag SSC0_CON BE, PE, RE, TE SSC0 error flags SSC1_CON BSY SSC1 busy flag SSC1_CON BE, PE, RE, TE SSC1 error flags ADC_CON/ ADC_CTR0 ADST, ADCRQ ADC start flag/injection request flag TFR TFR.15, 14, 13, 12 Class A trap flags TFR TFR.7, 4, 3, 2 Class B trap flags PECISNC C7IR … C0IR All channel interrupt request flags CC1_SEE SEE.15 … SEE.0 Single event enable bits CC2_SEE SEE.15 … SEE.0 Single event enable bits CC1_OUT CC15IO … CC0IO Compare output bits CC2_OUT CC15IO … CC0IO Compare output bits P1L P1L.7 Those bits of PORT1 used for CAPCOM2 P1H P1H.7-4, P1H.0 Those bits of PORT1 used for CAPCOM2 P2 P2.15 … P2.8 All bits of Port 2 used for CAPCOM1 P6 P6.7 … P6.0 All bits of Port 6 used for CAPCOM1 P7 P7.7 … P7.4 All bits of Port 7 used for CAPCOM2 P9 P9.5 … P9.0 All bits of Port 9 used for CAPCOM2 RTC_ISNC T14IR, CNT3IR … CNT0IR Interrupt node sharing request flags CC1_CC15-0IC CC15IR … CC0IR CAPCOM1 interrupt request flags CC2_CC31-16IC CC31IR … CC16IR CAPCOM2 interrupt request flags User’s Manual Architecture_X73, V2.1 2-32 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-4 XC167 Protected Bits (cont’d) Register Bit Name Notes CC1_T1-0IC T0IR, T1IR CAPCOM1 timer interrupt request flags CC2_T8-7IC T7IR, T8IR CAPCOM2 timer interrupt request flags CCU6_IC CIR CAPCOM6 interrupt request flag CCU6_EIC EIR CAPCOM6 error interrupt request flag CCU6_T12IC T12IR CAPCOM6 timer T12 interrupt request flag CCU6_T13IC T13IR CAPCOM6 timer T13 interrupt request flag GPT12E_T6-2IC T6IR … T2IR GPT timer interrupt request flags GPT12E_CRIC CRIR GPT2 CAPREL interrupt request flag ADC_CIC ADCIR ADC end-of-conversion intr. request flag ADC_EIC ADEIR ADC overrun interrupt request flag ASC0_T(B)IC TIR, TBIR ASC0 transmit (buffer) intr. request flags ASC0_RIC, ASC0_EIC RIR, EIR ASC0 receive/error interrupt request flags ASC0_ABIC ABIR ASC0 autobaud interrupt request flags ASC1_T(B)IC TIR, TBIR ASC1 transmit (buffer) intr. request flags ASC1_RIC, ASC1_EIC RIR, EIR ASC1 receive/error interrupt request flags ASC1_ABIC ABIR ASC1 autobaud interrupt request flags SSC0_TIC, SSC0_RIC TIR, RIR SSC0 transmit/receive intr. request flags SSC0_EIC EIR SSC0 error interrupt request flag SSC1_TIC, SSC1_RIC TIR, RIR SSC1 transmit/receive intr. request flags SSC1_EIC EIR SSC1 error interrupt request flag IIC_DTIC DIR IIC data transfer interrupt request flag IIC_PEIC PIR IIC error interrupt request flag PLLIC PLLIR PLL/OWD interrupt request flag EOPIC EOPIR End-of-PEC interrupt request flag CAN_7IC, CAN_0IC CAN7IR … CAN0IR TwinCAN interrupt request flags RTC_IC RTCIR RTC interrupt request flag ----- XX2IR … XX0IR “Unassigned node” interrupt request flags User’s Manual Architecture_X73, V2.1 2-33 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3 Memory Organization The memory space of the XC167 is configured in a “Von Neumann” architecture. This means that code and data are accessed within the same linear address space. All of the physically separated memory areas, including internal ROM/Flash/OTP (where integrated), internal RAM, the internal Special Function Register Areas (SFRs and ESFRs), the internal IO area, and external memory are mapped into one common address space. F F’F FF F 2 55...240 H 2 39...224 O n -C hip P ro gram M em o ry A reas E 0’0000 H 2 23...208 2 07...192 C 0’0 000 H 1 75...160 A 0’0000 H ~12 Mbytes External Addressing Capability 1 59...144 1 43...128 E xternal M em ory A rea 80’0 000 H 1 27...112 111...96 60’0 000 H 95...80 16 Mbytes Total Addressing Capability 1 91...176 79...64 40’0 000 H 63...48 E xternal IO A rea 47...32 20’0 000 E xternal M em ory A rea H 31...16 15...0 00’0 000 H Total A dd ress S pa ce 16 M byte s, S egm ents 255 ...0 m c_xc 16x_m m ap.vs d Figure 3-1 Address Space Overview User’s Manual Memory_X73, V2.1 3-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization The XC167 provides a total addressable memory space of 16 Mbytes. This address space is arranged as 256 segments of 64 Kbytes each, and each segment is again subdivided into four data pages of 16 Kbytes each (see Figure 3-1). Bytes are stored at even or odd byte addresses. Words are stored in ascending memory locations with the low byte at an even byte address being followed by the high byte at the next odd byte address. Double words (code only) are stored in ascending memory locations as two subsequent words. Single bits are always stored in the specified bit position at a word address. Bit position 0 is the least significant bit of the byte at an even byte address, and bit position 15 is the most significant bit of the byte at the next odd byte address. Bit addressing is supported for a part of the Special Function Registers, a part of the internal RAM and for the General Purpose Registers. xxxx6 H 15 14 Bits 8 xxxx5 H 7 6 Bits 0 xxxx4 H Byte xxxx3 H Byte xxxx2 H Word (High Byte) xxxx1 H Word (Low Byte) xxxx0 H xxxxF H MCD01996 Figure 3-2 Storage of Words, Bytes, and Bits in a Byte Organized Memory Note: Byte units forming a single word or a double word must always be stored within the same physical (internal, external, ROM, RAM) and organizational (page, segment) memory area. User’s Manual Memory_X73, V2.1 3-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.1 Address Mapping All the various memory areas and peripheral registers (see Table 3-1) are mapped into one contiguous address space. All sections can be accessed in the same way. The memory map of the XC167 contains some reserved areas, so future derivatives can be enhanced in an upward-compatible fashion. Table 3-1 XC167 Memory Map1) Address Area Start Loc. End Loc. Area Size2) Notes Flash register space FF’F000H FF’FFFFH 4 Kbytes 3) Reserved (Acc. trap) FE’0000H FF’EFFFH 60 Kbytes Minus Flash regs Reserved for EPSRAM F8’1800H FD’FFFFH 378 Kbytes – Emul. Program SRAM4) F8’0000H F8’17FFH 6 Kbytes 2nd way to PSRAM Reserved for PSRAM E0’1800H F7’FFFFH < 1.5 Mbytes Minus PSRAM Program SRAM E0’0000H E0’17FFH 6 Kbytes Maximum5) Reserved for pr. mem. C4’0000H DF’FFFFH < 2 Mbytes Minus Flash Program Flash C0’0000H C3’FFFFH 256 Kbytes – Reserved BF’0000H BF’FFFFH 64 Kbytes – External memory area 40’0000H BE’FFFFH < 8 Mbytes Minus res. seg. External IO area6) 20’0800H 3F’FFFFH < 2 Mbytes Minus TwinCAN TwinCAN registers 20’0000H 20’07FFH 2 Kbytes Accessed via EBC External memory area 01’0000H 1F’FFFFH < 2 Mbytes Minus segment 0 SFR area 00’FE00H 00’FFFFH 0.5 Kbyte – Dual-Port RAM 00’F600H 00’FDFFH 2 Kbytes – Reserved for DPRAM 00’F200H 00’F5FFH 1 Kbyte – ESFR area 00’F000H 00’F1FFH 0.5 Kbyte – XSFR area 00’E000H 00’EFFFH 4 Kbytes – Reserved 00’D000H 00’DFFFH 4 Kbytes – Data SRAM 00’C000H 00’CFFFH 4 Kbytes – Reserved for DSRAM 00’8000H 00’BFFFH 16 Kbytes – External memory area 00’0000H 00’7FFFH 32 Kbytes – 1) Accesses to the shaded areas generate external bus accesses. 2) The areas marked with “<” are slightly smaller than indicated, see column “Notes”. 3) Not defined register locations return a trap code. 4) The Emulation PSRAM (EPSRAM) realizes a 2nd access path to the PSRAM with a different timing. User’s Manual Memory_X73, V2.1 3-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 5) This is the maximum implemented in the derivatives described in this manual. 6) Several pipeline optimizations are not active within the external IO area. This is necessary to control external peripherals properly. 3.2 Special Function Register Areas The Special Function Registers (SFRs) controlling the system and peripheral functions of the XC167 can be accessed via three dedicated address areas: • • • 512-byte SFR area (located above the internal RAM: 00’FFFFH … 00’FE00H) 512-byte ESFR area (located below the internal RAM: 00’F1FFH … 00’F000H) 4-Kbyte XSFR area (located below the ESFR area: 00’EFFFH … 00’E000H) This arrangement provides upward compatibility with the derivatives of the C166 Family. User’s Manual Memory_X73, V2.1 3-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) SFR Area Memory Organization 00’FFFF SFRs H 00’FE00H DPRAM ESFR Area Reserved 00’F200 H ESFRs 00’F000 H EBC 00’EE00H XSFR Area Intr./PEC 00’EC00H 8 Kbytes Upper Half of Data Page 3 00’F600 H 00’EA00 H CAPCOM6 IIC 00’E800 H 00’E600 H 00’E400 H 00’E200 H 00’E000H mc_xc167_regareas.vsd Figure 3-3 Special Function Register Mapping Note: The upper 256 bytes of SFR area, ESFR area, and internal RAM are bit-addressable (see hashed blocks in Figure 3-3). Special Function Registers The functions of the CPU, the bus interface, the IO ports, and the on-chip peripherals of the XC167 are controlled via a number of Special Function Registers (SFRs). All Special Function Registers can be addressed via indirect and long 16-bit addressing modes. The (word) SFRs and their respective low bytes in the SFR/ESFR areas can be addressed using an 8-bit offset together with an implicit base address. However, this does not work for the respective high bytes! User’s Manual Memory_X73, V2.1 3-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Note: Writing to any byte of an SFR causes the not addressed complementary byte to be cleared. The upper half of the SFR-area (00’FFFFH … 00’FF00H) and the ESFR-area (00’F1FFH … 00’F100H) is bit-addressable, so the respective control/status bits can be modified directly or checked using bit addressing. When accessing registers in the ESFR area using 8-bit addresses or direct bit addressing, an Extend Register (EXTR) instruction is required beforehand to switch the short addressing mechanism from the standard SFR area to the Extended SFR area. This is not required for 16-bit and indirect addresses. The GPRs R15 … R0 are duplicated, i.e. they are accessible within both register blocks via short 2-, 4-, or 8-bit addresses without switching. ESFR_SWITCH_EXAMPLE: EXTR #4 MOV ODP9, #data16 BFLDL DP9, #mask, #data8 BSET DP1H.7 MOV T8REL, R1 ;---- ;--------------MOV T8REL, R1 ;Switch to ESFR area for next 4 instr. ;ODP9 uses 8-bit reg addressing ;Bit addressing for bitfields ;Bit addressing for single bits ;T8REL uses 16-bit mem address, ;R1 is duplicated into the ESFR space ;(EXTR is not required for this access) ;The scope of the EXTR #4 instruction … ;… ends here! ;T8REL uses 16-bit mem address, ;R1 is accessed via the SFR space In order to minimize the use of the EXTR instructions the ESFR area mostly holds registers which are mainly required for initialization and mode selection. Registers that need to be accessed frequently are allocated to the standard SFR area, wherever possible. Note: The tools are equipped to monitor accesses to the ESFR area and will automatically insert EXTR instructions, or issue a warning in case of missing or excessive EXTR instructions. Accesses to registers in the XSFR area use 16-bit addresses and require no specific addressing modes or precautions. General Purpose Registers The General Purpose Registers (GPRs) use a block of 16 consecutive words either within the global register bank or within one of the two local register banks. Bitfield BANK in register PSW selects the currently active register bank. The global register bank is mirrored to a section in the DPRAM, the Context Pointer (CP) register determines the base address of the currently active global register bank section. This register bank may consist of up to 16 Word-GPRs (R0, R1, … R15) and/or of up to 16 byte-GPRs (RL0, User’s Manual Memory_X73, V2.1 3-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization RH0, … RL7, RH7). The sixteen byte-GPRs are mapped onto the first eight Word-GPRs (see Table 3-2). In contrast to the system stack, a register bank grows from lower towards higher address locations and occupies a maximum space of 32 bytes. The GPRs are accessed via short 2-, 4-, or 8-bit addressing modes using the Context Pointer (CP) register as base address for the global bank (independent of the current DPP register contents). Additionally, each bit in the currently active register bank can be accessed individually. Table 3-2 Mapping of General Purpose Registers to DPRAM Addresses DPRAM Address High Byte Registers Low Byte Registers Word Register <CP> + 1EH – – R15 <CP> + 1CH – – R14 <CP> + 1AH – – R13 <CP> + 18H – – R12 <CP> + 16H – – R11 <CP> + 14H – – R10 <CP> + 12H – – R9 <CP> + 10H – – R8 <CP> + 0EH RH7 RL7 R7 <CP> + 0CH RH6 RL6 R6 <CP> + 0AH RH5 RL5 R5 <CP> + 08H RH4 RL4 R4 <CP> + 06H RH3 RL3 R3 <CP> + 04H RH2 RL2 R2 <CP> + 02H RH1 RL1 R1 <CP> + 00H RH0 RL0 R0 User’s Manual Memory_X73, V2.1 3-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization The XC167 supports fast register bank (context) switching. Multiple global register banks can physically exist within the DPRAM at the same time. Only the global register bank selected by the Context Pointer register (CP) is active at a given time, however. Selecting a new active global register bank is simply done by updating the CP register. A particular Switch Context (SCXT) instruction performs register bank switching by automatically saving the previous context and loading the new context. The number of implemented register banks (arbitrary sizes) is limited only by the size of the available DPRAM. Note: The local GPR banks are not memory mapped and the GPRs cannot be accessed using a long or indirect memory address. PEC Source and Destination Pointers The source and destination address pointers for data transfers on the PEC channels are located in the XSFR area. Each channel uses a pair of pointers stored in two subsequent word locations with the source pointer (SRCPx) on the lower and the destination pointer (DSTPx) on the higher word address (x = 7 … 0). An additional segment register stores the associated source and destination segments, so PEC transfers can move data from/to any location within the complete addressing range. Whenever a PEC data transfer is performed, the pair of source and destination pointers (selected by the specified PEC channel number) accesses the locations referred to by these pointers independently of the current DPP register contents. If a PEC channel is not used, the corresponding pointer locations can be used for other purposes. For more details about the use of the source and destination pointers for PEC data transfers see Section 5.4. Note: Writing to any byte of the PEC pointers causes the not addressed complementary byte to be cleared. User’s Manual Memory_X73, V2.1 3-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.3 Data Memory Areas The XC167 provides two on-chip RAM areas for data storage: • • The Dual Port RAM (DPRAM) can be used for global register banks (GPRs), system stack, storage of variables and other data, in particular for MAC operands. The Data SRAM (DSRAM) can be used for system stack (recommended), storage of variables and other data. Note: Data can also be stored in the PSRAM (see Section 3.4). However, the data memory areas provide the fastest access. 00’FF FF DPRAM 00’F600 R es. for D P R A M H E S FR s 00’F000 H X S FR s 00’E 80 0 00’E 00 0 R e served DSRAM H H 00’D 800 H 00’D 000 DSRAM H 16 Kbytes Data Page 3 DPRAM S FR s 00’C 800 H H 00’C 000 H m c _xc16x _dataram .v sd Figure 3-4 On-Chip Data RAM Mapping User’s Manual Memory_X73, V2.1 3-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Dual-Port RAM (DPRAM) The XC167 provides 2 Kbytes of DPRAM (00’F600H … 00’FDFFH). Any word or byte data in the DPRAM can be accessed via indirect or long 16-bit addressing modes, if the selected DPP register points to data page 3. Any word data access is made on an even byte address. The highest possible word data storage location in the DPRAM is 00’FDFEH. For PEC data transfers, the DPRAM can be accessed independent of the contents of the DPP registers via the PEC source and destination pointers. The upper 256 bytes of the DPRAM (00’FD00H through 00’FDFFH) are provided for single bit storage, and thus they are bitaddressable (see hashed block in Figure 3-4). Note: Code cannot be executed out of the DPRAM. An area of 3 Kbytes is dedicated to DPRAM (00’F200H … 00’FDFFH). The locations without implemented DPRAM are reserved. Data SRAM (DSRAM) The XC167 provides 4 Kbytes of DSRAM (00’C000H … 00’CFFFH). Any word or byte data in the DSRAM can be accessed via indirect or long 16-bit addressing modes, if the selected DPP register points to data page 3. Any word data access is made on an even byte address. The highest possible word data storage location in the DSRAM is 00’CFFEH. For PEC data transfers, the DSRAM can be accessed independent of the contents of the DPP registers via the PEC source and destination pointers. Note: Code cannot be executed out of the DSRAM. An area of 20 Kbytes is dedicated to DSRAM (00’8000H … 00’CFFFH). The locations without implemented DSRAM are reserved. User’s Manual Memory_X73, V2.1 3-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.4 Program Memory Areas The XC167 provides two on-chip program memory areas for code/data storage: • EPSRAM Flash Reg. • The Program Flash/ROM stores code and constant data. Flash memory is (re-) programmed by the application software, ROM is mask-programmed in the factory. The Program SRAM (PSRAM) stores temporary code sequences and other data. For example higher level bootloader software can be written to the PSRAM and then be executed to program the on-chip Flash memory. FF’FFFF H FF’F000 H Reserved F8’1800 H F8’0000 H F0’0000 H E0’1800 H E0’0000 H 4 Mbytes Segments 255 ... 192 PSRAM Reserved Reserved Flash/ROM D0’0000 H C4’0000 H C0’0000 H mc_xc16x32_progmem.vsd Figure 3-5 On-Chip Program Memory Mapping User’s Manual Memory_X73, V2.1 3-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Program/Data SRAM (PSRAM) The XC167 provides 6 Kbytes of PSRAM (E0’0000H … E0’17FFH). The PSRAM provides fast code execution without initial delays. Therefore, it supports non-sequential code execution, for example via the interrupt vector table. Any word or byte data in the PSRAM can be accessed via indirect or long 16-bit addressing modes, if the selected DPP register points to data page 896. Any word data access is made on an even byte address. The highest possible word data storage location in the PSRAM is E0’17FEH. For PEC data transfers, the PSRAM can be accessed independent of the contents of the DPP registers via the PEC source and destination pointers. Any data can be stored in the PSRAM. Because the PSRAM is optimized for code fetches, however, data accesses to the data memories provide higher performance. Note: The PSRAM is not bitaddressable. An area of 1.5 Mbytes is dedicated to PSRAM (E0’0000H … F7’FFFFH). The locations without implemented PSRAM are reserved. The Emulation PSRAM area (EPSRAM, F8’0000H … F8’17FFH) provides a second access path to the PSRAM with timing parameters that correspond to Flash-timing with WSFLASH = 00B. Using the EPSRAM area produces exactly the same timing as on an emulation device, for timing-sensitive applications. Non-Volatile Program Memory (Flash) The XC167 provides 256 Kbytes of program Flash (C0’0000H … C3’FFFFH). Code and data fetches are always 64-bit aligned, using byte select lines for word and byte data. Any word or byte data in the program memory can be accessed via indirect or long 16-bit addressing modes, if the selected DPP register points to one of the respective data pages. Any word data access is made on an even byte address. The highest possible word data storage location in the program memory is C3’FFFEH. For PEC data transfers, the program memory can be accessed independent of the contents of the DPP registers via the PEC source and destination pointers. Note: The program memory is not bitaddressable. An area of 2 Mbytes is dedicated to program memory (C0’0000H … DF’FFFFH). The locations without implemented program memory are reserved. User’s Manual Memory_X73, V2.1 3-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.5 System Stack The system stack may be defined anywhere within the XC167’s memory areas (including external memory). For all system stack operations the respective stack memory is accessed via a 24-bit stack pointer. The Stack Pointer (SP) register provides the lower 16 bits of the stack pointer (stack pointer offset), the Stack Pointer Segment (SPSEG) register adds the upper 8 bits of the stack pointer (stack segment). The system stack grows downward from higher towards lower locations as it is filled. Only word accesses are supported to the system stack. Register SP is decremented before data is pushed on the system stack, and incremented after data has been pulled from the system stack. Only word accesses are supported to the system stack. By using register SP for stack operations, the size of the system stack is limited to 64 Kbytes. The stack must be located in the segment defined by register SPSEG. The stack pointer points to the latest system stack entry, rather than to the next available system stack address. A stack overflow (STKOV) register and a stack underflow (STKUN) register are provided to control the lower and upper limits of the selected stack area. These two stack boundary registers can be used both for protection against data corruption. For best performance it is recommended to locate the stack to the DPRAM or to the DSRAM. Using the DPRAM may conflict with register banks or MAC operands. User’s Manual Memory_X73, V2.1 3-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.6 IO Areas The following areas of the XC167’s address space are marked as IO area: • • The external IO area is provided for external peripherals (or memories) and also comprises the on-chip LXBus-peripherals, such as the TwinCAN module. It is located from 20’0000H to 3F’FFFFH (2 Mbytes). The internal IO area provides access to the internal peripherals and is split into three blocks: – The SFR area, located from 00’FE00H to 00’FFFFH (512 bytes) – The ESFR area, located from 00’F000H to 00’F1FFH (512 bytes) – The XSFR area, located from 00’E000H to 00’EFFFH (4 Kbytes) Note: The external IO area supports real byte accesses. The internal IO area does not support real byte transfers, the complementary byte is cleared when writing to a byte location. The IO areas have special properties, because peripheral modules must be controlled in a different way than memories: • • • Accesses are not buffered and cached, the write back buffers and caches are not used to store IO read and write accesses. Speculative reads are not executed, but delayed until all speculations are solved (e.g. prefetching after conditional branches). Data forwarding is disabled, an IO read access is delayed until all IO writes pending in the pipeline are executed, because peripherals can change their internal state after a write access. User’s Manual Memory_X73, V2.1 3-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.7 External Memory Space The XC167 is capable of using an address space of up to 16 Mbytes. Only parts of this address space are occupied by internal memory areas or are reserved. A total area of approximately 12 Mbytes references external memory locations. This external memory is accessed via the XC167’s external bus interface. Selectable memory bank sizes are supported: The maximum size of a bank in the external memory space depends on the number of activated address bits. It can vary from 64 Kbytes (with A15 … A0 activated) to 12 Mbytes (with A23 … A0 activated). The logical size of a memory bank and its location in the address space is defined by programming the respective address window. It can vary from 4 Kbytes to 12 Mbytes. • • • • Non-segmented mode: – 64 Kbytes with A15 … A0 on PORT0 or PORT1 1-bit segmented mode: – 128 Kbytes with A16 on Port 4 – and A15 … A0 on PORT0 or PORT1 2-bit … 7-bit segmented mode: – with Ax … A16 on Port 4 – and A15 … A0 on PORT0 or PORT1 8-bit segmented mode: – 12 Mbytes with A23 … A16 on Port 4 – and A15 … A0 on PORT0 or PORT1 Each bank can be directly addressed via the address bus, while the programmable chip select signals can be used to select various memory banks. The XC167 also supports four different bus types: • • • • Multiplexed 16-bit Bus with address and data on PORT0 (default after Reset) Multiplexed 8-bit Bus with address and data on PORT0/P0L Demultiplexed 16-bit Bus with address on PORT1 and data on PORT0 Demultiplexed 8-bit Bus with address on PORT1 and data on P0L Memory model and bus mode are preselected during reset by pin EA and PORT0 pins. For further details about the external bus configuration and control please refer to Chapter 9. External word and byte data can only be accessed via indirect or long 16-bit addressing modes using one of the four DPP registers. There is no short addressing mode for external operands. Any word data access is made to an even byte address. For PEC data transfers the external memory can be accessed independent of the contents of the DPP registers via the PEC source and destination pointers. Note: The external memory is not bitaddressable. User’s Manual Memory_X73, V2.1 3-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.8 Crossing Memory Boundaries The address space of the XC167 is implicitly divided into equally sized blocks of different granularity and into logical memory areas. Crossing the boundaries between these blocks (code or data) or areas requires special attention to ensure that the controller executes the desired operations. Memory Areas are partitions of the address space assigned to different kinds of memory (if provided at all). These memory areas are the SFR areas, the on-chip program or data RAM areas, the on-chip ROM/Flash/OTP (if available), the on-chip LXBus-peripherals (if integrated), and the external memory. Accessing subsequent data locations which belong to different memory areas is no problem. However, when executing code, the different memory areas must be switched explicitly via branch instructions. Sequential boundary crossing is not supported and leads to erroneous results. Note: Changing from the external memory area to the on-chip RAM area takes place within segment 0. Segments are contiguous blocks of 64 Kbytes each. They are referenced via the Code Segment Pointer CSP for code fetches and via an explicit segment number for data accesses overriding the standard DPP scheme. During code fetching, segments are not changed automatically, but rather must be switched explicitly. The instructions JMPS, CALLS and RETS will do this. In larger sequential programs, make sure that the highest used code location of a segment contains an unconditional branch instruction to the respective following segment to prevent the prefetcher from trying to leave the current segment. Data Pages are contiguous blocks of 16 Kbytes each. They are referenced via the data page pointers DPP3 … DPP0 and via an explicit data page number for data accesses overriding the standard DPP scheme. Each DPP register can select one of the possible 1024 data pages. The DPP register which is used for the current access is selected via the two upper bits of the 16-bit data address. Therefore, subsequent 16-bit data addresses which cross the 16-Kbyte data page boundaries will use different data page pointers, while the physical locations need not be subsequent within memory. User’s Manual Memory_X73, V2.1 3-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9 The On-Chip Program Flash Module The XC167 incorporates 256 Kbytes of embedded Flash memory (starting at location C0’0000H, see Figure 3-5) for code or constant data. It is operated from the 5 V pad supply and requires no additional programming voltage. The on-chip voltage generators require a power stabilization time of approx. 250 µs. The Flash array is organized in eight sectors of 4 × 8 Kbytes, 1 × 32 Kbytes, and 3 × 64 Kbytes. It combines the advantages of very fast read accesses with protected but simple writing algorithms for programming and erasing. The 64-bit code read accesses realize maximum CPU performance by fetching two double word instructions (or four single word instructions) in a single access cycle. Data integrity is enhanced by an error correction code enabling dynamic correction of single bit errors. Additionally, special margin checks are provided to detect and correct problematic bits before they lead to actual malfunctions. All Flash operations are controlled by command sequences (according to the JEDEC single-power-supply Flash standard). The algorithms for programming and erasing are executed automatically by the internal Flash state control machine. This avoids inadvertent destruction of the Flash contents at a reasonably low software overhead. Command sequences consist of subsequent write (or read) accesses to virtual locations within the Flash space or the Flash register space. The virtual Flash locations are defined by special addresses (see command sequence table). For optimized programming efficiency, paging mode (burst mode) allows 128 bytes to be loaded into a page buffer with fast CPU accesses before this buffer is programmed into the Flash with one single store command (2 ms typical1)). Each sector can be erased separately (200 ms typical1)). Note: Erased Flash memory cells contain all ‘0’s, contrary to standard EPROMs. Security is provided by a general read/write protection (complete Flash array) and a sector-specific2) write protection. The temporary disabling of these hardware protection features is secured with a password check sequence. The lock information and the keywords used for the password check sequence are stored apart from the user’s code and data in a separate security sector (see Section 3.9.4). A dedicated Flash status register returns global and sector-specific status information. The correct execution of an operation and the general status of the Flash module can be checked via the Flash status register at any time. The physical address range of the Flash module covers byte addresses from 0’0000H to 3’FFFFH. These physical addresses are mapped to the XC167’s program memory area starting at C0’0000H. Also the separate security sector is mapped to this area. Access conflicts are avoided by special security commands. 1) For exact parameters please refer to the data sheet. 2) For write protection two 8-Kbyte sectors are combined to one lockable 16-Kbyte section. User’s Manual Memory_X73, V2.1 3-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization In-System-Programming is supported by the automatic program/erase algorithms and the large page buffer, which may be filled by a programming routine executed out of the Flash memory itself. During the actual program/erase algorithm Flash read accesses are stalled. Also completely erased Flash modules can be programmed within the system. The built-in bootstrap loader can load an initial programming routine via the serial interface, which in turn can then program the Flash module. This is useful for the initial programming (virgin Flash) as well as in case of a problem (e.g. power failure) during reprogramming, when no safety routines are provided. Note: Accesses to a protected Flash are totally disabled during bootstrap mode. Before any program/erase operation the protection must be temporarily disabled using the correct password sequence. DF’FFFF H Segments 196 ... 223 C4’0000 H H 64 Kbytes Segment 195 64 Kbytes 2 Mbytes C3’0000 H 256 Kbytes Flash Module Segment 194 C2’0000 H 64 Kbytes Flash/ROM Area 3’FFFF Segment 193 1’0000 H C1’0000 H 32 Kbytes 0’0000H Physical Address 8 Kbytes 8 Kbytes 8 Kbytes 8 Kbytes Segment 192 Flash Sectors Program Memory (Segments) C0’0000H Location mc_xc16x32_flashmap.vsd Figure 3-6 Mapping of the On-Chip Flash Module Sectors User’s Manual Memory_X73, V2.1 3-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.1 Flash Operating Modes Two basic operating modes of the on-chip Flash module can be distinguished: • • Standard read mode: code and data can be read from the Flash module Command mode: the Flash module executes a previously defined command Standard Read Mode In standard read mode (the normal operating mode) the Flash memory appears like a standard ROM, allowing code and data accesses in any addressing mode. Standard read mode is entered in the following cases: • • • • • After the deactivation of the system reset (after power stabilization) After execution of the reset command, if no program or erase operation is active After every completed command execution (program, erase, etc.) When a command sequence error is detected When a protection violation is detected (program or erase a protected sector) Note: Standard read mode is indicated by status bit BUSY = ‘0’. Standard read mode is terminated when the last command of a command sequence is decoded and a Flash array operation is started (program or erase). Therefore, all steps of a command sequence before the last command (in particular the loading of the page buffer) can be executed by code read from the Flash module itself. Each read access to the Flash memory activates the automatic error detection. Double bit errors are detected and indicated, single bit errors are detected, indicated, and automatically corrected (see Section 3.9.3). Note: Single bit errors can be located and avoided by a margin check operation. Command Mode All Flash operation except for standard read operations are initiated by command sequences written to the (virtual) Flash command register (a location within the Flash space). Protected commands additionally require four passwords for validation. Command mode is entered after the last command of a command sequence has been written. For all other command sequences, which activate a Flash array operation such as erase sector, the command execution and thus the command mode remains active for a defined time. While in command mode (busy) read accesses to the Flash array are delayed until the Flash module returns to standard read mode. Note: Command mode is indicated by status bit BUSY = ‘1’. Command mode is terminated by the correct execution of the command or by an error condition as indicated in the status register. Command sequences not starting Flash operations (e.g. Enter Page Mode) are executed immediately and command mode is not entered. User’s Manual Memory_X73, V2.1 3-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.2 Command Sequences All operations besides normal read operations are initiated and controlled by command sequences written to the Flash state machine. The different write cycles of command sequences define the intended command, but also establish a fail-safe mechanism to protect against inadvertent operations. Commands not directly controlling Flash array operations are single cycle commands for performance reasons, commands affecting the Flash array require several cycles, commands affecting security issues require a 64-bit security code (four passwords) to be accepted. Command cycles need not be consecutively received (pauses allowed). Command sequences can be performed simultaneously to instruction fetch operations, so instructions for command sequences also can be executed out of the on-chip Flash, as long as the Flash module is in read mode and not executing an erase or programming operation. Command sequences for polling the status register are allowed in any state, also during erase and programming operations, if they are executed out of memory outside the Flash module. Otherwise, instruction fetching is stalled. Writing incorrect address and data values or writing them in the improper sequence will abort the intended operation, reset the module to read mode, and set the sequence error flag in the status register. Read Status commands address the separate Flash register space and do not require command sequences. Register write cycles are only executed with a command cycle. Programming operations are supported by a 128-byte page buffer which can be loaded with maximum speed, and is then programmed with one single command sequence. Programming is done in three steps: • • • Initialize the page buffer with the Enter Page Mode command (this also defines the target page address). Load the page buffer with consecutive Load Page command (the page buffer offset is incremented automatically). Program the complete buffer with the Write Page command. Erase operations clear all bits of a selected sector or of a 256-byte wordline. Erase command sequences include the address of the target sector or wordline. After being requested the program/erase operation is executed automatically and requires no additional user control. The operation itself and its termination are indicated by status flags. A Power Down request is delayed until the termination of the program/erase operation. A reset aborts the program/erase operation within the power stabilization time, indicated by an operation error (OPER) in the Flash status register. The three tables below summarize the implemented command sequences for: • • • organizational Flash accesses (Table 3-3), programming and erasing (Table 3-4), protection control (Table 3-5). Note: Each command sequence lists the required address (A = …) and data (D = …). User’s Manual Memory_X73, V2.1 3-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Organizational Commands Cycle Table 3-3 Command Sequence Definitions (Organizational Accesses) Reset to Read Mode Clear Status Read Flash Status or Margin Write Margin 1 A = Cx’xxAAH D = xxF0H A = Cx’xxAAH D = xxF5H A = RLOC D = <status> A = Cx’xxAAH D = xxFAH 2 – – – A = FF’F00CH D = margin Notes: RLOC is the respective register offset (rr) within the Flash register area starting at FF’F000H (FF’F0rrH). <status> is the returned status word. margin is the control word used for margin control. The shown virtual address (Cx’xxAAH) must point to the Flash space (e.g. C0’00AAH). The “Read Flash status” command may be executed during command mode in order to check the BUSY bit of the Flash module. The Reset To Read command aborts not completed command sequences and clears the error flags in the status register FSR. The reset command can be issued at any point during the command sequence, except for parts of the password check sequence. It does not terminate command mode, i.e. abort busy state. The Clear Status command clears the error flags and the write status bits PROG and ERASE (the hardware-controlled indication flags are not affected). The clear status command is only accepted in Read Mode and otherwise generates a sequence error. The Read Register command returns the contents of the following registers: • • • The Flash Status Register FSR providing general Flash status information. The Protection Configuration Register PROCON indicating the protected sectors. The Margin Control Register MAR indicating the selected Flash read margin. The Write Margin Register command is used for verify operations and for usercontrolled refresh operations to identify and correct problematic bits (see Section 3.9.3). User’s Manual Memory_X73, V2.1 3-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Program/Erase Commands Table 3-4 Command Sequence Definitions (Programming & Erasing) Load Page Data Word2) Write Page3) Erase Sector1) Erase Wordline1) 1 A = Cx’xxAAH D = xx50H A = Cx’xxF2H D = WDAT A = Cx’xxAAH D = xxA0H A = Cx’xxAAH D = xx80H A = Cx’xxAAH D = xx80H 2 A = WLOC D = xxAAH – A = Cx’xx5AH D = xxAAH A = Cx’xx54H D = xxAAH A = Cx’xx54H D = xxAAH 3 – – – A = SLOC D = xx33H A = WLA D = xx03H Cycle Enter Page Mode1) 1) While protection is enabled, this command sequence is rejected. 2) Words written in excess of the buffer capacity of 128 bytes are lost. 3) This command sequence is only accepted if page mode has been entered before. Notes: WLOC is the first (lowest) location of the 128-byte block to which the 128-byte buffer shall be written, e.g. C0’AB80H or C0’AC00H (128-byte boundary). WDAT is the data word which shall be stored in the buffer. SLOC is the first (lowest) location within the target sector, e.g. C0’6000H for sector 3. WLA is the first (lowest) location of the 256-byte wordline to be erased, e.g. C1’FF00H for the uppermost 256 bytes (top of sector 5). The shown virtual addresses (Cx’xx..H) must point to the Flash space (e.g. C0’00AAH). The “Read Flash status” command may be executed during command mode in order to check the BUSY bit of the Flash module. Caution: Writing to a Flash page (space for the 128-byte buffer) more than once before erasing may destroy data stored in neighbor cells! This is especially important for programming algorithms that do not write to sequential locations. The Enter Page Mode command prepares the programming of a 128-byte page by clearing the page buffer and initializing the internal word assembly pointer. Bit PAGE in the status register FSR is set to indicate this. Issuing the Enter Page Mode command during page mode aborts the current operation and starts a new page operation. The data written to the page buffer during the aborted page operation are lost. The Enter Page Mode command also defines the location of the 128-byte page to be programmed. Note: The Enter Page Mode command is only accepted while protection is disabled. User’s Manual Memory_X73, V2.1 3-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization The Load Page Data Word command adds the accompanying data word to the page buffer. The offset within the page buffer is determined by the internal buffer pointer which is incremented after each load operation. Data words written in excess of the buffer capacity of 128 bytes are lost (no error indicated). Note: The Load Page Data Word command is only accepted while page mode is active. The Write Page command writes (programs) the contents of the 128-byte page buffer (including the error correction code) to the Flash array. The address of the programmed page is defined by the preceding Enter (Security) Page Mode command. After the Write Page command the Flash module enters command mode, indicated by PAGE = 0, PROG = 1, BUSY = 1. Read accesses to the Flash module are delayed until command mode is terminated. The programming operation itself is executed automatically and requires no additional user control. If a security page is written the new protection configuration (including keywords or protection confirmation code) is valid directly after execution of this command. Note: The Write Page command is only accepted while page mode is active. The Erase Sector command clears all bits within the selected sector (see SLOC). After the Erase Sector command the Flash module enters command mode, indicated by ERASE = 1, BUSY = 1. Read accesses to the Flash module are delayed until command mode is terminated. The erase operation itself is executed automatically and requires no additional user control. Note: The Erase Sector command is only accepted while protection is disabled. The Erase Wordline command clears all bits within the selected 256-byte wordline (see WLA). After the Erase Wordline command the Flash module enters command mode, indicated by ERASE = 1, BUSY = 1. Read accesses to the Flash module are delayed until command mode is terminated. The erase operation itself is executed automatically and requires no additional user control. Note: The Erase Wordline command is only accepted while protection is disabled. User’s Manual Memory_X73, V2.1 3-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Protection Control Commands Table 3-5 Command Sequence Definitions (Protection Control) Erase Security Wordline1) Enter Security Page Mode1) 1 A = Cx’xx3CH D = xx00H A = Cx’xx3CH D = xx00H A = Cx’xx5EH D = xx5EH A = Cx’xxAAH D = xx80H A = Cx’xxAAH D = xx55H 2 A = Cx’xx54H D = PW1 A = Cx’xx54H D = PW1 – A = Cx’xx54H D = xxA5H A = SECLOC D = xxAAH 3 A = Cx’xxAAH D = PW2 A = Cx’xxAAH D = PW2 – A = SECWLA D = xx53H – 4 A = Cx’xx54H D = PW3 A = Cx’xx54H D = PW3 – – – 5 A = Cx’xxAAH D = PW4 A = Cx’xxAAH D = PW4 – – – 6 A = Cx’xx5AH D = xx55H A = Cx’xx5AH D = xx05H – – – Cycle Disable Read Disable Write Re-Enable Protection Protection Protection 1) While protection is enabled, this command sequence is rejected. Note: A Reset-To-Read command cannot be executed while the 2nd or the 4th password is expected. In this case the command is taken as a password. Notes: SECLOC is the first (lowest) location of the 128-byte block within the security sector to which the 128-byte buffer shall be written, e.g. C0’0080H or C0’0100H. SECWLA is the first (lowest) location of the 256-byte security wordline to be erased, e.g. C0’0100H for the upper 256-byte wordline. PWn is one of the four passwords building the 64-bit security code (n = 1 … 4). The shown virtual addresses (Cx’xx..H) must point to the Flash space (e.g. C0’00AAH). The “Read Flash status” command may be executed during command mode in order to check the BUSY bit of the Flash module. The Disable Read Protection command temporarily disables the general Flash read protection (including the general write protection), indicated by PRODI = 1. Read protection remains disabled until the execution of the Re-Enable Protection command or until the next reset. While read protection is disabled, Flash read accesses including injected OCDS instructions are executed. Program/Erase operations can be executed as long as the respective sector is not locked by a sector-specific write protection. User’s Manual Memory_X73, V2.1 3-24 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Note: This command sequence can also be used to verify the programmed keywords before the protection is locked with the confirmation. A wrong keyword is indicated by bit PROER in the Flash Status Register FSR. This is a protected command sequence requiring the 64-bit security code (four userdefined passwords) for validation (see Section 3.9.4). The Disable Sector Write Protection command temporarily disables the sectorspecific write protection for all write-protected sectors, indicated by SUL = 1. Write protection remains disabled until the execution of the Re-Enable Protection command or until the next reset. While write protection is disabled, all Flash operations can be executed as long as the respective sector is not locked by the general read/write protection. Note: This command sequence can also be used to verify the programmed keywords before the protection is locked with the confirmation. A wrong keyword is indicated by bit PROER in the Flash Status Register FSR. This is a protected command sequence requiring the 64-bit security code (four userdefined passwords) for validation (see Section 3.9.4). The Re-Enable Protection command immediately resumes all installed but temporarily disabled protection features (general read/write protection and/or sector-specific write protection). The Erase Security Wordline command clears all bits within the selected wordline (see SECLOC). After the Erase Security Wordline command the Flash module enters command mode, indicated by ERASE = 1, BUSY = 1. Read accesses to the Flash module are delayed until command mode is terminated. The erase operation itself is executed automatically and requires no additional user control. After the erase operation, the protection configuration (including keywords or protection confirmation code) is valid directly after execution of this command (see Section 3.9.4). Note: The Erase Security Wordline command is only accepted while protection is disabled. The Enter Security Page Mode command prepares the programming of a 128-byte page within the security sector by clearing the page buffer and initializing the internal word assembly pointer. Bit PAGE in the status register FSR is set to indicate this. Issuing the Enter Security Page Mode command during page mode aborts the current operation and starts a new page operation. The data written to the page buffer during the aborted page operation are lost. The Enter Security Page Mode command also defines the location of the 128-byte page to be programmed. Also refer to Section 3.9.4. Note: The Enter Security Page Mode command is only accepted while any protection is disabled. User’s Manual Memory_X73, V2.1 3-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.3 Error Correction and Data Integrity Data integrity is supported by the Error Correction Code (ECC). This ECC is dynamically generated during Flash write operations and stored in the Flash array together with the corresponding data. For each read access the associated 8-bit ECC is fetched together with the 64-bit read data and is evaluated. Single-bit errors are detected and automatically corrected on-the-fly (during run-time). Therefore, single bit errors do not affect system operation. Double-bit errors are detected and trigger an Access Fault trap. This prevents erroneous instructions or data from being used. Each read error condition is indicated by a dedicated flag (SBER, DBER) in the Flash Status Register FSR. The probability of a double bit error (not automatically correctable by ECC) is extremely low. Double-bit errors can be avoided by performing a recovery operation after a single bit error has been detected. For the recovery operation the following steps must be done: • • • • Detect the wordline containing the erroneous bit Store the contents of the wordline temporarily Erase this wordline Reprogram the erased wordline (requires two write page operations) The wordline data copied to the temporary buffer are valid, because a single bit error during reading is automatically corrected via the ECC. Erasing and programming is done using standard command sequences. Verify Operation The violated wordline can be detected by a verify operation. After clearing bit SBER a certain area of the Flash memory is read. Since the Flash array always delivers 64-bit data, the read address can be incremented by 8 after every access, which minimizes the number of necessary read cycles. After reading the defined area bit SBER indicates if or if not this area contains the single bit error. The verify algorithm can gradually decrease the size of the checked area down to the size of a wordline, or can check all wordlines sequentially. Refresh Operation Even single bit errors can be avoided by detecting problematic (moving) bits before they lead to a read error (and a recovery operation during runtime) and by reprogramming (refreshing) them in advance. Problematic bits can be detected by combining the verify operation with margin check control. User’s Manual Memory_X73, V2.1 3-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Margin Check Control Flash cells store charges to represent bit levels. If the charge stored in a cell changes (e.g. due to charge coupling during operations on neighbor cells) the respective bit may be read wrong. As the charges change slowly this effect can be detected before a bit is actually read wrong. In this case also a preventive correction (via software) is possible. A problematic bit (i.e. a bit with a changed charge) can be detected by applying a more severe comparator margin when reading a Flash location. This margin is controlled by the Margin Control Register MAR, accessible with the special command sequences Read/Write Margin (see Table 3-3). A severe margin is selected by writing the value MARLEVSEL = 0001B or 0100B to register MAR. A bit that returns a 1 when read with low level margin, while returning a 0 when read with standard margin, represents a problematic bit, called weak zero. A bit that returns a 0 when read with high level margin, while returning a 1 when read with standard margin, represents a problematic bit, called weak one. Compare operations over a certain memory area using standard and severe margins reveal these problematic bits. Note: Read operations may directly follow a MAR change operation. MAR Margin Control Register SFR (FF’F00CH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 - - - - - - - - MAR WV - - - MARLEVSEL - - - - - - - - rw - - - rw 0 Field Bits Type Description MARWV 7 rw Margin Write Validation 0 Reset value. MARWV must not be written 0. 1 Must be set (MARWV = 1) with every write access to register MAR, independent of the purpose of the write access. MARLEVSEL [3:0] rw Margin Level Selection 0000 Standard read margin (regular operation) 0001 Low level margin (used to verify weak zeros) 0100 High level margin (used to verify weak ones) other Reserved Note: Margin values can only be written via the Write Margin command. Bit MARWV must be set with every write access. User’s Manual Memory_X73, V2.1 3-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.4 Protection and Security Features The Flash module provides powerful and flexible protection of data and code against destruction (i.e. erasure) and undesired modification (i.e. reprogramming) as well as against undesired read access to Flash contents. Two protection mechanisms can be activated: • • Sector-specific write protection protects individual sectors against erasing and programming. This is important for the integrity of boot software and also avoids modifications of code/data by malfunction or even manipulation. General read/write protection protects the complete program Flash area against all accesses from outside the module itself. This includes data read accesses, instruction fetches (i.e. jumps into the program Flash area), and OCDS operations. The general read/write protection also disables erasing and programming. Command sequences and register accesses are executed, however. Each protection feature is installed by user software. Protection features may be disabled temporarily to reprogram portions of the Flash memory or to call an external subroutine. Disabling and re-enabling is done under software control. However, after a reset all installed protection features are active (enabled) automatically. By combining the two protection features a flexible protection scheme can be installed to protect the Flash memory or parts of it against unauthorized programming or erasing according to the application’s requirements. Note: Protection is provided for the Program Flash only, there is no protection for the Program SRAM. Passwords and Security Code All protection feature control (install, disable, re-enable) is accomplished through command sequences similar to the program/erase sequences (see Table 3-5). The two command sequences that temporarily suspend the protection feature are additionally secured by a password check sequence (64-bit security code) to ensure maximum safety against undesired accesses. During password checking, the four passwords entered via the command sequence are compared to the four keywords (building the 64-bit security code) stored in the security sector. If any mismatch is detected the respective protection feature remains active, the sector(s) remain(s) locked, and a protection error (PROER) is indicated in the Flash status register. In this case, a new Disable Sector Write Protection command or a Disable Read Protection command is only accepted after the next system-reset. User’s Manual Memory_X73, V2.1 3-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Security Feature Installation The security features are installed by programming the following data (see Figure 3-7) into the security sector: • • • Security control bits, selecting the security feature(s) to be installed 64-bit security code (four keywords) 16-bit confirmation code Note: If any protection is enabled also the security sector itself is protected. The security control bits can be checked via register PROCON. The same bit-layout must be used when programming the security control bits. PROCON Protection Control Register SFR (FF’F004H) 15 14 13 12 11 10 9 8 7 6 R PRO - - - - - - - - - rh - - - - - - - - - Reset Value: xxxxH 5 4 3 2 1 0 SL5 SL4 SL3 SL2 SL1 SL0 rh rh rh rh rh Field Bits Type Description RPRO 15 rh Read/Write Protection Configuration 0 No general protection installed 1 General read/write protection is installed SLn (n = 5 … 0) 5, 4, 3, rh 2, 1, 0 rh Sector Lock Bit n 0 Sector is unprotected 1 Write protection installed for sector n Note: Each two 8-Kbyte sectors are combined to a 16-Kbyte region that can be jointly locked by bits SL1 and SL0. Note: The security configuration can be checked by reading register PROCON. To modify the security configuration the security sector must be modified. The 64-bit security code (e.g. 494E’4649’4E45’4F4EH) must be correctly entered for commands that temporarily disable security features. Any failure to enter all four words correctly aborts the command and freezes the current security state until the next system reset. The 16-bit confirmation code (8AFEH) is required to validate the security feature installation. The installed configuration can be verified prior to validating it. The security information and the confirmation code are stored in separate wordlines so they can be programmed and erased independently from each other. User’s Manual Memory_X73, V2.1 3-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Each byte of the security information is stored three times and completed with a zerobyte, so each 16-bit word to be stored uses the space of two doublewords (see example in Figure 3-7). All three copies of a data byte are used for evaluation which provides extreme reliability. S ecurity P age 3: R eserved S e curity P age 2: C on trol B its/W ords S ecu rity W ordline 1 S ecurity P age 0: R eserved S ecurity P a ge 1 : C onfirm ation S ecu rity W ordline 0 12 7 64 56 48 40 32 K e yw . 4 24 K eyw . 3 16 K e yw . 2 8 0 K eyw . 1 P R O C O N S ecu rity P age 2 12 7 64 56 48 40 32 24 16 8 0 C onfirm . S ecu rity P age 1 7 00H 6 8A H 5 8A H 4 8A H 3 00H 2 FEH 1 FE H 0 FE H S torage -S tructure fo r a W ord 1 8A H 0 FE H E xam ple-W ord = C onfirm ation m c_xc 16x _secs ector.vsd Figure 3-7 Security Sector Structure User’s Manual Memory_X73, V2.1 3-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Whenever the security configuration is modified (installation, modification, deinstallation) the following procedure should be performed: • • • • • Clear confirmation code by erasing security wordline 0. This uninstalls all protection features (PROIN = 0). Erase security wordline 1. Program the intended configuration and keywords into security page 2. Verify the programmed configuration and keywords. Program the confirmation code into security page 1. This installs the new protection features. Following these steps prevents dead-locks resulting for example from programming erroneous keywords (e.g. due to power problems during programming) with existing confirmation code. The security features would be immediately active in this case whereas the erroneous keywords are not known. Read/Write Protection Control Read protection can be activated for code fetches and data reads separately via the control bits DCF (Disable Code Fetch) and DDF (Disable Data Fetch) in register IMBCTR. Read accesses are blocked as long as the respective disable flag (DCF, DDF) is set and read protection is active, indicated by bit RPA (Read Protection Active) in register IMBCTR. An access to the protected Flash will deliver a dummy value of 1E9BH. While read protection is disabled (RPA = 0), bits DCF and DDF have no effect on read accesses. After a reset starting execution out of the on-chip Flash module bits DCF and DDF are cleared. This enables all accesses while code is executed from a safe source. Bit DDF can be set by user software to prevent data reads from the Flash module while still enabling code execution. After any other reset (including boot mode) both bits are set (if protection is installed). By entering the 64-bit security code the read protection can be disabled temporarily by software executed out of external sources. Note: Bits DCF and DDF can only be set via software, they cannot be cleared. Attention: Be sure not to set DCF while executing out of on-chip Flash with read protection active. Read/write protection is active (RPA = 1) if it has been installed (RPRO = 1) and is currently not disabled (PRODI = 0). User’s Manual Memory_X73, V2.1 3-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Read/Write Protection Handling After reset, bit RPA indicates if the read/write protection is installed or not. User software can disable the read/write protection temporarily (indicated by RPA = 0). Bits DCF and DDF prevent Flash read accesses while RPA = 1. Because DCF and DDF are cleared after starting from the on-chip Flash memory, the user software is responsible for the protection handling. If the read/write protection is enabled, the debug system is disabled to avoid notauthorized accesses to the Flash via the debug interface. Only if explicitly enabled by user software, the debug interface can be temporarily activated, even if the read/write protection is enabled. The following rules ensure a safe read/write protection: • • • • no JUMPs or CALLs to external memory locations no execution of code loaded via any interface set DCF and DDF before transferring control to external locations (no return!) leave the debug system disabled Note: Of course, external code can be executed intermediately while the read/write protection is disabled. Also the debug interface can be enabled, so protected devices can be debugged. However, this should only be done after validation (e.g. by a specific security key), because read/write protection does not work during these phases. User’s Manual Memory_X73, V2.1 3-32 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.5 Flash Status Information The Flash Status Register FSR provides status information about all functions of the Flash module: • • • Operating state Error conditions Security level The FSR should be read before and after the execution of command sequences. The FSR cannot be written directly. The “Clear Status” command clears the error flags and the status flags PROG and ERASE, the “Reset to Read” command clears the error flags. FSR Flash Status Register 15 14 13 12 - - SUL - - - rh - SFR (FF’F000H) 11 10 9 PRO PRO DB IN DI ER rh rh rh 8 7 5 4 SB PRO SQ ER ER ER - OP ER PA ERA PR GE SE OG BU SY rh - rh rh rh rh 6 Reset Value: 0xxxH rh 3 2 rh 1 0 rh Field Bits Type Description SUL 13 rh Sectors Unlocked 0 Sectors are protected according to the installation 1 All sectors are temporarily unlocked (check general protection) PROIN 11 rh Protection Installed 0 No security features installed 1 General read/write protection and/or sectorspecific write protection installed (see register PROCON) PRODI 10 rh Protection Disabled 0 General read/write protection active (if installed) 1 General read/write is temporarily disabled DBER 9 rh Double Bit Error (Cleared via “Clear status”, “Reset-to-read”) 0 No double bit error has occurred 1 A double bit error was detected (no correction possible) User’s Manual Memory_X73, V2.1 3-33 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Type Description SBER 8 rh Single Bit Error (Cleared via “Clear status”, “Reset-to-read”) 0 Read/fetch accesses executed without error 1 A single bit error was detected and automatically corrected PROER 7 rh Protection Error (Cleared via “Clear status”, “Reset-to-read”) 0 No protection error detected 1 Protection error has occurred: attempt to program/erase a locked sector or invalid security code1) SQER 6 rh Command Sequence Error (Cleared via “Clear status”, “Reset-to-read”) 0 No command sequence error detected 1 State machine operation aborted due to invalid command step Note: SQER is not set when a command sequence is aborted with a “Reset to Read” command. SQER is set when a “Clear Status” command is attempted while the Flash module is busy (PROG or ERASE are not cleared). OPER 4 rh Operation Error (Cleared via “Clear status”, “Reset-to-read”) 0 Flash operation successfully finished or currently in progress 1 Flash operation not successfully terminated (abortion) PAGE 3 rh Page Mode (Cleared via “Reset-to-read”) 0 Flash not in page mode 1 Flash in page mode, page buffer being filled Note: Page mode can be active during standard read mode. ERASE 2 User’s Manual Memory_X73, V2.1 rh Erase State (Cleared via “Clear status”, “Reset-to-read”) 0 There is no erase operation in progress 1 Flash busy with erase operation 3-34 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Type Description PROG 1 rh Programming State (Cleared via “Clear status”, “Reset-to-read”) 0 There is no programming operation in progress 1 Flash busy with programming operation (write page) BUSY 0 rh Flash Busy 0 Ready: Flash command execution is completed. Module is in standard read mode. 1 Busy: Embedded algorithm for command execution is in progress or Flash module is in ramp-up state2). Module not in read mode. 1) After the occurrence of a protection error the next password sequence is only accepted after a reset. 2) After a system reset BUSY will be active for approx. 250 µs until the internal voltages have settled. Note: By evaluating bits PROG and ERASE together with bits BUSY and OPER the control software can determine if an operation is in progress, has terminated, or has been aborted. User’s Manual Memory_X73, V2.1 3-35 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.6 Operation Control and Error Handling Command execution is started with the last command of the respective command sequence and is indicated by the respective state flag (PROG for programming, ERASE for erasing) as well as by the summarizing BUSY flag. While polling BUSY is sufficient to detect the end of a command execution it is recommended to check the error flags afterwards to find erroneous operations. The following general structure for command execution is recommended: • • • • • • • Clear status Write command sequence to Flash module Ensure correct sequence by checking bits SQER and PROER If error: clear flags via “Clear Status” or “Reset” and act upon it (e.g. with a retry operation) Check for the correct command by polling bits PROG and ERASE Poll BUSY to determine the command termination Check error flags The error bits in status register FSR are registered bits (flipflops) and indicate a fault condition as long as the error bit is set. It is therefore necessary to clear the error flags by commands. Table 3-6 gives examples of software actions to be taken after a specific error has been detected: Table 3-6 Software Reactions to Error Conditions Detected Error Fault Condition Software Reaction SQER Sequence Error Wrong register address, wrong command/sector/wordline address, wrong command code, illegal command sequence Check address or code and repeat with correct values OPER Operation Error Aborted programming or erase operation due to SW reset, WDT reset, or warm HW reset Repeat Flash operation (PROG and ERASE indicate the failed operation) PROER Protection Error Begin of write operation (Enter Page Mode) to protected sector, General password failure Retry operation after disabling protection, Retry operation after reset SBER Single Bit Error The Error Correction Code (ECC) has Refresh faulty wordline revealed a single bit error (see Section 3.9.3) DBER1) The Error Correction Code (ECC) has Double bit error triggers a trap Double Bit Error revealed a double bit error 1) Does not occur if a refresh operation is executed after a single bit error (see Section 3.9.3). User’s Manual Memory_X73, V2.1 3-36 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Reset and Power-Down Processing Upon a reset the Flash module resets its state machine and enters the standard read mode after the internal voltages have stabilized. The internal voltages need to ramp up (e.g. after power down) or to ramp down (e.g. after an interrupted programming or erase operation). This power stabilization phase is indicated by flag BUSY. Accesses during the power stabilization phase are delayed until power has stabilized. The Flash module is requested to ramp down its internal voltages by entering Power Down mode, Sleep mode or Idle mode (with Flash off), by disabling it via SYSCON3, or by executing a software reset. After completing execution and termination of the running operation (including program or erase operation) the request is acknowledged and the CPU can complete the intended action. Note: The delay caused by the stabilization phase must also be considered when calculating delays for wake-up from idle, sleep, or power down states. User’s Manual Memory_X73, V2.1 3-37 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10 Program Memory Control The internal program memory block IMB consists of an interface part (program memory interface PMI) to control the accesses to the memories and the following memory blocks: • • 256 Kbytes program Flash memory, starting at address C0’0000H (including error correction ECC) 6 Kbytes program SRAM, starting at address E0’0000H (second access path at address F8’0000H) The Flash memory block and the program SRAM block can contain the program code, but can also store data, which can be accessed by the CPU. Note: The access to addresses, which are not explicitly mentioned as valid memory/register area is forbidden. CFGS Program Memory Block Flash Array Program SRAM FIM Interface 22 64 Interface 16 22 64 16 1 21 RPA 15 SCU IMBCTR 64 Program Memory Interface (PMI) Control and Multiplexer 16 1 1 16 Ready 64 Flash 18 Acces Ready Flash Enable Startup ROM PMU Interface Wait States Flash Emulation Device (Overlay + Monitor) Address Data Figure 3-8 MCA05503 Overview of the Internal Program Memory Block IMB User’s Manual Memory_X73, V2.1 3-38 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Figure 3-8 shows the main blocks of the IMB, specific control signals are not mentioned for simplicity reasons. The behavior of the memories is adaptable to the requirements of the application. If the program is executed from the on-chip Flash memory or from the internal SRAM, the latencies have to be identical in some cases. To solve this problem, the access times of the SRAM can be programmed to be equal to the Flash timings. In the best case, the internal SRAM will allow single cycle accesses. A programmable wait state generation logic is part of the program memory interface (PMI) inside the IMB. The number of access cycles can be programmed independently for the SRAM and the Flash memory. 3.10.1 Flash Memory Access The internal functional structure of the interface between the PMU/PMI and the Flash memory is shown in Figure 3-9. The access is done in two phases: • • The Flash array delivers the accessed data within a fixed time of 50 ns maximum. The duration of the first access phase (1+WS) must cover the Flash Array’s access time. Waitstates must be selected accordingly (bitfield WSFLASH in register IMBCTRL). Example: Operating at 40 MHz results in a cycle time of 25 ns. Therefore, the access phase requires 2 cycles, so one waitstate must be selected (1+1). The error correction (ECC) and the PMU require one additional clock cycle each. The CPU receives requested data after 1+WS+2 cycles (4 cycles if 1 WS is selected). However, this delay only becomes effective for an isolated access (read from a nonlinear address). A prefetching mechanism overlaps phase 1 of a subsequent access with phase 2 of the previous access, so the sustained performance for linear accesses (e.g. code fetches) is considerably higher. Flash accesses can be serviced every 1+WS cycles, because the Flash array itself only requires phase 1. User’s Manual Memory_X73, V2.1 3-39 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Flash Address Access Phase 1 Flash Address Unit PMU + PMI Flash Array Flash Data CPU Address CPU PMU + PMI ECC CPU Data Access Phase 2 MCA05500 Figure 3-9 Flash - PMI Structure Note: The read-only Flash status registers can be accessed starting at address FF’F000H. Any write access to this address range must be avoided. Example for Flash Accesses with one Wait State (WS = 1) After the first access (e.g. after a jump to the first address delivered by the CPU), four clock cycles are necessary to fetch the corresponding data (1+1+2). This leads to the following sequence of clock cycles between the delivery of subsequent data words: 4 - 2 - 2 - 2 - 2 - … In the case that the CPU requests another address than the one proposed by the prefetcher (e.g. in case of a jump), the Flash address unit immediately changes to the new address and begins a new sequence (4 - …). Note: If the Flash access phase takes more than two cycles (more than 1 WS), prefetch accesses make no sense, so the Flash prefetching mechanism is disabled. User’s Manual Memory_X73, V2.1 3-40 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10.2 Program SRAM Access The internal functional structure of the interface between the PMU/PMI and the PSRAM is shown in Figure 3-10. The access is done in two phases: • • The PSRAM module delivers the accessed data within one cycle. Waitstates can be selected to emulate accesses to a Flash memory. The PMU requires one additional clock cycle. One more clock cycle is inserted if either Flash timing is selected (bit WSRAM in register IMBCTRL), or the accesses use the EPSRAM area. The CPU receives requested data after the following access delays: • • • 1+1 cycles with standard PSRAM timing 1+2 cycles with EPSRAM timing (using the emulation PSRAM area) 1+WS+2 cycles with Flash timing (WSRAM = 1). However, this delay only becomes effective for an isolated access (read from a nonlinear address). A prefetching mechanism overlaps phase 1 of a subsequent access with phase 2 of the previous access, so the sustained performance for linear accesses (e.g. code fetches) is considerably higher. PMU+ PMI RAM address access phase 1 PSRAM Array CPU address CPU PMU+ PMI ROM data CPU data access phase 2 PSRAM_PMI_structure.vsd Figure 3-10 PSRAM - PMI Structure User’s Manual Memory_X73, V2.1 3-41 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10.3 IMB Control Functions Wait State Generation The generation of wait states is handled by a wait state unit, which indicates when the requested data (or instruction word) is available. The address window for the Flash memory starts at the address C0’0000H and selects the address range of 2 Mbytes. The reset value defines a two cycle Flash memory access. The program SRAM is accessed with a one cycle read. IMB Control Register Register IMBCTR contains the bitfields controlling the wait state generation for the Flash memory and the other IMB memory blocks. One wait state represents one clock cycle. The wait states have to be introduced in order to adapt the memory access time in clock cycles (depending on the clock frequency) to the Flash access time. This register is protected against undesired modification by the register security mechanism. This register is only reset by a hardware reset, a SW reset or a WDT reset do not change the bits. IMBCTR IMB Control Register 15 14 13 12 ESFR (F0FEH/7FH) 11 10 9 8 7 6 Reset Value: xx01H 5 4 3 2 1 0 RPA - DDF DCF - WS RAM WS FLASH rh - rwh - rw rw Field Bits Type Description RPA 15 rh Read Protection Activated This bit monitors the status of the Flash-internal read protection. 0 The Flash-internal read protection is not activated. Bits DCF, DDF are not taken into account. 1 The Flash-internal read protection is activated. Bits DCF, DDF are taken into account. User’s Manual Memory_X73, V2.1 rwh 3-42 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Type Description DDF 9 rwh Disable Data Read from Flash Memory This bit enables/disables the data read access from the internal Flash memory area. Once set, this bit can only be cleared by a HW reset. 0 The data read access from the Flash memory area is allowed. 1 The data read access from the Flash memory area is not allowed. This bit is not taken into account while RPA = 0. DCF 8 rwh Disable Code Fetch from Flash Memory This bit enables/disables the code fetch from the internal Flash memory area. Once set, this bit can only be cleared by a HW reset. 0 The code fetch from the Flash memory area is allowed. 1 The code fetch from the Flash memory area is not allowed. This bit is not taken into account while RPA = 0. WSRAM 2 rw Wait State Control for Program RAM Access1) This bit defines the behavior of a memory in the program SRAM area in the IMB for a read access. This memory area is located in the address range from E0’0000H to F7’FFFFH. The write access to this memory area is always handled within one clock cycle for the memory. 0 The program SRAM area is accessed with the maximum access speed, which is a single cycle read access. 1 The program SRAM behaves exactly like the memory located in the Flash memory area. The pipelined structure and the access time are taken into account to rebuild the identical behavior. User’s Manual Memory_X73, V2.1 3-43 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Type Description WSFLASH [1:0] rw Wait States for the Flash Memory This bitfield defines the number of additional wait states, which are added for a read access from the Flash memory area, which is located in the address range from C0’0000H to DF’FFFFH. 00 No additional wait state is introduced for the Flash read access. This corresponds to a Flash read access in one clock cycle. 01 One additional wait state is introduced for the Flash read access. This corresponds to a Flash read access in two clock cycles. (default) 10 Two additional wait states are introduced for the Flash read access. This corresponds to a Flash read access in three clock cycles. 11 Three additional wait states are introduced for the Flash read access. This corresponds to a Flash read access in four clock cycles. 1) WSRAM selects the access timing for the standard PSRAM area. The access timing for the EPSRAM area (F8’0000H) equals Flash accesses with WSFLASH = 00B. User’s Manual Memory_X73, V2.1 3-44 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4 Central Processing Unit (CPU) Basic tasks of the Central Processing Unit (CPU) are to fetch and decode instructions, to supply operands for the Arithmetic and Logic unit (ALU) and the Multiply and Accumulate unit (MAC), to perform operations on these operands in the ALU and MAC, and to store the previously calculated results. As the CPU is the main engine of the XC167 microcontroller, it is also affected by certain actions of the peripheral subsystem. Because a five-stage processing pipeline (plus 2-stage fetch pipeline) is implemented in the XC167, up to five instructions can be processed in parallel. Most instructions of the XC167 are executed in one single clock cycle due to this parallelism. This chapter describes how the pipeline works for sequential and branch instructions in general, and the hardware provisions which have been made to speed up execution of jump instructions in particular. General instruction timing is described, including standard timing, as well as exceptions. While internal memory accesses are normally performed by the CPU itself, external peripheral or memory accesses are performed by a particular on-chip External Bus Controller (EBC) which is invoked automatically by the CPU whenever a code or data address refers to the external address space. Whenever possible, the CPU continues operating while an external memory access is in progress. If external data are required but are not yet available, or if a new external memory access is requested by the CPU before a previous access has been completed, the CPU will be held by the EBC until the request can be satisfied. The EBC is described in a separate chapter. The on-chip peripheral units of the XC167 work nearly independently of the CPU with a separate clock generator. Data and control information are interchanged between the CPU and these peripherals via Special Function Registers (SFRs). Whenever peripherals need a non-deterministic CPU action, an on-chip Interrupt Controller compares all pending peripheral service requests against each other and prioritizes one of them. If the priority of the current CPU operation is lower than the priority of the selected peripheral request, an interrupt will occur. There are two basic types of interrupt processing: • • Standard interrupt processing forces the CPU to save the current program status and return address on the stack before branching to the interrupt vector jump table. PEC interrupt processing steals only one machine cycle from the current CPU activity to perform a single data transfer via the on-chip Peripheral Event Controller (PEC). System errors detected during program execution (hardware traps) and external nonmaskable interrupts are also processed as standard interrupts with a very high priority. In contrast to other on-chip peripherals, there is a closer conjunction between the watchdog timer and the CPU. If enabled, the watchdog timer expects to be serviced by User’s Manual CPUSV2_X, V2.2 4-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) the CPU within a programmable period of time, otherwise it will reset the chip. Thus, the watchdog timer is able to prevent the CPU from going astray when executing erroneous code. After reset, the watchdog timer starts counting automatically but, it can be disabled via software, if desired. In addition to its normal operation state, the CPU has the following particular states: • • • • Reset state: Any reset (hardware, software, watchdog) forces the CPU into a predefined active state. IDLE state: The clock signal to the CPU itself is switched off, while the clocks for the on-chip peripherals keep running. SLEEP state: All of the on-chip clocks are switched off (RTC clock selectable), external interrupt inputs are enabled. POWER DOWN state: All of the on-chip clocks are switched off (RTC clock selectable), all inputs are disregarded. Transition to an active CPU state is forced by an interrupt (if in IDLE or SLEEP mode) or by a reset (if in POWER DOWN mode). The IDLE, SLEEP, POWER DOWN, and RESET states can be entered by specific XC167 system control instructions. A set of Special Function Registers is dedicated to the CPU core (CSFRs): • • • • • • • • • • CPU Status Indication and Control: PSW, CPUCON1, CPUCON2 Code Access Control: IP, CSP Data Paging Control: DPP0, DPP1, DPP2, DPP3 Global GPRs Access Control: CP System Stack Access Control: SP, SPSEG, STKUN, STKOV Multiply and Divide Support: MDL, MDH, MDC Indirect Addressing Offset: QR0, QR1, QX0, QX1 MAC Address Pointers: IDX0, IDX1 MAC Status Indication and Control: MCW, MSW, MAH, MAL, MRW ALU Constants Support: ZEROS, ONES The CPU also uses CSFRs to access the General Purpose Registers (GPRs). Since all CSFRs can be controlled by any instruction capable of addressing the SFR/CSFR memory space, there is no need for special system control instructions. However, to ensure proper processor operation, certain restrictions on the user access to some CSFRs must be imposed. For example, the instruction pointer (CSP, IP) cannot be accessed directly at all. These registers can only be changed indirectly via branch instructions. Registers PSW, SP, and MDC can be modified not only explicitly by the programmer, but also implicitly by the CPU during normal instruction processing. Note: Note that any explicit write request (via software) to an CSFR supersedes a simultaneous modification by hardware of the same register. User’s Manual CPUSV2_X, V2.2 4-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) All CSFRs may be accessed wordwise, or bytewise (some of them even bitwise). Reading bytes from word CSFRs is a non-critical operation. Any write operation to a single byte of a CSFR clears the non-addressed complementary byte within the specified CSFR. Attention: Reserved CSFR bits must not be modified explicitly, and will always supply a read value of 0. If a byte/word access is preferred by the programmer or is the only possible access the reserved CSFR bits must be written with 0 to provide compatibility with future versions. User’s Manual CPUSV2_X, V2.2 4-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.1 Components of the CPU The high performance of the CPU results from the cooperation of several units which are optimized for their respective tasks (see Figure 4-1). Prefetch Unit and Branch Unit feed the pipeline minimizing CPU stalls due to instruction reads. The Address Unit supports sophisticated addressing modes avoiding additional instructions needed otherwise. Arithmetic and Logic Unit and Multiply and Accumulate Unit handle differently sized data and execute complex operations. Three memory interfaces and Write Buffer minimize CPU stalls due to data transfers. PSRAM Flash/RO M PM U CPU P refetch U nit CSP IP VECSEG CPUCON1 CPUCON2 B ranch U nit R etu rn S tack FIFO ID X 0 ID X 1 QX0 QX1 QR0 QR1 + /- + /- M u ltip ly U nit MRW + /- MCW MSW MAH MAL 2-S tage P refetch P ipeline TF R Injection/ Exception Handler 5-S tage P ipeline IFU DPRAM IPIP DPP0 DPP1 DPP2 DPP3 SPSEG SP STKO V STKUN ADU D ivisio n U n it B it-M a sk-G e n . M u ltip ly U n it B a rre l-S h ifte r MDC CP RR1515 RR1414 R 15 R 14 R 15 R 14 GPRs GPRs GPRs GPRs RR1 1 RR0 0R 1 R0 R1 R0 RF PSW + /- MDH MDL ZE R O S ONES M AC B uffer ALU WB DSRAM EBC Peripherals DM U m ca04917_x.vsd Figure 4-1 CPU Block Diagram User’s Manual CPUSV2_X, V2.2 4-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) In general the instructions move through 7 pipeline stages, where each stage processes its individual task (see Section 4.3 for a summary): • • the 2-stage fetch pipeline prefetches instructions from program memory and stores them into an instruction FIFO the 5-stage processing pipeline executes each instruction stored in the instruction FIFO Because passing through one pipeline stage takes at least one clock cycle, any isolated instruction takes at least five clock cycles to be completed. Pipelining, however, allows parallel (i.e. simultaneous) processing of up to five instructions (with branches up to six instructions). Therefore, most of the instructions appear to be processed during one clock cycle as soon as the pipeline has been filled once after reset. The pipelining increases the average instruction throughput considered over a certain period of time. 4.2 Instruction Fetch and Program Flow Control The Instruction Fetch Unit (IFU) prefetches and preprocesses instructions to provide a continuous instruction flow. The IFU can fetch simultaneously at least two instructions via a 64-bit wide bus from the Program Management Unit (PMU). The prefetched instructions are stored in an instruction FIFO. Preprocessing of branch instructions enables the instruction flow to be predicted. While the CPU is in the process of executing an instruction fetched from the FIFO, the prefetcher of the IFU starts to fetch a new instruction at a predicted target address from the PMU. The latency time of this access is hidden by the execution of the instructions which have already been buffered in the FIFO. Even for a non-sequential instruction execution, the IFU can generally provide a continuous instruction flow. The IFU contains two pipeline stages: the Prefetch Stage and the Fetch Stage. During the prefetch stage, the Branch Detection and Prediction Logic analyzes up to three prefetched instructions stored in the first Instruction Buffer (can hold up to six instructions). If a branch is detected, then the IFU starts to fetch the next instructions from the PMU according to the prediction rules. After having been analyzed, up to three instructions are stored in the second Instruction Buffer (can hold up to three instructions) which is the input register of the Fetch Stage. In the case of an incorrectly predicted instruction flow, the instruction fetch pipeline is bypassed to reduce the number of dead cycles. User’s Manual CPUSV2_X, V2.2 4-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 24-bit Address 64-bit Data IFU Control IFU Pipeline Instruction Buffer (up to 6 Instr.) CSP +/- IP Branch Detection and Prediction Logic Prefetch Stage Return Stack CPUCON1 CPUCON2 Control Registers Instruction FIFO Injection and Exception Handler VECSEG TFR Bypass Fetch to Decode Branch Folding Unit Bypass Prefetch to Decode Instruction Buffer (up to 3 Instr.) Fetch Stage Instruction Buffer (up to 1 Instr.) Decode Stage MCA05501 Figure 4-2 IFU Block Diagram On the Fetch Stage, the prefetched instructions are stored in the instruction FIFO. The Branch Folding Unit (BFU) allows processing of branch instructions in parallel with preceding instructions. To achieve this the BFU preprocesses and reformats the branch instruction. First, the BFU defines (calculates) the absolute target address. This address — after being combined with branch condition and branch attribute bits — is stored in the same FIFO step as the preceding instruction. The target address is also used to prefetch the next instructions. For the Processing Pipeline, both instructions are fetched from the FIFO again and are executed in parallel. If the instruction flow was predicted incorrectly (or FIFO is empty), the two stages of the IFU can be bypassed. Note: Pipeline behavior in case of a incorrectly predicted instruction flow is described in the following sections. User’s Manual CPUSV2_X, V2.2 4-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.2.1 Branch Detection and Branch Prediction Rules The Branch Detection Unit preprocesses instructions and classifies detected branches. Depending on the branch class, the Branch Prediction Unit predicts the program flow using the following rules: Table 4-1 Branch Classes and Prediction Rules Branch Instruction Classes Instructions Prediction Rule (Assumption) Inter-segment branch instructions JMPS seg, caddr CALLS seg, caddr The branch is always taken Branch instructions with user programmable branch prediction JMPA- xcc, caddr JMPA+ xcc, caddr CALLA- xcc, caddr CALLA+ xcc, caddr User-specified1) via bit 8 (‘a’) of the instruction long word: …+: branch ‘taken’ (a = 0) …-: branch ‘not taken’ (a = 1) Indirect branch instructions JMPI cc, [Rw] CALLI cc, [Rw] Unconditional: branch ‘taken’ Conditional: ‘not taken’ Relative branch instructions with condition code JMPR cc, rel Unconditional or backward: branch ‘taken’ Conditional forward: ‘not taken’ Relative branch instructions without condition code CALLR rel The branch is always taken Branch instructions with bitcondition JB(C) bitaddr, rel JNB(S) bitaddr, rel Backward: branch ‘taken’ Forward: ‘not taken’ Return instructions RET, RETP RETS, RETI The branch is always taken 1) This bit can be also set/cleared automatically by the Assembler for generic JMPA and CALLA instructions depending on the jump condition (condition is cc_Z: ‘not taken’, otherwise: ‘taken’). 4.2.2 Correctly Predicted Instruction Flow Table 4-2 shows the continuous execution of instructions, assuming a 0-waitstate1) program memory. In this example, most of the instructions are executed in one CPU cycle while instruction In+6 takes two CPU cycles (general example for multicycle instructions). The diagram shows the sequential instruction flow through the different pipeline stages. Figure 4-3 shows the corresponding program memory section. The instructions for the processing pipeline are fetched from the Instruction FIFO while the IFU prefetches the next instructions to fill the FIFO. As long as the instruction flow is correctly predicted by the IFU, both processes are independent. 1) For the exact Flash memory access timing and the required waitstates please refer to Section 3.10.1. User’s Manual CPUSV2_X, V2.2 4-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) In this example with a fast Internal Program Memory, the Prefetcher is able to fetch more instructions than the processing pipeline can execute. In Tn+4, the FIFO and prefetch buffer are filled and no further instructions can be prefetched. The PMU address stays stable (Tn+4) until a whole 64-bit double word can be buffered (Tn+7) in the 96-bit prefetch buffer again. Table 4-2 Correctly Predicted Instruction Flow (Sequential Execution) Tn Tn+1 Tn+2 Tn+3 Tn+4 Tn+5 Tn+6 Tn+7 Tn+8 Ia+16 Ia+24 Ia+32 Ia+40 Ia+40 Ia+40 Ia+40 Ia+48 Ia+48 PMU Data 64bit Id+1 Id+2 Id+3 Id+4 Id+5 Id+5 Id+5 Id+5 Id+7 PREFETCH 96-bit Buffer In+6 … In+9 In+9 … In+11 In+12 In+13 In+14 In+15 In+15 … In+19 In+15 … In+19 In+16 … In+19 In+17 … In+19 In+18 … In+21 FETCH Instruction Buffer In+5 In+6 In+7 In+8 In+9 In+10 In+11 In+12 In+13 In+14 – In+15 In+16 In+17 FIFO contents In+3 … In+5 In+4 … In+8 In+5 … In+11 In+6 … In+13 In+7 … In+14 In+7 … In+14 In+8 … In+15 In+9 … In+16 In+10 … In+17 Fetch from FIFO In+4 In+5 In+6 In+7 In+7 In+8 In+9 In+10 In+11 DECODE In+3 In+4 In+5 In+6 In+6 In+7 In+8 In+9 In+10 ADDRESS In+2 In+3 In+4 In+5 In+6 In+6 In+7 In+8 In+9 MEMORY In+1 In+2 In+3 In+4 In+5 In+6 In+6 In+7 In+8 EXECUTE In In+1 In+2 In+3 In+4 In+5 In+6 In+6 In+7 WRITE BACK – In In+1 In+2 In+3 In+4 In+5 In+6 In+6 PMU Address User’s Manual CPUSV2_X, V2.2 4-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) In+21 In+21 In+20 In+20 In+19 In+18 In+17 In+16 In+16 In+15 In+15 In+14 In+14 In+13 In+12 In+12 In+11 In+11 In+10 In+10 In+9 In+8 In+7 In+6 Ia+40 Ia+32 Ia+24 Ia+16 Ia+8 Ia MCA04918 Figure 4-3 4.2.3 Program Memory Section for Correctly Predicted Flow Incorrectly Predicted Instruction Flow If the CPU detects that the IFU made an incorrect prediction of the instruction flow, then the pipeline stages and the Instruction FIFO containing the wrong prefetched instructions are canceled. The entire instruction fetch is restarted at the correct point of the program. Table 4-3 shows the restarted execution of instructions, assuming a 0-waitstate program memory. Figure 4-4 shows the corresponding program memory section. During the cycle Tn, the CPU detects an incorrectly prediction case which leads to a canceling of the pipeline. The new address is transferred to the PMU in Tn+1 which delivers the first data in the next cycle Tn+2. But, the target instruction crosses the 64-bit memory boundary and a second fetch in Tn+3 is required to get the entire 32-bit instruction. In Tn+4, the Prefetch Buffer contains two 32-bit instructions while the first instruction Im is directly forwarded to the Decode stage. The prefetcher is now restarted and prefetches further instructions. In Tn+5, the instruction Im+1 is forwarded from the Fetch Instruction Buffer directly to the Decode stage as well. The Fetch row shows all instructions in the Fetch Instruction Buffer and the instructions fetched from the Instruction FIFO. The instruction Im+3 is the first instruction fetched from the FIFO during Tn+6. During the same cycle, instruction Im+2 was still forwarded from the Fetch Instruction Buffer to the Decode stage. User’s Manual CPUSV2_X, V2.2 4-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 4-3 Incorrectly Predicted Instruction Flow (Restarted Execution) Tn Tn+1 Tn+2 Tn+3 Tn+4 Tn+5 Tn+6 Tn+7 Tn+8 I… Ia Ia+8 Ia+16 Ia+24 I… I… I… I… PMU Data 64bit I… – Id Id+1 Id+2 Id+3 I… I… I… PREFETCH 96-bit Buffer I… – – – Im Im+1 Im+2 Im+3 Im+4 Im+5 I… I… FETCH Instruction Buffer Inext+2 – – – – Im+1 Im+2 Im+3 Im+4 Im+5 I… Fetch from FIFO – – – – – – Im+3 Im+4 Im+5 DECODE Inext+1 – – – Im Im+1 Im+2 Im+3 Im+4 ADDRESS Inext – – – – Im Im+1 Im+2 Im+3 MEMORY Ibranch – – – – – Im Im+1 Im+2 EXECUTE In Ibranch – – – – – Im Im+1 WRITE BACK – In Ibranch – – – – – Im PMU Address I... I... Im+5 Im+5 Im+4 Im+4 Im+3 Im+3 Im+2 Im+2 Im+1 Im+1 Im Im I... I... Ia+24 Ia+16 Ia+8 Ia MCA04919 64-bit wide Program Memory with four 16 bit packages Figure 4-4 Program Memory Section for Incorrectly Predicted Flow User’s Manual CPUSV2_X, V2.2 4-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.3 Instruction Processing Pipeline The XC167 uses five pipeline stages to execute an instruction. All instructions pass through each of the five stages of the instruction processing pipeline. The pipeline stages are listed here together with the 2 stages of the fetch pipeline: 1st -> PREFETCH: This stage prefetches instructions from the PMU in the predicted order. The instructions are preprocessed in the branch detection unit to detect branches. The prediction logic decides if the branches are assumed to be taken or not. 2nd -> FETCH: The instruction pointer of the next instruction to be fetched is calculated according to the branch prediction rules. For zero-cycle branch execution, the Branch Folding Unit preprocesses and combines detected branches with the preceding instructions. Prefetched instructions are stored in the instruction FIFO. At the same time, instructions are transported out of the instruction FIFO to be executed in the instruction processing pipeline. 3rd -> DECODE: The instructions are decoded and, if required, the register file is accessed to read the GPR used in indirect addressing modes. 4th -> ADDRESS: All the operand addresses are calculated. Register SP is decremented or incremented for all instructions which implicitly access the system stack. 5th -> MEMORY: All the required operands are fetched. 6th -> EXECUTE: An ALU or MAC-Unit operation is performed on the previously fetched operands. The condition flags are updated. All explicit write operations to CPU-SFRs and all auto-increment/auto-decrement operations of GPRs used as indirect address pointers are performed. 7th -> WRITE BACK: All external operands and the remaining operands within the internal DPRAM space are written back. Operands located in the internal SRAM are buffered in the Write Back Buffer. Specific so-called injected instructions are generated internally to provide the time needed to process instructions requiring more than one CPU cycle for processing. They are automatically injected into the decode stage of the pipeline, then they pass through the remaining stages like every standard instruction. Program interrupt, PEC transfer, and OCE operations are also performed by means of injected instructions. Although these internally injected instructions will not be noticed in reality, they help to explain the operation of the pipeline. The performance of the CPU (pipeline) is decreased by bandwidth limitations (same resource is accessed by different stages) and data dependencies between instructions. The XC167’s CPU has dedicated hardware to detect and to resolve different kinds of dependencies. Some of those dependencies are described in the following section. Because up to five different instructions are processed simultaneously, additional hardware has been dedicated to deal with dependencies which may exist between instructions in different pipeline stages. This extra hardware supports ‘forwarding’ of the User’s Manual CPUSV2_X, V2.2 4-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) operand read and write values and resolves most of the possible conflicts — such as multiple usage of buses — in a time optimized way without performance loss. This makes the pipeline unnoticeable for the user in most cases. However, there are some rare cases in which the pipeline requires attention by the programmer. In these cases, the delays caused by the pipeline conflicts can be used for other instructions to optimize performance. Note: The XC167 has a fully interlocked pipeline, which means that these conflicts do not cause any malfunction. Instruction re-ordering is only required for performance reasons. The following examples describe the pipeline behavior in special cases and give principle rules to improve the performance by re-ordering the execution of instructions. User’s Manual CPUSV2_X, V2.2 4-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.3.1 Pipeline Conflicts Using General Purpose Registers The GPRs are the working registers of the CPU and there are a lot of possible dependencies between instructions using GPRs. A high-speed five-port register file prevents bandwidth conflicts. Dedicated hardware is implemented to detect and resolve the data dependencies. Special forwarding busses are used to forward GPR values from one pipeline stage to another. In most cases, this allows the execution of instructions without any delay despite of data dependencies. Conflict_GPRs_Resolved: In ADD R0,R1 ;Compute new value for R0 In+1 ADD R3,R0 ;Use R0 again In+2 ADD R6,R0 ;Use R0 again In+3 ADD R6,R1 ;Use R6 again In+4 ... Table 4-4 Resolved Pipeline Dependencies Using GPRs Tn+31) Tn+42) Stage Tn Tn+1 DECODE In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = ADD In+4 R3, R0 R6, R0 R6, R1 Tn+2 Tn+53) In+5 ADDRESS In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = ADD In+4 R3, R0 R6, R0 R6, R1 MEMORY In-2 In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = ADD R3, R0 R6, R0 R6, R1 EXECUTE In-3 In-2 In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD R3, R0 R6, R0 WR.BACK In-4 In-3 In-2 In-1 In = ADD R0, R1 In+1 = ADD R3, R0 1) R0 forwarded from EXECUTE to MEMORY. 2) R0 forwarded from WRITE BACK to MEMORY. 3) R6 forwarded from EXECUTE to MEMORY. User’s Manual CPUSV2_X, V2.2 4-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) However, if a GPR is used for indirect addressing the address pointer (i.e. the GPR) will be required already in the DECODE stage. In this case the instruction is stalled in the address stage until the operation in the ALU is executed and the result is forwarded to the address stage. Conflict_GPRs_Pointer_Stall: In ADD R0,R1 ;Compute new value for R0 ;Use R0 as address pointer In+1 MOV R3,[R0] In+2 ADD R6,R0 In+3 ADD R6,R1 In+4 ... Table 4-5 Pipeline Dependencies Using GPRs as Pointers (Stall) Tn+21) Stage Tn Tn+1 DECODE In = ADD R0, R1 In+1 = MOV In+2 R3, [R0] Tn+32) Tn+4 Tn+5 In+2 In+2 In+3 ADDRESS In-1 In = ADD R0, R1 In+1 = MOV In+1 = MOV In+1 = MOV In+2 R3, [R0] R3, [R0] R3, [R0] MEMORY In-2 In-1 In = ADD R0, R1 – – In+1 = MOV R3, [R0] EXECUTE In-3 In-2 In-1 In = ADD R0, R1 – – WR.BACK In-4 In-3 In-2 In-1 In = ADD R0, R1 – 1) New value of R0 not yet available. 2) R0 forwarded from EXECUTE to ADDRESS (next cycle). User’s Manual CPUSV2_X, V2.2 4-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) To avoid these stalls, one multicycle instruction or two single cycle instructions may be inserted. These instructions must not update the GPR used for indirect addressing. Conflict_GPRs_Pointer_NoStall: In ADD R0,R1 ;Compute new value for R0 In+1 ADD R6,R0 ;R0 is not updated, just read In+2 ADD R6,R1 ;Use R0 as address pointer In+3 MOV R3,[R0] In+4 ... Table 4-6 Pipeline Dependencies Using GPRs as Pointers (No Stall) Tn+31) Stage Tn Tn+1 DECODE In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = MOV In+4 R6, R0 R6, R1 R3, [R0] Tn+2 Tn+4 Tn+5 In+5 ADDRESS In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = MOV In+4 R6, R0 R6, R1 R3, [R0] MEMORY In-2 In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = MOV R6, R0 R6, R1 R3, [R0] EXECUTE In-3 In-2 In-1 In = ADD R0, R1 In+1 = ADD In+2 = ADD R6, R0 R6, R1 WR.BACK In-4 In-3 In-2 In-1 In = ADD R0, R1 In+1 = ADD R6, R0 1) R0 forwarded from EXECUTE to ADDRESS (next cycle). 4.3.2 Pipeline Conflicts Using Indirect Addressing Modes In the case of read accesses using indirect addressing modes, the Address Generation Unit uses a speculative addressing mechanism. The read data path to one of the different memory areas (DPRAM, DSRAM, etc.) is selected according to a history table before the address is decoded. This history table has one entry for each of the GPRs. The entries store the information of the last accessed memory area using the corresponding GPR. In the case of an incorrect prediction of the memory area, the read access must be restarted. It is recommended that the GPRs used for indirect addressing always point to the same memory area. If an updated GPR points to a different memory area, the next read operation will access the wrong memory area. The read access must be repeated, which leads to pipeline stalls. User’s Manual CPUSV2_X, V2.2 4-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Conflict_GPRs_Pointer_WrongHistory: In ADD R3,[R0] ;R0 points to DPRAM (e.g.) In+1 MOV R0,R4 ... Ii MOV DPPX, ... ;change DPPx ... Im ADD R6,[R0] ;R0 now points to SRAM (e.g.) Im+1 MOV R6,R1 Im+2 ... Table 4-7 Pipeline Dependencies with Pointers (Valid Speculation) Stage Tn Tn+1 DECODE In = ADD R3, [R0] In+1 = MOV In+2 R0, R4 Tn+2 Tn+3 Tn+4 Tn+5 In+3 In+4 In+5 In+3 In+4 In+3 ADDRESS In-1 In = ADD R3, [R0] In+1 = MOV In+2 R0, R4 MEMORY In-2 In-1 In = ADD R3, [R0] In+1 = MOV In+2 R0, R4 EXECUTE In-3 In-2 In-1 In = ADD R3, [R0] In+1 = MOV In+2 R0, R4 WR.BACK In-4 In-3 In-2 In-1 In = ADD R3, [R0] Table 4-8 In+1 = MOV R0, R4 Pipeline Dependencies with Pointers (Invalid Speculation) Tm+1 Tm+21) Tm+3 Stage Tm DECODE Im = ADD Im+1 = MOV Im+1 = MOV Im+2 R6, [R0] R6, R1 R6, R1 Tm+4 Tm+5 Im+3 Im+4 Im+3 ADDRESS Im-1 Im = ADD R6, [R0] Im = ADD R6, [R0] Im+1 = MOV Im+2 R6, R1 MEMORY Im-2 Im-1 – Im = ADD R6, [R0] Im+1 = MOV Im+2 R6, R1 EXECUTE Im-3 Im-2 Im-1 – Im = ADD R6, [R0] Im+1 = MOV R6, R1 WR.BACK Im-4 Im-3 Im-2 Im-1 – Im = ADD R6, [R0] 1) Access to location [R0] must be repeated due to wrong history (target area was changed). User’s Manual CPUSV2_X, V2.2 4-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.3.3 Pipeline Conflicts Due to Memory Bandwidth Memory bandwidth conflicts can occur if instructions in the pipeline access the same memory area at the same time. Special access mechanisms are implemented to minimize conflicts. The DPRAM of the CPU has two independent read/write ports; this allows parallel read and write operation without delays. Write accesses to the DSRAM can be buffered in a Write Back Buffer until read accesses are finished. All instructions except the CoXXX instructions can read only one memory operand per cycle. A conflict between the read and one write access cannot occur because the DPRAM has two independent read/write ports. Only other pipeline stall conditions can generate a DPRAM bandwidth conflict. The DPRAM is a synchronous pipelined memory. The read access starts with the valid addresses on the address stage. The data are delivered in the Memory stage. If a memory read access is stalled in the Memory stage and the following instruction on the Address stage tries to start a memory read, the new read access must be delayed as well. But, this conflict is hidden by an already existing stall of the pipeline. User’s Manual CPUSV2_X, V2.2 4-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The CoXXX instructions are the only instructions able to read two memory operands per cycle. A conflict between the two read and one pending write access can occur if all three operands are located in the DPRAM area. This is especially important for performance in the case of executing a filter routine. One of the operands should be located in the DSRAM to guarantee a single-cycle execution of the CoXXX instructions. Conflict_DPRAM_Bandwidth: In ADD op1,R1 In+1 ADD R6,R0 In+2 CoMAC [IDX0],[R0] In+3 MOV R3,[R0] In+4 ... Table 4-9 Pipeline Dependencies in Case of Memory Conflicts (DPRAM) Tn+41) Stage Tn Tn+1 DECODE In = ADD op1, R1 In+1 = ADD In+2 = In+3 = MOV In+4 R6, R0 CoMAC … R3, [R0] Tn+2 Tn+3 Tn+5 In+4 ADDRESS In-1 In = ADD op1, R1 In+1 = ADD In+2 = In+3 = MOV In+3 = MOV R6, R0 CoMAC … R3, [R0] R3, [R0] MEMORY In-2 In-1 In = ADD op1, R1 In+1 = ADD In+2 = In+2 = R6, R0 CoMAC … CoMAC … EXECUTE In-3 In-2 In-1 In = ADD op1, R1 In+1 = ADD – R6, R0 WR.BACK In-4 In-3 In-2 In-1 In = ADD op1, R1 In+1 = ADD R6, R0 1) COMAC instruction stalls due to memory bandwidth conflict. User’s Manual CPUSV2_X, V2.2 4-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The DSRAM is a single-port memory with one read/write port. To reduce the number of bandwidth conflict cases, a Write Back Buffer is implemented. It has three data entries. Only if the buffer is filled and a read access and a write access occur at the same time, must the read access be stalled while one of the buffer entries is written back. Conflict_DSRAM_Bandwidth: In ADD op1,R1 In+1 ADD R6,R0 In+2 ADD R6,op2 In+3 MOV R3,R2 In+4 ... Table 4-10 Pipeline Dependencies in Case of Memory Conflicts (DSRAM) Tn+2 Tn+3 Tn+41) Stage Tn Tn+1 DECODE In = ADD op1, R1 In+1 = ADD In+2 = ADD In+3 = MOV In+4 R6, R0 R6, op2 R3, R2 Tn+5 In+4 ADDRESS In-1 In = ADD op1, R1 In+1 = ADD In+2 = ADD In+3 = MOV In+3 = MOV R6, R0 R6, op2 R3, R2 R3, R2 MEMORY In-2 In-1 In = ADD op1, R1 In+1 = ADD In+2 = ADD In+2 = ADD R6, R0 R6, op2 R6, op2 EXECUTE In-3 In-2 In-1 In = ADD op1, R1 In+1 = ADD – R6, R0 WR.BACK In-4 In-3 In-2 In-1 In = ADD op1, R1 In+1 = ADD R6, R0 WB.Buffer full full full full full full 1) ADD R6, op2 instruction stalls due to memory bandwidth conflict. User’s Manual CPUSV2_X, V2.2 4-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.3.4 Pipeline Conflicts Caused by CPU-SFR Updates CPU-SFRs control the CPU functionality and behavior. Changes and updates of CSFRs influence the instruction flow in the pipeline. Therefore, special care is required to ensure that instructions in the pipeline always work with the correct CSFR values. CSFRs are updated late on the EXECUTE stage of the pipeline. Meanwhile, without conflict detection, the instructions in the DECODE, ADDRESS, and MEMORY stages would still work without updated register values. The CPU detects conflict cases and stalls the pipeline to guarantee a correct execution. For performance reasons, the CPU differentiates between different classes of CPU-SFRs. The flow of instructions through the pipeline can be improved by following the given rules used for instruction re-ordering. There are three classes of CPU-SFRs: • • • CSFRs not generating pipeline conflicts (ONES, ZEROS, MCW) CSFR result registers updated late in the EXECUTE stage, causing one stall cycle CSFRs affecting the whole CPU or the pipeline, causing canceling CSFR Result Registers The CSFR result registers MDH, MDL, MSW, MAH, MAL, and MRW of the ALU and MAC-Unit are updated late in the EXECUTE stage of the pipeline. If an instruction (except CoSTORE) accesses explicitly these registers in the memory stage, the value cannot be forwarded. The instruction must be stalled for one cycle on the MEMORY stage. User’s Manual CPUSV2_X, V2.2 4-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Conflict_CSFR_Update_Stall: In MUL R0,R1 In+1 MOV R6,MDL In+2 ADD R6,R1 In+3 MOV R3,[R0] In+4 ... Table 4-11 Pipeline Dependencies with Result CSFRs (Stall) Tn+31) Stage Tn Tn+1 DECODE In = MUL R0, R1 In+1 = MOV In+2 = ADD In+3 = MOV In+3 = MOV In+4 R6, MDL R6, R1 R3, [R0] R3, [R0] Tn+2 Tn+4 Tn+5 ADDRESS In-1 In = MUL R0, R1 In+1 = MOV In+2 = ADD In+2 = ADD In+3 = MOV R6, MDL R6, R1 R6, R1 R3, [R0] MEMORY In-2 In-1 In = MUL R0, R1 In+1 = MOV In+1 = MOV In+2 = ADD R6, MDL R6, MDL R6, R1 EXECUTE In-3 In-2 In-1 In = MUL R0, R1 – In+1 = MOV R6, MDL WR.BACK In-4 In-3 In-2 In-1 In = MUL R0, R1 – 1) Cannot read MDL here. User’s Manual CPUSV2_X, V2.2 4-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) By reordering instructions, the bubble in the pipeline can be filled with an instruction not using this resource. Conflict_CSFR_Update_Resolved: In MUL R0,R1 In+1 MOV R3,[R0] In+2 MOV R6,MDL In+3 ADD R6,R1 In+4 ... Table 4-12 Pipeline Dependencies with Result CSFRs (No Stall) Tn+41) Stage Tn Tn+1 DECODE In = MUL R0, R1 In+1 = MOV In+2 = MOV In+3 = ADD In+4 R3, [R0] R6, MDL R6, R1 Tn+2 Tn+3 Tn+5 In+5 ADDRESS In-1 In = MUL R0, R1 In+1 = MOV In+2 = MOV In+3 = ADD In+4 R3, [R0] R6, MDL R6, R1 MEMORY In-2 In-1 In = MUL R0, R1 In+1 = MOV In+2 = MOV In+3 = ADD R3, [R0] R6, MDL R6, R1 EXECUTE In-3 In-2 In-1 In = MUL R0, R1 In+1 = MOV In+2 = MOV R3, [R0] R6, MDL WR.BACK In-4 In-3 In-2 In-1 In = MUL R0, R1 In+1 = MOV R3, [R0] 1) MDL can be read now, no stall cycle necessary. User’s Manual CPUSV2_X, V2.2 4-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) CSFRs Affecting the Whole CPU Some CSFRs affect the whole CPU or the pipeline before the Memory stage. The CPUSFRs CPUCON1, CP, SP, STKUN, STKOV, VECSEG, TFR, and PSW affect the overall CPU function, while the CPU-SFRs IDX0, IDX1, QX1, QX0, DPP0, DPP1, DPP2, and DPP3 only affect the DECODE, ADDRESS, and MEMORY stage when they are modified explicitly. In this case the pipeline behavior depends on the instruction and addressing mode used to modify the CSFR. In the case of modification of these CSFRs by “POP CSFR” or by instructions using the reg,#data16 addressing mode, a special mechanism is implemented to improve performance during the initialization. For further explanation, the instruction which modifies the CSFR can be called “instruction_modify_CSFR”. This special case is detected in the DECODE stage when the instruction_modify_CSFR enters the processing pipeline. Further on, instructions described in the following list are held in the DECODE stage (all other instructions are not held): • • • • Instructions using long addressing mode (mem) Instructions using indirect addressing modes ([Rw], [Rw+]…), except JMPI and CALLI ENWDT, DISWDT, EINIT All CoXXX instructions If the CPUCON1, CP, SP, STKUN, STKOV, VECSEG, TFR, or the PSW are modified and the instruction_modify_CSFR reaches the EXECUTE stage, the pipeline is canceled. The modification affects the entire pipeline and the instruction prefetch. A clean cancel and restart mechanism is required to guarantee a correct instruction flow. In case of modification of IDX0, IDX1, QX1, QX0, DPP0, DPP1, DPP2, or DPP3 only the DECODE, ADDRESS, and MEMORY stages are affected and the pipeline needs not to be canceled. The modification does not affect the instructions in the ADDRESS, MEMORY stage because they are not using this resource. Other kinds of instructions are held in the DECODE stage until the CSFR is modified. The following example shows a case in which the pipeline is stalled. The instruction “MOV R6, R1” after the “MOV IDX1, #12” instruction which modifies the CSFR will be held in DECODE Stage until the IDX1 register is updated. The next example shows an optimized initialization routine. User’s Manual CPUSV2_X, V2.2 4-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Conflict_Canceling: In MOV IDX1,#12 In+1 MOV R6,mem In+2 ADD R6,R1 In+3 MOV R3,[R0] Table 4-13 Pipeline Dependencies with Control CSFRs (Canceling) Stage Tn Tn+1 Tn+2 Tn+3 Tn+4 Tn+5 DECODE In = MOV IDX1, #12 In+1 = MOV In+1 = MOV In+1 = MOV In+1 = MOV In+2 = ADD R6, mem R6, mem R6, mem R6, mem R6, R1 ADDRESS In-1 In = MOV IDX1, #12 – – – In+1 = MOV R6, mem MEMORY In-2 In-1 In = MOV IDX1, #12 – – – EXECUTE In-3 In-2 In-1 In = MOV IDX1, #12 – – WR.BACK In-4 In-3 In-2 In-1 In = MOV IDX1, #12 – Conflict_Canceling_Optimized: In MOV IDX1,#12 In+1 MOV MAH,#23 In+2 MOV MAL,#25 In+3 MOV R3,#08 In+4 ... Table 4-14 Pipeline Dependencies with Control CSFRs (Optimized) Stage Tn Tn+1 Tn+2 Tn+3 Tn+4 DECODE In = MOV IDX1, #12 In+1 = MOV In+2 = MOV In+3 = MOV In+4 MAH, #23 MAL, #25 R3, #08 Tn+5 In+5 ADDRESS In-1 In = MOV IDX1, #12 In+1 = MOV In+2 = MOV In+3 = MOV In+4 MAH, #23 MAL, #25 R3, #08 MEMORY In-2 In-1 In = MOV IDX1, #12 In+1 = MOV In+2 = MOV In+3 = MOV MAH, #23 MAL, #25 R3, #08 EXECUTE In-3 In-2 In-1 In = MOV IDX1, #12 In+1 = MOV In+2 = MOV MAH, #23 MAL, #25 WR.BACK In-4 In-3 In-2 In-1 In = MOV IDX1, #12 User’s Manual CPUSV2_X, V2.2 4-24 In+1 = MOV MAH, #23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) For all the other instructions that modify this kind of CSFR, a simple stall and cancel mechanism guarantees the correct instruction flow. A possible explicit write-operation to this kind of CSFRs is detected on the MEMORY stage of the pipeline. The following instructions on the ADDRESS and DECODE Stage are stalled. If the instruction reaches the EXECUTE stage, the entire pipeline and the Instruction FIFO of the IFU are canceled. The instruction flow is completely re-started. Conflict_Canceling_Completely: In MOV PSW,R4 In+1 MOV R6,R1 In+2 ADD R6,R1 In+3 MOV R3,[R0] In+4 ... Table 4-15 Pipeline Dependencies with Control CSFRs (Cancel All) Stage Tn+1 DECODE In+1 = MOV In+2 = ADD In+2 = ADD – R6, R1 R6, R1 R6, R1 Tn+2 Tn+3 Tn+4 Tn+5 Tn+6 – In+1 = MOV R6, R1 ADDRESS In = MOV PSW, R4 In+1 = MOV In+1 = MOV – R6, R1 R6, R1 – – MEMORY In = MOV PSW, R4 – – – – EXECUTE In-2 In-1 In = MOV PSW, R4 – – – WR.BACK In-3 In-2 In-1 In = MOV PSW, R4 – – In-1 User’s Manual CPUSV2_X, V2.2 4-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.4 CPU Configuration Registers The CPU configuration registers select a number of general features and behaviors of the XC167’s CPU core. In general, these registers must not be modified by application software (exceptions will be documented, e.g. in an errata sheet). Note: The CPU configuration registers are protected by the register security mechanism after the EINIT instruction has been executed. CPUCON1 CPU Control Register 1 SFR (FE18H/0CH) Reset Value: 0007H 15 14 13 12 11 10 9 8 7 6 5 - - - - - - - - - VECSC - - - - - - - - - rw 4 3 2 1 WDT SGT INTS BP CTL DIS CXT rw rw rw 0 ZCJ rw Field Bits Type Description VECSC [6:5] rw Scaling Factor of Vector Table 00 Space between two vectors is 2 words1) 01 Space between two vectors is 4 words 10 Space between two vectors is 8 words 11 Space between two vectors is 16 words WDTCTL 4 rw Configuration of Watchdog Timer 0 DISWDT executable only until End Of Init2) 1 DISWDT/ENWDT always executable (enhanced WDT mode) SGTDIS 3 rw Segmentation Disable/Enable Control 0 Segmentation enabled 1 Segmentation disabled INTSCXT 2 rw Enable Interruptibility of Switch Context 0 Switch context is not interruptible 1 Switch context is interruptible BP 1 rw Enable Branch Prediction Unit 0 Branch prediction disabled 1 Branch prediction enabled ZCJ 0 rw Enable Zero Cycle Jump Function 0 Zero cycle jump function disabled 1 Zero cycle jump function enabled rw 1) The default value (2 words) is compatible with the vector distance defined in the C166 Family architecture. 2) The DISWDT (executed after EINIT) and ENWDT instructions are internally converted in a NOP instruction. User’s Manual CPUSV2_X, V2.2 4-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) CPUCON2 CPU Control Register 2 15 14 13 12 SFR (FE1AH/0DH) 11 10 FIFODEPTH FIFOFED rw rw 8 7 6 5 4 3 BYP BYP EIO STE OV RET LFIC PF F IAEN N RUN ST rw rw rw rw rw rw rw 2 - 1 0 DAID SL rw rw Field Bits FIFODEPTH [15:12] rw FIFO Depth Configuration 0000 No FIFO (entries) 0001 One FIFO entry … … 1000 Eight FIFO entries 1001 reserved … … 1111 reserved FIFOFED [11:10] rw FIFO Fed Configuration 00 FIFO disabled 01 FIFO filled with up to one instruction per cycle 10 FIFO filled with up to two instructions per cycle 11 FIFO filled with up to three instruction per cycle BYPPF 9 rw Prefetch Bypass Control 0 Bypass path from prefetch to decode disabled 1 Bypass path from prefetch to decode available BYPF 8 rw Fetch Bypass Control 0 Bypass path from fetch to decode disabled 1 Bypass path from fetch to decode available EIOIAEN 7 rw Early IO Injection Acknowledge Enable 0 Injection acknowledge by destructive read not guaranteed 1 Injection acknowledge by destructive read guaranteed STEN 6 rw Stall Instruction Enable (for debug purposes) 0 Stall Instruction disabled 1 Stall Instruction enabled (see example below) LFIC 5 rw Linear Follower Instruction Cache 0 Linear Follower Instruction Cache disabled 1 Linear Follower Instruction Cache enabled User’s Manual CPUSV2_X, V2.2 Type 9 Reset Value: 8FBBH Description 4-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description OVRUN 4 rw Pipeline Control 0 Overrun of pipeline bubbles not allowed 1 Overrun of pipeline bubbles allowed RETST 3 rw Enable Return Stack 0 Return Stack is disabled 1 Return Stack is enabled DAID 1 rw Disable Atomic Injection Deny 0 Injection-requests are denied during Atomic 1 Injection-requests are not denied during Atomic SL 0 rw Enables Short Loop Mode 0 Short loop mode disabled 1 Short loop mode enabled Example for dedicated stall debug instructions: STALLAM da,ha,dm,hm STALLEW de,he,dw,hw ;Opcode: 44 dahadmhm ;Opcode: 45 dehedwhw ;Stalls the corresponding pipeline ;stage after “d” cycles for “h” cycles ;(“d” and “h” are 6-bit values) Note: In general, these registers must not be modified by application software (exceptions will be documented, e.g. in an errata sheet). User’s Manual CPUSV2_X, V2.2 4-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.5 Use of General Purpose Registers The CPU uses several banks of sixteen dedicated registers R0, R1, R2, … R15, called General Purpose Registers (GPRs), which can be accessed in one CPU cycle. The GPRs are the working registers of the arithmetic and logic units and many also serve as address pointers for indirect addressing modes. The register banks are accessed via the 5-port register file providing the high access speed required for the CPU’s performance. The register file is split into three independent physical register banks. There are two types of register banks: • • Two local register banks which are a part of the register file A global register bank which is memory-mapped and cached in the register file Core-RAM Registerfile Global Local AGU Write Port ALU Write Port Memory mapped GPR Bank R15 R0 CP R15 R15 R15 R14 R14 R13 R13 R12 R12 R11 R11 R10 R10 R9 R9 R8 R8 R7 R7 R6 R6 R5 R5 R4 R4 R3 R3 R2 R2 R1 R1 R0 R0 AGU Read Port ALU Read Port 1 ALU Read Port 2 MCD04873 Figure 4-5 Register File User’s Manual CPUSV2_X, V2.2 4-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Bitfield BANK in register PSW selects which of the three physical register banks is activated. The selected bank can be changed explicitly by any instruction which writes to the PSW, or implicitly by a RETI instruction, an interrupt or hardware trap. In case of an interrupt, the selection of the register bank is configured via registers BNKSELx in the Interrupt Controller ITC. Hardware traps always use the global register bank. The local register banks are built of dedicated physical registers, while the global register bank represents a cache. The banks of the memory-mapped GPRs (global bank) are located in the internal DPRAM. One bank uses a block of 16 consecutive words. A Context Pointer (CP) register determines the base address of the current selected bank. To provide the required access speed, the GPRs located in the DPRAM are cached in the 5-port register file (only one memory-mapped GPR bank can be cached at the time). If the global register bank is activated, the cache will be validated before further instructions are executed. After validation, all further accesses to the GPRs are redirected to the global register bank. Internal DPRAM R15 (CP) + 30 R14 (CP) + 28 Register File R13 R12 R11 15 0 16-Bit Context Pointer R10 R15 R9 R8 R7 R0 R6 R5 R4 Global R3 local R2 R1 (CP) + 2 R0 (CP) MCA04921 Figure 4-6 Register Bank Selection via Register CP User’s Manual CPUSV2_X, V2.2 4-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.5.1 GPR Addressing Modes Because the GPRs are the working registers and are accessed frequently, there are three possible ways to access a register bank: • • • Short GPR Address (mnemonic: Rw or Rb) Short Register Address (mnemonic: reg or bitoff) Long Memory Address (mnemonic: mem), for the global bank only Short GPR Addresses specify the register offset within the current register bank (selected via bitfield BANK). Short 4-bit GPR addresses can access all sixteen registers, short 2-bit addresses (used by some instructions) can access the lower four registers. Depending on whether a relative word (Rw) or byte (Rb) GPR address is specified, the short GPR address is either multiplied by two (Rw) or not (Rb) before it is used to physically access the register bank. Thus, both byte and word GPR accesses are possible in this way. Note: GPRs used as indirect address pointers are always accessed wordwise. For the local register banks the resulting offset is used directly, for the global register bank the resulting offset is logically added to the contents of register CP which points to the memory location of the base of the current global register bank (see Figure 4-7). Short 8-Bit Register Addresses within a range from F0H to FFH interpret the four least significant bits as short 4-bit GPR addresses, while the four most significant bits are ignored. The respective physical GPR address is calculated in the same way as for short 4-bit GPR addresses. For single bit GPR accesses, the GPR’s word address is calculated in the same way. The accessed bit position within the word is specified by a separate additional 4-bit value. 12-Bit Context Pointer 11 Specified by reg or bitoff 1 0 1 1 1 1 For byte GPR accesses 4-Bit GPR Address *1 *2 For word GPR accesses Internal DRAM Must be within the internal DPRAM area + GPRs MCA04922 Figure 4-7 Implicit CP Use by Logical Short GPR Addressing Modes User’s Manual CPUSV2_X, V2.2 4-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 24-Bit Memory Addresses can be directly used to access GPRs located in the DPRAM (not applicable for local register banks). In case of a memory read access, a hit detection logic checks if the accessed memory location is cached in the global register bank. In case of a cache hit, an additional global register bank read access is initiated. The data that is read from cache will be used and the data that is read from memory will be discarded. This leads to a delay of one CPU cycle (MOV R4, mem [CP ≤ mem ≤ CP + 31]). In case of a memory write access, the hit detection logic determines a cache hit in advance. Nevertheless, the address conversion needs one additional CPU cycle. The value is directly written into the global register bank without further delay (MOV mem, R4). Note: The 24-bit GPR addressing mode is not recommended because it requires an extra cycle for the read and write access. Table 4-16 Addressing Modes to Access GPRs Word Registers1) Short Address2) Byte Registers Name Mem. Addr.3) Name Mem. Addr.3) 8-Bit 4-Bit 2-Bit R0 (CP) + 0 RL0 (CP) + 0 F0H 0H 0H R1 (CP) + 2 RH0 (CP) + 1 F1H 1H 1H R2 (CP) + 4 RL1 (CP) + 2 F2H 2H 2H R3 (CP) + 6 RH1 (CP) + 3 F3H 3H 3H R4 (CP) + 8 RL2 (CP) + 4 F4H 4H --- R5 (CP) + 10 RH2 (CP) + 5 F5H 5H --- R6 (CP) + 12 RL3 (CP) + 6 F6H 6H --- R7 (CP) + 14 RH3 (CP) + 7 F7H 7H --- R8 (CP) + 16 RL4 (CP) + 8 F8H 8H --- R9 (CP) + 18 RH4 (CP) + 9 F9H 9H --- R10 (CP) + 20 RL5 (CP) + 10 FAH AH --- R11 (CP) + 22 RH5 (CP) + 11 FBH BH --- R12 (CP) + 24 RL6 (CP) + 12 FCH CH --- R13 (CP) + 26 RH6 (CP) + 13 FDH DH --- R14 (CP) + 28 RL7 (CP) + 14 FEH EH --- R15 (CP) + 30 RH7 (CP) + 15 FFH FH --- 1) The first 8 GPRs (R7 … R0) may also be accessed bytewise. Writing to a GPR byte does not affect the other byte of the respective GPR. 2) Short addressing modes are usable for all register banks. 3) Long addressing mode only usable for the memory mapped global GPR bank. User’s Manual CPUSV2_X, V2.2 4-32 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.5.2 Context Switching When a task scheduler of an operating system activates a new task or an interrupt service routine is called or terminated, the working context (i.e. the registers) of the left task must be saved and the working context of the new task must be restored. The CPU context can be changed in two ways: • • Switching the selected register bank Switching the context of the global register Switching the Selected Physical Register Bank By updating bitfield BANK in register PSW the active register bank is switched immediately. It is possible to switch between the current memory-mapped GPR bank cached in the global register bank (BANK = 00B), local register bank 1 (BANK = 10B), and local register bank 2 (BANK = 11B). In case of an interrupt service, the bank switch can be automatically executed by updating bitfield BANK from registers BNKSELx in the interrupt controller. By executing a RETI instruction, bitfield BANK will automatically be restored and the context will switched to the original register bank. The switch between the three physical register banks of the register file can also be executed by writing to bitfield BANK. Because of pipeline dependencies an explicit change of register PSW must cancel the pipeline. Global Bank Local Bank Global Bank Execution Task A Execution Task B ExecutionTask A Interrupt of Task B recognized Execution of RETI MCA04877 Figure 4-8 Context Switch by Changing the Physical Register Bank After a switch to a local register bank, the new bank is immediately available. After switching to the global register bank, the cached memory-mapped GPRs must be valid before any further instructions can be executed. If the global register bank is not valid at this time (in case if the context switch process has been interrupted), the cache validation process is repeated automatically. User’s Manual CPUSV2_X, V2.2 4-33 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Switching the Context of the Global Register Bank The contents of the global register bank are switched by changing the base address of the memory-mapped GPR bank. The base address is given by the contents of the Context Pointer (CP). After the CP has been updated, a state machine starts to store the old contents of the global register bank and to load the new one. The store and load algorithm is executed in nineteen CPU cycles: the execution of the cache validation process takes sixteen cycles plus three cycles to stall an instruction execution to avoid pipeline conflicts upon the completion of the validation process. The context switch process has two phases: • • Store phase: The contents of the global register bank is stored back into the DPRAM by executing eight injected STORE instructions. After the last STORE instruction the contents of the global register bank are invalidated. Load phase: The global register bank is loaded with the new context by executing eight injected LOAD instructions. After the last LOAD instruction the contents of the global register bank are validated. The code execution is stopped until the global register bank is valid again. A hardware interrupt can occur during the validation process. The way the validation process is completed depends on the type of register bank selected for this interrupt: • • If the interrupt also uses a global register bank the validation process is finished before executing the service routine (see Figure 4-9). If the interrupt uses a local register bank the validation process is interrupted and the service routine is executed immediately (see Figure 4-10). After switching back to the global register bank, the validation process is finished: – If the interrupt occurred during the store phase, the entire validation process is restarted from the very beginning. – If the interrupt occurred during the load phase, only the load phase is repeated. If a local-bank interrupt routine (Task B in Figure 4-11) is again interrupted by a globalbank interrupt (Task C), the suspended validation process must be finished before code of Task C can be executed. This means that the validation process of Task A does not affect the interrupt latency of Task B but the latency of Task C. Note: If Task C would immediately interrupt Task A, the register bank validation process of Task A would be finished first. The worst case interrupt latency is identical in both cases (see Figure 4-9 and Figure 4-11). User’s Manual CPUSV2_X, V2.2 4-34 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Global Bank Execution Task A Execution Task B Execution Task B Execution of SCXT CP Execution of SCXT CP Interrupt of Task B recognized Register Bank Validation Process Started Finished Figure 4-9 Global Bank Global Bank Execution Task B Execution Task A Execution of POP CP Execution of RETI Register Bank Validation Process Started Finished Register Bank Validation Process Started Finished MCA04874 Validation Process Interrupted by Global-Bank Interrupt Global Bank Global Bank Local Bank Execution Task A Execution Task A Execution Task B Execution of SCXT CP Interrupt of Task B recognized Execution of RETI Register Bank Validation Process Started Stopped Register Bank Validation Process Restarted Finished MCA04875 Figure 4-10 Validation Process Interrupted by Local-Bank Interrupt Local Bank Global Bank Execution Task A Global Bank Execution Task B Interrupt of Task C recognized Execution of SCXT CP Interrupt of Task B recognized Local Bank Global Bank Execution Task B Execution Task A Execution Task C Execution of RETI Register Bank Validation Process Restarted Finished Register Bank Validation Process Started Stopped Execution of RETI MCA04876 Figure 4-11 Validation Process Interrupted by Local- and Global-Bank Intr. User’s Manual CPUSV2_X, V2.2 4-35 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The Context Pointer (CP) This non-bit-addressable register selects the current global register bank context. It can be updated via any instruction capable of modifying SFRs. CP Context Pointer SFR (FE10H/08H) 11 10 9 8 7 6 Reset Value: FC00H 15 14 13 12 5 4 3 2 1 0 1 1 1 1 cp 0 r r r r rw r Field Bits Type Description cp [11:1] rw Modifiable Portion of Register CP Specifies the (word) base address of the current global (memory-mapped) register bank. When writing a value to register CP with bits CP[11:9] = 000B, bits CP[11:10] are set to 11B by hardware. Note: It is the user’s responsibility to ensure that the physical GPR address specified via CP register plus short GPR address is always an internal DPRAM location. If this condition is not met, unexpected results may occur. Do not set CP below the internal DPRAM start address. The XC167 switches the complete memory-mapped GPR bank with a single instruction. After switching, the service routine executes within its own separate context. The instruction “SCXT CP, #New_Bank” pushes the value of the current context pointer (CP) into the system stack and loads CP with the immediate value “New_Bank”, which selects a new register bank. The service routine may now use its “own registers”. This memory register bank is preserved when the service routine terminates, i.e. its contents are available on the next call. Before returning from the service routine (RETI), the previous CP is simply popped from the system stack which returns the registers to the original bank. Note: Due to the internal instruction pipeline, a write operation to the CP register stalls the instruction flow until the register file context switch is really executed. The instruction immediately following the instruction that updates CP register can use the new value of the changed CP. User’s Manual CPUSV2_X, V2.2 4-36 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.6 Code Addressing The XC167 provides a total addressable memory space of 16 Mbytes. This address space is arranged as 256 segments of 64 Kbytes each. A dedicated 24-bit code address pointer is used to access the memories for instruction fetches. This pointer has two parts: an 8-bit code segment pointer CSP and a 16-bit offset pointer called Instruction Pointer (IP). The concatenation of the CSP and IP results directly in a correct 24-bit physical memory address. Memory organized in segments 15 87 CSP 0 15 IP 0 255 FF'0000H 254 FE'0000H 1 01'0000H 0 00'0000H 23 16 15 Segment 0 Offset MCA04920 Figure 4-12 Addressing via the Code Segment and Instruction Pointer The Code Segment Pointer CSP selects the code segment being used at run-time to access instructions. The lower 8 bits of register CSP select one of up 256 segments of 64 Kbytes each, while the higher 8 bits are reserved for future use. The reset value is specified by the contents of the VECSEG register (Section 5.3). Note: Register CSP can only be read but cannot be written by data operations. In segmented memory mode (default after reset), register CSP is modified either directly by JMPS and CALLS instructions, or indirectly via the stack by RETS and RETI instructions. In non-segmented memory mode (selected by setting bit SGTDIS in register CPUCON1), CSP is fixed to the segment of the instruction that disabled segmentation. Modification by inter-segment CALLs or RETurns is no longer possible. For processing an accepted interrupt or a TRAP, register CSP is automatically loaded with the segment of the vector table (defined in register VECSEG). User’s Manual CPUSV2_X, V2.2 4-37 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Note: For the correct execution of interrupt tasks in non-segmented memory mode, the contents of VECSEG must select the same segment as the current value of CSP, i.e. the vector table must be located in the segment pointed to by the CSP. CSP Code Segment Pointer SFR (FE08H/04H) 7 6 Reset Value: xxxxH 15 14 13 12 11 10 9 8 5 4 3 - - - - - - - - SEGNR - - - - - - - - rh 2 1 0 Field Bits Type Description SEGNR [7:0] rh Specifies the code segment from which the current instruction is to be fetched. Note: After a reset, register CSP is automatically loaded from register VECSEG. The Instruction Pointer IP determines the 16-bit intra-segment address of the currently fetched instruction within the code segment selected by the CSP register. Register IP is not mapped into the XC167’s address space; thus, it is not directly accessible by the programmer. However, the IP can be modified indirectly via the stack by means of a return instruction. IP is implicitly updated by the CPU for branch instructions and after instruction fetch operations. IP Instruction Pointer 15 14 13 12 - - - (- - - -/- -) 11 10 9 8 7 Reset Value: 0000H 6 5 4 3 2 1 0 ip (r)(w)h Field Bits Type Description ip [15:1] h Specifies the intra segment offset from which the current instruction is to be fetched. IP refers to the current segment <SEGNR>. User’s Manual CPUSV2_X, V2.2 4-38 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.7 Data Addressing The Address Data Unit (ADU) contains two independent arithmetic units to generate, calculate, and update addresses for data accesses, the Standard Address Generation Unit (SAGU) and the DSP Address Generation Unit (DAGU). The ADU performs the following major tasks: • • • • Standard Address Generation (SAGU) DSP Address Generation (DAGU) Data Paging (SAGU) Stack Handling (SAGU) The SAGU supports linear arithmetic for the indirect addressing modes and also generates the address in case of all other short and long addressing modes. The DAGU contains an additional set of address pointers and offset registers which are used in conjunction with the CoXXX instructions only. The CPU provides a lot of powerful addressing modes (short, long, indirect) for word, byte, and bit data accesses. The different addressing modes use different formats and have different scopes. 4.7.1 Short Addressing Modes Short addressing modes allow access to the GPR, SFR or bit-addressable memory space. All of these addressing modes use an offset (8/4/2 bits) together with an implicit base address to specify a 24-bit physical address: Table 4-17 Short Addressing Modes Mnemo- Base nic Address1) Offset Short Address Scope of Access Range Rw (CP) 2 × Rw 0 … 15 GPRs (word) Rb (CP) 1 × Rb 0 … 15 GPRs (byte) reg 00’FE00H 00’F000H (CP) (CP) 2 × reg 2 × reg 2 × (reg ∧ 0FH) 1 × (reg ∧ 0FH) 00H … EFH 00H … EFH F0H … FFH F0H … FFH SFRs (word, low byte) ESFRs (word, low byte) GPRs (word) GPRs (bytes) bitoff 00’FD00H 00’FF00H 00’F100H (CP) 2 × bitoff 2 × (bitoff ∧ 7FH) 2 × (bitoff ∧ 7FH) 2 × (bitoff ∧ 0FH) 00H … 7FH 80H … EFH 80H … EFH F0H … FFH RAM Bit word offset SFR Bit word offset ESFR Bit word offset GPR Bit word offset bitaddr Bit word see bitoff Immediate bit position 0 … 15 Any single bit 1) Accesses to general purpose registers (GPRs) may also access local register banks, instead of using CP. User’s Manual CPUSV2_X, V2.2 4-39 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Physical Address = Base Address + ∆ × Short Address Note: ∆ is 1 for byte GPRs, ∆ is 2 for word GPRs. Rw, Rb: Specifies direct access to any GPR in the currently active context (global register bank or local register bank). Both ‘Rw’ and ‘Rb’ require four bits in the instruction format. The base address of the global register bank is determined by the contents of register CP. ‘Rw’ specifies a 4-bit word GPR address, ‘Rb’ specifies a 4-bit byte GPR address within a local register bank or relative to (CP). reg: Specifies direct access to any (E)SFR or GPR in the currently active context (global or local register bank). The ‘reg’ value requires eight bits in the instruction format. Short ‘reg’ addresses in the range from 00H to EFH always specify (E)SFRs. In that case, the factor ‘∆’ equates 2 and the base address is 00’FE00H for the standard SFR area or 00’F000H for the extended ESFR area. The ‘reg’ accesses to the ESFR area require a preceding EXT*R instruction to switch the base address. Depending on the opcode, either the total word (for word operations) or the low byte (for byte operations) of an SFR can be addressed via ‘reg’. Note that the high byte of an SFR cannot be accessed via the ‘reg’ addressing mode. Short ‘reg’ addresses in the range from F0H to FFH always specify GPRs. In that case, only the lower four bits of ‘reg’ are significant for physical address generation and, therefore, it is identical to the address generation described for the ‘Rb’ and ‘Rw’ addressing modes. bitoff: Specifies direct access to any word in the bit addressable memory space. The ‘bitoff’ value requires eight bits in the instruction format. The specified ‘bitoff’ range selects different base addresses to generate physical addresses (see Table 4-17). The ‘bitoff’ accesses to the ESFR area require a preceding EXT*R instruction to switch the base address. bitaddr: Any bit address is specified by a word address within the bit addressable memory space (see ‘bitoff’) and a bit position (‘bitpos’) within that word. Therefore, ‘bitaddr’ requires twelve bits in the instruction format. User’s Manual CPUSV2_X, V2.2 4-40 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.7.2 Long Addressing Modes Long addressing modes specify 24-bit addresses and, therefore, can access any word or byte data within the entire address space. Long addresses can be specified in different ways to generate the full 24-bit address: • • • Use one of the four Data Page Pointers (DPP registers): The used 16-bit pointer selects a DPP with bits 15 … 14, bits 13 … 0 specify the 14-bit data page offset (see Figure 4-13). Select the used data page directly: The data page is selected by a preceeding EXTP(R) instruction, bits 13 … 0 of the used 16-bit pointer specify the 14-bit data page offset. Select the used segment directly: The segment is selected by a preceeding EXTS(R) instruction, the used 16-bit pointer specifies the 16-bit segment offset. Note: Word accesses on odd byte addresses are not executed. A hardware trap will be triggered. 16-Bit Data Address Memory 15 14 0 255 FF'0000H 254 Selects DPP 9 DPP 0 FE'0000H DPP3 - 11 DPP2 - 10 DPP1 - 01 X DPP0 - 00 23 1 15 14 0 01'0000H 0 Page Page Offset Segment Segment Offset 00'0000H MCA04924 Figure 4-13 Data Page Pointer Addressing User’s Manual CPUSV2_X, V2.2 4-41 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Data Page Pointers DPP0, DPP1, DPP2, DPP3 These four non-bit-addressable registers select up to four different data pages to be active simultaneously at run-time. The lower 10 bits of each DPP register select one of the 1024 possible 16-Kbyte data pages; the upper 6 bits are reserved for future use. DPP0 Data Page Pointer 0 SFR (FE00H/00H) 15 14 13 12 11 10 - - - - - - DPP0PN - - - - - - rw DPP1 Data Page Pointer 1 9 8 7 6 Reset Value: 0000H 5 4 SFR (FE02H/01H) 8 7 6 14 13 12 11 10 - - - - - - DPP1PN - - - - - - rw 5 4 SFR (FE04H/02H) 8 7 6 14 13 12 11 10 - - - - - - DPP2PN - - - - - - rw 5 4 SFR (FE06H/03H) 9 8 7 6 1 3 2 1 14 13 12 11 10 - - - - - - DPP3PN - - - - - - rw 3 2 1 5 4 3 2 1 Field Bits Type Description DPPxPN [9:0] rw Data Page Number of DPPx Specifies the data page selected via DPPx. 4-42 0 0 Reset Value: 0003H 15 User’s Manual CPUSV2_X, V2.2 0 Reset Value: 0002H 15 DPP3 Data Page Pointer 3 9 2 Reset Value: 0001H 15 DPP2 Data Page Pointer 2 9 3 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The DPP registers allow access to the entire memory space in pages of 16 Kbytes each. The DPP registers are implicitly used whenever data accesses to any memory location are made via indirect or direct long 16-bit addressing modes (except for override accesses via EXTended instructions and PEC data transfers). After reset, the Data Page Pointers are initialized in such a way that all indirect or direct long 16-bit addresses result in identical 18-bit addresses. This allows access to data pages 3 … 0 within segment 0 as shown in Figure 4-13. If the user does not want to use data paging, no further action is required. Data paging is performed by concatenating the lower 14 bits of an indirect or direct long 16-bit address with the contents of the DPP register selected by the upper two bits of the 16-bit address. The contents of the selected DPP register specify one of the 1024 possible data pages. This data page base address together with the 14-bit page offset forms the physical 24-bit address (even if segmentation is disabled). The selected number of segment address bits (via bitfield SALSEL) of the respective DPP register is output on the respective segment address pins for all external data accesses. A DPP register can be updated via any instruction capable of modifying an SFR. Note: Due to the internal instruction pipeline, a write operation to the DPPx registers could stall the instruction flow until the DPP is actually updated. The instruction that immediately follows the instruction which updates the DPP register can use the new value of the changed DPPx. 15 14 13 EXTP(R): 0 16-Bit Long Address #pag 14-Bit Page Offset 24-Bit Physical Address 15 EXTS(R): 0 16-Bit Long Address #seg 16-Bit Segment Offset 24-Bit Physical Address MCA04925 Figure 4-14 Overriding the DPP Mechanism User’s Manual CPUSV2_X, V2.2 4-43 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Note: The overriding page or segment may be specified as a constant (#pag, #seg) or via a word GPR (Rw). Table 4-18 Long Addressing Modes Mnemonic Base Address1) Offset Scope of Access mem (DPPx) mem ∧ 3FFFH Any Word or Byte mem pag mem ∧ 3FFFH Any Word or Byte mem seg mem Any Word or Byte 1) Represents either a 10-bit data page number to be concatenated with a 14-bit offset, or an 8-bit segment number to be concatenated with a 16-bit offset. User’s Manual CPUSV2_X, V2.2 4-44 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.7.3 Indirect Addressing Modes Indirect addressing modes can be considered as a combination of short and long addressing modes. This means that the “long” 16-bit pointer is provided indirectly by the contents of a word GPR which itself is specified directly by a short 4-bit address (‘Rw’ = 0 … 15). There are indirect addressing modes, which add a constant value to the GPR contents before the long 16-bit address is calculated. Other indirect addressing modes can decrement or increment the indirect address pointers (GPR contents) by 2 or 1 (referring to words or bytes) or by the contents of the offset registers QR0 or QR1. Table 4-19 Generating Physical Addresses from Indirect Pointers Step Executed Action Calculation Notes see Table 4-17 1 Calculate the address of the GPR Address = indirect pointer (word GPR) 2 × Short Addr. [+ (CP)] from its short address 2 Pre-decrement indirect pointer (‘-Rw’) depending on datatype (∆ = 1 or 2 for byte or word operations) (GPR Address) = Optional step, executed only if (GPR Address) required by addressing mode -∆ 3 Adjust the pointer by a constant value (‘Rw + const16’) Pointer = (GPR Address) + Constant 4 Calculate the physical 24-bit Physical Addr. = Uses DPPs or page/segment address using the resulting Page/Segment + override mechanisms, pointer Pointer offset see Table 4-18 5 Post-in/decrement indirect (GPR Address) = Optional step, executed only if pointer (‘Rw±’) depending required by addressing mode (GPR Address) on datatype (∆ = 1 or 2 for ± ∆ byte or word operations), or depending on offset registers (∆ = QRx)1) Optional step, executed only if required by addressing mode 1) Post-decrement and QRx-based modification is provided only for CoXXX instructions. Note: Some instructions only use the lowest four word GPRs (R3 … R0) as indirect address pointers, which are specified via short 2-bit addresses in that case. The following indirect addressing modes are provided: User’s Manual CPUSV2_X, V2.2 4-45 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 4-20 Indirect Addressing Modes Mnemonic Particularities [Rw] Most instructions accept any GPR (R15 … R0) as indirect address pointer. Some instructions accept only the lower four GPRs (R3 … R0). [Rw+] The specified indirect address pointer is automatically post-incremented by 2 or 1 (for word or byte data operations) after the access. [-Rw] The specified indirect address pointer is automatically pre-decremented by 2 or 1 (for word or byte data operations) before the access. [Rw + #data16] The specified 16-bit constant is added to the indirect address pointer, before the long address is calculated. [Rw-] The specified indirect address pointer is automatically postdecremented by 2 (word data operations) after the access. [Rw + QRx] The specified indirect address pointer is automatically post-incremented by QRx (word data operations) after the access. [Rw - QRx] The specified indirect address pointer is automatically postdecremented by QRX (word data operations) after the access. The non-bit-addressable offset registers QR0 and QR1 are used with CoXXX instructions. For possible instruction flow stalls refer to Section 4.3.4. QR0 Offset Register 15 14 13 ESFR (F004H/02H) 12 11 10 QR1 Offset Register 15 14 13 9 8 7 6 Reset Value: 0000H 5 4 11 10 9 2 1 0 QR 0 rw r ESFR (F006H/03H) 12 3 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 QR 0 rw r Field Bits Type Description QR [15:1] rw Modifiable Portion of Register QRx Specifies the 16-bit word offset address for indirect addressing modes (LSB always zero). User’s Manual CPUSV2_X, V2.2 4-46 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.7.4 DSP Addressing Modes In addition to the Standard Address Generation Unit (SAGU), the DSP Address Generation Unit (DAGU) provides an additional set of pointer registers (IDX0, IDX1) and offset registers (QX0, QX1). The additional set of pointer registers IDX0 and IDX1 allows the execution of DSP specific CoXXX instructions in one CPU cycle. An independent arithmetic unit allows the update of these dedicated pointer registers in parallel with the GPR-pointer modification of the SAGU. The DAGU only supports indirect addressing modes that use the special pointer registers IDX0 and IDX1. The address pointers can be used for arithmetic operations as well as for the special CoMOV instruction. The generation of the 24-bit memory address is different: • For CoMOV instructions, the IDX pointers are concatenated with the DPPs or the selected page/segment address, as described for long addressing modes (see Figure 4-13 for a summary). For arithmetic CoXXX instructions, the IDX pointers are automatically extended to a 24-bit memory address pointing to the internal DPRAM area, as shown in Figure 4-15. • IDX0 Address Pointer 15 14 13 SFR (FF08H/84H) 12 11 10 IDX1 Address Pointer 15 14 13 9 8 7 6 Reset Value: 0000H 5 4 11 10 9 2 1 0 idx 0 rw r SFR (FF0AH/85H) 12 3 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 idx 0 rw r Field Bits Type Description idx [15:1] rw Modifiable Portion of Register IDXx Specifies the 16-bit word address pointer Note: During the initialization of the IDX registers, instruction flow stalls are possible. For the proper operation, refer to Section 4.3.4. User’s Manual CPUSV2_X, V2.2 4-47 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) There are indirect addressing modes which allow parallel data move operations before the long 16-bit address is calculated (see Figure 4-16 for an example). Other indirect addressing modes allow decrementing or incrementing the indirect address pointers (IDXx contents) by 2 or by the contents of the offset registers QX0 and QX1 (used in conjunction with the IDX pointers). QX0 Offset Register 15 14 13 ESFR (F000H/00H) 12 11 10 QX1 Offset Register 15 14 13 9 8 7 6 Reset Value: 0000H 5 4 11 10 9 2 1 0 qx 0 rw r ESFR (F002H/01H) 12 3 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 qx 0 rw r Field Bits Type Description qx [15:1] rw Modifiable Portion of Register QXx Specifies the 16-bit word offset for indirect addressing modes Note: During the initialization of the QX registers, instruction flow stalls are possible. For the proper operation, refer to Section 4.3.4. User’s Manual CPUSV2_X, V2.2 4-48 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 15 Memory 16-Bit IDX Pointer 12 11 0 12 11 0 2 02'0000H 1 01'0000H DPRAM in Data Page 3 0 23 00'0000H 15 000000001111 MCA04926 Figure 4-15 Arithmetic MAC Operations and Addressing via the IDX Pointers Table 4-21 Generating Physical Addresses from Indirect Pointers (IDXx) Step Executed Action Calculation Notes 1 Determine the used IDXx pointer --- – 2 Interm. Addr. = Calculate an intermediate long address for the parallel (IDXx Address) ±∆ data move operation and in/decrement indirect pointer (‘IDXx±’) by 2 (∆ = 2), or depending on offset registers (∆ = QXx) Optional step, executed only if required by instruction CoXXXM and addressing mode 3 Calculate long 16-bit address – 4 Calculate the physical 24-bit Physical Addr. = Uses DPPs or page/segment address using the resulting Page/Segment + override mechanisms, see Table 4-18 and Figure 4-15 Pointer offset pointer 5 Post-in/decrement indirect pointer (‘IDXx±’) by 2 (∆ = 2), or depending on offset registers (∆ = QXx) User’s Manual CPUSV2_X, V2.2 Long Address = (IDXx Pointer) (IDXx Pointer) = (IDXx Pointer) ±∆ 4-49 Optional step, executed only if required by addressing mode V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The following indirect addressing modes are provided: Table 4-22 DSP Addressing Modes Mnemonic Particularities [IDXx] Most CoXXX instructions accept IDXx (IDX0, IDX1) as an indirect address pointer. [IDXx+] The specified indirect address pointer is automatically post-incremented by 2 after the access. with parallel data move In case of a CoXXXM instruction, the address stored in the specified indirect address pointer is automatically pre-decremented by 2 for the parallel move operation. The pointer itself is not pre-decremented. Then, the specified indirect address pointer is automatically postincremented by 2 after the access. [IDXx-] The specified indirect address pointer is automatically postdecremented by 2 after the access. with parallel data move In case of a CoXXXM instruction, the address stored in the specified indirect address pointer is automatically pre-incremented by 2 for the parallel move operation. The pointer itself is not pre-incremented. Then, the specified indirect address pointer is automatically post-decremented by 2 after the access. [IDXx + QXx] The specified indirect address pointer is automatically post-incremented by QXx after the access. with parallel data move In case of a CoXXXM instruction, the address stored in the specified indirect address pointer is automatically pre-decremented by QXx for the parallel move operation. The pointer itself is not pre-decremented. Then, the specified indirect address pointer is automatically postincremented by QXx after the access. [IDXx - QXx] The specified indirect address pointer is automatically postdecremented by QXx after the access. with parallel data move In case of a CoXXXM instruction, the address stored in the specified indirect address pointer is automatically pre-incremented by QXx for the parallel move operation. The pointer itself is not pre-incremented. Then, the specified indirect address pointer is automatically post-decremented by QXx after the access. Note: An example for parallel data move operations can be found in Figure 4-16. User’s Manual CPUSV2_X, V2.2 4-50 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The CoREG Addressing Mode The CoSTORE instruction utilizes the special CoREG addressing mode for immediate storage of the MAC-Unit register after a MAC operation. The address of the MAC-Unit register is coded in the CoSTORE instruction format as described in Table 4-23: Table 4-23 Coding of the CoREG Addressing Mode Mnemonic Register Coding of wwww:w bits [31:27] MSW MAC-Unit Status Word 00000 MAH MAC-Unit Accumulator High Word 00001 MAS Limited MAC-Unit Accumulator High 00010 Word MAL MAC-Unit Accumulator Low Word 00100 MCW MAC-Unit Control Word 00101 MRW MAC-Unit Repeat Word 00110 The example in Figure 4-16 shows the complex operation of CoXXXM instructions with a parallel move operation based on the descriptions about addressing modes given in Section 4.7.3 (Indirect Addressing Modes) and Section 4.7.4 (DSP Addressing Modes). User’s Manual CPUSV2_X, V2.2 4-51 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) CoXXXMxx [IDX0+], [R2+] Address Operations 1) Calculate Pointer Addresses IDXx = IDX0 R2 Address = CP + 2 × 2 (Global Register Bank) 2) Intermediate Address of Write Pointer for the Parallel Move Operation Intermediate Address = (IDX0) - 2 3) Calculate Long 16-Bit Address Long Address 1 = (IDX0) Long Address 2 = (R2) 4) Calculate 24-Bit Physical Address Physical Address 1 = Page 3 + Page Offset Physical Address 2 = (DPPi) + Page Offset 5) Post Modify Address Pointer (IDX0)new = (IDX0) + 2 (R2)new = (R2) + 2 Data Operations 1) Read Operands op1 = (Physical Address 1) op2 = (Physical Address 2) 1) Write Operand op1 (Intermediate Address) = op1 (IDX0)new (Updated Pointer) (R2)new (Updated Pointer) op1 op2 (IDX0) (Read Pointer) Parallel Move (R2) (Read Pointer) Intermediate Address (Write Pointer for Parallel Move) MCA04928 Figure 4-16 Arithmetic MAC Operations with Parallel Move User’s Manual CPUSV2_X, V2.2 4-52 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.7.5 The System Stack The XC167 supports a system stack of up to 64 Kbytes. The stack can be located internally in one of the on-chip memories or externally. The 16-bit Stack Pointer register (SP) addresses the stack within a 64-Kbyte segment selected by the Stack Pointer Segment register (SPSG). A virtual stack (usually bigger than 64 Kbytes) can be implemented by software. This mechanism is supported by the Stack Overflow register STKOV and the Stack Underflow register STKUN (see descriptions below). The Stack Pointer Registers SP and SPSEG Register SPSEG (not bitaddressable) selects the segment being used at run-time to access the system stack. The lower eight bits of register SPSEG select one of up 256 segments of 64 Kbytes each, while the higher 8 bits are reserved for future use. The Stack Pointer SP (not bitaddressable) points to the top of the system stack (TOS). SP is pre-decremented whenever data is pushed onto the stack, and it is postincremented whenever data is popped from the stack. Therefore, the system stack grows from higher towards lower memory locations. System stack addresses are generated by directly extending the 16-bit contents of register SP by the contents of register SPSG, as shown in Figure 4-17. The system stack cannot cross a 64-Kbyte segment boundary. Stack Pointer Segment SPSEG 15 7 SPSEGNR 0 15 SP 0 255 FF'0000H 254 FE'0000H 1 01'0000H 0 00'0000H 23 16 15 0 MCA04929 Figure 4-17 Addressing via the Stack Pointer User’s Manual CPUSV2_X, V2.2 4-53 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) SP Stack Pointer Register 15 14 13 12 SFR (FE12H/09H) 11 10 9 8 7 6 Reset Value: FC00H 5 4 3 1 0 sp 0 rwh r Field Bits Type Description sp [15:1] rwh Modifiable Portion of Register SP Specifies the top of the system stack. SPSEG Stack Pointer Segment 2 SFR (FF0CH/86H) 7 6 Reset Value: 0000H 15 14 13 12 11 10 9 8 5 4 3 - - - - - - - - SPSEGNR - - - - - - - - rw 2 1 Field Bits Type Description SPSEGNR [7:0] rw Stack Pointer Segment Number Specifies the segment where the stack is located. 0 Note: SPSEG and SP can be updated via any instruction capable of modifying a 16-bit SFR. Due to the internal instruction pipeline, a write operation to SPSG or SP stalls the instruction flow until the register is really updated. The instruction immediately following the instruction updating SPSG or SP can use the new value. Extreme care should be taken when changing the contents of the stack pointer registers. Improper changes may result in erroneous system behavior. User’s Manual CPUSV2_X, V2.2 4-54 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The Stack Overflow/Underflow Pointers STKOV/STKUN These limit registers (not bit-addressable) supervise the stack pointer. A trap is generated when the stack pointer reaches its upper or lower limit. The Stack Pointer Segment Register SPSG is not taken into account for the stack pointer comparison. The system stack cannot cross a 64-Kbyte segment. STKOV is compared with SP before each implicit write operation which decrements the contents of SP (instructions CALLA, CALLI, CALLR, CALLS, PCALL, TRAP, SCXT, or PUSH). If the contents of SP are equal to the contents of STKOV a stack overflow trap is triggered. STKUN is compared with SP before each implicit read operation which increments the contents of SP (instructions RET, RETS, RETP, RETI, or POP). If the contents of SP are equal to the contents of STKUN a stack underflow trap is triggered. The Stack Overflow/Underflow Traps may be used in two different ways: • • Fatal error indication treats the stack overflow as a system error and executes the associated trap service routine. In case of a stack overflow trap, data in the bottom of the stack may have been overwritten by the status information stacked upon servicing the trap itself. Virtual stack control allows the system stack to be used as a ‘Stack Cache’ for a bigger external user stack: flush cache in case of an overflow, refill cache in case of an underflow. Scope of Stack Limit Control The stack limit control implemented by the register pair STKOV and STKUN detects cases in which the Stack Pointer (SP) crosses the defined stack area as a result of an implicit change. If the stack pointer was explicitly changed as a result of move or arithmetic instruction, SP is not compared to the contents of STKOV and STKUN. In this case, a stack violation will not be detected if the modified stack pointer is on or outside the defined limits, i.e. below (STKOV) or above (STKUN). Stack overflow/underflow is detected only in case of implicit SP modification. SP may be operated outside the permitted SP range without triggering a trap. However, if SP reaches the limit of the permitted SP range from outside the range as a result of an implicit change (PUSH or POP, for example), the respective trap will be triggered. Note: STKOV and STKUN can be updated via any instruction capable of modifying an SFR. If a stack overflow or underflow event occurs in an ATOMIC/EXT sequence, the stack operations that are part of the sequence are completed. The trap is issued after the completion of the entire ATOMIC/EXT sequence. User’s Manual CPUSV2_X, V2.2 4-55 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) STKOV Stack Overflow Reg. 15 14 13 12 SFR (FE14H/0AH) 11 10 9 8 7 6 Reset Value: FA00H 5 4 3 2 1 stkov 0 0 rw Field Bits Type Description stkov [15:1] rw Modifiable Portion of Register STKOV Specifies the segment offset address of the lower limit of the system stack. STKUN Stack Underflow Reg. 15 14 13 12 SFR (FE16H/0BH) 11 10 9 8 7 6 Reset Value: FC00H 5 4 3 2 1 stkun 0 rw r Field Bits Type Description stkun [15:1] rw Modifiable Portion of Register STKUN Specifies the segment offset address of the upper limit of the system stack. User’s Manual CPUSV2_X, V2.2 0 4-56 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.8 Standard Data Processing All standard arithmetic, shift-, and logical operations are performed in the 16-bit ALU. In addition to the standard functions, the ALU of the XC167 includes a bit-manipulation unit and a multiply and divide unit. Most internal execution blocks have been optimized to perform operations on either 8-bit or 16-bit numbers. After the pipeline has been filled, most instructions are completed in one CPU cycle. The status flags are automatically updated in register PSW after each ALU operation and reflect the current state of the microcontroller. These flags allow branching upon specific conditions. Support of both signed and unsigned arithmetic is provided by the user selectable branch test. The status flags are also preserved automatically by the CPU upon entry into an interrupt or trap routine. Another group of bits represents the current CPU interrupt status. Two separate bits (USR0 and USR1) are provided as general purpose flags. PSW Processor Status Word 15 14 13 12 SFRb 11 10 ILVL IEN HLD EN BANK rwh rw rw rwh Type 9 8 7 Reset Value: 0000H 6 5 USR USR MUL 1 0 IP rwh rwh r 4 3 2 1 0 E Z V C N rwh rwh rwh rwh rwh Field Bits Description ILVL [15:12] rwh CPU Priority Level 0H Lowest Priority … … FH Highest Priority IEN 11 rw Global Interrupt/PEC Enable Bit 0 Interrupt/PEC requests are disabled 1 Interrupt/PEC requests are enabled HLDEN 10 rw Hold Enable 0 External bus arbitration disabled 1 External bus arbitration enabled Note: The selected arbitration mode is activated when HLDEN is set for the first time. BANK User’s Manual CPUSV2_X, V2.2 [9:8] rwh Reserved for Register File Bank Selection 00 Global register bank 01 Reserved 10 Local register bank 1 11 Local register bank 2 4-57 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description USR1 7 rwh General Purpose Flag May be used by application USR0 6 rwh General Purpose Flag May be used by application MULIP 5 r Multiplication/Division in Progress Note: Always set to 0 (MUL/DIV not interruptible), for compatibility with existing software. E 4 rwh End of Table Flag 0 Source operand is neither 8000H nor 80H 1 Source operand is 8000H or 80H Z 3 rwh Zero Flag 0 ALU result is not zero 1 ALU result is zero V 2 rwh Overflow Flag 0 No Overflow produced 0 Overflow produced C 1 rwh Carry Flag 0 No carry/borrow bit produced 1 Carry/borrow bit produced N 0 rwh Negative Result 0 ALU result is not negative 1 ALU result is negative ALU/MAC Status (N, C, V, Z, E, USR0, USR1) The condition flags (N, C, V, Z, E) within the PSW indicate the ALU status after the most recently performed ALU operation. They are set by most of the instructions according to specific rules which depend on the ALU or data movement operation performed by an instruction. After execution of an instruction which explicitly updates the PSW register, the condition flags cannot be interpreted as described below because any explicit write to the PSW register supersedes the condition flag values which are implicitly generated by the CPU. Explicitly reading the PSW register supplies a read value which represents the state of the PSW register after execution of the immediately preceding instruction. Note: After reset, all of the ALU status bits are cleared. N-Flag: For most of the ALU operations, the N-flag is set to 1, if the most significant bit of the result contains a 1; otherwise, it is cleared. In the case of integer operations, the N-flag can be interpreted as the sign bit of the result (negative: N = 1, positive: N = 0). User’s Manual CPUSV2_X, V2.2 4-58 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Negative numbers are always represented as the 2’s complement of the corresponding positive number. The range of signed numbers extends from -8000H to +7FFFH for the word data type, or from -80H to +7FH for the byte data type. For Boolean bit operations with only one operand, the N-flag represents the previous state of the specified bit. For Boolean bit operations with two operands, the N-flag represents the logical XORing of the two specified bits. C-Flag: After an addition, the C-flag indicates that a carry from the most significant bit of the specified word or byte data type has been generated. After a subtraction or a comparison, the C-flag indicates a borrow which represents the logical negation of a carry for the addition. This means that the C-flag is set to 1, if no carry from the most significant bit of the specified word or byte data type has been generated during a subtraction, which is performed internally by the ALU as a 2’s complement addition, and, the C-flag is cleared when this complement addition caused a carry. The C-flag is always cleared for logical, multiply and divide ALU operations, because these operations cannot cause a carry. For shift and rotate operations, the C-flag represents the value of the bit shifted out last. If a shift count of zero is specified, the C-flag will be cleared. The C-flag is also cleared for a prioritize ALU operation, because a 1 is never shifted out of the MSB during the normalization of an operand. For Boolean bit operations with only one operand, the C-flag is always cleared. For Boolean bit operations with two operands, the C-flag represents the logical ANDing of the two specified bits. V-Flag: For addition, subtraction, and 2’s complementation, the V-flag is always set to 1 if the result exceeds the range of 16-bit signed numbers for word operations (-8000H to +7FFFH), or 8-bit signed numbers for byte operations (-80H to +7FH). Otherwise, the V-flag is cleared. Note that the result of an integer addition, integer subtraction, or 2’s complement is not valid if the V-flag indicates an arithmetic overflow. For multiplication and division, the V-flag is set to 1 if the result cannot be represented in a word data type; otherwise, it is cleared. Note that a division by zero will always cause an overflow. In contrast to the result of a division, the result of a multiplication is valid whether or not the V-flag is set to 1. Because logical ALU operations cannot produce an invalid result, the V-flag is cleared by these operations. The V-flag is also used as a ‘Sticky Bit’ for rotate right and shift right operations. With only using the C-flag, a rounding error caused by a shift right operation can be estimated up to a quantity of one half of the LSB of the result. In conjunction with the V-flag, the C-flag allows evaluation of the rounding error with a finer resolution (see Table 4-24). For Boolean bit operations with only one operand, the V-flag is always cleared. For Boolean bit operations with two operands, the V-flag represents the logical ORing of the two specified bits. User’s Manual CPUSV2_X, V2.2 4-59 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 4-24 Shift Right Rounding Error Evaluation C-Flag V-Flag Rounding Error Quantity 0 0 1 1 0 1 0 1 No rounding error 0 < Rounding error < 1/2 LSB Rounding error = 1/2 LSB Rounding error > 1/2 LSB Z-Flag: The Z-flag is normally set to 1 if the result of an ALU operation equals zero, otherwise it is cleared. For the addition and subtraction with carry, the Z-flag is only set to 1, if the Z-flag already contains a 1 and the result of the current ALU operation also equals zero. This mechanism is provided to support multiple precision calculations. For Boolean bit operations with only one operand, the Z-flag represents the logical negation of the previous state of the specified bit. For Boolean bit operations with two operands, the Z-flag represents the logical NORing of the two specified bits. For the prioritize ALU operation, the Z-flag indicates whether the second operand was zero. E-Flag: End of table flag. The E-flag can be altered by instructions which perform ALU or data movement operations. The E-flag is cleared by those instructions which cannot be reasonably used for table search operations. In all other cases, the E-flag value depends on the value of the source operand to signify whether the end of a search table is reached or not. If the value of the source operand of an instruction equals the lowest negative number which is representable by the data format of the corresponding instruction (8000H for the word data type, or 80H for the byte data type), the E-flag is set to 1; otherwise, it is cleared. General Control Functions (USR0, USR1, BANK, HLDEN) A few bits in register PSW are dedicated to general control functions. Thus, they are saved and restored automatically upon task switches and interrupts. USR0/USR1-Flags: These bits can be set automatically during the execution of repeated MAC instructions. These bits can also be used as general flags by an application. BANK: Bitfield BANK selects the currently active register bank (local or global). Bitfield BANK is updated implicitly by hardware upon entering an interrupt service routine, and by a RETI instruction. It can be also modified explicitly via software by any instruction which can write to PSW. HLDEN: Setting this bit for the first time activates the selected bus arbitration mode (see Section 9.3.9). Bus arbitration can be disabled by temporarily clearing bit HLDEN. In this case the bus is locked, while the bus arbitration mode remains selected. User’s Manual CPUSV2_X, V2.2 4-60 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) CPU Interrupt Status (IEN, ILVL) IEN: The Interrupt Enable bit allows interrupts to be globally enabled (IEN = 1) or disabled (IEN = 0). ILVL: The four-bit Interrupt Level field (ILVL) specifies the priority of the current CPU activity. The interrupt level is updated by hardware on entry into an interrupt service routine, but it can also be modified via software to prevent other interrupts from being acknowledged. If an interrupt level 15 has been assigned to the CPU, it has the highest possible priority; thus, the current CPU operation cannot be interrupted except by hardware traps or external non-maskable interrupts. For details refer to Chapter 5. After reset, all interrupts are globally disabled, and the lowest priority (ILVL = 0) is assigned to the initial CPU activity. 4.8.1 16-bit Adder/Subtracter, Barrel Shifter, and 16-bit Logic Unit All standard arithmetic and logical operations are performed by the 16-bit ALU. In case of byte operations, signals from bits 6 and 7 of the ALU result are used to control the condition flags. Multiple precision arithmetic is supported by a “CARRY-IN” signal to the ALU from previously calculated portions of the desired operation. A 16-bit barrel shifter provides multiple bit shifts in a single cycle. Rotations and arithmetic shifts are also supported. 4.8.2 Bit Manipulation Unit The XC167 offers a large number of instructions for bit processing. These instructions either manipulate software flags within the internal RAM, control on-chip peripherals via control bits in their respective SFRs, or control IO functions via port pins. Unlike other microcontrollers, the XC167 features instructions that provide direct access to two operands in the bit addressable space without requiring them to be moved to temporary locations. Multiple bit shift instructions have been included to avoid long instruction streams of single bit shift operations. These instructions require a single CPU cycle. The instructions BSET, BCLR, BAND, BOR, BXOR, BMOV, BMOVN explicitly set or clear specific bits. The bitfield instructions BFLDL and BFLDH allow manipulation of up to 8 bits of a specific byte at one time. The instructions JBC and JNBS implicitly clear or set the specified bit when the jump is taken. The instructions JB and JNB (also conditional jump instructions that refer to flags) evaluate the specified bit to determine if the jump is to be taken. Note: Bit operations on undefined bit locations will always read a bit value of ‘0’, while the write access will not affect the respective bit location. All instructions that manipulate single bits or bit groups internally use a read-modify-write sequence that accesses the whole word containing the specified bit(s). User’s Manual CPUSV2_X, V2.2 4-61 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) This method has several consequences: • • The read-modify-write approach may be critical with hardware-affected bits. In these cases, the hardware may change specific bits while the read-modify-write operation is in progress; thus, the writeback would overwrite the new bit value generated by the hardware. The solution is provided by either the implemented hardware protection (see below) or through special programming (see Section 4.3). Bits can be modified only within the internal address areas (internal RAM and SFRs). External locations cannot be used with bit instructions. The upper 256 bytes of SFR area, ESFR area, and internal DPRAM are bit-addressable; so, the register bits located within those respective sections can be manipulated directly using bit instructions. The other SFRs must be accessed byte/word wise. Note: All GPRs are bit-addressable independently from the allocation of the register bank via the Context Pointer (CP). Even GPRs which are allocated to non-bitaddressable RAM locations provide this feature. Protected bits are not changed during the read-modify-write sequence, such as when hardware sets an interrupt request flag between the read and the write of the readmodify-write sequence. The hardware protection logic guarantees that only the intended bit(s) is/are affected by the write-back operation. A summary of the protected bits implemented in the XC167 can be found in Section 2.7. Note: If a conflict occurs between a bit manipulation generated by hardware and an intended software access, the software access has priority and determines the final value of the respective bit. User’s Manual CPUSV2_X, V2.2 4-62 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.8.3 Multiply and Divide Unit The XC167’s multiply and divide unit has two separated parts. One is the fast 16 × 16-bit multiplier that executes a multiplication in one CPU cycle. The other one is a division subunit which performs the division algorithm in 18 … 21 CPU cycles (depending on the data and division types). The divide instruction requires four CPU cycles to be executed. For performance reasons, the rest of the division algorithm runs in the background during the following seventeen CPU cycles, while further instructions are executed in parallel. Interrupt tasks can also be started and executed immediately without any delay. If an instruction (from the original instruction stream or from the interrupt task) tries to use the unit while a division is still running, the execution of this new instruction is stalled until the previous division is finished. To avoid these stalls, the multiply and division unit should not be used during the first fourteen CPU cycles of the interrupt tasks. For example, this requires up to fourteen onecycle instructions to be executed between the interrupt entry and the first instruction which uses the multiply and divide unit again (worst case). Multiplications and divisions implicitly use the 32-bit multiply/divide register MD (represented by the concatenation of the two non-bit-addressable data registers MDH and MDL) and the associated control register MDC. This bit-addressable 16-bit register is implicitly used by the CPU when it performs a division or multiplication in the ALU. After a multiplication, MD represents the 32-bit result. For long divisions, MD must be loaded with the 32-bit dividend before the division is started. After any division, register MDH represents the 16-bit remainder, register MDL represents the 16-bit quotient. MDH Multiply/Divide High Reg. 15 14 13 12 11 SFR (FE0CH/06H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 mdh rwh Field Bits Type Description mdh [15:0] rwh High Part of MD The high order sixteen bits of the 32-bit multiply and divide register MD. User’s Manual CPUSV2_X, V2.2 4-63 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) MDL Multiply/Divide Low Reg. 15 14 13 12 11 SFR (FE0EH/07H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 mdl rwh Field Bits Type Description mdl [15:0] rwh Low Part of MD The low order sixteen bits of the 32-bit multiply and divide register MD. Whenever MDH or MDL is updated via software, the Multiply/Divide Register In Use flag (MDRIU) in the Multiply/Divide Control register (MDC) is set to ‘1’. The MDRIU flag is cleared, whenever register MDL is read via software. MDC Multiply/Divide Control Reg. SFR (FF0EH/87H) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - - - - - - - - - MDR IU - - - - - - - - - - - - - - - r(w)h - - - - Field Bits Type Description MDRIU 4 rwh Multiply/Divide Register In Use 0 Cleared when MDL is read via software. 1 Set when MDL or MDH is written via software, or when a multiply or divide instruction is executed. Note: The MDRIU flag indicates the usage of register MD (MDL and MDH). In this case MD must be saved prior to a new multiplication or division operation. User’s Manual CPUSV2_X, V2.2 4-64 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.9 DSP Data Processing (MAC Unit) The new CoXXX arithmetic instructions are performed in the MAC unit. The MAC unit provides single-instruction-cycle, non-pipelined, 32-bit additions; 32-bit subtraction; right and left shifts; 16-bit by 16-bit multiplication; and multiplication with cumulative subtraction/addition. The MAC unit includes the following major components, shown in Figure 4-18: • • • • • • • • 16-bit by 16-bit signed/unsigned multiplier with signed result1) Concatenation Unit Scaler (one-bit left shifter) for fractional computing 40-bit Adder/Subtracter 40-bit Signed Accumulator Data Limiter Accumulator Shifter Repeat Counter 16-Bit Input Operands 16 16 16 Repeat Counter MCW Register 16 Signed/ Unsigned Multiplier Concatenation Unit 32 32 32 Signed Ext. 40-Bit Adder/Subtracter 40 Round + Saturation ACCU-Shifter 40 40-Bit Signed Accumulator 40 40 MSW Register Limiter MCA04930 Figure 4-18 Functional MAC Unit Block Diagram 1) The same hardware-multiplier is used in the ALU. User’s Manual CPUSV2_X, V2.2 4-65 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The working register of the MAC unit is a dedicated 40-bit accumulator register. A set of consistent flags is automatically updated in status register MSW after each MAC operation. These flags allow branching on specific conditions. Unlike the PSW flags, these flags are not preserved automatically by the CPU upon entry into an interrupt or trap routine. All dedicated MAC registers must be saved on the stack if the MAC unit is shared between different tasks and interrupts. General properties of the MAC unit are selected via the MAC control word MCW. MCW MAC Control Word SFR (FFDCH/EEH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - - - MP MS - - - - - - - - - - - - - - rw rw - - - - - - - - - Field Bits Type Description MP 10 rw One-Bit Scaler Control 0 Multiplier product shift disabled 1 Multiplier product shift enabled for signed multiplications MS 9 rw Saturation Control 0 Saturation disabled 1 Saturation to 32-bit value enabled 4.9.1 Representation of Numbers and Rounding The XC167 supports the 2’s complement representation of binary numbers. In this format, the sign bit is the MSB of the binary word. This is set to zero for positive numbers and set to one for negative numbers. Unsigned numbers are supported only by multiply/multiply-accumulate instructions which specify whether each operand is signed or unsigned. In 2’s complement fractional format, the N-bit operand is represented using the 1.[N-1] format (1 signed bit, N-1 fractional bits). Such a format can represent numbers between -1 and +1 - 2-[N-1]. This format is supported when bit MP of register MCW is set. The XC167 implements 2’s complement rounding. With this rounding type, one is added to the bit to the right of the rounding point (bit 15 of MAL), before truncation (MAL is cleared). User’s Manual CPUSV2_X, V2.2 4-66 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.9.2 The 16-bit by 16-bit Signed/Unsigned Multiplier and Scaler The multiplier executes 16-bit by 16-bit parallel signed/unsigned fractional and integer multiplication in one CPU-cycle. The multiplier allows the multiplication of unsigned and signed operands. The result is always presented in a signed fractional or integer format. The result of the multiplication feeds a one-bit scaler to allow compensation for the extra sign bit gained in multiplying two 16-bit 2’s complement numbers. 4.9.3 Concatenation Unit The concatenation unit enables the MAC unit to perform 32-bit arithmetic operations in one CPU cycle. The concatenation unit concatenates two 16-bit operands to a 32-bit operand before the 32-bit arithmetic operation is executed in the 40-bit adder/subtracter. The second required operand is always the current accumulator contents. The concatenation unit is also used to pre-load the accumulator with a 32-bit value. 4.9.4 One-bit Scaler The one-bit scaler can shift the result of the concatenation unit or the output of the multiplier one bit to the left. The scaler is controlled by the executed instruction for the concatenation or by control bit MP in register MCW. If bit MP is set the product is shifted one bit to the left to compensate for the extra sign bit gained in multiplying two 16-bit 2’s-complement numbers. The enabled automatic shift is performed only if both input operands are signed. 4.9.5 The 40-bit Adder/Subtracter The 40-bit Adder/Subtracter allows intermediate overflows in a series of multiply/accumulate operations. The Adder/Subtracter has two input ports. The 40-bit port is the feedback of the accumulator output through the ACCU-Shifter to the Adder/Subtracter. The 32-bit port is the input port for the operand coming from the onebit Scaler. The 32-bit operands are signed and extended to 40 bits before the addition/subtraction is performed. The output of the Adder/Subtracter goes to the accumulator. It is also possible to round the result and to saturate it on a 32-bit value automatically after every accumulation. The round operation is performed by adding 00’0000’8000H to the result. Automatic saturation is enabled by setting the saturation control bit MS in register MCW. When the accumulator is in the overflow saturation mode and an overflow occurs, the accumulator is loaded with either the most positive or the most negative value representable in a 32-bit value, depending on the direction of the overflow as well as on the arithmetic used. The value of the accumulator upon saturation is either 00’7FFF’FFFFH (positive) or FF’8000’0000H (negative). User’s Manual CPUSV2_X, V2.2 4-67 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.9.6 The Data Limiter Saturation arithmetic is also provided to selectively limit overflow when reading the accumulator by means of a CoSTORE <destination>., MAS instruction. Limiting is performed on the MAC-Unit accumulator. If the contents of the accumulator can be represented in the destination operand size without overflow, then the data limiter is disabled and the operand is not modified. If the contents of the accumulator cannot be represented without overflow in the destination operand size, the limiter will substitute a “limited” data as explained in Table 4-25: Table 4-25 Limiter Output ME-flag MN-flag Output of Limiter 0 x unchanged 1 0 7FFFH 1 1 8000H Note: In this particular case, both the accumulator and the status register are not affected. MAS is readable by means of a CoSTORE instruction only. 4.9.7 The Accumulator Shifter The accumulator shifter is a parallel shifter with a 40-bit input and a 40-bit output. The source accumulator shifting operations are: • • • No shift (Unmodified) Up to 16-bit Arithmetic Left Shift Up to 16-bit Arithmetic Right Shift Notice that bits ME, MSV, and MSL in register MSW are affected by left shifts; therefore, if the saturation mechanism is enabled (MS) the behavior is similar to the one of the Adder/Subtracter. Note: Certain precautions are required in case of left shift with saturation enabled. Generally, if MAE contains significant bits, then the 32-bit value in the accumulator is to be saturated. However, it is possible that left shift may move some significant bits out of the accumulator. The 40-bit result will be misinterpreted and will be either not saturated or saturated incorrectly. There is a chance that the result of left shift may produce a result which can saturate an original positive number to the minimum negative value, or vice versa. User’s Manual CPUSV2_X, V2.2 4-68 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.9.8 The 40-bit Signed Accumulator Register The 40-bit accumulator consists of three concatenated registers MAE, MAH, and MAL. MAE is 8 bits wide, MAH and MAL are 16 bits wide. MAE is the Most Significant Byte of the 40-bit accumulator. This byte performs a guarding function. MAE is accessed as the lower byte of register MSW. When MAH is written, the value in the accumulator is automatically adjusted to signed extended 40-bit format. That means MAL is cleared and MAE will be automatically loaded with zeros for a positive number (the most significant bit of MAH is 0), and with ones for a negative number (the most significant bit of MAH is 1), representing the extended 40-bit negative number in 2’s complement notation. One may see that the extended 40-bit value is equal to the 32-bit value without extension. In other words, after this extension, MAE does not contain significant bits. Generally, this condition is present when the highest 9 bits of the 40-bit signed result are the same. During the accumulator operations, an overflow may happen and the result may not fit into 32 bits and MAE will change. The extension flag “E” in register MSW is set when the signed result in the accumulator has exceeded the 32-bit boundary. This condition is present when the highest 9 bits of the 40-bit signed result are not the same, i.e. MAE contains significant bits. Most CoXXX operations specify the 40-bit accumulator register as a source and/or a destination operand. MAH Accumulator High Word 15 14 13 12 11 SFR (FE5EH/2FH) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MAH rwh MAL Accumulator Low Word 15 14 13 12 11 SFR (FE5CH/2EH) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MAL rwh Field Bits Type Description MAH, MAL [15:0] rwh High and Low Part of Accumulator The 40-bit accumulator is completed by MAE User’s Manual CPUSV2_X, V2.2 4-69 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.9.9 The MAC Unit Status Word MSW The upper byte of register MSW (bit-addressable) shows the current status of the MAC Unit. The lower byte of register MSW represents the 8-bit MAC accumulator extension, building the 40-bit accumulator together with registers MAH and MAL. MSW MAC Status Word 15 14 13 12 SFR (FFDEH/EFH) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 - MV MSL ME MSV MC MZ MN MAE - rwh rwh rwh rwh rwh rwh rwh rwh 2 1 0 Field Bits Type Description MV 14 rwh Overflow Flag 0 No Overflow produced 1 Overflow produced MSL 13 rwh Sticky Limit Flag 0 Result was not saturated 1 Result was saturated ME 12 rwh MAC Extension Flag 0 MAE does not contain significant bits 1 MAE contains significant bits MSV 11 rwh Sticky Overflow Flag 0 No Overflow occurred 1 Overflow occurred MC 10 rwh Carry Flag 0 No carry/borrow produced 1 Carry/borrow produced MZ 9 rwh Zero Flag 0 MAC result is not zero 1 MAC result is zero MN 8 rwh Negative Result 0 MAC result is positive 1 MAC result is negative MAE [7:0] rwh MAC Accumulator Extension The most significant bits of the 40-bit accumulator, completing registers MAH and MAL User’s Manual CPUSV2_X, V2.2 4-70 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) MAC Unit Status (MV, MN, MZ, MC, MSV, ME, MSL) These condition flags indicate the MAC status resulting from the most recently performed MAC operation. These flags are controlled by the majority of MAC instructions according to specific rules. Those rules depend on the instruction managing the MAC or data movement operation. After execution of an instruction which explicitly updates register MSW, the condition flags may no longer represent an actual MAC status. An explicit write operation to register MSW supersedes the condition flag values implicitly generated by the MAC unit. An explicit read access returns the value of register MSW after execution of the immediately preceding instruction. Register MSW can be accessed via any instruction capable of accessing an SFR. Note: After reset, all MAC status bits are cleared. MN-Flag: For the majority of the MAC operations, the MN-flag is set to 1 if the most significant bit of the result contains a 1; otherwise, it is cleared. In the case of integer operations, the MN-flag can be interpreted as the sign bit of the result (negative: MN = 1, positive: MN = 0). Negative numbers are always represented as the 2’s complement of the corresponding positive number. The range of signed numbers extends from 80’0000’0000H to 7F’FFFF’FFFFH. MZ-Flag: The MZ-flag is normally set to 1 if the result of a MAC operation equals zero; otherwise, it is cleared. MC-Flag: After a MAC addition, the MC-flag indicates that a “Carry” from the most significant bit of the accumulator extension MAE has been generated. After a MAC subtraction or a MAC comparison, the MC-flag indicates a “Borrow” representing the logical negation of a “Carry” for the addition. This means that the MC-flag is set to 1 if no “Carry” from the most significant bit of the accumulator has been generated during a subtraction. Subtraction is performed by the MAC Unit as a 2’s complement addition and the MC-flag is cleared when this complement addition caused a “Carry”. For left-shift MAC operations, the MC-flag represents the value of the bit shifted out last. Right-shift MAC operations always clear the MC-flag. The arithmetic right-shift MAC operation can set the MC-flag if the enabled round operation generates a “Carry” from the most significant bit of the accumulator extension MAE. MSV-Flag: The addition, subtraction, 2’s complement, and round operations always set the MSV-flag to 1 if the MAC result exceeds the maximum range of 40-bit signed numbers. If the MSV-flag indicates an arithmetic overflow, the MAC result of an operation is not valid. The MSV-flag is a ‘Sticky Bit’. Once set, other MAC operations cannot affect the status of the MSV-flag. Only a direct write operation can clear the MSV-flag. ME-Flag: The ME-flag is set if the accumulator extension MAE contains significant bits, that means if the nine highest accumulator bits are not all equal. User’s Manual CPUSV2_X, V2.2 4-71 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) MSL-Flag: The MSL-flag is set if an automatic saturation of the accumulator has happened. The automatic saturation is enabled if bit MS in register MCW is set. The MSL-Flag can be also set by instructions which limit the contents of the accumulator. If the accumulator has been limited, the MSL-Flag is set. The MSL-Flag is a ‘Sticky Bit’. Once set, it cannot be affected by the other MAC operations. Only a direct write operation can clear the MSL-flag. MV-Flag: The addition, subtraction, and accumulation operations set the MV-flag to 1 if the result exceeds the maximum range of signed numbers (80’0000’0000H to 7F’FFFF’FFFFH); otherwise, the MV-flag is cleared. Note that if the MV-flag indicates an arithmetic overflow, the result of the integer addition, integer subtraction, or accumulation is not valid. 4.9.10 The Repeat Counter MRW The Repeat Counter MRW controls the number of repetitions a loop must be executed. The register must be pre-loaded before it can be used with -USRx CoXXX operations. MAC operations are able to decrement this counter. When a -USRx CoXXX instruction is executed, MRW is checked for zero before being decremented. If MRW equals zero, bit USRx is set and MRW is not further decremented. Register MRW can be accessed via any instruction capable of accessing a SFR. MRW MAC Repeat Word 15 14 13 12 SFR (FFDAH/EDH) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 REPEAT_COUNT rwh Field Bits Type Description REPEAT_ COUNT [15:0] rwh 16-bit loop counter All CoXXX instructions have a 3-bit wide repeat control field ‘rrr’ (bit positions [31:29]) in the operand field to control the MRW repeat counter. Table 4-26 lists the possible encodings. User’s Manual CPUSV2_X, V2.2 4-72 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 4-26 Encoding of MAC Repeat Word Control Code in ‘rrr’ Effect on Repeat Counter 000B regular CoXXX instruction 001B RESERVED 010B ‘-USR0 CoXXX’ instruction, decrements repeat counter and sets bit USR0 if MRW is zero 011B ‘-USR1 CoXXX’ instruction, decrements repeat counter and sets bit USR1 if MRW is zero 1XXB RESERVED Note: Bit USR0 has been a general purpose flag also in previous architectures. To prevent collisions due to using this flag by programmer or compiler, use ‘-USR0 C0XXX’ instructions very carefully. The following example shows a loop which is executed 20 times. Every time the CoMACM instruction is executed, the MRW counter is decremented. MOV loop01: -USR1 MRW, #19 ;Pre-load loop counter CoMACM [IDX0+], [R0+] ;Calculate and decrement MSW ADD R2,#0002H JMPA cc_nusr1, loop01 ;Repeat loop until USR1 is set Note: Because correctly predicted JMPA is executed in 0-cycle, it offers the functionality of a repeat instruction. User’s Manual CPUSV2_X, V2.2 4-73 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 4.10 Constant Registers All bits of these bit-addressable registers are fixed to 0 or 1 by hardware. These registers can be read only. Register ZEROS/ONES can be used as a register-addressable constant of all zeros or all ones, for example for bit manipulation or mask generation. The constant registers can be accessed via any instruction capable of addressing an SFR. ZEROS Zeros Register SFR (FF1CH/8EH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 r r r r r r r r r r r r r r r r ONES Ones Register SFR (FF1EH/8FH) Reset Value: FFFFH 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 r r r r r r r r r r r r r r r r User’s Manual CPUSV2_X, V2.2 4-74 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5 Interrupt and Trap Functions The architecture of the XC167 supports several mechanisms for fast and flexible response to service requests from various sources internal or external to the microcontroller. Different kinds of exceptions are handled in a similar way: • • • Interrupts generated by the Interrupt Controller (ITC) DMA transfers issued by the Peripheral Event Controller (PEC) Traps caused by the TRAP instruction or issued by faults or specific system states Normal Interrupt Processing The CPU temporarily suspends current program execution and branches to an interrupt service routine to service an interrupt requesting device. The current program status (IP, PSW, also CSP in segmentation mode) is saved on the internal system stack. A prioritization scheme with 16 priority levels allows the user to specify the order in which multiple interrupt requests are to be handled. Interrupt Processing via the Peripheral Event Controller (PEC) A faster alternative to normal software controlled interrupt processing is servicing an interrupt requesting device with the XC167’s integrated Peripheral Event Controller (PEC). Triggered by an interrupt request, the PEC performs a single word or byte data transfer between any two locations through one of eight programmable PEC Service Channels. During a PEC transfer, normal program execution of the CPU is halted. No internal program status information needs to be saved. The same prioritization scheme is used for PEC service as for normal interrupt processing. Trap Functions Trap functions are activated in response to special conditions that occur during the execution of instructions. A trap can also be caused externally by the Non-Maskable Interrupt pin, NMI. Several hardware trap functions are provided to handle erroneous conditions and exceptions arising during instruction execution. Hardware traps always have highest priority and cause immediate system reaction. The software trap function is invoked by the TRAP instruction that generates a software interrupt for a specified interrupt vector. For all types of traps, the current program status is saved on the system stack. External Interrupt Processing Although the XC167 does not provide dedicated interrupt pins, it allows connection of external interrupt sources and provides several mechanisms to react to external events including standard inputs, non-maskable interrupts, and fast external interrupts. Except for the non-maskable interrupt and the reset input, these interrupt functions are alternate port functions. User’s Manual ICU_X73, V2.1 5-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.1 Interrupt System Structure The XC167 provides 80 separate interrupt nodes assignable to 16 priority levels, with 8 sub-levels (group priority) on each level. In order to support modular and consistent software design techniques, most sources of an interrupt or PEC request are supplied with a separate interrupt control register and an interrupt vector. The control register contains the interrupt request flag, the interrupt enable bit, and the interrupt priority of the associated source. Each source request is then activated by one specific event, determined by the selected operating mode of the respective device. For efficient resource usage, multi-source interrupt nodes are also incorporated. These nodes can be activated by several source requests, such as by different kinds of errors in the serial interfaces. However, specific status flags which identify the type of error are implemented in the serial channels’ control registers. Additional sharing of interrupt nodes is supported via interrupt subnode control registers. The XC167 provides a vectored interrupt system. In this system specific vector locations in the memory space are reserved for the reset, trap, and interrupt service functions. Whenever a request occurs, the CPU branches to the location that is associated with the respective interrupt source. This allows direct identification of the source which caused the request. The Class B hardware traps all share the same interrupt vector. The status flags in the Trap Flag Register (TFR) can then be used to determine which exception caused the trap. For the special software TRAP instruction, the vector address is specified by the operand field of the instruction, which is a seven bit trap number. The reserved vector locations build a jump table in the low end of a segment (selected by register VECSEG) in the XC167’s address space. The jump table consists of the appropriate jump instructions which transfer control to the interrupt or trap service routines and which may be located anywhere within the address space. The entries of the jump table are located at the lowest addresses in the selected code segment. Each entry occupies 2, 4, 8, or 16 words (selected by bitfield VECSC in register CPUCON1), providing room for at least one doubleword instruction. The respective vector location results from multiplying the trap number by the selected step width (2(VECSC+2)). All pending interrupt requests are arbitrated. The arbitration winner is indicated to the CPU together with its priority level and action request. The CPU triggers the corresponding action based on the required functionality (normal interrupt, PEC, jump table cache, etc.) of the arbitration winner. An action request will be accepted by the CPU if the requesting source has a higher priority than the current CPU priority level and interrupts are globally enabled. If the requesting source has a lower (or equal) interrupt level priority than the current CPU task, it remains pending. User’s Manual ICU_X73, V2.1 5-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Interrupt and Peripheral Event Controller PEC Pointer Interrupt Request Lines SRCP0 DSTP0 PECSEG0 SRCP1 DSTP1 PECSEG1 SRCP7 DSTP7 PECSEG7 irq0 irq1 C166S V2 CPU PEC Request irq2 irq3 Arbitration irq n-3 irq n-1 irq n-2 1) Arbitr. Winner Peripheral Event Controller (PEC) EOP INT 2) Request Control Request Control Interrupt Handler Interrupt Request Arbitration Control PEC Control Interrupt Handler Control (Interrupt Control Registers) (PEC Control Registers) Fast Bank Switching Interrupt Request Injection Control (CPU Action Request) Injection Interface OCE/OCDS OCE Injection Request & Control BNKSEL0 irq0IC PECC0 irq1IC PECC1 BNKSEL3 Interrupt Jump Table Cache irq126IC PECC7 EOPIC PECISNC FINT0CSP FINT0ADDR FINT1CSP FINT1ADDR 1) 2) Number of interrupt nodes n (up to 128) End of PEC Interrupt (EOPINT) is connected to Interrupt request line irq n-1. Therefore, only n-1 interrupt lines (irq n-2 ... 0) are available for peripheral request handling. Figure 5-1 User’s Manual ICU_X73, V2.1 MCB04915 Block Diagram of the Interrupt and PEC Controller 5-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.2 Interrupt Arbitration and Control The XC167’s interrupt arbitration system handles interrupt requests from up to 80 sources. Interrupt requests may be triggered either by the on-chip peripherals or by external inputs. Interrupt processing is controlled globally by register PSW through a general interrupt enable bit (IEN) and the CPU priority field (ILVL). Additionally, the different interrupt sources are controlled individually by their specific interrupt control registers (… IC). Thus, the acceptance of requests by the CPU is determined by both the individual interrupt control registers and by the PSW. PEC services are controlled by the respective PECCx register and by the source and destination pointers which specify the task of the respective PEC service channel. An interrupt request sets the associated interrupt request flag xxIR. If the requesting interrupt node is enabled by the associated interrupt enable bit xxIE arbitration starts with the next clock cycle, or after completion of an arbitration cycle that is already in progress. All interrupt requests pending at the beginning of a new arbitration cycle are considered, independently from when they were actually requested. Figure 5-2 shows the three-stage interrupt prioritization scheme: Hardware Traps OCDS break request OCDS or OCE Interrupt Request Lines Request Lines Arbitration xxxx (ILVL)+ x.xx (XGLVL) PEC/ Interrupt Handler xxxxx (OCDS service request priority level) 0xxxx (ILVL extended with 0 in MSB) CPU CPU Action Control xxxxx (request priority level) 0xxxx (ILVL.PSW extended with 0 in MSB) CPU Arbitration PSW Stage 1: Compared 4-Bit ILVL+ 2/3-Bit XGLVL priority levels of interrupt sources (64/128 priority levels) Stage 2: 4-Bit IRQ/PEC priority level comparated with 5-Bit OCDS priority level Stage 3: 5-Bit request priority level comparated with 4-Bit PSW priority level MCD04913 Figure 5-2 User’s Manual ICU_X73, V2.1 Interrupt Arbitration 5-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions The interrupt prioritization is done in three stages: • • • Select one of the active interrupt requests Compare the priority levels of the selected request and an OCDS service request Compare the priority level of the final request with the CPU priority level The First Arbitration Stage compares the priority levels of the active interrupt request lines. The interrupt priority level of each requestor is defined by bitfield ILVL in the respective xxIC register. The extended group priority level XGLVL (combined from bitfields GPX and GLVL) defines up to eight sub-priorities within one interrupt level. The group priority level distinguishes interrupt requests assigned to the same priority level, so one winner can be determined. Note: All interrupt request sources that are enabled and programmed to the same interrupt priority level (ILVL) must have different group priority levels. Otherwise, an incorrect interrupt vector will be generated. The Second Arbitration Stage compares the priority of the first stage winner with the priority of OCDS service requests. OCDS service requests bypass the first stage of arbitration and go directly to the CPU Action Control Unit. The CPU Action Control Unit compares the winner’s 4-bit priority level (disregarding the group level) with the 5-bit OCDS service request priority. The 4-bit ILVL of the interrupt request is extended to a 5-bit value with MSB = 0. This means that any OCDS request with MSB = 1 will always win the second stage arbitration. However, if there is a conflict between an OCDS request and an interrupt request, the interrupt request wins. The Third Arbitration Stage compares the priority level of the second stage winner with the priority of the current CPU task. An action request will be accepted by the CPU only if the priority level of the request is higher than the current CPU priority level (bitfield ILVL in register PSW) and if interrupt and PEC requests are globally enabled by the global interrupt enable flag IEN in register PSW. To compare with the 5-bit priority level of the second stage winner, the 4-bit CPU priority level is extended to a 5-bit value with MSB = 0. This means that any request with MSB = 1 will always interrupt the current CPU task. If the requestor has a priority level lower than or equal to the current CPU task, the request remains pending. Note: Priority level 0000B is the default level of the CPU. Therefore, a request on interrupt priority level 0000B will be arbitrated, but the CPU will never accept an action request on this level. However, every individually enabled interrupt request (including all denied interrupt requests and priority level 0000B requests) triggers a CPU wake-up from idle state independent of the global interrupt enable bit IEN. User’s Manual ICU_X73, V2.1 5-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Both the OCDS break requests and the hardware traps bypass the arbitration scheme and go directly to the core (see also Figure 5-2). The arbitration process starts with an enabled interrupt request and stays active as long as an interrupt request is pending. If no interrupt request is pending the arbitration is stopped to save power. Interrupt Control Registers The control functions for each interrupt node are grouped in a dedicated interrupt control register (xxIC, where “xx” stands for a mnemonic for the respective node). All interrupt control registers are organized identically. The lower 9 bits of an interrupt control register contain the complete interrupt control and status information of the associated source required during one round of prioritization (arbitration cycle); the upper 7 bits are reserved for future use. All interrupt control registers are bit-addressable and all bits can be read or written via software. Therefore, each interrupt source can be programmed or modified with just one instruction. xxIC Interrupt Control Register (E)SFR (yyyyH/zzH) 15 14 13 12 11 10 9 - - - - - - - - - - - - - - 8 7 6 Reset Value: - 000H 5 GPX xxIR xxIE rw rwh rw 4 3 2 1 0 ILVL GLVL rw rw Field Bits Type Description GPX 8 rw Group Priority Extension Completes bitfield GLVL to the 3-bit group level xxIR1) 7 rwh Interrupt Request Flag 0 No request pending 1 This source has raised an interrupt request xxIE 6 rw Interrupt Enable Control Bit (individually enables/disables a specific source) 0 Interrupt request is disabled 1 Interrupt request is enabled ILVL [5:2] rw Interrupt Priority Level FH Highest priority level … … 0H Lowest priority level User’s Manual ICU_X73, V2.1 5-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Field Bits Type Description GLVL [1:0] rw Group Priority Level (Is completed by bit GPX to the 3-bit group level) 3H Highest priority level … … 0H Lowest priority level 1) Bit xxIR supports bit-protection. When accessing interrupt control registers through instructions which operate on word data types, their upper 7 bits (15 … 9) will return zeros when read, and will discard written data. It is recommended to always write zeros to these bit positions. The layout of the interrupt control registers shown below applies to each xxIC register, where “xx” represents the mnemonic for the respective source. The Interrupt Request Flag is set by hardware whenever a service request from its respective source occurs. It is cleared automatically upon entry into the interrupt service routine or upon a PEC service. In the case of PEC service, the Interrupt Request flag remains set if the COUNT field in register PECCx of the selected PEC channel decrements to zero and bit EOPINT is cleared. This allows a normal CPU interrupt to respond to a completed PEC block transfer on the same priority level. Note: Modifying the Interrupt Request flag via software causes the same effects as if it had been set or cleared by hardware. The Interrupt Enable Control Bit determines whether the respective interrupt node takes part in the arbitration process (enabled) or not (disabled). The associated request flag will be set upon a source request in any case. The occurrence of an interrupt request can so be polled via xxIR even while the node is disabled. Note: In this case the interrupt request flag xxIR is not cleared automatically but must be cleared via software. Interrupt Priority Level and Group Level The four bits of bitfield ILVL specify the priority level of a service request for the arbitration of simultaneous requests. The priority increases with the numerical value of ILVL: so, 0000B is the lowest and 1111B is the highest priority level. When more than one interrupt request on a specific level becomes active at the same time, the values in the respective bitfields GPX and GLVL are used for second level arbitration to select one request to be serviced. Again, the group priority increases with the numerical value of the concatenation of bitfields GPX and GLVL, so 000B is the lowest and 111B is the highest group priority. Note: All interrupt request sources enabled and programmed to the same priority level must always be programmed to different group priorities. Otherwise, an incorrect interrupt vector will be generated. User’s Manual ICU_X73, V2.1 5-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Upon entry into the interrupt service routine, the priority level of the source that won the arbitration and whose priority level is higher than the current CPU level, is copied into bitfield ILVL of register PSW after pushing the old PSW contents onto the stack. The interrupt system of the XC167 allows nesting of up to 15 interrupt service routines of different priority levels (level 0 cannot be arbitrated). Interrupt requests programmed to priority levels 15 … 8 (i.e., ILVL = 1XXXB) can be serviced by the PEC if the associated PEC channel is properly assigned and enabled (please refer to Section 5.4). Interrupt requests programmed to priority levels 7 through 1 will always be serviced by normal interrupt processing. Note: Priority level 0000B is the default level of the CPU. Therefore, a request on level 0 will never be serviced because it can never interrupt the CPU. However, an individually enabled interrupt request (independent of bit IEN) on level 0000B will terminate the XC167’s Idle mode and reactivate the CPU. General Interrupt Control Functions in Register PSW The acceptance of an interrupt request depends on the current CPU priority level (bitfield ILVL in register PSW) and the global interrupt enable control bit IEN in register PSW (see Section 4.8). CPU Priority ILVL defines the current level for the operation of the CPU. This bitfield reflects the priority level of the routine currently executed. Upon entry into an interrupt service routine, this bitfield is updated with the priority level of the request being serviced. The PSW is saved on the system stack before the request is serviced. The CPU level determines the minimum interrupt priority level which will be serviced. Any request on the same or a lower level will not be acknowledged. The current CPU priority level may be adjusted via software to control which interrupt request sources will be acknowledged. PEC transfers do not really interrupt the CPU, but rather “steal” a single cycle, so PEC services do not influence the ILVL field in the PSW. Hardware traps switch the CPU level to maximum priority (i.e. 15) so no interrupt or PEC requests will be acknowledged while an exception trap service routine is executed. Note: The TRAP instruction does not change the CPU level, so software invoked trap service routines may be interrupted by higher requests. Interrupt Enable bit IEN globally enables or disables PEC operation and the acceptance of interrupts by the CPU. When IEN is cleared, no new interrupt requests are accepted by the CPU (see also Section 4.3.4). When IEN is set to 1, all interrupt sources, which have been individually enabled by the interrupt enable bits in their associated control registers, are globally enabled. Traps are non-maskable and are, therefore, not affected by the IEN bit. Note: To generate requests, interrupt sources must be also enabled by the interrupt enable bits in their associated control register. User’s Manual ICU_X73, V2.1 5-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Register Bank Select bitfield BANK defines the currently used register bank for the CPU operation. When the CPU enters an interrupt service routine, this bitfield is updated to select the register bank associated with the serviced request: • • • • Requests on priority levels 15 … 12 use the register bank pre-selected via the respective bitfield GPRSELx in the corresponding BNKSEL register Requests on priority levels 11 … 1 always use the global register bank, i.e. BANK = 00B Hardware traps always use the global register bank, i.e. BANK = 00B The TRAP instruction does not change the current register bank User’s Manual ICU_X73, V2.1 5-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.3 Interrupt Vector Table The XC167 provides a vectored interrupt system. This system reserves a set of specific memory locations, which are accessed automatically upon the respective trigger event. Entries for the following events are provided: • • • Reset (hardware, software, watchdog) Traps (hardware-generated by fault conditions or via TRAP instruction) Interrupt service requests Whenever a request is accepted, the CPU branches to the location associated with the respective trigger source. This vector position directly identifies the source causing the request, with two exceptions: • • Class B hardware traps all share the same interrupt vector. The status flags in the Trap Flag Register (TFR) are used to determine which exception caused the trap. For details, see Section 5.11. An interrupt node may be shared by several interrupt requests, e.g. within a module. Additional flags identify the requesting source, so the software can handle each request individually. For details, see Section 5.7. The reserved vector locations build a vector table located in the address space of the XC167. The vector table usually contains the appropriate jump instructions that transfer control to the interrupt or trap service routines. These routines may be located anywhere within the address space. The location and organization of the vector table is programmable. The Vector Segment register VECSEG defines the segment of the Vector Table (can be located in all segments, except for reserved areas). Bitfield VECSC in register CPUCON1 defines the space between two adjacent vectors (can be 2, 4, 8, or 16 words). For a summary of register CPUCON1, please refer to Section 4.4. Each vector location has an offset address to the segment base address of the vector table (given by VECSEG). The offset can be easily calculated by multiplying the vector number with the vector space programmed in bitfield VECSC. Table 5-2 lists all sources capable of requesting interrupt or PEC service in the XC167, the associated interrupt vector locations, the associated vector numbers, and the associated interrupt control registers. Note: All interrupt nodes which are currently not used by their associated modules or are not connected to a module in the actual derivative may be used to generate software controlled interrupt requests by setting the respective IR flag. User’s Manual ICU_X73, V2.1 5-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions VECSEG Vector Segment Pointer SFR (FF12H/89H) 7 6 Reset Value: Table 5-1 15 14 13 12 11 10 9 8 5 4 3 - - - - - - - - vecseg - - - - - - - - rwh Field Bits Type Description vecseg [7:0] rwh Segment number of the Vector Table 2 1 0 The reset value of register VECSEG, that means the initial location of the vector table, depends on the reset configuration. Table 5-1 lists the possible locations. This is required because the vector table also provides the reset vector. Table 5-1 Reset Values for Register VECSEG Initial Value Reset Configuration 0000H Standard start from external memory 0041H Alternate start from external memory 00C0H Standard start from Internal Program Memory 00C1H Alternate start from Internal Program Memory 00E0H Execute bootstrap loader code User’s Manual ICU_X73, V2.1 5-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-2 XC167 Interrupt Nodes Source of Interrupt or PEC Service Request Control Register Vector Location1) Vector Number CAPCOM Register 0 CC1_CC0IC xx’0040H 10H / 16D CAPCOM Register 1 CC1_CC1IC xx’0044H 11H / 17D CAPCOM Register 2 CC1_CC2IC xx’0048H 12H / 18D CAPCOM Register 3 CC1_CC3IC xx’004CH 13H / 19D CAPCOM Register 4 CC1_CC4IC xx’0050H 14H / 20D CAPCOM Register 5 CC1_CC5IC xx’0054H 15H / 21D CAPCOM Register 6 CC1_CC6IC xx’0058H 16H / 22D CAPCOM Register 7 CC1_CC7IC xx’005CH 17H / 23D CAPCOM Register 8 CC1_CC8IC xx’0060H 18H / 24D CAPCOM Register 9 CC1_CC9IC xx’0064H 19H / 25D CAPCOM Register 10 CC1_CC10IC xx’0068H 1AH / 26D CAPCOM Register 11 CC1_CC11IC xx’006CH 1BH / 27D CAPCOM Register 12 CC1_CC12IC xx’0070H 1CH / 28D CAPCOM Register 13 CC1_CC13IC xx’0074H 1DH / 29D CAPCOM Register 14 CC1_CC14IC xx’0078H 1EH / 30D CAPCOM Register 15 CC1_CC15IC xx’007CH 1FH / 31D CAPCOM Register 16 CC2_CC16IC xx’00C0H 30H / 48D CAPCOM Register 17 CC2_CC17IC xx’00C4H 31H / 49D CAPCOM Register 18 CC2_CC18IC xx’00C8H 32H / 50D CAPCOM Register 19 CC2_CC19IC xx’00CCH 33H / 51D CAPCOM Register 20 CC2_CC20IC xx’00D0H 34H / 52D CAPCOM Register 21 CC2_CC21IC xx’00D4H 35H / 53D CAPCOM Register 22 CC2_CC22IC xx’00D8H 36H / 54D CAPCOM Register 23 CC2_CC23IC xx’00DCH 37H / 55D CAPCOM Register 24 CC2_CC24IC xx’00E0H 38H / 56D CAPCOM Register 25 CC2_CC25IC xx’00E4H 39H / 57D CAPCOM Register 26 CC2_CC26IC xx’00E8H 3AH / 58D CAPCOM Register 27 CC2_CC27IC xx’00ECH 3BH / 59D CAPCOM Register 28 CC2_CC28IC xx’00E0H 3CH / 60D CAPCOM Register 29 CC2_CC29IC xx’0110H 44H / 68D User’s Manual ICU_X73, V2.1 5-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-2 XC167 Interrupt Nodes (cont’d) Source of Interrupt or PEC Service Request Control Register Vector Location1) Vector Number CAPCOM Register 30 CC2_CC30IC xx’0114H 45H / 69D CAPCOM Register 31 CC2_CC31IC xx’0118H 46H / 70D CAPCOM Timer 0 CC1_T0IC xx’0080H 20H / 32D CAPCOM Timer 1 CC1_T1IC xx’0084H 21H / 33D CAPCOM Timer 7 CC2_T7IC xx’00F4H 3DH / 61D CAPCOM Timer 8 CC2_T8IC xx’00F8H 3EH / 62D GPT1 Timer 2 GPT12E_T2IC xx’0088H 22H / 34D GPT1 Timer 3 GPT12E_T3IC xx’008CH 23H / 35D GPT1 Timer 4 GPT12E_T4IC xx’0090H 24H / 36D GPT2 Timer 5 GPT12E_T5IC xx’0094H 25H / 37D GPT2 Timer 6 GPT12E_T6IC xx’0098H 26H / 38D GPT2 CAPREL Reg. GPT12E_CRIC xx’009CH 27H / 39D A/D Conversion Compl. ADC_CIC xx’00A0H 28H / 40D A/D Overrun Error ADC_EIC xx’00A4H 29H / 41D ASC0 Transmit ASC0_TIC xx’00A8H 2AH / 42D ASC0 Transmit Buffer ASC0_TBIC xx’011CH 47H / 71D ASC0 Receive ASC0_RIC xx’00ACH 2BH / 43D ASC0 Error ASC0_EIC xx’00B0H 2CH / 44D ASC0 Autobaud ASC0_ABIC xx’017CH 5FH / 95D SSC0 Transmit SSC0_TIC xx’00B4H 2DH / 45D SSC0 Receive SSC0_RIC xx’00B8H 2EH / 46D SSC0 Error SSC0_EIC xx’00BCH 2FH / 47D IIC Data Transfer Event IIC_DTIC xx’0100H 40H / 64D IIC Protocol Event IIC_PEIC xx’0104H 41H / 65D PLL/OWD PLL_IC xx’010CH 43H / 67D ASC1 Transmit ASC1_TIC xx’0120H 48H / 72D ASC1 Transmit Buffer ASC1_TBIC xx’0178H 5EH / 94D ASC1 Receive ASC1_RIC xx’0124H 49H / 73D ASC1 Error ASC1_EIC xx’0128H 4AH / 74D ASC1 Autobaud ASC1_ABIC xx’0108H 42H / 66D User’s Manual ICU_X73, V2.1 5-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-2 XC167 Interrupt Nodes (cont’d) Source of Interrupt or PEC Service Request Control Register Vector Location1) Vector Number End of PEC Subchannel EOPIC xx’0130H 4CH / 76D CAPCOM6 Timer T12 CCU6_T12IC xx’0134H 4DH / 77D CAPCOM6 Timer T13 CCU6_T13IC xx’0138H 4EH / 78D CAPCOM6 Emergency CCU6_EIC xx’013CH 4FH / 79D CAPCOM6 CCU6_IC xx’0140H 50H / 80D SSC1 Transmit SSC1_TIC xx’0144H 51H / 81D SSC1 Receive SSC1_RIC xx’0148H 52H / 82D SSC1 Error SSC1_EIC xx’014CH 53H / 83D CAN0 CAN_0IC xx’0150H 54H / 84D CAN1 CAN_1IC xx’0154H 55H / 85D CAN2 CAN_2IC xx’0158H 56H / 86D CAN3 CAN_3IC xx’015CH 57H / 87D CAN4 CAN_4IC xx’0164H 59H / 89D CAN5 CAN_5IC xx’0168H 5AH / 90D CAN6 CAN_6IC xx’016CH 5BH / 91D CAN7 CAN_7IC xx’0170H 5CH / 92D RTC RTC_IC xx’0174H 5DH / 93D Unassigned node --- xx’00FCH 3FH / 63D Unassigned node --- xx’012CH 4BH / 75D Unassigned node --- xx’0160H 58H / 88D 1) Register VECSEG defines the segment where the vector table is located to. Bitfield VECSC in register CPUCON1 defines the distance between two adjacent vectors. This table represents the default setting, with a distance of 4 (two words) between two vectors. User’s Manual ICU_X73, V2.1 5-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-3 lists the vector locations for hardware traps and the corresponding status flags in register TFR. It also lists the priorities of trap service for those cases in which more than one trap condition might be detected within the same instruction. After any reset (hardware reset, software reset instruction SRST, or reset by watchdog timer overflow) program execution starts at the reset vector at location xx’0000H. Reset conditions have priority over every other system activity and, therefore, have the highest priority (trap priority III). Software traps may be initiated to any defined vector location. A service routine entered via a software TRAP instruction is always executed on the current CPU priority level which is indicated in bitfield ILVL in register PSW. This means that routines entered via the software TRAP instruction can be interrupted by all hardware traps or higher level interrupt requests. Table 5-3 Hardware Trap Summary Exception Condition Trap Flag Reset Functions: • Hardware Reset • Software Reset • W-dog Timer Overflow – Vector Trap Trap Vector Vector 1) Location Number Priority RESET RESET RESET xx’0000H xx’0000H xx’0000H 00H 00H 00H III III III NMI STKOF STKUF SOFTBRK NMITRAP STOTRAP STUTRAP SBRKTRAP xx’0008H xx’0010H xx’0018H xx’0020H 02H 04H 06H 08H II II II II UNDOPC PACER PRTFLT BTRAP BTRAP BTRAP xx’0028H xx’0028H xx’0028H 0AH 0AH 0AH I I I ILLOPA BTRAP xx’0028H 0AH I Reserved – – [2CH 3CH] [0BH 0FH] – Software Traps • TRAP Instruction – – Any1) Any [00H 7FH] Current CPU Priority Class A Hardware Traps: • Non-Maskable Interrupt • Stack Overflow • Stack Underflow • Software Break Class B Hardware Traps: • Undefined Opcode • PMI Access Error • Protected Instruction Fault • Illegal Word Operand Access 1) Register VECSEG defines the segment where the vector table is located to. Bitfield VECSC in register CPUCON1 defines the distance between two adjacent vectors. This table represents the default setting, with a distance of 4 (two words) between two vectors. User’s Manual ICU_X73, V2.1 5-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Interrupt Jump Table Cache Servicing an interrupt request via the vector table usually incurs two subsequent branches: an implicit branch to the vector location and an explicit branch to the actual service routine. The interrupt servicing time can be reduced by the Interrupt Jump Table Cache (ITC, also called “fast interrupt”). This feature eliminates the second explicit branch by directly providing the CPU with the service routine’s location. The ITC provides two 24-bit pointers, so the CPU can directly branch to the respective service routines. These fast interrupts can be selected for two interrupt sources on priority levels 15 … 12. The two pointers are each stored in a pair of interrupt jump table cache registers (FINTxADDR, FINTxCSP), which store a pointer’s segment and offset along with the priority level it shall be assigned to (select the same priority that is programmed for the respective interrupt node). FINT0ADDR Fast Interrupt Address Reg. 0 15 14 13 12 11 10 FINT1ADDR Fast Interrupt Address Reg. 1 15 14 13 12 11 10 XSFR (EC02H/--) 9 8 7 6 Reset Value: 0000H 5 4 2 1 0 ADDR 0 rw r XSFR (EC06H/--) 9 3 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 ADDR 0 rw r Field Bits Type Description ADDR [15:1] rw Address of Interrupt Service Routine Specifies address bits 15 … 1 of the 24-bit pointer to the interrupt service routine. This word offset is concatenated with FINTxCSP.SEG. User’s Manual ICU_X73, V2.1 5-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions FINT0CSP Fast Interrupt Control Reg. 0 11 10 XSFR (EC00H/--) 15 14 13 12 EN - - GPX ILVL GLVL SEG rw - - rw rw rw rw FINT1CSP Fast Interrupt Control Reg. 1 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 XSFR (EC04H/--) 9 8 7 6 2 1 0 Reset Value: 0000H 15 14 13 12 5 4 3 EN - - GPX ILVL GLVL SEG rw - - rw rw rw rw 2 1 0 Field Bits Type Description EN 15 rw Fast Interrupt Enable 0 The interrupt jump table cache is not used 1 The interrupt jump table cache is enabled, the vector table entry for the specified request is bypassed, the cache pointer is used GPX 12 rw Group Priority Extension Used together with bitfield GLVL ILVL [11:10] rw Interrupt Priority Level This selects the interrupt priority (15 … 12) of the request this pointer shall be assigned to 00 Interrupt priority level 12 (1100B) 01 Interrupt priority level 13 (1101B) 10 Interrupt priority level 14 (1110B) 11 Interrupt priority level 15 (1111B) GLVL [9:8] rw Group Priority Level Together with bit GPX this selects the group priority of the request this pointer shall be assigned to SEG [7:0] rw Segment Number of Interrupt Service Routine Specifies address bits 23 … 16 of the 24-bit pointer to the interrupt service routine, is concatenated with FINTxADDR. User’s Manual ICU_X73, V2.1 5-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.4 Operation of the Peripheral Event Controller Channels The XC167’s Peripheral Event Controller (PEC) provides 8 PEC service channels which move a single byte or word between any two locations. A PEC transfer can be triggered by an interrupt service request and is the fastest possible interrupt response. In many cases a PEC transfer is sufficient to service the respective peripheral request (for example, serial channels, etc.). PEC transfers do not change the current context, but rather “steal” cycles from the CPU, so the current program status and context needs not to be saved and restored as with standard interrupts. The PEC channels are controlled by a dedicated set of registers which are assigned to dedicated PEC resources: • • • • A 24-bit source pointer for each channel A 24-bit destination pointer for each channel A Channel Counter/Control register (PECCx) for each channel, selecting the operating mode for the respective channel Two interrupt control registers to control the operation of block transfers The PECC registers control the action performed by the respective PEC channel. Transfer Size (bit BWT) controls whether a byte or a word is moved during a PEC service cycle. This selection controls the transferred data size and the increment step for the pointer(s) to be modified. Pointer Modification (bitfield INC) controls, which of the PEC pointers is incremented after the PEC transfer. If the pointers are not modified (INC = 00B), the respective channel will always move data from the same source to the same destination. Transfer Control (bitfield COUNT) controls if the respective PEC channel remains active after the transfer or not. Bitfield COUNT also generally enables a PEC channel (COUNT > 00H). The PECC registers also select the assignment of PEC channels to interrupt priority levels (bitfield PLEV) and the interrupt behavior after PEC transfer completion (bit EOPINT). Note: All interrupt request sources that are enabled and programmed for PEC service should use different channels. Otherwise, only one transfer will be performed for all simultaneous requests. When COUNT is decremented to 00H, and the CPU is to be interrupted, an incorrect interrupt vector will be generated. PEC transfers are executed only if their priority level is higher than the CPU level. User’s Manual ICU_X73, V2.1 5-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions PECCx PEC Control Reg. 13 12 SFR (FECyH/6zH, Table 5-4) 11 10 9 8 7 6 Reset Value: 0000H 15 14 5 4 3 - EOP INT PLEV CL INC BWT COUNT - rw rw rw rw rw rwh 2 1 0 Field Bits Type Description EOPINT 14 rw End of PEC Interrupt Selection 0 End of PEC interrupt on the same (PEC) level 1 End of PEC interrupt via separate node EOPIC PLEV [13:12] rw PEC Level Selection This bitfield controls the PEC channel assignment to an arbitration priority level (see section below) CL 11 rw Channel Link Control 0 PEC channels work independently 1 Pairs of PEC channels are linked together1) INC [10:9] rw Increment Control (Pointer Modification)2) 00 Pointers are not modified 01 Increment DSTPx by 1 or 2 (BWT = 1 or 0) 10 Increment SRCPx by 1 or 2 (BWT = 1 or 0) 11 Increment both DSTPx and SRCPx by 1 or 2 BWT 8 rw Byte/Word Transfer Selection 0 Transfer a word 1 Transfer a byte COUNT [7:0] rwh PEC Transfer Count Counts PEC transfers and influences the channel’s action (see Section 5.4.2) 1) For a functional description see “Channel Link Mode for Data Chaining”. 2) Pointers are incremented/decremented only within the current segment. Table 5-4 PEC Control Register Addresses Register Address PECC0 Register Address FEC0H / 60H SFR PECC4 FEC8H / 64H SFR PECC1 FEC2H / 61H SFR PECC5 FECAH / 65H SFR PECC2 FEC4H / 62H SFR PECC6 FECCH / 66H SFR PECC3 FEC6H / 63H SFR PECC7 FECEH / 67H SFR User’s Manual ICU_X73, V2.1 Reg. Space 5-19 Reg. Space V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions The PEC channel number is derived from the respective ILVL (LSB) and GLVL, where the priority band (ILVL) is selected by the channel’s bitfield PLEV (see Table 5-5). So, programming a source to priority level 15 (ILVL = 1111B) selects the PEC channel group 7 … 4 with PLEV = 00B; programming a source to priority level 14 (ILVL = 1110B) selects the PEC channel group 3 … 0 with PLEV = 00B; programming a source to priority level 10 (ILVL = 1010B) selects the PEC channel group 3 … 0 with PLEV = 10B. The actual PEC channel number is then determined by the group priority (levels 3 … 0, i.e. GPX = 0). Simultaneous requests for PEC channels are prioritized according to the PEC channel number, where channel 0 has lowest and channel 7 has highest priority. Note: All sources requesting PEC service must be programmed to different PEC channels. Otherwise, an incorrect PEC channel may be activated. Table 5-5 PEC Channel Assignment Selected PEC Channel Group Level Used Interrupt Priorities Depending on Bitfield PLEV PLEV = 00B PLEV = 01B PLEV = 10B PLEV = 11B 7 3 15 13 11 9 6 2 5 1 4 0 3 3 14 12 10 8 2 2 1 1 0 0 Table 5-6 shows in a few examples which action is executed with a given programming of an interrupt control register and a PEC channel. User’s Manual ICU_X73, V2.1 5-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-6 Interrupt Priority Examples Priority Level Type of Service Interr. Group COUNT = 00H, Level Level PLEV = XXB COUNT ≠ 00H, PLEV = 00B COUNT ≠ 00H, PLEV = 01B 1111 111 CPU interrupt, level 15, group prio 7 CPU interrupt, level 15, group prio 7 CPU interrupt, level 15, group prio 7 1111 011 CPU interrupt, level 15, group prio 3 PEC service, channel 7 CPU interrupt, level 15, group prio 3 1111 010 CPU interrupt, level 15, group prio 2 PEC service, channel 6 CPU interrupt, level 15, group prio 2 1110 010 CPU interrupt, level 14, group prio 2 PEC service, channel 2 CPU interrupt, level 14, group prio 2 1101 110 CPU interrupt, level 13, group prio 6 CPU interrupt, level 13, group prio 6 CPU interrupt, level 13, group prio 6 1101 010 CPU interrupt, level 13, group prio 2 CPU interrupt, level 13, group prio 2 PEC service, channel 6 0001 011 CPU interrupt, level 1, group prio 3 CPU interrupt, level 1, group prio 3 CPU interrupt, level 1, group prio 3 0001 000 CPU interrupt, level 1, group prio 0 CPU interrupt, level 1, group prio 0 CPU interrupt, level 1, group prio 0 0000 XXX No service! No service! No service! Note: PEC service is only achieved when bit GPX = 0 and COUNT ≠ 0. Requests on levels 7 … 1 cannot initiate PEC transfers. They are always serviced by an interrupt service routine: no PECC register is associated and no COUNT field is checked. User’s Manual ICU_X73, V2.1 5-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.4.1 The PEC Source and Destination Pointers The PEC channels’ source and destination pointers specify the locations between which the data is to be moved. Both 24-bit pointers are built by concatenating the 16-bit offset register (SRCPx or DSTPx) with the respective 8-bit segment bitfield (SRCSEGx or DSTSEGx, combined in register PECSEGx). PECSEGx SRCSEGx 15 DSTSEGx 8 7 SRCPx 0 DSTPx SRCPx 15 DSTPx 0 15 Source Pointer 23 16 15 Segment Address Destination Pointer 0 Segment Offset 0 23 16 15 Segment Address 0 Segment Offset Data Transfer MCD04916 x = 7 … 0, depending on PEC channel number Figure 5-3 PEC Data Pointers When a PEC pointer is automatically incremented after a transfer, only the offset part is incremented (SRCPx and/or DSTPx), while the respective segment part is not modified by hardware. Thus, a pointer may be incremented within the current segment, but may not cross the segment boundary. When a PEC pointer reaches the maximum offset (FFFEH for word transfers, FFFFH for byte transfers), it is not incremented further, but keeps its maximum offset value. This protects memory in adjacent segments from being overwritten unintentionally. No explicit error event is generated by the system in case of a pointer saturation; therefore, it is the user’s responsibility to prevent this condition. Note: PEC data transfers do not use the data page pointers DPP3 … DPP0. Unused PEC pointers may be used for general data storage. User’s Manual ICU_X73, V2.1 5-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions SRCPx PEC Source Pointer 15 14 13 12 XSFR (ECyyH/--, Table 5-7) 11 10 9 8 7 6 5 Reset Value: 0000H 4 3 2 1 0 srcpx rwh Field Bits Type Description srcpx [15:0] rwh Source Pointer Offset of Channel x Source address bits 15 … 0 DSTPx PEC Destination Pointer 15 14 13 12 11 XSFR (ECyyH/--, Table 5-7) 10 9 8 7 6 5 Reset Value: 0000H 4 3 2 1 0 dstpx rwh Field Bits Type Description dstpx [15:0] rwh Destination Pointer Offset of Channel x Destination address bits 15 … 0 PECSEGx PEC Segment Pointer 15 14 13 12 XSFR (ECyyH/--, Table 5-7) 11 10 9 8 7 6 5 Reset Value: 0000H 4 3 srcsegx dstsegx rw rw 2 1 Field Bits Type Description srcsegx [15:8] rw Source Pointer Segment of Channel x Source address bits 23 … 16 dstsegx [7:0] rw Destination Pointer Segment of Channel x Destination address bits 23 … 16 User’s Manual ICU_X73, V2.1 5-23 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-7 PEC Data Pointer Register Addresses Channel # 0 1 2 3 4 5 6 7 PECSEGx EC80H EC82H EC84H EC86H EC88H EC8AH EC8CH EC8EH SRCPx EC40H EC44H EC48H EC4CH EC50H EC54H EC58H DSTPx EC42H EC46H EC4AH EC4EH EC52H EC56H EC5AH EC5EH EC5CH Note: If word data transfer is selected for a specific PEC channel (BWT = 0), the respective source and destination pointers must both contain a valid word address which points to an even byte boundary. Otherwise, the Illegal Word Access trap will be invoked when this channel is used. 5.4.2 PEC Transfer Control The PEC Transfer Count Field COUNT controls the behavior of the respective PEC channel. The contents of bitfield COUNT select the action to be taken at the time the request is activated. COUNT may allow a specified number of PEC transfers, unlimited transfers, or no PEC service at all. Table 5-8 summarizes, how the COUNT field, the interrupt requests flag IR, and the PEC channel action depend on the previous contents of COUNT. Table 5-8 Influence of Bitfield COUNT Previous COUNT Modified COUNT IR after Service Action of PEC Channel and Comments FFH FFH 0 Move a Byte/Word Continuous transfer mode, i.e. COUNT is not modified FEH … 02H FDH … 01H 0 Move a Byte/Word and decrement COUNT 01H 00H 1 EOPINT = 0 (channel-specific interrupt) Move a Byte/Word Leave request flag set, which triggers another request 0 EOPINT = 1 (separate end-of-PEC interrupt) Move a Byte/Word Clear request flag, set the respective PEC subnode request flag CxIR instead1) – No PEC action! Activate interrupt service routine rather than PEC channel 00H 00H 1) Setting a subnode request flag also sets flag EOPIR if the subnode request is enabled (CxIE = 1). User’s Manual ICU_X73, V2.1 5-24 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions The PEC transfer counter allows service of a specified number of requests by the respective PEC channel, and then (when COUNT reaches 00H) activation of an interrupt service routine, either associated with the PEC channel’s priority level or with the general end-of-PEC interrupt. After each PEC transfer, the COUNT field is decremented (except for COUNT = FFH) and the request flag is cleared to indicate that the request has been serviced. When COUNT contains the value 00H, the respective PEC channel remains idle and the associated interrupt service routine is activated instead. This allows servicing requests on all priority levels by standard interrupt service routines. Continuous transfers are selected by the value FFH in bitfield COUNT. In this case, COUNT is not modified and the respective PEC channel services any request until it is disabled again. When COUNT is decremented from 01H to 00H after a transfer, a standard interrupt is requested which can then handle the end of the PEC block transfer (channel-specific interrupt or common end-of-PEC interrupt, see Table 5-8). User’s Manual ICU_X73, V2.1 5-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.4.3 Channel Link Mode for Data Chaining In channel link mode, every two PEC channels build a pair (channels 0+1, 2+3, 4+5, 6+7), where the two channels of a pair are activated in turn. Requests for the even channel trigger the currently active PEC channel (or the end-of-block interrupt), while requests for the odd channel only trigger its associated interrupt node. When the transfer count of one channel expires, control is switched to the other channel, and back. This mode supports data chaining where independent blocks of data can be transferred to the same destination (or vice versa), e.g. to build communication frames from several blocks, such as preamble, data, etc. Channel link mode for a pair of channels is enabled if at least one of the channel link control bits (bit CL in register PECCx) of the respective pair is set. A linked channel pair is controlled by the priority-settings (level, group) for its even channel. After enabling channel link mode the even channel is active. Channel linking is executed if the active channel’s link control bit CL is 1 at the time its transfer count decrements from 1 to 0 (count > 0 before) and the transfer count of the other channel is non-zero. In this case the active channel issues an EOP interrupt request and the respective other channel of the pair is automatically selected. Note: Channel linking always begins with the even channel. Channel linking is terminated if the active channel’s link control bit CL is 0 at the time its transfer count decrements from 1 to 0, or if the transfer count of the respective linked channel is zero. In this case an interrupt is triggered as selected by bit EOPINT (channelspecific or general EOP interrupt). A data-chaining sequence using PEC channel linking is programmed by setting bit CL together with a transfer count value (> 0). This is repeated, triggered by the channel link interrupts, for the complete sequence. For the last transfer, the interrupt routine should clear the respective bit CL, so, at the end of the complete transfer, either a standard or an END of PEC interrupt can be selected by bit EOPINT of the last channel. Note: To enable linking, initially both channels must receive a non-zero transfer count. For the rest of the sequence only the channel with the expired transfer count needs to be reconfigured. User’s Manual ICU_X73, V2.1 5-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.4.4 PEC Interrupt Control When the selected number of PEC transfers has been executed, the respective PEC channel is disabled and a standard interrupt service routine is activated instead. Each PEC channel can either activate the associated channel-specific interrupt node, or activate its associated PEC subnode request flag in register PECISNC, which then activates the common node request flag in register EOPIC (see Figure 5-4). PECISNC PEC Intr. Sub-Node Ctrl. Reg. 15 14 13 12 11 10 SFR (FFA8H/D4H) 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 C7IR C7IE C6IR C6IE C5IR C5IE C4IR C4IE C3IR C3IE C2IR C2IE C1IR C1IE C0IR C0IE rwh rw rwh rw rwh rw rwh rw rwh rw rwh rw rwh rw rwh rw Field Bits Type Description CxIR x=7…0 [2x+1] rwh Interrupt Request Flag of PEC Channel x 0 No request from PEC channel x pending 1 PEC channel x has raised an end-of-PEC interrupt request Note: These request flags must be cleared by SW. CxIE x=7…0 [2x] rw Interrupt Enable Control Bit of PEC Channel x (individually enables/disables a specific source) 0 End-of-PEC request of channel x disabled 1 End-of-PEC request of channel x enabled1) 1) It is recommended to clear an interrupt request flag (CxIR) before setting the respective enable flag (CxIE). Otherwise, former requests still pending cannot trigger a new interrupt request. EOPIC End-of-PEC Intr. Ctrl. Reg. ESFR (F180H/C0H) 15 14 13 12 11 10 9 8 7 - - - - - - - GPX - - - - - - - rw 6 EOP EOP IR IE rwh rw Reset Value: 0000H 5 4 3 2 1 0 ILVL GLVL rw rw Note: Please refer to the general Interrupt Control Register description for an explanation of the control fields. User’s Manual ICU_X73, V2.1 5-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions PECISNC C7IR C7IE C6IR C6IE C5IR C5IE C4IR C4IE C3IR C3IE C2IR C2IE C1IR C1IE C0IR C0IE & & & & & 1 & & 0 & 15 Interrupt Request Pulse Generator EOPIC 0 0 15 0 0 0 0 0 GPX EOP EOP IR IE 87 ILVL GLVL 0 MCD04914 Figure 5-4 End of PEC Interrupt Sub Node Note: The interrupt service routine must service and clear all currently active requests before terminating. Requests occurring later will set EOPIR again and the service routine will be re-entered. User’s Manual ICU_X73, V2.1 5-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.5 Prioritization of Interrupt and PEC Service Requests Interrupt and PEC service requests from all sources can be enabled so they are arbitrated and serviced (if they win), or they may be disabled, so their requests are disregarded and not serviced. Enabling and disabling interrupt requests may be done via three mechanisms: • • • Control Bits Priority Level ATOMIC and EXTended Instructions Control Bits allow switching of each individual source “ON” or “OFF” so that it may generate a request or not. The control bits (xxIE) are located in the respective interrupt control registers. All interrupt requests may be enabled or disabled generally via bit IEN in register PSW. This control bit is the “main switch” which selects if requests from any source are accepted or not. For a specific request to be arbitrated, the respective source’s enable bit and the global enable bit must both be set. The Priority Level automatically selects a certain group of interrupt requests to be acknowledged and ignores all other requests. The priority level of the source that won the arbitration is compared against the CPU’s current level and the source is serviced only if its level is higher than the current CPU level. Changing the CPU level to a specific value via software blocks all requests on the same or a lower level. An interrupt source assigned to level 0 will be disabled and will never be serviced. The ATOMIC and EXTend instructions automatically disable all interrupt requests for the duration of the following 1 … 4 instructions. This is useful for semaphore handling, for example, and does not require to re-enable the interrupt system after the inseparable instruction sequence. Interrupt Class Management An interrupt class covers a set of interrupt sources with the same importance, i.e. the same priority from the system’s viewpoint. Interrupts of the same class must not interrupt each other. The XC167 supports this function with two features: Classes with up to eight members can be established by using the same interrupt priority (ILVL) and assigning a dedicated group level to each member. This functionality is builtin and handled automatically by the interrupt controller. Classes with more than eight members can be established by using a number of adjacent interrupt priorities (ILVL) and the respective group levels (eight per ILVL). Each interrupt service routine within this class sets the CPU level to the highest interrupt priority within the class. All requests from the same or any lower level are blocked now, i.e. no request of this class will be accepted. User’s Manual ICU_X73, V2.1 5-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions The example shown below establishes 3 interrupt classes which cover 2 or 3 interrupt priorities, depending on the number of members in a class. A level 6 interrupt disables all other sources in class 2 by changing the current CPU level to 8, which is the highest priority (ILVL) in class 2. Class 1 requests or PEC requests are still serviced, in this case. In this way, the interrupt sources (excluding PEC requests) are assigned to 3 classes of priority rather than to 7 different levels, as the hardware support would do. Table 5-9 ILVL (Priority) Software Controlled Interrupt Classes (Example) Group Level Interpretation 7 6 5 4 3 2 1 0 15 PEC service on up to 8 channels 14 13 12 11 X X X X X X X X Interrupt Class 1 9 sources on 2 levels X 10 9 8 7 X X X X X X X X Interrupt Class 2 X X X X X X X X 17 sources on 3 levels 6 X 5 X X X X X X X X Interrupt Class 3 9 sources on 2 levels X 4 3 2 1 0 User’s Manual ICU_X73, V2.1 No service! 5-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.6 Context Switching and Saving Status Before an interrupt request that has been arbitrated is actually serviced, the status of the current task is automatically saved on the system stack. The CPU status (PSW) is saved together with the location at which execution of the interrupted task is to be resumed after returning from the service routine. This return location is specified through the Instruction Pointer (IP) and, in the case of a segmented memory model, the Code Segment Pointer (CSP). Bit SGTDIS in register CPUCON1 controls how the return location is stored. The system stack receives the PSW first, followed by the IP (unsegmented), or followed by CSP and then IP (segmented mode). This optimizes the usage of the system stack if segmentation is disabled. The CPU priority field (ILVL in PSW) is updated with the priority of the interrupt request to be serviced, so the CPU now executes on the new level. The register bank select field (BANK in PSW) is changed to select the register bank associated with the interrupt request. The association between interrupt requests and register banks are partly pre-defined and can partly be programmed. The interrupt request flag of the source being serviced is cleared. IP and CSP are loaded with the vector associated with the requesting source, and the first instruction of the service routine is fetched from the vector location which is expected to branch to the actual service routine (except when the interrupt jump table cache is used). All other CPU resources, such as data page pointers and the context pointer, are not affected. When the interrupt service routine is exited (RETI is executed), the status information is popped from the system stack in the reverse order, taking into account the value of bit SGTDIS. High Addresses Status of Interrupted Task SP --- SP -- PSW PSW IP CSP -- IP SP Low Addresses a) System Stack before Interrupt Entry b) System Stack after Interrupt Entry (Unsegmented) b) System Stack after Interrupt Entry (Segmented) MCD02226 Figure 5-5 User’s Manual ICU_X73, V2.1 Task Status Saved on the System Stack 5-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Context Switching An interrupt service routine usually saves all the registers it uses on the stack and restores them before returning. The more registers a routine uses, the more time is spent saving and restoring. The XC167 allows switching the complete bank of CPU registers (GPRs) either automatically or with a single instruction, so the service routine executes within its own separate context (see also Section 4.5.2). There are two ways to switch the context in the XC167 core: Switching Context of the Global Register Bank changes the complete global register bank of CPU registers (GPRs) by changing the Context Pointer with a single instruction, so the service routine executes within its own separate context. The instruction “SCXT CP, #New_Bank” pushes the contents of the context pointer (CP) on the system stack and loads CP with the immediate value “New_Bank”; this in turn, selects a new register bank. The service routine may now use its “own registers”. This register bank is preserved when the service routine terminates, i.e. its contents are available on the next call. Before returning (RETI), the previous CP is simply POPped from the system stack, which returns the registers to the original global bank. Resources used by the interrupting program, such as the DPPs, must eventually be saved and restored. Note: There are certain timing restrictions during context switching that are associated with pipeline behavior. Switching Context by changing the selected register bank automatically updates bitfield BANK to select one of the two local register banks or the current global register bank, so the service routine may now use its “own registers” directly. This local register bank is preserved when the service routine is terminated; thus, its contents are available on the next call. When switching to the global register bank, the service routine usually must also switch the context of the global register bank to get a private set of GPRs, because the global bank is likely to be used by several tasks. For interrupt priority levels 15 … 12 the target register bank can be pre-selected and then be switched automatically. The register bank selection registers BNKSELx provide a 2-bit field for each possible arbitration priority level. The respective bitfield is then copied to bitfield BANK in register PSW to select the register bank, as soon as the respective interrupt request is accepted. Table 5-10 identifies the arbitration priority level assignment to the respective bitfields within the four register bank selection registers. User’s Manual ICU_X73, V2.1 5-32 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions BNKSELx Register Bank Select Reg. x 15 14 13 12 11 XSFR (Table 5-10) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 GPRSEL7 GPRSEL6 GPRSEL5 GPRSEL4 GPRSEL3 GPRSEL2 GPRSEL1 GPRSEL0 rw rw rw rw rw rw Field Bits Type Description GPRSELy (y = 7 … 0) [2y+1 :2y] rw Register Bank Selection 00 Global register bank 01 Reserved 10 Local register bank 1 11 Local register bank 2 Table 5-10 rw rw Assignment of Register Bank Control Fields Bank Select Control Register Interrupt Node Priority Notes Register Name Bitfields Intr. Level Group Levels BNKSEL0 (EC20H/--) GPRSEL0 … 3 12 0…3 GPRSEL4 … 7 13 0…3 BNKSEL1 (EC22H/--) GPRSEL0 … 3 14 0…3 GPRSEL4 … 7 15 0…3 BNKSEL2 (EC24H/--) GPRSEL0 … 3 12 4…7 GPRSEL4 … 7 13 4…7 BNKSEL3 (EC26H/--) GPRSEL0 … 3 14 4…7 GPRSEL4 … 7 15 4…7 User’s Manual ICU_X73, V2.1 5-33 Lower group levels Upper group levels V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.7 Interrupt Node Sharing Interrupt nodes may be shared among several module requests if either the requests are generated mutually exclusively or the requests are generated at a low rate. If more than one source is enabled in this case, the interrupt handler will first need to determine the requesting source. However, this overhead is not critical for low rate requests. This node sharing is either controlled via interrupt sub-node control registers (ISNC) which provide separate request flags and enable bits for each supported request source, or the involved request sources are simply ORed to trigger the common node. The interrupt level used for arbitration is determined by the node control register (… IC). The specific request flags within ISNC registers must be reset by software, contrary to the node request bits which are cleared automatically. Table 5-11 Sub-Node Control Bit Allocation Interrupt Node Interrupt Sources Control EOPIC PEC channels 7 … 0 PECISNC RTC_IC RTC: overflow of T14, CNT0 … CNT3 RTC_ISNC ASC0_ABIC ASC0: autobaud detect start, error request ORed ASC1_ABIC ASC0: autobaud detect start, error request ORed IIC_DTIC IIC: data interrupt, end of data interrupt ORed User’s Manual ICU_X73, V2.1 5-34 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.8 External Interrupts Although the XC167 has no dedicated INTR input pins, it supports many possibilities to react to external asynchronous events. It does this by using a number of IO lines for interrupt input. The interrupt function may be either combined with the pin’s main function or used instead of it if the main pin function is not required. The Fast External Interrupt detection provides flexible wake-up signals even in sleep mode. This function can also generate additional interrupt requests from external input signals. Table 5-12 Pins Usable as External Interrupt Inputs Port Pin Original Function Control Register P7.7-4/CC31-28IO CAPCOM Register 31-28 Capture Input CC31-CC28 P1H.7-4/CC27-24IO CAPCOM Register 27-24 Capture Input CC27-CC24 P1H.0/CC23IO CAPCOM Register 23 Capture Input CC23 P1L.7/CC22IO CAPCOM Register 22 Capture Input CC22 P9.5-0/CC21-16IO CAPCOM Register 21-16 Capture Input CC21-CC16 P2.15-8/CC15-8IO CAPCOM Register 15-8 Capture Input CC15-CC8 P6.7-0/CC7-0IO CAPCOM Register 7-0 Capture Input CC7-CC0 P3.2/CAPIN GPT2 capture input pin T5CON P3.7/T2IN Auxiliary timer T2 input pin T2CON P3.5/T4IN Auxiliary timer T4 input pin T4CON For each of these pins, either a positive, a negative, or both a positive and a negative external transition can be selected to cause an interrupt or PEC service request. The edge selection is performed in the control register of the peripheral device associated with the respective port pin (separate control for fast external interrupts). The peripheral must be programmed to a specific operating mode to allow generation of an interrupt by the external signal. The priority of the interrupt request is determined by the interrupt control register of the respective peripheral interrupt source, and the interrupt vector of this source will be used to service the external interrupt request. Note: In order to use any of the listed pins as an external interrupt input, it must be switched to input mode via its direction control bit DPx.y in the respective port direction control register DPx. When port pins CCxIO are to be used as external interrupt input pins, bitfield CCMODx in the control register of the corresponding capture/compare register CCx must select capture mode. When CCMODx is programmed to 001B, the interrupt request flag CCxIR in register CCxIC will be set on a positive external transition at pin CCxIO. When User’s Manual ICU_X73, V2.1 5-35 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions CCMODx is programmed to 010B, a negative external transition will set the interrupt request flag. When CCMODx = 011B, both a positive and a negative transition will set the request flag. In all three cases, the contents of the allocated CAPCOM timer will be latched into capture register CCx, independent of whether or not the timer is running. When the interrupt enable bit CCxIE is set, a PEC request or an interrupt request for vector CCxINT will be generated. Pins T2IN or T4IN can be used as external interrupt input pins when the associated auxiliary timer T2 or T4 in block GPT1 is configured for capture mode. This mode is selected by programming the mode control fields T2M or T4M in control registers T2CON or T4CON to 101B. The active edge of the external input signal is determined by bitfields T2I or T4I. When these fields are programmed to X01B, interrupt request flags T2IR or T4IR in registers T2IC or T4IC will be set on a positive external transition at pins T2IN or T4IN, respectively. When T2I or T4I is programmed to X10B, then a negative external transition will set the corresponding request flag. When T2I or T4I is programmed to X11B, both a positive and a negative transition will set the request flag. In all three cases, the contents of the core timer T3 will be captured into the auxiliary timer registers T2 or T4 based on the transition at pins T2IN or T4IN. When the interrupt enable bits T2IE or T4IE are set, a PEC request or an interrupt request for vector T2INT or T4INT will be generated. Pin CAPIN differs slightly from the timer input pins as it can be used as external interrupt input pin without affecting peripheral functions. When the capture mode enable bit T5SC in register T5CON is cleared to ‘0’, signal transitions on pin CAPIN will only set the interrupt request flag CRIR in register CRIC, and the capture function of register CAPREL is not activated. So register CAPREL can still be used as reload register for GPT2 timer T5, while pin CAPIN serves as external interrupt input. Bitfield CI in register T5CON selects the effective transition of the external interrupt input signal. When CI is programmed to 01B, a positive external transition will set the interrupt request flag. CI = 10B selects a negative transition to set the interrupt request flag, and with CI = 11B, both a positive and a negative transition will set the request flag. When the interrupt enable bit CRIE is set, an interrupt request for vector CRINT or a PEC request will be generated. Note: The non-maskable interrupt input pin NMI and the reset input RSTIN provide another possibility for the CPU to react to an external input signal. NMI and RSTIN are dedicated input pins which cause hardware traps. User’s Manual ICU_X73, V2.1 5-36 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Fast External Interrupts The fast external interrupt pins are sampled every system clock cycle; that is, external events are scanned and detected in time frames of 1/fSYS. The arbitration and processing of these interrupt requests, however, is done with the normal timing. The External Interrupt Control register EXICON selects the trigger transition (rising, falling or both) individually for each of 8 fast external interrupts. These fast external interrupts use the interrupt nodes and vectors of the CAPCOM channels CC15 … CC8, so the capture/compare function cannot be used on the respective Port 2 pins (with EXIxES ≠ 00B). However, general purpose IO is possible in all cases. EXICON External Intr. Control Reg. 15 14 13 12 11 ESFR (F1C0H/E0H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 EXI7ES EXI6ES EXI5ES EXI4ES EXI3ES EXI2ES EXI1ES EXI0ES rw rw rw rw rw rw rw rw Field Bits Type Description EXIxES (x = 7 … 0) [15:14] … [1:0] rw External Interrupt x Edge Selection Field 00 Fast external interrupts disabled: std. mode 01 Interrupt on positive edge (rising) 10 Interrupt on negative edge (falling) 11 Interrupt on any edge (rising or falling) External Interrupt Source Control The input source for each of the fast external interrupts (controlled via register EXICON) can be derived from up to three associated port pins (standard pin EXnIN or two alternate sources). Activating an alternate input source, for example, allows the detection of transitions on the interface lines of disabled interfaces. Upon this trigger, the respective interface can be reactivated and respond to the detected activity. Source selection is controlled via registers EXISEL0 and EXISEL1. Besides selecting one of the three possible input pins, two or all of them can also be logically combined. This can be used to increase the number of wake-up lines or to define specific signal combinations to trigger a wake-up interrupt. User’s Manual ICU_X73, V2.1 5-37 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions EXISEL0 Ext. Interrupt Source Reg. 0 15 14 13 12 11 ESFR (F1DAH/EDH) 10 9 7 6 5 4 3 2 1 EXI3SS EXI2SS EXI1SS EXI0SS rw rw rw rw EXISEL1 Ext. Interrupt Source Reg. 1 15 8 Reset Value: 0000H 14 13 12 11 ESFR (F1D8H/ECH) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 EXI7SS EXI6SS EXI5SS EXI4SS rw rw rw rw Field Bits Type EXIxSS (x = 7 … 0) [15:12] rw … [3:0] 0 0 Description External Interrupt x Source Selection Field 0000 Input from associated EXxIN pin 0001 Input from alternate pin AltA 0010 Input from alternate pin AltB 0011 Input from pin EXxIN ORed with alternate pin AltA 0100 Input from pin EXxIN ANDed with alternate pin AltA 0101 Input from alternate pin AltA ORed with alternate pin AltB 0110 Input from alternate pin AltA ANDed with alternate pin AltB 0111 Input from pin EXxIN ORed with pin AltA ORed with pin AltB 1XXX Reserved, do not use The Table 5-13 summarizes the association of the bitfields of register EXISEL (i.e. the interrupt lines) with the respective input pins. User’s Manual ICU_X73, V2.1 5-38 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Table 5-13 Connection of Interrupt Inputs to External Interrupt Nodes Control Bitfield Std. Pin EXnIN Alternate Alternate Interrupt Associated Pin AltA Pin AltB Ctrl. Reg. Interface Notes EXI0SS P2.8 P1H.3 P1H.0 CC8IC SSC1 – EXI1SS P2.9 P3.1 P3.0 CC9IC ASC1 – EXI2SS P2.10 P3.11 P3.10 CC10IC ASC0 – EXI3SS P2.11 P3.13 P3.12 CC11IC SSC0 – EXI4SS P2.12 P4.7 P4.5 CC12IC CAN_A EXI5SS P2.13 P4.6 P4.4 CC13IC EXI6SS P2.14 P7.7 P7.5 CC14IC EXI7SS P2.15 P7.6 P7.4 CC15IC The actual interface CAN_B pin is – programmCAN_A, CAN_B able External Interrupts During Sleep Mode During Sleep Mode, all peripheral clock signals are deactivated. This also disables the standard edge detection logic for the fast external interrupts. However, transitions on these interrupt inputs must be recognized in order to initiate the wake-up. This is accomplished by a special edge detection logic for the fast external interrupts which requires no clock signal (therefore also works in Sleep mode) and is equipped with an analog noise filter. This filter suppresses spikes (generated by noise) up to 10 ns. Input pulses with a duration of 100 ns minimum are recognized and generate an interrupt request. This filter delays the recognition of an external wake-up signal by approximately 100 ns, but the spike suppression ensures safe and robust operation of the sleep/wake-up mechanism in an active environment. 100 ns 10 ns Input Signal Interrupt Request Rejected Recognized MCD04456 Figure 5-6 User’s Manual ICU_X73, V2.1 Input Noise Filter Operation 5-39 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions External Interrupt Pulse Timing External interrupt inputs are evaluated by a synchronous logic and by an asynchronous logic. The synchronous logic supports the recognition of short interrupt pulses at higher system frequencies, the asynchronous logic ensures recognition of interrupt pulses during sleep mode, when no system clock is available. An external interrupt signal is safely recognized in two cases: • • if it is active for more than 100 ns (async. logic with spike filter), or if it is active for more than 2 cycles of fSYS (sync. logic). The interrupt signal is recognized after whatever condition becomes true first. Note: After wake-up from Sleep mode, the time span until the PLL becomes locked is not critical for new external interrupt pulses to be correctly synchronized, because in this case the asynchronous logic will detect the external interrupt correctly, if it is active for at least 100 ns. Note: The NMI input features the same spike filter and the same timing requirements. 5.9 OCDS Requests The OCDS module issues high-priority break requests or standard service requests. The break requests are routed directly to the CPU (like the hardware trap requests) and are prioritized there. Therefore, break requests ignore the standard interrupt arbitration and receive highest priority. The standard OCDS service requests are routed to the CPU Action Control Unit together with the arbitrated interrupt/PEC requests. The service request with the higher priority is sent to the CPU to be serviced. If both the interrupt/PEC request and the OCDS request have the same priority level, the interrupt/PEC request wins. This approach ensures precise break control, while affecting the system behavior as little as possible. The CPU Action Control Unit also routes back request acknowledges and denials from the core to the corresponding requestor. User’s Manual ICU_X73, V2.1 5-40 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.10 Service Request Latency The numerous service requests of the XC167 (requests for interrupt or PEC service) are generated asynchronously with respect to the execution of the instruction flow. Therefore, these requests are arbitrated and are inserted into the current instruction stream. This decouples the service request handling from the currently executed instruction stream, but also leads to a certain latency. The request latency is the time from activating a request signal at the interrupt controller (ITC) until the corresponding instruction reaches the pipeline’s execution stage. Table 5-14 lists the consecutive steps required for this process. Table 5-14 Steps Contributing to Service Request Latency Description of Step Interrupt Response PEC Response Request arbitration in 3 stages, leads to acceptance by the CPU (see Section 5.2) 9 cycles 9 cycles Injection of an internal instruction into the pipeline’s instruction stream 4 cycles 4 cycles The first instruction fetched from the interrupt vector table reaches the pipeline’s execution stage 4 cycles / 01) --- Resulting minimum request latency 17/13 cycles 13 cycles 1) Can be saved by using the interrupt jump table cache (see Section 5.3). User’s Manual ICU_X73, V2.1 5-41 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Sources for Additional Delays Because the service requests are inserted into the current instruction stream, the properties of this instruction stream can influence the request latency. Table 5-15 Additional Delays Caused by System Logic Reason for Delay Interrupt Response PEC Response Interrupt controller busy, because it is just executing an arbitration cycle max. 9 cycles max. 9 cycles 2 × TACCmax Pipeline is stalled, because instructions preceding the injected instruction in the pipeline need to write/read data to/from a peripheral or memory 2 × TACCmax Pipeline cancelled, because instructions preceding the injected instruction in the pipeline update core SFRs 4 cycles 4 cycles Memory access for stack writes (if not to DPRAM or DSRAM) 2/3 × TACC1) --- Memory access for vector table read (except for intr. jump table cache) 2 × TACC --- 1) Depending on segmentation off/on. The actual response to an interrupt request may be delayed further depending on programming techniques used by the application. The following factors can contribute: • • Actual interrupt service routine is only reached via a JUMP from the interrupt vector table. Time-critical instructions can be placed directly into the interrupt vector table, followed by a branch to the remaining part of the interrupt service routine. The space between two adjacent vectors can be selected via bitfield VECSC in register CPUCON1. Context switching is executed before the intended action takes place (see Section 5.6) Time-critical instructions can be programmed “non-destructive” and can be executed before switching context for the remaining part of the interrupt service routine. User’s Manual ICU_X73, V2.1 5-42 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions 5.11 Trap Functions Traps interrupt current execution in a manner similar to standard interrupts. However, trap functions offer the possibility to bypass the interrupt system’s prioritization process for cases in which immediate system reaction is required. Trap functions are not maskable and always have priority over interrupt requests on any priority level. The XC167 provides two different kinds of trapping mechanisms: Hardware Traps are triggered by events that occur during program execution (such as illegal access or undefined opcode); Software Traps are initiated via an instruction within the current execution flow. Software Traps The TRAP instruction causes a software call to an interrupt service routine. The vector number specified in the operand field of the trap instruction determines which vector location in the vector table will be branched to. Executing a TRAP instruction causes an effect similar to the occurrence of an interrupt at the same vector. PSW, CSP (in segmentation mode), and IP are pushed on the internal system stack and a jump is taken to the specified vector location. When a trap is executed, the CSP for the trap service routine is loaded from register VECSEG. No Interrupt Request flags are affected by the TRAP instruction. The interrupt service routine called by a TRAP instruction must be terminated with a RETI (return from interrupt) instruction to ensure correct operation. Note: The CPU priority level and the selected register bank in register PSW are not modified by the TRAP instruction, so the service routine is executed on the same priority level from which it was invoked. Therefore, the service routine entered by the TRAP instruction uses the original register bank and can be interrupted by other traps or higher priority interrupts, other than when triggered by a hardware event. Hardware Traps Hardware traps are issued by faults or specific system states which occur during runtime of a program (not identified at assembly time). A hardware trap may also be triggered intentionally, for example: to emulate additional instructions by generating an Illegal Opcode trap. The XC167 distinguishes eight different hardware trap functions. When a hardware trap condition has been detected, the CPU branches to the trap vector location for the respective trap condition. The instruction which caused the trap is completed before the trap handling routine is entered. Hardware traps are non-maskable and always have priority over every other CPU activity. If several hardware trap conditions are detected within the same instruction cycle, the highest priority trap is serviced (see Table 5-3). User’s Manual ICU_X73, V2.1 5-43 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions PSW, CSP (in segmentation mode), and IP are pushed on the internal system stack and the CPU level in register PSW is set to the highest possible priority level (level 15), disabling all interrupts. The global register bank is selected. Execution branches to the respective trap vector in the vector table. A trap service routine must be terminated with the RETI instruction. The eight hardware trap functions of the XC167 are divided into two classes: Class A traps are: • • • • External Non-Maskable Interrupt (NMI) Stack Overflow Stack Underflow trap Software Break These traps share the same trap priority, but have individual vector addresses. Class B traps are: • • • • Undefined Opcode Program Memory Access Error Protection Fault Illegal Word Operand Access The Class B traps share the same trap priority and the same vector address. The bit-addressable Trap Flag Register (TFR) allows a trap service routine to identify the kind of trap which caused the exception. Each trap function is indicated by a separate request flag. When a hardware trap occurs, the corresponding request flag in register TFR is set to ‘1’. The reset functions (hardware, software, watchdog) may be regarded as a type of trap. Reset functions have the highest system priority (trap priority III). Class A traps have the second highest priority (trap priority II), on the 3rd rank are Class B traps, so a Class A trap can interrupt a Class B trap. If more than one Class A trap occur at a time, they are prioritized internally, with the NMI trap at the highest and the software break trap at the lowest priority. In the case where e.g. an Undefined Opcode trap (class B) occurs simultaneously with an NMI trap (class A), both the NMI and the UNDOPC flag is set, the IP of the instruction with the undefined opcode is pushed onto the system stack, but the NMI trap is executed. After return from the NMI service routine, the IP is popped from the stack and immediately pushed again because of the pending UNDOPC trap. Note: The trap service routine must clear the respective trap flag; otherwise, a new trap will be requested after exiting the service routine. Setting a trap request flag by software causes the same effects as if it had been set by hardware. User’s Manual ICU_X73, V2.1 5-44 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions TFR Trap Flag Register 15 14 13 12 SFR (FFACH/D6H) Reset Value: 0000H 11 10 9 8 7 6 5 4 3 2 1 0 NMI SOF STK STK TBR OF UF K - - - - UND OPC - - PAC PRT ILL ER FLT OPA - - rwh rwh - - - - rwh - - rwh - - rwh rwh rwh rwh Field Bits Type Description NMI 15 rwh Non Maskable Interrupt Flag 0 No non-maskable interrupt detected 1 A negative transition (falling edge) has been detected at pin NMI STKOF 14 rwh Stack Overflow Flag 0 No stack overflow event detected 1 The current stack pointer value falls below the contents of register STKOV STKUF 13 rwh Stack Underflow Flag 0 No stack underflow event detected 1 The current stack pointer value exceeds the contents of register STKUN SOFTBRK 12 rwh Software Break 0 No software break event detected 1 Software break event detected UNDOPC 7 rwh Undefined Opcode 0 No undefined opcode event detected 1 The currently decoded instruction has no valid XC167 opcode PACER 4 rwh Program Memory Access Error 0 No access error event detected 1 Illegal or erroneous access detected PRTFLT 3 rwh Protection Fault 0 No protection fault event detected 1 A protected instruction with an illegal format has been detected ILLOPA 2 rwh Illegal Word Operand Access 0 No illegal word operand access event detected 1 A word operand access (read or write) to an odd address has been attempted User’s Manual ICU_X73, V2.1 5-45 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Class A Traps Class A traps are generated by the high priority system NMI or by special CPU events such as the software break, a stack overflow, or an underflow event. Class A traps are not used to indicate hardware failures. After a Class A event, a dedicated service routine is called to react on the events. Each Class A trap has its own vector location in the vector table. Class A traps cannot interrupt atomic/extend sequences and I/O accesses in progress, because after finishing the service routine, the instruction flow must be further correctly executed. For example, an interrupted extend sequence cannot be restarted. All Class A traps are generated in the pipeline during the execution of instructions, except for NMI, which is an asynchronous external event. Class A trap events can be generated only during the memory stage of execution, so traps cannot be generated by two different instructions in the pipeline in the same CPU cycle. The execution of instructions which caused a Class A trap event is always completed. In the case of an atomic/extend sequence or I/O read access in progress, the complete sequence is executed. Upon completion of the instruction or sequence, the pipeline is canceled and the IP of the instruction following the last one executed is pushed on the stack. Therefore, in the case of a Class A trap, the stack always contains the IP of the first not-executed instruction in the instruction flow. Note: The Branch Folding Unit allows the execution of a branch instruction in parallel with the preceding instruction. The pre-processed branch instruction is combined with the preceding instruction. The branch is executed together with the instruction which caused the Class A trap. The IP of the first following not-executed instruction in the instruction flow is then pushed on the stack. If more than one Class A trap occur at the same time, they are prioritized internally. The NMI trap has the highest priority and the software break has the lowest. Note: In the case of two different Class A traps occurring simultaneously, both trap flags are set. The IP of the instruction following the last one executed is pushed on the stack. The trap with the higher priority is executed. After return from the service routine, the IP is popped from the stack and immediately pushed again because of the other pending Class A trap (unless the trap related to the second trap flag in TFR has been cleared by the first trap service routine). User’s Manual ICU_X73, V2.1 5-46 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Class B Traps Class B traps are generated by unrecoverable hardware failures. In the case of a hardware failure, the CPU must immediately start a failure service routine. Class B traps can interrupt an atomic/extend sequence and an I/O read access. After finishing the Class B service routine, a restoration of the interrupted instruction flow is not possible. All Class B traps have the same priority (trap priority I). When several Class B traps become active at the same time, the corresponding flags in the TFR register are set and the trap service routine is entered. Because all Class B traps have the same vector, the priority of service of simultaneously occurring Class B traps is determined by software in the trap service routine. The Access Error is an asynchronous external (to the CPU) event while all other Class B traps are generated in the pipeline during the execution of instructions. Class B trap events can be generated only during the memory stage of execution, so traps cannot be generated by two different instructions in the pipeline in the same CPU cycle. Instructions which caused a Class B trap event are always executed, then the pipeline is canceled and the IP of the instruction following the one which caused the trap is pushed on the stack. Therefore, the stack always contains the IP of the first following not-executed instruction in the instruction flow. Note: The Branch Folding Unit allows the execution of a branch instruction in parallel with the preceding instruction. The pre-processed branch instruction is combined with the preceding instruction. The branch is executed together with the instruction causing the Class B trap. The IP of the first following not-executed instruction in the instruction flow is pushed on the stack. A Class A trap occurring during the execution of a Class B trap service routine will be serviced immediately. During the execution of a Class A trap service routine, however, any Class B trap occurring will not be serviced until the Class A trap service routine is exited with a RETI instruction. In this case, the occurrence of the Class B trap condition is stored in the TFR register, but the IP value of the instruction which caused this trap is lost. Note: If a Class A trap occurs simultaneously with a Class B trap, both trap flags are set. The IP of the instruction following the one which caused the trap is pushed into the stack, and the Class A trap is executed. If this occurs during execution of an atomic/extend sequence or I/O read access in progress, then the presence of the Class B trap breaks the protection of atomic/extend operations and the Class A trap will be executed immediately without waiting for the sequence completion. After return from the service routine, the IP is popped from the system stack and immediately pushed again because of the other pending Class B trap. In this situation, the restoration of the interrupted instruction flow is not possible. User’s Manual ICU_X73, V2.1 5-47 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions External NMI Trap Whenever a high to low transition on the dedicated external NMI pin (Non-Maskable Interrupt) is detected, the NMI flag in register TFR is set and the CPU will enter the NMI trap routine. Stack Overflow Trap Whenever the stack pointer is implicitly decremented and the stack pointer is equal to the value in the stack overflow register STKOV, the STKOF flag in register TFR is set and the CPU will enter the stack overflow trap routine. For recovery from stack overflow, it must be ensured that there is enough excess space on the stack to save the current system state twice (PSW, IP, in segmented mode also CSP). Otherwise, a system reset should be generated. Stack Underflow Trap Whenever the stack pointer is implicitly incremented and the stack pointer is equal to the value in the stack underflow register STKUN, the STKUF flag is set in register TFR and the CPU will enter the stack underflow trap routine. Software Break Trap When the instruction currently being executed by the CPU is a SBRK instruction, the SOFTBRK flag is set in register TFR and the CPU enters the software break debug routine. The flag generation of the software break instruction can be disabled by the Onchip Emulation Module. In this case, the instruction only breaks the instruction flow and signals this event to the debugger, the flag is not set and the trap will not be executed. Undefined Opcode Trap When the instruction currently decoded by the CPU does not contain a valid XC167 opcode, the UNDOPC flag is set in register TFR and the CPU enters the undefined opcode trap routine. The instruction that causes the undefined opcode trap is executed as a NOP. This can be used to emulate unimplemented instructions. The trap service routine can examine the faulting instruction to decode operands for unimplemented opcodes based on the stacked IP. In order to resume processing, the stacked IP value must be incremented by the size of the undefined instruction, which is determined by the user, before a RETI instruction is executed. User’s Manual ICU_X73, V2.1 5-48 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Interrupt and Trap Functions Program Memory Access Error When a program memory access error is detected, the PACER flag is set in register TFR and the CPU enters the PMI access error trap routine. The access error is reported in the following cases: • • • • • access to Flash memory while it is disabled access to Flash memory from outside while read-protection is active double bit error detected when reading Flash memory access to reserved locations (see memory map in Table 3-1) access to Monitor RAM, if not in emulation mode In case of an access error, additionally the soft-trap code 1E9BH is issued. Protection Fault Trap Whenever one of the special protected instructions is executed where the opcode of that instruction is not repeated twice in the second word of the instruction and the byte following the opcode is not the complement of the opcode, the PRTFLT flag in register TFR is set and the CPU enters the protection fault trap routine. The protected instructions include DISWDT, EINIT, IDLE, PWRDN, SRST, ENWDT and SRVWDT. The instruction that causes the protection fault trap is executed like a NOP. Illegal Word Operand Access Trap Whenever a word operand read or write access is attempted to an odd byte address, the ILLOPA flag in register TFR is set and the CPU enters the illegal word operand access trap routine. User’s Manual ICU_X73, V2.1 5-49 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6 General System Control Functions The XC167 System Control Unit (SCU) summarizes a number of central control tasks and product specific features. These features include functional modules such as the Watchdog Timer (WDT) or the Clock Generation Unit (CGU), as well as basic functions such as the register protection mechanism or the reset generation. The following general functions are provided: • • • • • The System Reset is generated by the Reset Control Block and handles the reset and startup behavior (internal initialization) of the chip. It controls the reset triggers as well as the reset timing. This block controls also the basic configuration of the XC167 via external hardware. The Clock Generation Unit (CGU) provides the on-chip oscillator and the Phase Locked Loop (PLL). This block generates all clock signals for the XC167 and distributes them to the respective modules. Also the status of the clock generation system is indicated. The Central System Control Functions comprise all central control tasks like security level selection and system behavior in Sleep mode and Powerdown mode. Depending on the application state, different security levels (like protected and unprotected mode) are supported by the security level control state machine. The Watchdog Timer (WDT) represents one of the fail-safe mechanisms which have been implemented to prevent the controller from malfunctioning. It can detect long term malfunctions and is always enabled after chip initialization. The WDT can operate in Compatible mode or in Enhanced WDT mode. The Identification Control Block supports a set of six identification registers for identification of the most important silicon parameters (chip manufacturer, chip type and its properties). This information can be used for automatic test selection. User’s Manual SCU_X7, V2.1 6-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1 System Reset The internal system reset function provides initialization of the XC167 into a defined default state. The default state is invoked either by asserting a hardware reset signal on pin RSTIN (Hardware Reset Input), by executing the SRST instruction (Software Reset), or by an overflow of the watchdog timer. Whenever one of these conditions occurs, the microcontroller is reset into a predefined default state through an internal reset procedure. When a software reset is initiated, pending internal hold states are cancelled and the current internal access cycle (if any) is completed. An external bus cycle is completed, except for a READY-controlled bus cycle without a valid READY signal. Afterwards, the bus pin drivers and the IO pin drivers are switched off (tristate). Hardware reset and watchdog reset immediately abort all actions. The internal reset procedure is executed in several consecutive phases. The order of these phases depends on the reset source. In general, reset is triggered asynchronously (external) or synchronously (internal), it is always terminated synchronously. Table 6-1 Sequence of Reset Phases Phase Hardware Reset1) Watchdog Reset Software Reset 1 External Reset Phase -------skipped------Covers the time until the external trigger is removed (RSTIN = 1), the device is reset asynchronously 2 Internal Reset Phase The appropriate parts of the chip (peripheral system and/or CPU) are in reset state (except for the reset control block, of course). The internal reset phase covers the time specified by the reset event timer. 3 Initialization Phase The appropriate parts of the chip (peripheral system and/or CPU) are set up according to the default configuration: • External startup: the default configuration depends on the PORT0 settings • Internal startup: a fixed default configuration is used • Bootstrap loader: program code is loaded from the external system 4 Operation (Reset phases are terminated) The user software is executed from now on. Prereset Phase (Shut down) Covers the time until the running and pending actions of on-chip modules are completed 1) A hardware reset must always be asserted while the supply voltages are outside their defined operating ranges, for example, during Power-On. User’s Manual SCU_X7, V2.1 6-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.1 Reset Sources and Phases The XC167 executes a reset in several phases whose sequence depends on the reset trigger (see Table 6-1). External Reset Phase A hardware reset is asynchronously triggered when the reset input signal RSTIN is recognized low. A spike suppression input filter in the RSTIN line suppresses all signals shorter than 10 ns. To ensure the recognition of the RSTIN signal, it must be held low for at least 100 ns so it will safely pass the reset input filter. This is also required after the supply voltages have become stable. Note: The minimum duration of the external reset must ensure that the hardware configuration signals have reached their intended logic levels. RSTOUT External Hardware VDD b) Generated Warm Reset External Reset Sources RSTIN & Spike Filter (10 - 100 ns) Reset Control Block + System Control Unit a) Automatic Power-on Reset MCS05329 Figure 6-1 External Reset Circuitry A hardware reset on input RSTIN may be triggered in several ways (see Figure 6-1). • An external pull-up device connected to an external capacitor is sufficient for an automatic power-on reset. User’s Manual SCU_X7, V2.1 6-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions • • An external pull-up device connected to an external switch provides a manual reset. RSTIN may also be connected to the output of other logic for generating a warm reset. Note: During the external reset phase the complete chip is in reset state. The external reset phase is left synchronously, when the RSTIN level goes inactive (high). Pre-Reset Phase The pre-reset phase is triggered by a software reset. During the pre-reset phase, the CPU first runs its pipeline (including all write back buffers) empty, and then indicates the software reset request to the system control unit. The pipeline stays empty after this request trigger is activated. As soon as the software reset request occurs, the SCU requests a shutdown from the active modules equipped with shutdown handshake (see Section 6.3.3). The pre-reset phase is complete as soon as all modules acknowledge the shutdown state. Upon a shutdown request the EBC will finish the currently running bus cycle. If in a READY-controlled bus cycle the READY signal is not sampled active after the programmed number of waitstates, this bus cycle is aborted in order to prevent a deadlock situation. Internal Reset Phase At the beginning of the internal reset phase the internal reset condition becomes active, that means, the internal reset signal is actually applied to the modules. If the reset was triggered by hardware, it may be active already. Note: The reset control block (including the watchdog timer) is not reset, of course. The duration of the internal reset phase is determined by the reset-length-control bitfield RSTLEN in register RSTCON. The WDT low byte is used for counting the reset duration. When entering the internal reset phase, the timer is cleared and then counts up with frequency fWDT. The default count frequency after a hardware reset is fWDT = fSYS/2 = fMC/2. Internal reset triggers do not change the current clock setting and the value of bitfield WDTIN, so the previously selected fWDT is used. The actual duration of the internal reset sequence can therefore be calculated using the following formula: ( RSTLEN ) t RST = 2 -----------------------f WDT User’s Manual SCU_X7, V2.1 (6.1) 6-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Reset Termination (Initialization Phase) When the end of the internal reset phase has been reached, the following actions take place, before control is passed to the software: • • • • Set the reset indication flags in register SYSSTAT accordingly Select initial configuration and reset start address Deactivate the internal reset signals Execute bootstrap loader if selected Note: The WDT continues counting up from zero. User’s Manual SCU_X7, V2.1 6-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.2 Status After Reset Most units of the XC167 enter a well-defined default status after a reset is completed. This ensures repeatable start conditions and avoids spurious activities after reset. Reset Values for the XC167 Registers During the reset sequence, the registers of the XC167 are preset with a default value. Most SFRs, including system registers and peripheral control and data registers, are cleared to zero, so all peripherals and the interrupt system are off or idle after reset. A few exceptions to this rule provide a first pre-initialization, which is either fixed or controlled by input pins. A number of registers are reset only upon a hardware reset, after a software or WDT reset they retain their previous values (see Table 6-2). Table 6-2 Non-Zero Registers after Reset Register Name Initial Value Comments DPP1 0001H Points to data page 1 DPP2 0002H Points to data page 2 DPP3 0003H Points to data page 3 CP FC00H – STKUN FC00H – STKOV FA00H – SP FC00H – RSTCFG XXXXH Reset levels of PORT0, 0DFFH in single-chip mode RSTCON 00XXH Depends on configuration after a hardware reset PLLCON XXXXH Depends on selected clock configuration VECSEG 00XXH Depends on startup mode SYSSTAT XXXXH Depends on current status SYSCON3 9FD0H RTC, TwinCAN, Flash, GPT, SSC0, ASC0, ADC enabled EBCMOD0 XXXXH Depends on selected bus type EBCMOD1 00XXH Depends on selected bus type TCONCS0 7AXXH 7A68H for MUX bus, 7A40H for DEMUX bus FCONCS0 00X1H Depends on selected bus type ONES FFFFH Fixed value User’s Manual SCU_X7, V2.1 6-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Operation after Reset After the internal reset condition is removed, the XC167 fetches the first instruction from the selected program memory location (depending on the configuration). As a rule, this first location holds a branch instruction to the actual initialization routine that may be located anywhere in the address space. Note: If the Bootstrap Loader Mode was activated during a hardware reset, the XC167 does not fetch instructions from the program memory. The standard bootstrap loader expects data via serial interface ASC0. Watchdog Timer Operation after Reset The watchdog timer continues running after the internal reset is complete. It will be clocked with the currently selected clock signal fWDT. After a watchdog/software reset fWDT is not changed, after a hardware reset the frequency is fWDT = fSYS/2 = fMC/2. The default reload value is 00H. Thus, a watchdog timer overflow will occur 216 clock cycles (217 fMC cycles after a hardware reset) after completion of the internal reset (depending on the selected reset length), unless it is disabled, serviced, or reprogrammed in the meantime. If the system reset was caused by a watchdog timer overflow, the WDTR (Watchdog Timer Reset Indication) flag in register SYSSTAT will be set to 1. This indicates the cause of the internal reset to the software initialization routine. WDTR is reset to 0 after each other reset. After the internal reset is complete, the operation of the watchdog timer can be disabled by the DISWDT (Disable Watchdog Timer) instruction prior to the EINIT instruction if using compatibility mode, or anytime in enhanced mode. The On-Chip RAM Areas after Reset The contents of the major parts of the on-chip RAMs are preserved during a software reset and a WDT reset. There are two exceptions to this rule: • • A part of the DPRAM (the area 00’FBA0H … 00’FC1FH) may be altered during the initialization phase (see Table 6-1) and, therefore, should not store data to be preserved beyond a WDT/SW reset. During bootstrap loader operation the serially received data is stored in the PSRAM starting at location E0’0000H. Because a hardware reset can occur asynchronously to an internal operation, it may interrupt a current write operation and so inadvertently corrupt the contents of on-chip RAM. RAM contents are preserved if the hardware reset occurs during Power-Down mode, during Sleep mode, or during Idle mode with no PEC transfers enabled. Note: After a power-up hardware reset the RAM contents are undefined, of course. User’s Manual SCU_X7, V2.1 6-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions External Bus Interface after Reset If an external start is selected after reset the EBC is initialized accordingly and some of the port pins of the XC167 are controlled accordingly. Pin ALE is held low through an internal pull-down, and pins RD and WR are held high through internal pull-ups. Also, all pins which can be configured for CS output will be pulled high during each reset. The registers EBCMOD0, EBCMOD1, TCONCS0, and FCONCS0 are initialized according to the configuration selected via PORT0. When an external start is selected (pin EA = 0): • • • • • Bit ENCS in register FCONCS0 is set to 1 Bus Type field (BTYP) in register FCONCS0 is initialized according to P0L.7 and P0L.6 A default bus timing (depending on the bus mode) is selected via register TCONCS0 The required pins of PORT0 and PORT1 are assigned via register EBCMOD1 Bitfields WRCFG, CSPEN, and SAPEN in register EBCMOD0 are set as selected via PORT0 (WRC, CSSEL, SASEL) When an internal start is selected (pin EA = 1): • • Register EBCMOD0 is set to 7400H (EBC-pins disabled) Registers EBCMOD1, FCONCS0, and TCONCS0 are cleared Note: This initial configuration of the EBC may be changed by user software at any time. When the internal reset is complete, the configuration of PORT0, PORT1, Port 4, Port 6, and of the BHE signal (High Byte Enable, alternate function of P3.12) depends on the bus type selected during reset. If any of the external bus modes was selected during reset, PORT0 will operate in the selected bus mode. Port 4 will output the selected number of segment address lines (all zero after reset). Port 6 will drive the selected number of CS lines (CS0 will be 0, while the other active CS lines will be 1). If no memory accesses above 64 Kbytes are to be performed, segmentation may be disabled. Ports after Reset During the internal reset sequence, all port pins of the XC167 are configured as inputs by clearing the associated direction registers, and their pin drivers are switched to the high impedance state. This ensures that the XC167 and external devices will not try to drive the same pin to different levels. Pins assigned to the EBC become active after an external-start reset. Note: Pull-ups for configuration and possible CS signals are active during each reset. When the on-chip bootstrap loader was activated during reset, pin TxD0 (alternate port function) will be switched to output mode after the reception of the zero byte. All other pins remain in the high-impedance state until they are changed by software or peripheral operation. User’s Manual SCU_X7, V2.1 6-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Reset Output Pin The RSTOUT pin is dedicated to the generation of a reset signal for external system components such as peripherals or Flash memories. The behaviour of RSTOUT can be selected via software and can so be adapted to the respective external system. RSTOUT is activated asynchronously with an external hardware reset. It may also be activated (selectable) synchronously with an internal software or watchdog reset. RSTOUT is deactivated at a selectable time: • • • optionally at the end of reset (supports a reset signal to an external program Flash) upon the execution of the EINIT instruction (latest) at an earlier time via user software This allows the complete configuration of the controller including its on-chip peripheral units before releasing the reset signal for the external peripherals of the system. Note: RSTOUT will float during adapt mode (see register RSTCFG in Section 6.1.4). RSTOUT is controlled via several bits in registers RSTCON (see Table 6-3). The resulting output signal timing is shown in Figure 6-2. Table 6-3 Usage of RSTOUT Control Bits Control Bit Operation RODIS Deactivates RSTOUT when set by software. ROCON Lets the user software select if RSTOUT is activated upon any reset (0) or only upon an external reset (1). If ROCON = 0 bit RODIS is cleared and pin RSTOUT is activated after a software or WDT reset. The lower part of Figure 6-2 shows both cases (see “Automatic Activation”). ROCOFF Lets the user software select if RSTOUT is deactivated (1) automatically at the end of reset (after the initialization phase). See “Automatic Deactivation” in Figure 6-2. Bit RODIS is set in this case. If no automatic deactivation is selected (ROCOFF = 0) user software may set bit RODIS at any time. Automatic Deactivation can also be selected by hardware configuration. This supports an external start out of an external Flash memory. RORMV Lets the user remove the RSTOUT function from pin P20.12 and free it for general purpose IO. P20.12 IO can also be selected by hardware configuration. Note: At the very latest, RSTOUT is deactivated by execution of the EINIT instruction, independent of software selections (register RSTCON). User’s Manual SCU_X7, V2.1 6-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Hardware Reset RSTIN Appl. SW Ext. Reset Phase Int. Reset Phase Startup Sequence User Application Software RSTOUT User Control Software/WDT Reset Software Appl. SW Reset WDT Reset Pre-Reset Phase Int. Reset Phase Startup Sequence User Application Software User Application Software RSTOUT User Control Automatic Activation Automatic Deactivation EINIT MCT05330 Figure 6-2 User’s Manual SCU_X7, V2.1 Timings of Reset Signals and Pin RSTOUT 6-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.3 Application-Specific Initialization Routine After a reset, the modules of the XC167 must be initialized to enable their operation on a given application. This initialization depends on the task to be performed by the XC167 in that application and on some system properties such as operating frequency, external circuitry connected, etc. Typically, the following initializations should be done before the XC167 is prepared to run the actual application software: Bus Interface The external bus interface can be reconfigured after an external reset because the EBC registers are initialized to default values and may not represent the optimum bus configuration. The programmable address windows can be enabled in order to adapt the bus cycle characteristics to various memory areas or peripherals. Also, after a singlechip mode reset, the external bus interface can be enabled and configured. Programmable program memory (on-chip or external) can be programmed, for instance, with data received over a serial link. Note: Bootstrap loader mode can be used for initial Flash or OTP programming. System Stack The default setup for the system stack (size, stack pointer, upper and lower limit registers) can be adjusted to application-specific values. After reset, registers SP and STKUN contain the same reset value 00’FC00H, while register STKOV contains 00’FA00H. With the default reset initialization, 256 words of system stack are available in the DPRAM, where the system stack selected by the SP grows downwards from 00’FBFEH. The system stack may be moved to the DSRAM and its size can be adjusted to the application’s requirements. Note: The interrupt system, which is disabled upon completion of the internal reset, should remain disabled until the SP is initialized. Traps (including NMI) may occur, although the interrupt system is still disabled. Register Bank The location of a global register bank is defined by the context pointer (CP) and can be adjusted to an application-specific bank before the general purpose registers (GPRs) are used. After reset, register CP contains the value FC00H, i.e. the register bank selected by the CP grows upward from 00’FC00H. User’s Manual SCU_X7, V2.1 6-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions On-Chip RAM Depending on the application, the user may wish to initialize portions of the internal writable memory (DPRAM/DSRAM/PSRAM) before normal program operation. After the register bank has been selected by programming the CP register, the desired portions of the internal memory can easily be initialized via indirect addressing. Interrupt System After reset, the individual interrupt nodes and the global interrupt system are disabled. In order to enable interrupt requests, the nodes must be assigned to their respective interrupt priority levels and must be enabled. The vector table can be adjusted if the default properties do not fit. Register VECSEG defines the vector table’s location, bitfield VECSC in register CPUCON1 defines the vector spacing. The vector locations must receive pointers to the respective exception handlers. The interrupt system must globally be enabled by setting bit IEN in register PSW. To avoid such problems as the corruption of internal memory locations caused by stack operations using an uninitialized stack pointer, care must be taken not to enable the interrupt system before the initialization is complete. Ports Generally, all ports of the XC167 are switched to input upon reset activation. After reset, some pins may be automatically controlled, such as bus interface pins for an external start, TxD in Boot mode, etc. Pins to be used for general purpose IO must be initialized via software. The required mode (input/output, open drain/push pull, input threshold, etc.) depends on the intended function for a given pin. Peripherals Upon reset activation the XC167’s on-chip peripheral modules enter a defined default state (see respective peripheral description) in which they are disabled from operation. In order to use a certain peripheral it must be initialized according to its intended operation in the application. This includes enabling the peripheral, selecting the operating mode (such as counter/timer), operating parameters (such as baudrate), enabling interface pins (if required), assigning interrupt nodes to the respective priority levels, etc. After these standard initialization actions, application-specific actions may be required, such as asserting certain levels to output pins, sending codes via interfaces, latching input levels, etc. User’s Manual SCU_X7, V2.1 6-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Watchdog Timer After reset, the watchdog timer is active and counting its default period. If the watchdog timer is to remain active the desired period should be programmed by selecting the appropriate prescaler value and reload value. Otherwise, the watchdog timer must be disabled before EINIT. Termination of Initialization The software initialization routine should be terminated with the EINIT instruction. This instruction has been implemented as a protected instruction. Execution of the EINIT instruction has the following effects: • • • Disables the action of the DISWDT instruction (unless enhanced mode is selected), Switches the register security level to “write-protected mode” (see Section 6.3.5), Causes the RSTOUT pin to go high if it is not high already (this signal can be used to indicate the end of the initialization routine and the proper operation of the microcontroller to external hardware). User’s Manual SCU_X7, V2.1 6-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.4 System Startup Configuration Although most of the programmable features of the XC167 are selected by software either during the initialization phase or repeatedly during program execution, some features must be selected earlier because they are used for the first access of the program execution (for example, internal or external start selected via EA). These configurations are accomplished by latching the logic levels at a number of pins at the end of the internal reset sequence. During reset, internal pull-up/pull-down devices are active on those lines. They ensure inactive/default levels at pins which are not driven externally. External pull-down/pull-up devices may override the default levels in order to select a specific configuration. Many configurations can, therefore, be coded with a minimum of external circuitry. Note: The load on those pins to be latched for configuration must be small enough for the internal pull-up/pull-down device to sustain the default level, or external pull-up/pull-down devices must ensure this level. Those pins whose default level will be overridden must be pulled low/high externally. Ensure that the valid target levels are reached by the end of the reset sequence. There is a specific application note to illustrate this. A hardware reset can be terminated as soon as the target levels are reached. The XC167 automatically waits until oscillator, PLL, and Flash are up and operable. The levels on pins EA, RD, ALE, and WR are latched whenever the internal reset phase is left after a hardware reset (see Table 6-1). These four pins must be controlled externally in any case. In the case of an external start (EA = 0) also PORT0 is latched to provide additional configuration inputs. WR EA RD ALE PORT0 End of internal reset phase after a hardware reset RSTCFG MCA05331 Figure 6-3 User’s Manual SCU_X7, V2.1 Latching Configuration 6-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Table 6-4 Basic Startup Configuration via External Circuitry EA = 0 (Latch PORT0) EA = 1 (Use Default) RD = 0 RD = 1 RD = 0 RD = 1 ALE = 0 Standard start external, PLL/OWD off1) Standard start external, PLL/OWD ON Standard Boot Standard start internal ALE = 1 Reserved Reserved Alternate Boot Alternate start internal WR = 0 RORMV = 0 (RSTOUT is always active) WR = 1 RORMV = 12) RORMV = 0 1) Only effective in bypass mode. This option switches off the oscillator watchdog. 2) P20.12 enabled instead of signal RSTOUT, can be used in a single-chip system without external bus system. Pin WR is evaluated independently of pins EA, RD and ALE. Note: The pull-ups/pull-downs on the configuration pins are activated at the beginning of a hardware reset. They are deactivated after latching the configuration information. The further startup behavior of the XC167 can be configured in several ways: • • Read the configuration from PORT0 (EA = 0) Use a fixed default configuration (EA = 1) The respective configuration is stored into register RSTCFG and the XC167 is accordingly initialized. User software may read this register in order to determine the actual configuration. The user can change this startup configuration at any time via the dedicated configuration registers, for example with the EBCMOD0/1 registers or with the PLLCON register. Note: In Single-Chip Mode (internal start with EA = 1) the default configuration selects clock generation in bypass mode with factor 2:1 ensuring proper operation for the defined input frequency range of up to 50 MHz. User’s Manual SCU_X7, V2.1 6-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions RSTCFG Reset Configuration Register 15 14 13 12 11 ESFR (F108H/84H) 10 9 CLKCFG SALSEL CSSEL rh rh rh 7 6 WRC BUSTYP rh 5 4 3 2 1 SMOD rh 0 ADP ROC rh rh Field Bits CLKCFG [15:13] rh Clock Generation Mode Configuration1) 000 fMC = fOSC/2, fOSC = 1 - 50 MHz2) 001 fMC = fOSC × 2.5, fOSC = 12 - 16 MHz 010 fMC = fOSC × 2.5, fOSC = 8 - 12 MHz 011 fMC = fOSC, fOSC = 1 - 40 MHz2) 100 fMC = fOSC × 5, fOSC = 4 - 6 MHz 101 fMC = fOSC × 2, fOSC = 12.5 - 18.7 MHz 110 fMC = fOSC × 4.5, fOSC = 5.6 - 8.3 MHz 111 fMC = fOSC × 3, fOSC = 8.3 - 12.5 MHz SALSEL [12:11] rh Segment Address Line Select 00 4-bit segment address (A19 … A16) 01 No segment address lines 10 8-bit segment address (A23 … A16) 11 2-bit segment address (A17 … A16) CSSEL [10:9] rh Chip Select Line Select 00 3 CS lines: CS2 … CS0 01 2 CS lines: CS1 … CS0 10 No CS lines 11 5 CS lines: CS4 … CS0 WRC 8 rh Write Configuration 0 Pins WR and BHE operate as WRL/WRH 1 Pins WR and BHE operate as WR/BHE BUSTYP [7:6] rh External Bus Type 00 8-bit data, DEMUX address 01 8-bit data, MUX address 10 16-bit data, DEMUX address 11 16-bit data, MUX address User’s Manual SCU_X7, V2.1 Type 8 Reset Value: XXXXH rh Description 6-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Field Bits Type Description SMOD [5:2] rh Special Modes 0111 Alternate start 1001 Alternate bootstrap loader mode 1011 Standard bootstrap loader mode 1111 Standard start (default) All other combinations are reserved for future use. ADP 1 rh Adapt Mode 0 Adapt mode selected (device floats all pins) 1 Standard operation ROC 0 rh RSTOUT Control 0 RSTOUT is deactivated automatically at the end of reset 1 RSTOUT is deactivated by user software 1) The external clock input range indicates the operating range of the CGU. If a crystal is connected the oscillator frequency range is limited to 4 … 16 MHz. 2) This selection enables bypass mode. Note: The reset value of register RSTCFG depends on the values latched from PORT0 in case of an external reset. After a single-chip reset (EA = 1) register RSTCFG is loaded with the default value 0DFFH. The pins which control operation of the internal control logic, the clock configuration, and the reserved pins are evaluated only during a hardware triggered reset sequence. The configuration via PORT0 is latched in register RSTCFG for subsequent evaluation by software. The following descriptions refer to the various selections available for reset configuration. The default modes refer to pins at high level without external pull-down devices connected. Note: The initial configuration in single-chip mode is described in Section 6.1.6. User’s Manual SCU_X7, V2.1 6-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.5 Hardware Configuration in External Start Mode For an external start mode reset (indicated by EA = 0) the configuration value (register RSTCFG) is copied from PORT0. In this case, external circuitry (pull-ups/pull-downs) on PORT0 generate application-specific configuration values. Note: To support hardware configuration with minimum external circuitry, the PORT0 pins have internal pull-up devices activated upon each reset. Clock Generation Control Pins P0H.7, P0H.6, and P0H.5 (CLKCFG) select the initial clock generation mode during reset. The oscillator clock either directly feeds the CPU and peripherals (direct drive), is divided by 2 or is fed to the on-chip PLL which then provides the master clock signal (selectable multiple of the oscillator frequency, i.e. the input frequency). This coarse selection adapts the clock generation unit to the selected oscillator frequency. The user initialization code may then exactly select the required configuration by updating register PLLCON. Table 6-5 XC167 Clock Generation Modes (P0H.7-5) (CLKCFG) Master Clock fMC = fOSC × F External Clock Input Range1) PLLCON (initial) Notes 111 fOSC × 3 fOSC × 4.5 fOSC × 2 fOSC × 5 fOSC × 1 fOSC × 2.5 fOSC × 2.5 fOSC/2 8.3 to 12.5 MHz 6B03H Default configuration 5.6 to 8.3 MHz 7103H – 12.5 to 18.7 MHz 6F13H – 4 to 6 MHz 7804H – 1 to 40 MHz 2780H Direct drive 8 to 12 MHz 7814H – 12 to 16 MHz 7854H – 1 to 50 MHz 2790H 2:1 prescaler 110 101 100 011 010 001 000 1) The external clock input range indicates the operating range of the CGU. If a crystal is connected the oscillator frequency range is limited to 4 … 16 MHz. Default: On-chip PLL is active with a factor of 1:3 (fOSC is multiplied by 3). Watch the different requirements for frequency of the oscillator input clock for the possible selections. User’s Manual SCU_X7, V2.1 6-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Segment Address Lines Pins P0H.4 and P0H.3 (SALSEL) define the number of active segment address lines during reset. This allows selection of which Port 4 pins drive address lines. Depending on the system architecture, the required address space is chosen and accessible right from the start; so, the initialization routine can directly access all locations without prior programming. The required Port 4 pins are automatically switched to address output mode. The user initialization code may then exactly select the required configuration by updating register EBCMOD0. Table 6-6 Configuration of Segment Address Lines P0H.4-3 (SALSEL) Segment Address Lines Directly Accessible Addr. Space 11 Two: A17 … A16 256 Kbytes (Default without pull-downs) 10 Eight: A23 … A16 12 Mbytes (Maximum) 01 None 64 Kbytes (Minimum) 00 Four: A19 … A16 1 Mbyte Even if not all segment address lines are enabled on Port 4, the XC167 internally uses its complete 24-bit addressing mechanism. This allows restriction of the width of the effective address bus, while still deriving CS signals from the complete addresses. Default: 2-bit segment address (A17 … A16) allowing access to 256 Kbytes. Chip Select Lines Pins P0H.2 and P0H.1 (CSSEL) define the number of active chip select signals during reset. This allows selection of which Port 6 pins drive external CS signals. The user initialization code may then exactly select the required configuration by updating register EBCMOD0. Table 6-7 Configuration of Chip Select Lines P0H.2-1 (CSSEL) Chip Select Lines Note 11 Five: CS4 … CS0 Default without pull-downs 10 None – 01 Two: CS1 … CS0 – 00 Three: CS2 … CS0 – Default: All 5 chip select lines active (CS4 … CS0). User’s Manual SCU_X7, V2.1 6-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Write Configuration Pin P0H.0 (WRC) selects the initial operation of the control pins WR and BHE during reset. When high, this pin selects the standard function, i.e. WR control and BHE. When low, it selects the alternate configuration, i.e. WRH and WRL. Thus, even the first access after a reset can go to a memory controlled via WRH and WRL. The user initialization code may then exactly select the required configuration by updating register EBCMOD0. Default: Standard function (WR control and BHE). External Bus Type Pins P0L.7 and P0L.6 (BUSTYP) select the external bus type during reset, if an external start is selected via pin EA. This allows configuration of the external bus interface of the XC167 even for the first code fetch after reset. P0L.7 controls the data bus width, while P0L.6 controls the address output (multiplexed or demultiplexed). The user initialization code may then exactly select the required configuration by updating register FCONCS0. Table 6-8 Configuration of External Bus Type P0L.7-6 (BTYP) Encoding External Data Bus Width External Address Bus Mode 00 8-bit Data Demultiplexed Addresses 01 8-bit Data Multiplexed Addresses 10 16-bit Data Demultiplexed Addresses 11 16-bit Data Multiplexed Addresses PORT0 and PORT1 are automatically switched to the selected bus mode. In multiplexed bus modes, PORT0 drives both the 16-bit intra-segment address and the output data, while PORT1 remains in high impedance state as long as no demultiplexed bus is selected via one of the FCONCS registers. In demultiplexed bus modes, PORT1 drives the 16-bit intra-segment address, while PORT0 or P0L (according to the selected data bus width) drives the output data. For a 16-bit data bus, BHE is automatically enabled, for an 8-bit data bus, BHE is disabled via bit BYTDIS in register EBCMOD0. Default: 16-bit data bus with multiplexed addresses. User’s Manual SCU_X7, V2.1 6-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Special Operation Modes Pins P0L.5 to P0L.2 (SMOD) select special operation modes of the XC167 during reset (see Table 6-9). Make sure to select only valid configurations to ensure proper operation of the XC167. Table 6-9 Definition of Special Modes for Reset Configuration P0L.5-2 (SMOD) Special Mode Notes 1111 Standard Start Begin executing at location 00’0000H 1011 Standard Bootstrap Loader Load an initial boot routine of 32 Bytes via interface ASC0. 1001 Alternate Boot Operation not yet defined. Do not use! 0111 Alternate Start Begin executing at location 41’0000H All other combinations Reserved for future modes Do not select this configuration! The On-Chip Bootstrap Loader allows the start code to be moved into the internal PSRAM of the XC167 via the serial interface ASC0. The XC167 will then execute the loaded start code out of the PSRAM. Default: The XC167 starts fetching code from location 00’0000H, the bootstrap loader is off. Adapt Mode Pin P0L.1 (ADP) selects the Adapt Mode when latched low at the end of reset. In this mode, the XC167 goes into a passive state similar to its state during reset. The pins of the XC167 float to tristate or are deactivated via internal pull-up/pull-down devices, as described for the reset state. Additionally, the RSTOUT pin floats to tristate rather than being driven low. The on-chip oscillator and the realtime clock are disabled. This mode allows a XC167 mounted to a board to be virtually switched off. This enables an emulator to control the board’s circuitry even though the original XC167 remains in place. The original XC167 may resume control of the board after a reset sequence with P0L.1 high. Please note that adapt mode overrides any other configuration via PORT0. Default: Adapt Mode is off. Note: When XTAL1 is fed by an external clock generator (while XTAL2 is left open), this clock signal may also be used to drive the emulator device. However, if a crystal is used, the emulator device’s oscillator can use this crystal only if at least XTAL2 of the original device is disconnected from the circuitry (the output XTAL2 will be driven high in Adapt Mode). Adapt mode can be activated only upon an external reset (EA = 0). Pin P0L.1 is not evaluated upon a single-chip reset (EA = 1). User’s Manual SCU_X7, V2.1 6-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions RSTOUT Control Pin P0L.0 (ROC) selects the initial deactivation mode of pin RSTOUT after reset. When high, this pin selects the standard function, i.e. RSTOUT is deactivated by user software (or at the very latest after the execution of EINIT). When low, it selects immediate deactivation, i.e. RSTOUT is deactivated automatically at the end of the internal reset state. Thus, even the first access after a reset can go to a memory controlled by the RSTOUT signal (e.g. an external Flash). The user initialization code may select the required configuration for each subsequent reset. Default: Standard function (RSTOUT deactivated by user software). In addition, the RSTOUT signal can be disabled, so the pin can be used for general purpose IO. This is selected by pulling pin WR low during a single-chip-mode reset with EA = 1. The user initialization code may select the required configuration by updating bit RORMV in register RSTCON. Oscillator Watchdog Control The on-chip oscillator watchdog (OWD) may be disabled via hardware by (externally) pulling the RD line low upon a reset (see Table 6-4). This option is valid for bypass operation. The user initialization code may select the required configuration by updating register PLLCON. Default: Oscillator watchdog is active (PLL is on). Note: If direct drive or prescaler operation is selected as basic clock generation mode (see above) the PLL is switched off whenever bit OWDDIS is set (via software or via hardware configuration). User’s Manual SCU_X7, V2.1 6-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.6 Default Configuration in Single-Chip Mode For a single-chip mode reset (indicated by EA = 1) the additional configuration via PORT0 is ignored and a fixed configuration value is used instead (RSTCFG = 0DFFH). In this case, PORT0 needs no external circuitry (pull-ups/pull-downs). This fixed default configuration selects a safe worst-case configuration. The initialization software can then modify these parameters and select the intended configuration for a given application. Table 6-10 lists the respective default configuration values which are selected and the bitfields which permit software modification. Table 6-10 Default Configuration for Single-Chip Mode Reset Configuration Parameter Default Values (RSTCFG = 0DFFH) External Config.1) Software Access2) CLKCFG: Initial clock generation mode 000B = 2:1 prescaler mode, PLLCON = 2790H, fMC = fOSC/2, see Table 6-5 P0.15-13 PLLCON SALSEL: Number of active seg. addr. lines 01B = No segment address lines P0.12-11 EBCMOD0.3-0 CSSEL: Number of active CS lines 10B = No chip select lines P0.10-9 EBCMOD0.7-4 WRC: Write signal encoding RSTCFG.0 = 1, EBCMOD0.WRCFG = 0, i.e. WR and BHE P0.8 EBCMOD0.11 BTYP: Default bustype BUSCON0.BTYP = 11B (FCONCS0) i.e. 16-bit MUX bus P0.7-6 FCONCS0.5-4 SMOD: Special modes 1111B = Standard start; (start/boot modes) Startup modes selected via pins RD and ALE (see Table 6-4) P0.5-2 – ADP: Adapt Mode P0.1 – ROC: RSTOUT control 1 = Deactivation via user software P0.0 RSTCON.5 OWD disable RD PLLCON.14-13 1 = Not possible PLLCON.PLLCTRL = 01B i.e. OWD is active 1) Refers to the configuration pins which are replaced by the default values. 2) Software can modify the default values via these bitfields. Note: Single-chip mode reset cannot be selected on ROMless devices. The attempt to read the first instruction after reset will fail in such a case. User’s Manual SCU_X7, V2.1 6-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.1.7 Reset Behavior Control The reset behavior is controlled by a set of control/status registers. The status information can be used by the initialization code to execute different actions depending on the reset source. The reset control register RSTCON is used by the application to program the basic reset behavior like length of internal reset phase and behavior of the reset output pin RSTOUT (see Figure 6-3). RSTCON Reset Control Register Reset Value: 0000H1) mem (F1E0H/--) 15 14 13 12 11 10 9 8 7 6 - - - - - - - - RO ROC ROC RO DIS ON OFF RMV - RSTLEN - - - - - - - - rwh - rw rw 5 rwh 4 rwh 3 2 1 0 1) The reset value is only valid for a hardware reset. Field Bits Type Description RODIS 7 rwh RSTOUT Disable Control 0 RSTOUT is controlled by other mechanisms 1 RSTOUT is deactivated Note: Bit RODIS is cleared if automatic enabling is selected (ROCON = 0), bit RODIS is set if automatic disabling is selected (ROCOFF = 1). ROCON 6 rw RSTOUT Control Switching ON 0 RSTOUT is activated upon any reset 1 RSTOUT is only activated upon a hardware reset ROCOFF 5 rwh RSTOUT Control Switching Off 0 RSTOUT is deactivated by user software 1 RSTOUT is automatically deactivated at the end of reset (RODIS is set) Note: Automatic deactivation can also be requested by hardware configuration (ROCOFF is set if RSTCFG.ROC = 0). RORMV User’s Manual SCU_X7, V2.1 4 rwh RSTOUT Remove Control 0 Pin delivers the RSTOUT signal (default) 1 Pin operates as P20.12 (gen. purpose IO) (selected if EA = 1 and WR = 0 during reset) 6-24 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Field Bits Type Description RSTLEN [2:0] rw Reset Length Control1) The duration of the next internal reset phase is tRST = tWDT × 2RSTLEN. 000 1 tWDT: default duration after hardware reset … … 111 128 tWDT: maximum duration 1) RSTLEN is always valid for the next reset sequence. An initial power up reset, however, is controlled by external hardware and is expected to last considerably longer than any configurable reset sequence. Note: RSTCON is protected by the register security mechanism (see Section 6.3.5). RSTCON can only be accessed via its long (mem) address. User’s Manual SCU_X7, V2.1 6-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2 Clock Generation All activities of the XC167’s controller hardware and its on-chip peripherals are controlled by clock signals which are generated by the Clock Generation Unit (CGU). This reference clock is generated in three stages: Oscillators The on-chip Pierce oscillators (main oscillator and auxiliary oscillator) can either run with an external crystal and appropriate oscillator circuitry or they can be driven by an external oscillator or another clock source. Clock Generation and Frequency Control The input clock signal of the main oscillator feeds the controller hardware: • • directly, divided by a programmable prescaler factor (1 … 60), either providing phase-coupled operation (factor = 1) or operating the device at low frequencies to reduce power consumption (factor >> 1) via an on-chip Phase Locked Loop (PLL) providing maximum performance on low input frequency Clock Distribution The clock signals are distributed via separate clock drivers which feed the CPU itself and groups of peripheral modules. Certain sections of the device can be supplied with a prescaled clock signal. Note: The RTC is fed with the auxiliary oscillator clock or with the prescaled main oscillator clock via a separate clock driver, so it is not affected by the clock control functions. User’s Manual SCU_X7, V2.1 6-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2.1 Oscillators The main oscillator of the XC167 is a power optimized Pierce oscillator providing an inverter and a feedback element. Pins XTAL1 and XTAL2 connect the inverter to the external crystal. The standard external oscillator circuitry (see Figure 6-4) comprises the crystal, two low end capacitors and series resistor (Rx2) to limit the current through the crystal. A test resistor (RQ) may be temporarily inserted to measure the oscillation allowance (negative resistance) of the oscillator circuitry. XTAL1 XTAL2 RQ Rx2 mc_osc0100_extcirc.vsd Figure 6-4 External (Main) Oscillator Circuitry The on-chip oscillator is optimized for an input frequency range of 4 to 16 MHz. An external clock signal (e.g. from an external oscillator or from a master device) may be fed to the input XTAL1. The Pierce oscillator then is not required to support the oscillation itself but is rather driven by the input signal. In this case the input frequency range may be 0 to 50 MHz (please note that the maximum applicable input frequency is limited by the device’s maximum clock frequency). Note: Oscillator measurement within the final target system is strongly recommended to verify the input amplitude and to determine the actual oscillation allowance (margin or negative resistance) for the oscillator-crystal system. The measurement technique is provided in a specific application note about oscillators (available via your representative or www). The main oscillator is automatically switched off during Power-Down mode and Sleep mode, unless it provides the count clock for the RTC and the RTC remains on. Switching off the main oscillator is useful to further reduce the power consumption during phases where only minimum system life functions must be maintained. User’s Manual SCU_X7, V2.1 6-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Main Oscillator Gain Reduction The main oscillator starts with a high drive level (gain) during and after a hardware reset to ensure safe startup behavior in the beginning (force the crystal oscillation). The beginning of the crystal oscillation is indicated by bit OSCLOCK (= 1) in register SYSSTAT. When a stable oscillation has been reached after oscillator startup (amplitude more than 90% of its maximum), the gain of the main oscillator can be reduced. This reduces the oscillator’s power consumption which is especially important in power reduction modes. This gain reduction is induced by software and so is transparent in existing software. The oscillator gain is reduced by setting bit OSCGRED in register SYSCON0 (see Section 6.3). Because the oscillator amplitude is not measured directly, a delay of approximately 215 oscillator clock cycles is required before enabling the gain reduction. The occurrence of 215 consecutive oscillator clock cycles is indicated by bit OSCSTAB (= 1) in register SYSSTAT. The oscillator gain reduction is disabled while OSCSTAB = 0. Therefore, software can set bit OSCGRED at any time. If OSCGRED is set before OSCSTAB = 1 the gain reduction is automatically delayed. Note: After the delay indicated by OSCSTAB = 1 the oscillation has reached more than 90% of its maximum amplitude with an optimized oscillator circuitry. Oscillator measurement (margin or negative resistance) for the oscillator-crystal system must be executed also in reduced-gain mode if this mode is intended in the application. If the main oscillator is switched off during sleep mode, both bits OSCLOCK and OSCSTAB are cleared and the oscillator startup begins anew. This ensures a safe oscillator startup after wake-up. User’s Manual SCU_X7, V2.1 6-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions The auxiliary oscillator of the XC167 is a Pierce oscillator which is highly optimized for operation at a frequency of approximately 32 kHz. This narrow band optimization allows an extremely low power consumption of the auxiliary oscillator. Pins XTAL3 and XTAL4 connect the inverter to the external crystal. The recommendations given for the main oscillator apply accordingly. Note: When the auxiliary oscillator is not used, i.e. is not connected to an external clock signal or to a crystal, its input XTAL3 should be connected to GND. XTAL3 XTAL4 RQ mc_osc0101_extaux.vsd Figure 6-5 External (Auxiliary) Oscillator Circuitry The auxiliary oscillator is automatically switched off while it provides the count clock for the RTC and the RTC is not running. Note: Oscillator measurement within the final target system is strongly recommended to verify the input amplitude and to determine the actual oscillation allowance (margin or negative resistance) for the oscillator-crystal system. The measurement technique is provided in a specific application note about oscillators (available via your representative or www). User’s Manual SCU_X7, V2.1 6-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2.2 Clock Generation and Frequency Control The Clock Generation Unit uses a programmable on-chip PLL with multiple prescalers to generate the clock signals for the XC167 with high flexibility. The internal operation of the XC167 is controlled by the internal master clock fMC. The master clock fMC is the reference clock signal, and is used for TwinCAN and is output to the external system. CPU and EBC are clocked with the CPU clock signal fCPU. The CPU clock can have the same frequency as the master clock (fCPU = fMC) or can be the master clock divided by two: fCPU = fMC/2. This factor is selected by bit CPSYS in register SYSCON1. The other peripherals are supplied with the system clock signal fSYS which has the same frequency as the CPU clock signal (fSYS = fCPU). The oscillator clock frequency can be multiplied by the on-chip PLL (by a programmable factor) or can be divided by a programmable prescaler factor. With these options the master clock can be adjusted to a wide range of frequencies. PLL operation achieves maximum performance even from moderate crystal frequencies, dividing the oscillator clock runs the system at low frequency, greatly reducing its power consumption. Phase Locked Loop Operation (1:N) fOSC fMC Direct Clock Drive (1:1) fOSC fMC Prescaler Operation (N:1) fOSC fMC MCT05332 Figure 6-6 User’s Manual SCU_X7, V2.1 Generation Mechanisms for the Master Clock 6-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Note: The example for PLL operation shown in Figure 6-6 refers to a PLL factor of 1:4, the example for prescaler operation refers to a divider factor of 2:1. The Clock Generation Unit (CGU) summarizes the following required functions to generate the clock signals used in the XC167: • • • • Generation of the master clock signal from the oscillator clock according to userprogrammed mode and factor Generation of clock signals for specific functional areas Control the oscillator operation according to the XC167’s operating mode Generation of an interrupt request in case of detected malfunctions of the clock system CGU Main OSC fOSCm Osc. Watchdog PLL Unlock fMC PLL N:1 fOSCa fCPU fSYS Run Control Aux. OSC IRQ fRTCm 32:1 fRTCa N = CPSYS + 1 MCA05333 Figure 6-7 Basic Structure of the Clock Generation Unit Note: The divider factor for the CPU clock and the system clock is selected by bit CPSYS in register SYSCON1. The master clock signal is generated by the highly-flexible on-chip PLL. The PLL block can multiply the oscillator clock frequency by a programmable factor (1:0.15 … 1:10) to achieve high performance even from moderate crystal frequencies. In bypass mode the oscillator clock is divided by a factor of 1:1 … 60:1 to achieve direct coupling to the oscillator clock signal (1:1) or reduce the system frequency to save power. The used mechanism to generate the master clock and the respective parameters are selected via the PLL control register PLLCON. User’s Manual SCU_X7, V2.1 6-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions PLLCON PLL Control Register 15 14 13 12 ESFR (F1D0H/E8H) 11 10 9 8 7 6 Reset Value: XXXXH 5 4 3 2 1 PLL WRI PLL CTRL PLLMUL PLLVB PLLIDIV PLLODIV rh rw rw rw rw rw 0 Field Bits Type Description PLLWRI 15 rh PLLCON Write Ignore Flag 0 Register PLLCON may be written 1 Write cycles to register PLLCON are ignored PLLCTRL [14:13] rw PLL Operation Control 00 Bypass PLL clock mult., the VCO is off 01 Bypass PLL clock mult., the VCO is running 10 VCO clock used, input clock switched off 11 VCO clock used, input clock connected PLLMUL [12:8] rw PLL Multiplication Factor … by which the PLL multiplies its input frequency (valid values: 1’1111B … 0’0111B)1) fVCO = fIN × (PLLMUL+1) PLLVB [7:6] rw PLL VCO Band Select2) Value, VCO output frequency, Base frequency 00 100 … 150 MHz, 20 … 80 MHz 01 150 … 200 MHz, 40 … 130 MHz 10 200 … 250 MHz, 60 … 180 MHz 11 Reserved PLLIDIV [5:4] rw PLL Input Divider Adjusts the oscillator frequency to the defined input frequency range of the PLL (valid values: 11B … 00B) fIN = fOSC / (PLLIDIV+1) PLLODIV [3:0] rwh PLL Output Divider Scales the PLL output frequency to the desired CPU frequency (valid values: 1110B … 0000B)3) fMC = fVCO / (PLLODIV+1) 1) Multiplication factors below N = 8 (PLLMUL = 7) may affect stability. For example, this may lead to undesired VCO frequencies due to increased noise-susceptibility. 2) The VCO band must be selected to contain the intended VCO frequency (8 MHz × 20 = 160 MHz --> band 01B). 3) Value 1111B is reserved for emergency mode operation and cannot be entered via software. User’s Manual SCU_X7, V2.1 6-32 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Note: PLLCON is protected by the register security mechanism (see Section 6.3.5). The reset value depends on the initial system startup configuration (see Table 6-5). The clock generation path is controlled by a state machine according to the selection in register PLLCON. Bit PLLWRI = 0 indicates when this state machine is ready to accept a new selection value in PLLCON. To support monitoring, register PLLCON accepts the (new) selection for clock configuration when written, and returns the actual state of the clock generation mechanism when read. The PLL module contains the frequency multiplication logic, a set of prescalers, the lock detection, and the oscillator watchdog. PLL OWD Lock fOSC OSC P:1 fIN VCO fVCO MUX fMC K:1 1:N Bypass P = PLLIDIV + 1 N = PLLMUL + 1 K = PLLODIV + 1 MCB05334 Figure 6-8 PLL Block Diagram PLL Operation When PLL operation is configured (PLLCTRL = 11B), the XC167’s input clock is fed to the on-chip Phase Locked Loop circuit which can multiply its frequency by a factor of up to F = 10 and generates a master clock signal, i.e. fMC = fOSC × F. The on-chip PLL circuit allows operation of the XC167 on a low frequency external clock while still providing maximum performance. The PLL also provides fail safe mechanisms which allow the detection of frequency deviations and the execution of emergency actions in case of an external clock failure. User’s Manual SCU_X7, V2.1 6-33 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions When the PLL detects a missing input clock signal it generates an interrupt request. This warning interrupt indicates that the PLL frequency is no longer locked, i.e. no longer related to the oscillator frequency. This occurs when the input clock is unstable and especially when the input clock fails completely, such as due to a broken crystal. In this case the synchronization mechanism will reduce the PLL output frequency down to the PLL’s base frequency (depending on the VCO band selected by bitfield PLLVB) and select the safety output divider factor K = 16. The base frequency is still generated and allows the CPU to execute emergency actions in case of a loss of the external clock. The master clock in this emergency case is, therefore, fMCe = fVCObase/16. Note: During a hardware reset the lowest VCO band is selected together with factor 16. On power-up the PLL provides a stable clock signal, even if there is no external clock signal (in this case the PLL will run on its base frequency). The PLL starts synchronizing with the external clock signal as soon as it is available. After stable oscillations of the external clock within the specified frequency range the PLL locks to the external clock. This means the PLL will be synchronous with this clock at a frequency of F × fOSC. The PLL circuit constantly synchronizes the master clock to the input clock. This synchronization is done smoothly, i.e. the master clock frequency does not change abruptly. Due to the fact that the external frequency is 1/Fth of the PLL output frequency the output frequency may be slightly higher or lower than the desired frequency. The slight variation causes a jitter of fMC which also affects the duration of individual master clock periods. This jitter is irrelevant for longer time periods. For short periods (1 … 4 CPU clock cycles) it remains below 9%. The clock signal passes through several blocks (see Figure 6-8). The total clock multiplication factor F results from the input divider (P:1), the multiplication factor (1:N), and the output divider (K:1), so F = PLLMUL+1 / ((PLLIDIV+1) × (PLLODIV+1)). The input clock divider adjusts the oscillator clock frequency to the input frequency range for which the PLL is optimized (fIN = fOSC / (PLLIDIV+1) = 4 … 35 MHz). The PLL core multiplies the adjusted input frequency within the selected VCO band by a selectable factor (fVCO = fIN × (PLLMUL+1)). The valid VCO band (PLLVB) must be selected according to the intended VCO frequency. Note: This PLL core can be bypassed, e.g. while the PLL is locking to a given factor, to ensure a proper CPU clock signal, or if the PLL is not used for clock generation. The output clock divider scales the VCO’s output frequency by a selectable factor to generate the master clock signal (fMC = fVCO / (PLLODIV+1)). Adjusting this factor may be used to control the operating frequency of the XC167 without having to reprogram the PLL core itself. Figure 6-9 summarizes the subsequent steps the master clock generation. The maximum multiplication factor Fmax is employed when the highest possible master clock frequency (40 MHz) shall be generated from the lowest possible input clock frequency (4 MHz). Therefore, the maximum usable factor is Fmax = 10. User’s Manual SCU_X7, V2.1 6-34 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Application software can select the optimum clock generation mode via registers PLLCON and SYSCON1. After reset the XC167 enters a default clock generation mode, which can be coarsely adjusted to the actual oscillator frequency by external configuration in case of an external start (see Table 6-7). fIN 1:1 1:8 [MHz] 35 2:1 25 20 16 10 1:32 4 fOSC fVCO [MHz] [MHz] 50 30 16 100 4 150 200 250 10 PLLIDIV PLLMUL PLLVB 16:1 16 20 30 PLLODIV 8:1 40 1:1 Respective Control Factors 50 3:1 fMC [MHz] MCA05335 Figure 6-9 User’s Manual SCU_X7, V2.1 Valid Clock Frequency Bands 6-35 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Bypass Operation When bypass operation is configured (PLLCTRL = 0XB) the clock signal does not pass through the PLL core and the master clock is derived from the internal oscillator (input clock signal XTAL1) through the input- and output-prescalers: fMC = fOSC / ((PLLIDIV+1) × (PLLODIV+1)). If both divider factors are selected as 1 (PLLIDIV = PLLODIV = 0) the frequency of fMC directly follows the frequency of fOSC so the high and low time of fMC is defined by the duty cycle of the input clock fOSC. The lowest master clock frequency is achieved by selecting the maximum values for both divider factors: fMCmin = fOSC / ((3+1) × (14+1)) = fOSC / 60. Master Clock Duty Cycle The master clock signal fMC is formed by the output divider (factor K = PLLODIV+1). The duty cycle of fMC depends on the selected output divider factor. For all even factors the duty cycle is 50%, for all odd factors the duty cycle is (K-1)/(K×2)%. The worst case here is K = 3, which leads to a duty cycle of (3-1)/(3×2)% = 2/6% = 33%. User’s Manual SCU_X7, V2.1 6-36 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2.3 Clock Distribution The operating clock signals are distributed to the controller hardware via several clock drivers. This establishes the corresponding clock domains summarized in Table 6-11. The real time clock RTC is clocked via a separate clock driver which delivers the auxiliary oscillator clock or the prescaled main oscillator clock. Table 6-11 Clock Domains Clock Domain Domain Active Clock Mode Idle Mode Sleep, Connected Circuitry P. Down Module Clock LXBus fMC ON Off TwinCAN fCAN SCU1) – ON CPU fCPU ON Off Off CPU, DPRAM, EBC, OCDS, Flash, PSRAM, DSRAM – PDBus fSYS ON ON Off ADC fADC fASC fCC fCC6 fGPT fIIC fSSC ASC0, ASC1 CAPCOM1, CAPCOM2 CAPCOM6 GPT12 IIC SSC0, SSC1 Ports, RTC2), WDT, SCU – (Intr. Ctrl., Reg. access)1) RTC fRTC ON ON ON/Off RTC2) – 1) As the clock generation unit is part of the SCU, the SCU consequently belongs to more than one clock domain. 2) The RTC is part of the PDBus clock domain which provides its operating clock. The count clock signal is derived directly from the auxiliary oscillator or from the main oscillator (as selected). Note: All PDBus peripherals are provided with the clock signal fSYS. Within a peripheral description, however, this clock signal is called according to the peripheral’s name. Table 6-11 shows this in column “Module Clock”. User’s Manual SCU_X7, V2.1 6-37 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2.4 Oscillator Watchdog The XC167 provides an Oscillator Watchdog (OWD) which monitors the clock signal fed to input XTAL1 of the on-chip oscillator (either with a crystal or via external clock drive) in bypass mode (not if the PLL provides the master clock). For this operation, the PLL provides a VCO clock signal (base frequency) which is used to supervise transitions on the oscillator clock. This VCO clock is independent from the XTAL1 clock. When the expected oscillator clock transitions are missing, the OWD activates the PLL Unlock/OWD interrupt node and supplies the CPU with an emergency clock instead of the selected oscillator clock. Under these circumstances the VCO will oscillate with its base frequency in the selected VCO band. The emergency clock frequency is fVCO/16. If the oscillator clock fails while the PLL provides the master clock the system will be supplied with the PLL base frequency anyway. With this emergency clock signal the CPU can either execute a controlled shutdown sequence bringing the system into a defined and safe idle state, or it can provide an emergency operation of the system with reduced performance based on this (normally slower) emergency clock. Note: In special cases, when bypass mode is selected with a high prescaler factor, the emergency clock frequency may be higher than the originally intended frequency. The oscillator watchdog can be disabled by switching the VCO off (PLLCTRL = 00B). In this case the VCO remains idle and provides no clock signal, while the master clock signal is derived directly from the oscillator clock. This reduces power consumption, but also no interrupt request will be generated in case of a missing oscillator clock. Note: At the end of a hardware reset triggering a standard external start (EA = 0, ALE = 1) the VCO (and thus the oscillator watchdog) may also be disabled via hardware by (externally) pulling the RD line low, similar to the standard reset configuration via PORT0. 6.2.5 Interrupt Generation When the PLL leaves its locked state or when the OWD detects an improper clock input signal, the CGU issues an interrupt request. PLL_IC PLL Interrupt Ctrl. Reg. ESFR (F19EH/CFH) 15 14 13 12 11 10 9 8 7 - - - - - - - GPX - - - - - - - rw 6 PLL PLL IR IE rwh rw Reset Value: 0000H 5 4 3 2 1 0 ILVL GLVL rw rw Note: Please refer to the general Interrupt Control Register description for an explanation of the control fields. User’s Manual SCU_X7, V2.1 6-38 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.2.6 Generation of an External Clock Signal The external circuitry can be provided with a clock signal either to operate external peripherals or for reference purposes. Two types can be selected: • • CLKOUT directly outputs the master clock signal fMC and is mainly used as a timing reference FOUT outputs a clock signal with a programmable frequency and can be used to drive and control external circuitry The programmable frequency output signal fOUT can be controlled via software (contrary to CLKOUT), and so can be adapted to the requirements of the connected external circuitry. The programmability also extends the power management to a system level, as also circuitry (peripherals, etc.) outside the XC167 can be influenced, i.e. run at a scalable frequency or temporarily can be switched off completely. This clock signal is generated via a reload counter, so the output frequency can be selected in small steps. An optional toggle latch provides a clock signal with a 50% duty cycle. FORV FOEN Ctrl. fOUT MUX fCPU FOCNT FOTL FOSS MCA04480 Figure 6-10 Clock Output Signal Generation Signal fOUT always provides complete output periods (see Figure 6-11): • • When fOUT is started (FOEN --> 1), FOCNT is loaded from FORV When fOUT is stopped (FOEN --> 0), FOCNT is stopped when fOUT has reached (or is) 0. Register FOCON provides control over the output signal generation (output signal type, frequency, waveform, activation) as well as all status information (counter value, FOTL). User’s Manual SCU_X7, V2.1 6-39 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions FOCON Frequ. Output Control Reg. 15 14 13 12 11 10 FO EN FO SS FORV rw rw rw SFR (FFAAH/D5H) 9 8 7 6 Reset Value: 0000H 5 4 CLK FO EN TL rw rh 3 2 1 0 FOCNT rh Field Bits Type Description FOEN 15 rw Frequency Output Enable 0 Frequency output generation stops when signal fOUT is/becomes low. 1 FOCNT is running, fOUT is gated to pin. First reload after 0-1 transition. FOSS 14 rw Frequency Output Signal Select 0 Output of the toggle latch: duty cycle = 50%. 1 Output of the reload counter: duty cycle depends on FORV. FORV [13:8] rw Frequency Output Reload Value Is copied to FOCNT upon each underflow of FOCNT. CLKEN 7 rw CLKOUT Enable 0 CLKOUT signal disabled, P3.15 is IO or outputs FOUT (default) 1 P3.15 outputs signal CLKOUT (f = fMC) FOTL 6 rh Frequency Output Toggle Latch Is toggled upon each underflow of FOCNT. FOCNT [5:0] rh Frequency Output Counter Note: Bitfield FOCNT and bit FOTL cannot be written. This prevents the generation of invalid clock cycles when writing to register FOCON, for example to change the output frequency or to stop the output clock signal. FOCON is write protected after the execution of EINIT by the register security mechanism (see Section 6.3.5). During the generation of CLKOUT and fOUT the shared pin is automatically switched to output. While fOUT is disabled, the pin is controlled by the port latch and the direction latch. Pin FOUT must be switched to output and the port latch must be 0 in order to maintain the fOUT inactive level at the pin. User’s Manual SCU_X7, V2.1 6-40 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Signals CLKOUT and fOUT in the XC167 are alternate output pin functions. A priority ranking determines which function controls the shared pin: Table 6-12 Priority Ranking for Shared Output Pin Priority Function Control 1 CLKOUT CLKEN = 1, FOEN = x 2 FOUT CLKEN = 0, FOEN = 1 3 General purpose IO CLKEN = 0, FOEN = 0 fMC 1) fOUT (FORV = 0) 2) 1) fOUT (FORV = 2) 2) 1) fOUT (FORV = 5) 2) FOEN 1 1) FOSS = 1, Output of Counter 2) FOSS = 0, Output of Toggle Latch FOEN The counter starts here 0 The counter stops here mc_xc16x_foutwaves.vsd Figure 6-11 Signal Waveforms Note: The output signal (for FOSS = 1) is high for the duration of one fMC cycle for all reload values FORV > 0. For FORV = 0 the output signal corresponds to fMC. User’s Manual SCU_X7, V2.1 6-41 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Output Frequency Calculation The output frequency can be calculated as fOUT = fMC / ((FORV + 1) × 2(1 - FOSS)), so fOUTmin = fMC / 128 (FORV = 3FH, FOSS = 0), and fOUTmax = fMC / 1 (FORV = 00H, FOSS = 1). Selectable Output Frequency Range for fOUT Table 6-13 fMC fOUT in [kHz] for FORV = xx, FOSS = 1/0 FORV for fOUT = 1 MHz 00H 01H 02H 3EH 3FH FOSS = 0 FOSS = 1 4 MHz 4000 2000 2000 1000 1333.33 666.67 63.492 31.746 62.5 31.25 01H 03H 10 MHz 10000 5000 5000 2500 3333.33 1666.67 158.73 79.365 156.25 78.125 04H 09H 12 MHz 12000 6000 6000 3000 4000 2000 190.476 95.238 187.5 93.75 05H 0BH 16 MHz 16000 8000 8000 4000 5333.33 2666.67 253.968 126.984 250 125 07H 0FH 20 MHz 20000 10000 10000 5000 6666.67 3333.33 317.46 158.73 312.5 156.25 09H 13H 25 MHz 25000 12500 12500 6250 8333.33 4166.67 396.825 198.413 390.625 195.313 0BH (1.04167) 0CH (0.96154) 18H 33 MHz 33000 16500 16500 8250 11000 5500 523.810 261.905 515.625 257.813 0FH (1.03125) 10H (0.97059) 20H User’s Manual SCU_X7, V2.1 6-42 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.3 Central System Control Functions Most control functions and status information are tightly coupled to the respective peripheral modules of the XC167. However, some of these functions are valid for the complete device, rather than for a specific module. These functions, including the associated control- and status-bits, are part of the SCU. SYSCON0 General System Control Reg. 15 14 ESFR (F1BEH/DFH) Reset Value: 0000H 13 12 11 10 9 8 7 6 5 4 3 2 1 0 RTC RTC RST CM - OSC G RED - - - - - - - - - - - - rwh - rw - - - - - - - - - - - - rw Field Bits Type Description RTCRST 15 rwh RTC Reset Trigger 0 No action 1 The RTC module is reset1) Note: RTCRST returns to 0 one SCU clock after being set. RTCCM 14 rw RTC Clocking Mode1) 0 Synchronous mode: The RTC operates with the system clock. Registers can be read and written. 1 Asynchronous mode:2) The RTC operates with the (asynchronous) count clock. No write access is possible. OSCGRED 12 rw Oscillator Gain Reduction Control 0 No reduction, retain initial gain level 1 Reduce gain (see Section 6.2.1) 1) After an RTC reset, the RTC immediately enters the clocking mode currently selected by bit RTCCM. 2) Asynchronous mode is required if the system clock is slower than the 4 × fCOUNT. This is, of course, the case in Sleep mode or Powerdown mode, where the system clock is disabled, while the RTC shall continue to run (see also Chapter 15). Note: SYSCON0 is protected by the register security mechanism (see Section 6.3.5). The reset value is only valid for a hardware reset. User’s Manual SCU_X7, V2.1 6-43 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Register SYSCON1 selects the following functions: • • • • Master clock prescaler factor for the system Program Flash behavior in Idle/Sleep mode Port driver behavior during Sleep mode and Powerdown mode Selection of Idle mode or Sleep mode SYSCON1 System Control Reg. 1 ESFR (F1DCH/EEH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - - - - - CP SYS - - PF CFG PD CFG SLEEP CON - - - - - - - rw - - rw rw rw Field Bits Type Description CPSYS 8 rw Clock Prescaler for System (see Section 6.2.2) The clock signal for the CPU is prescaled: 0 fCPU = fMC 1 fCPU = fMC/2 PFCFG [5:4] rw Program Flash Configuration1) 00 Program Flash is always ON (default) 01 Program Flash is off in IDLE or Sleep mode 10 Reserved 11 Reserved PDCFG [3:2] rw Port Driver Configuration 00 Port drivers are always ON (default) 01 Port drivers are off in IDLE or Sleep mode 10 Port drivers are off in Powerdown mode 11 Reserved SLEEPCON [1:0] rw SLEEP Mode Configuration (mode entered upon the IDLE instruction) 00 Enter normal IDLE mode 01 Enter SLEEP mode 10 Reserved 11 Reserved 1) In Powerdown mode the Program Flash will always be off. Note: SYSCON1 is protected by the register security mechanism (see Section 6.3.5). User’s Manual SCU_X7, V2.1 6-44 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.3.1 Status Indication The system status register SYSSTAT indicates the status of the clock generation unit and the recent reset with a number of flags. SYSSTAT System Status Register 15 14 13 12 mem (F1E4H/--) 11 10 OSC PLL OSC CLK CLK PLL LOC LOC STA HIX LOX EM K K B rh rh rh rh rh rh Reset Value: XXXXH 9 8 7 6 5 4 3 2 1 - - - - - - - HW R - - - - - - - rh 0 SW WDT R R rh rh Field Bits Type Description OSCLOCK 15 rh Oscillator Signal Status Bit 0 The oscillator is unlocked 1 The oscillator is locked (2048 fOSC periods have been counted, so it is assumed as stable) PLLLOCK 14 rh PLL Signal Status Bit 0 PLL unlocked (base frequency or adjusting) 1 The PLL is locked (stable output frequency) CLKHIX 13 rh Input Clock High Limit Exceeded 0 The input clock frequency is below the upper limit of the monitored range 1 The input clock frequency is too high CLKLOX 12 rh Input Clock Low Limit Exceeded 0 The input clock frequency is above the lower limit of the monitored range 1 The input clock frequency is too low OSCSTAB 11 rh Oscillator Stable Flag 0 The oscillator is starting up 1 The oscillator counter has reached its upper threshold (215 fOSC periods). With default gain, this indicates that the oscillator has reached 90% of its maximum amplitude. OSCSTAB is cleared upon a hardware reset or after a walk-up trigger when the oscillator was off. User’s Manual SCU_X7, V2.1 6-45 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Field Bits Type Description PLLEM 10 rh PLL Emergency Mode Flag 0 No clock generation problem encountered 1 A clock generation problem has occurred Note: PLLEM is cleared automatically if the oscillator has locked after a wake-up from sleep mode. Otherwise it remains set until hardware reset. HWR 2 rh Hardware Reset Indication Flag 0 Last reset was no hardware reset 1 Last reset was a hardware reset SWR 1 rh Software Reset Indication Flag 0 Last reset was no software reset 1 Last reset was a software reset WDTR 0 rh Watchdog Timer Reset Indication Flag 0 Last reset was no watchdog timer reset 1 Last reset was a watchdog timer reset Note: The reset value of register SYSSTAT depends on the active status flags. 6.3.2 Reset Source Indication Reset indication flags in register SYSSTAT provide information about the source of the last reset. After the XC167 starts execution, the initialization software may check these flags to determine if the recent reset event was triggered by an external hardware signal (via RSTIN), by software, or by an overflow of the watchdog timer. The initialization and further operation of the microcontroller system can thus be adapted to the respective circumstances. For instance, a special routine may verify software integrity after a watchdog timer reset. The reset indication flags are mutually exclusive; only one flag is set after reset depending on its source. Hardware Reset is indicated when the RSTIN input is sampled low (active). Software Reset is indicated after a reset triggered by the execution of instruction SRST. Watchdog Timer Reset is indicated after a reset triggered by an overflow of the watchdog timer. User’s Manual SCU_X7, V2.1 6-46 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.3.3 Peripheral Shutdown Handshake When executing a software reset or when entering Powerdown mode the SCU requests a shutdown from those peripheral units that are currently active and provide the shutdown handshake mechanism. Upon this request the respective peripheral unit completes the currently active action (if any) and then acknowledges the shutdown request to the SCU. These units are: PMU, DMU, EBC, ADC, and Program Flash. The shutdown handshake sequence is completed as soon as all units have acknowledged the shutdown request. User’s Manual SCU_X7, V2.1 6-47 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.3.4 Debug System Control The debug and emulation circuitry is controlled by two registers: • • The Emulation Control register EMUCON controls basic OCDS functions. The OCE/OCDS Peripheral Suspend Enable register OPSEN selects the peripherals that will be halted by the suspend signal. OPSEN is structured identically to register SYSCON3. EMUCON Emulation Control Reg. SFR (FE0AH/05H) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - - - - - - - - - - - OC EN OCD SIO EN - - - - - - - - - - - - - - rw rw - Field Bits Type Description OCEN 2 rw OCDS/Cerberus Enable 0 OCDS and Cerberus are still in reset state 1 ODCS and Cerberus are operable OCDSIOEN 1 rw OCDS Break Input/Output Enable 0 OCDS break input/output BRKIN/BRKOUT are disabled. 1 OCDS break input/output BRKIN/BRKOUT are enabled. Note: EMUCON is write protected after the execution of EINIT by the register security mechanism (see Section 6.3.5). User’s Manual SCU_X7, V2.1 6-48 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions OPSEN OCE/OCDS P-Susp. En. Reg. 15 14 13 SSC RTC CAN 1 SEN SEN SEN rw rw rw ESFR (FE58H/2CH) 12 11 10 9 - I2C SEN ASC 1 SEN - - rw rw - 8 7 6 Reset Value: 0000H 5 CC6 CC2 CC1 PFM SEN SEN SEN SEN rw rw rw rw 4 3 - GPT SEN - rw 2 1 0 SSC ASC ADC 0 0 SEN SEN SEN rw Field Bits Type Description xxSEN [15:13], [11:10], [8:5], [3:0] rw Module xx Suspend Enable 0 Respective module remains active during suspend 1 Respective module halts operation during suspend rw rw Note: For most debugging scenarios it is recommended to keep bit PFMSEN cleared. OPSEN is write protected after the execution of EINIT by the register security mechanism (see Section 6.3.5). User’s Manual SCU_X7, V2.1 6-49 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.3.5 Register Security Mechanism There are some dedicated registers which control critical functions and modes. These registers are protected by a special register security mechanism so these vital system functions cannot be changed inadvertently. This security mechanism controls three different security levels: • • • Write Protected Mode (entered after the execution of EINIT) Protected registers are locked against any write access (read only). Secured Mode Protected registers can be written using a special command sequence. Unprotected Mode (entered after reset) No protection is active. Registers can be written at any time. Note: The selected security level applies to all protected registers throughout the XC167 (see Table 6-15). Controlling the Security Level Two registers build the interface for controlling the security level. The security level command register SCUSLC accepts the commands to control the state machine modifying the security level (the required command sequence is safeguarded with a password). The security level status register SCUSLS (read only) shows the actual password, the actual security level, and the state of the switching state machine. Any SCU register write access State 4 Command 2 or any other SCU register write access Command 1 or any other SCU register write access State 1 Command 1 State 3 Command 2 Command 4 and low protected mode Command 0 State 0 Reset Command 3 or any other SCU register write access State 2 MCA05336 Figure 6-12 State Machine for Security Level Switching User’s Manual SCU_X7, V2.1 6-50 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Two mechanisms can be used to control the actual security level: • Changing the security level can be done by executing the following command sequence: “command0-command1-command2-command3”. This sequence establishes a new security level and/or a new password. Access in secured mode can be achieved by preceding the intended write access with writing “command4” to register SCUSLC. This quick access is only possible while secured mode is selected. • Read accesses are always possible to all registers of the SCU and will not influence the command sequences. In register SCUSLS the actual status of the command state machine can always be read. Note: After writing command4 in secured mode the lock mechanism remains disabled until the next write access to an SCU register or a register on the PD bus, i.e. accesses to registers outside this area do not re-activate the protection. SCUSLS Sec. Level Status Reg. 15 14 13 12 ESFR (F0C2H/61H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 STATE SL - - - PASSWORD rh rh - - - rh 1 Field Bits STATE [15:13] rh Current State of Switching State Machine 000 Awaiting command0 or command4 (default) 001 Awaiting command1 010 Awaiting command2 011 Awaiting new security level and password 100 Next access granted in secured mode 101 Reserved 11X Reserved SL [12:11] rh Security Level 00 Unprotected mode (default after reset) 01 Secured mode 10 Reserved 11 Write protected mode (entered after EINIT) PASSWORD [7:0] Current Security Control Password Default after reset = 00H User’s Manual SCU_X7, V2.1 Type 2 rh 0 Description 6-51 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions SCUSLC Sec. Level Command Reg. 15 14 13 12 11 ESFR (F0C0H/60H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 COMMAND rw Field Bits Type Description COMMAND [15:0] rw Security Level Control Command The commands to control the security level must be written to this register (see Table 6-14) Note: Register SCUSLC is not protected by the security mechanism. This is required to be able to change the security level in any state. Table 6-14 Commands for Security Level Control Command Definition Note Command0 AAAAH – Command1 5554H – Command2 96H || <inverse password> – Command3 000B || <new level> || 000B || <new password> – Command4 8EH || <inverse password> Secured mode only Note: It is recommended to lock all command sequences with an atomic sequence. Programming Examples EXTR MOV MOV MOV MOV #4 SCUSLC, SCUSLC, SCUSLC, SCUSLC, #0AAAAH #05554H #096FFH #008EDH EXTR #1 MOV SCUSLC, #8E12H MOV register, data User’s Manual SCU_X7, V2.1 ;Sequence to change the security level ;Command0 ;Command1 ;Command2: current password = 00H ;Command3: level = 01, new password = EDH ;Access sequence in secured mode ;Command4: current password = EDH ;Access enabled by the preceding Command4 6-52 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions The Register Security Mechanism protects not only the SCU registers but also a number of registers in other modules. Table 6-15 summarizes these registers. Table 6-15 Registers Protected by the Security Mechanism Register Name Function RSTCON Reset control SYSCON0 General system control SYSCON1 Power management PLLCON Clock generation control SYSCON3 Peripheral management FOCON Peripheral management (CLKOUT/FOUT) IMBCTR Control of internal instruction memory block OPSEN Emulation control EMUCON Emulation control WDTCON Watchdog timer properties EXICON Ext. interrupt control EXISEL0, EXISEL1 Ext. interrupt control CPUCON1, CPUCON2 CPU configuration, protected after EINIT EBCMOD0, EBCMOD1 EBC mode selection TCONCSx EBC timing configuration FCONCSx EBC function configuration ADDRSELx EBC address window configuration User’s Manual SCU_X7, V2.1 6-53 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.4 Power Management The power consumption of the XC167 can be reduced by several mechanisms. The level of power reduction depends on the level of system performance that is required under these circumstances. The architecture of the XC167 provides three major means of reducing its power consumption under software control: • • • Power reduction modes (Idle, Sleep, Powerdown) to deactivate CPU, ports and control logic Reduction of the CPU frequency and system frequency Selection of the active peripheral modules (Flexible Peripheral Management) This enables the application (that is, the programmer) to choose the optimum constellation for each operating condition, so the power consumption can be adapted to conditions like maximum performance, partial performance, or standby. These three means can be applied independent from each other and thus provide a maximum of flexibility for each application. 6.4.1 Power Reduction Modes Three different general power reduction modes with different levels of power reduction have been implemented in the XC167, which are selected by dedicated instructions: IDLE selects Idle mode or Sleep mode, PWRDN selects Powerdown mode. Attention: Upon a reset all power reduction modes are terminated. Idle Mode In Idle mode all enabled peripherals, including the watchdog timer, continue to operate normally, only the CPU operation is halted. Note: Peripherals that have been disabled via software also remain disabled in Idle mode, of course. Idle mode is entered after the IDLE instruction has been executed (bitfield SLEEPCON in register SYSCON1 must be 00B) and the instruction before the IDLE instruction has been completed. To prevent unintentional entry into Idle mode, the IDLE instruction has been implemented as a protected 32-bit instruction. Idle mode is terminated by interrupt requests from any enabled interrupt source whose individual Interrupt Enable flag was set before the Idle mode was entered, regardless of bit IEN. User’s Manual SCU_X7, V2.1 6-54 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Denied CPU Interrupt Request Accepted Active Mode IDLE Instruction Idle Mode Denied PEC Request Executed PEC Request MCA04407 Figure 6-13 Transitions between Idle Mode and Active Mode For a request selected for CPU interrupt service the associated interrupt service routine is entered if the priority level of the requesting source is higher than the current CPU priority and the interrupt system is globally enabled. After the RETI (Return from Interrupt) instruction of the interrupt service routine is executed the CPU continues executing the program with the instruction following the IDLE instruction. Otherwise, if the interrupt request cannot be serviced because of a too low priority or a globally disabled interrupt system the CPU immediately resumes normal program execution with the instruction following the IDLE instruction. For a request which was programmed for PEC service a PEC data transfer is performed if the priority level of this request is higher than the current CPU priority and the interrupt system is globally enabled. After the PEC data transfer has been completed the CPU remains in Idle mode. Otherwise, if the PEC request cannot be serviced because of a too low priority or a globally disabled interrupt system the CPU does not remain in Idle mode but continues program execution with the instruction following the IDLE instruction. Idle mode can also be terminated by a Non-Maskable Interrupt, i.e. a high-to-low transition on the NMI pin. After Idle mode has been terminated by an interrupt or NMI request, the interrupt arbitration block performs a round of prioritization to determine the highest priority request. In the case of an NMI request, the NMI trap will always be entered. Any interrupt request whose individual Interrupt Enable flag was set before Idle mode was entered will terminate Idle mode regardless of the current CPU priority. The CPU will not go back into Idle mode when a CPU interrupt request is detected, even when the interrupt was not serviced because of a higher CPU priority or a globally disabled interrupt system (IEN = 0). The CPU will only go back into Idle mode when the interrupt system is globally enabled (IEN = 1) and a PEC service on a priority level higher than the current CPU level is requested and executed. User’s Manual SCU_X7, V2.1 6-55 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Note: An interrupt request which is individually enabled and assigned to priority level 0 will terminate Idle mode. The associated interrupt vector will not be accessed, however. The watchdog timer may be used to monitor the Idle mode: an internal watchdog timer reset will be generated if no interrupt or NMI request occurs before the watchdog timer overflows. To prevent the watchdog timer from overflowing during Idle mode it must be programmed to a reasonable time interval before Idle mode is entered. Sleep Mode In Sleep mode clocking of all internal blocks (including WDT) is stopped. The contents of the internal RAM, however, are preserved through the voltage supplied via the VDD pins. Sleep mode is entered after the IDLE instruction has been executed (bitfield SLEEPCON in register SYSCON1 must be 01B), the instruction before the IDLE instruction has been completed, and all individual enabled interrupt request flags are inactive. To prevent unintentional entry into Sleep mode, the IDLE instruction has been implemented as a protected 32-bit instruction. Sleep mode is terminated by an RTC interrupt (if enabled), by a transition on the external interrupt line or the respective alternate input line, when the respective level detection is enabled, and by an NMI request. The external interrupt level detection is controlled by register EXICON. The action, which is taken after the wake-up, depends on the individual interrupt enable flag’s setting of the fast external interrupt, which has triggered the wake-up, and on the setting of the global interrupt enable flag PSW.IEN. The XC167 switches from Sleep mode to Idle mode if the individual interrupt enable flag was not set by software before the wake-up has been triggered. Otherwise, the XC167 switches to active mode if the enable flag is 1. Whether a branch to the interrupt service routine or to the instruction following the IDLE is executed, depends on the setting of the global interrupt enable flag and on the interrupt level. This behavior is identical to the Idle mode. Note: The receive lines of serial interfaces may be internally routed to external interrupt inputs via registers EXISELn. All peripherals are stopped and hence cannot generate an interrupt request. The total power consumption in Sleep mode depends mainly on the current that flows through the port drivers. To minimize the consumed current all pin drivers can be disabled (pins switched to tristate) via a central control bit in register SYSCON1. If an application requires one or more port drivers to remain active even in Sleep mode also individual port drivers can be disabled simply by configuring them for input. User’s Manual SCU_X7, V2.1 6-56 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Powerdown Mode In Powerdown mode both the CPU and all peripherals are stopped. The contents of the internal RAM, however, are preserved through the voltage supplied via the VDD pins. Powerdown mode is entered after the PWRDN instruction has been executed, the instruction before the PWRDN instruction has been completed, and the NMI (Non Maskable Interrupt) pin is externally pulled low while the PWRDN instruction is executed. To prevent unintentional entry into Powerdown mode, the PWRDN instruction has been implemented as a protected 32-bit instruction and is additionally validated by the NMI signal. If pin NMI is not low at this time, Powerdown mode will not be entered and program execution continues. Powerdown mode is terminated only by a hardware reset (an internal reset cannot occur). The total power consumption in Powerdown mode depends mainly on the current that flows through the port drivers. To minimize the consumed current all pin drivers can be disabled (pins switched to tristate) via a central control bit in register SYSCON1. If an application requires one or more port drivers to remain active even in Powerdown mode also individual port drivers can be disabled simply by configuring them for input. 6.4.2 Reduction of Clock Frequencies The power consumption of the XC167 is linearly dependent on the logic’s switching frequency. The means to control this are described in Section 6.2, Clock Generation. 6.4.3 Flexible Peripheral Management The power consumed by the XC167 also depends on the amount of active logic. Peripheral management deactivates those on-chip peripherals that are not required in a given system status (e.g. a certain interface mode or standby). This reduces the amount of clocked circuitry. All modules that remain active, however, will still deliver their usual performance. Note: A read access to a register of a disabled peripheral returns the valid register content, whereas a write access to this register is ignored. A read access will not trigger any actions within a disabled peripheral. While a peripheral is disabled, its associated output pins remain in the state they had at the time of disabling. Software controls this flexible peripheral management via register SYSCON3 where each control bit is associated with an on-chip peripheral module. Writing SYSCON3 requests deactivation/activation of the respective peripheral(s). When writing to register SYSCON3, make sure that all undefined bits are written with 1s. Reading SYSCON3 returns the peripherals’ actual status according to the shutdown handshake mechanism (see Section 6.3.3). User’s Manual SCU_X7, V2.1 6-57 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions SYSCON3 System Control Reg. 3 15 14 13 12 11 10 - IIC DIS ASC 1 DIS - - rw rw - SSC RTC CAN 1 DIS DIS DIS rw rw ESFR (F1D4H/EAH) rw 9 8 7 Reset Value: 9FD0H 6 5 CC6 CC2 CC1 PFM DIS DIS DIS DIS rw rw rw rw 4 3 - GPT DIS - rw Field Bits Type Description SSC1DIS 15 rw Synchronous Serial Channel SSC1 RTCDIS 14 rw Real Time Clock CANDIS 13 rw On-chip CAN Module IICDIS 11 rw On-chip IIC Bus Module ASC1DIS 10 rw USART ASC1 CC6DIS 8 rw CAPCOM Unit 6 CC2DIS 7 rw CAPCOM Unit 2 CC1DIS 6 rw CAPCOM Unit 1 PFMDIS 5 rw Program Flash Module1) GPTDIS 3 rw General Purpose Timer Blocks SSC0DIS 2 rw Synchronous Serial Channel SSC0 ASC0DIS 1 rw USART ASC0 ADCDIS 0 rw Analog/Digital Converter 2 1 0 SSC ASC ADC 0 0 DIS DIS DIS rw rw rw 1) When the program flash module is deactivated in active mode (by setting bit PFMDIS), the next access to the program flash module will be answered with the trap code 1E9BH and produce a “Program Memory Access Error” trap. With the reset value 9FD0H, the program memory and a basic set of peripherals is enabled. The other peripherals of the XC167 must be enabled during initialization before they can be used. In ROM derivatives bit PFMDIS is not valid and, therefore, disabled after reset, so the reset value for ROM derivatives is 9FF0H. Note: The allocation of peripheral disable bits within register SYSCON3 is device specific and may be different in other derivatives than the XC167. SYSCON3 is write protected after the execution of EINIT by the register security mechanism (see Section 6.3.5). User’s Manual SCU_X7, V2.1 6-58 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.5 Watchdog Timer (WDT) To allow recovery from software or hardware failure, the XC167 provides a Watchdog Timer. If the software fails to service this timer before an overflow occurs, an internal reset sequence will be initiated. This internal reset can also pull the RSTOUT pin low, which in turn resets the peripheral hardware which might have caused the malfunction. If the watchdog timer is enabled and the software has been designed to service it regularly before it overflows, the watchdog timer will supervise the program execution so it will overflow only if the program does not progress properly. The watchdog timer will also time out if a software error was caused by hardware related failures. This prevents the controller from malfunctioning for a time longer than specified by the user. The watchdog timer provides two registers: • • a read-only timer register containing the current count, and a control register for initialization and reset source detection. R eset Indication P in RSTO UT D ata/S tatus R e gisters W DT C o ntrol R egisters W DTCO N SYSSTAT m SYSCON1 E CPUCON1 W D TC O N W DT W atchd og T im er C o ntrol R egister W ta chd og T im er C o unt R egister S Y S C O N 1 S ystem C ontro l R eg ister 1 S Y S S TA T S ystem S tatus R egister C P U C O N 1 C P U C ontrol R egiste r 1 m c a04381_x c.v sd Figure 6-14 SFRs and Port Pins Associated with the Watchdog Timer The watchdog timer is a 16-bit up counter which is clocked with the prescaled system clock (fSYS). The prescaler divides the system clock: • • • • by 2 (WDTIN = 00B), or by 4 (WDTIN = 10B), or by 128 (WDTIN = 01B), or by 256 (WDTIN = 11B). User’s Manual SCU_X7, V2.1 6-59 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions The 16-bit watchdog timer is implemented as two concatenated 8-bit timers (see Figure 6-15). The upper 8 bits of the watchdog timer can be preset to a userprogrammable value via a watchdog service access in order to vary the watchdog expire time. The lower 8 bits are reset after each service access. 1:2 MUX WDT Control WDTREL Clear 1:4 WDT Low Byte WDT High Byte WDTR 1:128 fSYS 1:256 WDT Enable WDTIN MCB05337 Figure 6-15 Watchdog Timer Block Diagram User’s Manual SCU_X7, V2.1 6-60 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Operation of the Watchdog Timer The current count value of the Watchdog Timer is contained in the Watchdog Timer Register WDT which is a non-bitaddressable read-only register. Operation of the Watchdog Timer is controlled by its bitaddressable Watchdog Timer Control Register WDTCON. This register specifies the reload value for the high byte of the timer, selects the input clock prescaling factor, and also provides flags to indicate the source of a reset. After any reset (except as noted) the watchdog timer is enabled and starts counting up from 0000H with the default frequency fWDT = fSYS/2. The default input frequency may be changed to another frequency (fWDT = fSYS/4, 128, 256) by programming the prescaler (bitfield WDTIN). The watchdog timer can be disabled by executing the instruction DISWDT (Disable Watchdog Timer). Instruction DISWDT is a protected 32-bit instruction. In compatible WDT mode instruction DISWDT will ONLY be executed during the time between a reset and execution of either the EINIT or the SRVWDT instruction. Either one of these instructions disables the execution of DISWDT. Once disabled, the WDT can only be enabled by a reset. In enhanced WDT mode the watchdog timer can be disabled and enabled at any time (independent of the EINIT instruction). This is controlled by executing instructions DISWDT and ENWDT, respectively. Instruction ENWDT is a protected 32-bit instruction. If the watchdog timer is re-enabled via instruction ENWDT it is implicitly serviced. The basic control mode (compatible/enhanced) is selected by bit WDTCTL in register CPUCON1. Note: After a hardware reset that activates the Bootstrap Loader the watchdog timer will be disabled. The WDT is enabled, when the loaded software begins executing. When the watchdog timer is not disabled via instruction DISWDT it will continue counting up, even in Idle Mode. If it is not serviced by the time the count reaches FFFFH the watchdog timer will overflow and cause an internal reset. This reset will pull the external reset indication pin RSTOUT low. The Watchdog Timer Reset Indication Flag (WDTR) in register SYSSTAT will be set in this case. Attention: A watchdog timer reset is unconditional. All current data/code accesses are aborted. To prevent the watchdog timer from overflowing, it must be serviced periodically by the user software. The watchdog timer is serviced by three different actions: • • • by executing instruction SRVWDT which is a protected 32-bit instruction by writing to register WDTCON by executing instruction ENWDT (in enhanced WDT mode) Servicing the watchdog timer clears the low byte and reloads the high byte of the watchdog timer register WDT with the preset value from bitfield WDTREL which is the User’s Manual SCU_X7, V2.1 6-61 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions high byte of register WDTCON. After servicing, the watchdog timer resumes counting up from the value (<WDTREL> × 28). Instruction SRVWDT has been encoded in such a way that the chance of unintentionally servicing the watchdog timer is minimized (such as by fetching and executing a bit pattern from a wrong location). When instruction SRVWDT does not match the format for protected instructions, the Protection Fault Trap will be entered, rather than executing the instruction. WDTCON WDT Control Register 15 14 13 12 SFR (FFAEH/D7H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 WDTREL - - - - - - WDTIN rw - - - - - - rw Field Bits Type Description WDTREL [15:8] rw Watchdog Timer Reload Value (for the high byte of WDT) WDTIN [1:0] rw Watchdog Timer Input Frequency Select 00 fWDT = fSYS/2 01 fWDT = fSYS/128 10 fWDT = fSYS/4 11 fWDT = fSYS/256 0 Note: WDTCON is protected by the register security mechanism (see Section 6.3.5). The time period for an overflow of the watchdog timer is programmable in two ways: • • Input frequency to the watchdog timer can be selected via a prescaler controlled by bitfield WDTIN in register WDTCON to be fSYS/2, fSYS/4, fSYS/128 or fSYS/256. Reload value WDTREL for the high byte of WDT can be programmed in register WDTCON. The period PWDT between servicing the watchdog timer and the next overflow can therefore be determined by the following formula: ( 1 + <WDTIN.1> + <WDTIN.0> × 6 ) P WDT 16 8 2 × ( 2 – <WDTREL> × 2 ) = ---------------------------------------------------------------------------------------------------------------------------------------- User’s Manual SCU_X7, V2.1 f SYS 6-62 (6.2) V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions Table 6-16 lists the possible ranges (depending on the prescaler bitfield WDTIN) for the watchdog time which can be achieved using a certain system clock. Table 6-16 Watchdog Time Ranges System Clock fSYS Prescaler WDTIN 10 MHz 00B 10B 01B 11B 20 MHz 00B 10B 01B 11B 30 MHz 00B 10B 01B 11B 40 MHz 00B 10B 01B 11B fWDT fSYS / 2 fSYS / 4 fSYS / 128 fSYS / 256 fSYS / 2 fSYS / 4 fSYS / 128 fSYS / 256 fSYS / 2 fSYS / 4 fSYS / 128 fSYS / 256 fSYS / 2 fSYS / 4 fSYS / 128 fSYS / 256 Reload Value in WDTREL FFH 7FH 00H 51.20 µs 6.61 ms 13.11 ms 102.4 µs 13.21 ms 26.21 ms 3.28 ms 422.7 ms 838.9 ms 6.55 ms 845.4 ms 1678 ms 25.60 µs 3.30 ms 6.55 ms 51.20 µs 6.61 ms 13.11 ms 1.64 ms 211.4 ms 419.4 ms 3.28 ms 422.7 ms 838.9 ms 17.07 µs 2.20 ms 4.37 ms 34.13 µs 4.40 ms 8.74 ms 1.09 ms 140.1 ms 279.6 ms 2.19 ms 281.8 ms 559.2 ms 12.80 µs 1.65 ms 3.28 ms 25.60 µs 3.30 ms 6.55 ms 0.82 ms 105.7 ms 209.7 ms 1.64 ms 211.4 ms 419.4 ms Note: The user is advised to rewrite WDTCON each time before the watchdog timer is serviced, particularly when the register security mechanism is disabled or when the software concept uses alternating watchdog periods. User’s Manual SCU_X7, V2.1 6-63 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions 6.6 Identification Control Block For identification of the most important silicon parameters a set of identification registers is defined that provide information on the chip manufacturer, the chip type and its properties. These ID registers can be used for automatic test selection as well as for identification of unknown silicon. Note: The not defined locations within the area 00’F070H … 00’F07EH are reserved for future identification features. IDMANUF Manufacturer Ident. Reg. 15 14 13 12 11 ESFR (F07EH/3FH) 10 9 8 7 6 Reset Value: 1820H 5 4 3 2 1 MANUF MANSEC r r 0 Field Bits Type Description MANUF [15:5] r Manufacturer This is the JEDEC normalized manufacturer code. 0C1H Infineon Technologies AG MANSEC [4:0] r Section within Manufacturer 00H Standard microcontroller IDCHIP Chip Identification Reg. 15 14 13 12 ESFR (F07CH/3EH) 11 10 9 8 7 6 Reset Value: 20XXH 5 4 3 CHIPID Revision r r 2 1 Field Bits Type Description CHIPID [15:8] r Device Identification Identifies the device name (reference via table). Revision [7:0] r Device Revision Code Identifies the device step. User’s Manual SCU_X7, V2.1 6-64 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) General System Control Functions IDMEM Program Mem. Ident. Reg. 15 14 13 12 11 ESFR (F07AH/3DH) 10 9 8 7 Reset Value: 3040H 6 5 Type Size r r Type 4 3 2 1 0 Field Bits Type [15:12] r Type of on-chip Program Memory Identifies the memory type on this silicon. 0H No on-chip program memory 1H Mask-programmable ROM 2H EEPROM memory 3H Flash memory 4H OTP memory Size [11:0] Size of on-chip Program Memory The size of the program memory in terms of 4-Kbyte blocks, i.e. Mem-size = <Size> × 4 Kbytes. r IDPROG Progr. Voltage Ident. Reg. 15 14 13 12 11 Description ESFR (F078H/3CH) 10 9 8 7 6 Reset Value: 4040H 5 4 3 PROGVPP PROGVDD r r 2 1 0 Field Bits Type Description PROGVPP [15:8] rw Programming VPP Voltage The voltage of the special programming power supply required to program or erase (if applicable) the on-chip program memory. Formula: VPP = 20 × <PROGVPP> / 256 [V]1) PROGVDD [7:0] r Programming VDD Voltage The voltage of the standard power supply required to program or erase the on-chip program memory. Formula: VDD = 20 × <PROGVDD> / 256 [V] 1) The XC167 needs no special programming voltage and PROGVPP = PROGVDD. User’s Manual SCU_X7, V2.1 6-65 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7 Parallel Ports This chapter describes the implementation details of the Parallel Ports in XC167. This includes the definition of registers associated with each Port, the assignment and control of alternate functions to each port pin, and configuration diagrams for each port pin. The XC167’s IO lines are organized into nine input/output ports and one input port. 15 Port Register 0 Direct. O.Drain AltSel P0L P0H P1L P1H P2 P3 P4 P5 P6 P7 P9 P20 mc_xc167_ports.vsd Figure 7-1 User’s Manual Ports_X7, V2.2 Port Overview of XC167 7-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.1 Input Threshold Control The standard inputs of the XC167 determine the status of input signals according to TTL levels. In order to accept and recognize noisy signals, CMOS-like input thresholds can be selected instead of the standard TTL thresholds for all pins of specific ports. These special thresholds are defined above the TTL thresholds and feature a defined hysteresis to prevent the inputs from toggling while the respective input signal level is near the thresholds. The Port Input Control register PICON allows to select these thresholds for each byte of the indicated ports, i.e. 8-bit ports are controlled by one bit each while 16-bit ports are controlled by two bits each. PICON Port Input Control Register 15 14 13 12 11 10 ESFR (F1C4H/E2H) 9 8 P20 P20 HIN LIN - rw Reset Value: 0000H 7 6 5 4 3 2 1 0 P9 LIN P7 LIN P6 LIN P4 LIN P3 HIN P3 LIN P2 HIN - rw rw rw rw rw rw rw - rw Field Bits Type Description PxHIN 1, 3, 9 rw Port x High Byte Input Level Selection 0 Pins Px[15-8] switch on standard TTL input levels 1 Pins Px[15-8] switch on special threshold input levels PxLIN 2, [8:4] rw Port x Low Byte Input Level Selection 0 Pins Px[7-0] switch on standard TTL input levels 1 Pins Px[7-0] switch on special threshold input levels All options for individual direction and output mode control are available for each pin independent from the selected input threshold. The input hysteresis provides stable inputs from noisy or slowly changing external signals. User’s Manual Ports_X7, V2.2 7-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Hysteresis Input Level Bit State MCT05338 Figure 7-2 7.2 Hysteresis for Special Input Thresholds Output Driver Control The output driver of a port pin is activated by switching the respective pin to output, i.e. DPx.y = ‘1’. The value that is driven to the pin is determined by the port output latch or by the associated alternate function (e.g. address, peripheral IO, etc.). The user software can control the characteristics of the output driver via the following mechanisms: • • • Open Drain Mode: The upper (push) transistor is always disabled. Only ‘0’ is driven actively, an external pull-up is required. Driver Characteristic: The driver strength can be selected. Edge Characteristic: The rise/fall time of an output signal can be selected. Open Drain Mode In the XC167 certain ports provide Open Drain Control, which allows to switch the output driver of a port pin from a push/pull configuration to an open drain configuration. In push/pull mode a port output driver has an upper and a lower transistor, thus it can actively drive the line either to a high or a low level. In open drain mode the upper transistor is always switched off, and the output driver can only actively drive the line to a low level. When writing a ‘1’ to the port latch, the lower transistor is switched off and the output enters a high-impedance state. The high level must then be provided by an external pull-up device. With this feature, it is possible to connect several port pins together to a Wired-AND configuration, saving external glue logic and/or additional software overhead for enabling/disabling output signals. This feature is controlled through the respective Open Drain Control Registers ODPx which are provided for each port that has this feature implemented. These registers allow the individual bit-wise selection of the open drain mode for each port line. If the respective control bit ODPx.y is ‘0’ (default after reset), the output driver is in the push/pull mode. If ODPx.y is ‘1’, the open drain configuration is selected. Note that all ODPx registers are located in the ESFR space. User’s Manual Ports_X7, V2.2 7-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Figure 7-3 Output Drivers in Push/Pull Mode and in Open Drain Mode Driver Characteristic This defines either the general driving capability of the respective driver, or if the driver strength is reduced after the target output level has been reached or not. Reducing the driver strength increases the output’s internal resistance which attenuates noise that is imported/exported via the output line. For driving LEDs or power transistors, however, a stable high output current may still be required. The controllable output drivers of the XC167 pins feature three differently sized transistors (strong, medium and weak) for each direction (push and pull). The time of activating/deactivating these transistors determines the output characteristics of the respective port driver. The strength of the driver can be selected to adapt the driver characteristics to the application’s requirements: In Strong Driver Mode, the medium and strong transistors are activated. In this mode the driver provides maximum output current even after the target signal level is reached. In Medium Driver Mode, only the medium transistors are activated while the other transistors remain off. In Weak Driver Mode, only the weak transistor is activated while the other transistors remain off. This results in smooth transitions with low current peaks (and reduced susceptibility for noise) on the cost of increased transition times, i.e. slower edges, depending on the capacitive load. User’s Manual Ports_X7, V2.2 7-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Edge Characteristic This defines the rise/fall time for the respective output, i.e. the output transition time. Soft edges reduce the peak currents that are drawn when changing the voltage level of an external capacitive load. For a bus interface, however, sharp edges may still be required. Edge characteristic effects the pre-driver which controls the final output driver stage. Open Drain Control Driver Control Edge Control Strong Push Medium Weak Data Signal Weak Pull Medium Strong Control Signals Driver Control Logic Driver Stage Pin MCA05339 Figure 7-4 Structure of Three-Level Output Driver with Edge Control Note: The upper (push) transistors are always off for output pins that operate in open drain mode. Figure 7-4 only shows the functional structure of the output drivers, not the real implementation. User’s Manual Ports_X7, V2.2 7-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports The Port Output Control registers POCONx provide the corresponding control bits. A 4-bit control field configures the driver strength and the edge shape. Word ports consume four control nibbles each, byte ports consume two control nibbles each, where each control nibble controls 4 pins of the respective port. Table 7-1 lists the defined POCON registers and the allocation of control bitfields and port pins. POCON* Port Output Ctrl. Reg.* 15 14 13 12 ESFR (F0xxH/yyH) 11 10 9 8 7 Reset Value: 0000H 6 5 4 3 2 1 PDM3N PDM2N PDM1N PDM0N rw rw rw rw Field Bits Type Description PDMxN [3:0], x=0 [7:4], x=1 [11:8], x=2 [15:12], x=3 rw Port Driver Mode, Nibble x Code, Driver strength1), Edge Shape2) 0000 Strong driver, Sharp edge mode 0001 Strong driver, Medium edge mode 0010 Strong driver, Soft edge mode 0011 Weak driver, Standard edge3) 0100 Medium driver, Standard edge3) 0101 Reserved, do not use! 0110 Reserved, do not use! 0111 Reserved, do not use! 1xxx Reserved, do not use! 0 1) Defines the current the respective driver can deliver to the external circuitry. 2) Defines the switching characteristics to the respective new output level. This also influences the peak currents through the driver when producing an edge, i.e. when changing the output level. 3) No additional edge shaping can be selected at this driver level. User’s Manual Ports_X7, V2.2 7-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-1 Port Output Control Register Allocation Control Register Address POCON20 Controlled Pins (by POCONx.y-z)1) Notes .15-12 .11-8 .7-4 .3-0 F0AAH / 55H P20.12, RSTOUT --- P20.5-4, ALE P20.2-0, WR, RD – POCON9 F094H / 4AH --- --- P9.5-4 P9.3-0 – POCON7 F090H / 48H --- --- P7.7-4 --- – POCON6 F08EH / 47H --- --- P6.7-4 P6.3-0 – POCON4 F08CH / 46H --- --- P4.7-4 P4.3-0 – POCON3 F08AH / 45H P3.15-12 P3.11-8 P3.7-4 P3.3-0 – POCON2 F088H / 44H P2.15-12 P2.11-8 --- --- – POCON1H F086H / 43H --- --- P1H.7-4 P1H.3-0 – POCON1L F084H / 42H --- --- P1L.7-4 P1L.3-0 – POCON0H F082H / 41H --- --- P0H.7-4 P0H.3-0 – POCON0L F080H / 40H --- --- P0L.7-4 P0L.3-0 – 1) x denotes the port number, while y-z represents the bitfield range. User’s Manual Ports_X7, V2.2 7-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.3 Alternate Port Functions In order to provide maximum flexibility for different applications and their specific IO requirements, port lines have programmable alternate input or output functions associated with them. If an alternate output function of a pin is to be used, the direction of this pin must be programmed for output (DPx.y = ‘1’), except for some signals that are used directly after reset and are configured automatically. Otherwise the pin remains in the high-impedance state and is not effected by the alternate output function. There are port lines, however, whose direction is switched automatically. For instance, in the multiplexed external bus modes of PORT0, the direction must be switched several times for an instruction fetch in order to output the addresses and to input the data. Obviously, this cannot be done through instructions. In these cases, the direction of the port line is switched automatically by hardware if the alternate function of such a pin is enabled. However, the software controlled output functions of a port are selected with one or two specific ALTSELnPx registers (n = 0 or 1). If an alternate input function is used, the direction of the pin must be programmed for input (DPx.y = ‘0’) if an external device is driving the pin. The input direction is the default after reset. Alternate inputs are supported for some peripherals and for the external interrupt inputs. Alternate inputs for a peripheral are selected with the peripheral’s PISEL register, for example the CAN_PISEL register. Alternate external interrupt inputs are selected with the registers EXISEL0 and EXISEL1 in the SCU. All port lines that are not used for these alternate functions may be used as general purpose IO lines. When using port pins for general purpose output, the initial output value should be written to the port latch prior to enabling the output drivers, in order to avoid undesired transitions on the output pins. This applies to single pins as well as to pin groups. In this case, the input operation reads the value stored in the port output latch. This can be used for testing purposes to allow a software trigger of an input function by writing to the port output latch. User’s Manual Ports_X7, V2.2 7-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.4 PORT0 PORT0 consists of two 8-bit ports P0H and P0L, representing the higher and lower byte respectively. Both halves of PORT0 can be written (e.g. via a PEC transfer) without effecting the other half. If this port is used for general purpose IO, the direction of each line can be configured via the corresponding direction registers DP0H and DP0L. P0L PORT0 Low Register 15 14 13 12 SFR (FF00H/80H) 11 10 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw P0H PORT0 High Register 15 14 13 12 9 8 Reset Value: 0000H SFR (FF02H/81H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description P0X.y 7…0 rw Port Data Register P0H or P0L Bit y User’s Manual Ports_X7, V2.2 7-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports DP0L P0L Direction Ctrl. Register 15 14 13 12 11 10 14 13 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw 12 11 10 9 8 Reset Value: 0000H 7 DP0H P0H Direction Ctrl. Register 15 ESFR (F100H/80H) ESFR (F102H/81H) 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description DP0X.y 7…0 rw Port Direction Register DP0H or DP0L Bit y 0 Port line P0X.y is an input (high-impedance) 1 Port line P0X.y is an output Alternate Functions of PORT0 In addition to GPIO functions, PORT0 is used as the address/data bus when an external bus is enabled. Figure 7-5 shows PORT0 and its alternate functions. In XC167, PORT0 pins are used as configuration pins for an external system startup (see Section 6.1.5). In this case, pull-downs and/or pull-ups may be required for the pins of PORT0. Note: To support hardware configuration with minimum external circuitry, the PORT0 pins have internal pull-up devices activated upon each reset. User’s Manual Ports_X7, V2.2 7-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Function P0H PORT0 P0L a) P0H.7 P0H.6 P0H.5 P0H.4 P0H.3 P0H.2 P0H.1 P0H.0 P0L.7 P0L.6 P0L.5 P0L.4 P0L.3 P0L.2 P0L.1 P0L.0 General Purpose Input/Output b) 16-bit DEMUX Bus d) A15 A14 A13 A12 A11 A10 A9 A8 AD7 AD6 AD5 AD4 AD3 AD2 AD1 AD0 D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 D7 D6 D5 D4 D3 D2 D1 D0 8-bit DEMUX Bus c) 8-bit MUX Bus AD15 AD14 AD13 AD12 AD11 AD10 AD9 AD8 AD7 AD6 AD5 AD4 AD3 AD2 AD1 AD0 16-bit MUX Bus MCA05340 Figure 7-5 PORT0 IO and Alternate Functions Table 7-2 lists the functions of each pin in PORT0 and also shows how they are configured. Table 7-2 PORT0 Functions Port Pin Pin Function Associated Register/ Module Alternate Function General purpose input P0L.x (x = 7-0) General purpose output P0L.x EBC inactive DP0L.Px = 0 (AltEN = 0) DP0L.Px = 1 EBC EBC active (AltEN = 1) P0H.x EBC inactive DP0H.Px = 0 (AltEN = 0) DP0H.Px = 1 EBC EBC active (AltEN = 1) Address/data bus: AD7-AD0 Data bus: D7-D0 P0H.x General purpose input (x = 7-0) General purpose output Address/data bus: AD15-AD8 Data bus: D15-D8 Control Direction controlled by EBC (AltDIR) controlled by EBC (AltDIR) Figure 7-6 shows the configuration of a PORT0 pin. User’s Manual Ports_X7, V2.2 7-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Port Output Latch Direction Latch 1 AltDIR Read Read Write Internal Bus 0 0 1 AltEN AltDataOut Pin 0 1 Driver Input Latch AltDataIn Figure 7-6 User’s Manual Ports_X7, V2.2 Clock MCA05341 P0L and P0H Port Configuration 7-12 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.5 PORT1 The two 8-bit ports P1H and P1L represent the higher and lower byte of PORT1, respectively. Both halves of PORT1 can be written (e.g. via a PEC transfer) without effecting the other half. If this port is used for general purpose IO, the direction of each line can be configured via the corresponding direction registers DP1H and DP1L. P1L PORT1 Low Register 15 14 13 12 SFR (FF04H/82H) 11 10 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rwh rw rw rw rw rw rw rw P1H PORT1 High Register 15 14 13 12 9 8 Reset Value: 0000H SFR (FF06H/83H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rwh rwh rwh rwh rw rw rw rwh Field Bits Type Description P1X.y 7…0 rw(h) Port Data Register P1H or P1L Bit y Note: Bits P1L.7, P1H.0 and P1H.4-7 are bit-protected for CAPCOM2 Output. User’s Manual Ports_X7, V2.2 7-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports DP1L P1L Direction Ctrl. Register 15 14 13 12 11 10 14 13 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw 12 11 10 9 8 Reset Value: 0000H 7 DP1H P1H Direction Ctrl. Register 15 ESFR (F104H/82H) ESFR (F106H/83H) 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description DP1X.y 7…0 rw Port Direction Register DP1H or DP1L Bit y 0 Port line P1X.y is an input (high-impedance) 1 Port line P1X.y is an output The alternate functions of the CAPCOM2 and the SSC1 are selected via the registers ALTSEL0P1L and ALTSEL0P1H. ALTSEL0P1L P1L Alternate Select Reg. 0 15 14 13 12 11 10 ESFR (F130H/98H) 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description ALTSEL0 P1L.y 7…0 rw P1L Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ALTSEL0P1H P1H Alternate Select Reg. 0 15 14 13 12 11 10 ESFR (F120H/90H) 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description ALTSEL0 P1H.y 7…0 rw P1H Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Functions of PORT1 PORT1 IO and alternate functions are shown in Figure 7-7. Alternate Function P1H PORT1 P1L a) b) c) P1H.7 P1H.6 P1H.5 A15 A14 A13 CC27IO CC26IO CC25IO P1H.4 P1H.3 P1H.2 P1H.1 P1H.0 A12 A11 A10 A9 A8 CC24IO CC6POS2 CC6POS1 CC6POS0, CC23IO P1L.7 P1L.6 P1L.5 P1L.4 P1L.3 A7 A6 A5 A4 A3 CTRAP, CC22IO COUT63 COUT62 CC62 COUT61 P1L.2 P1L.1 P1L.0 A2 A1 A0 CC61 COUT60 CC60 General Purpose Input/Output 8/16-bit DEMUX Bus SCLK1 CAPCOM Input/Output MTSR1 MRST1 SSC1 Input/Output MCA05342_X7 Figure 7-7 PORT1 IO and Alternate Functions Table 7-3 shows how the functions of each PORT1 pin can be set. Note: The compare output signals listed here are derived from the CAPCOM unit’s OUT register. If the CAPCOM unit controls the port latch directly, the output multiplexer must select general purpose output. User’s Manual Ports_X7, V2.2 7-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-3 PORT1 Functions Port Pin Pin Function Associated Register/ Module Alternate Function P1L.0 General purpose input P1L.0 DP1L.P0 = 0 EBC inactive (AltEN1 = 0) and DP1L.P0 = 1 ALTSEL0P1L. P0 = 0 Address line output A0 EBC EBC active (AltEN1 = 1) General purpose input P1L.1 EBC inactive DP1L.P1 = 0 (AltEN1 = 0) and DP1L.P1 = 1 ALTSEL0P1L. P1 = 0 Address line output A1 EBC EBC active (AltEN1 = 1) General purpose input P1L.2 DP1L.P2 = 0 EBC inactive (AltEN1 = 0) and DP1L.P2 = 1 ALTSEL0P1L. P2 = 0 Address line output A2 EBC EBC active (AltEN1 = 1) General purpose input P1L.3 DP1L.P3 = 0 EBC inactive (AltEN1 = 0) and DP1L.P3 = 1 ALTSEL0P1L. P3 = 0 Address line output A3 EBC EBC active (AltEN1 = 1) General purpose input P1L.4 DP1L.P4 = 0 EBC inactive (AltEN1 = 0) and DP1L.P4 = 1 ALTSEL0P1L. P4 = 1 EBC EBC active (AltEN1 = 1) General purpose output P1L.1 General purpose output P1L.2 General purpose output P1L.3 General purpose output P1L.4 General purpose output Address line output A4 User’s Manual Ports_X7, V2.2 7-17 Control Direction Output (AltDIR = 1) Output (AltDIR = 1) Output (AltDIR = 1) Output (AltDIR = 1) Output (AltDIR = 1) V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-3 PORT1 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P1L.5 General purpose input P1L.5 DP1L.P5 = 0 EBC inactive (AltEN1 = 0) and DP1L.P5 = 1 ALTSEL0P1L. P5 = 0 Address line output A5 EBC EBC active (AltEN1 = 1) General purpose input P1L.6 EBC inactive DP1L.P6 = 0 (AltEN1 = 0) and DP1L.P6 = 1 ALTSEL0P1L. P6 = 0 Address line output A6 EBC EBC active (AltEN1 = 1) General purpose input P1L.7 DP1L.P7 = 0 EBC inactive (AltEN1 = 0) and DP1L.P7 = 1 ALTSEL0P1L. P7 = 0 Address line output A7 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) CC22I Capture Input CAPCOM2 – DP1L.P7 = 0 General purpose output P1L.6 General purpose output P1L.7 General purpose output Output (AltDIR = 1) Output (AltDIR = 1) DP1L.P7 = 1 EBC inactive (AltEN1 = 0) and ALTSEL0P1L. P7 = 1 CC22O Compare Output User’s Manual Ports_X7, V2.2 Control Direction 7-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-3 PORT1 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P1H.0 General purpose input P1H.0 DP1H.P0 = 0 EBC inactive (AltEN1 = 0) and DP1H.P0 = 1 ALTSEL0P1H. P0 = 0 Address line output A8 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) CC23I Capture Input CAPCOM2 – DP1H.P0 = 0 General purpose output CC23O Compare Output P1H.1 Control Direction EBC inactive DP1H.P0 = 1 (AltEN1 = 0) and ALTSEL0P1H. P0 = 1 P1H.1 DP1H.P1 = 0 EBC inactive (AltEN1 = 0) and DP1H.P1 = 1 ALTSEL0P1H. P1 = 0 Address line output A9 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) SSC1 master receive input MRST1 SSC1 – DP1H.P1 = 0 General purpose input General purpose output DP1H.P1 = 1 EBC inactive (AltEN1 = 0) and ALTSEL0P1H. P1 = 1 SSC1 slave transmit output MRST1 User’s Manual Ports_X7, V2.2 7-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-3 PORT1 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P1H.2 General purpose input P1H.2 DP1H.P2 = 0 EBC inactive (AltEN1 = 0) and DP1H.P2 = 1 ALTSEL0P1H. P2 = 0 Address line output A10 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) SSC1 slave receive input MTSR1 SSC1 – DP1H.P2 = 0 General purpose output SSC1 master transmit output MTSR1 P1H.3 Control Direction EBC inactive DP1H.P2 = 1 (AltEN1 = 0) and ALTSEL0P1H. P2 = 1 P1H.3 DP1H.P3 = 0 EBC inactive (AltEN1 = 0) and DP1H.P3 = 1 ALTSEL0P1H. P3 = 0 Address line output A11 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) SSC1 clock input SCLK1 SSC1 – DP1H.P3 = 0 General purpose input General purpose output DP1H.P3 = 1 EBC inactive (AltEN1 = 0) and ALTSEL0P1H. P3 = 1 SSC1 clock output SCLK1 User’s Manual Ports_X7, V2.2 7-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-3 PORT1 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P1H.x (x = 7-4) General purpose input P1H.x DP1H.Px = 0 EBC inactive (AltEN1 = 0) and DP1H.Px = 1 ALTSEL0P1H. Px = 0 Address line output A15 to A12 EBC EBC active (AltEN1 = 1) Output (AltDIR = 1) CC27 to CC24 Capture Input CAPCOM2 – DP1H.Px = 0 General purpose output CC27 to CC24 Compare Output Control Direction EBC inactive DP1H.Px = 1 (AltEN1 = 0) and ALTSEL0P1H. Px = 1 The subsequent figures show the different configurations of a PORT1 pin. User’s Manual Ports_X7, V2.2 7-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Function Select Reg. 0 Direction Latch 1 Write Read Write Port Output Latch Read Read Write Internal Bus 0 00 10 ‘1’ X1 AltEN2 AltEN1 (EBC) 00 AltDataOut (SSC1) Pin 10 AltDataOut (EBC) X1 Driver Input Latch AltDataIn Clock MCA05343 Figure 7-8 P1L.0 to P1L.6 and P1H.1 to P1H.3 Port Configuration Table 7-4 P1L.0 to P1L.6 and P1H.1 to P1H.3 Alternate Function Control Pins Control Lines AltEN P1L.y (y = 0-6) P1H.x (x = 1-3) Registers 2 1 Alt DP1L/ ALTSEL0 DIR DP1H P1L/P1H 0 0 – 0 or 1 0 GPIO 0 – 0 1 X User’s Manual Ports_X7, V2.2 1 1 Function – 0 – SSC1 1 1 SSC1 – X EBC active: address line output 7-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 1 Port Output Latch Alternate Function Select Reg. 0 Direction Latch 1 Write Read Read Read Write CCx 1) Write Internal Bus 0 00 10 AltDIR = ‘1’ X1 AltEN2 AltEN1 (EBC) 00 AltDataOut (CAPCOM2) Pin 10 AltDataOut (EBC) X1 Driver AltDataIn (Latch) Input Latch Clock AltDataIn (Pin) 1) x = 27 - 22 MCA05344 Figure 7-9 P1L.7, P1H.0, P1H.4 to P1H.7 Port Configuration Table 7-5 P1L.7, P1H.0, P1H.4 to P1H.7 Alternate Function Control Pins Control Lines AltEN P1L.7 P1H.0 P1H.x (x = 4-7) Registers 2 1 AltDIR DP1L/ ALTSEL0 DP1H P1L/P1H 0 0 – – 0 1 X User’s Manual Ports_X7, V2.2 1 1 Function 0 or 1 0 GPIO 0 – CAPCOM2 capture input 1 1 CAPCOM2 compare output – X EBC active: address line output 7-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.6 Port 2 If this 8-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP2. Each port line can be switched into push/pull or open drain mode via the open drain control register ODP2. P2 Port 2 Data Register 15 14 12 8 P15 P14 P13 P12 P11 P10 P9 P8 - rwh rwh rwh - rwh rwh 11 rwh 10 rwh 7 6 Reset Value: 0000H 9 rwh 13 SFR (FFC0H/E0H) 5 4 Field Bits Type Description P2.y 15 … 8 rwh Port Data Register P2 Bit y 3 2 1 0 Note: Bits P2.8 - P2.15 are bit-protected for CAPCOM1 Output. DP2 P2 Direction Ctrl. Register 15 14 13 12 11 SFR (FFC2H/E1H) 10 P15 P14 P13 P12 P11 P10 rw rw rw rw rw rw 7 6 Reset Value: 0000H 9 8 5 4 3 P9 P8 - rw rw - Field Bits Type Description DP2.y 15 … 8 rw Port Direction Register DP2 Bit y 0 Port line P2.y is an input (high-impedance) 1 Port line P2.y is an output User’s Manual Ports_X7, V2.2 7-24 2 1 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP2 P2 Open Drain Ctrl. Reg. 15 14 13 12 11 ESFR (F1C2H/E1H) 10 P15 P14 P13 P12 P11 P10 rw rw rw rw rw rw 7 6 Reset Value: 0000H 9 8 5 4 3 P9 P8 - rw rw - 2 1 0 Field Bits Type Description ODP2.y 15 … 8 rw Port 2 Open Drain Control Register Bit y 0 Port line P2.y output driver in push/pull mode 1 Port line P2.y output driver in open drain mode The alternate functions of CAPCOM1 are selected via register ALTSEL0P2. ALTSEL0P2 P2 Alternate Select Reg. 0 15 14 13 12 11 ESFR (F122H/91H) 10 P15 P14 P13 P12 P11 P10 rw rw rw rw rw rw 7 6 Reset Value: 0000H 9 8 5 4 3 P9 P8 - rw rw - 2 1 0 Field Bits Type Description ALTSEL0 P2.y 15 … 8 rw P2 Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Functions of Port 2 Figure 7-10 shows the IO and alternate functions of Port 2. Alternate Function Port 2 P2.15 P2.14 P2.13 P2.12 P2.11 P2.10 P2.9 P2.8 General Purpose Input/Output a) b) CC15IO CC14IO CC13IO CC12IO CC11IO CC10IO CC9IO CC8IO CAPCOM1 Capt. Inp./ Comp. Output c) EX7IN EX6IN EX5IN EX4IN EX3IN EX2IN EX1IN EX0IN Fast External Interrupt Input T7IN CAPCOM2 Timer T7 Input MCA05345 Figure 7-10 Port 2 IO and Alternate Functions Port 2 functions are summarized in Table 7-6. Note: The compare output signals listed here are derived from the CAPCOM unit’s OUT register. If the CAPCOM unit controls the port latch directly, the output multiplexer must select general purpose output. User’s Manual Ports_X7, V2.2 7-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-6 Port 2 Functions Port Pin Pin Function Associated Register/ Module Alternate Function P2.x (x = 15-8) General purpose input P2.x ALTSEL0P2.Px DP2.Px = 0 =0 DP2.Px = 1 CAPCOM1 – General purpose output CC15I to CC8I Capture Input CC15O to CC8O Compare Output P2.15 Control Direction DP2.Px = 0 ALTSEL0P2.Px DP2.Px = 1 =1 Fast External Interrupt inputs EX7IN to EX0IN – – DP2.Px = 0 Timer 7 input T7IN CAPCOM2 – DP2.P15 = 0 The configuration of a Port 2 pins is shown in Figure 7-11. User’s Manual Ports_X7, V2.2 7-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 Port Output Latch Write Read Read Write Open Drain Latch Direction Latch 1 AltDataOut (CAPCOM1) Read Write 1 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 0 0 Pin 0 1 AltDataIn (Latch) AltDataIn (Pin) Driver Input Latch Clock EXzIN (z = 7 - 0) 1) x = 15 - 8 MCA05346 Figure 7-11 P2 Port Configuration User’s Manual Ports_X7, V2.2 7-28 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.7 Port 3 If this 15-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP3. Each port line can be switched into push/pull or open drain mode via the open drain control register ODP3 (pins P3.15 and P3.12 do not support open drain mode!). P3 Port 3 Data Register 15 14 P15 - rw - 13 12 SFR (FFC4H/E2H) 11 10 P13 P12 P11 P10 rw rw rw rw Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rw rw rw rw rw rw rw rw rw rw Field Bits Type Description P3.y 15, 13 … 0 rw Port Data Register P3 Bit y DP3 P3 Direction Ctrl. Register 15 14 P15 - rw - 13 12 11 SFR (FFC6H/E3H) 10 P13 P12 P11 P10 rw rw rw rw Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rw rw rw rw rw rw rw rw rw rw Field Bits Type Description DP3.y 15, 13 … 0 rw Port Direction Register DP3 Bit y 0 Port line P3.y is an input (high-impedance) 1 Port line P3.y is an output User’s Manual Ports_X7, V2.2 7-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP3 P3 Open Drain Ctrl. Reg. 15 14 13 12 - - P13 - - - rw - 11 ESFR (F1C6H/E3H) 10 P11 P10 rw rw Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rw rw rw rw rw rw rw rw rw rw Field Bits Type Description ODP3.y 13, 11 … 0 rw Port 3 Open Drain Control Register Bit y 0 Port line P3.y output driver in push/pull mode 1 Port line P3.y output driver in open drain mode Note: Due to pin limitations register bit P3.14 is not implemented. Pins P3.15 and P3.12 do not support open drain mode. The alternate functions of the SSC0, ASC0, ASC1 and GPT are selected via the registers ALTSEL0P3 and ALTSEL1P3. Note: For the exact selection of a peripheral output as alternate function, refer to Table 7-7. ALTSEL0P3 P3 Alternate Select Reg. 0 15 14 13 12 - - P13 - - - rw - 11 10 P11 P10 rw Field Bits ALTSEL0 P3.y rw 13, 11 … 8, 5, 3, 1, 0 User’s Manual Ports_X7, V2.2 ESFR (F126H/93H) rw Type Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 P9 P8 - - P5 - P3 - P1 P0 rw rw - - rw - rw - rw rw Description P3 Alternate Select Register 0 Bit y 0 Associated peripheral output is not selected as alternate function 1 Associated peripheral output is selected as alternate function 7-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ALTSEL1P3 P3 Alternate Select Reg. 1 ESFR (F128H/94H) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - - - - - - - - - - - - P1 - - - - - - - - - - - - - - - rw - Field Bits Type Description ALTSEL1 P3.y 1 rw P3 Alternate Select Register 1 Bit y 0 Associated peripheral output is not selected as alternate function 1 Associated peripheral output is selected as alternate function Alternate Functions of Port 3 During normal mode, Port 3 serves for various functions which include external timer control lines, the two serial interfaces, and the control lines BHE/WRH and CLKOUT/FOUT. The Port 3 IO and alternate functions are shown in Figure 7-12. Alternate Function a) b) P3.15 CLKOUT P3.13 P3.12 P3.11 P3.10 P3.9 P3.8 P3.7 P3.6 P3.5 P3.4 P3.3 P3.2 P3.1 P3.0 SCLK0 BHE RxDA0 TxDA0 MTSR0 MRST0 T2IN T3IN T4IN T3EUD T3OUT CAPIN T6OUT T0IN FOUT No Pin Port 3 General Purpose Input/Output WRH TxDA1 RxDA1 TxDA1 MCA05347 Figure 7-12 Port 3 IO and Alternate Functions User’s Manual Ports_X7, V2.2 7-31 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports The alternate output functions - TxDA1, T6OUT, T3OUT, MRST0, MTSR0, TxDA0, RxDA0 and SCLK0 - when selected, is ANDed with the port output latch line (general purpose output). A complete listing of Port 3 functions is found in Table 7-7. Table 7-7 Port 3 Functions Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P3.0 General purpose input P3.0 DP3.P0 = 0 General purpose output P3.1 DP3.P0 = 1 ASC1 transmitter output TxDA1 ASC1 ALTSEL0P3.P0 = 1 and P3.P0 = 1 DP3.P0 = 1 CAPCOM1 Timer T0 count input, T0IN CAPCOM1 – DP3.P0 = 0 General purpose input P3.1 ALTSEL0P3.P1 = 0 and ALTSEL1P3.P1 =0 DP3.P1 = 0 General purpose output DP3.P1 = 1 Timer T6 Toggle Latch output, T6OUT GPT ALTSEL0P3.P1 = 0 and ALTSEL1P3.P1 = 1 and P3.P1 = 1 DP3.P1 = 1 ASC1 receiver input RxDA1, used as input ASC1 – DP3.P1 = 0 ALTSEL0P3.P1 =1 DP3.P1 = 1 – DP3.P2 = 0 ASC1 receiver input RxDA1, used as output P3.2 ALTSEL0P3.P0 =0 General purpose input P3.2 General purpose output GPT12E Capture input CAPIN User’s Manual Ports_X7, V2.2 DP3.P2 = 1 GPT 7-32 DP3.P2 = 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-7 Port 3 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P3.3 General purpose input P3.3 DP3.P3 = 0 General purpose output P3.4 P3.5 P3.6 ALTSEL0P3.P3 =0 Timer 3 Toggle Latch output, T3OUT GPT ALTSEL0P3.P3 = 1 and P3.P3 = 1 DP3.P3 = 1 General purpose input P3.4 – DP3.P4 = 0 General purpose output DP3.P4 = 1 Timer 3 external up/down GPT input, T3EUD DP3.P4 = 0 General purpose input P3.5 General purpose output ALTSEL0P3.P5 =0 DP3.P5 = 1 Timer 4 count input, T4IN GPT – DP3.P5 = 0 ASC1 transmitter output TxDA1 ASC1 ALTSEL0P3.P5 =1 DP3.P5 = 1 General purpose input P3.6 – DP3.P6 = 0 General purpose output P3.7 DP3.P5 = 0 DP3.P6 = 1 Timer 3 count input, T3IN GPT – DP3.P6 = 0 General purpose input AltEN1.7 = 0 DP3.P7 = 0 P3.7 General purpose output P3.8 DP3.P3 = 1 DP3.P7 = 1 Timer 2 count input, T2IN GPT – DP3.P7 = 0 General purpose input ALTSEL0P3.P8 =0 DP3.P8 = 0 DP3.P8 = 1 – DP3.P8 = 0 ALTSEL0P3.P8 = 1 and P3.P8 = 1 DP3.P8 = 1 P3.8 General purpose output SSC0 master receive input, MRST0 SSC0 SSC0 slave transmit output, MRST0 User’s Manual Ports_X7, V2.2 7-33 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-7 Port 3 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P3.9 General purpose input P3.9 DP3.P9 = 0 ALTSEL0P3.P9 =0 DP3.P9 = 1 SSC0 slave receive input SSC0 MTSR0 – DP3.P9 = 0 SSC0 master transmit output, MTSR0 ALTSEL0P3.P9 = 1 and P3.P9 = 1 DP3.P9 = 1 ALTSEL0P3. P10 = 0 DP3.P10 = 0 General purpose output P3.10 General purpose input P3.10 General purpose output P3.11 ASC0 transmitter output TxDA0 ASC0 ALTSEL0P3. P10 = 1 and P3.P10 = 1 DP3.P10 = 1 General purpose input P3.11 ALTSEL0P3. P11 = 0 DP3.P11 = 0 DP3.P11 = 1 – DP3.P11 = 0 ALTSEL0P3. P11 = 1 and P3.P11 = 1 DP3.P11 = 1 Byte High and Write High are both disabled DP3.P12 = 0 General purpose output ASC0 receiver input RxDA0, used as input ASC0 ASC0 receiver input RxDA0, used as output P3.12 DP3.P10 = 1 General purpose input P3.12 General purpose output Byte High enable output BHE EBC Write High output WRH User’s Manual Ports_X7, V2.2 7-34 DP3.P12 = 1 Enabled by EBC Output (AltDIR = 1) after reset (AltEN1.12) V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-7 Port 3 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P3.13 General purpose input P3.13 DP3.P13 = 0 General purpose output SSC0 slave clock input, SCLK0 SSC0 SSC0 master clock output, SCLK0 P3.14 Not Implemented P3.15 General purpose input P3.15 General purpose output ALTSEL0P3. P13 = 0 DP3.P13 = 1 – DP3.P13 = 0 ALTSEL0P3. P13 = 1 and P3.P13 = 1 DP3.P13 = 1 Both CLKOUT and FOUT disabled DP3.P15 = 0 DP3.P15 = 1 System Clock output CLKOUT – CLKOUT enabled (AltEN1.15) Output (AltDIR = 1) Programmable Freq. output, FOUT – CLKOUT disabled and FOUT enabled (AltEN2.15) Output (AltDIR = 1) The subsequent figures show the different configurations of Port 3 pins. User’s Manual Ports_X7, V2.2 7-35 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Port Output Latch Open Drain Latch Direction Latch 1 Write Read Read Write Read Write Read Write Internal Bus Alternate Function Select Reg. 0 0 AltEN2 AltDataOut (SSC0, ASC0, ASC1, GPT) Pin 0 & 1 Driver Input Latch Clock AltDataIn MCA05348 Figure 7-13 P3.0, P3.3, P3.5, P3.8 to P3.11, and P3.13 Port Configuration P3.0, P3.31), P3.5, P3.8 to P3.11, P3.13 Alternate Function Control Table 7-8 Pins P3.0 P3.3 P3.5 P3.8-11 P3.13 Control Lines Registers AltEN AltDIR DP3L P3 – 0 or 1 0 or 1 0 GPIO 1 1 1 SSC0, ASC0, ASC1, GPT output 0 – – CAPCOM1, SSC0, ASC0 input 2 1 0 – 1 – – Function ALTSEL0 P3 1) Pin P3.3 has no alternate input function assigned. User’s Manual Ports_X7, V2.2 7-36 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Function Select Reg. 0 Write Read Write Read Open Drain Latch Direction Latch 1 Read Write Write Port Output Latch Read Read Write Internal Bus Alternate Function Select Reg. 1 0 AltEN2 AltEN3 00 & AltDataOut (GPT) Pin 10 X1 Driver AltDataOut (ASC1) Input Latch AltDataIn Clock MCA05349 Figure 7-14 P3.1 Port Configuration Table 7-9 Pins P3.1 Alternate Function Control Control Lines AltEN P3.1 Registers AltDIR DP3L P3 3 2 1 0 0 – 1 0 ALTSEL0 ALTSEL1 P3 P3 0 or 1 0 or 1 0 0 GPIO 0 1 1 0 1 GPT output 1 1 – 1 – ASC1 output User’s Manual Ports_X7, V2.2 – Function 7-37 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Open Drain Latch Direction Latch 1 Read Write Write Port Output Latch Read Read Write Internal Bus 0 Pin Driver Input Latch Clock AltDataIn MCA05350 Figure 7-15 P3.2, P3.4, P3.6 and P3.7 Port Configuration Table 7-10 P3.2, P3.4, P3.6, and P3.7 Alternate Function Control Pins Control Lines AltEN P3.2, P3.4, P3.6, P3.7 2 1 – – User’s Manual Ports_X7, V2.2 Registers AltDIR DP3L ALTSEL0 P3 – – 0 or 1 Function GPIO 0 GPT input 7-38 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Port Output Latch Read Read Write Internal Bus Direction Latch 1 0 0 AltDIR = ‘1’ 1 AltEN1 (EBC) Pin 0 AltDataOut (BHE, WRH) 1 Driver Input Latch Clock AltDataIn MCA05351 Figure 7-16 P3.12 Port Configuration Table 7-11 P3.12 Alternate Function Control Pins Control Lines AltEN P3.12 Registers AltDIR DP3L ALTSEL0 P3 – 2 1 – 0 – 0 or 1 – 1 1 – User’s Manual Ports_X7, V2.2 Function GPIO BHE/WRH output 7-39 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Port Output Latch Read Read Write Internal Bus Direction Latch 1 0 00 10 AltDIR = ‘1’ X1 AltEN2 (FOUT) AltEN1 (CLKOUT) 00 AltDataOut (FOUT) Pin 10 AltDataOut (CLKOUT) X1 Driver Input Latch Clock AltDataIn MCA05352 Figure 7-17 P3.15 Port Configuration Table 7-12 P3.15 Alternate Function Control Pins Control Lines AltEN P3.15 Registers AltDIR DP3L ALTSEL0 P3 – 2 1 0 0 – 0 or 1 x 1 1 – 1 0 User’s Manual Ports_X7, V2.2 Function GPIO CLKOUT FOUT 7-40 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.8 Port 4 If this 8-bit port is used for general purpose IO or alternate function, the direction of each line can be configured via the corresponding direction register DP4. Only if used for external bus its functions and directions are controlled by the EBC. P4 Port 4 Data Register 15 14 13 12 SFR (FFC8H/E4H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description P4.y 7…0 rw Port Data Register P4 Bit y DP4 P4 Direction Ctrl. Register 15 14 13 12 11 SFR (FFCAH/E5H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description DP4.y 7…0 rw Port Direction Register DP4 Bit y 0 Port line P4.y is an input (high-impedance) 1 Port line P4.y is an output User’s Manual Ports_X7, V2.2 7-41 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP4 P4 Open Drain Ctrl. Reg. 15 14 13 12 11 ESFR (F1CAH/E5H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Function ODP4.y 7…0 rw Port 4 Open Drain Control Register Bit y 0 Port line P4.y output driver in push/pull mode 1 Port line P4.y output driver in open drain mode Note: The alternate functions of the TwinCAN module are selected via the register ALTSEL0P4. For the exact selection of a peripheral output as alternate function, refer to Table 7-14. ALTSEL0P4 P4 Alternate Select Reg. 0 15 14 13 12 11 ESFR (F12AH/95H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 - - - - - - - rw rw - - - - - - Field Bits Type Description ALTSEL0 P4.y 7, 6 rw P4 Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-42 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports A lternate Functions of Port 4 Port 4 pins are utilized as segment address lines during external bus cycles that use segmentation (i.e. an address space above 64 Kbytes). The number of pins that is used for segment address output determines the external address space which is directly accessible. The required pins for segment address lines are configured at system startup. Note: Port 4 pins (if any) may also be used for general purpose IO or for the CAN interface. If segment address lines are selected, the alternate function of Port 4 may be necessary to access e.g. external memory directly after reset. For this reason Port 4 will be switched to this alternate function automatically. Table 7-13 summarizes the possible alternate functions of Port 4 depending on the number of selected segment address lines (coded via bitfield SALSEL and defined via SAPEN in EBCMOD0). Table 7-13 Alternate Functions of Port 4 Port 4 Std. Function Pin SALSEL = 01 64 KB Altern. Function SALSEL = 11 256 KB Altern. Function SALSEL = 00 1 MB Altern. Function SALSEL = 10 4 MB P4.0 GPIO Seg. Addr. A16 Seg. Addr. A16 Seg. Addr. A16 P4.1 GPIO Seg. Addr. A17 Seg. Addr. A17 Seg. Addr. A17 P4.2 GPIO GPIO Seg. Addr. A18 Seg. Addr. A18 P4.3 GPIO GPIO Seg. Addr. A19 Seg. Addr. A19 P4.4 GPIO/TwinCAN GPIO/TwinCAN GPIO/TwinCAN Seg. Addr. A20 P4.5 GPIO/TwinCAN GPIO/TwinCAN GPIO/TwinCAN Seg. Addr. A21 P4.6 GPIO/TwinCAN GPIO/TwinCAN GPIO/TwinCAN Seg. Addr. A22 P4.7 GPIO/TwinCAN GPIO/TwinCAN GPIO/TwinCAN Seg. Addr. A23 Note: Port 4 pins that are neither used for segment address output nor for the TwinCAN interface may be used for general purpose IO. If more than one function is selected for a Port 4 pin, the TwinCAN interface takes priority over the segment address lines. An overview of the Port 4 IO and alternate function assignment is shown in Figure 7-18. User’s Manual Ports_X7, V2.2 7-43 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Function Port 4 P4.7 P4.6 P4.5 P4.4 P4.3 P4.2 P4.1 P4.0 General Purpose Input/Output a) b) TxDCB TxDCA RxDCA RxDCB A19 A18 A17 A16 RxDCA Segment Address + Communication Alternate Routing for Communication A23 A22 A21 A20 A19 A18 A17 A16 Segment Address Output c) MCA05353_X7 Figure 7-18 Port 4 IO and Alternate Functions User’s Manual Ports_X7, V2.2 7-44 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-14 Port 4 Functions Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P4.x (x = 3-0) General purpose input P4.x DP4.Px = 0 P4.4 General purpose output EBC EBC active (AltEN1.x = 1) Output (AltDIR = 1) General purpose input P4.4 EBC inactive (AltEN1.4 = 0) DP4.P4 = 0 DP4.P4 = 1 Address line output A20 EBC EBC active (AltEN1.4 = 1) Output (AltDIR = 1) TwinCAN Receiver input RxDCB TwinCAN – DP4.P4 = 0 General purpose input P4.5 EBC inactive (AltEN1.5 = 0) DP4.P5 = 0 DP4.P5 = 1 General purpose output P4.6 DP4.Px = 1 Address line output A19 - A16 General purpose output P4.5 EBC inactive (AltEN1.x = 0) Address line output A21 EBC EBC active (AltEN1.5 = 1) Output (AltDIR = 1) TwinCAN Receiver input RxDCA TwinCAN – DP4.P5 = 0 General purpose input P4.6 EBC inactive (AltEN1.6 = 0) and ALTSEL0P4. P6 = 0 DP4.P6 = 0 General purpose output DP4.P6 = 1 Address line output A22 EBC EBC active (AltEN1.6 = 1) and ALTSEL0P4. P6 = 0 Output (AltDIR = 1) TwinCAN Transmitter output TxDCA TwinCAN ALTSEL0P4. P6 = 1 DP4.P6 = 1 User’s Manual Ports_X7, V2.2 7-45 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-14 Port 4 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P4.7 General purpose input P4.7 DP4.P7 = 0 General purpose output EBC inactive (AltEN1.7 = 0) and ALTSEL0P4. P7 = 0 DP4.P7 = 1 Address line output A23 EBC EBC active (AltEN1.7 = 1) and ALTSEL0P4. P7 = 0 Output (AltDIR = 1) TwinCAN Transmitter output, TxDCB TwinCAN ALTSEL0P4. P7=1 DP4.P7 = 1 TwinCAN Receiver input RxDCA TwinCAN – DP4.P7 = 0 The configurations of Port 4 pins are shown in the following diagrams. User’s Manual Ports_X7, V2.2 7-46 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Open Drain Latch Direction Latch 1 Read Write Write Port Output Latch Read Read Write Internal Bus 0 0 AltDIR = ‘1’ 1 AltEN1 (EBC for Addr.) Pin 0 AltDataOut (A[19:16]) 1 Driver Input Latch Clock MCA05354 Figure 7-19 P4.[3:0] Port Configuration Table 7-15 P4.[3:0] Alternate Function Control Pins Control Lines AltEN P4.[3:0] Registers AltDIR DP4L ALTSEL0 P4 – 3 2 1 – 0 0 – 0 or 1 X 1 1 – User’s Manual Ports_X7, V2.2 Function GPIO EBC: address 7-47 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Open Drain Latch Direction Latch 1 Read Write Write Port Output Latch Read Read Write Internal Bus 0 0 AltDIR = ‘1’ 1 AltEN1 (EBC) Pin 0 AltDataOut (A20, A21) 1 Driver Input Latch Clock AltDataIn MCA05355 Figure 7-20 P4.4 and P4.5 Port Configuration Table 7-16 P4.4 and P4.5 Alternate Function Control Pins Control Lines AltEN P4.4 P4.5 3 2 1 – – 0 1 User’s Manual Ports_X7, V2.2 Registers AltDIR DP4L ALTSEL0 P4 – – 1 0 Function CAN input 0 or 1 GPIO – EBC: address 7-48 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Read Open Drain Latch Direction Latch 1 Read Write Write Port Output Latch Read Read Write Internal Bus Alternate Function Select Reg. 0 0 00 AltDIR = ‘1’ 01 1X AltEN2 AltEN1 (EBC) 00 AltDataOut (EBC Address) Pin 01 AltDataOut (CAN) 1X Driver Input Latch Clock AltDataIn MCA05356_X4 Figure 7-21 P4.6 and P4.7 Port Configuration Table 7-17 P4.6 and P4.7 Alternate Function Control Pins Control Lines AltEN P4.6 P4.7 3 2 1 – 0 0 1 User’s Manual Ports_X7, V2.2 Registers AltDIR DP4L ALTSEL0 P4 – 0 0 Function CAN input (P4.7) 0 or 1 GPIO EBC: address 1 1 – X – 1 1 7-49 CAN output V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.9 Port 5 Port 5 is a 16-bit input port. There is no output latch and no direction register. Data written to P5 will be lost. P5 Port 5 Data Register 15 14 13 12 SFR (FFA2H/D1H) 11 10 P15 P14 P13 P12 P11 P10 r r r r r r Reset Value: XXXXH 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 r r r r r r r r r r Field Bits Type Description P5.y 15 … 0 r Port Data Register P5 Bit y (Read only) Alternate Functions of Port 5 Each line of Port 5 is connected to the input multiplexer of the Analog/Digital Converter. The upper 4 bits are also used as alternate input functions of the GPT. The IO and alternate functions of Port 5 are shown in Figure 7-22. Alternate Function Port 5 a) AN15 AN14 AN13 AN12 AN11 AN10 AN9 AN8 AN7 AN6 AN5 AN4 AN3 AN2 AN1 AN0 P5.15 P5.14 P5.13 P5.12 P5.11 P5.10 P5.9 P5.8 P5.7 P5.6 P5.5 P5.4 P5.3 P5.2 P5.1 P5.0 General Purpose Input b) A/D Converter Input T2EUD T4EUD T5IN T6IN T5EUD T6EUD Timer Control Input MCA05358_X7 Figure 7-22 Port 5 IO and Alternate Functions User’s Manual Ports_X7, V2.2 7-50 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Port 5 Digital Input Control Port 5 pins may be used for both digital and analog input. To use a Port 5 pin as an analog input, the Schmitt-trigger in its input stage must be disabled. This is achieved by setting the corresponding bit in the register P5DIDIS. After reset, Port 5 pins are enabled for digital inputs. P5DIDIS Port 5 Digital Inp. Disable Reg. SFR (FFA4H/D2H) 15 14 13 12 11 10 P15 P14 P13 P12 P11 P10 rw rw rw rw rw rw Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rw rw rw rw rw rw rw rw rw rw Field Bits Type Description P5DIDIS.y 15 … 0 rw Port 5 Bit y Digital Input Control 0 Digital input stage (Schmitt-trigger) is enabled. 1 Digital input stage (Schmitt-trigger) is disabled, necessary if pin is used as analog input. The functions of Port 5 pins are listed in Table 7-18. Table 7-18 Port 5 Functions Port Pin Pin Function Associated Register/ Module Alternate Function Control Direction P5.x (x = 9-0) General purpose input P5.x P5DIDIS.Px = 0 Analog input channel ANx ADC P5DIDIS.Px = 1 General purpose input P5.10 P5DIDIS.P10 = 0 Analog input channel AN10 ADC P5DIDIS.P10 = 1 P5.10 Timer 6 external up/down GPT12E input, T6EUD User’s Manual Ports_X7, V2.2 7-51 Input Only Input Only P5DIDIS.P10 = 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-18 Port 5 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function Control Direction P5.11 General purpose input P5.11 P5DIDIS.P11 = 0 Analog input channel AN11 ADC P5DIDIS.P11 = 1 P5.12 P5.13 P5.14 P5.15 Timer 5 external up/down GPT12E input, T5EUD P5DIDIS.P11 = 0 General purpose input P5.12 P5DIDIS.P12 = 0 Analog input channel AN12 ADC P5DIDIS.P12 = 1 Timer 6 count input, T6IN GPT P5DIDIS.P12 = 0 General purpose input P5.13 P5DIDIS.P13 = 0 Analog input channel AN13 ADC P5DIDIS.P13 = 1 Timer 5 count input, T5IN GPT P5DIDIS.P13 = 0 General purpose input P5.14 P5DIDIS.P14 = 0 Analog input channel AN14 ADC P5DIDIS.P14 = 1 Timer 4 external up/down GPT input, T4EUD P5DIDIS.P14 = 0 General purpose input P5.15 P5DIDIS.P15 = 0 Analog input channel AN15 ADC P5DIDIS.P15 = 1 Timer 2 external up/down GPT input, T2EUD P5DIDIS.P15 = 0 Input Only Input Only Input Only Input Only Input Only The configuration of Port 5 is shown in Figure 7-23. User’s Manual Ports_X7, V2.2 7-52 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Read Read Internal Bus P5DIDIS Input Latch Pin Clock AltDataIn AnalogInput MCA05359 Figure 7-23 P5 Port Configuration User’s Manual Ports_X7, V2.2 7-53 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.10 Port 6 If this 8-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP6. Each port line can be switched into push/pull or open drain mode via the open drain control register ODP6. P6 Port 6 Data Register 15 14 13 12 SFR (FFCCH/E6H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rwh rwh rwh rwh rwh rwh rwh rwh Field Bits Type Description P6.y 7…0 rwh Port Data Register P6 Bit y Note: Bits P6.0 - P6.7 are bit-protected for CAPCOM1 Output. DP6 P6 Direction Ctrl. Register 15 14 13 12 11 SFR (FFCEH/E7H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description DP6.y 7…0 rw Port Direction Register DP6 Bit y 0 Port line P6.y is an input (high-impedance) 1 Port line P6.y is an output User’s Manual Ports_X7, V2.2 7-54 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP6 P6 Open Drain Ctrl. Reg. 15 14 13 12 11 ESFR (F1CEH/E7H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description ODP6.y 7…0 rw Port 6 Open Drain Control Register Bit y 0 Port line P6.y output driver in push/pull mode 1 Port line P6.y output driver in open drain mode The alternate functions of CAPCOM1 are selected via the register ALTSEL0P6. ALTSEL0P6 P6 Alternate Select Reg. 0 15 14 13 12 11 ESFR (F12CH/96H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 P3 P2 P1 P0 - rw rw rw rw rw rw rw rw Field Bits Type Description ALTSEL0 P6.y 7…0 rw P6 Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-55 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Functions of Port 6 A total of 5 chip select lines (CS4 … CS0) can be selected for XC167 configuration during system startup configuration. Each potential chip select output of a Port 6 line has an internal weak pull-up device which is switched on during reset (external or single-chip). Furthermore, if the Port 6 line is configured as a chip select output with its pin driver set to push/pull mode (ODP6.x = ‘0’), then the weak pull-up device will also be switched on when the controller enters Hold mode. If the pin driver is set to open drain mode (ODP6.x = ‘1’), an external pull-up device is necessary. The above features ensure that multiple chip selection during reset is avoided, and also allows a second master to access the external memory via the same chip select lines (Wired-AND), while the controller is in Hold mode. Port 6 pins which are not configured as chip select outputs may be used for bus arbitration to accommodate additional masters. Alternatively, they may be programmed as the capture inputs or compare outputs of CAPCOM1, i.e. CC0IO - CC7IO. Alternate Function Port 6 a) P6.7 P6.6 P6.5 P6.4 P6.3 P6.2 P6.1 P6.0 General Purpose Input/Output b) c) BREQ HLDA HOLD CS4 CS3 CS2 CS1 CS0 CS4 CS3 CS2 CS1 CS0 Chip Select Output Bus Arbitration Input/Output CC7 CC6 CC5 CC4 CC3 CC2 CC1 CC0 CAPCOM1 Input/Output MCA05360 Figure 7-24 Port 6 IO and Alternate Functions The functions of Port 6 pins are summarized in Table 7-19. User’s Manual Ports_X7, V2.2 7-56 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-19 Port 6 Functions Port Pin Pin Function Associated Alternate Register/ Function Module Control Direction P6.x (x = 4-0) General purpose input P6.x Chip Select not enabled (AltEN1.x = 0) and ALTSEL0P6.Px =0 DP6.Px = 0 General purpose output DP6.Px = 1 Chip Select output CS4 to CS0 EBC Chip Select enabled (AltEN1.x = 1) Output (AltDIR = 1) CC4I to CC0I Capture Input CAPCOM1 – DP6.Px = 0 Chip Select not enabled (AltEN1.x = 0) and ALTSEL0P6.Px =1 DP6.Px = 1 CC4O to CC0O Compare Output User’s Manual Ports_X7, V2.2 7-57 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-19 Port 6 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module P6.5 General purpose input P6.5 DP6.P5 = 0 Chip Select (AltEN1.5 = 0) DP6.P5 = 1 and Bus arbitration (AltEN2 = 0) not enabled, and ALTSEL0P6.P5 =0 HOLD input EBC Chip Select not enabled (AltEN1.5 = 0) and Bus arbitration enabled (AltEN2 = 1) Input (AltDIR = 0) CC5I Capture Input CAPCOM1 – DP6.P5 = 0 General purpose output Chip Select not DP6.P5 = 1 enabled (AltEN1.5 = 0) and Bus arbitration not enabled (AltEN2 = 0) and ALTSEL0P6.P5 =1 CC5O Compare Output User’s Manual Ports_X7, V2.2 Control Direction 7-58 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-19 Port 6 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module P6.6 General purpose input P6.6 DP6.P6 = 0 Chip Select (AltEN1.6 = 0) DP6.P6 = 1 and Bus arbitration (AltEN2 = 0) not enabled, and ALTSEL0P6.P6 =0 HLDA output (master mode) EBC HLDA input (slave mode) EBC Chip Select not enabled (AltEN1.6 = 0) and Bus arbitration enabled (AltEN2 = 1) Input (AltDIR = 0) CC6I Capture Input CAPCOM1 – DP6.P6 = 0 General purpose output Output (AltDIR = 1) Chip Select not DP6.P6 = 1 enabled (AltEN1.6 = 0) and Bus arbitration not enabled (AltEN2 = 0) and ALTSEL0P6.P6 =1 CC6O Compare Output User’s Manual Ports_X7, V2.2 Control Direction 7-59 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-19 Port 6 Functions (cont’d) Port Pin Pin Function Associated Alternate Register/ Function Module P6.7 General purpose input P6.7 DP6.P7 = 0 Bus arbitration (AltEN2 = 0) not DP6.P7 = 1 enabled, and ALTSEL0P6.P7 =0 BREQ output EBC Bus arbitration enabled (AltEN2 = 1) Output (AltDIR = 1) CC7I Capture Input CAPCOM1 – DP6.P7 = 0 General purpose output CC7O Compare Output Control Direction Bus arbitration DP6.P7 = 1 not enabled (AltEN2 = 0) and ALTSEL0P6.P7 =1 The configuration of Port 6 pins are shown in the subsequent figures. User’s Manual Ports_X7, V2.2 7-60 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 1 Port Output Latch Open Drain Latch Direction Latch 1 Write Read Read Write Read Read Write CCx 1) Write Internal Bus Alternate Function Select Reg. 0 0 00 10 AltDIR = ‘1’ X1 AltEN3 AltEN1 (EBC) 00 AltDataOut (CAPCOM1) Pin 10 AltDataOut (EBC) X1 Driver Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 4 - 0 MCA05361 Figure 7-25 P6.[4:0] Port Configuration Table 7-20 P6.[4:0] Alternate Function Control Pins Control Lines AltEN P6.[4:0] Registers AltDIR DP6L ALTSEL0 P6 Function 3 2 1 0 – 0 – 0 or 1 0 GPIO X 1 1 – X EBC: chip select – 0 – 0 – CAPCOM1 input 1 1 CAPCOM1 output 1 User’s Manual Ports_X7, V2.2 7-61 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 1 Port Output Latch Open Drain Latch Direction Latch 1 Write Read Read Write Read Read Write CCx 1) Write Internal Bus Alternate Function Select Reg. 0 0 000 100 AltDIR = ‘0’ X10 AltDIR = ‘1’ XX1 AltEN2 (EBC for Bus Arb.) AltEN3 AltEN1 (EBC for CS) 000 AltDataOut (CAPCOM1) Pin 100 Don’t Care AltDataOut (EBC for CS) X10 Driver XX1 Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 5 MCA05362 Figure 7-26 P6.5 Port Configuration Table 7-21 P6.5 Alternate Function Control Pins Control Lines AltEN P6.5 Registers AltDIR DP6L ALTSEL0 P6 Function 3 2 1 0 0 0 – 0 or 1 0 GPIO X 1 0 0 – X EBC: bus arbitration – 0 0 – 0 – CAPCOM1 input 1 1 CAPCOM1 output 1 User’s Manual Ports_X7, V2.2 7-62 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 Port Output Latch Open Drain Latch Direction Latch 1 Write Read Read Write Read Write 1 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 0 0 000 100 AltDIR 2) X10 AltDIR = ‘1’ XX1 AltEN2 (EBC for Bus Arb.) AltEN3 AltEN1 (EBC for CS) 000 AltDataOut (CAPCOM1) Pin 100 AltDataOut (EBC for Bus Arb.) X10 AltDataOut (EBC for CS) Driver XX1 Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 6 2) AltDIR = ‘0’, if Slave Mode AltDIR = ‘1’, if Master Mode MCA05363 Figure 7-27 P6.6 Port Configuration Table 7-22 P6.6 Alternate Function Control Pins Control Lines AltEN P6.6 Registers AltDIR DP6L ALTSEL0 P6 Function 3 2 1 0 0 0 – 0 or 1 0 GPIO X 1 0 0 or 1 – X EBC: bus arbitration – 0 0 – 0 – CAPCOM1 input 1 1 CAPCOM1 output 1 User’s Manual Ports_X7, V2.2 7-63 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 1 Port Output Latch Open Drain Latch Direction Latch 1 Write Read Read Write Read Read Write CCx 1) Write Internal Bus Alternate Function Select Reg. 0 0 00 10 AltDIR = ‘1’ X1 AltEN3 AltEN2 (EBC for Bus Arb.) 00 AltDataOut (CAPCOM1) Pin 10 AltDataOut (EBC for Bus Arb.) X1 Driver Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 7 MCA05364 Figure 7-28 P6.7 Port Configuration Table 7-23 P6.7 Alternate Function Control Pins Control Lines AltEN P6.7 Registers Function AltDIR DP6L ALTSEL0 P6 – 0 or 1 0 GPIO 3 2 1 0 0 – X 1 1 – X EBC: bus arbitration – 0 – 0 – CAPCOM1 input 1 1 CAPCOM1 output 1 User’s Manual Ports_X7, V2.2 7-64 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.11 Port 7 If this 4-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP7. Each port line can be switched into push/pull or open drain mode via the open drain control register ODP7. P7 Port 7 Data Register 15 14 13 12 SFR (FFD0H/E8H) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 - - - - - rwh rwh rwh rwh - - - - Field Bits Type Description P7.y 7…4 rwh Port Data Register P7 Bit y Note: Bits P7.4 - P7.7 are bit-protected for CAPCOM2 Output. DP7 P7 Direction Ctrl. Register 15 14 13 12 11 SFR (FFD2H/E9H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 - - - - - rw rw rw rw - - - - Field Bits Type Description DP7.y 7…4 rw Port Direction Register DP7 Bit y 0 Port line P7.y is an input (high-impedance) 1 Port line P7.y is an output User’s Manual Ports_X7, V2.2 7-65 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP7 P7 Open Drain Ctrl. Reg. 15 14 13 12 11 ESFR (F1D2H/E9H) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 - - - - - rw rw rw rw - - - - Field Bits Type Description ODP7.y 7…4 rw Port 7 Open Drain Control Register Bit y 0 Port line P7.y output driver in push/pull mode 1 Port line P7.y output driver in open drain mode The alternate functions of the TwinCAN and CAPCOM2 modules are selected via the registers ALTSEL0P7 and ALTSEL1P7. ALTSEL0P7 P7 Alternate Select Reg. 0 15 14 13 12 11 ESFR (F13CH/9EH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 - - - - - rw rw rw rw - - - - Field Bits Type Description ALTSEL0 P7.y 7…4 rw P7 Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-66 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ALTSEL1P7 P7 Alternate Select Reg. 1 15 14 13 12 11 ESFR (F13EH/9FH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - P7 P6 P5 P4 - - - - - rw rw rw rw - - - - Field Bits Type Description ALTSEL1 P7.y 7…4 rw P7 Alternate Select Register 1 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function Alternate Functions of Port 7 Port 7 pins are used as receive data and transmit data lines for TwinCAN and SDLM interface. Alternatively, they can be used as capture inputs or compare outputs of the CAPCOM2. Its IO and alternate functions are shown in Figure 7-29. Alternate Function Port 7 P7.7 P7.6 P7.5 P7.4 General Purpose Input/Output a) b) CC31IO CC30IO CC29IO CC28IO TxDCA RxDCA TxDCB RxDCB CAPCOM2 Input/Output CAN Input/Output MCA05365_X32 Figure 7-29 Port 7 IO and Alternate Functions User’s Manual Ports_X7, V2.2 7-67 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports The functions of Port 7 pins are summarized in Table 7-24. Table 7-24 Port 7 Functions Port Pin Pin Function Associated Register/ Module Alternate Function P7.4 General purpose input P7.4 ALTSEL0P7.P4 DP7.P4 = 0 = 0 and DP7.P4 = 1 ALTSEL1P7.P4 =0 CAPCOM2 – General purpose output CC28I Capture Input P7.5 Control Direction DP7.P4 = 0 CC28O Compare Output ALTSEL0P7.P4 DP7.P4 = 1 = 0 and ALTSEL1P7.P4 =1 TwinCAN Receiver input, TwinCAN RxDCB – General purpose input P7.5 ALTSEL0P7.P5 DP7.P5 = 0 = 0 and DP7.P5 = 1 ALTSEL1P7.P5 =0 CAPCOM2 – General purpose output CC29I Capture Input User’s Manual Ports_X7, V2.2 DP7.P5 = 0 ALTSEL0P7.P5 DP7.P5 = 1 = 0 and ALTSEL1P7.P5 =1 CC29O Compare Output TwinCAN Transmitter output, TxDCB DP7.P4 = 0 TwinCAN 7-68 ALTSEL0P7.P5 DP7.P5 = 1 =1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-24 Port 7 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P7.6 General purpose input P7.6 ALTSEL0P7.P6 DP7.P6 = 0 = 0 and DP7.P6 = 1 ALTSEL1P7.P6 =0 CAPCOM2 – General purpose output CC30I Capture Input P7.7 Control Direction DP7.P6 = 0 CC30O Compare Output ALTSEL0P7.P6 DP7.P6 = 1 = 0 and ALTSEL1P7.P6 =1 TwinCAN Receiver input, TwinCAN RxDCA – General purpose input P7.7 ALTSEL0P7.P7 DP7.P7 = 0 = 0 and DP7.P7 = 1 ALTSEL1P7.P7 =0 CAPCOM2 – General purpose output CC31I Capture Input DP7.P7 = 0 ALTSEL0P7.P7 DP7.P7 = 1 = 0 and ALTSEL1P7.P7 =1 CC31O Compare Output TwinCAN Transmitter output, TxDCA DP7.P6 = 0 TwinCAN ALTSEL0P7.P7 DP7.P7 = 1 =1 The configuration of Port 7 pins are shown in the following figures. User’s Manual Ports_X7, V2.2 7-69 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 Port Output Latch Open Drain Latch Direction Latch 1 Alternate Function Select Reg. 0 Write Read Write Read Read Write Read Write 1 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 1 0 AltEN1 AltEN2 00 AltDataOut (CAPCOM2) Pin 10 ‘0’ 2) X1 AltDataIn (Latch) AltDataIn (Pin) Driver Input Latch Clock 1) x = 28 2) This option is prohibited by spec MCA05366 Figure 7-30 P7.4 Port Configuration User’s Manual Ports_X7, V2.2 7-70 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 0 Port Output Latch Open Drain Latch Direction Latch 1 Alternate Function Select Reg. 0 Write Read Write Read Read Write Read Write 1 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 1 0 AltEN1 AltEN2 AltDataOut (CAPCOM2) AltDataOut (CAN) 00 Pin 10 X1 Driver Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 31 - 29 MCA05367_X32 Figure 7-31 P7.[7:5] Port Configuration User’s Manual Ports_X7, V2.2 7-71 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.12 Port 9 If this 6-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP9. Each port line can be switched into push/pull or open drain mode via the open drain control register ODP9. P9 Port 9 Data Register 15 14 13 12 SFR (FF16H/8BH) 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - - - P5 P4 P3 P2 P1 P0 - - - rwh rwh rwh rwh rwh rwh Field Bits Type Description P9.y 5…0 rwh Port Data Register P9 Bit y Note: Bits P9.0 - P9.5 are bit-protected for CAPCOM2 Output. DP9 P9 Direction Ctrl. Register 15 14 13 12 11 SFR (FF18H/8CH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - - - P5 P4 P3 P2 P1 P0 - - - rw rw rw rw rw rw Field Bits Type Description DP9.y 5…0 rw Port Direction Register DP9 Bit y 0 Port line P9.y is an input (high-impedance) 1 Port line P9.y is an output User’s Manual Ports_X7, V2.2 7-72 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ODP9 P9 Open Drain Ctrl. Reg. 15 14 13 12 11 SFR (FF1AH/8DH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - - - P5 P4 P3 P2 P1 P0 - - - rw rw rw rw rw rw Field Bits Type Description ODP9.y 5…0 rw Port 9 Open Drain Control Register Bit y 0 Port line P9.y output driver in push/pull mode 1 Port line P9.y output driver in open drain mode The alternate functions of the IIC, TwinCAN and CAPCOM2 modules are selected via the register ALTSEL0P9 and ALTSEL1P9. ALTSEL0P9 P9 Alternate Select Reg. 0 15 14 13 12 11 ESFR (F138H/9CH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - - - P5 P4 P3 P2 P1 P0 - - - rw rw rw rw rw rw Field Bits Type Description ALTSEL0 P9.y 5…0 rw P9 Alternate Select Register 0 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-73 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports ALTSEL1P9 P9 Alternate Select Reg. 1 15 14 13 12 11 ESFR (F13AH/9DH) 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 - - - P5 P4 P3 P2 P1 P0 - - - rw rw rw rw rw rw Field Bits Type Description ALTSEL1 P9.y 5…0 rw P9 Alternate Select Register 1 Bit y 0 associated peripheral output is not selected as alternate function 1 associated peripheral output is selected as alternate function User’s Manual Ports_X7, V2.2 7-74 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Functions of Port 9 Port 9 pins can be used as the serial data and clock lines of the IIC interface, the transmitter and receiver lines of the TwinCAN or alternatively, the CAPCOM2 input/output lines. If IIC interface is configured, it is necessary to switch the respective pins to open drain mode (ODP9.y = ‘1’). Figure 7-32 shows the IO and alternate functions of Port 9. Alternate Function a) b) c) Port 9 P9.5 P9.4 P9.3 P9.2 P9.1 P9.0 General Purpose Input/Output SCL2 SDA2 SCL1 SDA1 SCL0 SDA0 IIC Input/Output TxDCA RxDCA TxDCB RxDCB CAN Input/Output CC21 CC20 CC19 CC18 CC17 CC16 CAPCOM2 Input/Output MCA05368_X32 Figure 7-32 Port 9 IO and Alternate Functions The functions of Port 9 pins are listed in Table 7-25. User’s Manual Ports_X7, V2.2 7-75 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-25 Port 9 Functions Port Pin Pin Function Associated Register/ Module Alternate Function P9.0 General purpose input P9.0 ALTSEL0P9.P0 DP9.P0 = 0 = 0 and DP9.P0 = 1 ALTSEL1P9.P0 =0 CAPCOM2 – General purpose output CC16I Capture input P9.1 Control Direction DP9.P0 = 0 CC16O Compare output ALTSEL0P9.P0 DP9.P0 = 1 = 0 and ALTSEL1P9.P0 =1 TwinCAN Receiver input, TwinCAN RxDCB – DP9.P0 = 0 IIC serial data line 0 input, IIC SDA0 – DP9.P0 = 0 IIC serial data line 0 output, SDA0 ALTSEL0P9.P0 DP9.P0 = 1 = 1 and ALTSEL1P9.P0 =X General purpose input P9.1 ALTSEL0P9.P1 DP9.P1 = 0 = 0, and DP9.P1 = 1 ALTSEL1P9.P1 =0 CAPCOM2 – General purpose output CC17I Capture input ALTSEL0P9.P1 DP9.P1 = 1 = 0, and ALTSEL1P9.P1 =1 CC17O Compare output TwinCAN Transmitter output, TxDCB User’s Manual Ports_X7, V2.2 DP9.P1 = 0 TwinCAN 7-76 ALTSEL0P9.P1 DP9.P1 = 1 = 1, and ALTSEL1P9.P1 =1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-25 Port 9 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function Control Direction P9.1 IIC serial clock line 0 input, SCL0 IIC – DP9.P1 = 0 ALTSEL0P9.P1 DP9.P1 = 1 = 1 and ALTSEL1P9.P1 =0 IIC serial clock line 0 output, SCL0 P9.2 General purpose input P9.2 ALTSEL0P9.P2 DP9.P2 = 0 = 0 and DP9.P2 = 1 ALTSEL1P9.P2 =0 CAPCOM2 – General purpose output CC18I Capture input DP9.P2 = 0 CC18O Compare output ALTSEL0P9.P2 DP9.P2 = 1 = 0 and ALTSEL1P9.P2 =1 TwinCAN Receiver input, TwinCAN RxDCA – DP9.P2 = 0 IIC serial data line 1 input, IIC SDA1 – DP9.P2 = 0 IIC serial clock line 1 output, SDA1 ALTSEL0P9.P2 DP9.P2 = 1 = 1 and ALTSEL1P9.P2 =0 User’s Manual Ports_X7, V2.2 7-77 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-25 Port 9 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function P9.3 General purpose input P9.3 ALTSEL0P9.P3 DP9.P3 = 0 = 0 and DP9.P3 = 1 ALTSEL1P9.P3 =0 CAPCOM2 – General purpose output CC19I Capture input CC19O Compare output DP9.P3 = 0 ALTSEL0P9.P3 DP9.P3 = 1 = 0 and ALTSEL1P9.P3 =1 TwinCAN Transmitter output, TxDCA TwinCAN ALTSEL0P9.P3 DP9.P3 = 1 = 1 and ALTSEL1P9.P3 =1 IIC serial clock line 1 input, SCL1 IIC – General purpose input P9.4 ALTSEL0P9.P4 DP9.P4 = 0 = 0 and DP9.P4 = 1 ALTSEL1P9.P4 =0 CAPCOM2 – General purpose output CC20I Capture input DP9.P4 = 0 ALTSEL0P9.P4 DP9.P4 = 1 = 0 and ALTSEL1P9.P4 =1 CC20O Compare output User’s Manual Ports_X7, V2.2 DP9.P3 = 0 ALTSEL0P9.P3 DP9.P3 = 1 = 1 and ALTSEL1P9.P3 =0 IIC serial clock line 1 output, SCL1 P9.4 Control Direction 7-78 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-25 Port 9 Functions (cont’d) Alternate Function Control Direction IIC serial data line 2 input IIC SDA2 – DP9.P4 = 0 IIC serial data line 2 output, SDA2 ALTSEL0P9.P4 DP9.P4 = 1 = 1 and ALTSEL1P9.P4 =X Port Pin Pin Function P9.4 P9.5 General purpose input Associated Register/ Module P9.5 ALTSEL0P9.P5 DP9.P5 = 0 = 0 and DP9.P5 = 1 ALTSEL1P9.P5 =0 CAPCOM2 – General purpose output CC21I Capture input ALTSEL0P9.P5 DP9.P5 = 1 = 0 and ALTSEL1P9.P5 =1 CC21O Compare output IIC serial clock line 2 input, SCL1 DP9.P5 = 0 IIC – DP9.P5 = 0 ALTSEL0P9.P5 DP9.P5 = 1 = 1 and ALTSEL1P9.P5 =X IIC serial clock line 2 output, SCL1 The configuration of Port 9 pins is shown in the subsequent figures. User’s Manual Ports_X7, V2.2 7-79 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 1 Port Output Latch Open Drain Latch Direction Latch 0 Alternate Function Select Reg. 0 Write Read Write Read Read Write Read Write 0 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 1 1 AltEN1 AltEN2 00 AltDataOut (IIC) Pin 01 AltDataOut (CAPCOM) 10 Driver 11 Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 16, 20, 21 MCA05369 Figure 7-33 P9.0, P9.4 and P9.5 Port Configuration Table 7-26 Pins P9.0, P9.2, P9.4 and P9.5 Alternate Function Control Control Lines AltEN P9.0, P9.4, P9.5 2 1 0 0 – Registers DP9 Function ALTSEL1 ALTSEL0 P9 P9 0 or 1 0 0 – 0 GPIO IIC, CAN, CAPCOM2 input X 1 1 X 1 IIC output 1 0 1 1 0 CAPCOM2 compare output User’s Manual Ports_X7, V2.2 7-80 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 1 Port Output Latch Open Drain Latch Direction Latch 0 Alternate Function Select Reg. 0 Write Read Write Read Read Write Read Write 0 Read CCx 1) Write Internal Bus Alternate Function Select Reg. 1 1 AltEN1 AltEN2 00 (IIC) Pin 01 AltDataOut (CAPCOM) 10 AltDataOut (CAN) Driver 11 Input Latch AltDataIn (Latch) Clock AltDataIn (Pin) 1) x = 17, 18, 19 MCA05370_X7 Figure 7-34 P9.1, P9.2 and P9.3 Port Configuration Table 7-27 Pins P9.1, P9.2 and P9.3 Alternate Function Control Control Lines AltEN P9.1, P9.2, P9.3 2 1 0 0 – Registers DP9 Function ALTSEL1 ALTSEL0 P9 P9 0 or 1 0 0 – 0 GPIO IIC, CAN, CAPCOM2 input 0 1 1 0 1 IIC output 1 1 0 1 1 CAN output 1 0 1 1 0 CAPCOM2 compare output User’s Manual Ports_X7, V2.2 7-81 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports 7.13 Port 20 If this 6-bit port is used for general purpose IO, the direction of each line can be configured via the corresponding direction register DP20. Port 20 adds general purpose IO functionality to a set of previously dedicated pins. P20 Port 20 Data Register SFR (FFB4H/DAH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - P12 - - - - - - P5 P4 - P2 P1 P0 - - - rw - - - - - - rw rw - rw rw rw Field Bits Type P20.y 12, 5, 4, rw 2…0 DP20 P20 Direction Ctrl. Register Description Port Data Register P20 Bit y SFR (FFB6H/DBH) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 - - - P12 - - - - - - P5 P4 - P2 P1 P0 - - - rw - - - - - - rw rw - rw rw rw Field Bits DP20.y 12, 5, 4, rw 2…0 User’s Manual Ports_X7, V2.2 Type Description Port Direction Register DP20 Bit y 0 Port line P20.y is an input (high-impedance) 1 Port line P20.y is an output 7-82 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Alternate Functions of Port 20 Figure 7-35 shows the IO and alternate functions of Port 20. The pins EA, ALE, RD and WR are used as configuration pins during system startup. They are shown as CFG_EA, CFG_ALE, CFG_RD and CFG_WR inputs in Figure 7-35. Alternate Function a) b) P20.12 RSTOUT P20.5 P20.4 ALE CFG_EA CFG_ALE P20.2 P20.1 P20.0 READY WR RD CFG_WR CFG_RD Port 20 General Purpose Input/Output During Operation Startup Configuration MCA05371 Figure 7-35 Port 20 IO and Alternate Functions During normal operation, Port 20 pins (P20.0, P20.1, P20.2, P20.4) which are not used for external bus function can be released for general purpose IO. Separate control bits (in the EBC) are available for each pin, thus partial release for general purpose IO function is possible even if an external bus is used. P20.5 is always available for general purpose IO during normal mode, its alternate function (EA) is only required during startup configuration. P20.12 is available if a singlechip reset without external bus system is selected at startup. For details concerning the Port 20 pins, please refer to Chapter 8. The functions of Port 20 pins are listed in Table 7-28. User’s Manual Ports_X7, V2.2 7-83 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-28 Port 20 Functions Port Pin Pin Function Associated Register/ Module Alternate Function Control Direction P20.0 General purpose input P20.0 EBC pins disabled (AltEN1.0 = 0) DP20.P0 = 0 General purpose output P20.1 Read command signal RD EBC pins enabled (AltEN1.0 = 1) Output (AltDIR = 1) Config. read input CFG_RD – Input (AltDIR = 0) EBC pins disabled (AltEN1.1 = 0) DP20.P1 = 0 General purpose input P20.1 General purpose output P20.2 EBC pins enabled (AltEN1.1 = 1) Output (AltDIR = 1) Config. write input CFG_WR – Input (AltDIR = 0) READY pin disabled (AltEN1.2 = 0) DP20.P2 = 0 General purpose input P20.2 Bus termination input signal, READY General purpose input P20.4 General purpose output P20.5 DP20.P1 = 1 Write command signal WR/WRL General purpose output P20.4 DP20.P0 = 1 DP20.P2 = 1 READY pin enabled (AltEN1.2 = 1) Input (AltDIR = 0) ALE pin disabled (AltEN1.4 = 0) DP20.P4 = 0 DP20.P4 = 1 Address latch enable signal, ALE ALE pin enabled (AltEN1.4 = 1) Output (AltDIR = 1) Config. ALE input CFG_ALE – Input (AltDIR = 0) AltEN1.5 = 0 DP20.P5 = 0 General purpose input P20.5 General purpose output DP20.P5 = 1 Config. EA input CFG_EA User’s Manual Ports_X7, V2.2 – 7-84 Input (AltDIR = 0) V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 7-28 Port 20 Functions (cont’d) Port Pin Pin Function Associated Register/ Module Alternate Function Control Direction P20.12 General purpose input P20.12 RSTOUT pin disabled (AltEN1.12 = 0) DP20.P12 = 0 General purpose output RSTOUT pin enabled (AltEN1.12 = 1) Reset indication output, RSTOUT DP20.P12 = 1 Output (AltDIR = 1) The configuration of Port 20 pins is shown in the subsequent figure. User’s Manual Ports_X7, V2.2 7-85 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Parallel Ports Write Port Output Latch Direction Latch 1 AltDIR Read Read Write Internal Bus 0 0 1 AltEN1 AltDataOut Pin 0 1 Driver Input Latch AltDataIn Clock MCA05372 Figure 7-36 P20 Port Configuration Note: For P20.5: AltEN1 = 0, for others: refer to Table 7-28. For P20.0, P20.1, P20.4, P20.12: AltDIR = 1, for P20.2, P20.5: AltDIR = 0. User’s Manual Ports_X7, V2.2 7-86 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Dedicated Pins 8 Dedicated Pins Most of the input/output or control signals of the functional the XC167 are realized as alternate functions of pins of the parallel ports. There is, however, a number of signals that use separate pins, including the oscillator, special control signals and, of course, the power supply. Table 8-1 summarizes the 41 dedicated pins of the XC167. Table 8-1 XC167 Dedicated Pins Pin(s) Function NMI Non-Maskable Interrupt Input XTAL1, XTAL2 Oscillator Input/Output (main oscillator) XTAL3, XTAL4 Oscillator Input/Output (auxiliary oscillator) RSTIN Reset Input TRST Test Reset Input for the Debug System TMS, TCK, TDI, TDO JTAG Interface used for On-Chip Debugging BRKIN, BRKOUT Break Interface for Debugging VAREF, VAGND VDDI VDDP VSS Power Supply for Analog/Digital Converter NC Not Connected Pins (6 pins) Digital Power Supply for Internal Logic (3 pins) Digital Power Supply for Port Drivers (7 pins) Digital Reference Ground (10 pins) The Non-Maskable Interrupt Input NMI allows to trigger a high priority trap via an external signal (e.g. a power-fail signal). It also serves to validate the PWRDN instruction that switches the XC167 into Power-Down mode. The NMI pin is sampled with every system clock cycle to detect transitions. The Oscillator Input XTAL1 and Output XTAL2 connect the internal Main Oscillator to the external crystal. The oscillator provides an inverter and a feedback element. The standard external oscillator circuitry (see Section 6.2.1) comprises the crystal, two low end capacitors and series resistor to limit the current through the crystal. The main oscillator is intended for the generation of the basic operating clock signal of the XC167. An external clock signal may be fed to the input XTAL1, leaving XTAL2 open or terminating it for higher input frequencies. The Oscillator Input XTAL3 and Output XTAL4 connect the internal Auxiliary Oscillator to the external crystal. The oscillator provides an inverter and a User’s Manual DediPins_X7, V2.0 8-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Dedicated Pins feedback element. The standard external oscillator circuitry (see Section 6.2.1) comprises the crystal, two low end capacitors and series resistor to limit the current through the crystal. The auxiliary oscillator is intended for the generation of a power saving backup clock signal for the XC167’s real time clock, especially during power saving modes. An external clock signal may be fed to the input XTAL3, leaving XTAL4 open. Note: In order to reduce its power consumption as much as possible the operating range of the auxiliary oscillator is optimized around 32 kHz (10 … 50 kHz when driven externally). The Reset Input RSTIN allows to put the XC167 into the well defined reset condition either at power-up or external events like a hardware failure or manual reset. The Test Reset Input TRST puts the XC167’s debug system into reset state. During normal operation this input should be held active. For debugging purposes the on-chip debugging system can be enabled by releasing pin TRST. The JTAG Interface Pins TMS, TCK, TDI, and TDO form the standard debugging interface used for the on-chip debug system, and also for device testing. Pins TDI and TDO input and output the serial data stream clocked by TCK. TMS provides mode control. The Break Interface Pins BRKIN and BRKOUT support on-chip debugging. Pin BRKIN accepts an external trigger to intermediately suspend the operation of the XC167. Pin BRKOUT can indicate the occurrence of a breakpoint. This can automatically stop other connected circuitry or can be used as a monitor signal. The Power Supply pins for the Analog/Digital Converter VAREF and VAGND provide a separate power supply (reference voltage) for the on-chip ADC. This reduces the noise that is coupled to the analog input signals from the digital logic sections and so improves the stability of the conversion results, when VAREF and VAGND are properly discoupled from VDD and VSS. The Power Supply pins VDDI/VDDP and VSS provide the power supply for the digital logic of the XC167. The respective VDD/VSS pairs should be decoupled as close to the pins as possible. The VDDI pins (2.5 V) supply the internal logic blocks of the XC167, while the VDDP pins (5.0 V) supply the output port drivers. Note: All VDD pins and all VSS pins must be connected to the power supplies and ground, respectively. The Not Connected pins (NC) are spare pins without a function. However, it is recommended to leave them unconnected on the p.c.-board to provide upward compatibility with further products that may have functions assigned to these pins. Table 8-2 summarizes 6 pins of Port 20 which are used for specific functions either during reset configuration or while the external bus interface is active. User’s Manual DediPins_X7, V2.0 8-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Dedicated Pins Table 8-2 XC167 Special Port 20 Pins Pin(s) Function ALE Address Latch Enable RD External Read Strobe WR/WRL External Write/Write Low Strobe READY Ready Input EA External Access Enable RSTOUT Reset Output The Address Latch Enable signal ALE controls external address latches that provide a stable address in multiplexed bus modes. ALE is activated for every external bus cycle independent of the selected bus mode, i.e. it is also activated for bus cycles with a demultiplexed address bus. ALE is not activated for internal accesses, i.e. accesses to ROM/OTP/Flash (if provided), the internal RAMs and the special function registers. During reset an internal pull-down ensures an inactive (low) level on the ALE output. At the end of a true single-chip mode reset (EA = 1) the current level on pin ALE is latched and is used for configuration. Pin ALE selects standard start/boot, when driven low (default) or alternate start/boot when driven high. For standard configuration pin ALE should be low or not connected. The External Read Strobe RD controls the output drivers of external memory or peripherals when the XC167 reads data from these external devices. During accesses to on-chip LXBus-Peripherals RD remains inactive (high). During reset an internal pull-up ensures an inactive (high) level on the RD output. At the end of reset the current level on pin RD is latched and is used for configuration. For a reset with external access (EA = 0) pin RD controls the oscillator watchdog. The default high level on pin RD leaves the oscillator watchdog active, while a low level disables the watchdog e.g. for testing purposes. For a true single-chip mode reset (EA = 1) pin RD enables the bootstrap loader, when driven low (pin ALE is evaluated together with pin RD). For standard configuration pin RD should be high or not connected. User’s Manual DediPins_X7, V2.0 8-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Dedicated Pins The External Write Strobe WR/WRL controls the data transfer from the XC167 to an external memory or peripheral device. This pin may either provide an general WR signal activated for both byte and word write accesses, or specifically control the low byte of an external 16-bit device (WRL) together with the signal WRH (alternate function of BHE). During accesses to on-chip LXBus-Peripherals WR/WRL remains inactive (high). During reset an internal pull-up ensures an inactive (high) level on the WR/WRL output. At the end of reset the current level on pin WR is latched and is used for configuration. For a true single-chip mode reset (EA = 1) pin WR enables the Port 20 IO mode, when driven low. The Ready Input READY receives a control signal from an external memory or peripheral device that is used to terminate an external bus cycle, provided that this function is enabled for the current bus cycle. READY may be used as synchronous READY or may be evaluated asynchronously. When waitstates are defined for a READY controlled address window the READY input is not evaluated during these waitstates. The polarity of signal READY is programmable. The External Access Enable Pin EA determines if the XC167 after reset starts fetching code from the on-chip program memory (EA = 1) or via the external bus interface (EA = 0). Be sure to hold this input low for ROMless devices. At the end of the internal reset sequence the EA signal is latched together with the configuration (PORT0, RD, WR, ALE). Note: The reset configuration is described in Section 6.1.4. The Reset Output RSTOUT provides a special reset signal for external circuitry. RSTOUT is activated at the beginning of the reset sequence, triggered via RSTIN, a watchdog timer overflow or by the SRST instruction. For internal resets the activation of RSTOUT can be disabled. RSTOUT remains active (low) until the end of the reset sequence, until disabled by user software, or latest until the EINIT instruction is executed. This allows to initialize the controller before the external circuitry is activated. Note: Section 6.1.2 describes the control mechanisms for RSTOUT. User’s Manual DediPins_X7, V2.0 8-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9 The External Bus Controller EBC All external memory accesses are performed by a particular on-chip External Bus Controller (EBC). It can be programmed either to Single Chip Mode when no external memory is required at all, or dynamically (depending on the selected address range, belonging to a chip-select signal) to one of four different external memory access modes, which are as follows: • • • • 16/17/18/19 … 24-bit Addresses, 16-bit Data, Demultiplexed 16/17/18/19 … 24-bit Addresses, 16-bit Data, Multiplexed 16/17/18/19 … 24-bit Addresses, 8-bit Data, Multiplexed 16/17/18/19 … 24-bit Addresses, 8-bit Data, Demultiplexed In the demultiplexed bus modes, addresses are output on PORT1 and data is input/output on PORT0. In the multiplexed bus modes both addresses and data use PORT0 for input/output. High order address (segment) lines use Port 4. For applications which do not use all address lines for external devices, the external address space can be restricted to 8 Mbytes, 4 Mbytes, 2 Mbytes, 1 Mbyte, 512 Kbytes, 256 Kbytes, 128 Kbytes or 64 Kbytes. In this case Port 4 outputs seven, six, five and so on, or no segment address lines at all. Up to 5 external CS signals can be generated in order to save external glue logic. Access to very slow memories is supported via a particular ‘Ready’ function. A HOLD/HLDA protocol is available for bus arbitration. The XC167 External Bus Controller (EBC) allows access to external peripherals/memories and to internal LXBus modules. The LXBus is an internal representation of the ExtBus and it controls accesses to integrated peripherals and modules in the same way as accesses to external components. Because some ExtBus control signals are generally configurable, related additional control signals are necessary for the internal LXBus to support its maybe different configuration. The function of the EBC is controlled via a set of configuration registers. The basic and general behaviour is programmed via the mode-selection registers EBCMOD0 and EBCMOD1. Additionally to the supported external bus chip-select channels, one LXBus chip select channel is provided (both types together handled as ‘external’ chip select channels). With one exception, each of these chip-select signals is programmable via a set of registers. The Function CONtrol register for CSx (FCONCSx) register specifies the external bus/LXBus cycles in terms of address (multiplexed/demultiplexed), data (16-bit/8-bit), READY control, and chip-select enable. The timing of the bus access is controlled by the Timing CONfiguration registers for CSx (TCONCSx), which specify the timing of the bus cycle with the lengths of the different access phases. All these parameters are used for accesses within a specific address area that is defined via the corresponding ADDRess SELect register ADDRSELx. The five register sets (FCONCSx/TCONCSx/ADDRSELx) define five independent and programmable “address windows”, whereas all external accesses outside these User’s Manual EBC_X8, V2.2 9-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC windows are controlled via registers FCONCS0 and TCONCS0. Chip Select signals CS0 … CS4 belong to accesses on external bus, the additional Chip Select CS7 is used for access to the internal TwinCAN module on LXBus. The external bus timing is related to the reference CLocK OUTput (CLKOUT). All bus signals are generated in relation to the rising edge of this clock. The external bus protocol is compatible with those of the standard C166 Family. However, the external bus timing is improved in terms of wait-state granularity and signal flexibility. These improvements are configured via an enhanced register set (see above) in comparison to C166 Family. The C16x registers SYSCON and BUSCONx are no longer used. But because the configuration of the external bus controller is done during the application initialization, only some initialization code has to be adapted for using the new EBC module instead of the C16x external bus controller. User’s Manual EBC_X8, V2.2 9-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.1 External Bus Signals The EBC is using the following I/O signals: Table 9-1 EBC Bus Signals Signal I/O Port Pins Description ALE O P20 Address Latch Enable; active high RD O P20 ReaD strobe: activated for every read access (active low) WR, WRL O P20 WRite/WRite Low byte strobe (active low) WR-mode: activated for every write access. WRL-mode: activated for low byte write accesses on a 16-bit bus and for every data write access on an 8-bit bus. BHE, WRH O P3 Byte High Enable/WRite High byte strobe (active low) BHE-mode: activated for every data access to the upper byte of the 16-bit bus (handled as additional address bit) WRH-mode: activated for high byte write accesses on a 16-bit bus. AD[15..0] I/O P0 Address/Data bus; in multiplexed mode this bus is used for both address and data, in demultiplexed mode it is data bus only A[23..0] O P4, P1 Address bus READY/ READY I P20 READY; used for dynamic wait state insertion; programmable active high or low CS[4..0] O P6 Chip Select; active low; CS7 is used for internal LXBus access to TwinCAN Table 9-2 Write Configurations (see Chapter 9.3.2) Written Byte General Write Configuration Low High WR BHE – – inactive don’t care 0/1 write – active inactive – write active write write active User’s Manual EBC_X8, V2.2 Separated Byte Low/High Writes ADDR[0] WRL WRH ADDR[0] inactive inactive 0/1 0 active inactive 0/1 active 1 inactive active 0/1 active 0 active active 0/1 9-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.2 Timing Principles 9.2.1 Basic Bus Cycle Protocols The external bus timing is defined by six different timing phases (A-F). These phases define all control signals needed for any access sequence to external devices. At the beginning of a phase, the output signals may change within a given output delay time. After the output delay time, the values of the control output signals are stable within this phase. The output delay times are specified in the AC characteristics. Each phase can occupy a programmable number of clock cycles. The number of clock cycles is programmed in the TCONCSx register selected via the related address range and CSx. Access n Phases Address FCONCSx A BCDEF Address n FCON of n Access n + 1 A BCDEF Address n + 1 Access n + 2 A BCDEF Address n + 2 FCON of n + 1 Access n + 3 A BCDEF A Address n + 3 FCON of n + 2 FCON of n + 3 MCA05373 Figure 9-1 Phases of a Sequence of Several Accesses Phase A is used for tristating databus drivers from the previous cycle (tristate wait states after CS switch). Phase A cycles are not inserted at every access cycle but only when changing the CS. If an access using one CS (CSx) was finished and the next access with a different CS (CSy) is started then Phase A cycle(s) are performed according to the control bits as set in the first CS (CSx). The A Phase cycles are inserted while the addresses and ALE of the next cycle are already applied. The following diagrams show the 6 timing phases for read and write accesses on the demultiplexed bus and the multiplexed bus. User’s Manual EBC_X8, V2.2 9-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.2.1.1 Demultiplexed Bus Phases A B C D E F ALE Valid ADDR, CS RD Valid Read DATA Programmable Clocks 0-3 1-2 0-3 0-1 1 - 32 0-3 MCT05374 Figure 9-2 Demultiplexed Bus Read Phases A B C D E F ALE Valid ADDR, CS WR Valid Write DATA Programmable Clocks 0-3 1-2 0-3 0-1 1 - 32 0-3 MCT05375 Figure 9-3 • • • • • • Demultiplexed Bus Write A phase: Addresses valid, ALE high, no command. CS switch tristate wait states B phase: Addresses valid, ALE high, no command. ALE length C phase: Addresses valid, ALE low, no command. R/W delay D phase: Write data valid, ALE low, no command. Data valid for write cycles E phase: Command (read or write) active. Access time F phase: Command inactive, address hold. Read data tristate time, write data hold time User’s Manual EBC_X8, V2.2 9-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.2.1.2 Multiplexed Bus Phases A B C D E F ALE Valid ADDR, CS RD Address Valid RD DATA Programmable Clocks 0-3 1-2 0-3 Data In 0-1 1 - 32 0-3 MCT05376 Figure 9-4 Multiplexed Bus Read Phases A B C D E F ALE Valid ADDR, CS WR WR DATA Programmable Clocks Address Valid 0-3 1-2 0-3 Data Out 0-1 1 - 32 Next Address 0-3 MCT05377 Figure 9-5 • • • • • • Multiplexed Bus Write A phase: addresses valid, ALE high, no command. CS switch tristate wait states B phase: addresses valid, ALE high, no command. ALE length C phase: addresses valid, ALE low, no command. Address hold, R/W delay D phase: address tristate for read cycles, data valid for write cycles, ALE low, no command E phase: command (read or write) active. Access time F phase: command inactive, address hold. Read data tristate time, write data hold time User’s Manual EBC_X8, V2.2 9-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.2.2 Bus Cycle Phases 9.2.2.1 A Phase - CS Change Phase The A phase can take 0-3 clocks. It is used for tristating databus drivers from the previous cycle (tristate wait states after chip select switch). A phase cycles are not inserted at every access cycle, but only when changing the CS. If an access using one CS (CSx) ends and the next access with a different CS (CSy) is started, then A phase cycles are performed according to the bits set in the first CS (CSx). This feature is used to optimize wait states with devices having a long turn-off delay at their databus drivers, such as EPROMs and flash memories. The A phase cycles are inserted while the addresses and ALE of the next cycle are already applied. If there are some idle cycles between two accesses, these clocks are taken into account and the A phase is shortened accordingly. For example, if there are three tristate cycles programmed and two idle cycles occur, then the A phase takes only one clock. 9.2.2.2 B Phase - Address Setup/ALE Phase The B phase can take 1-2 clocks. It is used for addressing devices before giving a command, and defines the length of time that ALE is active. In multiplexed bus mode, the address is applied for latching. 9.2.2.3 C Phase - Delay Phase The C phase is similar to the A an B phases but ALE is already low. It can take 0-3 clocks. In multiplexed bus mode, the address is held in order to be latched safely. Phase C cycles can be used to delay the command signals (RW delay). 9.2.2.4 D Phase - Write Data Setup/MUX Tristate Phase The D phase can take 0-1 clocks. It is used to tristate the address on the multiplexed bus when a read cycle is performed. For all write cycles, it is used to ensure that the data are valid on the bus before the command is applied. 9.2.2.5 E Phase - RD/WR Command Phase The E phase is the command or access phase, and takes 1-32 clocks. Read data are fetched, write data are put onto the bus, and the command signals are active. Read data are registered with the terminating clock of this phase. The READY function lengthens this phase, too. READY-controlled access cycles may have an unlimited cycle time. User’s Manual EBC_X8, V2.2 9-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.2.2.6 F Phase - Address/Write Data Hold Phase The F phase is at the end of an access. It can take 0-3 clocks. Addresses and write data are held while the command is inactive. The number of wait states inserted during the F phase is independently programmable for read and write accesses. The F phase is used to program tristate wait states on the bidirectional data bus in order to avoid bus conflicts. 9.2.3 Bus Cycle Examples: Fastest Access Cycles b e CLK ALE ADDR, CS Valid RD DATA In Valid MCT05378 Figure 9-6 Fastest Read Cycle Demultiplexed Bus b e CLK ALE ADDR, CS Valid WR DATA Out Valid MCT05379 Figure 9-7 User’s Manual EBC_X8, V2.2 Fastest Write Cycle Demultiplexed Bus 9-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC b d e f CLK ALE ADDR, CS Valid RD Muxed Address Out / Data In Addr. Valid Data Valid MCT05380 Figure 9-8 Fastest Read Cycle Multiplexed Bus b e CLK ALE ADDR, CS Valid RD Muxed Address Out / Data Out Addr. Valid Valid MCT05381 Figure 9-9 User’s Manual EBC_X8, V2.2 Fastest Write Cycle Multiplexed Bus 9-9 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3 Functional Description 9.3.1 Configuration Register Overview There are 3 groups of EBC registers: • • • EBC mode registers have influence on global functions. Chip-select-related registers control the functionality linked to one CS. TwinCAN and Startup Memory registers are used to control the access to the internal LXBus. CS0 is the default chip-select signal that is active whenever no other chip-select or internal address space is addressed. Therefore, CS0 has no ADDRSEL register. Note: All EBC registers are write-protected by the EINIT protection mechanism. Thus, after execution of the EINIT instruction, these registers are not writable any more. Table 9-3 EBC Configuration Register Overview Name CS1) Description Address 00EExxH Start-up Value EBCMOD0 all EBC MODe 0; alternate function of EBC pins 00 0xxxH EBCMOD1 all EBC MODe 1; alternate function of EBC pins 02 0000H TCONCS0 0 Timing CONtrol for CS0 10 6243H FCONCS0 0 Function CONtrol for CS0 12 0021H TCONCS1-71) 1-61), Timing CONtrol for CS1 … CS71) 7 18, 20, 28, 30, 38, 40, 48 0000H FCONCS1-71) 1-61), Function CONtrol for CS1 … CS71) 7 1A, 22, 2A, 32, 3A, 42, 4A 0000H 1E, 26, 2E, 36, 3E, 46, 4E 0000H ADDRSEL1-71) 1-61), ADDress window SELection 7 for CS1 … CS71) 1) CS5 and CS6 register sets are not available (reserved for future LXBus peripherals). A 128-byte address space is occupied/reserved by the EBC. User’s Manual EBC_X8, V2.2 9-10 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 00EE8E ADDRSEL7 00EE4E FCONCS7 00EE4A TCONCS7 00EE48 ADDRSEL4 00EE36 FCONCS4 TCONCS4 00EE32 00EE30 ADDRSEL3 00EE2E FCONCS3 00EE2A TCONCS3 00EE28 ADDRSEL2 00EE26 FCONCS2 00EE22 TCONCS2 00EE20 ADDRSEL1 00EE1E FCONCS1 TCONCS1 00EE1A 00EE18 FCONCS0 TCONCS0 00EE12 00EE10 EBCMOD1 00EE02 CS7 Channel Control CS4 Channel Control CS3 Channel Control CS2 Channel Control CS1 Channel Control CS0 Channel Control General EBC Control EBCMOD0 00EE00 MCA05382_XC Figure 9-10 Mapping of EBC Registers into the XSFR Space Note: CS5 and CS6 register sets are not available (reserved for future LXBus peripherals). User’s Manual EBC_X8, V2.2 9-11 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.2 The EBC Mode Register 0 EBCMODe Register 0 EBCMOD0 EBC Mode Register 0 15 14 13 12 XSFR (EE00H/--) 11 10 9 8 7 RDY RDY ALE BYT WR EBC SLA ARB POL DIS DIS DIS CFG DIS VE EN rw rw rw rw rw rw rw rw Reset Value: XXXXH 6 5 4 3 2 1 CSPEN SAPEN rw rw Field Bits Type Description RDYPOL 15 rw READY Pin Polarity 0 READY is active low 1 READY is active high RDYDIS 14 rw READY Pin Disable 0 READY enabled 1 READY disabled ALEDIS 13 rw ALE Pin Disable 0 ALE enabled 1 ALE disabled BYTDIS 12 rw BHE Pin Disable 0 BHE enabled 1 BHE disabled WRCFG1) 11 rw Configuration for Pins WR/WRL, BHE/WRH 0 WR and BHE 1 WRL and WRH EBCDIS 10 rw EBC Pins Disable 0 EBC is using the pins for external bus 1 EBC pins disabled SLAVE 9 rw SLAVE Mode Enable 0 Bus arbiter acts in master mode 1 Bus arbiter acts in slave mode ARBEN 8 rw BUS Arbitration Pins Enable 0 HOLD, HLDA and BREQ pins are disabled 1 Pins act as HOLD, HLDA, and BREQ User’s Manual EBC_X8, V2.2 9-12 0 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Field Bits Type Description CSPEN [7:4] rw CSx Pins Enable (only external CSx) 0000 All external Chip Select pins disabled. 0001 CS0 pin enabled 0010 CS1 and CS0 pin enabled … … 0101 Five CSx pins enabled: CS4 - CS0 Else not supported (reserved) SAPEN [3:0] rw Segment Address Pins Enable 0000 All segment address pins disabled 0001 One: A[16] enabled … … 1000 Eight: A[23:16] enabled Else not supported (reserved) 1) A change of the bit content is not valid before the next external bus access cycle. Notes: 1. Disabled pins are used for general purpose IO or for alternate functions (see port and pin descriptions). 2. Bitfield CSPEN controls the number of available CSx pins. The related address windows and bus functions are enabled with the specific ENCSx bits in the FCONCSx registers (see Page 9-16). There, an additional chip select (CS7) is defined for internal access to the LXBus peripheral TwinCAN. 3. The external bus arbitration pins have a separate ARBitration ENable bit (ARBEN) that has to be set in order to use the pins for arbitration and not for General Purpose IO (GPIO). If ARBEN is cleared, the arbitration inputs HLDA and HOLD are fixed internally to an inactive high state. Additionally, the master/slave setting of the arbiter is done with a separate bit (SLAVE). 4. The reset value depends on the selected startup configuration. User’s Manual EBC_X8, V2.2 9-13 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.3 The EBC Mode Register 1 EBC MODe register 1 controls the general use of port pins for external bus. EBCMOD1 EBC Mode Register 1 XSFR (EE02H/--) 15 14 13 12 11 10 9 8 - - - - - - - - - - - - - - - - 7 6 Reset Value: 0000H 5 4 WRP DHP ALP A0P DIS DIS DIS DIS rw rw rw 3 2 1 0 APDIS rw rw Field Bits Type Description WRPDIS 7 rw WR/WRL Pin Disable 0 WR/WRL pin of Port P20 enabled 1 WR/WRL pin of Port P20 disabled DHPDIS 6 rw Data High Port Pins Disable 0 Addr./Data bus pins 15-8 of P0H enabled 1 Addr./Data bus pins 15-8 of P0H disabled ALPDIS 5 rw Address Low Pins Disable 0 Address bus pins 7-0 of PORT1 generally enabled (depending on APDIS/A0PDIS) 1 Address bus pins 7-0 of PORT1 disabled A0PDIS 4 rw Address Bit 0 Pin Disable 0 Address bus pin 0 of PORT1 enabled 1 Address bus pin 0 of PORT1 disabled APDIS [3:0] rw Address Port Pins Disable 0000 Address bus pins 15-1 of PORT1 enabled 0001 Pin A15 disabled, A14-A1 enabled 0010 Pins A15-A14 disabled, A13-A1 enabled 0011 Pins A15-A13 disabled, A12-A1 enabled … … 1110 Pins A15-A2 disabled, A1 enabled 1111 Address bus pins 15-1 of PORT1 disabled Note: Disabled bus pins may be used for general purpose IO or for alternate functions (see port and pin descriptions). Note: After reset, the address and data bus pins are enabled, but in Idle state. User’s Manual EBC_X8, V2.2 9-14 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.4 The Timing Configuration Registers TCONCSx The timing control registers are used to program the described cycle timing for the different access phases. The timing control registers may be reprogrammed during code fetches from the affected address window. The new settings are first valid for the next access. TCONCS0 Timing Cfg. Reg. for CS0 15 14 13 12 11 XSFR (EE10H/--) 10 9 8 7 6 Reset Value: 7AXXH 5 4 3 2 1 0 - WRPHF RDPHF PHE PHD PHC PHB PHA - rw rw rw rw rw rw rw TCONCSx Timing Cfg. Reg. for CSx 15 14 13 12 11 XSFR (EEXXH/--) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 - WRPHF RDPHF PHE PHD PHC PHB PHA - rw rw rw rw rw rw rw x = 1 … 4, 7 Note: x = 7 belongs to the additional chip select (CS7) which is used and defined for internal access to the LXBus peripheral TwinCAN. Field Bits WRPHF [14:13] rw Write Phase F 00 0 clock cycles … … 11 3 clock cycles (default) RDPHF [12:11] rw Read Phase F 00 0 clock cycles (default) … … 11 3 clock cycles PHE [10:6] Phase E 00000: 1 clock cycle … … (default: 9 clock cycles) 11111: 32 clock cycles User’s Manual EBC_X8, V2.2 Type rw Description 9-15 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Field Bits Type Description PHD 5 rw Phase D 0 0 clock cycles (default) 1 1 clock cycle PHC [4:3] rw Phase C 00 0 clock cycles (default) … … 11 3 clock cycles PHB 2 rw Phase B 0 1 clock cycle (default) 1 2 clock cycles PHA [1:0] rw Phase A 00 0 clock cycles … … 11 3 clock cycles (default) 9.3.5 The Function Configuration Registers FCONCSx The Function Control registers are used to control the bus and READY functionality for a selected address window. It can be distinguished between 8 and 16-bit bus and multiplexed and demulitplexed accesses. Furthermore it can be defined whether the address window (and its chip select signal CSx) is generally enabled or not. FCONCS0 Function Cfg. Reg. for CS0 XSFR (EE12H/--) Reset Value: 00X1H 15 14 13 12 11 10 9 8 7 6 5 4 3 - - - - - - - - - - BTYP - - - - - - - - - - - rw - FCONCSx Function Cfg. Reg. for CSx XSFR (EEXXH/--) 2 1 0 RDY RDY EN MOD EN CS rw rw rw Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 - - - - - - - - - - BTYP - - - - - - - - - - - rw - 2 1 0 RDY RDY EN MOD EN CS rw rw rw x = 1 … 4, 7 Note: x = 7 belongs to the additional chip select (CS7) which is used and defined for internal access to the LXBus peripheral TwinCAN. User’s Manual EBC_X8, V2.2 9-16 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Field Bits Type Description BTYP [5:4] rw Bus Type Selection 00 8 bit Demultiplexed 01 8 bit Multiplexed 10 16 bit Demultiplexed 11 16 bit Multiplexed RDYMOD 2 rw Ready Mode 0 Asynchronous READY 1 Synchronous READY RDYEN 1 rw Ready Enable 0 Access time is controlled by bitfield PHEx 1 Access time is controlled by bitfield PHEx and READY signal ENCS1) 0 rw Enable Chip Select 0 Disable 1 Enable 1) Disabling a chip select not only effects the chip select output signal, it also deactivates the respective address window of the disabled chip select. A disabled address window is also ignored by an address window arbitration (see Chapter 9.3.6.2). Note: The specific ENCSx bits in the FCONCSx registers enable the related address windows and bus functions and the corresponding chip select signal CSx. But it depends on the definition of bitfield CSPEN in register EBCMOD0 how many CSx pins are available and used for the external system. If an address window is enabled but no external pin is available for the CSx, the external bus cycle is executed without chip select signal. Note: With ENCS7 the chip select CS7 and its related register set is enabled and defined for internal access to the LXBus peripheral TwinCAN. User’s Manual EBC_X8, V2.2 9-17 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.6 The Address Window Selection Registers ADDRSELx ADDRSELx Address Range/Size for CSx 15 14 13 12 11 XSFR (/--) 10 9 8 7 Reset Value: 0000H 6 5 4 3 2 1 RGSAD RGSZ rw rw 0 x = 1 … 4, 7 Field Bits Type Description RGSAD [15:4] rw Address Range Start Address Selection RGSZ [3:0] rw Address Range Size Selection (see Table 9-4) Note: There is no register ADDRSEL0, as register set FCONCS0/TCONCS0 controls all external accesses outside the address windows built by the enabled (by ENCS bit in FCONCSx) address selects ADDRSELx. 9.3.6.1 Definition of Address Areas The enabled register sets FCONCSx/TCONCSx/ADDRSELx (x = 1 … 4, 7) define separate address areas within the address space of the XC167. Within each of these address areas the conditions of external accesses and LXBus accesses (x = 7) can be controlled separately, whereby the different address areas (windows) are defined by the ADDRSELx registers. Each ADDRSELx register cuts out an address window, where the corresponding parameters of the registers FCONCSx and TCONCSx are used to control external accesses. The range start address of such a window defines the most significant address bits of the selected window which are consequently not needed to address the memory/module in this window (Table 9-4). The size of the window chosen by ADDRSELx.RGSZ defines the relevant bits of ADDRSELx.RGSAD (marked with ‘R’) which are used to select with the most significant bits of the request address the corresponding window. The other bits of the request address are used to address the memory locations inside this window. The lower bits of ADDRSELx.RGSAD (marked ‘x’) are disregarded. The address area from 00’8000H to 00’FFFFH (32 Kbytes) is reserved for CPU internal registers and data RAM, the area from BF’0000H to BF’7FFFH (32 Kbytes) for internal startup memory and the area from C0’0000H to FF’FFFFH (4 Mbytes) is used by the internal program memory. Therefore, these address areas cannot be used by external resources connected to the external bus. User’s Manual EBC_X8, V2.2 9-18 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Table 9-4 Address Range and Size for ADDRSELx ADDRSELx Address Window Range Size RGSZ Relevant (R) Bits of RGSAD Selected Address Range Range Start Address A[23:0] Selected with R-bits of RGSAD 3…0 15 … 4 Size A23 … A0 0000 0001 0010 0011 0100 0101 0110 0111 1000 1001 1010 1011 11xx RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRRx RRxx Rxxx xxxx 4 Kbytes 8 Kbytes 16 Kbytes 32 Kbytes 64 Kbytes 128 Kbytes 256 Kbytes 512 Kbytes 1 Mbytes 2 Mbytes 4 Mbytes 8 Mbytes reserved1) RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRRR RRR0 RR00 R000 ---- RRRR RRRR RRRR RRRR RRRR RRRx RRxx Rxxx xxxx xxxx xxxx xxxx xxxx RRRR RRRx RRxx Rxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx RRRR RRRR RRRR RRRR RRRR RRR0 RR00 R000 0000 0000 0000 0000 ---- RRRR RRR0 RR00 R000 0000 0000 0000 0000 0000 0000 0000 0000 ---- 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 ---- 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 ---- 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 ---- 1) The complete address space of 12 Mbytes can be selected by the default chip select CS0. Note: The range start address can only be on boundaries specified by the selected range size according to Table 9-4. User’s Manual EBC_X8, V2.2 9-19 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.6.2 Address Window Arbitration For each external access the EBC compares the current address with all address select registers (programmable ADDRSELx and hardwired address select registers for startup memory) of enabled windows. This comparison is done in four levels: Priority 1: Registers ADDRSELx [x = 2, 4] are evaluated first. A window match with one of these registers directs the access to the respective external area using the corresponding set of control registers FCONCSx/TCONCSx and ignoring registers ADDRSELy. An overlapping of windows of this group will lead to an undefined behaviour. Priority 2: A match with registers ADDRSELy [y = 1, 3, 7] directs the access to the respective external area using the corresponding set of control registers FCONCSy/TCONCSy. An overlapping of windows of this group will lead to an undefined behaviour. Overlaps with priority 2 ADDRSELx are only allowed for the (x, y) pairs (2, 1) and (4, 3). Priority 3: If there is no match with any address select register (neither the hardwired ones nor the programmable ADDRSEL) the access to the external bus uses the general set of control registers FCONCS0/TCONCS0 if enabled. User’s Manual EBC_X8, V2.2 9-20 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.7 Ready Controlled Bus Cycles 9.3.7.1 General In cases, where the response (access) time of a peripheral is not constant, or where the programmable wait states are not enough, the EBC provides external bus cycles that are terminated via a READY input signal. In this case during phase E the EBC first counts a programmable number of clock cycles (1 … 32) and then starts in the last wait cycle to monitor the internal READY line (see Figure 9-11) to determine the actual end of the current bus cycle. The external device drives READY active in order to indicate that data has been latched (write cycle) or is available (read cycle). The READY pin is generally enabled by setting the bit RDYDIS in EBCMOD0 to ‘0’ in order to switch the corresponding port pin. Also the polarity of the READY is defined inside the EBCMOD0 register on the RDYPOL bit. For a specific address window the READY function is enabled via the RDYEN bit in the FCONCSx register. With FCONCSx.RDYMOD the READY is handled either in synchronous or in asynchronous mode (see also Figure 9-11). When the READY function is enabled for a specific address window, each bus cycle within this window must be terminated with an active READY signal. Otherwise the controller hangs until the next reset. This is also the case for an enabled RDYEN but a disabled READY port pin. MUX READY Ext Async. 0 1 Sync. 1 EBCMOD0.RDYPOL MUX 0 1 READY Int FCONCSx.RDYMODx MCA05383 Figure 9-11 External to Internal READY Conversion User’s Manual EBC_X8, V2.2 9-21 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.7.2 The Synchronous/Asynchronous READY The synchronous READY provides the fastest bus cycles, but requires setup and hold times to be met. The CLKOUT signal should be enabled and may be used by the peripheral logic to control the READY timing in this case. The asynchronous READY is less restrictive, but requires one additional wait state caused by the internal synchronization. As the asynchronous READY is sampled earlier programmed wait states may be necessary to provide proper bus cycles. A READY signal (especially asynchronous READY) that has been activated by an external device may be deactivated in response to the trailing (rising) edge of the respective command (RD or WR). Bus Cycle with Active READY Programmed phase E wait states Bus Cycle Extended via READY Programmed phase E wait states ALE RD / WR Sync. READY Async. READY Sampling of READY Input Not Interesting READY Cycles MCT05384 Figure 9-12 READY Controlled Bus Cycles 9.3.7.3 Combining the READY Function with Predefined Wait States Typically an external wait state or READY control logic takes a while to generate the READY signal when a cycle was started. After a predefined number of clock cycles the EBC will start checking its READY line to determine the end of the bus cycle. When using the READY function with so-called ‘normally-ready’ peripherals, it may lead to erroneous bus cycles, if the READY line is sampled too early. These peripherals pull their READY output active, while they are idle. When they are accessed, they drive READY inactive until the bus cycle is complete, then drive it active again. If, however, the peripheral drives READY inactive a little late, after the first sample point of the XC167, the controller samples an active READY and terminates the current bus cycle User’s Manual EBC_X8, V2.2 9-22 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC too early. By inserting predefined wait states the first READY sample point can be shifted to a time, where the peripheral has safely controlled the READY line. 9.3.8 Access Control to TwinCAN Access control to LXBus is required for accesses to the TwinCAN module. In general, accesses to LXBus are not visible on external bus. During LXBus cycles, the external bus is still enabled, but driven to inactive states (control signals) or switched into the read mode (buses). For accesses to the TwinCAN, CS7 and its control registers, the ADDRSEL7, TCONCS7 and the FCONCS7 are used. The selection of LXBus is controlled with CS7. The address range, defined in ADDRSEL7, is recommended to be located in the ‘External IO Range’ (range from 20’0000H to 3F’0000H). Only for the External IO Range of the total external address range it is guaranteed that a read access is executed after a preceeding write access. After reset (controlled by the startup program sequence), the TwinCAN address range is adjusted per default to the area from address 20’0000H to 20’0FFFH (4 KB), resulting in the ADDRSEL7 default-code of 2000H. This initial value of ADDRSEL7 may be changed afterwards by the user. The initial default value of the bus function control register FCONCS7 is selected according to the requirements of the TwinCAN: 16-bit demultiplexed bus, access time controlled with synchronous READY. This function control is represented by the default value for FCONCS7 of 0027H. The initial LXBus cycle timing as controlled with register TCONCS7 after reset is the shortest possible timing using two clock cycles for one bus cycle. But this minimum timing will be lengthened with waitstate(s) controlled by the TwinCAN itself with the READY function. This timing control is controlled by the reset value of TCONCS7 (0000H). User’s Manual EBC_X8, V2.2 9-23 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.9 External Bus Arbitration The XC167 supports multi master systems on the external bus by its external bus arbitration. This bus arbitration allows an external master to request the external bus. The XC167 will release the external bus and will float the data and address bus lines and force the control signals via pull-ups/downs to their inactive state. 9.3.9.1 Initialization of Arbitration During reset all arbitration pins are tristate, except pin BREQ which is pulled inactive. After reset the XC167 EBC always starts in ‘init mode’ where the external bus is available but no arbitration is enabled. All arbitration pins are ignored in this state. Other to the external bus connected XC167 EBCs assume to have the bus also, so potential bus conflicts are not resolved. For a multimaster system the arbitration should be initialized first before starting any bus access. The EBC can either be chosen as arbitration master or as arbitration slave by programming the EBCMOD0 bit SLAVE. The selected mode and the arbitration gets active by the first setting of the HLDEN bit inside the CPUs PSW register. Afterwards a change of the slave/master mode is not possible without resetting the device. Of course for arbitration the dedicated pins have to be activated by setting EBCMOD0.ARBEN. 9.3.9.2 Arbitration Master Scheme If the XC167 EBC is configured as arbitration master, it is default owner of the external bus, controls the arbitration protocol and drives the bus also during idle phases with no bus requests. To perform the arbitration handshake a HOLD input allows the request of the external bus from the arbitration master. When the arbitration master hands over the bus to the requester this is signaled by driving the hold acknowledge pin HLDA low, which remains at this level until the arbitration slave frees the bus by releasing its request on the HOLD input. If the arbitration master is not the owner of the bus it treats the external bus interface as follows: • • • • Address and data bus(es) float to tristate Command lines are pulled high by internal pull-up devices (RD, WR/WRL, BHE/WRH) Address latch control line ALE is pulled low by an internal pull-down device CSx outputs are pulled high by internal pull-up devices. In this state the arbitration slave can take over the bus. If the arbitration master requires the bus again, it can request the bus via the bus request signal BREQ. As soon as the arbitration master regains the bus it releases the BREQ signal and drives HLDA to high. User’s Manual EBC_X8, V2.2 9-24 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Not fixed number of cycles (0 … n) HOLD HLDA Earliest Change BREQ CSx, WRH Pull Up WR/WRL, RD Not Active Driven ADD, DATA High Impedance BHE MCT05385 Figure 9-13 Releasing the Bus by the Arbitration Master Note: Figure 9-13 shows the first possibility for BREQ to get active. The XC167 will complete the currently running bus cycle before granting the external bus as indicated by the broken lines. User’s Manual EBC_X8, V2.2 9-25 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC HOLD HLDA No BREQ Request BREQ Latest Possible Change CSx, WRH WR/WRL, RD ADD, BHE Pull Up Not Active Driven High Impedance MCT05386 Figure 9-14 Regaining the Bus by the Arbitration Master Note: The falling BREQ edge shows the last chance for BREQ to trigger the indicated regain-sequence. Even if BREQ is activated earlier the regain-sequence is initiated by HOLD going high. Please note that HOLD may also be deactivated without the XC167 requesting the bus. 9.3.9.3 Arbitration Slave Scheme If the EBC is configured as arbitration slave it is by default not owner of the external bus and has to request the bus first. As long as it has not finished all its queued requests and the arbitration master is not requesting the bus the arbitration slave stays owner of the bus. For the description of the signal handling of the handshake see Chapter 9.3.9.2. For the arbitration slave the hold acknowledge pin HLDA is configured as input. User’s Manual EBC_X8, V2.2 9-26 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.9.4 Bus Lock Function If an application in a multimaster system requires a sequence of undisturbed bus access it has the possibility (independently of being arbitration slave or master) to lock1) the bus by setting the PSW bit HLDEN to ‘0’. In this case the locked EBC will not answer to HOLD requests from other external bus master until HLDEN is set to ‘1’ again. Of course a locked bus master not owning the bus can request the external bus. If a master and a slave are requesting the external bus at the same time for several accesses, they toggle the ownership after each access cycle if the bus is not locked. 9.3.9.5 Direct Master Slave Connection If one XC167 is configured as master and the other as slave and both are working on the same external bus as bus master, they can be connected directly together for bus arbitration as shown in Figure 9-15. As both EBCs assume after reset to own the external bus, the ‘slave’ CPU has to be released from reset and initialized first, before starting the ‘master’ CPU. The other way is to start both systems at the same time but then both EBC must be configured from internal memory and the PSW.HLDEN bits set before the first external bus request. EBC in Master Mode EBC in Slave Mode HOLD HOLD HLDA HLDA BREQ BREQ MCA05387 Figure 9-15 Connecting two XC167 Using Master/Slave Arbitration When multiple (more than two) bus masters (XC167 or other masters) shall share the same external resources an additional external bus arbiter logic is required that determines the currently active bus master and that controls the necessary signal sequences. 1) It is not allowed to lock the bus by resetting the EBCMOD0.ARBEN bit, as this can lead to bus conflicts. User’s Manual EBC_X8, V2.2 9-27 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.3.10 Shutdown Control In case of a shutdown request from the SCU it must be insured by the EBC that all the different functions of the EBC are in a non-active state before the whole chip is switched in a Idle, Powerdown, Sleep or Software Reset mode. A running bus cycle is finished, still requested bus cycles are executed. Depending on the master/slave configuration of EBC, the external bus arbiter is controlled for regaining the bus (master) before performing the requested cycles, or the external bus must be released after complete execution of still requested bus cycles (see Table 9-5). Only when this shutdown sequence is terminated, the shutdown acknowledge is generated from EBC (and from other modules, as described for SCU) and the chip can enter the requested mode. Table 9-5 gives an overview of the shutdown control in EBC depending on the EBC configuration. Table 9-5 Arbitration Mode EBC Shutdown Control Master Mode Slave Mode Bus Control With Control of Without the Bus Control of the Bus – User’s Manual EBC_X8, V2.2 Finish all pending cycle requests. Send shutdown acknowledge with the control of the bus. Ask for the bus. Finish all pending cycle requests. Send shutdown acknowledge with the control of the bus. 9-28 With Control of Without the Bus Control of the Bus Finish all pending requests. Send shutdown acknowledge after leaving the bus. Ask for the bus if needed and finish all requests. Send shutdown acknowledge after leaving the bus. V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC 9.4 LXBus Access Control and Signal Generation To connect on_chip peripherals via the EBC, the local system bus LXBus is provided. The LXBus is an internal (local) extension of the external bus. It is controlled by the External Bus Controller EBC identically to the external bus, using the select and cycle control functions as described for the external bus. The address range and chip select control with ADDRSELn registers, the function control with FCONCSn registers and the timing control with TCONCSn registers is identical to the external bus. Chip selects CS5 … CS7 are reserved for LXBus peripherals. In XC167, only one standard CSx, the CS7 is used for the LXBus, necessary for the TwinCAN module (see Chapter 9.3.8). Per default, the address range of this peripheral is located within the so-called ‘External IO Range’ (from 20’0000H to 3F’0000H). Accesses to the IO range are not buffered and not cached, and a read access is delayed until all IO writes pending in the pipeline are executed. Only internal accesses to LXBus peripherals are supported by the EBC. External accesses are not supported in this C166SV2 derivative. Accesses to LXBus peripherals and memories are not visible on external bus pads. 9.5 EBC Register Table Table 9-6 lists all EBC Configuration Registers which are implemented in the XC167 ordered by their physical address. The registers are all located in the XSFR space (internal IO space). Table 9-6 EBC Memory Table (ordered by physical address) Name Physic. Addr. Description Reset Value1) EBCMOD0 EE00H EBC Mode Register 0 XXXXH EBCMOD1 EE02H EBC Mode Register 1 0000H TCONCS0 EE10H CS0 Timing Configuration Register 7AXXH FCONCS0 EE12H CS0 Function Configuration Register 00X1H TCONCS1 EE18H CS1 Timing Configuration Register 0000H FCONCS1 EE1AH CS1 Function Configuration Register 0000H ADDRSEL1 EE1EH CS1 Address Size and Range Register 0000H TCONCS2 EE20H CS2 Timing Configuration Register 0000H FCONCS2 EE22H CS2 Function Configuration Register 0000H ADDRSEL2 EE26H CS2 Address Size and Range Register 0000H TCONCS3 EE28H CS3 Timing Configuration Register 0000H FCONCS3 EE2AH CS3 Function Configuration Register 0000H User’s Manual EBC_X8, V2.2 9-29 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The External Bus Controller EBC Table 9-6 EBC Memory Table (ordered by physical address) (cont’d) Name Physic. Addr. Description Reset Value1) ADDRSEL3 EE2EH CS3 Address Size and Range Register 0000H TCONCS4 EE30H CS4 Timing Configuration Register 0000H FCONCS4 EE32H CS4 Function Configuration Register 0000H ADDRSEL4 EE36H CS4 Address Size and Range Register 0000H TCONCS7 EE48H CS7 Timing Configuration Register 0000H FCONCS7 EE4AH CS7 Function Configuration Register 0000H ADDRSEL7 EE4EH CS7 Address Size and Range Register 0000H reserved EE50H EEFFH reserved - do not use – 1) NOTE: Reserved (and not listed) addresses are always read as FFFFH. However, for enabling future enhancements without any compatibility problems, these addresses should neither be written nor be used as read value by the software. User’s Manual EBC_X8, V2.2 9-30 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader 10 The Bootstrap Loader The built-in bootstrap loader of the XC167 provides a mechanism to load the startup program, which is executed after reset, via the serial interface. In this case no external memory or an internal ROM/OTP/Flash is required for the initialization code. The bootstrap loader moves code/data into the internal RAM, but it is also possible to transfer data via the serial interface into an external RAM using a second level loader routine. ROM memory (internal or external) is not necessary. However, it may be used to provide lookup tables or may provide “core-code”, i.e. a set of general purpose subroutines, e.g. for IO operations, number crunching, system initialization, etc. RSTIN P0L.4 or RD 1) 2) 4) RxD0 3) TxD0 5) CSP:IP 6) Int. Boot ROM BSL-routine 32 bytes User Software MCT04465_xx Figure 10-1 Bootstrap Loader Sequence The Bootstrap Loader may be used to load the complete application software into ROMless systems, it may load temporary software into complete systems for testing or calibration, it may also be used to load a programming routine for Flash devices. The BSL mechanism may be used for standard system startup as well as only for special occasions like system maintenance (firmware update) or end-of-line programming or testing. User’s Manual BSL_X, V2.0 10-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader 10.1 Entering the Bootstrap Loader The XC167 enters BSL mode triggered by external configuration during a hardware reset: • • when selected via bitfield SMOD at the end of an external reset (EA = 0) when pin RD is sampled low at the end of an internal reset (EA = 1). In this case the built-in bootstrap loader is activated independent of the selected bus mode. The bootstrap loader code is stored in a special Boot-ROM, no part of the standard mask ROM, OTP, or Flash memory area is required for this. The hardware that activates the BSL during reset may be a simple pull-down resistor for systems that use this feature upon every hardware reset. You may want to use a switchable solution (via jumper or an external signal) for systems that only temporarily use the bootstrap loader. The ASC0 receiver is only enabled after the identification byte has been transmitted. A half duplex connection to the host is therefore sufficient to feed the BSL. Note: The proper reset configuration for BSL mode requires a set of pins to be driven to defined logic levels (see Section 6.1.4). User’s Manual BSL_X, V2.0 10-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader Initial State in BSL Mode After entering BSL mode and the respective initialization the XC167 scans the RxD0 line to receive a zero byte, i.e. one start bit, eight 0 data bits and one stop bit. From the duration of this zero byte it calculates the corresponding baudrate factor with respect to the current CPU clock, initializes the serial interface ASC0 accordingly and switches pin TxD0 to output. Using this baudrate, an identification byte is returned to the host that provides the loaded data. This identification byte identifies the device to be booted. The following codes are defined: 55H: 8xC166. A5H: Previous versions of the C167 (obsolete). B5H: Previous versions of the C165. C5H: C167 derivatives. D5H: All devices equipped with identification registers. Note: The identification byte D5H does not directly identify a specific derivative. This information can in this case be obtained from the identification registers. When the XC167 has entered BSL mode, the following configuration is automatically set (values that deviate from the normal reset values, are marked): Watchdog Timer: Disabled ASC0_BG: XXXXH P3.10/TxD0: ‘1’ ASC0_CON: 8811H DP3.10: ‘1’ GPT12E_T6CON: 0880H ALTSEL0P3.10: ‘1’ GPT12E_T6: XXXXH Other than after a normal reset the watchdog timer is disabled, so the bootstrap loading sequence is not time limited. Pin TxD0 is configured as output, so the XC167 can return the identification byte. User’s Manual BSL_X, V2.0 10-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader 10.2 Loading the Startup Code After sending the identification byte the BSL enters a loop to receive 32 Bytes via ASC0. These bytes are stored sequentially into locations E0’0004H through E0’0023H of the internal PSRAM. So up to 16 instructions may be placed into the PSRAM area. The first two words of the PSRAM are loaded with the DISWDT instruction. To execute the loaded code the BSL then points register VECSEG to location E0’0000H, i.e. the first loaded instruction1). The bootstrap loading sequence terminates by executing a software reset. Most probably the initially loaded routine will load additional code or data, as an average application is likely to require substantially more than 16 instructions. This second receive loop may directly use the pre-initialized interface ASC0 to receive data and store it to arbitrary user-defined locations. This second level of loaded code may be the final application code. It may also be another, more sophisticated, loader routine that adds a transmission protocol to enhance the integrity of the loaded code or data. It may also contain a code sequence to change the system configuration and enable the bus interface to store the received data into external memory. This process may go through several iterations or may directly execute the final application. Note: Data fetches from a protected ROM will not be executed. 10.3 Exiting Bootstrap Loader Mode After the bootstrap loader has been activated, the watchdog timer and the debug system are disabled. The debug system is released automatically when the BSL terminates after having received the 32nd byte from the host. In order to activate the watchdog timer, if required, it must be enabled via instruction ENWDT (before executing the EINIT instruction). Also a reset will re-enable the WDT: • • a software reset (ignoring the external configuration) a hardware reset, not configuring BSL mode. After the (non-BSL) reset the XC167 will start executing out of user memory as externally configured via PORT0 or RD/ALE (depending on EA). 1) This includes the execution of the initial DISWDT instruction, ensuring that the 2nd level loader is not aborted by the watchdog timer. User’s Manual BSL_X, V2.0 10-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader 10.4 Choosing the Baudrate for the BSL The calculation of the serial baudrate for ASC0 from the length of the first zero byte that is received, allows the operation of the bootstrap loader of the XC167 with a wide range of baudrates. However, the upper and lower limits have to be kept, in order to ensure proper data transfer. f SYS B MC = ----------------------------------------------------32 × ( ASC0BG + 1 ) (10.1) The XC167 uses timer GPT12E_T6 to measure the length of the initial zero byte. The quantization uncertainty of this measurement implies the first deviation from the real baudrate, the next deviation is implied by the computation of the ASC0_BG reload value from the timer contents. Equation (10.2) shows the association: T6 – 36 ASC0BG = -------------------72 9 f SYS T6 = --- × -----------4 B Host (10.2) For a correct data transfer from the host to the XC167 the maximum deviation between the internal initialized baudrate for ASC0 and the real baudrate of the host should be below 2.5%. The deviation (FB, in percent) between host baudrate and XC167 baudrate can be calculated via Equation (10.3): B Contr – B Host F B = ------------------------------------ × 100% B Contr F B ≤ 2.5% (10.3) Note: Function (FB) does not consider the tolerances of oscillators and other devices supporting the serial communication. This baudrate deviation is a nonlinear function depending on the CPU clock and the baudrate of the host. The maxima of the function (FB) increase with the host baudrate due to the smaller baudrate prescaler factors and the implied higher quantization error (see Figure 10-2). User’s Manual BSL_X, V2.0 10-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader Ι FB 2.5% B Low B High B Host ΙΙ MCA02260 Figure 10-2 Baudrate Deviation between Host and XC167 The minimum baudrate (BLow in Figure 10-2) is determined by the maximum count capacity of timer GPT12E_T6, when measuring the zero byte, i.e. it depends on the system clock. The minimum baudrate is obtained by using the maximum GPT12E_T6 count 216 in the baudrate formula. Baudrates below BLow would cause GPT12E_T6 to overflow. In this case ASC0 cannot be initialized properly and the communication with the external host is likely to fail. The maximum baudrate (BHigh in Figure 10-2) is the highest baudrate where the deviation still does not exceed the limit, i.e. all baudrates between BLow and BHigh are below the deviation limit. BHigh marks the baudrate up to which communication with the external host will work properly without additional tests or investigations. Higher baudrates, however, may be used as long as the actual deviation does not exceed the indicated limit. A certain baudrate (marked I) in Figure 10-2) may e.g. violate the deviation limit, while an even higher baudrate (marked II) in Figure 10-2) stays very well below it. Any baudrate can be used for the bootstrap loader provided that the following three prerequisites are fulfilled: • • • the baudrate is within the specified operating range for the ASC0 the external host is able to use this baudrate the computed deviation error is below the limit. Table 10-1 Bootstrap Loader Baudrate Ranges fSYS [MHz] 10 12 16 20 25 33 BMAX 312,500 375,000 500,000 625,000 781,250 1,031,250 BHigh 9,600 19,200 19,200 19,200 38,400 38,400 BSTDmin 600 600 600 1,200 1,200 1,200 BLow 344 412 550 687 859 1,133 User’s Manual BSL_X, V2.0 10-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) The Bootstrap Loader Note: When the bootstrap loader mode is entered via an internal reset (EA = 1), the default configuration selects bypass mode with factor 2:1 for clock generation. In this case the bootstrap loader will begin to operate with fSYS = fOSC/2 which will limit the maximum baudrate for ASC0 at low input frequencies intended for PLL operation. Higher levels of the bootstrapping sequence can then switch the clock generation mode (via register PLLCON) e.g. to PLL in order to achieve higher baudrates for the download. User’s Manual BSL_X, V2.0 10-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System 11 Debug System 11.1 Introduction The XC167 includes an On-Chip Debug Support (OCDS) system, which provides convenient debugging, controlled directly by an external device via debug interface pins. On-Chip Debug Support (OCDS) The OCDS system supports a broad range of debug features including setting up breakpoints and tracing memory locations. Typical application of OCDS is to debug the user software running on the XC167 in the customer’s system environment. The OCDS system is controlled by an external debugging device via the Debug Interface, including an independent JTAG interface and a break interface (Figure 11-1). The debugger manages the debugging tasks through a set of OCDS registers accessible via the JTAG interface, and through a set of special debug IO instructions. Additionally, the OCDS system can be controlled by the CPU, e.g. by the monitor program. The OCDS system interacts with the core through an injection interface to allow execution of Cerberus-generated instructions, and through a break port. OCDS System break_in Trace Interface Break Interface OCDS Module break_out CPU Debugger JTAG Interface Debug Interface JTAG Module Controller Cerberus (IO Module) Injection Interface CPU Status Other Resources MCA05388 Figure 11-1 OCDS Overall Structure The OCDS system functions are represented and controlled by the Debug Interface, the OCDS Module and by the debug IO control module (Cerberus) which provides all the User’s Manual OCDS_X8, V2.1 11-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System functionality necessary to interact between the debug interface (the external debugger) and the internal system. The OCDS system provides the following basic features: • • • • • • • • Hardware, software and external pin breakpoints Reaction on break with CPU-Halt, monitor call, data transfer and external signal Read/write access to the whole address space Single stepping Debug Interface pins for JTAG interface and break interface Injection of arbitrary instructions Fast memory tracing through transfer to external bus Analysis and status registers 11.2 Debug Interface The Debug Interface is a channel to access XC167 On-Chip Debug Support (OCDS) resources. Through it data can be transferred to/from all on- and off-chip (if any) memories and control registers. Features and Functions • • • • • Independent interface for On-Chip Debug Support (OCDS) JTAG port based on the IEEE 1149 JTAG standard Break interface for external trigger and indication of breaks Generic memory access functionality Independent data transfer channel for e.g. programming of on-chip non volatile memory The Debug Interface is represented by: • • Standard JTAG Interface Two additional XC167 specific signals - OCDS Break-Interface User’s Manual OCDS_X8, V2.1 11-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System JTAG Interface The JTAG interface is a standardized and dedicated port usually used for boundary scan and for chip internal tests. Because both of these applications are not enabled during normal operation of the device in a system, the JTAG port is an ideal interface for debugging tasks. This interface holds the JTAG IEEE.1149-standard signals: • • • • • TDI - Serial data input TDO - Serial data output TCK - JTAG clock TMS - State machine control signal TRST - Reset/Module enable OCDS Break-Interface Two additional signals are used to implement a direct asynchronous-break channel between the Debugger and XC167 OCDS Module: • • BRKIN (BReaK IN request) allows the Debugger asynchronously to interrupt the CPU and force it to a predefined status/action. BRKOUT (BReaK OUT signal) can be activated by OCDS to notify the external world that some predefined debug event has happened, while not interrupting the CPU and using its pin(s). 11.3 OCDS Module The application of OCDS Module is to debug the user software running on the CPU in the customer’s system. This is done with an external debugger, that controls the OCDS Module via the independent Debug Interface. Features • • • • • • • Hardware, software and external pin breakpoints Up to 4 instruction pointer breakpoints Masked comparisons for hardware breakpoints The OCDS can also be configured by a monitor Support of multi CPU/master system Single stepping with monitor or CPU halt PC is visible in halt mode (IO_READ_IP instruction injection via Cerberus) User’s Manual OCDS_X8, V2.1 11-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System Basic Concept The on chip debug concept is split up into two parts. The first part covers the generation of debug events and the second part defines what actions are taken when a debug event is generated. • • Debug events: – Hardware Breakpoints – Decoding of a SBRK Instruction – Break Pin Input activated Debug event actions: – Halt Mode of the CPU – Call a Monitor – Trigger Transfer – Activate External Pin Output Debug Event Sources Hardware Triggers Debug Actions HALT the CPU Programmable Combination Debug Event Processing CALL a Monitor SBRK Instruction Transfer Triggered Break_In Pin Activated Break_Out Pin Activated MCB05389 Figure 11-2 OCDS Concept: Block Diagram User’s Manual OCDS_X8, V2.1 11-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System 11.3.1 Debug Events The Debug Events can come from a few different sources. Hardware Breakpoints The Hardware Breakpoint is a debug-event, raised when a single or a combination of multiple trigger-signals are matching with the programmed conditions. The following hardware trigger sources can be used: Table 11-1 Hardware Triggers Trigger Source Size Task Identifier 16 bits Instruction Pointer 24 bits Data address of reads (two busses monitored) 2 × 24 bits Data address of writes 24 bits Data value (reads or writes) 16 bits SBRK Instruction This is a mechanism through which the software can explicitly generate a debug event. It can be used for instance by a debugger to temporarily patch code held in RAM in order to implement Software Breakpoints. A special SBRK (Software BReaK) instruction is defined with opcode 0x8C00. When this instruction has been decoded and it reaches the Execute stage, the whole pipeline is canceled including the SBRK itself. Hence in fact the SBRK instruction is never “executed” by itself. The further behavior is dependent on how OCDS has been programmed: • • if the OCDS is enabled and the software breakpoints are also enabled, then the CPU goes into Halt Mode if the OCDS is disabled or the software breakpoints are disabled, then the Software Break Trap (SBRKTRAP) is executed-Class A Trap, number 08H Break Pin Input An external debug break pin (BRKIN) is provided to allow the debugger to asynchronously interrupt the processor. User’s Manual OCDS_X8, V2.1 11-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System 11.3.2 Debug Actions When the OCDS is enabled and a debug event is generated, one of the following actions is taken: Trigger Transfer One of the actions that can be specified to occur on a debug event being raised is to trigger the Cerberus: • • to execute a Data Transfer - this can be used in critical routines where the system cannot be interrupted to transfer a memory location to inject an instruction to the Core - using this mechanism, an arbitrary instruction can be injected into the XC167 pipeline Halt Mode Upon this Action the OCDS Module sends a Break-Request to the Core. The Core accepts this request, if the OCDS Break Level is higher than current CPU priority level. In case a Break-Request is accepted, the system suspends execution with halting the instruction flow. The Halt Mode can be still interrupted by higher priority user interrupts. It then relies on the external debugger system to interrogate the target purely through reading and updating via the debug interface. Call a Monitor One of the possible actions to be taken when a debug event is raised is to call a Monitor Program. This short entry to a Monitor allows a flexible debug environment to be defined which is capable of satisfying many of the requirements for efficient debugging of a real time system. In the common case the Monitor has the highest priority and can not be interrupted from any other requesting source. It is also possible to have an Interruptible Monitor Program. In such a case safety critical code can be still served while the Monitor (Debugger) is active, which gives a maximum flexibility to the user. Activate External Pin This action activates the external pin BRKOUT of the OCDS Break-Interface. It can be used in critical routines where the system cannot be interrupted to signal to the external world that a particular event has happened. The feature could also be useful to synchronize the internal and external debug hardware. User’s Manual OCDS_X8, V2.1 11-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System 11.4 Cerberus Cerberus is the module which provides and controls all the operations necessary to interact between the external debugger (via the Debug Interface), the OCDS Module and the internal system of XC167. Features • • • • • • • • • • • JTAG interface is used as control and data channel Generic memory read/write functionality (RW mode) with access to the whole address space Reading and writing of general-purpose registers (GPRs) Injection of arbitrary instructions External host controls all transactions All transactions are available at normal run time and in halt mode Priority of transactions can be configured Full support for communication between the monitor and an external host (debugger) Optional error protection Tracing memory locations through transferring values to the external bus Analysis Register for internal bus locking situations The target application of Cerberus is to use the JTAG interface as an independent port for On Chip Debug Support. The external debugger can access the OCDS registers and arbitrary memory locations with the injection mechanism. 11.4.1 Functional Overview Cerberus is operated by an external debugger across the JTAG Interface. The Debugger supplies Cerberus IO Instructions and performs bidirectional data-transfers. The Cerberus distinguishes between two main modes of operation: Read/Write Mode of Operation Read/Write (RW) Mode is the most typical way to operate Cerberus. This mode is used to read and write memory locations or to inject instructions. The injection interface to the core is actively used in this mode. In this mode an external Debugger (host), using JTAG Interface, can: • • read and write memory locations from the target system (data-transfer); inject arbitrary instructions to be executed by the Core. All Cerberus IO Instructions can be used in RW mode. The dedicated IO_READ_IP instruction is provided in RW mode to read the IP of the CPU while in Break. User’s Manual OCDS_X8, V2.1 11-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Debug System The access to any memory location is performed with injected instructions, as PEC transfer. The following Cerberus IO Instructions can be used in their generic meaning: • • • IO_READ_WORD, IO_WRITE_WORD IO_READ_BLOCK, IO_WRITE_BLOCK IO_WRITE_BYTE Within these instructions, the host writes/reads data to/from a dedicated register/memory, while the Cerberus itself takes care of the rest: to perform a PEC transfer by injection of the appropriate instructions to the Core. Communication Mode of Operation In this mode the external host (debugger) communicates with a program (Monitor) running on the CPU. The data-transfers are made via a PDBus+ register. The external host is master of all transactions, requesting the monitor to write or read a value. The difference to Read/Write Mode of Operation is that the read or write request now is not actively executed by the Cerberus, but it sets request bits in a CPU accessible register to signal the Monitor, that the host wants to send (IO_WRITE_WORD) or receive (IO_READ_WORD) a value. The Monitor has to poll this status register and perform respectively the proper actions Communication Mode is the default mode after reset. Only the IO_WRITE_WORD and IO_READ_WORD Instructions are effectively used in Communication Mode. The Host and the Monitor exchange data directly with the dedicated data-register. For a synchronization of Host (Debugger) and Monitor accesses, there are associated control bits in a Cerberus status register. User’s Manual OCDS_X8, V2.1 11-8 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary 12 Instruction Set Summary This chapter briefly summarizes the XC167’s instructions ordered by instruction classes. This provides a basic understanding of the XC167’s instruction set, the power and versatility of the instructions and their general usage. A detailed description of each single instruction, including its operand data type, condition flag settings, addressing modes, length (number of bytes) and object code format is provided in the “Instruction Set Manual” for the XC166 Family. This manual also provides tables ordering the instructions according to various criteria, to allow quick references. Summary of Instruction Classes Grouping the various instruction into classes aids in identifying similar instructions (e.g. SHR, ROR) and variations of certain instructions (e.g. ADD, ADDB). This provides an easy access to the possibilities and the power of the instructions of the XC167. Note: The used mnemonics refer to the detailed description. Table 12-1 Arithmetic Instructions Addition of two words or bytes: ADD ADDB Addition with Carry of two words or bytes: ADDC ADDCB Subtraction of two words or bytes: SUB SUBB Subtraction with Carry of two words or bytes: SUBC SUBCB 16 × 16 bit signed or unsigned multiplication: MUL MULU 16/16 bit signed or unsigned division: DIV DIVU 32/16 bit signed or unsigned division: DIVL DIVLU 1’s complement of a word or byte: CPL CPLB 2’s complement (negation) of a word or byte: NEG NEGB Bitwise ANDing of two words or bytes: AND ANDB Bitwise ORing of two words or bytes: OR ORB Bitwise XORing of two words or bytes: XOR XORB Table 12-2 Logical Instructions User’s Manual InstructionSummary_X, V2.0 12-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary Table 12-3 Compare and Loop Control Instructions Comparison of two words or bytes: CMP CMPB Comparison of two words with post-increment by either 1 or 2: CMPI1 CMPI2 Comparison of two words with post-decrement by either 1 or 2: CMPD1 CMPD2 Manipulation of a maskable bit field in either the high or the low byte of a word: BFLDH BFLDL Setting a single bit (to ‘1’): BSET – Clearing a single bit (to ‘0’): BCLR – Movement of a single bit: BMOV – Movement of a negated bit: BMOVN – ANDing of two bits: BAND – ORing of two bits: BOR – XORing of two bits: BXOR – Comparison of two bits: BCMP – Shifting right of a word: SHR – Shifting left of a word: SHL – Rotating right of a word: ROR – Rotating left of a word: ROL – Arithmetic shifting right of a word (sign bit shifting): ASHR – Table 12-4 Table 12-5 Table 12-6 Boolean Bit Manipulation Instructions Shift and Rotate Instructions Prioritize Instruction Determination of the number of shift cycles required to PRIOR normalize a word operand (floating point support): User’s Manual InstructionSummary_X, V2.0 12-2 – V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary Table 12-7 Data Movement Instructions Standard data movement of a word or byte: MOV MOVB Data movement of a byte to a word location with either MOVBS sign or zero byte extension: MOVBZ Note: The data movement instructions can be used with a big number of different addressing modes including indirect addressing and automatic pointer in/decrementing. Table 12-8 System Stack Instructions Pushing of a word onto the system stack: PUSH – Popping of a word from the system stack: POP – Saving of a word on the system stack, and then updating the old word with a new value (provided for register bank switching): SCXT – Table 12-9 Jump Instructions Conditional jumping to an either absolutely, indirectly, JMPA or relatively addressed target instruction within the current code segment: JMPI JMPR Unconditional jumping to an absolutely addressed target instruction within any code segment: – – JB Conditional jumping to a relatively addressed target instruction within the current code segment depending on the state of a selectable bit: JNB – JBC Conditional jumping to a relatively addressed target instruction within the current code segment depending on the state of a selectable bit with a post-inversion of the tested bit in case of jump taken (semaphore support): JNBS – User’s Manual InstructionSummary_X, V2.0 12-3 JMPS V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary Table 12-10 Call Instructions Conditional calling of an either absolutely or indirectly CALLA addressed subroutine within the current code segment: CALLI Unconditional calling of a relatively addressed subroutine within the current code segment: CALLR – Unconditional calling of an absolutely addressed subroutine within any code segment: CALLS – Unconditional calling of an absolutely addressed subroutine within the current code segment plus an additional pushing of a selectable register onto the system stack: PCALL – Unconditional branching to the interrupt or trap vector TRAP jump table in code segment <VECSEG>: – Table 12-11 Return Instructions Returning from a subroutine within the current code segment: RET – Returning from a subroutine within any code segment: RETS – Returning from a subroutine within the current code segment plus an additional popping of a selectable register from the system stack: RETP – Returning from an interrupt service routine: RETI – User’s Manual InstructionSummary_X, V2.0 12-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary Table 12-12 System Control Instructions Resetting the XC167 via software: SRST – Entering the Idle mode or Sleep mode: IDLE – Entering the Power Down mode: PWRDN – Servicing the Watchdog Timer: SRVWDT – Disabling the Watchdog Timer: DISWDT – Enabling the Watchdog Timer (can only be executed in WDT enhanced mode): ENWDT – Signifying the end of the initialization routine (pulls pin EINIT RSTOUT high, and disables the effect of any later execution of a DISWDT instruction in WDT compatibility mode): – Table 12-13 Miscellaneous Null operation which requires 2 Bytes of storage and the minimum time for execution: NOP – Definition of an unseparable instruction sequence: ATOMIC – Switch ‘reg’, ‘bitoff’ and ‘bitaddr’ addressing modes to EXTR the Extended SFR space: – Override the DPP addressing scheme using a specific EXTP data page instead of the DPPs, and optionally switch to ESFR space: EXTPR Override the DPP addressing scheme using a specific EXTS segment instead of the DPPs, and optionally switch to ESFR space: EXTSR Note: The ATOMIC and EXT* instructions provide support for uninterruptable code sequences e.g. for semaphore operations. They also support data addressing beyond the limits of the current DPPs (except ATOMIC), which is advantageous for bigger memory models in high level languages. User’s Manual InstructionSummary_X, V2.0 12-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Instruction Set Summary Table 12-14 MAC-Unit Instructions Multiply (and Accumulate): CoMUL CoMAC Add/Subtract: CoADD CoSUB Shift right/Shift left: CoSHR CoSHL Arithmetic Shift right: CoASHR – Load Accumulator: CoLOAD – Store MAC register: CoSTORE – Compare values: CoCMP – Minimum/Maximum: CoMIN CoMAX Absolute value: CoABS – Rounding: CoRND – Move data: CoMOV – Negate accumulator: CoNEG – Null operation: CoNOP – Protected Instructions Some instructions of the XC167 which are critical for the functionality of the controller are implemented as so-called Protected Instructions. These protected instructions use the maximum instruction format of 32 bits for decoding, while the regular instructions only use a part of it (e.g. the lower 8 bits) with the other bits providing additional information like involved registers. Decoding all 32 bits of a protected doubleword instruction increases the security in cases of data distortion during instruction fetching. Critical operations like a software reset are therefore only executed if the complete instruction is decoded without an error. This enhances the safety and reliability of a microcontroller system. User’s Manual InstructionSummary_X, V2.0 12-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Device Specification 13 Device Specification The device specification describes the electrical parameters of the device. It lists DC characteristics like input, output or supply voltages or currents, and AC characteristics like timing characteristics and requirements. Other than the architecture, the instruction set or the basic functions of the XC167 core and its peripherals, these DC and AC characteristics are subject to changes due to device improvements or specific derivatives of the standard device. Therefore these characteristics are not contained in this manual, but rather provided in a separate Data Sheet, which can be updated more frequently. Please refer to the current version of the Data Sheet of the respective device for all electrical parameters. Note: In any case the specific characteristics of a device should be verified, before a new design is started. This ensures that the used information is up to date. Figure 13-1 shows the pin diagram of the XC167. It shows the location of the different supply and IO pins. A detailed description of all the pins is also found in the Data Sheet. Note: Not all alternate functions shown in Figure 13-1 are supported by all derivatives. Please refer to the corresponding descriptions in the data sheets. User’s Manual DeviceSpec_X7, V2.0 13-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) P1L.7/A7/CTRAP/CC22IO P1L.6/A6/COUT63 P1L.5/A5/COUT62 P1L.4/A4/CC62 P1L.3/A3/COUT61 P1L.2/A2/CC61 P1L.1/A1/COUT60 P1L.0/A0/CC60 P0H.7/AD15 P0H.6/AD14 P0H.5/AD13 P0H.4/AD12 P0H.3/AD11 P0H.2/AD10 N.C. N.C. VSSP VDDP P1H.7/A15/CC27IO P1H.6/A14/CC26IO P1H.5/A13/CC25IO P1H.4/A12/CC24IO P1H.3/A11/SCLK1/E*) P1H.2/A10/C6POS2/MTSR1 P1H.1/A9/C6POS1/MRST1 P1H.0/A8/C6POS0/CC23IO/E*) VSSI VDDI XTAL1 XTAL2 144 143 142 141 140 139 138 137 136 135 134 133 132 131 130 129 128 127 126 125 124 123 122 121 120 119 118 117 116 115 114 113 112 111 110 109 VSSI BRKIN BRKOUT RSTIN XTAL4 XTAL3 Device Specification N.C. N.C. P20.12/RSTOUT NMI VSSP VDDP P6.0/CS0/CC0IO P6.1/CS1/CC1IO P6.2/CS2/CC2IO P6.3/CS3/CC3IO P6.4/CS4/CC4IO P6.5/HOLD/CC5IO P6.6/HLDA/CC6IO P6.7/BREQ/CC7IO P7.4/CC28IO/C*) P7.5/CC29IO/C*) P7.6/CC30IO/C*) P7.7/CC31IO/C*) VSSP VDDP P9.0/SDA0/CC16IO/C*) P9.1/SCL0/CC17IO/C*) P9.2/SDA1/CC18IO/C*) P9.3/SCL1/CC19IO/C*) P9.4/SDA2/CC20IO P9.5/SCL2/CC21IO VSSP VDDP 108 107 106 105 104 103 102 101 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 76 75 74 73 XC167 N.C. N.C. P0H.1/AD9 P0H.0/AD8 VSSP VDDP P0L.7/AD7 P0L.6/AD6 P0L.5/AD5 P0L.4/AD4 P0L.3/AD3 P0L.2/AD2 P0L.1/AD1 P0L.0/AD0 P20.5/EA P20.4/ALE P20.2/READY P20.1/WR/WRL P20.0/RD VSSP VDDP P4.7/A23/C*) P4.6/A22/C*) P4.5/A21/C*) P4.4/A20/C*) P4.3/A19 P4.2/A18 P4.1/A17 P4.0/A16 VSSI VDDI P3.15/CLKOUT/FOUT P3.13/SCLK0/E*) P3.12/BHE/WRH/E*) TMS TDO P3.0/T0IN/TxD1/E*) P3.1/T6OUT/RxD1/E*) P3.2/CAPIN P3.3/T3OUT P3.4/T3EUD P3.5/T4IN/TxD1 P3.6/T3IN P3.7/T2IN P3.8/MRST0 P3.9/MTSR0 P3.10/TxD0/E*) P3.11/RxD0/E*) TCK TDI VDDP P2.8/CC8IO/EX0IN P2.9/CC9IO/EX1IN P2.10/CC10IO/EX2IN P2.11/CC11IO/EX3IN P2.12/CC12IO/EX4IN P2.13/CC13IO/EX5IN P2.14/CC14IO/EX6IN P2.15/CC15IO/EX7IN/T7IN TRST VSSI VDDI P5.12/AN12/T6IN P5.13/AN13/T5IN P5.14/AN14/T4EUD P5.15/AN15/T2EUD P5.8/AN8 P5.9/AN9 P5.6/AN6 P5.7/AN7 VAREF VAGND 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 P5.0/AN0 P5.1/AN1 P5.2/AN2 P5.3/AN3 P5.4/AN4 P5.5/AN5 P5.10/AN10/T6EUD P5.11/AN11/T5EUD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 MCP05390_X7 Figure 13-1 Pin Configuration P-TQFP-144 Package (top view) Note: The CAN interface lines can be assigned to the indicated pins (C*)) of Port 4, Port 7, or Port 9. User’s Manual DeviceSpec_X7, V2.0 13-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index Keyword Index This section lists a number of keywords which refer to specific details of the XC167 in terms of its architecture, its functional units or functions. This helps to quickly find the answer to specific questions about the XC167. This User’s Manual consists of two Volumes, “System Units” and “Peripheral Units”. For your convenience this keyword index (and also the table of contents) refers to both volumes, so you can immediately find the reference to the desired section in the corresponding document ([1] or [2]). A BG 19-22 [2] RBUF 19-12 [2], 19-20 [2] TBUF 19-9 [2], 19-20 [2] Transmit FIFO 19-9 [2] ASCx_BG 19-42 [2] ASCx_CON 19-40 [2] ASCx_FDV 19-43 [2] Auto Scan conversion 16-12 [2] Autobaud Detection 19-27 [2] Acronyms 1-9 [1] Adapt Mode 6-21 [1] ADC 2-22 [1], 16-1 [2] ADC_CIC, ADC_EIC 16-21 [2] ADC_CON 16-3 [2] ADC_CON1 16-4 [2] ADC_CTR0 16-5 [2] ADC_CTR2 16-7 [2] ADC_CTR2IN 16-7 [2] Address Boundaries 3-16 [1] Mapping 3-3 [1] Segment 6-19 [1] Addressing Modes CoREG Addressing Mode 4-51 [1] DSP Addressing Modes 4-47 [1] Indirect Addressing Modes 4-45 [1] Long Addressing Modes 4-41 [1] Short Addressing Modes 4-39 [1] Alternate Port Functions 7-8 [1] ALU 4-58 [1] Analog/Digital Converter 16-1 [2] Arbitration of conversions 16-16 [2] ASC 19-1 [2] ASCx_EIC, ASCx_RIC 19-35 [2] ASCx_TIC, ASCx_TBIC 19-35 [2] Autobaud Detection 19-27 [2] Error Detection 19-34 [2] Features and Functions 19-1 [2] IrDA Frames 19-8 [2] Register User’s Manual B BANKSELx 5-33 [1] Baudrate ASC0 19-22 [2] Bootstrap Loader 10-5 [1] CAN 22-56 [2] Bit Handling 4-61 [1] Manipulation Instructions 12-2 [1] protected 2-32 [1], 4-62 [1] reserved 2-16 [1] Block Diagram ITC / PEC 5-3 [1] Bootstrap Loader 6-21 [1], 10-1 [1] Boundaries 3-16 [1] Bus ASC 19-1 [2] CAN 2-25 [1] IIC 2-26 [1], 21-1 [2] Mode Configuration 6-20 [1] SSC 20-1 [2] i-1 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index C CC63SR 18-41 [2] CC6xR 18-18 [2] CC6XSR 18-19 [2] CCU6xIC 18-81 [2] CCxIC 17-34 [2] Chip Select Configuration 6-19 [1] Clock generation 2-29 [1] generator modes 6-18 [1] output signal 6-39 [1] CMPMODIF 18-47 [2] CMPSTAT 18-46 [2] Command sequences (Flash) 3-20 [1] Concatenation of Timers 14-22 [2], 14-45 [2] Configuration Address 6-19 [1] Bus Mode 6-20 [1] Chip Select 6-19 [1] default 6-23 [1] PLL 6-18 [1] Reset 6-14 [1] Reset Output 6-22 [1] special modes 6-21 [1] Write Control 6-20 [1] Context Pointer Updating 4-34 [1] Switch 4-33 [1] Switching 5-32 [1] Conversion analog/digital 16-1 [2] Arbitration 16-16 [2] Auto Scan 16-12 [2] timing control 16-18 [2] Count direction 14-6 [2], 14-35 [2] Counter 14-20 [2], 14-43 [2] Counter Mode (GPT1) 14-10 [2], 14-39 [2] CP 4-36 [1] CPU 2-2 [1], 4-1 [1] CPUCON1 4-26 [1] CPUCON2 4-27 [1] Calibration 16-17 [2] CAN acceptance filtering 22-16 [2] analysing mode 22-7 [2] arbitration 22-16 [2] baudrate 22-56 [2] bit timing 22-9 [2], 22-56 [2] bus off recovery sequence 22-4 [2] status bit 22-51 [2] CAN siehe TwinCAN 22-1 [2] error counters 22-55 [2] error handling 22-11 [2] error warning level 22-55 [2] frame counter/time stamp 22-55 [2], 22-58 [2] Interface 2-25 [1] single data transfer 22-23 [2] CAPCOM 2-18 [1] CAPCOM12 2-16 [1] Capture Mode 17-13 [2] Counter Mode 17-8 [2] CAPREL 14-54 [2] Capture Mode GPT1 14-26 [2] GPT2 (CAPREL) 14-46 [2] Capture/Compare Registers 17-10 [2] CC1_DRM, CC2_DRM 17-23 [2] CC1_IOC, CC2_IOC 17-29 [2] CC1_M0-3 17-10 [2] CC1_OUT, CC2_OUT 17-25 [2] CC1_SEE, CC2_SEE 17-28 [2] CC1_SEM, CC2_SEM 17-27 [2] CC1_T01CON 17-5 [2] CC1_T0IC 17-9 [2] CC1_T1IC 17-9 [2] CC2_M4-7 17-11 [2] CC2_T78CON 17-5 [2] CC2_T7IC 17-9 [2] CC2_T8IC 17-9 [2] CC63R 18-41 [2] User’s Manual i-2 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index Error Detection ASC 19-34 [2] SSC 20-14 [2] EXICON 5-37 [1] EXISEL0 5-38 [1] EXISEL1 5-38 [1] External Bus 2-13 [1] Fast interrupts 5-37 [1] Interrupt pulses 5-40 [1] Interrupt source control 5-37 [1] Interrupts 5-35 [1] Interrupts during sleep mode 5-39 [1] CRIC 14-55 [2] CSP 4-38 [1] D Data Management Unit (Introduction) 2-9 [1] Data Page 4-42 [1] boundaries 3-16 [1] Data SRAM 3-10 [1] Default startup configuration 6-23 [1] Development Support 1-8 [1] Direction count 14-6 [2], 14-35 [2] Disable Interrupt 5-29 [1] Division 4-63 [1] Double-Register Compare 17-22 [2] DP0L, DP0H 7-10 [1] DP1L, DP1H 7-14 [1] DP20 7-82 [1] DP3 7-24 [1], 7-29 [1] DP4 7-41 [1], 22-85 [2] DP6 7-54 [1] DP7 7-65 [1], 22-86 [2] DP9 7-72 [1], 22-88 [2] DPP 4-42 [1] Driver characteristic (ports) 7-4 [1] DSTPx 5-23 [1] Dual-Port RAM 3-10 [1] F Fast external interrupts 5-37 [1] FINT0ADDR 5-16 [1] FINT0CSP 5-17 [1] FINT1ADDR 5-16 [1] FINT1CSP 5-17 [1] Flags 4-57 [1]–4-60 [1] Flash command sequences 3-20 [1] memory 3-12 [1] memory mapping 3-17 [1] waitstates 3-40 [1] FOCON 6-40 [1] Frequency output signal 6-39 [1] FSR 3-33 [1] E G EBC Bus Signals 9-3 [1] Memory Table 9-29 [1] EBCMOD0 9-12 [1] Edge characteristic (ports) 7-5 [1] EMUCON 6-48 [1] Enable Interrupt 5-29 [1] End of PEC Interrupt Sub Node 5-28 [1] EOPIC 5-27 [1] Erase command (Flash) 3-22 [1] Error correction 3-26 [1] User’s Manual Gated timer mode (GPT1) 14-9 [2] Gated timer mode (GPT2) 14-38 [2] GPR 3-6 [1] GPT 2-19 [1] GPT1 14-2 [2] GPT12E_CAPREL 14-54 [2] GPT12E_T2,-T3,-T4 14-29 [2] GPT12E_T2CON 14-15 [2] GPT12E_T2IC,-T3IC,-T4IC 14-30 [2] GPT12E_T3CON 14-4 [2] GPT12E_T4CON 14-15 [2] i-3 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index Pipeline 4-11 [1] protected 12-6 [1] Interface ASC 19-1 [2] CAN 2-25 [1] External Bus 9-1 [1] IIC 2-26 [1], 21-1 [2] SSC 20-1 [2] Interrupt Arbitration 5-4 [1] during sleep mode 5-39 [1] Enable/Disable 5-29 [1] External 5-35 [1] Fast external 5-37 [1] input timing 5-40 [1] Jump Table Cache 5-16 [1] Latency 5-41 [1] Node Sharing 5-34 [1] Priority 5-7 [1] Processing 5-1 [1] RTC 15-12 [2] source control 5-37 [1] Sources 5-12 [1] System 2-8 [1], 5-2 [1] Vectors 5-12 [1] Interrupt Handling CAN transfer 22-6 [2] IP 4-38 [1] IrDA Frames ASC 19-8 [2] IS 18-74 [2] ISR 18-78 [2] ISS 18-77 [2] GPT12E_T5,-T6 14-54 [2] GPT12E_T5CON 14-40 [2] GPT12E_T5IC,-T6IC,-CRIC 14-55 [2] GPT12E_T6CON 14-33 [2] GPT2 14-31 [2] H Hardware Traps 5-43 [1] I I2C 21-1 [2] IDCHIP 6-64 [1] Idle Mode 6-54 [1] IDMANUF 6-64 [1] IDMEM 6-65 [1] IDPROG 6-65 [1] IDX0, IDX1 4-47 [1] IEN 18-79 [2] IIC 21-1 [2] Programming 21-16 [2] Register Overview 21-12 [2] IIC Interface 2-26 [1] IIC_ADR 21-9 [2] IIC_CFG 21-10 [2] IIC_CON 21-5 [2] IIC_DIC 21-18 [2] IIC_PEIC 21-18 [2] IIC_RTBH 21-11 [2] IIC_RTBL 21-11 [2] IIC_ST 21-7 [2] IMB block diagram 3-38 [1] control functions 3-42 [1] memories wait states 3-42 [1] IMBCTR 3-42 [1] Incremental Interface Mode (GPT1) 14-11 [2] Indication of reset source 6-46 [1] INP 18-80 [2] Instruction 12-1 [1] Bit Manipulation 12-2 [1] User’s Manual L Latency Interrupt, PEC 5-41 [1] LXBus 2-13 [1] M MAH, MAL 4-69 [1] MAR 3-27 [1] Margin check 3-26 [1] Master mode i-4 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index P IIC Bus 21-12 [2] MCMCTR 18-62 [2] MCMOUT 18-60 [2] MCMOUTS 18-59 [2] MCW 4-66 [1] MDC 4-64 [1] MDH 4-63 [1] MDL 4-64 [1] Memory 2-10 [1] Areas (Data) 3-9 [1] Areas (Program) 3-11 [1] DPRAM 3-10 [1] DSRAM 3-10 [1] External 3-15 [1] Flash 3-12 [1] Program Flash 3-17 [1] PSRAM 3-12 [1] MODCTR 18-50 [2] MRW 4-72 [1] MSW 4-70 [1] Multimaster mode IIC Bus 21-12 [2] Multiplication 4-63 [1] P0L, P0H 7-9 [1] P1L, P1H 7-13 [1] P3 7-24 [1], 7-29 [1] P4 7-41 [1] P5 7-50 [1], 7-51 [1] P8 7-54 [1], 7-65 [1], 7-72 [1], 7-82 [1] PEC 2-10 [1], 5-18 [1] Latency 5-41 [1] Transfer Count 5-19 [1] PEC pointers 3-8 [1] PECCx 5-19 [1] PECISNC 5-27 [1] PECSEGx 5-23 [1] Peripheral Event Controller --> PEC 5-18 [1] Register Set 23-1 [2] Summary 2-14 [1] Phase Locked Loop (->PLL) 6-26 [1] PICON 7-2 [1] Pins 8-1 [1] Pipeline 4-11 [1] PLL 6-18 [1], 6-26 [1] PLL_IC 6-38 [1] PLLCON 6-32 [1] POCON* 7-6 [1] Port 2-27 [1] Ports Alternate Port Functions 7-8 [1] Driver characteristic 7-4 [1] Edge characteristic 7-5 [1] Power Management 2-29 [1], 6-54 [1] PROCON 3-29 [1] Program Management Unit (Introduction) 2-9 [1] Programming command (Flash) 3-22 [1] Protected Bits 2-32 [1], 4-62 [1] instruction 12-6 [1] Protection commands (Flash) 3-24 [1] features (Flash) 3-28 [1] N NMI 5-1 [1], 5-48 [1] Noise filter (Ext. Interrupts) 5-39 [1] O OCDS Requests 5-40 [1] ODP3 7-25 [1], 7-30 [1] ODP4 7-42 [1] ODP6 7-55 [1] ODP7 7-66 [1] ODP9 7-73 [1] ONES 4-74 [1] Open Drain Mode 7-3 [1] OPSEN 6-49 [1] Oscillator circuitry 6-27 [1], 6-29 [1] measurement 6-27 [1], 6-29 [1] Watchdog 6-22 [1], 6-38 [1] User’s Manual i-5 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index boundaries 3-16 [1] Segmentation 4-37 [1] Self-calibration 16-17 [2] Serial Interface 2-23 [1], 2-24 [1] ASC 19-1 [2] Asynchronous 19-5 [2] CAN 2-25 [1] IIC 2-26 [1], 21-1 [2] SSC 20-1 [2] Synchronous 19-19 [2] SFR 3-5 [1] Sharing Interrupt Nodes 5-34 [1] Slave mode IIC Bus 21-13 [2] Sleep mode 6-56 [1] Software Traps 5-43 [1] Source Interrupt 5-12 [1] Reset 6-46 [1] SP 4-54 [1] Special operation modes (config.) 6-21 [1] SPSEG 4-54 [1] SRAM Data 3-10 [1] SRCPx 5-23 [1] SSC 20-1 [2] Baudrate generation 20-12 [2] Block diagram 20-3 [2] Continous transfer operation 20-12 [2] Error detection 20-14 [2] Full duplex operation 20-8 [2] General Operation 20-1 [2] Half duplex operation 20-11 [2] Interrupts 20-14 [2] SSCx_CON 20-4 [2], 20-5 [2] Stack 3-13 [1], 4-53 [1] Startup Configuration 6-14 [1] STKOV 4-56 [1] STKUN 4-56 [1] SYSCON0 6-43 [1] SYSCON1 6-44 [1] PSLR 18-70 [2] PSW 4-57 [1] Q QR0 4-46 [1] QR1 4-46 [1] QX0, QX1 4-48 [1] R RAM data SRAM 3-10 [1] dual ported 3-10 [1] program/data 3-12 [1] status after reset 6-7 [1] Real Time Clock (->RTC) 2-21 [1], 15-1 [2] Register Areas 3-4 [1] Register map TwinCAN module 22-47 [2] Register Table LXBUS Peripherals 23-16 [2] PD+BUS Peripherals 23-1 [2] RELH, RELL 15-9 [2] Reserved bits 2-16 [1] Reset 6-2 [1] Configuration 6-14 [1] Output 6-9 [1] Source indication 6-46 [1] Values 6-6 [1] RSTCFG 6-16 [1] RSTCON 6-24 [1] RTC 2-21 [1], 15-1 [2] RTC_CON 15-5 [2] RTC_IC 15-13 [2] RTC_ISNC 15-13 [2] RTCH, RTCL 15-8 [2] S SCUSLC 6-52 [1] SCUSLS 6-51 [1] Security features (Flash) 3-28 [1] Segment Address 6-19 [1] User’s Manual i-6 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index TwinCAN FIFO base object 22-24 [2] circular buffer 22-26 [2] configuration 22-73 [2] for CAN messages 22-24 [2] gateway control 22-73 [2] slave objects 22-26 [2] frames counter 22-8 [2] handling 22-17 [2] gateway configuration 22-73 [2] normal mode 22-29 [2] shared mode 22-36 [2] with FIFO 22-33 [2] initialization 22-40 [2] interrupts indication/INTID 22-13 [2], 22-53 [2] node pointer/request compressor 22-5 [2] loop-back mode 22-44 [2] message handling 22-15 [2] FIFO 22-24 [2] gateway overview 22-28 [2] gateway+FIFO 22-33 [2] normal gateway 22-29 [2] shared gateway 22-36 [2] transfer control 22-41 [2] message interrupts 22-13 [2] message object configuration 22-71 [2] control bits 22-68 [2] interrupt indication 22-13 [2] interrupts 22-13 [2] register description 22-64 [2] transfer handling 22-17 [2] node control 22-7 [2] node interrupts 22-11 [2], 22-12 [2] node selection 22-71 [2] overview 22-1 [2] register map 22-47 [2] SYSCON3 6-58 [1] SYSSTAT 6-45 [1] T T0IC 17-9 [2] T12 18-6 [2] T12DTC 18-24 [2] T12MSEL 18-20 [2] T12PR 18-6 [2] T13 18-32 [2] T13PR 18-32 [2] T1IC 17-9 [2] T2, T3, T4 14-29 [2] T2CON 14-15 [2] T2IC, T3IC, T4IC 14-30 [2] T3CON 14-4 [2] T4CON 14-15 [2] T5, T6 14-54 [2] T5CON 14-40 [2] T5IC, T6IC 14-55 [2] T6CON 14-33 [2] T7IC 17-9 [2] T8IC 17-9 [2] TCTR0 18-42 [2] TCTR2 18-44 [2] TCTR4 18-45 [2] TFR 5-45 [1] Timer 14-2 [2], 14-31 [2] Auxiliary Timer 14-15 [2], 14-40 [2] Concatenation 14-22 [2], 14-45 [2] Core Timer 14-4 [2], 14-33 [2] Counter Mode (GPT1) 14-10 [2], 14-39 [2] Gated Mode (GPT1) 14-9 [2] Gated Mode (GPT2) 14-38 [2] Incremental Interface Mode (GPT1) 14-11 [2] Mode (GPT1) 14-8 [2] Mode (GPT2) 14-37 [2] Tools 1-8 [1] Transmit FIFO ASC 19-9 [2] Traps 5-43 [1] TRPCTR 18-65 [2] User’s Manual i-7 V1.0, 2004-06 XC167-32 Derivatives System Units (Vol. 1 of 2) Keyword Index BGINP 22-61 [2] BIMR0H 22-62 [2] BIMR0L 22-62 [2] BIMR4 22-63 [2] BIR 22-53 [2] BSR 22-51 [2] MSGAMRHn 22-67 [2] MSGAMRLn 22-67 [2] MSGARHn 22-66 [2] MSGARLn 22-66 [2] MSGCFGHn 22-71 [2] MSGCFGLn 22-71 [2] MSGCTRHn 22-68 [2] MSGCTRLn 22-68 [2] MSGDRH0 22-64 [2] MSGDRH4 22-65 [2] MSGDRL0 22-64 [2] MSGDRL4 22-65 [2] MSGFGCRHn 22-74 [2] MSGFGCRLn 22-74 [2] RXIPNDH 22-80 [2] RXIPNDL 22-80 [2] TXIPNDH 22-81 [2] TXIPNDL 22-81 [2] registers (global) receive interrupt pending 22-80 [2] transmit interrupt pending 22-81 [2] registers (message specific) acceptance mask 22-67 [2] arbitration (identifier) 22-66 [2] configuration 22-71 [2] control 22-68 [2] data 22-64 [2] registers (node specific) bit timing 22-56 [2] control 22-49 [2] error counter 22-55 [2] frame counter 22-58 [2] global interrupt node pointer 22-61 [2] interrupt pending 22-53 [2] INTID mask 22-62 [2] status 22-51 [2] single transmission 22-45 [2] single-shot mode 22-23 [2] transfer interrupts 22-6 [2] TwinCAN Registers (short names) ABTRH 22-56 [2] ABTRL 22-56 [2] ACR 22-49 [2] AECNTH 22-54 [2] AECNTL 22-54 [2] AFCRH 22-58 [2] AFCRL 22-58 [2] AGINP 22-61 [2] AIMR0H 22-62 [2] AIMR0L 22-62 [2] AIMR4 22-63 [2] AIR 22-53 [2] ASR 22-51 [2] BBTRH 22-56 [2] BBTRL 22-56 [2] BCR 22-49 [2] BECNTH 22-54 [2] BECNTL 22-54 [2] BFCRH 22-58 [2] BFCRL 22-58 [2] User’s Manual V VECSEG 5-11 [1] W Waitstates Flash 3-40 [1] Watchdog 2-26 [1], 6-59 [1] after reset 6-7 [1] Oscillator 6-22 [1], 6-38 [1] WDT 6-60 [1] WDTCON 6-62 [1] Z ZEROS 4-74 [1] i-8 V1.0, 2004-06 w w w . i n f i n e o n . c o m Published by Infineon Technologies AG