U s e r ’ s M a n u a l , V 1 . 2, Se p . 2 0 0 8 XC27x5X Derivatives 1 6 / 3 2 - B i t S i n g l e -C h i p M i c r o c o n t r o l l e r w i t h 32-Bit Performance V o l u m e 1 ( o f 2 ) : S y s t e m U n i ts M i c r o c o n t r o l l e rs Edition 2008-09 Published by Infineon Technologies AG 81726 Munich, Germany © 2008 Infineon Technologies AG All Rights Reserved. Legal Disclaimer The information given in this document shall in no event be regarded as a guarantee of conditions or characteristics. With respect to any examples or hints given herein, any typical values stated herein and/or any information regarding the application of the device, Infineon Technologies hereby disclaims any and all warranties and liabilities of any kind, including without limitation, warranties of non-infringement of intellectual property rights of any third party. Information For further information on technology, delivery terms and conditions and prices, please contact the 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 the nearest Infineon Technologies Office. Infineon Technologies components may be used in life-support devices or systems only 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 . 2, Se p . 2 0 0 8 XC27x5X Derivatives 1 6 / 3 2 - B i t S i n g l e -C h i p M i c r o c o n t r o l l e r w i t h 32-Bit Performance V o l u m e 1 ( o f 2 ) : S y s t e m U n i ts M i c r o c o n t r o l l e rs XC27x5X Derivatives System Units (Vol. 1 of 2) XC27x5X Revision History: V1.2, 2008-09 Previous Version(s): V1.0, 2008-06 (V1.1 skipped) Page Subjects (major changes since last revision) 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] User’s Manual V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Summary Of Chapters Summary Of Chapters This User’s Manual consists of two Volumes, “System Units” and “Peripheral Units”. For a quick overview this table of chapters summarizes both volumes, so you immediately can find the reference to the desired section in the corresponding document ([1] or [2]). Summary Of Chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0-1 [1] Table Of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0-3 [1] 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 [1] 2 Architectural Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 [1] 3 Memory Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 [1] 4 Memory Checker Module (MCHK) . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 [1] 5 Central Processing Unit (CPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 [1] 6 Memory Protection Unit (MPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 [1] 7 Interrupt and Exception Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 [1] 8 System Control Unit (SCU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 [1] 9 Parallel Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1 [1] 10 Dedicated Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 [1] 11 External Bus Controller (EBC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 [1] 12 Startup Configuration and Bootstrap Loading . . . . . . . . . . . . . . . 12-1 [1] 13 Debug System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1 [1] 14 Instruction Set Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1 [1] 15 Device Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1 [1] 16 The General Purpose Timer Units . . . . . . . . . . . . . . . . . . . . . . . . . 16-1 [2] 17 Real Time Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1 [2] 18 Analog to Digital Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-1 [2] 19 Capture/Compare Unit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-1 [2] 20 Capture/Compare Unit 6 (CCU6) . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 [2] 21 Universal Serial Interface Channel . . . . . . . . . . . . . . . . . . . . . . . . . 21-1 [2] 22 Controller Area Network (MultiCAN) Controller . . . . . . . . . . . . . . 22-1 [2] User’s Manual L-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Summary Of Chapters 23 Appendix: Functional and Operational Updates . . . . . . . . . . . . . 23-1 [2] Keyword Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-1 [2] Register Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-8 [2] User’s Manual L-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents Table Of Contents This User’s Manual consists of two Volumes, “System Units” and “Peripheral Units”. For your convenience this table of contents (and also the keyword and register index) lists both volumes, so you can immediately find the reference to the desired section in the corresponding document ([1] or [2]). Summary Of Chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0-1 [1] Table Of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0-3 [1] 1 1.1 1.2 1.3 1.4 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 [1] Members of the 16/32-bit Microcontroller Families . . . . . . . . . . . . . . . 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 2.1.7 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] Memory Protection Unit (MPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 [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 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-15 [1] Clock Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-33 [1] Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-34 [1] On-Chip Debug Support (OCDS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-35 [1] Protected Bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-36 [1] 3 3.1 3.2 3.3 3.4 3.4.1 3.4.2 3.5 3.6 3.7 3.8 Memory Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 [1] Address Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 [1] Register Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5 [1] Data Memory Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10 [1] Program Memory Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12 [1] Program/Data SRAM (PSRAM) . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13 [1] Non-Volatile Program Memory (Flash) . . . . . . . . . . . . . . . . . . . . . 3-14 [1] System Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-15 [1] IO Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-16 [1] External Memory Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-17 [1] Crossing Memory Boundaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-17 [1] User’s Manual L-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 3.9 3.9.1 3.9.2 3.9.3 3.9.4 3.9.5 3.9.6 3.9.7 3.9.8 3.9.9 3.9.10 3.9.11 3.9.12 3.10 3.10.1 3.10.2 3.10.3 3.11 Embedded Flash Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Details of Command Sequences . . . . . . . . . . . . . . . . . . . . . . . . . . Sequence Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instructions for Executing Program and Erase Jobs Concurrently Data Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protection Handling Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protection Handling Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . EEPROM Emulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recommendations for Optimized Flash Usage . . . . . . . . . . . . . . . On-Chip Program Memory Control . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Error Reporting Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Retention Memories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 4.1 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.2.7 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.4.1 4.5 Memory Checker Module (MCHK) . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 [1] Operational Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 [1] Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 [1] Principle of the LFSR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4 [1] Principle of the MISR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-6 [1] Commonly used Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7 [1] Architecture of the Memory Checker Module . . . . . . . . . . . . . . . . . 4-8 [1] Preferable Usage of the Memory Checker Module . . . . . . . . . . . . 4-10 [1] Calculation of Seed Values (Magic Word) . . . . . . . . . . . . . . . . . . . 4-10 [1] Example Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-12 [1] Memory Checker Module Registers . . . . . . . . . . . . . . . . . . . . . . . . . 4-14 [1] Memory Checker Input Register . . . . . . . . . . . . . . . . . . . . . . . . . . 4-15 [1] Memory Checker Result Registers . . . . . . . . . . . . . . . . . . . . . . . . 4-16 [1] Memory Checker Count Register . . . . . . . . . . . . . . . . . . . . . . . . . 4-17 [1] Memory Checker Polynomial Registers . . . . . . . . . . . . . . . . . . . . . 4-18 [1] General Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20 [1] ID Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20 [1] Interfaces of the MCHK Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20 [1] 5 5.1 5.2 5.2.1 5.2.2 5.2.3 Central Processing Unit (CPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Components of the CPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instruction Fetch and Program Flow Control . . . . . . . . . . . . . . . . . . . . Branch Detection and Branch Prediction Rules . . . . . . . . . . . . . . . . Zero-Cycle Jumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atomic and Extend Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . User’s Manual L-4 3-19 [1] 3-19 [1] 3-21 [1] 3-23 [1] 3-26 [1] 3-36 [1] 3-37 [1] 3-40 [1] 3-44 [1] 3-51 [1] 3-53 [1] 3-55 [1] 3-55 [1] 3-57 [1] 3-57 [1] 3-59 [1] 3-74 [1] 3-76 [1] 5-1 [1] 5-4 [1] 5-5 [1] 5-7 [1] 5-7 [1] 5-8 [1] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 5.3 5.3.1 5.3.2 5.4 5.5 5.5.1 5.5.2 5.6 5.7 5.7.1 5.7.2 5.7.3 5.7.4 5.7.5 5.8 5.8.1 5.8.2 5.8.3 5.9 5.9.1 5.9.2 5.9.3 5.9.4 5.9.5 5.9.6 5.9.7 5.9.8 5.9.9 5.9.10 5.9.11 5.10 Instruction Processing Pipeline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9 [1] Access to the IO Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10 [1] Pipeline Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-10 [1] CPU Configuration Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-21 [1] Use of General Purpose Registers . . . . . . . . . . . . . . . . . . . . . . . . . . 5-24 [1] GPR Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-26 [1] Context Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-28 [1] Code Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-33 [1] Data Addressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-36 [1] Short Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-36 [1] Long Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-38 [1] Indirect Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-41 [1] DSP Addressing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-43 [1] The System Stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-49 [1] Standard Data Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-53 [1] 16-bit Adder/Subtracter, Barrel Shifter, and 16-bit Logic Unit . . . . 5-57 [1] Bit Manipulation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-58 [1] Multiply and Divide Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-60 [1] DSP Data Processing (MAC Unit) . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-62 [1] MAC Unit Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-63 [1] Representation of Numbers and Rounding . . . . . . . . . . . . . . . . . . 5-63 [1] The 16-bit by 16-bit Signed/Unsigned Multiplier and Scaler . . . . . 5-64 [1] Concatenation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-64 [1] One-bit Scaler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-64 [1] The 40-bit Adder/Subtracter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-64 [1] The Data Limiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-65 [1] The Accumulator Shifter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-65 [1] The 40-bit Signed Accumulator Register . . . . . . . . . . . . . . . . . . . . 5-66 [1] The MAC Unit Status Word MSW . . . . . . . . . . . . . . . . . . . . . . . . . 5-68 [1] The Repeat Counter MRW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-70 [1] Constant Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-72 [1] 6 6.1 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.3 6.3.1 6.3.2 6.3.3 Memory Protection Unit (MPU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 [1] Functional Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 [1] Memory Protection Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3 [1] Protection Range Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3 [1] Protection Mode Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-5 [1] Protection Range Data Register . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-8 [1] Protection Range Address Register . . . . . . . . . . . . . . . . . . . . . . . . 6-8 [1] Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10 [1] Enabling Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10 [1] Protection Levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10 [1] Intersecting Memory Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-11 [1] User’s Manual L-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 6.3.4 6.3.5 6.3.6 6.3.7 6.3.8 6.3.9 6.3.10 6.4 6.4.1 6.4.2 6.4.3 6.4.4 6.4.5 6.4.6 Protection of the MPU registers . . . . . . . . . . . . . . . . . . . . . . . . . . . Accessing SFRs and GPRs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupts and PECs Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . Special handling of RETI instruction . . . . . . . . . . . . . . . . . . . . . . . Context Switch operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debugger Access Permissions . . . . . . . . . . . . . . . . . . . . . . . . . . . Invalid Access Traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Initializing and using the MPU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Installing Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Changing Protection Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Executing privileged code from non-privileged one . . . . . . . . . . . . Fast task switches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Bank Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debugger Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 7.1 7.2 7.3 7.3.1 7.3.2 7.3.3 7.3.4 7.4 7.5 7.6 7.7 7.8 7.9 7.9.1 7.9.2 7.10 7.10.1 7.10.2 7.10.3 7.10.4 7.11 7.11.1 7.11.2 7.12 7.13 7.14 7.14.1 Interrupt and Exception Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 [1] Interrupt System Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3 [1] Interrupt Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-5 [1] Interrupt Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-8 [1] Interrupt Priority Level and Group Level . . . . . . . . . . . . . . . . . . . . . 7-9 [1] General Interrupt Control Functions in Register PSW . . . . . . . . . . 7-10 [1] Selective Interrupt Disabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-11 [1] Interrupt Class Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12 [1] Interrupt Vector Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-14 [1] Interrupt Jump Table Cache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-16 [1] CPU Status Saving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19 [1] CPU Context Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-20 [1] Fast Bank Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-21 [1] Trap Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23 [1] Software Traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23 [1] Hardware Traps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-24 [1] Peripheral Event Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31 [1] PEC Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-32 [1] The PEC Source and Destination Pointer . . . . . . . . . . . . . . . . . . . 7-40 [1] PEC Interrupt Processing Summary . . . . . . . . . . . . . . . . . . . . . . . 7-42 [1] PEC Channel Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-43 [1] External Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45 [1] External Request Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45 [1] Using Peripheral Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-45 [1] OCDS Requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-47 [1] Service Request Latency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-48 [1] Interrupt Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-50 [1] Physical Interrupt Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-50 [1] User’s Manual L-6 6-11 [1] 6-12 [1] 6-12 [1] 6-13 [1] 6-13 [1] 6-14 [1] 6-14 [1] 6-15 [1] 6-15 [1] 6-16 [1] 6-17 [1] 6-17 [1] 6-17 [1] 6-17 [1] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 7.14.2 7.14.3 7.15 Interrupt Node Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-54 [1] Interrupt Source Select Registers . . . . . . . . . . . . . . . . . . . . . . . . . 7-56 [1] Interrupt and PEC Configuration Registers . . . . . . . . . . . . . . . . . . . . 7-58 [1] 8 8.1 8.1.1 8.1.2 8.1.3 8.1.4 8.1.5 8.1.6 8.1.7 8.2 8.2.1 8.3 8.3.1 8.3.2 8.4 8.4.1 8.4.2 8.4.3 8.4.4 8.4.5 8.4.6 8.4.7 8.5 8.5.1 8.5.2 8.5.3 8.6 8.6.1 8.6.2 8.6.3 8.6.4 8.7 8.7.1 8.7.2 8.8 8.8.1 8.9 8.9.1 8.9.2 System Control Unit (SCU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 [1] Clock Generation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2 [1] Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2 [1] Trimmed Current Controlled Wake-Up Clock (OSC_WU) . . . . . . . . 8-4 [1] High Precision Oscillator Circuit (OSC_HP) . . . . . . . . . . . . . . . . . . 8-4 [1] Phase-Locked Loop (PLL) Module . . . . . . . . . . . . . . . . . . . . . . . . . 8-6 [1] Clock Control Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-16 [1] External Clock Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-22 [1] CGU Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-25 [1] System Timer Function (STM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-46 [1] STM Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-47 [1] Wake-up Timer (WUT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-49 [1] Wake-up Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-49 [1] WUT Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-51 [1] Reset Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-54 [1] Reset Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-54 [1] General Reset Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-57 [1] Debug Reset Assertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-59 [1] Coupling of Reset Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-59 [1] Reset Request Trigger Sources . . . . . . . . . . . . . . . . . . . . . . . . . . 8-60 [1] Module Reset Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-62 [1] Reset Controller Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-64 [1] External Service Request (ESR) Pins . . . . . . . . . . . . . . . . . . . . . . . . 8-74 [1] General Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-74 [1] ESR Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-78 [1] ESR Data Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-83 [1] Power Supply and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-84 [1] Supply Watchdog (SWD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-86 [1] Monitoring the Voltage Level of a Core Domain . . . . . . . . . . . . . . 8-93 [1] Controlling the Voltage Level of a Core Domain . . . . . . . . . . . . . 8-100 [1] Handling the Power System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-109 [1] Global State Controller (GSC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-110 [1] GSC Control Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-110 [1] GSC Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-114 [1] Software Boot Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-128 [1] Start-up Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-128 [1] External Request Unit (ERU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-129 [1] Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-129 [1] ERU Input Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-131 [1] User’s Manual L-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 8.9.3 8.9.4 8.9.5 8.9.6 8.9.7 8.9.8 8.10 8.10.1 8.10.2 8.10.3 8.11 8.11.1 8.12 8.12.1 8.12.2 8.12.3 8.12.4 8.13 8.13.1 8.13.2 8.13.3 8.14 8.14.1 8.14.2 8.14.3 8.15 8.15.1 8.15.2 8.16 8.16.1 8.16.2 8.16.3 8.17 8.18 8.18.1 8.18.2 8.18.3 External Request Select Unit (ERSx) . . . . . . . . . . . . . . . . . . . . . Event Trigger Logic (ETLx) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Connecting Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Output Gating Unit (OGUy) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ERU Output Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ERU Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Interrupt Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature Compensation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature Compensation Registers . . . . . . . . . . . . . . . . . . . . Watchdog Timer (WDT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . WDT Kernel Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Trap Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Trap Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Trap Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Trap Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . Memory Content Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Memory Checking Control Register . . . . . . . . . . . . . . . . . . . . . . . Parity Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ECC Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Access Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Protection Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . Miscellaneous System Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . System Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identification Block . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marker Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SCU Register Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clock Generation Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Service Requests (ESR) . . . . . . . . . . . . . . . . . . . . . . . . External Request Unit (ERU) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 9.1 9.1.1 9.1.2 9.1.3 Parallel Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basic Port Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Input Stage Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Output Driver Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User’s Manual L-8 8-132 [1] 8-133 [1] 8-135 [1] 8-136 [1] 8-139 [1] 8-140 [1] 8-147 [1] 8-148 [1] 8-148 [1] 8-149 [1] 8-169 [1] 8-171 [1] 8-173 [1] 8-173 [1] 8-173 [1] 8-174 [1] 8-178 [1] 8-182 [1] 8-182 [1] 8-183 [1] 8-184 [1] 8-196 [1] 8-197 [1] 8-198 [1] 8-207 [1] 8-211 [1] 8-211 [1] 8-214 [1] 8-216 [1] 8-216 [1] 8-217 [1] 8-221 [1] 8-222 [1] 8-228 [1] 8-228 [1] 8-229 [1] 8-230 [1] 9-1 [1] 9-2 [1] 9-2 [1] 9-5 [1] 9-5 [1] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 9.2 9.2.1 9.2.2 9.2.3 9.2.4 9.2.5 9.2.6 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.3.5 9.3.6 9.3.7 9.3.8 9.3.9 9.3.10 9.3.11 9.3.12 9.3.13 9.4 Port Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-6 [1] Pad Driver Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-6 [1] Port Output Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-9 [1] Port Output Modification Register . . . . . . . . . . . . . . . . . . . . . . . . . 9-10 [1] Port Input Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-12 [1] Port Input/Output Control Registers . . . . . . . . . . . . . . . . . . . . . . . . 9-13 [1] Port Digital Input Disable Register . . . . . . . . . . . . . . . . . . . . . . . . . 9-16 [1] Port Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17 [1] Port 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18 [1] Port 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-19 [1] Port 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-20 [1] Port 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-22 [1] Port 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-23 [1] Port 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-24 [1] Port 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-25 [1] Port 7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-26 [1] Port 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-27 [1] Port 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-28 [1] Port 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-29 [1] Port 11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-31 [1] Port 15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-32 [1] Pin Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-33 [1] 10 Dedicated Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 [1] 11 11.1 11.2 11.3 11.4 11.4.1 11.4.2 11.4.3 11.4.4 11.5 11.5.1 11.5.2 11.6 11.6.1 11.6.2 11.6.3 11.7 11.7.1 11.7.2 External Bus Controller (EBC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 [1] Summary of Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 [1] Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 [1] Naming Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-2 [1] Timing Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-2 [1] Bus Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-2 [1] Demultiplexed Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-5 [1] Multiplexed Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-6 [1] Fastest Access Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-7 [1] Address Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-9 [1] Window Allocation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-9 [1] Window Overlap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-10 [1] Ready Controlled Bus Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-11 [1] Enabling the Ready Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-11 [1] Synchronous and Asynchronous READY . . . . . . . . . . . . . . . . . . 11-11 [1] Combining the READY function with predefined wait states . . . . 11-12 [1] External Bus Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-13 [1] Initialization of Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-13 [1] Arbitration Master Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-13 [1] User’s Manual L-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 11.7.3 11.7.4 11.7.5 11.8 11.9 11.9.1 11.9.2 11.9.3 11.9.4 11.10 11.10.1 11.10.2 11.10.3 11.10.4 Arbitration Slave Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bus Lock Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Direct Master Slave Connection . . . . . . . . . . . . . . . . . . . . . . . . . EBC Idle State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . EBC Mode Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Timing Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Function Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Address Window Selection Registers . . . . . . . . . . . . . . . . . . . . . EBC Implementation in XC27x5X . . . . . . . . . . . . . . . . . . . . . . . . . . Unused Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Access Control to LXBUS Modules . . . . . . . . . . . . . . . . . . . . . . . Shutdown Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dedicated Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 12.1 12.2 12.2.1 12.2.2 12.2.3 12.2.4 12.3 12.3.1 12.3.2 12.3.3 12.4 12.5 12.5.1 12.6 12.6.1 12.6.2 12.6.3 12.6.4 12.6.5 Startup Configuration and Bootstrap Loading . . . . . . . . . . . . . . . 12-1 [1] Startup Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1 [1] Device Status after Startup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-2 [1] Registers modified by the Startup Procedure . . . . . . . . . . . . . . . . 12-2 [1] System Frequency after Startup . . . . . . . . . . . . . . . . . . . . . . . . . . 12-3 [1] Watchdog Timer handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-4 [1] Startup Error state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-5 [1] Supported Startup Modes and Options . . . . . . . . . . . . . . . . . . . . . . . 12-6 [1] Basic Startup Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-7 [1] Startup Modes with Debug Support . . . . . . . . . . . . . . . . . . . . . . . . 12-8 [1] Special Startup Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-11 [1] Internal Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-15 [1] External Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-15 [1] Specific Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-17 [1] Bootstrap Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-18 [1] General Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-18 [1] Bootstrap Loaders using UART Protocol . . . . . . . . . . . . . . . . . . . 12-20 [1] Synchronous Serial Channel Bootstrap Loader . . . . . . . . . . . . . . 12-27 [1] CAN Bootstrap Loader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-30 [1] Summary of Bootstrap Loader Modes . . . . . . . . . . . . . . . . . . . . . 12-33 [1] 13 13.1 13.1.1 13.2 13.2.1 13.2.2 13.3 13.3.1 Debug System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Routing of Debug Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OCDS Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debug Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cerberus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functional Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User’s Manual L-10 11-15 [1] 11-15 [1] 11-16 [1] 11-17 [1] 11-18 [1] 11-18 [1] 11-20 [1] 11-22 [1] 11-23 [1] 11-24 [1] 11-24 [1] 11-24 [1] 11-24 [1] 11-25 [1] 13-1 [1] 13-2 [1] 13-3 [1] 13-5 [1] 13-6 [1] 13-7 [1] 13-8 [1] 13-9 [1] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 13.4 13.4.1 13.4.2 13.5 Emulation Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MCDS Use Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MCDS Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Boundary-Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Instruction Set Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1 [1] 15 Device Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1 [1] 16 16.1 16.1.1 16.1.2 16.1.3 16.1.4 16.1.5 16.1.6 16.1.7 16.2 16.2.1 16.2.2 16.2.3 16.2.4 16.2.5 16.2.6 16.2.7 16.2.8 16.3 16.4 The General Purpose Timer Units . . . . . . . . . . . . . . . . . . . . . . . . . 16-1 [2] Timer Block GPT1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-2 [2] GPT1 Core Timer T3 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-4 [2] GPT1 Core Timer T3 Operating Modes . . . . . . . . . . . . . . . . . . . . . 16-8 [2] GPT1 Auxiliary Timers T2/T4 Control . . . . . . . . . . . . . . . . . . . . . 16-15 [2] GPT1 Auxiliary Timers T2/T4 Operating Modes . . . . . . . . . . . . . 16-19 [2] GPT1 Clock Signal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-28 [2] GPT1 Timer Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-31 [2] Interrupt Control for GPT1 Timers . . . . . . . . . . . . . . . . . . . . . . . . 16-32 [2] Timer Block GPT2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-33 [2] GPT2 Core Timer T6 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-35 [2] GPT2 Core Timer T6 Operating Modes . . . . . . . . . . . . . . . . . . . . 16-39 [2] GPT2 Auxiliary Timer T5 Control . . . . . . . . . . . . . . . . . . . . . . . . 16-42 [2] GPT2 Auxiliary Timer T5 Operating Modes . . . . . . . . . . . . . . . . . 16-45 [2] GPT2 Register CAPREL Operating Modes . . . . . . . . . . . . . . . . . 16-49 [2] GPT2 Clock Signal Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-55 [2] GPT2 Timer Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-58 [2] Interrupt Control for GPT2 Timers and CAPREL . . . . . . . . . . . . . 16-59 [2] Miscellaneous Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-60 [2] Interfaces of the GPT Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-65 [2] 17 17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8 Real Time Clock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-1 [2] Defining the RTC Time Base . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-2 [2] RTC Run Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-5 [2] RTC Operating Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-7 [2] 48-bit Timer Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-11 [2] System Clock Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-11 [2] Cyclic Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-12 [2] RTC Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-13 [2] Miscellaneous Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-15 [2] 18 18.1 18.1.1 18.1.2 18.1.3 Analog to Digital Converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADC Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Feature Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User’s Manual L-11 13-10 [1] 13-10 [1] 13-11 [1] 13-12 [1] 18-1 [2] 18-1 [2] 18-2 [2] 18-3 [2] 18-4 [2] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 18.1.4 18.1.5 18.1.6 18.1.7 18.1.8 18.1.9 18.2 18.2.1 18.2.2 18.2.3 18.2.4 18.2.5 18.2.6 18.2.7 18.2.8 18.2.9 18.2.10 18.2.11 18.2.12 18.2.13 18.2.14 18.2.15 18.2.16 18.2.17 18.2.18 18.2.19 18.2.20 18.2.21 18.3 18.3.1 18.3.2 18.3.3 18.3.4 ADC Kernel Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-5 [2] Conversion Request Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-7 [2] Conversion Result Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-9 [2] Interrupt Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-10 [2] Electrical Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-12 [2] Transfer Characteristics and Error Definitions . . . . . . . . . . . . . . . 18-15 [2] Operating the ADC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-16 [2] Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-17 [2] Mode Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-20 [2] Module Activation and Power Saving Modes . . . . . . . . . . . . . . . 18-22 [2] Clocking Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-23 [2] General ADC Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-24 [2] Request Source Arbiter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-34 [2] Arbiter Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-39 [2] Scan Request Source Handling . . . . . . . . . . . . . . . . . . . . . . . . . . 18-41 [2] Scan Request Source Registers . . . . . . . . . . . . . . . . . . . . . . . . . 18-45 [2] Sequential Request Source Handling . . . . . . . . . . . . . . . . . . . . . 18-49 [2] Sequential Source Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-54 [2] Channel-Related Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-65 [2] Channel-Related Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-70 [2] Conversion Result Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-80 [2] Conversion Result-Related Registers . . . . . . . . . . . . . . . . . . . . . 18-88 [2] Multiplexer Test Mode for CH7 . . . . . . . . . . . . . . . . . . . . . . . . . . 18-98 [2] External Multiplexer Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-99 [2] Synchronized Conversions for Parallel Sampling . . . . . . . . . . . 18-102 [2] Equidistant Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-106 [2] Broken Wire Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-108 [2] Additional Feature Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-110 [2] Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-117 [2] Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-117 [2] Interrupt Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-117 [2] Analog Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-119 [2] Digital Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-122 [2] 19 19.1 19.2 19.3 19.3.1 19.4 19.5 19.5.1 19.5.2 Capture/Compare Unit 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-1 [2] The CAPCOM2 Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-4 [2] CAPCOM2 Timer Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-10 [2] Capture/Compare Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-11 [2] Capture/Compare Registers for the CAPCOM2 (CC31 … CC16) 19-11 [2] Capture Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-14 [2] Compare Mode Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-15 [2] Compare Mode 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-16 [2] Compare Mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-16 [2] User’s Manual L-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 19.5.3 19.5.4 19.5.5 19.6 19.7 19.8 19.9 19.10 19.10.1 19.11 Compare Mode 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compare Mode 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Double-Register Compare Mode . . . . . . . . . . . . . . . . . . . . . . . . Compare Output Signal Generation . . . . . . . . . . . . . . . . . . . . . . . . Single Event Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Staggered and Non-Staggered Operation . . . . . . . . . . . . . . . . . . . . CAPCOM2 Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . External Input Signal Requirements . . . . . . . . . . . . . . . . . . . . . . . . Miscellaneous Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Interfaces of the CAPCOM Units . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 20.1 20.1.1 20.1.2 20.1.3 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 20.2.9 20.3 20.3.1 20.3.2 20.3.3 20.3.4 20.3.5 20.3.6 20.4 20.5 20.6 20.6.1 20.6.2 20.6.3 20.6.4 20.7 20.7.1 20.7.2 Capture/Compare Unit 6 (CCU6) . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 [2] Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-1 [2] Feature Set Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-2 [2] Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-3 [2] Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-4 [2] Operating Timer T12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-7 [2] T12 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-8 [2] T12 Counting Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-10 [2] T12 Compare Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-14 [2] Compare Mode Output Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-21 [2] T12 Capture Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-26 [2] T12 Shadow Register Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . 20-30 [2] Timer T12 Operating Mode Selection . . . . . . . . . . . . . . . . . . . . . 20-31 [2] T12 related Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-32 [2] Capture/Compare Control Registers . . . . . . . . . . . . . . . . . . . . . . 20-37 [2] Operating Timer T13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-49 [2] T13 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-49 [2] T13 Counting Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-52 [2] T13 Compare Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-57 [2] Compare Mode Output Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-59 [2] T13 Shadow Register Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . 20-60 [2] T13 related Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-62 [2] Trap Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-65 [2] Multi-Channel Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-67 [2] Hall Sensor Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-70 [2] Hall Pattern Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-71 [2] Hall Pattern Compare Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-73 [2] Hall Mode Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-74 [2] Hall Mode for Brushless DC-Motor Control . . . . . . . . . . . . . . . . . 20-76 [2] Modulation Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-78 [2] Modulation Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-78 [2] Trap Control Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-80 [2] User’s Manual L-13 19-19 [2] 19-19 [2] 19-24 [2] 19-27 [2] 19-29 [2] 19-31 [2] 19-36 [2] 19-38 [2] 19-39 [2] 19-42 [2] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 20.7.3 20.7.4 20.8 20.8.1 20.8.2 20.9 20.9.1 20.9.2 20.9.3 20.10 20.10.1 20.10.2 20.10.3 20.10.4 Passive State Level Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-83 [2] Multi-Channel Mode Registers . . . . . . . . . . . . . . . . . . . . . . . . . . 20-84 [2] Interrupt Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-89 [2] Interrupt Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-89 [2] Interrupt Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-91 [2] General Module Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-103 [2] Mode Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-103 [2] Input Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-106 [2] General Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-107 [2] Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-115 [2] Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-115 [2] Interrupt Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-116 [2] Synchronous Start Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-117 [2] Digital Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20-118 [2] 21 21.1 21.1.1 21.1.2 21.1.3 21.1.4 21.1.5 21.1.6 21.1.7 21.2 21.2.1 21.2.2 21.2.3 21.2.4 21.2.5 21.2.6 21.2.7 21.2.8 21.2.9 21.2.10 21.2.11 21.2.12 21.2.13 21.2.14 21.3 21.3.1 21.3.2 21.3.3 Universal Serial Interface Channel . . . . . . . . . . . . . . . . . . . . . . . . . 21-1 [2] Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1 [2] Feature Set Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-2 [2] Channel Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-5 [2] Input Stages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-6 [2] Output Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-7 [2] Baud Rate Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-8 [2] Channel Events and Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-9 [2] Data Shifting and Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-9 [2] Operating the USIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-13 [2] Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-13 [2] Operating the USIC Communication Channel . . . . . . . . . . . . . . . 21-18 [2] Channel Control and Configuration Registers . . . . . . . . . . . . . . . 21-25 [2] Protocol Related Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-33 [2] Operating the Input Stages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-36 [2] Input Stage Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-38 [2] Operating the Baud Rate Generator . . . . . . . . . . . . . . . . . . . . . . 21-41 [2] Baud Rate Generator Registers . . . . . . . . . . . . . . . . . . . . . . . . . 21-46 [2] Operating the Transmit Data Path . . . . . . . . . . . . . . . . . . . . . . . . 21-51 [2] Operating the Receive Data Path . . . . . . . . . . . . . . . . . . . . . . . . 21-55 [2] Transfer Control and Status Registers . . . . . . . . . . . . . . . . . . . . 21-57 [2] Data Buffer Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-69 [2] Operating the FIFO Data Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . 21-79 [2] FIFO Buffer and Bypass Registers . . . . . . . . . . . . . . . . . . . . . . . 21-89 [2] Asynchronous Serial Channel (ASC = UART) . . . . . . . . . . . . . . . . 21-110 [2] Signal Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-110 [2] Frame Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-111 [2] Operating the ASC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-114 [2] User’s Manual L-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 21.3.4 21.3.5 21.4 21.4.1 21.4.2 21.4.3 21.4.4 21.4.5 21.4.6 21.5 21.5.1 21.5.2 21.5.3 21.5.4 21.5.5 21.6 21.6.1 21.6.2 21.6.3 21.6.4 21.6.5 21.7 21.7.1 21.7.2 21.7.3 21.7.4 21.7.5 21.7.6 ASC Protocol Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hardware LIN Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synchronous Serial Channel (SSC) . . . . . . . . . . . . . . . . . . . . . . . Signal Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating the SSC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating the SSC in Master Mode . . . . . . . . . . . . . . . . . . . . . . Operating the SSC in Slave Mode . . . . . . . . . . . . . . . . . . . . . . . SSC Protocol Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SSC Timing Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inter-IC Bus Protocol (IIC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating the IIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Symbol Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Flow Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IIC Protocol Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IIS Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating the IIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Operating the IIS in Master Mode . . . . . . . . . . . . . . . . . . . . . . . Operating the IIS in Slave Mode . . . . . . . . . . . . . . . . . . . . . . . . IIS Protocol Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . USIC Implementation in XC27x5X . . . . . . . . . . . . . . . . . . . . . . . . . Implementation Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Channel Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Module Identification Registers . . . . . . . . . . . . . . . . . . . . . . . . . Interrupt Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Input/Output Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 22.1 22.1.1 22.1.2 22.2 22.2.1 22.2.2 22.2.3 22.2.4 22.2.5 22.2.6 22.2.7 22.2.8 22.2.9 Controller Area Network (MultiCAN) Controller . . . . . . . . . . . . . . 22-1 [2] MultiCAN Short Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 [2] Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1 [2] CAN Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-2 [2] CAN Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-3 [2] Conventions and Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-3 [2] Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-3 [2] CAN Node Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-9 [2] Message Object List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-13 [2] CAN Node Analysis Features . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-18 [2] Message Acceptance Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-21 [2] Message Postprocessing Interface . . . . . . . . . . . . . . . . . . . . . . . 22-24 [2] Message Object Data Handling . . . . . . . . . . . . . . . . . . . . . . . . . . 22-28 [2] Message Object Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-35 [2] User’s Manual L-15 21-123 [2] 21-129 [2] 21-131 [2] 21-131 [2] 21-139 [2] 21-143 [2] 21-150 [2] 21-152 [2] 21-158 [2] 21-161 [2] 21-161 [2] 21-165 [2] 21-171 [2] 21-174 [2] 21-179 [2] 21-185 [2] 21-185 [2] 21-189 [2] 21-194 [2] 21-198 [2] 21-199 [2] 21-205 [2] 21-205 [2] 21-206 [2] 21-207 [2] 21-208 [2] 21-210 [2] 21-212 [2] V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Table Of Contents 22.3 22.3.1 22.3.2 22.3.3 22.3.4 22.4 22.4.1 22.4.2 22.4.3 22.4.4 22.5 22.5.1 22.5.2 22.5.3 22.5.4 22.5.5 22.5.6 22.5.7 MultiCAN Kernel Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-44 [2] Register Address Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-44 [2] Global MultiCAN Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-49 [2] CAN Node Specific Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-62 [2] Message Object Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-79 [2] General Control and Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-102 [2] Clock Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-102 [2] Port Input Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-103 [2] Suspend Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-104 [2] Interrupt Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-105 [2] MultiCAN Module Implementation . . . . . . . . . . . . . . . . . . . . . . . . . 22-106 [2] Interfaces of the CAN Module . . . . . . . . . . . . . . . . . . . . . . . . . . 22-106 [2] Module Clock Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-107 [2] Mode Control Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-116 [2] Mode Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-117 [2] Mode Control Register Description . . . . . . . . . . . . . . . . . . . . . . 22-119 [2] Connection of External Signals . . . . . . . . . . . . . . . . . . . . . . . . . 22-122 [2] MultiCAN Module Register Address Map . . . . . . . . . . . . . . . . . 22-127 [2] 23 Appendix: Functional and Operational Updates . . . . . . . . . . . . . 23-1 [2] Keyword Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-1 [2] Register Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25-8 [2] User’s Manual L-16 V1.2, 2008-09 XC27x5X 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 CPU performance and peripheral functionality. To achieve this high performance goal, Infineon has decided to develop its families of 16/32-bit CMOS microcontrollers without the constraints of backward compatibility wih previous architectures. Nonetheless the architectures of these microcontroller families pursue successful hardware and software concepts, which have been established in Infineon’s popular 8-bit controller families, while delivering 32-bit performance. This established functionality, which has been the basis for system solutions in a wide range of application areas, is amended with flexible peripheral modules and effective power control features. The sum of this provides the prerequisites for powerful, yet efficient systems-on-chip. Solutions for PowerTrain Systems The XC27x5X derivatives address the increasing requirements for value, performance, and features in PowerTrain applications. The combination of 32-bit features with the well established C166SV2 core supports competitive engine- and transmission-applications: • The scaleable XC27x5X product portfolio fits perfectly to your requirements • Compatibility provides easy adaptation to changing requirements in performance and functionality • 32-bit (system) performance and features ensure long term usability • Flexible peripherals such as USIC, MultiCAN, ADC maximize re-use across your developments • Enhanced value through single voltage supply and embedded voltage regulator Soon to come variants enlarging the portfolio, and Infineon’s high quality standards make the XC27x5X Controllers an ideal choice for performing and upgradable systems. User’s Manual Intro, V1.1 1-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction About this Manual This manual describes the functionality of a number of microcontroller types of the Infineon XC2000 Family, the XC27x5X derivatives. 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. A detailled list of features of the various derivatives is provided in the Data Sheets of these products. 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 XC27x5X 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 XC27x5X 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 Intro, V1.1 1-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction 1.1 Members of the 16/32-bit Microcontroller Families The microcontrollers in the Infineon 16/32-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/LXBus 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. 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 Intro, V1.1 1-3 V1.2, 2008-09 XC27x5X 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. It also adds 32-bit operations and a 40-bit accumulator. The 5-stage pipeline, single-cycle execution of most instructions, and PEC-transfers within the complete addressing range increase system performance. The previous XBUS is replaced by the optimized LXBus. Debugging the target system is supported by integrated functions for On-Chip Debug Support (OCDS). The present XC2000 Family of microcontrollers builds the fifth generation of 16-bit microcontrollers which provides 32-bit performance and takes users and applications a considerable step towards industry’s target of systems on chip. Integrated memories and peripherals allow compact systems, the integrated core power supply and control reduces system requirements to one single voltage supply, the powerful combination of CPU and MAC-unit is unleashed by optimized compilers. This leaves no performance gap towards 32-bit systems. 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 XC27x5X, descriptions within this manual that refer to the “XC27x5X” also apply to the other variations, unless otherwise noted. User’s Manual Intro, V1.1 1-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction 1.2 Summary of Basic Features The XC27x5X devices are enhanced members of the Infineon XC2000 Family of full featured single-chip CMOS microcontrollers. Note: The various derivates are referred to as XC27x5X throughout this manual. The XC27x5X combines the extended functionality and performance of the C166SV2 Core with powerful on-chip peripheral subsystems and on-chip memory units and provides several means for power reduction. The following key features contribute to the high performance of the XC27x5X: Intelligent On-Chip Peripheral Subsystems • Two synchronizable A/D Converters with programmable resolution (10-bit or 8-bit) and conversion time (down to below 1 µs), up to 24 analog input channels, auto scan modes, channel injection, data reduction features • One Capture/Compare Unit with 2 independent time bases, very flexible PWM unit/event recording unit with different operating modes, includes two 16-bit timers/counters, maximum resolution fSYS • Up to four Capture/Compare Units for flexible PWM Signal Generation (CCU6) (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 • Eight Serial Channels with baud rate generator, receive/transmit FIFOs, programmable data length and shift direction, usable as UART, SPI-like, IIC, IIS, and LIN interface • Controller Area Network (MultiCAN) Module, Rev. 2.0B active, two nodes operating independently or exchanging data via a gateway function, FullCAN/Basic-CAN • Real Time Clock with alarm interrupt • Watchdog Timer with programmable time intervals • Bootstrap Loaders for flexible system initialization • Protection management for system configuration and control registers Integrated On-Chip Memories • • • • • 8 Kbytes on-chip Stand-By RAM (SBRAM) for data 2 Kbytes Dual-Port RAM (DPRAM) for variables, register banks, and stacks 16 Kbytes on-chip high-speed Data SRAM (DSRAM) for variables and stacks Up to 32 Kbytes on-chip high-speed Program/Data SRAM (PSRAM) for code and data Up to 832 Kbytes on-chip Flash Program Memory for instruction code or constant data Note: The system stack can be located in any memory area within the complete addressing range. User’s Manual Intro, V1.1 1-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction 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 32-bit addition and 32-bit subtraction (MAC unit) 40-bit barrel shifter and 40-bit accumulator 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 • • • • • Two IO power domains fulfill system requirements from 3 V to 5 V Gated clock concept for improved power consumption and EMC Programmable system slowdown via clock generation unit Flexible management of peripherals, can be individually disabled Programmable frequency output 16-Priority-Level Interrupt System • • • • • 96 interrupt nodes with separate interrupt vectors on 15 priority levels (8 group levels) 7 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) User’s Manual Intro, V1.1 1-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction 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 On-Chip Debug Support • • • • • • • • Communication through DAP interface (2-wire) or JTAG interface (5-wire) On-chip debug controller with optional break interface 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 DAP/JTAG interface Access to any internal register or memory location via DAP/JTAG interface Single step support and watchpoints with MOV-injection Input/Output Lines With Individual Bit Addressability • • • • Tri-stated in input mode Push/pull or open drain output mode Programmable port driver control Two I/O power domains with a supply voltage range from 3.0 V to 5.5 V (core-logic and oscillator input voltage is 1.5 V) Various Temperature Ranges • -40 to +85 °C • -40 to +125 °C1) Infineon CMOS Process • Low power CMOS technology enables power saving modes with flexible power management. Green Plastic Low-Profile Quad Flat Pack (LQFP) Packages • PG-LQFP-144, 20 × 20 mm body, 0.5 mm (19.7 mil) lead spacing, surface mount technology 1) Not all derivatives are offered in all temperature ranges. User’s Manual Intro, V1.1 1-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction • PG-LQFP-100, 14 × 14 mm body, 0.5 mm (19.7 mil) lead spacing, surface mount technology • PG-LQFP-64, 10 × 10 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 collaborates with third party tool suppliers. There is a wide range of tool suppliers world-wide, ranging from local niche manufacturers to multinational companies with broad product portfolios, offering powerful development tools for the complete variety of Infineon microcontroller families, providing a remarkable selection 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/C++) Macro-assemblers, linkers, locators, library managers, format-converters Architectural simulators HLL debuggers Real-time operating systems In-circuit emulators Logic analyzer disassemblers Starter kits and evaluation boards Chip configuration code generation tool (DAvE) User’s Manual Intro, V1.1 1-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction 1.3 Abbreviations The following acronyms and terms are used within this document: 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 DAP Device Access Port DMU Data Management Unit EBC External Bus Controller ESFR Extended Special Function Register EVVR Embedded Validated Voltage Regulator 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) IIS Inter Integrated Circuit Sound (Bus) IO Input/Output JTAG Joint Test Access Group LIN Local Interconnect Network LPR Low Power Reference LQFP Low Profile Quad Flat Pack LXBus Internal representation of the external bus MAC Multiply/Accumulate (unit) MPU Memory Protection Unit OCDS On-Chip Debug Support User’s Manual Intro, V1.1 1-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Introduction OTP One-Time Programmable memory PEC Peripheral Event Controller PLA Programmable Logic Array PLL Phase Locked Loop PMU Program Management Unit PVC Power Validation Circuit PWM Pulse Width Modulation RAM Random Access Memory RISC Reduced Instruction Set Computing ROM Read Only Memory RTC Real Time Clock SFR Special Function Register SSC Synchronous Serial Channel SWD Supply Watchdog UART Universal Asynchronous Receiver/Transmitter USIC Universal Serial Interface Channel 1.4 Naming Conventions The diverse bitfields used for control functions and status indication and the registers housing them are equipped with unique names wherever applicable. Thereby these control structures 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 ADC0_GLOBCTR clearly identifies register GLOBCTR as part of module ADC0, the term SYSCON0.CLKSEL clearly identifies bitfield CLKSEL as part of register SYSCON0. User’s Manual Intro, V1.1 1-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2 Architectural Overview The architecture of the XC27x5X 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 XC27x5X 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 XC27x5X 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 XC27x5X. PSRAM 32 Kbytes DPRAM 2 Kbytes DSRAM 16 Kbytes OCDS Debug Support Program Flash 0 256 Kbytes DMU C166SV2 - Core MCHK System Functions Clock, Reset, Power Control, StandBy RAM WDT Interrupt & PEC LXBus XTAL Program Flash 2 256 Kbytes Program Flash 3 64 Kbytes EBC LXBus Control External Bus Control CPU/MPU PMU IMB Program Flash 1 256 Kbytes RTC ADC1 ADC0 8-/10- 8-/10Bit, Bit, 8 16 Chan. Chan. GPT T2 T3 T4 ... CCU60 CC2 CCU63 T7 T12 T12 T8 T13 T13 Peri pheral Data B us Interrupt Bus T5 T6 BRGen P15 8/5 Port 5 16/11 P11 6 P10 16 P9 P8 8 P7 P6 7 5 USIC1 2 Ch., 64 x Buffer USIC0 Multi 2 Ch., CAN 64 x Buffer RS232, LIN, SPI, IIC, IIS RS232, 2 LIN, chan. SPI, IIC, IIS P4 4/3 8/4 P3 8 P2 14 P1 P0 8 8 MC_XC27X5X_BLOCKDIAGRAM Figure 2-1 XC27x5X Functional Block Diagram User’s Manual ArchitectureX27, V1.1 2-1 V1.2, 2008-09 XC27x5X 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/ROM PMU CPU Prefetch Unit Branch Unit CSP IP VECSEG CPUCON1 CPUCON2 Return Stack FIFO IDX0 IDX1 QX0 QX1 QR0 QR1 +/- +/- Multiply Unit MRW +/- MCW MSW MAH MAL 2-Stage Prefetch Pipeline TFR Injection/ Exception Handler 5-Stage Pipeline IFU DPP0 DPP1 DPP2 DPP3 DPRAM IPIP SPSEG SP STKOV STKUN ADU Division Unit Bit-Mask-Gen. Multiply Unit Barrel-Shifter MDC R15 R15 R14 R15 R14 R14 R15 R14 GPRs GPRs GPRs GPRs R1 R1 R0 R0R1 R0 R1 R0 RF PSW +/- MDH MDL ZEROS ONES MAC CP Buffer ALU WB DSRAM EBC Peripherals DMU mca04917_x.vsd Figure 2-2 CPU Block Diagram User’s Manual ArchitectureX27, V1.1 2-2 V1.2, 2008-09 XC27x5X 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 – 32-bit addition and 32-bit subtraction – 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 Built-in memory protection unit (MPU) 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. 2.1.1 Memory Protection Unit (MPU) The Memory Protection Unit (MPU) provides the hardware mechanisms needed for implementing memory protection. The MPU allows detection of unauthorized accesses (read, write or instruction fetch) in user-defined memory ranges. It offers protection for the complete address space, including the peripheral area. The MPU can be used to support the encapsulation of different applications or software components running on the processor. This encapsulation provides the means to ensure integrity and fault isolation capabilities in today’s complex systems relying on multiplesources software. User’s Manual ArchitectureX27, V1.1 2-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.2 High Instruction Bandwidth/Fast Execution Based on the hardware provisions, most of the XC27x5X’s instructions can be executed in just one clock cycle (1/fSYS). This includes arithmetic instructions, logic instructions, and move instructions with most addressing modes. Special instructions such as JMPS 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 5.2), 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 ArchitectureX27, V1.1 2-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.3 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 ArchitectureX27, V1.1 2-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.4 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 5.2.1) 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 ArchitectureX27, V1.1 2-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.5 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 XC27x5X 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 ArchitectureX27, V1.1 2-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.6 Programmable Multiple Priority Interrupt System The XC27x5X provides 96 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 XC27x5X 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 XC27x5X is capable of reacting very quickly to non-deterministic events because its interrupt response time is within a very narrow range of typically 7 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 XC27x5X 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 ArchitectureX27, V1.1 2-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.1.7 Interfaces to System Resources The CPU of the XC27x5X 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 XC27x5X can execute code from external resources. No code can be fetched from the Data RAM (DSRAM). • DMU via IMB: Data accesses can also be executed to on-chip resources normally 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 ArchitectureX27, V1.1 2-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.2 On-Chip System Resources The XC27x5X 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 XC27x5X has eight PEC channels, each of which offers such fast interruptdriven data transfer capabilities. Memory Areas The memory space of the XC27x5X 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. Note: The actual memory sizes depend on the selected device type. This overview describes the maximum block sizes. User’s Manual ArchitectureX27, V1.1 2-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Up to 832 Kbytes of on-chip Flash memory store code or constant data. The on-chip Flash memory consists of 4 Flash modules, each built up from 4-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 user-defined password sequence temporarily unlocks protected areas. The Flash modules combine 128-bit 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. Accesses to different Flash modules can be executed in parallel. Note: Program execution from on-chip program memory is the fastest of all possible alternatives and results in maximum performance. The size of the on-chip program memory depends on the chosen derivative. On-chip program memory also includes the PSRAM. Up to 32 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. A section of the PSRAM with programmable size can be write-protected. 16 Kbytes of on-chip Data SRAM (DSRAM) are provided as a storage for general user data. The DSRAM is accessed via a separate interface 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. 8 Kbytes of on-chip Stand-By SRAM (SBRAM) is provided as a storage for systemrelevant user data that must be preserved while the major part of the device is powered down. The SBRAM is accessed via a specific interface and is powered via domain M. 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 1) To save control bits, sectors are clustered for protection purposes, they remain separate for programming/ erasing. User’s Manual ArchitectureX27, V1.1 2-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 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. 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 XC2000 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 XC27x5X Memory Map Address Area Start Loc. End Loc. Area Size1) Notes IMB register space FF’FF00H FF’FFFFH 256 Bytes – Reserved (Access trap) F0’0000H FF’FEFFH <1 Mbyte Minus IMB reg. Reserved for EPSRAM E8’8000H EF’FFFFH 480 Kbytes Mirrors EPSRAM Emulated PSRAM E8’0000H E8’7FFFH 32 Kbytes Flash timing Reserved for PSRAM E0’8000H E7’FFFFH 480 Kbytes Mirrors PSRAM Program SRAM E0’0000H E0’7FFFH 32 Kbytes Maximum speed Reserved for pr. mem. CC’4000H DF’FFFFH <1.25 Mbytes – Program Flash 3 CC’0000H CC’3FFFH 64 Kbytes – Program Flash 2 C8’0000H CB’FFFFH 256 Kbytes – Program Flash 1 C4’0000H C7’FFFFH 256 Kbytes – Program Flash 0 C0’0000H C3’FFFFH 256 Kbytes 2) External memory area 40’0000H BF’FFFFH 8 Mbytes – Available Ext. IO area3) 20’C000H 3F’FFFFH < 2 Mbytes Minus USIC/CAN MultiCAN/USIC regs. 20’8000H 20’BFFFH 16 Kbytes Alternate location Available Ext. IO area3) 20’6000H 20’7FFFH 8Kbytes – User’s Manual ArchitectureX27, V1.1 2-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-1 XC27x5X Memory Map (cont’d) Address Area Start Loc. End Loc. Area Size1) Notes USIC registers 20’4000H 20’5FFFH 8 Kbytes Accessed via EBC MultiCAN registers 20’0000H 20’3FFFH 16 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 – Data SRAM 00’A000H 00’DFFFH 16 Kbytes – Reserved for DSRAM 00’8000H 00’9FFFH 8 Kbytes – External memory area 00’0000H 00’7FFFH 32 Kbytes – 1) The areas marked with “<” are slightly smaller than indicated, see column “Notes”. 2) The uppermost 4-Kbyte sector of the first Flash segment is reserved for internal use (C0’F000H to C0’FFFFH). 3) Several pipeline optimizations are not active within the external IO area. This is necessary to control external peripherals properly. Note: For an overview of the available memory sections for the different derivatives, please refer to the corresponding Data Sheets. User’s Manual ArchitectureX27, V1.1 2-13 V1.2, 2008-09 XC27x5X 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/ROM/Flash or peripherals can be connected to the XC27x5X 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 an external bus mode with the following possible selections1): • Address Bus Width with a range of 0 … 24-bit • Data Bus Width 8-bit or 16-bit • Bus Operation Multiplexed or Demultiplexed In the demultiplexed bus modes, addresses are output on Port 0 and Port 1 and data is input/output on Port 10 and Port 2. In the multiplexed bus modes both addresses and data use Port 10 and Port 2 for input/output. The high order address (segment) lines use Port 2. The number of active address lines is selectable, so the external address space can be restricted. This is required when interface lines are assigned to Port 2. 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, the upper Port 2 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 MultiCAN module and the USIC modules are connected to and accessed via the LXBus. 1) Bus modes are switched dynamically if several address windows with different mode settings are used. User’s Manual ArchitectureX27, V1.1 2-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.3 On-Chip Peripheral Blocks The XC2000 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 XC27x5X generic peripherals are: • • • • • • • • Memory Checker Unit General Purpose Timer Unit (GPT1, GPT2) Watchdog Timer Capture/Compare Unit (CAPCOM2) Four Capture/Compare Units CCU6 (CCU60, CCU61, CCU62, CCU63) Two 10-bit Analog/Digital Converters (ADC0, ADC1) Real Time Clock (RTC) Thirteen/Nine Parallel Ports with a total of 119/76 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: • MultiCAN Module with 2 CAN nodes and gateway functionality • Two Universal Serial Interface Channel Modules (USIC) 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. Note: For an overview of the available peripherals for the different derivatives, please refer to “Summary of Basic Features” on Page 1-5. User’s Manual ArchitectureX27, V1.1 2-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Peripheral Interfaces The on-chip peripherals generally have two different types of interfaces: a bus interface to the CPU and interface signals to other on-chip peripherals or 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 system clock (fSYS). The clock generation unit uses external (e.g. a crystal) or internal clock sources to generate the system clock signal. Peripherals can be disconnected from the clock signal either temporarily to save energy or permanently if they are not used in a specific application. 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: The SFRs reside in various data pages 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 the corresponding data page. – 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). • 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 User’s Manual ArchitectureX27, V1.1 2-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 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. • Write Operations to Write-Only Bits/Registers usually modify bits within other registers. In some cases this modification is controlled by state machines. Therefore, the effect of the write operation may not be visible, when the modified register is read immediately after the write access that triggers the modification. • Reserved Bits: Some of the bits which are contained in the XC27x5X’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 Unit (CAPCOM2) The CAPCOM units support generation and control of timing sequences on up to 16 channels with a maximum resolution of 1 system clock cycle (8 cycles in staggered mode). The CAPCOM unit is 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. Two 16-bit timers (T7/T8) with reload registers provide two independent time bases for each capture/compare register. 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 timer T7 allow event scheduling for the capture/compare registers relative to external events. The capture/compare register array contains 16 dual purpose capture/compare registers, each of which may be individually allocated to either CAPCOM timer T7 or T8 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 ArchitectureX27, V1.1 2-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-2 Compare Modes (CAPCOM2) 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 ArchitectureX27, V1.1 2-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Capture/Compare Units CCU6 The CCU6 units support generation and control of timing sequences on up to three 16bit capture/compare channels plus one independent 16-bit compare channel. In compare mode, the CCU6 units provide 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 and T13 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. Note: The number of available CCU6 units and channels depends on the selected device type. User’s Manual ArchitectureX27, V1.1 2-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview General Purpose Timer Unit (GPT1, GPT2) 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 is User’s Manual ArchitectureX27, V1.1 2-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 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 CAPCOM2 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 XC27x5X 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 ArchitectureX27, V1.1 2-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Real Time Clock (RTC) The Real Time Clock (RTC) module of the XC27x5X is directly clocked wih a separate clock signal. Several internal and external clock sources can be selected via register RTCCLKCON. It is, therefore, independent from the selected clock generation mode of the XC27x5X. The RTC basically consists of a chain of divider blocks: • Selectable 32:1 and 8:1 dividers (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. Note: The registers associated with the RTC are not affected by an application reset in order to maintain the contents even when intermediate resets are executed. The RTC module can be used for different purposes: • System clock to determine the current time and date • Cyclic time based interrupt, to provide a system time tick independent of CPU frequency and other resources • 48-bit timer for long term measurements User’s Manual ArchitectureX27, V1.1 2-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Analog/Digital Converters (ADC0, ADC1) For analog signal measurement, two 10-bit A/D converters (ADC0, ADC1) with 16 (or 8) multiplexed input channels including a sample and hold circuit have been integrated onchip. They use the method of successive approximation. The sample time (for loading the capacitors) and the conversion time are programmable and can thus be adjusted to the external circuitry. The A/D converters can also operate in 8-bit conversion mode, where the conversion time is further reduced. Several independent conversion result registers, selectable interrupt requests, and highly flexible conversion sequences provide a high degree of programmability to fulfill the requirements of the respective application. Both modules can be synchronized to allow parallel sampling of two input channels. For applications that require more analog input channels, external analog multiplexers can be controlled automatically. For applications that require less analog input channels, the remaining channel inputs can be used as digital input port pins. The A/D converters of the XC27x5X support two types of request sources which can be triggered by several internal and external events. • Scan requests are activated at the same time and then executed in a predefined sequence. • Queued requests are executed in a user-defined sequence. In addition, the conversion of a specific channel can be inserted into a running sequence without disturbing this sequence. All requests are arbitrated according to the priority level that has been assigned to them. Data reduction features, such as limit checking or result accumulation, reduce the number of required CPU accesses and so allow the precise evaluation of analog inputs (high conversion rate) even at low CPU speed. The Peripheral Event Controller (PEC) may be used to control the A/D converters or to automatically store conversion results into a table in memory for later evaluation, without requiring the overhead of entering and exiting interrupt routines for each data transfer. Therefore, each A/D converter contains 8 result registers which can be concatenated to build a result FIFO. Wait-for-read mode can be enabled for each result register to prevent loss of conversion data. 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 input stages under software control. This can be selected for each pin separately via registers P5_DIDIS and P15_DIDIS (Port x Digital Input Disable). The Auto-Power-Down feature of the A/D converters minimizes the power consumption when no conversion is in progress. Note: The number of available analog channels depends on the selected device type. User’s Manual ArchitectureX27, V1.1 2-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Universal Serial Interface Channel Modules (USIC) Each USIC channel can be individually configured to match the application needs, e.g. the protocol can be selected or changed during run time without the need for a reset. The following protocols are supported: • UART (ASC, asynchronous serial channel) - module capability: receiver/transmitter with max. baud rate fsys/4 - application target baud rate range: 1.2 kBaud to 3.5 MBaud - number of data bits per data frame 1 to 63 - MSB or LSB first • LIN Support by HW (low-cost network, baud rate up to 20 kBaud) - data transfers based on ASC protocol - baud rate detection possible by built-in capture event of baud rate generator - checksum generation under SW control for higher flexibility • SSC/SPI (synchronous serial channel with or without slave select lines) - module capability: slave mode with max. baud rate fsys - module capability: master mode with max. baud rate fsys /2 - application target baud rate range: 2 kBaud to 10 MBaud - number of data bits per data frame 1 to 63, more with explicit stop condition - MSB or LSB first • IIC (Inter-IC Bus) - application baud rate 100 kBaud to 400 kBaud - 7-bit and 10-bit addressing supported - full master and slave device capability • IIS (infotainment audio bus) - module capability: receiver with max. baud rate fSYS - module capability: transmitter with max. baud rate fSYS /2 - application target baud rate range: up to 26 MBaud In addition to the flexible choice of the communication protocol, the USIC structure has been designed to reduce the system load (CPU load) allowing efficient data handling. The following aspects have been considered: • Data buffer capability The standard buffer capability includes a double word buffer for receive data and a single word buffer for transmit data. This allows longer CPU reaction times (e.g. interrupt latency). • Additional FIFO buffer capability In addition to the standard buffer capability, the received data and the data to be transmitted can be buffered in a FIFO buffer structure. The size of the receive and the transmit FIFO buffer can be programmed independently. Depending on the application needs, a total buffer capability of 64 data words can be assigned to the receive and transmit FIFO buffers of a USIC module (the two channels of the USIC module share the 64 data word buffer). User’s Manual ArchitectureX27, V1.1 2-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview • • • • • • • • In addition to the FIFO buffer, a bypass mechanism allows the introduction of highpriority data without flushing the FIFO buffer. Transmit control information For each data word to be transmitted, a 5-bit transmit control information has been added to automatically control some transmission parameters, such as word length, frame length, or the slave select control for the SPI protocol. The transmit control information is generated automatically by analyzing the address where the user SW has written the data word to be transmitted (32 input locations = 2^5 = 5 bit transmit control information). This feature allows individual handling of each data word, e.g. the transmit control information associated to the data words stored in a transmit FIFO can automatically modify the slave select outputs to select different communication targets (slave devices) without CPU load. Alternatively, it can be used to control the frame length. Flexible frame length control The number of bits to be transferred within a data frame is independent of the data word length and can be handled in two different ways. The first option allows automatic generation of frames up to 63 bits with a known length. The second option supports longer frames (even unlimited length) or frames with a dynamically controlled length. Interrupt capability The events of each USIC channel can be individually routed to one of 4 service request outputs, depending on the application needs. Furthermore, specific start and end of frame indications are supported in addition to protocol-specific events. Flexible interface routing Each USIC channel offers the choice between several possible input and output pins connections for the communications signals. This allows a flexible assignment of USIC signals to pins that can be changed without resetting the device. Input conditioning Each input signal is handled by a programmable input conditioning stage with programmable filtering and synchronization capability. Baud rate generation Each USIC channel contains an own baud rate generator. The baud rate generation can be based either on the internal module clock or on an external frequency input. This structure allows data transfers with a frequency that can not be generated internally, e.g. to synchronize several communication partners. Transfer trigger capability In master mode, data transfers can be triggered events generated outside the USIC module, e.g. at an input pin or a timer unit (transmit data validation). This feature allows time base related data transmission. Debugger support The USIC offers specific addresses to read out received data without interaction with the FIFO buffer mechanism. This feature allows debugger accesses without the risk of a corrupted receive data sequence. User’s Manual ArchitectureX27, V1.1 2-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview To reach a desired baud rate, two criteria have to be respected, the module capability and the application environment. The module capability is defined with respect to the module’s input clock frequency fsys, being the base for the module operation. Although the module’s capability being much higher (depending on the module clock and the number of module clock cycles needed to represent a data bit), the reachable baud rate is generally limited by the application environment. In most cases, the application environment limits the maximum reachable baud rate due to driver delays, signal propagation times, or due to EMI reasons. Note: Depending on the selected additional functions (such as digital filters, input synchronization stages, sample point adjustment, data structure, etc.), the maximum reachable baud rate can be limited. Please also take care about additional delays, such as (internal or external) propagation delays and driver delays (e.g. for collision detection in ASC mode, for IIC, etc.). SRx Baud Rate Generator User Interface Data Buffer Data Shift Unit To Interrupt Registers f SYS PPP Input Stages (ASC, SSC. ..) Channel 0 f SYS Baud Rate Generator Data Buffer Data Shift Unit PPP Pins UxC0 UxC1 Signal Distribution Interrupt Generation Input Stages (ASC, SSC. ..) Channel 1 Optional : FIFO Data Buffer shared between UxC 0 and UxC1 USIC Module x USIC_Module Figure 2-3 USIC Channel Structure User’s Manual ArchitectureX27, V1.1 2-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview The USIC module contains two independent communication channels, with structure shown in Figure 2-3. The data shift unit and the data buffering of each channel support full-duplex data transfers. The protocol-specific actions are handled by protocol pre-processors (PPP). In order to simplify data handling, an additional FIFO data buffer is optionally available for each USIC module to store transmit and receive data for each channel. This FIFO data buffer is not necessarily available in all devices (please refer to USIC implementation chapter for details). Due to the independent channel control and baud rate generation, the communication protocol, baud rate and the data format can be independently programmed for each communication channel. User’s Manual ArchitectureX27, V1.1 2-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview MultiCAN Module The MultiCAN module contains two independently operating CAN nodes with Full-CAN functionality which are able to exchange Data and Remote Frames via a gateway function. Transmission and reception of CAN frames is handled in accordance with CAN specification V2.0 B (active). Each CAN node can receive and transmit standard frames with 11-bit identifiers as well as extended frames with 29-bit identifiers. Note: The number of available CAN nodes depends on the selected device type. All CAN nodes share a common set of 128 message objects. Each message object can be individually allocated to one of the CAN nodes. Besides serving as a storage container for incoming and outgoing frames, message objects can be combined to build gateways between the CAN nodes or to setup a FIFO buffer. The message objects are organized in double-chained linked lists, where each CAN node has its own list of message objects. A CAN node stores frames only into message objects that are allocated to its own message object list, and it transmits only messages belonging to this message object list. A powerful, command-driven list controller performs all message object list operations. MultiCAN Module Kernel Clock Control CAN Node 1 fCAN Message Object Buffer Address Decoder Linked List Control Port Control CAN Node 0 Interrupt Control TXDC1 RXDC2 TXDC0 RXDC0 CAN Control mc_multican_block2.vsd Figure 2-4 Block Diagram of the MultiCAN Module User’s Manual ArchitectureX27, V1.1 2-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview MultiCAN Features • CAN functionality conforms to CAN specification V2.0 B active for each CAN node (compliant to ISO 11898) • Two independent CAN nodes • Up to 128 independent message objects (shared by the CAN nodes) • Dedicated control registers for each CAN node • Data transfer rate up to 1 Mbit/s, individually programmable for each node • Flexible and powerful message transfer control and error handling capabilities • Full-CAN functionality for message objects: – Can be assigned to one of the CAN nodes – Configurable as transmit or receive objects, or as message buffer FIFO – Handle 11-bit or 29-bit identifiers with programmable acceptance mask for filtering – Remote Monitoring Mode, and frame counter for monitoring • Automatic Gateway Mode support • 16 individually programmable interrupt nodes • Analyzer mode for CAN bus monitoring User’s Manual ArchitectureX27, V1.1 2-29 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 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 and enabled at any time by executing instructions DISWDT and ENWDT. 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 a reset request. The Watchdog Timer is a 16-bit timer, clocked with the system clock divided by 16,384 or 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 3.2 µs and 13.4 s can be monitored (@ 80 MHz). The default Watchdog Timer interval after reset is 6.5 ms (@ 10 MHz). Memory Checker Unit The memory checker module (MCHK) of the XC27x5X supports checking the data consistency of memories, registers (e.g. configuration registers), or communication channels. It calculates a checksum on a block of data, often called cyclic redundancy code (CRC). It is implemented as a parallel signature generation based on a multi input linear feedback shift register (MISR). Being based on a linear feedback shift register (LFSR), it also can generate pseudo-random numbers and cyclic codes. From the programmer's point of view, the MCHK is a set of registers associated with this peripheral. To communicate respective error or operation events a port pin may be used for the signal “MATCH” to generate an external event and an interrupt line may be used for the signal “MISMATCH” to generate an internal event. User’s Manual ArchitectureX27, V1.1 2-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Parallel Ports The XC27x5X derivatives are available in two different packages: • In LQFP-144, they provide up to 119 I/O lines which are organized into 11 input/output ports and 2 input ports. • In LQFP-100, they provide up to 76 I/O lines which are organized into 7 input/output ports and 2 input ports. All port lines are bit-addressable, and all input/output lines can be individually (bit-wise) configured via port control registers. This configuration selects the direction (input/ output), push/pull or open-drain operation, and activation of pull devices for each pin. Edge characteristics (shape) and driver characteristics (output current) of the port drivers can be selected for groups of 4 pins. The I/O ports are true bidirectional ports which are switched to high impedance state when configured as inputs. During the internal reset, all port pins are configured as inputs without pull devices active. All port lines have programmable alternate input or output functions associated with them. These alternate fucntions can be assigned to various port pins to support the optimal utilization for a given application. For this reason, certain functions appear several times in Table 2-3. 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 XC27x5X’s Parallel Ports Port Width 1441) Width 1001) Alternate Functions Port 0 8 8 Address lines, Serial interface lines of USIC and CAN, Input/Output lines for CCU6 Port 1 8 8 Address lines, Serial interface lines of USIC, Input/Output lines for CCU6, OCDS control, interrupts Port 2 14 14 Address and/or data lines, bus control, Serial interface lines of USIC and CAN, Input/Output lines for CCU6 and CAPCOM2, Timer control signals, DAP/JTAG, interrupts, system clock output Port 3 8 --- Bus arbitration signals, Serial interface lines of USIC and CAN User’s Manual ArchitectureX27, V1.1 2-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-3 Summary of the XC27x5X’s Parallel Ports (cont’d) Port Width 1441) Width 1001) Alternate Functions Port 4 8 4 Chip select signals, Serial interface lines of USIC and CAN, Input/Output lines for CAPCOM2, Timer control signals Port 5 16 11 Analog input channels to ADC0, Input/Output lines for CCU6, Timer control signals, DAP/JTAG, OCDS control, interrupts Port 62) 4 3 ADC control lines, Serial interface lines of CAN, Timer control signals, OCDS control Port 7 5 5 ADC control lines, Serial interface lines of USIC and CAN, Input/Output lines for CCU6, Timer control signals, DAP/JTAG, OCDS control, system clock output Port 8 7 --- Serial interface lines of USIC, Input/Output lines for CCU6, DAP/JTAG, OCDS control Port 9 8 --- Serial interface lines of USIC and CAN, Input/Output lines for CCU6, DAP/JTAG, OCDS control Port 10 16 16 Address and/or data lines, bus control, Serial interface lines of USIC and CAN, Input/Output lines for CCU6, DAP/JTAG, OCDS control Port 11 6 --- Serial interface lines of USIC and CAN, Input/Output lines for CCU6 Port 15 8 5 Analog input channels to ADC1, Timer control signals 1) These columns describe the availability of port pins in the different packages. 2) The drivers of these pins are supplied by VDDPA. User’s Manual ArchitectureX27, V1.1 2-32 V1.2, 2008-09 XC27x5X 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 XC27x5X with high flexibility. The system clock fSYS is the reference clock signal, which can be output to the external system. The system clock fSYS can be derived from several internal and external clock sources. The on-chip high-precision oscillator (OSC_HP) 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. An internal clock source can provide a clock signal without requiring an external crystal. The Oscillator Watchdog (OWD) supervises the input clock and enables an emergency clock if the input clock appears as not reliable. User’s Manual ArchitectureX27, V1.1 2-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.5 Power Management The XC27x5X can operate within a wide supply voltage range from 3 V to 5 V. The internal core supply voltage is generated via on-chip Embedded Voltage Regulators and is supervised by on-chip Power Validation Circuits. Two IO power domains help to reduce heat dissipation by supplying the major part of the device with a low voltage (3 V), while still connecting analog 5 V sensor signals to the ADCs (5 V). The XC27x5X provides several means to control the power it consumes either at a given time or averaged over a certain timespan. Three mechanisms can be used (partly in parallel): • Supply Voltage Management allows to switch off the supply voltage. This drastically reduces the power consumed because of leakage current, in particular at high temperature. A power-on reset restarts the system. • Clock Generation Management controls the distribution and the frequency of internal and external clock signals. While the clock signals for currently inactive parts of logic are disabled automatically, the user can reduce the XC27x5X’s CPU clock frequency which drastically reduces the consumed power. External circuitry can be controlled via the programmable frequency output EXTCLK. • Peripheral Management permits temporary disabling of peripheral modules. Each peripheral can separately be disabled/enabled. Note: When selecting the supply voltage and the clock source and generation method, the required parameters must be carefully written to the respective bitfields, to avoid unintended intermediate states. Recommended sequences are provided which ensure the intended operation of power supply system and clock system. User’s Manual ArchitectureX27, V1.1 2-34 V1.2, 2008-09 XC27x5X 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 XC27x5X. The user software running on the XC27x5X can thus be debugged within the target system environment. The OCDS is controlled by an external debugging device via the debug interface and an optional break interface. The debugger controls the OCDS via a set of dedicated registers accessible via the debug 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 debug interface or via the external bus interface for increased performance. For the debug interface two variants can be used: • Debug interface through the DAP port. This interface uses 2 DAP lines. • Debug interface through the IEEE-1149-conforming JTAG port. This interface uses 4 JTAG lines. The optional break interface uses another 2 lines. User’s Manual ArchitectureX27, V1.1 2-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview 2.7 Protected Bits The XC27x5X 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 5.8.2). The following bits are protected: Table 2-4 XC27x5X Protected Bits Register Bit Name Notes RTC_ISNC T14IR, CNT3IR … CNT0IR Interrupt node sharing request flags U0C0_0IC … U0C0_2IC IR USIC0 channel 0 interrupt request flags U0C1_0IC … U0C1_2IC IR USIC0 channel 1 interrupt request flags U1C0_0IC … U1C0_2IC IR USIC1 channel 0 interrupt request flags U1C1_0IC … U1C1_2IC IR USIC1 channel 1 interrupt request flags CAN_0IC … CAN_15IC IR MultiCAN interrupt request flags CCU60_0IC … CCU60_3IC IR CCU60 interrupt request flags CCU61_0IC … CCU61_3IC IR CCU61 interrupt request flags CCU62_0IC … CCU62_3IC IR CCU62 interrupt request flags CCU63_0IC … CCU63_3IC IR CCU63 interrupt request flags SCU_0IC, SCU_1IC IR SCU interrupt request flags RTC_IC RTCIR RTC interrupt request flag EOPIC EOPIR End-of-PEC interrupt request flag CC2_CC16IC … CC2_CC31IC CC16IR … CC31IR CAPCOM2 interrupt request flags CC2_OUT CC0IO … CC15IO Compare output bits GPT12E_T2CON T2CHDIR, T2EDGE GPT1 timer T2 flags User’s Manual ArchitectureX27, V1.1 2-36 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview Table 2-4 XC27x5X Protected Bits (cont’d) Register Bit Name Notes GPT12E_T3CON T3CHDIR, T3EDGE, T3OTL GPT1 timer T3 flags and output toggle latch GPT12E_T4CON T4CHDIR, T4EDGE GPT1 timer T4 flags GPT12E_T6CON T6OTL GPT2 timer T6 output toggle latch GPT12E_T2IC … GPT12E_T6IC T2IR … T6IR GPT timer interrupt request flags GPT12E_CRIC CRIR GPT2 CAPREL interrupt request flag CC2_T7IC, CC2_T8IC T7IR, T8IR CAPCOM2 timer interrupt request flags ADC_0IC … ADC_7IC IR AD Converter interrupt request flags TFR SR0, STKOF, STKUF, SOFTBRK Class A trap flags TFR SR1, MPR, MPW, MPX, UNDOPC, ACER, PRTFLT, ILLOPA Class B trap flags PECISNC C7IR … C0IR All channel interrupt request flags ERU_0IC … ERU_3IC IR External request unit interrupt request flags ----- IR “Unassigned node” interrupt request flags User’s Manual ArchitectureX27, V1.1 2-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Architectural Overview User’s Manual ArchitectureX27, V1.1 2-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3 Memory Organization The memory space of the XC27x5X 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 and Flash, 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. 239...224 On-Chip Program Memory Areas 223...208 207...192 191...176 175...160 ~12 Mbytes External Addressing Capability 159...144 143...128 External Memory Area FF’FFFF H E0’0000 H C0’0000H A0’0000 H 80’0000 127...112 111...96 95...80 79...64 60’0000H 40’0000 63...48 External IO Area 47...32 External Memory Area 20’0000 31...16 15...0 H 16 Mbytes Total Addressing Capability 255...240 H H 00’0000H Total Address Space 16 Mbytes, Segments 255...0 mc_xc16x_mmap.vsd Figure 3-1 Address Space Overview User’s Manual MemoryX2K, V1.2, 2008-09 3-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization The XC27x5X 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 (“little endian”). 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. xxxx’xxxAH xxxx’xxx9H 7 6 … Bits ... 0 xxxx’xxx8H Byte xxxx’xxx7H Byte xxxx’xxx6H Word (High Byte) xxxx’xxx5H Word (Low Byte) xxxx’xxx4H Double Word (High Byte) xxxx’xxx3H Double Word (Third Byte) xxxx’xxx2H Double Word (Second Byte) xxxx’xxx1H Double Word (Low Byte) xxxx’xxx0H xxxx’xxxFH imb_endianess.vsd:byte_orga 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 MemoryX2K, V1.2, 2008-09 3-2 V1.2, 2008-09 XC27x5X 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 XC27x5X contains some reserved areas, so future derivatives can be enhanced in an upward-compatible fashion. Note: Table 3-1 shows the maximum available memory areas. The actual available memory areas depend on the selected device type. Table 3-1 XC27x5X Memory Map 1) Address Area Start Loc. End Loc. Area Size2) IMB register space FF’FF00H FF’FFFFH 256 Bytes Reserved F0’0000H FF’FEFFH < 1 MByte Minus IMB registers Reserved for EPSRAM E8’8000H EF’FFFFH 480 KBytes Mirrors EPSRAM Emulated PSRAM E8’0000H E8’7FFFH up to 32 KBytes With Flash timing Reserved for PSRAM E0’8000H E7’FFFFH 480 KBytes Mirrors PSRAM PSRAM E0’0000H E0’7FFFH up to 32 KBytes Program SRAM Reserved for Flash CD’0000H DF’FFFFH Flash 3 CC’0000H CC’FFFFH 64 KBytes Flash 2 C8’0000H CB’FFFFH 256 KBytes Flash 1 C4’0000H C7’FFFFH 256 KBytes Flash 0 C0’0000H C3’FFFFH 252 KBytes3) External memory area 40’0000H BF’FFFFH 8 MBytes External IO area4) 21’0000H 3F’FFFFH 1984 KBytes Reserved 20’C000H 20’FFFFH 16 KBytes USIC0–3 alternate regs. 20’B000H 20’BFFFH 4 KBytes Accessed via EBC MultiCAN alternate regs. 20’8000H 20’AFFFH 12 KBytes Accessed via EBC Reserved 20’6000H 20’7FFFH 8 KBytes USIC0–3 registers 20’4000H 20’5FFFH 8 KBytes Accessed via EBC MultiCAN registers 20’0000H 20’3FFFH 16 KBytes Accessed via EBC External memory area 01’0000H 1F’FFFFH 1984 KBytes SFR area 00’FE00H 00’FFFFH 0.5 KBytes Dualport RAM (DPRAM) 00’F600H 00’FDFFH 2 KBytes User’s Manual MemoryX2K, V1.2, 2008-09 3-3 Notes 1216 KBytes Minus res. seg. V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Table 3-1 XC27x5X Memory Map (cont’d)1) Address Area Start Loc. End Loc. Area Size2) Reserved for DPRAM 00’F200H 00’F5FFH 1 KBytes ESFR area 00’F000H 00’F1FFH 0.5 KBytes XSFR area 00’E000H 00’EFFFH 4 KBytes Data SRAM (DSRAM) 00’A000H 00’DFFFH 16 KBytes Reserved for DSRAM 00’8000H 00’9FFFH 8 KBytes External memory area 00’0000H 00’7FFFH 32 KBytes Notes 1) Accesses to the shaded areas are reserved. In devices with external bus interface these accesses generate external bus accesses. 2) The areas marked with “<” are slightly smaller than indicated, see column “Notes”. 3) The uppermost 4-Kbyte sector of the first Flash segment is reserved for internal use (C0’F000H to C0’FFFFH). 4) Several pipeline optimizations are not active within the external IO area. This is necessary to control external peripherals properly. User’s Manual MemoryX2K, V1.2, 2008-09 3-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.2 Register Areas The registers controlling the system and peripheral functions of the XC27x5X can be accessed through five address areas. The address areas differ in their access properties. Please refer to Chapter 3.6 and the CPU chapter for further details. The first three areas provide Special Function Registers (SFRs) access capabilities for controlling the system and peripheral functions of the XC27x5X: • • • 512-byte SFR area (located above the DPRAM: 00’FFFFH … 00’FE00H). 512-byte ESFR area (located below the DPRAM: 00’F1FFH … 00’F000H). 4-Kbyte XSFR area (located below the ESFR area: 00’EFFFH … 00’E000H). The USIC and MultiCAN registers are located within the external IO area: • 64-Kbyte external IO area (located in: 20’0000H … 20’FFFFH). The IMB registers are located within a regular memory area. CPU pipeline effects must be regarded for access in this area: • 256-byte IMB SFR area (located in: FF’FF00H … FF’FFFFH). This arrangement provides upward compatibility with the derivatives of the C166 and XC166 families. User’s Manual MemoryX2K, V1.2, 2008-09 3-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) SFR Area Memory Organization SFRs 00'FE00 H 00'FC00 H DPRAM 00'FA00 H 00'F800 H ESFRs EBC Interrupt/PEC XSFR Area CC6 Ports Reserved Reserved Reserved ADC 00'F400 H 00'F200 H 00'F000 H 00'EE00 H 00'EC00 H Upper Half of Data Page 3 Reserved for DPRAM 8 KBytes ESFR Area 00'F600 H 00'EA00 H 00'E800 H 00'E600 H 00'E400 H 00'E200 H 00'E000 H x c2000_regareas. v sd:UM Figure 3-3 Special Function Register Mapping Note: The upper 256 bytes of SFR area, ESFR area, and internal RAM are bitaddressable (see hatched 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 XC27x5X 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! 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. User’s Manual MemoryX2K, V1.2, 2008-09 3-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 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 ;Switch to ESFR area for next 4 instr. MOV STMREL, #data16 ;STMREL uses 8-bit reg addressing BFLDL STMCON, #mask, #data8 ;Bit addressing for bitfields BSET WUCR.CLRTRG ;Bit addressing for single bits MOV T8REL, R1 ;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! MOV T8REL, R1 ;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. The bit-field 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,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 User’s Manual MemoryX2K, V1.2, 2008-09 3-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 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 Registers <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 The XC27x5X 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. User’s Manual MemoryX2K, V1.2, 2008-09 3-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 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. Note: Writing to any byte of the PEC pointers causes the not addressed complementary byte to be cleared. User’s Manual MemoryX2K, V1.2, 2008-09 3-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.3 Data Memory Areas The XC27x5X provides two on-chip RAM areas exclusively 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 3.10). However, both data memory areas provide the fastest access. Depending on the device additional on-chip memory areas may exist with the special purpose to retain data while the system power domain is switched off. The XC27x5X contains: • The Marker Memory (MKMEM). Dual-Port RAM (DPRAM) The XC27x5X 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. 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 bit addressable. Note: Code cannot be executed out of the DPRAM. Note: The locations 00’FBFEH … 00’FC01H of the DPRAM may be altered during the initialization phase after a reset. This area, therefore, should not store data to be preserved beyond a reset. An area of 3 Kbytes is dedicated to DPRAM (00’F200H … 00’FDFFH). The locations without implemented DPRAM are reserved. Data SRAM (DSRAM) The XC27x5X provides 16 Kbytes of DSRAM (00’A000H … 00’DFFFH). 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 (for the range 00’C000H … 00’DFFFH) or to data page 2 (for the range 00’A000H … 00’BFFFH). Any word data access is made on an even byte address. 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. User’s Manual MemoryX2K, V1.2, 2008-09 3-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization An area of 24 Kbytes is dedicated to DSRAM (00’8000H … 00’DFFFH). The locations without implemented DSRAM are reserved. Marker Memory (MKMEM) The MKMEM provides 4 bytes of memory. It can be used to store system state information during power down. The MKMEM consists of 2 16-bit SFRs that are accessible as all other SFRs. Details are described in the SCU chapter. Note: Code cannot be executed out of the MKMEM. User’s Manual MemoryX2K, V1.2, 2008-09 3-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.4 Program Memory Areas The XC27x5X provides two on-chip program memory areas for code/data storage: • • The Program Flash/ROM stores code and constant data. Flash memory is (re-) programmed by the application software or flash loaders, ROM is mask-programmed in the factory. The Program SRAM (PSRAM) stores temporary code sequences and other data. For example higher level boot loader software can be written to the PSRAM and then be executed to program the on-chip Flash memory. IMB Reg. FF'FFFFH FF'FF00 H FF'FFFFH FF'FF00 H Reserved Reserved F0'0000 H Reserved PSRAM Reserved PSRAM F0'0000 H E8'0000 H PSRAM (32 KB) Flash Access Timing E8'8000 H E8'0000 H E0'0000 H E0'8000 H Reserved Flash Area PSRAM (32 KB) SRAM Timing E0'0000 H D0'0000 H Flash 3 (64 KB) Flash 2 (256 KB) C4'0000 H Flash 1 (256 KB) Flash 0 (252 KB) Flash 0 (192 KB) C0'0000 H Reserved (4 KB) Flash 0 (60 KB) C1'0000 H C0'F000 H No software access to this Flash range . C0'0000 H imb _memory_map. v sd Figure 3-4 On-Chip Program Memory Mapping User’s Manual MemoryX2K, V1.2, 2008-09 3-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.4.1 Program/Data SRAM (PSRAM) The XC27x5X provides up to 32 Kbytes of PSRAM (E0’0000H … E0’7FFFH). The PSRAM provides fast code execution without initial delays. Therefore, it supports nonsequential 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 one of its data pages. Any word data access is made on an even byte address. 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 bit-addressable. Note: The upper 256 Bytes of the PSRAM may be altered during the initialization phase after a reset. This area, therefore, should not store data to be preserved beyond a reset. Also, during bootstrap loader operation, the serially received data is stored in the PSRAM starting at location E0’0000H. An area of 512 Kbytes is dedicated to PSRAM (E0’0000H … F7’FFFFH). The locations without implemented PSRAM are reserved. Flash Emulation During code development the PSRAM will often be used for storing code or data that the production chip will later contain in the flash memory. In order to ensure similar execution time the PSRAM supports a second access path in the range E8’0000H … EF’FFFFH with timing parameters that correspond to Flash timing. The number of wait-cycles is determined by the flash access timing configuration (see IMB_IMBCTRL.WSFLASH). Writes are always performed without wait-cycles. This flash access timing imitation is nearly cycle accurate because the same read logic as for reading the flash memory is used1). Discrepancies might occur if the software uses the PSRAM for flash emulation and directly as PSRAM. During emulation access conflicts can cause a slightly different timing as in the product chip where these conflicts do not occur. Another source of timing differences can be access conflicts at the flash modules in the product chip. Data reads and instruction fetches that target different flash modules can be executed concurrently whereas if they target the same flash module they are 1) The dual use of the flash read logic might cause unexpected behavior: while the IMB Core is busy with updating the protection configuration (after startup or after changing the security pages) read accesses to the flash emulation range of the PSRAM are blocked because Flash data reads would be blocked also. User’s Manual MemoryX2K, V1.2, 2008-09 3-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization executed sequentially with the data access as first. In the flash emulation this type of conflict can not occur. The data and the instruction access will both incur the defined number of wait-cycles (as if they would target different flash modules) and if they collide at the PSRAM interface the instruction fetch will see an additional wait-cycle. Data Integrity The PSRAM contains its own error control which can be switched between ECC and parity. Details are described in the SCU chapter. Write Protection As the PSRAM is often used to store timing critical code or constant data it is supplied with a write protection. After storing critical data in the PSRAM the register field IMB_IMBCTRH.PSPROT can be used to split the PSRAM into a read-only and a writable part. Write accesses to the read-only part are blocked and a trap can be activated. 3.4.2 Non-Volatile Program Memory (Flash) The XC27x5X provides up to up to 828 Kbytes of program Flash starting at address C0’0000H. 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 16bit 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. 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 bit-addressable. An area of 2 Mbytes is dedicated to program memory (C0’0000H … DF’FFFFH). The locations without implemented program memory are reserved. A more detailed description can be found in “Embedded Flash Memory” on Page 3-19. User’s Manual MemoryX2K, V1.2, 2008-09 3-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.5 System Stack The system stack may be defined anywhere within the XC27x5X’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 MemoryX2K, V1.2, 2008-09 3-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.6 IO Areas The following areas of the XC27x5X’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 MultiCAN or USIC modules. 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. pre-fetching 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 MemoryX2K, V1.2, 2008-09 3-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.7 External Memory Space The XC27x5X 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 XC27x5X’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. 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 XC27x5X also supports four different bus types: • • • • Multiplexed 16-bit Bus (default after Reset). Multiplexed 8-bit Bus. Demultiplexed 16-bit Bus. Demultiplexed 8-bit Bus. For further details about the external bus configuration and control refer to the External Bus Controller chapter. 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 bit addressable. 3.8 Crossing Memory Boundaries The address space of the XC27x5X 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 (if available), the on-chip LXBusperipherals (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. User’s Manual MemoryX2K, V1.2, 2008-09 3-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 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. 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-Kbytes data page boundaries will use different data page pointers, while the physical locations need not be subsequent within memory. User’s Manual MemoryX2K, V1.2, 2008-09 3-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9 Embedded Flash Memory This chapter describes the embedded flash memory of the XC27x5X: • • • • • • • Section 3.9.1 defines the flash specific nomenclature and the structure of the flash memory. Section 3.9.2 describes the operating modes. Section 3.9.3 contains all operations. Section 3.9.4 gives the details of operating sequences. The three sections Section 3.9.7, Section 3.9.8 and Section 3.9.9 look more into depth of maintaining data integrity and protection issues. Section 3.9.10 discusses Flash EEPROM emulation. Section 3.9.11 describes interrupt generation by the flash memory. The Chapter 3.10 describes how the flash memory is embedded into the memory architecture of the XC27x5X and lists all SFRs that affect its behavior. 3.9.1 Definitions This section defines the nomenclature and some abbreviations as a base for the rest of the document. The used flash memory is a non-volatile memory (“NVM”) based on a floating gate one-transistor cell. It is called “non-volatile” because the memory content is kept when the memory power supply is shut off. Logical and Physical States Flash memory content can not be changed directly as in SRAMs. Changing data is a complicated process with a typically much longer duration than reading. • • Erasing: The erased state of a cell is logical 0. Forcing an flash cell to this state is called “erasing”. Erasing is possible with a minimum granularity of one page (see below). A device is delivered with completely erased flash memory. Programming: The programmed state of a cell is logical 1. Changing an erased cell to this state is called “programming”. A page must only be programmed once and has to be erased before it can be programmed again. The above listed processes have certain limitations: • • Retention: This is the time during which the data of a flash cell can be read reliably. The retention time is a statistical figure that depends on the operating conditions of the flash array (temperature profile) and the accesses to the flash array. With an increasing number of program/erase cycles (see endurance) the retention is lowered. Drain and gate disturbs decrease data retention as well. Endurance: As described above the data retention is reduced with an increasing number of program/erase cycles. A flash cell incurs one cycle whenever its page or sector is erased. This number is called “endurance”. As said for the retention it is a statistical figure that depends on operating conditions and the use of the flash cells and not to forget on the required quality level. User’s Manual MemoryX2K, V1.2, 2008-09 3-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization • Drain Disturb: Because of using a so called “one-transistor” flash cell each program access disturbs all pages of the same sector slightly. Over long these “drain disturbs” make 0 and 1 values indistinguishable and thus provoke read errors. This effect is again interrelated with the retention. A cell that incurred a high number of drain disturbs will have a lower retention. The physical sectors of the flash array are isolated from each other. So pages of a different sector do not incur a drain disturb. This effect must be therefor considered when the page erase feature is used. The durations of programming and erasing as well as the limits for endurance, retention and drain disturbs are documented in the data sheet. Attention: No means exist in the device that prevent the application from violating these limitation. Array Structure The flash memory is hierarchically structured: • • • • • Block: A block consists of 128 user data bits (i.e. 16 bytes) and 9 ECC bits. One read access delivers one block. Page: A page consists of 8 blocks (i.e. 128 bytes). Programming changes always complete pages. Sector: A sector consists of 32 pages (i.e. 4096 bytes). The pages of one sector are affected by drain disturb as described above. The pages of different sectors are isolated from each other. Array: Each 256 KB array has 64 sectors1) and the 64 KB array has 16 sectors. Usually when referring to an “array” this contains as well all accompanying logic as assembly buffer, high voltage logic and the digital logic that allows to operate them in parallel. Memory: The complete flash memory of the XC27x5X consists of 4 flash arrays. This structure of the 256 KB array is visualized in Figure 3-5. The structure of the 64 KB array is analog. 1) In the Flash0 one sector is reserved for device internal purposes. It is not accessible by software. User’s Manual MemoryX2K, V1.2, 2008-09 3-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 256 KB Array Sector Number 63 Page Number Sector Block Number 31 Page 7 2 2 2 1 1 1 Sector 0 0 Page Sector [5:0] Page [4: 0] Block [ 2:0] Block 137 Bits Combined flash memory byte address Array [1:0] 0 Byte [ 3:0] [ .. ][. . . . . .][. . . . .][. . .][. . . .] 9 Bits ECC 128 Bits Data flas h_array_us erview_diagram .vs d Figure 3-5 3.9.2 Flash Structure Operating Modes The IMB and the flash memory and each flash module have certain modes of operation. Some modes define clocking and power supply and the operating state of the analog logic as oscillators and voltage pumps. Overall system modes (e.g. startup mode) influence the behavior or the flash memory as well. Other modes define the functional behavior. These will be discussed here. 3.9.2.1 Standard Read Mode After reset and after performing a clean startup the flash memory with all its modules is in “standard read mode”. In this mode it behaves as an on-chip ROM. This mode is entered: • • • After reset when the complete start-up has been performed. After completion of a longer lasting command like “erase” or “program” which is acknowledged by clearing the “busy” flag. Immediately after each other command execution. User’s Manual MemoryX2K, V1.2, 2008-09 3-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization • In case of detecting an execution error like attempting to write to a write protected range, sending a wrong password, after all sequence errors. For the long lasting commands the read mode stays active until the last command of the sequence is received and the operation is started. 3.9.2.2 Command Mode After receiving the last command of a command sequence the addressed flash module (not the whole flash memory!) is placed into command mode. For most commands this will not be noticed by the user as the command executes immediately and afterwards the flash module is placed again into read mode. For the long lasting commands the flash module stays in command mode for several milliseconds. This is reported by setting the corresponding “busy” flag. The data of a busy flash module cannot be read but other not busy flash modules stay readable. New command sequences are generally not accepted and cause a sequence error until the running operation has finished. In certain cases however new command sequences are accepted in order to enable concurrent programming and erase of independent flash modules. Read accesses to busy flash modules stall the CPU until the read mode is entered again. A stalled CPU responds only to the reset. As no interrupts can be handled this state must be avoided. Nevertheless this feature can be used to execute code from a flash module that erases or programs data in the same flash module. Note: Because command sequences to busy flash memory are not always rejected by the hardware with a sequence error it is necessary to handle all commands more careful than in previous device generations that didn’t support concurrent processes. A new command sequence shall be only be issued to a flash module after checking that it is not busy anymore. This is especially vital when using the “stall CPU when reading busy flash” feature. Further advice can be found in Section 3.9.5 (sequence errors) and Section 3.9.6 (concurrent processes). 3.9.2.3 Page Mode The page mode is entered with the “Enter Page Mode” command. Please find its description below. A flash module that is in page mode can still be read (so it is concurrently in “read mode”). At a time only one flash module can be in page mode. When the flash memory is in page mode — i.e. one of the flash modules is in page mode — some command sequences are not allowed. These are all erase sequences and the “change read margin” sequence. These are ignored and a sequence error is reported. User’s Manual MemoryX2K, V1.2, 2008-09 3-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.3 Operations The flash memory supports the following operations: • • • Instruction fetch. Data read. Command sequences to change data and control the protection. 3.9.3.1 Instruction Fetch from Flash Memory Instructions are fetched by the PMU in groups of aligned 64 bits. These code requests are forwarded to the flash memory. It needs a varying number of cycles (depending on the system clock frequency) to perform the read access. The number of cycles must be known to the IMB Core because the flash does not signal data availability. The number of wait states is therefore stored in the IMB_IMBCTRL register. The complete duration of a flash read access is: IMB_IMBCTRL.WSFLASH + 1 cycles. Consult the data sheet for correct values of WSFLASH dependent on the system clock frequency and device. One read access to the flash memory delivers 128 data bits and a 9-bit ECC value. The ECC value is used to detect and possibly correct errors. The addressed 64-bit part of the 128-bit chunk is sent to the PMU. The complete 128 data bits and the 9 ECC bits are stored in the IMB Core with their address. If a succeeding fetch request matches this address the data is delivered from the buffer without performing a read access in the flash memory. The delivery from the buffer happens after one cycle. The flash read waitcycles are not waited. The stored data are a kind of instruction cache. In order to support self-modifying code (e.g. boot loaders) this cache is invalidated when the corresponding address is written (i.e. erased or programmed). In addition to this fetch buffer the IMB Core has an additional performance increasing feature — the Linear Code Pre-Fetch. When this feature is enabled with IMB_IMBCTRL.DLCPF = 0 the IMB Core fetches autonomously the following instructions while the CPU executes from its own buffers or the fetch buffer. As this feature is fetching only the linear successors (it does not analyze the code stream) it is most effective for code with longer linear sequences. For code with a high density of jumps and calls it can even cause a reduction of performance and should be switched off. User’s Manual MemoryX2K, V1.2, 2008-09 3-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.3.2 Data Reads from Flash Memory Data reads are issued by the DMU. Data is always requested in 16-bit words. The flash memory delivers for every read request 128 bits plus ECC as described in “Instruction Fetch from Flash Memory” on Page 3-23. The IMB Core has to get all 128 bits to evaluate the ECC data. The requested 16 bits will be delivered to the DMU. All data and ECC bits are kept in the data register and their address is kept in the address register. For all following data reads the address is compared with the address register and in case of a match the data is delivered after one cycle from the data register. Every data read that is not delivered from this cache invalidates the cache content. When the requested data arrives the cache contains again valid data. This small data cache is invalidated when a write (i.e. erase or program) access to this address happens. For data reads the IMB Core does not perform any autonomous pre-fetching. 3.9.3.3 Data Writes to Flash Memory Flash memory content can not be changed by directly writing data to this memory. Command sequences are used to execute all other operations in the flash except reading. Command sequences consist of data writes with certain data to the flash memory address range. All data moves targeting this range are interpreted as command sequences. If they do not match a defined one or if the IMB Core cannot accept a new one because it is busy a sequence error is reported. User’s Manual MemoryX2K, V1.2, 2008-09 3-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.3.4 Command Sequences As described before changing data in the flash memory is performed with command sequences. Table 3-3 Command Sequence Overview Command Sequence Description Details on Page Reset to Read Reset Flash into read mode and clear error flags. Page 3-27 Clear Status Clear error and status flags. Page 3-27 Change Read Margin Change read margins. Page 3-27 Enter Page Mode Prepare page for programming. Page 3-28 Enter Security Page Mode Prepare security page for programming. Page 3-29 Load Page Word Load page with data. Page 3-30 Program Page Start page programming process. Page 3-31 Erase Sector Start sector erase process. Page 3-32 Erase Page Start page erase process. Page 3-33 Erase Security Page Start security page erase process. Page 3-33 Disable Read Protection Disable temporarily read protection with password. Page 3-34 Disable Write Protection Disable temporarily write protection with password. Page 3-35 Re-Enable Read/Write Protection Re-enable protection. Page 3-36 User’s Manual MemoryX2K, V1.2, 2008-09 3-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.4 Details of Command Sequences The description defines the command sequence with pseudo assembler code. It is “pseudo” because all addresses are direct addresses which is generally not possible in real assembler code. The commands are called by a sequence of one to six data moves into the flash memory range. The data moves must be of the “word” type, i.e. not byte move instructions. The following sections describe each command. The following abbreviations for addresses and data will be used: • • • • • • • • PA: “Page Address”. This is the base address of the destination page. For example the very first page has the address C0’0000H. The page 13 of the second array has the PA = C0’0000H + 1·256·1024 (for the array) + 0·4·1024 (for the sector) + 13·128 (for the page) = C4’0680H. SECPA: “Security Page Address”. This is the virtual address of a security page. It is “virtual” because SECPA is just used as argument of the command sequence to identify the security page but the physical storage of the security page is hidden. Two security pages are defined: SecP0: address C0’0000H. SecP1: address C0’0080H. WD: “Write Data”. This is a 16-bit data word that is written into the assembly buffer. SA: “Sector Address”. This is the physical sector number as defined in Figure 3-6 based on the address of the flash module. Two examples as clarification: 1. Physical sector number 16 of the first array that is based on C0’0000H is addressed with SA = C0’0000H + 16·4·1024 = C1’0000H. 2. The second 256 KB array has the base address C4’0000H (as shown in Table 3-1). So its physical sector number 3 has the SA = C4’0000H + 3·4·1024 = C4’3000H. PWD: “Password”. This is a 64-bit password. It is transferred in 4 16-bit data words PWD0 = PWD[15:0], PWD1 = PWD[31:16], PWD2 = PWD[47:32] and PWD3 = PWD[63:48]. Address XX followed by two hexadecimal digits, for example “XXAAH”. If the command targets a certain flash module the XX must be translated to its base address. So “XXAAH” means C0’00AAH for all commands addressing flash 0, C4’00AAH for flash 1 and C8’00AAH for flash 2. If a command (e.g. “Clear Status”) addresses the complete flash memory the base address of flash module 0 must be used. Data XX followed by two hexadecimal digits, e.g. XXA5H. This is a “don’t care” data word where only the low byte must match a certain pattern. So in this example all data words like 12A5H or 79A5H can be used. MR: “Margin”. This 8-bit number defines the read margin. MR can take the values 00H (normal read), 01H (hard read 0), 02H (alternate hard read 0), 05H (hard read 1), 06H (alternate hard read 1). All other values of MR are reserved. User’s Manual MemoryX2K, V1.2, 2008-09 3-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Reset to Read Arguments: – Definition: MOV XXAAH, XXF0H Timing: One cycle command that does not set any “BUSY” flags. But note that an immediately following write access to the IMB Core is stalled for a few clock cycles during which the IMB Core is busy with aborting a previous command. Description: The internal command state machine is reset to initial state and returns to read mode. An already started programming or erase operation is not affected and will be continued (the “Reset to Read” command — i.e. all commands — will anyhow not be accepted while the IMB Core is busy). The “Reset to Read” command is a single cycle command. It can be used during a command sequence to reset the command interpreter and return the IMB Core into its initial state. It clears also all error flags in the Flash Status Register IMB_FSR and an active page mode is aborted. “Reset to Read” can not be used to abort an active command mode. When at least one flash module is busy this command is rejected with SQER1). This command clears: PROER, PAGE, SQER, OPER, ISBER, IDBER, DSBER, DDBER. Clear Status Arguments: – Definition: MOV XXAAH, XXF5H Timing: 1-cycle command that does not set any busy flags. Description: The flags OPER, SQER, PROER, ISBER, IDBER, DSBER, DDBER in Flash status register are cleared. Additionally, the process status bits (PROG, ERASE, POWER, MAR) are cleared. This command must not be issued when any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. Change Read Margin Arguments: MR 1) In the XC27x5X there is one exception to this rule: when one flash module is busy with program or erase and the FAPI has received some but not all command cycles of a concurrently executable command sequence (“Erase Sector”, “Erase Page”, “Enter Page Mode”, “Load Page Word”, “Program Page”) then a Reset to Read is performed without issuing a sequence error. User’s Manual MemoryX2K, V1.2, 2008-09 3-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Definition: MOV XXAAH, XXB0H MOV XX54H, XXMRH Timing: 2-cycle command that sets “BUSY” of the addressed flash module for around 30 micro seconds. Description: This command sequence changes the read margin of one flash module. The address XX of the second move identifies the targeted flash module. The flash module needs some time to change its read voltage. During this time BUSY is set and this flash module cannot be accessed. The other flash modules stay readable. The argument “MR” defines the read margin: • • • • • • 00H: normal read margin. 01H: hard read 0 margin. 02H: alternate hard read 0 margin. 05H: hard read 1 margin. 06H: alternate hard read 1 margin. Other values: reserved. For understanding the read margins please refer to “Margin Reads” on Page 3-41. This command must not be issued when the flash memory is in page mode or any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. Note: As noted in “Margin Control” on Page 3-68 the command sequences “Program Page”, “Erase Sector”, “Erase Page” and “Erase Security Page” reset the read margin back to 00H, i.e. to the normal read margin. The same happens in case of a flash wake-up. Enter Page Mode Arguments: PA Definition: MOV XXAAH, XX50H MOV PA, XXAAH Timing: 2-cycle command that sets “BUSY” of the addressed flash module for around 20 clock cycles1). Description: The page mode is entered to prepare a page programming operation on page address PA. (Write data are accepted only with the “Load Page Word” command.) 1) When this command is used to abort a page mode of an other flash module the duration increases to around 30 clock cycles. User’s Manual MemoryX2K, V1.2, 2008-09 3-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization With this command, the IMB Core initializes the write pointer of its block assembly register to zero so that it points to the first word. The page mode is indicated in the status register IMB_FSR_BUSY with the PAGE bit, separately for each flash module. The page mode and the read mode are allowed in parallel at the same time and in the same flash module so the flash module stays readable. When the addressed page PA is read the content of the flash memory is delivered. The page mode can be aborted and the related PAGE bit in IMB_FSR_BUSY be cleared with the “Reset to Read” command. A new “Enter Page Mode” command during page mode aborts the actual page mode, which is indicated with the error flag SQER, and restarts a new page operation. So as mentioned above only one of the flash modules can be in page mode at a time. If one of the erase commands or the “Change Read Margin” command are received while in page mode it is ignored and a sequence error is reported. The page mode can be entered in one flash module while others are busy with executing a user data erase or program command, i.e. not while programming or erasing security pages or other blocking sequences. If write protection is installed for the sector to be programmed, the “Enter Page Mode” command is only accepted when write protection has before been disabled using the unlock command sequence “Disable Write Protection” with four passwords. If global write protection is installed with read protection, also the command “Disable Read Protection” can be used if no sector specific protection is installed. If write protection is not disabled when the “Enter Page Mode” command is received, the command is not executed, and the protection error flag PROER is set in the IMB_FSR_PROT. Note: In previous device families (e.g. XC16x) the “Enter Page Mode” did not set “BUSY”. In these devices the “Load Page Word” could be sent directly after issuing “Enter Page Mode”. In XC27x5X it must be waited until “BUSY” clears before sending the “Load Page Word” command sequence. Enter Security Page Mode Arguments: SECPA Definition: MOV XXAAH, XX55H MOV SECPA, XXAAH Timing: 2-cycle command that sets “BUSY” of flash module 0 for around 100 clock cycles. Description: This command is identical to the “Enter Page Mode” command (see above), with the following exceptions: The addressed page (SECPA) belongs to the security pages of the flash memory and not to the user flash range. This command can only be executed when neither flash write protection nor read protection are active (RPA = 0 and WPA = 0), otherwise it fails with PROER. User’s Manual MemoryX2K, V1.2, 2008-09 3-29 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization This command is refused with SQER when any of the flash modules is in command mode. The use of this command to install passwords and to disable them again is described in “Protection Handling Details” on Page 3-44. Load Page Word Arguments: WD Definition: MOV XXF2H, WD Timing: 1-cycle command that does not set any “BUSY” flags. But note that an immediately following write access to the IMB Core or read from the flash memory is stalled for a few clock cycles if it arrives while the IMB Core is busy with copying its block assembly register content into the flash module assembly buffer. During this stall time the CPU can not perform any action! So either the user software can accept this stall time (which must be taken into account for the worst-case interrupt latency) or the software must avoid the blocking accesses. Description: Load the IMB Core block assembly register with a 16-bit word and increment the write pointer. The 128 byte assembly buffer (i.e. a complete page) is filled by a sequence of 64 “Load Page Word” commands. The word address is not determined by the command but the “Enter Page Mode” command sets a write word pointer to zero which is incremented after each “Load Page Word” command. This (sequential) data write access to the block assembly register belongs to and is only accepted in Page Mode. The command address of this single cycle command is always the same (F2H). These low order address bits also identify the “Load Page Word” command and the sequential write data to be loaded into the block assembly register. The high order bits XX should address the target page. The IMB Core takes always the page address that was used by the last “Enter Page Mode” command. When the 128-bit block assembly register of the IMB Core is filled completely after 8 “Load Page Word” commands the IMB Core calculates the 9 ECC bits and transfers the block into the assembly buffer of the flash module. After that it sets the write pointer of the block assembly register back to zero. The following 8 “Load Page Word” commands fill again the block. After all 8 blocks are filled the “Program Page” command can be used to trigger the program process that transfers the assembly buffer content into the flash array. While the IMB Core transfers the completed block assembly register to the flash module it can not accept new data for a few cycles. A “Load Page Word” command arriving during this time is stalled by the IMB Core. If “Program Page” is called before all blocks of the assembly buffer have received new data then the remaining bits are cleared. User’s Manual MemoryX2K, V1.2, 2008-09 3-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization If more than 8 times 8 commands are used the additional data is lost. The overflow condition is indicated by the sequence error flag, but the execution of a following “Program Page” command is not suppressed (the page mode is not aborted). When a “Load Page Word” command is received and the flash is not in page mode, a sequence error is reported in IMB_FSR_OP with SQER flag. In case of a new “Enter Page Mode” command or a “Reset to Read” command during page mode, or in case of an Application Reset, the write data in the assembly buffer is lost. The current page mode is aborted and in case of a new “Enter Page Mode” command entered again for the new address. Program Page Arguments: – Definition: MOV XXAAH, XXA0H MOV XX5AH, XXAAH Timing: 2-cycle command that sets “BUSY” of the selected flash module for the whole programming duration. The IMB Core is blocked a few clock cycles after receiving this command and again a few clock cycles before finishing the programming. Write accesses to the flash memory range to execute another command sequence during these times stall the CPU. Description: The assembly buffer of the flash module is programmed into the flash array. If the last block of data was not filled completely this command finalizes its ECC calculation and copies its data into the assembly buffer before it starts the program process. The selection of the flash module and the page to be programmed depends on the page address used by the last “Enter Page Mode” command. The user software should always address the targeted page. The programming process is autonomously performed by the selected flash module. The CPU is not occupied and can continue with its application. The “Program Page” command is only accepted if the addressed flash module is in Page Mode (otherwise, a sequence error is reported instead of execution). With the “Program Page” command, the page mode is terminated, indicated by resetting the related PAGE flag and the command mode is entered and the PROG flag in the status register IMB_FSR_OP is activated and the related BUSY flag is set in IMB_FSR_BUSY. When the program process has finished BUSY is cleared but PROG stays set. It indicates which operation has finished and will be cleared by a Power-On Reset or by “Clear Status”. Read accesses to the busy flash module are not possible. Reading a busy flash module stalls until the flash module becomes ready again. User’s Manual MemoryX2K, V1.2, 2008-09 3-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization If write protection is active for the sector to be programmed, the “Program Page” command is not accepted because the Flash is not in Page Mode (see description of the “Enter Page Mode” command). If the page to be programmed is a security page (accepted only in security page mode), the new protection configuration (including keywords or protection confirmation code) is valid directly after execution of this command. During its execution all commands are rejected with a sequence error. While the IMB Core reads the new protection configuration all DMU accesses to any flash module are stalled. Erase Sector Arguments: SA Definition: MOV XXAAH, XX80H MOV XX54H, XXAAH MOV SA, XX33H Timing: 3-cycle command that sets BUSY of the addressed flash module for the whole erasing duration. The IMB Core is blocked a few clock cycles after receiving this command and again a few clock cycles before finishing the erasing. Write accesses to the flash memory range during these times stall the CPU. Description: The addressed physical sector in the flash array is erased. Following data reads deliver all-zero data with correct ECC. The erasing process is autonomously performed by the selected flash module. The CPU is not occupied and can continue with its application. The sector to be erased is addressed by SA (sector address) in the last command cycle. With the last cycle of the “Erase Sector” command, the command mode is entered, indicated by activation of the ERASE flag in IMB_FSR_OP and after start of erase operation also by the related busy flag in the status register IMB_FSR_BUSY. The BUSY flag is cleared after finishing the operation but ERASE stays set. It can be cleared by a Power-On Reset or the “Clear Status” command. Read accesses to the busy flash module are not possible. Read accesses to the not busy flash module are especially supported. Reading a busy flash module stalls until the flash module becomes ready again. If write protection is installed for the sector to be erased, the Erase Sector command is only accepted when write protection has before been disabled using the unlock command sequence “Disable Write Protection”. If global write protection is installed with read protection, also the command “Disable Read Protection” can be used if no sector specific protection is installed. If write protection is not disabled when the “Erase User’s Manual MemoryX2K, V1.2, 2008-09 3-32 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Sector” command is received, the command is not executed, and the protection error flag PROER is set in the IMB_FSR_PROT. This command must not be issued when the flash memory is in page mode. In this case it is ignored and a sequence error is reported. Erase Page Arguments: PA Definition: MOV XXAAH, XX80H MOV XX54H, XXAAH MOV PA, XX03H Timing: 3-cycle command that sets BUSY of the addressed flash module for the whole erasing duration. The IMB Core is blocked a few clock cycles after receiving this command and again a few clock cycles before finishing the erasing. Write accesses to the flash memory range during these times stall the CPU. Description: The addressed page is erased. Following data reads deliver all-zero data with correct ECC. With the last cycle of the “Erase Page” command, the command mode is entered, indicated by activation of the ERASE flag in IMB_FSR_OP and after start of erase operation also by the related BUSY flag in the status register IMB_FSR_BUSY. BUSY is cleared automatically after finishing the operation but ERASE stays set. It is cleared by a Power-On Reset or the “Clear Status” command. Read accesses to the busy flash array are not possible. Read accesses to the not busy flash modules are especially supported. Reading a busy flash module stalls until the flash module becomes ready again. If the page to be erased belongs to a sector which is write protected, the command is only executed when write protection has before been disabled (see “Erase Sector” command). In case of using the page erase care must be taken not to exceed the drain disturb limit of the other pages of the same sector. This command must not be issued when the flash memory is in page mode. In this case it is ignored and a sequence error is reported. Erase Security Page Arguments: SECPA Definition: MOV XXAAH, XX80H MOV XX54H, XXA5H MOV SECPA, XX53H User’s Manual MemoryX2K, V1.2, 2008-09 3-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Timing: 3-cycle command that sets BUSY of flash module 0 for the whole erasing duration. Description: The addressed security page is erased. This command is identical to the “Erase Page” command with the following exceptions: The addressed page (SecP0 or SecP1) belongs not to the user visible flash memory range. This command can only be executed after disabling of read protection and of sector write protection. See “Protection Handling Examples” on Page 3-51 for a detailed description of reprogramming security pages. The structure of the two security pages (SecP0 and SecP1) is described in “Layout of the Security Pages” on Page 3-50. After erasing a security page the new protection configuration (including keywords or protection confirmation code) is valid directly after execution of this command. While the IMB Core reads the protection configuration all DMU accesses to any flash module are stalled. This command must not be issued when the flash memory is in page mode or any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. Disable Read Protection Arguments: PWD Definition: MOV MOV MOV MOV MOV MOV XX3CH, XX54H, XXAAH, XX54H, XXAAH, XX5AH, XXXXH PWD0 PWD1 PWD2 PWD3 XX55H Timing: 6-cycle command that does not set any busy flag. Description: Disable temporarily Flash read protection and — if activated — global write protection of the whole flash memory. The RPA bit in IMB_IMBCTRH is reset. This is a protected command sequence, using four user defined passwords to release this command or to check the programmed keywords. For every password one command cycle is required. If the second or fourth password represents the code of the “Reset to Read” command, it is interpreted as password and the reset is not executed. The 16-bit passwords are internally compared with the keywords out of the “Security Page 0”. If one or more passwords are not identical to their related keywords, the protected sectors remain in the locked state and a protection error (PROER) is indicated in the Flash status register. In this case, a new “Disable Read Protection” command or User’s Manual MemoryX2K, V1.2, 2008-09 3-34 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization a “Disable Write Protection” command is only accepted after the next Application Reset. Note: During execution of the “Disable Read” (or Write) Protection command a password compare error is only indicated after all four passwords have been compared with the related keywords. Note: This command sequence is also used to check the correctness of keywords before the protection is confirmed in the Security Page 1. A wrong keyword is indicated by the IMB_FSR_PROT flag PROER. After correct execution of this command, the whole flash memory is unlocked and the read protection disable bit RPRODIS is set in the Flash Status Register (IMB_FSR_PROT). Erase and program operations on all sectors are then possible, if the flash memory was also globally write protected (WPA=1), and if they are not separately write protected. The read protection (including global write protection, if so selected) remains disabled until the command “Re-Enable Read/Write Protection” is executed, or until the next Application Reset (including HW and SW reset). This command must not be issued when any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. Disable Write Protection Arguments: PWD Definition: MOV MOV MOV MOV MOV MOV XX3CH, XX54H, XXAAH, XX54H, XXAAH, XX5AH, XXXXH PWD0 PWD1 PWD2 PWD3 XX05H Timing: 6-cycle command that does not set any busy flag. Description: Disable temporarily the global flash write protection or/and the sector write protection of all protected sectors. The WPA bit in IMB_IMBCTRH is reset. This is a protected command sequence, using four user defined passwords to release this command (as described above for the “Disable Read Protection” command). After correct execution of this command, all write-protected sectors are unlocked, which is indicated in the Flash Status Register (IMB_FSR_PROT) with the WPRODIS bit. Erase and program operations on all sectors are now possible, until • • The command “Re-Enable Read/Write Protection” is executed, or The next Application Reset (including HW and SW reset) is received. This command must not be issued when any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. User’s Manual MemoryX2K, V1.2, 2008-09 3-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Re-Enable Read/Write Protection Arguments: – Definition: MOV XX5EH, XXXXH Timing: 1-cycle command that does not set any busy flags. Description: Flash read and write protection is resumed. This single-cycle command clears RPRODIS and WPRODIS. The IMB Core is triggered to restore the protection states RPA and WPA from the content of the security page 0 as defined in Table 3-5 ““Flash State” Determining RPA and WPA” on Page 3-47. So in effect this command resumes all kinds of temporarily disabled protection installations. This command is released immediately after execution. This command must not be issued when any of the flash modules is in command mode. In this case it is ignored and a sequence error is reported. 3.9.5 Sequence Errors A word (i.e. 16-bit) data move into the flash address range is interpreted by the command interpreter as command sequence. All byte moves are ignored and cause a sequence error which is reported by setting the bit SQER. As soon the command interpreter detects that the data moves can’t be executed as legal sequence it reports the sequence error. Note: Data moves addressing not implemented flash areas or powered-down flash modules don’t enter the command interpreter and consequently can’t cause a sequence error. Usually the next correct command sequence will cause the sequence error because it is interpreted as continuation of the previous one. So instead of checking only for the absence of SQER the other flags (e.g. PAGE, PROG, ERASE) can be further evaluated. For an example see Section 3.9.6. Generally each data move received while at least one flash module is BUSY causes a sequence error. But in order to support concurrent execution of command sequences this is under certain conditions not done. A SQER is reported under the following conditions: • • • If one of the flash modules is in command mode and the running command does not allow concurrent execution a SQER is reported immediately. If at least one of the flash modules is in command mode and the running command allows concurrent execution SQER is only reported when the new command targets a busy flash module. If at least one of the flash modules is in command mode SQER is reported as soon as a command cycle is detected that can not belong to a command sequence that allows concurrent execution (i.e. when the received data does not belong to “Enter Page Mode”, “Load Page Word”, “Program Page” or “Erase Page”). User’s Manual MemoryX2K, V1.2, 2008-09 3-36 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization The concurrency issues are summarized in Table 3-4. Table 3-4 Concurrency Issues New sequence while any module is in mode: Page Mode Busy with normal erase or program Busy with blocking sequence1) Reset to Read Resets page mode SQER2) SQER2) Enter Page Mode SQER and Reenters page mode OK3) SQER Enter Sec. Page Mode SQER and Reenters page mode SQER SQER Load Page Word OK OK in page mode SQER/–4) Program Page OK OK in page mode SQER/–4) Erase Page/Sector SQER OK3) SQER Erase Sec. Page SQER SQER SQER *Protection OK SQER SQER Clear Status OK SQER SQER Change Read Margin SQER SQER SQER 1) “Blocking sequences” are: “Erase Security Page”, “Program Page” for a security page, “Change Read Margin”, “Enter Page Mode”, “Enter Security Page Mode” only while these set busy. 2) As described in “Reset to Read” on Page 3-27 there is one exception to this rule. 3) If the new command sequence targets a different flash module that is in read mode else SQER. 4) Situation can not occur because “Program Page” is only allowed in page mode and page mode could not be entered. Other conditions that cause a sequence error were mentioned above in the command descriptions. 3.9.6 Instructions for Executing Program and Erase Jobs User’s Manual MemoryX2K, V1.2, 2008-09 3-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Concurrently All flash modules1) can be programmed and erased concurrently. This is however an exceptional case for high-speed flash programming. In the normal case at most one flash module shall be busy while the others can be read. The limitations reported above in the command sequence descriptions enforce certain behavior for concurrent processes: • • • • A programming task shall be started in one not interrupted sequence: “Enter Page Mode”, then 64 “Load Page Word” and finally the “Program Page”. No other command sequence on any other flash module shall interrupt this sequence. All security page handling shall be done while all flash modules are in read mode. Clearing of error and status flags is as well only possible when all flash modules are in read mode. An exception is the flash module specific handling via IMB_ECC_STAT. The IMB Core can only finish an ongoing program or erase task successfully when it is not busy with interpreting a command sequence (i.e. the busy of the ongoing tasks is only cleared when the IMB Core is ready to accept a new command sequence and no new command sequence has been started but not completed). So the required sequence for programming flash modules concurrently is as follows: 1. Send the “Erase Sector” command sequence to each flash module. 2. Wait until all “BUSY” flags are cleared. During this time the data for programming can be read from external. 3. Send “Enter Page Mode”, 64 “Load Page Word” and the “Program Page” to the first flash module. Continue this sequence with the other flash modules. 4. Wait until all “BUSY” flags are cleared. This time can be used to read the data for programming the next pages from external. 5. Verify the programmed data of all flash modules. 6. Continue the steps 3 to 5 until all pages of the erased sectors are programmed. 7. Continue the steps 1 to 6 until all sectors are programmed. The recommend sequence which detects incorrect sequences as early as possible is as follows: 1. “Clear Status” and check that SQER is 0. 2. Send the “Erase Sector” command sequence to each flash module. Check for SQER after issuing each sequence. 3. Wait until all “BUSY” flags are cleared. During this time the data for programming can be read from external. 4. Check for SQER which would indicate an incorrectly issued command sequence. 1) Additional constraints may apply due to power supply and other device specific reasons. The allowed concurrent processes (including read) are described in the data sheet. This section describes only the logic hardware capabilities. User’s Manual MemoryX2K, V1.2, 2008-09 3-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 5. “Clear Status” and check again for SQER. If SQER would be set after “Clear Status” a previous “Erase Sector” hasn’t been completed. 6. Send “Enter Page Mode”, check if the PAGE flag was set and check if SQER stays 0, send the 64 “Load Page Word” and the “Program Page” to the first flash module. Check if the PAGE flag is cleared and SQER stays cleared after “Program Page” is accepted. Continue this sequence with the other flash modules. 7. Wait until all “BUSY” flags are cleared. This time can be used to read the data for programming the next pages from external. 8. Verify the programmed data of all flash modules. 9. Continue the steps 6 to 8 until all pages of the erased sectors are programmed. 10. Continue the steps 1 to 9 until all sectors are programmed. User’s Manual MemoryX2K, V1.2, 2008-09 3-39 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.7 Data Integrity This section describes means for detecting and preventing the inadvertent modification of data in the flash memory. 3.9.7.1 Error Correcting Codes (ECC) With very low probability a flash cell can become disturbed or lose its data value faster than specified. In order to reach the defined overall device reliability each 128-bit block of flash data is accompanied with a 9-bit ECC value. This redundancy supplies SECDED capability, meaning “single error correction and double error detection”. All single bit errors are corrected (and the incident is detected), all double bit errors are detected and even most triple bit errors are detected but some of these escape as valid data or corrected data. A detected error is reported in the register IMB_FSR_PROT and IMB_ECC_STAT. Software can select which type of error should trigger a trap by the means of register IMB_INTCTR. In the system control further means exist to modify the handling of errors (see “SCU Trap Control Registers” on Page 8-184). The enabled trap requests by the flash module are handled there as “Flash Access Trap”. In case of a double-bit error the read data is always replaced with a dummy data word. 3.9.7.2 Aborted Program/Erase Detection Where the ECC should protect from intrinsic failures of the flash memory that affect usually only single bits; an interruption of a running program or erase process might cause massive data corruption: • • The erase process programs first all cells to 1 before it erases them. So depending on the time when it is interrupted the data might be in a different state. This can be the old data, all-one, a random value, a weak all-zero or finally all-zero. The program process programs all bits concurrently from 0 to 1. If it is interrupted not all set bits might read as 1 or contain a weak 1. The register IMB_FSR_OP contains the bits ERASE and PROG. These bits stay set until the next “Clear Status” command or Power-On Reset. So if an erase or program process is interrupted by an Application Reset one of these bits is still set which allows to detect the interruption. It lies in the responsibility of the software to send the “Clear Status” command after a finalized program/erase process to enable this evaluation. Another possible measure against aborted program/erase processes is to prevent resets by configuring the SCU appropriately. If a program or erase process was aborted by a Power-On Reset (e.g. due to a power failure) there do not exist reliable means to detect this by reading the affected flash range. Even with margin reads an early or late aborted process might go unnoticed although it might in the long-term affect reliability. User’s Manual MemoryX2K, V1.2, 2008-09 3-40 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Therefore the application must ensure that flash processes can perform uninterrupted and under the defined operating conditions, e.g. by early brown-out warning that prevents the software from starting flash processes. After a flash process aborted the affected address range must be erased and reprogrammed. 3.9.7.3 Margin Reads Margin reads can be used to verify that flash data is readable with a certain margin. This is typically used as additional check directly after end-of-line programming. As explained above this is not a reliable method for detecting interrupted program or erase processes but the probability of detecting such cases can be increased. Reading with “hard read 0 margin” returns weak 0s as 1s and reading with “hard read 1 margin” returns weak 1s as 0s. Changing the read margin is done with the command sequence “Change Read Margin” and is reported by the status register “IMB_MAR0”. 3.9.7.4 Protection Overview The flash memory supports read and write protection for the whole memory and separate write protection for each logical sector. The logical sector structure is depicted in Figure 3-6 for a 256 KB array. The logical sector structure of a 64 KB array is equivalent, only the logical sector number 7 to 12 do not exist. User’s Manual MemoryX2K, V1.2, 2008-09 3-41 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 256 KB Array Phys . Sector Number Logical Sector Address Logical Sector Number 63 12 = 60 KB 49 11 = 64 KB Logical Grouping 33 10 = 32 KB 25 9 = 16 KB 21 8 = 16 KB 17 16 15 Phys Sect 15 Reserved in Flash 0 7 = 4 KB 6 = 12/16 KB 12 5 = 16 KB 8 4 = 16 KB 4 0 0 0 - 3 = 4 * 4 KB flash_array_logs ec tors_diagram .v s d Figure 3-6 Logical Sectors If read protection is installed and active, any flash read access is disabled in case of start after reset from external memory or from internal RAM. Debug access is as well disabled and thus the execution of injected OCDS instructions. In case of start after reset in internal flash, all flash access operations are controlled by the flash-internal user code and are therefore allowed, as long as not especially disabled by the user, e.g. before enabling the debug interface. Per default, the read protection includes a full (global) flash memory write protection covering all flash modules. This is necessary to eliminate the possibility to program a dump routine into the Flash, which reads the whole Flash and writes it out via the external bus or a serial interface. Program and erase accesses to the flash during active read protection are only possible, if write protection is separately disabled. Flash write User’s Manual MemoryX2K, V1.2, 2008-09 3-42 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization and read protection can be temporarily disabled, if the user authorizes himself with correct passwords. The device also features a sector specific write protection. Software locking of flash memory sectors is provided to protect code and data. This feature disables both program and erase operations for all protected sectors. With write protection it is supported to protect the flash memory or parts of it from unauthorized programming or erase accesses and to provide virus-proof protection for all sectors. Read and write protection is installed by specific security configuration words which are programmed by the user directly into two “Security Pages” (SecP0/1). After any reset, the security configuration is checked by the command state machine (IMB Core) and installations are stored (and indicated) in related registers. If any protection is enabled also the security pages are especially protected. For authorization of short-term disabling of read protection or/and of write protection a password checking feature is provided. Only with correct 64-bit password a temporary unprotected state is taken and the protected command sequences are enabled. If not finished by the command “Re-Enable Read/Write Protection”, the unprotected state is terminated with the next reset. Password checking is based on four 16-bit keywords (together 64 bits) which are programmed by the user directly into the “Security Page 0” (SecP0). Special support is provided to protect also the protection installation itself against any stressing or beaming aggressors. The codes of configuration bits are selected, so that in case of any violation in the flash array, on the read path or in registers the protected state is taken per default. In registers and security pages, protection control bits are coded always with two bits, having both codes, “00B” and “11B” as indication of illegal and therefore protected state. User’s Manual MemoryX2K, V1.2, 2008-09 3-43 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.8 Protection Handling Details As shortly described in “Protection Overview” on Page 3-41 the flash memory can be in different protection states. The protection handling can be separated into different layers that interact which each other (see Figure 3-7). • • • The lowest layer consists of the physical content of the security pages SecP0 and SecP1. This information is used to initialize the protection system during startup. The next layer consists of registers that report the state of the physical layer (IMB_PROCONx) and the protection state (IMB_FSR_PROT). The protection state can be temporarily changed with command sequences which is reflected in the IMB_FSR_PROT. The highest layer is represented by 4 fields of the IMB_IMBCTR register. These fields define the protection rights of the customer software (are read or write accesses currently allowed or not). The IMB Core controls the protection state of all connected flash modules centrally. In this position it can supervise all accesses that are issued by the CPU. User’s Manual MemoryX2K, V1.2, 2008-09 3-44 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Boot Mode Upper Layer IMB_CTRL DDF DCF IMB_CTRH RPA WPA Write to DDF/DCF Middle Layer IMB_PROCONx IMB_FSR IMB_FSR Disable/ ReEnable Protection PROCONs WPRODIS RPRO RPRODIS PROIN PROINER Physical Layer Security Page 0 Passwords Security Page 1 Lock Code RPRO PROCONs Erase/ Program Sec. Page copied influences influences indirectly flash_protection.vsd Figure 3-7 3.9.8.1 Protection Layers The Lower Layer “Physical State” After reset the protection state of the device is restored from the following information: • • The security page 1 contains a “lock code”. This consists of two words of data (32 bits). If it has the value AA55AA55H then security page 0 determines the protection state. Otherwise (i.e. the lock code was not found) the device is in the “non-protected state”. The content of the security page 0 is still copied into the registers as described in “The Middle Layer “Flash State”” on Page 3-46 but their values are ignored in the non-protected state. The security page 0 contains the RPRO double bit, the write protection bits SnU and 4 passwords. If the field RPRO contains a valid 01B or 10B entry the page is valid and the device is in the “protection installed state”. The page content determines the security settings after startup. If SecP0 contains an invalid RPRO entry the device is in the “errored protection” state. User’s Manual MemoryX2K, V1.2, 2008-09 3-45 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization To summarize: the content of the security pages determines if the device is in the “nonprotected state”, “protection installed state” or “errored protection state”. These states are reflected in the register settings of the next layer. The device is usually delivered in the “non-protected state”. The exact layout of the security pages is described in “Layout of the Security Pages” on Page 3-50. 3.9.8.2 The Middle Layer “Flash State” The middle layer consists of the registers IMB_PROCONx and IMB_FSR_PROT and commands that manipulate them and the content of the security pages. During startup the physical state is examined by the IMB Core and it is reflected in the following bit settings of IMB_FSR_PROT: • • • “non-protected state”: PROIN = 0, PROINER = 0. “protection installed state”: PROIN = 1, PROINER = 0. “errored protection state”: PROIN = 0, ROINER = 1. The fourth possible setting PROIN=1 and PROINER=1 is invalid and can not occur. The IMB_PROCONx registers are initialized during startup with the content of the security page 0. The bits DSBER and DDBER indicate if an ECC error occurred. The customer software has thus the possibility to detect disturbed security pages and it can refresh their content. Commands Other bits of the IMB_FSR_PROT: RPRODIS, WPRODIS, PROER can be manipulated with command sequences and define together with the other bits the protection effective for the next layer. All three bits are 0 after system startup. The command “Disable Read Protection” sets RPRODIS to 1 if the correct passwords that are stored in SecP0 are supplied. If incorrect passwords are entered the bit PROER is set and RPRODIS stays unchanged. As protection against “brute force attacks” that search the correct password the password detection is locked. So after supplying the first incorrect password all following passwords even the correct ones are rejected with PROER. This state is only left by an Application Reset or by erasing SecP0. The disabled protection can be enabled again by the Application Reset or by the command “Re-Enable Read/Write Protection” which clears RPRODIS again. The bit PROER can be reset by an Application Reset or by the commands “Reset to Read” and “Clear Status”. The command “Disable Write Protection” sets WPRODIS to 1 if the correct passwords are supplied. It behaves analog to RPRODIS as described above. The command “Re-Enable Read/Write Protection” clears RPRODIS and WPRODIS. User’s Manual MemoryX2K, V1.2, 2008-09 3-46 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization The commands “Enter Page Mode”, “Enter Security Page Mode”, “Erase Page”, “Erase Security Page” and “Erase Sector” set PROER if the write access to the addressed range is not allowed. If a write access is allowed or not is determined by the next level. Table 3-5 summarizes how the “Flash State” of protection determines the RPA and WPA fields of IMB_IMBCTRH. For the double bits a short notation is used here and in the following sections: 1 means active, 0 means inactive, ‘#’ means invalid and ‘–’ means do not care including invalid states. The symbol ‘|’ means logic or. Table 3-5 “Flash State” Determining RPA and WPA IMB_ FSR. PROI N IMB_ FSR. PROI NER IMB_ FSR. RPR O IMB_ FSR. RPR ODIS IMB_ FSR. WPR ODIS Resulting Security Level in RPA and WPA 0 0 – – – Non-protected state: RPA = 0, WPA = 0. 1 0 0 3.9.8.3 Protection installed state (possibly disabled, see below): 0 – 0 RPA = 0, WPA = 1. 0 0 1 RPA = 0, WPA = 0. 1|# 0 0 RPA = 1, WPA = 1. – 1 1 RPA = 0, WPA = 0 (all disabled). 1|# 0 1 RPA = 1, WPA = 0. 1|# 1 0 RPA = 0, WPA = 1. 1 Errored protection state (see below): – 0 0 RPA = 1, WPA = 1. – 0 1 RPA = 1, WPA = 0. – 1 0 RPA = 0, WPA = 1. – 1 1 RPA = 0, WPA = 0. The Upper Layer “Protection State” This layer consists mainly of the 4 fields DCF, DDF, WPA and RPA of the IMB_IMBCTRH register. These determine the effective protection state together with registers of the lower layers. Some of the above mentioned command sequences directly influence these fields as well. In order to increase the resistance against beaming or power supply manipulation all 4 fields are coded with 2 bits. Generally “01” User’s Manual MemoryX2K, V1.2, 2008-09 3-47 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization means active, “10” inactive and the two other states “00” and “11” are invalid and are recognized as “attacked” state. Effective Security Level The effective security level based on these 4 double-bits is summarized in Table 3-6 and Table 3-7. For the double bits the same short notation is used as before: 1 means active, 0 means inactive, ‘#’ means invalid and ‘–’ means do not care including invalid states. Table 3-6 Effective Read Security RPA DCF DDF Security Level 0 – – No read protection. 1|# 0 0 No read protection. – 1|# Data reads prohibited. 1|# – Code fetches prohibited. Table 3-7 Effective Write Security WPA RPA Security Level 0 – No write protection 1|# 1|# Global write protection. 1|# 0 Sector specific write protection depending on IMB_PROCONx. To summarize: • • Read protection is always globally affecting the whole flash memory range. Code fetches and data reads can be separately controlled. Write protection can be global when the read protection is effective or it can be specific for each logical sector. The lower and the middle security layers determine how the 4 effective IMB_IMBCTR fields are preset, changed and how software can access them. This is discussed in the following paragraphs. Initialization of the Effective Security Level After Application Reset protection is activated so that RPA, WPA, DDF and DCF are set. During startup the IMB Core determines the stored security level as described in “The Lower Layer “Physical State”” on Page 3-45 and sets IMB_FSR_PROT.PROIN and IMB_FSR_PROT.PROINER and IMB_PROCONx as described in “The Middle Layer User’s Manual MemoryX2K, V1.2, 2008-09 3-48 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization “Flash State”” on Page 3-46. The IMB Core further initializes the IMB_IMBCTRH fields RPA and WPA according to the rules of Table 3-5. The bits DDF and DCF of the IMB_IMBCTRL are not initialized by the IMB Core. During system startup they are initialized depending on the startup condition. If code fetching starts in the flash memory then they are set to the inactive state. In all other cases they are activated to prevent read access to the flash memory without proving password knowledge. Changing the Effective Security Level During run-time the effective security level can be changed. This can be done by directly writing to the IMB_IMBCTRL register or indirectly by changing the bits of the middle layer by commands as “Disable Write Protection” or even double indirectly by changing the content of the security pages which changes bits in the middle layer and influences the effective security level. Writing directly to IMB_IMBCTRL: • DCF and DDF can be deactivated only if RPA is inactive. They can always be activated. Indirectly by using a command sequence: • • • A successful “Disable Read Protection” sets RPRODIS and clears RPA. A successful “Disable Write Protection” sets WPRODIS and clears WPA. “Re-Enable Read/Write Protection” clears RPRODIS and WPRODIS and sets RPA and WPA according to Table 3-5 depending on PROIN, PROINER and RPRO. Double indirect by changing security pages. After executing a command sequence that changed the content of a security page the IMB Core immediately reads back the pages and determines all resulting security data as described for system startup in “Initialization of the Effective Security Level” on Page 3-48. The examples in “Protection Handling Examples” on Page 3-51 will show how this can be used for installing and removing protection or changing passwords. 3.9.8.4 Reaction on Protection Violation If software tries to violate the protection rules the following happens: • • • Reading data when read protection is effective: The bit IMB_FSR_PROT.PROER is set and the Flash access trap can be triggered via the SCU if IMB_INTCTR.DPROTRP is 0. Default data is delivered. Fetching code when read protection is effective: the trap code “TRAP 15D” is delivered instead. Programming or erasing memory ranges when they are write protected: PROER is set. User’s Manual MemoryX2K, V1.2, 2008-09 3-49 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.8.5 Layout of the Security Pages The previous sections just mentioned the content of the security pages. This section depicts their exact layout. Figure 3-8 depicts symbolically the layout of the security pages 0 and 1. Security Page 0 FF'00FFH Rest unused FF'007FH Security Page 1 unused FF'0010 H FF'0008 H 4 PassWords FF'0000 H RPRO PW3 PW2 PW1 PW0 Rest unused Block 1 PROCON Words Pn ... P2 P1 P0 Block 0 unused FF'0020 H Lock Code (2 Words) FF'0080 H CH CL flas h _s ecurity_page_layout.v sd Figure 3-8 Layout of Security Pages Generally the 16-bit words are stored as always in the XC27x5X in little endian format. • • • • The PWx words contain the passwords. The double bit RPRO is stored as in the related ISFR IMB_FSR_PROT in the bits 15 and 14. The other bits of this word are unused and should be kept all-zero. The PROCON data is stored as defined in the IMB_PROCONx (x=0-3) ISFR. The lock code consists of the two words CL and CH. Both contain “AA55H” to form the correct lock code. User’s Manual MemoryX2K, V1.2, 2008-09 3-50 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization All bytes of the used blocks of the security pages (block 0 and 1 of SecP0 and block 0 of SecP1) are to be considered as “reserved” and must be kept erased, i.e. with all-zero content. The unused blocks of the security pages (blocks 2 to 7 of SecP0 and blocks 1 to 7 of SecP1) shall be programmed with all-one data. 3.9.9 Protection Handling Examples Some examples on how to work with the protection system. Delivery State The device is delivered in the “non-protected state”. Security page 1 is erased (so it does not contain the “lock code” AA55AA55H). Security page 0 is erased and so “invalid” but because SecP1 is erased this data is anyhow not evaluated. Only its content is copied into corresponding the registers. During startup the bits DDF and DCF are set depending on the start mode but as RPA and WPA are inactive all accesses to the flash memory are allowed. The data sectors of the flash memory are delivered in the erased state as well. All sectors can be programmed. After uploading the software the customer can install write and read protection. First Time Password Installation In order to install a password generally the lock code in SecP1 has to be erased. In this case the code is not present. After that SecP0 must be erased with “Erase Security Page” in order to be able to change RPRO. Erasing SecP0 clears RPRO to “00B” which is an invalid state. After finishing the erase command the IMB Core restores the IMB_FSR_PROT and IMB_IMBCTRH fields from the flash data. Because no lock code is present in SecP1 the invalid state of RPRO has no effect on the user visible protection. Still all parts of the flash memory can be written. The second step is to program the information of SecP0 with the required security information. Again the IMB Core reads immediately back the stored data and initializes the security system. As SecP1 still does not contain the lock code the device stays in the “non-protected” mode. The security pages cannot be read directly by customer software. The data programmed into SecP0 can therefore only be verified indirectly. The data of the RPRO and SnU fields can be checked by reading the IMB_PROCON and IMB_FSR_PROT registers. The passwords can be verified with the command “Disable Read Protection”. If the password does not match the bit PROER is set. But because of the erased SecP1 the flash memory stays writable. So after erasing SecP0 the correct password can be programmed again. User’s Manual MemoryX2K, V1.2, 2008-09 3-51 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization After the SecP0 was verified successfully SecP1 gets programmed with the lock code AA55AA55H which enables the security settings of SecP0. Because the password validation left RPRODIS set the command “Re-Enable Read/ Write Protection” must be used to finally activate the new protection. Changing Passwords or Security Settings Changing the passwords is a delicate operation. The interrelation of the two security pages must be kept in mind. Usually in the protected state the SecP1 contains the lock code. First write protection must be disabled with the correct passwords. Then the lock code in SecP1 is erased. If this operation was successful PROIN will be cleared by the IMB Core. Now SecP0 can be safely erased. From this point on the security pages are in the factory delivery state and the new passwords and security settings can be installed as described above. Attention: The number of times a security page may be changed is noted in the data sheet. User’s Manual MemoryX2K, V1.2, 2008-09 3-52 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.10 EEPROM Emulation The flash memory of the XC27x5X is used for three purposes: 1. Storage of program code. Updates happen usually very seldom. The main criteria to be fulfilled is a retention of the life-time of the product. 2. Storage of constant data: this data is stored together with program code. So this data is very seldom updated. Endurance is of no issue here but retention identical to the code memory is required. 3. Data updated during run-time: this might be data with a very high frequency of updates like a mileage counter or access keys for key-less entry. Other data might be changed only in case of failures and other data might only be transferred from RAM to non-volatile memory before the system is powered down. Especially for the third type of data the non-volatile memory needs EEPROM like characteristics: • • • Fine program/erase granularity which is in EEPROMs typically 1 byte. Higher endurance than the intrinsic endurance of flash cells. Short program and erase duration per byte. Especially for storing data in an emergency (e.g. power failure) short latencies might be required. A basic requirement for changing data during run-time is that code execution can still resume, especially interrupt requests must still be serviced. This requirement is fulfilled in the XC27x5X because all four flash modules work independently. If one is busy with program or erase then code can still be executed from the other. The other requirements are more difficult to fulfill because the XC27x5X does not have an EEPROM available but only the flash memory with the already frequently mentioned limitations: big program/erase granularity, moderately long program/erase duration, limited cell endurance with reduced retention at high number of program/erase cycles, pages not isolated but affected by drain disturbs. In order to alleviate these effects on run-time storage of data software is used to emulate EEPROM. There is quite a number of algorithms for efficiently using flash memory as EEPROM. The following section describes one (the most simple) of these algorithms. It should be noted that the XC27x5X does not offer the customer any hardware means for EEPROM emulation. All of the following must be realized by software. 3.9.10.1 The Traditional EEPROM Emulation The key point is to solve the limited endurance by storing data in N different physical places. In XC27x5X the algorithm could use N sequential pages or groups of pages. If data is currently stored in the page group “x” then the next program happens to the page group “(x+1) mod N”. After boot up the last correct page group must be found. This could be done by either evaluating a counter (from 0 to 2*N-1) or the old entries are invalidated by erasing the User’s Manual MemoryX2K, V1.2, 2008-09 3-53 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization page after programming the new one. Additionally a CRC check could be performed over the group. As all involved pages are re-used cyclically the endurance from customer perspective is increased by the factor N. N must be chosen high enough to fulfill endurance and retention requirements. Disturbs in the group of N pages are no issue because they incur at most N-1 disturbs before they get written with new data. Care must be taken however if one sector accommodates different groups of pages with different update behavior. In this case the updates of one group of pages could exceed the disturb limits of the other group. So generally one sector should be used only by one such EEPROM cyclic buffer. The algorithm keeps the old data until the new data is verified so power failure during programming can only destroy the last update but the older data is still available. There are still some issues with power failure that need special treatment: • • Power is cut during programming: the following boot-up might find an apparently correctly programmed page. However the cells might be not fully programmed and thus have a much lower retention or the read data is unstable (e.g. changing operating conditions cause read errors). If the power is cut early the page can appear as erased although some cells are partly programmed. When programming different data to this apparently erased page read errors might occur. Power is cut during erase: the same as above can happen. Data may appear as erased but the retention is lowered. A power failure during a page-erase can inhibit readability of all data in its physical sector. Therefore an algorithm is advantageous that performs erases only in sectors that don’t contain anymore current data. The algorithm can be improved to be more robust against such cases, e.g. program always two pages, mark the end of an erase process by programming a page. But generally aborting flash processes is a forbidden “operating condition”. The main deficiency of the described algorithm is that the software designer is required to plan the use of the flash memory thoroughly. The user has to choose the correct value of N. Then all data has to be allocated to pages. Data sharing one page should have a similar or better identical update pattern (otherwise unchanged data is unnecessarily written). If one set of data does not fill a complete sector the available pages must be possibly left unused because they might incur too many drain disturbs. There are other algorithms that try to alleviate these efforts by monitoring the flash usage and adapt automatically the assignment of data to flash cells. User’s Manual MemoryX2K, V1.2, 2008-09 3-54 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.9.11 Interrupt Generation Long lasting processes (these are mainly: program page, erase page, erase sector and margin changes, but also enter page mode) set the IMB_FSR_BUSY.BUSY flag of one flash module when accepting the request and reset this flag after finishing the process. Software is required to poll the busy flag in order to determine the end of the operation. In order to release the software from this burden an interrupt can be generated. If the interrupt is enabled by IMB_INTCTRL.IEN then all transitions from 1 to 0 of one of the IMB_FSR_BUSY.BUSY flags send an interrupt request to the SCU. In the SCU (see “SCU Interrupt Generation” on Page 8-147 in the SCU chapter) the interrupt request (noted as “PFI” Program Flash Interrupt) is multiplexed with other interrupt sources and is forwarded to one of four interrupt nodes. The selection of the interrupt node is done with the register field INTNP1.PF. The SCU contains its own set of interrupt status flags (INTSTAT), interrupt disable control (INTDIS) and registers for setting this interrupt (INTSET) and clearing it (INTCLR). The “Enter Page Mode” command sets BUSY only for a few clock cycles. It is usually not advisable to enable the interrupt for this command. The register IMB_INTCTR contains fields for the interrupt status “ISR”, an enable for the interrupt request “IEN” and fields for clearing the status flag “ICLR” or setting if “ISET”. It should be noted that the interrupt request is only sent when ISR becomes 1 and IEN was already 1. No interrupt is sent when IEN becomes 1 when ISR was already 1 or both are set to 1 at the same time. 3.9.12 Recommendations for Optimized Flash Usage This section describes best practices for using the flash in certain application scenarios, e.g. how to use effectively ECC and margin reads. For a description of the hardware features consult “Data Integrity” on Page 3-40. 3.9.12.1 Programming Code and Constant Data Code and constant data are programmed only few times during life-time of a device, e.g. end-of-line in ECU production or when service updates are performed. As the readability of this data is decisive for the product quality customers might want to implement the elaborate “best practice” advice. Basic Advice Always ensure correct operating conditions and prevent power failures during flash operation. As basic protection against handling errors all data should be verified after programming. Single-bit ECC errors should be ignored. The appearance of small numbers of single-bit errors is a consequence of known physical effects. User’s Manual MemoryX2K, V1.2, 2008-09 3-55 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Best Practice This approach offers best possible quality but risks that programming steps need to repeated even unnecessarily (“false negatives”): • • • • • • Use “Erase Sector” to erase complete sectors. Program the sector with data. A common protection against software crashes is to fill the unused part of the sector with trap codes. Change the read level to hard margin 0. Verify the programmed data, note comparison errors and double-bit ECC errors and count single-bit ECC errors. Take care to evaluate the ECC error flags only once per 128-bit data block and clear them afterwards. Repeat this check with hard margin 1. After programming all sectors: – Erase and re-program all sectors with comparison or double-bit ECC errors. – If a flash module contained more than a certain number (e.g. 10) of single-bit ECC errors it is recommended to erase and re-program the affected sectors (i.e. those containing at least one single-bit error). – Attention: a high number of single-bit errors indicates usually a violation of operating conditions. The threshold of allowed single-bit errors could be increased for in-service updates in order to reduce the risk of false negatives. 3.9.12.2 EEPROM Emulation For EEPROM emulation the goal is usually not readability over device life-time but highest possible robustness (against violated operating conditions, power failures, even failing flash pages e.g. due to over-cycling). The risk of false negatives should be minimized. A good robustness is achieved with the following approach: • • • Verify data after programming with the normal read level. Single-bit ECC errors should be ignored. In case of comparison error or double-bit ECC error the data should be programmed again to the next flash range (e.g. next page or sector). The number of re-programming trials should be limited (e.g. to 3) to protect against violated operating conditions. Obviously this jumping over failed pages can be only used optimally when the algorithm does not expect data on fixed addresses. Failing pages can prevent “Erase Sector” from erasing any data in the affected sector. The “Erase Page” command however could still erase all other pages. These other pages stay readable and programmable. User’s Manual MemoryX2K, V1.2, 2008-09 3-56 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10 On-Chip Program Memory Control The internal memory block “IMB” contains all memories of the so called “on-chip program memory area” in the address range from C0’0000H to FF’FFFFH. Included are the program SRAM, the embedded flash memories and central control logic called “IMB Core”. In the XC27x5X device the IMB contains the following memories: • • up to 828 KB flash memory in four independent modules. up to 32 KB program SRAM (see Section 3.4.1). The IMB connects these memories to the CPU data bus and the instruction fetch bus. Each memory can contain instruction code, data or a mixture of both. The IMB manages accesses to the memories and supports flash programming and erase. 3.10.1 Overview The Figure 3-9 shows how the IMB and its memories are integrated into the device architecture. Only the main data streams are included. The data buses are usually accompanied by address and control signals and check-sum data like parity or ECC. IMB C166SV2 DMU (Data access ) CPU PMU (Instr fetch ) Data IMB Core 16 PSRAM 64 (Program SRAM) Instructions Flash Memory 64 Flash Module 0 128 Flash Module n 128 imb _bloc k_diagram .vs d Figure 3-9 IMB Block Diagram The CPU has two independent busses. The instruction fetch bus is controlled by the program management unit “PMU” of the CPU. It fetches instructions in aligned groups of 64 bits. The instruction fetch unit of the CPU predicts the outcome of jumps and fetches instructions on the predicted branch in advance. In case of a misprediction this interface User’s Manual MemoryX2K, V1.2, 2008-09 3-57 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization can abort outstanding requests and continues fetching on the correct branch. As the CPU can consume up to one 32-bit instruction per clock cycle the performance of this interface determines the CPU performance. The data bus is controlled by the data management unit “DMU” of the CPU. It reads data in words of 16 bits. Write accesses address as well 16-bit words but additional byte enables allow changing single bytes. Because of the CPU’s “von Neumann” architecture data and instructions (and “special function registers” to complete the list) share a common address range. When instructions are used as data (e.g. when copying code from an IO interface to the PSRAM) they are accessed via the data bus. The pipelined behavior of the CPU can cause that code fetches and data accesses are requested simultaneously. The IMB takes care that accesses can perform concurrently if they address different memories or flash modules. Additional connections of the IMB to central system control units exist. These are not shown in the block diagram. User’s Manual MemoryX2K, V1.2, 2008-09 3-58 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10.2 Register Interface The “IMB Registers” on Page 3-59 describes the special function registers of the IMB. 3.10.2.1 IMB Registers The section describes all IMB special function registers. Table 3-8 Registers Overview Register Short Name Register Long Name Offset Address Page Number IMB_IMBCTRL IMB Control Low FF FF00H Page 3-60 IMB_IMBCTRH IMB Control High FF FF02H Page 3-61 IMB_INTCTR Interrupt Control FF FF04H Page 3-63 IMB_FSR_BUSY Flash State Busy FF FF06H Page 3-64 IMB_FSR_OP Flash State Operations FF FF08H Page 3-65 IMB_FSR_PROT Flash State Protection FF FF0AH Page 3-67 IMB_MAR0 Margin 0 FF FF0CH Page 3-68 IMB_PROCON0 Protection Configuration 0 FF FF10H Page 3-69 IMB_PROCON1 Protection Configuration 1 FF FF12H Page 3-69 IMB_PROCON2 Protection Configuration 2 FF FF14H Page 3-69 IMB_PROCON3 Protection Configuration 3 FF FF16H Page 3-69 IMB_ECC_TRAP ECC Trap Control FF FF20H Page 3-70 IMB_ECC_STAT ECC Status FF FF22H Page 3-72 IMB Control Global IMB control. Both IMB_IMBCTRL and IMB_IMBCTRH are reset by an Application Reset. The write access to both registers is controlled by the register security mechanism as defined in the SCU chapter “Register Control” on Page 8-211. Please note that the register write-protection is not activated automatically again after an access to IMB_IMBCTR because this happens only for SCU internal registers. User’s Manual MemoryX2K, V1.2, 2008-09 3-59 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization IMB_IMBCTRL IMB Control Low 15 14 13 ISFR (FF FF00H) 12 Reset value: 55ACH 11 10 9 8 7 6 5 4 3 2 1 0 DDF DCF - - - - - - - - DLC PF WSFLASH rw rw - - - - - - - - rw rw Field Bits Typ Description WSFLASH [2:0] rw Wait States for Flash Access Number of wait cycles after which the IMB expects read data from the flash memory is: NWS = WSFLASH. This field determines as well the read timing of the PSRAM in the flash emulation address range. See “Flash Emulation” on Page 3-13. The correct setting of this field depends on the system clock frequency. The data sheet of the device describes this relation. Note: WSFLASH must not be 0. This value is forbidden! DLCPF 3 User’s Manual MemoryX2K, V1.2, 2008-09 rw Disable Linear Code Pre-Fetch “High Speed Mode”: When the next read 0B request will be delivered from the buffer and so the flash memory would be idle, the IMB Core autonomously increments the last address and reads the next 128-bit block from the flash memory. 1B “Low Power Mode”: This feature is disabled. Usually for code with power minimization requirements or for code with short linear code sections this feature should be disabled (DLCPF = 1). Enabling this feature is only advantageous for code section with longer linear sequences. With lower values of WSFLASH the performance gain of DLCPF=0 is reduced. In case of low WSFLASH settings DLCPF=1 might even lead to better performance than with linear code pre-fetch. 3-60 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description DCF [13:12] rw Disable Code Fetch from Flash Memory 01B Short notation DCF = 1. If RPA = 1 instructions cannot be fetched from flash memory. If RPA = 0 this field has no effect. 10B Short notation DCF = 0. Instructions can be fetched independent of RPA. 00B Illegal state. 11B Illegal state. Both illegal states have the same effect as “01”. This state can only be left by an Application Reset. When RPA = 0 software can change this field to any value. Otherwise code fetch can only be disabled but not enabled anymore until the next Application Reset. DDF [15:14] rw Disable Data Read from Flash Memory 01B Short notation DDF = 1. If RPA = 1 data cannot be read from flash memory. If RPA = 0 this field has no effect. 10B Short notation DDF = 0. Data can be read independent of RPA. 00B Illegal state. 11B Illegal state. Both illegal states have the same effect as “01”. This state can only be left by an Application Reset. When RPA = 0 software can change this field to any value. Otherwise data reads can only be disabled but not enabled anymore until the next Application Reset. IMB control high word. The WPA and RPA fields are described in “Protection Handling Details” on Page 3-44. IMB_IMBCTRH IMB Control High 15 14 13 ISFR (FF FF02H) 12 11 7 6 5 4 PSPROT – – – – RPA WPA rw – – – – rh rh User’s Manual MemoryX2K, V1.2, 2008-09 10 9 8 Reset value: 0005H 3-61 3 2 1 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description WPA [1:0] rh Write Protection Activated 01B Short notation WPA = 1. The write protection of the flash memory is activated. 10B Short notation WPA = 0. The write protection is not activated. 00B Illegal state. 11B Illegal state. Both illegal states have the same effect as “01”. This state can only be left by an Application Reset. This field is only changed by the IMB Core. Software writes are ignored. RPA [3:2] rh Read Protection Activated 01B Short notation RPA = 1. The read protection of the flash memory is activated. 10B Short notation RPA = 0. The read protection is not activated. 00B Illegal state. 11B Illegal state. Both illegal states have the same effect as “01”. This state can only be left by an Application Reset. This field is only changed by the IMB Core. Software writes are ignored. PSPROT [15:8] rw PSRAM Write Protection This 8-bit field determines the address up to which the PSRAM is write protected. The start address of the writable range is E0’0000H + 1000H*PSPROT. The end address is determined by the implemented memory. The equivalent range in the PSRAM area with flash access timing is protected as well. Here the writable range starts at E8’0000H + 1000H*PSPROT. So with PSPROT=00H the complete PSRAM is writable. Interrupt Control Interrupt control and status. Reset by Application Reset. User’s Manual MemoryX2K, V1.2, 2008-09 3-62 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization IMB_INTCTR Interrupt Control ISFR (FF FF04H) 15 14 13 12 11 ISR PSE R – – – rh rh – – – 10 9 8 PSE RCL ISET ICLR R w w w Reset value: 0000H 7 6 5 4 – – – – – – – – 3 2 1 0 DPR DDD DIDT OTR IEN TRP RP P rw rw rw rw Field Bits Typ Description IEN 0 rw Interrupt Enable If set, the interrupt signal of the IMB gets activated when ISR is set. DIDTRP 1 rw Disable Instruction Fetch Double Bit Error Trap If set, a double bit ECC error does not cause the replacement of the fetched data by a trap instruction. See also IMB_ECC_TRAP.DITRPx. DDDTRP 2 rw Disable Data Read Double Bit Error Trap If set, a double bit ECC error during data read does not send a “Flash Access Error” request to the SCU, i.e. no HW trap is generated and the read data is not replaced with default data. The error flags are still set in IMB_FSR_PROT and IMB_ECC_STAT. See also IMB_ECC_TRAP.DDTRPx. DPROTRP 3 rw Disable Protection Trap If set, a read request from read protected flash memory does not generate a “Flash Access Error” request to the SCU, i.e. no HW trap is generated. ICLR 8 w Interrupt Clear When written with 1 the ISR is cleared. Reading this bit delivers always 0. Writing a 0 is ignored. ISET 9 w Interrupt Set When written with 1 the ISR is set and if IEN is set the interrupt signal is activated. Reading this bit delivers always 0. Writing a 0 is ignored. PSERCLR 10 w Clear PSRAM Error Flag When written with 1 the PSER is cleared. Reading this bit delivers always 0. Writing a 0 is ignored. User’s Manual MemoryX2K, V1.2, 2008-09 3-63 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description PSER 14 rh PSRAM Error Flag This flag is set when write requests to the write protected or not implemented PSRAM range are detected. This flag can be cleared by writing 1 to PSERCLR. ISR 15 rh Interrupt Service Request If set, it indicates that at least one IMB_FSR_BUSY.BUSY bit changed from 1 to 0. If IEN was set an interrupt request is sent to the interrupt controller. After servicing the interrupt the software handler clears this flag by writing a 1 to ICLR. Flash State Flash state. Split into 3 registers IMB_FSR_BUSY, IMB_FSR_OP, and IMB_FSR_PROT. The protection relevant fields of IMB_FSR_PROT are described in “Protection Handling Details” on Page 3-44. The registers are reset by the Application Reset with the exception of “ERASE”, “PROG”, and “OPER”. These three fields are only reset by a Power-On Reset. IMB_FSR_BUSY Flash State Busy ISFR (FF FF06H) 15 14 13 12 – – – – – – – – 11 10 9 8 PAG PAG PAG PAG E3 E2 E1 E0 rh rh rh rh 7 6 5 4 – – – – – – – – Field Bits Typ Description BUSY0 0 rh User’s Manual MemoryX2K, V1.2, 2008-09 Reset value: 0000H 3 2 1 0 BUS BUS BUS BUS Y3 Y2 Y1 Y0 rh rh rh rh Busy Flash 0 Flash module 0 is busy with a task. The task is indicated by the bits MAR, POWER, ERASE or PROG of IMB_FSR_OP. BUSY0 is automatically cleared when the task has finished. The corresponding task indication is not cleared in order to allow an interrupt handler to determine the finished task. 3-64 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description BUSY1 1 rh Busy Flash 1 Same as BUSY0 for flash module 1. BUSY2 2 rh Busy Flash 2 Same as BUSY0 for flash module 2. BUSY3 3 rh Busy Flash 3 Same as BUSY0 for flash module 3. PAGE0 8 rh Page Mode Indication Flash 0 Set as long the flash module 0 is in page mode. Page mode is entered by the “Enter Page Mode” commands and finished by a “Program Page” command. The page mode can be also left by a “Reset to Read” command. Also an Application Reset clears this bit. PAGE1 9 rh Page Mode Indication Flash 1 Same as PAGE0 for flash module 1. PAGE2 10 rh Page Mode Indication Flash 2 Same as PAGE0 for flash module 2. PAGE3 11 rh Page Mode Indication Flash 3 Same as PAGE0 for flash module 3. IMB_FSR_OP Flash State Operations ISFR (FF FF08H) 15 14 13 12 11 10 9 8 7 6 – – – – – – – – – – – – – – – – – – – – Field Bits Typ Description PROG 0 rh User’s Manual MemoryX2K, V1.2, 2008-09 Reset value: 0000H 5 4 3 2 1 0 OPE SQE POW ERA PRO MAR R R ER SE G rh rh rh rh rh rh Program Task Indication This bit is set when a program task is started. The affected flash module is indicated by a BUSY bit. The PROG bit is not automatically reset but must be cleared by a “Clear Status” command. This bit is not cleared by an Application Reset but only by a PowerOn Reset. 3-65 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description ERASE 1 rh Erase Task Indication This bit is set when an erase task is started. The affected flash module is indicated by a BUSY bit. The ERASE bit is not automatically reset but must be cleared by a “Clear Status” command. This bit is not cleared by an Application Reset but only by a PowerOn Reset. POWER 2 rh Power Change Indication This bit indicates that a flash module is in its startup phase or in a shutdown phase. The BUSY bits indicate which flash module is busy. This bit is not automatically reset but must be cleared by a “Clear Status” command. MAR 3 rh Margin Change Indication If a read margin modification is requested this bit is set together with the corresponding BUSY bit. The BUSY bit is cleared when the margin change is effective and the flash module can be read again. The MAR bit must be cleared by a “Clear Status” command. SQER 4 rh Sequence Error This bit is set by a errored command sequence or a command that is not accepted. It is cleared by “Clear Status” and “Reset to Read”. OPER 5 rh Operation Error The IMB Core maintains internal bits that are set when starting a program or erase process. They are cleared when this process finishes. These bits are not reset by an Application Reset but only by a Power-On Reset. If one of these bits is set after Application Reset the IMB Core sets OPER. So this signals that a running erase or program process was interrupted by an Application Reset. The OPER is cleared by “Reset to Read”, “Clear Status” or a Power-On Reset. User’s Manual MemoryX2K, V1.2, 2008-09 3-66 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization IMB_FSR_PROT Flash State Protection 15 14 13 12 RPRO – – rh – – 11 ISFR (FF FF0AH) 10 9 8 DDB DSB IDBE ISBE ER ER R R rh rh rh rh Reset value: x000H 7 6 5 – – – – – – 4 3 2 1 0 PRO WPR RPR PROI PROI ER ODIS ODIS NER N rh rh rh rh rh Field Bits Typ Description PROIN 0 rh Flash Protection Installed Modified by the IMB Core. Cleared by Application Reset. PROINER 1 rh Flash Protection Installation Error Modified by the IMB Core. Cleared by Application Reset. RPRODIS 2 rh Read Protection Disabled The read protection was temporarily disabled with the “Disable Read Protection” command. Modified by the IMB Core. Cleared by Application Reset. WPRODIS 3 rh Write Protection Disabled The write protection was temporarily disabled with the “Disable Write Protection” command. Modified by the IMB Core. Cleared by Application Reset. PROER 4 rh Protection Error Set by a violation of the installed protection. Reset by the “Clear Status” and “Reset to Read” commands or an Application Reset. ISBER 8 rh Instruction Fetch Single Bit Error Set if during instruction fetch a single-bit ECC error was detected (and corrected). Reset by “Clear Status” or “Reset to Read” commands or an Application Reset. IDBER 9 rh Instruction Fetch Double Bit Error Set if during instruction fetch a double-bit ECC error was detected (and not corrected). Reset by “Clear Status” or “Reset to Read” commands or an Application Reset. User’s Manual MemoryX2K, V1.2, 2008-09 3-67 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description DSBER 10 rh Data Read Single Bit Error Same as ISBER for data reads. DDBER 11 rh Data Read Double Bit Error Same as IDBER for data reads. RPRO [15:14] rh Read Protection Configuration This field is copied by the IMB Core from the corresponding field in the security page 0. After Application Reset read protection is activated. See Table 3-5 and ff for interpreting this and other protection bit fields. 00B Invalid. 01B Active. 10B Inactive. 11B Invalid. Margin Control Read margin control. Each field corresponds to one flash module. A hard read 0 detects not completely erased cells. These are read as “1”. A hard read 1 detects not completely programmed cells. These are read as “0”. Read margin changes are caused by the command sequence “Change Read Margin”. The resulting read margin is reflected in this status register. The command sequences “Program Page”, “Erase Sector”, “Erase Page” and “Erase Security Page” resets the read margin back to “normal”. The same happens in case of a flash wake-up. Reset by Application Reset. IMB_MAR0 Margin Control 0 15 14 13 12 ISFR (FF FF0CH) 11 10 9 8 7 6 Reset value: 0000H 5 4 3 2 1 – – HREAD3 HREAD2 HREAD1 HREAD0 – – rh rh rh rh User’s Manual MemoryX2K, V1.2, 2008-09 3-68 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description HREAD0 [2:0] rh Hard Read 0 Active read margin of flash module 0. “000”:Normal read. “001”:Hard read 0. “010”: Alternate hard read 0 (usually harder than 001). “101”:Hard read 1. “110”: Alternate hard read 1 (usually harder than 101). other codes:Reserved. HREAD1 [5:3] rh Hard Read 1 Same as HREAD0 for flash module 1. HREAD2 [8:6] rh Hard Read 2 Same as HREAD0 for flash module 2. HREAD3 [11:9] rh Hard Read 3 Same as HREAD0 for flash module 3. Protection Configuration Protection configuration register of each implemented flash module. The logical sector numbering is depicted in Figure 3-6. Each bit of the PROCONs is related to a logical sector. If it is cleared the write access to the corresponding logical sector (this means to the range of physical sectors) is locked under the conditions that are documented in “Protection Handling Details” on Page 3-44. The PROCON registers are exclusively modified by the IMB Core which copies them from the security page 0. For flash modules smaller than 256 KB the SsU bits corresponding to the not implemented flash range are reserved and shall be programmed to 0 in the security page. Reset by Application Reset. IMB_PROCONx (x=0-3) Protection Configuration. 15 14 13 – – – – – – 12 11 ISFR (FF FF10H+2*x) 10 9 8 7 6 Reset value: 0000H 5 4 3 2 1 0 S12 S11 S10 S9U S8U S7U S6U S5U S4U S3U S2U S1U S0U U U U rh rh User’s Manual MemoryX2K, V1.2, 2008-09 rh rh rh rh 3-69 rh rh rh rh rh rh rh V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description SsU (s=0-12) s rh Sector s Unlock 0B: Logical sector s of flash module x is writeprotected. 1B: Logical sector s of flash module x is not writeprotected. Note: In previous device families and the TriCore™ based products these are “lock” bits and not “unlock” bits! ECC Trap Control ECC trap control register. Reset by Application Reset. The register IMB_ECC_TRAP allows to disable the double bit ECC error trap generation for selected flash modules in contrast to IMB_INTCTR which allows to switch this only globally. This selective control enables to operate part of the flash memory as quasi ROM with enabled error traps. But while a flash module is programmed or erased its trap generation can be switched off without affecting the “ROM” modules. Without this facility the traps would have to be globally disabled and the flash driver had to work from SRAM and all interrupts would have to be blocked. IMB_ECC_TRAP ECC Trap Control 15 14 13 12 0 0 0 0 r r r r ISFR (FF FF20H) 11 10 9 8 DDT DDT DDT DDT RP3 RP2 RP1 RP0 rw User’s Manual MemoryX2K, V1.2, 2008-09 rw rw rw 3-70 Reset value: 0000H 7 6 5 4 0 0 0 0 r r r r 3 2 1 0 DITR DITR DITR DITR P3 P2 P1 P0 rw rw rw rw V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description DITRP0 0 rw Disable Instruction Fetch Double Bit ECC Trap 0 0B Replacing instructions by a trap code for double bit ECC errors when fetching from flash module 0 is handled globally via IMB_INTCTR.DIDTRP. 1B If set, a double bit ECC error does not cause the replacement of the fetched data by a trap instruction for fetches from flash module 0 independent of IMB_INTCTR.DIDTRP. Additionally IMB_FSR_PROT.ISBER/IDBER are not set for ECC errors from flash module 0. DITRP1 1 rw Disable Instruction Fetch Double Bit ECC Trap 1 Same as DITRP0 for flash module 1. DITRP2 2 rw Disable Instruction Fetch Double Bit ECC Trap 2 Same as DITRP0 for flash module 2. DITRP3 3 rw Disable Instruction Fetch Double Bit ECC Trap 3 Same as DITRP0 for flash module 3. DDTRP0 8 rw Disable Data Read Double Bit ECC Trap 0 Double bit ECC error trap for data reads from 0B flash module 0 are handled globally via IMB_INTCTR.DDDTRP. 1B Double bit ECC error for data reads from flash module 0 does not trigger the “Flash Access Error” trap independent of IMB_INTCTR.DDDTRP. Additionally IMB_FSR_PROT.DSBER/DDBER are not set for ECC errors from flash module 0 and the data from flash memory is delivered not default data. But the bits IMB_ECC_STAT.xBERx are still set for ECC errors. DDTRP1 9 rw Disable Data Read Double Bit ECC Trap 1 Same as DDTRP0 for flash module 1. DDTRP2 10 rw Disable Data Read Double Bit ECC Trap 2 Same as DDTRP0 for flash module 2. DDTRP3 11 rw Disable Data Read Double Bit ECC Trap 3 Same as DDTRP0 for flash module 3. User’s Manual MemoryX2K, V1.2, 2008-09 3-71 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization ECC Status ECC status register. Reset by Application Reset. This register reports ECC data read single and double bit errors selectively per flash module. Each bit can be cleared independently. This enables to use part of the flash memory quasi as “ROM”. In this part all errors trigger traps that are handled by a trap handler and trigger typically a reset of the application. However while flash modules are programmed or erased all ECC errors can be handled by a low-level driver without necessarily affecting the complete system. IMB_ECC_STAT ECC Status ISFR (FF FF22H) 15 14 13 12 0 0 0 0 r r r r 11 10 7 6 5 4 DBE DBE DBE DBE R3 R2 R1 R0 0 0 0 0 SBE SBE SBE SBE R3 R2 R1 R0 rwh r r r r rwh rwh 9 rwh 8 Reset value: 0000H rwh 3 2 1 rwh rwh 0 rwh Field Bits Typ Description SBER0 0 rwh Data Read Single Bit Error 0 Set when a single bit ECC errors occurs when reading data from flash module 0. Cleared by Application Reset or by writing 1 to this bit. SBER1 1 rwh Data Read Single Bit Error 1 Same as SBER0 for flash module 1. SBER2 2 rwh Data Read Single Bit Error 2 Same as SBER0 for flash module 2. SBER3 3 rwh Data Read Single Bit Error 3 Same as SBER0 for flash module 3. DBER0 8 rwh Data Read Double Bit Error 0 Set when a double bit ECC errors occurs when reading data from flash module 0. Cleared by Application Reset or by writing 1 to this bit. DBER1 9 rwh Data Read Double Bit Error 1 Same as DBER0 for flash module 1. User’s Manual MemoryX2K, V1.2, 2008-09 3-72 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Field Bits Typ Description DBER2 10 rwh Data Read Double Bit Error 2 Same as DBER0 for flash module 2. DBER3 11 rwh Data Read Double Bit Error 3 Same as DBER0 for flash module 3. User’s Manual MemoryX2K, V1.2, 2008-09 3-73 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.10.3 Error Reporting Summary The Table 3-9 summarizes the types of detected errors and the possible reactions. Table 3-9 IMB Error Reporting Error Reaction Data read from PSRAM with parity error. If PECON.PEENPS: HW trap (see Section 8.13). Instruction fetch from PSRAM with parity error. If PECON.PEENPS: HW trap (see Section 8.13). Data read from flash memory with single bit Silently corrected. Bit error. IMB_FSR_PROT.DSBER set. Data read from flash memory with double bit error. Bit IMB_FSR_PROT.DDBER set. If IMB_INTCTR.DDDTRP = 0: Flash access trap1) and default data is delivered. Instruction fetch from flash memory with single bit error. Silently corrected. Bit IMB_FSR_PROT.ISBER set. Instruction fetch from flash memory with double bit error. Bit IMB_FSR_PROT.IDBER set. If IMB_INTCTR.DIDTRP = 0: “TRAP 15D” delivered instead of corrupted data. Data read from protected flash memory. IMB_FSR_PROT.PROER set. If IMB_INTCTR.DPROTRP = 0: Flash access trap1) and default data is delivered. Instruction fetch from protected flash memory. “TRAP 15D” delivered. Program/erase request of write protected flash range. Only bit PROER in IMB_FSR_PROT set. Data read or instruction fetch from busy flash memory. Read access stalled until end of busy state. Instruction fetch from ISFR addresses. Default data (“TRAP 15D”) delivered. Data read from not implemented ISFRs. Default data delivered. Data writes to not implemented ISFRs. Silently ignored. Data read from not implemented address range. Unpredictable. Mirrored data from other memories might be returned or default values. User’s Manual MemoryX2K, V1.2, 2008-09 3-74 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization Table 3-9 IMB Error Reporting (cont’d) Error Reaction Instruction fetch from not implemented address range. Unpredictable. Mirrored data from other memories might be returned or default values. Data written to not implemented PSRAM or Bit IMB_INTCTR.PSER set. write protected PSRAM address range Flash access trap1) and no data is changed in the PSRAM. (both determined by IMB_IMBCTRH.PSPROT). Program or erase command targeting not implemented flash memory. Unpredictable. Access is ignored2) or mirrored into implemented flash memory3). Data read from powered-down flash modules. Considered as access to not-implemented memory range. Default data or data from implemented flash modules will be returned. Instruction fetch from powered-down flash modules. Considered as access to not-implemented memory range. Default data (“TRAP 15D”) will be returned or data from implemented flash modules. Program or erase command targeting powered-down flash modules. Silently ignored2). Shutdown or power-down request received The command interpreter is reset and a while the command sequence interpreter is “Reset to Read” command sequence is executed. waiting for the last words of a command sequence. 1) More information about the Flash Access Trap can be found in chapter “SCU”. 2) Attention: when an access (i.e. MOV) is ignored, the command sequence interpreter will still wait for this outstanding MOV. So the next command sequence might cause a SQER because it delivers an unexpected MOV. 3) The flash protection can not be by-passed by accessing the reserved memory ranges. User’s Manual MemoryX2K, V1.2, 2008-09 3-75 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Organization 3.11 Data Retention Memories This section describes the usage of the special purpose data memories Standby RAM (SBRAM) and Marker Memory (MKMEM). Depending on the device not all of them are available. The XC27x5X contains: • MKMEM. User’s Manual MemoryX2K, V1.2, 2008-09 3-76 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4 Memory Checker Module (MCHK) The memory checker module (MCHK) of the XC27x5X supports checking the data consistency of memories, registers (e.g. configuration registers), or communication channels. It calculates a checksum on a block of data, often called cyclic redundancy code (CRC). It is implemented as a parallel signature generation based on a multi input linear feedback shift register (MISR). Being based on a linear feedback shift register (LFSR), it also can generate pseudo-random numbers and cyclic codes. This chapter is structured as follows: • • • • • An operational overview of the Memory Checker Module (see Section 4.1) Functional description of the Memory Checker Module (see Section 4.2) Description of the Memory Checker Module registers (see Section 4.3) Description of the general registers (see Section 4.4) Interfaces of the Memory Checker Module (see Section 4.5) 4.1 Operational Overview From the programmer's point of view, the MCHK is a set of registers associated with this peripheral. To communicate respective error or operation events a port pin may be used for the signal “MATCH” to generate an external event and an interrupt line may be used for the signal “MISMATCH” to generate an internal event. The MCHK is reset together with the CPU so it can be used as a CPU coprocessor. This ensures a deterministic state of the MCHK after the CPU exits the reset state. SCU EINIT Clock Control f MCHK MCHK Module (Kernel) PD+ Bus Interrupt Control Core Interface MCHKINT MATCH Port Control Product Interface Module MCHK_INTERFACE Figure 4-1 User’s Manual MCHK, V1.10 Interface Diagram 4-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Note: The MATCH output is connected to an external port in packaged devices with 144 or more pins only. 4.2 Functional Description Conventional digital processing systems generally are configured around volatile and non-volatile memory elements. These memories provide (store) the data and instructions to the CPU doing the main computing of an embedded system. This includes the administration of the system by coordinating the operation of various system units to perform system tasks. Faults within these memories are in general critical for the safety and reliability of an embedded system. Therefore these memories have mechanisms to check for data consistency, e.g. parity or ECC (error correction code). These mechanisms can detect faults up to a certain amount (e.g. double bit faults) per word (bit line faults). Concatenated codes (block codes, word codes) can also detect multibit faults per word (word line faults), which increases the fault coverage. The MCHK is a parallel signature compression circuitry that calculates a concatenated CRC block code to increase test coverage by code concatenation. This enables error detection within a block of data stored in memory, registers, or communicated e.g. via serial communication lines. The MCHK reduces the probability of error masking due to repeated error patterns by compressing parallel test inputs from a block of data to be tested. Furthermore, it can generate pseudo random numbers. The MCHK uses a multiple input linear feedback shift register to generate a checksum (signature) of a block of data. A multiple input linear feedback shift register (MISR) is a shift register whose internal feedback input bit is driven by the exclusive-or (XOR) of some bits of the overall shift register value (LFSR: linear feedback shift register). This generator includes an arithmetic circuitry to calculate the block code. This circuitry is implemented as an independent piece of hardware and, therefore, does not rely on the memories to be tested. Only for the configuration it requires initialization data out of the memories. To avoid the need of a multi master system (CPU, DMA, etc.), the CPU (e.g. PEC, subroutine) is used to handle the data read and write transactions. These transactions rely on the memories to be tested because they may contain the respective CPU instruction code. Therefore the MCHK implements additional measures to enable detection of erroneous data move operations by the CPU. The following error scenarios are detected: • • • • The CPU configures the MCHK erroneously The CPU does not provide the data from the respective address range The CPU does not provide the correct amount of data The CPU is not able to check correctly the match of the online generated CRC block code and the expected offline (during development) generated CRC block code User’s Manual MCHK, V1.10 4-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) The principle of this circuitry is to generate an external coded life signal of the CRC block code check. Furthermore the configuration of the circuitry cannot be changed without an external notification. The life signal is not a static signal, but changes polarity in a predefined manner to avoid static faults, e.g. open and short circuit in the output stages. The circuitry consists of the following components: • • • • • • An arithmetic circuitry calculating the CRC block code out of the data transferred into an input register of this circuitry. A compare unit to check if the value of the calculated CRC block code is correct. The MCHK compares the content of the CRC block code result registers to a fixed value (FADE’EDDAH). Before calculating a CRC value, the result register is initialized with a specific value (magic word, seed), which results in a specific value after the CRC calculation. This so called magic word must be selected in a way that the block code ends up with the fixed value (FADE’EDDAH). This works fine for linear code, e.g. the CRC block code. A method granting the CRC block code calculation over a given amount of data. Therefore functional redundancy is used to grant this. A local count register within the MCHK initiates the compare of the calculated CRC block code after a given amount of input data. Secondly the CPU reloads the magic word (seed) of the CRC block code when initiating a new CRC block code generation (loop variable within data move subroutine or count register within PEC). An internal service request generation to enable software recovery in case of a fault. This could be a software routine running out of a different flash block than the one that produced the error, e.g. to support a limb home function. There is a residual risk: The CPU could write dummy data into the memory checker within this error interrupt routine and then rewrite the COUNT register. All configuration registers are protected by a time redundant mechanism. So modifications of configuration registers are only possible following a specific sequence of write operations (EINIT protected). Additionally, the COUNT register is protected by a content dependent access scheme. An external MATCH signal is generated on every successful CRC block code generation and may be used to trigger an external window watchdog. This window watchdog may generate a reset in case too many or too few MATCH signal toggles fall into a specific time window. To grant a correctly working block code unit, the application must also perform from time to time a block code generation outside the watchdog time window by having an incorrect compare (no MATCH signal toggle). This will check the correct function of the compare circuitry, because in case of an erroneous compare circuitry a MATCH signal toggle is generated outside the watchdog time window. User’s Manual MCHK, V1.10 4-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.2.1 Principle of the LFSR The list of the bit positions that affect the internal feedback bit is called the tap sequence. Figure 4-2 shows the principle of an LFSR. It assumes a tap sequence of [32, 26, 23, 22, 16, 12, 11, 10, 8, 7, 5, 4, 2, 1]. On activation, all bits in the LFSR are shifted left one bit position (in direction of the most significant bit). All bits on the tap position are exclusive-ORed and the result is fed into bit 0 as a feedback. On the next activation the same procedure is repeated. The LFSR outputs that influence the internal feedback input are called taps (marked gray in Figure 4-2). T=t 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 + 0 + + 0 1 T = t+1 + 1 + + + 1 0 1 + + 0 + + 0 0 1 + + 0 feedback 0 1 1 1 1 0 0 0 1 0 1 1 0 1 1 0 0 1 1 0 0 1 1 1 1 0 0 0 1 0 1 1 1 shift 7 6 5 4 3 2 1 0 1 1 1 1 0 0 0 1 0 1 1 0 1 1 0 0 1 1 0 0 1 1 1 1 0 0 0 1 0 1 1 1 + 1 + + 0 1 T = t+2 + 0 + + + 1 0 1 + + 1 1 + + 0 0 + + 1 0 feedback 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 0 shift 7 6 5 4 3 2 1 0 1 1 1 1 0 0 0 1 0 1 1 0 1 1 0 0 1 1 0 0 1 1 1 1 0 0 0 1 0 1 1 1 0 + 0 + + 1 1 + 0 + + + 1 0 1 + + 1 1 + + 1 0 + + 1 feedback 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 1 LSFR_OVERVIEW Figure 4-2 User’s Manual MCHK, V1.10 Principle of an LFSR 4-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) The tap sequence of an LFSR can be represented as a polynomial mod 2. This means that the coefficients of the polynomial must be 0 in case a respective feedback is not implemented or 1 in case a feedback tap is implemented. This is called the feedback polynomial or characteristic polynomial. For example, if the taps are at the 32nd, 26th, 23rd, 22nd, 16th, 12th, 11th, 10th, 8th, 7th, 5th, 4th, and the 2nd bits (as below), the resulting LFSR polynomial is G32 = x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5 + x4 + x2 + x + 1 (4.1) The '+1' in the polynomial does not correspond to a tap. The powers of the terms represent the tapped bits, counting from the least significant bit. The polynomial may be represented by a binary number (binary representation). Every power of the terms represent a 1 in the binary format counting from the most significant bit. So for the polynomial listed above, the number would be: G32 = 1000 0010 0110 0000 1000 1110 1101 1011B = 8260’8EDBH The polynomial used by the MCHK is defined in the tap polynomial registers TPRH and TPRL. User’s Manual MCHK, V1.10 4-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.2.2 Principle of the MISR In parallel to the internal feedback input bit of the LFSR, 16 external bits may be loaded into the LFSR (multiple input). These 16 input bits are exclusive-ORed with the bits shifted or fed back in the LFSR. MISR Input T= t 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 0 0 0 1 0 0 0 1 1 1 0 0 0 1 1 LFSR 30 28 26 24 22 20 18 16 31 29 27 25 23 21 19 17 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0 1 11 0 0 0 1 0 11 01 1 00 1 + 1 + 0 + 0 + 1 + 1 + 1 + 1 + 0 + 0 + 0 + 1 + 0 + 1 + 1 + 0 + 0 + 0 ++ 1 0 + + 1 0 + 1 + 0 + 0 + 0 + 1 + 1 + 0 + 0 T = t+1 LFSR 30 28 26 24 22 20 18 16 31 29 27 25 23 21 19 17 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1 11 0 0 01 0 11 0 1 10 0 10 + 0 + 0 + 1 + 0 + 1 + 1 + 0 + 1 + 1 + 0 + 0 + 1 + 1 + 1 + 1 + 1 + 0 ++ 1 1 + 1 + 1 + 0 + 1 + 0 + 1 + 1 + 0 + 1 + 0 MISR_OVERVIEW Figure 4-3 Principle of a MISR The initial value of the LFSR/MISR is called the seed and may be defined by an initial write to the result registers RRL and RRH. Because the operation of the MISR is deterministic, the sequence of values produced by the MISR is completely determined by its current (or previous) states and inputs. User’s Manual MCHK, V1.10 4-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.2.3 Commonly used Polynomials Polynomials for cyclic codes as used in globally standardized systems have not been fully standardized themselves. Most cyclic codes in current use have some weakness with respect to strength or construction. Standardization of cyclic codes would allow for better designed cyclic codes to come into common use. The following table provides a list of common polynomials used for sequential CRC signature generation. Table 4-1 Some Commonly used Polynomials Name Polynomial Maximum Data Width Normal (Reverse) of Reciprocal CRC-8-ATM x8+x2+x+1 8-bit 0000’0083H (0000’00C1H) CRC-8-CCITT x8+x7+x3+x2+1 8-bit 0000’00C6H CRC-8-Dallas x8+x5+x4+1 8-bit 0000’0098H CRC-8 x8+x7+x6+x4+x2+1 8-bit 0000’00EAH CRC-8 SAE J1850 x8+x4+x3+x2+1 8-bit 0000’008EH CRC-1 (parity) x8+x7+x6+x5+x4+x3+x2+x+1 8-bit 0000’00FFH CRC-10 x10+x9+x5+x4+x+1 10-bit 0000’0319H (0000’0263H) CRC-12 x12+x11+x3+x2+x+1 12-bit 0000’0C07H (0000’0E03H) CRC-15-CAN x15+x14+x10+x8+x7+x4+x3+1 15-bit (13-bit) 0000’62CCH (0000’19A3H) CRC-1 (parity) x16+x15+x14+x13+x12+x11+x10+x9+x8+x7+ x6+x5+x4+x3+x2+x+1 16-bit 0000’FFFFH CRC-16-CCITT x16+x12+x5+1 16-bit 0000’8810H CRC-16-IBM x16+x15+x2+1 16-bit 0000’C002H CRC-24Radix-64 x24+x23+x18+x17+x14+x11+x10+x7+x6+x5+ x4+x3+x+1 16-bit 00C3’267DH (24-bit (BE64’C300H) polynomial) User’s Manual MCHK, V1.10 4-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Table 4-1 Some Commonly used Polynomials (cont’d) Name Polynomial Maximum Data Width Normal (Reverse) of Reciprocal CRC-32IEEE802.3/ MPEG2 x32+x26+x23+x22+x16+x12+x11+x10+x8+x7+ 16-bit 8260’8ED6H 5 4 2 x +x +x +x+1 (32-bit (DB71’0641H) polynomial) CRC-32C x32+x28+x27+x26+x25+x23+x22+x20 +x19+x18+x14+x13+x11+x10+x9+x8+x6+1 16-bit 8F6E’37A0H (32/27-bit (05EC’76F1H) polynomial) Note: The polynomials above are in general used for sequential signature generation (in general named as CRC), resulting in different signatures than the parallel signature generation algorithm used by the MCHK. 4.2.4 Architecture of the Memory Checker Module The LFSR is represented by the Result Register High (RRH) and the Result Register Low (RRL). These may be used to initialize the MCHK with a seed value. When writing to register RRL, the MISR count register COUNT is reloaded with the last value written to register COUNT. The result registers RRH and RRL may be used to read the final or intermediate signature of a block of data loaded into the MCHK. Data may be loaded into the MCHK by writing either 8-bit or 16-bit data into the MISR input register IR. Each write access to register IR decrements the content of the MISR count register COUNT. The polynomial is defined by writing the binary normal reciprocal value into the TAP Polynomial Register High (TPRH) and TAP Polynomial Register Low (TPRL). The TAP polynomial registers and the result registers are combined by a binary AND. If the amount of ones in the result of this AND operation is odd, a 1B is fed back, else a 0B. The effectiveness of the Memory Checker Module is significantly reduced if a polynomial is used with a the most significant 1B bit position in the TAP Polynomial register being smaller than the most significant 1B in the data fed into the Memory Checker Module. So in general the content of the TAP polynomial register must be larger than 80H for 8-bit data and larger than 8000H for 16-bit data. If the content of the MISR count register COUNT is decremented from 0001H to 0000H, a service request signal is generated in case the content of the LFSR result register high (RRH) is not equal FADEH or the content of the LFSR result register low (RRL) is not equal EDDAH. If the content of the LFSR result registers equals FADE’EDDAH, the external MATCH signal is toggled. Figure 4-4 summarizes the architecture of the checksum circuit. User’s Manual MCHK, V1.10 4-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) PD+ Bus Write IR Memory Checker Input Register MCHKIN 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 … Q D Q D Q D Q D Q D Q D ………………………….. =1 … Q D =1 Q D … =1 Q D =1 Q D =1 … ………………………….. RRH Result Register HIGH MCHKRH 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 RRL Result Register Low MCHKRL 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 TPRH TAP Polynomial Register HIGH MCHKPH 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 TPRL TAP Polynomial Register Low MCHKPL 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 & User’s Manual MCHK, V1.10 … & & Figure 4-4 =1 =1 & & =1 =1 … =1 … =1 MCHK_PRINCIPLE Implementation of the MCHK Checksum Circuit 4-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.2.5 Preferable Usage of the Memory Checker Module Preferably the MCHK is used together with the CPU. The CPU reads the data block from the selected address area and writes it to the input register IR. Alternatively the PEC may be configured to move the data block to input register IR using 8-bit or 16-bit moves (PECCx.BWT = 0: 16-bit; PECCx.BWT = 1: 8-bit). Each write operation to register IR triggers a intermediate polynomial checksum calculation and the result of the calculation is stored in the result registers RRL and RRH. Furthermore, every write operation to register IR decrements the content of count register COUNT. In order to start a memory check sequence, the result register must be initialized with a seed (e.g. written with the desired start value) and a CPU or PEC transaction must be set up (start address, length, etc.). When the defined data block is completely written to register IR, an interrupt may be generated if the contents of the LFSR result registers RRH and RRL does not equal FADE’EDDAH. The MCHK may use e.g. the standard Ethernet (IEEE802.3/MPEG2) polynomial, which is given by: G32 = x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5 + x4 + x2 + x + 1 (4.2) Note: Although the polynomial above is used for generation, the result of the parallel signature generation (MISR) differs from the sequential signature generation (LFSR) used by the Ethernet protocol. 4.2.6 Calculation of Seed Values (Magic Word) To achieve a successful CRC calculation and MATCH or MISMATCH signal on a block of non volatile data, a data specific seed value, the so called magic word, has to be loaded into the LFSR result registers RRH and RRL prior to the CRC calculation. This magic word should be calculated during development of the respective data. Such a magic word can only be generated, if the order of the TAP polynomial is equal order 32 (most significant bit of TAP polynomial TPRH must be equal 1). Otherwise the higher order bits are non equal to the required end result FADE EDDAH. The following program sketches the principle of the program. It uses VBA (Microsoft Visual Basic) syntax. The Data_Array contains all the 16-bit data the CRC is to be calculated. COUNT passes the number of data the CRC is to be calculated, to the subroutine, as defined in the Memory Checker count register COUNT. The magical word is passed back to the calling routine through RRH and RRL, as it has to be written into the Result Register High (RRH) and the LFSR Result Register Low (RRL) as seed value. Because VBA has no unsigned integer format, a long integer format has been used within this demonstration code. Sub MagicWord(ByRef Data_Array() As Long, _ ByVal COUNT, TPRH, TPRL as Long, _ ByRef RRH, RRL As Long) User’s Manual MCHK, V1.10 4-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Dim i, j, order, feedback_bit As Integer Dim temp As Long RRH = &HFADE RRL = &HEDDA For j = COUNT To 1 Step -1 If TPRH <> 0 Then ' order of polynomial > 16 order = 31 Do While TPRH< 2 ^(order - 16)' calculate order of polynomial, order = order - 1 ' determines bit position Loop ' “rolled” out of LFSR Else ' order of polynomial < 17 order = 15 Do While TPRL< 2 ^order ' calculate order of polynomial, order = order - 1 ' determines bit position Loop ' “rolled” out of LFSR End If RRL = RRL Xor Data_Array(j) ' MISR XOR Input feedback_bit = RRL And 1 ' Extract CRC feedback bit RRL = RRL \ 2 ' 32 bit shift right (LFSR) RRL = RRL + (RRH And 1) * 2 ^ 15 RRH = RRH \ 2 temp = (RRH And TPRH) _ Xor (RRL And TPRL) ' generate 32 TAP Bits For i = 0 To 15 feedback_bit = feedback_bit _ Xor ((temp \ (2 ^ i)) And 1) Next i ' XOR TAP bits to bit If feedback_bit <> 0 Then ' TAP feedback bit is equal 1 If order > 16 Then RRH = RRH Or (2 ^ (order - 16)) Else RRL = RRL Xor (2 ^ order) End If End If Next j ' calculate CRC of all data End Sub User’s Manual MCHK, V1.10 4-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.2.7 Example Application Assuming MCHK is to be used to detect faults within a set of twenty one 16 bit Data. The Memory Checker Module uses the standard Ethernet (IEEE802.3/MPEG2) polynomial, which is given by: G32 = x32 + x26 + x23 + x22 + x16 + x12 + x11 + x10 + x8 + x7 + x5 + x4 + x2 + x + 1 (4.3) The Polynomial is therefore 8260 8EDBH and is written to TPRH and TPRL. Next the magical word of this set of data has to be calculated offline using a respective program as described in Section 4.2.6: “Calculation of Seed Values (Magic Word)” on Page 4-10. For the given data in Table 4-2, a magical word = AA1F ED4EH is calculated and written to RRH and RRL. Table 4-2 Example for a CRC Check User Action Data Value Content of Register COUNT RRH RRL TPRH TPRL TAP Polynomial written to TPRH 8260H xxxx xxxx xxxx 8260H xxxx TAP Polynomial written to TPRL 8EDBH xxxx xxxx xxxx 8260H 8EDBH Magical Word written into RRH AA1FH 0015H AA1FH xxxx 8260H 8EDBH Magical Word written into RRL ED4EH 0015H AA1FH ED4EH 8260H 8EDBH Data Amount written to COUNT 0015H 0015H xxxx xxxx 8260H 8EDBH Data 1 written into IR 8BEDH 0014H 543FH 5171H 8260H 8EDBH Data 2 written into IR AA61H 0013H A87EH 0883H 8260H 8EDBH Data 3 written into IR C64EH 0012H 50FCH D749H 8260H 8EDBH Data 4 written into IR 17E4H 0011H A1F9H B976H 8260H 8EDBH Data 5 written into IR A329H 0010H 43F3H D1C5H 8260H 8EDBH Data 6 written into IR 66B5H 000FH 87E7H C53EH 8260H 8EDBH Data 7 written into IR 422AH 000EH 0FCFH C857H 8260H 8EDBH Data 8 written into IR 4FF6H 000DH 1F9FH DF58H 8260H 8EDBH Data 9 written into IR 4046H 000CH 3F3FH FEF6H 8260H 8EDBH Data 10 written into IR 911CH 000BH 7E7FH 6CF0H 8260H 8EDBH Data 11 written into IR 1FA0H 000AH FCFEH C640H 8260H 8EDBH User’s Manual MCHK, V1.10 4-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Table 4-2 User Action Example for a CRC Check (cont’d) Data Value Content of Register COUNT RRH RRL TPRH TPRL Data 12 written into IR BF38H 0009H F9FDH 33B9H 8260H 8EDBH Data 13 written into IR 9FE3H 0008H F3FAH F891H 8260H 8EDBH Data 14 written into IR 44DDH 0007H E7F5H B5FEH 8260H 8EDBH Data 15 written into IR 749AH 0006H CFEBH 1F67H 8260H 8EDBH Data 16 written into IR 8C09H 0005H 9FD6H B2C7H 8260H 8EDBH Data 17 written into IR D0F5H 0004H 3FADH B57AH 8260H 8EDBH Data 18 written into IR DC5FH 0003H 7F5BH B6ABH 8260H 8EDBH Data 19 written into IR DB06H 0002H FEB7H B651H 8260H 8EDBH Data 20 written into IR 4604H 0001H FD6FH 2AA7H 8260H 8EDBH Data 21 written into IR B894H 0000H FADEH EDDAH 8260H 8EDBH User’s Manual MCHK, V1.10 4-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.3 Memory Checker Module Registers From the programmer's point of view, the MCHK is composed of a set of SFRs as summarized below. Memory Checker Control Registers Memory Checker Data Registers MCHK_IR MCHK_COUNT MCHK_RRL MCHK_TPRL MCHK_RRH MCHK_TPRH mchk_reg_its Figure 4-5 Memory Checker Module Kernel Registers The following tables show the MCHK registers and their addresses. Table 4-3 Registers Address Space Module Base Address MCHK 0000H Table 4-4 End Address Note Registers Overview Register Register Long Name Short Name Offset Page Address Number ID Module Identification Register FFE0H Page 4-20 IR Memory Checker Input Register FE58H Page 4-15 RRL Memory Checker Result Register Low F058H Page 4-16 RRH Memory Checker Result Register High F05AH Page 4-16 COUNT Memory Checker Count Register FE5AH Page 4-18 TPRL Memory Checker Polynomial Register Low F05CH Page 4-19 TPRH Memory Checker Polynomial Register High F05EH Page 4-19 Note: All registers are reset by the same reset class as the CPU is reset. User’s Manual MCHK, V1.10 4-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.3.1 Memory Checker Input Register The input register receives the data written to the MCHK for checksum calculation. If the CPU moves to register MCHK_IR are 8-bit wide, the unused register bits of the 16-bit MCHKIN value are taken as 0s for the current result calculation. IR Input Register 15 14 13 SFR(FE58H/2CH) 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MCHKIN w Field Bits Type Description MCHKIN [15:0] w Memory Checker Module Input The value written to MCHKIN is used for the next checksum calculation. Any read action will deliver 0000H. Note: MCHK_IR is a write-only register. Any read action will deliver 0000H. User’s Manual MCHK, V1.10 4-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.3.2 Memory Checker Result Registers The result registers contain the signature (result) of the memory check operation. Before starting a checksum calculation operation, they should be written with the initial checksum calculation value (seed). RRL Result Register Low 15 14 13 12 ESFR(F058H/2CH) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MCHKRL rwh Field Bits Type Description MCHKRL [15:0] rwh Memory Checker Result Low This bit field contains the least significant 16 bits of the current result of the 32-bit checksum calculation operation. Note: Writing to the RRL.MCHKRL will reset (reload) the MCHKCNT.COUNT register to the last data written to this register MCHKCNT.COUNT value. Therefore writing to RRL will immediately initialize a new CRC calculation cycle. RRH Result Register High 15 14 13 12 ESFR(F05AH/2DH) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MCHKRH rwh Field Bits Type Description MCHKRH [15:0] rwh User’s Manual MCHK, V1.10 Memory Checker Result High This bit field contains the most significant 16 bits of the current result of the 32-bit checksum calculation operation. 4-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.3.3 Memory Checker Count Register The count register COUNT is decremented on each write access to the input register. If the count register is decremented to 0000H, a service request (interrupt) is generated if the content of the result registers is non equal FADE’EDDAH, or instead the output signal MATCH is toggled if the result registers are equal FADE’EDDAH. The count register is reloaded with the last value written to it, when the CPU transfers a new seed value (magic word) to the LFSR result register low (RRL). When the CPU or PEC writes to register COUNT and its content is not equal to the last value written to it, the service request (interrupt) is generated and the output signal MATCH is toggled. This enables detection of software not correctly handling the MCHK, e.g. due to an erroneous program memory. The timely correct toggling of the MATCH signal may be used as a life signal e.g. by an external window watchdog. The reset value of the MATCH signal = 0B. Because register COUNT controls a safety critical system function, it is protected by a special register security mechanism so this vital system function cannot be changed inadvertently after executing the EINIT instruction. User’s Manual MCHK, V1.10 4-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) COUNT Count Register 15 14 13 SFR(FE5AH/2DH) 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 MCHKCNT rwh Field Bits Type Description MCHKCNT [15:0] rwh Memory Checker Count MCHKCNT indicates the number of remaining data in the current data block to be entered into MCHK. 0001H One remaining data to be written to register IR to trigger the compare for the MATCH signal, interrupt signal NOMATCH. 0002H Two remaining data to be written to register IR. …H … FFFEH 65534 remaining data to be written to register IR. FFFFH 65535 remaining data to be written to register IR. 0000H 65536 remaining data to be written to register IR to trigger the compare for the MATCH signal, interrupt signal NOMATCH. Note: Register COUNT should only be written if MCHKCNT is equal to the last value written to this register. Otherwise, a service request (interrupt) will be triggered and the MATCH signal will be toggled. Modify register COUNT only after writing to register RRL, which reloads COUNT to the previously written value (see also Table 4-2). Note: COUNT is write protected after the execution of EINIT by the register security mechanism. 4.3.4 Memory Checker Polynomial Registers The polynomial registers contain the LFSR polynomial of the checksum calculation operation. User’s Manual MCHK, V1.10 4-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Because the polynomial registers control a safety critical system function, they are protected by a special register security mechanism so this vital system function cannot be changed inadvertently after executing the EINIT instruction. TPRL Tap Polynomial Register Low 15 14 13 12 11 10 ESFR(F05CH/2EH) 9 8 7 6 Reset Value: FFFFH 5 4 3 2 1 0 MCHKPL rwh Field Bits Type Description MCHKPL [15:0] rwh Memory Checker Polynomial Low This bit field contains the least significant 16 bits of the binary tap polynomial format. TPRH Tap Polynomial Register High ESFR(F05EH/2FH) 15 14 13 12 11 10 9 8 7 6 Reset Value: FFFFH 5 4 3 2 1 0 MCHKPH rwh Field Bits Type Description MCHKPH [15:0] rwh Memory Checker Polynomial High This bit field contains the most significant 16 bits of the binary tap polynomial format. Note: TPRH and TPRL is write protected after the execution of EINIT by the register security mechanism. User’s Manual MCHK, V1.10 4-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) 4.4 General Registers The ID register is a read-only register used for MCHK module identification purposes. It provides 8 bits for module identification and 8 bits for revision numbering. 4.4.1 ID Register ID Module Identification Register 15 14 13 12 11 10 MEM(FFE0H) 9 8 7 Reset Value: 3BXXH 6 5 4 3 MOD_NUMBER MOD_REV r r 2 1 0 Field Bits Type Description MOD_REV [7:0] r Module Revision Number Value Bits 7-0 bits are used for module revision numbering. The value of the module revision number starts with 01H (first revision). MOD_NUMBER [15:8] r Module Identification Number Value Bits 15-8 are used for module identification. The MCHK has the module number 3BH. 4.5 Interfaces of the MCHK Module The MCHK module can generate an interrupt request and an external life signal. The interrupt request signal is connected to the SCU and can be routed to one of the SCU’s interrupt nodes. The MATCH output is connected to an external pin in packaged devices with 144 or more pins only. Table 4-5 MCHK Digital Connections in XC27x5X Signal from/to Module I/O to MCHK MATCH P8.6 O INT (MISMATCH) SCU O MCHKIN_Write_Access Write trigger from MCHKIN I1) User’s Manual MCHK, V1.10 4-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) Table 4-5 Signal MCHK Digital Connections in XC27x5X (cont’d) from/to Module I/O to MCHK MCHKCNT_Write_Access Write trigger from MCHKCNT I1) EINIT I SCU 1) This signal is generated within the module itself and is not present at the module boundary. User’s Manual MCHK, V1.10 4-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Checker Module (MCHK) User’s Manual MCHK, V1.10 4-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5 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 XC27x5X 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 XC27x5X, up to five instructions can be processed in parallel. Most instructions of the XC27x5X 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 XC27x5X operate independently of the CPU. Data and control information are interchanged between the CPU and these peripherals via Special Function Registers (SFRs) or shared memory areas. 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. User’s Manual C166S-V2 CPU, V2.7 5-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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 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. The CPU provides a set of instructions for enabling (ENWDT), disabling (DISWDT) and servicing (SRVWDT) the watchdog timer. In addition to its active operation state, the CPU can enter idle mode by executing the IDLE instruction. In idle mode the CPU stops program execution but still reacts to interrupt or PEC requests. Transition to the active state can be forced by an interrupt request or a hardware reset. The PWRDN instruction is not enabled in the XC27x5X. If executed a NOP will be performed instead. System power state transitions are controlled by the System Control Unit (SCU). 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 CPU identification: CPUID 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 C166S-V2 CPU, V2.7 5-2 V1.2, 2008-09 XC27x5X 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 C166S-V2 CPU, V2.7 5-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.1 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 5-1). Prefetch Unit and Branch Unit feed the pipeline minimizing CPU stalls due to instruction reads. The Address Unit (ADU) supports sophisticated addressing modes avoiding additional instructions needed otherwise. Arithmetic and Logic Unit (ALU) and Multiply and Accumulate Unit (MAC) handle differently sized data and execute complex operations. Three memory interfaces and Write Buffer (WB) minimize CPU stalls due to data transfers. PSRAM Flash/ROM PMU CPU Prefetch Unit Branch Unit FIFO CSP IP VECSEG CPUCON1 CPUCON2 Return Stack IDX0 IDX1 QX0 QX1 QR0 QR1 +/- +/- Multiply Unit MRW 2- Stage Prefetch Pipeline TFR 5-Stage Pipeline Injection/ Exception Handler IFU DPP0 DPP1 DPP2 DPP3 DSRAM PIPE SPSEG SP STKOV STKUN ADU Division Unit Bit-Mask-Gen. Multiply Unit Barrel-Shifter MCW MDC +/- MSW PSW +/- MAH MAL MDH MDL ZEROS ONES MAC DPRAM CP R15 R15 R14R15 R14 R14 local bank GPRs GPRs GPRs R1 R1 R0 R1 R0 R0 R15 R14 global bank GPRs R1 R0 RF Buffer ALU WB EBC Peripherals DMU MCA4917_X Figure 5-1 CPU Block Diagram User’s Manual C166S-V2 CPU, V2.7 5-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) In general the instructions move through 7 pipeline stages (Section 5.3). The stages can be grouped as follows: • • 2 stages fetch pipeline - receives instructions from program memory and stores them into an instruction FIFO. Fetch pipeline stages can be bypassed. 5 stages processing pipeline - executes each instruction received from fetch stages. Because passing through one pipeline stage takes at least one clock cycle and because the fetch pipeline stages can be bypassed, 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. 5.2 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 C166S-V2 CPU, V2.7 5-5 V1.2, 2008-09 XC27x5X 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 5-2 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 C166S-V2 CPU, V2.7 5-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.2.1 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 5-1 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 5.2.2 Zero-Cycle Jumps The “Zero-Cycle Jumps” are one of the advanced XC27x5X specifics, which becomes possible due to the complex pipelined structure for processing instruction-flow. This feature allows, under some circumstances, jumps to be executed in “null time”. In fact, a jump is “hooked” to the previous instruction and the two instructions pass through the pipeline as one instruction. This can be only possible, if the jump instruction does not need any of the pipeline resources needed by the predecessor. Hence, the following rules are essential: • a jump can not be hooked onto another jump instruction, as the pipeline resource “target IP” can not be shared between the two; User’s Manual C166S-V2 CPU, V2.7 5-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) • a jump can not be executed in zero-cycle if it requires any memory access, as basically any predecessor instruction might access a memory. The above are only preliminary conditions, needed to make a jump zero-cycle. But would this really happen, it’s not reliable enough to predict: it also depends on the exact instruction sequence, speed of the program memory etc. What can be summarized is: • • only JMPA, JMPR and JMPS Instructions can be converted to zero-cycle; if the immediately preceding instruction is not a branch (any JMP, CALL or RET). If a Jump is executed as zero-cycle, in fact the address of this Jump will not be assigned to the Instruction Pointer. Note: No IP-Breakpoint must be set over an instruction, which satisfies the two prepositions above for a zero-cycle Jump. Otherwise, if set, it is very possible this Breakpoint will be missed by the debug module. 5.2.3 Atomic and Extend Instructions The atomic and extend instructions (ATOMIC, EXTR, EXTP, EXTS, EXTPR, EXTSR) disable standard and PEC interrupts and class A traps until completion of the immediately following sequence of instructions. The number of instructions in the sequence may vary from 1 to 4. It is coded in the 2-bit constant field #irang2 and takes values from 0 to 3. The EXTended instructions additionally change the addressing mechanism during this sequence (see instruction description). ATOMIC and EXTended instructions become active immediately, so no additional NOPs are required. All instructions requiring multi cycles or hold states for execution are considered to be one instruction. The ATOMIC and EXTended instructions can be used with any instruction type. If a branch instruction following immediately after an atomic sequence is executed as zero-cycle jump, then this branch is part of the atomic sequence as well. If the branch instruction is not a part of the ATOMIC sequence, it should not be hooked on to the atomic sequence, a NOP could be inserted in between. Note: If a class B trap interrupt occurs during an ATOMIC or EXTended sequence, then the sequence is terminated, an interrupt lock is removed, and the standard condition is restored before the trap routine is executed. The remaining instructions of the terminated sequence executed after returning from the trap routine will run under standard conditions. Note: When using nested ATOMIC and EXTended instructions. There is only one counter to control the length of the sequence, i.e. issuing an ATOMIC or EXTended instruction within a sequence will reload the counter with the value of the new instruction. User’s Manual C166S-V2 CPU, V2.7 5-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.3 Instruction Processing Pipeline The XC27x5X 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 debug 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 XC27x5X’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 operand read and write values and resolves most of the possible conflicts — such as User’s Manual C166S-V2 CPU, V2.7 5-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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 cases in which the pipeline requires attention by the programmer. 5.3.1 Access to the IO Area Read or write accesses to the IO Areas of the XC27x5X memory space enforce particular pipeline behavior. Thus the requirements of peripheral devices with registers located in these areas are handled appropriately. The following typical properties of peripheral device registers are considered: • • Upon a write to a peripheral register the contents of any (also multiple) peripheral register(s) may change as a consequence of the write. Upon a read from a peripheral register the contents of the same register may change as a consequence of the read (e.g. read buffer of a serial channel) These cases are handled by following pipeline measures: Write before read execution enforced If the instructions in the pipeline contain a write action followed by a read action both to the IO areas then the read action is delayed (held in memory stage) until the write action has passed through the writeback stage. Thus the write action will always be scheduled before a read action. Attention: Due to additional system delay this does not guarantee that a write will become effective before a read at the target registers. Additional system delay is accumulated by the bus system or caused by the peripheral itself. In case the additional read delay differs from the write delay the read may overtake the write. However since the on-chip delays are similar the programmer must take care about this in particular when using off-chip peripherals allocating IO area through the EXTBUS. Prevention of buffered writes Write access to the IO area is not buffered in the writeback buffer. 5.3.2 Pipeline Conflicts The following examples describe the pipeline behavior in special cases and give provide rules to optimize performance by instruction re-ordering. Note: The XC27x5X has a fully interlocked pipeline, which means that pipeline conflicts do not cause any malfunction. Instruction re-ordering is only required for performance reasons. User’s Manual C166S-V2 CPU, V2.7 5-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.3.2.1 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 buses 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 Table 5-2 Resolved Pipeline Dependencies Using GPRs Tn+41) 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+3 Tn+52) 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 WRITE BACK to MEMORY. 2) R6 forwarded from EXECUTE to MEMORY. 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 In+1 MOV R3,[R0] ;Use R0 as address pointer In+2 ADD R6,R0 In+3 ADD R6,R1 User’s Manual C166S-V2 CPU, V2.7 5-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-3 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). 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: ADD R0,R1 ;Compute new value for R0 In In+1 ADD R6,R0 ;R0 is not updated, just read In+2 ADD R6,R1 In+3 MOV R3,[R0] ;Use R0 as address pointer Table 5-4 Pipeline Dependencies Using GPRs as Pointers (No Stall) Tn+2 Tn+31) Stage Tn Tn+1 Tn+4 DECODE In = ADD R0, R1 In+1 = ADD In+2 = ADD In+3 = MOV In+4 R6, R0 R6, R1 R3, [R0] 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). User’s Manual C166S-V2 CPU, V2.7 5-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.3.2.2 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. Conflict_GPRs_Pointer_WrongHistory: ADD R3,[R0] ;R0 points to DPRAM (e.g.) In 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 Table 5-5 Pipeline Dependencies with Pointers (Valid Speculation) Stage Tn Tn+1 Tn+2 DECODE In = ADD R3, [R0] In+1 = MOV In+2 R0, R4 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] User’s Manual C166S-V2 CPU, V2.7 5-13 In+1 = MOV R0, R4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-6 Pipeline Dependencies with Pointers (Invalid Speculation) Tm+1 Tm+21) Stage Tm Tm+3 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). 5.3.2.3 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. 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: ADD op1,R1 In In+1 ADD R6,R0 In+2 CoMAC [IDX0],[R0] User’s Manual C166S-V2 CPU, V2.7 5-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) In+3 MOV R3,[R0] Table 5-7 Pipeline Dependencies in Case of Memory Conflicts (DPRAM) Tn+2 Tn+3 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+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. 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 Table 5-8 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 User’s Manual C166S-V2 CPU, V2.7 5-15 In+1 = ADD – R6, R0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-8 Pipeline Dependencies in Case of Memory Conflicts (DSRAM) Tn+1 Tn+2 Tn+3 Tn+41) Tn+5 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 Stage Tn 1) ADD R6, op2 instruction stalls due to memory bandwidth conflict. 5.3.2.4 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 a pipeline cancelation 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 these registers in the MEMORY stage, the value cannot be forwarded. The instruction must be stalled for one cycle on the MEMORY stage. Conflict_CSFR_Update_Stall: In MUL R0,R1 In+1 MOV R6,MDL In+2 ADD R6,R1 In+3 MOV R3,[R0] User’s Manual C166S-V2 CPU, V2.7 5-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-9 Pipeline Dependencies with Result CSFRs (Stall) Tn+2 Tn+31) Stage Tn Tn+1 Tn+4 Tn+5 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] 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. 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 Table 5-10 Pipeline Dependencies with Result CSFRs (No Stall) Tn+2 Tn+3 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+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 C166S-V2 CPU, V2.7 5-17 V1.2, 2008-09 XC27x5X 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 CPUCON1/2, 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 CPUCON1/2, 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. Conflict_Canceling: MOV IDX1,#12 In In+1 MOV R6,mem In+2 ADD R6,R1 In+3 MOV R3,[R0] User’s Manual C166S-V2 CPU, V2.7 5-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-11 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 Table 5-12 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 In+1 = MOV MAH, #23 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. User’s Manual C166S-V2 CPU, V2.7 5-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Conflict_Canceling_Completely: In MOV PSW,R4 In+1 MOV R6,R1 In+2 ADD R6,R1 In+3 MOV R3,[R0] Table 5-13 Pipeline Dependencies with Control CSFRs (Cancel All) Stage Tn+1 Tn+2 Tn+3 Tn+4 DECODE In+1 = MOV In+2 = ADD In+2 = ADD – R6, R1 R6, R1 R6, R1 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 C166S-V2 CPU, V2.7 5-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.4 CPU Configuration Registers The CPU configuration registers select a number of general features and behaviors of the XC27x5X’s CPU core. In general these registers are only written by the startup software and not altered during application software run time. Note: The CPU configuration registers are protected by the register security mechanism after the EINIT instruction has been executed. CPUCON1 CPU Control Register 1 15 14 13 12 SFR (FE18H/0CH) 11 10 9 8 7 0 r r r r Reset Value: 0007H 6 5 VECSC r r r r r rw 4 3 2 1 WDT SGT INTS BP CTL DIS CXT rw rw rw rw 0 ZCJ Field Bits Type Description 0 [15:7] r Reserved Read as 0, should be written 0 VECSC [6:5] rw Scaling Factor of Vector Table 00B Space between two vectors is 2 words1) 01B Space between two vectors is 4 words 10B Space between two vectors is 8 words 11B Space between two vectors is 16 words WDTCTL 4 rw Configuration of Watchdog Timer 0B DISWDT executable only until End Of Init2) 1B DISWDT/ENWDT always executable (enhanced WDT mode) SGTDIS 3 rw Segmentation Disable/Enable Control 0B Segmentation enabled Segmentation disabled 1B INTSCXT 2 rw Enable Interruptibility of Switch Context 0B Switch context is not interruptible 1B Switch context is interruptible BP 1 rw Enable Branch Prediction Unit 0B Branch prediction disabled Branch prediction enabled 1B ZCJ 0 rw Enable Zero-Cycle Jump Function 0B Zero-cycle jump function disabled 1B Zero-cycle jump function enabled User’s Manual C166S-V2 CPU, V2.7 5-21 rw V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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. CPUCON2 CPU Control Register 2 15 14 13 12 SFR (FE1AH/0DH) 11 10 FIFODEPTH FIFOFED rw rw 8 7 BYP BYP PF F rw rw 1 r 6 5 4 3 2 STE OV RET FAS LFIC N RUN ST TBL rw rw rw rw rw 1 0 1 SL r rw Field Bits FIFODEPTH [15:12] rw FIFO Depth Configuration 0H No FIFO (entries) One FIFO entry 1H … … 8H Eight FIFO entries 9H reserved … … FH reserved FIFOFED [11:10] rw FIFO Fed Configuration 00B FIFO disabled 01B FIFO filled with up to one instruction per cycle 10B FIFO filled with up to two instructions per cycle 11B FIFO filled with up to three instruction per cycle BYPPF 9 rw Prefetch Bypass Control 0B Bypass path from prefetch to decode disabled Bypass path from prefetch to decode available 1B BYPF 8 rw Fetch Bypass Control 0B Bypass path from fetch to decode disabled 1B Bypass path from fetch to decode available 1 7 r Reserved Read as 1, should be written 1 STEN 6 rw Stall Instruction Enable (for debug purposes) 0B Stall Instruction disabled 1B Stall Instruction enabled (see example below) LFIC 5 rw Linear Follower Instruction Cache 0B Linear Follower Instruction Cache disabled 1B Linear Follower Instruction Cache enabled User’s Manual C166S-V2 CPU, V2.7 Type 9 Reset Value: 8FBBH Description 5-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description OVRUN 4 rw Pipeline Control 0B Overrun of pipeline bubbles not allowed 1B Overrun of pipeline bubbles allowed RETST 3 rw Enable Return Stack Return Stack is disabled 0B Return Stack is enabled 1B FASTBL 2 rw Enables the fast injection of block transfers 0B Direct injection disabled 1B Direct injection enabled 1 1 r Reserved Read as 1, should be written 1 SL 0 rw Enables Short Loop Mode 0B Short loop mode disabled 1B Short loop mode enabled Note: This register must only be modified when explicitly documented - e.g. in an errata sheet. User’s Manual C166S-V2 CPU, V2.7 5-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.5 Use of General Purpose Registers The CPU provides three 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 One global register bank which is memory-mapped and cached in the register file Registerfile DPRAM AGU Write Port ALU Write Port R15 R14 ... R15 R15 R14 R14 ... ... R15 R14 ... Memory image Memory Global register of global mapping bank register bank ... R1 R0 Local register Local register banks banks ... ... R1 R1 R0 R0 ... R1 R0 AGU Read Port ALU Read Port 1 ALU Read Port 2 mcd04873 .vsd Figure 5-3 Register File 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. User’s Manual C166S-V2 CPU, V2.7 5-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The local register banks are built of dedicated physical registers, while the global register bank represents a cache. Multiple global banks can be mapped to the internal DPRAM. Each of these banks 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 R14 (CP) + 30 (CP) + 28 R13 Register File R12 R11 15 0 16-Bit Context Pointer R10 R15 R9 R8 R7 R6 R0 R5 R4 Global R3 local R2 R1 (CP) + 2 R0 (CP) MCA04921 Figure 5-4 Register Bank Selection via Register CP User’s Manual C166S-V2 CPU, V2.7 5-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.5.1 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 register word (Rw) or byte (Rb) 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 5-5). 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. 11 12-Bit Context Pointer 1 0 Specified by reg or bitoff 1 1 1 1 For byte GPR accesses 4-Bit GPR Address *1 *2 For word GPR accesses Internal DPRAM Must be within the internal DPRAM area + GPRs MCA04922 Figure 5-5 Implicit CP Use by Logical Short GPR Addressing Modes User’s Manual C166S-V2 CPU, V2.7 5-26 V1.2, 2008-09 XC27x5X 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 the read is redirected to the global register bank. The data that is read from cache will be used and the 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 requires an extra cycle for the read and write access. Table 5-14 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 bank. User’s Manual C166S-V2 CPU, V2.7 5-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.5.2 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 bank 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 5-6 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 started automatically. User’s Manual C166S-V2 CPU, V2.7 5-28 V1.2, 2008-09 XC27x5X 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 bank1) 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 5-7). If the interrupt uses a local register bank the validation process is interrupted and the service routine is executed immediately (see Figure 5-8). 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 5-9) 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 5-7 and Figure 5-9). 1) During the store phase of the context switch the complete register bank is written to the DPRAM even if the application only uses a part of this register bank. A register bank must not be located above FDE0H, otherwise the store phase will overwrite SFRs (beginning at FE00H). User’s Manual C166S-V2 CPU, V2.7 5-29 V1.2, 2008-09 XC27x5X 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 5-7 Global Bank Global Bank Execution Task B Execution Task A Execution of POP CP Register Bank Validation Process Started Finished Execution of RETI 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 5-8 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 Execution Task C Register Bank Validation Process Restarted Finished Register Bank Validation Process Started Stopped Figure 5-9 Local Bank Global Bank Execution Task B Execution Task A Execution of RETI Execution of RETI MCA04876 Validation Process Interrupted by Local- and Global-Bank Intr. User’s Manual C166S-V2 CPU, V2.7 5-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.5.2.1 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 15 14 13 12 1 1 1 1 CP 0 r r r r rw r Type 9 8 7 6 Reset Value: FC00H 5 4 3 2 1 0 Field Bits Description 1 15, 14, r 13, 12 Fixed part of CP Read as 1 CP [11:1] rw Modifiable part of CP Specifies bits [11:1] of the 16-bit 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. 0 0 r Fixed part of CP Read as 0 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. Do not set CP above FDE0H, otherwise the store phase will overwrite SFRs (beginning at FE00H). The XC27x5X 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 User’s Manual C166S-V2 CPU, V2.7 5-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) instruction immediately following the instruction that updates CP register can use the new value of the changed CP. User’s Manual C166S-V2 CPU, V2.7 5-32 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.6 Code Addressing The XC27x5X 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 8 7 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 5-10 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 hardware reset value is 0000H, but immediately after reset it is loaded with the contents of the VECSEG register due to an injected MOVCSIP instruction. 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. User’s Manual C166S-V2 CPU, V2.7 5-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) For processing an accepted interrupt or a TRAP, register CSP is automatically loaded with the segment of the vector table (defined in register VECSEG). 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 15 14 13 12 SFR (FE08H/04H) 11 10 9 8 7 6 Reset Value: 0000H 5 0 r r r r 4 3 2 1 0 SEGNR r r r r rh Field Bits Type Description 0 [15:8] r Reserved Read as 0, should be written 0 SEGNR [7:0] rh Segment Number 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 XC27x5X’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 User’s Manual C166S-V2 CPU, V2.7 (not addressable) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 IP 0 h r 5-34 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description IP [15:1] h Instruction Pointer Specifies bits [15:1] of the intra segment offset from which the current instruction is to be fetched. IP refers to the current segment <SEGNR>. 0 0 r Fixed part of IP Read as 0 User’s Manual C166S-V2 CPU, V2.7 5-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7 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. 5.7.1 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 5-15 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 C166S-V2 CPU, V2.7 5-36 V1.2, 2008-09 XC27x5X 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 5-15). 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 C166S-V2 CPU, V2.7 5-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.2 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 5-11). Select the used data page directly: The data page is selected by a preceding 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 preceding 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 01'0 000H 0 Page Page Offset Segment Segment Offset 00'0 000H MCA 04924 Figure 5-11 Data Page Pointer Addressing User’s Manual C166S-V2 CPU, V2.7 5-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.2.1 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 DPP1 Data Page Pointer 1 DPP2 Data Page Pointer 2 DPP3 Data Page Pointer 3 15 14 13 12 11 10 SFR (FE00H/00H) Reset Value: 0000H SFR (FE02H/01H) Reset Value: 0001H SFR (FE04H/02H) Reset Value: 0002H SFR (FE06H/03H) Reset Value: 0003H 9 8 7 0 r r r 6 5 4 3 2 1 0 PN r r r rw Field Bits Type Description 0 [15:8] r Reserved Read as 0, should be written 0 PN [9:0] rw Data Page Number of DPPx Specifies the data page selected via DPPx. 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 5-11. 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). A DPP register can be updated via any instruction capable of modifying an SFR. User’s Manual C166S-V2 CPU, V2.7 5-39 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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 MCA 04925 Figure 5-12 Overriding the DPP Mechanism Note: The overriding page or segment may be specified as a constant (#pag, #seg) or via a word GPR (Rw). Table 5-16 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 C166S-V2 CPU, V2.7 5-40 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.3 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 5-17 Generating Physical Addresses from Indirect Pointers Step Executed Action Calculation Notes 1 Calculate the address of the GPR Address = indirect pointer (word GPR) 2 × Short Addr. from its short address [+ (CP)] see Table 5-15 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 5-16 5 Post-in/decrement indirect (GPR Address) = Optional step, executed only if pointer (‘Rw±’) depending (GPR Address) required by addressing mode 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 C166S-V2 CPU, V2.7 5-41 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-18 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. 5.7.3.1 Offset Registers QR0 and QR1 The non-bit-addressable offset registers QR0 and QR1 are used with CoXXX instructions. For possible instruction flow stalls refer to Section 5.3.2.4. QR0 Offset Register QR1 Offset Register 15 14 13 12 11 10 ESFR (F004H/02H) Reset Value: 0000H ESFR (F006H/03H) Reset Value: 0000H 9 8 7 6 5 4 3 2 1 QR 0 rw r Field Bits Type Description QR [15:1] rw Modifiable part of QRx Specifies the 16-bit offset address for indirect addressing modes (LSB always zero). 0 0 r Fixed part of QRx Read as 0 User’s Manual C166S-V2 CPU, V2.7 0 5-42 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.4 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 5-11 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 5-13. IDX0 Address Pointer IDX1 Address Pointer 15 14 13 12 11 10 SFR (FF08H/84H) Reset Value: 0000H SFR (FF0AH/85H) Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 IDX 0 rw r Field Bits Type Description IDX [15:1] rw Modifiable part of IDXx Specifies the 16-bit address pointer 0 0 r Fixed part of IDXx Read as 0 Note: During the initialization of the IDX registers, instruction flow stalls are possible. For the proper operation, refer to Section 5.3.2.4. User’s Manual C166S-V2 CPU, V2.7 5-43 V1.2, 2008-09 XC27x5X 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 5-14 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 QX1 Offset Register 15 14 13 12 11 10 ESFR (F000H/00H) Reset Value: 0000H ESFR (F002H/01H) Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 QX 0 rw r Field Bits Type Description QX [15:1] rw Modifiable part of QXx Specifies the 16-bit word offset for indirect addressing modes 0 0 r Fixed part of QXx Read as 0 Note: During the initialization of the QX registers, instruction flow stalls are possible. For the proper operation, refer to Section 5.3.2.4. User’s Manual C166S-V2 CPU, V2.7 5-44 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Memory 15 16-Bit IDX Pointer 12 11 0 12 11 0 2 02'0000H 1 01'0000H DPRAM in Data Page3 0 23 00'0000H 15 000 000 0011 11 MCA04926 Figure 5-13 Arithmetic MAC Operations and Addressing via the IDX Pointers Table 5-19 Generating Physical Addresses from Indirect Pointers (IDXx) Step Executed Action Calculation Notes 1 Determine the used IDXx pointer --- – 2 Calculate an intermediate Interm. Addr. = 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 – User’s Manual C166S-V2 CPU, V2.7 Long Address = (IDXx Pointer) 5-45 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-19 Generating Physical Addresses from Indirect Pointers (IDXx) (cont’d) Step Executed Action Calculation Notes 4 Calculate the physical 24bit address using the resulting pointer Physical Addr. = Page/Segment + Pointer offset Uses DPPs or page/segment override mechanisms, see Table 5-16 and Figure 5-13 5 Post-in/decrement indirect pointer (‘IDXx±’) by 2 (∆ = 2), or depending on offset registers (∆ = QXx) (IDXx Pointer) = (IDXx Pointer) ± ∆ Optional step, executed only if required by addressing mode The following indirect addressing modes are provided: Table 5-20 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. User’s Manual C166S-V2 CPU, V2.7 5-46 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-20 DSP Addressing Modes (cont’d) Mnemonic Particularities [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 5-14. 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 5-21: Table 5-21 Coding of the CoREG Addressing Mode Mnemonic Register Coding of wwww:w bits [31:27] MSW MAC-Unit Status Word 00000B MAH MAC-Unit Accumulator High Word 00001B MAS Limited MAC-Unit Accumulator High 00010B Word MAL MAC-Unit Accumulator Low Word 00100B MCW MAC-Unit Control Word 00101B MRW MAC-Unit Repeat Word 00110B The example in Figure 5-14 shows the complex operation of CoXXXM instructions with a parallel move operation based on the descriptions about addressing modes given in Section 5.7.3 (Indirect Addressing Modes) and Section 5.7.4 (DSP Addressing Modes). User’s Manual C166S-V2 CPU, V2.7 5-47 V1.2, 2008-09 XC27x5X 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 Address1 = (IDX0) Long Address2 = (R2) 4) Calculate 24-Bit Physical Address Physical Address1 = Page 3 + Page Offset Physical Address 2 = (DPPi) + Page Offset 5) Post Modify Address Pointer (IDX0) n e w = (IDX0) + 2 (R2) n e w = (R2) + 2 Data Operations 1) Read Operands op1 = (Physical Address 1) op2 = (Physical Address2) 1) Write Operand op1 (Intermediate Address) = op1 (IDX0)n e w (Updated Pointer) op1 Parallel Move (R2)n e w (Updated Pointer) op2 (IDX0) (Read Pointer) (R2) (Read Pointer) Intermediate Address (Write Pointer for Parallel Move) MCA 04928 Figure 5-14 Arithmetic MAC Operations with Parallel Move User’s Manual C166S-V2 CPU, V2.7 5-48 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.5 The System Stack The XC27x5X 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). 5.7.5.1 The Stack Pointer Registers SP and SPSEG Register SPSEG (not bit addressable) 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 bit addressable) 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 SPSEG, as shown in Figure 5-15. The system stack cannot cross a 64-Kbyte segment boundary. SPSEG Stack Pointer Segment 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 5-15 Addressing via the Stack Pointer User’s Manual C166S-V2 CPU, V2.7 5-49 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) SP Stack Pointer 15 14 13 SFR (FE12H/09H) 12 11 10 9 8 7 6 Reset Value: FC00H 5 4 3 2 1 0 SP 0 rwh r Field Bits Type Description SP [15:1] rwh Modifiable part of SP Specifies bits [15:1] of the 16-bit system stack pointer intra segment address 0 0 r Fixed part of SP Read as 0 SPSEG Stack Pointer Segment 15 14 13 12 SFR (FF0CH/86H) 11 10 9 8 7 0 r r r r 6 Reset Value: 0000H 5 4 3 2 1 0 SPSEGNR r r r r rw Field Bits Type Description 0 [15:8] r Reserved Read as 0, should be written 0 SPSEGNR [7:0] rw Stack Pointer Segment Number Specifies the segment where the stack is located. 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 SPSEG or SP stalls the instruction flow until the register is really updated. The instruction immediately following the instruction updating SPSEG or SP can use the new value. User’s Manual C166S-V2 CPU, V2.7 5-50 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.7.5.2 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 C166S-V2 CPU, V2.7 5-51 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) STKOV Stack Overflow Pointer 15 14 13 12 SFR (FE14H/0AH) 11 10 9 8 7 6 Reset Value: FA00H 5 4 3 2 1 STKOV 0 rw r Field Bits Type Description STKOV [15:1] rw Modifiable part of STKOV Specifies the segment offset address of the lower limit of the system stack. 0 0 r Fixed part of STKOV Read as 0 STKUN Stack Underflow Pointer 15 14 13 12 11 SFR (FE16H/0BH) 10 9 8 7 6 Reset Value: FC00H 5 4 3 2 1 0 STKUN 0 rw r Field Bits Type Description STKUN [15:1] rw Modifiable part of STKUN Specifies the segment offset address of the upper limit of the system stack. 0 0 r Fixed part of STKUN Read as 0 User’s Manual C166S-V2 CPU, V2.7 0 5-52 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.8 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 XC27x5X 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 SFR (FF10H/88H) 11 10 HLD IEN EN_ PL1 rw rwh ILVL rwh 8 BANK rwh 7 6 5 USR USR PL0 1 0 rwh rwh rwh 4 3 2 1 0 E Z V C N rwh rwh rwh rwh rwh Field Bits ILVL [15:12] rwh CPU Priority Level 0H Lowest Priority … … FH Highest Priority IEN 11 rw Global Interrupt/PEC Enable Bit 0B Interrupt/PEC requests are disabled 1B Interrupt/PEC requests are enabled HLDEN_PL1 10 rwh Hold Enable/Protection Level selection 1 0B external bus arbitration disabled or protection level 0/1 (refer to Table 5-23) 1B external bus arbitration enabled or protection level 2/3 (refer to Table 5-23) BANK [9:8] rwh Reserved for Register File Bank Selection 00B Global register bank 01B Reserved 10B Local register bank 1 11B Local register bank 2 User’s Manual C166S-V2 CPU, V2.7 Type 9 Reset Value: 0000H Description 5-53 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description USR1 7 rwh General Purpose Flag Can be used by application software. Also set when using repeated MAC instructions (Section 5.9.11) USR0 6 rwh General Purpose Flag Can be used by application software. Also set when using repeated MAC instructions (Section 5.9.11) PL0 5 rwh Protection Level selection 0 Protection level 0/2 (refer to Table 5-23) 0B 1B Protection level 1/3 (refer to Table 5-23) E 4 rwh End of Table Flag 0B Source operand is neither 8000H nor 80H 1B Source operand is 8000H or 80H Z 3 rwh Zero Flag 0B ALU result is not zero 1B ALU result is zero V 2 rwh Overflow Flag 0B No Overflow produced 1B Overflow produced C 1 rwh Carry Flag 0B No carry/borrow bit produced Carry/borrow bit produced 1B N 0 rwh Negative Result 0B ALU result is not negative 1B ALU result is negative ALU/MAC Status (N, C, V, Z, E) 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. User’s Manual C166S-V2 CPU, V2.7 5-54 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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). 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 5-22). User’s Manual C166S-V2 CPU, V2.7 5-55 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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. Table 5-22 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. User’s Manual C166S-V2 CPU, V2.7 5-56 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) HLDEN: Setting this bit for the first time activates the selected bus arbitration mode. Bus arbitration can be disabled by temporarily clearing bit HLDEN. In this case the bus is locked, while the bus arbitration mode remains selected. Please refer to the External Bus Controller (EBC) chapter for functional details. Note that the HLDEN bit can be accessed only when memory protection (MPU) is disabled. Protection Level (PL0, PL1) These flags specify the current protection level of the system. This information is needed for systems implementing memory protection (i.e. MPU). Four different protection levels are defined according to the table below. Refer to the Memory Protection (MPU) chapter for more information on how the protection system works. Table 5-23 Decoding of Protection Level PL1 PL0 0 0 1 1 0 1 0 1 Protection Level Protection Level 0 Protection Level 1 Protection Level 2 Protection Level 3 A write into bit PSW.10 will be interpreted as a write into PL1 when the MPU is enabled or as a write into HLDEN when the MPU is disabled. Considering this fact it is possible to use both the EBC arbitration and MPU functionalities. Note that software made to support the external master functionality, i.e. trying to write into this HLDEN bit, may not have write permission when the MPU is enabled unless it runs in privileged mode. 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. After reset, all interrupts are globally disabled, and the lowest priority (ILVL = 0) is assigned to the initial CPU activity. 5.8.1 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 User’s Manual C166S-V2 CPU, V2.7 5-57 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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. 5.8.2 Bit Manipulation Unit The XC27x5X offers a large number of instructions for bit processing. These instructions are typically used to • • manipulate software bit flags within CPU registers, GPRs or DPRAM control on-chip PD+Bus peripherals and port logic via control bits of their respective bit addressable (E)SFRs. The bit manipulation instructions allow short addressing mode with bitoff operands only (see Chapter 5.7.1). 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. Instructions BSET, BCLR, BAND, BOR, BXOR, BMOV, BMOVN explicitly set or clear specific bits. The bitfield instructions BFLDL and BFLDH allow masked 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 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. Bit protection of PD+Bus register bits All instructions that manipulate single bits or bit groups use a read-modify-write sequence that accesses the whole word containing the specified bit(s). The read-modifywrite approach may be critical with hardware affected bits of type ’rwh’ or ’wh’. In these cases, the hardware may change other bits of the register while the read-modify-write operation is in progress. Thus the writeback could overwrite the new bit value generated by the hardware. To handle this side effect the bit addressable PD+Bus registers support a protection mechanism using a protection mask. Example: BCLR EOPIC.EOPIE ; disable ’end of PEC’ interrupts The instruction will clear the interrupt enable bit EOPIE while the ’rwh’ bit EOPIR will be mask protected. This ensures that an EOP interrupt occuring exactly at the same time will be correctly flagged. Note: For the BFLD(LH) instructions the protection mask must be supplied by the programmer. User’s Manual C166S-V2 CPU, V2.7 5-58 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Note: If a direct conflict occurs between a bit manipulation generated by hardware and an intended software access on the same bit, the software access has priority and determines the final value of the respective bit. User’s Manual C166S-V2 CPU, V2.7 5-59 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.8.3 Multiply and Divide Unit The XC27x5X’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 sub-unit 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 Word 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 C166S-V2 CPU, V2.7 5-60 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) MDL Multiply Divide Low Word 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 15 14 13 12 SFR (FF0EH/87H) 11 10 9 8 7 6 Reset Value: 0000H 5 r r r r r 3 2 MDR IU 0 r 4 r r r r r rwh 1 0 r r 0 r r Field Bits Type Description MDRIU 4 rwh Multiply/Divide Register In Use 0B Cleared when MDL is read via software. 1B Set when MDL or MDH is written via software, or when a multiply or divide instruction is executed. 0 [15:5], [3:0] r Reserved Read as 0, should be written 0 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 C166S-V2 CPU, V2.7 5-61 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.9 DSP Data Processing (MAC Unit) The CoXXX arithmetic instructions are executed by the MAC unit. It provides singleinstruction-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 also shown in Figure 5-16: • • • • • • • • 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 Concatenation Unit 32 16 Repeat Counter MCW Register 16 Signed/ Unsigned Multiplier 32 32 Signed Ext. 40-Bit Adder/Subtracter Round + Saturation 40 40-Bit Signed Accumulator MSW Register 40 ACCU-Shifter 40 40 Limiter MCA 04930 Figure 5-16 Functional MAC Unit Block Diagram 1) The same hardware-multiplier is used in the ALU. User’s Manual C166S-V2 CPU, V2.7 5-62 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.9.1 MAC Unit Control 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 15 14 13 12 SFR (FFDCH/EEH) 11 0 r r r r r 10 9 MP MS rw rw 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 r r r r 0 r r r r r Field Bits Type Description MP 10 rw One-Bit Scaler Control 0B Multiplier product shift disabled 1B Multiplier product shift enabled for signed multiplications MS 9 rw Saturation Control 0B Saturation disabled 1B Saturation to 32-bit value enabled 0 [15:11] r , [8:0] 5.9.2 Reserved Read as 0, should be written 0 Representation of Numbers and Rounding The XC27x5X 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. User’s Manual C166S-V2 CPU, V2.7 5-63 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) The XC27x5X 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). 5.9.3 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. 5.9.4 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. 5.9.5 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. 5.9.6 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 User’s Manual C166S-V2 CPU, V2.7 5-64 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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). 5.9.7 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 5-24: Table 5-24 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. 5.9.8 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 C166S-V2 CPU, V2.7 5-65 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.9.9 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. 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 MAL [15:0] rwh Low Part of Accumulator The 40-bit accumulator is completed by the accumulator high word (MAH) and bitfield MAE User’s Manual C166S-V2 CPU, V2.7 5-66 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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 Field Bits Type Description MAH [15:0] rwh High Part of Accumulator The 40-bit accumulator is completed by the accumulator low word (MAL) and bitfield MAE User’s Manual C166S-V2 CPU, V2.7 5-67 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.9.10 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 0 r 13 8 MV MSL ME MSV MC MZ MN MAE rwh rwh rwh rwh rwh 11 rwh 10 rwh 7 6 Reset Value: 0000H 9 rwh 12 SFR (FFDEH/EFH) 5 4 3 2 1 Field Bits Type Description 0 15 r Reserved Read as 0, should be written 0 MV 14 rwh Overflow Flag 0B No Overflow produced 1B Overflow produced MSL 13 rwh Sticky Limit Flag 0B Result was not saturated 1B Result was saturated ME 12 rwh MAC Extension Flag 0B MAE does not contain significant bits 1B MAE contains significant bits MSV 11 rwh Sticky Overflow Flag 0B No Overflow occurred Overflow occurred 1B MC 10 rwh Carry Flag 0B No carry/borrow produced 1B Carry/borrow produced MZ 9 rwh Zero Flag 0B MAC result is not zero 1B MAC result is zero MN 8 rwh Negative Result 0B MAC result is positive 1B MAC result is negative User’s Manual C166S-V2 CPU, V2.7 5-68 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits Type Description MAE [7:0] rwh MAC Accumulator Extension The most significant bits of the 40-bit accumulator, completing registers MAH and MAL 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. User’s Manual C166S-V2 CPU, V2.7 5-69 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 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. 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. 5.9.11 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 REPEATCOUNT rwh Field Bits Type Description REPEATCOUNT [15:0] rwh MAC repeat 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 5-25 lists the possible encodings. User’s Manual C166S-V2 CPU, V2.7 5-70 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Table 5-25 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 C166S-V2 CPU, V2.7 5-71 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) 5.10 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 15 14 13 SFR (FF1CH/8EH) 12 11 10 9 8 7 Reset Value: 0000H 6 5 4 3 2 1 0 r r r r r r r ZERO r r r r r r r r Field Bits Type Description ZERO [15:0] r Constant Zero Bits ONES Ones Register 15 14 13 r SFR (FF1EH/8FH) 12 11 10 9 8 7 6 Reset Value: FFFFH 5 4 3 2 1 0 ONE r Field Bits Type Description ONE [15:0] r Constant One Bits CPUID CPU Identification Register 15 14 13 12 11 ESFR (F00CH/06H) 10 9 8 7 6 MODULENUMBER r r r r r 5 4 3 2 1 0 r r VERSIONNUMBER r r r r Field Bits Type Description MODULENUMBER [15:8] r User’s Manual C166S-V2 CPU, V2.7 Reset Value: 0313H r r r r r C166 Family CPU Module Number (C166S-V2) 5-72 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) Field Bits VERSIONNUMBER [7:0] User’s Manual C166S-V2 CPU, V2.7 Type Description r C166S-V2 CPU Version Number 5-73 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Central Processing Unit (CPU) User’s Manual C166S-V2 CPU, V2.7 5-74 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) 6 Memory Protection Unit (MPU) The Memory Protection Unit (MPU) provides the hardware mechanisms needed for implementing memory protection. The MPU allows detection of unauthorized accesses (read, write or instruction fetch) in user-defined memory ranges. It offers protection for the complete address space, including the peripheral area. The MPU can be used to support the encapsulation of different applications or software components running on the processor. This encapsulation provides the means to ensure integrity and fault isolation capabilities in today’s complex systems relying on multiplesources software. 6.1 Functional Overview Different protection levels are usually needed to support a programming system where for example an operating system or software kernel runs and controls different application and low level drivers parts. One level can be associated to the operating system and for the other tasks that need protection against each other or against the operating system, other levels can be used. For every protection level different address ranges with different access permissions for instructions and/or data can be defined. When a piece of code is executed and the memory protection is enabled, the permissions associated to its protection level are selected and every time a memory access is performed it will be checked if the access is outside of the specified ranges or violates the access permissions. In this case the access may not be performed but marked as invalid and a protection trap routine can be executed. The basic MPU functionality is shown in Figure 6-1. PSW PL1/0 Level n r/w/e Range z uppbnd lowbnd r/w/e … Range y uppbnd lowbnd Memory access (r/w/e) Within range? no Protection Violation r/w/e … Range x uppbnd lowbnd MPU_RUN Figure 6-1 User’s Manual MPU, V1.7 MPU operation 6-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) Four Protection Levels can coexist during run time in this architecture. Two bits in the Processor Status Word (PSW) are used to select which protection level is active at a given time. If an application requires more than 4 protection levels, a re-mapping of all the levels to the 4 possible values has to be performed and during run time reprogramming of the protection register sets when switching levels is needed. A protection register set is associated to every protection level, every set contains all the address ranges and the access permissions associated to the corresponding protection level. Every protection register set can contain a programmable number of range registers. All together, a maximum of 12 ranges is supported. Associated to every code or data range, a protection mode register defines the permissions for this range. Refer to the next chapters for a detailed explanation of the MPU registers needed for the protection system and its usage. User’s Manual MPU, V1.7 6-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) 6.2 Memory Protection Registers A protection register set consists of a variable number of Protection Range register pairs (PRUx/PRLx) and the corresponding number of Protection Mode registers (PMx). The PMx registers are located in the SFRs area and are accessed through the Peripheral Data Bus -PD-Bus-. The PRUx/PRLx registers are not memory mapped, their access mechanism is supported through the memory mapped registers Protection Range Address register (PRA) and Protection Range Data register (PRD). Table 6-1 Registers Address Space Module Base Address MPU 0H Table 6-2 End Address Note Registers Overview Register Short Name Register Long Name Offset Page Address Number PRUx (x =0-11) Protection Range Register x Upper Bound noneH 6-4 PRLx (x =0-11) Protection Range Register x Lower Bound noneH 6-4 PM5 Protection Mode Register 5 FFD2H 6-7 PM4 Protection Mode Register 4 FFD0H 6-7 PM3 Protection Mode Register 3 FFCEH 6-7 PM2 Protection Mode Register 2 FFCCH 6-7 PM1 Protection Mode Register 1 FFCAH 6-7 PM0 Protection Mode Register 0 FFC8H 6-7 PRD Protection Range Data FFC6H 6-8 PRA Protection Range Address FFC4H 6-9 6.2.1 Protection Range Registers The PRUx/PRLx pairs are 16-bits registers and specify the upper 16 bits of the physical addresses, upper and lower bound, for data and/or code for all the allowed ranges (12 is the maximum supported). Only these upper 16 bits of the physical addresses are considered in the address comparisons, as a consequence, the minimum granularity of the ranges is 256 bytes and all the ranges are aligned to this size. The PRUx and PRLx registers specify respectively the Upper and Lower addresses of a Range. If due to a programming error PRLx specifies a value bigger than PRUx, the User’s Manual MPU, V1.7 6-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) corresponding range will not specify a correct address range, and as a consequence the corresponding range is useless (i.e. ignored). For programming a protection range in the PRUx/PRLx registers, it has to be selected first which range is going to be written by programming the address into PRA, then the data write operation can be performed by writing the data into PRD. In a similar way, a read operation has to be performed by selecting first which range in going to be read (by programming the address into PRA) and then the read operation can be performed by reading PRD. Programming a PRUx/PRLx register requires then two write operations. Similarly, reading a PRUx/PRLx register requires also two operations (one write and one read). For continuous accesses and when using the auto increment feature only one initialization into PRA is needed, afterwards only the PRD register needs to be written/read every time. Registers PRD and PRA are described in Chapter 6.2.3 and Chapter 6.2.4 respectively. PRUx (x =0-11) Protection Range Register x Upper Bound 15 14 13 12 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 UPPBDN Field Bits Type Description UPPBND [15:0] rw Upper Boundary Address (upper 16 bits) PRLx (x =0-11) Protection Range Register x Lower Bound 15 14 13 12 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 LOWBND Field Bits Type Description LOWBND [15:0] rw User’s Manual MPU, V1.7 Lower Boundary Address (upper 16 bits) 6-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) 6.2.2 Protection Mode Registers All the control information associated to every address range is contained in the Protection Mode registers. Access permissions (execute, read and/or write) are defined here and also the range-to-level mapping. Every range can be individually enabled to be used for any protection level, even can be used for more than one level (but with the same access permissions). Also the field used the enable the protection system is implemented in one of the protection mode registers. Note that no hardware mechanism is implemented to flush the pipeline upon a modification of these registers. This is usually not a problem because a (re-) programming of the MPU configuration registers should be anyhow performed having the protection disabled. Also the configuration affecting a particular protection level will be usually (re-)programmed from another level meaning that even at the point when protection is enabled the software currently running will not be affected by the configuration change (the configuration change is usually seen once the protection level is changed according to the procedure described in Chapter 6.4.2). For special cases where the change will and needs to be immediately seen, the software has to take care that the write is effective before executing the next affected instruction (by reading for example the latest written register). The bit fields of the PMx registers in the description below use generic Range names (A, B), their mapping to the physical ranges is given after the PMx register name where they belong to. Given a Protection Mode register x, the range named A is addressing the physical range 2*x and range named B the range 2*x +1. The PMx registers are EINIT protected. PM0 Protection Mode Register 0 15 14 13 12 11 10 SFR (FFC8H) 9 L3E L2E L1E L0E WEB REB XEB B B B B rw rw rw rw rw rw rw 8 0 7 6 5 4 3 2 1 0 L3E L2E L1E L0E PRO WEA REA XEA A A A A TEN r rw Field Bits Type Description PROTEN 0 rw User’s Manual MPU, V1.7 Reset Value: 0000H rw rw rw rw rw rw rw Protection Enable bit This bit enables the Protection mechanism 0B Protection not enabled 1B Protection enabled 6-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) Field Bits Type Description XEA, XEB 1, 9 rw Execute Enable 0B Instruction fetch accesses to associated address range (A, B) not permitted 1B Instruction fetch accesses to associated address range (A, B) permitted REA, REB 2, 10 rw Read Enable 0B Data read accesses to associated address range (A, B) not permitted 1B Data read accesses to associated address range (A, B) permitted WEA, WEB 3, 11 rw Write Enable Data write accesses to associated address 0B range (A, B) not permitted 1B Data write accesses to associated address range (A, B) permitted L0EA, L0EB 4, 12 rw Level 0 Enable 0B Range (A, B) not enabled for Protection Level 0 1B Range (A, B) enabled for Protection Level 0 L1EA, L1EB 5, 13 rw Level 1 Enable 0B Range (A, B) not enabled for Protection Level 1 1B Range (A, B) enabled for Protection Level 1 L2EA, L2EB 6, 14 rw Level 2 Enable 0B Range (A, B) not enabled for Protection Level 2 1B Range (A, B) enabled for Protection Level 2 L3EA, L3EB 7, 15 rw Level 3 Enable 0B Range (A, B) not enabled for Protection Level 3 1B Range (A, B) enabled for Protection Level 3 0 8 r Reserved field The field PROTEN exists only in the Protection Mode Register 0. User’s Manual MPU, V1.7 6-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) PMx (x =1-5) Protection Mode Register x 15 14 13 12 11 10 SFR (FFC8H+2*x) 9 L3E L2E L1E L0E WEB REB XEB B B B B rw rw rw rw rw rw rw 8 0 7 6 Reset Value: 0000H 5 4 3 2 1 L3E L2E L1E L0E WEA REA XEA A A A A r rw rw rw rw rw rw rw 0 0 r Field Bits Type Description 0 0 r Reserved field XEA, XEB 1, 9 rw Execute Enable 0B Instruction fetch accesses to associated address range (A, B) not permitted 1B Instruction fetch accesses to associated address range (A, B) permitted REA, REB 2, 10 rw Read Enable 0B Data read accesses to associated address range (A, B) not permitted 1B Data read accesses to associated address range (A, B) permitted WEA, WEB 3, 11 rw Write Enable 0B Data write accesses to associated address range (A, B) not permitted 1B Data write accesses to associated address range (A, B) permitted L0EA, L0EB 4, 12 rw Level 0 Enable 0B Range (A, B) not enabled for Protection Level 0 1B Range (A, B) enabled for Protection Level 0 L1EA, L1EB 5, 13 rw Level 1 Enable 0B Range (A, B) not enabled for Protection Level 1 1B Range (A, B) enabled for Protection Level 1 L2EA, L2EB 6, 14 rw Level 2 Enable 0B Range (A, B) not enabled for Protection Level 2 1B Range (A, B) enabled for Protection Level 2 User’s Manual MPU, V1.7 6-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) Field Bits Type Description L3EA, L3EB 7, 15 rw Level 3 Enable 0B Range (A, B) not enabled for Protection Level 3 1B Range (A, B) enabled for Protection Level 3 0 8 r Reserved field 6.2.3 Protection Range Data Register The Protection Range Data register contains the 16 bits data value needed to program the content of the Protection Range Registers. It also contains the data read during the last read access on the Protection Range Registers. A write into PRD triggers immediately a write into the corresponding PRUx/PRLx register (the one that is currently selected by the write pointer -in PRA register-). Also a read into PRD delivers the corresponding PRUx/PRLx data immediately (the one that is currently selected by the read pointer -in PRA register-). The PRD register is EINIT protected. PRD Protection Range Data 15 14 13 12 11 SFR (FFC6H) 10 9 8 7 Reset Value: 0000H 6 5 4 3 2 1 0 DATA rwh Field Bits Type Description DATA [15:0] rwh 6.2.4 Data Value for/from PRUx/PRLx Protection Range Address Register The Protection Range Address register contains two access pointers, one used for write operations and the other for read operations. With every 5-bit pointer it is possible to select a PRUx/PRLx register from a set of 24 register (the 24 PRUx/PRLx registers needed to implement 12 protection ranges). An auto increment capability can be enabled for the access pointers (controlled by WMOD and RMOD fields), after every write or read into/from PRD the write or read pointers are incremented respectively. This feature enables a faster programming of the protection range registers. When the auto increment mode is active, the access pointers automatically do a wrap around (i.e. initialized to 0) after reaching its maximum value. User’s Manual MPU, V1.7 6-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) The occurrence of a wrap around is shown in the status bits WWA or RWA. The software can then check if this situation has happened taking the corresponding action and resetting the corresponding flag. Special care has to be taken when programming the PRA register in order not to modify one of the pointers unintentionally. It is recommended to use bit instructions for that (bit field instructions for example). Also when using the auto increment feature and during debugging it has to be considered that a debugger access can also modify the pointer values, the debugger software should then take care of restoring the original status of this register. The PRA register is EINIT protected. PRA Protection Range Address 15 14 13 RMO RWA D rw rwh 12 11 SFR (FFC4H) 10 0 RPTR r rwh 9 8 7 Reset Value: 0000H 6 WM WW OD A rw rwh 5 4 3 2 1 0 WPTR r rwh 0 Field Bits Type Description WPTR [4:0] rwh Write Pointer Selects the Protection Range Register to be written 00000B Selects PRL0 00001B Selects PRU0 00010B Selects PRL1 00011B Selects PRU1 00100B Selects PRL2 00101B Selects PRU2 ... 11110B Selects PRL15 11111B Selects PRU15 0 5 r Reserved field WWA 6 rwh Write Wrap Around Status 0B No WPTR Wrap Around occurred on last Write A WPTR Wrap Around occurred on last Write 1B Bit to be cleared by SW WMOD 7 rw Auto increment Write Mode 0B No increment WPTR on every Write 1B Auto increment WPTR on every Write User’s Manual MPU, V1.7 6-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) Field Bits Type Description RPTR [12:8] rwh Read Pointer Selects the Protection Range Register to be read 00000B Selects PRL0 00001B Selects PRU0 00010B Selects PRL1 00011B Selects PRU1 00100B Selects PRL2 00101B Selects PRU2 ... 11110B Selects PRL15 11111B Selects PRU15 0 13 r Reserved field RWA 14 rwh Read Wrap Around Status 0B No RPTR Wrap Around occurred on last Read 1B A RPTR Wrap Around occurred on last Read Bit to be cleared by SW RMOD 15 rw Auto increment Read Mode 0B No increment RPTR on every Read 1B Auto increment RPTR on every Read 6.3 Functional Description 6.3.1 Enabling Protection Protection has to be globally enabled per software, bit PM0.PROTEN implements this functionality, refer to chapter Chapter 6.2.2. 6.3.2 Protection Levels The bits PSW.PL1/0 select the current protection level, i.e the protection register set currently active. The decoding of PL1/0 is as follows: Table 6-3 Decoding of Protection Level PL1 PL0 0 0 1 1 0 1 0 1 User’s Manual MPU, V1.7 Protection Level Protection Level 0 Protection Level 1 Protection Level 2 Protection Level 3 6-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) PL1 and PL0 bits are mapped into PSW.10 and PSW.5 bits respectively. Note that due to the shared functionality implemented on the bit PSW.10, a write on this bit will be interpreted as a write on the PSW.PL1 only when the MPU is itself enabled (PM0.PROTEN is 1). When the MPU is not enabled a write on this bit will be interpreted as a write on the HLDEN flag. For consistency, the flag PSW.PL0 is handled in a similar way, a write on PSW.PL0 is only effective when the MPU is enabled. 6.3.2.1 Protection Level 0 For the protection mechanism to work properly, the MPU has to be operated under a kind of privileged mode, programming and changing the protection information should only be allowed during this mode. Even if the C166 family architecture does not support directly this operation mode (only the one associated to the initialization phase ended by the EINIT execution), the privileged mode can be defined in this context as the mode entered when the processor runs with protection Level 0. This is the level entered after reset and the level automatically entered after an interrupt/trap is taken. Level 0 should be then the level used by the operating system, software kernel or the software components needing access to the whole system resources (specially to system control registers and peripheral area). But note that defining and programming address ranges and permissions is still needed for Level 0 (even if it is the whole space). Per default (i.e. after reset), no access in any address range is allowed, also not for this level. Note: The need to program Level 0 allows in special occasions to give restricted access also to this level. Restriction sometimes needed to probe reliability of the software running under this level. 6.3.3 Intersecting Memory Ranges The permission to access a memory location is the OR of the memory range permissions. When two or more ranges intersect, the intersecting region has the permission of the most permissive range. 6.3.4 Protection of the MPU registers As mentioned in Chapter 6.3.2.1, the MPU registers need to be protected. A protection mechanism comes automatically with the use of the MPU and the fact that the whole address space, including SFRs, is under control of the MPU, see also Chapter 6.3.5. Once protection is enabled, changing protection information can then only be performed from a protection level which has access to the corresponding SFR area (i.e. to the protection registers). In addition to this inherited protection mechanism, the protection control registers are also EINIT protected. The EINIT protection creates some overhead during dynamic reprogramming, however it adds an additional protection level that may be needed in case User’s Manual MPU, V1.7 6-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) different software component need to be executed at the same protection level (the one having access to these control registers -usually level 0-). 6.3.5 Accessing SFRs and GPRs Once the protection system is activated, a task is not free anymore to access per software any special function register (SFR) unless this is explicitly covered by the address ranges and permissions assigned to this task. This applies to the internal IO area (SFR, ESFR, XSFR) and also to the external IO area (on chip LXBus peripherals or external peripherals). Since the minimum granularity of the address ranges is 256 bytes, the IO space is partitioned into blocks. A task will have access either to one of these blocks with its full set of registers or to none. For example the SFR/ESFR area (1 Kbyte), is divided into four blocks (F000h...F0FFh, F100h...F1FFh, FE00h...FEFFh, FF00h...FFFFh). For the XSFRs area, 4 Kbyte, the space is divided in 16 blocks. CSFRs are also handled by the protection scheme, but exceptions are required for those CSFRs that are user registers. CSFR that are kept under the protection scheme are: • PSW (partly), CPUCON1/2, CP, CSP, SP, SPSEG, STKUN, STKOV, TFR, VECSEG. The USR0/1 bits of PSW, that are user bits, are excluded from the protection scheme. Also the PSW condition flags are excluded. Instructions like JBC/JNBS on the PSW conditions flags can then still be used in user mode. Read accesses to all the PSW fields are allowed. CSFR that are excluded from the protection mechanism are: • DPP0/1/2/3, MDL, MDH, MSW, MDC, MAH, MAL, MRW, MCW, QR0/1, QX0/1, IDX0/1, ZEROS, ONES, CPUID. The DPP registers are handled as user registers to support its re-programming during run time (practice needed for code optimization purposes). When used in this way, it will be responsibility of the software to ensure their right handling, for example saving and restoring them in task switches. CPUID is not strictly a user register, however it is not required to define it as protected since it is anyhow not writable. GPRs are excluded from the memory protection mechanism. Protection on the DPRAM is however guaranteed since the CP itself is protected. Similarly, GPRs mapped into the Local Register Banks are excluded from the protection mechanism. 6.3.6 Interrupts and PECs Handling Any interrupt taken by the CPU will switch automatically the protection level to 0. This is valid for peripheral interrupts, debugger interrupts, hardware and software traps. As a consequence Interrupt Service Routines (ISRs) are always started with protection level 0, having usually access to all the system resources. The ISR itself can afterwards reduce the protection level and execute user code with protection restrictions. User’s Manual MPU, V1.7 6-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) Interrupt requests can be also serviced through PEC transfers, that is, fast data transfers between two memory locations. PEC transfers will be executed by the CPU without protection. Protection can still be ensured through the programming of the PEC control registers that should be only performed under the right protection level, usually in privileged mode (i.e. protection level 0). At the configuration time the software should then check for the correctness of the PEC source and destination pointers (according to the permissions allowed) and the PEC control register. Special care has to be taken when using continuous mode, in this case the software can not take care at the configuration time if the PEC will not violate an area in the future. Additional run time checks may be needed to support this mode (executed by the privileged software) or this mode will have to be avoided. 6.3.7 Special handling of RETI instruction The PSW and specially the Protection Level selection flags (PSW.PL0/1) are handled under the protection scheme: explicit writes on the PSW are detected by the hardware and checked if they are triggered under the right protection level. In case the access is not allowed, a trap will be generated and the modification of the PSW will be avoided by the protection logic. But the PSW can also be modified implicitly by the hardware and this hardware update can hardly be managed by the protection logic. Hardware updates on the PSW.PL0/1 field are triggered by the execution of a RETI instruction. These PSW hardware updates are in principle not critical as long as the PSW (and PSW.PL0/1) value that is taken from the Stack has not been manipulated by any user code. But since there is no possibility to prohibit user code from this possible manipulations (user code may make use of local stacks with write access to it) the only work around is to prohibit un-trusted user code from using the RETI instruction. RETI will be then specially handled as a kind of protected instruction that can only be executed when the protection level 0. This handling is consistent with the fact that interrupts are handled under protection level 0, returning from interrupts should then also be performed under the same level. 6.3.8 Context Switch operations The Context Switch mechanism is executed in the core with the help of internal instructions that are auto-injected in the pipeline. Usually auto-injected instructions should run with the same protection level as the instruction causing the auto injection. However, due to the fact that the context switch is an interruptible operation and its completion may be delayed in certain situations, these context switch auto injected instructions have to be executed without considering protection. That means, while they are executed, the protection checks are not performed and/or are ignored. As a consequence only the CP update operation, that is not performed by the auto-injected instruction but by the context switch instruction itself, is performed under the protection scheme. The saving procedure of registers into DPRAM or the read of GPRs from User’s Manual MPU, V1.7 6-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) DPRAM is not performed under the protection scheme. DPRAM protection in this case will have to be ensured, in case it is needed, by the software, the software can check if the region addressed by the values programmed into CP are allowed. 6.3.9 Debugger Access Permissions The debugger must be able to access all the memory space even if memory protection is active, this includes also the IO space (i.e. SFRs). The OCDS/Cerberus implements basically 2 mechanisms for accessing the system resources: • • triggering the CPU to execute a Monitor Routine that contains the code to access the resources (Call a Monitor) Injecting any instruction that can by itself access any resource When using the first mechanism, that is started by the injection of an ITRAP instruction, the debugger will automatically run in privileged mode, i.e. with protection level 0. As defined in Chapter 6.3.2.1, this is the level automatically entered after an interrupt (in this case after the injection of the debug TRAP instruction). When using the second mechanism (also if the CPU is halted) the injected instruction will run without protection. The CPU keeps track of the fact that an instruction was injected by the Debugger and disables the protection check for that instruction. With respect to accessing OCDS/Cerberus/MCDS SFRs by the debugger it just needs to be ensured that the debug monitor (used to program the debug logic) can access these registers with minimum overhead and without any impact on the user code. Since the debug monitor routine will always be executed with protection level 0, it is expected that all the memory space is then allowed. Also accessing these registers via injecting instructions can be performed without restrictions as explained above. 6.3.10 Invalid Access Traps If an access is performed in a protected area an invalid access trap will be generated. Three traps are defined for this purpose: MPR Memory Protection, Read MPW Memory Protection, Write MPX Memory Protection, Execute They are defined as Class B traps. They are mapped to TFR.10,9,8 respectively (MPR is TFR.10, MPW is TFR.9 and MPX is TFR.8). Refer to the Hardware Traps description chapter for the complete description of the TFR register. Note that no trap must be performed on accesses that are performed speculatively, this is why these traps can just be generated when it is known that the instruction is not cancelled anymore (this is, when the instruction goes into the Execute stage). User’s Manual MPU, V1.7 6-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) The already existing trap PRTFLT Protection Fault Trap is also used to indicate the execution of a RETI instruction from a protection level different to 0. Even when RETI causes a protection fault trap, it is normally executed. 6.3.10.1 Cancelling operations Instructions causing a protection violation will be detected by the MPU but its full execution can not be suppressed, only the writes operations causing a protection fault or derived from an instruction causing a protection fault will be cancelled. Read operations can not be cancelled since they are triggered very soon in the pipeline (sometimes speculatively), however, the read data will not be written by the corresponding instruction in any memory mapped address. There are some exceptions to the above general rule of cancelling writes operations, in particular, for instructions performing 2 write operations sometimes the first write can not be cancelled. These are the concrete cases: • • SCXT instruction. The write into the Stack can not be avoided when a Read protection violation on the mem operand is detected (for SCXT reg, mem) or a Write protection violation on the reg operand. CALLS, PCALL instruction. The first write into the Stack (CSP, or reg in case of PCALL) can not be avoided when a Write protection violation on the second Stack address is detected (where the IP should be pushed). This situation assumes that the Stack has grown over the limit of an allowed area right while executing this instruction (first stack push in an allowed area, second stack push in a non-allowed area). As a consequence of the fact that Read operations can not be cancelled, destructive reads on the IO space can be still performed even if the MPU detects a protection violation. As a consequence of the fact that Execute operations can not be cancelled, system instructions and their corresponding actions may be still executed even if they trigger an Execute protection violation. For example an IDLE instruction may still put the CPU in idle mode before the corresponding hardware trap routine can be executed (once the idle mode is left). 6.4 Initializing and using the MPU 6.4.1 Installing Protection This chapter describes briefly the SW sequences needed for initializing and using the protection system. It also analyses the overhead created (real time performance). The implementation with 12 ranges is analyzed. The initialization sequence that can be used for installing protection is: User’s Manual MPU, V1.7 6-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) • • • • Disable Protection in case it is not (after reset protection is disabled), 1 write into PM0. Program the Range Registers, 1 PRA write, 24 writes into PRD (absolute maximum value, assumes that all ranges are used). Program Protection Mode Registers, 6 PMx writes. Enable protection, 1 write into PM0. This last write can be performed together with the write into the PM0 above, but in this case care should be taken to write this register at the end. When the applications or software components using different protection levels can exactly be mapped to the protection sets implemented, this code sequence would set up the system and no additional overhead when using the MPU would exist during run time. After the initialization phase, whenever a change in the protection level is needed, the corresponding protection set has to be selected (changing PSW.PL). This is the only additional operation during run time. This initialization, and in general any change in the protection registers, should be performed always having the protection disabled and usually will be executed from. protection level 0. Note that in cases where the protection configuration and its activation needs to be immediately seen, the software has to take care that the latest write activating the protection is effective before executing the next affected instruction (by reading for example the latest written register). This is because as explained in Chapter 6.2.2, there is no hardware mechanism to flush the pipeline when the protection is activated. In case the protection needs can not be mapped into the implemented protection sets, some re-programming during run time is needed. The worst case scenario is that the ranges have to be re-programmed, then a sequence similar to the one during the initialization is needed. However it may be that only some already defined Ranges needs to be activated/deactivated, in this case only the Protection Mode Registers will need to be re-programmed. This assumes that at every moment it is known which ranges are used, so the PSW.PL1/0 has to be read before deciding what to change. An additional overhead during reprogramming is coming from the EINIT protection. After EINIT execution, reprogramming of the PMU registers is only allowed by releasing temporarily this protection by going to an unprotected mode (a command sequence of 4 write instructions with the use of a password is needed for that). After the re-programming the EINIT-protection has to be of course restored (again a command sequence of 4 write instructions). Note that reprogramming of protection registers that are currently not active (i.e. selected through PSW.PL) and do not become active through the reprogramming, is still possible without having to disable protection. 6.4.2 Changing Protection Level Special care has to be taken when changing the protection level by writing explicitly into PSW. This is because any write into PSW takes effect immediately. If the privileged code User’s Manual MPU, V1.7 6-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) handling protection would write into the PSW before performing a code/task switch, the level of the privileged code would be itself changed and eventually the code/task switch (function call for example) couldn’t be performed. For avoiding this immediate effect when writing into the PSW some tricks have to be used: stack manipulation and calling functions using the RETI/RETP instructions. The value of the new PSW with the new protection level has to be then stored in the stack, instead of writing the PSW explicitly. The RET instruction will then update the PSW associated to the new task with the correct protection level and at the right time. 6.4.3 Executing privileged code from non-privileged one It is possible for a non-privileged (un-trusted) software to invoke a privileged software component (trusted). The non-privileged part has to give control to a privileged part, this can be performed by executing a software TRAP instruction. Automatically this instruction will change the protection level to 0. This mechanism allows for example invoking low level drivers from an application software and also returning from the user part of an ISR to the ISR itself, i.e. to the part handle by the OS or software kernel. 6.4.4 Fast task switches Any task switch that is not controlled by the OS or software kernel will be handled as a trusted task with respect to the software that is invoking it. This is because if the switch is not performed under a software running with enough protection level (usually level 0), there is no possibility to change the protection level explicitly for this task. 6.4.5 Register Bank Selection Since PSW.BANK field is now handled within the protection scheme, the register bank selection (global or local 1/2) will have to be settled by the software running with enough access permissions, i.e. enough protection level (usually level 0). The bank selection should then be done in the part of the ISR running on level 0 (starting level of all ISRs). 6.4.6 Debugger Use Cases This chapter documents how to debug the system when memory protection is in use. The following 4 use cases are identified: • • • • user wants to find the reason for a MPU trap user wants to debug the MPU trap routine user doesn't want to debug, he just wants to poll a variable with Cerberus user wants to debug without any irregular influence from the protection system For the first use case, the standard debugging resources (OCDS) can be used for setting an IP breakpoint on the ISR/s handling the trap. Once there, it will be known which access type causes the trap, either implicitly because there are different trap routines User’s Manual MPU, V1.7 6-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) depending on the exception type or explicitly by reading the TFR flags. Also the protection level causing the violation can be obtained by reading the stacked PSW. With respect to the IP causing the trap, there is no direct access to it but to the linear following one, that is also stored in the stack. The second use case, i.e. the debugging of the MPU trap routine, can be done similarly to the debugging of any trap routine and will be started probably also by setting an IP breakpoint on the corresponding ISR. With respect the third use case, the variable polling action can be performed at any time independently of if the MPU is enabled or not. The debugger has always access to all the system resources even if the MPU is enabled. The fourth use case, i.e. debugging without influence from the PMU, can be covered by disabling explicitly the MPU (PM0.PROTEN) via the debugger. Since the debugger does not know when the application will enable the MPU after the reset of the system, the debugger will have to monitor the status of the MPU (PMU0.PROTEN) and re-disable it once enabled. This use case is however rather strange because debugging is intended to be done with the real system behavior, if an application causes a MPU exception, this should also be seen during debugging. User’s Manual MPU, V1.7 6-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Memory Protection Unit (MPU) User’s Manual MPU, V1.7 6-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7 Interrupt and Exception Control The architecture of the XC27x5X supports several mechanisms for fast and flexible response to service requests from various sources internal or external to the micro controller. 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 the current program execution and branches to an interrupt service routine in order to service an interrupt requesting device. As a result, the current program status (IP, PSW, and, in segmentation mode, also CSP) is saved on the system stack. A prioritization scheme with sixteen priority levels specifies the execution order of multiple interrupt requests. PEC Interrupt Processing A faster alternative to normal interrupt processing is the use of the XC27x5X’s integrated Peripheral Event Controller (PEC) to service an interrupt requesting device. Triggered by an interrupt request, the PEC performs a single word or byte data transfer between any two memory locations. During a PEC transfer, the normal program execution of the CPU is interrupted only for the data transfer. 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 External Service Request pins ESRx (e.g. used to implement NMI like behavior). Several hardware trap functions are provided for handling erroneous conditions and exceptions that arise during the program 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 The XC27x5X does not provide dedicated external interrupt input pins but rather allows to configure a subset of its input pins as interrupt inputs. Interrupt (trap) input pins can be chosen from standard inputs or External Service Request pins ESRx. The available options are detailed in the External Interrupts section. User’s Manual Interrupts, V3.2 7-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Interrupt Sources and Routing To activate and correctly route an interrupt source programming of the following on-chip components must be considered: • • • • Interrupt control of each peripheral IMB memory controller Interrupt Generation SCU External Request Unit (ERU) SCU Interrupt Generation Additionally the port programming must be considered if external interrupt sources are to be used. User’s Manual Interrupts, V3.2 7-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.1 Interrupt System Structure The XC27x5X provides 96 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 respective peripheral control registers. Additional sharing of interrupt nodes is supported via Interrupt Node Sharing. The XC27x5X 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. 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 XC27x5X’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 double word 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 Interrupts, V3.2 7-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Interrupt and Peripheral Event Controller PEC Pointer per_irq0 PECSEG0 DSTP1 PECSEG1 SRCP7 DSTP7 PECSEG7 irq 1 per_irq2 per_irq3 per_irq n-1 DSTP0 SRCP1 irq 0 per_irq1 per_irq n-2 SRCP0 C166SV2 CPU irq 2 Interrupt Source Selection (ISSR) irq 3 Arbitration PEC Request Arbitr. Winner irq n-1 EOP irq n irq x INT per_irq m Peripheral Event Controller (PEC) 1) Request Request Control Control Interrupt Handler Injection Control (CPU Action Request) Interrupt Interrupt Request Request Injection Interface per_irq n per_irq n+1 OCDS Arbitration Control PEC Control (Interrupt Control Registers) (PEC Control Registers) irq0IC PECC0 irq1IC PECC1 irqnIC PECC7 EOPIC PECISNC Interrupt Handler Control OCDS Injection Request & Control Fast Bank Switching BNKSEL0 BNKSEL3 Interrupt Jump Table Cache FINT0CSP FINT0ADDR FINT1CSP FINT1ADDR 1) End of PEC Interrupt (EOPINT) is connected to an interrupt request line . Figure 7-1 User’s Manual Interrupts, V3.2 Block Diagram of the Interrupt and PEC Controller 7-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.2 Interrupt Arbitration The XC27x5X interrupt arbitration system can handle interrupt requests from up to 96 sources. Interrupt requests may be triggered either by the internal peripherals or by external inputs. The “End of PEC” interrupt for supporting enhanced PEC functionality is connected internally to one of the interrupt request lines. The arbitration process starts with an enabled interrupt request and stays active for as long as an enabled interrupt request is pending. Each interrupt request line is controlled by its interrupt control register xxIC (here and below xx stands for the mnemonic of the respective interrupt source). An interrupt request event sets the interrupt request flag in the corresponding interrupt control register (bit xxIC.IR). The interrupt request can also be triggered by the software if the program sets the respective interrupt request bit. If the request bit has been set and this interrupt request is enabled by the interrupt enable bit of the same control register (bit xxIC.IE), an arbitration cycle starts with the next clock cycle. However, if an arbitration cycle is currently in progress, the new interrupt request will be delayed until the next arbitration cycle. If an interrupt request (or PEC request) is accepted by the core, the respective interrupt request flag is cleared automatically. All interrupt requests pending at the beginning of a new arbitration cycle are considered simultaneous. Within the arbitration cycle, the arbitration is independent of the actual request time. The XC27x5X uses a three-stage interrupt prioritization scheme for interrupt arbitration as shown in Figure 7-2. User’s Manual Interrupts, V3.2 7-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Interrupt Arbitration 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) xxxxx (request priority level) CPU Action Control CPU Arbitration 0xxxx (ILVL. PSW extended with 0 in MSB) PSW CPU 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 compared with 5-bit OCDS priority level Stage 3: 5-bit request priority level compared with 4-bit PSW priority level *): OCE requests are identically serviced as OCDS requests Figure 7-2 Interrupt Arbitration The first arbitration stage compares the priority levels of interrupt request lines. The priority level of each requestor consists of interrupt priority level and group priority level. An interrupt priority level is programmed for each interrupt request line by the 4-bit bitfield ILVL of respective xxIC register. The group priority level is programmed for each interrupt request line by the 2-bit bitfield GLVL and the extension bit GPX of the register xxIC. Both together, GPX and GLVL form the 3-bit (extended) group priority level XGLVL, controlling up to eight interrupt sub-priorities within one of the 16 interrupt levels. 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 may 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 disregards the group priority level of interrupt/PEC requests and deals only with interrupt priority levels (ILVL). To compare with OCDS service request priority programmed by 5-bit value, the 4-bit ILVL of the interrupt/PEC request is extended to a 5-bit value with MSB equal to 0. This means that any OCDS request with MSB=1 will always win the second stage arbitration. However, if there is an OCDS request with MSB=0 conflicting with the same priority interrupt/PEC request, the latter is sent to the CPU. User’s Manual Interrupts, V3.2 7-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control On the third arbitration stage, the priority level of the second stage winner is compared with the priority of the current CPU task. An action request will be accepted by the CPU if the requesting source has a higher priority level than the current CPU priority level (bits ILVL of the PSW register) and interrupts are globally enabled by the global interrupt enable flag IEN in PSW. The CPU denies all requests in case of a cleared IEN flag. To compare with the 5-bit priority level of the second stage winner, the 4-bit ILVL.PSW is extended to a 5-bit value with MSB equal to 0. This means that any request with MSB=1 will always win the arbitration against any CPU level. If the requester has a lower or equal priority level than current CPU task, the request stays 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 enabled interrupt request (including a denied interrupt request and a priority level 0000B request) triggers a CPU wakeup from idle state independent of setting the global interrupt enable bit PSW.IEN. User’s Manual Interrupts, V3.2 7-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.3 Interrupt Control All interrupt control registers are organized identically. The lower nine 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 seven bits of the respective register are reserved. All interrupt control registers are bit addressable and all control bits can be read or written via software. Therefore, each interrupt source can be programmed or modified with just one instruction. In the case of reading the interrupt control registers with instructions that operate with word data types, the upper 7 bits (15...9) will return zeros. It is recommended to always write zeros to these bit positions. The IR bit of any IC register is of type “rwh” and is set by hardware upon occurrence of an interrupt. If the software requires to write to the IC register while the interrupt source is enabled the software write may conflict with a hardware access to bit IR. To address this conflict scenario all xxIC registers are located in the bit addressable memory area. The use of C166 bit modification instructions is therefore possible and recommended. These instructions provide a special “protection mask” feature which allows to protect IR bit from unintended software write. Refer to CPU Bit Manipulation Unit chapter for details. The layout of the interrupt control registers shown below is applicable to all xxIC registers. xxIC Interrupt Control Register (E)SFR (xxxxH) 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 GPX IR IE ILVL GLVL r r r r r r r rw rwh rw rw rw Field Bits Type Description GPX 8 rw Group Priority Extension Defines the value of high-order group level bit IR 7 rwh Interrupt Request Flag 0B No request pending 1B This source has raised an interrupt request IE 6 rw Interrupt Enable Control Bit (individually enables/disables a specific source) 0B Interrupt request is disabled 1B Interrupt request is enabled User’s Manual Interrupts, V3.2 7-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description ILVL [5:2] rw Interrupt Priority Level FH Highest priority level ...... 0HLowest priority level GLVL [1:0] rw Group Priority Level Highest priority level 3H ... ... 0H Lowest priority level 0 [15:9] r Reserved read as 0; should be written with 0. 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 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. 7.3.1 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. User’s Manual Interrupts, V3.2 7-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 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. 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 XC27x5X 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 7.10.4). Interrupt requests programmed to priority levels 7 through 1 will always be serviced by normal interrupt processing. 7.3.2 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 5.8). 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. 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 Interrupts, V3.2 7-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 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 7.3.3 Selective Interrupt Disabling Interrupt requests may be temporarily disabled and enabled during the execution of the software. This may be required to exclude specific interrupt sources based on the current status of the application. In particular, this is necessary to achieve a deterministic execution of time-critical code sequences. Interrupt requests in the XC27x5X can be disabled and enabled on three different levels: • • • Disable all interrupt requests for a certain code sequence Disable all interrupt requests globally Disable single interrupt requests 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. Global interrupt control is achieved with a single instruction: BCLR IEN ;Clear IEN flag (causes pipeline restart) Specific interrupt control is achieved by controlling the enable bits in the associated interrupt control registers. BCLR T2IE ;Clear enable flag to disable intr.node Due to pipeline effects, however, an interrupt request may be executed after the corresponding node was disabled, if the request coincides with clearing the enable flag. If the application must avoid this, the following sequence can be used, ensuring that no interrupt requests from this source will be serviced after disabling the interrupt node: BCLR IEN BCLR T2IE JNB T2IE, Next Next: BSET IEN User’s Manual Interrupts, V3.2 ;Globally disable interrupts ;Disable Timer 2 interrupt node ;Any instruction reading T2IC can be used ;(assures that T2IC is written by BCLR ;before being read by JNB or other instr.) ;Globally enable interrupts again 7-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Please note that the sequence above blindly controls the global enable flag. If the global setting must not be changed, the code sequence can be enhanced, as shown below: JNB IEN, GlobalIntOff BCLR IEN ;Globally disable interrupts BCLR T2IE ;Disable Timer 2 interrupt node JNB T2IE, Next ;Any instruction reading T2IC can be used Next: ;(assures that T2IC is written by BCLR ;before being read by JNB or other instr.) BSET IEN ;Globally enable interrupts again JMPR cc_uc, Continue GlobalIntOff: ;Interrupts are globally disabled anyway BCLR T2IE ;Disable Timer 2 interrupt node JNB T2IE, Continue ;Reading T2IC can be omitted if the next Continue: ;few instructions do not set IEN ... The same function can easily be implemented as a C macro: #define Disable_One_Interrupt(IE_bit) \\ {if(IEN) {IEN=0; IE_bit=0; while (IE_bit); IEN=1;} else {IE_bit=0; while IE_bit);}} Usage Example: Disable_One_Interrupt(T2IE) ; // T2 interrupt enable flag ATOMIC or EXTend sequences preserve the status of the interrupt arbitration when they begin. An accepted request is processed after the ATOMIC/EXTend sequence. Therefore, the following code sequence may not produce the desired result: AvoidThis: ATOMIC #3 NOP BCLR T2IE NOP 7.3.4 ;Disable Timer 2 interrupt node ;Timer 2 request may be processed ;after this instruction!!! 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 XC27x5X 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 built-in and handled automatically by the interrupt controller. User’s Manual Interrupts, V3.2 7-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control • 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. 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 7-1 ILVL (Priority) Software Controlled Interrupt Classes (Example) Group Level 7 15 Interpretation 6 5 4 3 2 1 0 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 Interrupts, V3.2 No service! 7-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.4 Interrupt Vector Table The XC27x5X 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 the following 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 7.9. 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 7.14.2. The interrupt jump cache feature is used. For details, see Section 7.5 The reserved vector locations build a vector table located in the address space of the XC27x5X. 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 5.4. 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 7-9 lists all sources capable of requesting interrupt or PEC service in the XC27x5X, the associated interrupt vector locations, the associated vector numbers, and the associated interrupt control registers. Note: Interrupt nodes which are 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 Interrupts, V3.2 7-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control VECSEG Vector Segment Pointer SFR(FF12H) 7 Reset Value: 00XXH 15 14 13 12 11 10 9 8 6 5 4 3 0 0 0 0 0 0 0 0 VECSEG r r r r r r r r rwh 2 1 Field Bits Type Description VECSEG [7:0] rwh Segment number of the Vector Table 0 [15:8] r Reserved read as 0; should be written with 0. 0 The initial user value of register VECSEG is configured according to settings made for Startup Configuration and Bootstrap Loading. User’s Manual Interrupts, V3.2 7-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.5 Interrupt Jump Table Cache The mechanism that uses the vector table location as the entry point for the interrupt service routines can be overwritten by the Interrupt Controller (ITC). For a very fast interrupt response time, the XC27x5X offers the Interrupt Jump Table Cache (also called “fast interrupt”). The ITC can transfer to the CPU a 24-bit vector which is directly used as a start address for the service routine. This feature skips the path through the vector table which normally saves the execution of at least one branch. Therefore, avoiding the vector table may significantly improve interrupt response time. However, the number of 24-bit vectors in the ITC is limited. Fast interrupt is available for two interrupt sources with interrupt priority levels greater than or equal to 12. The Interrupt Jump Table Cache skips the instruction fetches from the interrupt vector table and executes a direct jump to the interrupt service routines entry point. This feature is controlled by a set of two interrupt jump table cache registers (FINTxCSP, FINTxADDR) for each of the two jump table entries. Every interrupt jump table cache entry contains an enable bit, an associated arbitration priority level (ILVL and GLVL), and the 24-bit address of the interrupt service routine. Note that only the two lower bits of the interrupt priority level are selectable in the respective control registers. The two upper bits of the interrupt priority level are fixed to 11B’, which limits the allowed interrupt priority level to be greater than or equal to 12. FINT0CSP Fast Interrupt Control 0 FINT1CSP Fast Interrupt Control 1 11 10 XSFR(EC00H) Reset Value: 0000H XSFR(EC04H) Reset Value: 0000H 15 14 13 12 9 8 7 EN 0 0 GPX ILVL GLVL SEG rw r r rw rw rw rw Field Bits Type Description EN 15 rw User’s Manual Interrupts, V3.2 6 5 4 3 2 1 0 Fast Interrupt Enable 0B The interrupt jump table cache is disabled. No fast interrupt is used. 1B The interrupt jump table cache is enabled. A fast interrupt (direct jump to the interrupt service routine) is used instead of the normal fetch from the interrupt vector table. 7-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description GPX 12 rw ILVL [11:10] rw Group Priority Extension This bit together with bitfield GLVL selects the group priority level (XGLVL) of the associated interrupt jump table cache entry. Interrupt Priority Level This bitfield selects the lower two bits of the interrupt priority level associated with this interrupt jump table cache entry. Note: The two upper bits of the interrupt priority level are fixed to 11B, which ends in an interrupt priority level greater than or equal to 12. GLVL [9:8] rw Group Priority Level This bitfield together with GPX-bit selects the group priority level (XGLVL) of the associated interrupt jump table cache entry. SEG [7:0] rw Segment Number of Interrupt Service Routine Address bits 23:16 of the interrupt service routine´s entry point. 0 [14:13] r Reserved read as 0; should be written with 0. FINT0ADDR Fast Interrupt Address 0 FINT1ADDR Fast Interrupt Address 1 15 14 13 12 11 10 9 XSFR (EC02H) Reset Value: 0000H XSFR (EC06H) Reset Value: 0000H 8 7 5 4 3 2 1 0 ADDR 0 rw r Field Bits Type Description ADDR [15:1] rw User’s Manual Interrupts, V3.2 6 Address of Interrupt Service Routine Address bits 15:1 of the interrupt service routine’s entry point. 7-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description 0 0 r User’s Manual Interrupts, V3.2 Interrupt Service Routine Address Bit 0 LSB of the interrupt service routine’s entry point address. This address bit is always 0 because of the program code’s word alignment. 7-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.6 CPU Status Saving 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. Status of Interrupted Task SP ---System Stack before Interrupt Entry -- PSW CSP IP System Stack after Interrupt Entry System Stack after Interrupt Entry SP PSW IP SP Task Status saved on the System Stack User’s Manual Interrupts, V3.2 7-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.7 CPU Context Switch An interrupt service routine usually saves all the registers it uses on the stack and restores them before returning. To ease this process the XC27x5X allows switching the complete bank of CPU registers (GPRs) either automatically or with a single instruction, so that the service routine executes within its own separate context (see also Section 5.5.2). There are two ways to switch context: 1. Context switch on interrupt automatically updates bitfield PSW.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. For interrupt priority levels 15 … 12 the target register bank can be pre-selected. The register bank selection registers BNKSELx provide a 2-bit field for each 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. 2. Explicit context switch by software is initiated by a write to CP or PSW registers. a) A write to PSW.BANK bitfields allows to switch between global and local banks. b) A write to CP allows to relocate the memory mapped global bank to another memory location. For example 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”. The new CP value sets a new global register bank. The service routine may now use its “own registers”. This global register bank is preserved when the service routine is terminated, i.e. its contents are available for the next call. Before returning (RETI), the previous CP simply be restored from the system stack using “POP CP”. Note: Other resources used by an interrupting program (like DPP registers) must be saved and restored separately. Note: There are certain timing restrictions during context switching associated with pipeline behavior. For details, see Section 5.5.2. User’s Manual Interrupts, V3.2 7-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.8 Fast Bank Switching The interrupt handler supports an additional enhanced feature (compared to the C166 family) for normal interrupts called Fast Bank Switching. To speed up interrupt handling, the core can use fast General Purpose Register (GPR) bank switching for interrupts with an interrupt level greater or equal than 12. For every arbitration priority level with [ILVL = ’15D’-’12D and XGLVL = ’7D’-’0D’], the register bank can be selected with two bits. The select-bits are located in the four register bank selection registers BNKSELx (x = 0...3). The following table identifies the arbitration priority level assignment to the respective bit fields within the four register bank selection registers: Table 7-2 Register Bank Assignment ILVL XGLVL Assigned GPRSELx Register ILVL XGLVL Assigned GPRSELx Register 15 7 BNKSEL3.GPRSEL7 13 7 BNKSEL2.GPRSEL7 15 6 BNKSEL3.GPRSEL6 13 6 BNKSEL2.GPRSEL6 15 5 BNKSEL3.GPRSEL5 13 5 BNKSEL2.GPRSEL5 15 4 BNKSEL3.GPRSEL4 13 4 BNKSEL2.GPRSEL4 15 3 BNKSEL1.GPRSEL7 13 3 BNKSEL0.GPRSEL7 15 2 BNKSEL1.GPRSEL6 13 2 BNKSEL0.GPRSEL6 15 1 BNKSEL1.GPRSEL5 13 1 BNKSEL0.GPRSEL5 15 0 BNKSEL1.GPRSEL4 13 0 BNKSEL0.GPRSEL4 14 7 BNKSEL3.GPRSEL3 12 7 BNKSEL2.GPRSEL3 14 6 BNKSEL3.GPRSEL2 12 6 BNKSEL2.GPRSEL2 14 5 BNKSEL3.GPRSEL1 12 5 BNKSEL2.GPRSEL1 14 4 BNKSEL3.GPRSEL0 12 4 BNKSEL2.GPRSEL0 14 3 BNKSEL1.GPRSEL3 12 3 BNKSEL0.GPRSEL3 14 2 BNKSEL1.GPRSEL2 12 2 BNKSEL0.GPRSEL2 14 1 BNKSEL1.GPRSEL1 12 1 BNKSEL0.GPRSEL1 14 0 BNKSEL1.GPRSEL0 12 0 BNKSEL0.GPRSEL0 User’s Manual Interrupts, V3.2 7-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control BNKSEL0 Register Bank Selection 0 BNKSEL1 Register Bank Selection 1 BNKSEL2 Register Bank Selection 2 BNKSEL3 Register Bank Selection 3 15 14 13 12 11 10 XSFR(EC20H) Reset Value: 0000H XSFR(EC22H) Reset Value: 0000H XSFR(EC24H) Reset Value: 0000H XSFR(EC26H) Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 GPRSEL7 GPRSEL6 GPRSEL5 GPRSEL4 GPRSEL3 GPRSEL2 GPRSEL1 GPRSEL0 rw rw rw rw Field Bits GPRSEL0, GPRSEL1, GPRSEL2, GPRSEL3, GPRSEL4, GPRSEL5, GPRSEL6, GPRSEL7 [1:0], rw [3:2], [5:4], [7:6], [9:8], [11:10], [13:12], [15:14] rw rw rw rw Type Description Register Bank Selection 00B Global register bank 01B Reserved 10B Local register bank 1 11B Local register bank 2 Note: The GPRSELx value of the current triggered interrupt is automatically transferred into the Program Status Word (PSW). : User’s Manual Interrupts, V3.2 7-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.9 Trap Functions The C166SV2 CPU supports software and hardware trap functions. 7.9.1 Software Traps The TRAP instruction is used to cause a software call to an interrupt service routine. The trap number specified in the operand field of the trap instruction determines which vector location of the vector table will be used. The TRAP instruction has an effect similar to an interrupt request at the same vector. PSW, CSP (in segmentation mode), and IP are pushed into the system stack and then a jump is taken to the specified vector location. When a software trap is executed, the CSP for the trap service routine is loaded with the value of the VECSEG register. 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 PSW register are not modified by the TRAP instruction; so, the service routine is executed with the same priority level as the interrupt task. Therefore, the service routine entered by the TRAP instruction can be interrupted by other traps or by higher priority interrupts, unless triggered by a real hardware event. The service routine also works with an unchanged register bank. If the hardware triggers the same service routine, register bank can be selected by the ITC and may be different. Note: Software traps are also generated and issued, when data reads from the internal program memory space are requested which are not allowed, e.g. a user-read access to the protected Flash. User’s Manual Interrupts, V3.2 7-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.9.2 Hardware Traps Hardware Traps are issued by faults or specific system states that occur during runtime (not identified at assembly time). The XC27x5X distinguishes twelve 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 causing the trap event is completed before the trap handling routine is entered. Hardware traps are not-maskable and always have a priority higher than any other CPU task. If several hardware trap conditions are detected within the same instruction cycle, the highest priority trap is serviced (see Table 7-3). In case of a hardware trap, the injection unit injects a ITRAP instruction into the pipeline. The ITRAP instruction performs the following actions: • • • • Push PSW, CSP (in segmented mode) and IP into the System Stack Set CPU level in the PSW register to the highest possible priority level, which disables all interrupts and PEC transfers Select the global register bank for the trap service routine Branch to the trap vector location specified by the trap number of the trap condition The hardware trap functions of the core are divided in two classes. Class A traps are: • • • • System Request 0 (SR0) Stack Overflow Stack Underflow Software Break These traps share the same trap priority, but have an individual vector address. Class B traps are: • • • • • • System Request 1 (SR1) Memory Protection Undefined Opcode Memory Access Error Protection Fault Illegal Word Operand Access The Class B traps share the same interrupt node and interrupt vector. The bit addressable Trap Flag Register (TFR) allows a trap service routine to identify the trap which caused the exception. 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. The reset functions (hardware, software, watchdog) may be also regarded as a type of trap. Reset functions have the highest priority (trap priority III). Class A traps have the User’s Manual Interrupts, V3.2 7-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control second highest priority (trap priority II), on the 3rd rank are class B traps (trap priority I); thus, a class A trap can interrupt a class B trap (for priority see also Table 7-3). Class A Traps Class A traps are generated by the high priority system request SR0 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 to the events. Each class A trap has its own vector location in the vector table. After finishing the service routine, the instruction flow must be further correctly executed. This explains why class A traps cannot interrupt atomic/ extend sequences and IO accesses in progress. For example, an interrupted extend sequence cannot be restarted. All class A traps are generated in the pipeline during the execution of instructions, with the exception of SR0, which is an asynchronous external event. It is not possible for two different instructions in the pipeline to generate traps in the same CPU cycle. Class A trap events can be generated only during the memory stage of execution. An execution of instruction which caused a class A trap event is always completed. In the case of a class A trap, the pipeline is directly canceled and the IP of the instruction following the last executed one is pushed into the stack. In the case of an atomic/extend sequence or IO read access in progress, the execution continues till the sequence completion. Upon completion of the sequence, the IP of the instruction following the last one executed is pushed into 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 an execution of branch instructions 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 into the stack. If more than one Class A trap occurs at a same time, they are prioritized internally. The SR0 trap has the highest priority and the software break has the lowest. Note: In the case of two different class A trap occurring simultaneously, both trap flags are set. The IP of the instruction following the last one executed is pushed into 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 elated to the second trap flag in TFR has been cleared by the first trap service routine). 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 User’s Manual Interrupts, V3.2 7-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control can interrupt an atomic/extend sequence and an IO 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 the software in the trap service routine. All class B traps are synchronous to instruction execution; most of them are generated in the pipeline during the execution of instructions. It is not possible for two different instructions in the pipeline to generate class A and class B traps in the same CPU cycle. Class B trap events can be generated only during memory stage execution. SR1 and ACER are exceptions, because they are generated by the SCU. Instructions which caused a class B trap event are always executed. In the case of a class B trap, the pipeline is directly 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 branch instructions 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 into the stack. During execution of a class A trap service routine, any class B trap will not be serviced until the class A trap service routine is exited with a RETI instruction. In this case, the class B trap condition is stored in the TFR register, but the IP value of the instruction which caused this trap will be 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 IO 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. • • System Request 0 Trap (A): The control signal is generated by the SCU. See chapter SCU Trap Generation. Stack Overflow Trap (A): Whenever the stack pointer is implicitly decremented and if the stack pointer was 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. User’s Manual Interrupts, V3.2 7-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control • • • • • • • • Stack Underflow Trap (A): Whenever the stack pointer is implicitly incremented and if the stack pointer was 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 (A): 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 an On-chip 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. System Request 1 Trap (B): The control signal is generated by the SCU. See chapter SCU Trap Generation. Memory Protection Traps (B): When an access violation outside the permitted address ranges is detected. Depending on the access type it is differentiated between Read (MPR), Write (MPW) and Execute (MPX) violations. Undefined Opcode Trap (B): When the instruction currently being decoded by the CPU does not contain a valid C166SV2 opcode, the UNDOPC flag is set in register TFR and the CPU enters the undefined opcode trap routine. The instruction which causes the undefined opcode trap is executed as a NOP. Memory Access Error (B): The control signal is generated by the SCU. See chapter SCU Trap Generation. Protection Fault Trap (B): 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 which causes the protection fault trap is executed as a NOP. For products supporting MPU, RETI is also defined as a protected instruction in the sense that its execution is only allowed for privileged code, i.e. code executed with protection level 0. This flag is then used to indicate that a RETI instruction was tried to be executed from a protection level different to 0. Note that RETI will be still executed even if it causes a protection fault trap (it is not executed as a NOP). Illegal Word Operand Access Trap (B): Whenever a word operand read or write access (including Flash commands!) 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. Trap Vector Locations Table 7-3 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 User’s Manual Interrupts, V3.2 7-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control (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 7-3 Hardware Trap Summary Exception Condition Trap Flag Trap Vector Vector Vector Trap 1) Number Priority Location Application Reset – RESET xx’0000H 00H III Class A Hardware Traps: • System Request 0 • Stack Overflow • Stack Underflow • Software Break SR0 STKOF STKUF SOFTBRK SR0TRAP STOTRAP STUTRAP SBRKTRAP xx’0008H xx’0010H xx’0018H xx’0020H 02H 04H 06H 08H II II II II SR1 MPR/W/X UNDOPC ACER PRTFLT BTRAP BTRAP BTRAP BTRAP BTRAP xx’0028H xx’0028H xx’0028H xx’0028H xx’0028H 0AH 0AH 0AH 0AH 0AH I I I I I ILLOPA BTRAP xx’0028H 0AH I Class B Hardware Traps: • System Request 1 • Memory Protection • Undefined Opcode • Memory 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 Interrupts, V3.2 7-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.9.2.1 The Trap Flag Register TFR The XC27x5X provides a number of trap vectors (class A and class B) which are indicated in the trap flag register TFR. TFR Trap Flag Register 15 14 12 11 10 9 8 7 Reset Value: 0000H 6 5 STK STK SR0 OF UF 0 0 rwh r r rwh 13 SFR(FFACH/D6H) SOF UND T SR1 MPR MPW MPX OPC BRK rwh rwh rwh rwh rwh rwh rwh 4 3 1 0 AC PRT ILL ER FLT OPA 0 0 rwh r r rwh 2 rwh Field Bits Type Description SR0 15 rwh System request flag 0 0B No trigger detected 1B The selected condition has been detected STKOF 14 rwh Stack overflow flag 0B No stack overflow event detected 1B The current stack pointer value falls below the contents of register STKOV STKUF 13 rwh Stack underflow flag 0B No stack underflow event detected The current stack pointer value exceeds the 1B contents of register STKUN SOFTBRK 12 rwh Software Break 0B No software break event detected 1B Software break event detected SR1 11 rwh System request flag 1 0B No trigger detected 1B The selected condition has been detected MPR 10 rwh Memory Protection Read No read protection violation detected 0B 1B Read protection violation detected MPW 9 rwh Memory Protection Write 0B No write protection violation detected 1B Write protection violation detected User’s Manual Interrupts, V3.2 7-29 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description MPX 8 rwh Memory Protection Execute 0B No execute protection violation detected 1B Execute protection violation detected UNDOPC 7 rwh Undefined Opcode No undefined opcode event detected 0B The currently decoded instruction has no valid 1B opcode ACER 4 rwh Memory Access Error 0B No access error event detected 1B Illegal or erroneous access detected PRTFLT 3 rwh Protection Fault 0B No protection fault event detected A protected instruction with an illegal format 1B has been detected ILLOPA 2 rwh Illegal word operand access 0B No illegal word operand access event detected 1B A word operand access (read or write) to an odd address has been attempted 0 [6:5], [1:0] r Reserved read as 0; should be written with 0. Note: Flags TFR.15, TFR.11 and TFR.4 are generated via SCU. TFR.8, TFR9 and TFR.10 are generated via MPU. Other flags are generated by the CPU. User’s Manual Interrupts, V3.2 7-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.10 Peripheral Event Controller The XC27x5X’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 can perform the following actions: • • • • Byte or word transfer Continuous data transfer PEC channel-specific interrupt request upon data transfer completion or common for all channels “End of PEC” interrupt for enhanced handling Automatic increment of source or/and destination pointers with support of memory to memory transfer Note: PEC transfer is executed if its priority level is higher than current CPU priority level. User’s Manual Interrupts, V3.2 7-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.10.1 PEC Control Registers Each PEC channel is controlled by the respective PEC channel Control register (PECCx) and a set of source and destination pointers (SRCPx, DSTPx and PECSEGx), where x stands for the PEC channel number. The PECCx registers control the arbitration priority level assignment to the PEC channels and the action to be performed. PECCx (x=0-7) PEC Channel Control x 15 0 r 14 EOP INT rw 13 12 11 SFR(FEC0H+2*x) 10 9 8 7 6 Reset Value: 0000H 5 4 3 PLEV CL INC BWT COUNT rw rw rw rw rwh 2 1 0 Field Bits Type Description EOPINT 14 rw End of PEC Interrupt Selection 0B End of PEC interrupt with the same level as the PEC transfer is triggered 1B End of PEC interrupt is serviced by a separate interrupt node with programmable interrupt level (EOPIC) and interrupt sharing control register (PECISNC) PLEV [13:12] rw Programmable PEC Interrupt Level 00B Standard (compatible mode): Levels 15 and 14 01B Levels 13 and 12 10B Levels 11 and 10 11B Levels 9 and 8 CL 11 Channel Link Control 0B PEC channels work independent 1B Pairs of channels are linked together User’s Manual Interrupts, V3.2 rw 7-32 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description INC [10:9] rw Increment Control (Modification of source and destination pointer after PEC transfer) 00B No modification 01B Increment of destination pointer DSTPx by 1 (BWT = 1) or by 2 (BWT = 0) 10B Increment of source pointer SRCPx by 1 (BWT = 1) or by 2 (BWT = 0) 11B Increment of destination pointer DSTPx and source pointer SRCPx by 1 (BWT = 1) or by 2 (BWT = 0) BWT 8 rw Byte/Word Transfer Selection 0B Transfer a word 1B Transfer a byte COUNT [7:0] rwh PEC Transfer Count Counts PEC transfers and influences the channel´s action (see Table 7-4) 0 15 r Reserved read as 0; should be written with 0. The Byte/Word Transfer Bit (BWT) of the PECCx register selects if a byte or a word is to be moved during a PEC service cycle and defines an increment step size for the pointer(s) to be modified. The PEC Transfer Count Field (COUNT) of the PECCx directly controls the action of the respective PEC channel. The contents of the bitfield COUNT may specify a certain number of PEC transfers, unlimited transfers, or no PEC service at all. • If the PEC transfer counter COUNT value is set to 00H, the normal interrupt requests are processed instead of PEC data transfers and the corresponding PEC channel remains idle. • Continuous data transfers are selected by setting the bitfield COUNT to FFH value. In this case, COUNT is not decremented by the transfers and the respective PEC channel can serve unlimited number of PEC requests until it is modified by the program. • If the bitfield COUNT is set to service a specified number of requests by the respective PEC channel, it is decremented with each PEC transfer and the request flag is cleared to indicate that the request has been serviced. When COUNT reaches 00H, it activates the interrupt service routine which has the same priority level User’s Manual Interrupts, V3.2 7-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control (EOPINT = 0) or triggers the “End of PEC” interrupt with a different priority level (EOPINT = 1). When COUNT is decremented from 01H to 00H after a data transfer, the request flag will be cleared if EOPINT is set to 1. If EOPINT is 0, the request flag will not be cleared and another interrupt request will be generated on the same priority level. The respective PEC channel remains idle and the associated interrupt service routine is activated instead of PEC transfer because COUNT contains the 00H value. (see Section 7.10.3). The EOPIC register is the interrupt control register of the End Of PEC interrupt. EOPIC End Of PEC Interrupt Control ESFR(F19EH/CFH) 15 14 13 12 11 10 9 8 7 0 0 0 0 0 0 0 GPX r r r r r r r rw 6 Reset Value: 0000H 5 4 EOP EOP IR IE rwh rw 3 2 1 0 ILVL GLVL rw rw Field Bits Type Description GPX 8 rw Group Priority Extension Defines the value of high-order group level bit EOPIR 7 rwh Interrupt Request Flag 0B No request pending 1B The source has raised an interrupt request EOPIE 6 rw Interrupt Enable Control Bit 0B Interrupt request is disabled Interrupt request is enabled 1B ILVL [5:2] rw Interrupt Priority Level FH Highest priority level ...H ... 0H Lowest priority level GLVL [1:0] rw Group Priority Level 3H Highest priority level ... ... 0H Lowest priority level 0 [15:9] r Reserved read as 0; should be written with 0. User’s Manual Interrupts, V3.2 7-34 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Note: The concatenation of the group priority extension bit GPX and the group priority level bitfield GLVL builds the 3-bit Extended Group Priority Level XGLVL, where 7H is the highest priority level and 0H is the lowest priority level. The Register PECISNC contains flags of the “End of PEC” interrupt node. This node is used when enhanced “End of PEC” interrupt feature was invoked and control bit EOPINT is set to 1 in the corresponding PECCx. PECISNC PEC Interrupt Sub Node ControlSFR(FFD8H/ECH) 15 14 13 12 11 10 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=0-7) 2*x+1 rwh Interrupt Sub Node Request Flag of PEC Channel x 1) 0B No special end of PEC interrupt request is pending for PEC channel x 1B PEC channel x has raised an end of PEC interrupt request CxIE (x=0-7) 2*x rw Interrupt Sub Node Enable Control Bit of PEC Channel x 2) (individually enables/disables a specific source) 0B End of PEC interrupt request of PEC channel x is disabled 1B End of PEC interrupt request of PEC channel x is enabled 1) x = 7...0, depending on PEC channel number 2) It is recommended to clear an interrupt request flag (CxIR) before setting the respective enable flag (CxIE). Otherwise, former requests still pending will immediately trigger an interrupt request after setting the enable bit. Note: The “End of PEC” sub-node interrupt request flags are not cleared by hardware when entering the interrupt service routine (interrupt has been accepted by the CPU), unlike the interrupt request flags of the interrupt nodes (request flags xxIC.xxIR). The interrupt service routine must check the request flags and clear them before executing the RETI instruction. The following figure shows the usage of the “End of PEC” interrupt subnode: User’s Manual Interrupts, V3.2 7-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control PECISNC C7IR C7IE C6IR C6IE C5IR C5IE C4IR C4IE C3IR C3IE C2IR C2IE C1IR C1IE C0IR C0IE 15 0 Interrupt Request Pulse Generator EOPIC 0 0 15 Figure 7-3 0 0 0 0 0 GPX EOP EOP IR IE 87 ILVL GLVL 0 End of PEC Interrupt Sub Node The table below summarizes the values of the bitfield COUNT and the corresponding PEC channel actions User’s Manual Interrupts, V3.2 7-36 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-4 PEC Channel Actions Previous COUNT field value Modified COUNT field value Action of PEC Channel and Comments FFH FFH Move a Byte/Word Continuous transfer mode, i.e. COUNT is not modified FEH...02H FDH...01H Move a Byte/Word and decrement COUNT 01H 00H Move a Byte/Word Depending on bit EOPINT, one of two different actions are taken: EOPINT = 0 (compatible mode) The service request flag (xxIR) of the respective interrupt remains set (it is cleared for all other COUNT values). Therefore, an additional interrupt request is triggered on the next arbitration cycle with a COUNT field value of 00H (see next raw) EOPINT = 1 The service request flag (xxIR) of the respective interrupt is cleared. Additionally, the interrupt request flag of the EOP sub node (PECISNC.CxIR) is set. Furthermore, the interrupt request flag of the end of PEC interrupt node (EOPIC.EOPIR) is automatically set if the sub node request is enabled (PECISNC.CxIE = 1’). (see also Section 7.10.3) 00H 00H No PEC action! A normal interrupt is requested instead of a PEC data transfer (see also Section 7.10.3). The Increment Control Field (INC) of the PECCx register defines whether one or both of the PEC pointers must be incremented after the PEC transfer. If the pointers are not to be modified (INC=’00’), the respective channel will always move data from the same source to the same destination. Channel Link Mode for Data Chaining Channel linking, if enabled, links two channels together to serve the data transfer requests of one peripheral. The whole data transfer (for example a message) is divided into separately controlled and chained block transfers. The two PEC channels which are linked together, handle chained block transfers alternately to each other. At the end of a User’s Manual Interrupts, V3.2 7-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control data block transfer, controlled by one PEC channel, automatically the other (linked) PEC channel is started to continue the transfer with the next data block. Channel linking and thus data (block) chaining is supported within pairs of PEC channels (channels 0&1, 2&3 a.s.o.). Each data block is controlled by one PEC channel of the channel pair. While one of the two channels is active, the CPU can update the pointer and counter values of the other channel to prepare it for continuation of data transfer after next channel linking. Channel linking is enabled, if in the active PEC channel of the channel pair the Channel Link Control Bit “CL” in it’s PECCx register is set to 1. The data transfer of linked channels is started always with the even numbered channel of the channel pair. If in Channel Link mode (at least one CL bit of the pair is set) the channel’s data block is completely transferred the PEC service request processing is automatically switched to the other PEC channel of the channel-pair. Channel linking and thus the switching from one channel to the other channel is performed, when the CL bit of first (active) channel is set (in its PEC control register) and its transfer count is changed from one to zero with last transfer. If the channel link flag CL of the first (terminated) PEC control register is found to be zero or the count field of linked channel is zero, the whole data transfer is finished. Note: The CL-flags are fully controlled by software and should be cleared by SW when the whole data transfer shall be finished and the termination of transfer shall be executed. Because termination can also be entered with a zero-value of the transfer count field of linked channel, termination of whole data transfer is automatically performed if the channels count field was not updated after the last channel link interrupt for this channel. When a data block of a linked channel is completely transferred and PEC servicing switches to the other channel of channel pair, a channel specific channel link interrupt is generated (for old channel) to inform the CPU that the channel is inactive now and may be configured for its next block transfer. The channel link interrupt is requested, indicated and enabled in the respective PEC Interrupt Subnode Control Register (PECISNC), which is also used for the channel’s End of PEC interrupt (see chapter above). Thus, all channel link interrupts are also controlled with the one EOP interrupt control register EOPIC and therefore with the same interrupt priority level as the EOP interrupt. This service request node requests CPU interrupt service in case of one or more pending channel link interrupt requests or End of PEC interrupt requests, if the respective enable control bit(s) is (are) set in the PEC interrupt subnode control register PECISNC and in the interrupt control register EOPIC. Note: The generation of Channel Link/EOP interrupt is automatically enabled, if the CLbit of the active (terminated) channel is set. If it is not set, either a standard interrupt or an EOP interrupt is initiated according to the EOPINT bit in the channel’s PEC Control Register. The channel is not switched in this case, because a missing CL flag defines the last block of data transfer. User’s Manual Interrupts, V3.2 7-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Note: If Channel Link mode is active (at least one of the pair’s CL bits is set), interrupt requests connected to the odd channel (via priority levels) will trigger only a standard interrupt but no PEC transfer. Note: The start of data transfer on linked channels is always performed with the even numbered PEC channel of the channel pair. User’s Manual Interrupts, V3.2 7-39 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.10.2 The PEC Source and Destination Pointer The PEC channels source and destination pointers specify the locations between which the data is to be moved. All pointers are 24-bits wide. The 24-bit source address is stored in the register SRCPx (lower 16 bits of address) and in the high byte of register PECSEGx (highest 8 address bits). PECSEGx SRCSEGx DSTSEGx 8 7 15 0 SRCPx DSTPx SRCPx 15 DSTPx 0 15 Source Pointer 23 16 15 Segment Address 0 Destination Pointer 0 23 Segment Offset 16 15 Segment Address 0 Segment Offset Data Transfer x = 7...0, depending on PEC channel number Figure 7-4 PEC Pointer Address Handling The 24-bit destination address is stored in the register DSTPx (lower 16 bits of address) and in the low byte of register PECSEGx (highest 8 address bits). Only the lower 16 bits of the PEC address pointers (segment offset) can be modified (incremented) by the PEC transfer mechanism. The highest 8 bits, which represent the segment number, are not modified by hardware. Therefore, the PEC pointers may be incremented within the address space of one segment and may not cross the segment border. If the offset address pointer gets the FFFFH value in case of byte transfers (BWT = 1) or FFFEH in case of word transfers (BWT = 0), the next increment will be disregarded. The address register will keep one of these maximum values and no overflow will happen. The described behavior protects the subsequent memory from unintentional overwriting. No User’s Manual Interrupts, V3.2 7-40 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control explicit error event is generated by the system in case of address pointer(s) 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. Note: If a word data transfer is selected for a specific PEC channel (i.e. 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 Operand Access trap will be invoked when this channel is used. SRCPx (x=0-7) PEC Source Pointer x 15 14 13 12 XSFR(EC40H+4*x) 11 10 9 8 7 6 Reset Value: 0000H 5 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 x = 7...0, depending on PEC channel number DSTPx (x=0-7) PEC Destination Pointer x 15 14 13 12 11 XSFR(EC42H+4*x) 10 9 8 7 6 Reset Value: 0000H 5 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 x = 7...0, depending on PEC channel number For XSFR addresses see Table 7-11. User’s Manual Interrupts, V3.2 7-41 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control PECSEGx (x=0-7) PEC Segment Pointer x 15 14 13 12 11 XSFR(EC80H+2*x) 10 9 8 7 6 Reset Value: 0000H 5 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 Address of Channel x Destination address bits 23 … 16 0 x = 7...0, depending on PEC channel number 7.10.3 PEC Interrupt Processing Summary As described above, two different kinds of interrupts can be triggered by the PEC handler depending on the status of the bitfield COUNT. • • PEC channel is enabled1) and the bitfield COUNT has a value higher then 01H. – Control bit EOPINT = 0 or 1 ACTIONS: – PEC request is proceeded – No other interrupt activity PEC channel is enabled and the bitfield COUNT gets a decrement from 01H to 00H triggered by a service request. – Control bit EOPINT = 0 (compatible with C166) ACTIONS: – PEC request is proceeded – Interrupt request flag (xxIR) of the requesting interrupt node (arbitration winner) is not cleared, participates on the next arbitration cycle, and triggers a normal interrupt on the same level as the PEC request is served. – Control bit EOPINT = 1 (enhanced end of PEC interrupt feature) ACTIONS: – PEC request is proceeded 1) Every PEC channel is automatically enabled, when its COUNT value is greater than 00H. User’s Manual Interrupts, V3.2 7-42 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control – Interrupt request flag (xxIR) of requesting interrupt node (arbitration winner) is cleared and will not trigger more actions. – Interrupt request flag of the end of PEC interrupt subnode will be set (PECISNC.CxIR = 1) – If the respective interrupt enable flag of the end of PEC interrupt subnode was set before by software (PECISNC.CxIE = 1), an end of PEC interrupt is requested (EOPIC.EOPIR = 1). This end of PEC interrupt participates on the next arbitration cycle with its priority (selected via EOPIC.ILVL and EOPIC.GLVL), if this interrupt source was enabled before by software (EOPIC.EOPIR = 1). With this behavior an end of PEC interrupt can be triggered on a lower level than the respective PEC requests have been serviced. • PEC channel is disabled if the bitfield COUNT is cleared (either by hardware or by software) – Control bit EOPINT = 0 or 1 ACTIONS: – A normal interrupt service routine is requested on the PEC channel priority level. 7.10.4 PEC Channel Assignment The PEC channel used for executing the transfer depends on the programming of the interrupt, group and PEC level. The following table lists the channel assignments. Table 7-5 ICx. ILVL PEC Channel Assignment ICx. PECCy. Assigned GLVL PLEV PEC Channel 1111B 11B 1111B ICx. PECCy. Assigned GLVL PLEV PEC Channel 7 1011B 11B 10B 6 1011B 10B 6 1111B 01B 5 1011B 01B 5 1111B 00B 4 1011B 00B 4 1110B 11B 3 1010B 11B 3 1110B 10B 2 1010B 10B 2 1110B 01B 1 1010B 01B 1 1110B 00B 0 1010B 00B 0 User’s Manual Interrupts, V3.2 00B ICx. ILVL 7-43 10B 7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-5 ICx. ILVL PEC Channel Assignment ICx. PECCy. Assigned GLVL PLEV PEC Channel 1101B 11B 1101B 01B ICx. ILVL ICx. PECCy. Assigned GLVL PLEV PEC Channel 7 1001B 11B 11B 7 10B 6 1001B 10B 6 1101B 01B 5 1001B 01B 5 1101B 00B 4 1001B 00B 4 1100B 11B 3 1000B 11B 3 1100B 10B 2 1000B 10B 2 1100B 01B 1 1000B 01B 1 1100B 00B 0 1000B 00B 0 All interrupt requests not assigned to a PEC channel go directly to the interrupt handler. User’s Manual Interrupts, V3.2 7-44 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.11 External Interrupts Although the XC27x5X has no dedicated interrupt input pins, it supports many possibilities to react to external asynchronous events by providing a number of IO lines which can be selected as interrupt inputs. 7.11.1 External Request Unit Please refer to the External Request Unit (ERU) chapter. The ERU provides routing capabilities and allows to define advanced trigger conditions for the interrupt input signals. The resulting ERU interrupt requests are forwarded to the interrupt controller registers SCU_ERU_0IC ... SCU_ERU_3IC. 7.11.2 Using Peripheral Pins The interrupt function of some peripheral pins may be either combined with the pin’s main function or used instead of it if the main pin function is not required. Table 7-6 Pins Usable as External Interrupt Inputs Port Pin Original Function Control Register P4.7-0/CC31-24IO CAPCOM Register 31-24 Capture Input CC31-CC241) P2.10-3/CC23-16IO CAPCOM Register 23-16 Capture Input2) CC23-CC161) P4.2/T2IN Auxiliary timer T2 input pin T2CON P4.6/T4IN Auxiliary timer T4 input pin T4CON P2.10/CAPIN GPT2 capture input pin2) T5CON 1) Must be enabled by Interrupt Node Sharing. 2) Pin P2.10 overlays two possible input functions. 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. 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 port control register. 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 User’s Manual Interrupts, V3.2 7-45 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control in register CCxIC will be set on a positive external transition at pin CCxIO. When 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. User’s Manual Interrupts, V3.2 7-46 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.12 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 Interrupts, V3.2 7-47 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.13 Service Request Latency The numerous service requests of the XC27x5X (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 7-7 lists the consecutive steps required for this process. Table 7-7 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 7.2) 3 cycles 3 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 11/7 cycles 7 cycles 1) Can be saved by using the interrupt jump table cache (see Section 7.5). User’s Manual Interrupts, V3.2 7-48 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 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 7-8 Additional Delays Caused by System Logic Reason for Delay Interrupt Response PEC Response Interrupt controller busy, max. 7 cycles because the previous interrupt request is still in process max. 7 cycles Pipeline is stalled, 2 × TACCmax1) 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 × TACC2) --- Memory access for vector table read (except for intr. jump table cache) 2 × TACC --- 1) This is the longest possible access time within the XC27x5X system. 2) 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 7.6) 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 Interrupts, V3.2 7-49 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.14 Interrupt Nodes 7.14.1 Physical Interrupt Nodes The full set of enabled and used modules integrated in the XC27x5X would require more than the 96 interrupt nodes provided by the C166SV2 interrupt controller. Therefore some of the physically available interrupt nodes are shared between selected modules. The following table summarizes the 96 physical interrupt nodes with their related • • • • trap number vector location control register name and address node sharing information Table 7-9 XC27x5X Interrupt Nodes Source of Interrupt or PEC Trap Service Request Number Vector1) Location Control Register Register Address selected by ISSR.ISS0 10H xx’0040H CC2_CC16IC F1C0H selected by ISSR.ISS1 11H xx’0044H CC2_CC17IC F1C2H selected by ISSR.ISS2 12H xx’0048H CC2_CC18IC F1C4H selected by ISSR.ISS3 13H xx’004CH CC2_CC19IC F1C6H selected by ISSR.ISS4 14H xx’0050H CC2_CC20IC F1C8H selected by ISSR.ISS5 15H xx’0054H CC2_CC21IC F1CAH selected by ISSR.ISS6 16H xx’0058H CC2_CC22IC F1CCH selected by ISSR.ISS7 17H xx’005CH CC2_CC23IC F1CEH selected by ISSR.ISS8 18H xx’0060H CC2_CC24IC F1D0H selected by ISSR.ISS9 19H xx’0064H CC2_CC25IC F1D2H selected by ISSR.ISS10 1AH xx’0068H CC2_CC26IC F1D4H selected by ISSR.ISS11 1BH xx’006CH CC2_CC27IC F1D6H selected by ISSR.ISS12 1CH xx’0070H CC2_CC28IC F1D8H selected by ISSR.ISS13 1DH xx’0074H CC2_CC29IC F1DAH selected by ISSR.ISS14 1EH xx’0078H CC2_CC30IC F1DCH selected by ISSR.ISS15 1FH xx’007CH CC2_CC31IC F1DEH GPT1 Timer 2 20H xx’0080H GPT12E_T2IC FF60H GPT1 Timer 3 21H xx’0084H GPT12E_T3IC FF62H GPT1 Timer 4 22H xx’0088H GPT12E_T4IC FF64H User’s Manual Interrupts, V3.2 7-50 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-9 XC27x5X Interrupt Nodes Source of Interrupt or PEC Trap Service Request Number Vector1) Location Control Register Register Address GPT2 Timer 5 23H xx’008CH GPT12E_T5IC FF66H GPT2 Timer 6 24H xx’0090H GPT12E_T6IC FF68H GPT2 CAPREL 25H xx’0094H GPT12E_CRIC FF6AH CAPCOM2 Timer 7 26H xx’0098H CC2_T7IC FF6CH CAPCOM2 Timer 8 27H xx’009CH CC2_T8IC FF6EH A/D Converter Request 0 28H xx’00A0H ADC_0IC FF70H A/D Converter Request 1 29H xx’00A4H ADC_1IC FF72H A/D Converter Request 2 2AH xx’00A8H ADC_2IC FF74H A/D Converter Request 3 2BH xx’00ACH ADC_3IC FF76H A/D Converter Request 4 2CH xx’00B0H ADC_4IC FF78H A/D Converter Request 5 2DH xx’00B4H ADC_5IC FF7AH A/D Converter Request 6 2EH xx’00B8H ADC_6IC FF7CH A/D Converter Request 7 2FH xx’00BCH ADC_7IC FF7EH CCU60 Request 0 30H xx’00C0H CCU60_0IC F160H CCU60 Request 1 31H xx’00C4H CCU60_1IC F162H CCU60 Request 2 32H xx’00C8H CCU60_2IC F164H CCU60 Request 3 33H xx’00CCH CCU60_3IC F166H CCU61 Request 0 34H xx’00D0H CCU61_0IC F168H CCU61 Request 1 35H xx’00D4H CCU61_1IC F16AH CCU61 Request 2 36H xx’00D8H CCU61_2IC F16CH CCU61 Request 3 37H xx’00DCH CCU61_3IC F16EH CCU62 Request 0 38H xx’00E0H CCU62_0IC F170H CCU62 Request 1 39H xx’00E4H CCU62_1IC F172H CCU62 Request 2 3AH xx’00E8H CCU62_2IC F174H CCU62 Request 3 3BH xx’00ECH CCU62_3IC F176H CCU63 Request 0 3CH xx’00F0H CCU63_0IC F178H CCU63 Request 1 3DH xx’00F4H CCU63_1IC F17AH CCU63 Request 2 3EH xx’00F8H CCU63_2IC F17CH CCU63 Request 3 3FH xx’00FCH CCU63_3IC F17EH CAN0 40H xx’0100H CAN_0IC F140H User’s Manual Interrupts, V3.2 7-51 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-9 XC27x5X Interrupt Nodes Source of Interrupt or PEC Trap Service Request Number Vector1) Location Control Register Register Address CAN1 41H xx’0104H CAN_1IC F142H CAN2 42H xx’0108H CAN_2IC F144H CAN3 43H xx’010CH CAN_3IC F146H CAN4 44H xx’0110H CAN_4IC F148H CAN5 45H xx’0114H CAN_5IC F14AH CAN6 46H xx’0118H CAN_6IC F14CH CAN7 47H xx’011CH CAN_7IC F14EH CAN8 48H xx’0120H CAN_8IC F150H CAN9 49H xx’0124H CAN_9IC F152H CAN10 4AH xx’0128H CAN_10IC F154H CAN11 4BH xx’012CH CAN_11IC F156H CAN12 4CH xx’0130H CAN_12IC F158H CAN13 4DH xx’0134H CAN_13IC F15AH CAN14 4EH xx’0138H CAN_14IC F15CH CAN15 4FH xx’013CH CAN_15IC F15EH USIC0 CH0 SR0 50H xx’0140H U0C0_0IC F120H USIC0 CH0 SR1 51H xx’0144H U0C0_1IC F122H USIC0 CH0 SR2 52H xx’0148H U0C0_2IC F124H USIC0 CH1 SR0 53H xx’014CH U0C1_0IC F126H USIC0 CH1 SR1 54H xx’0150H U0C1_1IC F128H USIC0 CH1 SR2 55H xx’0154H U0C1_2IC F12AH USIC1 CH0 SR0 56H xx’0158H U1C0_0IC F12CH USIC1 CH0 SR1 57H xx’015CH U1C0_1IC F12EH USIC1 CH0 SR2 58H xx’0160H U1C0_2IC F130H USIC1 CH1 SR0 59H xx’0164H U1C1_0IC F132H USIC1 CH1 SR1 5AH xx’0168H U1C1_1IC F134H USIC1 CH1 SR2 5BH xx’016CH U1C1_2IC F136H USIC2 CH0 SR0 5CH xx’0170H U2C0_0IC F138H USIC2 CH0 SR1 5DH xx’0174H U2C0_1IC F13AH USIC2 CH0 SR2 5EH xx’0178H U2C0_2IC F13CH User’s Manual Interrupts, V3.2 7-52 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-9 XC27x5X Interrupt Nodes Source of Interrupt or PEC Trap Service Request Number Vector1) Location Control Register Register Address USIC2 CH1 SR0 5FH xx’017CH U2C1_0IC F13EH USIC2 CH1 SR1 60H xx’0180H U2C1_1IC F180H USIC2 CH1 SR2 61H xx’0184H U2C1_2IC F182H USIC3 CH0 SR0 62H xx’0188H U3C0_0IC F184H USIC3 CH0 SR1 63H xx’018CH U3C0_1IC F186H USIC3 CH0 SR2 64H xx’0190H U3C0_2IC F188H USIC3 CH1 SR0 65H xx’0194H U3C1_0IC F18AH USIC3 CH1 SR1 66H xx’0198H U3C1_1IC F18CH USIC3 CH1 SR2 67H xx’019CH U3C1_2IC F18EH SCU External Request 0 68H xx’01A0H SCU_ERU_0IC F190H SCU External Request 1 69H xx’01A4H SCU_ERU_1IC F192H SCU External Request 2 6AH xx’01A8H SCU_ERU_2IC F194H SCU Interrupt 1 6BH xx’01ACH SCU_1IC F196H SCU Interrupt 0 6CH xx’01B0H SCU_0IC F198H SCU External Request 3 6DH xx’01B4H SCU_ERU_3IC F19AH RTC 6EH xx’01B8H RTC_IC F19CH End of PEC Subchannel 6FH xx’01BCH EOPIC F19EH 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 Interrupts, V3.2 7-53 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.14.2 Interrupt Node Sharing Interrupt source selection is adjustable to the application focus. The concept described in this chapter allows to adjust the focus to be more on control (CAPCOM2) or on communication (USIC) side. Interrupt node sharing is controlled by SCU register ISSR (see Section 7.14.3). Shared Nodes controlled by ISSR The following figure visualizes the sharing principle controlled by ISSR register. The default interrupt source is CAPCOM2. The alternate source selections are the SCU interrupts 2 and 3 and the SR3 requests of all USIC channels. Default Source x(0) ISSR MUX Alternate Source x(1) Physical Node (Control Register) ISSx ISSR Figure 7-5 Node Sharing Principle controlled by ISSR The table below lists the possible selections of interrupt request sources to the physical interrupt nodes (Control Register). Selection is controlled through ISSx bits of the ISSR register. Table 7-10 Nodes Sharing controlled by ISSR Control Register Select Bit Default Source (ISSx=0) Alternate Source (ISSx=1) CC2_CC16IC ISS0 CAPCOM2 Request 16 (not assigned) CC2_CC17IC ISS1 CAPCOM2 Request 17 (not assigned) CC2_CC18IC ISS2 CAPCOM2 Request 18 USIC3 CH0 SR3 CC2_CC19IC ISS3 CAPCOM2 Request 19 USIC3 CH1 SR3 CC2_CC20IC ISS4 CAPCOM2 Request 20 USIC0 CH0 SR3 CC2_CC21IC ISS5 CAPCOM2 Request 21 USIC0 CH1 SR3 CC2_CC22IC ISS6 CAPCOM2 Request 22 USIC1 CH0 SR3 CC2_CC23IC ISS7 CAPCOM2 Request 23 USIC1 CH1 SR3 User’s Manual Interrupts, V3.2 7-54 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-10 Nodes Sharing controlled by ISSR Control Register Select Bit Default Source (ISSx=0) Alternate Source (ISSx=1) CC2_CC24IC ISS8 CAPCOM2 Request 24 (not assigned) CC2_CC25IC ISS9 CAPCOM2 Request 25 (not assigned) CC2_CC26IC ISS10 CAPCOM2 Request 26 (not assigned) CC2_CC27IC ISS11 CAPCOM2 Request 27 (not assigned) CC2_CC28IC ISS12 CAPCOM2 Request 28 USIC2 CH0 SR3 CC2_CC29IC ISS13 CAPCOM2 Request 29 USIC2 CH1 SR3 CC2_CC30IC ISS14 CAPCOM2 Request 30 SCU Interrupt 2 CC2_CC31IC ISS15 CAPCOM2 Request 31 SCU Interrupt 3 User’s Manual Interrupts, V3.2 7-55 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.14.3 Interrupt Source Select Registers In order to map the interrupt request sources in the complete system to the available interrupt nodes, interrupt nodes are shared between selected modules. SCU_ISSR Interrupt Source Select Register SFR (FF2EH/97H) Reset Value: 0000H 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 ISS 15 ISS 14 ISS 13 ISS 12 ISS 11 ISS 10 ISS 9 ISS 8 ISS 7 ISS 6 ISS 5 ISS 4 ISS 3 ISS 2 ISS 1 ISS 0 rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw Field Bits Type Description ISS0 0 rw Interrupt Source Select for CC2_CC16IC 0B CC2 channel 16 interrupt is selected 1B Reserved, do not use this combination ISS1 1 rw Interrupt Source Select for CC2_CC17IC 0B CC2 channel 17 interrupt is selected Reserved, do not use this combination 1B ISS2 2 rw Interrupt Source Select for CC2_CC18IC 0B CC2 channel 18 interrupt is selected 1B USIC3 channel 0 SR3 is selected ISS3 3 rw Interrupt Source Select for CC2_CC19IC 0B CC2 channel 19 interrupt is selected 1B USIC3 channel 1 SR3 is selected ISS4 4 rw Interrupt Source Select for CC2_CC20IC 0B CC2 channel 20 interrupt is selected 1B USIC0 channel 0 SR3 is selected ISS5 5 rw Interrupt Source Select for CC2_CC21IC 0B CC2 channel 21 interrupt is selected 1B USIC0 channel 1 SR3 is selected ISS6 6 rw Interrupt Source Select for CC2_CC22IC 0B CC2 channel 22 interrupt is selected 1 B USIC1 channel 0 SR3 is selected ISS7 7 rw Interrupt Source Select for CC2_CC23IC 0B CC2 channel 23 interrupt is selected 1B USIC1 channel 1 SR3 is selected User’s Manual Interrupts, V3.2 7-56 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Field Bits Type Description ISS8 8 rw Interrupt Source Select for CC2_CC24IC 0B CC2 channel 24 interrupt is selected 1B Reserved, do not use this combination ISS9 9 rw Interrupt Source Select for CC2_CC25IC CC2 channel 25 interrupt is selected 0B Reserved, do not use this combination 1B ISS10 10 rw Interrupt Source Select for CC2_CC26IC 0B CC2 channel 26 interrupt is selected 1B Reserved, do not use this combination ISS11 11 rw Interrupt Source Select for CC2_CC27IC 0B CC2 channel 27 interrupt is selected 1B Reserved, do not use this combination ISS12 12 rw Interrupt Source Select for CC2_CC28IC 0B CC2 channel 28 interrupt is selected 1B USIC2 channel 0 SR3 is selected ISS13 13 rw Interrupt Source Select for CC2_CC29IC 0B CC2 channel 29 interrupt is selected 1B USIC2 channel 1 SR3 is selected ISS14 14 rw Interrupt Source Select for CC2_CC30IC 0B CC2 channel 30 interrupt is selected SCU Interrupt 2 is selected 1B ISS15 15 rw Interrupt Source Select for CC2_CC31IC 0B CC2 channel 31 interrupt is selected 1B SCU Interrupt 3 is selected User’s Manual Interrupts, V3.2 7-57 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control 7.15 Interrupt and PEC Configuration Registers The following table lists all registers used to configure the interrupt and PEC behavior of the XC27x5X. Registers are ordered by address. The Interrupt Control registers xxIC, assigned to each interrupt request, are listed separately (see Section 7.14). Bit addressable SFRs are marked with the letter “b” in column “Name”. Table 7-11 Register Overview Interrupt and PEC - ordered by address Name Physical Address 8-bit Description Address Reset Value FINT0CSP EC00H -- Fast Interrupt 0 CSP Register 0000H FINT0ADDR EC02H -- Fast Interrupt 0 Address Register 0000H FINT1CSP EC04H -- Fast Interrupt 1 CSP Register 0000H FINT1ADDR EC06H -- Fast Interrupt 1 Address Register 0000H BNKSEL0 EC20H -- Bank Selection Register 0 0000H BNKSEL1 EC22H -- Bank Selection Register 1 0000H BNKSEL2 EC24H -- Bank Selection Register 2 0000H BNKSEL3 EC26H -- Bank Selection Register 3 0000H SRCP0 EC40H -- PEC Channel 0 Source Pointer 0000H DSTP0 EC42H -- PEC Channel 0 Destination Pointer 0000H SRCP1 EC44H -- PEC Channel 1 Source Pointer 0000H DSTP1 EC46H -- PEC Channel 1 Destination Pointer 0000H SRCP2 EC48H -- PEC Channel 2 Source Pointer 0000H DSTP2 EC4AH -- PEC Channel 2 Destination Pointer 0000H SRCP3 EC4CH -- PEC Channel 3 Source Pointer 0000H DSTP3 EC4EH -- PEC Channel 3 Destination Pointer 0000H SRCP4 EC50H -- PEC Channel 4 Source Pointer 0000H DSTP4 EC52H -- PEC Channel 4 Destination Pointer 0000H SRCP5 EC54H -- PEC Channel 5 Source Pointer 0000H DSTP5 EC56H -- PEC Channel 5 Destination Pointer 0000H SRCP6 EC58H -- PEC Channel 6 Source Pointer 0000H DSTP6 EC5AH -- PEC Channel 6 Destination Pointer 0000H SRCP7 EC5CH -- PEC Channel 7 Source Pointer 0000H DSTP7 EC5EH -- PEC Channel 7 Destination Pointer 0000H User’s Manual Interrupts, V3.2 7-58 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Interrupt and Exception Control Table 7-11 Register Overview Interrupt and PEC - ordered by address Name Physical Address 8-bit Description Address Reset Value PECSEG0 EC80H -- PEC Pointer 0 Segment Address 0000H PECSEG1 EC82H -- PEC Pointer 1 Segment Address 0000H PECSEG2 EC84H -- PEC Pointer 2 Segment Address 0000H PECSEG3 EC86H -- PEC Pointer 3 Segment Address 0000H PECSEG4 EC88H -- PEC Pointer 4 Segment Address 0000H PECSEG5 EC8AH -- PEC Pointer 5 Segment Address 0000H PECSEG6 EC8CH -- PEC Pointer 6 Segment Address 0000H PECSEG7 EC8EH -- PEC Pointer 7 Segment Address 0000H PECISNC b FFD8H ECH PEC Interrupt Subnode Control 0000H PECC0 FEC0H 60H PEC Channel 0 Control Register 0000H PECC1 FEC2H 61H PEC Channel 1 Control Register 0000H PECC2 FEC4H 62H PEC Channel 2 Control Register 0000H PECC3 FEC6H 63H PEC Channel 3 Control Register 0000H PECC4 FEC8H 64H PEC Channel 4 Control Register 0000H PECC5 FECAH 65H PEC Channel 5 Control Register 0000H PECC6 FECCH 66H PEC Channel 6 Control Register 0000H PECC7 FECEH 67H PEC Channel 7 Control Register 0000H User’s Manual Interrupts, V3.2 7-59 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8 System Control Unit (SCU) The System Control Unit (SCU) of the XC27x5X handles all system control tasks besides the debug related tasks which are controlled by the OCDS/Cerberus. All functions described in this chapter are tightly coupled, thus, they are conveniently handled by one unit, the SCU. The SCU contains the following functional sub-blocks: • • • • • • • • • • • • • • • • • Clock Generation (see Chapter 8.1) System Timer (see Chapter 8.2) Wake-up Timer (see Chapter 8.3) Reset Operation (see Chapter 8.4) External Service Requests (see Chapter 8.5) Power Supply and Control (see Chapter 8.6) Global State Control (see Chapter 8.7) Software Boot Support (see Chapter 8.8) External Request Unit (see Chapter 8.9) Interrupt Generation (see Chapter 8.10) Temperature Compensation (see Chapter 8.11) Watchdog Timer (see Chapter 8.12) Trap Generation (see Chapter 8.13) Memory Content Protection (see Chapter 8.14) Register Access Control (see Chapter 8.15) Miscellaneous System Registers (see Chapter 8.16) SCU Registers and Address map (see Chapter 8.17) Important Information: Register Programming The System Control Unit contains special function registers, which can not be programmed in an arbitrary order in particular due to the usage of an internal voltage regulator. In order to prevent critical system conditions because of an improper setup and to provide means for easy and quick configuration and control of sensitve features such as power supply and clock generation, recommendations and examples for the programming sequence of the registers will be given in the Programmer’s Guide. In particular the registers listed below have to be updated with care: • • • Clock Generation Unit: WUOSCCON, HPOSCCON, PLLOSCCON, PLLCONx Power Supply: EVR1CON0, EVR1SET15VHP, EVRMCON0,EVRMSET15VHP, PVC1CON0, PVCMCON0, SWDCON0 System: SYSCON0 User’s Manual SCU, V1.15 8-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1 Clock Generation Unit The Clock Generation Unit (CGU) allows a very flexible clock generation for the XC27x5X. During user program execution the frequency can be programmed for an optimal ratio between performance and power consumption in the actual application state. 8.1.1 Overview The CGU can convert a low-frequency external clock to a high-speed system clock or can create a high-speed system clock without external input. The CGU consists of a Clock Generator and a Clock Control Unit (CCU). fOSC XTAL1 fWUT fWU XTAL2 Clock Generator fSYS to System fPLL fCLKIN1 CLKIN1 fSYS to fRTC RTC CCU fSTM fCLKIN2 CLKIN2 to WUT to STM EXTCLK Clock Generation Unit (CGU) CGU_Block_MR+_EXT .vsd Figure 8-1 Clock Generation Unit Block Diagram The input connections of the CGU are described in Chapter 8.18.1. The following clock signals are generated: • • • • • System clock fSYS RTC count clock fRTC ( Wake-Up Timer (WUT) clock fWUT STM clock fSTM External clock fEXT Chapter 8.1.5 and Chapter 8.1.6 describe which clock signals is generated out of which selectable clocks. User’s Manual SCU, V1.15 8-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Register Overview The CGU is controlled by a number of registers shown in the following figure. Oscillator Control PLL Control System Control Output Control WUOSCCON PLLSTAT SYSCON0 RTCCLKCON HPOSCCON STATCLR1 STATCLR 0 EXTCON PLLOSCCON PLLCON 0 PLLCON 1 PLLCON 2 PLLCON 3 WUOSCCON HPOSCCON PLLOSCCON PLLSTAT STATCLR 1 PLLCON 0 PLLCON 1 PLLCON 2 Wake-up OSC Control Register High Precision OSC Control Register PLL OSC Configuration Register PLL Status Register PLL Status Clear 1 Register PLL Configuration 0 Register PLL Configuration 1 Register PLL Configuration 2 Register PLLCON 3 PLL Configuration 3 Register SYSCON 0 System Control 0 Register STATCLR 0 Status Clear 0 Register RTCCLKCON EXTCON Figure 8-2 RTC Clock Control Register External Clock Control Register CGU_Register _Overview .vsd Clock Generation Unit Register Overview The following sections describe the different parts of the CGU. User’s Manual SCU, V1.15 8-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.2 Trimmed Current Controlled Wake-Up Clock (OSC_WU) The trimmed current controlled wake-up clock source provides a clock to control internal operations independent of the standard clock supplies and requires no external components. Its output frequency fWU is configured via bit field WUOSCCON.FREQSEL and has a typical range from 130 kHz to 500 kHz. 8.1.3 High Precision Oscillator Circuit (OSC_HP) The high precision oscillator circuit can drive an external crystal or accepts an external clock source. It consists of an inverting amplifier with XTAL1 as input, and XTAL2 as output. Figure 8-4 and Figure 8-3 show the recommended external circuitries for both operating modes, External Crystal Mode and External Input Clock Mode. 8.1.3.1 External Input Clock Mode An external clock signal is supplied directly not using an external crystal and bypassing the amplifier of the oscillator. The maximum allowed input frequency depends on the characteristics of pin XTAL1. When using an external clock signal it must be connected to XTAL1. XTAL2 is left open (unconnected). Note: Voltages on XTAL1 must comply to the voltage defined in the data sheet. XTAL1 External Clock Signal XTAL2 OSC_HP fOSC leave unconnected VSS CGU_OSC _HP_ExtIn.vsd Figure 8-3 8.1.3.2 XC27x5X External Clock Input Mode for the High-Precision Oscillator External Crystal Mode An external oscillator load circuitry must be used, connected to both pins, XTAL1 and XTAL2. It consists normally of the two load capacitances C1 and C2. For some crystals User’s Manual SCU, V1.15 8-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) a series damping resistor might be necessary. The exact values and related operating range depend on the crystal and have to be determined and optimized together with the crystal vendor using the negative resistance method. Fundamental Mode Crystal XTAL1 OSC_HP XTAL2 C1 fOSC C2 VSS CGU_OSC _HP_Crystal.vsd Figure 8-4 XC27x5X External Crystal Mode Circuitry for the High-Precision Oscillator Support for Start-up Control of an External Crystal The first time before the system clock is generated based on an external crystal 1000 cycles of the crystal clock should be waited before the clock control system is changed to External Crystal Mode. Bit PLLSTAT.OSCLOCK indicates if the oscillator OSC_HP operates for at least 211 periods. Bit PLLSTAT.OSCSTAB indicates if OSC_HP operates for at least 215 periods. Oscillator Gain Control The oscillator starts with a high drive level (gain) during and after a Power-on Reset to ensure safe start-up behavior in the beginning (force the crystal oscillation). When a stable oscillation has been reached after oscillation start-up (PLLSTAT.OSCSTAB = 1), the gain of the oscillator can be reduced. This reduces the power consumption of the oscillator, which is especially important in the power saving modes. This gain reduction is selected by HPOSCCON.GAINSEL. Note: Choosing the gain setting is only possible with detailed consideration of parasitics, external circuitry, frequency range and quality of the applied crystal and has to be verified by testing togeher with the crystal manufacturer. User’s Manual SCU, V1.15 8-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.4 Phase-Locked Loop (PLL) Module The PLL can convert a low-frequency external clock signal to a high-speed system clock for maximum performance. The PLL also has fail-safe logic that detects degenerate external clock behavior such as abnormal frequency deviations or a total loss of the external clock. It can execute emergency actions if it loses its lock on the external clock. This module is a phase locked loop for integer frequency synthesis. It allows the use of input and output frequencies of a wide range by varying the different divider factors. 8.1.4.1 Features Here is a brief overview of the functions that are offered by the PLL. • • • • • • • • • • • VCO lock detection 4-bit input divider P: (divide by PDIV+1) 6-bit feedback divider N: (multiply by NDIV+1) 10-bit output divider K2: (divide K2DIV+1) 10-bit VCO bypass divider K1: (divide by either by K1DIV+1) Oscillator run detection and Watchdog Different operating modes – Prescaler Mode – Unlocked Mode – Normal Mode Different power saving modes – Power Down – Sleep Mode (VCO Power Down) Glitchless programming of output divider K2 and VCO bypass divider K1 Glitchless switching between Normal Mode and Prescaler Mode Trimmed current controlled clock source 8.1.4.2 PLL Functional Description The PLL consists of a Voltage Controlled Oscillator (VCO) with a feedback path. A divider in the feedback path (N-Divider) divides the VCO frequency. The resulting frequency is then compared with the divided external frequency (P-Divider). The phase detection logic determines the difference between the two clocks and accordingly controls the frequency of the VCO (fVCO). A PLL lock detection unit monitors and signals this condition. The phase detection logic continues to monitor the two clocks and adjusts the VCO clock if required. The PLL output clock fPLL is derived from the VCO clock using the K2-Divider or from the oscillator clockusing the K1-Divider. The following figure shows the PLL block structure. User’s Manual SCU, V1.15 8-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON 1. OSCSEL 0 fIN 1 PLLSTAT . FINDIS PLLCON0. VCOBY K1Divider M U X fINT fR fP PDivider fREF Clock Lock Source Detect. VCO Core 1 K20 Divider fK2 M U X fPLL NfDIV Divider Osc. WDG PLL Block HPOSCCON. HPOSCCON. PLLCON1. PLLSTAT . OSCWDTRST PLLV RESLD VCOLOCK Figure 8-5 fVCO fK1 PLL _Block _Diagram.vsd PLL Block Diagram Clock Source Control The reference frequency fR can be selected to be either taken from the trimmed current controlled clock source fINT or from an external clock source fIN. PLL Modes The PLL clock fPLL is generated from fR in one of the following software selectable modes: • • • Normal Mode Prescaler Mode Unlocked Mode In Normal Mode the reference frequency fR is divided by a factor P, multiplied by a factor N and then divided by a factor K2. The output frequency is given by (8.1) N f PLL = P-------- ⋅ fR ⋅ K2 User’s Manual SCU, V1.15 8-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) In Prescaler Mode the reference frequency fR is divided by a factor K1. The output frequency is given by (8.2) fR f PLL = --K1 In Unlocked Mode the base output frequency of the Voltage Controlled Oscillator (VCO) fVCObase is divided by a factor K2. The output frequency is given by (8.3) f VCObase f PLL = --------------K2 PLL Power Saving Modes PLL Power Down Mode The PLL offers a Power Down Mode to save power if the PLL is not needed at all. While the PLL is in Power Down Mode no PLL output frequency is generated. PLL Sleep Mode The PLL offers a Sleep Mode (also called VCO Power Down Mode) to save power within the PLL. While the PLL is in Sleep Mode only the Prescaler Mode can be used. 8.1.4.3 Configuration and Operation of the PLL Modes The following section describes the configuration and the operation of the different PLL modes. Further information can be found in the Programmer’s Guide. Configuration and Operation of the Unlocked Mode In Unlocked Mode, the PLL is running at its VCO base frequency and fPLL is derived from fVCO by the K2-Divider. User’s Manual SCU, V1.15 8-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLSTAT . FINDIS = 1 PLLCON0. VCOBY 1 ´0´ Lock Detect. fREF fDIV VCO Core fVCO 0 K2f Divider K2 M U X fPLL NDivider PLL Block PLL_Unlocked_Mode. vsd Figure 8-6 PLL Unlocked Mode Diagram The Unlocked Mode is selected by the following settings: • • STATCLR1.SETFINDIS = 1 PLLCON0.VCOBY = 0 The Unlocked Mode is entered when all following conditions are true: • • PLLSTAT.FINDIS = 1 PLLSTAT.VCOBYST = 1 Operation in Unlocked Mode does not require an input clock fIN. The Unlocked Mode is automatically entered on a PLL VCO Loss-of-Lock event if bit PLLCON1.EMFINDISEN is cleared. This mechanism allows a fail-safe operation of the PLL as in emergency cases still a clock is available. The frequency of the Unlocked Mode fVCObase is listed in the Data Sheet. Note: Changing the system operation frequency by changing the value of the K2-Divider or the VCO range has a direct influence on the power consumption of the device. Therefore, this has to be done carefully. Configuration and Operation of the Normal Mode In Normal Mode, the PLL is running at frequency fPLL, where fR is divided by a factor P, multiplied by a factor N and then divided by a factor K2. User’s Manual SCU, V1.15 8-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON 1. OSCSEL fIN 0 1 M U X fINT fR PLLSTAT . FINDIS fP PDivider Clock Lock Source Detect. Osc. WDG HPOSCCON. HPOSCCON. PLLCON1. OSCWDTRST PLLV RESLD Figure 8-7 fREF fDIV PLLCON0. VCOBY 1 VCO Core fVCO K20 Divider f K2 M U X fPLL NDivider PLL Block PLLSTAT . VCOLOCK PLL _Normal_Mode.vsd PLL Normal Mode Diagram The Normal Mode is selected by the following settings: • • PLLCON0.VCOBY = 0 STATCLR1.CLRFINDIS = 1 The Normal Mode is entered when all following conditions are true: • • • • PLLSTAT.FINDIS = 0 PLLSTAT.VCOBYST = 1 PLLSTAT.VCOLOCK = 1 HPOSCCON.PLLV = 1 Operation in Normal Mode requires a clock frequency of fR. When fIN is selected as source for fR it is recommended to check and monitor if an input frequency fR is available at all by checking HPOSCCON.PLLV. The system operation frequency in Normal Mode is controlled by the values of the three dividers: P, N, and K2. A modification of the two dividers P and N has a direct influence on the VCO frequency and leads to a loss of the VCO Lock status. A modification of the K2-divider has no impact on the VCO Lock status but changes the PLL output frequency. Note: Changing the system operation frequency by changing the value of the K2-Divider has a direct influence on the power consumption of the device. Therefore, this has to be done carefully. To modify or enter the Normal Mode frequency, follow the sequence described below: Configure and enter Prescaler Mode. For more details see the Prescaler Mode. Disable the trap generation for the VCO Lost-of-Lock. User’s Manual SCU, V1.15 8-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) While the Prescaler Mode is used the Normal Mode can be configured and checked for a positive VCO Lock status. The first target frequency of the Normal Mode should be selected in a way that it matches or is only slightly higher as the one used in the Prescaler Mode. This avoids big changes in the system operation frequency and, therefore, the power consumption when switching later from Prescaler Mode to Normal Mode. The P and N dividers should be selected in the following way: • • Selecting P and N in a way that fVCO is in the lower area of its allowed values leads to a slightly reduced power consumption but to a slightly increased jitter Selecting P and N in a way that fVCO is in the upper area of its allowed values leads to a slightly increased power consumption but to a slightly reduced jitter After the P, and N dividers are updated for the first configuration, the indication of the VCO Lock status (PLLSTAT.VCOLOCK = 1) should be awaited. Note: It is recommended to reset the VCO Lock detection (PLLCON1.RESLD = 1) after the new values of the dividers have been configured to get a defined VCO lock check time. When this happens the switch from Prescaler Mode to Normal Mode can be done. Normal Mode is requested by clearing PLLCON0.VCOBY. The Normal Mode is entered when the status bit PLLSTAT.VCOBYST is set. Now the Normal Mode is entered. The trap status flag for the VCO Lock trap should be cleared and then enabled again. The intended PLL output target frequency can be configured by changing only the K2Divider. Depending on the selected divider value of the K2-Divider, the duty cycle of the clock is selected. This can have an impact on the operation with an external communication interface. In order to avoid too big frequency changes it might be neccessary to change the K2-Divider in multiple steps. When the value of the K2-Diver was changed the next update of this value should not be done before bit PLLSTAT.K2RDY is set. Note: The Programmers’s Guide describes a smooth frequency stepping to achieve an appropriate load regulation of the internal voltage regulator. PLL VCO Lock Detection The PLL has a lock detection that supervises the VCO part of the PLL in order to detect instable VCO circuit behavior. The lock detector marks the VCO circuit and therefore the output fVCO of the VCO as instable if the two inputs fREF and fDIV differ too much. Changes in one or both input frequencies below a level are not marked by a loss of lock because the VCO can handle such small changes without any problem for the system. PLL VCO Loss-of-Lock Event The PLL may become unlocked, caused by a break of the crystal or the external clock line. In such a case, a trap is generated if the according trap is enabled. Additionally, the clock fR is disconnected from the PLL VCO to avoid unstable operation due to noise or User’s Manual SCU, V1.15 8-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) sporadic clock pulses coming from the oscillator circuit. Without a clock input fR, the PLL gradually slows down to its VCO base frequency and remains there. The automatic disconnection of the VCO from its input clock fR in case of a VCO Loss-of-Lock event can be enabled by setting bit PLLCON1.EMFINDISEN. If this bit is cleared the clock fR remains connected to the VCO. Configuration and Operation of the Prescaler Mode In Prescaler Mode, the PLL is running at frequency fPLL, where fR is divided by the K1Divider. PLLCON 1. OSCSEL fIN 0 1 PLLCON0. VCOBY K1Divider M U X fINT fR 1 0 M U X fPLL Clock Source Osc. WDG PLL Block HPOSCCON. HPOSCCON. OSCWDTRST PLLV Figure 8-8 fK1 PLL _Prescaler _Mode.vsd PLL Prescaler Mode Diagram The Prescaler Mode is selected by the following setting: • PLLCON0.VCOBY = 1 The Prescaler Mode is entered when all following conditions are true: • • PLLSTAT.VCOBYST = 0 HPOSCCON.PLLV = 1 Operation in Prescaler Mode requires an input clock frequency fR. If fIN is selected as clock source for fR it is recommended to check and monitor if an input frequency fOSC is available at all by checking HPOSCCON.PLLV. There are no requirements regarding the frequency of fR. The system operation frequency in Prescaler Mode is controlled by the value of the K1Divider. When the value of PLLCON1.K1DIV was changed the next update of this value should not be done before bit PLLSTAT.K1RDY is set. User’s Manual SCU, V1.15 8-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Note: Changing the system operation frequency by changing the value of the K1-Divider has a direct influence on the power consumption of the device. Therefore, this has to be done carefully. The duty cycle of the clock signal depends on the selected value of the K1-Divider. This can have an impact for the operation with an external communication interface. The Prescaler Mode is requested from the Unlocked or Normal Mode by setting bit PLLCON0.VCOBY. The Prescaler Mode is entered when the status bit PLLSTAT.VCOBYST is cleared. Before the Prescaler Mode is requested the K1-Divider should be configured with a value generating a PLL output frequency fPLL that matches the one generated by the Unlocked or Normal Mode as much as possible. In this way the frequency change resulting out of the mode change is reduced to a minimum. The Prescaler Mode is requested to be left by clearing bit PLLCON0.VCOBY. The Prescaler Mode is left when the status bit PLLSTAT.VCOBYST is set. Configuration and Operation of the PLL Power Down Mode The Power Down Mode is entered by setting bit PLLCON0.PLLPWD. While the PLL is in Power Down Mode no PLL output frequency is generated. Configuration and Operation of the PLL Sleep Mode The Sleep Mode (also called VCO Power Down Mode) is entered by setting bit PLLCON0.VCOPWD. While the PLL is in Sleep Mode only the Prescaler Mode is operable. Selecting the Sleep Mode does not automatically switch to the Prescaler Mode. Therefore, before the Sleep Mode is entered the Prescaler Mode must be active. 8.1.4.4 Power Regulator The analog parts of the PLL (VCO, trimmed current controlled clock source) are running on a dedicated supply generated by a dedicated regulator integrated within the PLL unit. The regulator has to be enabled separately before the analog blocks of the PLL are activated, i.e. trimmed currenct controlled clock source and VCO must be kept off until the supply is stable. After activation, the PLL regulator will need its ramp-up time to properly ramp-up the analog PLL supply. When the regulator shall be disabled in conjunction with a power down of the PLL digital part, it has to be taken into account that the digital part needs an active clock at the output of the PLL to ramp down. In case this clock is generated by one of the PLL oscillators, power down of PLL must be entered before the regulator is disabled. VCO and trimmed currenct controlled clock source may be activated or switched off together. User’s Manual SCU, V1.15 8-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.4.5 Divider Handshake The PLL provides several handshake interfaces for dividers. This section describes how a handshake is to be conducted upon a change of configuration. The general conduction of the handshake is the same for all interfaces. However, a sample sequence is described here in conjunction with re-programming of a divider. Note: The described handshake only works if the new setting (e.g. divider value) changes the current value upon the handshake. The handshake should be done in the the following steps: 1. 2. 3. 4. Clear acknowledge bit together with setting the new divider value Poll on ready bit to be 0 Set acknowledge bit Poll on ready bit to be 1 This approach will even work in case the handshake has not been properly served before, and ready is already at 0 from the beginning. In any case, a change of the divider value will set ready to 0. 8.1.4.6 Trimmed Current Controlled Clock The trimmed current controlled clock source provides a clock fINT for the PLL. This is configured via bit PLLCON1.OSCSEL. Note: The clock fINT is also required for the operation of the oscillator watchdog. 8.1.4.7 Oscillator Watchdog The oscillator watchdog continuously monitors the input clock fIN. If the input frequency becomes too low or if the input clock fails, this oscillator fail condition is indicated by HPOSCCON.PLLV = 0 and an interrupt request is generated. By setting bit HPOSCCON.OSCWDTRST the detection can be restarted without a reset of the complete PLL, e.g. in case of a VCO loss-of-lock condition. Note: The oscillator watchdog requires the trimmed currenct controlled clock fINT as a reference. Therefore, it can only be used (HPOSCCON.PLLV is valid) while the clock source is active. User’s Manual SCU, V1.15 8-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.4.8 Switching PLL Parameters The following restriction applies when changing PLL parameters inside the PLLCON0 to PLLCON3 registers: • • • • • The VCO bypass switch may be used at any time, however, it has to be ensured that the maximum operating frequency of the device (see data sheet) will not be exceeded. Prescaler Mode should be selected. After switching to Prescaler Mode, NDIV and PDIV can be adjusted. Before deselecting the Prescaler Mode, the RESLD bit has to be set and then the VCOLOCK flag has to be checked. Only when the VCOLOCK flag is set again, the Prescaler Mode may be deselected. Before changing VCOSEL, the Prescaler Mode must be selected. Note: PDIV and NDIV can also be switched in Normal Mode. When changing NDIV, it must be regarded that the VCO clock fVCO may exceed the target frequency until the PLL becomes locked. After changing PDIV or NDIV, it must be waited for the PLL lock condition. This procedure is typically used for increasing the VCO clock step-by-step. User’s Manual SCU, V1.15 8-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.5 Clock Control Unit The Clock Control Unit (CCU) selects the current clock sources for the clock signals used in the XC27x5X. It generates the following clocks: • • • • • System clock fSYS RTC count clock fRTC WUT clock fWUT System timer clock fSTM Output clock fEXT The following clock signals can be selected: • • • • • PLL clock fPLL The oscillator clock (OSC_HP) fOSC Wake-up clock fWU Input CLKIN1 as Direct Clock Input fCLKIN1 Input CLKIN2 as Direct Clock Input fCLKIN2 8.1.5.1 Clock Generation Different clock sources can be selected for the generated clock signals. Note: The selected clock sources are affected by the start-up procedure. See chapter Device Status after Start-up for the register values set by the different start-up procedures. System Clock Generation The system clock fSYS can be selected from the following clock sources in the CCU: • • • • Wake-up clock fWU The oscillator clock (OSC_HP) fOSC PLL clock fPLL Input CLKIN1 as Direct Clock Input fCLKIN1 User’s Manual SCU, V1.15 8-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) SYSCON 0. CLKSEL fWU 00 fOSC 01 fPLL 10 fCLKIN1 11 M U X OSCWDT Emergency Event Master Clock Multiplexer (MCM) 10 System Clock Selection SYSCON 0.EMCLKSELEN PLLCON 1.EMCLKEN HPOSCCON.EMCLKEN 0 1 00 01 VCOLCK Emergency Event M U X Emergency Clock M U X fSYS to EXTCLK selection 11 SYSCON 0. EMCLKSEL Figure 8-9 User’s Manual SCU, V1.15 CCU_SYSCLK _EXT .vsd Clock Control Unit, System Clock Generation 8-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) RTC Clock Generation For the RTC module it is possible to select the operation in synchronous or asynchronous mode in the module itself. The asynchronous clock for the RTC can be selected out of following clock sources in the CCU: • • • • PLL clock fPLL The oscillator clock (OSC_HP) fOSC Input CLKIN2 as Direct Clock Input fCLKIN2 Wake-up clock fWU RTCCLKCON. RTCCLKSEL fPLL 00 fOSC 01 fWU 10 fCLKIN2 11 RTCCLKCON. RTCCM M U X fRTC 0 1 M U X to RTC block fSYS CCU CCU_RTCCLK_EXT.vsd Figure 8-10 Clock Control Unit, RTC Clock Generation User’s Manual SCU, V1.15 8-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) System Timer (STM) Clock Generation The system timer clock can be selected out of following clock sources: • • • • • The Direct Clock from oscillator OSC_HP fOSC PLL clock fPLL Input CLKIN1 as Direct Clock Input fCLKIN1 Input CLKIN2 as Direct Clock Input fCLKIN2 Wake-Up Oscillator fWU Then the selected clock can be divided by the factor defined in STMCON.CLKDIV (see Chapter 8.2.1.2). SYSCON 0. STMCLKSEL fCLKIN2 000 fOSC 001 fPLL 010 fCLKIN1 011 fWU 100 Reserved 101 Reserved 110 Reserved 111 CCU CLKDIV fSTM to STM M U X STMCON. CLKDIV CCU_STMCLK_EXT .vsd Figure 8-11 Clock Control Unit, STM Clock Generation User’s Manual SCU, V1.15 8-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Wake-up Timer (WUT) Clock Generation The wake-up timer clock can be selected out of following clock sources in the CCU: • • • • Wake-Up Oscillator fWU The Direct Clock from oscillator OSC_HP fOSC PLL clock fPLL Input CLKIN1 as Direct Clock Input fCLKIN1 Then the selected clock can be divided by the factor defined in WUCR.CLKDIV (see Chapter 8.3.2.2). SYSCON 0. WUTCLKSEL fWU 00 fOSC 01 fPLL 10 fCLKIN1 11 CCU M U X fWUT CLKDIV WUCR. CLKDIV to WUT CCU_WUTCLK_EXT .vsd Figure 8-12 Clock Control Unit, WUT Clock Generation 8.1.5.2 Selecting and Changing the Operating Frequency When selecting the clock source and the clock generation method, the required parameters must be carefully written to the respective bit fields, to avoid unintended intermediate states. Many applications change the frequency of the system clock fSYS during operation to optimize performance and power consumption of the system. Modifying the operating frequency changes the consumed switching current, which influences the power supply. Therefore, while the core voltage is generated by the on-chip Embedded Voltage Regulators (EVRs), the operating frequency may only be changed according to the rules given in the data sheet. Note: To avoid the indicated problems, specific sequences are recommened that ensure the intended operation of the clock system interacting with the power system. Please refer to the document “Programmer’s Guide”. User’s Manual SCU, V1.15 8-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.5.3 System Clock Emergency Handling The generation of the system clock fSYS can be affected, if either the PLL is no more locked to its input signal fIN, or if the input clock fIN is no more active. Both events can be detected and are indicated to the application software. The clock system takes appropriate actions where necessary, so the device and the application is never left without an alternate clock signal. Oscillator Watchdog Event If the clock frequency of the external source drops below a limit value the oscillator watchdog (OSCWDT) (see Chapter 8.1.4.7) then the clock source for the system clock fSYS is switched to an alternate clock source, if enabled (HPOSCCON.EMCLKEN = 1). In this case following information is available: • • • • The oscillator watchdog trap flag (TRAPSTAT.OSCWDTT) is set and a trap request to the CPU is activated, if enabled (TRAPDIS.OSCWDTT = 0). Bit HPOSCCON.PLLV = 0, while the clock fIN is missing Bit SYSCON0.EMSOSC is set, if SYSCON0.EMCLKSELEN is set The source of the system clock fSYS is switched to alternate clock source selected by SYSCON0.EMCLKSEL, if enabled (SYSCON0.EMCLKSELEN = 1). This is indicated by bit SYSCON0.SELSTAT = 1. PLL VCO Loss-of-Lock Event If the PLL output frequency is no longer locked to its input frequency fIN, the PLL switches from PLL Normal mode to the Unlocked mode, if enabled (PLLCON1.EMFINDISEN = 1). In this case following information is available: • • • • The PLL VCO loss of lock trap flag (TRAPSTAT.VCOLCKT) is set and a trap request to the CPU is activated, if enabled (TRAPDIS.VCOLCKT = 0). Bit PLLSTAT.VCOLOCK = 0, while the PLL is not locked Bit SYSCON0.EMSVCO is set, if SYSCON0.EMCLKSELEN is set The PLL VCO clock input is disconnected (PLLSTAT.FINDIS = 1) and the PLL clock slows down to its VCO base frequency. System Behavior Emergency routines can be executed with the alternate clock (emergency clock or VCO base frequency). The application can then enter a safe status and stop operation, or it can switch to an emergency operating mode, where a reduced performance and/or feature set is provided. The Programmer’s Guide describes both, how to enable these features, and how to react properly on each of the two events. User’s Manual SCU, V1.15 8-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.6 External Clock Output An external clock output can be provided via pin EXTCLK to clock an external system or to observe one of the selectable device clocks. This external clock is enabled by setting bit EXTCON.EN and by selecting the clock signal as alternate output function at pin EXTCLK. Following clocks can be selected by EXTCON.SEL for external clock fEXT: • • • • • • • System clock fSYS Programmable clock output fOUT Direct Clock from oscillator OSC_HP fOSC Direct Clock Input fCLKIN1 PLL clock fPLL Wake-up clock fWU RTC clock fRTC Note: Changing bit field EXTCON.SEL can lead to spikes at pin EXTCLK. EXTCON .SEL 0000 fSYS Reload fOUT Counter 0001 fPLL 0010 fOSC 0011 fWU 0100 fCLKIN1 0101 Reserved 0110 Reserved 0111 EXTCON.EN '0' M U X 0 1 M U X EXTCLK fEXT 1000 fRTC '0' 1001 ... 1111 CCU_EXTCLK.vsd Figure 8-13 EXTCLK Generation 8.1.6.1 Programmable Frequency Output The programmable frequency output fOUT can be selected as clock output (EXTCLK). This clock can be controlled via software, 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 XC27x5X can be run at a scalable frequency or can temporarily be left without a clock. User’s Manual SCU, V1.15 8-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Clock fOUT is generated via a reload counter, so the output frequency can be selected in small steps. EXTCON.FORV EXTCON. FOEN Ctrl. Reload 1 fSYS Counter 0 FOTL M U X fOUT EXTCON.FOSS EXTCON.FOTL Reload Counter CCU_EXTCLK_Counter.vsd Figure 8-14 Programmable Frequency Output Generation fOUT always provides complete output periods (provided fSYS is available): • When fOUT is started (EXTCON.FOEN is set) counter FOCNT is loaded from • EXTCON.FORV When OUT is stopped (EXTCON.FOEN is cleared) counter FOCNT is stopped when fOUT has reached (or is) ’0’. Register EXTCON provides control over the output generation (frequency, waveform, activation) as well as all status information (EXTCON.FOTL). User’s Manual SCU, V1.15 8-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) f SYS 1) f OUT (FORV = 0) 2) 1) f OUT (FORV = 2) 2) 1) f OUT (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 8-15 Output Waveforms Examples Note: The output (for EXTCON.FOSS= 1) is high for the duration of one fSYS cycle for all reload values EXTCON.FORV > 0. For EXTCON.FORV = 0 the output frequency corresponds to fSYS. When a reference clock is required (e.g. for the bus interface), fSYS must be selected directly. User’s Manual SCU, V1.15 8-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7 CGU Registers 8.1.7.1 Wake-up Clock Register This register controls the settings of OSC_WU. WUOSCCON Wake-up OSC Control Register ESFR (F1AEH/D7H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 DIS 0 FREQSEL r rw rw rw Field Bits Type Description FREQSEL [1:0] rw Frequency Selection 00B fWU is approximately 500 kHz 01B fWU is approximately 300 kHz 10B fWU is approximately 200 kHz 11B fWU is approximately 130 kHz Note: This value must not be changed while fWU is used as clock source for any logic. 0 [3:2] rw Reserved Must be written with reset value 00B. DIS 4 rw Clock Disable 0B The oscillator is switched on and the clock is enabled 1B The oscillator is swiched off and the clock is disabled 0 [15:5] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.2 High Precision Oscillator Register This register controls the setting of OSC_HP. HPOSCCON High Precision OSC Control Register ESFR (F1B4H/DAH) 15 14 13 0 r 12 11 10 9 8 7 EM OSC OSC EM FIN SH 2 2 CLK DIS BY L0 L1 EN EN rh rh rw rw rw 6 Reset Value: 053CH 5 4 X1D X1D GAINSEL EN rw rh rw 3 2 MODE rw 1 0 OSC PLL WDT V RST w rh Field Bits Type Description PLLV 0 rh Oscillator for PLL Valid Status Bit This bit indicates whether the frequency output of OSC_HP is usable. This is checked by the Oscillator Watchdog of the PLL. 0B The OSC_HP frequency is not usable. The frequency is below the limit. 1B The OSC_HP frequency is usable. The frequency is not below the limit. For more information see Chapter 8.1.4.7. OSCWDTRST 1 w Oscillator Watchdog Reset No action 0B 1B The Oscillator Watchdog of the PLL is reset and restarted MODE [3:2] rw Oscillator Mode 00B The oscillator is active (crystal or ext. input) 01B Reserved, do not use 10B Reserved, do not use 11B OSC_HP is disabled and in power-saving mode GAINSEL [5:4] rw Oscillator Gain Selection 00B Supply current is typically 300 µA (not tested) 01B Supply current is typically 530 µA (not tested) 10B Supply current is typically 450 µA (not tested) 11B Supply current is typically 610 µA (not tested) User’s Manual SCU, V1.15 8-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description X1D 6 rh XTAL1 Data Value This bit reflects the inverted level of pin XTAL1. This bit is sampled with fSYS while X1DEN is set. Note: Voltages on XTAL1 must comply to the voltage defined in the data sheet. X1DEN 7 rw XTAL1 Data Enable Bit X1D is not updated 0B 1B Bit X1D can be updated SHBY 8 rw Shaper Bypass The shaper forms a proper signal from the input signal. This bit must be 0 for proper operation. 0B The shaper is not bypassed 1B The shaper is bypassed EMCLKEN 9 rw OSCWDT Emergency System Clock Source Select Enable This bit requests the master clock multiplexer (MCM) to switch to an alternate clock (selected by bit field SYSCON0.EMCLKSEL) in an OSCWDT emergency case. 0B MCM remains controlled by SYSCON0.CLKSEL 1B MCM is controlled by SYSCON0.EMCLKSEL EMFINDISEN 10 rw Emergency Input Clock Disconnect Enable This bit defines if bit PLLSTAT.FINDIS is set in an OSCWDT emergency case. 0B No action 1B PLLSTAT.FINDIS is set in an emergency case Note: Please refer to the Programmer’s Guide for a description of the proper handling. OSC2L1 User’s Manual SCU, V1.15 11 rh OSC_HP Not Usable Frequency Event This sticky bit indicates if bit PLLV has been cleared since OSC2L1 has last been cleared (by writing 1 to bit STATCLR1.OSC2L1CLR). 0B No change of PLLV detected Bit PLLV has been cleared at least once 1B 8-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description OSC2L0 12 rh OSC_HP Usable Frequency Event This sticky bit indicates if bit PLLV has been set since OSC2L0 has last been cleared (by writing 1 to bit STATCLR1.OSC2L0CLR). 0B No change of PLLV detected 1B PLLV has been set at least once 0 [15:13] r User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.3 PLL Control Register This register controls the trimmed current controlled clock source. PLLOSCCON PLL OSC Control Register ESFR (F1B6H/DBH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 OSCTRIM OSC PD r rw rw Field Bits Type Description OSCPD 0 rw Clock Source Power Saving Mode 0B Trimmed current controlled clock source is active 1B Trimmed current controlled clock source is off OSCTRIM [9:1] rw Clock Source Trim Configuration This value is used to adjust the frequency range of the current controlled clock source. Do not change this value when writing to this register. 0 [15:10] r User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-29 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.4 PLL Registers These registers control the settings of the PLL. PLLSTAT PLL Status Register 15 14 13 OSC OSC LOC STA B K rh rh 0 r 12 ESFR (F0BCH/5EH) 11 10 9 8 7 6 Reset Value: 0000H 5 REG VCO VCO FIN K2 K1 N P STA L1 L0 DIS RDY RDY RDY RDY T rh rh rh rh rh rh rh rh 4 0 r 3 2 1 0 VCO OSC PWD VCO LOC SEL STA BY ST T ST K rh rh rh rh Field Bits Type Description VCOBYST 0 rh VCO Bypass Status 0B The PLL clock is derived from divider K1 (Prescaler Mode) 1B The PLL clock is derived from divider K2 (Normal / Unlocked Mode) Note: Coding of PLLCON0.VCOBY and VCOBYST are different. PWDSTAT 1 rh PLL Power-saving Mode Status 0B The PLL is operable The digital part of the PLL is disabled 1B OSCSELST 2 rh Oscillator Input Selection Status 0B External input clock source for the PLL (fIN) 1B Internal input clock source for the PLL VCOLOCK 3 rh PLL VCO Lock Status 0B The frequency difference of fREF and fDIV is greater than allowed. The PLL cannot lock. 1B The PLL clock fPLL is locked to fREF and is stable. Note: In case of a loss of lock, the VCO frequency fVCO approaches to the upper/lower boundary of the selected VCO band if the reference frequency is higher/lower than possible for locking. User’s Manual SCU, V1.15 8-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PRDY 5 rh P-Divider Ready Status 0B Bit field PLLCON1.PDIV has been changed, new K1 divider value not yet used. 1B The P-Divider operates with the value defined in bit field PLLCON1.PDIV. NRDY 6 rh N-Divider Ready Status 0B Bit field PLLCON0.NDIV has been changed, new K1 divider value not yet used. 1B The P-Divider operates with the value defined in bit field PLLCON0.NDIV. K1RDY 7 rh K1-Divider Ready Status Bit field PLLCON2.K1DIV has been changed, 0B new K1 divider value not yet used. 1B The K1-Divider operates with the value defined in bit field PLLCON2.K1DIV. K2RDY 8 rh K2-Divider Ready Status 0B Bit field PLLCON3.K2DIV has been changed, new K2 divider value not yet used. 1B The K2-Divider operates with the value defined in bit field PLLCON3.K2DIV. FINDIS 9 rh Input Clock Disconnect Select Status 0B The VCO is connected to the reference clock 1B The VCO is disconnected from the reference clock Note: Software can control this bit by writing 1 to bits SETFINDIS or CLRFINDIS in register STATCLR1. VCOL0 User’s Manual SCU, V1.15 10 rh VCO Lock Detection Lost Status This sticky bit indicates if bit VCOLOCK has been cleared since VCOL0 has last been cleared (by writing 1 to bit STATCLR1.VCOL0CLR). 0B No falling edge detected 1B PLLV has been cleared at least once (VCO lock was lost) 8-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description VCOL1 11 rh VCO Lock Detection Reached Status This sticky bit indicates if bit VCOLOCK has been set since VCOL1 has last been cleared (by writing 1 to bit STATCLR1.VCOL1CLR). 0B No rising edge detected 1B VCO lock was reached REGSTAT 12 rh PLL Power Regulator Status The PLL is powered by a separate internal regulator. 0B The PLL is not powered (off) The PLL is powered (operation possible) 1B Note: Software can control this bit by writing 1 to bits REGENSET or REGENCLR in register PLLCON0. OSCSTAB 14 rh OSC_HP Stable 0B The oscillator is starting up. None or less than 215 clock cycles have been counted 1B At least 215 clock cycles have been counted Note: This bit is cleared when HPOSCCON.MODE is 1XB. OSCLOCK 15 rh OSC_HP Lock 0B The oscillator is unlocked. None or less than 211 clock cycles have been counted. 1B The oscillator is locked. At least 211 clock cycles have been counted. Note: This bit is cleared when HPOSCCON.MODE is 1XB. 0 User’s Manual SCU, V1.15 4, 13 r Reserved Read as 0; should be written with 0. 8-32 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) STATCLR1 PLL Status Clear 1 Register 15 14 13 12 11 10 ESFR (F0E2H/71H) 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 CLR SET OSC OSC VCO VCO FIN FIN 2L0 2L1 L1 L0 DIS DIS CLR CLR CLR CLR w w w w w w 0 r Field Bits Type Description VCOL0CLR 0 w VCOL0 Clear Trigger No action 0B 1B Bit PLLSTAT.VCOL0 is cleared VCOL1CLR 1 w VCOL1 Clear Trigger 0B No action Bit PLLSTAT.VCOL1 is cleared 1B OSC2L1CLR 2 w OSC2L1 Clear Trigger 0B No action 1B Bits HPOSCCON.OSC2L1 is cleared OSC2L0CLR 3 w OSC2L0 Clear Trigger 0B No action 1B Bit HPOSCCON.OSC2L0 is cleared SETFINDIS 4 w Set Status Bit PLLSTAT.FINDIS 0B No action 1B Bit PLLSTAT.FINDIS is set. The VCO input clock is disconnected. CLRFINDIS 5 w Clear Status Bit PLLSTAT.FINDIS 0B No action 1B Bit PLLSTAT.FINDIS is cleared. The VCO input clock is connected. 0 [15:6] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 0 8-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON0 PLL Configuration 0 Register 15 14 13 12 11 10 ESFR (F1B8H/DCH) 9 8 7 6 N ACK 0 NDIV 0 IN SEL rw r rw r rw Reset Value: 1302H 5 4 3 2 1 0 REG REG VCO VCO EN EN VCOSEL PWD BY SET CLR w w rw rw rw Field Bits Type Description VCOBY 0 rw VCO Bypass Select divider K2 for PLL clock 0B (Normal / Unlocked Mode) 1B Select divider K1 for PLL clock (Prescaler Mode, i.e. VCO is bypassed) Bit PLLSTAT.VCOBYST shows the actually selected divider. Note: Coding of VCOBY and PLLSTAT.VCOBYST are different. VCOPWD 1 rw VCO Power Saving Mode 0B Normal behavior The VCO is put into a power saving mode and 1B can no longer be used. Only the Prescaler Mode is active if previously selected. VCOSEL [3:2] rw VCO Range Select The values for the different settings are listed in the data sheet. REGENCLR 4 w Power Regulator Enable Clear 0B No action 1B Switch off the PLL’s power regulator. The PLL is not powered (no operation possible). REGENSET 5 w Power Regulator Enable Set 0B No action 1B Switch on the PLL’s power regulator. The PLL is powered (operation possible). INSEL 6 rw Input Select 0B fOSC is selected as input for the PLL 1B fCLKIN1 is selected as input for the PLL User’s Manual SCU, V1.15 8-34 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description NDIV [13:8] rw N-Divider Value The value the N-Divider operates is NDIV+1. Only values between N = 8 and N = 28 are allowed for VCOSEL = 00B. Only values between N = 16 and N = 40 are allowed for VCOSEL = 01B. Outside of this range, stable operation cannot be ensured. NACK 15 rw N-Divider Ready Acknowledge Setting this bit provides the acknowledge signal to NRDY. 0 7, 14 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON1 PLL Configuration 1 Register ESFR (F1BAH/DDH) 15 14 13 12 11 10 9 8 7 P ACK 0 PDIV 0 rw r rw r 6 Reset Value: 000AH 5 EM EM FIN CLK DIS EN EN rw rw 4 0 r 3 2 1 0 A RES OSC PLL OSC LD SEL PWD SEL rw w rw rw Field Bits Type Description PLLPWD 0 rw PLL Power Saving Mode 0B Normal behavior 1B Complete PLL block is put into a power saving mode and no longer operates OSCSEL 1 rw Oscillator Input Selection 0B Select external clock as input for PLL 1B Select trimmed current controlled clock as input for PLL RESLD 2 w Restart VCO Lock Detection Setting this bit will reset bit PLLSTAT.VCOLOCK and restart the VCO lock detection. AOSCSEL 3 rw Asynchronous Oscillator Input Selection This bit overrules the setting of bit OSCSEL. 0B Configuration is controlled via bit OSCSEL Select asynchronously trimmed current 1B controlled clock as input for PLL EMCLKEN 5 rw VCOLCK Emergency System Clock Source Select Enable This bit requests the master clock multiplexer (MCM) to switch to an alternate clock (selected by bit field SYSCON0.EMCLKSEL) in a VCOLCK emergency case. 0B MCM remains controlled by SYSCON0.CLKSEL 1B MCM is controlled by SYSCON0.EMCLKSEL User’s Manual SCU, V1.15 8-36 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description EMFINDISEN 6 rw Emergency Input Clock Disconnect Enable This bit defines if bit PLLSTAT.FINDIS is set in a VCOLCK emergency case. 0B No action 1B PLLSTAT.FINDIS is set in a VCOLCK emergency case Note: Please refer to the Programmer’s Guide for a description of the proper handling. PDIV [11:8] rw P-Divider Value The value the P-Divider operates is PDIV+1. PACK 15 rw P-Divider Ready Acknowledge Setting this bit provides the acknowledge to PRDY. 0 4, 7, r [14:12] User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON2 PLL Configuration 2 Register ESFR (F1BCH/DEH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0001H 5 4 3 2 1 0 K1 ACK 0 K1DIV rw r rw Field Bits Type Description K1DIV [9:0] rw K1-Divider Value The value the K1-Divider operates is K1DIV+1. K1ACK 15 rw K1-Divider Ready Acknowledge1) Setting this bit provides the acknowledge to K1RDY. 0 [14:10] r 1) Reserved Read as 0; should be written with 0. Please refer to the Programmer’s Guide for a description of the proper handling. User’s Manual SCU, V1.15 8-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PLLCON3 PLL Configuration 3 Register 15 14 13 12 11 10 ESFR (F1BEH/DFH) 9 8 7 6 Reset Value: 00CBH 5 4 3 2 1 0 K2 ACK 0 K2DIV rw r rw Field Bits Type Description K2DIV [9:0] rw K2-Divider Value The value the K2-Divider operates is K2DIV+1. K2ACK 15 rw K2-Divider Ready Acknowledge1) Setting this bit provides the acknowledge to K2RDY. 0 [14:10] r 1) Reserved Read as 0; should be written with 0. Please refer to the Programmer’s Guide for a description of the proper handling. User’s Manual SCU, V1.15 8-39 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.5 System Clock Control Registers These registers control the system level clock behavior. SYSCON0 System Control 0 Register 15 14 SEL STA T 0 rh r 13 12 EMS EMS VCO OSC rh 11 SFR (FF4AH/A5H) 10 9 8 7 STM CLKSEL WUT CLKSEL rw rw rh 6 EM CLK SEL EN rw Reset Value: 0000H 5 4 3 2 1 0 0 EM CLKSEL 0 CLKSEL r rw r rw Field Bits Type Description CLKSEL [1:0] rw Clock Select This bit field defines the clock source that is used as system clock for normal operation. 00B The Wake-up clock fWU is used 01B The oscillator clock (OSC_HP) fOSC is used 10B The PLL clock fPLL is used 11B CLKIN1 as direct input clock fCLKIN1 is used EMCLKSEL [4:3] rw Emergency Clock Select This bit field defines the clock source that is used as system clock in case of an OSCWDT or VCOLCK emergency event. 00B The Wake-up clock fWU is used 01B The oscillator clock (OSC_HP) fOSC is used 10B The PLL clock fPLL is used 11B CLKIN1 as direct input clock fCLKIN1 is used rw Emergency Clock Select Enable Controls switching the system clock to an alternate source in case of an OSCWDT or VCOLCK event. The switching is disabled 0B 1B The switching is enabled EMCLKSELEN 6 User’s Manual SCU, V1.15 8-40 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WUTCLKSEL [8:7] rw WUT Clock Select This bit field defines the clock source that is used as wake-up timer clock for operation. 00B The Wake-up clock fWU is used 01B The oscillator clock (OSC_HP) fOSC is used 10B The PLL clock fPLL is used 11B CLKIN1 as direct Input clock fCLKIN1 is used STMCLKSEL [11:9] rw STM Clock Select This bit field defines the clock source that is used as STM clock for operation. 000B CLKIN2 as direct input clock fCLKIN2 is used 001B The oscillator clock (OSC_HP) fOSC is used 010B The PLL clock fPLL is used 011B CLKIN1 as direct Input clock fCLKIN1 is used 100B The Wake-up clock fWU is used 101B Reserved, do not use this combination 110B Reserved, do not use this combination 111B Reserved, do not use this combination EMSOSC 12 rh OSCWDT Emergency Event Source Status 0B No OSCWDT emergency event occurred since EMSOSC has been cleared last 1B An OSCWDT emergency event has occurred Note: This bit is only set if EMCLKSELEN is set. EMSVCO 13 rh VCOLCK Emergency Event Source Status No VCOLCK emergency event occurred since 0B EMSVCO has been cleared last 1B A VCOLCK emergency event has occurred Note: This bit is only set if EMCLKSELEN is set. SELSTAT 15 rh Clock Select Status 0B The standard configuration from bit field CLKSEL is used currently 1B The configuration from bit field EMCLKSEL is used currently 0 2, 5, 14 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-41 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) STATCLR0 Status Clear 0 Register 15 14 0 13 12 ESFR (F0E0H/70H) 11 10 9 8 7 EMC EMC VCO OSC r w Reset Value: 0000H 6 5 4 3 2 1 0 w r Field Bits Type Description EMCOSC 12 w EMSOSC Clear Trigger 0B No action 1B Bit SYSCON0.EMSOSC is cleared EMCVCO 13 w EMSVCO Clear Trigger 0B No action 1B Bit SYSCON0.EMSVCO is cleared 0 [11:0], r [15:14] User’s Manual SCU, V1.15 0 Reserved Read as 0; should be written with 0. 8-42 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.6 RTC Clock Control Register Note: Only change register RTCCLKCON while the RTC is off. RTCCLKCON RTC Clock Control Register 15 14 13 12 11 10 SFR (FF4EH/A7H) 9 8 7 6 Reset Value: 0006H 5 4 3 2 1 0 0 RTC CM RTC CLKSEL r rw rw Field Bits Type Description RTCCLKSEL [1:0] rw RTC Clock Select This bit field defines the count clock source for the RTC. 00B The PLL clock fPLL is used 01B The oscillator clock (OSC_HP) fOSC is used 10B The Wake-up clock signal fWU is used 11B CLKIN2 as direct input clock fCLKIN2 is used RTCCM 2 rw RTC Clocking Mode 0B Asynchronous Mode: The RTC internally operates with fRTC. No register access is possible. 1B Synchronous Mode: The RTC internally operates with fSYS clock. Registers can be read and written. 0 [15:3] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-43 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.1.7.7 External Clock Control Register This register control the setting of external clock for pin 2.8 and 7.1. EXTCON External Clock Control Register SFR (FF5EH/AFH) 15 14 FO EN FO SS rw rw 13 12 11 10 9 8 Reset Value: 0000H 7 6 5 4 3 2 1 0 FORV 0 FO TL 0 SEL EN rw r rh r rw rw Field Bits Type Description EN 0 rw External Clock Enable 0B No external clock signal is provided. The signal is tied to zero. 1B The configured external clock signal is provided as alternate output signal SEL [4:1] rw External Clock Select Selects the clock signal to be routed to the EXTCLK pin: 0000BSystem clock fSYS 0001BProgrammable clock signal fOUT 0010BPLL output clock fPLL 0011BOscillator clock fOSC 0100BWake-up clock fWU 0101BDirect Input clock fCLKIN1 1000BRTC count clock fRTC All other combination are reserved, do not use. FOTL 6 rh Frequency Output Toggle Latch Toggled upon each underflow of FOCNT. FORV [13:8] rw Frequency Output Reload Value Copied to FOCNT upon each underflow of FOCNT. FOSS 14 rw Frequency Output Signal Select 0B Output of the toggle latch 1B Output of the reload counter: duty cycle depends on FORV User’s Manual SCU, V1.15 8-44 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description FOEN 15 rw Frequency Output Enable 0B Frequency output generation stops when fOUT is/becomes low. 1B FOCNT is running, fOUT is gated to pin. First reload after 0 - 1 transition. 0 5, 7 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-45 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.2 System Timer Function (STM) The System Timer equips the device with a real time counter function The STM function can operate on the clock sources described in System Timer (STM) Clock Generation. The STM consists of a 16-bit counter that is able to generate up to two interrupts. Driven by a clock source the counter can be used to count time based events and upon an interrupt trigger based on a time generated out of a clock different than the remaining of the system. A clock function can easily be implemented based on these interrupts in software. : M U X STMCON. CLKDIV SYSCON 0. STMCLKSEL RELLOW RELUP Sync. Sync. fSTM Counter 0 9 1 5 run CGU Sync. Run Control System Timer Sync . STM1 Interrupt Trigger Sync . STM0 Interrupt Trigger STM_Block _noSB.vsd Figure 8-16 STM Block Diagram User’s Manual SCU, V1.15 8-46 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.2.1 STM Registers 8.2.1.1 Register STMREL Via this register, the reload value and therefore the period of the STM is defined. STMREL STM Reload Register 15 14 13 12 ESFR (F1A8H/D4H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 RELUP RELLOW rw rw 3 2 1 0 Field Bits Type Description RELLOW [9:0] rw Reload Lower Value The counter counts up and issues an interrupt trigger when bit 9 changes from 1B to 0B. Upon this trigger the counter is loaded with the reload value defined by this bit field. RELUP [15:10] rw Reload Upper Value The counter counts up and issues an interrupt trigger when bit 15 changes from 1B to 0B. Upon this trigger the counter is loaded with the reload value defined by this bit field and by bit field RELLOW. User’s Manual SCU, V1.15 8-47 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.2.1.2 Register STMCON This register holds the status and control bits for the STM. STMCON STM Control Register 15 14 13 12 ESFR (F1AAH/D5H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 CLKDIV RUN r rw rw Field Bits Type Description RUN 0 rw Run Control 0B STM is stopped 1B STM is operating By setting this bit the STM is started and the reload value STMREL.REL is loaded into the counter. CLKDIV [4:1] rw Clock Divider for the STM Clock This bit field defines the divider factor of the STM clock input. The selected input clock is divided by 2<CLKDIV>. 0 [15:5] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-48 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.3 Wake-up Timer (WUT) The Wake-up Timer provides an additional resource to trigger system functions after a specific period of time. The Wake-up Timer function can operate on the clock sources described in Wake-up Timer (WUT) Clock Generation. The wake-up timer clock fWUT drives a simple counter. All functions are controlled by register WUCR. WUCR. TTSTAT fWU OSC_WU M U X WUCR. CLKDIV SYSCON 0. WUTCLKSEL Trim Interrupt Trigger Sync. fWUT Counter run CGU WUTREL Wake-up Trigger Run Control WUCR. AONCON Wake-up Interrupt Trigger WUCR. WUTRG WUCR. RUNCON To CCU60 WUCR. ASPCON Wake-Up Timer WUT_MR+_noSB .vsd Figure 8-17 Wake-up Timer Logic 8.3.1 Wake-up Timer Operation The Wake-up Timer start and stop is controlled by the Run Control logic. The timer can be started in the following way: • bit WUCR.RUN is set When the timer is started the prescaler is reset and the counter starts to count down. The wake-up interval counter is clocked with fWUT and counts down until it reaches zero. It then generates a wake-up trigger and sets bit WUCR.WUTRG. The timer is stopped in the following ways: • • bit WUCR.RUN is cleared bit WUCR.ASP is set AND a wake-up trigger is generated If the counter is not stopped by its zero trigger it continues counting down from WUTREL. User’s Manual SCU, V1.15 8-49 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Determination of Wake-up Period The actual frequency of the trimmed current controlled wake-up clock (OSC_WU) can be measured prior to entering power-save mode in order to adjust the number of clock cycles to be counted (value written to the counter) and so to define the time until wakeup. The period of the the OSC_WU can be measured by evaluating the (synchronized) trigger that can generate interrupt requests or can be monitored with bit WUCR.TTSTAT. As using an interrupt together with software contain some uncertainty there is a second way to determine the wake-up period. The wake-up triggers generated by the WUT are forwarded to the CCU60 and can there be evaluated compared to the accurate system clock. User’s Manual SCU, V1.15 8-50 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.3.2 WUT Registers 8.3.2.1 Register WUTREL This register configures the reload value of the counter. WUTREL Wake-up Timer Reload Register ESFR (F0B0H/58H) 15 14 13 12 11 10 9 8 7 6 Reset Value: FFFFH 5 4 3 2 1 0 RELVAL rw Field Bits Type Description RELVAL [15:0] rw Wake-up Timer Reload Value The WUT counter is reloaded with this value and starts to count down when the timer is started. User’s Manual SCU, V1.15 8-51 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.3.2.2 Register WUCR This register the status and control bits for the WUT. WUCR Wake-up Control Register 15 14 13 WU TTS TRG TAT rh rh 12 11 CLKDIV ESFR (F1B0H/D8H) 10 9 8 ASP AON RUN rw rh rh rh Reset Value: 0000H 7 6 5 4 3 2 1 CLR TRG 0 ASP CON AON CON RUN CON w r w w w Field Bits Type Description RUNCON [1:0] w Control Field for RUN 00B No action 01B Set bit RUN 10B Clear bit RUN 11B Reserved, do not use this combination AONCON [3:2] w Control Field for AON 00B No action 01B Reserved, do not use this combination 10B Clear bit AON 11B Reserved, do not use this combination ASPCON [5:4] w Control Field for ASP 00B No action 01B Set bit ASP 10B Clear bit ASP 11B Reserved, do not use this combination CLRTRG 7 w Clear Bit WUTRG 0B No action 1B Clear bit WUTRG RUN 8 rh Run Indicator 0B Wake-up counter is stopped Wake-up counter is counting down 1B 0 Note: Clearing this bit via a write action to bit field RUNCON stops the WUT after four cycles of fWUT. User’s Manual SCU, V1.15 8-52 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description AON 9 rh Auto-Start Indicator 0B Wake-up counter is started by software only 1B Reserved, do not use this combination Note: This bit is cleared by writing 01B to bit field AONCON. ASP 10 rh CLKDIV [13:11] rw Clock Divider for the WUT Clock This bit field defines the divider factor of the WUT clock input. The selected input clock is divided by 2<CLKDIV>. TTSTAT 14 Trim Trigger Status 0B No trim trigger event is active. No trim interrupt trigger is generated. 1B A trim trigger event is active. A trim interrupt trigger is generated. rh Auto-Stop Indicator Wake-up counter runs continuously 0B 1B Wake-up counter stops after generating a trigger when reaching zero Note: This bit is not valid if fWUT = fWU is configured by SYSCON0.WUTCLKSEL WUTRG 15 rh WUT Trigger Indicator 0B No trigger event has occurred since WUTRG has been cleared last. No interrupt trigger is generated. 1B A wake-up trigger event has occurred. A wakeup interrupt trigger is generated. 0 6 r Reserved Read as 0; should be written with 0. Note: The bits in the upper byte of register WUCR indicate the current status of the wake-up counter logic. They are not influenced by a write access, but are controlled by their associated control fields (lower byte) or by hardware. The control bit(field)s in the lower byte of register WUCR determine the state of the status bits (upper byte) of the wake-up counter logic. Setting bits by software triggers the associated action, writing 0 has no effect. User’s Manual SCU, V1.15 8-53 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.4 Reset Operation All resets are generated by the Reset Control Block. It handles the control of the reset triggers as well as the length of a reset and the reset timing. A reset leads the system, or a part of the system depending on the reset, to a initialization into a defined state. 8.4.1 Reset Architecture The XC27x5X contains a very sophisticated reset architecture to offer the greatest amount of flexibility for the support of different applications. The reset architecture supports the different power domains. Different reset types for the complete system are supported. 8.4.1.1 Device Reset Hierarchy The device reset hierarchy is divided according to the power domains (see Chapter 8.6) into following linked levels: Level 1: I/O domain (power domain DMP_B) Level 2: System power domain DMP_M Level 3: System power domain DMP_1 If a power domain (level) is deactivated all resets of the deactivated level and all resets of all lower power domains are asserted. 8.4.1.2 Reset Types The following summary shows the different reset types. Power Reset • • Power-on Reset This reset leads to a defined state of the complete system. This reset should only be requested on a real power-on event and not by any non power related event. Power Reset for DMP_M and DMP_1 power domains This reset regains data consistency upon a power fail in the DMP_M or DMP_1 power domains. Functional / User Reset • • Debug Reset This reset leads to a defined state of the complete debug system. Internal Application Reset This reset leads to a defined state of the complete application system with the following parts: all peripherals (except the RTC), the CPU and partially the SCU and the flash memory. User’s Manual SCU, V1.15 8-54 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) • Application Reset This reset leads to a defined state of the complete application system with the following parts: all peripherals (except the Ports and the RTC), the CPU and partially the SCU and the flash memory. After a reset has been executed, the Reset Status registers RSTSTATx indicate the latest reset that has occured. To identify the type and the trigger of the latest reset registers RSTSTATx and SWDCON1 may be evaluated according to Table 8-1. The latest reset that has occured is always the reset of the highest type. If two reset triggers of the same type are indicated, this means that the two triggers have been active at the same time. If two or more reset triggers of a different type are reported, always the reset of the highest type is the latest one. Table 8-1 Identification of a Reset Type of Reset (in hierarchical order, highest on top) Identification Power-on Reset SWDCON1.PON = 1B RSTSTAT1.STM = 11B RSTSTAT1.ST1 = 11B Further action: clear PON bit to be able to identify a Power Reset for DMP_M and DMP_1 power domains. Power Reset for DMP_M and DMP_1 power domains SWDCON1.PON = 0B (unchanged after clearing) RSTSTAT1.STM = 11B RSTSTAT1.ST1 = 11B Internal Application Reset RSTSTAT1.ST1 = 00B any bit field y in RSTSTATx.y = 10B Application Reset RSTSTAT1.ST1 = 00B any bit field y in RSTSTATx.y = 11B (except RSTSAT1.ST1 and RSTSTAT1.STM) The algorithm depicted in Figure 8-18 shows a sequence to detect the type of the reset comprising the conditions in Table 8-1. Further information can be found in the Programmer’s Guide. User’s Manual SCU, V1.15 8-55 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Detect Reset Type no ? yes RSTSTAT 1.ST 1=11 yes ? no User Reset Internal Application Reset any bit field YY of RSTSTATx. YY = 11 b SWDCON1.PON = 1 no ? yes yes ? no User Reset Application Reset invalid type of reset Power Reset DMP_M and DMP_1 power domains any bit field YY of RSTSTATx. YY = 10 b clear SWDCON1.PON Power Reset Power-on exit reset specific actions return Flowchart_DetectResetType _noSB.vsd Figure 8-18 Algorithm for the Detection of the Type of a Reset User’s Manual SCU, V1.15 8-56 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.4.2 General Reset Operation A reset is generated if an enabled reset request trigger is asserted. Most reset request triggers can be configured for the reset type it should initiate. No action (disabled) is one possible configuration and can be selected for a reset request trigger by setting the respective bit field in a Reset Configuration Register to 00B. The debug reset can only be requested by dedicated reset request triggers and can not be selected via a Reset Configuration Register. For more information see also registers RSTCON0 and RSTCON1. The duration of a reset is defined by two independent counters. One counter for the System and Application Reset types and one separate counter for the debug reset. A separate counter for the debug reset was implemented to allow a non-intrusive adaptation of the reset length to the debugger needs without modification of the application setting. 8.4.2.1 Reset Counters (RSTCNTA and RSTCNTD) RSTCNTA is the reset counter that controls the reset length for all application relevant resets (Internal Application Reset, and Application Reset). RSTCNTD is the reset counter that controls the reset length for the debug reset. The reset counters control the length of the internal resets. This can be used to configure the duration of a reset output via the ESRx pins, so this matches with the reset input requirements of external blocks connected to these signals. A reset counter RSTCNT is an 8-bit counter counting down from the reload value defined by RSTCNTCON.RELx (x = A or D). The counter is started by the reset control block as soon as a reset request trigger condition becomes active (for more information see Table 8-2 and Table 8-3). Whether the counter has to be started or not depends on the reset request trigger and whether the counter is already active or not. In case of that the counter is inactive, not counting down, it is always started. While the counter is already active it depends on the reset type of the new reset request trigger that was asserted anew if the counter is restarted or not. This behavior is summarized in Table 8-2 and Table 8-3. User’s Manual SCU, V1.15 8-57 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-2 Restart of RSTCNTA Reset Active New Reset Trigger Power-On Debug Reset Internal Application Reset Application Reset Internal Application Reset Restart with default delay No Change No Change No Change Application Reset Restart with default delay No Change Restart with defined delay No Change Table 8-3 Restart of RSTCNTD Reset Active New Reset Trigger Power-On Debug Reset Restart with default delay Debug Reset No Change Internal Application Reset No Change Application Reset No Change The reset counters RSTCNTx ensure a configurable minimum duration of a generated reset. If a reset request trigger remains asserted after the respective counter has counted down, the counter is not started again, instead the reset control block keeps the reset asserted until the reset request trigger is deasserted. 8.4.2.2 De-assertion of a Reset The reset of a dedicated type is de-asserted when all of the following conditions are fulfilled: • • The reset counter has been expired (reached zero). No reset request trigger that is configured to generate a reset of the dedicated type (or higher) is currently asserted. Example1 Reset request trigger A is asserted and leads to an Application Reset. If the reset request trigger is de-asserted before RSTCNTA reached zero the Application Reset is deasserted when RSTCNTA reaches zero. If the reset request trigger is de-asserted after RSTCNTA reached zero the Application Reset is de-asserted when the reset request trigger is de-asserted. User’s Manual SCU, V1.15 8-58 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Example2 Reset request trigger A is asserted and leads to an Application Reset. Reset request trigger A is de-asserted before RSTCNTA reached zero. Reset request trigger B is asserted after reset request trigger A but before RSTCNTA reaches zero. Reset request trigger B is also configured to result in a Application Reset. If the reset request trigger B is de-asserted before RSTCNTA reached zero the Application Reset is de-asserted when RSTCNTA reaches zero. If the reset request trigger B is de-asserted after RSTCNTA reached zero the Application Reset is de-asserted when the reset request trigger B is de-asserted. 8.4.3 Debug Reset Assertion Unlike the other reset types a Debug Reset can only be asserted if the following two conditions are valid: • • A reset request trigger is asserted that request a debug reset An Application Reset is already active in the system 8.4.4 Coupling of Reset Types The different reset types are coupled for a better usage: • • The assertion of a Power-on Reset automatically asserts also the following reset types: – Debug Reset – Internal Application Reset – Application Reset The assertion of an Internal Application Reset automatically asserts also the following reset type: – Application Reset User’s Manual SCU, V1.15 8-59 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.4.5 Reset Request Trigger Sources The following overview summarizes the different reset request trigger sources within the system. Power-On Reset Pin PORST A Power-on Reset is requests asynchronously, by driving the PORST pin low. Supply Watchdog (SWD) If the power supply for I/O domain is below the value required for proper functionality, a non-synchronized reset request trigger is generated if the SWD reset generation is enabled. This ensures a reproducible behavior in the case of power-fail. This can also be used to restart the system without the usage of the PORST pin. As long as the I/O power domain does not get the required voltage level the system is held in the reset. Core Power Validation (PVC_M and PVC_1) If the core power supply is below the value required for proper functionality of the main power domain (PVC_M), a reset request trigger can be forwarded to the system. The generation of a Power-on Reset is configured by bit PVCMCON0.L1RSTEN = 1B. If the bit PVCMCON0.L1RSTEN = 1B a request trigger is asserted for PVC_M1 upon a level check match. If the bit PVCMCON0.L2RSTEN = 1B a request trigger is asserted for PVC_M2 upon a level check match. If the core power supply is below the value required for proper functionality of the application power domain (PVC_1), a reset request trigger can be forwarded to the system. The generation of a Power-on Reset (Application Power Domain only) is configured by bit PVC1CON0.L1RSTEN = 1B. If bit PVC1CON0.L1RSTEN = 1B a request trigger is asserted for PVC_11 upon a level check match. If the bit PVC1CON0.L2RSTEN = 1B a request trigger is asserted for PVC_12 upon a level check match. For more information about the Power Validation Circuit see Chapter 8.6.2. ESRx An ESRx reset request trigger leads to a configurable reset. The type of reset can be configured via RSTCON1.ESRx. The pins ESRx can serve as an external reset input as well as a reset output (open drain) for Internal Application and Application Resets. Furthermore, several GPIO pad triggers, that can be enabled additionally via register ESREXCONx (x = 1, 2), interfere with the ESR pin function. GPIO and ESRx pin triggers can be enabled/disabled individually and are combined for the reset trigger generation. User’s Manual SCU, V1.15 8-60 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) If pin ESRx is enabled as reset output and the input level is low while the output stage is disabled (indicating that it is still driven low externally), the reset circuitry holds the chip in reset until a high level is detected on ESRx. Minimum value for RSTCNTCON.RELA must be the reset value. Note: The reset output is only driven low for the duration the reset counter RSTCNTA is active. During a possible reset extension the reset output is no longer driven. Software A software reset request trigger leads to a configurable reset. The type of reset can be configured via RSTCON0.SW. Watchdog Timer A WDT reset request trigger leads to a configurable reset. The type of reset can be configured via RSTCON1.WDT. A WDT reset is requested on a WDT overflow event. For more information see Chapter 8.12. CPU A CPU reset request trigger leads to a configurable reset. The type of reset can be configured via RSTCON0.CPU. A CPU reset is requested when instruction SRST is executed. Memory Parity A MP reset request trigger leads to a configurable reset. The type of reset can be configured via RSTCON1.MP. For more information see Chapter 8.14.2. OCDS Block The OCDS block has several options to request different reset types: 1. A Debug Reset either via the OCDS reset function or via bit CBS_OJCONF.RSTCL1 AND CBS_OJCONF.RSTCL3 2. An Internal Application Reset via bit CBS_OJCONF.RSTCL2 3. An Application Reset via bit CBS_OJCONF.RSTCL3 8.4.5.1 Reset Sources Overview The connection of the reset sources and the activated reset types are shown in Table 8-4. User’s Manual SCU, V1.15 8-61 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-4 Effects of Reset Types for Reset Activation Reset Request Trigger Application Reset Internal Application Reset Debug Reset PORST Activated Activated Activated SWD Activated Activated Activated PVC_M1 Activated Activated Activated PVC_M2 Activated Activated Activated PVC_11 Activated Activated Activated PVC_12 Activated Activated Activated ESR0 Configurable Configurable Not Activated ESR1 Configurable Configurable Not Activated ESR2 Configurable Configurable Not Activated WDT Configurable Configurable Not Activated SW Configurable Configurable Not Activated CPU Configurable Configurable Not Activated MP Configurable Configurable Not Activated OCDS Reset Not Activated Not Activated Activated1) CBS_OJCONF.RSTCL1 Not Activated Not Activated Activated1) CBS_OJCONF.RSTCL2 Activated Activated Not Activated CBS_OJCONF.RSTCL3 Activated Not Activated Not Activated 1) Only if an Application Reset is active or is requested in parallel. 8.4.6 Module Reset Behavior Table 8-5 lists how the various functions of the XC27x5X are affected through a reset depending on the reset type. A “X” means that this block has at least some register/bits that are affected by this reset type. User’s Manual SCU, V1.15 8-62 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-5 Effect of Reset on Device Functions Module / Function Application Reset Internal Application Reset Debug Reset CPU Core X X X Peripherals (except SCU and RTC) X X X SCU X Not affected Not affected RTC Not affected Not affected X On-chip DPRAM 1) Static RAMs Not affected, reliable Not affected, reliable Not affected, reliable PSRAM Not affected, reliable Not affected, reliable Not affected, reliable DSRAM Not affected, reliable Not affected, reliable Not affected, reliable Flash Memory X X 2) 2) Not affected, reliable JTAG Interface Not affected Not affected Not affected OCDS Not affected Not affected X Oscillator, PLL Not affected Not affected Not affected Port Pins Not affected X Not affected Pins ESRx Not affected X Not affected 1) Reliable here means that also the redundancy is not affected by the reset. 2) Parts of the flash memory block are only reset by a Power-on Reset. For more detail see the flash chapter. User’s Manual SCU, V1.15 8-63 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.4.7 Reset Controller Registers 8.4.7.1 Status Registers After a reset has been executed, the Reset Status registers provide information on the type of the last reset. The reset status registers are updated upon each reset. RSTSTAT0 Reset Status 0 Register 15 14 13 12 ESFR (F0B2H/59H) 11 10 9 8 7 Reset Value: 0000H 6 5 SW CPU 0 rh rh r 3 2 1 0 Field Bits CPU [13:12] rh CPU Reset Type Status 00B The CPU reset trigger was not relevant for the last reset 01B Reserved 10B The CPU reset trigger was relevant for the last reset. Internal Application and Application Resets were generated. 11B The CPU reset trigger was relevant for the last reset. Application Reset was generated. SW [15:14] rh Software Reset Type Status 00B The Software reset trigger was not relevant for the last reset 01B Reserved 10B The Software reset trigger was relevant for the last reset. Internal Application and Application Resets were generated. 11B The Software reset trigger was relevant for the last reset. Application Reset was generated. 0 [11:0] Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 Type 4 r Description 8-64 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) RSTSTAT1 Reset Status 1 Register 15 14 13 12 ESFR (F0B4H/5AH) 11 10 9 8 7 6 Reset Value: F000H 5 4 3 2 1 0 ST1 STM 0 MP WDT ESR2 ESR1 ESR0 rh rh r rh rh rh rh rh Field Bits Type Description ESR0 [1:0] rh ESR0 Reset Status 00B The ESR0 reset trigger was not relevant for the last reset 01B Reserved 10B The ESR0 reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The ESR0 reset trigger was relevant for the last reset. Application Reset was generated. ESR1 [3:2] rh ESR1 Reset Status 00B The ESR1 reset trigger was not relevant for the last reset 01B Reserved 10B The ESR1 reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The ESR1 reset trigger was relevant for the last reset. Application Reset was generated. ESR2 [5:4] rh ESR2 Reset Status 00B The ESR2 reset trigger was not relevant for the last reset 01B Reserved 10B The ESR2 reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The ESR2 reset trigger was relevant for the last reset. Application Reset was generated. User’s Manual SCU, V1.15 8-65 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WDT [7:6] rh WDT Reset Status 00B The WDT reset trigger was not relevant for the last reset 01B Reserved 10B The WDT reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The WDT reset trigger was relevant for the last reset. Application Reset was generated. MP [9:8] rh MP Reset Status 00B The MP reset trigger was not relevant for the last reset 01B Reserved 10B The MP reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The MP reset trigger was relevant for the last reset. Application Reset was generated. STM [13:12] rh Power-on for DMP_M Reset Status 00B The power-on reset for DMP_M reset trigger was not relevant for the last reset 01B The power-on reset for DMP_M reset trigger was not relevant for the last reset 10B The power-on reset for DMP_M reset trigger was not relevant for the last reset 11B The power-on reset for DMP_M reset trigger was relevant for the last reset ST1 [15:14] rh Power-on for DMP_1 Reset Status 00B The power-on reset for DMP_1 reset trigger was not relevant for the last reset 01B The power-on reset for DMP_1 reset trigger was not relevant for the last reset 10B The power-on reset for DMP_1 reset trigger was not relevant for the last reset 11B The power-on reset for DMP_1 reset trigger was relevant for the last reset 0 [11:10] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-66 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) RSTSTAT2 Reset Status 2 Register 15 14 13 12 ESFR (F0B6H/5BH) 11 10 0 9 8 7 6 Reset Value: 0000H 5 4 OJCONF3 OJCONF2 OJCONF1 r rh rh rh 3 2 1 0 0 DB r rh Field Bits Type Description DB [1:0] rh Debug Reset Status 00B The DB reset trigger was not relevant for the last reset 01B The DB reset trigger was not relevant for the last reset 10B The DB reset trigger was not relevant for the last reset 11B The DB reset trigger was relevant for the last reset OJCONF1 [5:4] rh OJCONF1 Reset Status 00B The OJCONF1 reset trigger was not relevant for the last reset 01B The OJCONF1 reset trigger was not relevant for the last reset 10B The OJCONF1 reset trigger was not relevant for the last reset 11B The OJCONF1 reset trigger was relevant for the last reset. Debug Reset was generated. OJCONF2 [7:6] rh OJCONF2 Reset Status 00B The OJCONF2 reset trigger was not relevant for the last reset 01B The OJCONF2 reset trigger was not relevant for the last reset 10B The OJCONF2 reset trigger was relevant for the last reset. Internal Application, and Application Resets were generated. 11B The OJCONF2 reset trigger was not relevant for the last reset User’s Manual SCU, V1.15 8-67 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description OJCONF3 [9:8] rw OJCONF3 Reset Status 00B The OJCONF3 reset trigger was not relevant for the last reset 01B The OJCONF3 reset trigger was not relevant for the last reset 10B The OJCONF3 reset trigger was not relevant for the last reset 11B The OJCONF3 reset trigger was relevant for the last reset. Application Reset was generated. 0 [3:2], r [15:10] User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-68 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.4.7.2 Configuration Registers These registers allow the behavioral configuration for the various reset trigger sources. RSTCON0 Reset Configuration 0 Register ESFR (F0B8H/5CH) 15 14 13 12 11 10 9 8 7 Reset Value: 0000H 6 5 SW CPU 0 rw rw rw 3 2 1 0 Field Bits CPU [13:12] rw CPU Reset Type Selection This bit field defines which reset types are generated by a CPU reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated SW [15:14] rw Software Reset Type Selection This bit field defines which reset types are generated by a software reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated 0 [11:0] Reserved Must be written with reset value 0. User’s Manual SCU, V1.15 Type 4 rw Description 8-69 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) RSTCON1 Reset Configuration 1 Register ESFR (F0BAH/5DH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0002H 5 4 3 2 1 0 0 MP WDT ESR2 ESR1 ESR0 rw rw rw rw rw rw Field Bits Type Description ESR0 [1:0] rw ESR0 Reset Type Selection This bit field defines which reset types are generated by a ESR0 reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated ESR1 [3:2] rw ESR1 Reset Type Selection This bit field defines which reset types are generated by a ESR1 reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated ESR2 [5:4] rw ESR2 Reset Type Selection This bit field defines which reset types are generated by a ESR2 reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated User’s Manual SCU, V1.15 8-70 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WDT [7:6] rw WDT Reset Type Selection This bit field defines which reset types are generated by a WDT reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated MP [9:8] rw MP Reset Type Selection This bit field defines which reset types are generated by a MP reset request trigger. 00B No reset is generated 01B Reserved, do not use this combination 10B Internal Application, and Application Resets are generated 11B Application Reset is generated 0 [15:10] rw User’s Manual SCU, V1.15 Reserved Should be written with 0. 8-71 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) RSTCNTCON Reset Counter Control RegisterESFR (F1B2H/D9H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0A0AH 5 4 3 RELD RELA rw rw 2 1 0 Field Bits Type Description RELA [7:0] rw Application Reset Counter Reload Value This bit field defines the reload value of RSTCNTA. This value is always used when counter RSTCNTA is started. This counter value is used for Internal Application, and Application Resets. In case of an ESRx reset the counter value must be not less than the reset value. RELD [15:8] rw Debug Reset Counter Reload Value This bit field defines the reload value of RSTCNTD. This value is always used when counter RSTCNTD is started. This counter value is used for the Debug Reset. In case of an ESRx reset the counter value must be not less than the reset value. User’s Manual SCU, V1.15 8-72 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Software Reset Control Register This register controls the software reset operation. SWRSTCON Software Reset Control RegisterESFR (F0AEH/57H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 SWCFG 0 rw r 3 2 1 0 SW SW RST BOO REQ T w rw Field Bits Type Description SWBOOT 0 rw Software Boot Configuration Selection 0B Bit field STSTAT.HWCFG is not updated with the content of SWCFG upon an Application Reset 1B Bit field STSTAT.HWCFG is updated with the content of SWCFG upon an Application Reset SWRSTREQ 1 w Software Reset Request 0B No software reset is requested 1B A software reset request trigger is generated SWCFG [15:8] rw Software Boot Configuration A valid software boot configuration (also different from the external applied hardware configuration) can be specified with these bits. The configuration encoding is equal to the HWCFG encoding in register STSTAT. 0 [7:2] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-73 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.5 External Service Request (ESR) Pins The ESR pins serve as multi-functional pins for an amount of different options: • • • • • Act as reset trigger input Act as reset output Act as trap input Act as trigger input for the GSC Independent pad configuration 8.5.1 General Operation Each ESR pin is equipped with an edge detection that allows the selection of the edges used as triggers. One, both, or no edge can be selected via bit field ESRCFGx.SEDCON if a clock is active. Additionally, there is a digital (3-stage median) filter (DF) to suppress spikes. The signal at ESRx pin has to be held at the active signal level for at least 2 system clock cycles (fSYS) in order to generate a trigger. The digital filter can be disabled by clearing bit ESRCFGx.DFEN. Each ESRx pin can be individually configured. If an ESR trigger is generated please note that triggers for all purposes (reset, trap, GSC, and non SCU module functions) are generated. If some of the actions resulting out of such a trigger should not occur this has to be disabled by each feature for its own. The pins that should be used as trigger input for an ESR operation have to be configured as input pin. User’s Manual SCU, V1.15 8-74 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) DMPMIT. ESRxT Edge Detection ESRCFGx. SEDCON ESRx Input Selection Trap GSC Digital Filter to System ESRCFGx. DFEN ESRx Control Reset Control ESRx ESR _control _noSB .vsd Figure 8-19 ESRx Control Up to three ESR pins (ESR0/ESR1/ESR2) are available. The availability of pins ESR1 and ESR2 is device and package dependent and is described in the data sheet. User’s Manual SCU, V1.15 8-75 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.5.1.1 ESR as Reset Input The pins ESRx can serve as an external reset input as well as a reset output (open drain) for Internal Application and Application Resets. Additionally several GPIO pad triggers that can be enabled additionally via register ESREXCONx interfere with the ESR pin function. GPIO and ESR pin triggers can be enabled/disabled individually and are combined for the reset trigger generation. For more information about the reset system see Chapter 8.4. Note: The reset output is only asserted for the duration the reset counter RSTCNTA is active. During a possible reset extension the reset output is not longer asserted. 8.5.1.2 ESR as Reset Output If pin ESRx is enabled as reset output and the input level is low while the output stage is disabled (indicating that it is still driven low externally), the reset circuitry holds the chip in reset until a high level is detected on ESRx. The internal output stage drives a low level during reset only while RSTCNTA is active. It deactivates the output stage when the time defined by RSTCNTCON.RELA has passed. For more information about the reset system see Chapter 8.4. 8.5.1.3 ESR as Trap Trigger The ESR can request traps. The control mechanism if and which trap is requested is located in the trap control logic. For more information see Chapter 8.13. 8.5.1.4 ESR as Trigger Input for the GSC The ESR can be used to request a change in the Control Mode. For more information see Chapter 8.7. User’s Manual SCU, V1.15 8-76 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.5.1.5 Pad Configuration for ESR Pads The configuration is selected via bit field ESRCFGx.PC. The pad functionality control can be configured independently for each pin, comprising: • • • A selection of the driver type (open-drain or push-pull) An enable function for the output driver (input and/or output capability) An enable function for the pull-up/down resistance The following table defines the coding of the bit fields PC in registers ESRCFG0, ESRCFG1, and ESRCFG2. Note: The coding is the same as for the port register bit fields Pn_IOCRx.PC. Table 8-6 PC Coding PCx[3:0] Selected Pull-up/Pull-down / Selected Output Function I/O 0000B No pull device activated 0001B Pull-down device activated 0010B Pull-up device activated Input is not inverted, the input stage is active in power-down mode 0011B No pull device activated 0100B No pull device activated 0101B Pull-down device activated 0110B Pull-up device activated 0111B No pull device activated 1000B Output of ESRCFGx.OUT 1001B Output of ESRCFGx.OUT 1010B Output drives a 0 for an Internal Application Reset, a 1 otherwise. 1011B Output drives a 0 for an Application Reset, a 1 otherwise. 1100B Output of ESRCFGx.OUT 1101B Output of ESRCFGx.OUT 1110B Output drives a 0 for an Internal Application Reset 1111B Output drives a 0 for an Application Reset User’s Manual SCU, V1.15 Output Characteristics Input is inverted, the input stage is active in power-down mode 8-77 Output, Push-pull the input stage is not inverted and active in power-down mode Open-drain, a pull-up device is activated while the output is not driving a 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.5.2 ESR Control Registers 8.5.2.1 Configuration Registers ESR External Control Register ESREXCON1 ESR1 External Control Register ESREXCON2 ESR2 External Control Register 15 14 13 12 11 10 SFR (FF32H/99H) Reset Value: 0001H SFR (FF34H/9AH) Reset Value: 0001H 9 8 7 6 5 4 3 2 1 0 ESR EN rw rw Field Bits Type Description ESREN 0 rw ESRy Pin Enable This bit enables/disables the ESRy pin for the activation of all ESRy related actions. 0B The input from pin ESRy is disabled 1B The input from pin ESRy is enabled 0 [15:1] rw Reserved Must be written with reset value 0. User’s Manual SCU, V1.15 0 8-78 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) ESREXSTAT1 ESR1 External Status Register SFR (FF36H/9BH) ESREXSTAT2 ESR2 External Status Register SFR (FF38H/9CH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H Reset Value: 0000H 5 4 3 2 1 0 ESR r rh Field Bits Type Description ESR 0 rh Input ESRy Status This bit is set upon a trigger on input x if ESREXCONy.ESREN was set. This bit can be cleared only by software. 0B No trigger for input ESRy occurred 1B A trigger for ESRy occurred since it was cleared last time 0 [15:1] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 0 8-79 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) CLRESREXSTAT1 Clear ESR1 External Status RegisterSFR (FF3AH/9DH) CLRESREXSTAT2 Clear ESR2 External Status RegisterSFR (FF3CH/9EH) 15 14 13 12 11 10 9 8 7 6 5 Reset Value: 0000H Reset Value: 0000H 4 3 2 1 0 0 ESR r w Field Bits Type Description ESR 0 w Clear Input ESRy Status Setting this bit clears the bit ESREXSTATy.ESR. This bit always read as zero. 0B No effect 1B Bit ESREXSTATy.ESR is cleared 0 [15:1] w Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-80 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) ESR Configuration Register The ESR configuration registers contains bits required for the behavioral control of the ESR pins. ESRCFG0 ESR0 Configuration Register ESRCFG1 ESR1 Configuration Register ESRCFG2 ESR2 Configuration Register 15 14 13 12 11 10 ESFR (F100H/80H) Reset Value: 000EH ESFR (F102H/81H) Reset Value: 0002H ESFR (F104H/82H) Reset Value: 0002H 9 8 7 6 5 4 3 2 1 0 0 SEDCON IN OUT DF EN PC r rw rw rh rh rw rw 0 Field Bits Type Description PC [3:0] rw Pin Control of ESRx This bit field controls the behavior of the associated ESRx pin. The coding is described in Table 8-6. DFEN 4 rw Digital Filter Enable This bit defines if the 3-stage median filter of the ESRx is used or bypassed. 0B The filter is bypassed 1B The filter is used OUT 5 rh Data Output This bit can be used as output value for the associated ESRx pin. 0B If selected, the output level is 0 1B If selected, the output level is 1 This bit is controlled via bit field ESRDAT.MOUTx. IN 6 rh Data Input This bit monitors the input value at the associated ESRx pin. User’s Manual SCU, V1.15 8-81 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description SEDCON [8:7] rw Synchronous Edge Detection Control This bit field defines the edges that lead to an ESRx trigger of the synchronous path. 00B No trigger is generated 01B A trigger is generated upon a raising edge 10B A trigger is generated upon a falling edge 11B A trigger is generated upon a raising AND falling edge 0 [10:9] rw Reserved Must be written with reset value 00B. 0 [15:11] r User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-82 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.5.3 ESR Data Register 8.5.3.1 ESRDAT The ESR data register contains bits required if ESRx are used as data ports. ESRDAT ESR Data Register 15 14 13 12 ESFR (F106H/83H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 MOUT2 MOUT1 MOUT0 r w w w Field Bits Type Description MOUT0 [1:0] w Modification of ESRCFG0.OUT Writing to this bit field can modify the content of bit ESRCFG0.OUT for ESR0. It always reads 0. 00B Bit ESRCFG0.OUT is unchanged 01B Bit ESRCFG0.OUT is set 10B Bit ESRCFG0.OUT is cleared 11B Reserved, do not use this combination MOUT1 [3:2] w Modification of ESRCFG1.OUT Writing to this bit field can modify the content of bit ESRCFG1.OUT for ESR1. It always reads 0. 00B Bit ESRCFG1.OUT is unchanged 01B Bit ESRCFG1.OUT is set 10B Bit ESRCFG1.OUT is cleared 11B Reserved, do not use this combination MOUT2 [5:4] w Modification of ESRCFG2.OUT Writing to this bit field can modify the content of bit ESRCFG2.OUT for ESR2. It always reads 0. 00B Bit ESRCFG2.OUT is unchanged 01B Bit ESRCFG2.OUT is set 10B Bit ESRCFG2.OUT is cleared 11B Reserved, do not use this combination 0 [15:6] w Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-83 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6 Power Supply and Control The XC27x5X can run from a single external power supply. The core supply voltages can be generated by on-chip Embedded Voltage Regulators (EVRs). Power Domains The I/O part is divided in two parts DMP_A and DMP_B. DMP_A contains all ADC related I/Os and DMP_B the remaining system and communication I/Os. The major part of the on-chip logic is located in an independent core power domain (DMP_1). A second power domain (DMP_M), marked grey in the figure below, controls important device infrastructure plus a Standby RAM (SBRAM). ESR STM CPU WDT PEC/ INT PSRAM IMB Clock Temp. Control Comp. PLL DPRAM PMU OSC_ HP DSRAM GSC MPU OSC_ Reset Power WU Control Control Program Flash DMU SCU ERU Stand-By RAM LXBus MCHK EBC LXBus Control External Bus Control OCDS Debug Support Peripheral Data Bus ADC x GPT CC x LXBus CCU 6x RTC USIC x Multi CAN Ports BLOCKDIAGRAM_MR+ Figure 8-20 XC27x5X Power Domain Structure Power Supply and Control Functions The power supply and control is divided into following parts: • • monitoring of the supply voltage controlling and adjusting the supply voltage User’s Manual SCU, V1.15 8-84 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) The supply voltage of pad IO domain for system and communication I/Os (power domain DMP_B) is monitored by a Supply WatchDog (SWD, see Chapter 8.6.1). The core voltage for each of the two core supply domains is supervised by a separate Power Validation Circuit (PVC) that provides two monitoring levels. Each monitoring level can request an interrupt (e.g. power-fail warning) or a reset depending on the voltage level. A PVC is used to detect under voltage due to an external short (see Chapter 8.6.2). By controlling the regulator, the core power can be switched off to save the leakage current (see Chapter 8.6.3). Table 8-7 XC27x5X Power Domains Supply and Control Power Domain Supply Source Pad IO domain (DMP_B) External supply VDDPB: 3.0 … 5.5 typ See data sheet SWD ADC IO domain (DMP_A) External supply VDDPA: 3.0 … 5.5 typ See data sheet - Core domain (DMP_M and DMP_1) EVR_M EVR_1 PVC_1, PVC_M User’s Manual SCU, V1.15 Supply Voltage [V] VDDIM, VDDI1: 1.5 typ See data sheet 8-85 Supply Checked by V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.1 Supply Watchdog (SWD) The supply voltage of the pad I/O domain for systems and communication I/Os (DMP_B) is monitored to validate the overall power supply. The external supply voltage is monitored for following purposes: • • • • POR Detecting the ramp-up of the external supply voltage, so the device can be started without requiring an external power-on reset (PORST). Brown-out Detecting the ramp-down of the external supply voltage, so the device can be brought into a save state without requiring an external power-on reset (PORST). Monitoring the external power supply allows the usage of a low-cost regulator without additional status signals (standard 3-pin device). Guarantee that the supply voltage for the EVRs is sufficient to generate a valid core voltage under every operating condition Feature list The following list is a summary of the SWD functions. • • • • Trigger a power-on reset whenever the supply falls and as long as the supply remains below VVAL Two completely independent threshold levels and comparators 16 selectable threshold levels Power Saving Mode (only VVAL detection active) User’s Manual SCU, V1.15 8-86 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Operating the SWD VOP VLEV2 VLEV1 VVAL VSS 0V start-up reset L2OK L1OK Power-Off Power-On Operation Operation Operation Fail Warning Power-Off Figure 8-21 SWD Power Validation Example The lower fix threshold VVAL defines the absolute minimum operation voltage for the IO domain. If VVAL has not been reached the device is held in reset. When VDDPB raises above VVAL, bit SWDCON1.PON is set. Note: The physical value for VVAL can found in the XC27x5X data sheet. The SWD provides two adjustable threshold levels (LEV1 and LEV2) that can be individually programmed, via SWDCON0.LEV1V and SWDCON0.LEV2V, and deliver a compare value each. The two compare results can be monitored via bits SWDCON0.L1OK and SWDCON0.L2OK. A reset or interrupt request can be generated while the voltage level is below or equal/above the configured level of a threshold. If an action and which action is triggered by each threshold can be configured via bits User’s Manual SCU, V1.15 8-87 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) SWDCON0.LxRSTEN and SWDCON0.LxINTEN and bit field SWDCON0.LxALEV (x = 1,2). The SWD control (programming of the threshold levels) is done by software only. With these features, an external supply watchdog, e.g. integrated in some external VR, can be replaced. It detects the minimum specified supply voltage level and can be configured to monitor other voltage levels. Note: If the PORST pin is used it has the same functionality as the min-power detection of the SWD. Power-Saving Mode of the SWD The two configurable thresholds can be disabled if not needed. This is called the SWD Power Saving Mode. The minimum operating voltage detection (POR/Brown-out detection) can not be disabled and it is always active. The SWD Power Saving Mode is entered by setting bit SWDCON1.POWENSET and exit by setting bit SWDCON1.POWENCLR. If the SWD Power Saving Mode is active is indicated by bit SWDCON1.POWEN. Note: The reset request and interrupt request action should be switched off before entering power-save mode. User’s Manual SCU, V1.15 8-88 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.1.1 SWD Control Registers The following registers are the software interface for the SWD. SWDCON0 SWD Control 0 Register 15 14 13 L2 L2 L2 A RST INT LEV EN EN rw rw rw 12 11 ESFR (F080H/40H) 10 9 L2 OK LEV2V rh rw 8 7 6 Reset Value: 0941H 5 L1 L1 L1 A RST INT LEV EN EN rw rw rw 4 3 2 1 L1 OK LEV1V rh rw 0 Field Bits Type Description LEV1V [3:0] rw Level Threshold 1 Voltage This bit field defines the voltage level that is used as threshold 1 check level. The values of the level thresholds are listed in the data sheet. L1OK 4 rh Level Threshold 1 Check Result 0B The supply voltage is below the Level Threshold 1 voltage LEV1V 1B The supply voltage is equal or above the Level Threshold 1 voltage LEV1V L1INTEN 5 rw Level Threshold 1 Interrupt Request Enable This bit field defines if an interrupt is requested if the supply voltage comparison matches the action level L1ALEV. 0B No interrupt is requested 1B An interrupt is requested L1RSTEN 6 rw Level Threshold 1 Reset Request Enable This bit field defines if an interrupt is requested if the supply voltage comparison matches the action level L1ALEV. 0B No reset is requested 1B An reset is requested User’s Manual SCU, V1.15 8-89 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L1ALEV 7 rw Level Threshold 1 Action Level 0B When the supply voltage is below the Level Threshold 1 voltage LEV1V the actions configured by bits L1INTEN and L1RSTEN are requested 1B When the supply voltage is equal or above the Level Threshold 1 voltage LEV1V the actions configured by bits L1INTEN and L1RSTEN are requested LEV2V [11:8] rw Level Threshold 2 Voltage This bit field defines the voltage level that is used as check level threshold 2. The values of the level thresholds are listed in the data sheet. L2OK 12 rh Level Threshold 2 Check Result 0B The supply voltage is below the Level Threshold 2 voltage LEV2V 1B The supply voltage is equal or above the Level Threshold 2 voltage LEV2V L2INTEN 13 rw Level Threshold 2 Interrupt Request Enable This bit field defines if an interrupt is requested if the supply voltage comparison matches the action level L2ALEV. 0B No interrupt is requested 1B An interrupt is requested L2RSTEN 14 rw Level Threshold 2 Reset Request Enable This bit field defines if an interrupt is requested if the supply voltage comparison matches the action level L2ALEV. 0B No reset is requested 1B An reset is requested User’s Manual SCU, V1.15 8-90 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L2ALEV 15 rw Level Threshold 2 Action Level 0B When the supply voltage is below the Level Threshold 2 voltage LEV2V the actions configured by bits L2INTEN and L2RSTEN are requested 1B When the supply voltage is equal or above the Level Threshold 2 voltage LEV2V the actions configured by bits L2INTEN and L2RSTEN are requested User’s Manual SCU, V1.15 8-91 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) SWDCON1 SWD Control 1 Register 15 14 13 12 11 ESFR (F082H/41H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 POW POW CLR POW PON EN EN PON EN SET CLR w rh rh w w 0 r Field Bits Type Description POWENCLR 0 w SWD Power Saving Mode Enable Clear No action 0B 1B Bit POWEN is cleared POWENSET 1 w SWD Power Saving Mode Enable Set 0B No action Bit POWEN is set 1B POWEN 2 rh SWD Power Saving Mode Enable 0B All SWD functions are enabled 1B The SWD Power Saving Mode is enabled. Comparators are disabled. PON 3 rh Power-On Status Flag No power-on event occurred 0B 1B A power-on event occurred (VDDP became greater than VVAL). CLRPON 4 w Clear Power-On Status Flag 0B No action 1B Bit PON is cleared 0 [15:5] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 0 8-92 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.2 Monitoring the Voltage Level of a Core Domain A Power Validation Circuit (PVC) monitors the internal core supply voltage of a core domain. It can be configured to monitor two programmable independent voltage levels. The voltage of the core domain is monitored by PVC_1 and PVC_M. Feature list The following list summarizes the features of a PVC. • • • • Two independent comparators Threshold levels selectable Shut-off, which disables the complete module Configurable action level A PVC provides two adjustable threshold levels (LEV1 and LEV2) that can be individually programmed via PVCxCON0.LEV1V and PVCxCON0.LEV2V (x = M or 1)PVC1CON0.LEV1V and PVC1CON0.LEV2V. The current supply level of a domain is compared with the threshold values. The two compare results can be monitored via bits PVCxCON0.LEV1OK and PVCxCON0.LEV2OK (x = M or 1) PVC1CON0.LEV1OK and PVC1CON0.LEV2OK A reset or interrupt request can be generated in case the core domain voltage level is below or equal / above the configured threshold level. An interrupt is requested if bit PVCxCON0.L1INTEN and / or PVCxCON0.L2INTEN (x = M or 1) PVC1CON0.L1INTEN and / or PVC1CON0.L2INTEN is set. A reset is requested if bit PVCxCON0.L1RSTEN and / or PVCxCON0.L2RSTEN (x = M or 1) PVC1CON0.L1RSTEN and / or PVC1CON0.L2RSTEN is set. Note: For a single threshold both interrupt and reset request generation should not be enabled at the same time. User’s Manual SCU, V1.15 8-93 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.2.1 PVC Status and Control Registers These registers are the software interface for PVC_1 and PVC_M. PVC1CON0 PVC_1 Control Step 0 Register ESFR (F014H/0AH) 15 L2 AS EN rw 14 13 12 11 10 L2 L2 L2 LEV RST INT A 2 EN EN LEV OK rw rw rw rh 9 LEV2V rw 8 7 L1 AS EN rw 6 Reset Value: 0504H 5 4 3 L1 L1 L1 LEV RST INT A 1 EN EN LEV OK rw rrw rw rh 2 1 0 LEV1V rw Field Bits Type Description LEV1V [2:0] rw Level Threshold 1 Voltage This bit field defines the Level Threshold 1 that is compared with the DMP_1 core voltage. The values for the different configurations are listed in the data sheet. LEV1OK 3 rh Level Threshold 1 Check Result 0B The core supply voltage of the DMP_1 is below Level Threshold 1 voltage LEV1V 1B The core supply voltage of the DMP_1 is equal or above the Level Threshold 1 voltage LEV1V L1ALEV 4 rw Level Threshold 1 Action Level 0B When the core supply voltage is below Level Threshold 1 voltage LEV1V the action configured by bits L1INTEN, L1RSTEN, and L1ASEN are requested 1B When the core supply voltage is equal or above Level Threshold 1 voltage LEV1V the actions configured by bits L1INTEN, L1RSTEN, and L1ASEN are requested L1INTEN 5 rw Level Threshold 1 Interrupt Request Enable This bit defines if an interrupt request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No interrupt is requested 1B An interrupt is requested User’s Manual SCU, V1.15 8-94 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L1RSTEN 6 rw Level Threshold 1 Reset Request Enable This bit defines if a reset request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No reset is requested 1B An reset is requested L1ASEN 7 rw Level Threshold 1 Asynchronous Action Enable This bit defines if asynchronous action can be performed if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No asynchronous actions are performed 1B Asynchronous actions can be performed LEV2V [10:8] rw Level Threshold 2 Voltage This bit field defines the level of threshold 2 that is compared with the DMP_1 core voltage.. The values for the different configurations are listed in the data sheet. LEV2OK 11 rh Level Threshold 2 Check Result 0B The core supply voltage of the DMP_1 is below the Level Threshold 2 LEV2V 1B The core supply voltage of the DMP_1 is equal or above the Level Threshold 2 LEV2V L2ALEV 12 rw Level Threshold 2 Action Level 0B When the core supply voltage is below the Level Threshold 2 voltage LEV2V the action configured by bits L2INTEN, L2RSTEN, and L2ASEN are requested When the core supply voltage is equal or 1B above the Level Threshold 2 voltage LEV2V the action configured by bits L2INTEN, L2RSTEN, and L2ASEN are requested L2INTEN 13 rw Level Threshold 2 Interrupt Request Enable This bit defines if an interrupt request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No interrupt is requested 1B An interrupt is requested User’s Manual SCU, V1.15 8-95 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L2RSTEN 14 rw Level Threshold 2 Reset Request Enable This bit defines if a reset request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No reset is requested 1B An reset is requested L2ASEN 15 rw Level Threshold 2 Asynchronous Action Enable This bit defines if asynchronous action can be performed if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No asynchronous actions are performed 1B Asynchronous actions can be performed User’s Manual SCU, V1.15 8-96 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PVCMCON0 PVC_M Control Step 0 Register MEM (F1E4H/--) 15 L2 AS EN rw 14 13 12 11 10 L2 L2 L2 LEV RST INT A 2 EN EN LEV OK rw rw rw rh 9 LEV2V rw 8 7 L1 AS EN rw 6 Reset Value: 0544H 5 4 3 L1 L1 L1 LEV RST INT A 1 EN EN LEV OK rw rw rw rh 2 1 0 LEV1V rw Field Bits Type Description LEV1V [2:0] rw Level Threshold 1 Voltage This bit field defines the Level Threshold 1 that is compared with the DMP_M core supply voltage. The values for the different configurations are listed in the data sheet. LEV1OK 3 rh Level Threshold 1 Check Result 0B The core supply voltage of the DMP_M is below Level Threshold 1 voltage LEV1V 1B The core supply voltage of the DMP_M is equal or above the Level Threshold 1 voltage LEV1V L1ALEV 4 rw Level Threshold 1 Action Level 0B When the core supply voltage is below Level Threshold 1 voltage LEV1V the action configured by bits L1INTEN, L1RSTEN, and L1ASEN are requested 1B When the core supply voltage is equal or above Level Threshold 1 voltage LEV1V the actions configured by bits L1INTEN, L1RSTEN, and L1ASEN are requested L1INTEN 5 rw Level Threshold 1 Interrupt Request Enable This bit defines if an interrupt request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No interrupt is requested 1B An interrupt is requested User’s Manual SCU, V1.15 8-97 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L1RSTEN 6 rw Level Threshold 1 Reset Request Enable This bit defines if a reset request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No reset is requested 1B An reset is requested L1ASEN 7 rw Level Threshold 1 Asynchronous Action Enable This bit defines if asynchronous action can be performed if the comparison check was successful. When a check is successful is defined via bit L1ALEV. 0B No asynchronous actions are performed 1B Asynchronous actions can be performed LEV2V [10:8] rw Level Threshold 2 Voltage This bit field defines the Level Threshold 2 that is compared with the DMP_M core supply voltage. The values for the different configurations are listed in the data sheet. LEV2OK 11 rh Level Threshold 2 Check Result 0B The core supply voltage of the DMP_M is below Level Threshold 2 voltage LEV2V 1B The core supply voltage of the DMP_M is equal or above the Level Threshold 2 voltage LEV2V L2ALEV 12 rw Level Threshold 2 Action Level 0B When the core supply voltage is below the Level Threshold 2 voltage LEV2V the action configured by bits L2INTEN, L2RSTEN, and L2ASEN are requested When the core supply voltage is equal or 1B above the Level Threshold 2 voltage LEV2V the action configured by bits L2INTEN, L2RSTEN, and L2ASEN are requested L2INTEN 13 rw Level Threshold 2 Interrupt Request Enable This bit defines if an interrupt request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No interrupt is requested 1B An interrupt is requested User’s Manual SCU, V1.15 8-98 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description L2RSTEN 14 rw Level Threshold 2 Reset Request Enable This bit defines if a reset request trigger is requested if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No reset is requested 1B An reset is requested L2ASEN 15 rw Level Threshold 2 Asynchronous Action Enable This bit defines if asynchronous action can be performed if the comparison check was successful. When a check is successful is defined via bit L2ALEV. 0B No asynchronous actions are performed 1B Asynchronous actions can be performed User’s Manual SCU, V1.15 8-99 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.3 Controlling the Voltage Level of a Core Domain The core powercan be controlled within certain limits. The voltage level is controlled by two Embedded Voltage Regulators (EVR). The power domain is controlled by both EVR_M and EVR_1. 8.6.3.1 Embedded Voltage Regulator The main part of the device logic operates at a typical voltage level of 1.5 V. This supply voltage is generated by the Embedded Power Regulators (EVRs) out of the pad voltage. External buffer caps are required for stable regulation. Feature list: • • • • Multiple core voltage levels including zero Core voltage generation based on a High Precision Bandgap External supply possible via capacitor-pin while EVR is switched-off Core current limit The EVR configurations to select the desired voltage and reference pair are combined within EVR settings EVRxSETyyV (x = M or 1 and yy = 15). Each setting contains a bit field (VRSEL) to select the voltage level and reference and a bit field to fine-tune the voltage level (VLEV). One out of the possible settings is used to control each of the EVRs, but only in the allowed combinations for the two EVRs. The EVRsusea High Precision Bandgap (HP) as reference The BG voltage of each setting can be adjusted to compensate application and environmental influences by the bit field EVRxSETyyV.VLEV. VLEV is set by default or trimmed by each device during production test to reach the default setting targets. High Precision Bandgap (HP) The HP bandgap of the system is used for following purposes: • • Provide a very stable reference for the two EVRs Provide an accurate reference for the flash memory. For more information see the flash memory description. The HP bandgap can be enabled / disabled via the bit EVRMCON1.HPEN. User’s Manual SCU, V1.15 8-100 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) EVR Status and Control Registers EVR1CON0 EVR_1 Control 0 Register ESFR (F088H/44H) 7 6 Reset Value: DF20H 15 14 13 12 11 10 9 8 5 4 3 2 1 EVR DIS - 0 CC DIS CCLEV - RES 1 0 RES0 0 rh - rw rh rw - rw rw rw r 0 Field Bits Type Description RES0 [5:3] rw Reserved Do not change this value when writing to this register. 0 [7:6] rw Reserved Must be written with reset value 00B. RES1 8 rw Reserved Must be written with reset value 1B. CCLEV [11:10] rw Current Control Level This bit field is required for enabling/disabling the current control (CCDIS). Valid values are described in the Programmer's Guide. CCDIS 12 rh Current Control Disable 0B The current control is enabled 1B The current control is disabled This bit is updated by bit EVR1SETy.CCDIS. 0 13 rw Reserved Must be written with reset value 0. EVRDIS 15 rh EVR_1 Disable 0B The EVR_1 is enabled 1B The EVR_1 is disabled This bit is updated by bit EVR1SETy.EVRDIS. 0 [2:0] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-101 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) EVR1SET15VHP EVR_1 Setting for 1.5 V HP Register ESFR (F09EH/4FH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 001BH 5 4 3 2 EVR DIS 0 CC DIS 0 RES 0 VRSEL VLEV rw rw rw rw rw rw rw rw 1 0 Field Bits Type Description VLEV [5:0] rw Voltage Level Adjust This bit field adjusts the BG voltage and is trimmed by each device during production test to reach the default setting targets. Do not change this value when writing to this register. VRSEL [7:6] rw Voltage Reference Selection 00B 15VHP - Full Voltage with high precision bandgap selected 01B Reserved, do not use this combination 10B Reserved, do not use this combination 11B Reserved, do not use this combination Note: The reset value should always be written to this bit field. RES 9 rw Reserved Must be written with 1B. CCDIS 12 rw Current Control Disable 0B The current control is enabled 1B The current control is disabled This bit updates bit EVR1CON0.CCDIS. Note: Before switching off the current control the CCLEV setting in EVR1CON0 has to be set to 00B. EVRDIS User’s Manual SCU, V1.15 15 rw EVR_1 Disable 0B The EVR_1 is enabled 1B The EVR_1 is disabled This bit updates bit EVR1CON0.EVRDIS. 8-102 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits 0 8, rw [11:10], [14:13] User’s Manual SCU, V1.15 Type Description Reserved Must be written with reset value 0. 8-103 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) EVRMCON0 EVR_M Control 0 Register 15 14 13 ESFR (F084H/42H) 12 11 10 9 8 7 6 Reset Value: 0D20H 5 4 3 2 1 EVR DIS 0 CC DIS CCLEV - RES 1 0 RES0 0 rh r rh rw - rw rw rw r 0 Field Bits Type Description RES0 [5:3] rw Reserved Do not change this value when writing to this register. 0 [7:6] rw Reserved Must be written with reset value 00B. RES1 8 rw Reserved Must be written with 1. CCLEV [11:10] rw Current Control Level This bit field is required for enabling/disabling the current control (CCDIS). Valid values are described in the Programmer's Guide. CCDIS 12 rh Current Control Disable 0B The current control is enabled 1B The current control is disabled This bit is updated by bit EVRMSETy.CCDIS. EVRDIS 15 rh EVR_M Disable 0B The EVR_M is enabled 1B The EVR_M is disabled This bit is updated by bit EVRMSETy.EVRDIS. 0 [2:0], 8 r [14:13] User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-104 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) EVRMCON1 EVR_M Control 1 Register 15 14 13 12 11 ESFR (F086H/43H) 10 9 8 7 6 Reset Value: 0101H 5 4 3 0 HP EN HPADJUST r rw rw 2 1 0 Field Bits Type Description HPADJUST [7:0] rw HP Bandgap Adjustment This bit field is a device specific trimmvalue for the HP bandgap. Do not change this value when writing to this register. HPEN 8 rw HP Bandgap Enable 0B The HP bandgap is disabled 1B The HP bandgap is enabled 0 [15:9] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-105 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) EVRMSET15VHP EVR_M Setting for 1.5 V HP Register ESFR (F096H/4BH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 001BH 5 4 3 2 EVR DIS 0 CC DIS 0 RES 0 0 VRSEL VLEV rw rw rw rw rw rw rw rw 1 0 Field Bits Type Description VLEV [5:0] rw Voltage Level Adjust This bit field adjusts the BG voltage and is trimmed by each device during production test to reach the default setting targets. Do not change this value when writing to this register. VRSEL [7:6] rw Voltage Reference Selection 00B 15VHP - Full Voltage with high precision bandgap selected 01B Reserved, do not use this combination 10B Reserved, do not use this combination 11B Reserved, do not use this combination Note: The reset value should always be written to this bit field. RES0 9 rw Reserved Must be written with 1B. CCDIS 12 rw Current Control Disable 0B The current control is enabled 1B The current control is disabled This bit updates bit EVRMCON0.CCDIS. Note: Before switching off the current control the CCLEV setting in EVRMCON0 has to be set to 00B. EVRDIS User’s Manual SCU, V1.15 15 rw EVR_M Disable 0B The EVR_M is enabled 1B The EVR_M is disabled This bit updates bit EVR1CON0.EVRDIS. 8-106 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits 0 8, rw [11:10], [14:13] User’s Manual SCU, V1.15 Type Description Reserved Must be written with reset value 0. 8-107 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.3.2 Sources for Core Supply Voltage The on-chip EVRs can generate the XC27x5X’s core supply voltage from the (externally supplied) IO voltage. Core Supply via On-chip EVRs Generating the core supply voltage via the integrated EVRs is the preferred operating mode, because it saves an additional external voltage regulator. The integrated EVRs are fed from supply voltage VDDPB. Proper operation of the EVRs requires external buffer capacitances. Please refer to the respective Data Sheet for the recommended values. The current is delivered by the integrated pass devices. Microcontroller Ext. Voltage Reg. VDDPB Onchip EVR VDDI CEVR V SS MC_XC2X_POWER _EVR Figure 8-22 Selecting the EVR for Core Supply Generating the core supply voltage with on-chip resources provides full control of power reduction modes, so the application can control and minimize the energy consumption of the XC27x5X using built-in mechanisms without requiring additional external circuitry. User’s Manual SCU, V1.15 8-108 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.6.4 Handling the Power System Using the power system correctly is the key to power saving. Depending on the application different operating states can be defined in order to save maximal power. The XC27x5X supports following power saving mechanisms: • • • • Reduction of the system performance – the power consumption depends directly on the frequency of the system – the system performance is controlled with the clock operation mechanism Stopping single unused peripheral – a peripheral not needed for an application can be disabled – the module operation is controlled via register MOD_KSCCFG Stopping multiple unused peripherals – peripherals not needed for an application can be disabled – system peripheral operation is controlled via the Global State Controller (GSC) Stopping single unused analog parts – an analog part not needed for an application can be stopped – the operation is controlled via register either located in the SCU (PLL, clocks, PVCs, SWD, Temperature Compensation) or the ADC User’s Manual SCU, V1.15 8-109 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.7 Global State Controller (GSC) Mode Control for the system peripherals provides besides power saving modes and the clock management an additional opportunity for configuring the system to the application needs. Mode Control is described in detail in this chapter and is implemented by the Global State Controller (GSC). The GSC enables the user to configure one operating mode in a fast and easy way, reacting fast and explicit to needs of an application. Feature Overview The following issues are handled by the GSC: • • • Control of peripheral clock operation Suspend control for debugging Arbitration between the different request sources According to the requests coming from the OCDS, the SWD pre-warning detection or other blocks, the GSC does an internal prioritization. The result is forwarded as command request broadcast to all peripherals. The GSC internal prioritization scheme for the implemented request sources is shown in Table 8-8. 8.7.1 GSC Control Flow The sequence begins when at least one request source asserts its trigger in order to request a mode change in the SoC. If several requests are pending there is an arbitration mechanism that treats this issue. Request triggers are not stored by the GSC, therefore a trigger source has to assert its trigger until the trigger is no longer valid or needed. A request trigger is kept asserted as long as either the request is still pending or the resulting command of the request was entered and acknowledged by the system. The communication of the GSC and the peripherals is based on commands. Three different commands are defined resulting in three modes: • • • Wake-up command: requests Normal Mode Clock-off command: requests Stop Mode Debug command: requests Suspend Mode The specific behavior in these three modes is defined for each peripheral in its module register mod_KSCCFG. 8.7.1.1 Request Source Arbitration The highest priority for the arbitration is zero (see Table 8-8). Each system clock cycle a new arbitration round is started. The winner of an arbitration round requests the next command towards the SoC. Please note that winning an arbitration does not lead automatically to a new command raised. Only if currently no command is broadcast in the SoC a new command can be generated and broadcast. If User’s Manual SCU, V1.15 8-110 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) the winner of the arbitration round is the same request trigger as in the previous round or if no winner was detected no new command request is generated. Table 8-8 Connection of the Request Sources Request Source Priority OCDS exit 4 ESR0 5 ESR1 6 ESR2 7 WUT 8 ITC 9 GPT12E 10 SW1 11 SW2 12 OCDS entry 14 8.7.1.2 Generation of a New Command When a new request trigger was detected and arbitrated a new command request is generated if currently no command request is broadcast that is not received by all slaves. Table 8-9 Request Source and Command Request Coupling Request Source Command Description OCDS exit Wake-up; Normal Mode ESR0 Wake-up; Normal Mode ESR1 Wake-up; Normal Mode ESR2 Clock-off; Stop Mode WUT Wake-up; Normal Mode ITC Wake-up; Normal Mode GPT12E Wake-up; Normal Mode SW1 Wake-up; Normal Mode SW2 Clock-off; Stop Mode OCDS entry Debug; Suspend Mode User’s Manual SCU, V1.15 8-111 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.7.1.3 Usage of Commands The complete control mechanism for the different operation modes of the various slaves is divided into two parts: • • A central control and configuration part; the Global State Controller (GSC) One local control part in each slave; the Kernel State Controller (KSC) Via the GSC either different hardware sources (e.g. the WUT or the OCDS) or the software can request the system to enter a specific mode. The parts that are affected by the mode can be pre-defined locally for each part via the KSC. For each command a specific reaction can be pre-configured in each KSC for each individual part. Note: Requesting a peripheral to be permanently shut off by clearing mod_KSCCFG.MODEN to 0 does not start a GSC run. However, a GSC run triggered in parallel to the ramp down of this peripheral will have the FSM of the GSC waiting for an acknowledge also of this peripheral as long as the peripheral does no deliver its acknowledge or the respective bit in GSCPERSTATEN is cleared. The proposal is either to disable automatic GSC runs (by setting GSCEN respectively) in case the application needs the information of the shutdown acknowledge of the peripheral or to disbale the respective bit in GSCPERSTATEN, so that the missing acknowlegde is not taken into account. Note: When a GSC mode request has been successfully entered, the GSC arbiter is open for any new request. In case a request occurs to enter the current mode, this request is pipelined and remains pending. It is recommended to request a command by a software trigger. In particular the clock-off command should be triggered by SW2. The usage of commands requested by hardware has to be done carefully. Only hardware resources requesting Normal Mode should be selected. If the software has detected a wakeup then pending mode change requests can be removed by clearing the bits of the selected sources in GSCEN and then enabling the bits in GSCEN again. 8.7.1.4 Terminating a Request Trigger A request trigger is no longer taken into account for the arbitration after the de-assertion of the request trigger, if it is not enabled or when its respective enable bit is cleared. 8.7.1.5 Suspend Control Flow The suspend feature is controlled by the OCDS block. The GSC operates only as control and communication interface towards the system. The suspend feature is composed out of two requirements: The mode that has to be entered when the Suspend Mode is requested. The mode that has to be entered when the Suspend Mode is left. User’s Manual SCU, V1.15 8-112 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) The request to enter Suspend Mode is forwarded from the OCDS. When the Suspend Mode is requested the system is expected to be stopped as soon as possible in an idle state where no internal process is pending and in a way that this system state does not lead to any damage internally or externally and can also be left without any damage. Therefore all peripherals in the system are requested to enter a mode where the clock can be stopped. This is done by sending a debug command. Leaving the Suspend Mode should serve the goal that debugging is a non-intrusive operation. Therefore leaving the Suspend Mode can not lead to only one dedicated system mode, instead it leads to the system mode the system left when it was requested to exit the Suspend Mode. The system mode is stored when a Suspend Mode request is detected by the GSC and is used as target system mode when a leave Suspend Mode trigger is detected by the GSC. 8.7.1.6 Error Feedback for a Mode Transition In case at least one peripheral reports an error the error flag in register GSCSTAT is set. If no error is currently detected upon a new assertion of a system mode by the GSC the error flag is cleared. To inform the system of this erroneous state an interrupt can be generated. User’s Manual SCU, V1.15 8-113 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.7.2 GSC Registers 8.7.2.1 GSC Control and Status Registers The following register control and configure the behavior of the GSC. GSCSWREQ GSC Software Request Register SFR (FF14H/8AH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 SWT SWT RG2 RG1 0 r rwh rwh Field Bits Type Description SWTRG1 0 rwh Software Trigger 1 (SW1) 0B No SW1 request trigger is generated 1B A SW1 request trigger is generated This bit is automatically cleared if the SW1 request trigger wins the arbitration and was broadcast to the system. SWTRG2 1 rwh Software Trigger 2 (SW2) 0B No SW2 request trigger is generated A SW2 request trigger is generated 1B This bit is automatically cleared if the SW2 request trigger wins the arbitration and was broadcast to the system. 0 [15:2] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-114 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) GSCEN GSC Enable Register 15 0 r 14 13 12 SFR (FF16H/8BH) 11 10 9 8 7 6 Reset Value: 7FFFH 5 4 OCD ESR ESR ESR OCD RES SW2 SW1 GPT ITC WUT SEN 2 1 0 SEX 1 EN EN EN EN EN EN EN EN EN EN rw rw rw rw rw rw rw rw rw rw rw 3 2 1 0 RES0 rw Field Bits Type Description RES0 [3:0] rw Reserved Must be written with 0000B. OCDSEXEN 4 rw OCDS Exit Request Trigger Enable 0B OCDS exit request trigger is not taken into account (disabled) 1B OCDS exit request trigger is taken into account (enabled) ESR0EN 5 rw ESR0 Request Trigger Enable 0B ESR0 request trigger is not taken into account (disabled) 1B ESR0 request trigger is taken into account (enabled) ESR1EN 6 rw ESR1 Request Trigger Enable 0B ESR1 request trigger is not taken into account (disabled) 1B ESR1 request trigger is taken into account (enabled) ESR2EN 7 rw ESR2 Request Trigger Enable 0B ESR2 request trigger is not taken into account (disabled) 1B ESR2 request trigger is taken into account (enabled) WUTEN 8 rw WUT Request Trigger Enable 0B WUT request trigger is not taken into account (disabled) 1B WUT request trigger is taken into account (enabled) User’s Manual SCU, V1.15 8-115 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ITCEN 9 rw ITC Request Trigger Enable 0B ITC request trigger is not taken into account (disabled) 1B ITC request trigger is taken into account (enabled) GPTEN 10 rw GTP12E Request Trigger Enable 0B GPT12E request trigger is not taken into account (disabled) 1B GPT12E request trigger is taken into account (enabled) SW1EN 11 rw Software 1 Request Trigger Enable SW1 request trigger is not taken into account 0B (disabled) 1B SW1 request trigger is taken into account (enabled) SW2EN 12 rw Software 2 Request Trigger Enable 0B SW2 request trigger is not taken into account (disabled) 1B SW2 request trigger is taken into account (enabled) RES1 13 rw Reserved Read as 1 after reset; returns the value that is written. OCDSENEN 14 rw OCDS Entry Request Trigger Enable 0B OCDS entry request trigger is not taken into account (disabled) 1B OCDS entry request trigger is taken into account (enabled) OCDS entry is the request source belonging to the according connector interface. 0 15 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-116 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) GSCSTAT GSC Status Register 15 14 13 12 SFR (FF18H/8CH) 11 0 SOURCE r rh 10 9 8 7 PEN ERR rh rh 6 Reset Value: 3C00H 5 4 3 2 1 0 0 NEXT 0 CURRENT r rh r rh Field Bits Type Description CURRENT [1:0] rh Currently used Command This bit field states the currently used system mode. NEXT [5:4] rh Next to use Command This bit field states the next to be used system mode. ERR 8 rh Error Status Flag This bit flags if with the last command that was broadcast was acknowledge with at least one error. This bit is automatically cleared when a new command is broadcast. PEN 9 rh Command Pending Flag This flag states if currently a command is pending or not. A command is pending after the broadcast as long as no all blocks acknowledge that they finished the operation requested by the command. User’s Manual SCU, V1.15 8-117 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits SOURCE [13:10] rh Requesting Source Status This bit field monitors the source that triggered the last request. 0000B Reserved 0001B Reserved 0010B Reserved 0011B Reserved 0100B OCDS exit 0101B ESR0 0110B ESR1 0111B ESR2 1000B WUT 1001B ITC 1010B GPT12E 1011B SW1 1100B SW2 1101B Reserved, do not use this combination 1110B OCDS entry 1111B Reserved, do not use this combination 0 [3:2], r [7:6], [15:14] Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 Type Description 8-118 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) GSCPERSTATEN GSC Peripheral Status Enable Register SFR (FF04H/82H) 15 14 13 12 11 10 9 8 7 6 Reset Value: FFFFH 5 4 3 2 1 0 USIC USIC USIC USIC CCU CCU CCU CCU M RES GPT FL MEM RTC CC2 ADC 3 2 1 0 63 62 61 60 CAN 0 12E rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw rw Field Bits Type Description ADC 0 rw ADC Acknowledge Enable This bit defines if the acknowledge status of ADC modules is taken into account and displayed or ignored. 0B The ADC modules acknowledge is ignored, it is treated as always asserted 1B The ADC modules acknowledge is used GPT12E 1 rw GPT12E Acknowledge Enable This bit defines if the acknowledge status of GPT12E module is taken into account and displayed or ignored. 0B The GPT12E module acknowledge is ignored, it is treated as always asserted 1B The GPT12E module acknowledge is used RES0 2 rw Reserved Must be written with 0B. CC2 3 rw CC2 Acknowledge Enable This bit defines if the acknowledge status of CC2 module is taken into account and displayed or ignored. The CC2 module acknowledge is ignored, it is 0B treated as always asserted 1B The CC2 module acknowledge is used User’s Manual SCU, V1.15 8-119 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description MCAN 4 rw MultiCAN Acknowledge Enable This bit defines if the acknowledge status of MultiCAN module is taken into account and displayed or ignored. 0B The MultiCAN module acknowledge is ignored, it is treated as always asserted 1B The MultiCAN module acknowledge is used CCU60 5 rw CCU60 Acknowledge Enable This bit defines if the acknowledge status of CCU60 module is taken into account and displayed or ignored. 0B The CCU60 module acknowledge is ignored, it is treated as always asserted 1B The CCU60 module acknowledge is used CCU61 6 rw CCU61 Acknowledge Enable This bit defines if the acknowledge status of CCU61 module is taken into account and displayed or ignored. 0B The CCU61 module acknowledge is ignored, it is treated as always asserted 1B The CCU61 module acknowledge is used CCU62 7 rw CCU62 Acknowledge Enable This bit defines if the acknowledge status of CCU62 module is taken into account and displayed or ignored. 0B The CCU62 module acknowledge is ignored, it is treated as always asserted 1B The CCU62 module acknowledge is used CCU63 8 rw CCU63 Acknowledge Enable This bit defines if the acknowledge status of CCU63 module is taken into account and displayed or ignored. 0B The CCU63 module acknowledge is ignored, it is treated as always asserted 1B The CCU63 module acknowledge is used User’s Manual SCU, V1.15 8-120 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description RTC 9 rw RTC Acknowledge Enable This bit defines if the acknowledge status of RTC module is taken into account and displayed or ignored. 0B The RTC module acknowledge is ignored, it is treated as always asserted 1B The RTC module acknowledge is used MEM 10 rw C166SV2 Subsystem Acknowledge Enable This bit defines if the acknowledge status of C166SV2 Subsystem module is taken into account and displayed or ignored. 0B The C166SV2 Subsystem module acknowledge is ignored, it is treated as always asserted 1B The C166SV2 Subsystem module acknowledge is used FL 11 rw Flash Acknowledge Enable This bit defines if the acknowledge status of Flash module is taken into account and displayed or ignored. The Flash module acknowledge is ignored, it is 0B treated as always asserted 1B The Flash module acknowledge is used USIC0 12 rw USIC0 Acknowledge Enable This bit defines if the acknowledge status of USIC0 module is taken into account and displayed or ignored. 0B The USIC0 module acknowledge is ignored, it is treated as always asserted 1B The USIC0 module acknowledge is used USIC1 13 rw USIC1 Acknowledge Enable This bit defines if the acknowledge status of USIC1 module is taken into account and displayed or ignored. 0B The USIC1 module acknowledge is ignored, it is treated as always asserted 1B The USIC1 module acknowledge is used User’s Manual SCU, V1.15 8-121 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description USIC2 14 rw USIC2 Acknowledge Enable This bit defines if the acknowledge status of USIC2 module is taken into account and displayed or ignored. 0B The USIC2 module acknowledge is ignored, it is treated as always asserted 1B The USIC2 module acknowledge is used USIC3 15 rw USIC3 Acknowledge Enable This bit defines if the acknowledge status of USIC3 module is taken into account and displayed or ignored. 0B The USIC3 module acknowledge is ignored, it is treated as always asserted 1B The USIC3 module acknowledge is used User’s Manual SCU, V1.15 8-122 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) GSCPERSTAT GSC Peripheral Status Register SFR (FF1AH/8DH) 15 14 13 12 11 10 9 8 7 6 Reset Value: FFFFH 5 4 3 USIC USIC USIC USIC CCU CCU CCU CCU M FL MEM RTC CC2 3 2 1 0 63 62 61 60 CAN rh rh rh rh rh rh rh rh rh rh rh rh rh 2 1 r 1 0 GPT ADC 12E rh rh Field Bits Type Description ADC 0 rh ADC Acknowledge Status This bit shows the acknowledge status of the modules ADC. The modules ADC change currently their 0B kernel state. Their acknowledge has not been received. 1B Acknowledge of the modules ADC is taken into account and has been received or is not relevant. GPT12E 1 rh GPT12E Acknowledge Status This bit shows the acknowledge status of the modules GPT12E. 0B The module GPT12E changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module GPT12E is taken into account and has been received or is not relevant. CC2 3 rh CC2 Acknowledge Status This bit shows the acknowledge status of the module CC2. The module CC2 changes currently its kernel 0B state. Its acknowledge has not been received. 1B Acknowledge of the module CC2 is taken into account and has been received or is not relevant. User’s Manual SCU, V1.15 8-123 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description MCAN 4 rh MultiCAN Acknowledge Status This bit shows the acknowledge status of the module MultiCAN. 0B The module MultiCAN changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module MultiCAN is taken into account and has been received or is not relevant. CCU60 5 rh CCU60 Acknowledge Status This bit shows the acknowledge status of the module CCU60. 0B The module CCU60 changes currently its kernel state. Its acknowledge has not been received. Acknowledge of the module CCU60 is taken 1B into account and has been received or is not relevant. CCU61 6 rh CCU61 Acknowledge Status This bit shows the acknowledge status of the module CCU61. 0B The module CCU61 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module CCU61 is taken into account and has been received or is not relevant. CCU62 7 rh CCU62 Acknowledge Status This bit shows the acknowledge status of the module CCU62. 0B The module CCU62 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module CCU62 is taken into account and has been received or is not relevant. User’s Manual SCU, V1.15 8-124 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description CCU63 8 rh CCU63 Acknowledge Status This bit shows the acknowledge status of the module CCU63. 0B The module CCU63 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module CCU63 is taken into account and has been received or is not relevant. RTC 9 rh RTC Acknowledge Status This bit shows the acknowledge status of the module RTC. 0B The module RTC changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module RTC is taken into account and has been received or is not relevant. MEM 10 rh C166SV2 Subsystem Acknowledge Status This bit shows the acknowledge status of the module C166SV2 Subystem. The module C166SV2 Subystem changes 0B currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module C166SV2 Subystem is taken into account and has been received or is not relevant. FL 11 rh Flash Acknowledge Status This bit shows the acknowledge status of the module Flash. 0B The module Flash changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module Flash is taken into account and has been received or is not relevant. User’s Manual SCU, V1.15 8-125 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description USIC0 12 rh USIC0 Acknowledge Status This bit shows the acknowledge status of the module USIC0. 0B The module USIC0 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module USIC0 is taken into account and has been received or is not relevant. USIC1 13 rh USIC1 Acknowledge Status This bit shows the acknowledge status of the module USIC1. 0B The module USIC1 changes currently its kernel state. Its acknowledge has not been received. Acknowledge of the module USIC1 is taken 1B into account and has been received or is not relevant. USIC2 14 rh USIC2 Acknowledge Status This bit shows the acknowledge status of the module USIC2. 0B The module USIC2 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module USIC2 is taken into account and has been received or is not relevant. USIC3 15 rh USIC3 Acknowledge Status This bit shows the acknowledge status of the module USIC3. 0B The module USIC3 changes currently its kernel state. Its acknowledge has not been received. 1B Acknowledge of the module USIC3 is taken into account and has been received or is not relevant. 1 2 r Reserved Read as 1 User’s Manual SCU, V1.15 8-126 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) The acknowledge status bit is set, if acknowledge of the module x is taken into account and has been received or is not relvant. The acknowledge of the module x is not relevant if the acknowledge of the module x is not taken into account or the module x currently does not undergo a change of kernel state mode. In these cases the acknowledge is assumed to be received. User’s Manual SCU, V1.15 8-127 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.8 Software Boot Support In order to determine the correct starting point of operation for the software a minimum of hardware support is required. As much as possible is done via software. Some decisions have to be done in hardware because they must be known before any software is operational. 8.8.1 Start-up Registers 8.8.1.1 Start-up Status Register Register STSTAT contains the information required by the boot software to identify the different start-up settings that can be selected. STSTAT Start-up Status Register 15 14 13 12 11 MEM (F1E0H/--) 10 9 8 7 6 Reset Value: 8000H 5 4 3 1 0 HWCFG r r rh 2 1 0 Field Bits Type Description HWCFG [7:0] rh Hardware Configuration Setting This bit field contains the value that is used by the boot software. This bit field is updated in case of an Application Reset with the content by register SWRSTCON.SWCFG if bit SWRSTCON.SWBOOT is set. 0 [14:8] r Reserved Read as 0; should be written with 0. 1 15 r Reserved Read as 1; should be written with 1. User’s Manual SCU, V1.15 8-128 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9 External Request Unit (ERU) The External Request Unit (ERU) is a versatile event and pattern detection unit. Its major task is the generation of interrupts based on selectable trigger events at different inputs, e.g. to generate external interrupt requests if an edge occurs at an input pin. The detected events can also be used by other modules to trigger or to gate modulespecific actions, such as conversions of the ADC module. 8.9.1 Introduction The ERU of the XC27x5X can be split in three main functional parts: • • • 4 independent Input Channels x for input selection and conditioning of trigger or gating functions Event distribution: A Connecting Matrix defines the events of the Input Channel x that lead to a reaction of an Output Channel y. 4 independent Output Channels y for combination of events, definition of their effects and distribution to the system (interrupt generation, ADC conversion triggers) External Request Unit from pins or modules ERU_0A[3:0] ERU_0B[3:0] 4 4 External Request Select Units ERS0 Event Trigger Logic Units Output Gating Units ETL0 OGU 0 Output Channel 0 4 4 ERS1 Connecting Matrix ERU_1B[3:0] ETL1 Input Channel 1 ERU_2A[3:0] ERU_2B[3:0] 4 4 ERS2 ETL2 ERU_3B[3:0] ERU_PDOUT1 ERU_GOUT1 ERU_IOUT1 ERU_TOUT1 OGU 1 Output Channel 1 ERU_PDOUT2 ERU_GOUT2 ERU_IOUT2 OGU 2 ERU_TOUT2 Output Channel 2 Input Channel 2 ERU_3A[3:0] ERU_PDOUT0 ERU_GOUT0 ERU_IOUT0 ERU_TOUT0 Input Channel 0 ERU_1A[3:0] to interrupt controller, ADC 4 4 ERS3 ERU_PDOUT3 ERU_GOUT3 ERU_IOUT3 ERU_TOUT3 OGU 3 ETL3 Input Channel 3 Output Channel 3 Peripheral Triggers ERU_Block _Overview .vsd Figure 8-23 External Request Unit Overview These tasks are handled by the following building blocks: • An External Request Select Unit (ERSx) per Input Channel allows the selection of one out of two or a logical combination of two input signals (ERU_xA, ERU_xB) to a User’s Manual SCU, V1.15 8-129 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) • • • common trigger. For each of these two signals, an input vector of 4 possible inputs is available (e.g. the actual input ERU_xA can be selected from one of the ERU inputs ERU_xA[3:0], similar for ERU_xB). An Event Trigger Logic (ETLx) per Input Channel allows the definition of the transition (edge selection, or by software) that lead to a trigger event and can also store this status. Here, the input levels of the selected signals are translated into events (event detected = event flag becomes set, independent of the polarity of the original input signals). The Connecting Matrix distributes the events and status flags generated by the Input Channels to the Output Channels. Additionally, some Peripheral Trigger signals from other modules (e.g. CC2) are made available and can be combined with the triggers generated by the Input Channels of the ERU. An Output Gating Unit (OGUy) per Output Channel that combines the available trigger events and status information from the Input Channels. An event of one Input Channel can lead to reactions of several Output Channels, or also events of several Input Channels can be combined to a reaction of one Output Channel (pattern detection). Different types of reactions are possible, e.g. interrupt generation (based on signals ERU_IOUTy), triggering of ADC conversions (based on signals ERU_TOUTy), or gating of ADC conversions (based on signals ERU_GOUTy). The ERU is controlled by a number of registers, shown in Figure 8-24, and described in Section 8.9.8. Input Selection Input & Trigger Control Output Control EXISEL EXICON0 EXOCON0 EXICON1 EXOCON1 EXICON2 EXOCON2 EXICON3 EXOCON3 EXISEL: EXICON0..3: EXOCON0..3: Input Selection Register Input and Trigger Control Registers Output Control Registers ERU_Register_Overview.vsd Figure 8-24 ERU Registers Overview User’s Manual SCU, V1.15 8-130 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.2 ERU Input Connections The inputs to the ERU can be selected from a large number of input signals. While some of the inputs come directly from a pin, other inputs use signals from various peripheral modules, such as the USIC (signals named with prefix UxCy to indicate the communication channel) and the MultiCAN modules. These signals are input signals from the pin that has been selected as input for a USIC or MultiCAN function. The selection of the input is made within the respective USIC or MultiCAN module. In the ERU input connections are described in Section 8.18.3.1. Usually, such signals would be selected for an ERU function when the input function to the USIC or MultiCAN module is not used otherwise, or the module is not used at all. However, it is also possible to select a input which is actually needed in a USIC or MultiCAN module, and to use it also in the ERU to provide a certain trigger functions, eventually combined with other signals (e.g. to generate an interrupt trigger in case a start of frame is detected at a selected communication). With this structure, the number of possible input pins is significantly increased, because not only the selection capability of the ERU is used, but also the selection capability of the communication modules. Note: All functional inputs of the ERU are synchronized to fSYS before they can affect the internal logic. The resulting delay of 2/fSYS and an uncertainty of 1/fSYS have to be taken into account for precise timing calculation. An edge of an input can only be correctly detected if both, the high phase and the low phase of the input are each longer than 1/fSYS. User’s Manual SCU, V1.15 8-131 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.3 External Request Select Unit (ERSx) For each Input Channel x (x = 0-3), an ERSx unit handles the input selection for the associated ETLx unit. Each ERSx performs a logical combination of two signals (Ax, Bx) to provide one combined output ERSxO to the associated ETLx. Input Ax can be selected from 4 options of the input vector ERU_xA[3:0] and can be optionally inverted. A similar structure exists for input Bx (selection from ERU_xB[3:0]). In addition to the direct choice of either input Ax or Bx or their inverted values, the possible logical combinations for two selected inputs are a logical AND or a logical OR. ERU_xA0 ERU_xA1 ERU_xA2 ERU_xA3 EXISEL. EXSxA EXICONx. NA Select Input Ax Select Polarity Ax EXICONx. SS Ax Bx 1 Ax OR Bx ERU_xB0 ERU_xB1 ERU_xB2 Select Input Bx Select Polarity Bx EXISEL. EXSxB EXICONx. NB & Ax AND Bx Select Source for ERSxO ERSxO ETLx ERU_xB3 External Request Select Unit x (ERSx) ERU_ERS.vsd Figure 8-25 External Request Select Unit Overview The ERS units are controlled via register EXISEL (one register for all four ERSx units) and registers EXICONx (one register for each ERSx and associated ETLx unit, e.g. EXICON0 for Input Channel 0). User’s Manual SCU, V1.15 8-132 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.4 Event Trigger Logic (ETLx) For each Input Channel x (x = 0-3), an event trigger logic ETLx derives a trigger event and a status from the input ERUxO delivered by the associated ERSx unit. Each ETLx is based on an edge detection block, where the detection of a rising or a falling edge can be individually enabled. Both edges lead to a trigger event if both enable bits are set (e.g. to handle a toggling input). Each of the four ETLx units has an associated EXICONx register, that controls all options of an ETLx (the register also holds control bits for the associated ERSx unit, e.g. EXICON0 to control ESR0 and ETL0). EXICONx. FE EXICONx. LD Modify Status Flag ERSx ERSxO Detect Event (edge) Event Trigger Logic x (ETLx) set clear Status Flag FL edge event Enable Trigger Pulse EXICONx.FL to all OGUy TRx0 to OGU0 trigger pulse Select Trigger Output TRx1 to OGU1 TRx2 to OGU2 TRx3 to OGU3 EXICONx. RE EXICONx. PE EXICONx. OCS ERU_ETL.vsd Figure 8-26 Event Trigger Logic Overview When the selected event (edge) is detected, the status flag EXICONx.FL becomes set. This flag can also be modified by software (set or clear). Two different operating modes are supported by this status flag. It can be used as “sticky” flag, that is set by hardware when the desired event has been detected and has to be cleared by software. In this operating mode, it indicates that the event has taken place, but without indicating the actual status of the input. In the second operating mode, it is cleared automatically if the “opposite” event is detected. For example, if only the falling edge detection is enabled to set the status flag, it is cleared when the rising edge is detected. In this mode, it can be used for pattern User’s Manual SCU, V1.15 8-133 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) detection where the actual status of the input is important (enabling both edge detections is not useful in this mode). The output of the status flag is connected to all following Output Gating Units (OGUy) in parallel (see Figure 8-27) to provide pattern detection capability of all OGUy units based on different or the same status flags. In addition to the modification of the status flag, a trigger pulse output TRxy of ETLx can be enabled (by bit EXICONx.PE) and selected to trigger actions in one of the OGUy units. The target OGUy for the trigger is selected by bit field EXICON.OCS. The trigger becomes active when the selected edge event is detected, independently from the status flag EXICONx.FL. User’s Manual SCU, V1.15 8-134 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.5 Connecting Matrix The connecting matrix distributes the trigger signals (TRxy) and status signals (EXICONx.FL) from the different ETLx units between the OGUy units. In addition, it receives peripheral trigger signals that can be OR-combined with the ETLx trigger signals in the OGUy units. Figure 8-27 provides a complete overview of the connections between the ETLx and the OGUy units. EXICON0.FL TR00 ETL0 TR01 TR02 TR10 TR11 TR12 TR13 ERU_IOUT0 ERU_GOUT0 TR20 TR21 TR22 Peripheral Triggers Pattern Detection Inputs ERU_PDOUT1 OGU1 ERU_IOUT1 Trigger Inputs TRx1 EXICON2.FL ERU_GOUT1 ERU_TOUT1 Peripheral Triggers Pattern Detection Inputs ERU_PDOUT2 OGU2 ERU_IOUT2 ERU_GOUT2 ERU_TOUT2 TR23 EXICON3.FL TR30 ETL3 OGU0 Trigger Inputs TRx0 EXICON1.FL ETL2 ERU_PDOUT0 ERU_TOUT0 TR03 ETL1 Pattern Detection Inputs TR31 TR32 Trigger Inputs TRx2 Peripheral Triggers Pattern Detection Inputs ERU_PDOUT3 OGU3 ERU_IOUT3 Trigger Inputs TRx3 TR33 ERU_GOUT3 ERU_TOUT3 Peripheral Triggers ERU_ETL_OGU_Overview.vsd Figure 8-27 Connecting Matrix between ETLx and OGUy User’s Manual SCU, V1.15 8-135 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.6 Output Gating Unit (OGUy) Each OGUy (y = 0-3) unit combines the available trigger events and status flags from the Input Channels and distributes the results to the system. Figure 8-28 illustrates the logic blocks within an OGUy unit. All functions of an OGUy unit are controlled by its associated EXOCONy register, e.g. EXOCON0 for OGU0. The function of an OGUy unit can be split into two parts: • • Trigger combination (see Section 8.9.6.1): All trigger signals TRxy from the Input Channels that are enabled and directed to OGUy, a selected peripheral-related trigger event, and a pattern change event (if enabled) are logically OR-combined. Pattern detection (see Section 8.9.6.2): The status flags EXICONx.FL of the Input Channels can be enabled to take part in the pattern detection. A pattern match is detected while all enabled status flags are set. Status Flags EXICON0.FL EXOCONy. IPEN0 EXOCONy. GEEN EXICON1.FL EXOCONy. IPEN1 EXICON2.FL ERU_PDOUTy Detect Pattern EXOCONy. PDR EXOCONy. IPEN2 Select Gating Scheme EXICON3.FL Triggers from Input Channels EXOCONy. IPEN3 TR0y ERU_GOUTy Combine OGU Triggers (OR) TR1y TR2y TR3y EXOCONy. GP Interrupt Gating (AND) ERU_IOUTy ERU_TOUTy ERU_OGUy1 Peripheral Triggers ERU_OGUy2 ERU_OGUy3 Select Periph. Triggers EXOCONy. ISS Output Gating Unit y (OGUy) ERU_OGU.vsd Figure 8-28 Output Gating Unit for Output Channel y User’s Manual SCU, V1.15 8-136 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Each OGUy unit generates 4 output signals that are distributed to the system (not all of them are necessarily used, please refer to Section 8.9.7): • • • • ERU_PDOUTy to directly output the pattern match information for gating purposes in other modules (pattern match = 1). ERU_GOUTy to output the pattern match or pattern miss information (inverted pattern match), or a permanent 0 or 1 under software control for gating purposes in other modules. ERU_TOUTy as combination of a peripheral trigger, a pattern detection result change event, or the ETLx trigger outputs TRxy to trigger actions in other modules. ERU_IOUTy as gated trigger output (ERU_GOUTy logical AND-combined with ERU_TOUTy) to trigger interrupts (e.g. the interrupt generation can be gated to allow interrupt activation during a certain time window). 8.9.6.1 Trigger Combination The trigger combination logically OR-combines different trigger inputs to form a common trigger ERU_TOUTy. Possible trigger inputs are: • • • In each ETLx unit of the Input Channels, the trigger output TRxy can be enabled and the trigger event can be directed to one of the OGUy units. One out of three peripheral trigger signals per OGUy can be selected as additional trigger source. These peripheral triggers are generated by on-chip peripheral modules, such as capture/compare or timer units. The selection is done by bit field EXOCONy.ISS. In the case that at least one pattern detection input is enabled (EXOCONy.IPENx) and a change of the pattern detection result from pattern match to pattern miss (or vice-versa) is detected, a trigger event is generated to indicate a pattern detection result event (if enabled by ECOCONy.GEEN). The trigger combination offers the possibility to program different trigger criteria for several input signals (independently for each Input Channel) or peripheral signals, and to combine their effects to a single output, e.g. to generate an interrupt or to start an ADC conversion. This combination capability allows the generation of an interrupt per OGU that can be triggered by several inputs (multitude of request sources -> one reaction). Section 8.18.3.2 describes the peripheral trigger connections for the OGUy stages. The selection is defined by the bit fields ISS in registers EXOCON0 (for OGU0), EXOCON1 (for OGU1), EXOCON2 (for OGU2), or EXOCON3 (for OGU3). 8.9.6.2 Pattern Detection The pattern detection logic allows the combination of the status flags of all ETLx units. Each status flag can be individually included or excluded from the pattern detection for each OGUy, via control bits EXOCONy.IPENx. The pattern detection block outputs the following pattern detection results: User’s Manual SCU, V1.15 8-137 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) • • Pattern match (EXOCONy.PDR = 1 and ERU_PDOUTy = 1): A pattern match is indicated while all status flags FL that are included in the pattern detection are 1. Pattern miss (EXOCONy.PDR = 0 and ERU_PDOUTy = 0): A pattern miss is indicated while at least one of the status flags FL that are included in the pattern detection is 0. In addition, the pattern detection can deliver a trigger event if the pattern detection result changes from match to miss or vice-versa (if enabled by EXOCONy.GEEN = 1). The pattern result change event is logically OR-combined with the other enabled trigger events to support interrupt generation or to trigger other module functions (e.g. in the ADC). The event is indicated when the pattern detection result changes and EXOCONy.PDR becomes updated. The interrupt generation in the OGUy is based on the trigger ERU_TOUTy that can be gated (masked) with the pattern detection result ERU_PDOUTy. This allows an automatic and reproducible generation of interrupts during a certain time window, where the request event is elaborated by the trigger combination block and the time window information (gating) is given by the pattern detection. For example, interrupts can be issued on a regular time base (peripheral trigger input from capture/compare unit is selected) while a combination of input signals occurs (pattern detection based on ETLx status bits). A programmable gating scheme introduces flexibility to adapt to application requirements and allows the generation of interrupt requests ERU_IOUTy under different conditions: • • • • Pattern match (EXOCONy.GP = 10B): An interrupt request is issued when a trigger event occurs while the pattern detection shows a pattern match. Pattern miss (EXOCONy.GP = 11B): An interrupt request is issued when the trigger event occurs while the pattern detection shows a pattern miss. Independent of pattern detection (EXOCONy.GP = 01B): In this mode, each occurring trigger event leads to an interrupt request. The pattern detection output can be used independently from the trigger combination for gating purposes of other peripherals (independent use of ERU_TOUTy and ERU_PDOUTy with interrupt requests on trigger events). No interrupts (EXOCONy.GP = 00B, default setting) In this mode, an occurring trigger event does not lead to an interrupt request. The pattern detection output can be used independently from the trigger combination for gating purposes of other peripherals (independent use of ERU_TOUTy and ERU_PDOUTy without interrupt requests on trigger events). User’s Manual SCU, V1.15 8-138 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.7 ERU Output Connections Section 8.18.3.3 describes the connections of the ERU output signals for gating or triggering other module functions, as well as the connections to the interrupt control registers. User’s Manual SCU, V1.15 8-139 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.8 ERU Registers 8.9.8.1 External Input Selection Register EXISEL This register selects the A and B inputs for all four ERS units. The possible input signals are given in Table 8-19. EXISEL External Input Select Register ESFR (F1A0H/D0H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 EXS3B EXS3A EXS2B EXS2A EXS1B EXS1A EXS0B EXS0A rw rw rw rw rw rw rw rw Field Bits Type Description EXS0A [1:0] rw External Source Select for A0 (ERS0) This bit field defines which input is selected for A0. 00B Input ERU_0A0 is selected 01B Input ERU_0A1 is selected 10B Input ERU_0A2 is selected 11B Input ERU_0A3 is selected EXS0B [3:2] rw External Source Select for B0 (ERS0) This bit field defines which input is selected for B0. 00B Input ERU_0B0 is selected 01B Input ERU_0B1 is selected 10B Input ERU_0B2 is selected 11B Input ERU_0B3 is selected EXS1A [5:4] rw External Source Select for A1 (ERS1) This bit field defines which input is selected for A1. 00B Input ERU_1A0 is selected 01B Input ERU_1A1 is selected 10B Input ERU_1A2 is selected 11B Input ERU_1A3 is selected EXS1B [7:6] rw External Source Select for B1 (ERS1) This bit field defines which input is selected for B1. 00B Input ERU_1B0 is selected 01B Input ERU_1B1 is selected 10B Input ERU_1B2 is selected 11B Input ERU_1B3 is selected User’s Manual SCU, V1.15 8-140 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits EXS2A [9:8] Type Description rw External Source Select for A2 (ERS2) This bit field defines which input is selected for A2. 00B Input ERU_2A0 is selected 01B Input ERU_2A1 is selected 10B Input ERU_2A2 is selected 11B Input ERU_2A3 is selected EXS2B [11:10] rw External Source Select for B2 (ERS2) This bit field defines which input is selected for B2. 00B Input ERU_2B0 is selected 01B Input ERU_2B1 is selected 10B Input ERU_2B2 is selected 11B Input ERU_2B3 is selected EXS3A [13:12] rw External Source Select for A3 (ERS3) This bit field defines which input is selected for A3. 00B Input ERU_3A0 is selected 01B Input ERU_3A1 is selected 10B Input ERU_3A2 is selected 11B Input ERU_3A3 is selected EXS3B [15:14] rw External Source Select for B3 (ERS3) This bit field defines which input is selected for B3. 00B Input ERU_3B0 is selected 01B Input ERU_3B1 is selected 10B Input ERU_3B2 is selected 11B Input ERU_3B3 is selected User’s Manual SCU, V1.15 8-141 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.8.2 External Input Control Registers EXICONx These registers control the inputs of the ERSx unit and the trigger functions of the ETLx units (x = 0..3). EXICON0 External Input Control 0 Register ESFR (F030H/18H) EXICON1 External Input Control 1 Register ESFR (F032H/19H) EXICON2 External Input Control 2 Register ESFR (F034H/1AH) EXICON3 External Input Control 3 Register ESFR (F036H/1CH) 15 14 13 12 11 10 9 8 7 0 NB NA SS FL r rw rw rw rwh 6 Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H 5 4 3 2 1 0 OCS FE RE LD PE rw rw rw rw rw Field Bits Type Description PE 0 rw Output Trigger Pulse Enable for ETLx This bit enables the generation of an output trigger pulse at TRxy when the selected edge is detected (set condition for the status flag FL). 0B The trigger pulse generation is disabled 1B The trigger pulse generation is enabled User’s Manual SCU, V1.15 8-142 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description LD 1 rw Rebuild Level Detection for Status Flag for ETLx This bit selects if the status flag FL is used as “sticky” bit or if it rebuilds the result of a level detection. 0B The status flag FL is not cleared by hardware and is used as “sticky” bit. Once set, it is not influenced by any edge until it becomes cleared by software. 1B The status flag FL rebuilds a level detection of the desired event. It becomes automatically set with a rising edge if RE = 1 or with a falling edge if FE = 1. It becomes automatically cleared with a rising edge if RE = 0 or with a falling edge if FE = 0. RE 2 rw Rising Edge Detection Enable ETLx This bit enables/disables the rising edge event as edge event as set condition for the status flag FL or as possible trigger pulse for TRxy. A rising edge is not considered as edge event 0B 1B A rising edge is considered as edge event FE 3 rw Falling Edge Detection Enable ETLx This bit enables/disables the falling edge event as edge event as set condition for the status flag FL or as possible trigger pulse for TRxy. 0B A falling edge is not considered as edge event 1B A falling edge is considered as edge event OCS [6:4] rw Output Channel Select for ETLx Output Trigger Pulse This bit field defines which Output Channel OGUy is targeted by an enabled trigger pulse TRxy. 000B Trigger pulses are sent to OGU0 001B Trigger pulses are sent to OGU1 010B Trigger pulses are sent to OGU2 011B Trigger pulses are sent to OGU3 1XXB Reserved, do not use this combination FL 7 rwh Status Flag for ETLx This bit represents the status flag that becomes set or cleared by the edge detection. 0B The enabled edge event has not been detected 1B The enabled edge event has been detected User’s Manual SCU, V1.15 8-143 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description SS [9:8] rw Input Source Select for ERSx This bit field defines which logical combination is taken into account as ESRxO. 00B Input A without additional combination 01B Input B without additional combination 10B Input A OR input B 11B Input A AND input B NA 10 rw Input A Negation Select for ERSx This bit selects the polarity for the input A. Input A is used directly 0B 1B Input A is inverted NB 11 rw Input B Negation Select for ERSx This bit selects the polarity for the input B. 0B Input B is used directly 1B Input B is inverted 0 [15:12] r User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-144 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.9.8.3 Output Control Registers EXOCONy These registers control the outputs of the Output Gating Unit y (y = 0..3). EXOCON0 External Output Trigger Control 0 Register SFR (FE30H/18H) EXOCON1 External Output Trigger Control 1 Register SFR (FE32H/19H) EXOCON2 External Output Trigger Control 2 Register SFR (FE34H/1AH) EXOCON3 External Output Trigger Control 3 Register SFR (FE36H/1BH) 15 14 13 12 11 IPEN IPEN IPEN IPEN 3 2 1 0 rw rw rw rw 10 9 8 7 6 Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H 5 4 3 2 1 0 0 GP PDR GE EN ISS r rw rh rw rw Field Bits Type Description ISS [1:0] rw Internal Trigger Source Selection This bit field defines which input is selected as peripheral trigger input for OGUy. The possible input signals are given in Table 8-20. 00B The peripheral trigger function is disabled 01B Input ERU_OGUy1 is selected 10B Input ERU_OGUy2 is selected 11B Input ERU_OGUy3 is selected GEEN 2 rw Gating Event Enable Bit GEEN enables the generation of a trigger event when the result of the pattern detection changes from match to miss or vice-versa. 0B The event detection is disabled 1B The event detection is enabled PDR 3 rh Pattern Detection Result Flag This bit represents the pattern detection result. 0B A pattern miss is detected 1B A pattern match is detected User’s Manual SCU, V1.15 8-145 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description GP [5:4] rw Gating Selection for Pattern Detection Result This bit field defines the gating scheme for the interrupt generation (relation between the OGU output ERU_PDOUTy and ERU_GOUTy). 00B ERU_GOUTy is always disabled and ERU_IOUTy can not be activated 01B ERU_GOUTy is always enabled and ERU_IOUTy becomes activated with each activation of ERU_TOUTy 10B ERU_GOUTy is equal to ERU_PDOUTy and ERU_IOUTy becomes activated with an activation of ERU_TOUTy while the desired pattern is detected (pattern match PDR = 1) 11B ERU_GOUTy is inverted to ERU_PDOUTy and ERU_IOUTy becomes activated with an activation of ERU_TOUTy while the desired pattern is not detected (pattern miss PDR = 0) IPENx (x = 0-3) 12+x rw Pattern Detection Enable for ETLx Bit IPENx defines whether the trigger event status flag EXICONx.FL of ETLx takes part in the pattern detection of OGUy. Flag EXICONx.FL is excluded from the pattern 0B detection 1B Flag EXICONx.FL is included in the pattern detection 0 [11:6] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-146 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10 SCU Interrupt Generation The interrupt structure of the SCU is shown in Figure 8-29. INTCLR.x Interrupt Event DMPMIT.x clear INTSTAT.x INTSET.x 1 1 request INTNPy.x & 1 set 1 disable 1 INTDIS.x SCU int 0 SCU int 1 SCU int 2 SCU int 3 other interrupt sources to the same ITC node SCU Interrupt Structure to ITC node 6C H to ITC node 6B H * to ITC node 1E * to ITC node 1F H H * Connection can be enabled via register ISSR SCU_Int _Struct _MR+.vsd Figure 8-29 SCU Interrupt Structure If enabled by the corresponding bit in register INTDIS, an interrupt is triggered either by the incoming interrupt request line, or by a software set of the respective bit in register INTSET. The trigger sets the respective flag in register INTSTAT and is gated to one of the interrupt nodes, selected by the node pointer registers INTNP0 or INTNP1. The interrupt flag can be cleared by software by writing to the corresponding bit in register INTCLR. If more than one interrupt source is connected to the same interrupt node pointer (in register INTNPx), the requests are combined to one common line. Interrupt Node Assignment The interrupt sources of the SCU module can be mapped to the dedicated interrupt node 6CH or 6BH by programming the interrupt node pointer registers INTNP0 and INTNP1. Furthermore, If the CAPCOM2 interrupts for channels 30 or 31 are not used the SCU interrupts can be mapped via register ISSR to the interrupt nodes 1EH or 1FH which are assigned to the CAPCOM2 interrupts. So for the SCU interrupts can be selected the interrupt node 6CH, 6BH, or in addition via register ISSR the node 1EH or 1FH. The default assignment of the interrupt sources to the nodes and their corresponding control registers are shown in Table 8-10. User’s Manual SCU, V1.15 8-147 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.1 Interrupt Support Some of the interrupt requests are first fed through a sticky flag register in the DMP_M domain. These flags are set with a trigger and if set trigger the interrupt generation in the DMP_1.. Please note that the disable control of register INTDIS also influences the sticky bit in register DMPMIT (see Section 8.10.3.6). Which of the interrupt requests have a sticky flag in register DMPMIT is listed in Table 8-10. Note: When servicing an SCU interrupt request, make sure that all related request flags are cleared after the identified request has been handled. To clear an interrupt request that is stored in register DMPMIT, first clear the request source of the source (e.g. WUTRG), clear the request within DMP_M via DMPMITCLR, and then clear the request within DMP_1 via INTCLR. 8.10.2 SCU Interrupt Sources The SCU receives the interrupt request lines listed in Table 8-10. Table 8-10 SCU Interrupt Overview Source of Interrupt Short Name Sticky Flag Default Interrupt Node in DMPMIT Assignment in INTNPx SWD OK 1 Interrupt SWDI1 yes 6CH SWD OK 2 Interrupt SWDI2 yes 6BH PVC_M OK 1 Interrupt PVCMI1 yes 6CH PVC_M OK 2 Interrupt PVCMI2 yes 6BH PVC_1 OK 1 Interrupt PVC1I1 yes 6CH PVC_1 OK 2 Interrupt PVC1I2 yes 6BH Wake-up Timer Interrupt WUTI yes 6BH Wake-up Trim Interrupt WUI yes 6CH Watchdog Timer Interrupt WDTI --- 6BH GSC Interrupt GSCI yes 6CH STM0 Interrupt STM0I yes 6BH STM1 Interrupt STM1I yes 6CH MCHK Interrupt MCHKI --- 6BH Program Flash Interrupt PFI --- 6CH User’s Manual SCU, V1.15 8-148 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3 Interrupt Control Registers 8.10.3.1 Register INTSTAT This register contains the status flags for all interrupt request trigger sources of the SCU. For setting and clearing of these status bits by software see registers INTSET and INTCLR, respectively. INTSTAT Interrupt Status Register 15 14 0 rh 13 12 11 SFR (FF00H/80H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 M STM STM PVC PVC PVC PVC PF GSC WDT WU WUT SWD SWD CHK 1 0 1 1 M M I I I I I I2 I1 I I I I2 I1 I2 I1 rh rh rh rh rh rh rh rh rh rh rh rh rh rh Field Bits Type Description SWDI1 0 rh SWD Interrupt Request Flag 1 This bit is set if bit DMPMIT.SWDI1 is set. 0B No SWDI1 interrupt trigger has occured since this bit was cleared the last time 1B A SWDI1 interrupt trigger has occured since this bit was cleared the last time SWDI2 1 rh SWD Interrupt Request Flag 2 This bit is set if bit DMPMIT.SWDI2 is set. 0B No SWDI2 interrupt trigger has occured since this bit was cleared the last time 1B A SWDI2 interrupt trigger has occured since this bit was cleared the last time PVCMI1 2 rh PVC_M Interrupt Request Flag 1 This bit is set if bit DMPMIT.PVCMI1 is set. 0B No PVCMI1 interrupt trigger has occured since this bit was cleared the last time 1B A PVCMI1 interrupt trigger has occured since this bit was cleared the last time User’s Manual SCU, V1.15 8-149 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PVCMI2 3 rh PVC_M Interrupt Request Flag 2 This bit is set if bit DMPMIT.PVCMI2 is set. 0B No PVCMI2 interrupt trigger has occured since this bit was cleared the last time A PVCMI2 interrupt trigger has occured since 1B this bit was cleared the last time PVC1I1 4 rh PVC_1 Interrupt Request Flag 1 This bit is set if bit DMPMIT.PVC1I1 is set. 0B No PVC1I1 interrupt trigger has occured since this bit was cleared the last time A PVCI1 interrupt trigger has occured since 1B this bit was cleared the last time PVC1I2 5 rh PVC_1 Interrupt Request Flag 2 This bit is set if bit DMPMIT.PVC1I2 is set. 0B No PVC1I2 interrupt trigger has occured since this bit was cleared the last time 1B A PVC1I2 interrupt trigger has occured since this bit was cleared the last time WUTI 6 rh Wake-up Timer Trim Interrupt Request Flag This bit is set if the WUT trim trigger event occur and bit is INTDIS.WUTI = 0. 0B No WUT interrupt trigger has occured since this bit was cleared the last time 1B A WUT interrupt trigger has occured since this bit was cleared the last time WUI 7 rh Wake-up Timer Interrupt Request Flag This bit is set if the WU trigger event occur and bit is INTDIS.WUI = 0. No WU interrupt trigger has occured since this 0B bit was cleared the last time 1B A WU interrupt trigger has occured since this bit was cleared the last time WDTI 8 rh Watchdog Timer Interrupt Request Flag This bit is set if the WDT Prewarning Mode is entered and bit is INTDIS.WDTI = 0. 0B No WDT interrupt trigger has occured since this bit was cleared the last time 1B A WDT interrupt trigger has occured since this bit was cleared the last time User’s Manual SCU, V1.15 8-150 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description GSCI 9 rh GSC Interrupt Request Flag This bit is set if the GSC error bit is set and bit is INTDIS.GSCI = 0. 0B No GSC interrupt trigger has occured since this bit was cleared the last time 1B A GSC interrupt trigger has occured since this bit was cleared the last time STM0I 10 rh STM Interrupt 0 Request Flag This bit is set if the STM interrupt trigger 0 is set and bit is INTDIS.STM0I = 0. 0B No STM0 interrupt trigger has occured since this bit was cleared the last time 1B A STM0 interrupt trigger has occured since this bit was cleared the last time STM1I 11 rh STM Interrupt 1 Request Flag This bit is set if the STM interrupt trigger 1 is set and bit is INTDIS.STM1I = 0. No STM1 interrupt trigger has occured since 0B this bit was cleared the last time 1B A STM1 interrupt trigger has occured since this bit was cleared the last time MCHKI 12 rh MCHK Interrupt Request Flag This bit is set if the MCHK interrupt trigger is set and bit is INTDIS.MCHKI = 0. 0B No MCHK interrupt trigger has occured since this bit was cleared the last time 1B A MCHK interrupt trigger has occured since this bit was cleared the last time PFI 13 rh Program Flash Interrupt Request Flag This bit is set if the Program Flash interrupt trigger is set and bit is INTDIS.PFI = 0. No PF interrupt trigger has occured since this 0B bit was cleared the last time 1B A PF interrupt trigger has occured since this bit was cleared the last time 0 [15:14] rh User’s Manual SCU, V1.15 Reserved Read as 0; should be written with 0. 8-151 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.2 Register INTCLR This register contains the software clear option for all status flags of all interrupt request trigger sources of the SCU. INTCLR Interrupt Clear Register 15 14 0 w 13 12 SFR (FE82H/41H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 M STM STM PVC PVC PVC PVC PF GSC WDT WU WUT SWD SWD CHK 1 0 1 1 M M I I I I I I2 I1 I I I I2 I1 I2 I1 w w w w w w w w w w w w w w Field Bits Type Description SWDI1 0 w Clear SWD Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.SWDI1 is cleared SWDI2 1 w Clear SWD Interrupt Request Flag 2 No action 0B 1B Bit INTSTAT.SWDI2 is cleared PVCMI1 2 w Clear PVC_M Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.PVCMI1 is cleared PVCMI2 3 w Clear PVC_M Interrupt Request Flag 2 0B No action 1B Bit INTSTAT.PVCMI2 is cleared PVC1I1 4 w Clear PVC_1 Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.PVC1I1 is cleared PVC1I2 5 w Clear PVC_1 Interrupt Request Flag 2 0B No action 1B Bit INTSTAT.PVC1I2 is cleared WUTI 6 w Clear Wake-up Trim Interrupt Request Flag 0B No action 1B Bit INTSTAT.WUTI is cleared WUI 7 w Clear Wake-up Interrupt Request Flag 0B No action 1B Bit INTSTAT.WUI is cleared User’s Manual SCU, V1.15 8-152 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WDTI 8 w Clear Watchdog Timer Interrupt Request Flag 0B No action 1B Bit INTSTAT.WDTI is cleared GSCI 9 w Clear GSC Interrupt Request Flag No action 0B Bit INTSTAT.GSCI is cleared 1B STM0I 10 w Clear STM0 Interrupt Request Flag 0B No action 1B Bit INTSTAT.STM0I is cleared STM1I 11 w Clear STM1 Interrupt Request Flag 0B No action 1B Bit INTSTAT.STM1I is cleared MCHKI 12 w Clear MCHK Interrupt Request Flag 0B No action 1B Bit INTSTAT.MCHKI is cleared PFI 13 w Clear Program Flash Interrupt Request Flag 0B No action 1B Bit INTSTAT.PFI is cleared 0 [15:14] w User’s Manual SCU, V1.15 Reserved Must be written with 0. 8-153 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.3 Register INTSET This register contains the software set option for all status flags of all interrupt request trigger sources of the SCU. INTSET Interrupt Set Register 15 14 0 w 13 12 SFR (FE80H/40H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 M STM STM PVC PVC PVC PVC PF GSC WDT WU WUT SWD SWD CHK 1 0 1 1 M M I I I I I I2 I1 I I I I2 I1 I2 I1 w w w w w w w w w w w w w w Field Bits Type Description SWDI1 0 w Set SWD Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.SWDI1 is set SWDI2 1 w Set SWD Interrupt Request Flag 2 No action 0B 1B Bit INTSTAT.SWDI2 is set PVCMI1 2 w Set PVC_M Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.PVCMI1 is set PVCMI2 3 w Set PVC_M Interrupt Request Flag 2 0B No action 1B Bit INTSTAT.PVCMI2 is set PVC1I1 4 w Set PVC_1 Interrupt Request Flag 1 0B No action 1B Bit INTSTAT.PVC1I1 is set PVC1I2 5 w Set PVC_1 Interrupt Request Flag 2 0B No action 1B Bit INTSTAT.PVC1I2 is set WUTI 6 w Set Wake-up Trim Interrupt Request Flag 0B No action 1B Bit INTSTAT.WUTI is set WUI 7 w Set Wake-up Interrupt Request Flag 0B No action 1B Bit INTSTAT.WUI is set User’s Manual SCU, V1.15 8-154 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WDTI 8 w Set Watchdog Timer Interrupt Request Flag 0B No action 1B Bit INTSTAT.WDTI is set GSCI 9 w Set GSC Interrupt Request Flag No action 0B Bit INTSTAT.GSCI is set 1B STM0I 10 w Set STM0 Interrupt Request Flag 0B No action 1B Bit INTSTAT.STM0I is set STM1I 11 w Set STM1 Interrupt Request Flag 0B No action 1B Bit INTSTAT.STM1I is set MCHKI 12 w Set MCHK Interrupt Request Flag 0B No action 1B Bit INTSTAT.MCHKI is set PFI 13 w Set Program Flash Interrupt Request Flag 0B No action 1B Bit INTSTAT.PFI is set 0 [15:14] w User’s Manual SCU, V1.15 Reserved Must be written with 0. 8-155 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.4 Register INTDIS This register contains the software disable control for all interrupt request trigger sources of the SCU. INTDIS Interrupt Disable Register 15 14 13 12 11 SFR (FE84H/42H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 M STM STM PVC PVC PVC PVC PF GSC WDT WU WUT SWD SWD CHK 1 0 1 1 M M I I I I I I2 I1 I I I I2 I1 I2 I1 rw rw rw rw rw rw rw rw rw rw rw rw rw rw 0 rw Field Bits Type Description SWDI1 0 rw Disable SWD Interrupt Request 1 0B SWDI1 interrupt request enabled 1B SWDI1 interrupt request disabled SWDI2 1 rw Disable SWD Interrupt Request 2 SWDI2 interrupt request enabled 0B 1B SWDI2 interrupt request disabled PVCMI1 2 rw Disable PVC_M Interrupt Request 1 0B PVCMI1 interrupt request enabled 1B PVCMI1 interrupt request disabled PVCMI2 3 rw Disable PVC_M Interrupt Request 2 0B PVCMI2 interrupt request enabled 1B PVCMI2 interrupt request disabled PVC1I1 4 rw Disable PVC_1 Interrupt Request 1 0B PVC1I1 interrupt request enabled 1B PVC1I1 interrupt request disabled PVC1I2 5 rw Disable PVC_1 Interrupt Request 2 0B PVC1I2 interrupt request enabled 1B PVC1I2 interrupt request disabled WUTI 6 rw Disable Wake-up Trim Interrupt Request 0B WUT interrupt request enabled 1B WUT interrupt request disabled WUI 7 rw Disable Wake-up Interrupt Request 0B WU interrupt request enabled 1B WU interrupt request disabled User’s Manual SCU, V1.15 8-156 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description WDTI 8 rw Disable Watchdog Timer Interrupt Request 0B WDT interrupt request enabled 1B WDT interrupt request disabled GSCI 9 rw Disable GSC Interrupt Request GSC interrupt request enabled 0B GSC interrupt request disabled 1B STM0I 10 rw Disable STM0 Interrupt Request 0B STM0 interrupt request enabled 1B STM0 interrupt request disabled STM1I 11 rw Disable STM1 Interrupt Request 0B STM1 interrupt request enabled 1B STM1 interrupt request disabled MCHKI 12 rw Disable MCHK Interrupt Request 0B MCHK interrupt request enabled 1B MCHK interrupt request disabled PFI 13 rw Disable Program Flash Interrupt Request 0B PF interrupt request enabled 1B PF interrupt request disabled 0 [15:14] rw User’s Manual SCU, V1.15 Reserved Should be written with 0. 8-157 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.5 Registers INTNP0 and INPNP1 These registers contain the control for the interrupt node pointers of all interrupt request trigger sources of the SCU. INTNP0 Interrupt Node Pointer 0 Register SFR (FE86H/43H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 4444H 5 4 3 2 1 0 WU WUT PVC12 PVC11 PVCM2 PVCM1 SWD2 SWD1 rw rw rw rw rw rw rw rw Field Bits Type Description SWD1 [1:0] rw Interrupt Node Pointer for SWD 1 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.SWDI1 (if enabled by bit INTDIS.SWDI1). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) SWD2 [3:2] rw Interrupt Node Pointer for SWD 2 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.SWDI2 (if enabled by bit INTDIS.SWDI2). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) User’s Manual SCU, V1.15 8-158 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PVCM1 [5:4] rw Interrupt Node Pointer for PVC_M 1 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.PCVMI1 (if enabled by bit INTDIS.PVCMI1). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) PVCM2 [7:6] rw Interrupt Node Pointer for PVC_M 2 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.PCVMI2 (if enabled by bit INTDIS.PVCMI2). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) PVC11 [9:8] rw Interrupt Node Pointer for PVC_1 1 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.PCV1I1 (if enabled by bit INTDIS.PVC1I1). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) PVC12 [11:10] rw Interrupt Node Pointer for PVC_1 2 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.PCV1I2 (if enabled by bit INTDIS.PVC1I2). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) User’s Manual SCU, V1.15 8-159 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits WUT [13:12] rw Interrupt Node Pointer for WU Trim Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.WUTI (if enabled by bit INTDIS.WUTI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) WU [15:14] rw Interrupt Node Pointer for WU Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.WUI (if enabled by bit INTDIS.WUI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) User’s Manual SCU, V1.15 Type Description 8-160 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) INTNP1 Interrupt Node Pointer 1 Register SFR (FE88H/44H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 1111H 5 4 3 2 1 0 - PF MCHK STM1 STM0 GSC WDT - rw rw rw rw rw rw Field Bits Type Description WDT [1:0] rw Interrupt Node Pointer for WDT Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.WDTI (if enabled by bit INTDIS.WDTI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) GSC [3:2] rw Interrupt Node Pointer for GSC Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.GSCI (if enabled by bit INTDIS.GSCI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) STM0 [5:4] rw Interrupt Node Pointer for STM0 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.STM0I (if enabled by bit INTDIS.STM0I). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) User’s Manual SCU, V1.15 8-161 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description STM1 [7:6] rw Interrupt Node Pointer for STM1 Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.STM1I (if enabled by bit INTDIS.STM1I). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) MCHK [9:8] rw Interrupt Node Pointer for MCHK Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.MCHKI (if enabled by bit INTDIS.MCHKI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) [11:10] rw Interrupt Node Pointer for Program Flash Interrupts This bit field defines the interrupt node, which is requested due to the set condition for bit INTSTAT.PFI (if enabled by bit INTDIS.PFI). 00B Interrupt node 6CH is selected 01B Interrupt node 6BH is selected 10B Interrupt node 1EH is selected if enabled by bit ISSR.ISS14 (bit is set) 11B Interrupt node 1FH is selected if enabled by bit ISSR.ISS15 (bit is set) PF User’s Manual SCU, V1.15 8-162 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.6 Register DMPMIT This register contains additional sticky interrupt and trap flags within the DMP_M power domain. DMPMIT DMP_M Interrupt and Trap Trigger Register SFR (FE96H/4BH) 15 14 RA T 0 rh r 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 ESR ESR ESR PVC PVC PVC PVC STM STM WU WUT SWD SWD 2 1 0 GSC 1 1 M M 1 0 I I I2 I1 T T T I2 I1 I2 I1 rh rh rh rh rh rh rh rh rh rh rh rh rh rh Field Bits Type Description SWDI1 0 rh SWD Interrupt Request Flag 1 This bit is set if bit SWDCON0.L1OK is cleared and SWDCON0.L1INTEN = 1B and INTDIS.SWDI1 = 0. 0B No SWDI1 interrupt was requested since this bit was cleared the last time 1B A SWDI1 interrupt was requested since this bit was cleared the last time SWDI2 1 rh SWD Interrupt Request Flag 2 This bit is set if bit SWDCON0.L2OK is cleared and SWDCON0.L2INTEN = 1B and INTDIS.SWDI2 = 0. 0B No SWDI2 interrupt was requested since this bit was cleared the last time 1B A SWDI2 interrupt was requested since this bit was cleared the last time PVCMI1 2 rh PVC_M Interrupt Request Flag 1 This bit is set if bit PVCMCON0.LEV1OK is cleared and PVCMCON0.L1INTEN = 1B and INTDIS.PVCMI1 = 0. 0B No PVCMI1 interrupt was requested since this bit was cleared the last time 1B A PVCMI1 interrupt was requested since this bit was cleared the last time User’s Manual SCU, V1.15 8-163 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PVCMI2 3 rh PVC_M Interrupt Request Flag 2 This bit is set if bit PVCMCON0.LEV2OK is cleared and PVCMCON0.L2INTEN = 1B and INTDIS.PVCMI2 = 0. No PVCMI2 interrupt was requested since this bit was 0B cleared the last time 1B A PVCMI2 interrupt was requested since this bit was cleared the last time PVC1I1 4 rh PVC_1 Interrupt Request Flag 1 This bit is set if bit PVC1CON0.LEV1OK is cleared and PVC1CON0.L1INTEN = 1B and INTDIS.PVC1I1 = 0. 0B No PVC1I1 interrupt was requested since this bit was cleared the last time 1B A PVC1I1 interrupt was requested since this bit was cleared the last time PVC1I2 5 rh PVC_1 Interrupt Request Flag 2 This bit is set if bit PVC1CON0.LEV2OK is cleared and PVC1CON0.L2INTEN = 1B and INTDIS.PVC1I2 = 0. No PVC1I2 interrupt was requested since this bit was 0B cleared the last time 1B A PVC1I2 interrupt was requested since this bit was cleared the last time WUTI 6 rh Wake-up Trim Interrupt Request Flag This bit is set if a wake-up trim trigger occurs and INTDIS.WUTI = 0. 0B No WUT interrupt was requested since this bit was cleared the last time 1B A WUT interrupt was requested since this bit was cleared the last time WUI 7 rh Wake-up Interrupt Request Flag This bit is set if a wake-up trigger occurs and INTDIS.WUI = 0. 0B No WU interrupt was requested since this bit was cleared the last time 1B A WU interrupt was requested since this bit was cleared the last time User’s Manual SCU, V1.15 8-164 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description GSC 8 rh GSC Interrupt Request Flag This bit is set if a GSC trigger occurs and INTDIS.GSCI = 0. 0B No GSC interrupt was requested since this bit was cleared the last time A GSC interrupt was requested since this bit was 1B cleared the last time STM0 9 rh STM0 Interrupt Request Flag This bit is set if a STM0 trigger occurs and INTDIS.STM0I = 0. 0B No STM0 interrupt was requested since this bit was cleared the last time A STM0 interrupt was requested since this bit was 1B cleared the last time STM1 10 rh STM1 Interrupt Request Flag This bit is set if a STM1 trigger occurs and INTDIS.STM1I = 0. 0B No STM1 interrupt was requested since this bit was cleared the last time 1B A STM1 interrupt was requested since this bit was cleared the last time ESR0T 11 rh ESR0 Trap Request Flag This bit is set if pin ESR0 is asserted. 0B No ESR0 trap was requested since this bit was cleared the last time 1B An ESR0 trap was requested since this bit was cleared the last time ESR1T 12 rh ESR1 Trap Request Flag This bit is set if pin ESR1 is asserted. 0B No ESR1 trap was requested since this bit was cleared the last time 1B An ESR1 trap was requested since this bit was cleared the last time ESR2T 13 rh ESR2 Trap Request Flag This bit is set if pin ESR2 is asserted. 0B No ESR2 trap was requested since this bit was cleared the last time 1B An ESR2 trap was requested since this bit was cleared the last time User’s Manual SCU, V1.15 8-165 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description RAT 15 rh Register Access Trap Request Flag This bit is set a protected register is written by an nonauthorized access. 0B No RA trap was requested since this bit was cleared the last time 1B A RA trap was requested since this bit was cleared the last time 0 14 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-166 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.10.3.7 Register DMPMITCLR This register contains the software clear option for all sticky status flags of all interrupt and trap request trigger sources of the DMP_M power domain. DMPMITCLR DMP_M Interrupt and Trap Clear Register SFR (FE98H/4CH) 15 14 RA T 0 w r 13 12 11 10 9 8 ESR ESR ESR STM STM 2 1 0 GSC 1 0 T T T w w w w w w 7 6 W UI WUT I w w Reset Value: 0000H 5 4 3 Bits Type Description SWDI1 0 w Clear SWD1 Interrupt Request Flag 1 0B No action 1B Bit DMPMIT.SWDI1 is cleared SWDI2 1 w Clear SWD Interrupt Request Flag 2 0B No action 1B Bit DMPMIT.SWDI2 is cleared PVCMI1 2 w Clear PVC_M Interrupt Request Flag 1 0B No action Bit DMPMIT.PVCMI1 is cleared 1B PVCMI2 3 w Clear PVC_M Interrupt Request Flag 2 0B No action 1B Bit DMPMIT.PVCMI2 is cleared PVC1I1 4 w Clear PVC_1 Interrupt Request Flag 1 0B No action 1B Bit DMPMIT.PVC1I1 is cleared PVC1I2 5 w Clear PVC_1 Interrupt Request Flag 2 0B No action 1B Bit DMPMIT.PVC1I2 is cleared WUTI 6 w Clear Wake-up Trim Interrupt Request Flag 0B No action 1B Bit DMPMIT.WUTI is cleared WUI 7 w Clear Wake-up Interrupt Request Flag 0B No action 1B Bit DMPMIT.WUI is cleared 8-167 1 0 PVC PVC PVC PVC SWD SWD 1 1 M M I2 I1 I2 I1 I2 I1 w w w w w w Field User’s Manual SCU, V1.15 2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description GSC 8 w Clear GSC Interrupt Request Flag 0B No action 1B Bit DMPMIT.GSCI is cleared STM0 9 w Clear STM0 Interrupt Request Flag No action 0B Bit DMPMIT.STM0I is cleared 1B STM1 10 w Clear STM1 Interrupt Request Flag 0B No action 1B Bit DMPMIT.STM1I is cleared ESR0T 11 w Clear ESR0 Trap Request Flag 0B No action 1B Bit DMPMIT.ESR0T is cleared ESR1T 12 w Clear ESR1 Trap Request Flag 0B No action 1B Bit DMPMIT.ESR1T is cleared ESR2T 13 w Clear ESR2 Trap Request Flag 0B No action 1B Bit DMPMIT.ESR2T is cleared RAT 15 w Clear Register Access Trap Request Flag 0B No action Bit DMPMIT.RAT is cleared 1B 0 14 r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-168 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.11 Temperature Compensation Unit The temperature compensation for the port drivers provides driver output characteristics which are stable (within a certain band of parameter variation) over the specified temperature range. The temperature compensation oscillator (sensor) provides a reference clock from a free-running temperature-dependent oscillator. An enable trigger is used to define counting cycles where the reference clock pulses are accumulated to build the sensor value TCLR.THCOUNT. The enable trigger is derived from the system clock by a prescaler and a programmable divider (see Figure 8-30). The value for the programmable divider must be written by the user according to the selected system frequency. After the count cycle, the resulting count value, i.e. the number of reference clock cycles, is copied to bit field TCLR.THCOUNT. Thus, TCLR.THCOUNT is updated after every count cycle while the temperature compensation is enabled. Software can compare the temperature-related count value (TCLR.THCOUNT) to several thresholds (temperature levels) in order to determine the control values TCCR.TCC. fREF OSC_TC fSYS Figure 8-30 32:1 N:1 Prescaler Programmable Divider N = (TCDIV+1) fEnable Temperature Compensation Clock Generation The clock divider is programmed via bit field TCCR.TCDIV. The value that should be used for bit field TCCR.TCDIV can be calculated using the following formula documented in the data sheet. Generally, temperature compensation is a user-controlled feature. The Temperature Compensation Control Register TCCR provides access to the actual compensation value (generated by the sensor) and allows software control of the pads. During operation the device (i.e. the pads) can be controlled by the value of the on-chip sensor, or by externally provided compensation values. Register TCCR also provides the programmable divider value. Note: The relation between the counter value and the temperature can differ between two devices and need to be evaluated for each device individually. User’s Manual SCU, V1.15 8-169 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Note: The temperature compensation circuit does not generate temperature compensation values continously. The idea is, that the SW frequently updates the pad control with the value currently found in the tempcomp register (e.g. by an interrupt generated by a timer). Since temparature is a continous function it is not relevant, whether the temperature value read is new or the value of a previous meaurement. User’s Manual SCU, V1.15 8-170 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.11.1 Temperature Compensation Registers 8.11.1.1 TCCR This register contains the control options. TCCR Temperature Compensation Control Register ESFR (F1ACH/D6H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0003H 5 4 3 2 1 0 0 TCE TCDIV TCC r rw rw rw Field Bits Type Description TCC [1:0] rw Temperature Compensation Control The value which controls the temperature compensation inputs of the pads. 00B Maximum reduction = min. driver strength, i.e. very low temperature 11B No reduction = max. driver strength, i.e. very high temperature TCDIV [6:2] rw Temperature Compensation Clock Divider This value adjusts the temperature compensation logic to the selected operating frequency. TCE 7 rw Temperature Compensation Enable 0B No action 1B Enable counting to generate new temperature values. Clearing this bit also stops the temperature compensation oscillator. 0 [15:8] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-171 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) TCLR Temperature Comp. Level Register ESFR (F0ACH/56H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 0 THCOUNT r rh 2 1 0 Field Bits Type Description THCOUNT [7:0] rh Threshold Counter Returns the result of the most recent count cycle of the temperature sensor, to be compared with the thresholds. 0 [15:8] r Reserved Read as 0; should be written with 0. Note: The threshold counter will not overflow but rather stop at count 255. User’s Manual SCU, V1.15 8-172 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.12 Watchdog Timer (WDT) The following part describes the Watchdog Timer (WDT) and its functionality. 8.12.1 Introduction The Watchdog Timer (WDT) is a secure mechanism to overcome life- and dead-locks. An enabled WDT generates a reset for the system if not serviced in a configured time frame. Features The following list is a summary of the WDT functions: • • • • • 16-bit Watchdog Timer Selectable operating frequency: fIN / 256 or fIN / 16384 Timer overflow error detection Individual disable for timer functionality Double Reset Detection Figure 8-31 provides an overview on the registers of the Watchdog Timer. W DT Control & Status Register W DT Reload Register W DT Timer Register WDTCS WDTREL WDTTIM WDTCS WDT Control and Status Register WDTREL WDT Reload Register WDTTIM WDT Timer Register WDT_Reg_Overview.vsd Figure 8-31 Watchdog Timer Register Overview 8.12.2 Overview The Watchdog Timer (WDT) provides a highly reliable and secure way to detect and recover from software or hardware failure. The WDT helps to abort an accidental malfunction of the XC27x5X in a user-specified time period. When enabled, the WDT will cause the XC27x5X system to be reset if the WDT is not serviced within a userprogrammable time period. The CPU must service the WDT within this time interval to prevent the WDT from causing a WDT reset request trigger. Hence, regular service of the WDT confirms that the system is functioning properly. A further feature of the Watchdog Timer is its reset prewarning operation. Instead of immediately resetting the device on the detection of an error, a prewarning output is given to the system via an interrupt request. This makes it possible to bring the system into a defined and predictable status, before the reset is finally issued. User’s Manual SCU, V1.15 8-173 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.12.3 Functional Description The following part describes all functions of the WDT. 8.12.3.1 Timer Operation The timer is enabled when instruction ENWDT (Enable Watchdog Timer) is executed correctly. The WDT uses the input clock fIN which is equal to the system clock fsys. A clock divider in front of the WDT timer provides two output frequencies, fIN / 256 and fIN / 16384. The selection of the counting rate is done via bit WDTCS.IR. WDT Periods The general formula to calculate a Watchdog period is: period = ⎛ 2 16 – startvalue⎞ ⋅ 256 ⋅ 2 ( 1 – IR ) ⋅ 6 ⎝ ⎠ ----------------------------------------------------------------------------------------------------------------------------------IN (8.4) f The parameter <startvalue> represents either the user-programmable reload value WDTREL.RELV (default value FFFCH) for the calculation of the period in Normal Mode or the fixed value FFFFH for the calculation of the period in Prewarning Mode. WDT Timer Reload The counter is reloaded and the prescaler is cleared when one of the following conditions occurs: • • • • • A successful access to register WDTREL The WDT is serviced via instruction SRVWDT A WDT overflow condition (Prewarning Mode is entered). The different reload value for the counter in Prewarning Mode is FFFFH. The Disable Mode is entered (when instruction DISWDT is executed) Upon any reset 8.12.3.2 Timer Modes The Watchdog Timer provides following modes: • • • Normal Mode Disable Mode Prewarning Mode Figure 8-32 provides a state diagram of the different Timer Modes and the transition possibilities. Please refer to the description of the conditions for changing from one mode to the other. User’s Manual SCU, V1.15 8-174 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Internal Application Reset Application Reset Normal Mode Timer overflow / DISWDT WDT interrupt trigger Reload WDT with FFFFH ENWDT Pre-Warning Mode Disable Mode Overflow and Internal Flag = 0 and (Internal Application Reset or Application Reset configured) / Overflow and Internal Flag = 1 and (Internal Application Reset or Application Reset configured) / WDT Reset Trigger Internal Flag = 1 Permanent WDT Reset Trigger Permanent Reset WDT Stopped WDT_modes _MR+.vsd Figure 8-32 State Diagram of the Timer Modes Normal Mode Normal Mode is the default mode after an Application Reset or an Internal Application Reset. Normal Mode can be entered from Disable Mode only when instruction ENWDT is executed. The timer is loaded with RELV when the Normal Mode is entered, and it starts counting upwards. After reset the timer is loaded with FFFCH (default value of RELV). It has to be serviced before the counter overflows. Servicing is performed by the CPU via instructions SRVWDT and/or ENWDT. If the WDT is not serviced before the timer overflows, a system malfunction is assumed, and following operations are done: • • An WDT interrupt trigger is issued Prewarning Mode is entered User’s Manual SCU, V1.15 8-175 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) • Timer is reloaded with FFFFH Disable Mode Disable Mode is provided for applications that do not require the Watchdog Timer function. Disable Mode is entered when instruction DISWDT is executed, either before End-of-Init, if CPUCON1.WDTCTL = 0, or at any time, if CPUCON1.WDTCTL = 1. The timer is reloaded with the value of WDTREL.RELV when Disable Mode is entered. A transition from Disable Mode to Normal Mode is performed when instruction ENWDT is executed while CPUCON1.WDTCTL = 1. Prewarning Mode Prewarning Mode is entered always when a Watchdog error is detected. This is an overflow in Normal Mode. Instead of immediately requesting a reset of the device, the WDT enables the system to enter a secure state by issuing the prewarning output before the reset occurs. Receiving the prewarning, the CPU and the system are requested to finish all pending transaction requests and to not generate new ones. The prewarning is signalled via an interrupt. The CPU can recognize the WDT prewarning interrupt via register INTSTAT. After finishing all pending transactions, the CPU should execute the IDLE instruction to stop all further processing before the coming reset. In Prewarning Mode, the WDT starts counting from FFFFH upwards, and then requests a WDT reset on the overflow of the WDT from FFFFH to 0000H. A reset request of the type as configured in RSTCON1.WDT can not be avoided. No reset will be requested If RSTCON1.WDT is cleared. The WDT does not react anymore to accesses to its registers and to the ENWDT or DISWDT instruction, nor will it change its state until it is reset. A further feature of the WDT detects double errors and sets the whole system into a permanent WDT reset. This feature prevents the XC27x5X from executing random wrong code for longer than the occurence of the overflow, and prevents the XC27x5X from being repeatedly reset by the WDT. Double WDT Reset If the Watchdog induced reset (Application or Internal Application Reset) occurs twice, a severe system malfunction is assumed and the XC27x5X is held in a reset of the type as configured in RSTCON1.WDT (or just not) until a Power-on Reset occurs. This prevents the device from being periodically reset if, for instance, connection to the external memory has been lost such that even system initialization could not be performed. Note: Triggering a PORST upon a WDT reset will never result in a double WDT overflow. If the WDT is configured by RSTCON1.WDT to request an Application Reset or an Internal Application Reset the second reset request will be permanently asserted User’s Manual SCU, V1.15 8-176 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) resulting (without any change in the reset configuration) in a permanent reset of the type configured by RSTCON1.WDT. The information about the first WDT reset request is stored in an internal flag. The internal flag is set when a WDT overflow has occured in Prewarning Mode and the reset request is generated. If the internal flag is already set then a double WDT reset event has occurred and a permanent reset request is generated. This internal flag is cleared by any Power-on Reset or when bit WDTCS.CLRIRF is set. A correct service of the WDT does not clear this internal flag nor do any access to the WDT related registers or commans. Therefore, if correct WDT-servicing has been done after the first WDT reset and a next WDT reset must not immediately lead to a double error state, application software has to clear the internal flag. Note: Regarding the handling of the internal flag It does not matter whether a reset was generated on a WDT reset request or if the reset configuration was changed between the two reset requests. Note: After the double WDT reset request trigger is generated the counter is stopped after the overflow. Port Configuration during WDT Reset The behavior of the ESRx ports can be defined with respect to the reset type by bit field ESRCFGx.PC. For the coding of PC see Table 8-6. The allows to signal the occurence of a reset. The configuration of the GPIOS ports depends on the reset type. In case of an Application Reset the pad configuration is unchanged, in case of an Internal Application Reset the ports are configured for input. 8.12.3.3 Suspend Mode Support In an enabled and active debug session, the Watchdog functionality can lead to unintended resets. Therefore, to avoid these resets, the OCDS can control whether the WDT is enabled or disabled (default after reset). This is done via bit CBS_IOSR.DB. Table 8-11 OCDS Behavior of WDT WDTCS.DS CBS_DBGSR.DBGEN CBS_IOSR.DB WDT Action 1 X X Stopped 0 0 X Running 0 1 0 Stopped 0 1 1 Running User’s Manual SCU, V1.15 8-177 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.12.4 WDT Kernel Registers 8.12.4.1 WDT Reload Register This register defines the WDT reload value. WDTREL WDT Reload Register 15 14 13 12 ESFR (F0C8H/64H) 11 10 9 8 7 6 Reset Value: FFFCH 5 4 3 2 1 0 RELV rw Field Bits Type Description RELV [15:0] rw Reload Value for the Watchdog Timer This bit field defines the reload value for the WDT. User’s Manual SCU, V1.15 8-178 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.12.4.2 WDT Control and Status Register The Control and Status Register can only be accessed in Secured Mode. WDTCS WDT Control and Status Register ESFR (F0C6H/63H) 15 14 13 12 11 10 9 8 7 Reset Value: 0000H 6 5 0 IR 0 r rw r 4 3 2 CLR PR IRF w rh 1 0 DS OE rh rh Field Bits Type Description OE 0 rh Overflow Error Status Flag 0B No WDT overflow error 1B A WDT overflow error has occurred. This bit is set by hardware when the Watchdog Timer overflows from FFFFH to 0000H. This bit is only cleared through: • any Power-on Reset • an executed SRVWDT or ENWDT instruction Note: It is not possible to clear this bit in Prewarning Mode with the SRVWDT or ENWDT instruction. DS 1 rh Timer Enable/Disable Status Flag Timer is enabled (default after reset) 0B 1B Timer is disabled This bit is cleared when instruction ENWDT was executed and CPUCON1.WDTCTL = 1. This bit is set when instruction DISWDT was executed before EINIT or CPUCON1.WDTCTL = 1. Note: ENWDT and DISWDT instruction will be reflected in this bit but in Prewarning Mode the WDT mode is not changed. PR User’s Manual SCU, V1.15 2 rh Prewarning Mode Flag 0B Normal Mode (default after reset) Prewarning Mode 1B 8-179 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description CLRIRF 3 w Clear Internal Reset Flag This bit is used to request a clear of the internal flag storing the information about the first WDT reset request. 0B No action 1B Request to clear the internal flag This bit always read as 0. IR 8 rw Input Frequency Request Bit 0B Request to set input frequency to fIN / 16384 Request to set input frequency to fIN / 256 1B An update of this bit is taken into account after the next successful execution of instruction SRVWDT or ENWDT, on a write to register WDTREL, and always when the WDT is in Disable Mode. 0 [7:4], [15:9] r Reserved Read as 0; should be written with 0; User’s Manual SCU, V1.15 8-180 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.12.4.3 WDT Timer Register WDTTIM WDT Timer Register 15 14 13 12 ESFR (F0CAH/65H) 11 10 9 8 7 6 Reset Value: FFFCH 5 4 3 2 1 0 TIM rh Field Bits Type Description TIM [15:0] rh Timer Value Reflects the current contents of the Watchdog Timer. User’s Manual SCU, V1.15 8-181 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13 SCU Trap Generation The basic trap structure of the SCU is shown in Figure 8-33. TRAPCLR.x Trap Event DMPMIT.x clear TRAPSTAT.x TRAPSET.x 1 1 request TRAPNPn.y & set 1 1 disable SCU TRQ0 SCU TRQ1 SCU TRQ2 to TFR.ACER to TFR.SR1 to TFR.SR0 Reserved TRAPDIS.x other trap sources to the same trap SCU Trap Structure SCU_Trap_Struct .vsd Figure 8-33 SCU Trap Structure If enabled by the corresponding bit in register TRAPDIS, a trap is triggered either by a pulse on the incoming trap line, or by a software set of the respective bit in register TRAPSET. The trigger sets the respective flag in register TRAPSTAT and is gated to one of the trap nodes, selected by the node pointer registers TRAPNP and TRAPNP1. The trap flag in register TRAPSTAT can be cleared by software by writing to the corresponding bit in register TRAPCLR. If more than one trap source is connected to the same trap (via registers TRAPNP and TRAPNP1), the requests are combined to one common line. Trap Node Assignment The trap sources of the system can be mapped to three trap nodes by programming the trap node pointer registers TRAPNP and TRAPNP1. The default assignment of the trap sources to the nodes and their corresponding control register is listed in Table 8-12. 8.13.1 Trap Support Some of the trap requests are first fed through a sticky flag register in the DMP_M domain. These flags are set with a trigger and if set trigger the trap generation in the User’s Manual SCU, V1.15 8-182 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) DMP_1. . Please note that the disable control of register TRAPDIS also influences the sticky bit in register DMPMIT (see Section 8.10.3.6). Which of the trap requests have a sticky flag in register DMPMIT is listed in Table 8-12. Note: When servicing an SCU trap request, make sure that all related request flags are cleared after the identified request has been handled. To clear a trap request that is stored in register DMPMIT, first clear the request source of the source, clear the request within DMP_M via DMPMITCLR, and then clear the request within DMP_1 via TRAPCLR. 8.13.2 SCU Trap Sources The SCU receives the trap lines listed in Table 8-12. Table 8-12 SCU Trap Request Overview Source of Trap Short Name Sticky Flag in DMPMIT Default Trap Flag Assignment in Register TFR Flash Access Trap FAT --- TFR.ACER (SCU_TRQ0) ESR0 Trap ESR0T yes TFR.SR1 (SCU_TRQ1) ESR1 Trap ESR1T yes TFR.SR1 (SCU_TRQ1) ESR2 Trap ESR2T yes TFR.SR1 (SCU_TRQ1) PLL Trap OSCWDTT --- TFR.SR1 (SCU_TRQ1) Register Access Trap RAT yes TFR.ACER (SCU_TRQ0) Parity Error Trap PET --- TFR.ACER (SCU_TRQ0) VCO Lock Trap VCOLCKT --- TFR.SR0 (SCU_TRQ2) ECC Error Trap ECCT --- TFR.ACER (SCU_TRQ0) User’s Manual SCU, V1.15 8-183 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13.3 SCU Trap Control Registers 8.13.3.1 Register TRAPSTAT This register contains the status flags for all trap request trigger sources of the SCU. For setting and clearing of these status bits by software see registers TRAPSET and TRAPCLR, respectively. TRAPSTAT Trap Status Register 15 14 13 12 SFR (FF02H/81H) 11 10 9 8 7 6 VCO ECC PE LCK T T T rh rh rh 0 r Reset Value: 0000H 5 4 3 2 1 OSC ESR ESR ESR RA WDT 2 1 0 T T T T T rh rh rh rh rh 0 FA T rh Field Bits Type Description FAT 0 rh Flash Access Trap Request Flag TRAPSTAT.FAT is set when a flash access violation occurs and TRAPDIS.FAT = 0. No FA trap trigger has occured since this bit was 0B cleared the last time 1B A FA trap trigger has occured since this bit was cleared the last time ESR0T 1 rh ESR0 Trap Request Flag TRAPSTAT.ESR0T is set when bit DMPMIT.ESR0T is set and TRAPDIS.ESR0T = 0. 0B No ESR0 trap trigger has occured since this bit was cleared the last time 1B An ESR0 trap trigger has occured since this bit was cleared the last time ESR1T 2 rh ESR1 Trap Request Flag TRAPSTAT.ESR1T is set when bit DMPMIT.ESR1T is set and TRAPDIS.ESR1T = 0. 0B No ESR1 trap trigger has occured since this bit was cleared the last time 1B An ESR1 trap trigger has occured since this bit was cleared the last time User’s Manual SCU, V1.15 8-184 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ESR2T 3 rh ESR2 Trap Request Flag TRAPSTAT.ESR2T is set when bit DMPMIT.ESR0T is set and TRAPDIS.ESR2T = 0. 0B No ESR2 trap trigger has occured since this bit was cleared the last time 1B An ESR2 trap trigger has occured since this bit was cleared the last time OSCWDTT 4 rh OSCWDT Trap Request Flag TRAPSTAT.OSCWDTT is set when an OSCWDT emergency event occurs and TRAPDIS.OSCWDTT = 0. 0B No OSCWDT trap trigger has occured since this bit was cleared the last time 1B An OSCWDT trap trigger has occured since this bit was cleared the last time RAT 5 rh Register Access Trap Request Flag TRAPSTAT.RAT is set when bit DMPMIT.RAT is set and TRAPDIS.RAT = 0. 0B No RA trap trigger has occured since this bit was cleared the last time 1B A RA trap trigger has occured since this bit was cleared the last time PET 6 rh Parity Error Trap Request Flag TRAPSTAT.PET is set when a memory parity error occurs and TRAPDIS.PET = 0. 0B No PE trap trigger has occured since this bit was cleared the last time 1B A PE trap trigger has occured since this bit was cleared the last time VCOLCKT 7 rh VCOLCK Trap Request Flag TRAPSTAT.VCOLCKT is set when a VCOLCK emergency event occurs and TRAPDIS.VCOLCKT = 0. 0B No VCOLCK trap trigger has occured since this bit was cleared the last time 1B A VCOLCK trap trigger has occured since this bit was cleared the last time User’s Manual SCU, V1.15 8-185 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ECCT 8 rh ECC Error Trap Request Flag TRAPSTAT.ECCT is set when a memory ECC error occurs and TRAPDIS.ECCT = 0. 0B No ECC trap trigger has occured since this bit was cleared the last time 1B An ECC trap trigger has occured since this bit was cleared the last time 0 [15:9] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-186 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13.3.2 Register TRAPCLR This register contains the software clear control for the trap status flags in register TRAPSTAT. Clearing a bit in this register has no effect, reading a bit always returns zero. TRAPCLR Trap Clear Register 15 14 13 12 SFR (FE8EH/47H) 11 10 9 8 7 6 VCO ECC PE LCK T T T w w w 0 r Reset Value: 0000H 5 4 3 2 1 OSC ESR ESR ESR RA WDT 2 1 0 T T T T T w w w w w Field Bits Type Description FAT 0 w Clear Flash Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.FAT is cleared ESR0T 1 w Clear ESR0 Trap Request Flag No action 0B 1B Flag TRAPSTAT.ESR0T is cleared ESR1T 2 w Clear ESR1 Trap Request Flag 0B No action 1B Flag TRAPSTAT.ESR1T is cleared ESR2T 3 w Clear ESR2 Trap Request Flag 0B No action 1B Flag TRAPSTAT.ESR2T is cleared OSCWDTT 4 w Clear OSCWDT Trap Request Flag 0B No action 1B Flag TRAPSTAT.OSCWDTT is cleared RAT 5 w Clear Register Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.RAT is cleared PET 6 w Clear Parity Error Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.PET is cleared VCOLCKT 7 w Clear VCOLCK Trap Request Flag 0B No action 1B Flag TRAPSTAT.VCOLCKT is cleared User’s Manual SCU, V1.15 8-187 0 FA T w V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ECCT 8 w Clear ECC Error Trap Request Flag 0B No action 1B Flag TRAPSTAT.ECCT is cleared 0 [15:9] r Reserved Read as 0; should be written with 0 User’s Manual SCU, V1.15 8-188 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13.3.3 Register TRAPSET This register contains the software set control for the trap status flags in register TRAPSTAT. Clearing a bit in this register has no effect, reading a bit always returns zero. TRAPSET Trap Set Register 15 14 13 12 SFR (FE8CH/46H) 11 10 9 8 7 6 VCO ECC PE LCK T T T w w w 0 r Reset Value: 0000H 5 4 3 2 1 OSC ESR ESR ESR RA WDT 2 1 0 T T T T T w w w w w Field Bits Type Description FAT 0 w Set Flash Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.FAT is set ESR0T 1 w Set ESR0 Trap Request Flag No action 0B 1B Flag TRAPSTAT.ESR0T is set ESR1T 2 w Set ESR1 Trap Request Flag 0B No action 1B Flag TRAPSTAT.ESR1T is set ESR2T 3 w Set ESR2 Trap Request Flag 0B No action 1B Flag TRAPSTAT.ESR2T is set OSCWDTT 4 w Set OSCWDT Trap Request Flag 0B No action 1B Flag TRAPSTAT.OSCWDTT is set RAT 5 w Set Register Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.RAT is set PET 6 w Set Parity Error Access Trap Request Flag 0B No action 1B Flag TRAPSTAT.PET is set VCOLCKT 7 w Set VCOLCK Trap Request Flag 0B No action 1B Flag TRAPSTAT.VCOLCKT is set User’s Manual SCU, V1.15 8-189 0 FA T w V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ECCT 8 w Set ECC Error Trap Request Flag 0B No action 1B Flag TRAPSTAT.ECCT is set 0 [15:9] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-190 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13.3.4 Register TRAPDIS This register contains the software disable control for all trap request trigger sources. Note that the bits ESRxT and RAT in this register also disable the setting of the respective flags in register DMPMIT (see Section 8.10.1). TRAPDIS Trap Disable Register 15 14 13 12 SFR (FE90H/48H) 11 10 9 8 7 6 VCO ECC PE LCK T T T rw rw rw 0 r Reset Value: 009EH 5 4 3 2 1 OSC ESR ESR ESR RA WDT 2 1 0 T T T T T rw rw rw rw rw Field Bits Type Description FAT 0 rw Disable Flash Access Trap Request 0B FA trap request enabled 1B FA trap request disabled ESR0T 1 rw Disable ESR0 Trap Request 0B ESR0 trap request enabled 1B ESR0 trap request disabled ESR1T 2 rw Disable ESR1 Trap Request 0B ESR1 trap request enabled ESR1 trap request disabled 1B ESR2T 3 rw Disable ESR2 Trap Request 0B ESR2 trap request enabled 1B ESR2 trap request disabled OSCWDTT 4 rw Disable OSCWDT Trap Request 0B OSCWDT trap request enabled 1B OSCWDT trap request disabled RAT 5 rw Disable Register Access Trap Request 0B RA trap request enabled 1B RA trap request disabled PET 6 rw Disable Parity Error Trap Request 0B PE trap request enabled 1B PE trap request disabled VCOLCKT 7 rw Disable VCOLCK Trap Request 0B VCOLCK trap request enabled 1B VCOLCK trap request disabled User’s Manual SCU, V1.15 8-191 0 FA T rw V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ECCT 8 rw Disable ECC Error Trap Request 0B ECC trap request enabled 1B ECC trap request disabled 0 [15:9] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-192 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.13.3.5 Register TRAPNP and TRAPNP1 These register contain the control for the trap node pointers of all SCU trap request trigger sources. TRAPNP Trap Node Pointer Register 15 14 13 12 11 10 SFR (FE92H/49H) 9 8 7 6 Reset Value: 8254H 5 4 3 2 1 0 VCOLCK PE RA OSCWDT ESR2 ESR1 ESR0 FA rw rw rw rw rw rw rw rw Field Bits Type Description FA [1:0] rw Trap Node Pointer for Flash Access Traps TRAPNP.FA selects the trap request output for an enabled FAT trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination ESR0 [3:2] rw Trap Node Pointer for ESR0 Traps TRAPNP.ESR0 selects the trap request output for an enabled ESR0 trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination ESR1 [5:4] rw Trap Node Pointer for ESR1 Traps TRAPNP.ESR1 selects the trap request output for an enabled ESR1 trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination User’s Manual SCU, V1.15 8-193 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description ESR2 [7:6] rw Trap Node Pointer for ESR2 Traps TRAPNP.ESR2 selects the trap request output for an enabled ESR2 trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination OSCWDT [9:8] rw Trap Node Pointer for OSCWDT Traps TRAPNP.OSCWDT selects the trap request output for an enabled OSCWDT trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination RA [11:10] rw Trap Node Pointer for Register Access Traps TRAPNP.RA selects the trap request output for an enabled RAT trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination PE [13:12] rw Trap Node Pointer for Parity Error Traps TRAPNP.PE selects the trap request output for an enabled PET trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination VCOLCK [15:14] rw Trap Node Pointer for VCOLCK Traps TRAPNP.VCOLCK selects the trap request output for an enabled VCOLCK trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination User’s Manual SCU, V1.15 8-194 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) TRAPNP1 Trap Node Pointer 1 Register 15 14 13 12 11 10 SFR (FE94H/4AH) 9 8 7 6 Reset Value:0000H 5 4 3 2 1 0 0 ECC r rw Field Bits Type Description ECC [1:0] rw Trap Node Pointer for ECC Error Traps TRAPNP.ECC selects the trap request output for an enabled ECCT trap request. 00B Select request output SCU_TRQ0 (TFR.ACER) 01B Select request output SCU_TRQ1 (TFR.SR1) 10B Select request output SCU_TRQ2 (TFR.SR0) 11B Reserved, do not use this combination 0 [15:2] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-195 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14 Memory Content Protection For supervising the content of the on-chip memories (Flash memory is not considered here) two mechanisms are provided: • • Error Correction Control Parity Checking Table 8-13 Available Memory Protection Mechanisms Memory Parity ECC Error Detection ECC Error Correction ECC code Program SRAM (PS) yes SED SEC 4 bits per byte Data SRAM (DS) yes SED SEC 4 bits per byte Dual Port SRAM (DP) yes SED SEC 4 bits per byte Standby RAM (SB) yes SED SEC 4 bits per byte USICx SRAM yes No No - MultiCAN SRAM no SED SEC 7 bits per 32 bits User’s Manual SCU, V1.15 8-196 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14.1 Memory Checking Control Register Only one of both methods can be activated via register MCHKCON. By default the ECC mechanism is selected. After a switch to parity checking all memories needs to be initialized again. MCHKCON Memory Checking Control RegisterESFR (F0DCH/6EH) 15 14 13 12 11 10 9 8 7 6 5 Reset Value: 0000H 4 3 2 1 0 SEL SEL SEL SEL SB PS DS DP 0 rw rw rw rw rw Field Bits Type Description SELDP 0 rw Select Memory Check for Dual Port Memory 0B ECC checking is selected for dual port memory 1B Parity checking is selected for dual port memory SELDS 1 rw Select Memory Check for Data SRAM 0B ECC checking is selected for data SRAM 1B Parity checking is selected for data SRAM SELPS 2 rw Select Memory Check for Program SRAM 0B ECC checking is selected for program SRAM 1B Parity checking is selected for program SRAM SELSB 3 rw Select Memory Check for Standby Memory 0B ECC checking is selected for Standby memory 1B Parity checking is selected for Standby memory 0 [15:4] rw Reserved Should be written with 0. User’s Manual SCU, V1.15 8-197 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14.2 Parity Error Handling The on-chip RAM modules check parity information during read accesses and in case of an error a signal can be generated if enabled. These signals are combined and trigger a trap. If a parity error is detected during the trap handler routine a reset request trigger is generated. Registers PEEN and PECON control the behavior of parity errors. U1RAM Parity error & U0RAM Parity error & SBRAM Parity error & PSRAM Parity error & Parity DSRAM error & Parity error & DPRAM & & & & p_trap >1 pf_trap & & & & & PEEN.PEENDP Parity error PEEN.PEENDS U2RAM PEEN.PEENPS & PEEN.PEENSB Parity error PEEN.PEENU0 U3RAM PEEN.PEENU1 PEEN.PEENU2 PEEN.PEENU3 PMTSR. PESEN p_rst_req & PECON.PEFDP PECON.PEFDS PECON.PEFPS PECON.PEFSB PECON.PEFU0 PECON.PEFU1 PECON.PEFU2 PECON.PEFU3 parity_error_MR+ Figure 8-34 Parity Error Control Logic A parity error, detected while the respective trap flag TFR.ACER is set, generates a reset request. The second error trap cannot be detected and handled by the CPU. User’s Manual SCU, V1.15 8-198 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Note: The parity trap trigger should activate the Access Error trap (ACER). The parity reset request trigger (p_rst_req) is generated when a parity error trap is request AND flag TFR.ACER is set. 8.14.2.1 Parity Software Testing Support To support testing algorithms for the parity error trap routines a memory parity test logic is implemented for the C166SV2 subsystem memories (PSRAM, DSRAM, and DPRAM) and SBRAM. The logic is controlled by registers PMTSR and PMTPR. Via bit field PMTPR.PWR a parity value can be writing to any address of every memory. The parity control software test update has to be enabled with bit PMTSR.MTEx for each memory individually. Otherwise a write to the parity control has no effect. With each read access to a memory the parity from the memory parity control is stored in register PMTPR.PRD. The width and therefore the valid bits in register PMTPR is listed in Table 8-14. Table 8-14 Memory Widths Memory Number of Parity Bits Valid Bits in PWR/PRD Dual Port (DP) Memory 2 PWR[1:0]/PRD[9:8] Data SRAM (DS) Memory 2 PWR[1:0]/PRD[9:8] Program SRAM (PS) Memory 8 PWR[7:0]/PRD[15:8] Standby RAM (SB) Memory 2 PWR[1:0]/PRD[9:8] Test software should be located in external memory and should be written in a way that no pre-fetching is performed. User’s Manual SCU, V1.15 8-199 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14.2.2 Parity Error Registers The following register enables the functional parity check mechanism. PEEN Parity Error Enable Register 15 14 13 12 11 10 ESFR (F0C4H/41H) 9 8 PE EN SB rw 0 rw 7 0 rw Reset Value: 0000H 6 5 4 3 2 1 0 PE EN U3 rw PE EN U2 rw PE EN U1 rw PE EN U0 rw PE EN PS rw PE EN DS rw PE EN DP rw Field Bits Type Description PEENDP 0 rw Parity Error Trap Enable for Dual Port Memory 0B No Parity trap is requested for dual port memory parity errors 1B A Parity trap is requested for dual port memory parity errors PEENDS 1 rw Parity Error Trap Enable for Data SRAM No Parity trap is requested for data SRAM 0B parity errors 1B A Parity trap is requested for data SRAM parity errors PEENPS 2 rw Parity Error Trap Enable for Program SRAM 0B No Parity trap is requested for program SRAM parity errors 1B A Parity trap is requested for program SRAM parity errors PEENU0 3 rw Parity Error Trap Enable for USIC0 Memory 0B No Parity trap is requested for USIC0 memory parity errors 1B A Parity trap is requested for USIC0 memory parity errors PEENU1 4 rw Parity Error Trap Enable for USIC1 Memory 0B No Parity trap is requested for USIC1 memory parity errors 1B A Parity trap is requested for USIC1 memory parity errors User’s Manual SCU, V1.15 8-200 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PEENU2 5 rw Parity Error Trap Enable for USIC2 Memory 0B No Parity trap is requested for USIC2 memory parity errors 1B A Parity trap is requested for USIC2 memory parity errors PEENU3 6 rw Parity Error Trap Enable for USIC3 Memory 0B No Parity trap is requested for USIC3 memory parity errors 1B A Parity trap is requested for USIC3 memory parity errors PEENSB 8 rw Parity Error Trap Enable for Standby Memory No Parity trap is requested for Standby 0B memory parity errors 1B A Parity trap is requested for Standby memory parity errors 0 7, [15:9] rw Reserved Should be written with 0. User’s Manual SCU, V1.15 8-201 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) The following register controls the functional parity check mechanism. PECON Parity Error Control Register 15 14 13 12 11 10 ESFR (F0DAH/6DH) 9 8 7 0 PEF SB 0 rwh rwh rwh 6 Reset Value: 0000H 5 4 3 2 1 0 PEF PEF PEF PEF PEF PEF PEF U3 U2 U1 U0 PS DS DP rwh rwh rwh rwh rwh rwh rwh Field Bits Type Description PEFDP 0 rwh Parity Error Flag for Dual Port Memory 0B No Parity errors have been detected for dual port memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for dual port memory The bit is only set by the enabled parity error from the dual port memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. PEFDS 1 rwh Parity Error Flag for Data SRAM No Parity errors have been detected for data 0B SRAM 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for data SRAM The bit is only set by the enabled parity error from the data SRAM. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. PEFPS 2 rwh Parity Error Flag for Program SRAM 0B No Parity errors have been detected for program SRAM 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for program SRAM The bit is only set by the enabled parity error from the program SRAM. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. User’s Manual SCU, V1.15 8-202 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PEFU0 3 rwh Parity Error Flag for USIC0 Memory 0B No Parity errors have been detected for USIC0 memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for USIC0 memory The bit is only set by the enabled parity error from the USIC0 memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. PEFU1 4 rwh Parity Error Flag for USIC1 Memory 0B No Parity errors have been detected for USIC1 memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for USIC1 memory The bit is only set by the enabled parity error from the USIC1 memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. PEFU2 5 rwh Parity Error Flag for USIC2 Memory 0B No Parity errors have been detected for USIC2 memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for USIC2 memory The bit is only set by the enabled parity error from the USIC2 memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. PEFU3 6 rwh Parity Error Flag for USIC3 Memory 0B No Parity errors have been detected for USIC3 memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for USIC3 memory The bit is only set by the enabled parity error from the USIC3 memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. User’s Manual SCU, V1.15 8-203 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Field Bits Type Description PEFSB 8 rwh Parity Error Flag for Standby Memory 0B No Parity errors have been detected for Standby memory 1B A Parity error is indicated and can trigger a trap request trigger, if enabled for Standby memory The bit is only set by the enabled parity error from the Standby memory. This bit can only be cleared via software. Writing a zero to this bit does not change the content. Writing a one to this bit does clear the bit. 0 7, [15:9] rwh Reserved Should be written with 0. User’s Manual SCU, V1.15 8-204 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) PMTPR Parity Memory Test Pattern RegisterESFR (F0E4H/72H) 15 14 13 12 11 10 9 8 7 6 5 Reset Value: 0000H 4 3 PRD PWR rh rw 2 1 0 Field Bits Type Description PRD [15:8] rh Parity Read Values for Memory Test For each byte of a memory module the parity bits generated during the most recent read access are indicated here. PWR [7:0] rw Parity Write Values for Memory Test For each byte of a memory module the parity bits corresponding to the next write access are stored here. User’s Manual SCU, V1.15 8-205 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Note: Only one bit MTENxx should be set at the same time in register PMTSR. Otherwise the result of the parity software test is not reliable. PMTSR Parity Memory Test Select RegisterESFR (F0E6H/73H) 15 14 13 12 11 PES EN 0 rw r 10 9 8 7 MT EN SB rw 6 5 Reset Value: 0000H 4 3 0 rw 2 1 0 MT EN PS rw MT EN DS rw MT EN DP rw Field Bits Type Description MTENDP 0 rw Memory Test Enable Control for Dual Port Memory Controls the test multiplexer for the dual port memory. Standard operation 0B 1B Test parity bits used (from PMTPR) MTENDS 1 rw Memory Test Enable Control for Data SRAM Controls the test multiplexer for the data SRAM. 0B Standard operation 1B Test parity bits used (from PMTPR) MTENPS 2 rw Memory Test Enable Control for Program SRAM Controls the test multiplexer for the program SRAM. 0B Standard operation 1B Test parity bits used (from PMTPR) MTENSB 8 rw Memory Test Enable Control for Standby Memory Controls the test multiplexer for the Standby memory. 0B Standard operation 1B Test parity bits used (from PMTPR) PESEN 15 rw Parity Error Sensitivity Enable 0B Parity errors have no effect 1B Parity errors are indicated and can trigger a trap, if enabled 0 [7:3] rw Reserved Should be written with 0. 0 [14:9] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-206 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14.3 ECC Error Handling The on-chip RAM check ECC information during read accesses and in case of an error a signal is generated. These signals are combined and trigger a trap. Register ECCCON controls the behavior of ECC mechanism. Note: Before the ECC logic of an on-chip RAM can be used the first time after a poweron reset operation (before setting its ECCCON enable bit), the corresponding memories must be completely initialized by writing every memory location of it once. Otherwise, unpredictable ECC errors may occur after setting its ECCCON enable bit. Note: When using the autoincrement mode to read the SBRAM and having ECC enabled, it is necessary not only to initialize the used addresses but also the one following the last used address. This problem does not exist if the whole SBRAM is initialized. Note: The handling of Flash ECC errors can not be done in the SCU but in the IMB only. MCRAM ECC error SBRAM ECC error PSRAM ECC error DSRAM ECC error DPRAM ECC error >1 ecc_trap ECC_error_MR+ Figure 8-35 ECC Error Control Logic User’s Manual SCU, V1.15 8-207 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.14.3.1 ECC Registers ECCCON ECC Control Register 15 14 13 12 ESFR (F0A8H/54H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 SB EN MC EN 0 PS EN DS EN DP EN rw rw rw rw rw rw rw Field Bits Type Description DPEN 0 rw Enable for Dual Port Memory 0B ECC check is disabled for dual port memory 1B ECC check is enabled for dual port memory DSEN 1 rw Enable for Data SRAM 0B ECC check is disabled for data SRAM 1B ECC check is enabled for data SRAM PSEN 2 rw Enable for Program SRAM 0B ECC check is disabled for program SRAM 1B ECC check is enabled for program SRAM MCEN 4 rw Enable for MultiCAN Memory 0B ECC check is disabled for MultiCAN memory 1B ECC check is enabled for MultiCAN memory SBEN 5 rw Enable for Standby Memory 0B ECC check is disabled for Standby memory 1B ECC check is enabled for Standby memory 0 3, [15:6] rw Reserved Should be written with 0. User’s Manual SCU, V1.15 8-208 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) ECCSTAT ECC Status Register 15 14 13 12 ESFR (F0AAH/55H) 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 SB MC 0 PS DS DP rh rh rh r rh rh rh Field Bits Type Description DP 0 rh Dual Port Memory ECC Error Status 0B No ECC error was detected for dual port memory 1B An ECC error was detected for dual port memory DS 1 rh Data SRAM ECC Error Status 0B No ECC error was detected for data SRAM 1B An ECC error was detected for data SRAM PS 2 rh Program SRAM ECC Error Status 0B No ECC error was detected for program SRAM An ECC error was detected for program SRAM 1B MC 4 rh MultiCAN Memory ECC Error Status 0B No ECC error was detected for MultiCAN memory 1B An ECC error was detected for MultiCAN memory SB 5 rh Standby Memory ECC Error Status 0B No ECC error was detected for Standby memory An ECC error was detected for Standby 1B memory 0 3, [15:6] rh Reserved Read as 0. User’s Manual SCU, V1.15 8-209 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) ECCCLRSTAT ECC Clear Status Register 15 14 13 12 11 ESFR (F0DEH/6FH) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 0 0 SB MC 0 PS DS DP w w w w w w w Field Bits Type Description DP 0 w Clear Dual Port Memory ECC Error Status 0B No action 1B Setting this bit clears bit ECCSTAT.DP This bit always read as 0. DS 1 w Clear Data SRAM ECC Error Status 0B No action 1B Setting this bit clears bit ECCSTAT.DS This bit always read as 0. PS 2 w Clear Program SRAM ECC Error Status 0B No action Setting this bit clears bit ECCSTAT.PS 1B This bit always read as 0. MC 4 w Clear MultiCAN Memory ECC Error Status 0B No action 1B Setting this bit clears bit ECCSTAT.MC This bit always read as 0. SB 5 w Clear Standby Memory ECC Error Status 0B No action 1B Setting this bit clears bit ECCSTAT.SB This bit always read as 0. 0 3, [15:6] w Reserved Read as 0. User’s Manual SCU, V1.15 8-210 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.15 Register Control This block handles the register accesses of the SCU and the register access control for all system register that use one of the following protection modes: • • • Unprotected Mode Write Protection Mode Secured Mode 8.15.1 Register Access Control There are some dedicated registers that control critical system functions and modes. These registers are protected by a special register security mechanism so these vital system functions cannot be changed inadvertently after the executing of the EINIT instruction. However, as these registers control central system behavior they need to be accessed during operation. The system control software gets this access via a special security state machine. If an access violation is detected a trap trigger request is generated. This security mechanism controls the following security levels wich can be configured via register SLC: • • • Unprotected Mode No protection is active. Registers can be written at any time. This mode is entered after the Application Reset. Write Protected Mode Protected registers are locked against any write access. Write accesses have no effect on these registers. This mode is entered automatically after the EINIT instruction is executed. Secured Mode Protected registers can be written using a special command. Registers that are protected by this mode are marked in Table 8-17 as Sec protected. Access in Secured Mode can be achieved by preceding the intended write access with writing “command 4” to register SLC. After writing “command 4” to register SLC the register protection mechanism remains disabled until the next write to a register on the PD+Bus (SFR, ESFR, XSFR area), i.e. accesses to registers (e.g. CSFR) outside this area do not enable the protection again automatically. Therefore, the lock mechanism after writing “command 4” works differently depending on the register address. Normally one single write access to a protected register is enabled. After this write access the protected registers are locked again automatically. Thereafter, “command 4” has to be written again in order to enable the next write to a protected register. The lock mechanism is not enabled again after a write access to a CSFR register or to a LXBus peripheral register (XLOC area, e.g. USIC, CAN, IMB). Note: In Secured Mode the re-enabling of register protection with respect to the write address after “command 4” can lead to an unexpected, not obvious behaviour of User’s Manual SCU, V1.15 8-211 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) an application: In case the succeeding write to a protected register is delayed due to an interrupt and the ISR itself uses the “command 4” mechanism. After writing “command 4” inside the ISR the protection is expectedly re-installed instead of released and the following write will lead to an ACER trap within the ISR. An ATOMIC instruction, which couples the unlock with the write to the protected register could be used. In case the succeeding write is to a register which does not re-enable the protection mechanism again then the write itself will succeed, but in a following “command 4” sequence the write to SLC register re-locks the protection again and the write to a protected register fails. All registers that are equipped with this protection mechanism have additional to normal access parameters (e.g. read only, bit type r or rh) the access limitations defined by the selected security level. Independently of the security level all protected registers can also be read. 8.15.1.1 Controlling the Security Level The two registers Security Level Command register (SLC) and Security Level Status register (SLS) control the security level. The SLC register accepts the commands to control the state machine modifying the security level, while the SLS register shows the actual password, the actual security level, and the state of the 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 8-36 State Machine for Security Level Controlling The following mechanism is used to control the actual security level: • Changing the security level can be done by executing the following command sequence: User’s Manual SCU, V1.15 8-212 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) “command 0 - command 1 - command 2 - command 3”. This sequence establishes a new security level and/or a new password. Table 8-15 Commands for Security Level Control Command Definition Command 0 AAAAH Command 1 5554H Command 2 96H +1) <inverse password> Command 3 000B + <new level> + 000B + <new password> Command 4 8EH + <inverse password> 1) Note Secured Mode only ’+’ denotes a bit field concatenation Note: It is recommended to lock all command sequences with an atomic sequence. User’s Manual SCU, V1.15 8-213 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.15.2 Register Protection Registers Register SLC This register is the interface for the protection commands. SLC Security Level Command RegisterESFR (F0C0H/60H) 15 14 13 12 11 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) User’s Manual SCU, V1.15 8-214 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Register SLS This register monitors the status of the register protection. SLS Security Level Status Register ESFR (F0C2H/61H) 15 14 13 12 11 10 9 8 7 6 Reset Value: 0000H 5 4 3 STATE SL 0 PASSWORD rh rh r rh 2 1 0 Field Bits Type Description PASSWORD [7:0] rh Current Security Control Password Default after reset = 00H SL [12:11] rh Security Level 1) 00B Unprotected Mode (default) 01B Secured Mode 10B Reserved, Do not use this combination 11B Write Protected Mode STATE [15:13] rh Current State of Switching State Machine 000B Awaiting command 0 or command 4 (default) 001B Awaiting command 1 010B Awaiting command 2 011B Awaiting new security level and password 100B Next access granted in Secured Mode 101B Reserved, do not use this combination 11XB Reserved, do not use this combination 0 [10:8] Reserved Read as 0; should be written with 0; 1) r While the security level is “unprotected” after reset, it changes to “write protected” after the execution of instruction EINIT. User’s Manual SCU, V1.15 8-215 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.16 Miscellaneous System Registers This chapter acts as container for various register that are not connected to one specific application topic. 8.16.1 System Registers 8.16.1.1 System Control Register The following register serve several different system tasks. SYSCON1 System Control 1 Register 15 14 13 12 11 SFR (FF4CH/A6H) 10 9 8 7 6 Reset Value: 0000H 5 4 3 2 1 GLC OCD CST SEN 0 r rw rw 0 0 r Field Bits Type Description OCDSEN 2 rw OCDS/Cerberus Enable 0B OCDS and Cerberus are still in reset state 1B ODCS and Cerberus are operable GLCCST 3 rw Global CAPCOM Start This bit starts all CAPCOM units synchronously if enabled. 0B CAPCOM timer start is controlled locally in each unit 1B All CAPCOM timers are started synchronously This bit needs to be cleared via software before setting starts a new CAPCOM start. 0 [1:0], [15:4] r Reserved Read as 0; should be written with 0. User’s Manual SCU, V1.15 8-216 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.16.2 Identification 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. 8.16.2.1 Identification Registers Register IDMANUF This register contains information about the manufacturer. IDMANUF Manufacturer Identification Register ESFR (F07EH/3FH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 1820H 5 4 3 2 1 MANUF DEPT r r 0 Field Bits Type Description DEPT [4:0] r Department Indicates the department within Infineon. 00HAIM MC MANUF [15:5] r Manufacturer This is the JEDEC normalized manufacturer code. 0C1HInfineon Technologies AG User’s Manual SCU, V1.15 8-217 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Register IDCHIP This register contains information about the device. IDCHIP Chip Identification Register 15 14 13 12 11 10 ESFR (F07CH/3EH) 9 8 7 6 Reset Value: XXXXH 5 4 3 CHIPID Revision r r 2 1 0 Field Bits Type Description Revision [7:0] r Device Revision Code Identifies the device step. Please refer to the data sheet for the device specific value. CHIPID [15:8] r Device Identification Identifies the device name. Please refer to the data sheet for the device specific value. User’s Manual SCU, V1.15 8-218 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Register IDMEM This register contains information about the program memory. IDMEM Program Memory Identification Register ESFR (F07AH/3DH) 15 14 13 12 11 10 9 8 7 Reset Value: 3XXXH 6 5 TYPE SIZE r rw 4 3 2 1 0 Field Bits Type Description SIZE [11:0] rw Size of on-chip Program Memory The size of the implemented program memory in terms of 4 K blocks, i.e. memory size = <SIZE>*4 Kbyte. Please refer to the data sheet for the device specific value. TYPE [15:12] r User’s Manual SCU, V1.15 Type of on-chip Program Memory Identifies the memory type on this silicon. Please refer to the data sheet for the device specific value. 8-219 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Register IDPROG This register contains information about the flash programming voltage. IDPROG Programming Voltage Id. Register ESFR (F078H/3CH) 15 14 13 12 11 10 9 8 7 6 Reset Value: 1313H 5 4 3 PROGVPP PROGVDD r r 2 1 0 Field Bits Type Description PROGVDD [7:0] r Programming VDD Voltage The voltage of the standard power supply required to program or erase (if applicable) the on-chip program memory. Please refer to the data sheet for the device specific value. PROGVPP [15:8] r Programming VPP Voltage The voltage of the special programming power supply (if existent) required to program or erase (if applicable) the on-chip program memory. Please refer to the data sheet for the device specific value. User’s Manual SCU, V1.15 8-220 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.16.3 Marker Memory 8.16.3.1 Marker Memory Registers The marker memory consists of following SFRs located in the DMP_M for free usage of the user software. MKMEM0 Marker Memory 0 Register MKMEM1 Marker Memory 1 Register 15 14 13 12 11 10 SFR (FED0H/68H) Reset Value: 0000H SFR (FED2H/69H) Reset Value: 0000H 9 8 7 6 5 4 3 2 1 0 MARKER rw Field Bits Type Description MARKER [15:0] rw Marker Content User’s Manual SCU, V1.15 8-221 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.17 SCU Register Addresses The SCU registers are within the (E)SFR space of the XC27x5X. Therefore, their specified addresses equal an offset from zero. Table 8-16 Registers Address Space Module Base Address End Address SCU 00 0000H 00 FFFEH Note SCU Register Overview Table 8-17 Register Overview of SCU Protection1) Reset 2) Short Name Register Long Name Offset Addr. WUOSCCON Wake-up OSC Control Register F1AEH Sec Power-on Reset HPOSCCON High Precision Oscillator Configuration Register F1B4H Sec Power-on Reset PLLOSCCON PLL Control Register F1B6H Sec Power-on Reset PLLSTAT PLL Status Register F0BCH - STATCLR1 PLL Status Clear 1 Register F0E2H Sec Power-on Reset PLLCON0 PLL Configuration 0 Register F1B8H Sec Power-on Reset PLLCON1 PLL Configuration 1 Register F1BAH Sec Power-on Reset PLLCON2 PLL Configuration 2 Register F1BCH Sec Power-on Reset PLLCON3 PLL Configuration 3 Register F1BEH Sec Power-on Reset SYSCON0 System Configuration 0 Register FF4AH Sec Power-on Reset STATCLR0 Status Clear 0 Register F0E0H Sec Power-on Reset RTCCLKCON RTC Clock Control Register FF4EH Sec Power-on Reset User’s Manual SCU, V1.15 8-222 Power-on Reset V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-17 Register Overview of SCU Short Name Register Long Name Offset Addr. Protection1) Reset 2) EXTCON External Clock Control Register FF5EH Sec Power-on Reset STMREL STM Reload Register F1A8H Sec Power-on Reset STMCON STM Control Register F1AAH Sec Power-on Reset WUTREL Wake-up Timer Reload Register F0B0H Sec Power-on Reset WUCR Wake-up Control Register F1B0H Sec Power-on Reset RSTSTAT0 Reset Status 0 Register F0B2H - Power-on Reset RSTSTAT1 Reset Status 1 Register F0B4H - Power-on Reset RSTSTAT2 Reset Status 2 Register F0B6H - Power-on Reset RSTCON0 Reset Configuration 0 Register F0B8H Sec Power-on Reset RSTCON1 Reset Configuration 1 Register F0BAH Sec Power-on Reset RSTCNTCON Reset Counter Configuration Register F1B2H Sec Power-on Reset SWRSTCON SW Reset Control Register F0AEH Sec Power-on Reset ESREXCON1 ESR 1 External Control Register FF32H Sec Power-on Reset ESREXCON2 ESR 2 External Control Register FF34H Sec Power-on Reset ESREXSTAT1 ESR 1 External Status Register FF36H - Power-on Reset ESREXSTAT2 ESR 2 External Status Register FF38H - Power-on Reset FF3AH Sec Power-on Reset CLRESREXSTAT1 Clear ESR 1 External Status Register User’s Manual SCU, V1.15 8-223 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-17 Register Overview of SCU Short Name Register Long Name Offset Addr. Protection1) Reset 2) CLRESREXSTAT2 Clear ESR 2 External Status Register FF3CH Sec Power-on Reset ESRCFG0 ESR 0 Configuration Register F100H Sec Power-on Reset ESRCFG1 ESR 1 Configuration Register F102H Sec Power-on Reset ESRCFG2 ESR 2 Configuration Register F104H Sec Power-on Reset ESRDAT ESR Data Register F106H Sec Power-on Reset SWDCON0 SWD Control 0 Register F080H Sec Power-on Reset SWDCON1 SWD Control 1 Register F082H Sec Power-on Reset PVC1CON0 PVC_1 Control for Step 0 Register F014H Sec Power-on Reset PVCMCON0 PVC_M Control for Step 0 Register F1E4H Sec Power-on Reset EVR1CON0 EVR_1 Control 0 Register F088H Sec Power-on Reset EVR1SET15VHP EVR_1 Setting for 1.5V HP F09EH Register Sec Power-on Reset EVRMCON0 EVR_M Control 0 Register F084H Sec Power-on Reset EVRMCON1 EVR_M Control 1 Register F086H Sec Power-on Reset EVRMSET15VHP EVR_M Setting for 1.5V HP F096H Register Sec Power-on Reset GSCSWREQ GSC SW Request Register FF14H Sec Application Reset GSCEN GSC Enable Register FF16H Sec Application Reset GSCSTAT GSC Status Register FF18H - Application Reset User’s Manual SCU, V1.15 8-224 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-17 Register Overview of SCU Short Name Register Long Name Offset Addr. Protection1) Reset 2) GSCPERSTATEN GSC Peripheral Status Enable Register FF04H Sec Application Reset GSCPERSTAT GSC Peripheral Status Register FF1AH - Application Reset STSTAT Start-up Status Register F1E0H - Application Reset EXISEL External Interrupt Input Select Register F1A0H Sec Application Reset EXICON0 External Interrupt Input Trigger Control 0 Register F030H Sec Application Reset EXICON1 External Interrupt Input Trigger Control 1 Register F032H Sec Application Reset EXICON2 External Interrupt Input Trigger Control 2 Register F034H Sec Application Reset EXICON3 External Interrupt Input Trigger Control 3 Register F036H Sec Application Reset EXOCON0 External Output Trigger Control 0 Register FE30H Sec Application Reset EXOCON1 External Output Trigger Control 1 Register FE32H Sec Application Reset EXOCON2 External Output Trigger Control 2 Register FE34H Sec Application Reset EXOCON3 External Output Trigger Control 3 Register FE36H Sec Application Reset INTSTAT Interrupt Status Register FF00H - Application Reset INTCLR Interrupt Clear Register FE82H Sec Application Reset INTSET Interrupt Set Register FE80H Sec Application Reset INTDIS Interrupt Disable Register FE84H Sec Application Reset INTNP0 Interrupt Node Pointer 0 Register FE86H Sec Application Reset User’s Manual SCU, V1.15 8-225 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-17 Register Overview of SCU Short Name Register Long Name Offset Addr. Protection1) Reset 2) INTNP1 Interrupt Node Pointer 1 Register FE88H Sec Application Reset DMPMIT DMP_M Interrupt and Trap FE96H Trigger Register - Power-on Reset DMPMITCLR DMP_M Interrupt and Trap FE98H Clear Register Sec Power-on Reset TCCR Temperature Compensation Control Register F1ACH Sec Application Reset TCLR Temperature Compensation Level Register F0ACH Sec Application Reset WDTREL WDT Reload Register F0C8H Sec Application Reset WDTCS WDT Control and Status Register F0C6H Sec Application Reset WDTTIM WDT Timer Register F0CAH Sec Application Reset TRAPSTAT Trap Status Register FF02H Power-on Reset TRAPCLR Trap Clear Register FE8EH Sec Power-on Reset TRAPSET Trap Set Register FE8CH Sec Power-on Reset TRAPDIS Trap Disable Register FE90H Sec Power-on Reset TRAPNP Trap Node Pointer Register FE92H Sec Power-on Reset TRAPNP1 Trap Node Pointer 1 Register FE94H Sec Power-on Reset MCHKCON Memory Checking Control Register F0DCH Sec Power-on Reset PEEN Parity Error Enable Register F0C4H Power-on Reset User’s Manual SCU, V1.15 8-226 - Sec V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-17 Register Overview of SCU Protection1) Reset 2) Short Name Register Long Name Offset Addr. PECON Parity Error Control Register F0DAH Sec PMTPR Parity Memory Test Pattern F0E4H Register Sec Application Reset PMTSR Parity Memory Test Select Register F0E6H Sec Application Reset ECCCON ECC Control Register F0A8H Sec Application Reset ECCSTAT ECC Status Register F0AAH - ECCCLRSTAT ECC Clear Status Register F0DEH Sec Application Reset SLC Security Level Command Register F0C0H - Application Reset SLS Security Level Status Register F0C2H - Application Reset SYSCON1 System Control 1 Register FF4CH Sec IDMANUF Manufacturer Identification F07EH Register - Power-on Reset IDCHIP Chip Identification Register F07CH - Power-on Reset IDMEM Program Memory Identification Register F07AH - Power-on Reset IDPROG Programming Voltage Identification Register F078H - Power-on Reset MKMEM0 Marker Memory 0 Register FED0H Sec Power-on Reset MKMEM1 Marker Memory 1 Register FED2H Sec Power-on Reset Power-on Reset Application Reset Application Reset 1) Register write protection mechanism: “Sec” = register security mechanism, “-” = always accessible (no protection), otherwise no access is possible. 2) Reset types are defined in Chapter 8.4.1.2. User’s Manual SCU, V1.15 8-227 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.18 Implementation This section shows the connections of the module to the system. 8.18.1 Clock Generation Unit The following table shows the input connection of the Clock Genation Unit (see Chapter 8.1). Table 8-18 CGU Input Connection Input Connected to XTAL 1 XTAL 1 XTAL 2 XTAL 2 CLKIN1 Port 2.9 CLKIN2 Port 4.4 User’s Manual SCU, V1.15 8-228 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.18.2 External Service Requests (ESR) The availability of pins ESR1 and ESR2 is device and package dependent. It is described in the data sheet. User’s Manual SCU, V1.15 8-229 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.18.3 External Request Unit (ERU) The connections of the ERU (see Chapter 8.9) are device specific. In the following the connections of the ERU in XC27x5X are described. 8.18.3.1 ERU Input Connections The following figure shows the ERU input connections, either directly with pins or via communication modules, such as USIC or MultiCAN. These communication modules provide their input signals (e.g. CAN receive input, or USIC data, clock, or control inputs) that have been selected in these modules. ERS0 ERS1 P2.1 ERU_0A0 P2.2 ERU_1A0 ESR 1 ERU_0A1 ESR2 ERU_1A1 U0C0_DX0INS, selected data input ERU_0A2 U1C0_DX0INS, selected data input ERU_1A2 U0C0_DX2INS, selected control input ERU_0A3 U1C0_DX2INS, selected control input ERU_1A3 P1.0 ERU_0B0 P5.13 ERU_0B1 MultiCAN_CAN4INS, selected receive input CAN 4 ERU_1B1 U0C1_DX0INS, selected data input ERU_0B2 U1C1_DX0INS, selected data input ERU_1B2 U0C1_DX2INS, selected control input ERU_0B3 U1C1_DX2INS, selected control input ERU_1B3 P1.1 ERS2 P1.2 ERU_1B0 ERS3 P1.3 ERU_2A0 ERU_3A0 MultiCAN_CAN3INS, selected receive input CAN 3 ERU_2A1 MultiCAN_CAN1INS, selected receive input CAN 1 ERU_3A1 U2C0_DX0INS, selected data input ERU_2A2 GPT12E_T3OUT ERU_3A2 U2C0_DX2INS, selected control input ERU_2A3 reserved ERU_3A3 U2C0_DX1INS, selected clock input ERU_2B0 U1C0_DX1INS, selected clock input ERU_3B0 MultiCAN_CAN2INS, selected receive input CAN 2 ERU_2B1 MultiCAN_CAN0INS, selected receive input CAN 0 ERU_3B1 U2C1_DX0INS, selected data input ERU_2B2 reserved ERU_3B2 U2C1_DX2INS, selected control input ERU_2B3 reserved ERU_3B3 ERU_Inputs _Overview_MR+.vsd Figure 8-37 ERU Inputs Overview User’s Manual SCU, V1.15 8-230 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) The following table describes the ERU input connections for the ERSx stages. The selection is defined by the bit fields in register EXISEL. Table 8-19 ERSx Connections in XC27x5X Input from/to Module I/O to ERSx Can be used to/as ERU_0A0 P2.1 I ERS0 input A ERU_0A1 ESR1 I ERU_0A2 U0C0_DX0INS I ERU_0A3 U0C0_DX2INS I ERU_0B0 P1.0 I ERU_0B1 P5.13 I ERU_0B2 U0C1_DX0INS I ERU_0B3 U0C1_DX2INS I ERU_1A0 P2.2 I ERU_1A1 ESR2 I ERU_1A2 U1C0_DX0INS I ERU_1A3 U1C0_DX2INS I ERU_1B0 P1.1 I ERU_1B1 MultiCAN_CAN4INS I ERU_1B2 U1C1_DX0INS I ERU_1B3 U1C1_DX2INS I ERU_2A0 P1.2 I ERU_2A1 MultiCAN_CAN3INS I ERU_2A2 U2C0_DX0INS I ERU_2A3 U2C0_DX2INS I ERS0 Inputs ERS0 input B ERS1 Inputs ERS1 input A ERS1 input B ERS2 Inputs User’s Manual SCU, V1.15 ERS2 input A 8-231 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-19 ERSx Connections in XC27x5X (cont’d) Input from/to Module I/O to ERSx Can be used to/as ERU_2B0 U2C0_DX1INS I ERS2 input B ERU_2B1 MultiCAN_CAN2INS I ERU_2B2 U2C1_DX0INS I ERU_2B3 U2C1_DX2INS I ERU_3A0 P1.3 I ERU_3A1 MultiCAN_CAN1INS I ERU_3A2 GPT12E_T3OUT I ERU_3A3 0 I ERU_3B0 U1C0_DX1INS I ERU_3B1 MultiCAN_CAN0INS I ERU_3B2 0 I ERU_3B3 0 I ERS3 Inputs User’s Manual SCU, V1.15 ERS3 input A ERS3 input B 8-232 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.18.3.2 Output Gating Unit (OGUy) The following table describes the peripheral trigger connections for the OGUy stages. Table 8-20 Input OGUy Peripheral Trigger Connections in XC27x5X from/to Module I/O to OGUy Can be used to/as ERU_OGU01 CCU60_MCM_ST I Peripheral triggers for OGU0 ERU_OGU02 CCU60_T13_PM I ERU_OGU03 CC2_28 I OGU0 Inputs OGU1 Inputs ERU_OGU11 CCU61_MCM_ST I ERU_OGU12 CCU61_T13_PM I ERU_OGU13 CC2_29 I Peripheral triggers for OGU1 OGU2 Inputs ERU_OGU21 CCU62_MCM_ST I ERU_OGU22 CCU62_T13_PM I ERU_OGU23 CC2_30 I Peripheral triggers for OGU2 OGU3 Inputs ERU_OGU31 CCU63_MCM_ST I ERU_OGU32 CCU63_T13_PM I ERU_OGU33 CC2_31 I User’s Manual SCU, V1.15 Peripheral triggers for OGU3 8-233 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) 8.18.3.3 ERU Output Connections The following table describes the connections of the ERU output signals for gating or triggering other module functions, as well as the connections to the interrupt control registers. Table 8-21 ERU Output Connections in XC27x5X Output from/to Module I/O to OGUy Can be used to/as ERU_PDOUT0 ADC0 (REQGT0E) ADC0 (REQGT1E) ADC0 (REQGT2E) ADC1 (REQGT0E) ADC1 (REQGT1E) ADC1 (REQGT2E) O Pattern detection output ERU_GOUT0 not connected O Gated pattern detection output ERU_GOUT0 not connected O Gated pattern detection output ERU_TOUT0 not connected O Trigger output ERU_IOUT0 ITC (SCU_ERU_0IC) O Interrupt output ERU_PDOUT1 ADC0 (REQGT0F) ADC0 (REQGT1F) ADC0 (REQGT2F) ADC1 (REQGT0F) ADC1 (REQGT1F) ADC1 (REQGT2F) O Pattern detection output ERU_GOUT1 not connected O Gated pattern detection output ERU_TOUT1 ADC0 (REQTR0B) ADC0 (REQTR1B) ADC0 (REQTR2B) ADC1 (REQTR0B) ADC1 (REQTR1B) ADC1 (REQTR2B) O Trigger output ERU_IOUT1 ITC (SCU_ERU_1IC) O Interrupt output OGU0 Outputs OGU1 Outputs OGU2 Outputs User’s Manual SCU, V1.15 8-234 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) Table 8-21 ERU Output Connections in XC27x5X (cont’d) Output from/to Module I/O to OGUy Can be used to/as ERU_PDOUT2 CCU60 (CTRAPD) O Pattern detection output ERU_GOUT2 not connected O Gated pattern detection output ERU_TOUT2 not connected O Trigger output ERU_IOUT2 ITC (SCU_ERU_2IC) O Interrupt output ERU_PDOUT3 CCU63 (CTRAPD) O Pattern detection output ERU_GOUT3 not connected O Gated pattern detection output ERU_TOUT3 not connected O Trigger output ERU_IOUT3 ITC (SCU_ERU_3IC) O Interrupt output OGU3 Outputs User’s Manual SCU, V1.15 8-235 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) System Control Unit (SCU) User’s Manual SCU, V1.15 8-236 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9 Parallel Ports The XC27x5X provides a set of General Purpose Input/Output (GPIO) ports that can be controlled by software and by the on-chip peripheral units: Table 9-1 Ports of the XC27x5X Group Width I/O Connected Modules P0 8 I/O EBC (A7...A0), CCU6, USIC, CAN P1 8 I/O EBC (A15...A8), CCU6, USIC P2 14 I/O EBC (READY, BHE, A23...A16, AD15...AD13, D15...D13), CAN, CC2, GPT12E, USIC, DAP/JTAG P3 8 I/O EBC arbitration (BREQ, HLDA, HOLD), CAN, USIC P4 8 I/O EBC (CS4...CS0), CC2, CAN, GPT12E, USIC P5 16 I Analog Inputs, CCU6, DAP/JTAG, GPT12E, CAN P6 4 I/O ADC, CAN, GPT12E P7 5 I/O CAN, GPT12E, SCU, DAP/JTAG, CCU6, ADC, USIC P8 7 I/O CCU6, DAP/JTAG, USIC P9 8 I/O CCU6, DAP/JTAG, CAN P10 16 I/O EBC(ALE, RD, WR, AD12...AD0, D12...D0), CCU6, USIC, DAP/JTAG, CAN P11 6 I/O CCU6, USIC, CAN P15 8 I Analog Inputs, GPT12E Note: The availability of ports and port pins depends on the selected device type. This chapter describes the maximum set of ports. User’s Manual Parallel Ports, V1.16 9-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.1 General Description This chapter describes the architecture of the digital control circuit for a single port pin. 9.1.1 Basic Port Operation There are three types of digital control circuits: with/without hardware override for digital GPIOs, and for one for analog inputs. Each port pin contains one of them. POCON.PPS A N D PWS Pn_IOCR control 4 Pn_IN 2 1 1 FF ALTIN X O R 0 1 pull devices OD, DIR ALTSEL0, 1 Pn_OUT 2 ENABQ Alternate Data signals or other Access to port registers control lines from Peripherals or SCU by PD Bus Pn_OMR control 4 2 Input stage INV ALT1 1 ALT2 1 ALT3 1 DQ1 ENDQ1 TC[1:0] PD[2:0] output stage pad pin 1 1 2 3 Standard _port_structure_4_um. vsd Figure 9-1 Structure of the Ports without Hardware Override Functionality Note: INV signal is derived from Pn_IOCR.PC[3:2]. User’s Manual Parallel Ports, V1.16 9-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports POCON.PPS A N D PWS HW_DIR Alternate Data signals or other control lines from Peripherals 3 Pn_IN 1 1 FF ALTIN X O R control 4 2 2 2 ALTSEL0,1 Pn_OMR Pn_OUT control pull devices O D, DIR 4 ENABQ Access to port registers by PD Bus Pn_IOCR 0 1 2 input stage HW_EN msb INV* ALT1 1 ALT2 1 ALT3 1 HW_OUT DQ1 ENDQ1 TC[1:0] PD[2:0] output stage pad pin 1 1 1 2 3 Standard _EBCport _structure_5_um.vsd Figure 9-2 Structure of the Ports with Hardware Override Functionality Note: If HW_EN is activated, INV* signal is always zero. Note: When HW_EN is disabled, the respective ports go to Power Save Mode as all other ports. When HW_EN is active, then the user should set the POCON.PPSx=0. User’s Manual Parallel Ports, V1.16 9-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Access to port registers by PD Bus Parallel Ports ENABQ Pn_DIDIS Pn_IN Input stage Analog Input pad pin Analog_port_digital_structure_2.vsd Figure 9-3 Structure of Port 5 and Port 15 Note: There is always a standard digital input connected in parallel to each analog input. User’s Manual Parallel Ports, V1.16 9-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.1.2 Input Stage Control An input stage consists of a Schmitt trigger, which can be enabled or disabled via software, and an input multiplexer that by default selects the output of the input Schmitt trigger. A disabled input driver drives high logical level. During and after reset, all input stages are enabled by default. 9.1.3 Output Driver Control An output stage consists of an output driver, output multiplexer, and register bit fields for their control. 9.1.3.1 Active Mode Behavior Each output driver can be configured in a push-pull or an open-drain mode, or it can be deactivated (three-stated). An output multiplexer in front of the output driver selects the signal source, choosing either the appropriate bit of the Pn_OUT register, or one of maximum three lines coming from a peripheral unit, see Figure 9-1. The selection is done via the Pn_IOCR register. Software can set or clear the bit Pn_OUT.Px, which drives the port pin in case it is selected by the output multiplexer. An output driver with hardware override can select an additional output signal coming from a peripheral. While the hardware override is activated, this signal has higher priority than all other output signals and can not be deselected by the port. In this case, the peripheral controls the direction of the pin. 9.1.3.2 Power Saving Mode Behavior In Power Saving Mode (core and IO supply voltages available), the behavior of a pin depends on the setting of the POCONx.PPSx bit. Basically, groups of four pins within a port can be configured to react to Power Save Mode Request or to ignore it. In case a pin group is configured to react to a Power Save Mode Request, each pin within a group reacts according to its own configuration according to the Table 9-4. 9.1.3.3 Reset Behavior During reset, all output stages of GPIO pins go to tri-state mode without any pull-up or pull-down device. 9.1.3.4 Power-fail Behavior When the core supply fails while the pad supply remains stable, the output stages go into tri-state mode. User’s Manual Parallel Ports, V1.16 9-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2 Port Register Description 9.2.1 Pad Driver Control The pad structure used in this device offers the possibility to select the output driver strength and the slew rate. These selections are independent from the output port functionality, such as open-drain, push/pull or input only. In order to minimize EMI problems, the driver strength can be adapted to the application requirements by bit fields PDMx. The selection is done in groups of four pins. The Port Output Control registers POCON 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. Note: P2_POCON register in the XC27x5X contains an exception regarding the additional strong output driver connected in parallel to the standard output driver of the P2.8 pin. See the respective port section for details. User’s Manual Parallel Ports, V1.16 9-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Px_POCON (x=0-4) Port x Output Control Register XSFR (E8A0H+2*x) Px_POCON (x=6-11) Port x Output Control Register XSFR (E8A0H+2*x) 15 14 13 12 11 10 9 8 7 Reset Value: 0000H Reset Value: 0000H 6 5 4 3 2 1 PPS 3 PDM3 PPS 2 PDM2 PPS 1 PDM1 PPS 0 PDM0 rw rwr rw rw rw rw rw rw 0 Field Bit Type Description PDM0, PDM1, PDM2, PDM3 [2:0], [6:4], [10:8], [14:12] rw Port Driver Mode x Code Driver strength 1) 000B Strong driver 001B Strong driver 010B Strong driver 011B Weak driver 100B Medium driver 101B Medium driver 110B Medium driver 111B Weak driver PPS0, PPS1, PPS2, PPS3 3, 7, 11, 15 rw Pin Power Save 0B Pin behaves like in the Active Mode. Power Save Management is ignored. 1B Behavior in the Power Save Mode described in the Table 9-4. Edge Shape2) Sharp edge mode Medium edge mode Soft edge mode 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. User’s Manual Parallel Ports, V1.16 9-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Mapping of the POCON Registers to Pins and Ports The table below lists the defined POCON registers and the allocation of control bit fields and port pins. Table 9-2 Port Output Control Register Allocation Controlled Pins (by Px_POCON.[y:z])1) Control Register [15:12] [11:8] [7:4] [3:0] P0_POCON --- --- P0.[7:4] P0.[3:0] 8 P1_POCON --- --- P1.[7:4] P1.[3:0] 8 P2_POCON CLOCKOUT driver at P2.82) P2.[11:8] + P2.[13:12]3) P2.[7:4] P2.[3:0] 14 P3_POCON --- --- P3.[7:4] P3.[3:0] 8 P4_POCON --- --- P4.[7:4] P4.[3:0] 8 P6_POCON --- --- --- P6.[3:0] 4 P7_POCON --- --- P7.4 P7.[3:0] 5 P8_POCON --- --- P8.[6:4] P8.[3:0] 7 P9_POCON --- --- P9.[7:4] P9.[3:0] 8 P10_POCON P10.[15:12] P10.[11:8] P10.[7:4] P10.[3:0] 16 P11_POCON --- --- P11.[5:4] P11.[3:0] 6 1) x denotes the port number, while [y:z] represents the bit field range. 2) The control of the additional driver is described in Chapter 9.3.3. 3) The output control of P2.[13:12] deviates from the standard definition, see Chapter 9.3.3. Port Width Note: When assigning functional signals to port pins, please consider the fact that the driver strength is selected for pin groups. Assign functions with similar requirements to pins within the same POCON control group. User’s Manual Parallel Ports, V1.16 9-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2.2 Port Output Register The port output register defines the values of the output pins if the pin is used as GPIO output. Pn_OUT (n=0-4) Port n Output Register Pn_OUT (n=6-11) Port n Output Register 15 14 6 5 4 3 2 1 0 P15 P14 P13 P12 P11 P10 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rwh rwh rwh rwh rwh rwh rwh rwh rwh rwh rwh rwh 10 Reset Value: 0000H 7 rwh 11 SFR (FFA2H+2*n) 8 rwh 12 Reset Value: 0000H 9 rwh 13 SFR (FFA2H+2*n) rwh Field Bits Type Description Px (x = 0-15) x rwh Port Output Bit x This bit defines the level at the output pin of port Pn, pin x if the output is selected as GPIO output. 0B The output level of Pn.x is 0. 1B The output level of Pn.x is 1. User’s Manual Parallel Ports, V1.16 9-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2.3 Port Output Modification Register The port output modification register contains the bits to individually set, clear, or toggle the value of the port n output register. P2_OMRH Port 2 Output Modification Register HighXSFR (E9CAH) P10_OMRH Port 10 Output Modification Register HighXSFR (E9EAH) Reset Value: XXXXH Reset Value: XXXXH 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 PC 15 PC 14 PC 13 PC 12 PC 11 PC 10 PC 9 PC 8 PS 15 PS 14 PS 13 PS 12 PS 11 PS 10 PS 9 PS 8 w w w w w w w w w w w w w w w w Field Bits Type Description PSx (x = 8-15) x-8 w Port Set Bit x Setting this bit sets or toggles the corresponding bit in the port output register Pn_OUT (see Table 9-3). On a read access, this bit returns an undefined value. PCx (x = 8-15) x w Port Clear Bit x Setting this bit clears or toggles the corresponding bit in the port output register Pn_OUT. (see Table 9-3). On a read access, this bit returns an undefined value. Pn_OMRL (n=0-4) Port n Output Modification Register LowXSFR (E9C0H+4*n) Reset Value: XXXXH Pn_OMRL (n=6-11) Port n Output Modification Register LowXSFR (E9C0H+4*n) Reset Value: XXXXH 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 PC 7 PC 6 PC 5 PC 4 PC 3 PC 2 PC 1 PC 0 PS 7 PS 6 PS 5 PS 4 PS 3 PS 2 PS 1 PS 0 w w w w w w w w w w w w w w w w User’s Manual Parallel Ports, V1.16 9-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Field Bits Type Description PSx (x = 0-7) x w Port Set Bit x Setting this bit sets or toggles the corresponding bit in the port output register Pn_OUT (see Table 9-3). On a read access, this bit returns an undefined value. PCx (x = 0-7) x+8 w Port Clear Bit x Setting this bit clears or toggles the corresponding bit in the port output register Pn_OUT. (see Table 9-3). On a read access, this bit returns an undefined value. Function of the PCx and PSx bit fields Table 9-3 Function of the Bits PCx and PSx PCx PSx Function 0 or no write access 0 or no write access Bit Pn_OUT.Px is not changed. 0 or no write access 1 Bit Pn_OUT.Px is set. 1 0 or no write access Bit Pn_OUT.Px is cleared. 1 1 Bit Pn_OUT.Px is toggled. Note: If a bit position is not written (one out of two bytes not targeted by a byte write), the corresponding value is considered as 0. Toggling a bit requires one 16-bit write. User’s Manual Parallel Ports, V1.16 9-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2.4 Port Input Register The port input register contains the values currently read at the input pins, also if a port line is assigned as output. Pn_IN (n=0-11) Port n Input Register P15_IN Port 15 Input Register 15 14 13 12 11 10 P15 P14 P13 P12 P11 P10 rh 1) rh rh rh rh rh SFR (FF80H+2*n) Reset Value: 0000H1) SFR (FF9EH) Reset Value: 0000H1) 9 8 7 6 5 4 3 2 1 0 P9 P8 P7 P6 P5 P4 P3 P2 P1 P0 rh rh rh rh rh rh rh rh rh rh Px bits for non implemented I/O lines are always read as 0. Field Bits Type Description Px (x = 0-15) x rh User’s Manual Parallel Ports, V1.16 Port Input Bit x This bit indicates the level at the input pin of port Pn, pin x. 0B The input level of Pn.x is 0. 1B The input level of Pn.x is 1. 9-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2.5 Port Input/Output Control Registers The port input/output control registers contain the bit fields to select the digital output and input driver characteristics, such as pull-up/down devices, port direction (input/output), open-drain and alternate output selections. The coding of the options is shown in Table 9-4. Depending on the port functionality not all of the input/output control registers may be implemented. The structure with one control bit field for each port pin located in different register offers the possibility to configure port pin functionality of a single pin without accessing some other PCx in the same register by word-oriented writes. P0_IOCRx (x=00-07) Port 0 Input/Output Control Register x XSFR (E800H+2*x) P1_IOCRx (x=00-07) Port 1 Input/Output Control Register x XSFR (E820H+2*x) P2_IOCRx (x=00-13) Port 2 Input/Output Control Register x XSFR (E840H+2*x) P3_IOCRx (x=00-07) Port 3 Input/Output Control Register x XSFR (E860H+2*x) P4_IOCRx (x=00-07) Port 4 Input/Output Control Register x XSFR (E880H+2*x) P6_IOCRx (x=00-03) Port 6 Input/Output Control Register x XSFR (E8C0H+2*x) P7_IOCRx (x=00-04) Port 7 Input/Output Control Register x XSFR (E8E0H+2*x) P8_IOCRx (x=00-06) Port 8 Input/Output Control Register x XSFR (E900H+2*x) P9_IOCRx (x=00-07) Port 9 Input/Output Control Register x XSFR (E920H+2*x) P10_IOCRx (x=00-15) Port 10 Input/Output Control Register x XSFR (E940H+2*x) P11_IOCRx (x=00-05) Port 11 Input/Output Control Register x XSFR (E960H+2*x) 15 14 13 12 User’s Manual Parallel Ports, V1.16 11 10 9 8 7 6 5 4 Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H Reset Value: 0000H 3 2 1 0 PC 0 r rw r 9-13 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Field Bits Type Description PC [7:4] rw Port Input/Output Control Bit see Table 9-4 0 [3:0], [15:8] r reserved Coding of the PC bit field The coding of the GPIO port behavior is done by the bit fields in the port control registers Pn_IOCRx. There’s a control bit field PC for each port pin. The bit fields PC are located in separate control registers in order to allow modifying a port pin (without influencing the others) with simple move operations. Note: When the pin direction is switched to output and the mode is test mode, the output characteristic must be push-pull only. Table 9-4 PC Coding PC[3:0] I/O Selected Pull-up/down / Selected Output Function Behavior in Power Saving Mode1) 0000B Direct Input No pull device connected Input value = Pn_OUT; no pull 0001B Pull-down device connected Input value = 0; pull-down 0010B Pull-up device connected Input value = 1; pull-up 0011B No pull device connected. In this mode Pn_OUT samples the pad input value continuously. Input value = Pn_OUT; Pn_OUT always samples input value while not in power save mode = freeze of input value; no pull 0100B 0101B Inverted No pull device connected Input value = Pn_OUT; no pull Input Pull-down device connected Input value = 1; pull-down 0110B Pull-up device connected Input value = 0; pull-up 0111B No pull device connected In this mode Pn_OUT samples the pad input value continuously. Input value = Pn_OUT; Pn_OUT always samples input value while not in power saving mode = freeze of input value; no pull2) User’s Manual Parallel Ports, V1.16 9-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-4 PC Coding PC[3:0] I/O Selected Pull-up/down / Selected Output Function Behavior in Power Saving Mode1) 1000B Output (Direct input) Pushpull General purpose Output Output driver off. Input Schmitt trigger off. Pn_OUT delivered to the internal logic; no pull 1001B 1010B 1011B 1100B 1101B 1110B 1111B Output (Direct input) Opendrain Output function ALT1 Output function ALT2 Output function ALT3 General purpose Output Output function ALT1 Output function ALT2 Output function ALT3 1) In power saving mode, the input Schmitt trigger is always switched off. A defined input value is driven to the internal circuitry instead of the level detected at the input pin. 2) If the IOCR setting is “inverted input”, then an inverted signal Pn_OUT is driven internally. The Pn_OUT register itself always contains the real, non-inverted input value of the pin. See Figure 7-1 and Figure 7-2. User’s Manual Parallel Ports, V1.16 9-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.2.6 Port Digital Input Disable Register Ports 5 and 15 have, additionally to the analog input functionality, digital input functionality too. In order to save switching of the internal Schmitt triggers of the digital inputs, they can be disabled by means of Px_DIDIS Register. P5_DIDIS Port 5 Digital Input Disable RegisterSFR (FE8AH) P15_DIDIS Port 15 Digital Input Disable RegisterSFR (FE9EH) 15 14 13 12 11 10 P15 P14 P13 P12 P11 P10 rw rw rw rw rw rw Reset Value: 0000H 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 Bit Type Description Py (y = 0-15) y rw Bit y Digital Input Control 0B Digital input stage (schmitt trigger) is enabled. 1B Digital input stage (schmitt trigger) is disabled, necessary if pin is used as analog input. User’s Manual Parallel Ports, V1.16 9-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3 Port Description The bit positions in the port registers always start right-aligned. For example, a port comprising only 8 pins only uses the bit positions [7:0] of the corresponding register. The remaining bit positions are filled with 0 (r). The pad driver mode registers may be different for each port. As a result, they are described independently for each port in the corresponding chapter. User’s Manual Parallel Ports, V1.16 9-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.1 Port 0 Port 0 is an 8-bit GPIO port. The registers of Port 0 are shown in Figure 9-4. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P0_IOCR00 : P0_IOCR07 P0_OMRL Data Registers P0_OUT P0_IN P0_POCON Port0_Regs.vsd Figure 9-4 Port 0 Register Overview Table 9-5 Port 0 Registers Register Short Name Register Long Name Address Offset Reset Value P0_OUT Port 0 Output Register FFA2H 0000H P0_IN Port 0 Input Register FF80H 0000H P0_OMRL Port 0 Output Modification Register Low E9C0H XXXXH P0_POCON Port 0 Output Control Register E8A0H 0000H P0_IOCR00 Port 0 Input/Output Control Register 0 E800H 0000H P0_IOCR01 Port 0 Input/Output Control Register 1 E802H 0000H P0_IOCR02 Port 0 Input/Output Control Register 2 E804H 0000H P0_IOCR03 Port 0 Input/Output Control Register 3 E806H 0000H P0_IOCR04 Port 0 Input/Output Control Register 4 E808H 0000H P0_IOCR05 Port 0 Input/Output Control Register 5 E80AH 0000H P0_IOCR06 Port 0 Input/Output Control Register 6 E80CH 0000H P0_IOCR07 Port 0 Input/Output Control Register 7 E80EH 0000H User’s Manual Parallel Ports, V1.16 9-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.2 Port 1 Port 1 is an 8-bit GPIO port. Theregisters of Port 1 are shown in Figure 9-5. Control Registers Modification Registers P1_IOCR00 : P1_IOCR07 P1_OMRL Data Registers P1_OUT P1_IN P1_POCON Port1_Regs.vsd Figure 9-5 Port 1 Register Overview For this port, all pins can be read as GPIO, from the Port Input Register. Table 9-6 Port 1 Registers Register Short Name Register Long Name Address Offset Reset Value P1_OUT Port 1 Output Register FFA4H 0000H P1_IN Port 1 Input Register FF82H 0000H P1_OMRL Port 1 Output Modification Register Low E9C4H XXXXH P1_POCON Port 1 Output Control Register E8A2H 0000H P1_IOCR00 Port 1 Input/Output Control Register 0 E820H 0000H P1_IOCR01 Port 1 Input/Output Control Register 1 E822H 0000H P1_IOCR02 Port 1 Input/Output Control Register 2 E824H 0000H P1_IOCR03 Port 1 Input/Output Control Register 3 E826H 0000H P1_IOCR04 Port 1 Input/Output Control Register 4 E828H 0000H P1_IOCR05 Port 1 Input/Output Control Register 5 E82AH 0000H P1_IOCR06 Port 1 Input/Output Control Register 6 E82CH 0000H P1_IOCR07 Port 1 Input/Output Control Register 7 E82EH 0000H User’s Manual Parallel Ports, V1.16 9-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.3 Port 2 Port 2 is an 14-bit GPIO port. The registers of Port 2 are shown in Figure 9-6. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers Data Registers P2_IOCR00 : P2_IOCR13 P2_OMRL P2_OMRH P2_OUT P2_IN P2_POCON Port 2_Regs.vsd Figure 9-6 Port 2 Register Overview Table 9-7 Port 2 Registers Register Short Name Register Long Name Address Offset Reset Value P2_OUT Port 2 Output Register FFA6H 0000H P2_IN Port 2 Input Register FF84H 0000H P2_OMRL Port 2 Output Modification Register Low E9C8H XXXXH P2_OMRH Port 2 Output Modification Register High E9CAH XXXXH P2_POCON Port 2 Output Control Register E8A4H 0000H P2_IOCR00 Port 2 Input/Output Control Register 0 E840H 0000H P2_IOCR01 Port 2 Input/Output Control Register 1 E842H 0000H P2_IOCR02 Port 2 Input/Output Control Register 2 E844H 0000H P2_IOCR03 Port 2 Input/Output Control Register 3 E846H 0000H P2_IOCR04 Port 2 Input/Output Control Register 4 E848H 0000H P2_IOCR05 Port 2 Input/Output Control Register 5 E84AH 0000H P2_IOCR06 Port 2 Input/Output Control Register 6 E84CH 0000H P2_IOCR07 Port 2 Input/Output Control Register 7 E84EH 0000H P2_IOCR08 Port 2 Input/Output Control Register 8 E850H 0000H P2_IOCR09 Port 2 Input/Output Control Register 9 E852H 0000H User’s Manual Parallel Ports, V1.16 9-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-7 Port 2 Registers (cont’d) Register Short Name Register Long Name Address Offset Reset Value P2_IOCR10 Port 2 Input/Output Control Register 10 E854H 0000H P2_IOCR11 Port 2 Input/Output Control Register 11 E856H 0000H P2_IOCR12 Port 2 Input/Output Control Register 12 E858H 0000H P2_IOCR13 Port 2 Input/Output Control Register 13 E85AH 0000H The CLKOUT Pad P2.8 In order to drive high frequency clock signals, a strong driver is connected in parallel to the normal output driver of P2.8. It is enabled instead of the standard driver while bitfield P2_POCON.PDM3 = xx1B. The strong clock driver only operates in strong driver sharp edge mode, i.e. it is not controlled by the driver-strength settings (P2_POCON.PDM2) for the standard driver. It has no pull devices but can be switched to input or output via register P2_IOCR08. Output Control for Pins P2.[13:12] Because bitfield P2_POCON.PDM3 controls the strong clock driver of P2.8, the driver mode of pins P2.[13:12] is controlled by the bitfield P2_POCON.PDM2, together with pins P2.[11:8]. The power saving behaviour of pins P2.[13:12] is controlled by bit P2_POCON.PPS3. User’s Manual Parallel Ports, V1.16 9-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.4 Port 3 Port 3 is an 8-bit GPIO port. The registers of Port 3 are shown in Figure 9-7. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P3_IOCR00 : P3_IOCR07 P3_OMRL Data Registers P3_OUT P3_IN P3_POCON Port3_Regs.vsd Figure 9-7 Port 3 Register Overview Table 9-8 Port 3 Registers Register Short Name Register Long Name Address Offset Reset Value P3_OUT Port 3 Output Register FFA8H 0000H P3_IN Port 3 Input Register FF86H 0000H P3_OMRL Port 3 Output Modification Register Low E9CCH XXXXH P3_POCON Port 3 Output Control Register E8A6H 0000H P3_IOCR00 Port 3 Input/Output Control Register 0 E860H 0000H P3_IOCR01 Port 3 Input/Output Control Register 1 E862H 0000H P3_IOCR02 Port 3 Input/Output Control Register 2 E864H 0000H P3_IOCR03 Port 3 Input/Output Control Register 3 E866H 0000H P3_IOCR04 Port 3 Input/Output Control Register 4 E868H 0000H P3_IOCR05 Port 3 Input/Output Control Register 5 E86AH 0000H P3_IOCR06 Port 3 Input/Output Control Register 6 E86CH 0000H P3_IOCR07 Port 3 Input/Output Control Register 7 E86EH 0000H User’s Manual Parallel Ports, V1.16 9-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.5 Port 4 Port 4 is an 8-bit GPIO port. The registers of Port 4 are shown in Figure 9-8. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P4_IOCR00 : P4_IOCR07 P4_OMRL Data Registers P4_OUT P4_IN P4_POCON Port4_Regs.vsd Figure 9-8 Port 4 Register Overview Table 9-9 Port 4 Registers Register Short Name Register Long Name Address Offset Reset Value P4_OUT Port 4 Output Register FFAAH 0000H P4_IN Port 4 Input Register FF88H 0000H P4_OMRL Port 4 Output Modification Register Low E9D0H XXXXH P4_POCON Port 4 Output Control Register E8A8H 0000H P4_IOCR00 Port 4 Input/Output Control Register 0 E880H 0000H P4_IOCR01 Port 4 Input/Output Control Register 1 E882H 0000H P4_IOCR02 Port 4 Input/Output Control Register 2 E884H 0000H P4_IOCR03 Port 4 Input/Output Control Register 3 E886H 0000H P4_IOCR04 Port 4 Input/Output Control Register 4 E888H 0000H P4_IOCR05 Port 4 Input/Output Control Register 5 E88AH 0000H P4_IOCR06 Port 4 Input/Output Control Register 6 E88CH 0000H P4_IOCR07 Port 4 Input/Output Control Register 7 E88EH 0000H User’s Manual Parallel Ports, V1.16 9-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.6 Port 5 Port 5 is a 16-bit analog or digital input port. To use the Port 5 as an analog input, the Schmitt trigger in the input stage must be disabled. This is achieved by setting the corresponding bit in the register P5_DIDIS. Data Registers Control Registers P5_DIDIS P5_IN Port5_Regs.vsd Figure 9-9 Port 5 Register Overview Table 9-10 Port 5 Registers Register Short Name Register Long Name Address Offset Reset Value P5_IN Port 5 Input Register FF8AH 0000H P5_DIDIS Port 5 Digital Input Disable Register FE8AH 0000H User’s Manual Parallel Ports, V1.16 9-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.7 Port 6 Port 6 is an 4-bit GPIO port. The registers of Port 6 are shown in Figure 9-10. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P6_IOCR00 : P6_IOCR03 P6_OMRL Data Registers P6_OUT P6_IN P6_POCON Port6_Regs.vsd Figure 9-10 Port 6 Register Overview Table 9-11 Port 6 Registers Register Short Name Register Long Name Address Offset Reset Value P6_OUT Port 6 Output Register FFAEH 0000H P6_IN Port 6 Input Register FF8CH 0000H P6_OMRL Port 6 Output Modification Register Low E9D8H XXXXH P6_POCON Port 6 Output Control Register E8ACH 0000H P6_IOCR00 Port 6 Input/Output Control Register 0 E8C0H 0000H P6_IOCR01 Port 6 Input/Output Control Register 1 E8C2H 0000H P6_IOCR02 Port 6 Input/Output Control Register 2 E8C4H 0000H P6_IOCR03 Port 6 Input/Output Control Register 3 E8C6H 0000H User’s Manual Parallel Ports, V1.16 9-25 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.8 Port 7 Port 7 is a 5-bit GPIO port. The port registers of Port 7 are shown in Figure 9-11. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P7_IOCR00 : P7_IOCR04 P7_OMRL Data Registers P7_OUT P7_IN P7_POCON Port7_Regs.vsd Figure 9-11 Port 7 Register Overview Table 9-12 Port 7 Registers Register Short Name Register Long Name Address Offset Reset Value P7_OUT Port 7 Output Register FFB0H 0000H P7_IN Port 7 Input Register FF8EH 0000H P7_OMRL Port 7 Output Modification Register Low E9DCH XXXXH P7_POCON Port 7 Output Control Register E8AEH 0000H P7_IOCR00 Port 7 Input/Output Control Register 0 E8E0H 0000H P7_IOCR01 Port 7 Input/Output Control Register 1 E8E2H 0000H P7_IOCR02 Port 7 Input/Output Control Register 2 E8E4H 0000H P7_IOCR03 Port 7 Input/Output Control Register 3 E8E6H 0000H P7_IOCR04 Port 7 Input/Output Control Register 4 E8E8H 0000H User’s Manual Parallel Ports, V1.16 9-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.9 Port 8 Port 8 is an 7-bit GPIO port. The registers of Port 8 are shown in Figure 9-12. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P8_IOCR00 : P8_IOCR06 P8_OMRL Data Registers P8_OUT P8_IN P8_POCON Port8_Regs.vsd Figure 9-12 Port 8 Register Overview Table 9-13 Port 8 Registers Register Short Name Register Long Name Address Offset Reset Value P8_OUT Port 8 Output Register FFB2H 0000H P8_IN Port 8 Input Register FF90H 0000H P8_OMRL Port 8 Output Modification Register Low E9E0H XXXXH P8_POCON Port 8 Output Control Register E8B0H 0000H P8_IOCR00 Port 8 Input/Output Control Register 0 E900H 0000H P8_IOCR01 Port 8 Input/Output Control Register 1 E902H 0000H P8_IOCR02 Port 8 Input/Output Control Register 2 E904H 0000H P8_IOCR03 Port 8 Input/Output Control Register 3 E906H 0000H P8_IOCR04 Port 8 Input/Output Control Register 4 E908H 0000H P8_IOCR05 Port 8 Input/Output Control Register 5 E90AH 0000H P8_IOCR06 Port 8 Input/Output Control Register 6 E90CH 0000H User’s Manual Parallel Ports, V1.16 9-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.10 Port 9 Port 9 is an 8-bit GPIO port. The port registers of Port 9 are shown in Figure 9-13. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P9_IOCR00 : P9_IOCR07 P9_OMRL Data Registers P9_OUT P9_IN P9_POCON Port9_Regs.vsd Figure 9-13 Port 9 Register Overview Table 9-14 Port 9 Registers Register Short Name Register Long Name Address Offset Reset Value P9_OUT Port 9 Output Register FFB4H 0000H P9_IN Port 9 Input Register FF92H 0000H P9_OMRL Port 9 Output Modification Register Low E9E4H XXXXH P9_POCON Port 9 Output Control Register E8B2H 0000H P9_IOCR00 Port 9 Input/Output Control Register 0 E920H 0000H P9_IOCR01 Port 9 Input/Output Control Register 1 E922H 0000H P9_IOCR02 Port 9 Input/Output Control Register 2 E924H 0000H P9_IOCR03 Port 9 Input/Output Control Register 3 E926H 0000H P9_IOCR04 Port 9 Input/Output Control Register 4 E928H 0000H P9_IOCR05 Port 9 Input/Output Control Register 5 E92AH 0000H P9_IOCR06 Port 9 Input/Output Control Register 6 E92CH 0000H P9_IOCR07 Port 9 Input/Output Control Register 7 E92EH 0000H User’s Manual Parallel Ports, V1.16 9-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.11 Port 10 Port 10 is a 16-bit GPIO port. The registers of Port 10 are shown in Figure 9-14. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers Data Registers P10_IOCR00 : P10_IOCR15 P10_OMRL P10_OUT P10_OMRH P10_IN P10_POCON Port10_Regs.vsd Figure 9-14 Port 10 Register Overview Table 9-15 Port 10 Registers Register Short Name Register Long Name Address Offset Reset Value P10_OUT Port 10 Output Register FFB6H 0000H P10_IN Port 10 Input Register FF94H 0000H P10_OMRL Port 10 Output Modification Register Low E9E8H XXXXH P10_OMRH Port 10 Output Modification Register High E9EAH XXXXH P10_POCON Port 10 Output Control Register E8B4H 0000H P10_IOCR00 Port 10 Input/Output Control Register 0 E940H 0000H P10_IOCR01 Port 10 Input/Output Control Register 1 E942H 0000H P10_IOCR02 Port 10 Input/Output Control Register 2 E944H 0000H P10_IOCR03 Port 10 Input/Output Control Register 3 E946H 0000H P10_IOCR04 Port 10 Input/Output Control Register 4 E948H 0000H P10_IOCR05 Port 10 Input/Output Control Register 5 E94AH 0000H P10_IOCR06 Port 10 Input/Output Control Register 6 E94CH 0000H P10_IOCR07 Port 10 Input/Output Control Register 7 E94EH 0000H P10_IOCR08 Port 10 Input/Output Control Register 8 E950H 0000H P10_IOCR09 Port 10 Input/Output Control Register 9 E952H 0000H User’s Manual Parallel Ports, V1.16 9-29 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-15 Register Short Name Port 10 Registers (cont’d) Register Long Name Address Offset Reset Value P10_IOCR10 Port 10 Input/Output Control Register 10 E954H 0000H P10_IOCR11 Port 10 Input/Output Control Register 11 E956H 0000H P10_IOCR12 Port 10 Input/Output Control Register 12 E958H 0000H P10_IOCR13 Port 10 Input/Output Control Register 13 E95AH 0000H P10_IOCR14 Port 10 Input/Output Control Register 14 E95CH 0000H P10_IOCR15 Port 10 Input/Output Control Register 15 E95EH 0000H User’s Manual Parallel Ports, V1.16 9-30 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.12 Port 11 Port 11 is a 6-bit GPIO port. The registers of Port 11 are shown in Figure 9-15. For this port, all pins can be read as GPIO, from the Port Input Register. Control Registers Modification Registers P11_IOCR00 : P11_IOCR05 P11_OMRL Data Registers P11_OUT P11_IN P11_POCON Port11_Regs.vsd Figure 9-15 Port 11 Register Overview Table 9-16 Port 11 Registers Register Short Name Register Long Name Address Offset Reset Value P11_OUT Port 11 Output Register FFB8H 0000H P11_IN Port 11 Input Register FF96H 0000H P11_OMRL Port 11 Output Modification Register Low E9ECH XXXXH P11_POCON Port 11 Output Control Register E8B6H 0000H P11_IOCR00 Port 11 Input/Output Control Register 0 E960H 0000H P11_IOCR01 Port 11 Input/Output Control Register 1 E962H 0000H P11_IOCR02 Port 11 Input/Output Control Register 2 E964H 0000H P11_IOCR03 Port 11 Input/Output Control Register 3 E966H 0000H P11_IOCR04 Port 11 Input/Output Control Register 4 E968H 0000H P11_IOCR05 Port 11 Input/Output Control Register 5 E96AH 0000H User’s Manual Parallel Ports, V1.16 9-31 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.3.13 Port 15 Port 15 is an 8-bit analog or digital input port. To use the Port 15 as an analog input, the Schmitt trigger in the input stage must be disabled. This is achieved by setting the corresponding bit in the register P15_DIDIS. Data Registers Control Registers P15_DIDIS P15_IN Port15_Regs.vsd Figure 9-16 Port 15 Register Overview Table 9-17 Port 15 Registers Register Short Name Register Long Name Address Offset Reset Value P15_IN Port 15 Input Register FF9EH 0000H P15_DIDIS Port 15 Digital Input Disable Register FE9EH 0000H User’s Manual Parallel Ports, V1.16 9-32 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports 9.4 Pin Description Each port pin of the XC27x5X can serve several functions of different modules. Also, most functions are available on several port pins. This enables an application so select the optimal connections for its specific circumstances. A pin can output its own port output signal or one of up to three signals coming from the peripherals. Its input signal is available in its own input register and at several peripherals. Note: Output signals are selected at the respective port pin, input signals are selected at the respective peripheral. Optionally a pin can be fully controlled by a peripheral, in case the peripheral is enabled (for example, EBC). Table 9-18 summarizes the various functions of each port and pin of the XC27x5X. The ’Pin’ column references to the PG-LQFP-144 package. Notes to Pin Definitions 1. Ctrl.: The output signal for a port pin is selected via bitfield PC in the associated register Px_IOCRy. Output O0 is selected by setting the respective bitfield PC to 1x00B, output O1 is selected by 1x01B, etc. Output signal OH is controlled by hardware. 2. Type: Indicates the employed pad type (St=standard pad, Sp=special pad, DP=double pad, In=input pad, PS=power supply) and its power supply domain (A, B, M, 1). Table 9-18 Pin Definitions and Functions Pin Symbol Ctrl. Type Function 3 TESTM I In/B 4 P7.2 O0 / I St/B Bit 2 of Port 7, General Purpose Input/Output EMUX0 O1 St/B External Analog MUX Control Output 0 (ADC1) CCU62_CCP I OS0A St/B CCU62 Position Input 0 TDI_C St/B JTAG Test Data Input User’s Manual Parallel Ports, V1.16 I Testmode Enable Enables factory test modes, must be held HIGH for normal operation (connect to VDDPB). An internal pullup device will hold this pin high when nothing is driving it. 9-33 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. Type Function 5 P8.4 O0 / I St/B Bit 4 of Port 8, General Purpose Input/Output CCU60_COU O1 T61 St/B CCU60 Channel 1 Output CCU62_CC6 1 O2 St/B CCU62 Channel 1 Output TMS_D I St/B JTAG Test Mode Selection Input CCU62_CC6 1INB I St/B CCU62 Channel 1 Input 6 TRST I In/B Test-System Reset Input For normal system operation, pin TRST should be held low. A high level at this pin at the rising edge of PORST activates the XC27x5X’s debug system. In this case, pin TRST must be driven low once to reset the debug system. An internal pulldown device will hold this pin low when nothing is driving it. 7 P8.3 O0 / I St/B 8 Bit 3 of Port 8, General Purpose Input/Output CCU60_COU O1 T60 St/B CCU60 Channel 0 Output CCU62_CC6 0 O2 St/B CCU62 Channel 0 Output TDI_D I St/B JTAG Test Data Input CCU62_CC6 0INB I St/B CCU62 Channel 0 Input P7.0 O0 / I St/B Bit 0 of Port 7, General Purpose Input/Output T3OUT O1 St/B GPT12E Timer T3 Toggle Latch Output T6OUT O2 St/B GPT12E Timer T6 Toggle Latch Output TDO_A OH St/B DAP1/JTAG Test Data Output ESR2_1 I St/B ESR2 Trigger Input 1 User’s Manual Parallel Ports, V1.16 9-34 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 9 P7.3 O0 / I St/B Bit 3 of Port 7, General Purpose Input/Output EMUX1 O1 St/B External Analog MUX Control Output 1 (ADC1) U0C1_DOUT O2 St/B USIC0 Channel 1 Shift Data Output U0C0_DOUT O3 St/B USIC0 Channel 0 Shift Data Output CCU62_CCP I OS1A St/B CCU62 Position Input 1 TMS_C I St/B JTAG Test Mode Selection Input U0C1_DX0F I St/B USIC0 Channel 1 Shift Data Input P8.2 O0 / I St/B Bit 2 of Port 8, General Purpose Input/Output CCU60_CC6 2 O1 St/B CCU60 Channel 2 Output TxDC1 O2 St/B CAN Node 1 Transmit Data Output U1C1_DOUT O3 St/B USIC1 Channel 1 Shift Data output CCU60_CC6 2INB I St/B CCU60 Channel 2 Input P7.1 O0 / I St/B Bit 1 of Port 7, General Purpose Input/Output EXTCLK O1 St/B Programmable Clock Signal Output CCU62_CTR APA I St/B CCU62 Emergency Trap Input BRKIN_C I St/B OCDS Break Signal Input P7.4 O0 / I St/B Bit 4 of Port 7, General Purpose Input/Output EMUX2 O1 St/B External Analog MUX Control Output 2 (ADC1) U0C1_DOUT O2 St/B USIC0 Channel 1 Shift Data Output U0C1_SCLK OUT St/B USIC0 Channel 1 Shift Clock Output CCU62_CCP I OS2A St/B CCU62 Position Input 2 TCK_C I St/B DAP0/JTAG Clock Input U0C0_DX0D I St/B USIC0 Channel 0 Shift Data Input U0C1_DX1E I St/B USIC0 Channel 1 Shift Clock Input 10 11 12 User’s Manual Parallel Ports, V1.16 O3 Type Function 9-35 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 13 P8.1 O0 / I St/B Bit 1 of Port 8, General Purpose Input/Output CCU60_CC6 1 O1 St/B CCU60 Channel 1 Output CCU60_CC6 1INB I St/B CCU60 Channel 1 Input RxDC1F I St/B CAN Node 1 Receive Data Input P8.0 O0 / I St/B Bit 0 of Port 8, General Purpose Input/Output CCU60_CC6 0 O1 St/B CCU60 Channel 0 Output CCU60_CC6 0INB I St/B CCU60 Channel 0 Input P6.0 O0 / I St/A Bit 0 of Port 6, General Purpose Input/Output EMUX0 O1 St/A External Analog MUX Control Output 0 (ADC0) BRKOUT O3 St/A OCDS Break Signal Output ADCx_REQG I TyG St/A External Request Gate Input for ADC0/1 U1C1_DX0E I St/A USIC1 Channel 1 Shift Data Input P6.1 O0 / I St/A Bit 1 of Port 6, General Purpose Input/Output EMUX1 O1 St/A External Analog MUX Control Output 1 (ADC0) T3OUT O2 St/A GPT12E Timer T3 Toggle Latch Output U1C1_DOUT O3 St/A USIC1 Channel 1 Shift Data Output ADCx_REQT I RyE St/A External Request Trigger Input for ADC0/1 ESR1_6 I St/A ESR1 Trigger Input 6 P6.2 O0 / I St/A Bit 2 of Port 6, General Purpose Input/Output EMUX2 O1 St/A External Analog MUX Control Output 2 (ADC0) T6OUT O2 St/A GPT12E Timer T6 Toggle Latch Output U1C1_SCLK OUT O3 St/A USIC1 Channel 1 Shift Clock Output U1C1_DX1C I St/A USIC1 Channel 1 Shift Clock Input 14 16 17 18 User’s Manual Parallel Ports, V1.16 Type Function 9-36 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 19 P6.3 O0 / I St/A Bit 3 of Port 6, General Purpose Input/Output T3OUT O2 St/A GPT12E Timer T3 Toggle Latch Output U1C1_SELO 0 O3 St/A USIC1 Channel 1 Select/Control 0 Output U1C1_DX2D I St/A USIC1 Channel 1 Shift Control Input ADCx_REQT I RyF St/A External Request Trigger Input for ADC0/1 P15.0 I In/A Bit 0 of Port 15, General Purpose Input ADC1_CH0 I In/A Analog Input Channel 0 for ADC1 P15.1 I In/A Bit 1 of Port 15, General Purpose Input ADC1_CH1 I In/A Analog Input Channel 1 for ADC1 P15.2 I In/A Bit 2 of Port 15, General Purpose Input ADC1_CH2 I In/A Analog Input Channel 2 for ADC1 T5INA I In/A GPT12E Timer T5 Count/Gate Input P15.3 I In/A Bit 3 of Port 15, General Purpose Input ADC1_CH3 I In/A Analog Input Channel 3 for ADC1 T5EUDA I In/A GPT12E Timer T5 External Up/Down Control Input P15.4 I In/A Bit 4 of Port 15, General Purpose Input ADC1_CH4 I In/A Analog Input Channel 4 for ADC1 T6INA I In/A GPT12E Timer T6 Count/Gate Input P15.5 I In/A Bit 5 of Port 15, General Purpose Input ADC1_CH5 I In/A Analog Input Channel 5 for ADC1 T6EUDA I In/A GPT12E Timer T6 External Up/Down Control Input P15.6 I In/A Bit 6 of Port 15, General Purpose Input ADC1_CH6 I In/A Analog Input Channel 6 for ADC1 P15.7 I In/A Bit 7 of Port 15, General Purpose Input ADC1_CH7 I In/A Analog Input Channel 7 for ADC1 VAREF1 VAREF0 - PS/A Reference Voltage for A/D Converter ADC1 - PS/A Reference Voltage for A/D Converter ADC0 21 22 23 24 25 26 27 28 29 30 User’s Manual Parallel Ports, V1.16 Type Function 9-37 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. Type Function 31 VAGND - PS/A Reference Ground for A/D Converters ADC0/1 32 P5.0 I In/A Bit 0 of Port 5, General Purpose Input ADC0_CH0 I In/A Analog Input Channel 0 for ADC0 P5.1 I In/A Bit 1 of Port 5, General Purpose Input ADC0_CH1 I In/A Analog Input Channel 1 for ADC0 P5.2 I In/A Bit 2 of Port 5, General Purpose Input ADC0_CH2 I In/A Analog Input Channel 2 for ADC0 TDI_A I In/A JTAG Test Data Input P5.3 I In/A Bit 3 of Port 5, General Purpose Input ADC0_CH3 I In/A Analog Input Channel 3 for ADC0 T3INA I In/A GPT12E Timer T3 Count/Gate Input P5.4 I In/A Bit 4 of Port 5, General Purpose Input ADC0_CH4 I In/A Analog Input Channel 4 for ADC0 CCU63_T12 HRB I In/A External Run Control Input for T12 of CCU63 T3EUDA I In/A GPT12E Timer T3 External Up/Down Control Input TMS_A I In/A JTAG Test Mode Selection Input P5.5 I In/A Bit 5 of Port 5, General Purpose Input ADC0_CH5 I In/A Analog Input Channel 5 for ADC0 CCU60_T12 HRB I In/A External Run Control Input for T12 of CCU60 P5.6 I In/A Bit 6 of Port 5, General Purpose Input ADC0_CH6 I In/A Analog Input Channel 6 for ADC0 P5.7 I In/A Bit 7 of Port 5, General Purpose Input ADC0_CH7 I In/A Analog Input Channel 7 for ADC0 33 34 35 39 40 41 42 User’s Manual Parallel Ports, V1.16 9-38 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. Type Function 43 P5.8 I In/A Bit 8 of Port 5, General Purpose Input ADC0_CH8 I In/A Analog Input Channel 8 for ADC0 CCU6x_T12H I RC In/A External Run Control Input for T12 of CCU60/1/ 2/3 CCU6x_T13H I RC In/A External Run Control Input for T13 of CCU60/1/ 2/3 U2C0_DX0F I In/A USIC2 Channel 0 Shift Data Input P5.9 I In/A Bit 9 of Port 5, General Purpose Input ADC0_CH9 I In/A Analog Input Channel 9 for ADC0 CC2_T7IN I In/A CAPCOM2 Timer T7 Count Input P5.10 I In/A Bit 10 of Port 5, General Purpose Input ADC0_CH10 I In/A Analog Input Channel 10 for ADC0 BRKIN_A I In/A OCDS Break Signal Input U2C1_DX0F I In/A USIC2 Channel 1 Shift Data Input CCU61_T13 HRA I In/A External Run Control Input for T13 of CCU61 P5.11 I In/A Bit 11 of Port 5, General Purpose Input ADC0_CH11 I In/A Analog Input Channel 11 for ADC0 P5.12 I In/A Bit 12 of Port 5, General Purpose Input ADC0_CH12 I In/A Analog Input Channel 12 for ADC0 P5.13 I In/A Bit 13 of Port 5, General Purpose Input ADC0_CH13 I In/A Analog Input Channel 13 for ADC0 CCU63_T13 HRF I In/A External Run Control Input for T13 of CCU63 P5.14 I In/A Bit 14 of Port 5, General Purpose Input ADC0_CH14 I In/A Analog Input Channel 14 for ADC0 P5.15 I In/A Bit 15 of Port 5, General Purpose Input ADC0_CH15 I In/A Analog Input Channel 15 for ADC0 44 45 46 47 48 49 50 User’s Manual Parallel Ports, V1.16 9-39 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 51 P2.12 O0 / I St/B Bit 12 of Port 2, General Purpose Input/Output U0C0_SELO 4 O1 St/B USIC0 Channel 0 Select/Control 4 Output U0C1_SELO 3 O2 St/B USIC0 Channel 1 Select/Control 3 Output READY I St/B External Bus Interface READY Input P2.11 O0 / I St/B Bit 11 of Port 2, General Purpose Input/Output U0C0_SELO 2 O1 St/B USIC0 Channel 0 Select/Control 2 Output U0C1_SELO 2 O2 St/B USIC0 Channel 1 Select/Control 2 Output BHE/WRH OH St/B External Bus Interf. High-Byte Control Output Can operate either as Byte High Enable (BHE) or as Write strobe for High Byte (WRH). P11.5 O0 / I St/B Bit 5 of Port 11, General Purpose Input/Output CCU61_CC6 0 O1 St/B CCU61 Channel 0 Output CCU61_COU O2 T63 St/B CCU61 Channel 3 Output CCU61_CC6 0INB I St/B CCU61 Channel 0 Input P2.0 O0 / I St/B Bit 0 of Port 2, General Purpose Input/Output CCU63_CC6 0 O2 CCU63 Channel 0 Output AD13 OH / I St/B External Bus Interface Address/Data Line 13 RxDC0C I St/B CAN Node 0 Receive Data Input CCU63_CC6 0INB I St/B CCU63 Channel 0 Input T5INB I St/B GPT12E Timer T5 Count/Gate Input 52 53 55 User’s Manual Parallel Ports, V1.16 Type Function St/B 9-40 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 56 P2.1 O0 / I St/B Bit 1 of Port 2, General Purpose Input/Output TxDC0 O1 St/B CAN Node 0 Transmit Data Output CCU63_CC6 1 O2 St/B CCU63 Channel 1 Output AD14 OH / I St/B External Bus Interface Address/Data Line 14 CCU63_CC6 1INB I St/B CCU63 Channel 1 Input T5EUDB I St/B GPT12E Timer T5 External Up/Down Control Input ESR1_5 I St/B ESR1 Trigger Input 5 P11.4 O0 / I St/B Bit 4 of Port 11, General Purpose Input/Output CCU61_CC6 2 O1 St/B CCU61 Channel 2 Output CCU61_CC6 2INB I St/B CCU61 Channel 0 Input P2.2 O0 / I St/B Bit 2 of Port 2, General Purpose Input/Output TxDC1 O1 St/B CAN Node 1 Transmit Data Output CCU63_CC6 2 O2 St/B CCU63 Channel 2 Output AD15 OH / I St/B External Bus Interface Address/Data Line 15 CCU63_CC6 2INB I St/B CCU63 Channel 2 Input ESR2_5 I St/B ESR2 Trigger Input 5 P11.3 O0 / I St/B 57 58 59 60 Type Function Bit 3 of Port 11, General Purpose Input/Output CCU61_COU O1 T63 St/B CCU61 Channel 3 Output CCU61_COU O2 T62 St/B CCU61 Channel 2 Output CCU61_T13 HRF I St/B External Run Control Input for T13 of CCU61 P4.0 O0 / I St/B Bit 0 of Port 4, General Purpose Input/Output CC2_CC24 O3 / I St/B CAPCOM2 CC24IO Capture Inp./ Compare Out. CS0 OH External Bus Interface Chip Select 0 Output User’s Manual Parallel Ports, V1.16 St/B 9-41 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 61 P2.3 O0 / I St/B 62 63 Type Function Bit 3 of Port 2, General Purpose Input/Output U0C0_DOUT O1 St/B USIC0 Channel 0 Shift Data Output CCU63_COU O2 T63 St/B CCU63 Channel 3 Output CC2_CC16 O3 / I St/B CAPCOM2 CC16IO Capture Inp./ Compare Out. A16 OH St/B External Bus Interface Address Line 16 ESR2_0 I St/B ESR2 Trigger Input 0 U0C0_DX0E I St/B USIC0 Channel 0 Shift Data Input U0C1_DX0D I St/B USIC0 Channel 1 Shift Data Input RxDC0A I St/B CAN Node 0 Receive Data Input P11.2 O0 / I St/B Bit 2 of Port 11, General Purpose Input/Output CCU61_CC6 1 O1 St/B CCU61 Channel 1 Output CCU63_CCP I OS2A St/B CCU63 Position Input 2 CCU61_CC6 1INB I St/B CCU61 Channel 1 Input P4.1 O0 / I St/B Bit 1 of Port 4, General Purpose Input/Output CC2_CC25 O3 / I St/B CAPCOM2 CC25IO Capture Inp./ Compare Out. CS1 OH St/B External Bus Interface Chip Select 1 Output CCU62_CCP I OS0B St/B CCU62 Position Input 0 T4EUDB I St/B GPT12E Timer T4 External Up/Down Control Input ESR1_8 I St/B ESR1 Trigger Input 8 User’s Manual Parallel Ports, V1.16 9-42 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 64 P2.4 O0 / I St/B 65 66 67 Type Function Bit 4 of Port 2, General Purpose Input/Output U0C1_DOUT O1 St/B USIC0 Channel 1 Shift Data Output TxDC0 O2 St/B CAN Node 0 Transmit Data Output CC2_CC17 O3 / I St/B CAPCOM2 CC17IO Capture Inp./ Compare Out. A17 OH St/B External Bus Interface Address Line 17 ESR1_0 I St/B ESR1 Trigger Input 0 U0C0_DX0F I St/B USIC0 Channel 0 Shift Data Input RxDC1A I St/B CAN Node 1 Receive Data Input P11.1 O0 / I St/B Bit 1 of Port 11, General Purpose Input/Output CCU61_COU O1 T61 St/B CCU61 Channel 1 Output TxDC0 St/B CAN Node 0 Transmit Data Output CCU63_CCP I OS1A St/B CCU63 Position Input 1 CCU61_CTR APD I St/B CCU61 Emergency Trap Input P11.0 O0 / I St/B O2 Bit 0 of Port 11, General Purpose Input/Output CCU61_COU O1 T60 St/B CCU61 Channel 0 Output CCU63_CCP I OS0A St/B CCU63 Position Input 0 RxDC0F I St/B CAN Node 0 Receive Data Input ESR1_7 I St/B ESR1 Trigger Input 7 P2.5 O0 / I St/B Bit 5 of Port 2, General Purpose Input/Output U0C0_SCLK OUT O1 St/B USIC0 Channel 0 Shift Clock Output TxDC0 O2 St/B CAN Node 0 Transmit Data Output CC2_CC18 O3 / I St/B CAPCOM2 CC18IO Capture Inp./ Compare Out. A18 OH St/B External Bus Interface Address Line 18 U0C0_DX1D I St/B USIC0 Channel 0 Shift Clock Input ESR1_10 I St/B ESR1 Trigger Input 10 User’s Manual Parallel Ports, V1.16 9-43 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 68 P4.2 O0 / I St/B Bit 2 of Port 4, General Purpose Input/Output CC2_CC26 O3 / I St/B CAPCOM2 CC26IO Capture Inp./ Compare Out. CS2 OH St/B External Bus Interface Chip Select 2 Output T2INA I St/B GPT12E Timer T2 Count/Gate Input CCU62_CCP I OS1B St/B CCU62 Position Input 1 69 70 71 Type Function P2.6 O0 / I St/B Bit 6 of Port 2, General Purpose Input/Output U0C0_SELO 0 O1 St/B USIC0 Channel 0 Select/Control 0 Output U0C1_SELO 1 O2 St/B USIC0 Channel 1 Select/Control 1 Output CC2_CC19 O3 / I St/B CAPCOM2 CC19IO Capture Inp./ Compare Out. A19 OH St/B External Bus Interface Address Line 19 U0C0_DX2D I St/B USIC0 Channel 0 Shift Control Input RxDC0D I St/B CAN Node 0 Receive Data Input ESR2_6 I St/B ESR2 Trigger Input 6 P4.4 O0 / I St/B Bit 4 of Port 4, General Purpose Input/Output CC2_CC28 O3 / I St/B CAPCOM2 CC28IO Capture Inp./ Compare Out. CS4 OH St/B External Bus Interface Chip Select 4 Output CLKIN2 I St/B Clock Signal Input 2 P4.3 O0 / I St/B U0C1_DOUT O1 St/B Bit 3 of Port 4, General Purpose Input/Output USIC0 Channel 1 Shift Data Output CC2_CC27 O3 / I St/B CAPCOM2 CC27IO Capture Inp./ Compare Out. CS3 OH St/B External Bus Interface Chip Select 3 Output T2EUDA I St/B GPT12E Timer T2 External Up/Down Control Input CCU62_CCP I OS2B St/B CCU62 Position Input 2 User’s Manual Parallel Ports, V1.16 9-44 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 75 P0.0 O0 / I St/B 76 77 78 Type Function Bit 0 of Port 0, General Purpose Input/Output U1C0_DOUT O1 St/B USIC1 Channel 0 Shift Data Output CCU61_CC6 0 O3 St/B CCU61 Channel 0 IOutput A0 OH St/B External Bus Interface Address Line 0 U1C0_DX0A I St/B USIC1 Channel 0 Shift Data Input CCU61_CC6 0INA I St/B CCU61 Channel 0 Input ESR1_11 I St/B ESR1 Trigger Input 11 P4.5 O0 / I St/B Bit 5 of Port 4, General Purpose Input/Output CC2_CC29 O3 / I St/B CAPCOM2 CC29IO Capture Inp./Compare Out. CCU61_CCP I OS0A St/B CCU61 Position Input 0 ESR2_10 I St/B ESR2 Trigger Input 10 P4.6 O0 / I St/B Bit 6 of Port 4, General Purpose Input/Output CC2_CC30 O3 / I St/B CAPCOM2 CC30IO Capture Inp./ Compare Out. T4INA I St/B GPT12E Timer T4 Count/Gate Input CCU61_CCP I OS1A St/B CCU61 Position Input 1 P2.7 O0 / I St/B Bit 7 of Port 2, General Purpose Input/Output U0C1_SELO 0 O1 St/B USIC0 Channel 1 Select/Control 0 Output U0C0_SELO 1 O2 St/B USIC0 Channel 0 Select/Control 1 Output CC2_CC20 O3 / I St/B CAPCOM2 CC20IO Capture Inp./ Compare Out. A20 OH St/B External Bus Interface Address Line 20 U0C1_DX2C I St/B USIC0 Channel 1 Shift Control Input RxDC1C I St/B CAN Node 1 Receive Data Input ESR2_7 I St/B ESR2 Trigger Input 7 User’s Manual Parallel Ports, V1.16 9-45 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 79 P0.1 O0 / I St/B 80 81 82 Type Function Bit 1 of Port 0, General Purpose Input/Output U1C0_DOUT O1 St/B USIC1 Channel 0 Shift Data Output TxDC0 O2 St/B CAN Node 0 Transmit Data Output CCU61_CC6 1 O3 St/B CCU61 Channel 1 Output A1 OH St/B External Bus Interface Address Line 1 U1C0_DX0B I St/B USIC1 Channel 0 Shift Data Input CCU61_CC6 1INA I St/B CCU61 Channel 1 Input U1C0_DX1A I St/B USIC1 Channel 0 Shift Clock Input P2.8 O0 / I DP/B Bit 8 of Port 2, General Purpose Input/Output U0C1_SCLK OUT O1 DP/B USIC0 Channel 1 Shift Clock Output EXTCLK O2 DP/B Programmable Clock Signal Output CC2_CC21 O3 / I DP/B CAPCOM2 CC21IO Capture Inp./ Compare Out. A21 OH DP/B External Bus Interface Address Line 21 U0C1_DX1D I DP/B USIC0 Channel 1 Shift Clock Input P4.7 O0 / I St/B Bit 7 of Port 4, General Purpose Input/Output CC2_CC31 O3 / I St/B CAPCOM2 CC31IO Capture Inp./ Compare Out. T4EUDA I St/B GPT12E Timer T4 External Up/Down Control Input CCU61_CCP I OS2A St/B CCU61 Position Input 2 P2.9 1) O0 / I St/B Bit 9 of Port 2, General Purpose Input/Output U0C1_DOUT O1 St/B USIC0 Channel 1 Shift Data Output TxDC1 O2 St/B CAN Node 1 Transmit Data Output CC2_CC22 O3 / I St/B CAPCOM2 CC22IO Capture Inp./ Compare Out. A22 OH St/B External Bus Interface Address Line 22 CLKIN1 I St/B Clock Signal Input 1 TCK_A I St/B DAP0/JTAG Clock Input User’s Manual Parallel Ports, V1.16 9-46 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 83 P0.2 O0 / I St/B Bit 2 of Port 0, General Purpose Input/Output U1C0_SCLK OUT O1 St/B USIC1 Channel 0 Shift Clock Output TxDC0 O2 St/B CAN Node 0 Transmit Data Output CCU61_CC6 2 O3 St/B CCU61 Channel 2 Output A2 OH St/B External Bus Interface Address Line 2 U1C0_DX1B I St/B USIC1 Channel 0 Shift Clock Input CCU61_CC6 2INA I St/B CCU61 Channel 2 Input P10.0 O0 / I St/B 84 85 Type Function Bit 0 of Port 10, General Purpose Input/Output U0C1_DOUT O1 St/B USIC0 Channel 1 Shift Data Output CCU60_CC6 0 O2 St/B CCU60 Channel 0 Output AD0 OH / I St/B External Bus Interface Address/Data Line 0 CCU60_CC6 0INA I St/B CCU60 Channel 0 Input ESR1_2 I St/B ESR1 Trigger Input 2 U0C0_DX0A I St/B USIC0 Channel 0 Shift Data Input U0C1_DX0A I St/B USIC0 Channel 1 Shift Data Input P3.0 O0 / I St/B Bit 0 of Port 3, General Purpose Input/Output U2C0_DOUT O1 St/B USIC2 Channel 0 Shift Data Output BREQ OH St/B External Bus Request Output ESR1_1 I St/B ESR1 Trigger Input 1 U2C0_DX0A I St/B USIC2 Channel 0 Shift Data Input U2C0_DX1A I St/B USIC2 Channel 0 Shift Clock Input User’s Manual Parallel Ports, V1.16 9-47 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 86 P10.1 O0 / I St/B 87 88 Bit 1 of Port 10, General Purpose Input/Output U0C0_DOUT O1 St/B USIC0 Channel 0 Shift Data Output CCU60_CC6 1 O2 St/B CCU60 Channel 1 Output AD1 OH / I St/B External Bus Interface Address/Data Line 1 CCU60_CC6 1INA I St/B CCU60 Channel 1 Input U0C0_DX0B I St/B USIC0 Channel 0 Shift Data Input U0C0_DX1A I St/B USIC0 Channel 0 Shift Clock Input P0.3 O0 / I St/B Bit 3 of Port 0, General Purpose Input/Output U1C0_SELO 0 O1 St/B USIC1 Channel 0 Select/Control 0 Output U1C1_SELO 1 O2 St/B USIC1 Channel 1 Select/Control 1 Output CCU61_COU O3 T60 St/B CCU61 Channel 0 Output A3 OH St/B External Bus Interface Address Line 3 U1C0_DX2A I St/B USIC1 Channel 0 Shift Control Input RxDC0B I St/B CAN Node 0 Receive Data Input P3.1 O0 / I St/B U2C0_DOUT O1 89 Type Function St/B Bit 1 of Port 3, General Purpose Input/Output USIC2 Channel 0 Shift Data Output HLDA OH / I St/B External Bus Hold Acknowledge Output/Input Output in master mode, input in slave mode. U2C0_DX0B I USIC2 Channel 0 Shift Data Input P10.2 O0 / I St/B Bit 2 of Port 10, General Purpose Input/Output U0C0_SCLK OUT O1 St/B USIC0 Channel 0 Shift Clock Output CCU60_CC6 2 O2 St/B CCU60 Channel 2 Output AD2 OH / I St/B External Bus Interface Address/Data Line 2 CCU60_CC6 2INA I St/B CCU60 Channel 2 Input U0C0_DX1B I St/B USIC0 Channel 0 Shift Clock Input User’s Manual Parallel Ports, V1.16 St/B 9-48 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 90 P0.4 O0 / I St/B Bit 4 of Port 0, General Purpose Input/Output U1C1_SELO 0 O1 St/B USIC1 Channel 1 Select/Control 0 Output U1C0_SELO 1 O2 St/B USIC1 Channel 0 Select/Control 1 Output CCU61_COU O3 T61 St/B CCU61 Channel 1 Output A4 OH St/B External Bus Interface Address Line 4 U1C1_DX2A I St/B USIC1 Channel 1 Shift Control Input RxDC1B I St/B CAN Node 1 Receive Data Input ESR2_8 I St/B ESR2 Trigger Input 8 P2.13 O0 / I St/B Bit 13 of Port 2, General Purpose Input/Output U2C1_SELO 2 O1 USIC2 Channel 1 Select/Control 2 Output P3.2 O0 / I St/B Bit 2 of Port 3, General Purpose Input/Output U2C0_SCLK OUT O1 St/B USIC2 Channel 0 Shift Clock Output U2C0_DX1B I St/B USIC2 Channel 0 Shift Clock Input HOLD I St/B External Bus Master Hold Request Input P2.10 O0 / I St/B 92 93 94 Type Function St/B Bit 10 of Port 2, General Purpose Input/Output U0C1_DOUT O1 St/B USIC0 Channel 1 Shift Data Output U0C0_SELO 3 O2 St/B USIC0 Channel 0 Select/Control 3 Output CC2_CC23 O3 / I St/B CAPCOM2 CC23IO Capture Inp./ Compare Out. A23 OH St/B External Bus Interface Address Line 23 U0C1_DX0E I St/B USIC0 Channel 1 Shift Data Input CAPINA I St/B GPT12E Register CAPREL Capture Input User’s Manual Parallel Ports, V1.16 9-49 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 95 P10.3 O0 / I St/B CCU60_COU O2 T60 96 97 98 Type Function St/B Bit 3 of Port 10, General Purpose Input/Output CCU60 Channel 0 Output AD3 OH / I St/B External Bus Interface Address/Data Line 3 U0C0_DX2A I St/B USIC0 Channel 0 Shift Control Input U0C1_DX2A I St/B USIC0 Channel 1 Shift Control Input P0.5 O0 / I St/B Bit 5 of Port 0, General Purpose Input/Output U1C1_SCLK OUT O1 St/B USIC1 Channel 1 Shift Clock Output U1C0_SELO 2 O2 St/B USIC1 Channel 0 Select/Control 2 Output CCU61_COU O3 T62 St/B CCU61 Channel 2 Output A5 OH St/B External Bus Interface Address Line 5 U1C1_DX1A I St/B USIC1 Channel 1 Shift Clock Input U1C0_DX1C I St/B USIC1 Channel 0 Shift Clock Input P3.3 O0 / I St/B Bit 3 of Port 3, General Purpose Input/Output U2C0_SELO 0 O1 St/B USIC2 Channel 0 Select/Control 0 Output U2C1_SELO 1 O2 St/B USIC2 Channel 1 Select/Control 1 Output U2C0_DX2A I St/B USIC2 Channel 0 Shift Control Input P10.4 O0 / I St/B Bit 4 of Port 10, General Purpose Input/Output U0C0_SELO 3 O1 St/B USIC0 Channel 0 Select/Control 3 Output CCU60_COU O2 T61 St/B CCU60 Channel 1 Output AD4 OH / I St/B External Bus Interface Address/Data Line 4 U0C0_DX2B I St/B USIC0 Channel 0 Shift Control Input U0C1_DX2B I St/B USIC0 Channel 1 Shift Control Input ESR1_9 I St/B ESR1 Trigger Input 9 User’s Manual Parallel Ports, V1.16 9-50 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 99 P3.4 O0 / I St/B Bit 4 of Port 3, General Purpose Input/Output U2C1_SELO 0 O1 St/B USIC2 Channel 1 Select/Control 0 Output U2C0_SELO 1 O2 St/B USIC2 Channel 0 Select/Control 1 Output U0C0_SELO 4 O3 St/B USIC0 Channel 0 Select/Control 4 Output U2C1_DX2A I St/B USIC2 Channel 1 Shift Control Input P10.5 O0 / I St/B Bit 5 of Port 10, General Purpose Input/Output U0C1_SCLK OUT O1 St/B USIC0 Channel 1 Shift Clock Output CCU60_COU O2 T62 St/B CCU60 Channel 2 Output U2C0_DOUT O3 St/B USIC2 Channel 0 Shift Data Output 100 101 Type Function AD5 OH / I St/B External Bus Interface Address/Data Line 5 U0C1_DX1B I USIC0 Channel 1 Shift Clock Input P3.5 O0 / I St/B Bit 5 of Port 3, General Purpose Input/Output U2C1_SCLK OUT O1 St/B USIC2 Channel 1 Shift Clock Output U2C0_SELO 2 O2 St/B USIC2 Channel 0 Select/Control 2 Output U0C0_SELO 5 O3 St/B USIC0 Channel 0 Select/Control 5 Output U2C1_DX1A I St/B USIC2 Channel 1 Shift Clock Input User’s Manual Parallel Ports, V1.16 St/B 9-51 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 102 P0.6 O0 / I St/B 103 104 Type Function Bit 6 of Port 0, General Purpose Input/Output U1C1_DOUT O1 St/B USIC1 Channel 1 Shift Data Output TxDC1 O2 St/B CAN Node 1 Transmit Data Output CCU61_COU O3 T63 St/B CCU61 Channel 3 Output A6 OH St/B External Bus Interface Address Line 6 U1C1_DX0A I St/B USIC1 Channel 1 Shift Data Input CCU61_CTR APA I St/B CCU61 Emergency Trap Input U1C1_DX1B I St/B USIC1 Channel 1 Shift Clock Input P10.6 O0 / I St/B Bit 6 of Port 10, General Purpose Input/Output U0C0_DOUT O1 St/B USIC0 Channel 0 Shift Data Output U1C0_SELO 0 O3 St/B USIC1 Channel 0 Select/Control 0 Output AD6 OH / I St/B External Bus Interface Address/Data Line 6 U0C0_DX0C I St/B USIC0 Channel 0 Shift Data Input U1C0_DX2D I St/B USIC1 Channel 0 Shift Control Input CCU60_CTR APA I St/B CCU60 Emergency Trap Input P3.6 O0 / I St/B Bit 6 of Port 3, General Purpose Input/Output U2C1_DOUT O1 St/B USIC2 Channel 1 Shift Data Output U0C0_SELO 6 O3 St/B USIC0 Channel 0 Select/Control 6 Output U2C1_DX0A I St/B USIC2 Channel 1 Shift Data Input U2C1_DX1B I St/B USIC2 Channel 1 Shift Clock Input User’s Manual Parallel Ports, V1.16 9-52 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 105 P10.7 O0 / I St/B 106 107 Type Function Bit 7 of Port 10, General Purpose Input/Output U0C1_DOUT O1 St/B USIC0 Channel 1 Shift Data Output CCU60_COU O2 T63 St/B CCU60 Channel 3 Output AD7 OH / I St/B External Bus Interface Address/Data Line 7 U0C1_DX0B I St/B USIC0 Channel 1 Shift Data Input CCU60_CCP I OS0A St/B CCU60 Position Input 0 T4INB I St/B GPT12E Timer T4 Count/Gate Input P0.7 O0 / I St/B Bit 7 of Port 0, General Purpose Input/Output U1C1_DOUT O1 St/B USIC1 Channel 1 Shift Data Output U1C0_SELO 3 O2 St/B USIC1 Channel 0 Select/Control 3 Output A7 OH St/B External Bus Interface Address Line 7 U1C1_DX0B I St/B USIC1 Channel 1 Shift Data Input CCU61_CTR APB I St/B CCU61 Emergency Trap Input P3.7 O0 / I St/B Bit 7 of Port 3, General Purpose Input/Output U2C1_DOUT O1 St/B USIC2 Channel 1 Shift Data Output U2C0_SELO 3 O2 St/B USIC2 Channel 0 Select/Control 3 Output U0C0_SELO 7 O3 St/B USIC0 Channel 0 Select/Control 7 Output U2C1_DX0B I St/B USIC2 Channel 1 Shift Data Input User’s Manual Parallel Ports, V1.16 9-53 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 111 P1.0 O0 / I St/B Bit 0 of Port 1, General Purpose Input/Output U1C0_MCLK OUT O1 St/B USIC1 Channel 0 Master Clock Output U1C0_SELO 4 O2 St/B USIC1 Channel 0 Select/Control 4 Output A8 OH St/B External Bus Interface Address Line 8 ESR1_3 I St/B ESR1 Trigger Input 3 CCU62_CTR APB I St/B CCU62 Emergency Trap Input T6INB I St/B GPT12E Timer T6 Count/Gate Input P9.0 O0 / I St/B Bit 0 of Port 9, General Purpose Input/Output CCU63_CC6 0 O1 St/B CCU63 Channel 0 Output CCU63_CC6 0INA I St/B CCU63 Channel 0 Input T6EUDB I St/B GPT12E Timer T6 External Up/Down Control Input P10.8 O0 / I St/B Bit 8 of Port 10, General Purpose Input/Output U0C0_MCLK OUT O1 St/B USIC0 Channel 0 Master Clock Output U0C1_SELO 0 O2 St/B USIC0 Channel 1 Select/Control 0 Output U2C1_DOUT O3 St/B USIC2 Channel 1 Shift Data Output 112 113 AD8 Type Function OH / I St/B External Bus Interface Address/Data Line 8 CCU60_CCP I OS1A St/B CCU60 Position Input 1 U0C0_DX1C I St/B USIC0 Channel 0 Shift Clock Input BRKIN_B I St/B OCDS Break Signal Input T3EUDB I St/B GPT12E Timer T3 External Up/Down Control Input User’s Manual Parallel Ports, V1.16 9-54 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 114 P9.1 O0 / I St/B Bit 1 of Port 9, General Purpose Input/Output CCU63_CC6 1 O1 St/B CCU63 Channel 1 Output CCU63_CC6 1INA I St/B CCU63 Channel 1 Input P10.9 O0 / I St/B Bit 9 of Port 10, General Purpose Input/Output U0C0_SELO 4 O1 St/B USIC0 Channel 0 Select/Control 4 Output U0C1_MCLK OUT O2 St/B USIC0 Channel 1 Master Clock Output AD9 OH / I St/B 115 116 117 Type Function External Bus Interface Address/Data Line 9 CCU60_CCP I OS2A St/B CCU60 Position Input 2 TCK_B I St/B DAP0/JTAG Clock Input T3INB I St/B GPT12E Timer T3 Count/Gate Input P1.1 O0 / I St/B Bit 1 of Port 1, General Purpose Input/Output CCU62_COU O1 T62 St/B CCU62 Channel 2 Output U1C0_SELO 5 O2 St/B USIC1 Channel 0 Select/Control 5 Output U2C1_DOUT O3 St/B USIC2 Channel 1 Shift Data Output A9 OH St/B External Bus Interface Address Line 9 ESR2_3 I St/B ESR2 Trigger Input 3 U2C1_DX0C I St/B USIC2 Channel 1 Shift Data Input P10.10 O0 / I St/B Bit 10 of Port 10, General Purpose Input/Output U0C0_SELO 0 O1 St/B USIC0 Channel 0 Select/Control 0 Output CCU60_COU O2 T63 St/B CCU60 Channel 3 Output AD10 OH / I St/B External Bus Interface Address/Data Line 10 U0C0_DX2C I St/B USIC0 Channel 0 Shift Control Input TDI_B I St/B JTAG Test Data Input U0C1_DX1A I St/B USIC0 Channel 1 Shift Clock Input User’s Manual Parallel Ports, V1.16 9-55 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 118 P10.11 O0 / I St/B Bit 11 of Port 10, General Purpose Input/Output U1C0_SCLK OUT O1 St/B USIC1 Channel 0 Shift Clock Output BRKOUT O2 St/B OCDS Break Signal Output AD11 OH / I St/B External Bus Interface Address/Data Line 11 U1C0_DX1D I St/B USIC1 Channel 0 Shift Clock Input TMS_B I St/B JTAG Test Mode Selection Input P9.2 O0 / I St/B Bit 2 of Port 9, General Purpose Input/Output CCU63_CC6 2 O1 St/B CCU63 Channel 2 Output CCU63_CC6 2INA I St/B CCU63 Channel 2 Input CAPINB I St/B GPT12E Register CAPREL Capture Input P1.2 O0 / I St/B Bit 2 of Port 1, General Purpose Input/Output CCU62_CC6 2 O1 St/B CCU62 Channel 2 Output U1C0_SELO 6 O2 St/B USIC1 Channel 0 Select/Control 6 Output U2C1_SCLK OUT O3 St/B USIC2 Channel 1 Shift Clock Output A10 OH St/B External Bus Interface Address Line 10 ESR1_4 I St/B ESR1 Trigger Input 4 CCU61_T12 HRB I St/B External Run Control Input for T12 of CCU61 CCU62_CC6 2INA I St/B CCU62 Channel 2 Input U2C1_DX0D I St/B USIC2 Channel 1 Shift Data Input U2C1_DX1C I St/B USIC2 Channel 1 Shift Clock Input 119 120 User’s Manual Parallel Ports, V1.16 Type Function 9-56 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 121 P10.12 O0 / I St/B 122 123 124 Type Function Bit 12 of Port 10, General Purpose Input/Output U1C0_DOUT O1 St/B USIC1 Channel 0 Shift Data Output TDO_B OH St/B DAP1/JTAG Test Data Output AD12 OH / I St/B External Bus Interface Address/Data Line 12 U1C0_DX0C I St/B USIC1 Channel 0 Shift Data Input U1C0_DX1E I St/B USIC1 Channel 0 Shift Clock Input P9.3 O0 / I St/B Bit 3 of Port 9, General Purpose Input/Output CCU63_COU O1 T60 St/B CCU63 Channel 0 Output BRKOUT O2 St/B OCDS Break Signal Output P10.13 O0 / I St/B Bit 13 of Port 10, General Purpose Input/Output U1C0_DOUT O1 St/B USIC1 Channel 0 Shift Data Output U1C0_SELO 3 O3 St/B USIC1 Channel 0 Select/Control 3 Output WR/WRL OH St/B External Bus Interface Write Strobe Output Active for each external write access, when WR, active for ext. writes to the low byte, when WRL. U1C0_DX0D I St/B USIC1 Channel 0 Shift Data Input P1.3 O0 / I St/B Bit 3 of Port 1, General Purpose Input/Output CCU62_COU O1 T63 St/B CCU62 Channel 3 Output U1C0_SELO 7 O2 St/B USIC1 Channel 0 Select/Control 7 Output U2C0_SELO 4 O3 St/B USIC2 Channel 0 Select/Control 4 Output A11 OH St/B External Bus Interface Address Line 11 ESR2_4 I St/B ESR2 Trigger Input 4 CCU62_T12 HRB I St/B External Run Control Input for T12 of CCU62 User’s Manual Parallel Ports, V1.16 9-57 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 125 P9.4 O0 / I St/B 126 128 129 Type Function Bit 4 of Port 9, General Purpose Input/Output CCU63_COU O1 T61 St/B CCU63 Channel 1 Output U2C0_DOUT O2 St/B USIC2 Channel 0 Shift Data Output CCU62_COU O3 T63 St/B CCU62 Channel 3 Output P9.5 O0 / I St/B Bit 5 of Port 9, General Purpose Input/Output CCU63_COU O1 T62 St/B CCU63 Channel 2 Output U2C0_DOUT O2 St/B USIC2 Channel 0 Shift Data Output CCU62_COU O3 T62 St/B CCU62 Channel 2 Output U2C0_DX0E I St/B USIC2 Channel 0 Shift Data Input CCU60_CCP I OS2B St/B CCU60 Position Input 2 P10.14 O0 / I St/B Bit 14 of Port 10, General Purpose Input/Output U1C0_SELO 1 O1 St/B USIC1 Channel 0 Select/Control 1 Output U0C1_DOUT O2 St/B USIC0 Channel 1 Shift Data Output RD OH St/B External Bus Interface Read Strobe Output ESR2_2 I St/B ESR2 Trigger Input 2 U0C1_DX0C I St/B USIC0 Channel 1 Shift Data Input P1.4 O0 / I St/B Bit 4 of Port 1, General Purpose Input/Output CCU62_COU O1 T61 St/B CCU62 Channel 1 Output U1C1_SELO 4 O2 St/B USIC1 Channel 1 Select/Control 4 Output U2C0_SELO 5 O3 St/B USIC2 Channel 0 Select/Control 5 Output A12 OH St/B External Bus Interface Address Line 12 U2C0_DX2B I St/B USIC2 Channel 0 Shift Control Input User’s Manual Parallel Ports, V1.16 9-58 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 130 P10.15 O0 / I St/B Bit 15 of Port 10, General Purpose Input/Output U1C0_SELO 2 O1 St/B USIC1 Channel 0 Select/Control 2 Output U0C1_DOUT O2 St/B USIC0 Channel 1 Shift Data Output U1C0_DOUT O3 St/B USIC1 Channel 0 Shift Data Output ALE OH St/B External Bus Interf. Addr. Latch Enable Output U0C1_DX1C I St/B USIC0 Channel 1 Shift Clock Input P1.5 O0 / I St/B 131 132 Type Function Bit 5 of Port 1, General Purpose Input/Output CCU62_COU O1 T60 St/B CCU62 Channel 0 Output U1C1_SELO 3 O2 St/B USIC1 Channel 1 Select/Control 3 Output BRKOUT O3 St/B OCDS Break Signal Output A13 OH St/B External Bus Interface Address Line 13 U2C0_DX0C I St/B USIC2 Channel 0 Shift Data Input P9.6 O0 / I St/B Bit 6 of Port 9, General Purpose Input/Output CCU63_COU O1 T63 St/B CCU63 Channel 3 Output CCU63_COU O2 T62 St/B CCU63 Channel 2 Output CCU62_COU O3 T61 St/B CCU62 Channel 1 Output CCU63 _CTRAPA I St/B CCU63 Emergency Trap Input CCU60_CCP I OS1B St/B CCU60 Position Input 1 User’s Manual Parallel Ports, V1.16 9-59 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 133 P1.6 O0 / I St/B Bit 6 of Port 1, General Purpose Input/Output CCU62_CC6 1 O1 / I St/B CCU62 Channel 1 Output U1C1_SELO 2 O2 St/B USIC1 Channel 1 Select/Control 2 Output U2C0_DOUT O3 St/B USIC2 Channel 0 Shift Data Output A14 OH St/B External Bus Interface Address Line 14 U2C0_DX0D I St/B USIC2 Channel 0 Shift Data Input CCU62_CC6 1INA I St/B CCU62 Channel 1 Input P9.7 O0 / I St/B 134 135 136 Type Function Bit 7 of Port 9, General Purpose Input/Output CCU62_COU O1 T60 St/B CCU62 Channel 0 Output CCU62_COU O2 T63 St/B CCU62 Channel 3 Output CCU63_CTR APB I St/B CCU63 Emergency Trap Input U2C0_DX1D I St/B USIC2 Channel 0 Shift Clock Input CCU60_CCP I OS0B St/B CCU60 Position Input 0 P1.7 O0 / I St/B Bit 7 of Port 1, General Purpose Input/Output CCU62_CC6 0 O1 St/B CCU62 Channel 0 Output U1C1_MCLK OUT O2 St/B USIC1 Channel 1 Master Clock Output U2C0_SCLK OUT O3 St/B USIC2 Channel 0 Shift Clock Output A15 OH St/B External Bus Interface Address Line 15 U2C0_DX1C I St/B USIC2 Channel 0 Shift Clock Input CCU62_CC6 0INA I St/B CCU62 Channel 0 Input XTAL2 O Sp/1 Crystal Oscillator Amplifier Output User’s Manual Parallel Ports, V1.16 9-60 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. Type Function 137 XTAL1 I Sp/1 Crystal Oscillator Amplifier Input To clock the device from an external source, drive XTAL1, while leaving XTAL2 unconnected. Voltages on XTAL1 must comply to the core supply voltage VDDI1. ESR2_9 I St/B ESR2 Trigger Input 9 138 PORST I In/B Power On Reset Input A low level at this pin resets the XC27x5X completely. A spike filter suppresses input pulses <10 ns. Input pulses >100 ns safely pass the filter. The minimum duration for a safe recognition should be 120 ns. An internal pullup device will hold this pin high when nothing is driving it. 139 ESR1 O0 / I St/B External Service Request 1 RxDC0E I St/B CAN Node 0 Receive Data Input U1C0_DX0F I St/B USIC1 Channel 0 Shift Data Input U1C0_DX2C I St/B USIC1 Channel 0 Shift Control Input U1C1_DX0C I St/B USIC1 Channel 1 Shift Data Input U1C1_DX2B I St/B USIC1 Channel 1 Shift Control Input U2C1_DX2C I St/B USIC2 Channel 1 Shift Control Input User’s Manual Parallel Ports, V1.16 9-61 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. 140 ESR2 O0 / I St/B External Service Request 2 RxDC1E I St/B CAN Node 1 Receive Data Input CCU60_CTR APC I St/B CCU60 Emergency Trap Input CCU61_CTR APC I St/B CCU61 Emergency Trap Input CCU62_CTR APC I St/B CCU62 Emergency Trap Input CCU63_CTR APC I St/B CCU63 Emergency Trap Input U1C1_DX0D I St/B USIC1 Channel 1 Shift Data Input U1C1_DX2C I St/B USIC1 Channel 1 Shift Control Input U2C1_DX0E I St/B USIC2 Channel 1 Shift Data Input U2C1_DX2B I St/B USIC2 Channel 1 Shift Control Input ESR0 O0 / I St/B 141 Type Function External Service Request 0 Note: After power-up, ESR0 operates as opendrain bidirectional reset with a weak pull-up. 142 U1C0_DX0E I St/B USIC1 Channel 0 Shift Data Input U1C0_DX2B I St/B USIC1 Channel 0 Shift Control Input P8.6 O0 / I St/B Bit 6 of Port 8, General Purpose Input/Output CCU60_COU O1 T63 St/B CCU60 Channel 3 Output MCHK_MAT CH O3 St/B Memory Checker Match Output CCU60_CTR APB I St/B CCU60 Emergency Trap Input BRKIN_D I St/B OCDS Break Signal Input CCU62_CTR APD I St/B CCU62 Emergency Trap Input User’s Manual Parallel Ports, V1.16 9-62 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Parallel Ports Table 9-18 Pin Definitions and Functions (cont’d) Pin Symbol Ctrl. Type Function 143 P8.5 O0 / I St/B Bit 5 of Port 8, General Purpose Input/Output CCU60_COU O1 T62 St/B CCU60 Channel 2 Output CCU62_CC6 2 O2 St/B CCU62 Channel 2 Output TCK_D I St/B DAP0/JTAG Clock Input CCU62_CC6 2INB I St/B CCU62 Channel 2 Input 15 VDDIM - PS/M Digital Core Supply Voltage for Domain M Decouple with a ceramic capacitor, see Data Sheet for details. 54, 91, 127 VDDI1 - PS/1 Digital Core Supply Voltage for Domain 1 Decouple with a ceramic capacitor, see Data Sheet for details. All VDDI1 pins must be connected to each other. 20 VDDPA - PS/A Digital Pad Supply Voltage for Domain A Connect decoupling capacitors to adjacent VDDP/ VSS pin pairs as close as possible to the pins. Note: The A/D_Converters and ports P5, P6 and P15 are fed from supply voltage VDDPA. 2, VDDPB 36, 38, 72, 74, 108, 110, 144 1, 37, 73, 109 1) VSS - PS/B Digital Pad Supply Voltage for Domain B Connect decoupling capacitors to adjacent VDDP/ VSS pin pairs as close as possible to the pins. Note: The on-chip voltage regulators and all ports except P5, P6 and P15 are fed from supply voltage VDDPB. - PS/-- Digital Ground All VSS pins must be connected to the ground-line or ground-plane. To generate the reference clock output for bus timing measurement, fSYS must be selected as source for EXTCLK and P2.8 must be selected as output pin. Also the high-speed clock pad must be enabled. This configuration is referred to as reference clock output signal CLKOUT. User’s Manual Parallel Ports, V1.16 9-63 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Dedicated Pins 10 Dedicated Pins Most of the input/output or control signals of the functional the XC27x5X 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 10-1 summarizes the dedicated pins of the XC27x5X. Table 10-1 XC27x5X Dedicated Pins Pin(s) Function PORST Power-On Reset Input ESR0 External Service Request Input 0 ESR1 External Service Request Input 1 ESR2 External Service Request Input 2 XTAL1, XTAL2 Oscillator Input/Output (main oscillator) TESTM Test Mode Enable TRST Test-System Reset Input VAREFx, VAGND VDDIM VDDI1 VDDPA VDDPB VSS Reference voltages for the Analog/Digital Converter(s) Digital Core Supply for Domain M (1 pin) Digital Core Supply for Domain 1 (3 pins) Digital Pad Supply for Domain A including ADCs (1 pin) Digital Pad Supply for Domain B (8 pins) Digital Ground (4 pins) The Power-On Reset Input PORST allows to put the XC27x5X into the well defined reset condition either at power-up or external events like a hardware failure or manual reset. The External Service Request Inputs ESR0, ESR1, and ESR2 can be used for several system-related functions: • trigger interrupt or trap (Class A or Class B) requests via an external signal (e.g. a power-fail signal) • generate wake-up request signals • generate hardware reset requests (ESR0 is bidirectional by default, ESR1 and ESR2 can optionally output a reset signal) • data/control input for CCU6x, MultiCAN, and USIC (ESR1 or ESR2) • software-controlled input/output signal User’s Manual DediPins_X7, V2.0 10-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Dedicated Pins 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 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 a high-precision operating clock signal for the XC27x5X. An external clock signal may be fed to the input XTAL1, leaving XTAL2 open. The current logic state of input XTAL1 can be read via a status flag, so XTAL1 can be used as digital input if neither the oscillator interface nor the clock input is required. Note: Pin XTAL1 belongs to the core power domain DMP_M. All input signals, therefore, must be within the core voltage range. The Test Mode Input TESTM puts the XC27x5X into a test mode, which is used during the production tests of the device. In test mode, the XC27x5X behaves different from normal operation. Therefore, pin TESTM must be held HIGH (connect to VDDPB) for normal operation in an application system. The Test Reset Input TRST puts the XC27x5X’s debug system into reset state. During normal operation this input should be held low. For debugging purposes the on-chip debugging system can be enabled by driving pin TRST high at the rising edge of PORST. The Analog Reference Voltage Supply pins VAREFx and VAGND provide separate reference voltage for the on-chip Analog/Digital-Converter(s). 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. Also, because conversion results are generated in relation to the reference voltages, ratiometric conversions are easily achieved. Note: Channel 0 of each module can be used as an alternate reference voltage input. The Core Supply pins VDDIM/VDDI1 serve two purposes: While the on-chip EVVRs provide the power for the core logic of the XC27x5X these pins connect the EVVRs to their external buffer capacitors. For external supply, the core voltage is applied to these pins. The respective VDDI/VSS pairs should be decoupled as close to the pins as possible. Use ceramic capacitors and observe their values recommended in the respective Data Sheet. The Power Supply pins VDDPA/VDDPB provide the power supply for all the analog and digital logic of the XC27x5X. Each power domain (DMP_A and DMP_B) can be supplied with an arbitrary voltage within the specified supply voltage range (please refer to the corresponding Data Sheets). These pins supply the output drivers as well as the on-chip EVVRs (VDDPB), except for external core voltage supply. The respective VDDP/VSS pairs should be decoupled as close to the pins as possible. User’s Manual DediPins_X7, V2.0 10-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Dedicated Pins The Ground Reference pins VSS provide the ground reference voltage for the power supplies as well as the reference voltage for the input signals. Note: All VDDx pins and all VSS pins must be connected to the power supplies and ground, respectively. User’s Manual DediPins_X7, V2.0 10-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11 External Bus Controller (EBC) The EBC enables the C166SV2 CPU access to chip external and internal peripherals and memories. The access can be of program fetch type or data exchange. If used to interface to the chip external world the external bus is also referred to as EXTBUS if used with internal (local) components it is also referred to as LXBUS. 11.1 Summary of Features The EBC functional and timing behavior is widely configurable so that it can be tailored to fit into a large range of applications. • • • • • • • • Demultiplexed and multiplexed mode of operation Up to 24 address lines 8-bit or 16-bit data bus Synchronous and asynchronous ready External bus arbitration support Up to 8 bus channels Address window programmable for up to 7 channels Bus timing and function programmable for each channel 11.2 Overview The function and timing characteristics of the EBC are controlled by a set of configuration registers. The basic and general behavior is programmed via the mode selection registers EBCMOD0 and EBCMOD1. The EBC supports up to eight (x=0…7) bus channels linked to a dedicated chip select (CSx). Each channel is programmable by a dedicated set of registers. The FCONCSx registers specify the functional characteristics while the TCONCSx registers specify the cycle timing. The address area assigned to a channel is definable for seven channels by the ADDRSEL(1…7) registers. The remaining uncovered address areas are assigned to CS0. External CSx signals can be used in order to save external glue logic. Access to non timing deterministic external devices is supported by a particular READY functionality. A HOLD/HLDA protocol is available for bus arbitration. The external bus timing is related to the reference CLocK OUTput (CLKOUT) signal. All bus signals are generated in relation to the rising edge of this clock. The external bus protocol is compatible with those of the C166 family. However, the external bus timing is improved in terms of wait state granularity and signal flexibility. User’s Manual C166S-V2 EBC, V2.7 11-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.3 Naming Conventions For description of EBC timing and functions the following bus signal names will be used. Control signals: • • • • • • • • • ALE - Address Latch Enable (high active). Indicates that the applied address is valid. CS - Chip select. WR/WRL - Write Strobe, Write Low Byte Strobe (low active). Configured either to a general write request or a write request for the low byte. BHE/WRH - Byte High Enable, Write High Byte Strobe (low active). Configured either to an enable for the high byte or a write request for the high byte. RD - Read Strobe (low active). READY - Ready to indicate end of actions (programmable polarity). HOLD - Hold input for foreign bus requests (low active). HLDA - Hold Acknowledge (low active), master output to grant bus. BREQ - Bus Request (low active). Bus signals: • • • A or ADDR - Address bus. D or DATA - Data bus. AD - Shared Data/Address[15:0] bus. 11.4 Timing Description Bus characteristics can be programmed to the following access modes: • • • • 16-bit data bus with address not multiplexed 16-bit data bus with address multiplexed 8-bit data bus with address not multiplexed 8-bit data bus with address multiplexed Multiplexed mode means that the data bus is used in a time multiplex mode for address (16 LSB only) and for data. In demultiplexed mode the data bus is used for data only and an additional address bus must be made available. 11.4.1 Bus Phases The external bus timing is defined by six different timing phases (A-F). These phases influence the control signals needed for any access sequence to a bus device. At the beginning of a phase the output signals change within a defined output delay time. The particular delay times are specified in the XC27x5X data book. After the output delay the values of the control output signals are stable within this phase. Each phase can occupy a programmable number of CLKOUT cycles. User’s Manual C166S-V2 EBC, V2.7 11-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 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. 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. 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). 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. 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. 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 User’s Manual C166S-V2 EBC, V2.7 11-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) accesses. The F phase is used to program tristate wait states on the bidirectional data bus in order to avoid bus conflicts. User’s Manual C166S-V2 EBC, V2.7 11-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.4.2 Demultiplexed Bus General timing diagrams of a read and a write demultiplexed access are shown below. 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 11-1 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 11-2 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 C166S-V2 EBC, V2.7 11-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.4.3 Multiplexed Bus General timing diagrams of a read and a write multiplexed access are shown below. 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 11-3 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 11-4 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 User’s Manual C166S-V2 EBC, V2.7 11-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) • F phase: command inactive, address hold. Read data tristate time, write data hold time. 11.4.4 Fastest Access Cycles The fastest possible bus cycle in a system depends also on the pad timing. Therefore, the number of required cycles for a bus access depends on the current system frequency. The minimum bus cycles shown below cannot be achieved at very high system frequencies. b e CLK ALE ADDR, CS Valid RD DATA In Valid MCT05378 Figure 11-5 Fastest Read Cycle Demultiplexed Bus b e CLK ALE ADDR, CS Valid WR DATA Out Valid MCT05379 Figure 11-6 Fastest Write Cycle Demultiplexed Bus User’s Manual C166S-V2 EBC, V2.7 11-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) 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 11-7 Fastest Read Cycle Multiplexed Bus b e CLK ALE ADDR, CS Valid RD Muxed Address Out / Data Out Addr. Valid Valid MCT05381 Figure 11-8 Fastest Write Cycle Multiplexed Bus User’s Manual C166S-V2 EBC, V2.7 11-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.5 Address Windows The EBC provides up to 8 logical bus channels. For 7 of these (assigned to CS1…CS7) the allocated address window can be programmed by the ADDRSEL(1…7) registers. The remaining channel (assigned to CS0) has no address select register assigned and covers the remaining EBC address space. Note that not the complete XC27x5X address space can be accessed by the EBC. For an overview on allocation of the complete address space please refer to the “Memory Organization” chapter. 11.5.1 Window Allocation The size and start address of an address window is defined according to Table 11-1. The size of the window is chosen by ADDRSELx.RGSZ. Depending on the size the relevant bits of ADDRSELx.RGSAD (marked with ‘R’) are used to select the corresponding window start address. The lower bits of ADDRSELx.RGSAD (marked ‘x’) are disregarded. A programmed address windows must additionally be enabled by setting the corresponding FCONCSx.ENCS bit. Table 11-1 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 reserved 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 User’s Manual C166S-V2 EBC, V2.7 RRRR RRRx RRxx Rxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx xxxx 11-9 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 ---- V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.5.2 Window Overlap Since it is possible to program overlapping address areas an arbitratration scheme is defined to handle these cases. For each access directed to the EBC it compares the current address with all address select registers of enabled windows. This comparison is done in three levels. Priority 1: Registers ADDRSELx [x = 2, 4, 6] of enabled windows are evaluated first. Upon a match the access starts on the respective channel. Overlap of windows within this group will lead to undefined behavior. Priority 2: Registers ADDRSELy [y = 1, 3, 5, 7] of enabled windows are evaluated in this step. Upon a match the access starts on the respective channel. Overlap of windows within this group will lead to undefined behavior. Overlaps with priority 1 ADDRSELx are only allowed for the (x,y) pairs (2,1), (4,3) and (6, 5). Priority 3: If channel 0 is enabled the access is directed to it. Otherwise no bus action occurs. User’s Manual C166S-V2 EBC, V2.7 11-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.6 Ready Controlled Bus Access In cases, where the response time of a peripheral is not constant, or where the programmable wait states are not enough, the XC27x5X EBC provides the so called READY controlled bus access scheme. In this scheme bus accesses are terminated by the READY input signal. 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 (“READY Int” in Figure 11-9) 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). A READY controlled bus cycle requires one synchronization cycle to terminate. Programmed phase F cycles include this synchronization cycle. Therefore setting TCONCSx phase F to 0 clock cycles will have the same effect as setting to 1 clock cycle. 11.6.1 Enabling the Ready Control The READY pin is enabled by setting the bit RDYDIS in EBCMOD0 to ’0’ in order to activate the corresponding port pin. The polarity of the READY is also defined inside the EBCMOD0 register by the RDYPOL bit. For a specific address window the READY function is enabled via the RDYEN bit in the FCONCSx register. By programming of FCONCSx.RDYMOD the READY is handled either in synchronous or in asynchronous mode (see also Figure 11-9). When the READY function is enabled for a specific address window, each bus cycle within this window must be cleanly terminated with an active READY signal. Otherwise the EBC will be completely blocked by this pending access. MUX READY Ext Async. 0 1 Sync. 1 EBCMOD0.RDYPOL MUX 0 1 READY Int FCONCSx.RDYMODx MCA05383 Figure 11-9 External to internal READY conversion 11.6.2 Synchronous and 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. User’s Manual C166S-V2 EBC, V2.7 11-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 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 should be deactivated in response to the trailing (rising) edge of the respective command (RD or WR). Bus Cycle with active READY 3 programmed phase E wait states(TCONCSx.PHE=2) Bus Cycle extended via READY 3 programmed phase E wait states(TCONCSx.PHE=2) ALE RD/WR sync. READY async. READY sampling of READY input not interesting READY cycles Figure 11-10 Ready controlled bus cycles 11.6.3 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 XC27x5X 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 XC27x5X, the controller samples an active READY and terminates the current bus cycle 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. User’s Manual C166S-V2 EBC, V2.7 11-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.7 External Bus Arbitration The XC27x5X 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 XC27x5X 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. Attention: This feature is only available if the memory protection unit (MPU) is disabled. 11.7.1 Initialization of Arbitration During reset all arbitration pins are tristate, except pin BREQ which is pulled inactive. After reset the XC27x5X 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 XC27x5X EBCs assume to have the bus also, so potential bus conflicts are not resolved. For a multi master 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. 11.7.2 Arbitration Master Scheme If the XC27x5X 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 C166S-V2 EBC, V2.7 11-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) 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 11-11 Releasing the Bus by the Arbitration Master Note: The figure above shows the first possibility for BREQ to get active. The XC27x5X will complete the currently running bus cycle before granting the external bus as indicated by the broken lines. User’s Manual C166S-V2 EBC, V2.7 11-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) 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 11-12 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 XC27x5X requesting the bus. 11.7.3 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 11.7.2. For the arbitration slave the hold acknowledge pin HLDA is configured as input. 11.7.4 Bus Lock Function If an application in a multi master 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 looked EBC will not answer to HOLD requests from other external bus master until HLDEN is set to ‘1’ again. Of course 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 C166S-V2 EBC, V2.7 11-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 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. 11.7.5 Direct Master Slave Connection If one XC27x5X 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 11-13. 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 11-13 Connecting two XC27x5X using Master/Slave Arbitration When multiple (more than two) bus masters (XC27x5X 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. User’s Manual C166S-V2 EBC, V2.7 11-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.8 EBC Idle State When the external bus interface of EBC is enabled, but no internal or external access is currently executed, the EBC is idle. As long as only on-chip resources such as RAM, peripherals (excluding LXBUS peripherals connected via EBC), registers, etc. are used, the external bus interface does not change (see table 11-2). The external control signals (RD and WR or WRL/WRH if enabled) remain inactive (high) during EBC idle state. Table 11-2 Status of the External Bus Outputs During EBC Idle State Pins Status of Pins During EBC Idle AD15…AD0 tristate (floating) A15…A0 undefined address (if used for the bus interface) A23…A16 undefined segment address (on selected pins) CS7…CS0 inactive (high) BHE level corresponding to last external access ALE inactive (low) RD, WR, WRL, WRH inactive (high) User’s Manual C166S-V2 EBC, V2.7 11-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.9 Register Description The EBC registers are located in the internal IO area. Registers located there use the shorthand XSFR. Table 11-3 Registers Address Space Module Base Address End Address EBC 00EE00H 00EEFFH 11.9.1 Note EBC Mode Registers The two mode registers control the XC27x5X EXTBUS pin usage and configuration. Disabled EXTBUS pins may be usable as general purpose IO as described in the “Parallel Ports” chapter. EBCMOD0 EBC Mode Register 0 15 14 13 12 XSFR(00H) 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 Reset value: 5000H 6 5 4 3 2 1 CSPEN SAPEN rw rw rw Field Bits Type Description RDYPOL 15 rw READY Pin Polarity 0B READY is active low 1B READY is active high RDYDIS 14 rw READY Pin Disable 0B READY enabled 1B READY disabled ALEDIS 13 rw ALE Pin Disable 0B ALE enabled 1B ALE disabled BYTDIS 12 rw BHE Pin Disable 0B BHE enabled 1B BHE disabled WRCFG1) 11 rw Configuration for Pins WR/WRL, BHE/WRH 0B WR and BHE 1B WRL and WRH User’s Manual C166S-V2 EBC, V2.7 11-18 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) Field Bits Type Description EBCDIS 10 rw EBC Pins Disable 0B EBC is using the pins for external bus 1B EBC pins disabled SLAVE 9 rw SLAVE Mode Enable Bus arbiter acts in master mode 0B Bus arbiter acts in slave mode 1B ARBEN 8 rw BUS Arbitration Pins Enable 0B HOLD, HLDA and BREQ pins are disabled 1B pins act as HOLD, HLDA and BREQ CSPEN [7:4] rw CSx Pins Enable (only external CSx) 0H All external Chip Select pins disabled. 1H CS0 pin enabled CS1 and CS0 pin enabled 2H … … 5H Five CSx pins enabled: CS4-CS0 other bit combinations not supported (reserved) SAPEN [3:0] rw Segment Address Pins Enable 0H All segment address pins disabled 1H One: A[16] enabled … … 8H Eight: A[23:16] enabled other bit combinations not supported (reserved) 1) A change of the bit content is not valid before the next external bus access cycle. Byte Write Configurations For 16-bit data bus configurations the byte write characteristics are programmable by bitfield WRCFG. The following table illustrates the related signals function. Table 11-4 Byte Write Configurations written byte WRCFG=0 WRCFG=1 low high WR BHE - - inactive don’t care 0/1 write - active inactive - write active write write active User’s Manual C166S-V2 EBC, V2.7 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 11-19 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) EBCMOD1 EBC Mode Register 1 15 14 13 12 XSFR(02H) 11 10 9 8 r r r 6 5 4 WRP DHP ALP A0P DIS DIS DIS DIS 0 r 7 Reset value: 003FH r r r r rw rw rw rw 3 2 1 APDIS rw Field Bits Type Description 0 [15:8] r Reserved Read as 0, should be written 0 WRPDIS 7 rw WR/WRL Pin Disable 0B WR/WRL pin enabled 1B WR/WRL pin disabled DHPDIS 6 rw Data High Port Pins Disable 0B Address/Data bus pins 15-8 enabled 1B Address/Data bus pins 15-8 disabled. ALPDIS 5 rw Address Low Pins Disable 0B Address bus pins 7-0 generally enabled (depending on APDIS/A0PDIS) 1B Address bus pins 7-0 disabled. A0PDIS 4 rw Address Bit 0 Pin Disable 0B Address bus pin 0 enabled 1B Address bus pin 0 disabled. APDIS [3:0] rw Address Port Pins Disable 0H Address bus pins A15-A1 enabled 1H Pin A15 disabled, A14-A1 enabled 2H Pins A15-A14 disabled, A13-A1 enabled 3H Pins A15-A13 disabled, A12-A1 enabled … … EH Pins A15-A2 disabled, A1 enabled FH Address bus pins 15-1 disabled. 11.9.2 0 Timing Control Registers The timing control registers are used to program the 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 become valid for the following access. User’s Manual C166S-V2 EBC, V2.7 11-20 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) TCONCS0 Timing Control for CS0 TCONCSx (x=1-4) Timing Control for CSx 13 12 11 10 XSFR(10H) Reset value: 7C3DH XSFR(10H+x*8) Reset value: 0000H 15 14 9 8 7 6 5 0 WRPHF RDPHF PHE PHD PHC PHB PHA r rw rw rw rw rw rw rw Field Bits Type Description 0 15 r WRPHF [14:13] rw Write Phase F 00B 0 clock cycles … … 11B 3 clock cycles RDPHF [12:11] rw Read Phase F 00B 0 clock cycles … … 11B 3 clock cycles PHE [10:6] rw Phase E 0H 1 clock cycle … … 1FH 32 clock cycles PHD 5 rw Phase D 0B 0 clock cycles 1B 1 clock cycle PHC [4:3] rw Phase C 00B 0 clock cycles … … 11B 3 clock cycles PHB 2 rw Phase B 0B 1 clock cycle 1B 2 clock cycles PHA [1:0] rw Phase A 00B 0 clock cycles … … 11B 3 clock cycles User’s Manual C166S-V2 EBC, V2.7 4 3 2 1 0 Reserved Read as 0, should be written 0 11-21 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.9.3 Function Control Registers 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 demultiplexed accesses. Furthermore it can be defined whether the address window (and its chip select signal CSx) is generally enabled or not. FCONCS0 Function Control for CS0 FCONCSx (x = 1-4) Function Control for CSx XSFR(12H) Reset value: 0011H XSFR(12H+x*8) Reset value: 0000H 15 14 13 12 11 10 9 8 7 6 0 0 0 0 0 0 0 0 0 r r r r r r r r r 5 4 3 0 BTYP 0 r rw r 2 1 0 RDY RDY EN MOD EN CS rw rw rw Field Bits Type Description 0 [15:6] r Reserved Read as 0, should be written 0 BTYP [5:4] rw Bus Type Selection 00B 8 bit Demultiplexed 01B 8 bit Multiplexed 10B 16 bit Demultiplexed 11B 16 bit Multiplexed 0 3 r Reserved Read as 0, should be written 0 RDYMOD 2 rw Ready Mode 0B asynchronous READY 1B synchronous READY RDYEN 1 rw Ready enable 0B access time is controlled by bitfield PHEx 1B access time is controlled by bitfield PHEx and READY signal ENCS 0 rw Enable Chip Select 0B disable 1B enable Note: The specific ENCSx bits in the FCONCSx registers enable the related address windows and bus functions and the corresponding chip select signal CSx. Additionally it depends on the definition of bitfield CSPEN in register EBCMOD0 User’s Manual C166S-V2 EBC, V2.7 11-22 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) how many CSx pins are made available for the external system. If an address window is enabled but no external pin is available for the CSx, the bus cycle is executed without assertion of the external chip select signal. 11.9.4 Address Window Selection Registers ADDRSELx (x = 1 - 4) Address Range/Size for CSx 15 14 13 12 11 XSFR(16H+x*8) 10 9 8 7 6 Reset value: 0000H 5 4 3 2 1 RGSAD RGSZ rw rw 0 Field Bits Type Description RGSAD [15:4] rw Range Start Address Address Range Start Address Selection RGSZ [3:0] rw Range Size Address Range Size Selection (see Table 11-1) User’s Manual C166S-V2 EBC, V2.7 11-23 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) 11.10 EBC Implementation in XC27x5X The EBC within the XC27x5X automatically affects the behaviour of other device components. In particular with memory mapping and ports allocation the EBC takes a priority role. To understand the particular relations to memory and ports the following chapters are recommended for additional reading: • • Memory Organization Parallel Ports 11.10.1 Unused Registers The XC27x5X external chip select signals are limited due to chip packaging. According to this limitation the corresponding EBC channels are not available. The channels x=(5, 6) with related set of registers ADDRSELx, TCONCSx, FCONCSx are therefore not available for use. For software compatibility reasons the corresponding register address space must not be read or written. 11.10.2 Access Control to LXBUS Modules In the XC27x5X the EBC channel 7 is reserved for access to chip internal LXBUS peripherals. In general accesses to LXBUS are not visible on the external bus EXTBUS. During LXBUS cycles the EXTBUS remains enabled but is driven to inactive states (control signals) or switched into the read mode (busses). 11.10.3 Shutdown Control In case of a shutdown request from the SCU the EBC ensures that all the different functions of the EBC are in a non-active state before the whole chip is switched to a power save 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. 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 11-5 gives an overview of the shutdown control in EBC depending on the EBC configuration. User’s Manual C166S-V2 EBC, V2.7 11-24 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) Table 11-5 EBC Shutdown Control Arbitration Mode Master Mode Bus control 11.10.4 Slave Mode With control of the bus Without control With control of the bus of the bus Without control of the bus 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. Ask for the bus if needed and finish all requests. Send shutdown acknowledge after leaving the bus Finish all pending requests. Send shutdown acknowledge after leaving the bus Dedicated Registers The dedicated EBC registers are located in the internal IO area. Registers located there use the shorthand XSFR. 11.10.4.1 Registers dedicated to LXBUS modules For accesses to MultiCAN and USICs CS7 and its control registers ADDRSEL7, TCONCS7 and FCONCS7 are used. The selection of LXBUS is chip internally controlled with CS7. TCONCS7 The LXBUS cycle timing is controlled with register TCONCS7. It uses the shortest possible timing with two clock cycles for one bus cycle. But this minimum timing will be lengthened with waitstate(s) controlled by the modules using the READY function. This timing is reflected by the reset value of TCONCS7. TCONCS7 Timing Control for CS7 13 12 10 9 8 7 Reset value: 0000H 15 14 0 WRPHF RDPHF PHE PHD PHC PHB PHA r r r r r r r r User’s Manual C166S-V2 EBC, V2.7 11 XSFR(48H) 11-25 6 5 4 3 2 1 0 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) Field Bits Type Description 0 15 r WRPHF [14:13] rw Write Phase F 00B 0 clock cycles … … 11B 3 clock cycles RDPHF [12:11] rw Read Phase F 00B 0 clock cycles … … 11B 3 clock cycles PHE [10:6] rw Phase E 0H 1 clock cycle … … 1FH 32 clock cycles PHD 5 rw Phase D 0B 0 clock cycles 1B 1 clock cycle PHC [4:3] rw Phase C 00B 0 clock cycles … … 11B 3 clock cycles PHB 2 rw Phase B 0B 1 clock cycle 1B 2 clock cycles PHA [1:0] rw Phase A 00B 0 clock cycles … … 11B 3 clock cycles Reserved Read as 0, should be written 0 FCONCS7 The value of this dedicated bus function control register is selected according to the requirements of the internally attached modules: 16-bit demultiplexed bus, access time controlled with synchronous READY. User’s Manual C166S-V2 EBC, V2.7 11-26 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) FCONCS7 Function Control for CS7 XSFR(4AH) Reset value: 0027H 15 14 13 12 11 10 9 8 7 6 0 0 0 0 0 0 0 0 0 r r r r r r r r r 5 4 3 0 BTYP 0 r r r 2 1 0 RDY RDY EN MOD EN CS r r r Field Bits Type Description 0 [15:6] r Reserved Read as 0, should be written 0 BTYP [5:4] rw Bus Type Selection 00B 8 bit Demultiplexed 01B 8 bit Multiplexed 10B 16 bit Demultiplexed 11B 16 bit Multiplexed 0 3 r Reserved Read as 0, should be written 0 RDYMOD 2 rw Ready Mode 0B asynchronous READY 1B synchronous READY RDYEN 1 rw Ready enable 0B access time is controlled by bitfield PHEx 1B access time is controlled by bitfield PHEx and READY signal ENCS 0 rw Enable Chip Select 0B disable 1B enable Register ADDRSEL7 The value of the dedicated address select register allocates the address range 20’0000H to 20’FFFFH in the external IO area for the attached chip internal modules. User’s Manual C166S-V2 EBC, V2.7 11-27 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) External Bus Controller (EBC) ADDRSEL7 Address Range/Size for CS7 15 14 13 12 11 XSFR(4EH) 10 9 8 7 Reset value: 2004H 6 5 4 3 2 1 RGSAD RGSZ r r 0 Field Bits Type Description RGSAD [15:4] rw Range Start Address Address Range Start Address Selection RGSZ [3:0] rw Range Size Address Range Size Selection (see Table 11-1) User’s Manual C166S-V2 EBC, V2.7 11-28 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12 Startup Configuration and Bootstrap Loading After startup, the XC27x5X executes code out of an on-chip or off-chip program memory. The initial code source can be selected (refer to the next Chapter 12.1 and Chapter 12.3to find out how) between the following options: • • • Internal Start Mode: executes code out of the on-chip program Flash. External Start Mode: executes code out of an off-chip memory connected to the External Bus Interface. Bootstrap Loading Modes: execute code out of the on-chip program SRAM (PSRAM). This code is downloaded beforehand via a selectable serial interface. 12.1 Startup Mode Selection After any startup the currently valid startup configuration is indicated in bitfield HWCFG of register SCU_STSTAT. The startup configurations and modes supported in XC27x5X are described in Chapter 12.3. A startup configuration can be selected basically in two ways: 1. Via an externally applied hardware configuration upon a Power-on reset The hardware configuration is applied to the dedicated TRST-pin and to Port 10 pins (P10[6:0]). The hardware that activates a startup configuration during reset may be simple pull resistors for systems that use this feature upon every reset. You may want to use a switchable solution (via jumpers or an external signal) for systems that only temporarily use a hardware configuration. 2. By executing the following software sequence (using SCU_SWRSTCON and SCU_RSTCON1 registers): a) Write respective configuration value (refer to Table 12-2 and Table 12-3) to bitfield SCU_SWRSTCON.SWCFG; b) Assign desired type of reset to the software request trigger by writing into SCU_RSTCON1.SW bitfield (by default SCU_RSTCON1.SW=00B meaning no reset generated by software request trigger) c) Set Software Boot Configuration bit: SCU_SWRSTCON.SWBOOT = 1; d) Trigger a software reset by activating Software Reset Request: SCU_SWRSTCON.SWRSTREQ = 1. Additionally, several specific cases of startup configuration handling must be noted: 1. Application reset triggered by hardware request (for example WDT, ESRx) TRST and P10 pins are not evaluated but the same startup configuration is used as after the previous reset; 2. Application or Internal Application reset triggered by software (with setting SCU_SWRSTCON.SWRSTREQ=1) can have different consequences: User’s Manual SCFG/BSL, PH2 V1.5d1 12-1 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading a) if Software Boot Configuration is selected (SCU_SWRSTCON.SWBOOT=1) - the startup configuration (SCU_STSTAT.HWCFG) is updated from SCU_SWRSTCON.SWCFG bitfield; b) otherwise - the same startup configuration is used as after the previous reset. 3. Internal Application reset triggered by hardware request only TRST pin is evaluated (refer to Page 12-7): a) if TRST=0 - Internal Start from Flash is selected, no debugging is possible; b) otherwise - the same startup configuration is used as after the previous reset. 12.2 Device Status after Startup The main parameters of XC27x5X-status at the point of time when the first user instruction is executed are summarized below. 12.2.1 Registers modified by the Startup Procedure Table 12-1 shows the XC27x5X registers which are initialized during the startup procedure with values different from their reset-content (defined into respective registerdescriptions). There are two groups of registers regarding the way they are affected by startup procedure: 1. registers initialized after any startup; 2. registers initialized after startup triggered by a power-on in DMP_1 power domain; Note: Power-on in DMP_M domain means power-on event also in DMP_1. The registers in Table 12-1 are grouped in accordance to the above differentiation. Two additional points regarding register-content after startup must be taken into account: • • The register-modifications shown in Table 12-1 happen independently on the startup mode currently selected, which means also in Internal Start mode. Next to these, in other modes - External Start and Bootstrap Loading (Chapter 12.5, Chapter 12.6) - more registers are additionally modified during startup, as described into respective Specific Settings chapters for any of the modes. The values seen in some bits after startup can be affected not only by the reset procedure itself but also by other events during and even before the last startup - for example an Emergency Event can change the clock-system status. Therefore occasional exceptions are possible from the above values (as well as from the default register content after reset), mainly for some clock control/status flags. For more information on such special cases and their handling - refer to XC27x5X Programmer’s Guide. User’s Manual SCFG/BSL, PH2 V1.5d1 12-2 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading Table 12-1 XC27x5X Registers installed by the Startup Procedure Register Value Comments 1. After any startup: TRAPDIS 019FH RSTCON1 UU: Internal Application Reset request generated by WDT 10uu:UH IMBCTRL 558CH In External or Bootstrap Loader mode with protected Flash A58CH In Internal Start mode or Flash not protected XXXX GPRs from Local Bank 1 - used by startup procedure R8..R15 All SCU-controlled traps disabled except PET and RAT 2. After power-on in DMP_1: PLLCON0 0F00H PLL in Normal Mode, N-divider = 16 PLLCON3 0007H K2-divider = 8 SYSCON0 0002H The PLL output (fPLL) used as system clock WUOSCCON 0000H Wake up Oscillator enabled with fWU approx. 500kHz HPOSCCON U:u0uu: PLLSTAT.FINDIS bit will not be set in an OSCWDT emergency case UUH PLLOSCCON XXXXH Device-specific value (chip-to-chip trimming) EVRMCON0 0100H EVR_M Control 0 register EVR1CON0 0D00H EVR_1 Control 0 register EVR1SET15VHP 401BH EVR_1 Setting for 1.5V HP register EVR1SET10V EVR_1 Setting for 1.0V register 405BH EVR1SET15VLP 40DBH EVR_1 Setting for 1.5V LP register PVCMCON0 2544H PVC_M Control for Step 0 register PVC1CON0 2544H PVC_1 Control for Step 0 register 12.2.2 System Frequency after Startup The system clock which is active when the first user instruction is executed, depends on the currently selected startup mode and the last startup trigger: • • after power-on in all modes except CAN Bootstrap Loader (Chapter 12.6.4) - 10MHz (nominal value) from the XC27x5X internal oscillator (doubled frequency); after power-on in CAN Bootstrap Loader (Chapter 12.6.4) - the frequency of an external crystal connected to XTAL-pins, 4MHz minimum; User’s Manual SCFG/BSL, PH2 V1.5d1 12-3 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading • after any functional (not power-on) reset - the clock system configuration is not changed by device startup, respectively the system frequency remains as before the reset. 12.2.3 Watchdog Timer handling The Watchdog Timer (WDT) in XC27x5X is always enabled by the startup procedure and configured to generate Internal Application Reset. Therefore, the user software must: • • if WDT-usage is foreseen by the code - service it for a first time within approx. 65500 system clock cycles after startup; otherwise - disable it within the same time frame as above but before to execute End of Init (EINIT). The reset requested by WDT serves as response to a device malfunction, due to which malfunction user software can not be anymore executed correctly - respectively the WDT is not regularly served. This reset causes a new device startup followed by user software restart. The Internal Application Reset - default for WDT - affects not all the modules in XC27x5X - refer to “Module Reset Behavior” in SCU Chapter. The unaffected modules do not change their state, so if this state is “wrong” it will not be recovered to “correct” due to Internal Application Reset, also when triggered by WDT. Therefore, the default WDT configuration and usage can resolve well purely software malfunction but not other failures - for example in clock- or power- system. One reset-request which puts all the XC27x5X modules into a known - and correctly functioning - initial state is from PORST-pin. Therefore, if an application requires that correct device restart upon WDT reset is guaranteed, the implementation must be done according to the next Chapter 12.2.3.1. 12.2.3.1 Triggering Power-on Reset by WDT This feature requires that either ESR1 or ESR2 pin is dedicated to it, e.g. not available for another functionality. The following must be done by the user to have power-on reset triggered by WDT: • • in hardware: tie the selected ESRx (x=1,2) pin to PORST pin of the device; in software: at its very beginning, install 1110B into bits[3:0] of the respective ESRCFGx register (x=1,2). Note: Keep the WDT reset configuration as installed by startup procedure - Internal Application Reset. With the above preparation, any Internal Application Reset - including triggered by WDT - will drive the selected ESRx pin low so leading to power-on and a next device restart. User’s Manual SCFG/BSL, PH2 V1.5d1 12-4 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading Attention: When using this solution, Internal Application reset is no more available as separate reset type. Respectively upon this reset all the device resources will be initialized as upon power-on and any previous information will be lost. Therefore, when controlling this feature by WDT do not assign Internal Application Reset to any other trigger if prevention of previous information/status is needed. For additional information on this feature - refer to Application Note AP16146 . 12.2.4 Startup Error state To prevent possible negative consequences for the device and/or the system, upon unrecoverable error during startup XC27x5X is put onto a stable, passive and neutral to the external world state - power-save mode with DMP_1 shut down and DMP_M powered with 1V. This state can be exited with power-on reset only. User’s Manual SCFG/BSL, PH2 V1.5d1 12-5 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.3 Supported Startup Modes and Options XC27x5X supports variety of startup modes, allowing the user to make selections in three aspects: • • • main functionality - where from the user code will be started (on-chip Flash, PSRAM, external memory); optionally - a way for initial code-downloading into PSRAM before to start it: – from an external host via a communication interface - UART, CAN, SSC; – from Stand-by RAM (SBRAM) - after exiting a power-saving mode; debug-related - either debugging will be possible, and if Yes - which debug-interface to use (JTAG, DAP, selectable pin-assignments). Following from the above differentiation, the startup modes in XC27x5X are divided into several groups, described in Chapter 12.3.1, Chapter 12.3.2 and Chapter 12.3.3. User’s Manual SCFG/BSL, PH2 V1.5d1 12-6 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.3.1 Basic Startup Modes These modes (refer to Table 12-2) have no debug support and no special features. Table 12-2 Basic XC27x5X Startup Modes STSTAT.HWCFG Configuration pins 2) Value 1) TRST P10 [6 : 0] Startup Mode Internal Start from Flash 00H 0 x x x x x x x UART Bootloader 2.x 3) 02H 1 x x x x 0 1 0 UART Bootloader 7.x 4) 06H 1 x x x x 1 1 0 SSC Bootloader 09H 1 x x x 1 0 0 1 CAN Bootloader 0DH 1 x x x 1 1 0 1 UART Enhanced Bootloader 2.x 3) 10H 1 0 0 1 0 0 0 0 70H 1 1 1 1 0 0 0 0 External Start 5) 1) Bitfield HWCFG can be loaded from Port 10 or from bitfield SWCFG in register SWRSTCON. 2) x means that the level on the corresponding pin is irrelevant. 3) 2.x means: TxD (transmit data) at P2.3 pin, RxD (receive data) at P2.4 pin. 4) 7.x means: TxD (transmit data) at P7.3 pin, RxD (receive data) at P7.4 pin. Not available in 64-pin package. 5) Not available in 64-pin package. The XC27x5X functionality in different modes - Internal start, External start, Bootstrap loading - is described further in Chapter 12.4, Chapter 12.5, Chapter 12.6 respectively. 0-pin Configuration This is a new feature for XC27x5X, meaning usage of no General-purpose Input-output (GPIO) pins - including no assignment of any alternate function to a pin - to select the startup configuration which is supposed to the most used one - Internal Start from Flash. One dedicated pin - TRST - is used as a checkpoint, as follows: • • TRST=0 during reset - no other pins (also from Port 10) are evaluated, the user code is started from Internal Flash memory - refer to the first row in Table 12-2; TRST=1 during reset - some of the Port 10 pins are evaluated to determine the further device-functioning. The number of P10 pins used for that purpose varies from 3 up to 7- refer to Table 12-2 and Table 12-3. User’s Manual SCFG/BSL, PH2 V1.5d1 12-7 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.3.2 Startup Modes with Debug Support These startup selections (refer to Table 12-3) lead to user-code start either from Internal Flash or from External memory. So from functional point of view they are similar to two from the Basic Startup Modes in Table 12-2, but additionally these selections allow an external tool (debugger) to be connected and used during the development process. Table 12-3 XC27x5X Startup Mode Configurations with debug support Startup Mode Internal Start from Flash Debug Interface External Start 2) JTAG pos.B 03H x x x x 0 1 1 DAP pos.1 04H x x x x 1 0 0 07H x x x x 1 1 1 DAP pos.0 01H x x x 0 0 0 1 DAP pos.24) 05H x x x 0 1 0 1 JTAG pos.C 4) 40H 1 0 0 0 0 0 0 JTAG pos.D 5) 50H 1 0 1 0 0 0 0 60H 1 1 0 0 0 0 0 from Flash 4) STSTAT.HWCFG CFG pins P10 [6:0] Value 1) TRST=1 from Flash 3) 3) 1) Bitfield HWCFG can be loaded from Port 10 or from bitfield SWCFG in register SWRSTCON. 2) x means that the level on the corresponding pin is irrelevant. 3) A defined location in Flash (C0’01F0H) must contain a value (2 Bytes) for DBGPRR register and the next wordlocation (C0’01F2H) must contain the inverse value. From the 16-bit value in Flash the four most-significant bits are don’t care - they are handled by the startup procedure itself. If the inverse-condition does not match - the value is considered as invalid and JTAG pins at position A are configured by default. 4) Not available in 64-pin package. 5) Not available in 64 and 100-pin package. The variety of startup configurations (Table 12-3) in this case serve to select the type (DAP or JTAG) and pin-location of the debug interface which will be enabled. Besides of this selection - done either via P10-pins or by SWRSTCON.HWCFG-bitfield - two basic conditions exist for debug-interface handling: • • debug interface configuration is a part of the complete device startup configuration. Therefore debug interface enabling/disabling/(re)configuration takes place only when the startup configuration (e.g. the startup mode selection) is updated in some of the ways described in Chapter 12.1. For example, upon an application reset triggered by hardware source (let say WDT) the debug interface will not be touched. debug interface will be always disabled, if Internal Start is selected and the on-chip Flash is protected User’s Manual SCFG/BSL, PH2 V1.5d1 12-8 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading The exact meaning - interface types and pin-assignments - of different debug interface selections is shown in Table 12-4. The last column of this table shows the value, which must be written into Flash location C0 01F0H, if Debug Interface Configuration from Flash has been selected as startup option (refer to Table 12-3). Table 12-4 Debug-selections in XC27x5X: interface types and pin assignments Debug Interface Type DAP JTAG Pins used for Debugging: Position main interface (obligatory) DBGPRR value in Flash addr. BRKIN (optional) C0 01F0H pos.0 P2.9, P7.0 P5.10 1000H pos.1 P10.9, P10.12 P10.8 1155H pos.2 1) P7.0, P7.4 P7.1 12AAH pos.A P2.9, P5.2, P5.4, P7.0 P5.10 1800H P10.9, P10.10, P10.11, P10.12 P10.8 1955H P7.0, P7.2, P7.3, P7.4 P7.1 1AAAH pos.D 2) P8.3, P8.4, P8.5, P10.12 P8.6 1BFFH --- --- --- 0000H pos.B pos.C not available 1) 1) Not available in 64-pin package. 2) Not available in 64- and 100-pin packages. There are two types of interface signals/pins related to debugging: Main Debug Interface These are 2 (in case of DAP) or 4 (in case of JTAG) pins listed in the third column of Table 12-4. If debugging is enabled, these pins are always assigned to the debug-interface, therefore the application software must never use any of them. Optional Break Interface The Break Interface of XC27x5X Debug System includes two signals: BRKIN and BRKOUT. The usage of this interface is optional, also selectively either only one out of the two signals or both of them can be utilized. The Break Interface usage requires additional preparation which will be done - when requested - by the external debugger once the main interface is available. User’s Manual SCFG/BSL, PH2 V1.5d1 12-9 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading As long as this preparation and the activation of “Break-In”/”Break-Out” feature has not happened, the respective pin(s) selected to host (potentially) BRKIN/BRKOUT-signal(s) can be still used for other functionality by the application software. The two Break-signals/pins are handled some differently to each other: • • BRKIN - the exact pin which will be used (in case) for this purpose is determined uniquely by the startup selection - refer to the BRKIN-column in Table 12-4. BRKOUT - no pin-selection for this signal is done during startup. Few pins are potentially available for BRKOUT-selection - done by the external tool before to activate the “Break-Out” feature: – P6.0; – P1.5 - not available in 64-pin package; – P9.3 - not available in 64- and 100-pin packages; – P10.11 - can not be used, if JTAG interface at position B is selected. User’s Manual SCFG/BSL, PH2 V1.5d1 12-10 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.3.3 Special Startup Features XC27x5X supports some special features, which allow the user software to influence the device startup, providing additional functionality next to the above (in Chapter 12.3.1 and Chapter 12.3.2) described. Attention: The correct usage of these features requires good and detailed understanding of the XC27x5X structure, behavior and programming. The special startup features are dedicated to advanced users, being familiar with device as a whole and especially with the System Control Unit - both as hardware (described in SCU Chapter of this User Manual) and how to control it properly by software (described in the XC27x5X Programmer’s Guide). 12.3.3.1 Supplementary Startup Information from/to the User The special startup features require/provide additional information from/to the application software, using a dedicated register inside the System Control Unit STMEM0. STMEM0 Register The SCU_STMEM0 register is located in DMP_M power-supply domain and is Securityprotected. The following startup information can be exchanged with application software using this register: 1. the user software can influence the next device startup by writing into STMEM0 bits[15:6] The supported features are described in Chapter 12.3.3.2. 2. if STMEM0[4]=0 after startup - this startup has been triggered by a Functional (i.e. Application or Internal Application) Reset; In such a case the emergency-status flags indicated in SCU_SYSCON0 bits[15:12] upon device startup can be read by user software from SCU_STMEM0 bits[3:0] 3. if STMEM0[4]=1 after startup - this startup has been triggered by a Power-On; In such a case, additional information is provided by SCU_STMEM0[3] bit: a) STMEM0[3]=0 - Power-On in DMP_1 domain only b) STMEM0[3]=1 - Power-On in both DMP_1 and DMP_M domains User’s Manual SCFG/BSL, PH2 V1.5d1 12-11 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading STMEM0 Startup Memory 0 Register 15 14 USSET 0 rw rw 13 12 ESFR (F0A0H/50H) 11 10 9 6 5 0 0 0 STSIND rw rw rw rw RINDP RINDS RINPS rw rw rw 8 7 Reset Value: 0000H 4 3 2 1 0 Field Bits Typ Description STSIND [4:0] rw Startup Status Indication to the user: 0SPVW Functional reset - bits[3:0] show status flags: S - Clock Select status 1) P - PVC1 Emergency Event Source status 1) V - VCOLCK Emergency Event Source status 1) W - OSCWDT Emergency Event Source status 1) 10000 Power-on in DMP_1 domain only 11000 Power-on in DMP_M and DMP_1 domains else Reserved 0 5 rw Reserved, must be written with reset value 0 0 6,[10:7] rw Reserved, must be written with reset value 0 RINPS 11 rw Initialization of the PSRAM: 0 not requested 1 will be performed upon startup RINDS 12 rw Initialization of the DSRAM: 0 not requested 1 will be performed upon startup RINDP 13 rw Initialization of the DPRAM: 0 not requested 1 will be performed upon startup 0 14 rw Reserved, must be written with reset value 0 USSET 15 rw RAM Initialization upon startup: 0 not requested 1 requested in STMEM0 [13:11] 1) Bit copied from SYSCON0 register upon startup. User’s Manual SCFG/BSL, PH2 V1.5d1 12-12 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.3.3.2 Preparing to activate Memory Content Protection XC27x5X supports two mechanisms for Memory Content protection: ECC (Error Correction Code) and Parity, both are disabled by default. Any of these mechanisms can be only activated by the user, using the sequence shown at Figure 12-1. The processing according to this sequence includes: • • • upon power-on of the device (indicated by STMEM0[4]=1, STMEM0[15] will be 0 in this case ) - RAM initialization is needed: – optionally - if the application will run with system clock faster than 10MHz (system frequency after power-on) - the clock reconfiguration can be done still here to use increased speed for a faster RAM initialization; – if parity will be used for some memory - write 1 into the respective bit(s) of SCU_MCHKCON register (by default ECC is the selected protection type); – install request for RAMs initialization by setting STMEM0[15:11]=10111B; It is also possible to set selectively only some of the bits[13:11] corresponding to the memories in which the Content Protection will be activated (refer to STMEM0description and Figure 12-1). This will not bring too much - in sense of a faster startup - because all the memories are initialized in parallel and the time-variation if processing one only or all the RAMs will be not so big. – trigger an application reset to cause a new device startup During this new device startup the RAMs are initialized as requested in STMEM0[13:11]. if STMEM0[15]=1 after startup - meaning RAMs have been just initialized: – RAM-initialization request is cleared - STMEM0[15:11]=00000B; – ECC/parity is configured/enabled as required by the application – continue with further system initialization (if any) and starting the application if STMEM0[15]=0 after functional reset (not power-on) - RAM initialization is not needed and the request is not active: – ECC/parity is configured/enabled as required by the application - this is needed because some control registers are reset upon any startup; – continue with further system initialization (if any) and starting the application. The read-operations from initialized memories produce no errors, the data delivered is: • • if ECC is active - 0600H; if parity is active - 3000H. User’s Manual SCFG/BSL, PH2 V1.5d1 12-13 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading Power-On or Functional reset Device Startup ECC/parity will be used ? Yes ECC/Parity Preparation SCU_STMEM0[15] = 0 ? Yes RAM Initialization has not been requested SCU_STMEM0[4] = 1 ? Functional Reset Yes Power-on – RAM Initialization is needed Set-up the system clock as needed for the application ( optionally here ) No If parity will be used for a memory – set the respective bit in SCU_MCHKCON (by default ECC selected–no action needed) RAM Initialization has been performed No Clear RAM Initialization Request: ---------------------------------------------------------SCU_STMEM 0:= SCU_STMEM0 & 07FF H Enable ECC/parity as needed No Request RAM Initialization during Start-up: ---------------------------------------------------------SCU_STMEM0 := 10CD P000 0000 0000B where bit=1 is to initialize respectively: C – DPRAM; D – DSRAM; P - PSRAM Trigger an Application (Software) Reset: ---------------------------------------------------------SCU_RSTCON0 := 0300H Execute SRST instruction Application-specific system initialization... EINIT ( Register protection activated ) Application Software P11-RAM_Init_Request-v3.vsd Figure 12-1 Software sequence to prepare ECC/Parity usage User’s Manual SCFG/BSL, PH2 V1.5d1 12-14 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.4 Internal Start When internal start mode is configured, the XC27x5X immediately begins executing code out of the on-chip Flash memory (first instruction from location C0’0000H). Because internal start mode without debug-support is expected to be the configuration used in most cases, this mode can be selected by pulling low the dedicated (e.g. not available for application-purposes) TRST-pin only - so-called 0-pin Configuration. If debug-support is needed - additional configuration options are available, refer to Chapter 12.3.2. Note: A read-protected Flash is readable for applications started in Internal mode without disabling the protection. 12.5 External Start When external start mode is configured, the XC27x5X begins executing code out of an off-chip memory (first instruction from location 00’0000H), connected to the XC27x5X’s external bus interface. Attention: External Startup mode is not supported in 64-pin package devices. The External Bus Controller is adjusted to the employed external memory by evaluating additional configuration pins. Seven pins of P10 are used to select the EBC mode (P10.[10:8]), the address width (P10.[12:11]), and the number of chip select lines (P10.[14:13]). The following tables summarize the available options. User’s Manual SCFG/BSL, PH2 V1.5d1 12-15 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading Table 12-5 EBC Configuration: EBC Mode EBC Startup Mode Cfg. Pins P10[10:8] Pins Used by the EBC 8-Bit Data, Multiplexed 0 0 0 P2.0 … P2.2, P10.0 … P10.15 8-Bit Data, Demultiplexed 0 0 1 P0.0 … P0.7, P1.0 … P1.7, P2.0 … P2.2, P10.0 … P10.7, P10.13, P10.14 16-Bit Data, MUX, BHE mode 0 1 0 P2.0 … P2.2, P2.11, P10.0 … P10.15 16-Bit Data, MUX, WRH mode 0 1 1 P2.0 … P2.2, P2.11, P10.0 … P10.15 16-Bit Data, DeMUX, BHE mode, A0 1 0 0 P0.0 … P0.7, P1.0 … P1.7, P2.0 … P2.2, P2.11, P10.0 … P10.14 16-Bit Data, DeMUX, WRH mode, A0 1 0 1 P0.0 … P0.7, P1.0 … P1.7, P2.0 … P2.2, P2.11, P10.0 … P10.14 16-Bit Data, DeMUX, BHE mode, A1 1 1 0 P2.0 … P2.2, P10.0 … P10.15 16-Bit Data, DeMUX, WRH mode, A1 1 1 1 P0.0 … P0.7, P1.0 … P1.7, P2.0 … P2.2, P10.0 … P10.7, P10.13, P10.14 Table 12-6 EBC Configuration: Address Width Available Address Lines Cfg. Pins Additional Address Pins P10[12:11] A15 … A0 0 0 None A17 … A0 0 1 P2.3, P2.4 A19 … A0 1 0 P2.3 … P2.6 A23 … A0 1 1 P2.3 … P2.10 Table 12-7 EBC Configuration: Chip Select Lines Available Chip Select Lines Cfg. Pins Used Pins P10[14:13] CS0 … CS4 0 0 P4.0 … P4.4 CS0 0 1 P4.0 CS0 … CS1 1 0 P4.0, P4.1 None 1 1 None User’s Manual SCFG/BSL, PH2 V1.5d1 12-16 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.5.1 Specific Settings When the XC27x5X has entered External Start mode, the configuration is automatically set according to Table 12-8 and Table 12-9. Note, that the startup procedure does not configure any address window within ADDRSELx registers. Therefore, even if some CS signal is configured (refer to Table 12-7), the startup procedure only makes the proper settings to assure the adequate pin-functionality in regard to the selected EBC mode. The user software must take care: • • to configure the address window (in ADDRSELx register) for the CSx pin(s) which will be used; to enable those pins by setting FCONCSx.ENCS. Table 12-8 External start mode-Specific State in EBC Registers Configuration at P10[10:8] EBCMOD0 [15:8] EBCMOD1 FCONCSx 1) Comment (EBC Mode) 000B 30H 001FH 0011H 8-Bit Multiplexed 001B 70H 0020H 0001H 8-Bit Demultiplexed 010B 40H 0000H 0031H 16-Bit MUX, BHE 011B 48H 0000H 0031H 16-Bit MUX, WRH 100B 60H 0000H 0021H 16-Bit DeMUX, BHE, A0 101B 61H 0000H 0021H 16-Bit DeMUX, WRH, A0 110B 60H 0010H 0021H 16-Bit DeMUX, BHE, A1 111B 61H 0010H 0021H 16-Bit DeMUX, WRH, A1 1) Which FCONCSx registers are affected is dependant on the configuration at P10[14:13] as follows: 11B or 01B - FCONCS0 is affected 10B - FCONCS0 and FCONCS1 are affected 00B - FCONCS0..FCONCS4 are affected The other (unaffected) FCONCS registers retain their default values - refer to the EBC Chapter of this Manual. Table 12-9 External start mode-Specific State in EBCMOD0[7:0] Configuration at P10[14:13] Configuration at P10[12:11] 00B (0 Segm.) 01B (2 Segm.) 10B (4 Segm.) 11B (8 Segm.) 00B (5 CS) 50H 52H 54H 58H 01B (1 CS) 10H 12H 14H 18H 10B (2 CS) 20H 22H 24H 28H 11B (0 CS) 00H 02H 04H 08H User’s Manual SCFG/BSL, PH2 V1.5d1 12-17 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading 12.6 Bootstrap Loading Bootstrap Loading is the technique of transferring code to the XC27x5X via a certain interface (usually serial) before the regular code execution out of non-volatile program memory commences. Instead, the XC27x5X executes the previously received code. This boot-code may be complete (e.g. temporary software for testing or calibration), amend existing code in non-volatile program memory (e.g. with product-specific data or routines), or load additional code (e.g. using higher or more secure protocols). A possible application for bootstrap loading is the programming of virgin Flash memory at the end of a production line, with no external memory or internal Flash required for the initialization code. 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. The XC27x5X supports bootstrap loading using several protocols/modes: • • • • Standard UART protocol, loading 32 bytes (see Section 12.6.2.1) UART protocol, Enhanced bootstrap loader transferring arbitrary number of bytes (see Section 12.6.2.2) Synchronous serial protocol (see Section 12.6.3) CAN protocol (see Section 12.6.4) For a summary of these modes, see also Table 12-16. 12.6.1 General Functionality Even though each bootstrap loader has its particular functionality, the general handling is the same for all of them. Entering a Bootstrap Loader Bootstrap loaders are enabled by selecting a specific startup configuration (see Section 12.1). The required configuration patterns are described in Table 12-16 for the bootstrap loaders, and are summarized in Table 12-2. User’s Manual SCFG/BSL, PH2 V1.5d1 12-18 V1.2, 2008-09 XC27x5X Derivatives System Units (Vol. 1 of 2) Startup Configuration and Bootstrap Loading Loading the Startup Code After establishing communication, the BSL enters a loop to receive the respective number of bytes. These bytes are stored sequentially into the on-chip PSRAM, starting at location E0’0000H. To execute the loaded code the BSL then points register VECSEG to location E0’0000H, i.e. the first loaded instruction, and then jumps to this instruction. The loaded code may be the final application code or another, more sophisticated, loader routine that adds a transmission protocol to enh